PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of blackwellopenThis ArticleFor AuthorsLearn MoreSubmit
The Cochrane Database of Systematic Reviews
 
Cochrane Database Syst Rev. 2015 September 14; (9): CD011045.
PMCID: PMC4579823

Portion, package or tableware size for changing selection and consumption of food, alcohol and tobacco

Abstract

Background

Overeating and harmful alcohol and tobacco use have been linked to the aetiology of various non-communicable diseases, which are among the leading global causes of morbidity and premature mortality. As people are repeatedly exposed to varying sizes and shapes of food, alcohol and tobacco products in environments such as shops, restaurants, bars and homes, this has stimulated public health policy interest in product size and shape as potential targets for intervention.

Objectives

1) To assess the effects of interventions involving exposure to different sizes or sets of physical dimensions of a portion, package, individual unit or item of tableware on unregulated selection or consumption of food, alcohol or tobacco products in adults and children.

2) To assess the extent to which these effects may be modified by study, intervention and participant characteristics.

Search methods

We searched CENTRAL, MEDLINE, EMBASE, PsycINFO, eight other published or grey literature databases, trial registries and key websites up to November 2012, followed by citation searches and contacts with study authors. This original search identified eligible studies published up to July 2013, which are fully incorporated into the review. We conducted an updated search up to 30 January 2015 but further eligible studies are not yet fully incorporated due to their minimal potential to change the conclusions.

Selection criteria

Randomised controlled trials with between-subjects (parallel-group) or within-subjects (cross-over) designs, conducted in laboratory or field settings, in adults or children. Eligible studies compared at least two groups of participants, each exposed to a different size or shape of a portion of a food (including non-alcoholic beverages), alcohol or tobacco product, its package or individual unit size, or of an item of tableware used to consume it, and included a measure of unregulated selection or consumption of food, alcohol or tobacco.

Data collection and analysis

We applied standard Cochrane methods to select eligible studies for inclusion and to collect data and assess risk of bias. We calculated study-level effect sizes as standardised mean differences (SMDs) between comparison groups, measured as quantities selected or consumed. We combined these results using random-effects meta-analysis models to estimate summary effect sizes (SMDs with 95% confidence intervals (CIs)) for each outcome for size and shape comparisons. We rated the overall quality of evidence using the GRADE system. Finally, we used meta-regression analysis to investigate statistical associations between summary effect sizes and variant study, intervention or participant characteristics.

Main results

The current version of this review includes 72 studies, published between 1978 and July 2013, assessed as being at overall unclear or high risk of bias with respect to selection and consumption outcomes. Ninety-six per cent of included studies (69/72) manipulated food products and 4% (3/72) manipulated cigarettes. No included studies manipulated alcohol products. Forty-nine per cent (35/72) manipulated portion size, 14% (10/72) package size and 21% (15/72) tableware size or shape. More studies investigated effects among adults (76% (55/72)) than children and all studies were conducted in high-income countries - predominantly in the USA (81% (58/72)). Sources of funding were reported for the majority of studies, with no evidence of funding by agencies with possible commercial interests in their results.

A meta-analysis of 86 independent comparisons from 58 studies (6603 participants) found a small to moderate effect of portion, package, individual unit or tableware size on consumption of food (SMD 0.38, 95% CI 0.29 to 0.46), providing moderate quality evidence that exposure to larger sizes increased quantities of food consumed among children (SMD 0.21, 95% CI 0.10 to 0.31) and adults (SMD 0.46, 95% CI 0.40 to 0.52). The size of this effect suggests that, if sustained reductions in exposure to larger-sized food portions, packages and tableware could be achieved across the whole diet, this could reduce average daily energy consumed from food by between 144 and 228 kcal (8.5% to 13.5% from a baseline of 1689 kcal) among UK children and adults. A meta-analysis of six independent comparisons from three studies (108 participants) found low quality evidence for no difference in the effect of cigarette length on consumption (SMD 0.25, 95% CI -0.14 to 0.65).

One included study (50 participants) estimated a large effect on consumption of exposure to differently shaped tableware (SMD 1.17, 95% CI 0.57 to 1.78), rated as very low quality evidence that exposure to shorter, wider bottles (versus taller, narrower bottles) increased quantities of water consumed by young adult participants.

A meta-analysis of 13 independent comparisons from 10 studies (1164 participants) found a small to moderate effect of portion or tableware size on selection of food (SMD 0.42, 95% CI 0.24 to 0.59), rated as moderate quality evidence that exposure to larger sizes increased the quantities of food people selected for subsequent consumption. This effect was present among adults (SMD 0.55, 95% CI 0.35 to 0.75) but not children (SMD 0.14, 95% CI -0.06 to 0.34).

In addition, a meta-analysis of three independent comparisons from three studies (232 participants) found a very large effect of exposure to differently shaped tableware on selection of non-alcoholic beverages (SMD 1.47, 95% CI 0.52 to 2.43), rated as low quality evidence that exposure to shorter, wider (versus taller, narrower) glasses or bottles increased the quantities selected for subsequent consumption among adults (SMD 2.31, 95% CI 1.79 to 2.83) and children (SMD 1.03, 95% CI 0.41 to 1.65).

Authors' conclusions

This review found that people consistently consume more food and drink when offered larger-sized portions, packages or tableware than when offered smaller-sized versions. This suggests that policies and practices that successfully reduce the size, availability and appeal of larger-sized portions, packages, individual units and tableware can contribute to meaningful reductions in the quantities of food (including non-alcoholic beverages) people select and consume in the immediate and short term. However, it is uncertain whether reducing portions at the smaller end of the size range can be as effective in reducing food consumption as reductions at the larger end of the range. We are unable to highlight clear implications for tobacco or alcohol policy due to identified gaps in the current evidence base.

PLAIN LANGUAGE SUMMARY

Portion, package or tableware size for changing selection and consumption of food, alcohol and tobacco

Review question

We reviewed the evidence to establish by how much the amounts of food, alcohol or tobacco adults and children select or consume change in response to being presented with larger or smaller-sized (or differently shaped) portions or packages of these products, or of items of tableware (such as plates or glasses) used to consume them.

Study characteristics

This review includes 72 randomised controlled trials (RCTs) published up to July 2013 that compared at least two groups of participants, each presented with a different size of a portion, package or item of tableware. Included studies measured the amounts of food, alcohol or tobacco selected and/or consumed by participants, typically over a period of one day or less. Almost all of the included studies investigated food, with only three tobacco studies and no alcohol studies found. Almost all assessed participants' responses to different sizes rather than different shapes. The average age of participants in the different studies ranged from three to 55 years, with more studies involving adults than children and most conducted in the USA. Sources of funding were reported for the majority of studies and there was no evidence of study funding by agencies with commercial interests in their results.

Key findings and quality of evidence

Effects of size on consumption: We found evidence that people consistently ate more food or drank more non-alcoholic drinks when offered larger-sized portions, packages or items of tableware than when offered smaller-sized versions. We estimate the size of this effect to be small to moderate among both children and adults. If an effect of this size were sustained across the whole diet it would be equivalent to around a 12% to 16% change in average daily energy intake from food among UK adults. We rated the overall quality of the evidence for this effect as moderate, due to concern about study limitations arising from incomplete or unclear reporting of methods and procedures. From three tobacco studies, we found no effect of longer compared with shorter cigarettes on the amounts of tobacco consumed. We rated the overall quality of evidence for this effect as low due to concerns about study limitations and not having enough evidence.

Effects of shape on consumption: One study found that adults provided with shorter, wider bottles drank larger amounts of water from them, having already poured more, compared with those provided with taller, narrower bottles. However, we rated the quality of this evidence as very low, due to very serious concerns about study limitations and not having enough evidence (only one study with outcome data from 50 participants).

Effects of size on selection: We further found that adults, but not children, consistently chose (selected) more food (including non-alcoholic drinks) when offered larger-sized portions, packages or items of tableware than when offered smaller-sized versions. The estimated size of this effect was again small to moderate. We rated the overall quality of the evidence for this effect as moderate, due to concern about study limitations.

Effects of shape on selection: Evidence from three studies suggested that adults and children provided with shorter, wider bottles or glasses selected increased quantities of non-alcoholic beverages for subsequent consumption, compared with those provided with taller, narrower bottles or glasses. We rated the quality of this evidence as low, again due to concerns about study limitations and unexplained variation in effects between the three studies.

Conclusions

Overall, this review provides the most conclusive evidence to date that acting to reduce the size, availability and appeal of larger-sized portions, packages and tableware has potential to reduce the quantities of food that people select and consume by meaningful amounts. However, it is uncertain whether reducing portions at the smaller end of the size range can be as effective in reducing food consumption as reductions at the larger end of the range. Our findings highlight the need for further research that aims to reduce uncertainties about these effects and address identified gaps in the evidence base, including not having enough evidence for longer-term effects and the absence of evidence about alcohol products.

SUMMARY OF FINDINGS FOR THE MAIN COMPARISON [Explanation]

Table thumbnail

Background

Description of the condition

Non-communicable diseases, principally cardiovascular diseases, diabetes, certain forms of cancer and chronic respiratory diseases, accounted for an estimated 62% of all deaths worldwide in 2012 (World Health Organization 2014a), and globally the proportion of years of life lost as a result of non-communicable diseases increased from 38% in 2000 to 47% in 2012 (World Health Organization 2014b). Major risk factors for these conditions are in part determined by patterns of behaviour that are in principle modifiable, including consumption of food, alcohol and tobacco products (United Nations 2014). Identifying interventions that are effective in achieving sustained health behaviour change has therefore become one of the most important public health challenges of the 21st century.

Description of the intervention

It is increasingly recognised that the physical environments that surround us can exert considerable influences on our health behaviour and that altering these environments may provide a catalyst for behaviour change (Das 2012). In a recent scoping review, we described a class of interventions that involve altering the properties or placement of objects or stimuli within micro-environments such as shops, restaurants, bars or homes, with the intention of changing health-related behaviours (Hollands 2013a; Hollands 2013b).

The size of a portion or package is a modifiable property of food, alcohol and tobacco products that may influence their selection and consumption. In the case of food and alcohol products, the size or shape of an item of tableware used to consume such products may similarly influence their selection and consumption. Examples include the portion size of alcoholic beverages served in bars or of foods served in restaurants, at a buffet or in the home, such as portions of a dish served to restaurant customers (Diliberti 2004), the size or shape of plates or glasses used to serve products (Shah 2011), and the number or length of cigarettes in packets sold in shops (Russell 1980). In this context, the intervention involves manipulation of the size or physical dimensions of a food, alcohol or tobacco product, its packaging or the tableware used in its consumption. Comparisons of interest are between products, packages or items of tableware that differ only in terms of these properties.

How the intervention might work

There are considerable influences on behaviour that are beyond individuals' deliberative control. Indeed, it has been suggested that most human behaviour occurs outside of awareness, cued by stimuli in environments and resulting in actions that may be largely unaccompanied by conscious reflection (Marteau 2012; Neal 2006). This proposition has led to increasing policy and research attention being placed on interventions with mechanisms of action that are less dependent on the conscious engagement of the recipients, including interventions that involve altering properties of objects or stimuli within the small-scale environments that surround and cue behaviour (Hollands 2013a).

A number of mechanisms of action have been proposed to explain how the size of products may affect their consumption (Herman 2015; Steenhuis 2009). It has been suggested that as the amount of a product made available for consumption is increased, individuals will continue to perceive each increasing amount as an appropriate quantity to consume. This phenomenon may be explained by several mediating factors including personal and social norms about what constitutes a suitable amount of a product to consume. Such norms can be influenced by the amounts that are presented for consumption, and larger portions of food have become increasingly prevalent, making it increasingly unlikely that smaller portions are viewed as normal or appropriate for a single serving (Young 2002). There is also a tendency for individuals to engage most comfortably with a product as a single entity independent of its size. This 'unit bias' means that they are predisposed to consume the entirety of a product even as it changes size (Geier 2006). In addition, the way in which products are presented can influence their consumption. The presentation of food and alcohol products often entails the use of tableware, such as plates, glasses or cutlery. Not only does the size of tableware have the potential to directly influence the amount of a product available for consumption (Pratt 2012), but its physical dimensions can elicit various cognitive biases (Wansink 2005), which may influence perceptions of quantity and in turn determine levels of consumption. Similarly, sub-dividing a fixed portion of a food into smaller pieces also affects perceptions of quantity (Scisco 2012). All of these mechanisms may also influence product selection (with or without purchasing), which is an important intermediate outcome in pathways to consumption.

Extant research involving the experimental manipulation of portion, package or tableware size has focused on food (including non-alcoholic beverage) products to a much greater extent than tobacco products (Hollands 2013a). Whilst the causal mechanisms of underlying potential effects of such manipulations on selection or consumption of tobacco may be assumed to be broadly similar to food, smokers are known to titrate their received dose of nicotine to regulate the level in the body, with the potential to attenuate the effects of interventions to alter the size of tobacco products (Kozlowski 1986).

Why it is important to do this review

A recent scoping review of evidence for the effects of choice architecture interventions identified a substantial number of randomised controlled trials that have investigated the effects of exposure to different portion, package or tableware sizes on selection and consumption behaviours (Hollands 2013a). The majority of these studies focused on food products, but because both tobacco and alcohol use also involve the selection and consumption of products, similar interventions may have the potential to change these behaviours via similar mechanisms. To our knowledge, evidence from these studies has yet to be synthesised using rigorous systematic review methods that include assessment of risk of bias and investigation of potential effect modifiers, nor to encompass alcohol and tobacco use. As such, we do not yet have reliable estimates of the effects of altering the sizes of portions, packages or tableware on product selection and consumption, nor of the influence of factors that may modify any such effects. Both are necessary to inform the selection and design of effective public health interventions.

Interventions that aim to reduce people's exposure to larger or smaller food portions, as opposed to those that involve providing information to encourage health behaviour change, may also have the potential to reduce health inequalities if they rely less on recipients' levels of literacy, numeracy and cognitive control, which have been found to be lower in population subgroups experiencing higher levels of social and material deprivation (Kutner 2006; Marteau 2012; Spears 2010; Williams 2003). Despite evidence that behaviours with the potential to undermine health are socially patterned (for example, that people in lower socioeconomic groups tend to consume less fruit and vegetables (Giskes 2010)), potential differences in behavioural responses to product sizing interventions between socioeconomic subgroups remain unclear. Also, to our knowledge (prior to conducting this review), no studies of the effects of product size had been conducted in low or middle-income (LMIC) country populations (Hollands 2013a). This review therefore includes a focus on identifying evidence for differential effects of exposure to different sizes of these products between socioeconomic subgroups (and between studies conducted in LMIC and high-income countries (HIC)), highlight any identified gaps in this aspect of the evidence base, and seek to draw implications for the potential of such interventions to affect health inequalities.

This systematic review is also timely given current interest in the topic within public health policy circles. There is evidence from the USA and Europe that portion sizes have been increasing since the 1970s (Young 2002; Young 2012). There have also been recent attempts to regulate the size of products in order to reduce consumption levels and improve public health, such as New York City Mayor Michael Bloomberg's proposed ban on the sale of sugary drinks larger than 16 oz (473 ml) (Gabbatt 2013). In the UK, there are recent examples of companies reducing the portion sizes of confectionery and sugary drinks as part of the Public Health Responsibility Deal in England. This systematic review can contribute to a better evidence-based understanding of the potential impact of such policies.

Objectives

  1. To assess the effects of interventions involving exposure to different sizes or sets of physical dimensions of a portion, package, individual unit or item of tableware on unregulated (ad libitum) selection or consumption of food, alcohol or tobacco products in adults and children.
  2. To assess the extent to which the effects of such interventions may be modified by:
    1. study characteristics, such as target product type (food, alcohol, tobacco) or whether the target of the manipulation is a portion, package, individual unit or item of tableware;
    2. intervention characteristics, such as magnitude of the difference in size; and
    3. participant characteristics, such as age, gender or socioeconomic status (to facilitate an assessment of social differentiation in effects relevant to health equity).

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials with between-subjects (parallel-group) or within-subjects (cross-over) designs, conducted in laboratory or field settings. We excluded non-randomised studies because our recent scoping review indicated that a sufficient number of eligible randomised controlled trials would be available to address our aim to synthesise evidence for intervention effects (Hollands 2013a). A key issue is that, compared with randomised controlled trials, non-randomised studies rely on more stringent and sometimes non-verifiable assumptions in order to confer confidence that, with successful implementation of the study design, the risk of systematic differences between comparison groups beyond the intervention of interest (i.e. confounding) is sufficiently low to permit valid inferences about causal effects.

Types of participants

Adults and children directly engaged with the manipulated products. We set no exclusion criteria in relation to demographic, socioeconomic or clinical characteristics or prognostic factors. We excluded studies involving non-human participants (animal studies).

Types of interventions

Interventions eligible to be considered in this review were those that involved comparison of the effects of exposure to at least two sizes or sets of visible physical dimensions (that is volume, shape, height, width or depth) of either a portion of the same food (including non-alcoholic beverages), alcohol or tobacco product, its package or individual unit size, or an item of tableware used to consume it. An eligible study could therefore include multiple eligible comparisons. For example, in a three-arm between-subjects study comparing the effects of exposure to a 200 g, 300 g or 400 g portion of pasta with sauce, eligible comparisons are: 200 g versus 300 g; 300 g versus 400 g; and 200 g versus 400 g (see also Data synthesis).

'Portion' refers to the overall amount (volume, weight or both) of a product that is presented for selection or consumption (for example, 200 g versus 300 g of pasta, 275 ml versus 440 ml of beer, or a packet of 10 versus 20 cigarettes). 'Package' refers to the different ways of packaging a specific portion, including that used for service, consumption or storage (for example, boxes, bags, cans or bottles). For example, the same portion of a food could be served within one large bag or multiple smaller bags. 'Individual unit' refers to the unit of a product that is presented within a given portion (for example, individual sweets or candies, biscuits or cookies, or cigarettes). 'Tableware' refers to crockery, cutlery or glassware used for serving or consuming food or drink (for example, plates, bowls, knives, forks, spoons or glasses). Packages and tableware as defined in this way have the capacity to limit or increase the portion or individual unit size of the consumed product and may therefore influence any corollary effects on selection and consumption.

We excluded the following:

  • Interventions in which product size and/or shape may have been altered indirectly as a result of a higher-level intervention but were not directly manipulated, to safeguard implementation fidelity (e.g. organisational-level interventions to encourage the introduction of small-scale environmental changes to alter product selection or consumption).
  • Interventions in which the behavioural responses of participants (that is, selection or consumption levels or rates) were regulated by either explicit instructions to participants or some other action of the researcher (e.g. participants exposed to a product were given instructions on how much they should consume or a target rate of consumption). In such cases, selection or consumption of the manipulated product cannot be considered unregulated (ad libitum).
  • Studies that compared packages, portions, individual units or tableware of different types or with different functions. For example, we excluded studies that made comparisons between different, differently sized eating utensils (e.g. straw versus spoon; chopsticks versus fork) whilst studies that made comparisons between different sizes of the same eating utensil were included (e.g. small spoon versus large spoon).
  • Studies in which there were concurrent interventions unrelated to sizing that were intrinsically confounded with the comparison(s) of interest. For example, we excluded two-arm studies in which one comparison group received a specified portion size and the other group received a smaller portion plus a concurrent nutritional labelling intervention.

Types of outcome measures

Primary outcomes

Behavioural endpoints

Eligible studies had to incorporate one or more measures of unregulated (ad libitum) consumption or selection (with or without purchasing) of food, alcohol or tobacco products. By unregulated, we refer to behaviour of participants that is not regulated by either explicit instructions or some other action of the researcher. Eligible studies may have measured consumption or selection in terms of quantities of manipulated products and/or quantities of non-manipulated products. For example, a study investigating the effects of exposure to a large versus small portion of a pasta entrée, provided as part of a lunch meal, may have measured consumption in terms of energy intake from the entrée itself, or from a non-manipulated vegetable side dish served with the entrée, or from the total lunch meal (that is, both manipulated and non-manipulated components), or from all meals taken over the course of a whole day. Similarly, quantities consumed or selected may have been measured over a time period less than (immediate) or exceeding one day (longer-term).

Our choice of eligible outcome constructs reflected a focus on the assessment of the effects of eligible interventions in terms of the types and amounts of food, alcohol and tobacco people consume, coupled with recognition that amount selected (with or without purchasing) is an important intermediate endpoint in pathways to consumption. We anticipated encountering a range of measures of these outcome constructs within included studies, and presented the following examples in the published protocol for this review.

1. Consumption (intake) of a product

We assessed the amount of energy (e.g. calories), substances (e.g. carbon monoxide, alcohol, saturated fat), or products (e.g. food, drink or tobacco) consumed, measured in applicable natural units (e.g. kcals, kilojoules, grams). Objective measurement may involve calculating the amount of a product consumed by subtracting the amount remaining after consumption from the amount presented to the participant. Alternatively, it may involve direct observation of the individual by outcome assessors. Subjective measurement would involve participant self report.

2. Selection of a product

a) Without purchase

b) With purchase

As per consumption, we assessed the amount of energy, substances or products selected for consumption, measured in applicable natural units. Depending on the study setting, a product may be selected with or without this act enjoining a purchase (that is, a transfer of money to the vendor).

Conceptual model

To supplement study eligibility criteria, we developed a provisional conceptual model that was published in the protocol for this review (Hollands 2014). This conceptual model was design-oriented in the sense that its purpose was to help direct the review process by providing a simplified visual representation of the causal system of interest: the proposed causal pathway between eligible interventions and their outcomes (behavioural endpoints), and potential moderators of that relationship (effect modifiers) given that differential effects were plausible (Anderson 2011; Anderson 2013). We used the provisional conceptual model to inform the development of search strategies, data extraction forms and a provisional framework for the statistical analysis of outcome data collected from the eligible studies (see Search methods for identification of studies and Data collection and analysis). We iteratively revised the provisional conceptual model based on theory and evidence encountered in eligible studies during the course of the review process, and documented all revisions including the rationale for each revision and supporting evidence (see Data collection and analysis). We used the provisional and subsequent iterations of the conceptual model as a reference point for the design (in the protocol) and conduct (post-protocol) of all stages of the systematic review up to and including data synthesis, and as a conceptual basis for explicit reporting of the methods and assumptions employed within the synthesis (Anderson 2013). In practice, iterative refinement of the conceptual model primarily involved incorporating further potential effect modifiers identified from theory and evidence presented in included study reports, which became candidates for consideration in the meta-regression analysis (see Data collection and analysis). The final version of the conceptual model is shown Figure Figure11.

Figure 1
Final conceptual model. The 28 constructs included in the provisional conceptual model (Hollands 2014) and retained in this final version are shown in plain type. The 22 constructs added to this final conceptual model based on theory and evidence encountered ...

Within the conceptual model (Figure (Figure1)1) we distinguished between three sets of potential effect modifiers: study characteristics; intervention characteristics; and participant characteristics. Within our analytic framework for quantitative synthesis of outcome data collected from the included studies (see Data collection and analysis), potential effect-modifying impacts of participant characteristics could in practice only be investigated based on between-study comparisons, due to lack of reporting of results by participant subgroups within the included studies.

Search methods for identification of studies

We initiated an original search, applying the methods described below in this section, in November 2012. We conducted an updated search, applying the same methods, prior to publication of the current version of the review, with a search date up to and including 30 January 2015. We have added eligible studies identified by the updated search (with subsequent title/abstract and full-text screening) to Characteristics of studies awaiting classification, provisionally analysed them and will fully incorporate them into the review at the next update (see also Results of the search, Appendix 1 and Appendix 2).

Electronic searches

We conducted electronic searches for eligible studies within each of the following databases:

  • Cochrane Central Register of Controlled Trials (CENTRAL 2015, Issue 1) (1992 to 30 January 2015);
  • MEDLINE (OvidSP) (including MEDLINE In-Process) (1946 to 30 January 2015);
  • EMBASE (OvidSP) (1980 to 30 January 2015);
  • PsycINFO (OvidSP) (1806 to 30 January 2015);
  • Applied Social Sciences Index and Abstracts (ProQuest) (1987 to 30 January 2015);
  • Food Science and Technology Abstracts (Web of Knowledge) (1969 to 22 November 2012);
  • Science Citation Index Expanded (Web of Knowledge) (1900 to 30 January 2015);
  • Social Sciences Citation Index (Web of Knowledge) (1956 to 30 January 2015);
  • Trials Register of Promoting Health Interventions (EPPI Centre) (2004 to 30 January 2015).

We developed a MEDLINE search strategy by combining sets of controlled vocabulary and free-text search terms based on the eligibility criteria described above (see Criteria for considering studies for this review). This was externally peer-reviewed by an information retrieval specialist and Co-convenor of the Cochrane Information Retrieval Methods Group and revised based on their peer-review comments. We tested the MEDLINE search strategy for its sensitivity to retrieve a reference set of 48 records of reports of potentially eligible studies known to be indexed in MEDLINE that were identified by our preceding scoping review (Hollands 2013a). We adapted the final MEDLINE search strategy for use to search each of the other databases listed above based on close examination of database thesauri and scope notes if available. We imposed no restrictions for publication date, publication format or language and incorporated no study design filters. Full details of final search strategies for each database, along with search dates and yields (for both the original search and the updated search), are provided in Appendix 1.

Searching other resources

We conducted electronic searches of two grey literature resources using search strategies adapted from the final MEDLINE search strategy:

  • Conference Proceedings Citation Index - Science (Web of Knowledge) (1990 to 30 January 2015);
  • Conference Proceedings Citation Index - Social Science & Humanities (Web of Knowledge) (1990 to 30 January 2015);
  • Open Grey - http://www.opengrey.eu (1980 to 30 January 2015).

We also searched trial registers (ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform (ICTRP)) to identify registered trials, and the websites of the following key organisations in the area of health and nutrition:

  • Centers for Disease Control and Prevention, USA;
  • EU Platform for Action on Diet, Physical Activity and Health;
  • International Obesity Task Force;
  • Rudd Centre for Food Policy and Obesity, USA;
  • UK Department of Health;
  • World Health Organization.

In addition, we searched the reference lists of all eligible study reports that had been identified using the other search methods described above and undertook forward citation tracking (using Google Scholar and PubMed) to identify further eligible studies or study reports.

Data collection and analysis

Selection of studies

We imported title-abstract records retrieved by the electronic searches to EPPI Reviewer 4 (ER4) systematic review software (Thomas 2010). We identified, reviewed manually and removed duplicate records using ER4's automatic de-duplication feature with the similarity threshold set initially to 0.85 and finally to 0.80 following satisfactory manual checks of incomplete duplicate groups. Two researchers working independently (GJH, IS) undertook duplicate screening of title-abstract records. We coded title-abstract records as 'provisionally eligible', 'excluded' or 'duplicate' by applying the eligibility criteria described above (see Criteria for considering studies for this review). Disagreements in the coding of title-abstract records were identified and resolved by discussion to reach consensus between the two researchers (GJH, IS).

We obtained copies of corresponding full-text study reports for all title-abstract records coded as 'provisionally eligible'. Two researchers working independently (GJH, IS) undertook duplicate screening of full-text study reports. We coded full-text study reports as 'eligible' or 'excluded' by applying the eligibility criteria described above (see Criteria for considering studies for this review). Coding disagreements were again identified and resolved by discussion to reach consensus between the two researchers, with a third researcher (DO) acting as arbiter when needed. We recorded bibliographic details of study reports excluded at the full-text screening stage, along with the primary reason for exclusion, in a Characteristics of excluded studies table. We identified and linked multiple full-text reports of the same study. We also identified full-text reports comprising multiple eligible studies. We documented the flow of records and studies through the systematic review process using a PRISMA flow diagram (Moher 2009).

Data extraction and management

We developed an electronic data extraction form based on the Cochrane Public Health Review Group's template (http://ph.cochrane.org/review-authors). We piloted an initial draft form using a selection of 10 included studies and then amended this in consultation with other members of the review team. One researcher (GJH or IS) extracted data on characteristics of included studies, while two researchers working independently (GJH, IS) extracted outcome data in duplicate. We only collected outcome data relating to comparison groups eligible for consideration in this review, but Characteristics of included studies tables record details of all study arms (conditions). Discrepancies in extracted outcome data were identified and resolved by checking against the study report, discussion and consensus between two researchers (GJH, IS). We sought key data missing from reports of included studies by contacting study authors.

At the protocol stage, we intended to collect the data summarised immediately below in this section. This represented the core data set (comprising 28 pre-specified moderator constructs for potential examination using meta-regression analyses; see Data synthesis) that we could reasonably anticipate would need to be collected based on our study eligibility criteria (see Criteria for considering studies for this review) and provisional conceptual model (Hollands 2014).

Study characteristics
  • Study design: between-subjects design, within-subjects design
  • Study (intervention) setting: laboratory, field; for consumption at home or away from home
  • Product type: food (including non-alcoholic beverages), alcohol, tobacco
  • Product healthiness: Food Standards Agency (FSA) score (Rayner 2005) at level of specific product or, if not possible, at level of product category
  • Target of manipulation: portion, package, individual unit, tableware
  • Type of manipulation: size (including volume) or shape
  • Manipulation from a standard size: no or yes*
  • If applicable, direction of the change relative to standard size: smaller or larger*
  • If applicable, selection with purchasing or selection without purchasing
  • Concurrent intervention components (e.g. nutritional labelling)
  • Socioeconomic status context (low, high)

Intervention characteristics
  • Magnitude of the absolute difference in size (e.g. difference in quantity): smaller size always coded as Intervention 1 and larger size as Intervention 2
  • Magnitude of the relative difference in size (e.g. percentage difference in quantity): smaller size always coded as Intervention 1 and larger size as Intervention 2

Participant characteristics
  • Age/age group
  • Gender: male, female
  • Ethnicity
  • Body mass index (BMI); body weight; body weight status
  • Behavioural characteristics (e.g. dietary restraint; susceptibility to hunger)
  • Biological state (e.g. hunger)
  • Other clinical characteristics (e.g. morbidities such as cardiovascular diseases, diabetes, psychiatric disorders)
  • Socioeconomic status (e.g. occupational status; education; income; food insecurity; welfare receipt)
  • Summary risk of bias

These participant characteristics cover several categories of social differentiation relevant to health equity, namely: age, ethnicity, gender, occupation, education, income and other proxy measures of socioeconomic status. The incorporation of study-level data on these participant characteristics into our proposed meta-regression analysis (see 'Data synthesis') was in part intended to enable us to interpret any differential effects through a health equity lens (Welch 2012) (see also Objectives 2c).

As anticipated, our conceptual model - and consequently the core data set - evolved as the review process progressed. First, we excluded a pair of potential effect modifiers (study characteristics) included in our provisional conceptual model that express studied portion size manipulations relative to a standard size (see asterisked characteristics '*' in the list of 'Study characteristics', above), since it was not judged feasible to define standard sizes based on information reported in included studies. Second, the process of collecting data from included studies identified 22 additional potential effect modifiers (moderator constructs) that were added to the conceptual model. These additional constructs were included in the current, published review version of the conceptual model (Figure (Figure1)1) and are listed below:

Study characteristics
  • Product energy density
  • Duration of exposure
  • Relationship between manipulated product(s) and outcome(s)

Intervention characteristics

None added.

Participant characteristics
  • Behavioural characteristics (susceptibility to hunger; external eating; emotional eating; plate cleaning tendency; consumption monitoring; binge eating; dieting behaviour; mood; habitual dietary energy intake; habitual dietary macronutrient intake (carbohydrate; protein; fat); physical activity; energy expenditure; physical exercise)
    • Biological state (fullness; satiety; prospective consumption)
    • Other clinical characteristics (depression)

We coded 28 variables that measured these constructs from included studies (as well as coding 43 variables that measured constructs included in the initial conceptual model). The current, published review version of our conceptual model (Figure (Figure1)1) therefore comprised 48 moderator constructs, with 72 corresponding variables, for potential examination using meta-regression analyses. Table 1 traces this iterative conceptual model development process, documenting all revisions made between the protocol (Hollands 2014) and final versions (Figure (Figure1),1), together with the rationale and supporting evidence for each revision.

Table 1
Record of conceptual model development

Outcome data

As anticipated, eligible primary studies frequently included more than one measure of each target outcome construct, specifically: (a) more than one measure of selection for a given comparison, (b) more than one measure of consumption for a given comparison, or both. For each included study in which (a) or (b) applied, we extracted outcome data for use in meta-analysis for the (a) primary selection or (b) primary consumption outcome(s) as (pre-)specified by the study authors. If the study authors did not (pre-)specify a single (primary) (a) selection or (b) consumption outcome, we applied the following criteria to select the (a) selection or (b) consumption measure for which outcome data would be extracted for use in meta-analysis from a list of all available measures. We selected the measure of (a) selection or (b) consumption most proximal to health outcomes in the context of the specific intervention at hand. For example, if a study reported measures of both energy intake and the amount of food eaten (in grams), we selected energy intake as the measure of the target outcome construct most proximal to diet-related health outcomes. We also selected the largest-scale measure of the target outcome construct. For example, if a study manipulated the size of a portion of vegetable served as one component of a plated entrée, and measured the effects of a large versus a small vegetable portion size in terms of: (i) the amount of that vegetable consumed from the plated entrée, and (ii) the total amount of food consumed from the plated entrée, then we selected (ii) as the consumption outcome measure for which we extracted data. We made each selection in advance of data extraction, blinded to the outcome data. We recorded details of selection and consumption outcomes measures available in each included study and documented these in Characteristics of included studies.

For included studies that investigated a size manipulation, we always coded exposure to the larger of the two portions, packages, individual units or items of tableware as the intervention, whilst we always coded exposure to the smaller of the two as the comparator. For included studies that investigated a shape manipulation, we always coded exposure to the shorter, wider of the two items of tableware as the intervention, whilst we always coded exposure to the taller, narrower of the two as the comparator.

For all outcome data we collected information on: outcome variable type (in practice, this was invariably continuous); outcome variable definition; unit of measurement (natural units); specific metric (final values, change from baseline); method of aggregation (mean); timing of measurement (immediate (that is, ≤ 1 day) or longer-term (that is, = 1 day)); and type of measure (objective, self report). For continuous outcomes, we extracted mean differences, or mean changes in final measurements from baseline measurements, for each comparison group along with associated standard deviations (or, if standard deviations were missing, standard errors, 95% confidence intervals or relevant t-statistics, f-statistics or exact P values that we used to calculate standard deviations); we also indicated whether a high or low value is favourable from a public health perspective. For included studies with factorial designs, we combined comparison groups so that any independent or interactive effects of the co-occurring manipulation were averaged across the comparison groups of interest, in order to allow investigation of the independent effects of the size or shape manipulation.

Assessment of risk of bias in included studies

We assessed risk of bias in the included studies using the Cochrane 'Risk of bias' tool addressing eight specific domains, namely: random sequence generation and allocation concealment (selection bias); blinding of participants and personnel (performance bias); blinding of outcome assessors (detection bias); incomplete outcome data (attrition bias); selective outcome reporting (reporting bias); and baseline comparability of participant characteristics between groups and consistency in intervention delivery (other bias) (Higgins 2011b). The last domain refers to whether information and specific instructions provided to participants were standardised between conditions and whether participant (non-)compliance with the study protocol was appropriately managed.

Two researchers working independently (GJH, IS) applied the Cochrane 'Risk of bias' tool to each included study. We recorded supporting information for judgements of risk of bias (high, low or unclear) in the form of verbatim text extracted from study reports, supplemented with reviewer comments. We identified and resolved discrepancies between the two researchers' judgements or supporting information by discussion to reach consensus. We derived a summary risk of bias judgement (high, low or unclear) for each specific outcome, for inclusion as a study-level covariate in the final stage of the meta-regression analysis (see Data synthesis). We also considered summary risk of bias in determining the strength of inferences drawn from the results of the data synthesis and in developing conclusions and recommendations concerning the design and conduct of future research. We derived the summary risk of bias judgement from the four domains judged to be most critical in this specific review, namely: random sequence generation (selection bias); allocation concealment (selection bias); blinding of participants and personnel (performance bias); and baseline comparability of participant characteristics between groups (other bias). It was derived using an algorithm suggested in Section 8.7 (Table 8.7a) of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011b). Specifically, if the judgement in at least one of these four domains was 'high risk of bias' then we determined summary risk of bias to be high. If no judgements of 'high' risk were made in these four domains, but the judgement in at least one of these domains was 'unclear risk of bias' then we determined the summary risk of bias to be unclear. We only judged summary risk of bias 'low' if judgements in all four of these domains were 'low risk of bias'.

Measures of treatment effect

We calculated the standardised mean difference (SMD) with 95% confidence intervals to express the size of the intervention effect in each study relative to the variability observed in that study. We classified included study results according to two categories of timing of outcome measurement: immediate outcomes (that is ≤ 1 day) versus longer-term outcomes (that is = 1 day).

Unit of analysis issues

In the case of cluster-randomised controlled trials, where an analysis was reported that accounted for the clustered study design, we estimated the effect on this basis. Where this was not possible and the information was not available from the authors, then we carried out an 'approximately correct' analysis according to current guidelines (Higgins 2011a). We imputed estimates of the intra-cluster correlation (ICC) using estimates derived from similar studies included in the review. We also computed inflated standard errors for outcome data from cluster-randomised controlled trials based on reported test statistics (f values, t values or P values) and used these data in all statistical analyses. Where test statistics were not available, we imputed inflated standard errors from unadjusted standard errors based on ratios of adjusted to unadjusted standard errors obtained from similar studies included in the review.

For included studies with a within-subjects design, we calculated the standardised mean difference for continuous outcomes using the methods described in Section 16.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). Similar to our approach for cluster-randomised controlled trials, we sought to compute deflated standard errors for outcome data from studies with a within-subjects design based on reported test statistics, or on ratios of inflated to unadjusted standard errors obtained from similar studies included in the review. However, in studies with a within-subjects design, these ratios exceeded one, which is counter-intuitive and suggests there was no statistical advantage in using within-subjects designs in this area. We therefore reverted to use of unadjusted standard errors for studies with a within-subjects design in all statistical analyses.

Final outcome values served as the primary unit of analysis. Only one included study reported outcome data using changes from baseline as the metric (Ahn 2010). For this study we computed final values based on reported data, supplemented with additional information supplied by the authors.

Dealing with missing data

Where data were missing due to participant dropout we conducted available case analyses and recorded any issues of missing data within the assessments conducted using the Cochrane 'Risk of bias' tool.

Assessment of heterogeneity

We assessed statistical heterogeneity in results by inspection of a graphical display of the estimated treatment effects from included studies along with their 95% confidence intervals, and by formal statistical tests of homogeneity (Chi2) and measures of inconsistency (I2) and heterogeneity (τ2).

Assessment of reporting biases

We drew funnel plots (plots of effect estimates versus the inverse of their standard errors) to inform assessment of reporting biases. We conducted statistical tests to formally investigate the degree of asymmetry using the method proposed by Egger et al (Egger 1997). We interpreted the results of statistical tests based on visual inspection of the funnel plots. Asymmetry of the funnel plot may indicate publication bias or other biases related to sample size, though it may also represent a true relationship between trial size and effect size.

Data synthesis

We described and summarised the findings of included studies to address the two stated objectives of the review. We provide a narrative synthesis describing the interventions, participants, study characteristics and effects of eligible interventions upon pre-specified outcomes (see Criteria for considering studies for this review).

Our statistical analysis of the results of included studies used a series of random-effects and fixed-effect models to estimate summary effect sizes as SMDs with 95% confidence intervals. We determined the final configuration of our statistical analysis based on the final version conceptual model (Figure (Figure1).1). We conducted the statistical analysis using STATA (StataCorp, College Station, TX, 2014) and it comprised the following stages:

Stage 1. A standard meta-analysis to estimate summary effect sizes for all eligible interventions versus all comparators, using metan (Harris 2008).

Stage 2. A meta-regression analysis with type of product (food, alcohol, tobacco) as a covariate.

Stage 3. A meta-regression analysis with study characteristics as additional covariates.

Stage 4. A meta-regression analysis with intervention characteristics as covariates. At the protocol stage, we considered the option of conducting multivariate analysis to deal with studies with multiple treatment arms in order for direct comparisons between each treatment arm and a control condition to be modelled, using mvmeta (White 2011). In practice, we did not judge this appropriate and we conducted all meta-regression analyses using metareg (Harbord 2008).

Stage 5. A meta-regression analysis with participant characteristics and 'Risk of bias' assessment as covariates.

We only incorporated outcome data from independent comparisons into the statistical analysis. For example, from an included study that measured energy consumed from a lunch meal in four groups of participants served with a 275 g, a 367 g, a 458 g or a 550 g sandwich (Rolls 2004a), available pairwise comparisons are: 275 g versus 367 g, 275 g versus 458 g, 275 g versus 550 g, 367 g versus 458 g, 367 g versus 550 g, and 458 g versus 550 g. However, since these comparisons are not independent from one another, only the incremental comparisons (which are independent) were incorporated: 275 g versus 367 g, 367 g versus 458 g, and 458 g versus 550 g. Our decision to incorporate only outcome data from incremental comparisons into the statistical analysis effectively assumes a linear 'dose-response' relationship between portion size and consumption/selection for portions of the sizes investigated in included studies. This assumption was judged reasonable by topic expert members of the review team and it is also conservative in terms of its impact on estimates of summary effect sizes. Some groups of study participants feature in two incremental comparisons (e.g. the 367 g group features in both the 275 g versus 367 g comparison and the 367 g versus 458 g comparison), therefore we halved sample sizes for groups featuring in two incremental comparisons to adjust their weighting in the analysis for this non-independence.

Preliminary examination of outcome data revealed substantive variation in effect sizes between comparisons identified from studies that manipulated portion, package, individual unit or tableware size and those identified from studies that manipulated tableware shape. We did not judge comparisons of size conceptually comparable to comparisons of shape among the set of studies included in this review: size comparisons consisted in larger versus smaller sizes (of a portion, package, individual unit or item of tableware), whilst shape comparisons consisted in shorter, wider versus taller, narrower glasses or bottles (tableware). We therefore took the post-hoc decision to conduct separate meta-analyses for size and shape respectively, for both consumption and selection outcomes. (This decision effectively removed the covariate that differentiated between size and shape manipulations from subsequent meta-regression analyses - see below and Table 1). Preliminary analyses also revealed substantive variation in effect sizes between those measured in children and those measured in adults (as well as variation in effect sizes between adults of different ages), and between comparisons involving food products and those involving tobacco products. We therefore estimated supplementary summary effect sizes for these subgroups to illustrate these variations in effects. In describing the effects of size and shape interventions on selection and consumption, our narrative synthesis is disaggregated as appropriate to reflect these variations and to incorporate supplementary effect sizes estimated to illustrate them (see Effects of interventions).

We used the following procedures for meta-regression analyses. First, for each of the two outcomes (consumption and selection), we conducted a series of univariable analyses using random-effects models to test for a statistical association between each covariate and the study-level effect size (SMD). All variables identified in the final version of the conceptual model (see Table 1) were candidate covariates for univariable analyses. Blinded to data extracted for covariates from study reports by two researchers (GJH, IS), topic experts within the review team selected six baseline participant characteristics to be prioritised when contacting study authors to request data on potential effect modifiers that appeared to have been measured but were missing from study reports. This selection was based on what were expected to be the most important modifiers of the effects of the intervention, primarily based on topic experts' knowledge of theory and evidence for determinants of between-person variation in levels of food and energy intake (since the majority of studies included in this review focused on food - see Description of studies). The six selected covariates (variable type) were: age (continuous), gender (categorical), BMI (continuous), dietary restraint (continuous), dietary disinhibition (continuous) and hunger (continuous). All six had been pre-specified in the original version of the conceptual model (Figure (Figure1)1) and had been measured at baseline in at least one included study. We decided in advance of conducting univariable meta-regression analyses that candidate covariates would be excluded if they had been measured in fewer than 10 independent comparisons feeding into an analysis (insufficient data) or if there was no variation in the value of the covariate between independent comparisons feeding into an analysis (absence of variation, which precluded estimation). Based on these exclusion criteria, we conducted two series of univariable meta-regression analyses to investigate potential modifiers of the effects of larger versus smaller portions, packages, individual units or tableware on: (a) consumption of food and tobacco; and (b) on the selection (without purchase) of food. We did not conduct other planned series of univariable meta-regression analyses due to insufficient data following application of the exclusion criteria outlined above.

Second, we estimated random-effects models to identify the collections of study-level covariates that best explained the between-studies component of the variance in study-level estimates of effect size. As with univariable analyses, it proved possible in practice to implement this analysis to investigate potential modifiers of the effects of larger versus smaller portions, packages, individual units or tableware on: (a) consumption of food and tobacco; and (b) on the selection (without purchase) of food. We did not conduct other planned second stage analyses due to insufficient data. We selected variables for inclusion in models using a stepwise forward selection procedure. We selected first the covariate which had the largest value of R2 (a measure of the proportion of the between-studies component of the variance explained by the model) based on the results of the preceding series of univariable analyses. Next, we added each of the other covariates observed to be statistically associated with the study-level effect size in the results of the preceding univariable analyses to the model in sequence (in an order corresponding to Stages 2 to 4 of the statistical analysis plan, outlined above in this section). Each covariate was retained in the final model if its incorporation contributed to an increase in the value of the R2 but was otherwise dropped from the model. Consequently, once this procedure was completed, the final model specification maximised the value of R2.

To facilitate interpretation of estimated effect sizes (Schünemann 2011), we re-expressed a series of SMD values ranging between 0.1 and 2.5 in terms of selected metrics of food or tobacco selection/consumption. Baseline values (SMD = 0.0) reflect estimated average (mean) consumption levels among representative samples of UK adults or children and associated among-participant variation (that is, the standard deviation). Two researchers (IS and HBL) estimated average (mean) food energy intake, non-alcoholic beverage consumption and cigarette consumption (among smokers) using unweighted data from the UK National Diet and Nutrition Survey Years 1-4, collected using 24-hour dietary recall in a nationally representative UK population sample (National Centre for Social Research 2012). One researcher (IS) also estimated an alternative estimate of average cigarette consumption (among smokers) based on unweighted data from the UK Opinions and Lifestyle Survey 2012 (Office for National Statistics 2012). We used these data to re-express SMD values in terms of the proportionate (%) and absolute changes from baseline values in terms of each selected metric and tabulated these data for illustrative purposes (see Effects of interventions). We also compared re-expressed values among UK adults and children to those based on published estimates among equivalent US samples.

'Summary of findings' table

We used the standard GRADE system to rate the quality of the respective bodies of evidence for (1) consumption and (2) selection (with or without purchasing) outcomes in terms of the extent of our confidence in (summary) estimates of effects. GRADE criteria for assessing quality of evidence encompass study limitations, inconsistency, imprecision, indirectness, publication bias and other considerations. We recorded the justifications underpinning these assessments. We present this information in a series of 'Summary of findings' tables developed using GRADEpro GDT (Brozek 2008), alongside a summary of the estimated intervention effect and details of the numbers of studies (independent comparisons) and participants that underpinned each estimate. Our decision to present a series of 'Summary of findings' tables rather than a single table reflects our decisions to conduct separate meta-analyses for size and shape respectively (for both consumption and selection outcomes) and to present separate summary effect sizes for food products and tobacco products (see above in this section - in both cases preliminary examination of outcome data had identified substantial variation in effect sizes between studies with these variant characteristics). Separate 'Summary of findings' tables are therefore presented to summarise evidence for the (differential) effects of exposure to larger-sized portions, packages and tableware (by product - food and tobacco) and exposure to differently shaped tableware (by product - food only). Within each 'Summary of findings' table, findings are grouped by outcome (consumption and selection). In addition to presenting the overall summary effect size for each outcome, we also present disaggregated summary effect sizes for subgroups of studies involving children and adults respectively (again, due to identified variation in effect sizes between those measured in children and those measured in adults - see above in this section).

Sensitivity analysis

We conducted a sensitivity analysis to explore the impact of outcome data imputed due to missing data. In practice, standard deviations were the only component of outcome data that needed to be imputed for some independent comparisons due to missing data. Therefore, this sensitivity analysis in practice involved re-estimating fixed-effect and random-effects meta-analyses (for both selection and consumption outcomes – all comparisons) using imputed values for standard deviations that were (1) double and (2) half those used in the 'base case' analyses reported in the Effects of interventions section. At the protocol stage, we had also planned to conduct a sensitivity analysis to explore the separate analysis of studies of food and tobacco products. In practice, we estimated supplementary summary effect sizes for these subgroups of studies (see Data synthesis), which was functionally equivalent to this planned sensitivity analysis.

Results

Description of studies

Results of the search

The flow of studies through the systematic review process is shown in Figure Figure2.2. Electronic database searches retrieved a total of 76,279 study records, including duplicates. Searches of other resources identified 23 additional study records not retrieved by electronic database searches, comprising 15 records identified by searching reference lists of eligible study reports or forward citation tracking and eight records identified within our preceding, broader scoping review (Hollands 2013a). Automatic and manual de-duplication identified 24,624 duplicate records, which we discarded. Therefore, 51,655 unique records entered title/abstract screening. Of these, we excluded 51,472 records and obtained corresponding full-text study reports for the remaining 183 records assessed as potentially eligible.

Figure 2
PRISMA study flow diagram.

We excluded 101 study reports based on full-text screening. Primary reasons for exclusion are summarised in Figure Figure22 (PRISMA flow diagram) and in the Characteristics of excluded studies table. A further four full-text study reports were conference abstracts with insufficient information to enable confident assessment of eligibility (Loney 2010, Martinez 2010, Schmidt 2013, Skov 2013). Brief details of these four studies are provided in Characteristics of studies awaiting classification tables. Therefore, following exclusions, identification and linking of multiple eligible study reports of the same study and identification of study reports comprising multiple eligible studies, we have identified a total of 83 studies as meeting the eligibility criteria for this review (from 78 full-text study reports). The number of included studies exceeds the number of included study reports due to the comparative incidences of study reports that report multiple studies (i.e. two or more studies reported in the same publication) and studies reported in single or multiple study reports among studies/reports that we identified as meeting eligibility criteria for this review.

Eligible studies included in the review

Seventy-two of the 83 eligible studies (66 study reports) were identified by the original search initiated in November 2012 (see Search methods for identification of studies). These 72 studies, published between 1978 and July 2013, are described in the Included studies section below (with further details of each study provided in Characteristics of included studies tables) and are recorded as 'studies included in the review' in Figure Figure2.2. All remaining sub-sections of the Results section of the current version of this review (i.e. Included studies, Excluded studies, Risk of bias in included studies and Effects of interventions), as well as its Discussion and Authors' conclusions sections, are based exclusively on evidence collected from these 72 included studies. We sought to establish contact with authors of 36 of 72 included studies to request data missing from study reports (Argo 2012 (S5); Burger 2011; Cavanagh 2013; Coelho do Vale 2008 (S2); DiSantis 2013; Fisher 2013; Flood 2006; Goldstein 2006; Jeffery 2007; Kral 2004a; Kral 2010; Levitsky 2004; Marchiori 2012a; Marchiori 2012c; Mishra 2012 (S1); Mishra 2012 (S2); Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2007b (S1); Rolls 2007b (S3); Rolls 2010a (E1); Rolls 2010b (E2); Russell 1980; Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); Spill 2010; Spill 2011b; Wansink 1996a (S1); Wansink 2001; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2011a (S4)). We were able to establish contact with authors of 32 of these 36 studies (Burger 2011; Cavanagh 2013; Coelho do Vale 2008 (S2); DiSantis 2013; Fisher 2013; Flood 2006; Jeffery 2007; Kral 2004a; Kral 2010; Levitsky 2004; Marchiori 2012a; Marchiori 2012c; Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2007b (S1); Rolls 2007b (S3); Rolls 2010a (E1); Rolls 2010b (E2); Russell 1980; Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); Spill 2010; Spill 2011b; Wansink 1996a (S1); Wansink 2001; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2011a (S4)), of which 20 supplied the requested information (Burger 2011; Cavanagh 2013; Coelho do Vale 2008 (S2); DiSantis 2013; Flood 2006; Kral 2010; Levitsky 2004; Marchiori 2012a; Marchiori 2012c; Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2007b (S1); Rolls 2007b (S3); Rolls 2010a (E1); Rolls 2010b (E2); Spill 2010; Spill 2011b). Including data supplied by study authors, 70 of 72 included studies provided useable data for meta-analyses (104 independent comparisons) - the exceptions were the studies by Argo 2012 (S5) and Goldstein 2006.

Eligible studies accepted into the review and awaiting full integration

The other 11 of the 83 eligible studies (12 study reports) were identified by the updated search (30 January 2015) (Bajaj 2014; Haire 2014; Kral 2014; Marchiori 2014; Rolls 2014a; Smith 2013a; van Ittersum 2013; van Kleef 2014; Wansink 2013; Wansink 2014; Williams 2014). These 11 studies, published during 2013 and 2014, are described in Characteristics of studies awaiting classification tables and are recorded as 'studies accepted into the review and awaiting full integration' in Figure Figure2.2. As well as describing key characteristics of each of these 11 further eligible studies, the Characteristics of studies awaiting classification tables also include provisional study-level effect sizes (SMDs and 95% CIs) computed based on useable data provisionally extracted from 12 corresponding study reports.

It was important to establish whether the full integration of these 11 eligible studies could change the interpretation of the results of this review, and hence its conclusions, as reported below in Results, Discussion and Authors' conclusions. We therefore conducted preliminary analyses to investigate this issue using outcome data that could provisionally be extracted from each of the 11 further eligible studies. These preliminary analyses are summarised in Appendix 2. Their results establish that there is minimal potential for full integration of these 11 studies to change the interpretation of the results of this review, and hence its conclusions, as reported below in Results, Discussion and Authors' conclusions. On this basis we took the pragmatic decision (in consultation with the Cochrane Public Health Review Group) to defer full integration of these 11 studies until the first major update of this review. Therefore, as highlighted above, all results and findings presented in the remainder of the main text of this review are based exclusively on evidence collected from the 72 included studies identified by the original search up to and including 20 November 2012.

Included studies

The majority of the 72 included studies were conducted in the USA (58 of 72), with five studies from Canada (Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Argo 2012 (S5); Koh 2009), three from Belgium (Marchiori 2011; Marchiori 2012a; Marchiori 2012c), two from the Netherlands (Coelho do Vale 2008 (S2); Hermans 2012), two from the UK (Kelly 2009; Russell 1980), and one study each from Australia (Cavanagh 2013) and South Korea (Ahn 2010). We identified no eligible studies conducted in low- or middle-income countries (LMICs). The majority of included studies were conducted in laboratory settings (50 of 72) and the others (22 of 72) were conducted in field settings - predominantly restaurants or school or workplace cafeterias (Ahn 2010; Diliberti 2004; DiSantis 2013; Ebbeling 2007; Huss 2013; Jeffery 2007; Leahy 2008; Looney 2011; Marchiori 2012c; Mishra 2012 (S1); Raynor 2007; Raynor 2009; Russell 1980; Spill 2010; Spill 2011b; Stroebele 2009; Wansink 2001; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005b; Wansink 2006; Wansink 2011b).

Study participants were adults (16 years or more) in 55 of 72 studies (predominantly younger adults aged 19 to 30 years), children in 16 studies (predominantly younger children aged three to six years) (DiSantis 2013; Ebbeling 2007; Fisher 2003; Fisher 2007b; Fisher 2007c; Fisher 2013; Huss 2013; Kral 2010; Leahy 2008; Looney 2011; Marchiori 2012c; Mathias 2012; Rolls 2000; Spill 2010; Spill 2011b; Wansink 2003 (S1)), and both adults and children in one study (Fisher 2007a). In the median study, participants' mean age was 22.2 years (Rolls 2002), ranging between 2.6 years (Fisher 2007c) and 55.2 years (Ahn 2010). Data on the sex of participants was available in 65 of 72 studies. The median study included 55% female participants, ranging from 0% to 100% female (interquartile range (IQR): 49 to 84). Seventy of 72 studies were conducted in low deprivation contexts, whilst the other two were conducted in high deprivation contexts (DiSantis 2013; Fisher 2007a).

In the median studies, participants' mean body mass indexes (BMIs) were 23.5 (Flood 2006; Raynor 2007) and, across all included studies, mean BMI ranged between 17.0 (Kral 2010) and 34.0 (Fisher 2007a). Mean dietary restraint score (Stunkard 1985) in the median studies was 5.8 (Flood 2006, Rolls 2006a), with a range of 4.3 (Raynor 2007) to 9.8 (Burger 2011), while mean dietary disinhibition score (Stunkard 1985) in the median studies was 4.3 (Rolls 2007b (S1); Rolls 2007b (S2)), with a range of 3.5 (Rolls 2002) to 5.3 (Burger 2011; Kral 2004a). Mean baseline hunger score (Stunkard 1985) in the median study was 4.5 (Flood 2006), with a range of 3.6 (Rolls 2007a) to 5.6 (Rolls 2004b). These results suggest that included studies examined effects in participants who were mainly unrestrained eaters (Stunkard 1985).

Sixty-nine of 72 studies involved manipulations of food products, with the other three focused on tobacco (Jarvik 1978 (E1); Jarvik 1978 (E2); Russell 1980). No eligible studies of alcohol products were identified. The target of manipulation was the portion size in 35 of 72 studies (Burger 2011; Cavanagh 2013; Diliberti 2004; Fisher 2003; Fisher 2007a; Fisher 2007b; Fisher 2007c; Flood 2006; Goldstein 2006; Hermans 2012; Huss 2013; Jeffery 2007; Kelly 2009; Kral 2004a; Kral 2010; Leahy 2008; Levitsky 2004; Looney 2011; Mathias 2012; Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2006b; Rolls 2007a; Rolls 2010a (E1); Rolls 2010b (E2); Spill 2010; Spill 2011b; van Kleef 2013; Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2001; Wansink 2005b). In 10 studies the target of manipulation was the package size (Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Argo 2012 (S5); Coelho do Vale 2008 (S2); Ebbeling 2007; Raynor 2009; Stroebele 2009; Wansink 1996a (S1); Wansink 2011a (S4)), in six studies it was the size of individual units of a product (including in the three included tobacco studies, which all manipulated the length of cigarettes) (Devitt 2004; Jarvik 1978 (E1); Jarvik 1978 (E2); Marchiori 2011; Marchiori 2012c; Russell 1980), and in 15 studies it was the size or shape of tableware (Ahn 2010; DiSantis 2013; Koh 2009; Mishra 2012 (S1); Mishra 2012 (S2); Rolls 2007b (S1); Rolls 2007b (S2); Rolls 2007b (S3); Shah 2011; van Kleef 2012; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005d; Wansink 2006; Wansink 2011b). One study incorporated separate manipulations of both portion size and tableware size (Fisher 2013), and two studies incorporated separate manipulations of both portion size and package size (Marchiori 2012a; Raynor 2007). Three studies incorporated concurrent manipulations of package size and individual unit size, applied simultaneously and were therefore inherently confounded (Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4)).

Sixty-nine of 72 studies manipulated size, whilst the other three manipulated shape (Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005d). Among studies that manipulated size, the larger of the two compared portions, packages, individual units or items of tableware was, on average (median) 167% (IQR: 140 to 200) of the size of the smaller version, and the mode was 200%. The larger of the two compared portions, packages, individual units or items of tableware was 200% of the size of the smaller version in one-third of included food studies (independent comparisons) and fell between 120% and 159% in half of the included food studies, indicating a bimodal distribution. Absolute sizes investigated in included food studies also tended to be large compared with reference portion sizes (defined here as the size that is recommended to be consumed, or that is customarily consumed, in a single eating occasion, by one or more schemes for communicating portion size messages to consumers (Lewis 2012)) derived from a published report on typical portion sizes in the UK in 2002 (Food Standards Agency 2002). For example, the pairs of portion, package or individual unit sizes compared within included food studies both exceeded the reference portion size in 81% (34 of 42) of those independent comparisons for which these data were available and applicable (42 of 86), whilst only 5% (2 of 42) compared a (larger) portion that was 100% of the reference portion size with a (smaller) portion that was < 100% of the reference portion size (Food Standards Agency 2002). Reference portion sizes could not be coded for approximately half of the pairs of food product sizes compared within included studies (44 of 86) due to them manipulating tableware (for example, DiSantis 2013), or multiple products simultaneously (for example, Kelly 2009), or due to missing data.

Further details on characteristics of interventions and comparators are provided in Characteristics of included studies.

Consumption outcomes only were reported in 59 of 72 included studies (Ahn 2010; Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Argo 2012 (S5); Burger 2011; Cavanagh 2013; Coelho do Vale 2008 (S2); Devitt 2004; Diliberti 2004; Ebbeling 2007; Fisher 2007a; Fisher 2007b; Fisher 2007c; Flood 2006; Goldstein 2006; Hermans 2012; Huss 2013; Jarvik 1978 (E1); Jarvik 1978 (E2); Jeffery 2007; Kelly 2009; Kral 2004a; Kral 2010; Leahy 2008; Levitsky 2004; Looney 2011; Marchiori 2011; Marchiori 2012a; Marchiori 2012c; Mathias 2012; Mishra 2012 (S1); Mishra 2012 (S2); Raynor 2007; Raynor 2009; Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2006b; Rolls 2007a; Rolls 2007b (S1); Rolls 2007b (S2); Rolls 2007b (S3); Rolls 2010a (E1); Rolls 2010b (E2); Russell 1980; Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); Shah 2011; Spill 2010; Spill 2011b; Stroebele 2009; van Kleef 2013; Wansink 2001; Wansink 2005b; Wansink 2011b). Selection outcomes only were reported in seven other studies (Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2006; Wansink 2011a (S4)), whilst both selection and consumption outcomes were reported in six other studies (DiSantis 2013; Fisher 2003; Fisher 2013; Koh 2009; van Kleef 2012; Wansink 2005d). Outcomes were measured objectively rather than by participant self report in almost all included studies with two exceptions (Ahn 2010; Jeffery 2007), and were typically measured over a period of one day or less (60 of 72 studies). Those studies that measured outcomes over a period exceeding one day were Ahn 2010, Fisher 2013, Huss 2013, Jeffery 2007, Kelly 2009, Raynor 2007, Raynor 2009, Rolls 2006a, Rolls 2006b, Rolls 2007a, Russell 1980 and Stroebele 2009.

In line with the eligibility criteria, all 72 included studies were randomised controlled trials (see Types of studies). Thirty-eight had a within-subjects (cross-over) design (Burger 2011; Devitt 2004; DiSantis 2013; Ebbeling 2007; Fisher 2003; Fisher 2007a; Fisher 2007b; Fisher 2007c; Fisher 2013; Flood 2006; Huss 2013; Jarvik 1978 (E1); Jarvik 1978 (E2); Jeffery 2007; Kelly 2009; Kral 2004a; Kral 2010; Leahy 2008; Levitsky 2004; Looney 2011; Mathias 2012; Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2006b; Rolls 2007a; Rolls 2007b (S1); Rolls 2007b (S2); Rolls 2007b (S3); Rolls 2010a (E1); Rolls 2010b (E2); Russell 1980; Shah 2011; Spill 2010; Spill 2011b; Stroebele 2009), and the remaining 34 had a between-subjects (parallel-group) design (Ahn 2010; Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Argo 2012 (S5); Cavanagh 2013; Coelho do Vale 2008 (S2); Diliberti 2004; Goldstein 2006; Hermans 2012; Koh 2009; Marchiori 2011; Marchiori 2012a; Marchiori 2012c; Mishra 2012 (S1); Mishra 2012 (S2); Raynor 2007; Raynor 2009; Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); van Kleef 2012; van Kleef 2013; Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2001; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005b; Wansink 2005d; Wansink 2006; Wansink 2011b; Wansink 2011a (S4)). There was no evidence of funding of included studies by agencies that may have commercial interests in their results.

Excluded studies

We excluded 81 of 149 study reports identified by the original search from this review at the full-text screening stage. We further excluded 20 of 34 study reports identified by the updated search at the full-text screening stage. Details of the combined total of 101 excluded study reports (of 183 screened in full-text) are provided in Characteristics of excluded studies, along with the primary reason for exclusion in each case (in two cases - Just 2014 and Scisco 2012 - the excluded study report comprised two ineligible studies (denoted as S1 and S2 in Characteristics of excluded studies tables), both excluded).

The most common reasons for exclusion were the lack of an eligible intervention, and the lack of an eligible study design. Illustrative examples of studies with no eligible intervention include Attwood 2012, in which participants were instructed to drink all of the product presented to them, rather than the quantity that they freely chose to drink. Bohnert 2011 examined the effects of using a specially designed plate (which gave visual indications of suggested portion size) versus a plain plate. There was no difference in the size or shape of the different plates, and the only difference was in its surface design, therefore there was no eligible intervention.

Illustrative examples of studies with an ineligible study design include Leidy 2010, in which participants were not randomly assigned between the two portion size conditions. The comparison was between two different experiments, as confirmed by correspondence with the senior author. Freedman 2010 again did not randomly assign participants, but instead appeared to report a study with a case series or uncontrolled longitudinal design.

Risk of bias in included studies

Following the procedures outlined in Assessment of risk of bias in included studies, we made a summary 'Risk of bias' assessment for each outcome. We classified seven studies from the 65 that measured consumption as at overall high risk of bias with respect to this outcome (Ahn 2010; Diliberti 2004; Goldstein 2006; Huss 2013; Mishra 2012 (S1); Raynor 2009; Wansink 2005d), with the remaining 58 studies classified as at overall unclear risk of bias. We classified nine of the 13 studies that measured selection (without purchase) as at overall unclear risk of bias with respect to this outcome (DiSantis 2013; Fisher 2003; Fisher 2013; Koh 2009; van Kleef 2012; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2006; Wansink 2011a (S4)), with four at high risk of bias (Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2005d).

Decisions regarding individual domains within the Cochrane 'Risk of bias' tool are summarised below. Figure Figure33 summarises risk of bias judgements across included studies and full details of review authors' judgements and support for judgements are provided for each study in 'Risk of bias' tables in Characteristics of included studies.

Figure 3
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all eligible studies (N = 83. 'Risk of bias' assessments completed for 72 eligible studies included in the review. White spaces in the bars of ...

Allocation

We judged the risk of allocation bias due to the procedures used to generate a randomised sequence of assignments to be unclear in 59 of 72 studies because insufficient information was provided about these procedures to permit a judgement of low or high risk. We judged the risk of bias from this source to be low in 10 studies (Ahn 2010; Ebbeling 2007; Looney 2011; Raynor 2009, Russell 1980; Spill 2010; Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2005d) and high in the remaining three studies (Goldstein 2006; Huss 2013; Mishra 2012 (S1)).

We judged risk of bias due to procedures used to conceal the allocation sequence from those involved in the enrolment and assignment of participants to be unclear in 58 studies, again due to insufficient information to permit a judgement of low or high risk. We judged risk of bias from this source to be low in five studies (DiSantis 2013; Ebbeling 2007; Huss 2013; Mathias 2012; Wansink 2011b), and high in the other nine studies (Ahn 2010; Diliberti 2004; Goldstein 2006; Mishra 2012 (S1); Raynor 2009; Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2005d).

Blinding

Blinding of participants and personnel

Among the 13 studies that reported selection outcomes, we judged risk of bias to be unclear in this domain due to insufficient information in eight studies (DiSantis 2013; Fisher 2003; Fisher 2013; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005d; Wansink 2006; Wansink 2011a (S4)), and low in the remaining five studies (Koh 2009; van Kleef 2012; Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4)).

Among the 65 studies that reported consumption outcomes, we judged risk of bias to be high in this domain in one study (Ahn 2010), low in 20 studies (Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Argo 2012 (S5); Cavanagh 2013; Coelho do Vale 2008 (S2); Goldstein 2006; Hermans 2012; Koh 2009; Marchiori 2011; Marchiori 2012a; Marchiori 2012c; Raynor 2007; Raynor 2009; Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); van Kleef 2012; van Kleef 2013; Wansink 2011b), and unclear due to insufficient information in the remaining 44 studies.

Blinding of outcome assessment

We judged all 13 studies that reported selection outcomes to be at low risk of bias in this domain (DiSantis 2013; Fisher 2003; Fisher 2013; Koh 2009; van Kleef 2012; Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005d; Wansink 2006; Wansink 2011a (S4)).

Among the 65 studies that reported consumption outcomes, we judged the risk of bias to be high in this domain in one study (Ahn 2010). In this study, we regarded it possible that the outcome measurement may have been influenced by a lack of blinding, because participants were instructed to keep dietary records of their own intake. We judged two other studies to be at unclear risk of bias due to insufficient information (Jeffery 2007; Stroebele 2009). We judged the remaining 62 studies to be at low risk of bias.

Incomplete outcome data

Among the 13 studies that reported selection outcomes, we judged two to be at high risk of bias for this domain (Fisher 2003; Fisher 2013), with the remaining 11 studies judged to be at low risk of bias. Of the 65 studies that reported consumption outcomes, we judged eight to be at high risk of bias (Coelho do Vale 2008 (S2); Fisher 2003; Fisher 2007c; Fisher 2013; Leahy 2008; Looney 2011; Marchiori 2011; Mathias 2012), with four studies assessed as at unclear risk of bias (Mishra 2012 (S1); Mishra 2012 (S2); Rolls 2007a; Russell 1980). We judged the remaining 53 studies as at low risk of bias. We judged studies to be at high risk of bias for this domain if = 10% of participants' data had been excluded from the analysis due to low (or zero) levels of selection or consumption, or due to being outliers.

Selective reporting

We judged 67 of 72 studies to be at unclear risk of bias in this domain. This was determined by searching for record(s) containing details of the study protocol in online trial registries (ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP)) and finding no corresponding records. As such, there was insufficient information to permit judgement of 'low risk' or 'high risk'. We assessed this domain to be at low risk of bias in four studies for which records were found and the comparison of the trial registry entries and published studies confirmed no selective outcome reporting (Ebbeling 2007; Fisher 2007b; Looney 2011; Raynor 2009). We classified one study as being at high risk of bias due to a discrepancy between the trial registry entry and the published study regarding the specified primary outcomes (Raynor 2007).

Other potential sources of bias

We assessed two additional potential sources of bias that we had pre-specified as potentially important for this review: baseline comparability of participant characteristics between groups and consistency in intervention delivery.

Regarding baseline comparability of participant characteristics between groups, we judged 29 studies to be at low risk of bias (Ahn 2010; Burger 2011; Cavanagh 2013; Ebbeling 2007; Fisher 2003; Fisher 2007a; Fisher 2007c; Fisher 2013; Hermans 2012; Huss 2013; Jeffery 2007; Kelly 2009; Koh 2009; Kral 2010; Levitsky 2004; Looney 2011; Marchiori 2011; Marchiori 2012a; Marchiori 2012c; Raynor 2007; Raynor 2009; Rolls 2010a (E1); Rolls 2010b (E2); Russell 1980; Stroebele 2009; van Kleef 2012; van Kleef 2013; Wansink 2005b; Wansink 2011b). We assessed studies as being at low risk of bias in this domain if there were no differences in terms of baseline characteristics between comparison groups (study arms in the case of between-subjects designs and condition orders in the case of within-subjects designs), or where any observed differences in characteristics had been controlled for in the statistical analysis, or were judged by the review team to be unlikely to impact on key outcomes. We judged risk of bias to be high in this domain in the other 43 studies.

Regarding consistency in intervention delivery, we judged one study to be at high risk of bias because the bowl that was being manipulated was placed in a different location and at a different distance from participants in each comparison group (van Kleef 2012). We judged risk of bias unclear in this domain in 31 studies (Burger 2011; Devitt 2004; DiSantis 2013; Ebbeling 2007; Fisher 2003; Fisher 2007b; Fisher 2007c; Fisher 2013; Hermans 2012; Huss 2013; Koh 2009; Kral 2004a; Kral 2010; Levitsky 2004; Looney 2011; Mathias 2012; Mishra 2012 (S2); Raynor 2009; Rolls 2006a; Rolls 2006b; Rolls 2007a; Rolls 2007b (S1); Rolls 2007b (S2); Rolls 2007b (S3); Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); Shah 2011; Spill 2010; Spill 2011b; Stroebele 2009). We judged the remaining 40 studies to be at low risk of bias in this domain since information and instructions appeared to be standardised between comparison groups.

Effects of interventions

See: Summary of findings for the main comparison Food: Larger versus smaller-sized portions, packages or tableware for changing quantity consumed or selected; Summary of findings 2 Alcohol: Larger versus smaller-sized portions, packages or tableware for changing quantity consumed or selected; Summary of findings 3 Tobacco: Longer versus shorter cigarettes for changing quantity consumed or selected; Summary of findings 4 Food: Shorter, wider versus taller, narrower glasses or plastic bottles (shape) for changing quantity of non-alcoholic beverages consumed or selected

This section presents the results of our statistical analyses of outcome data collected from included studies. Results of meta-analyses are presented as standardised mean differences (SMDs) with 95% confidence intervals (CIs). A rule of thumb for interpreting these effect sizes (SMDs) is as follows: 0.2 represents a small effect, 0.5 a moderate effect and 0.8 a large effect (Cohen 1988; Schünemann 2011).

However, it is perhaps more intuitive to interpret SMDs once they have been re-expressed using a familiar metric (Schünemann 2011). Figure Figure44 is intended as an illustrative guide to help readers interpret the estimated effect sizes (SMDs) presented below in this section. Figure Figure44 re-expresses a series of SMD values ranging between 0.1 and 2.5 in terms of selected measures of food or tobacco selection/consumption (for example, 'Equivalent change in average daily energy intake from food (kcal) selected or consumed' in the first column). Baseline values (SMD = 0.0) reflect estimated average (mean) consumption levels among representative samples of UK adults or children (see Data synthesis). For example, mean (standard deviation (SD)) daily energy intake from food among UK adults is estimated to be 1727 (± 537) kcal (National Centre for Social Research 2012). Each column of Figure Figure44 re-expresses SMD values in terms of proportionate (%) and absolute changes from baseline values (reflecting observed among-participant variation in consumption-levels within each corresponding UK sample). For example, a SMD of 0.4 can be re-expressed as equivalent to a 12.4% (215 kcal) increase in average daily energy intake from food, or a 27.2% (67 g) increase in the average single-serve quantity of energy-containing non-alcoholic beverage, or a three to four cigarette increase in the average daily number of cigarettes, selected or consumed by UK adults.

Figure 4
Effect sizes re-expressed using familiar metrics

It is important to use Figure Figure44 judiciously. First, end users of this review should consider the extent to which average (mean) baseline values and SDs reflect consumption patterns in their own country or region. For example, at 1727 (± 537) kcal, estimated mean (SD) daily energy intake from food among UK adults is slightly lower than among US adults with a smaller standard deviation (1834 ± 1013 kcal - Drewnowski 2013). As such, if SMDs were re-expressed based on data for US adults, proportionate (%) and absolute changes from baseline values would be larger than among UK adults (that is, a SMD of 0.4 would be re-expressed as equivalent to a 22.1% (405 kcal) increase in average daily energy intake from food among US adults). Likewise, at 459 ± 370 g, estimated mean (SD) daily consumption of energy-containing non-alcoholic beverages among UK children is lower than daily sugar-sweetened beverage (SSB) consumption among US children, with a smaller standard deviation (551 ± 1257 g - Wang 2009). As such, if SMDs were re-expressed based on US children's data, proportionate (%) and absolute changes from baseline values would again be larger than among UK children (that is, a SMD of 0.2 would be re-expressed as equivalent to a 45.7% (251 g) increase in average daily SSB consumption among US children). Moreover, the inclusion of Figure Figure44 for illustrative purposes does not restrict the applicability of the results of this review to the UK population, nor is it intended to generalise the results to the UK population.

Second, none of the metrics shown in Figure Figure44 were actually measured as outcomes in the studies that were incorporated into meta-analyses presented in this section (and we are not aware of any representative observational studies that include estimates of among-participant variation in any of the specific measures of consumption/selection that were actually used to assess outcomes in these studies). Re-expressing SMDs estimated using meta-analyses as equivalent changes in other metrics therefore makes an implicit assumption that our estimates of effect size are directly transferable to these other metrics. For example, it assumes that the estimated size of the effect of (larger) size on consumption of food - typically measured in included studies of food products as the quantity of food or energy consumed from a single meal (or single course within a meal) - would produce the same size of effect on a person's energy intake over the course of a whole day. It is therefore important to recognise that, whilst Figure Figure44 offers illustrations to help guide interpretation of effect sizes estimated using meta-analyses, it also extrapolates beyond the scope of the outcome data and source studies incorporated into those analyses.

1. Consumption

Ninety-seven comparisons identified from 64 eligible studies assessed the effect of exposure to different sizes or shapes of portions, packages, individual units or tableware on consumption of food or tobacco by exposed participants.

1.1 Effect of larger size on consumption

We conducted a meta-analysis to investigate the effect of exposure to larger size on unregulated consumption. Based on characteristics of the studies it incorporated, this meta-analysis effectively investigated the effect of exposure to larger portions, packages, individual units or tableware on participants' unregulated consumption of food or tobacco. Usable outcome data were available for 92 independent comparisons, involving 6711 participants, identified from 61 eligible food or tobacco studies (Ahn 2010; Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Burger 2011; Cavanagh 2013; Coelho do Vale 2008 (S2); Devitt 2004; Diliberti 2004; DiSantis 2013; Ebbeling 2007; Fisher 2003; Fisher 2007a; Fisher 2007b; Fisher 2007c; Flood 2006; Hermans 2012; Huss 2013; Jarvik 1978 (E1); Jarvik 1978 (E2); Jeffery 2007; Kelly 2009; Koh 2009; Kral 2004a; Kral 2010; Leahy 2008; Levitsky 2004; Looney 2011; Marchiori 2011; Marchiori 2012a; Marchiori 2012c; Mathias 2012; Mishra 2012 (S1); Mishra 2012 (S2); Raynor 2007; Raynor 2009; Rolls 2000; Rolls 2002; Rolls 2004a; Rolls 2004b; Rolls 2006a; Rolls 2006b; Rolls 2007a; Rolls 2007b (S1); Rolls 2007b (S2); Rolls 2007b (S3); Rolls 2010a (E1); Rolls 2010b (E2); Russell 1980; Scott 2008b (S2); Scott 2008c (S3); Scott 2008d (S4); Shah 2011; Spill 2010; Spill 2011b; Stroebele 2009; van Kleef 2012; van Kleef 2013; Wansink 2001; Wansink 2005b; Wansink 2011b).

Random-effects meta-analysis showed a summary mean effect size (SMD) of 0.37 (95% CI 0.29 to 0.45, P value < 0.001), suggesting that exposure to larger-sized portions, packages, individual units or tableware increased the quantities of food or tobacco people consumed and that the relative effect size was small to moderate (Figure (Figure5).5). This result was consistent between random-effects and fixed-effect models with the fixed-effect model generating a SMD of 0.40 (95% CI 0.35 to 0.45). The I2 statistic shows that 58.4% of the total variance in study-level estimates of this effect was due to statistical heterogeneity (variation in true effect sizes across studies) rather than sampling error (chance). This represents substantial heterogeneity. A 95% interval for prediction of an effect in a new study similar to the included studies ranges from SMD -0.21 to SMD 0.96, reflecting effects ranging from a moderate reduction to a large increase in consumption. An Egger test for funnel plot asymmetry did not identify evidence consistent with the presence of publication bias (P value = 0.20) (Figure (Figure66).

Figure 5
Forest plot of the standardised mean difference in unregulated consumption of food or tobacco between participants exposed to larger (intervention) versus smaller (control) sized portions, packages, individual units and/or tableware
Figure 6
Assessing publication bias. Funnel plots including all studies reporting the selection outcome (left) and consumption outcome (right) do not show asymmetry (Egger test P value = 0.20 and P value = 0.18 respectively)

The results of a sensitivity analysis, in which standard deviations imputed for five independent comparisons (five studies: Argo 2012 (S1); Argo 2012 (S2); Argo 2012 (S4); Mishra 2012 (S1); Mishra 2012 (S2)) were (1) doubled and (2) halved (see Sensitivity analysis), indicated that the interpretation of the results of this meta-analysis is not influenced by changes in the values of imputed standard deviations. Summary mean effect sizes (SMDs) estimated for this sensitivity analysis using random-effects models were (1) 0.36 (95% CI 0.28 to 0.44, P value < 0.001) and (2) 0.37 (95% CI 0.29 to 0.46, P value < 0.001), respectively. Corresponding summary mean effect sizes (SMDs) from fixed-effect models were (1) 0.37 (95% CI 0.32 to 0.42) and (2) 0.50 (95% CI 0.45 to 0.54).

Potential modifiers of the effect of larger size on consumption

We conducted a series of meta-regression analyses to investigate the extent to which this substantial heterogeneity could be explained by study-level covariates. Of 71 candidate study-level covariates, 40 were excluded due to either insufficient data (< 10 included studies) or were not estimable due to the absence of variability in data values between studies. Univariable meta-regression analysis results for the 31 remaining study-level covariates are presented in Appendix 3. We observed six of these covariates to be associated with the effect of larger-sized portions, packages, individual units or tableware on the quantities of food or tobacco people consume. Below, we report results from each stage of our meta-regression analyses (as described in the Data synthesis section) and for each stage highlight any variables that we observed to be associated with the intervention effect. We also report on any variables that the review team pre-specified as potential effect modifiers, but which were not observed in our univariable meta-regression analyses to be associated with the intervention effect.

Type of product (food, alcohol, tobacco)
  • Meta-regression analysis did not find evidence that the effect of larger-sized portions, packages, individual units or tableware on consumption differed by the type of product studied (i.e. between food and tobacco products - there were no outcome data for alcohol products). However, based on overall low quality evidence from tobacco studies comprising 108 total participants (effective sample size), exposure to longer versus shorter cigarettes was not found to influence the quantity consumed (SMD 0.25, 95% CI -0.14 to 0.65) in tobacco studies, while moderate quality evidence for a small to moderate effect of exposure to larger versus smaller-sized portions, packages or tableware was found among food studies (SMD 0.38, 95% CI 0.29 to 0.46) based on data collected from 6603 total participants (effective sample size).

Study characteristics
  • Effect sizes were smaller in studies with a within-subjects design than in those with a between-subjects design. Specifically, increases in the amount of food or tobacco consumed by participants exposed to larger-sized portions, packages, individual units or tableware were, on average, 0.40 units smaller (95% CI -0.55 to -0.25) in studies with a within-subjects design than in those with a between-subjects design. Effect sizes for each of these subgroups are presented in Figure Figure7,7, showing that exposure to larger sizes increased consumption among participants in both within-subjects and between-subjects studies.
    Figure 7
    Summary effect sizes (standardised mean differences) in subgroups of studies (consumption outcome)
  • Effect sizes were larger in studies of less healthy food products. Specifically, each 10-point increase in Food Standards Agency (FSA) nutrient profile score corresponded to a 0.06 unit increase (95% CI 0.04 to 0.22) in the amount of additional food consumed as a result of exposure to larger sizes.
  • Effect sizes were larger in studies of more energy-dense food products. Specifically, each one-point increase in energy density score (a component of the FSA nutrient profile score) corresponded to a 0.04 unit increase (95% CI 0.00 to 0.08) in the amount of additional food consumed as a result of exposure to larger sizes.
  • Effect sizes were larger in studies of food products in which the manipulated food(s) comprised all of those available in the study and all were consumed ad libitum than in the other studies of food products. Specifically, increases in the amount of food consumed as a result of exposure to larger sizes were, on average, 0.22 units larger (95% CI 0.02 to 0.41) in studies of food products in which the manipulated food(s) comprised all of those in the study and all were consumed ad libitum than in studies of food products that did not have these characteristics.
  • Effect sizes were larger in studies of food products in which outcome data mapped directly onto the manipulated food(s), as opposed to a wider set of foods including, but not limited to, the manipulated food(s). Specifically, increases in the amount of food consumed as a result of exposure to larger sizes were, on average, 0.32 units larger (95% CI 0.16 to 0.48) in studies of food products in which outcome data mapped directly onto the manipulated food(s) than in studies of food products in which outcome data mapped to a wider set of foods including, but not limited to, the manipulated food(s).
  • Meta-regression analysis did not find evidence that the size of the effect of larger size on consumption was associated with the target of the manipulation (i.e. whether this was a portion, package, individual unit or tableware). Effect sizes for each of these subgroups are presented in Figure Figure7.7. While no evidence was found for an effect of exposure to larger-sized packages and individual units on consumption within the 'package with individual unit' subgroup, this analysis was likely underpowered. We found evidence for this effect in all other subgroups (see Figure Figure77).

Intervention characteristics
  • In meta-regression analysis, we observed neither the absolute nor the relative difference in size between the two portions, packages, individual units or items of tableware being compared to be associated with the effect of larger size on consumption. This pre-planned analysis explored the relationship between relative difference in size and the effect of larger size on consumption using a linear regression that (as can be inferred from the null result) showed no convincing evidence of a linear relationship. On visual examination of the relationship, however, a pattern was apparent, with a bimodal distribution of the variable that captures the relative difference in size (that is, the variable that expresses the larger size as a proportion of the smaller size within each independent pairwise comparison - see also Included studies). We therefore undertook a post-hoc analysis in order to characterise this relationship among studies of food products (that is, limited to independent pairwise comparisons between food portion, package, individual unit or tableware sizes). Specifically, we conducted a meta-analysis to investigate the effect of larger size on consumption among two subgroups of studies (independent comparisons) clustered around each mode of the identified bimodal distribution (see also Included studies): (1) those in which the larger-sized portion, package, individual unit of food or item of tableware was in the range between 120% and 160% of the smaller size; and (2) those in which the larger-sized portion, package or individual unit of food was 200% of the smaller size. This analysis therefore excluded outliers (that is, excluding nine independent comparisons in which the larger-sized portion, package, individual unit of food or item of tableware was = 202% of the smaller size, from Coelho do Vale 2008 (S2), Devitt 2004, Marchiori 2012a, Raynor 2007, Raynor 2009, Shah 2011, van Kleef 2013 and Wansink 2011b - range 243% to 2607%). Summary effect sizes (SMDs), estimated using random-effects models for each subgroup, were: (1) 0.25 (95% CI 0.15 to 0.35), I2 = 22% (based on 39 independent comparisons, 2415 participants); and (2) 0.50 (95% CI 0.31 to 0.69), I2 = 66% (based on 25 independent comparisons, 1414 participants).

Participant characteristics
  • Effect sizes were larger in studies comprising older participants. Specifically, each 10-year increase in the mean age of participants corresponded, on average, to a 0.09 unit increase (95% CI 0.00 to 0.18) in the incremental amount of food or tobacco consumed as a result of exposure to larger sizes. This result is set in the context of overall moderate quality evidence that the effect of exposure to larger size on consumption of food was present among both children (SMD 0.21, 95% CI 0.10 to 0.31 - moderate quality evidence - 1421 participants) and adults (SMD 0.46, 95% CI 0.40 to 0.52 - moderate quality evidence - 5182 participants) - see Figure Figure77 and Table 1. We also identified variation in this effect size between studies comprising adult participants of different ages.
  • We did not observe the following participant characteristics to be associated with the effect of larger size on consumption: gender, BMI, hunger, dietary restraint and dietary disinhibition.

Final regression model

A meta-regression model was estimated to identify the collection of study-level covariates that best explained the between-studies component of the total variance in estimates of the effect of larger sizes on consumption. The final random-effects model explained 91% of the between-studies variance in effect sizes for the consumption outcome (R2 = 90.77%, P value = 0.001), leaving 9% unexplained. This model incorporated the following five covariates, each of which had been identified as a potential modifier of the effect of larger sizes on consumption based on observed associations in univariable meta-regression analyses: study design (within-subjects or between-subjects); FSA 'nutrient profile score'; FSA 'energy density score'; participants' mean age; and a variable differentiating studies of food products in which the manipulated food(s) comprised all of those available in the study and all were consumed ad libitum from other food studies. The variable differentiating food studies, in which outcome data mapped directly onto the manipulated food(s) as opposed to a wider set of foods, was excluded from the final model for two reasons: first, its addition did not increase the adjusted R2 and second, due to its collinearity with the study design covariate (within-subjects or between-subjects). Not all of the five incorporated covariates were independently predictive of effect size (consumption) in the final model. Figure Figure88 comprises three bubble plots that show associations between study-level effect sizes (effect of larger size on consumption) and each of the three continuous variables identified as potential effect modifiers: FSA 'nutrient profile score'; FSA 'energy density score'; and participants' mean age.

Figure 8
Bubble plots. Fitted meta-regression lines showing associations between study-level effect sizes for consumption and study characteristics (continuous variables) identified as effect modifiers: a) FSA score; b) energy density; c) age.

1.2. Effect of shape on consumption

One food study involving 50 adult participants investigated the effect of shape on unregulated consumption (Wansink 2005d). This study investigated the effect of being provided with shorter, wider (versus taller, narrower) empty clear plastic bottles on the quantities of water selected and consumed one hour after vigorous physical activity in a sample of US Army and Marine Reserve Officer's Training Corps students. It reported an effect size (SMD) of 1.17 (95% CI 0.57 to 1.78), assessed as very low quality evidence for a large effect of shorter, wider bottles on quantities of water consumed, given that participants provided with shorter, wider bottles had more water available for consumption than those provided with taller, narrower bottles due to having selected (poured) more in the first place (see Potential modifiers of the effect of shape on selection without purchase, below).

Potential modifiers of the effect of shape on consumption

Investigation of potential modifiers of the effect of shape on consumption was not possible as only one study (comprising one comparison) investigated this effect (Wansink 2005d).

2. Selection

Seventeen comparisons identified from 14 eligible studies assessed the effect of exposure to different sizes or shapes of portions, packages or tableware on quantities of food selected for consumption by exposed participants. No studies investigated this effect in relation to alcohol or tobacco products. None of the 17 comparisons involved purchasing of the food selected for consumption (that is, all measured unregulated selection without purchase).

2.1. Effect of larger size on selection without purchase

We conducted a meta-analysis to investigate the effect of exposure to larger size on unregulated selection without purchase. Based on characteristics of the studies it incorporated, this meta-analysis effectively investigated the effect of exposure to larger-sized portions or tableware on participants' unregulated selection without purchase of food. Usable outcome data were available for 13 comparisons, involving 1164 participants, identified from 10 eligible food studies that we assessed as being at unclear or high risk of bias (DiSantis 2013; Fisher 2003; Fisher 2013; Koh 2009; van Kleef 2012; Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2006; Wansink 2011a (S4).

Random effects meta-analysis showed a mean summary effect size (SMD) of 0.42 (95% CI 0.24 to 0.59, P value = 0.011), providing overall moderate quality evidence that exposure to larger-sized portions, packages, individual units or tableware increased the quantities of food people selected for consumption and that the relative effect size was on average small to moderate (Figure (Figure9).9). This result was consistent between random-effects and fixed-effect models, with the fixed-effect model generating a SMD of 0.40 (95% CI 0.28 to 0.52). The I2 statistic indicated that 53.5% of the total variance in study-level estimates of this effect was due to statistical heterogeneity (substantial heterogeneity). A 95% interval for prediction of an effect in a new study similar to the included studies ranges from SMD -0.14 to SMD 0.97, reflecting effects ranging from a small reduction to a large increase in quantity of food selected. An Egger test for funnel plot asymmetry did not identify evidence consistent with the presence of publication bias (P value = 0.18) (Figure (Figure66).

Figure 9
Forest plot of the standardised mean difference in unregulated selection (without purchase) of food between participants exposed to larger (intervention) versus smaller (control) sized portions, packages and/or tableware

The results of a sensitivity analysis, in which standard deviations imputed for one independent comparison (one study: Wansink 1996c (S4)) were (1) doubled and (2) halved (see Sensitivity analysis), indicated that the interpretation of the results of this meta-analysis is robust to changes in the value of the imputed standard deviation. Summary mean effect sizes (SMDs) estimated for this sensitivity analysis using random-effects models were (1) 0.42 (95% CI 0.23 to 0.60, P value < 0.001) and (2) 0.41 (95% CI 0.25 to 0.58, P value < 0.001) respectively. Corresponding summary mean effect sizes (SMDs) from fixed-effect models were (1) 0.42 (95% CI 0.28 to 0.52) and (2) 0.40 (95% 0.30 to 0.50).

Potential modifiers of the effect of larger size on selection without purchase

We conducted a series of meta-regression analyses to investigate the extent to which this substantial heterogeneity in effect sizes could be explained by study-level covariates. These analyses were limited by low statistical power. Most of the 71 candidate study-level covariates were excluded due to either insufficient data (< 10 included studies) or were not estimable due to the absence of variability in data values between studies. A full set of results of these univariable meta-regression analyses is provided in Appendix 4. Of 15 study-level covariates investigated in these analyses, we observed two to be associated with the effect of larger-sized portions, packages and/or tableware on the quantities of food participants selected for consumption. Below, we report results from each stage of our meta-regression analyses (as described in the Data synthesis section) and for each stage highlight any variables that we observed to be associated with the intervention effect. We also report on any variables that the review team pre-specified as potential effect modifiers, but which were not observed in our univariable meta-regression analyses to be associated with the intervention effect.

Type of product (food, alcohol, tobacco)
  • This was excluded due to absence of variation in product type between included comparisons: all comparisons related to food products.

Study characteristics
  • Effect sizes were smaller in studies with a within-subjects design than in those with a between-subjects design. Specifically, increases in the quantities of food selected as a result of exposure to larger-sized portions or tableware were, on average, -0.41 units smaller (95% CI -0.76 to -0.06) among studies with a within-subjects design than among those with a between-subjects design. Effect sizes for each of these subgroups presented in Figure Figure1010 further indicate that exposure to larger sizes was observed to be associated with increased selection of food among participants in between-subjects studies but not among participants in within-subjects studies.
    Figure 10
    Summary effect sizes (standardised mean differences) in subgroups of studies (selection outcome)
  • Effect sizes were larger in studies of food products in which outcome data mapped directly onto the manipulated food(s), as opposed to a wider set of foods including (but not limited to) the manipulated food(s). Specifically, increases in the quantities of food selected as a result of exposure to larger sizes were, on average, 0.41 units larger (95% CI 0.06 to 0.76) in the former subgroup than in the latter.
  • Meta-regression analysis did not find evidence that the size of the effect of larger size on selection of food was associated with the target of the manipulation (i.e. whether this was a portion or an item of tableware). Effect sizes for each of these subgroups are presented in Figure Figure10,10, which shows that evidence for this effect was found in both studies manipulating portion size (SMD 0.30, 95% CI 0.09 to 0.50) and those manipulating tableware size (SMD 0.51, 95% CI 0.21 to 0.81).

Intervention characteristics
  • In meta-regression analysis, we did not observe the relative difference in size between the two portions or items of tableware being compared to be associated with the effect of larger size on selection without purchase. The potential association between this effect and absolute difference in size could not be investigated due to insufficient data.

Participant characteristics
  • Potential associations between the effect of larger size on selection and the following participant characteristics could not be investigated using meta-regression analysis due to insufficient data: age, BMI, hunger, dietary restraint and dietary disinhibition. We observed no association between this effect and participants' gender. The results of an illustrative analysis presented in Figure Figure1010 indicate that the effect of exposure to larger size on selection of food was present among adults (SMD 0.55, 95% CI 0.35 to 0.75 - moderate quality evidence - 782 participants) but not among children (SMD 0.14, 95% CI -0.06 to 0.34 - low quality evidence - 382 participants) - see also Table 1.

Final regression model

Variation in study design (within-subjects versus between-subjects) alone explained 79% of the statistical heterogeneity observed in the effect of (larger) size on selection of food (R2 = 79.46%), leaving 21% unexplained. The covariate of outcome data mapping directly onto the manipulated food(s) also explained 79% of this statistical heterogeneity (R2 = 78.77%), leaving 21% unexplained. A meta-regression model containing both of these covariates identified as potential effect modifiers could not be estimated due to perfect collinearity. As such the independent effect modifying influences of these two covariates cannot be disentangled. There are at least two plausible complementary explanations for the result that variation in study design explained a large proportion of this statistical heterogeneity. First, all those studies included in the meta-analysis of the effect of larger size on selection that had a within-subjects design measured this effect in children, whilst all those with a between-subjects design measured it in adults. As highlighted above, the results presented in Figure Figure1010 provide an indication that the effect of exposure to larger-sized portions or items of tableware on quantities of food selected was found in studies of adults but not in studies of children. Second, all source studies included in this meta-analysis that had a within-subjects design were conducted by teams from one research centre, as (largely) were source studies that had a between-subjects design.

2.2. Effect of shape on selection without purchase

We conducted a meta-analysis to investigate the effect of shape on unregulated selection. Given the characteristics of studies included in this meta-analysis, it effectively investigated the effect of being provided with shorter, wider empty glasses or plastic bottles on participants' unregulated selection (without purchase) of fruit juices or water in a single, self serve setting. Usable outcome data for this meta-analysis were available for three comparisons, involving 232 participants, identified from three eligible food studies assessed as being at unclear or high risk of bias (Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005d).

Random-effects meta-analysis showed a mean summary effect size (SMD) of 1.47 with wide confidence intervals (95% CI 0.52 to 2.43). This result provides overall low quality evidence that exposure to shorter, wider glasses or plastic bottles increased the quantities of fruit juices or water people selected for consumption and that the relative size of this effect was very large (Figure (Figure11).11). This result was consistent between random-effects and fixed-effect models with the fixed-effect model generating a SMD of 1.39 (95% CI 1.10 to 1.69). Although 95% confidence intervals were wide, the lower bound of 0.52 based on the random-effects model still represents a moderate effect size. The I2 statistic from the random-effects model shows that 90.1% of the total variance in study-level estimates of this effect was due to statistical heterogeneity (considerable heterogeneity).

Figure 11
Forest plot of the standardised mean difference in unregulated selection without purchase of fruit juices or water between participants exposed to shorter, wider (intervention) versus taller, narrower (control) empty glasses or plastic bottles

Potential modifiers of the effect of shape on selection without purchase

We conducted no meta-regression analyses to investigate the extent to which this statistical heterogeneity could be explained by study-level covariates, due to insufficient data. However, it is likely that the considerable between-studies variance in estimates of this effect may be attributable to the influence of variations between the three source studies providing data incorporated into this meta-analysis in terms of their participants, interventions, comparisons and settings. Although Wansink 2003 (S1) and Wansink 2003 (S2) both investigated the effect of being provided with shorter, wider (versus taller, narrower) empty glasses on quantities of fruit juices selected by participants from a cafeteria line for consumption at breakfast, the former investigated this effect in a sample of adolescents (aged 12 to 17 years) attending a six-week health and fitness camp who were motivated as a group to lose weight as well as trained to monitor how much they consumed, whilst the latter investigated the effect in a convenience sample of adults attending a weekend camp on jazz improvisation. The third source study, Wansink 2005d, investigated the effect of being provided with shorter, wider (versus taller, narrower) empty clear plastic bottles on the quantities of water selected for consumption one hour after vigorous physical activity in a sample of US Army and Marine Reserve Officer's Training Corps students. The study conducted in children, Wansink 2003 (S1), comprised 96 participants and found a SMD of 2.31 (95% CI 1.79 to 2.83 - low quality evidence), whilst the estimated summary effect size in the subgroup of two studies conducted in adults, Wansink 2003 (S2) and Wansink 2005d, comprising 136 participants, was SMD 1.03 (95% CI 0.41 to 1.65 - low quality evidence).

ADDITIONAL SUMMARY OF FINDINGS [Explanation]

Table thumbnail

Table thumbnail

Table thumbnail

Discussion

Summary of main results

Main effects of size and shape on consumption and selection

Size

A clear finding of this review is that people exposed to larger-sized portions, packages, individual units or tableware consistently consumed larger quantities of food compared with those exposed to smaller sizes. We rated the overall quality of evidence for a small to moderate effect of portion, package, individual unit or tableware size on food consumption among both children and adults as moderate. This quality rating confers confidence that the true effect is likely to be close to the estimated effect size (that is, small to moderate), but leaves open the possibility that it may be substantially different.

If sustained across the whole diet, the summary effect size attributable to these differences in product size would be equivalent to an absolute change in average daily energy intake from food (that is, energy intake from food and non-alcoholic beverages, but excluding energy intake from alcoholic beverages and dietary supplements) of 215 to 279 kcal among UK adults (a 12% to 16% change from a baseline of 1727 kcal per day - see Figure Figure4)4) (National Centre for Social Research 2012). Sustained reductions in daily energy intake from food of this size would have the potential to make meaningful contributions to the prevention and treatment of major risk factors for non-communicable diseases. For example, 10-year weight gain between 1999 and 2009 among adults in England (that is, 9 kg at the 90th percentile) has been estimated to be equivalent to extra energy intake of around 24 kcal per day over the same period (Department of Health 2011). Any sustained reductions in daily energy intake exceeding this level are therefore likely to be effective in helping to prevent further weight gain in the population (Department of Health 2011). In relation to the treatment of obesity, the UK National Institute for Health and Care Excellence recommends that adults should lose no more than 0.5 to 1 kg (1 to 2 lb) a week (NICE 2014). This rate of weight loss equates to an energy deficit of 500 to 1000 kcal per day. Although this target energy deficit is some way beyond the effect sizes that could feasibly be achieved by interventions to reduce portion size alone (based on our summary estimate of this effect among studies included in the review), our result suggests that interventions of this kind could meaningfully contribute to helping patients achieve such a target if their effects were sustained. Whilst these illustrations highlight the promise of interventions to reduce exposure to larger portion sizes, it is important to highlight that the sustainability of effects remains to be established, since studies included in this review were limited to the investigation of one-off or repeated exposures over short time periods (see also Implications for practice and Implications for research). Moreover, very few studies included in this review investigated effects among samples of participants motivated to lose weight, further limiting inferences that can be drawn with respect to obesity treatment.

We also found overall moderate quality evidence for a small to moderate effect of portion or tableware size on food selection among adults. Adults consistently selected larger quantities of food for consumption when exposed to larger sizes (compared with exposure to smaller sizes). This result is consistent with the role of food selection as an important intermediate endpoint in pathways to consumption. If we assumed that all food selected for consumption were consumed and that this effect size were sustained over time (noting again that we found no evidence for sustainability of effects), it would be equivalent to an absolute change in average daily energy intake from food of 188 to 403 kcal among UK adults (an 11% to 23% change from a baseline of 1727 kcal per day - see Figure Figure4)4) (National Centre for Social Research 2012). Whilst we did not find an effect of portion or tableware size on food selection among children, this result was based on overall low quality evidence from a small number of studies (independent comparisons), which confers limited confidence in our estimate of this effect (that is, the true effect among children may be substantially different from our estimate).

We did not find evidence for an effect of individual unit size on consumption of tobacco, based on a meta-analysis of data collected from studies that investigated exposure to longer versus shorter cigarettes among adult smokers. However, this finding was again based on overall low quality evidence from a small number of older studies. We did not identify any eligible studies that investigated the effects of exposure to differently sized cigarette packs (for example, packs of 20 cigarettes versus packs of 10 cigarettes). Nor did we identify any eligible studies that investigated the effects of exposure to differently sized alcoholic beverage products (or tableware, such as glasses, used to consume such products).

Shape

This review found overall very low quality evidence from a single included study for a large effect of exposure to shorter, wider (versus taller, narrower) plastic bottles on the quantities of water participants consumed in a single-serve context (Wansink 2005d). In this study, participants provided with shorter, wider bottles had more water available for consumption in the first place (due to having already selected more by pouring more into their bottles from a 10 gallon container) than participants provided with taller, narrower bottles. The 'very low quality' rating means that we have little confidence in the estimate of this effect (that is, the true effect is likely to be substantially different from our estimate).

We also found overall low quality evidence for a large to very large effect of exposure to shorter, wider (versus taller, narrower) glasses or plastic bottles on the quantities of fruit juice or water participants selected for consumption in a single-serve context. If the effect size we estimated were transferable to energy-containing non-alcoholic beverages (Figure (Figure4),4), it would be equivalent to an absolute change of 292 to 462 grams in the average quantity of these beverages selected in a single-serve context among UK children (a 128% to 203% change from a baseline of 228 grams per serve) or 68 to 274 grams among UK adults (a 28% to 112% change from a baseline of 245 grams per serve) respectively (National Centre for Social Research 2012). We rated the quality of evidence as low with respect to our estimates of this effect, which again confers limited confidence in their accuracy. The findings are, however, consistent with long-established psychological theory and evidence concerning the perceptual biases associated with exposure to differently shaped receptacles (Piaget 1969). While it seems unlikely that interventions that successfully reduced exposure to shorter, wider drinking receptacles (or conversely, increased exposure to taller, narrower versions) could in practice achieve sustained reductions in self served quantities of energy-containing non-alcoholic beverages (or increases in self served quantities of healthier alternatives) of this magnitude, this awaits study.

Moderators of main effects

As reflected in the discussion of main effects, our results indicated that the effects of portion, package, individual unit or tableware size may be modified by the age of those exposed to such manipulations. Whilst there was evidence that children and young people exposed to larger sizes still consumed more food, the size of this effect was found to be larger among adults, also increasing (albeit by very small incremental amounts) with the age of those exposed. These results suggest that intervening to reduce exposure to larger sizes of portions, packages, individual units or tableware may be more effective in influencing food consumption among adults than among children. This finding appears consistent with suggestions in the literature that as people age, external cues to consumption play an increasingly important role in the regulation of energy intake relative to internal cues, such as hunger and satiety (Ello-Martin 2005). This phenomenon has been observed in children, but we are not aware of any current evidence for whether this process continues over the adult life course.

It is noteworthy that, with the exception of age, no evidence was found in this review to support claims that the effects of exposure to different portion, package, individual unit or tableware sizes vary between men and women, between individuals with a different body mass index, or between those with different baseline levels of dietary restraint, dietary disinhibition or hunger (that is, those participant characteristics identified in advance as most likely to modify effects). With respect to gender and body mass index, we note that these findings differ from those suggested by the results of another recent review of food portion size effects (Zlatevska 2014). In relation to gender and amounts consumed, Zlatevska and colleagues found that female participants responded less to a doubling of portion size than did male participants (Zlatevska 2014). In relation to body mass index and amounts consumed, they found that overweight participants responded less to a doubling of portion size than did non-overweight participants (Zlatevska 2014) - a result which the authors highlight was unexpected since it challenges previous research suggesting that overweight people may be less sensitive to satiation and more sensitive to external cues than those who are not overweight (Wansink 2007b).

We were unable to examine effect moderation by study participants' socioeconomic status in this review due to the infrequency of reporting of such measures across included studies (this was one component of analysis intended to inform assessment of social differentiation in effects relevant to health equity - see Objectives and further, related discussion in Overall completeness and applicability of evidence). Socioeconomic status therefore remains an important potential moderator of the effects of sizing interventions that deserves closer attention in future research (see Implications for research).

We did, however, find evidence that this effect of size on consumption may be moderated by the type of food, specifically characterised by the healthiness and energy density of the manipulated food(s), with larger effects found in studies that manipulated less healthy products and in those that manipulated more energy-dense products (albeit by very small incremental amounts) (see Implications for practice for further discussion of these tentative findings).

We found little evidence consistent with the proposal that the observed effects of size on consumption or selection may differ depending on whether it is the size of a portion, package, individual unit or item of tableware size that is altered. This finding indicates that interventions that successfully reduce exposure to larger sizes can be effective across a range of targets for manipulation.

However, we did identify some evidence to indicate that between-study variation in the effect of larger size on food consumption may be attributable in part to between-study differences in the relative size of the two portions, packages, individual units or items of tableware being compared. Although this finding is based on the results of a post-hoc subgroup analysis (see Effects of interventions), we note that the results are consistent with our prior assumptions that the dose-response relationship between portion size and consumption or selection would be linear at many of the sizes investigated (see Data synthesis), but that at extremes a non-linear relationshipcould be expected due to a ceiling effect: external cues, such as social norms or perceptual biases that indicate a given amount of a product is appropriate, will eventually give way to internal cues to stop consuming, such as satiety. A recent analysis that plotted the absolute portion size served to each group of participants among included studies against the average (mean) amount of food they consumed from that portion also found a relationship of this kind (Zlatevska 2014). We reiterate (as stated in Included studies) that absolute sizes investigated in included food studies tended to be large compared with reference portion sizes, derived from a published report on typical portion sizes in the UK in 2002 (Food Standards Agency 2002). Knowledge of how the sizes of portions, packages and tableware investigated among included studies compare with reference portion sizes for those foods in different settings was not fully elucidated by this review due to the limited scope and availability of data (from included studies and external sources) to fully address it. However, this remains a critical issue for determining the policy implications of our findings concerning the effects of larger size on selection and consumption (see further commentary on this issue in Overall completeness and applicability of evidence and Implications for practice).

Meta-regression analyses identified two further variables as potential moderators of the main effects of size on both consumption and selection, both methodological variables. The first variable delimits studies with a within-subjects design and those with a between-subjects design (effect sizes were larger in between-subjects studies). We cannot fully explain this result. It may be an artefact of the different methods used to measure effects in between-subjects and within-subjects designs respectively: there are two independent groups in the former but only one group (with repeated measures for each participant) in the latter. Alternatively, the result may be due to factors related to the choice of design, including other methods and procedures applied by research centres using different study designs. The second variable distinguishes studies of food products in which the manipulated food(s) comprised all of those available in the study from all other studies (effect sizes were larger in the former studies). Providing additional foods for study participants to consume beyond those that were manipulated may result in additional energy consumption in either or both comparison groups, with the potential to modify the effect of larger sizes due to the same ceiling effect described above.

It is important to avoid over-interpretation of the results of the meta-regression analyses we conducted due to their observational nature, limited statistical power and multiple tests, which meant heightened probability of type I (obtaining a false positive result) and type II (obtaining a false negative result) errors. These results should therefore be viewed primarily as generating hypotheses about potential effect modifiers that will need to be investigated in further studies, with patterns of results replicated, before more confident inferences can be drawn.

Overall completeness and applicability of evidence

The evidence synthesised in this review was collected from 72 included studies that featured 107 eligible independent comparisons between two different sizes or shapes of portions, packages, individual units or tableware used to consume food products (69 of 72 included studies), or between two different sizes (lengths) of individual units of tobacco products (cigarettes) (3 of 72 included studies). The effective sample sizes feeding into meta-analyses of outcome data collected from included food studies typically exceeded numbers generated by a conventional sample size calculation for a single adequately powered trial (that is, the optimal information size), which strengthens confidence that these studies were sufficient to enable us to address our first objective to assess the effects of eligible interventions on unregulated selection or consumption of food products in adults and children. Moreover, included food studies encompassed a range of participants in terms of their age, gender and other trait or state characteristics, a range of specific manipulations (for example, various types of foods), and a variety of eating or drinking contexts (encompassing both laboratory and naturalistic field settings). This confers a degree of confidence that our findings concerning food are likely to be widely applicable. It was also possible to exploit variations between included studies to investigate and attempt to explain observed variations in effects, addressing the second objective of this review to assess potential effect modifiers. This allowed us to report observed associations that, if confirmed by further research, may prove useful in configuring and targeting sizing interventions for maximum effectiveness (see Implications for practice).

Eligible studies typically investigated exposures that were one-off or, if repeated, were repeated over relatively short time periods, and participants' selection and consumption responses were typically measured over correspondingly immediate or short time periods. In addition, the laboratory and naturalistic field settings in which participants were exposed and had their selection and consumption responses measured were often highly controlled by the researchers. These findings highlight the current lack of evidence to establish whether meaningful changes in the quantities of food people consume can be sustained over the longer term in response to prolonged or repeated exposures, under free-living conditions.

In terms of intervention characteristics, the distribution of evidence for effects on selection and consumption of food was skewed towards pairwise comparisons in which the difference in relative size of the portions, packages, individual units or tableware was large. In addition, the absolute sizes investigated in food studies tended to be large. Therefore, while included food studies did cover a range of absolute and relative sizes, further studies focusing on smaller incremental changes at the smaller end of the portion size continuum are needed to strengthen the evidence base in this respect.

As highlighted above (see Summary of main results), knowledge of how the absolute sizes of food portions and packages investigated among studies included in this review compare with reference portion sizes for those specific foods (defined here as the size that is recommended to be consumed, or that is customarily consumed, in a single eating occasion, by one or more schemes for communicating portion size messages to consumers (Lewis 2012)) is critical to the interpretation of the results of this review. However, this relationship is both complex and dynamic. Alongside variation between specific food products within each scheme, there is also variation between reference portion sizes for comparable products between schemes and jurisdictions (for example, recommended amounts may be defined by food manufacturers, food retailers, government agencies or non-governmental organisations, and may provide general advice or weight-loss advice (Institute of Grocery Distribution 2008; Lewis 2012)). Schemes that provide reference portion size information based on amounts customarily consumed are also typically based on analysis of dietary intake within a defined population, which will also vary between population subgroups and over time; estimates from some schemes still in current use may therefore diverge from current dietary intakes due to their age (for example, the US Food and Drug Administration's Reference Amounts Customarily Consumed are largely based on data published in 1993 (USFDA 2014)). It is therefore important to highlight that our discussion of potential policy actions that would be consistent with the evidence in this review concerning the effects of size on consumption of food (see Implications for practice, below) is necessarily tempered by consideration of where this body of evidence may be located on the 'absolute size continuum'. Our observation that the absolute sizes investigated in food studies tended to be large is based primarily on comparison with external data, derived from ranges of typical dietary intakes (amounts customarily consumed in a single eating occasion), that were published in 2002 (Food Standards Agency 2002), which may not be transferable to the present day or other settings. The key message is that we urge caution in extrapolating the results of this review beyond the range of relative size differences between, and/or the absolute sizes of, portions, packages and tableware sizes investigated among included studies.

Specifically, the limited body of evidence identified for the consumption effects of exposure to different portion, package and tableware sizes at the smaller end of the size continuum means that we cannot be certain whether reducing portions at the smaller end of the size range can be as effective in reducing food consumption as reductions at the larger end of the range. There may also be some potential for unintended effects of exposure to small portions. Exposure to smaller portions than those typically encountered could sometimes lead to increased consumption. One possibility is that people may avoid selecting or consuming larger portions of products they perceive as unhealthy, but allow themselves to indulge when those products are presented in small sizes, thereby shifting from no consumption to some. The potential for unintended compensatory effects (that is, compensating for smaller portions by eating more later in the day), whilst not evident from individual studies we have encountered (Jeffery 2007; Kral 2004a; Lewis 2015; Vermeer 2011), is another related issue that deserves close attention.

We judged few participant samples in included food studies to be characterised by high levels of material or social deprivation; few studies measured participants' socioeconomic status and no studies reported effects disaggregated by socioeconomic subgroup. Moreover, evidence for effects on selection and consumption of food was derived mainly from studies conducted in US samples, with no included studies conducted in low or middle-income countries (LMICs). These factors largely precluded any assessment of social differentiation in effects relevant to health equity (with the exception of gender - see Effects of interventions, 'Potential modifiers of the effect of larger size on consumption') (see also Objectives). We have no reasons to expect that cognitive biases proposed as mechanisms by which exposure to these interventions may influence food selection and consumption (for example, 'unit bias') will differ substantively between people living in high-income countries (HICs) and those living in LMICs (see How the intervention might work). However, people living in HICs are likely to have different personal and social (descriptive and injunctive) norms about what constitutes a suitable amount of food to consume than those living in LMICs and such factors have been proposed to influence the effects of exposure to larger sizes on food selection and consumption. A range of other social, cultural, economic and contextual differences surrounding diet-related behaviours between people living in HICs and LMICs may also plausibly modify these effects. For these reasons, the predominance of US evidence may limit the applicability of findings of this review to LMICs (and also to other HICs) to some extent.

This review identified three studies that investigated the effects of exposure to longer versus shorter cigarettes on tobacco consumption (Jarvik 1978 (E1); Jarvik 1978 (E2); Russell 1980). We did not identify any tobacco studies investigating the effects of exposure to different sizes (or shapes) of cigarette packs, which may be an alternative target for interventions to reduce exposure to single cigarettes or packs containing smaller than standard numbers of cigarettes. Applicability of the evidence derived from the three included tobacco studies we did find, published in 1978 and 1980, may be limited by its age. The small effective sample size (six independent comparisons, 108 participants) contributing to our meta-analysis from these studies further weakens confidence that they provided sufficient evidence to allow us to address the first objective of this review with respect to tobacco products. The true effect of exposure to longer versus shorter cigarettes on tobacco consumption is likely to be substantially different from our summary estimate. Results based on evidence from tobacco studies should therefore be interpreted with caution.

The most notable gap in this evidence base, however, was the absence of any randomised controlled trials investigating effects on unregulated selection or consumption of alcoholic beverage products. This finding is in keeping with the small proportion of studies on alcohol, compared with food products, which we found in a large scoping review of interventions that involve altering the properties or placement of objects or stimuli within small-scale micro-environments to change health behaviour, of which 'sizing interventions' was just one type (Hollands 2013a; Hollands 2013b). One possible reason for the current dearth of studies on alcohol is that this reflects the focus of recent alcohol policies on reducing consumption in harmful and hazardous drinkers through individual-level interventions (Kaner 2009). Interventions that target price can reduce consumption of alcohol across populations (Holmes 2014; Wagenaar 2009), but such interventions are generally unacceptable to industry, politicians and the general public (Diepeveen 2013). More recent evidence regarding the harmful effects on population health of alcohol consumption at moderate levels (Rehm 2015) may extend the research focus to include interventions in micro-environments such as those pertaining to size.

Quality of the evidence

Ratings of the overall quality of evidence incorporated into this review ranged between moderate and very low, which leaves open the possibility that our estimates of intervention effects differ substantially from true effects. Confidence in estimates of effects was diminished by serious concerns about study limitations, which were primarily raised by unclear and incomplete reporting of study methods and procedures by authors of included studies. Indeed, we identified limitations in study reporting and/or conduct with respect to each of the domains judged most critical to 'Risk of bias' assessment in this review: random sequence generation (selection bias); allocation concealment (selection bias); blinding of participants and personnel (performance bias); and baseline comparability of participant characteristics between groups (other bias). Given the nature of the included studies, we could not identify any obvious reason to prevent the straightforward implementation of unbiased methods and procedures for random sequence generation and allocation concealment. The use of within-subjects designs precluded the blinding of participants in over half of the included studies, but we did not judge lack of blinding to place studies at high risk of bias in this domain due to a general lack of evidence for the presence and potential influence of carry-over effects among included studies. We did not consider blinding of personnel (that is, intervention providers) to be a relevant consideration in assessing risk of bias in included studies because personnel were not judged instrumental in delivery of the intervention. Finally, while it may not always be practical to test such differences in applied field settings, in many instances baseline comparability of participant characteristics between comparison groups can and should be examined.

We identified few concerns regarding inconsistency in study results, since in general large amounts of unexplained inconsistency did not remain following planned investigations of potential effect modifiers using meta-regression analyses. There were no serious concerns about the directness of the assembled evidence either, since it was all derived from studies that directly compared the interventions in which we were interested, in groups of eligible participants, and incorporated direct (and typically objective) measures of unregulated selection or consumption.

We had no serious concerns about imprecision in relation to our estimates of the effects of exposure to larger (versus smaller) portion, package, individual unit or tableware size on unregulated selection or consumption of food, since (as noted above) effective sample sizes comfortably exceeded the numbers generated by conventional sample size calculations for single adequately powered trials (optimal information sizes). However, we did have serious concerns about imprecision in relation to our estimates of the effect of exposure to longer (versus shorter) cigarettes on consumption of tobacco, and of the effect of exposure to shorter, wider (versus taller, narrower) glasses or plastic bottles on consumption of non-alcoholic beverages, based on consideration of both threshold optimal information sizes and confidence intervals.

Potential biases in the review process

Whilst it is possible that we may have failed to identify every study eligible for inclusion in this review, we took several steps to minimise this risk, including our use of highly sensitive search strategies and backward and forward citation searches. We therefore consider it improbable that we have failed to identify sufficient relevant evidence to substantively alter our conclusions. The scope, scale and complexity of this review and its analysis meant that we took the pragmatic decision (in consultation with the Cochrane Public Health Review Group) to defer full integration of 11 further eligible studies identified by the updated search (30 January 2015) (Bajaj 2014; Haire 2014; Kral 2014; Marchiori 2014; Rolls 2014a; Smith 2013a; van Ittersum 2013; van Kleef 2014; Wansink 2013; Wansink 2014; Williams 2014), until the first major update of this review. However, the results of preliminary analyses of outcome data that could provisionally be extracted from each of these 11 further eligible studies (see Appendix 2) establish that there is minimal potential for the full integration of these studies to change the interpretation of the results of this review, and hence its conclusions, as currently reported in the Results, Discussion and Authors' conclusions.

Agreements and disagreements with other studies or reviews

In a review of the effects of portion sizing published in 2014, Zlatevska and colleagues found that increasing portion size led to a small to moderate increase in consumption, reporting an effect size of d = 0.45 (Zlatevska 2014). This point estimate was similar to those we found in the current review and within its 95% confidence intervals. Results of moderator analyses conducted in Zlatevska and colleagues' review were again broadly consistent with our results. First, Zlatevska and colleagues similarly reported that the intervention effect was greater in adults than in children. Second, consistent with our findings regarding moderation by healthiness and by energy density of food, they reported a larger effect for snack foods (which are typically less healthy and more energy-dense) than non-snack foods. Contrary to the results of our analysis, however, they reported finding a larger effect among men than among women and a smaller effect among overweight participants than among participants who were not overweight. Discrepancies between the results of these analyses are expected since they used different data sets as a consequence of differences in their respective eligibility criteria, procedures and analytic methods. Although criteria for considering studies in Zlatevska and colleagues' review were broadly similar to those applied in this review, the former focused exclusively on food, did not appear to exclude studies in which participants' consumption was regulated by either explicit instructions or some other action of the researcher, and additionally included studies that measured intended but not actual consumption. Zlatevska and colleagues' review did not include coverage of evidence for the effects of package, individual unit or tableware size on consumption and did not investigate food selection as an outcome. Indeed, we are not aware of any relevant, previously published reviews that investigate either the effects of exposure to food packages or to individual food units of varying size (and only one that investigates dishware size - see below in this section), nor that investigate food selection as an outcome.

We are aware of only one other systematic review, published in 2013 (Small 2013), which - like ours and Zlatevska and colleagues' reviews (Zlatevska 2014) - encompassed evidence for the effects of exposure to food portions of varying size on energy intake among well and normally developing children. Small and colleagues aggregated evidence from six eligible primary studies - all randomised controlled trials that are fully incorporated into our review (Fisher 2003; Fisher 2007a; Fisher 2007b; Fisher 2007c; Rolls 2000; Spill 2010) - using a narrative synthesis and reported a similar finding: that larger served portions resulted in greater daily energy intake among participants (Small 2013).

In a review of the effect of dishware size on consumption of food published in 2014, Robinson and colleagues reported results consistent with no effect of dishware size on consumption (standardised mean difference (SMD) -0.18, 95% confidence interval (Cl) -0.35 to 0.00, P value = 0.05) - although we note that the authors reported "a small effect that was not statistically significant", with exposure to larger dishware leading to greater consumption (Robinson 2014). Although this review again differed from ours with respect to its inclusion criteria (for example, non-randomised studies were eligible and targets of the manipulation were restricted to bowl size or plate only), its estimate of this effect overlaps considerably with our corresponding estimate for the effect of tableware size on consumption (see Figure Figure77).

Authors' conclusions

Implications for practice

Due to limitations in the scope, quality and quantity of relevant research evidence that is currently available (including in the case of alcohol, a complete absence of evidence), the key implications of this review for public health policy and practice, set out below, concern food. We are unable to highlight any clear implications for alcohol or tobacco policy. In addition, all of the currently available evidence derives from studies conducted in high-income countries (HICs) (predominantly in the USA), with no evidence from studies conducted in low and middle-income countries (LMICs). The applicability of our findings to public health decision-making in LMICs therefore remains uncertain. Moreover, we found insufficient evidence to indicate whether portion size effects may vary in HICs between people according to their socioeconomic status or levels of social or material deprivation. As such, it is unknown whether and how interventions that reduce, or moderate the effects of, exposure to larger-sized portions, packages, individual units and tableware would impact on existing inequalities between socioeconomic groups in health-related behaviours or corollary health outcomes.

The principal finding of this review is that people consistently consume more food and drink when offered larger-sized portions, packages or tableware than when offered smaller-sized versions. This suggests that policies and practices that successfully reduce, or moderate the effects of, exposure to larger-sized portions, packages, individual units and tableware – in and outside the home – can contribute to meaningful reductions in the quantities of food and non-alcoholic beverages people select and consume in the immediate and short term. Actions to halt, reverse or mitigate the effects of recent trends towards larger portions (Young 2002; Young 2012) may therefore be justified on public health grounds. The portion sizes investigated in included food studies were typically at the larger end of the absolute size continuum, therefore the evidence in this review confers confidence that reducing the sizes of portions and packages that are large in absolute terms can achieve effects of the magnitude estimated. However, the evidence in this review neither convincingly supports, nor undermines, claims that making sizes smaller than have become typical or standard can be expected to have similarly meaningful impacts on food selection or consumption. In response to these findings, possible intervention strategies targeting the physical environment (in public sector and/or commercial sector settings) include: regulatory and legislative frameworks, or voluntary agreements with the food industry, which result in alterations in portion size (Bryden 2013; Hsiao 2013); reducing default serving sizes of energy-dense foods and drinks where these are large in absolute terms, or providing smaller crockery, cutlery and glasses for use in their consumption; and various 'choice architecture' interventions in micro-environments such as restaurants or supermarkets (Hollands 2013a). Examples of the latter may include, for example, reducing the availability of larger portion, package and tableware sizes; placement of larger portion sizes further away from purchasers; or demarcation of single portion sizes in packaging through wrapping or a visual cue.

Other potential intervention strategies targeting the economic environment include eliminating pricing practices whereby larger portion and package sizes cost less in relative (and sometimes absolute) monetary terms than smaller sizes and thus offer more value for money to consumers (Steenhuis 2009) and restricting price promotions on larger-sized packages. There is limited and equivocal evidence for the effectiveness of interventions that do not seek to directly alter the availability or cost of larger sizes, but instead aim to educate people about appropriate portion sizes - for example, by providing information about the portion size effect or the number of portions in a serving (Cavanagh 2013; Spanos 2015; Versluis 2015). This does not, however, rule out a potential role for social marketing campaigns to raise awareness and engender public acceptability of the public health case for interventions to reduce or moderate the effects of exposure to larger-sized portions of food and drink. Such approaches may help to create the social and political conditions necessary to enable effective interventions to be implemented. The design of interventions targeting physical or economic environments, or aiming to educate or otherwise create enabling social, cultural and political conditions for effective intervention of this kind, will need to remain sensitive to local cultural and socioeconomic circumstances in different implementation settings (Huang 2015; Rychetnik 2002).

With the exception of directly controlling the sizes of the foods people consume, assessment of the effectiveness of possible intervention strategies was beyond the scope of this review. However, findings from relevant published evidence syntheses present a mixed picture. For example, a recent economic analysis ranked interventions comprising reductions in portion size of foods and beverages in various contexts highest, among a portfolio of evaluated policy levers, for reducing the population health burden of obesity (McKinsey Global Institute 2014). However, the portion size component of this economic analysis, based on a smaller, overlapping set of studies compared to the current review, assumed that the same sizes of effects estimated in source studies (which measured consumption effects over immediate or short time periods in response to one-off or short-term exposures) will be sustained and cumulative over people's lifetimes in response to repeated exposures (Corrine Sawyers, personal communication 2015). In addition, a 2009 review of interventions aiming to address the negative influences of portion size effects on consumption that formed part of the evidence base used in this economic analysis found few studies, and these showed mixed effects (Steenhuis 2009) (see also Implications for research).

This review suggested that the effect of larger size on consumption may be robust to variation between interventions in terms of several of their key characteristics and those of their participants. For example, we did not find evidence that the intervention effect varied substantively between men and women, nor by people's body mass index, susceptibility to hunger, or tendency to consciously control their eating behaviour. These findings are essentially observational, should be interpreted with caution and would need to be confirmed by future studies before they can be distilled into clear policy implications. However, if confirmed, these null findings would add credence to the claim that people are susceptible to environmental influences on food consumption that operate independently of individual characteristics that are often portrayed as the main drivers of over-consumption; and indicate the potential for effective interventions targeting portion, package and tableware size to reduce consumption among a broad range of people. Other tentative findings suggested that such interventions may be particularly effective in reducing consumption among adults and that reductions in exposure to larger portion sizes of less healthy and of more energy-dense foods - those foods whose over-consumption is most damaging to health - might usefully be the principal target for policy action. We cannot readily explain these results but note that they replicate those of another recent review of food portion size effects (Zlatevska 2014). It may be that people have reduced ability to regulate their consumption of less healthy and more energy-dense foods in response to external cues - either due to these properties or other associated properties (for example, palatability) - thereby increasing the potential for size to influence quantity consumed. However, studies included in this review that experimentally manipulated both size and energy density variables did not find interaction effects consistent with this proposal (Devitt 2004; Rolls 2006b; Rolls 2010a (E1); Rolls 2010b (E2)).

Irrespective of uncertainty regarding the mechanism of this moderation, these findings would be encouraging from a public health perspective if replicated by further research for two reasons. First, they highlight the possibility that the largest reductions in consumption might be achieved by reducing exposure to larger sizes of those products for which a reduction is likely to be most beneficial for health. Second, they are consistent with the proposal that a 'portion size effect' is still present when people are exposed to larger sizes of healthier and less energy-dense foods, suggesting that interventions that successfully increase people's exposure to larger portion sizes of healthier, low energy-dense foods such as vegetables may still be an effective strategy for increasing consumption of these foods (Rolls 2014b).

Whilst this review found evidence of moderate overall quality indicating that people select and consume more food when exposed to larger-sized portions, packages, individual units and tableware, it is important to highlight that these findings were derived from studies that typically investigated exposures that were one-off, or if repeated at all, were repeated over relatively short time periods, often under highly controlled experimental conditions. The longer-term sustainability of the effects of prolonged or repeated exposures, and effects under free-living conditions, therefore remain to be established. This underscores that the long-term effectiveness of interventions introduced with the aim of reducing people's exposure to larger portion, package and tableware sizes is currently unknown (worldwide) and will be subject to all the challenges and complexities of achieving effective and sustained implementation at scale.

One such complexity is the actual and perceived monetary costs (prices) of food products, which have been proposed to modify the effects of portion or package size on food consumption (Steenhuis 2009). Evidence to inform understanding of potential interactions between product size and cost appears to be lacking (that is, no studies eligible for inclusion in this review investigated such interactions). Another is that scaling up interventions of this kind (that is, increasing their geographic coverage and scope with the corollary potential to influence the behaviour of large numbers of people in a wider range of eating and drinking contexts) would involve their introduction into a complex food environment populated by a multitude of available food products other than those having their sizes directly or indirectly altered. For example, in homes, shops and restaurants people have access to additional quantities of a wide variety of foods. The potential for compensatory consumption of other foods is not elucidated by this review.

A further set of challenges to implementing policies to reduce exposure to larger-sized portions of food and non-alcoholic beverages is provided by the commercial and legal contexts in which these products are sold. The likely strength of resistance among food and beverage industry representatives was evident in an unsuccessful attempt in New York to cap the portion sizes of sugar-sweetened beverages sold in restaurants and other venues serving food (Gabbatt 2013; Grynbaum 2012). However, policies of this kind appear to be more acceptable among the general public (Diepeveen 2013; Petrescu under review), which raises the possibility of pursuing alternative strategies such as engaging civil and other organisations at local, national and international levels to advocate for reconfiguration of systems of production and consumption (Freudenberg 2014; Jackson 2009; Skidelsky 2013).

In summary, this review provides the most conclusive evidence to date that people consistently consume more food and drink when offered larger-sized portions, packages or tableware than when offered smaller-sized versions. This suggests that policies and practices that reduce, or moderate the effects of, exposure to larger sizes can contribute to meaningful reductions in the quantities of food and non-alcoholic beverages people select and consume. This may justify actions to reduce the size, availability and appeal of food portion, package and tableware sizes that are large in absolute terms. However, it is uncertain whether reducing portions at the smaller end of the size range can be as effective in reducing food consumption as reductions at the larger end of the range. We are unable to highlight clear implications for tobacco or alcohol policy due to identified gaps and limitations in the current evidence base.

Implications for research

The implications for research set out below are based on gaps and uncertainties identified by reviewing the current evidence base, which (as highlighted above - see Implications for practice) derives exclusively from studies conducted in HICs. Although it is feasible that the implications may also be applicable to research in LMICs, the lack of experience of conducting studies of this kind in LMICs leaves open the possibility that LMIC-specific research issues may emerge if such experience accumulates.

This review found no evidence from randomised controlled trials for the effects of altering size or shape on selection or consumption of alcoholic beverages and identified only five eligible studies that included a focus on non-alcoholic beverages. More evidence for intervention effects on unregulated selection and consumption is needed with respect to both of these product categories to inform the design of interventions to reduce their consumption and ameliorate associated impacts on health inequalities. The social patterning of harmful alcohol use and its health consequences is well documented (Fone 2013), whilst sugar-sweetened beverage consumption, which represents the largest source of added sugar in UK and US diets (Tedstone 2014; Welsh 2011), is also socially patterned, with heavy consumption being more likely among adults and children from lower socioeconomic status backgrounds (Han 2013). Furthermore, few eligible tobacco studies were identified and those we did find compared the effects of exposure to longer versus shorter cigarettes, the most recent published in 1980 (Russell 1980). We found no studies of other conceivable tobacco product size or shape manipulations, such as cigarette packs sized to contain different numbers of cigarettes. This is notable given the European Union decision (Tobacco Products Directive: European Union 2014) to ban smaller cigarette packs containing fewer than 20 cigarettes from 2016. This decision was based on factors related to both harmonisation of trade and public health, including implementation of the WHO Framework Convention on Tobacco Control (WHO FCTC), which entered into force in 2005 (World Health Organization 2003). Article 16 of the WHO FCTC prohibits the sale of cigarettes individually or in small packets on the basis that this increases their affordability to children, which aligns with evidence indicating that price is an important factor in determining smoking initiation among children and young people (Godfrey 2009; NICE 2008; Pierce 2012). As such, most of the evidence incorporated into this review relates to the effect of exposure to larger versus smaller-sized portions, packages, individual units and tableware on the selection and consumption of food (including non-alcoholic beverages, although as noted above, these were underrepresented). However, several of the implications for research that we highlight below in relation to food studies may be transferable for consideration in the development of future research on alcohol and tobacco products.

The body of evidence in this review clearly indicates a potential role for interventions that successfully reduce exposure to larger portion, package or tableware sizes, or mitigate the effects of such exposure, to help change people's food, energy and nutrient intake. As noted above (see Implications for practice) the range of possible intervention strategies includes regulatory and legislative frameworks that mandate alterations in size, voluntary agreements with industry, choice architecture interventions, interventions targeting price, and educational and social marketing interventions (all of which fell outside the scope of this systematic review). Whilst we are not currently aware of any systematic reviews that have aimed to assess the effectiveness of such interventions, a traditional literature review of interventions designed to address the negative influence of portion size on energy intake, published in 2009, identified only five relevant primary studies (all conducted in HIC settings) investigating different specific interventions involving: provision of nutritional information on product labelling; nutritional labelling with price promotion; and restrictions placed on customers' purchasing of larger portions (Steenhuis 2009).

These observations point to the need for further research in two specific areas. First, further new primary studies of the effects of exposure to larger versus smaller-sized portions, packages, individual units and tableware on selection and consumption of food (that is, studies meeting the eligibility criteria for this review) are needed. Second, a systematic review of evidence for the effectiveness of interventions to reduce exposure to larger sizes, or to mitigate the effects of exposure to larger sizes (that is, studies outside the scope of this review), may be needed, possibly followed by further, new primary studies of such interventions and policies. Critically, in order to generate evidence for effectiveness and the sustainability of effects, future primary studies in both of these identified areas of research should evaluate people's selection and consumption responses over longer time periods in 'real world' environments (such as homes, shops and restaurants) and under free-living conditions as far as possible (that is, with minimal research-imposed constraints on target behaviours and environments). This may mean, for example, studying interventions implemented within otherwise unaltered restaurant or shop environments in which participants are able to freely select and consume from a typically wide range of products and over a number of weeks or months. Moreover, the studies need to be designed to contribute to summary estimates of corollary impacts on health inequalities. This would not only ensure that policies found to be effective do not cause "intervention generated inequalities" (Lorenc 2013), but would also increase understanding of their potential to reduce inequalities arising from excessive consumption of less healthy products by more socially and materially deprived people, such as those with low levels of education or income. None of the included studies assessed (or indeed were powered to assess) the moderation of intervention effects by socioeconomic status, or potential interactions between product size and cost in influencing selection with purchasing.

With respect to the first specific area in which research is needed, further new primary studies of intervention effects on selection and consumption of food could feed into an updated synthesis that would have the potential to increase our confidence in summary estimates of these effect sizes and reduce associated uncertainty. This would have the potential to strengthen our qualified finding that portion, package, individual unit and tableware size represent promising targets for public health intervention to change the quantities of food, energy and nutrients people select consume. Any such studies should include further investigation of the tentative findings of this review in relation to potential effect modifiers.

There is also considerable scope for any such further studies to help fill gaps in the current evidence base that we have identified in this review. As well as the critical need to generate evidence for the effectiveness of prolonged or repeated exposures over longer time periods and with minimal research-imposed constraints on behaviour, this could usefully include investigations of effects in a wider range of participant subgroups, such as adolescents and older adults. New primary studies could also expand the current evidence base by investigating effects in a wider set of field settings than were represented among studies included in this review, which were predominantly conducted in restaurants or in school or workplace cafeterias. Given that most food and drink is purchased in shops for consumption in the home (DEFRA 2013; Harnack 2000; Smith 2013b), research to examine intervention effects in these contexts is especially needed.

Critically, any further primary studies of this kind should also feature smaller absolute sizes, and smaller magnitudes of size difference between the compared portions, packages, individual units or items of tableware. More evidence from studies presenting participants with smaller absolute sizes is needed to confer a higher degree of confidence than can be derived from the body of evidence in this review that reducing sizes to amounts smaller than have become typical or standard has the potential to be an effective intervention strategy (see Overall completeness and applicability of evidence and Implications for practice).

With respect to the second specific area in which research is needed, it would be useful - especially given the age of Steenhuis and colleagues' traditional literature review of interventions to address negative influences of portion sizing (Steenhuis 2009) - to conduct a preliminary scoping exercise to ascertain whether sufficient primary studies of various possible interventions to reduce, or mitigate the effects of, exposure to larger food sizes have been conducted to warrant a new systematic review. If not, new primary studies of the effectiveness of a broader range of possible interventions than were identified in the earlier review (Steenhuis 2009) should be undertaken, encompassing regulatory, non-regulatory and pricing strategies (highlighted above in this section). The appropriate balance between the two areas of primary research we have highlighted will depend in part on the extent to which overall moderate quality evidence for a small to moderate effect of size on consumption is regarded as a sufficient basis for policy action to mitigate the undesirable consequences of such effects.

Finally, the evidence base for the effects of these kinds of interventions would be substantively improved by better-conducted and reported primary studies. In the process of conducting this review we encountered some egregious examples of study reporting - such as reports lacking basic descriptive statistics for outcome data, or key details of study methods and procedures - and unwillingness or inability of some study authors to provide additional data missing from study reports. This may be attributable in part to the age of some of the included studies and the slow diffusion of study reporting guidelines that have become established in medical research into the psychology and nutrition literatures (Grant 2013; Mayo-Wilson 2013). Primary researchers should ensure that their study reporting complies with CONSORT-SPI – a forthcoming extension of the Consolidated Standards of Reporting Trials (CONSORT) Statement, which has specifically been developed for randomised controlled trials of social and psychological interventions (Montgomery 2013) – and that it includes descriptions of interventions (exposures) sufficiently detailed to allow their replication (Hoffmann 2014). To maximise the optimal use and reuse of primary research, new study authors and those of existing studies will ideally ultimately provide open access to their complete, anonymised individual participant-level data sets in machine-readable format. In principle it would be possible to synthesise these data using individual participant data meta-analysis methods (Stewart 2011), with the potential to reduce current levels of uncertainty concerning main effects and effect modifiers, and to generate findings with much sharper implications for policy concerning portion, package and tableware size interventions.

In summary, this review highlights the potential value of further research to establish sizes of effects of exposure to differently sized alcoholic beverage products. Further research may also be conducted to reduce uncertainty about the sizes of effects of exposure to differently sized portions and packages of food and (in particular) non-alcoholic beverages, and of tableware used in their consumption, especially with regards to smaller absolute sizes and magnitudes of difference in relative sizes, and the sustainability of such effects, in 'real world' environments. Finally, effect sizes of interventions to reduce, or mitigate the effects of, exposure to larger-sized food portions, packages and tableware, need to be established. Such interventions encompass a range of potential strategies, including changes to physical and economic environments designed to reduce the size, availability and/or appeal of larger food portions.

Acknowledgments

Production of this Cochrane review was funded by the UK Department of Health Policy Research Programme (107/0001-Policy Research Unit in Behaviour and Health). The views expressed are those of the authors and not necessarily those of the UK Department of Health. We would like to acknowledge the contributions of Julie Glanville (York Health Economics Consortium, University of York, UK) who reviewed a draft of our MEDLINE search strategy, Claire Stansfield (EPPI-Centre, University of London, UK) who helped to develop and ran our searches of the TRoPHI database, and Helen Morgan (TSC, Cochrane Public Health Review Group) who initially reviewed our overall search strategy. We would like to thank Jodie Doyle (Managing Editor), Liz Waters (Co-ordinating Editor) and their colleagues in the Cochrane Public Health Review Group - especially Jonathan Shepherd (our Contact Editor), Anke Rowher (Methods Advisor), Ruth Turley (Author) and Daniel Francis (Editor). We also greatly appreciated input from Sadequa Shahrook (Research Fellow, Department of Health Policy, National Center for Child Health and Development, Tokyo, Japan) and anonymous external referees who provided helpful comments on the penultimate and final draft full review, as well as other editors, peer reviewers and critical friends who have provided helpful feedback at various stages. Finally, we would like to acknowledge all the authors of included studies who responded to our requests for clarifications or data - especially those who responded with respect to several included studies.

APPENDICES

Appendix 1. Search strategies, search dates and yields

Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library, 1992 to 30 January 2015

Original search executed: 20 November 2012; Retrieved: 3192 records

Updated search executed: 30 January 2015; Retrieved 1269 records

drink* OR drunk* OR alcohol* OR beverage* OR beer* OR lager* OR wine* OR cider* OR alcopop* OR alco-pop* OR spirit OR spirits OR liquor* OR liquer* OR liqueur* OR whisky OR whiskey OR whiskies OR whiskeys OR schnapps OR brandy OR brandies OR gin OR gins OR rum OR rums OR tequila* OR vodka* OR cocktail* OR cigar* OR smoke OR smokes OR smoking OR smoker OR smokers OR smoked OR tobacco* OR nutri* OR calori* OR food* OR eat OR eats OR eaten OR eating OR ate OR meal* OR snack*

AND

siz* OR dimension* OR capacit* OR volume* OR shap* OR height* OR width* OR length* OR depth* OR divide*

AND

portion* OR serving* OR product* OR packag* OR packet* OR unit* OR tableware OR drinkware OR dinnerware OR crockery OR plate* OR platter* OR tureen* OR tajine* OR tagine* OR bowl* OR charger* OR cup* OR saucer* OR glass OR glasses OR mug OR mugs OR beaker* OR pitcher* OR jug* OR decanter* OR receptacle* OR container* OR dish* OR pot OR pots OR cutlery OR flatware OR utensil* OR knife OR *knife OR knives OR fork* OR spoon* OR *spoon OR tongs OR ladle* OR chopstick* OR box* OR bag* OR can* OR carton* OR bottle* OR straw*

NOT

rat OR rats OR mouse OR mice OR murine OR rodent OR rodents OR hamster OR hamsters OR pig OR pigs OR porcine OR rabbit OR rabbits OR animal OR animals OR dog OR dogs OR cat OR cats OR cow OR cows OR bovine OR sheep OR ovine OR monkey OR monkeys

MEDLINE (OvidSP - including MEDLINE In-Process), 1946 to November Week 1 2012

Original search executed: 13 November 2012; Retrieved: 17,085 records

Updated search executed: 30 January 2015; Retrieved 4205 records

1 exp Beverages/ 87429

2 exp Drinking Behavior/ 52972

3 exp Alcohol Drinking/ 47670

4 exp Food Industry/ 91946

5 exp Alcohol-Related Disorders/ 92856

6 (drink$ or drunk$ or alcohol$ or beverage$1 or beer$1 or lager$1 or wine$1 or cider$1 or alcopop$1 or alco-pop$1 or spirit or spirits or liquor$1 or liquer$1 or liqueur$1 or whisky or whiskey or whiskies or whiskeys or schnapps or brandy or brandies or gin or gins or rum or rums or tequila$1 or vodka$1 or cocktail$1).ti,ab. 286166

7 exp Tobacco/ 23931

8 exp Smoking/ 113243

9 exp "Tobacco Use Disorder"/ 7270

10 (cigar$ or smoke or smokes or smoking or smoker or smokers or smoked or tobacco$).ti,ab. 196390

11 exp Diet/ 178322

12 exp Food Industry/ 91946

13 exp Food/ 985939

14 exp Food Habits/ 18591

15 exp Food Preferences/ 8909

16 exp Eating/ 55571

17 exp Feeding Behavior/ 111521

18 exp Eating Disorders/ 20715

19 (nutri$ or calori$ or food$ or eat or eats or eaten or eating or ate or meal$ or snack$ or drink$ or drunk$ or beverage$1).ti,ab. 583819

20 exp Food Packaging/ 4321

21 exp Food Storage/ 249

22 exp Cooking/ and Eating Utensils/ 104

23 exp Product Packaging/ 15467

24 ((siz$ or dimension$ or capacit$ or volume$ or shap$ or height$ or width$ or length$ or depth$ or divide$) adj4 (portion$ or serving$ or product$ or packag$ or packet$ or unit$ or cigar$ or food$ or drink$ or alcohol$ or tableware or drinkware or dinnerware or crockery or plate$1 or platter$1 or tureen$1 or tajine$1 or tagine$1 or bowl$1 or charger$1 or cup$1 or saucer$1 or glass or glasses or mug or mugs or beaker$1 or pitcher$1 or jug$1 or decanter$1 or receptacle$1 or container$1 or dish$ or pot or pots or cutlery or flatware or utensil$1 or knife or $knife or knives or fork$1 or spoon$ or $spoon or tongs or ladle$1 or chopstick$1 or box$ or bag$ or can$ or carton$1 or bottle$ or straw$1)).ti,ab. 94119

25 or/1-6 465421

26 or/7-10 229371

27 or/11-19 1554173

28 or/20-24 109600

29 25 and 28 10916

30 26 and 28 2480

31 27 and 28 18704

32 or/29-31 22530

33 animals/ 5087545

34 (rat or rats or mouse or mice or murine or rodent or rodents or hamster or hamsters or pig or pigs or porcine or rabbit or rabbits or animal or animals or dog or dogs or cat or cats or cow or cows or bovine or sheep or ovine or monkey or monkeys).ti,ab.

3089377

35 or/33-34 5362242

36 humans/ and animals/ 1372372

37 35 not 36 3989870

38 32 not 37 17590

39 (editorial or case reports or in vitro).pt. 2288418

40 38 not 39 17085

EMBASE (OvidSP), 1980 to 30 January 2015

Original search executed: 14 November 2012; Retrieved: 22,308 records

Updated search executed: 30 January 2015; Retrieved 6922 records

1 exp beverage/ 121492

2 exp Drinking Behavior/ 32744

3 exp alcohol consumption/ 61917

4 exp food industry/ 18653

5 exp alcohol abuse/ 19149

6 (drink$ or drunk$ or alcohol$ or beverage$1 or beer$1 or lager$1 or wine$1 or cider$1 or alcopop$1 or alco-pop$1 or spirit or spirits or liquor$1 or liquer$1 or liqueur$1 or whisky or whiskey or whiskies or whiskeys or schnapps or brandy or brandies or gin or gins or rum or rums or tequila$1 or vodka$1 or cocktail$1).ti,ab. 380427

7 exp tobacco/ 28053

8 exp smoking/ 154998

9 exp tobacco dependence/ 11151

10 (cigar$ or smoke or smokes or smoking or smoker or smokers or smoked or tobacco$).ti,ab. 247027

11 exp diet/ 174704

12 exp food industry/ 18653

13 exp food/ 566656

14 exp food habits/ 103715

15 exp food preferences/ 8309

16 exp eating/ 19350

17 exp feeding behavior/ 103715

18 exp eating disorder/ 32352

19 (nutri$ or calori$ or food$ or eat or eats or eaten or eating or ate or meal$ or snack$ or drink$ or drunk$ or beverage$1).ti,ab. 737112

20 exp food packaging/ 5102

21 exp food storage/ 3444

22 exp kitchen/ 1553

23 exp packaging/ 16183

24 ((siz$ or dimension$ or capacit$ or volume$ or shap$ or height$ or width$ or length$ or depth$ or divide$) adj4 (portion$ or serving$ or product$ or packag$ or packet$ or unit$ or cigar$ or food$ or drink$ or alcohol$ or tableware or drinkware or dinnerware or crockery or plate$1 or platter$1 or tureen$1 or tajine$1 or tagine$1 or bowl$1 or charger$1 or cup$1 or saucer$1 or glass or glasses or mug or mugs or beaker$1 or pitcher$1 or jug$1 or decanter$1 or receptacle$1 or container$1 or dish$ or pot or pots or cutlery or flatware or utensil$1 or knife or $knife or knives or fork$1 or spoon$ or $spoon or tongs or ladle$1 or chopstick$1 or box$ or bag$ or can$ or carton$1 or bottle$ or straw$1)).ti,ab. 120594

25 or/1-6 494774

26 or/7-10 290348

27 or/11-19 1272638

28 or/20-24 140907

29 25 and 28 9711

30 26 and 28 3061

31 27 and 28 22322

32 or/29-31 27278

33 animals/ 1800693

34 (rat or rats or mouse or mice or murine or rodent or rodents or hamster or hamsters or pig or pigs or porcine or rabbit or rabbits or animal or animals or dog or dogs or cat or cats or cow or cows or bovine or sheep or ovine or monkey or monkeys).ti,ab.

3381652

35 or/33-34 4408920

36 humans/ and animals/ 454714

37 35 not 36 3954206

38 32 not 37 22488

39 (editorial or case reports or in vitro).pt. 415728

40 38 not 39 22308

PsycINFO (OvidSP), 1806 to 30 January 2015

Original search executed: 14 November 2012; Retrieved: 4099 records

Updated search executed: 30 January 2015; Retrieved 1079 records

1 exp Alcoholic Beverage/ 1884

2 exp "Beverages (Nonalcoholic)"/ 772

3 exp Drinking Behavior/ 54223

4 exp Alcohol Drinking Patterns/ 49383

5 exp Alcohol Abuse/ 36125

6 (drink$ or drunk$ or alcohol$ or beverage$1 or beer$1 or lager$1 or wine$1 or cider$1 or alcopop$1 or alco-pop$1 or spirit or spirits or liquor$1 or liquer$1 or liqueur$1 or whisky or whiskey or whiskies or whiskeys or schnapps or brandy or brandies or gin or gins or rum or rums or tequila$1 or vodka$1 or cocktail$1).ti,ab. 111663

7 exp Tobacco Smoking/ 20293

8 (cigar$ or smoke or smokes or smoking or smoker or smokers or smoked or tobacco$).ti,ab. 38912

9 exp diets/ 8007

10 exp eating behavior/ 11578

11 exp food/ 8002

12 exp food intake/ 11118

13 exp food preferences/ 3193

14 exp eating/ 11578

15 exp feeding behavior/ 8236

16 exp eating disorder/ 21015

17 (nutri$ or calori$ or food$ or eat or eats or eaten or eating or ate or meal$ or snack$ or drink$ or drunk$ or beverage$1).ti,ab. 123754

18 ((siz$ or dimension$ or capacit$ or volume$ or shap$ or height$ or width$ or length$ or depth$ or divide$) adj6 (portion$ or serving$ or product$ or packag$ or packet$ or unit$ or cigar$ or food$ or drink$ or alcohol$ or tableware or drinkware or dinnerware or crockery or plate$1 or platter$1 or tureen$1 or tajine$1 or tagine$1 or bowl$1 or charger$1 or cup$1 or saucer$1 or glass or glasses or mug or mugs or beaker$1 or pitcher$1 or jug$1 or decanter$1 or receptacle$1 or container$1 or dish$ or pot or pots or cutlery or flatware or utensil$1 or knife or $knife or knives or fork$1 or spoon$ or $spoon or tongs or ladle$1 or chopstick$1 or box$ or bag$ or can$ or carton$1 or bottle$ or straw$1)).ti,ab. 24137

19 or/1-6 115188

20 or/7-8 39235

21 or/9-17 139533

22 18 and 19 3224

23 18 and 20 503

24 18 and 21 4019

25 or/22-24 5627

26 limit 25 to human 4099

Applied Social Sciences Index and Abstracts (ProQuest), 1987 to 30 January 2015

Original search executed: 20 November 2012; Retrieved: 949 records

Updated search executed: 30 January 2015; Retrieved 178 records

all(drink* OR drunk* OR alcohol* OR beverage[*1] OR beer[*1] OR lager[*1] OR wine[*1] OR cider[*1] OR alcopop[*1] OR alco-pop[*1] OR spirit OR spirits OR liquor[*1] OR liquer[*1] OR liqueur[*1] OR whisky OR whiskey OR whiskies OR whiskeys OR schnapps OR brandy OR brandies OR gin OR gins OR rum OR rums OR tequila[*1] OR vodka[*1] OR cocktail[*1] OR cigar* OR smoke OR smokes OR smoking OR smoker OR smokers OR smoked OR tobacco* OR nutri* OR calori* OR food* OR eat OR eats OR eaten OR eating OR ate OR meal* OR snack*)

AND

all((siz* OR dimension* OR capacit* OR volume* OR shap* OR height* OR width* OR length* OR depth* OR divide*) NEAR/6 (portion* OR serving* OR product* OR packag* OR packet* OR unit* OR cigar* OR food* OR drink* OR alcohol* OR tableware OR drinkware OR dinnerware OR crockery OR plate[*1] OR platter[*1] OR tureen[*1] OR tajine[*1] OR tagine[*1] OR bowl[*1] OR charger[*1] OR cup[*1] OR saucer[*1] OR glass OR glasses OR mug OR mugs OR beaker[*1] OR pitcher[*1] OR jug[*1] OR decanter[*1] OR receptacle[*1] OR container[*1] OR dish* OR pot OR pots OR cutlery OR flatware OR utensil[*1] OR knife OR *knife OR knives OR fork[*1] OR spoon* OR *spoon OR tongs OR ladle[*1] OR chopstick[*1] OR box* OR bag* OR can* OR carton[*1] OR bottle* OR straw[*1]))

NOT

all(rat OR rats OR mouse OR mice OR murine OR rodent OR rodents OR hamster OR hamsters OR pig OR pigs OR porcine OR rabbit OR rabbits OR animal OR animals OR dog OR dogs OR cat OR cats OR cow OR cows OR bovine OR sheep OR ovine OR monkey OR monkeys)

Food Science and Technology Abstracts (Web of Knowledge), 1969 to 22 November 2012

Original search executed: 20 November 2012; Retrieved: 6437 records

Topic=(drink* OR drunk* OR alcohol* OR beverage* OR beer* OR lager* OR wine* OR cider* OR alcopop* OR alco-pop* OR spirit OR spirits OR liquor* OR liquer* OR liqueur* OR whisky OR whiskey OR whiskies OR whiskeys OR schnapps OR brandy OR brandies OR gin OR gins OR rum OR rums OR tequila* OR vodka* OR cocktail* OR cigar* OR smoke OR smokes OR smoking OR smoker OR smokers OR smoked OR tobacco* OR nutri* OR calori* OR food* OR eat OR eats OR eaten OR eating OR ate OR meal* OR snack*) AND Topic=((siz* OR dimension* OR capacit* OR volume* OR shap* OR height* OR width* OR length* OR depth* OR divide*) NEAR/6 (portion* OR serving* OR product* OR packag* OR packet* OR unit* OR cigar* OR food* OR drink* OR alcohol* OR tableware OR drinkware OR dinnerware OR crockery OR plate* OR platter* OR tureen* OR tajine* OR tagine* OR bowl* OR charger* OR cup* OR saucer* OR glass OR glasses OR mug OR mugs OR beaker* OR pitcher* OR jug* OR decanter* OR receptacle* OR container* OR dish* OR pot OR pots OR cutlery OR flatware OR utensil* OR knife OR *knife OR knives OR fork* OR spoon* OR *spoon OR tongs OR ladle* OR chopstick* OR box* OR bag* OR can* OR carton* OR bottle* OR straw*)) NOT Topic=(rat or rats or mouse or mice or murine or rodent or rodents or hamster or hamsters or pig or pigs or porcine or rabbit or rabbits or animal or animals or dog or dogs or cat or cats or cow or cows or bovine or sheep or ovine or monkey or monkeys)

Refined by: [excluding] Document Types=( PATENT OR REVIEW OR LEGISLATION OR BOOK ) AND [excluding] Research Areas=( PHYSICS OR BIOTECHNOLOGY APPLIED MICROBIOLOGY OR CHEMISTRY OR TOXICOLOGY ) AND [excluding] Descriptors=( FREEZING OR OXIDATION OR DRYING OR FOOD FACTORIES OR TEMP OR PHENOLS OR MOISTURE CONTENT OR STARCH OR ANTIOXIDATIVE ACTIVITY OR ANALYTICAL TECHNIQUES OR DISEASES OR STERILIZATION OR MODELLING OR TEMPERATURE OR PARTICLES OR MICROORGANISMS OR FLAVOUR OR PROCESSING THERMAL OR FOOD SAFETY OR EXTRUSION OR HEATING )

We also ran a supplementary search for the FSTA index term ‘portion sizes’. Executed: 20 November 2012; Retrieved: 72 records

Descriptors=(portion sizes)

Refined by: [excluding] Document Types=( REVIEW ) AND [excluding] FSTA Section=( PATENTS )

Web of Knowledge (Science Citation Index Expanded, 1900 to 30 January 2015 Social Sciences Citation Index, 1956 to 30 January 2015; Conference Proceedings Citation Index - Science, 1990 to 30 January 2015; Conference Proceedings Citation Index - Social Science & Humanities, 1990 to 30 January 2015)

Original search executed: 20 November 2012; Retrieved: 5298 records

Updated search executed: 30 January 2015; Retrieved 2194 records

Topic=(drink* OR drunk* OR alcohol* OR beverage* OR beer* OR lager* OR wine* OR cider* OR alcopop* OR alco-pop* OR spirit OR spirits OR liquor* OR liquer* OR liqueur* OR whisky OR whiskey OR whiskies OR whiskeys OR schnapps OR brandy OR brandies OR gin OR gins OR rum OR rums OR tequila* OR vodka* OR cocktail* OR cigar* OR smoke OR smokes OR smoking OR smoker OR smokers OR smoked OR tobacco* OR nutri* OR calori* OR food* OR eat OR eats OR eaten OR eating OR ate OR meal* OR snack*) AND Topic=((siz* OR dimension* OR capacit* OR volume* OR shap* OR height* OR width* OR length* OR depth* OR divide*) NEAR/6 (portion* OR serving* OR product* OR packag* OR packet* OR unit* OR cigar* OR food* OR drink* OR alcohol* OR tableware OR drinkware OR dinnerware OR crockery OR plate* OR platter* OR tureen* OR tajine* OR tagine* OR bowl* OR charger* OR cup* OR saucer* OR glass OR glasses OR mug OR mugs OR beaker* OR pitcher* OR jug* OR decanter* OR receptacle* OR container* OR dish* OR pot OR pots OR cutlery OR flatware OR utensil* OR knife OR *knife OR knives OR fork* OR spoon* OR *spoon OR tongs OR ladle* OR chopstick* OR box* OR bag* OR can* OR carton* OR bottle* OR straw*)) NOT Topic=(rat OR rats OR mouse OR mice OR murine OR rodent OR rodents OR hamster OR hamsters OR pig OR pigs OR porcine OR rabbit OR rabbits OR animal OR animals OR dog OR dogs OR cat OR cats OR cow OR cows OR bovine OR sheep OR ovine OR monkey OR monkeys)

Refined by: [excluding] Web of Science Categories=( ECOLOGY OR ENTOMOLOGY OR CLINICAL NEUROLOGY OR ORNITHOLOGY OR MATERIALS SCIENCE CERAMICS OR MARINE FRESHWATER BIOLOGY OR SOIL SCIENCE OR PEDIATRICS OR CHEMISTRY PHYSICAL OR EVOLUTIONARY BIOLOGY OR AGRICULTURAL ENGINEERING OR ENERGY FUELS OR DENTISTRY ORAL SURGERY MEDICINE OR ENVIRONMENTAL SCIENCES OR LIMNOLOGY OR CELL BIOLOGY OR PHYSICS ATOMIC MOLECULAR CHEMICAL OR BIOPHYSICS OR ENGINEERING CHEMICAL OR ENGINEERING ELECTRICAL ELECTRONIC OR PHYSICS MULTIDISCIPLINARY OR MATERIALS SCIENCE MULTIDISCIPLINARY OR SURGERY OR MECHANICS OR OCEANOGRAPHY OR FORESTRY OR CARDIAC CARDIOVASCULAR SYSTEMS OR GASTROENTEROLOGY HEPATOLOGY OR PERIPHERAL VASCULAR DISEASE OR ZOOLOGY OR GEOSCIENCES MULTIDISCIPLINARY OR METEOROLOGY ATMOSPHERIC SCIENCES OR BIOTECHNOLOGY APPLIED MICROBIOLOGY OR PHYSICS CONDENSED MATTER OR CHEMISTRY INORGANIC NUCLEAR OR POLYMER SCIENCE OR ELECTROCHEMISTRY OR FISHERIES OR TOXICOLOGY OR CHEMISTRY MULTIDISCIPLINARY OR NEUROSCIENCES OR VETERINARY SCIENCES OR PLANT SCIENCES OR PSYCHOLOGY CLINICAL OR SPORT SCIENCES OR CHEMISTRY APPLIED OR GENETICS HEREDITY OR ENGINEERING CIVIL OR CHEMISTRY ANALYTICAL OR BIOCHEMISTRY MOLECULAR BIOLOGY OR THERMODYNAMICS OR COMPUTER SCIENCE INTERDISCIPLINARY APPLICATIONS OR PSYCHIATRY OR OPTICS OR ENGINEERING BIOMEDICAL OR AGRONOMY OR AGRICULTURE DAIRY ANIMAL SCIENCE OR BUSINESS OR ONCOLOGY OR BIOCHEMICAL RESEARCH METHODS OR PHARMACOLOGY PHARMACY OR NANOSCIENCE NANOTECHNOLOGY OR ANTHROPOLOGY OR AGRICULTURE MULTIDISCIPLINARY OR METALLURGY METALLURGICAL ENGINEERING OR MANAGEMENT OR WATER RESOURCES OR ECONOMICS OR SPECTROSCOPY OR PHYSIOLOGY OR NUCLEAR SCIENCE TECHNOLOGY OR MICROBIOLOGY OR RESPIRATORY SYSTEM OR CRITICAL CARE MEDICINE OR BIOLOGY OR INSTRUMENTS INSTRUMENTATION OR AGRICULTURAL ECONOMICS POLICY OR ENGINEERING ENVIRONMENTAL OR RADIOLOGY NUCLEAR MEDICINE MEDICAL IMAGING OR CRYSTALLOGRAPHY OR BIODIVERSITY CONSERVATION OR ENGINEERING MANUFACTURING OR HORTICULTURE OR ENGINEERING MECHANICAL OR OPERATIONS RESEARCH MANAGEMENT SCIENCE OR PHYSICS APPLIED OR CHEMISTRY ORGANIC OR IMMUNOLOGY OR ENDOCRINOLOGY METABOLISM ) AND [excluding] Web of Science Categories=( EDUCATION EDUCATIONAL RESEARCH OR MEDICAL INFORMATICS OR WOMEN S STUDIES OR ASTRONOMY ASTROPHYSICS OR COMMUNICATION OR STATISTICS PROBABILITY OR COMPUTER SCIENCE INFORMATION SYSTEMS OR COMPUTER SCIENCE THEORY METHODS OR CRIMINOLOGY PENOLOGY OR ENVIRONMENTAL STUDIES OR MATHEMATICAL COMPUTATIONAL BIOLOGY OR HEMATOLOGY OR TROPICAL MEDICINE OR PHYSICS MATHEMATICAL OR VIROLOGY OR GERONTOLOGY OR CHEMISTRY MEDICINAL OR MEDICINE LEGAL OR PSYCHOLOGY DEVELOPMENTAL OR UROLOGY NEPHROLOGY OR SOCIAL ISSUES OR IMAGING SCIENCE PHOTOGRAPHIC TECHNOLOGY OR OBSTETRICS GYNECOLOGY OR TRANSPORTATION OR LAW OR GEOCHEMISTRY GEOPHYSICS OR DERMATOLOGY OR MINERALOGY OR PHYSICS FLUIDS PLASMAS OR PHYSICS NUCLEAR OR GERIATRICS GERONTOLOGY OR ERGONOMICS OR SOCIAL SCIENCES MATHEMATICAL METHODS OR OPHTHALMOLOGY OR HOSPITALITY LEISURE SPORT TOURISM OR NURSING OR SOCIAL WORK OR FAMILY STUDIES OR EDUCATION SCIENTIFIC DISCIPLINES OR ANESTHESIOLOGY OR EMERGENCY MEDICINE OR MATERIALS SCIENCE PAPER WOOD OR GEOLOGY OR INFORMATION SCIENCE LIBRARY SCIENCE OR PARASITOLOGY OR POLITICAL SCIENCE OR PALEONTOLOGY OR MATHEMATICS INTERDISCIPLINARY APPLICATIONS OR ORTHOPEDICS OR RHEUMATOLOGY OR SOCIOLOGY OR REHABILITATION OR DEMOGRAPHY OR REPRODUCTIVE BIOLOGY OR MICROSCOPY OR ANATOMY MORPHOLOGY OR TELECOMMUNICATIONS OR OTORHINOLARYNGOLOGY OR ENGINEERING INDUSTRIAL OR AUTOMATION CONTROL SYSTEMS OR PHYSICS PARTICLES FIELDS OR MATHEMATICS OR DEVELOPMENTAL BIOLOGY OR PATHOLOGY OR ENGINEERING MULTIDISCIPLINARY OR INTEGRATIVE COMPLEMENTARY MEDICINE OR INFECTIOUS DISEASES OR PRIMARY HEALTH CARE OR ROBOTICS OR MATHEMATICS APPLIED OR MATERIALS SCIENCE TEXTILES OR URBAN STUDIES OR GEOGRAPHY OR MYCOLOGY OR INTERNATIONAL RELATIONS OR MEDICAL LABORATORY TECHNOLOGY OR COMPUTER SCIENCE SOFTWARE ENGINEERING OR MINING MINERAL PROCESSING OR COMPUTER SCIENCE ARTIFICIAL INTELLIGENCE OR MATERIALS SCIENCE COMPOSITES OR REMOTE SENSING OR PLANNING DEVELOPMENT ) AND [excluding] Web of Science Categories=( ACOUSTICS OR ENGINEERING MARINE OR MATERIALS SCIENCE CHARACTERIZATION TESTING OR ETHICS OR HISTORY OR HUMANITIES MULTIDISCIPLINARY OR INDUSTRIAL RELATIONS LABOR OR PSYCHOLOGY EDUCATIONAL OR MATERIALS SCIENCE BIOMATERIALS OR ALLERGY OR MEDICAL ETHICS OR MATERIALS SCIENCE COATINGS FILMS OR PHILOSOPHY OR CONSTRUCTION BUILDING TECHNOLOGY OR PSYCHOLOGY MATHEMATICAL OR AREA STUDIES OR PUBLIC ADMINISTRATION OR AUDIOLOGY SPEECH LANGUAGE PATHOLOGY OR TRANSPLANTATION OR COMPUTER SCIENCE HARDWARE ARCHITECTURE OR TRANSPORTATION SCIENCE TECHNOLOGY OR ENGINEERING GEOLOGICAL OR BUSINESS FINANCE OR ENGINEERING PETROLEUM OR CULTURAL STUDIES OR ETHNIC STUDIES OR ENGINEERING OCEAN OR GEOGRAPHY PHYSICAL OR HISTORY OF SOCIAL SCIENCES OR RELIGION OR HISTORY PHILOSOPHY OF SCIENCE OR ANDROLOGY OR MUSIC OR ENGINEERING AEROSPACE OR ARCHAEOLOGY OR NEUROIMAGING )

Trials Register of Promoting Health Interventions (EPPI Centre), 2004 to 30 January 2015

Original search executed: 23 November 2012; Retrieved: 477 records

Updated search executed: 30 January 2015; Retrieved 167 records

110 Focus of the report: alcohol OR healthy eating OR tobacco

111 Type(s) of intervention: environmental modification

112 110 AND 111

113 Freetext (item record) "unit*"

114 Freetext (item record) "portion*"

115 Freetext (item record) "serving*"

116 Freetext (item record) "product*"

117 Freetext (item record) "packag*"

118 Freetext (item record) "packet*"

119 Freetext (item record) "tableware"

120 Freetext (item record) "drinkware"

121 Freetext (item record) "dinnerware"

122 Freetext (item record) "crockery"

123 Freetext (item record) "plate*"

124 Freetext (item record) "platter*"

125 Freetext (item record) "tureen*"

126 Freetext (item record) "tajine*"

127 Freetext (item record) "tagine*"

128 Freetext (item record) "bowl*"

129 Freetext (item record) "charger*"

130 Freetext (item record) "cup*"

131 Freetext (item record) "saucer*"

132 Freetext (item record) "glass"

133 Freetext (item record) "glasses"

134 Freetext (item record) "mug"

135 Freetext (item record) "mugs"

136 Freetext (item record) "beaker*"

137 Freetext (item record) "pitcher*"

138 Freetext (item record) "jug*"

139 Freetext (item record) "decanter*"

140 Freetext (item record) "receptacle*"

141 Freetext (item record) "container*"

142 Freetext (item record) "dish*"

143 Freetext (item record) "pot"

144 Freetext (item record) "pots"

145 Freetext (item record) "cutlery"

146 Freetext (item record) "flatware"

147 Freetext (item record) "utensil*"

148 Freetext (item record) "knife"

149 Freetext (item record) "*knife"

150 Freetext (item record) "knives"

151 Freetext (item record) "fork"

152 Freetext (item record) "fork*"

153 Freetext (item record) "spoon*"

154 Freetext (item record) "*spoon"

155 Freetext (item record) "tongs"

156 Freetext (item record) "ladle*"

157 Freetext (item record) "chopstick*"

158 Freetext (item record) "box*"

159 Freetext (item record) "bag*"

160 Freetext (item record) "cans"

161 Freetext (item record) "carton*"

162 Freetext (item record) "bottle*"

163 Freetext (item record) "straw*"

164 113 OR 114 OR 115 OR 116 OR 117 OR 118 OR 119 OR 120 OR 121 OR 122 OR 123 OR 124 OR 125 OR 126 OR 127 OR 128 OR 129 OR 130 OR 131 OR 132 OR 133 OR 134 OR 135 OR 136 OR 137 OR 138 OR 139 OR 140 OR 141 OR 142 OR 143 OR 144 OR 145 OR 146 OR 147 OR 148 OR 149 OR 150 OR 151 OR 152 OR 153 OR 154 OR 155 OR 156 OR 157 OR 158 OR 159 OR 160 OR 161 OR 162 OR 163

165 Freetext (item record) "drink*"

166 Freetext (item record) "drunk*"

167 Freetext (item record) "alcohol*"

168 Freetext (item record) "beverage*"

169 Freetext (item record) "beer*"

170 Freetext (item record) "lager*"

171 Freetext (item record) "wine*"

172 Freetext (item record) "cider*"

173 Freetext (item record) "alcopop*"

174 Freetext (item record) "alco-pop*"

175 Freetext (item record) "spirit"

176 Freetext (item record) "spirits"

177 Freetext (item record) "liquor*"

178 Freetext (item record) "liquer*"

179 Freetext (item record) "liqueur*"

180 Freetext (item record) "whisk*"

181 Freetext (item record) "schnapps"

182 Freetext (item record) "brandy"

183 Freetext (item record) "brandies"

184 Freetext (item record) "gin"

185 Freetext (item record) "gins"

186 Freetext (item record) "rum"

187 Freetext (item record) "rums"

188 Freetext (item record) "tequila*"

189 Freetext (item record) "vodka*"

190 Freetext (item record) "cocktail*"

191 Freetext (item record) "cigar*"

192 Freetext (item record) "smoke"

193 Freetext (item record) "smokes"

194 Freetext (item record) "smoking"

195 Freetext (item record) "smoker"

196 Freetext (item record) "smokers"

197 Freetext (item record) "smoked"

198 Freetext (item record) "tobacco*"

199 Freetext (item record) "nutri*"

200 Freetext (item record) "calori*"

201 Freetext (item record) "food*"

202 Freetext (item record) "eat"

203 Freetext (item record) "eats"

204 Freetext (item record) "eaten"

205 Freetext (item record) "eating"

206 Freetext (item record) "ate"

207 Freetext (item record) "meal"

208 Freetext (item record) "meal*"

209 Freetext (item record) "snack*"

210 165 OR 166 OR 167 OR 168 OR 169 OR 170 OR 171 OR 172 OR 173 OR 174 OR 175 OR 176 OR 177 OR 178 OR 179 OR 180 OR 181 OR 182 OR 183 OR 184 OR 185 OR 186 OR 187 OR 188 OR 189 OR 190 OR 191 OR 192 OR 193 OR 194 OR 195 OR 196 OR 197 OR 198 OR 199 OR 200 OR 201 OR 202 OR 203 OR 204 OR 205 OR 206 OR 207 OR 208 OR 209

211 164 AND 210

212 112 OR 211

213 114 OR 115 OR 116 OR 117 OR 118 OR 119 OR 120 OR 121 OR 122 OR 123 OR 124 OR 125 OR 126 OR 127 OR 128 OR 129 OR 130 OR 131 OR 132 OR 133 OR 134 OR 135 OR 136 OR 137 OR 138 OR 139 OR 140 OR 141 OR 142 OR 143 OR 144 OR 145 OR 146 OR 147 OR 148 OR 149 OR 150 OR 151 OR 152 OR 153 OR 154 OR 155 OR 156 OR 157 OR 158 OR 159 OR 160 OR 161 OR 162 OR 163

Open Grey (http://www.opengrey.eu), 1980 to 30 January 2015

Search executed: 30 January 2015; Retrieved 367 records

(drink* OR drunk* OR alcohol* OR beverage* OR beer* OR lager* OR wine* OR cider* OR alcopop* OR alco-pop* OR spirit OR spirits OR liquor* OR liquer* OR liqueur* OR whisky OR whiskey OR whiskies OR whiskeys OR schnapps OR brandy OR brandies OR gin OR gins OR rum OR rums OR tequila* OR vodka* OR cocktail* OR cigar* OR smoke OR smokes OR smoking OR smoker OR smokers OR smoked OR tobacco* OR nutri* OR calori* OR food* OR eat OR eats OR eaten OR eating OR ate OR meal* OR snack*) AND ((siz* OR dimension* OR capacit* OR volume* OR shap* OR height* OR width* OR length* OR depth* OR divide*) NEAR/6 (portion* OR serving* OR product* OR packag* OR packet* OR unit* OR cigar* OR food* OR drink* OR alcohol* OR tableware OR drinkware OR dinnerware OR crockery OR plate* OR platter* OR tureen* OR tajine* OR tagine* OR bowl* OR charger* OR cup* OR saucer* OR glass OR glasses OR mug OR mugs OR beaker* OR pitcher* OR jug* OR decanter* OR receptacle* OR container* OR dish* OR pot OR pots OR cutlery OR flatware OR utensil* OR knife OR *knife OR knives OR fork* OR spoon* OR *spoon OR tongs OR ladle* OR chopstick* OR box* OR bag* OR can* OR carton* OR bottle* OR straw*))

Appendix 2. Preliminary analyses of minimum data extracted from 11 eligible studies identified by the updated search

Introduction

The updated search conducted up to 30 January 2015 identified 11 further eligible studies published during 2013 and 2014 (see also Search methods for identification of studies, Results of the search and Appendix 1). Key characteristics of each of these 11 eligible studies (Bajaj 2014; Haire 2014; Kral 2014; Marchiori 2014; Rolls 2014a; Smith 2013a; van Ittersum 2013; van Kleef 2014; Wansink 2013; Wansink 2014; Williams 2014) are described in Characteristics of studies awaiting classification (the information in Characteristics of studies awaiting classification is based on the minimum data set that we provisionally extracted from the 12 corresponding study reports - see below in this section).

All 11 further eligible studies have been accepted into the review and currently await full integration, which is scheduled for the first major update. At that stage we will: collect the maximum data set for each study (comprising = 1000 variables) from the 12 corresponding study reports (including supplementary coding based on external data sources and contacts with study authors to request data that are not available in study reports); conduct 'Risk of bias' assessments; update meta-analyses; update meta-regression analyses; update GRADE assessments; and make corollary updates to the Results, Discussion and Authors' conclusions sections of the review, including 'Summary of findings' tables (see also Data collection and analysis).

However, in advance of their full integration into this review, it was important to establish whether the pending full integration of these 11 eligible studies has any potential to change the interpretation of the results of this review, and hence its conclusions, as these are currently reported in the Results, Discussion and Authors' conclusions. These sections are currently based exclusively on evidence collected from the 72 included studies identified by the original search and published between 1978 and July 2013 (see also Search methods for identification of studies, Results of the search and Figure Figure22).

We therefore conducted preliminary statistical analyses to investigate this issue based on outcome data that could provisionally be extracted from each of the 11 further eligible studies (i.e. in advance of contacting study authors, with one exception - see 'Potential impact of studies with no useable data', below).

Procedure

We provisionally extracted useable outcome data with respect to each eligible independent within-study comparison identified in these 11 studies (Bajaj 2014; Haire 2014; Kral 2014; Marchiori 2014; Rolls 2014a; Smith 2013a; van Ittersum 2013; van Kleef 2014; Wansink 2013; Wansink 2014; Williams 2014). We then provisionally computed study-level effect sizes for each eligible independent within-study comparison as the standardised difference in means (SMD) and its standard error, with respect to consumption and selection outcomes (as applicable). We then integrated provisional study-level effect sizes that could be computed from these 11 studies with those previously computed from 70 of 72 studies included studies identified by the original search, using random-effects meta-analysis (i.e. we applied the same procedures described in Data collection and analysis to provisionally update meta-analyses). Finally, we assessed the potential for full integration of these 11 studies to change current quality of evidence ratings with respect to provisionally updated estimates of summary effect sizes using the GRADE system (see Data synthesis).

Results

We identified a total of 17 eligible independent within-study comparisons (i.e. measurement of at least one of our specified outcomes) in the 11 further eligible studies (Bajaj 2014; Haire 2014; Kral 2014; Marchiori 2014; Rolls 2014a; Smith 2013a; van Ittersum 2013; van Kleef 2014; Wansink 2013; Wansink 2014; Williams 2014):

  • 16 comparisons assessed the effect of larger versus smaller-sized portions, packages or tableware on consumption of food; and
  • six comparisons assessed the effect of larger versus smaller-sized portions, packages or tableware on selection of food.

This established that full integration of these 11 studies could only influence the results of two meta-analyses (and related findings), which investigated:

  • the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food consumed (Table 1); and
  • the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food selected (see Table 1).

Table A2.1 shows effect sizes provisionally computed for each eligible independent within-study comparison identified in the 11 studies used in these preliminary analyses. For the consumption outcome, we extracted useable data with respect to 14 of 16 independent comparisons (nine of 11 studies). No useable consumption outcome data could be extracted from van Ittersum 2013. This was a paired study and the corresponding study report does not provide sufficient information (notably, the correlation coefficient) to enable estimation of the correct standard deviation or SMD based on reported F-statistics. In addition, no useable consumption outcome data could be extracted from Wansink 2013 due to unclear reporting of results from the relevant intention-to-treat (ITT) analysis. For the selection outcome, we extracted useable data with respect to four of six independent comparisons (four of six studies). No useable selection outcome data could be extracted from van Ittersum 2013 or Wansink 2013 for the same reasons given above.

Table A2.1 Study-level effect sizes

ConsumptionSelection

ComparisonSMD (95% CI)SEInterpretationSMD (95% CI)SEInterpretation
Bajaj 20140.23 (0.01 to 0.45)0.11Larger size increased consumptionNot measured--

Haire 20140.23 (-0.26 to 0.72)0.25No differenceNot measured--

Kral 2014 [1]0.43 (-0.05 to 0.91)0.25No differenceNot measured

Kral 2014 [2]-0.02 (-0.50 to 0.46)0.24No differenceNot measured--

Marchiori 20140.81 (0.42 to 1.20)0.20Larger size increased consumptionNot measured--

Rolls 2014a [1]-0.32 (-0.85 to 0.21)0.27No difference-0.35 (-0.88 to 0.18)0.27No difference

Rolls 2014a [2]-0.35 (-0.97 to 0.27)0.32No difference-0.36 (-0.98 to 0.26)0.32No difference

Rolls 2014a [3]-0.15 (-0.68 to 0.38)0.27No difference-0.32 (-0.86 to 0.22)0.28No difference

Smith 2013a [1]-0.96 (-1.33 to -0.59)0.19Larger size reduced consumptionNot measured--

Smith 2013a [2]1.04 (0.67 to 1.41)0.19Larger size increased consumptionNot measured--

Smith 2013a [3]0.67 (0.27 to 1.07)0.20Larger size increased consumptionNot measured--

Smith 2013a [4]0.61 (0.22 to 1.00)0.20Larger size increased consumptionNot measured--

van Ittersum 2013No useable data--No useable data--

van Kleef 20140.48 (0.17 to 0.79)0.16Larger size increased consumptionNot measured--

Wansink 2013No useable data--No useable data--

Wansink 2014Not measured--1.41 (0.88 to 1.94)0.27Larger size increased selection

Williams 20140.46 (0.05 to 0.87)0.21Larger size increased consumptionNot measured--

The first row of Table A2.2 (below) reproduces the result of the meta-analysis that we conducted to investigate (1) the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food consumed (see also Table 1). This meta-analysis was based on outcome data from a total of 6603 participants (86 independent comparisons). The second row of Table A2.2 shows the provisional result from a preliminary meta-analysis that integrates outcome data from an additional 1591 participants (15 independent comparisons); a combined total N of 9785 participants (101 independent comparisons).

Table A2.2. Effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food consumed

Independent comparisons (N)Total participants (N)SMD95% CI lower bound95% CI upper boundI2
8666030.380.290.4661%

10097480.350.270.4468%

The first row of Table A2.3 reproduces the result of the meta-analysis that was conducted to investigate (2) the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food selected (see also Table 1). This meta-analysis was based on outcome data from a total of 1164 participants (13 independent comparisons). The second row of Table A2.3 shows the provisional result from a preliminary meta-analysis that integrates outcome data from an additional 194 participants (four independent comparisons); a combined total N of 1358 participants (17 independent comparisons).

Table A2.3. Effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food selected

Independent comparisons (N)Total participants (N)SMD95% CI lower bound95% CI upper boundI2
1311640.420.240.5954%

1713580.360.150.5773%

As shown in Tables A2.2 and A2.3, point estimates and 95% confidence intervals from random-effects models are similar between the current and provisionally updated results of these meta-analyses. Critically, provisionally updated results remain consistent with the current findings of this review (see Discussion and Authors' conclusions) that exposure to larger versus smaller-sized portions, packages or tableware increased both quantities of food consumed and quantities of food selected for consumption, and that the sizes of these effects were small to moderate in relative terms.

Table A2.4 summarises the results of our quality of evidence ratings with respect to current and provisionally updated estimates of the summary effect size for (1) the effect of exposure to larger versus smaller sized portions, packages or tableware on quantities of food consumed.

Table A2.4 Review of quality of evidence ratings: consumption

Independent comparisons (N)Total participants (N)Risk of biasInconsistencyIndirectnessImprecisionOther considerationsOverall quality rating
Current866603Serious limitationsNot seriousNot seriousNot seriousNoneModerate

Provisionally updated1009748Serious limitationsNot seriousNot seriousNot seriousNoneModerate

With respect to risk of bias, we already rated current evidence (86 independent comparisons) down by one level (i.e. serious limitations) due to all study-level estimates of this effect having been judged to be at 'unclear or high risk of bias'. Therefore, even in the extreme hypothetical scenarios that all further eligible studies are in due course judged to be either at 'low' or 'unclear' or 'high' risk of bias with respect to their study-level estimates of this effect, integration of these assessments (with respect to 16 further independent comparisons) cannot change the current rating (i.e. serious limitations).

With respect to inconsistency, we did not rate down current evidence (86 independent comparisons) based on our judgement that large inconsistency (heterogeneity) in study results did not remain after exploration of a priori hypotheses that might explain heterogeneity (i.e. potential effect modifiers) using meta-regression analysis (see Data synthesis). Whilst the full integration of data concerning potential effect modifiers yet to be collected from further eligible studies (independent comparisons) into updated meta-regression analyses will inevitably influence the detailed results of those analyses, we judge that the likelihood of the current rating (i.e. 'Not serious') could change as a consequence is minimal.

With respect to indirectness, we did not rate down current evidence (86 independent comparisons) based on our judgement that all included studies (within-study comparisons) assessed interventions, comparators and outcomes that met eligibility criteria for this review in participant samples that also met eligibility criteria, and were all direct head-to-head comparisons. As such, there were no differences between the populations, interventions or outcomes measured among included studies and those under consideration in the current review. The same is also true of the 10 of 11 further eligible studies accepted into the review and currently awaiting full integration that measured the consumption outcome (see Characteristics of studies awaiting classification). Therefore, full integration of these further eligible studies cannot change the current rating (i.e. 'Not serious').

With respect to imprecision, we did not rate down current evidence (86 independent comparisons) based on examination of the upper and lower bounds of 95% confidence intervals associated with the estimated summary effect size, coupled with the consideration that the number of participants (effective sample size) incorporated into this meta-analysis exceeded the number of participants generated by a conventional sample size calculation for a single adequately powered trial (optimal information size). Since full integration of further eligible studies will increase the number of participants (effective sample size) incorporated into an updated version of this meta-analysis, this cannot change the current rating (i.e. 'Not serious').

With respect to other considerations, we judged that there were 'None' associated with current evidence (86 independent comparisons) on the basis that none of the primary reasons suggested by the GRADE system for rating up quality of evidence (Guyatt 2011) were applicable in this case. Based on provisional results of the relevant preliminary analysis reported above (see Table A2.2), we judge the likelihood that the current rating (i.e. 'None') could change as a consequence of full integration of data from 10 of 11 further eligible studies that measured the consumption outcome is minimal.

In summary, our review of quality of evidence ratings establishes that full integration of 10 further eligible studies accepted into the review and currently awaiting full integration that measured the consumption outcome cannot change the overall quality of evidence rating with respect to the provisionally updated estimate of the summary effect size for (1) the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food consumed.

Table A2.5 summarises the results of our quality of evidence ratings with respect to current and provisionally updated estimates of the summary effect size for (2) the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food selected.

Table A2.5 Review of quality of evidence ratings: selection

Independent comparisons (N)Total participants (N)Risk of biasInconsistencyIndirectnessImprecisionOther considerationsOverall quality rating
Current131164Serious limitationsNot seriousNot seriousNot seriousNoneModerate

Provisionally updated171358Serious limitationsNot seriousNot seriousNot seriousNoneModerate

Identical considerations to those described above in the case of the effect on consumption apply here with respect to ratings of risk of bias, inconsistency, indirectness, imprecision and other considerations that collectively determine confidence in estimates of the effect of exposure to larger versus smaller size on food selection. In summary, this review of quality of evidence ratings establishes that full integration of six further eligible studies accepted into the review and currently awaiting full integration that measured the selection outcome cannot change the overall quality of evidence rating with respect to the provisionally updated estimate of the summary effect size for the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food selected.

Potential impact of studies with no useable data

As stated above no useable data could be extracted from the Wansink 2013 study with respect to either the consumption or the selection outcome due to unclear reporting of results from the relevant intention-to-treat (ITT) analysis. As noted in Characteristics of studies awaiting classification the Wansink 2013 study was a between-subjects cluster-randomised controlled trial that included investigation of the effects of 'exposure to whole apples available for purchase in the school lunchroom' (larger individual unit size), versus 'exposure to apples sliced into six symmetric pieces available for purchase in the school lunchroom' (smaller individual unit size). The study randomised six middle schools (clusters) comprising a total of 2150 participants (students) to these two comparison groups: 'whole apple schools' (larger individual unit size) and 'sliced apple schools' (smaller individual unit size).

Outcomes in this study included measures of both selection and consumption that are eligible for inclusion in meta-analyses (1) and (2) respectively. The selection outcome appears to have been measured as the numbers of students who purchased (and did not purchase) an apple on study days in 'whole apple schools' and 'sliced apple schools' respectively. Based on these data it would in principle be possible to construct a 2 x 2 table in order to compute a log odds ratio and its standard error, which could then be converted into a useable SMD and its SE using the formula provided in Section 9.4.6 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011). However, in order to apply this procedure we would first need confirmation from study authors of the following data, which are currently unclear in the corresponding study report (Wansink 2013): the numbers of participants in schools randomised to each comparison group (i.e. 'whole apple schools' and 'sliced apple schools'); and the numbers of participants who purchased and did not purchase an apple on study days in 'whole apple schools' and 'sliced apple schools' respectively. The consumption outcome appears to have been measured as the amount of apple consumed in grams per student on study days in 'whole apple schools' and 'sliced apple schools' respectively. However, in order to compute a SMD and its standard error based on these data, we need both the standard deviations and denominators (i.e. numbers of participants in 'whole apple schools' and 'sliced apple schools') associated with reported mean gram amounts of consumption in 'whole apple schools' and 'sliced apple schools' respectively. These numerical data are (respectively) not reported and ambiguous in the corresponding study report (in the latter case it is also unclear whether or not the denominators reflect the randomised allocation).

Since Wansink 2013 was a large study (with an effective sample size of 4300 participants), we sought these numerical results by contacting the corresponding author, but to date of publication of this review we have received a response but not the necessary data. This is consistent with previous contacts with the author that we initiated to request numerical results that are missing from, or unclear in, published reports of several of their other 11 studies already included in this review (Wansink 1996a (S1); Wansink 1996b (S2); Wansink 1996c (S4); Wansink 2001; Wansink 2003 (S1); Wansink 2003 (S2); Wansink 2005b; Wansink 2005d; Wansink 2006; Wansink 2011a (S4); Wansink 2011b). Whilst we have received responses to our previous contacts, the author was unable or unwilling to provide the requested data. As such, no useable outcome data have to date been collected from the Wansink 2013 that could be incorporated into the preliminary analyses presented above.

Therefore, whilst the potential impact of integrating data from Wansink 2013 into further updated meta-analyses of (1) and (2) the effects of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food consumed and selected may be substantive, this cannot currently be established with any confidence and we judge the likelihood of obtaining useable data from the study authors to be low. To illustrate, with respect to the selection outcome, if we assumed that: (a) there were equal numbers of participants in schools randomised to each comparison group, (b) the denominator reported in Wansink 2013, Table 1, Row 1 ("n=334") was the 'total number of apples purchased' on study days in 'whole apple schools' and 'sliced apple schools' combined; and (c) the figures 6% and 10% in Wansink 2013 Table 1, Row 1 reflect the relative numbers of apples purchased on study days in 'whole apple schools' and 'sliced apple schools' respectively – then it would be possible to estimate a SMD and its standard error using the procedure described above as SMD -0.31 (SE 0.0647226) (were the latter estimate integrated into meta-analysis (2), the summary effect size would be SMD 0.01 (95% CI -0.01 to 0.16)). However, it is important to highlight assumptions (a), (b) and (c) have not been verified and are likely to be incorrect, and moreover that this estimate of the study level SMD and its standard error are sensitive to variation in these assumptions. With respect to the consumption outcome, it was not judged credible to make assumptions needed to enable provisional estimation of a SMD and its standard error, due to the level of ambiguity in the reporting of these outcome data and the lack of scope for imputing data from similar studies in this specific case. On the latter point, Wansink 2013 has distinctive characteristics that differentiate it from the other studies included and accepted for inclusion in this review. For example, this is the only eligible study identified to date which included a measure of the effect on purchasing (i.e. selection with purchase) and that this is the only cluster-randomised trial identified to date that includes a measure of selection (with or without purchase). Based on these considerations, we may propose to produce further updates of meta-analyses (1) and (2) for the first major update of this review both without outcome data from Wansink 2013 (primary analyses) and with outcome data from Wansink 2013 (sensitivity analysis), subject to being able to obtain useable data from the study authors.

The second study with no useable data was van Ittersum 2013. Since this was a small study (effective sample size of 36), we judge that full integration of outcome data from this study into meta-analyses (1) and (2) will have no substantive impact on current estimates of summary effect sizes.

Conclusions

The results of the preliminary analyses reported here in Appendix 2 (see also Characteristics of studies awaiting classification) establish that there is minimal potential for full integration 11 further eligible studies identified by the updated search to change the interpretation of the results of this review, and hence its conclusions, as these are currently reported in the Results, Discussion and Authors' conclusions. This conclusion is based on the following key findings:

  • Interpretation of the result of an updated meta-analysis of (1) the effect of exposure to larger versus smaller-sized portions, packages or tableware on quantities of food consumed will not change: there will still be overall moderate quality evidence that larger portion, package and tableware size increased consumption of food, with a small to moderate effect size.
  • Interpretation of the result of an updated meta-analysis of (2) the effect of exposure to larger versus smaller sized portions, packages or tableware on quantities of food selected will not change: there will still be overall moderate quality evidence that larger portion, package and tableware size increased selection of food, with a small to moderate effect size.
  • Overall quality of evidence ratings cannot change with respect to updated summary estimates of (1) and (2) the effects of exposure to larger versus smaller sized portions, packages or tableware on quantities of food consumed and selected.

Finally (as described above), we plan to fully integrate these 11 further eligible studies (Bajaj 2014; Haire 2014; Kral 2014; Marchiori 2014; Rolls 2014a; Smith 2013a; van Ittersum 2013; van Kleef 2014; Wansink 2013; Wansink 2014; Williams 2014) into this review as part of the process of conducting its first major update.

Appendix 3. Full results of meta-regression analyses conducted to investigate modifiers of the effect of larger size on consumption

Variable namenumincl_exclcoefcoef1coef2coef3coef4coef5
Sel_Pur4Only one categoryNANANANANANA

Prod_Type92Not significantNA-0.13[-0.65,0.38]NANANANA

Soc_Setting92Not significantNA-0.30[-0.64,0.05]-0.14 [-0.50,0.21]-0.30 [-0.97,0.37]NANA

FSA_Meth57Not significant0.02 [-0.21,0.24]NANANANANA

FSA_Score57Included0.01 [0.00,0.02]NANANANANA

En_Density57Included0.04 [-0.00,0.08]NANANANANA

Manip_Target92Not significantNA0.21 [-0.22,0.64]-0.11 [-0.62,0.40]0.04 [-0.33,0.40]-0.04 [-0.46,0.37]NA

Manip_Type92Only one categoryNANANANANANA

Dur_Exposure92Not significant0.23 [-0.02,0.48]NANANANANA

Conc_Int92Not significant-0.22 [-0.54,0.09]NANANANANA

SES_Context92Not significantNA0.15[-0.27,0.57]NANANANA

F_O_173Included0.22 [0.02,0.41]NANANANANA

F_O_273Not significant-0.12 [-0.38,0.15]NANANANANA

F_O_373Not significant-0.13 [-0.32,0.05]NANANANANA

F_O_486Included0.32 [0.16,0.48]NANANANANA

Size_Abs52Not significant0.00 [-0.00,0.00]NANANANANA

Size_Rel80Not significant-0.00 [-0.00,0.00]NANANANANA

Age_Mean74Included0.01 [-0.00,0.02]NANANANANA

Female_Percent86Not significant0.00 [-0.00,0.01]NANANANANA

Eth_White_Percent21Not significant0.00 [-0.00,0.00]NANANANANA

BMI_Mean52Not significant-0.01 [-0.05,0.04]NANANANANA

BMIz_Mean5Insufficient dataNANANANANANA

BodFat_Mean2Insufficient dataNANANANANANA

Weight_Mean41Not significant0.00 [-0.00,0.01]NANANANANA

Overweight_Percent19Not significant0.00 [-0.01,0.01]NANANANANA

Obese_Percent10Not significant0.01 [-0.02,0.05]NANANANANA

Overweight_Obese_Percent6Insufficient dataNANANANANANA

Restraint_1_Mean32Not significant0.01 [-0.09,0.10]NANANANANA

Restraint_2_Mean4Insufficient dataNANANANANANA

Restraint_3_Mean3Insufficient dataNANANANANANA

Disinhib_1_Mean29Not significant-0.05 [-0.27,0.17]NANANANANA

Disinhib_2_Mean1Insufficient dataNANANANANANA

ExEat_Mean4Insufficient dataNANANANANANA

EmEat_Mean3Insufficient dataNANANANANANA

PClean_Mean2Insufficient dataNANANANANANA

PClean_Ad_Percent3Insufficient dataNANANANANANA

PClean_Ch_Percent3Insufficient dataNANANANANANA

ConsMon_Mean2Insufficient dataNANANANANANA

Binge_1_Mean2Insufficient dataNANANANANANA

Binge_2_Mean1Insufficient dataNANANANANANA

Diet_Mean14Not significant-0.07[-0.15,0.01]NANANANANA

Mood_Mean2Insufficient dataNANANANANANA

EnInt_Mean2Insufficient dataNANANANANANA

Carb_Mean1Insufficient dataNANANANANANA

Prot_Mean1Insufficient dataNANANANANANA

Fat_Mean1Insufficient dataNANANANANANA

Step_Mean1Insufficient dataNANANANANANA

EnExp_Mean16Not significant-0.00[-0.00,0.00]NANANANANA

Exerc_Mean1Insufficient dataNANANANANANA

Hunger_1_Mean29Not significant-0.13[-0.33,0.07]NANANANANA

Hunger_2_Mean8Insufficient dataNANANANANANA

Hunger_3_Mean1Insufficient dataNANANANANANA

Hunger_4_Mean1Insufficient dataNANANANANANA

Fullness_Mean1Insufficient dataNANANANANANA

Sat_Mean1Insufficient dataNANANANANANA

ProsCon_Mean1Insufficient dataNANANANANANA

Depress_Mean12Not significant-0.22[-0.50,0.07]NANANANANA

Employ_Percent2Insufficient dataNANANANANANA

Par_Employ_Percent7Insufficient dataNANANANANANA

EduYears_Mean1Insufficient dataNANANANANANA

EduHigh_Percent2Insufficient dataNANANANANANA

Par_EduHigh_Percent8Insufficient dataNANANANANANA

Par_EduDeg_Percent5Insufficient dataNANANANANANA

Inc50_Percent1Insufficient dataNANANANANANA

FamInc50_Percent5Insufficient dataNANANANANANA

Insec_Percent3Insufficient dataNANANANANANA

NSLP_Percent1Insufficient dataNANANANANANA

SNAP_Percent0Insufficient dataNANANANANANA

ROBSum_Sel92Not significantNA-0.10[-0.47,0.27]NANANANA

ROBSum_Con92Not significantNA-0.24[-0.61,0.13]NANANANA

design192Not significant-0.14 [-0.38,0.09]NANANANANA

design292Included-0.40 [-0.55,-0.25]NANANANANA

design392Not significant0.07 [-0.13,0.26]NANANANANA

Appendix 4. Full results of meta-regression analyses conducted to investigate modifiers of the effect of larger size on selection

Variable namenumincl_exclcoefcoef1coef2coef3coef4coef5
Sel_Pur13Only one categoryNANANANANANA

Prod_Type13Only one categoryNANANANANANA

Soc_Setting13Not significantNA0.15 [-0.27,0.58]NANANANA

FSA_Meth11Not significant-0.49 [-1.14,0.16]NANANANANA

FSA_Score11Not significant-0.01 [-0.06,0.04]NANANANANA

En_Density11Not significant-0.02 [-0.23,0.19]NANANANANA

Manip_Target13Not significantNA0.22 [-0.63,1.07]0.21 [-0.25,0.68]NANANA

Manip_Type13Only one categoryNANANANANANA

Dur_Exposure13Not significant-0.51 [-1.33,0.31]NANANANANA

Conc_Int13Not significant-0.22 [-1.03,0.60]NANANANANA

SES_Context13Not significantNA0.22 [-0.60,1.03]NANANANA

F_O_17Insufficient dataNANANANANANA

F_O_27Insufficient dataNANANANANANA

F_O_37Insufficient dataNANANANANANA

F_O_413Included0.41 [0.06,0.76]NANANANANA

Size_Abs4Insufficient dataNANANANANANA

Size_Rel11Not significant-0.00 [-0.02,0.01]NANANANANA

Age_Mean6Insufficient dataNANANANANANA

Female_Percent13Not significant0.00 [-0.01,0.01]NANANANANA

Eth_White_Percent4Insufficient dataNANANANANANA

BMI_Mean2Insufficient dataNANANANANANA

BMIz_Mean2Insufficient dataNANANANANANA

BodFat_Mean0Insufficient dataNANANANANANA

Weight_Mean0Insufficient dataNANANANANANA

Overweight_Percent0Insufficient dataNANANANANANA

Obese_Percent0Insufficient dataNANANANANANA

Overweight_Obese_Percent1Insufficient dataNANANANANANA

Restraint_1_Mean0Insufficient dataNANANANANANA

Restraint_2_Mean0Insufficient dataNANANANANANA

Restraint_3_Mean1Insufficient dataNANANANANANA

Disinhib_1_Mean0Insufficient dataNANANANANANA

Disinhib_2_Mean0Insufficient dataNANANANANANA

ExEat_Mean0Insufficient dataNANANANANANA

EmEat_Mean0Insufficient dataNANANANANANA

PClean_Mean0Insufficient dataNANANANANANA

PClean_Ad_Percent0Insufficient dataNANANANANANA

PClean_Ch_Percent0Insufficient dataNANANANANANA

ConsMon_Mean0Insufficient dataNANANANANANA

Binge_1_Mean0Insufficient dataNANANANANANA

Binge_2_Mean0Insufficient dataNANANANANANA

Diet_Mean0Insufficient dataNANANANANANA

Mood_Mean0Insufficient dataNANANANANANA

EnInt_Mean0Insufficient dataNANANANANANA

Carb_Mean0Insufficient dataNANANANANANA

Prot_Mean0Insufficient dataNANANANANANA

Fat_Mean0Insufficient dataNANANANANANA

Step_Mean0Insufficient dataNANANANANANA

EnExp_Mean0Insufficient dataNANANANANANA

Exerc_Mean0Insufficient dataNANANANANANA

Hunger_1_Mean0Insufficient dataNANANANANANA

Hunger_2_Mean0Insufficient dataNANANANANANA

Hunger_3_Mean0Insufficient dataNANANANANANA

Hunger_4_Mean0Insufficient dataNANANANANANA

Fullness_Mean0Insufficient dataNANANANANANA

Sat_Mean0Insufficient dataNANANANANANA

ProsCon_Mean0Insufficient dataNANANANANANA

Depress_Mean0Insufficient dataNANANANANANA

Employ_Percent0Insufficient dataNANANANANANA

Par_Employ_Percent4Insufficient dataNANANANANANA

EduYears_Mean0Insufficient dataNANANANANANA

EduHigh_Percent0Insufficient dataNANANANANANA

Par_EduHigh_Percent4Insufficient dataNANANANANANA

Par_EduDeg_Percent0Insufficient dataNANANANANANA

Inc50_Percent1Insufficient dataNANANANANANA

FamInc50_Percent0Insufficient dataNANANANANANA

Insec_Percent1Insufficient dataNANANANANANA

NSLP_Percent1Insufficient dataNANANANANANA

SNAP_Percent2Insufficient dataNANANANANANA

ROBSum_Sel13Not significantNA0.02 [-0.45,0.49]NANANANA

ROBSum_Con13Not significantNA0.15 [-0.27,0.58]NANANANA

design113Not significant-0.32 [-0.76,0.12]NANANANANA

design213Included-0.41 [-0.76,-0.06]NANANANANA

design313Not significant0.08 [-0.39,0.56]NANANANANA

CHARACTERISTICS OF STUDIES

Characteristics of included studies [ordered by study ID]

Table thumbnail
Ahn 2010

Table thumbnail
Argo 2012 (S1)

Table thumbnail
Argo 2012 (S2)

Table thumbnail
Argo 2012 (S4)

Table thumbnail
Argo 2012 (S5)

Table thumbnail
Burger 2011

Table thumbnail
Cavanagh 2013

Table thumbnail
Coelho do Vale 2008 (S2)

Table thumbnail
Devitt 2004

Table thumbnail
Diliberti 2004

Table thumbnail
DiSantis 2013

Table thumbnail
Ebbeling 2007

Table thumbnail
Fisher 2003

Table thumbnail
Fisher 2007a

Table thumbnail
Fisher 2007b

Table thumbnail
Fisher 2007c

Table thumbnail
Fisher 2013

Table thumbnail
Flood 2006

Table thumbnail
Goldstein 2006

Table thumbnail
Hermans 2012

Table thumbnail
Huss 2013

Table thumbnail
Jarvik 1978 (E1)

Table thumbnail
Jarvik 1978 (E2)

Table thumbnail
Jeffery 2007

Table thumbnail
Kelly 2009

Table thumbnail
Koh 2009

Table thumbnail
Kral 2004a

Table thumbnail
Kral 2010

Table thumbnail
Leahy 2008

Table thumbnail
Levitsky 2004

Table thumbnail
Looney 2011

Table thumbnail
Marchiori 2011

Table thumbnail
Marchiori 2012a

Table thumbnail
Marchiori 2012c

Table thumbnail
Mathias 2012

Table thumbnail
Mishra 2012 (S1)

Table thumbnail
Mishra 2012 (S2)

Table thumbnail
Raynor 2007

Table thumbnail
Raynor 2009

Table thumbnail
Rolls 2000

Table thumbnail
Rolls 2002

Table thumbnail
Rolls 2004a

Table thumbnail
Rolls 2004b

Table thumbnail
Rolls 2006a

Table thumbnail
Rolls 2006b

Table thumbnail
Rolls 2007a

Table thumbnail
Rolls 2007b (S1)

Table thumbnail
Rolls 2007b (S2)

Table thumbnail
Rolls 2007b (S3)

Table thumbnail
Rolls 2010a (E1)

Table thumbnail
Rolls 2010b (E2)

Table thumbnail
Russell 1980

Table thumbnail
Scott 2008b (S2)

Table thumbnail
Scott 2008c (S3)

Table thumbnail
Scott 2008d (S4)

Table thumbnail
Shah 2011

Table thumbnail
Spill 2010

Table thumbnail
Spill 2011b

Table thumbnail
Stroebele 2009

Table thumbnail
van Kleef 2012

Table thumbnail
van Kleef 2013

Table thumbnail
Wansink 1996a (S1)

Table thumbnail
Wansink 1996b (S2)

Table thumbnail
Wansink 1996c (S4)

Table thumbnail
Wansink 2001

Table thumbnail
Wansink 2003 (S1)

Table thumbnail
Wansink 2003 (S2)

Table thumbnail
Wansink 2005b

Table thumbnail
Wansink 2005d

Table thumbnail
Wansink 2006

Table thumbnail
Wansink 2011a (S4)

Table thumbnail
Wansink 2011b

Characteristics of excluded studies [ordered by study ID]

Table thumbnail

Characteristics of studies awaiting assessment [ordered by study ID]

Table thumbnail
Bajaj 2014

Table thumbnail
Haire 2014

Table thumbnail
Kral 2014

Table thumbnail
Loney 2010

Table thumbnail
Marchiori 2014

Table thumbnail
Martinez 2010

Table thumbnail
Rolls 2014a

Table thumbnail
Schmidt 2013

Table thumbnail
Skov 2013

Table thumbnail
Smith 2013a

Table thumbnail
van Ittersum 2013

Table thumbnail
van Kleef 2014

Table thumbnail
Wansink 2013

Table thumbnail
Wansink 2014

Table thumbnail
Williams 2014

DATA AND ANALYSES

This review has no analyses.

CONTRIBUTIONS OF AUTHORS

Draft the protocol - all authors

Develop a search strategy - GJH, IS

Search for trials - GJH, IS

Obtain copies of trials - GJH, IS

Select which studies to include - GJH, IS, DO

Extract data from studies - GJH, IS, HBL, YW, JPTH

Enter data into RevMan - GJH, IS

Carry out the analysis - YW, JPTH, IS, GJH

Interpret the analysis - all authors

Draft the final review - all authors

DECLARATIONS OF INTEREST

Gareth Hollands declares no financial or other conflicts of interest.

Ian Shemilt declares no financial or other conflicts of interest.

Theresa Marteau declares no financial or other conflicts of interest.

Susan Jebb is Chair of the Public Health Responsibility Deal Food Network, which develops voluntary agreements with industry to improve health, including reductions in portion size of foods high in fat, saturated fat, sugar and salt. She has also led research projects in which foods have been provided by a range of commercial companies as part of dietary intervention studies funded by public bodies. She was also a co-author of a published study (completed 2010) funded by the Coca-Cola Institute for Health & Wellness, which showed no effect on weight loss of a putative functional beverage.

Hannah Lewis declares no financial or other conflicts of interest.

Yinghui Wei declares no financial or other conflicts of interest.

Julian Higgins declares no financial or other conflicts of interest.

David Ogilvie declares no financial or other conflicts of interest.

SOURCES OF SUPPORT

Internal sources

  • Kings College London, UK.
    Database access
  • University of Cambridge, UK.
    Computer provision, database access
  • University of East Anglia, UK.
    Database access
  • University of Bristol, UK.
    Computer provision
  • Plymouth University, UK.
    Computer provision

External sources

  • Funded by UK Department of Health Policy Research Programme (107/0001- Policy Research Unit in Behaviour and Health), UK.
  • YW was supported by the UK Medical Research Council (MRC) grant to the MRC Clinical Trials Unit Hub for Trials Methodology Research [Grant number MSA7355QP21], UK.

DIFFERENCES BETWEEN PROTOCOL AND REVIEW

A difference between the protocol (Hollands 2014) and review is that the proposed search of the Cochrane Public Health Group Specialised Register was not, in practice, conducted. This omission is unlikely to have had any impact on the review. Study records on the Cochrane Public Health Group Specialised Register are submitted for inclusion in the Cochrane Central Register of Controlled Trials (CENTRAL) on a quarterly basis and we conducted searches of CENTRAL for this review up to 30 January 2015. Also, at the protocol stage, we intended to use the most commonly available measure of participants' socioeconomic status to construct the socioeconomic status context variable (see Data extraction and management). We were unable to do this in practice because no single proxy measure of participants' socioeconomic status, such as education or income, was commonly measured in and reported by included studies. Therefore we instead coded a binary study-level covariate based on authors' explicit descriptors of the study sample and/or setting (e.g. "Low income Hispanic or non-Hispanic African American children and their mothers", or "Faculty, graduate students, and staff members of the Department of Food Science and Nutritional Science of a large Midwestern university". Unless explicitly described as being of low socioeconomic status, we coded the context of included studies as high socioeconomic status.

References

References to studies included in this review

  • Ahn HJ, Han KA, Kwon HR, Min KW. The small rice bowl-based meal plan was effective at reducing dietary energy intake, body weight, and blood glucose levels in Korean women with type 2 diabetes mellitus. Korean Diabetes Journal. 2010;34(6):340–9. [PMC free article] [PubMed]
  • Argo JJ, White K. When do consumers eat more? The role of appearance self-esteem and food packaging cues (Study 1) Journal of Marketing. 2012;76(2):67–80.
  • Argo JJ, White K. When do consumers eat more? The role of appearance self-esteem and food packaging cues (Study 2) Journal of Marketing. 2012;76(2):67–80.
  • Argo JJ, White K. When do consumers eat more? The role of appearance self-esteem and food packaging cues (Study 4) Journal of Marketing. 2012;76(2):67–80.
  • Argo JJ, White K. When do consumers eat more? The role of appearance self-esteem and food packaging cues (Study 5) Journal of Marketing. 2012;76(2):67–80.
  • Burger KS, Fisher JO, Johnson SL. Mechanisms behind the portion size effect: visibility and bite size. Obesity. 2011;19(3):546–51. [PubMed]
  • Cavanagh K, Vartanian LR, Herman CP, Polivy J. The effect of portion size on food intake is robust to brief education and mindfulness exercises. Journal of Health Psychology. 2013;19(6):730–9. [PubMed]
  • Coelho do Vale R, Pieters R, Zeelenberg M. Flying under the radar: perverse package size effects on consumption self-regulation (study 2) Journal of Consumer Research. 2008;35(3):380–90.
  • Devitt AA, Mattes RD. Effects of food unit size and energy density on intake in humans. Appetite. 2004;42(2):213–20. [PubMed]
  • Diliberti N, Bordi PL, Conklin MT, Roe LS, Rolls BJ. Increased portion size leads to increased energy intake in a restaurant meal. Obesity Research. 2004;12(3):562–8. [PubMed]
  • DiSantis KI, Birch LL, Davey A, Serrano EL, Zhang J, Bruton Y, et al. Plate size and children's appetite: effects of larger dishware on self-served portions and intake. Pediatrics. 2013;131(5):e1451–8. [PubMed]
  • Ebbeling CB, Garcia-Lago E, Leidig MM, Seger-Shippee LG, Feldman HA, Ludwig DS. Altering portion sizes and eating rate to attenuate gorging during a fast food meal: effects on energy intake. Pediatrics. 2007;119(5):869–75. [PubMed]
  • Fisher JO, Rolls BJ, Birch LL. Children's bite size and intake of an entree are greater with large portions than with age-appropriate or self-selected portions. American Journal of Clinical Nutrition. 2003;77(5):1164–70. [PMC free article] [PubMed]
  • Fisher JO, Arreola A, Birch LL, Rolls BJ. Portion size effects on daily energy intake in low-income Hispanic and African American children and their mothers. American Journal of Clinical Nutrition. 2007;86(6):1709–16. [PubMed]
  • Fisher JO, Liu Y, Birch LL, Rolls BJ. Effects of portion size and energy density on young children's intake at a meal. American Journal of Clinical Nutrition. 2007;86(1):174–9. [PMC free article] [PubMed]
  • Fisher JO. Effects of age on children's intake of large and self-selected food portions. Obesity. 2007;15(2):403–12. [PubMed]
  • Fisher JO, Birch LL, Zhang J, Grusak MA, Hughes SO. External influences on children's self-served portions at meals. International Journal of Obesity. 2013;37:954–60. [PubMed]
  • Flood JE, Roe LS, Rolls BJ. The effect of increased beverage portion size on energy intake at a meal. Journal of the American Dietetic Association. 2006;106(12):1984–90. [PubMed]
  • Goldstein RB. Mindless Eating: How Differences in Portion Size Influence Popcorn Consumption for Males and Females. 2006. Bethlehem, PA: Lehigh University.
  • Hermans RCJ, Larsen JK, Herman CP, Engels RCME. Effects of portion size and social modeling on food intake of young women. Appetite. 2010;54(3):649.
  • Hermans RCJ, Larsen JK, Herman CP, Engels RCME. How much should I eat? Situational norms affect young women's food intake during meal time. British Journal of Nutrition. 2012;107(4):588–94. [PubMed]
  • Huss LR. Timing of Dessert but Not Portion Size Affects Young Children's Intake at Lunchtime (College of Health and Human Sciences Honors Program Undergraduate Theses, Paper 6) West Lafayette, IN: Purdue University; 2012.
  • Huss LR, Laurentz S, Fisher JO, McCabe GP, Kranz S. Timing of serving dessert but not portion size affects young children's intake at lunchtime. Appetite. 2013;68:158–63. [PubMed]
  • Jarvik ME, Popek P, Schneider NG, Baer-Weiss V, Gritz ER. Can cigarette size and nicotine content influence smoking and puffing rates? (Experiment 1) Psychopharmacology. 1978;58(3):303–6. [PubMed]
  • Jarvik ME, Popek P, Schneider NG, Baer-Weiss V, Gritz ER. Can cigarette size and nicotine content influence smoking and puffing rates? (Experiment 2) Psychopharmacology. 1978;58(3):303–6. [PubMed]
  • Jeffery RW, Rydell S, Dunn CL, Harnack LJ, Levine AS, Pentel PR, et al. Effects of portion size on chronic energy intake. International Journal of Behavioral Nutrition and Physical Activity. 2007;4:27. [PMC free article] [PubMed]
  • Kelly MT. Investigation of the contribution made by food portion size to food and energy intake (PhD Thesis) Coleraine: University of Ulster; 2008.
  • Kelly MT, Wallace JMW, Robson PJ, Rennie KL, Welch RW, Hannon-Fletcher MP, et al. Increased portion size leads to a sustained increase in energy intake over 4 d in normal-weight and overweight men and women. British Journal of Nutrition. 2009;102(3):470–7. [PubMed]
  • Koh J, Pliner P. The effects of degree of acquaintance, plate size, and sharing on food intake. Appetite. 2009;52(3):595–602. [PubMed]
  • Kral TVE, Roe LS, Rolls BJ. Combined effects of energy density and portion size on energy intake in women. American Journal of Clinical Nutrition. 2004;79(6):962–8. [PubMed]
  • Kral TVE, Kabay AC, Roe LS, Rolls BJ. Effects of doubling the portion size of fruit and vegetable side dishes on children's intake at a meal. Obesity. 2010;18(3):521–7. [PubMed]
  • Leahy KE, Birch LL, Fisher JO, Rolls BJ. Reductions in entree energy density increase children's vegetable intake and reduce energy intake. Obesity. 2008;16(7):1559–65. [PubMed]
  • Levitsky DA, Youn T. The more food young adults are served, the more they overeat. Journal of Nutrition. 2004;134(10):2546–9. [PubMed]
  • Looney SM, Raynor HA. Impact of portion size and energy density on snack intake in preschool-aged children. Journal of the American Dietetic Association. 2011;111(3):414–8. [PubMed]
  • Marchiori D, Klein O. Size Matters! The Joint Influence of the Size of Portion, Food Item and Container on Food Intake [Dissertation thesis] Brussels: Universite Libre de Bruxelles; 2012.
  • Marchiori D, Waroquier L, Klein O. Smaller food item sizes of snack foods influence reduced portions and caloric intake in young adults. Journal of the American Dietetic Association. 2011;111(5):727–31. [PubMed]
  • Marchiori D, Corneille O, Klein O. Container size influences snack food intake independently of portion size. Appetite. 2012;58(3):814–7. [PubMed]
  • Marchiori D, Corneille O, Klein O. Corrigendum to ‘Container size influences snack food intake independently of portion size’ [Appetite 2012a; 58(3): 814-817] Appetite. 2012;59(2):616. [PubMed]
  • Marchiori D, Klein O. Size Matters! The Joint Influence of the Size of Portion, Food Item and Container on Food Intake [Dissertation thesis] Brussels: Universite Libre de Bruxelles; 2012.
  • Marchiori D, Klein O. Size Matters! The Joint Influence of the Size of Portion, Food Item and Container on Food Intake [Dissertation thesis] Brussels: Universite Libre de Bruxelles; 2012.
  • Marchiori D, Waroquier L, Klein O. "Split them!" smaller item sizes of cookies lead to a decrease in energy intake in children. Journal of Nutrition Education and Behavior. 2012;44(3):251–5. [PubMed]
  • Mathias KC, Rolls BJ, Birch LL, Kral TVE, Hanna EL, Davey A, et al. Serving larger portions of fruits and vegetables together at dinner promotes intake of both foods among young children. Journal of the Academy of Nutrition and Dietetics. 2012;112(2):266–70. [PMC free article] [PubMed]
  • Mishra A, Mishra H, Masters TM. The influence of bite size on quantity of food consumed: A field study (Study 1) Journal of Consumer Research. 2012;38(5):791–5.
  • Mishra A, Mishra H, Masters TM. The influence of bite size on quantity of food consumed: A field study (Study 2) Journal of Consumer Research. 2012;38(5):791–5.
  • Raynor HA, Wing RR. Package unit size and amount of food: do both influence intake? Obesity. 2007;15(9):2311–9. [PubMed]
  • Raynor HA, Van Walleghen EL, Niemeier H, Butryn ML, Wing RR. Do food provisions packaged in single-servings reduce energy intake at breakfast during a brief behavioral weight-loss intervention? Journal of the American Dietetic Association. 2009;109(11):1922–5. [PMC free article] [PubMed]
  • Rolls BJ, Engell D, Birch LL. Serving portion size influences 5-year-old but not 3-year-old children's food intakes. Journal of the American Dietetic Association. 2000;100(2):232–4. [PubMed]
  • Rolls BJ, Morris EL, Roe LS. Portion size of food affects energy intake in normal-weight and overweight men and women. American Journal of Clinical Nutrition. 2002;76(6):1207–13. [PubMed]
  • Rolls BJ, Roe LS, Meengs JS, Wall DE. Increasing the portion size of a sandwich increases energy intake. Journal of the American Dietetic Association. 2004;104(3):367–72. [PubMed]
  • Rolls BJ, Roe LS, Kral TVE, Meengs JS, Wall DE. Increasing the portion size of a packaged snack increases energy intake in men and women. Appetite. 2004;42(1):63–9. [PubMed]
  • Rolls BJ, Roe LS, Meengs JS. Larger portion sizes lead to a sustained increase in energy intake over 2 days. Journal of the American Dietetic Association. 2006;106(4):543–9. [PubMed]
  • Rolls BJ, Roe LS, Meengs JS. Reductions in portion size and energy density of foods are additive and lead to sustained decreases in energy intake. American Journal of Clinical Nutrition. 2006;83(1):11–7. [PMC free article] [PubMed]
  • Rolls BJ, Roe LS, Meengs JS. The effect of large portion sizes on energy intake is sustained for 11 days. Obesity. 2007;15(6):1535–43. [PubMed]
  • Rolls BJ, Roe LS, Halverson HH, Meengs JS. Using a smaller plate did not reduce energy intake at meals (Study 1) Appetite. 2007;49(3):652–60. [PMC free article] [PubMed]
  • Rolls BJ, Roe LS, Halverson KH, Meengs JS. Using a smaller plate did not reduce energy intake at meals (Study 2) Appetite. 2007;49(3):652–60. [PMC free article] [PubMed]
  • Rolls BJ, Roe LS, Halverson KH, Meengs JS. Using a smaller plate did not reduce energy intake at meals (Study 3) Appetite. 2007;49(3):652–60. [PMC free article] [PubMed]
  • Rolls BJ, Roe LS, Meengs JS. Portion size can be used strategically to increase vegetable consumption in adults (Experiment 1) American Journal of Clinical Nutrition. 2010;91(4):913–22. [PubMed]
  • Rolls BJ, Roe LS, Meengs JS. Portion size can be used strategically to increase vegetable consumption in adults (Experiment 2) American Journal of Clinical Nutrition. 2010;91(4):913–22. [PubMed]
  • Russell MAH, Sutton SR, Feyerabend C, Saloojee Y. Smokers' response to shortened cigarettes: dose reduction without dilution of tobacco smoke. Clinical Pharmacology and Therapeutics. 1980;27(2):210–8. [PubMed]
  • Scott ML. The effect of reduced food and package sizes on the consumption behavior of restrained and unrestrained eaters. Dissertation Abstracts International Section A: Humanities and Social Sciences. 2008;69(2-A):681.
  • Scott ML, Nowlis SM, Mandel N, Morales AC. The effects of reduced food size and package size on the consumption behavior of restrained and unrestrained eaters (Study 2) Journal of Consumer Research. 2008;35(3):391–405.
  • Scott ML. The effect of reduced food and package sizes on the consumption behavior of restrained and unrestrained eaters. Dissertation Abstracts International Section A: Humanities and Social Sciences. 2008b;69(2-A):681.
  • Scott ML, Nowlis SM, Mandel N, Morales AC. The effects of reduced food size and package size on the consumption behavior of restrained and unrestrained eaters (Study 3) Journal of Consumer Research. 2008;35(3):391–405.
  • Scott ML. The effect of reduced food and package sizes on the consumption behavior of restrained and unrestrained eaters. Dissertation Abstracts International Section A: Humanities and Social Sciences. 2008b;69(2-A):681.
  • Scott ML, Nowlis SM, Mandel N, Morales AC. The effects of reduced food size and package size on the consumption behavior of restrained and unrestrained eaters (Study 4) Journal of Consumer Research. 2008;35(3):391–405.
  • Shah M, Schroeder R, Winn W, Adams-Huet B. A pilot study to investigate the effect of plate size on meal energy intake in normal weight and overweight/obese women. Journal of Human Nutrition and Dietetics. 2011;24(6):612–5. [PubMed]
  • Spill MK, Birch LL, Roe LS, Rolls BJ. Eating vegetables first: the use of portion size to increase vegetable intake in preschool children. American Journal of Clinical Nutrition. 2010;91(5):1237–43. [PubMed]
  • Spill MK, Birch LL, Roe LS, Rolls BJ. Serving large portions of vegetable soup at the start of a meal affected children's energy and vegetable intake. Appetite. 2011;57(1):213–9. [PMC free article] [PubMed]
  • Stroebele N, Ogden LG, Hill JO. Do calorie-controlled portion sizes of snacks reduce energy intake? Appetite. 2009;52(3):793–6. [PMC free article] [PubMed]
  • van Kleef E, Shimizu M, Wansink B. Serving bowl selection biases the amount of food served. Journal of Nutrition Education and Behavior. 2012;44(1):66–70. [PubMed]
  • van Kleef E, Shimizu M, Wansink B. Just a bite: considerably smaller snack portions satisfy delayed hunger and craving. Food Quality and Preference. 2013;27(1):96–100.
  • Wansink B. Can package size accelerate usage volume? (Study 1) Journal of Marketing. 1996;60(3):1–14.
  • Wansink B. Can package size accelerate usage volume? (Study 2) Journal of Marketing. 1996;60(3):1–14.
  • Wansink B. Can package size accelerate usage volume? (Study 4) Journal of Marketing. 1996;60(3):1–14.
  • Wansink B, Park SB. At the movies: how external cues and perceived taste impact consumption volume. Food Quality and Preference. 2001;12(1):69–74.
  • Wansink B, Van Ittersum K. Bottoms up! The influence of elongation on pouring and consumption volume (Study 1) Journal of Consumer Research. 2003;30(3):455–63.
  • Wansink B, Van Ittersum K. Bottoms up! The influence of elongation on pouring and consumption volume (Study 2) Journal of Consumer Research. 2003;30(3):455–63.
  • Wansink B, Kim J. Bad popcorn in big buckets: portion size can influence intake as much as taste. Journal of Nutrition Education and Behavior. 2005;37(5):242–5. [PubMed]
  • Wansink B, Cardello A, North J. Fluid consumption and the potential role of canteen shape in minimizing dehydration. Military Medicine. 2005;170(10):871–3. [PubMed]
  • Wansink B, van Ittersum K, Painter JE. Ice cream illusions bowls, spoons, and self-served portion sizes. American Journal of Preventive Medicine. 2006;31(3):240–3. [PubMed]
  • Wansink B, Payne CR, Shimizu M. The 100-calorie semi-solution: sub-packaging most reduces intake among the heaviest. Obesity. 2011;19(5):1098–100. [PubMed]
  • Wansink B, Just DR, Smith LE, Wallace CE. Lunch line redesign: making school lunchrooms smarter. FASEB Journal. Conference: Experimental Biology. 2011;25:342–8.
  • Wansink B, van Ittersum K. Portion size me: plate-size induced consumption norms and win-win solutions for reducing food intake and waste. Journal of Experimental Psychology: Applied. 2013;19(4):320–32. [PubMed]

References to studies excluded from this review

  • Andrade AM, Greene GW, Melanson KJ. Eating slowly led to decreases in energy intake within meals in healthy women. Journal of the American Dietetic Association. 2008;108(7):1186–91. [PubMed]
  • Ashton H, Stepney R, Thompson JW. Smoking Behaviour - Physiological and Psychological Influences. Edinburgh: Churchill Livingstone; 1978. Smoking behaviour and nicotine intake in smokers presented with a "two-thirds" cigarette.
  • Attwood AS, Scott-Samuel NE, Stothart G, Munafo MR. Glass shape influences consumption rate for alcoholic beverages. PloS One. 2012;7(8):e43007. [PMC free article] [PubMed]
  • Balagura S, Harrell LE. Effect of size of food on food-consumption - some neurophysiological considerations. Journal of Comparative and Physiological Psychology. 1974;86(4):658–63. [PubMed]
  • Bell EA, Roe LS, Rolls BJ. Sensory-specific satiety is affected more by volume than by energy content of a liquid food. Physiology & Behavior. 2003;78:593–600. [PubMed]
  • Blum JEW, Davee AM, Devore RL, Beaudoin CM, Jenkins PL, Kaley LA, et al. Implementation of low-fat, low-sugar, and portion-controlled nutrition guidelines in competitive food venues of Maine public high schools: research article. Journal of School Health. 2007;77(10):687–93. [PubMed]
  • Bohnert AM, Randall ET, Tharp S, Germann J. The development and evaluation of a portion plate for youth: a pilot study. Journal of Nutrition Education and Behavior. 2011;1(4):268–73. [PubMed]
  • Boyer LE, Laurentz S, McCabe GP, Kranz S. Shape of snack foods does not predict snack intake in a sample of preschoolers: a cross-over study. International Journal of Behavioral Nutrition and Physical Activity. 2012;9:94. [PMC free article] [PubMed]
  • Brown D. Travel sizes bring portions to the forefront. Journal of the American Dietetic Association. 2006;106(6):793. [PubMed]
  • Caljouw SR, Van W. Is the glass half full or half empty? How to reverse the effect of glass elongation on the volume poured. PloS One. 2014;9:e109374. [PMC free article] [PubMed]
  • Campbell MK, Polhamus B, McClelland JW, Bennett K, Kalsbeek W, Coole D, et al. Assessing fruit and vegetable consumption in a 5 A Day study targeting rural blacks: the issue of portion size. Journal of the American Dietetic Association. 1996;96(10):1040–2. [PubMed]
  • Chait LD, Griffiths RR. Smoking behavior and tobacco smoke intake: response of smokers to shortened cigarettes. Clinical Pharmacology and Therapeutics. 1982;32(1):90–7. [PubMed]
  • Chait LD, Griffiths RR. Differential control of puff duration and interpuff interval in cigarette smokers. Pharmacology Biochemistry and Behavior. 1982;17(1):155–8. [PubMed]
  • Chandler C, Hietpas F, Clark H, Smead K. Effect of straw diameter on bolus volume and muscle activity. Dysphagia. 2009;24(4):471.
  • Chandon P, Ordabayeva N. Supersize in one dimension, downsize in three dimensions: effects of spatial dimensionality on size perceptions and preferences. Journal of Marketing Research. 2009;46(6):739–53.
  • Chang UJ, Suh HJ, Yang SO, Hong YH, Kim YS, Kim JM, et al. Distinct foods with smaller unit would be an effective approach to achieve sustainable weight loss. Eating Behaviors. 2012;13(1):74–7. [PubMed]
  • Cleghorn CL, Evans CE, Kitchen MS, Cade JE. Details and acceptability of a nutrition intervention programme designed to improve the contents of children's packed lunches. Public Health Nutrition. 2010;13(8):1254–61. [PubMed]
  • Cluskey M, Dunton N. Serving meals of reduced portion size did not improve appetite among elderly in a personal-care section of a long-term-care community. Journal of the American Dietetic Association. 1999;99(6):733–5. [PubMed]
  • Collings AS. An experiment analysis of the impact of advertising and food packaging on women's eating behavior. 2008;138 Master's Theses and Doctoral Dissertations.
  • Cullen KW, Thompson DI. Texas school food policy changes related to middle school a la carte/snack bar foods: potential savings in kilocalories. Journal of the American Dietetic Association. 2005;105(12):1952–4. [PubMed]
  • Cunningham E. What impact does plate size have on portion control? Journal of the American Dietetic Association. 2011;111(9):1438. [PubMed]
  • Divert C, Laghmaoui R, Crema C, Issanchou S, Van W, Virginie SRC. Improving meal context in nursing homes. Impact of four strategies on food intake and meal pleasure. Appetite. 2015;84:139–47. [PubMed]
  • Edelman B, Engell D, Bronstein P, Hirsch E. Environmental effects on the intake of overweight and normal-weight men. Appetite. 1986;7(1):71–83. [PubMed]
  • Ello-Martin JA, Ledikwe JH, Rolls BJ. The influence of food portion size and energy density on energy intake: implications for weight management. American Journal of Clinical Nutrition. 2005;82:236S–41S. [PubMed]
  • Etten ML, Higgins ST, Bickel WK. Effects of response cost and unit dose on alcohol self-administration in moderate drinkers. Behavioural Pharmacology. 2005;7:754–8. [PubMed]
  • Farleigh CA, Shepherd R, Wharf SG. The effect of manipulation of salt pot hole size on table salt use. Food Quality and Preference. 1990;2(1):13–20.
  • Faucher MA, Mobley J. A community intervention on portion control aimed at weight loss in low-income Mexican American women. Journal of Midwifery & Women's Health. 2010;55:60–4. [PubMed]
  • Freedman MR, Brochado C. Reducing portion size reduces food intake and plate waste. Obesity. 2010;18(9):1864–6. [PubMed]
  • French SA, Mitchell NR, Wolfson J, Harnack LJ, Jeffery RW, Gerlach AF, et al. Portion size effects on weight gain in a free living setting. Obesity. 2014;22:1400–5. [PMC free article] [PubMed]
  • Garber LL, Hyatt EM, Boya UO. Does visual package clutter obscure the communicability of food package shape? Journal of Food Products Marketing. 2008;14(4):21–32.
  • Geaney F, Scotto Di M, Kelly C, Fitzgerald AP, Harrington JM, Kirby A, et al. The food choice at work study: effectiveness of complex workplace dietary interventions on dietary behaviours and diet-related disease risk - study protocol for a clustered controlled trial. Trials. 2013;14:370. [PMC free article] [PubMed]
  • Geier AB, Rozin P, Doros G. Unit bias: a new heuristic that helps explain the effect of portion size on food intake. Psychological Science. 2006;17(6):521–5. [PubMed]
  • Gillis B, Mobley C, Stadler DD, Hartstein J, Virus A, Volpe SL, et al. Healthy Study Group. Rationale, design and methods of the HEALTHY study nutrition intervention component. International Journal of Obesity. 2009;33(Suppl 4):S29–S36. [PMC free article] [PubMed]
  • Goldfarb TL, Jarvik ME. Accommodation to restricted tobacco smoke intake in cigarette smokers. International Journal of the Addictions. 1972;7(3):559–65. [PubMed]
  • Gosnell BA, Mitchell JE, Lancaster KL, Burgard MA, Wonderlich SA, Crosby RD. Food presentation and energy intake in a feeding laboratory study of subjects with binge eating disorder. International Journal of Eating Disorders. 2001;30(4):441–6. [PubMed]
  • Greenfield H, Maples J, Wills RBH. Salting of food - a function of hole size and location of shakers. Nature. 1983;301(5898):331–2. [PubMed]
  • Greenfield H, Smith AM, Wills RB. Influence of multi-holed shakers on salting on food. Human Nutrition. 1984;38(3):199–201. [PubMed]
  • Gritz ER, Baer-Weiss V, Jarvik ME. Titration of nicotine intake with full-length and half-length cigarettes. Clinical Pharmacology & Therapeutics. 1976;20(5):552–6. [PubMed]
  • Hackbart SJ, LeCheminant JD, Smith JD, Lox CL. The influence of an environmental cue and exercise on food consumption in college students. International Journal of Exercise Science. 2009;2(2):3. [PMC free article] [PubMed]
  • Haisfield L, Fisher JO, Savage JS, Marini M, Birch LL. Influence of family-style meals on young children's self-selected portions and intake. Obesity. 2011;19:S66–7.
  • Hartstein J, Cullen KW, Reynolds KD, Harrell J, Resnicow K, Kennel P. Impact of portion-size control for school a la carte items: changes in kilocalories and macronutrients purchased by middle school students. Journal of the American Dietetic Association. 2008;108(1):140–4. [PMC free article] [PubMed]
  • Head MK, Weeks RJ. Conventional vs. formulated foods in school lunches. I. Comparison of students' food and nutrient intakes. Journal of the American Dietetic Association. 1977;71(2):116–23. [PubMed]
  • Hirst K, Baranowski T, DeBar L, Foster GD, Kaufman F, Kennel P, et al. Healthy Study Group HEALTHY study rationale, design and methods: moderating risk of type 2 diabetes in multi-ethnic middle school students. International Journal of Obesity. 2009;33(Suppl 4):S4–S20. [PMC free article] [PubMed]
  • Mobley CC, Stadler DD, Staten MA, El Ghormli L, Gillis B, Hartstein J, et al. Healthy Study Group Effect of nutrition changes on foods selected by students in a middle school-based diabetes prevention intervention program: the HEALTHY experience. Journal of School Health. 2012;82(2):82–90. [PMC free article] [PubMed]
  • Higgins IT. Length of cigarette ends and inhaling. British Journal of Industrial Medicine. 1964;21:321–3. [PMC free article] [PubMed]
  • Huyghe E, Van Kerckhove A. Can fat taxes and package size restrictions stimulate healthy food choices? International Journal of Research in Marketing. 2013;30:421–3.
  • Jaeger SR, Harker R, Triggs CM, Gunson A, Campbell RL, Jackman R, et al. Determining consumer purchase intentions: the importance of dry matter, size, and price of kiwifruit. Journal of Food Science. 2011;76(3):177–84. [PubMed]
  • Just DR, Wansink B. One man's tall is another man's small: how the framing of portion size influences food choice. Health Economics. 2014;23:776–91. [PubMed]
  • Just DR, Wansink B. One man's tall is another man's small: how the framing of portion size influences food choice. Health Economics. 2014;23:776–91. [PubMed]
  • Kallbekken S, Saelen H. 'Nudging' hotel guests to reduce food waste as a win-win environmental measure. Economics Letters. 2013;119:325–7.
  • Kesman RL, Ebbert JO, Harris KI, Schroeder DR. Portion control for the treatment of obesity in the primary care setting. BMC Research Notes. 2011;9(4):346. [PMC free article] [PubMed]
  • Kildegaard H, Olsen A, Gabrielsen G, Moller P, Thybo AK. A method to measure the effect of food appearance factors on children's visual preferences. Food Quality and Preference. 2011;22(8):763–71.
  • Kozlowski LT, Heatherton TF, Ferrence RG. Pack size, reported cigarette-smoking rates, and the heaviness of smoking. Canadian Journal of Public Health-Revue Canadienne De Sante Publique. 1989;80(4):266–70. [PubMed]
  • Kral TVE, Rolls BJ. Energy density and portion size: their independent and combined effects on energy intake. Physiology & Behavior. 2004;82(1):131–8. [PubMed]
  • Lawless HT, Bender S, Oman C, Pelletier C. Gender, age, vessel size, cup vs. straw sipping, and sequence effects on sip volume. Dysphagia. 2003;18(3):196–202. [PubMed]
  • Leidy HJ, Apolzan JW, Mattes RD, Campbell WW. Food form and portion size affect postprandial appetite sensations and hormonal responses in healthy, nonobese, older adults. Obesity. 2010;18(2):293–9. [PMC free article] [PubMed]
  • Levitsky DA, Pacanowski C. Losing weight without dieting. Use of commercial foods as meal replacements for lunch produces an extended energy deficit. Appetite. 2011;57(2):311–7. [PubMed]
  • Lewis HB, Solis-Trapala I, Jebb SA. The effect of covertly reducing portion size of a single meal on day-long energy intake in overweight and obese adults. Obesity Facts. 2013;6:139–40.
  • Libotte E, Siegrist M, Bucher T. The influence of plate size on meal composition. Literature review and experiment. Appetite. 2014;82:91–6. [PubMed]
  • Liem DG, Zandstra LH. Children's liking and wanting of snack products: influence of shape and flavour. International Journal of Behavioral Nutrition and Physical Activity. 2009;6:38. [PMC free article] [PubMed]
  • Lieux EM, Manning CK. Evening meals selected by college students: impact of the foodservice system. Journal of the American Dietetic Association. 1992;92(5):560–6. [PubMed]
  • Lin H-M, Lo H-Y, Liao Y-S. More than just a utensil: the influence of drinking straw size on perceived consumption. Marketing Letters. 2013;24:381–6.
  • Meguid MM, Laviano A, Rossi-Fanelli F. Food intake equals meal size times mean number. Appetite. 1998;31(3):404. [PubMed]
  • Mendoza JA, Watson K, Cullen KW. Change in dietary energy density after implementation of the Texas public school nutrition policy. Journal of the American Dietetic Association. 2010;110(3):434–40. [PMC free article] [PubMed]
  • Olsen A, Ritz C, Kramer L, Moller P. Serving styles of raw snack vegetables. What do children want? Appetite. 2012;59:556–62. [PubMed]
  • Pornpitakpan C. How package sizes, fill amounts, and unit costs influence product usage amounts. Journal of Global Marketing. 2010;23(4):275–87.
  • Raghubir P, Krishna A. Vital dimensions in volume perception: can the eye fool the stomach? Journal of Marketing Research. 1999;36(3):313–26.
  • Rolls BJ, Rowe EA, Rolls ET. How sensory properties of foods affect human feeding behavior. Physiology & Behavior. 1982;29(3):409–17. [PubMed]
  • Rolls BJ. Experimental analyses of the effects of variety in a meal on human feeding. American Journal of Clinical Nutrition. 1985;42:932–9. [PubMed]
  • Rolls BJ, Kim S, Fedoroff IC. Effects of drinks sweetened with sucrose or aspartame on hunger, thirst and food intake in men. Physiology & Behavior. 1990;48(1):19–26. [PubMed]
  • Rolls BJ. High satiety: avoiding obesity in a super-sized world. Obesity Research and Clinical Practice. 2012;6:1.
  • Savage JS, Fisher JO, Marini M, Birch LL. Serving smaller age-appropriate entree portions to children aged 3-5 y increases fruit and vegetable intake and reduces energy density and energy intake at lunch. American Journal of Clinical Nutrition. 2012;95(2):335–41. [PubMed]
  • Saylor JH. Volume of a swallow: role of orifice size and viscosity. Veterinary & Human Toxicology. 1987;29(1):79–83. [PubMed]
  • Scheibehenne B, Todd PM, Wansink B. Dining in the dark. The importance of visual cues for food consumption and satiety. Appetite. 2010;55(3):710–3. [PubMed]
  • Scisco JL, Blades C, Zielinski MJ, Muth ER. Dividing a fixed portion into more pieces leads to larger portion size estimates of JELL-O (R) squares. Perception. 2012;41:988–90. [PMC free article] [PubMed]
  • Scisco JL, Blades C, Zielinski MJ, Muth ER. Dividing a fixed portion into more pieces leads to larger portion size estimates of JELL-O (R) squares. Perception. 2012;41:988–90. [PMC free article] [PubMed]
  • Sharafi M. Children's Behavioral Responses to Portion Size [Master of Science Thesis] Pennsylvania State University; 2010.
  • Spanos S, Kenda AS, Vartanian LR. Can serving-size labels reduce the portion-size effect? A pilot study. Eating Behaviors. 2015;16:40–2. [PubMed]
  • Spiegel TA, Kaplan JM, Tomassini A, Stellar E. Bite size, ingestion rate, and meal size in lean and obese women. Appetite. 1993;21(2):131–45. [PubMed]
  • Spill MK, Birch LL, Roe LS, Rolls BJ. Hiding vegetables to reduce energy density: an effective strategy to increase children's vegetable intake and reduce energy intake. American Journal of Clinical Nutrition. 2011;94(3):735–41. [PubMed]
  • Stepney R. Behavioural regulation of nicotine intake in cigarette smokers presented with a ‘shortened’ cigarette. British Journal of Clinical Pharmacology (Proceedings of the British Pharmacological Society) 1977;4(5):653P. [PubMed]
  • Tapsell LC, Batterham MJ, Thorne RL, O'Shea JE, Grafenauer SJ, Probst YC. Weight loss effects from vegetable intake: a 12-month randomised controlled trial. European Journal of Clinical Nutrition. 2014;68:778–85. [PMC free article] [PubMed]
  • Ueland O, Cardello AV, Merrill EP, Lesher LL. Effect of portion size information on food intake. Journal of the American Dietetic Association. 2009;109(1):124–7. [PubMed]
  • Van Ittersum K, Wansink B. Plate size and color suggestibility: the Delboeuf illusion's bias on serving and eating behavior. Journal of Consumer Research. 2012;39(2):215–28.
  • Vermeer WM, Steenhuis IHM, Leeuwis FH, Heymans MW, Seidell JC. Small portion sizes in worksite cafeterias: do they help consumers to reduce their food intake? International Journal of Obesity. 2011;35(9):1200–7. [PMC free article] [PubMed]
  • Vermeer WM, Leeuwis FH, Koprulu S, Zouitni O, Seidell JC, Steenhuis IHM. The process evaluation of two interventions aimed at portion size in worksite cafeterias. Journal of Human Nutrition and Dietetics. 2012;25(2):180–8. [PubMed]
  • Walker D, Smarandescu L, Wansink B. Half full or empty: cues that lead wine drinkers to unintentionally overpour. Substance Use & Misuse. 2014;49:295–302. [PubMed]
  • Wansink B, van Ittersum K. Shape of glass and amount of alcohol poured: comparative study of effect of practice and concentration. BMJ. 2005;331(7531):1512–4. [PMC free article] [PubMed]
  • Wansink B, Painter JE, North J. Bottomless bowls: why visual cues of portion size may influence intake. Obesity Research. 2005;13(1):93–100. [PubMed]
  • Wansink B, Cheney MM. Super bowls: serving bowl size and food consumption. JAMA. 2005;293(14):1727–8. [PubMed]
  • Wansink B, van Ittersum K. Portion size me: downsizing our consumption norms. Journal of the American Dietetic Association. 2007;107(7):1103–6. [PubMed]
  • Weijzen PLG, Liem DG, Zandstra EH, de Graaf C. Sensory specific satiety and intake: the difference between nibble- and bar-size snacks. Appetite. 2008;50:435–42. [PubMed]
  • Weijzen PL, Smeets PA, Graaf C. Sip size of orangeade: effects on intake and sensory-specific satiation. British Journal of Nutrition. 2009;7:1091–7. [PubMed]
  • White AM, Kraus CL, McCracken LA, Swartzwelder H. Do college students drink more than they think? Use of a free-pour paradigm to determine how college students define standard drinks. Alcoholism: Clinical and Experimental Research. 2003;27(11):1750–6. [PubMed]
  • Williams RA, Roe LS, Rolls BJ. Comparison of three methods to reduce energy density: effects on daily energy intake. Appetite. 2013;66:75–83. [PMC free article] [PubMed]
  • Wilson BM, Stolarz-Fantino S, Fantino E. Regulating the way to obesity: unintended consequences of limiting sugary drink sizes. PloS One. 2013;8(4):e61081. [PMC free article] [PubMed]
  • Woodson PP, Griffiths RR. Control of cigarette smoking topography: smoke filtration and draw resistance. Behavioural Pharmacology. 1992;3(2):99–111. [PubMed]
  • Yamauchi K, Katayama T, Yamauchi T, Kotani K, Tsuzaki K, Takahashi K, et al. Efficacy of a 3-month lifestyle intervention program using a Japanese-style healthy plate on body weight in overweight and obese diabetic Japanese subjects: a randomized controlled trial. Nutrition Journal. 2014;13:108. [PMC free article] [PubMed]
  • Yang S, Raghubir P. Can bottles speak volumes? The effect of package shape on how much to buy. Journal of Retailing. 2005;81(4):269–81.
  • Yee RW. An analysis of beer consumption as a function of glass size and pitcher presence. Dissertation Abstracts International. 1979;39:4081.
  • Yeomans MR, Gould NJ, Leitch M, Mobini S. Effects of energy density and portion size on development of acquired flavour liking and learned satiety. Appetite. 2009;52(2):469–78. [PubMed]
  • Yip W, Wiessing KR, Budgett S, Poppitt SD. Using a smaller dining plate does not suppress food intake from a buffet lunch meal in overweight, unrestrained women. Appetite. 2013;69:102–7. [PubMed]
  • Zijlstra N, de Wijk RA, Mars M, Stafleu A, de Graaf C. Effect of bite size and oral processing time of a semisolid food on satiation. American Journal of Clinical Nutrition. 2009;90(2):269–75. [PubMed]

References to studies awaiting assessment

  • Bajaj D. Effect of number of food pieces on food selection and consumption in animals and humans. 2014;74 Dissertation Abstracts International: Section B: The Sciences and Engineering.
  • Haire C, Raynor HA. Weight status moderates the relationship between package size and food Intake. Journal of the Academy of Nutrition and Dietetics. 2014;114:1251–6. [PubMed]
  • Kral TVE, Remiker AM, Strutz EM, Moore RH. Role of child weight status and the relative reinforcing value of food in children's response to portion size increases. Obesity. 2014;22:1716–22. [PMC free article] [PubMed]
  • Loney T, Lawton K, Allen D, Carter JM. Size matters! Effect of a school canteen portion size intervention on weight loss in obese Emirati adolescents. Obesity Reviews. 2010;11(1) 239 (T3: PO 54)
  • Marchiori D, Papies EK. A brief mindfulness intervention reduces unhealthy eating when hungry, but not the portion size effect. Appetite. 2014;75:40–5. [PubMed]
  • Martinez AG, Lopez-Espinola A, Beltran C, Franco K, Diaz FJ, Cardenas A, et al. Portion size affects how much students consume in an eating occasion. Appetite. 2010;54(3):661.
  • Rolls BJ, Meengs JS, Roe LS. Variations in cereal volume affect the amount selected and eaten for breakfast. Journal of the Academy of Nutrition and Dietetics. 2014;114:1411–6. [PMC free article] [PubMed]
  • Rolls BJ, Roe LS, Meengs JS. Reshaping breakfast: the smaller the cereal flake, the greater the intake. FASEB Journal. 2013;27:273.3.
  • Schmidt K, Rohden S, Guldborg H, Maaloe J, Perez-Cueto FJA, Egberg M. Smaller plates, less food waste-a choice architectural experiment in a self-service eating setting. Annals of Nutrition and Metabolism. 2013;63:1754.
  • Skov LR, Schmidt K, Guldborg H, Lund S, Egberg M, Perez-Cueto FJA. The smaller the piece the healthier consumption-a choice architectural experiment in behavioural nutrition. Annals of Nutrition and Metabolism. 2013;63:1754.
  • Smith L, Conroy K, Wen H, Rui L, Humphries D. Portion size variably affects food intake of 6-year-old and 4-year-old children in Kunming, China. Appetite. 2013;69:31–8. [PMC free article] [PubMed]
  • Van Ittersum K, Wansink B. Extraverted children are more biased by bowl sizes than introverts. PloS One. 2013;8:e78224. [PMC free article] [PubMed]
  • van Kleef E, Kavvouris C, van Trijp HCM. The unit size effect of indulgent food: how eating smaller sized items signals impulsivity and makes consumers eat less. Psychology & Health. 2014;29:1081–103. [PubMed]
  • Wansink B, Just DR, Hanks AS, Smith LE. Pre-sliced fruit in school cafeterias: children's selection and intake. American Journal of Preventive Medicine. 2013;44:477–80. [PubMed]
  • Wansink B, van Ittersum K, Payne CR. Larger bowl size increases the amount of cereal children request, consume, and waste. Journal of Pediatrics. 2014;164:323–6. [PubMed]
  • Williams RA, Roe LS, Rolls BJ. Assessment of satiety depends on the energy density and portion size of the test meal. Obesity. 2014;22:318–24. [PMC free article] [PubMed]

Additional references

  • Anderson LM, Petticrew M, Rehfuess E, Armstrong R, Ueffing E, Baker P, et al. Using logic models to capture complexity in systematic reviews. Research Synthesis Methods. 2011;2:33–42. [PubMed]
  • Anderson LM, Oliver SR, Michie S, Rehfuess E, Noyes J, Shemilt I. Investigating complexity in systematic reviews of interventions by using a spectrum of methods. Journal of Clinical Epidemiology. 2013;66:1223–9. [PubMed]
  • Beasley JM, Ange BA, Anderson CA, Miller ER, Erlinger TP, Holbrook JT, et al. Associations between macronutrient intake and self-reported appetite and fasting levels of appetite hormones: results from the Optimal Macronutrient Intake Trial to Prevent Heart Disease. American Journal of Epidemiology. 2009;169(7):893–900. [PMC free article] [PubMed]
  • Bell EA, Castellanos VH, Pelkman CL, Thorwart ML, Rolls BJ. Energy density of foods affects energy intake in normal-weight women. American Journal of Clinical Nutrition. 1998;67:412–20. [PubMed]
  • Birch LL, Johnson SL, Andresen G, Peters JC, Schulte MC. The variability of young children's energy intake. New England Journal of Medicine. 1991;324(4):232–5. [PubMed]
  • Mela D, Salah D, Schuring E, van der Knaap H, Westerterp M. Appetite control: methodological aspects of the evaluation of foods. Obesity Reviews. 2010;11(3):251–70. [PMC free article] [PubMed]
  • Brennan IM, Luscombe-Marsh ND, Seimon RV, Otto B, Horowitz M, Wishart JM, et al. Effects of fat, protein, and carbohydrate and protein load on appetite, plasma cholecystokinin, peptide YY, and ghrelin, and energy intake in lean and obese men. American Journal of Physiology - Gastrointestinal and Liver Physiology. 2012;303(1):G129–40. [PubMed]
  • Brozek J, Oxman A, Schünemann H. GRADEpro. 2008. 3.2 for Windows. Brozek J, Oxman A, Schünemann H.
  • Bryden A, Petticrew M, Mays N, Eastmure E, Knai C. Voluntary agreements between government and business—a scoping review of the literature with specific reference to the Public Health Responsibility Deal. Health Policy. 2013;110(2-3):186–97. [PubMed]
  • Burton P, Smit HJ, Lightowler HJ. The influence of restrained and external eating patterns on overeating. Appetite. 2007;49(1):191–7. [PubMed]
  • Cohen J. Statistical Power Analysis in the Behavioral Sciences. 2nd Edition. Hillsdale (NJ): Lawrence Erlbaum Associates, Inc.; 1988.
  • Cooper Z, Fairburn CG. Refining the definition of binge eating disorder and non-purging bulimia nervosa. International Journal of Eating Disorders. 2003;34(S1):S89–S95. [PubMed]
  • Das P, Horton R. Rethinking our approach to physical activity. Lancet. 2012;380(9838):189–90. [PubMed]
  • Deeks JJ, Higgins JPT, Altman Dg Chapter 9: Analysing data and undertaking meta-analyses. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from http://www.cochrane-handbook.org.
  • Food, Rural Affairs, Department for Environment . Family Food 2013. London: Department for Environment, Food and Rural Affairs; 2014.
  • Department of Health. Statement of the Calorie Reduction Expert Group (Policy Paper) 2011. https://www.gov.uk/government/publications/statement-of-the-calorie-reduction-expert-group (accessed 7 January 2015)
  • Diepeveen S, Ling T, Suhrcke M, Roland M, Marteau TM. Public acceptability of government intervention to change health-related behaviours: a systematic review and narrative synthesis. BMC Public Health. 2013;13:756. [PMC free article] [PubMed]
  • Diliberti N, Bordi PL, Conklin MT, Roe LS, Rolls BJ. Increased portion size leads to increased energy intake in a restaurant meal. Obesity Research. 2004;12(3):562–8. [PubMed]
  • Doucet E, Laviolette M, Imbeault P, Strychar I, Rabasa-Lhoret R, Prud'homme D. Total peptide YY is a correlate of postprandial energy expenditure but not of appetite or energy intake in healthy women. Metabolism: Clinical and Experimental. 2008;57(10):1458–64. [PubMed]
  • Drewnowski A, Rehm CD, Constant F. Water and beverage consumption among adults in the United States: cross-sectional study using data from NHANES 2005–2010. BMC Public Health. 2013;13:1068. [PMC free article] [PubMed]
  • Egger M, Smith GD, Schneider M, Minde C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–34. [PMC free article] [PubMed]
  • Ello-Martin JA, Ledikwe JH, Rolls BJ. The influence of food portion size and energy density on energy intake: implications for weight management. American Journal of Clinical Nutrition. 2005;82:236S–41S. [PubMed]
  • European Union. Tobacco Products Directive (2014/40/EU) Available from: http://ec.europa.eu/health/tobacco/products/index_en.htm.
  • Fairburn CG, Cooper Z. The eating disorder examination. In: Fairburn CG, Wilson GT, editors. Binge Eating: Nature, Assessment, and Treatment. 12th Edition. New York, NY: Guilford; 1993. pp. 317–32.
  • Fone DL, Farewell DM, White J, Lyons RA, Dunstan FD. Socioeconomic patterning of excess alcohol consumption and binge drinking: a cross-sectional study of multilevel associations with neighbourhood deprivation. BMJ Open. 2014;3:e002337. [PMC free article] [PubMed]
  • Food Standards Agency. Food Portion Sizes. 3rd Edition. Food Standards Agency; 2002.
  • Freudenberg N. Lethal But Legal: Corporations, Consumption, and Protecting Public Health. Oxford: Oxford University Press; 2014.
  • Fyfe CL, Stewart J, Murison SD, Jackson DM, Rance K, Speakman JR, et al. Evaluating energy intake measurement in free-living subjects: when to record and for how long? Public Health Nutrition. 2010;13(2):172–80. [PubMed]
  • Gabbatt A. New York City soda ban struck down by judge in eleventh-hour ruling. The Guardian 2013 Mar 11.
  • Gardner MP, Wansink B, Kim J, Park S-B. Better moods for better eating? How mood influences food choice. Journal of Consumer Psychology. 2014;24(3):320–35.
  • Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The Eating Attitudes Test: psychometric features and clinical correlates. Psychological Medicine. 1982;12:871–8. [PubMed]
  • Geier AB, Rozin P, Doros G. A new heuristic that helps explain the effect of portion size on food intake. Psychological Science. 2006;17(6):521–5. [PubMed]
  • Giskes K, Avendano M, Brug J, Kunst AE. A systematic review of studies on socioeconomic inequalities in dietary intakes associated with weight gain and overweight/obesity conducted among European adults. Obesity Reviews. 2010;11(6):413–29. [PubMed]
  • Godfrey C, Rice N, Slack R, Sowden A, Worthy G. A Systematic Review of the Effects of Price on the Smoking Behaviour of Young People. York: Public Health Research Consortium, University of York; 2009.
  • Gormally J, Black S, Daston S, Rardin D. The assessment of binge eating severity among obese persons. Addictive Behaviors. 1982;7(1):47–55. [PubMed]
  • Grant SP, Mayo-Wilson E, Melendez-Torres GJ, Montgomery P. Reporting quality of social and psychological intervention trials: a systematic review of reporting guidelines and trial publications. PloS One. 2013;8:e65442. [PMC free article] [PubMed]
  • Grossniklaus DA, Dunbar SB, Tohill BC, Gary R, Higgins MK, Frediani J. Psychological factors are important correlates of dietary pattern in overweight adults. Journal of Cardiovascular Nursing. 2010;25(6):450–60. [PMC free article] [PubMed]
  • Grynbaum M. Soda makers begin their push against New York ban. New York Times. 2012 http://www.nytimes.com/2012/07/02/nyregion/in-fight-against-nyc-soda-ban-industry-focuses-on-personal-choice.html.
  • Guyatt GH, Oxman AD, Sultan S, Glasziou P, Akl EA, Alonso-Coello P, et al. and The GRADE Working Group GRADE guidelines: 9. Rating up the quality of evidence. Journal of Clinical Epidemiology. 2011;64:1311–6. [PubMed]
  • Han E, Powell LM. Consumption patterns of sugar sweetened beverages in the United States. Journal of the Academy of Nutrition and Dietetics. 2013;113(1):43–53. [PMC free article] [PubMed]
  • Harbord RM, Higgins JPT. Meta–regression in Stata. The Stata Journal. 2008;8(4):493–519.
  • Harnack LJ, Jeffery RW, Boutelle KN. Temporal trends in energy intake in the United States: an ecologic perspective. American Journal of Clinical Nutrition. 2000;71:1478–84. [PubMed]
  • Harris RJ, Bradburn MJ, Deeks JJ, Harbord RM, Altman DG, Sterne JAC. metan: fixed- and random-effects meta-analysis. The Stata Journal. 2008;8(1):3–28.
  • Herman CP, Polivy J. Restrained eating. In: Stunkard A, editor. Obesity. Philadelphia: Saunders; 1980. pp. 208–25.
  • Herman CP. Obese externality. In: Darrity WA, editor. International Encyclopedia of the Social Sciences. 2nd Edition. Vol. 6. Farmington, MI: Thomas/Gale Publishers; 2008.
  • Herman CP, Polivy J, Pliner P, Vartanian LR. Mechanisms underlying the portion-size effect. Physiology & Behavior. 2015;144:129–36. [PubMed]
  • Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from http://www.cochrane-handbook.org.
  • Higgins JPT, Altman DG, Sterne Jac Chapter 8: Assessing risk of bias in included studies. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from http://www.cochrane-handbook.org.
  • Hoffmann T, Glasziou P, Boutron I, Milne R, Perera R, Moher D, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. 2014;348:g1687. [PubMed]
  • Hollands GJ, Shemilt I, Marteau TM, Jebb SA, Kelly MP, Nakamura R, et al. Altering micro-environments to change population health behaviour: towards an evidence base for choice architecture interventions. BMC Public Health. 2013;13:1218. [PMC free article] [PubMed]
  • Hollands GJ, Shemilt I, Marteau TM, Jebb SA, Kelly MP, Nakamura R, et al. Altering Choice Architecture to Change Population Health Behaviour: a Large-Scale Conceptual and Empirical Scoping Review of Interventions Within Micro-Environments. Cambridge: University of Cambridge; 2013.
  • Hollands GJ, Shemilt I, Marteau TM, Jebb SA, Lewis HB, Wei Y, et al. Portion, package or tableware size for changing selection and consumption of food, alcohol and tobacco. Cochrane Database of Systematic Reviews. 2014;(4) [DOI: 10.1002/14651858.CD011045] [PMC free article] [PubMed]
  • Holmes J, Meng Y, Meier PS, Brennan A, Angus C, Campbell-Burton A, et al. Effects of minimum unit pricing for alcohol on different income and socioeconomic groups: a modelling study. Lancet. 2014;383(9929):1655–64. [PMC free article] [PubMed]
  • Hsiao A, Wang YC. Reducing sugar-sweetened beverage consumption: evidence, policies, and economics. Current Obesity Reports. 2013;2:191–9.
  • Huang TTK, Cawley JH, Ashe M, Costa SA, Frerichs LM, Zwicker L, et al. Mobilisation of public support for policy actions to prevent obesity. Lancet. 2015;385(9985):2422–31. [PubMed]
  • Institute of Grocery Distribution. Portion Size: A Review of Existing Approaches. Watford: Institute of Grocery Distribution; 2008.
  • Jackson T. Prosperity Without Growth? The Transition to a Sustainable Economy. London: Sustainable Development Commission; 2014.
  • Kaner EFS, Dickinson HO, Beyer F, Pienaar E, Schlesinger C, Campbell F, et al. The effectiveness of brief alcohol interventions in primary care settings: a systematic review. Drug and Alcohol Review. 2009;28(3):301–23. [PubMed]
  • Kozlowski LT. Pack size, reported cigarette smoking rates, and public health. American Journal of Public Health. 1986;76(11):1337–8. [PubMed]
  • Kutner M, Greenberg E, Jin Y, Paulsen C. The health literacy of America's adults: results from the 2003 National Assessment of Adult Literacy. NCES 2006–483; U.S. Department of Education. 2006. Washington, DC: National Center for Education Statistics.
  • Lemmens SG, Martens EA, Born JM, Martens MJ, Westerterp-Plantenga MS. Staggered meal consumption facilitates appetite control without affecting postprandial energy intake. Journal of Nutrition. 2011;141(3):482–8. [PubMed]
  • Lewis HB, Ahern AL, Jebb SA. How much should I eat? A comparison of suggested portion sizes in the UK. Public Health Nutrition. 2012;15(11):2110–7. [PMC free article] [PubMed]
  • Lewis HB, Ahern AL, Solis-Trapala I, Walker CG, Reimann F, Gribble FM, et al. Effect of reducing portion size at a compulsory meal on later energy intake, gut hormones, and appetite in overweight adults. Obesity. 2015;23(7):1362–70. [DOI: 10.1002/oby.21105] [PubMed]
  • Lindroos AK, Lissner L, Mathiassen ME, Karlsson J, Sullivan M, Bengtsson C, et al. Dietary intake in relation to restrained eating, disinhibition, and hunger in obese and nonobese Swedish women. Obesity Research. 1997;5(3):175–82. [PubMed]
  • Lorenc T, Petticrew M, Welch V, Tugwell P. What types of interventions generate inequalities? Evidence from systematic reviews. Journal of Epidemiology and Community Health. 2013;67(2):190–3. [PubMed]
  • Marteau TM, Hollands GJ, Fletcher PC. Changing human behaviour to prevent disease: the importance of targeting automatic processes. Science. 2012;337(6101):1492–5. [PubMed]
  • Martins C, Truby H, Morgan LM. Short-term appetite control in response to a 6-week exercise programme in sedentary volunteers. British Journal of Nutrition. 2007;98(4):834–42. [PubMed]
  • Mayo-Wilson E, Grant S, Hopewell S, Macdonald G, Moher D, Montgomery P. Developing a reporting guideline for social and psychological intervention trials. Trials. 2013;14:242. [PMC free article] [PubMed]
  • Dobbs R, Sawers C, Thompson F, Manyika J, Woetzel J, Child P, et al. Overcoming obesity: an initial economic analysis. 2014. McKinsey Global Institute.
  • Moher D, Liberati A, Tetzlaff J, Altman DG, The Prisma Group Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Medicine. 2009;6(7):e1000097. [PMC free article] [PubMed]
  • Monteleone P, Bencivenga R, Longobardi N, Serritella C, Maj M. Differential responses of circulating ghrelin to high-fat or high-carbohydrate meal in healthy women. Journal of Clinical Endocrinology & Metabolism. 2003;88(11):5510–4. [PubMed]
  • Montgomery P, Grant S, Hopewell S, Macdonald G, Moher D, Michie S, et al. Protocol for CONSORT-SPI: an extension for social and psychological interventions. Implementation Science. 2013;8:99. [PMC free article] [PubMed]
  • National Centre for Social Research. National Diet and Nutrition Survey Years 1-4, 2008/09-2011/12. 2012. London: National Centre for Social Research.
  • Neal DT, Wood W, Quinn JM. Habits—a repeat performance. Current Directions in Psychological Science. 2006;15:198–202.
  • National Institute for Health and Care Excellence. Preventing the uptake of smoking by children and young people (NICE public health guidance 14, updated November 204) London: National Institute for Health and Care Excellence; 2008.
  • National Institute of Health and Care Excellence. Obesity: Guidance on the prevention of overweight and obesity in adults and children (NICE Guidelines [CG43] - Updated 2014) London: National Institute of Health and Care Excellence; 2014.
  • Office for National Statistics. Opinions and Lifestyle Survey, December 2012. 2012. London: Office for National Statistics.
  • Petrescu D, Hollands GH, Ng Y, Marteau TM. Public acceptability in the UK and USA of nudging to reduce obesity: the example of reducing sugary drinks consumption. PLoS One Under review. [PMC free article] [PubMed]
  • Piaget J. The Mechanisms of Perception. London: Rutledge & Kegan Paul; 1969.
  • Pierce JP, White VM, Emery SL. What public health strategies are needed to reduce smoking initiation? Tobacco Control. 2012;21:258–64. [PubMed]
  • Polivy J, Herman CP, Hackett R, Kuleshnyk I. The effects of self-attention and public attention on eating in restrained and unrestrained subjects. Journal of Perspectives in Social Psychology. 1986;50:1203–24. [PubMed]
  • Pratt IS, Croager EJ, Rosenberg M. The mathematical relationship between dishware size and portion size. Appetite. 2012;58(1):299–302. [PubMed]
  • Provencher V, Drapeau V, Tremblay A, Després JP, Lemieux S. Eating behaviors and indexes of body composition in men and women from the Québec family study. Obesity Research. 2003;11(6):783–92. [PubMed]
  • Rayner M, Scarborough P, Stockley L, Boxer A. Nutrient Profiles: Further Refinement and Testing of Model SSCg3d. 2005. London: Food Standards Agency. London.
  • Rehm J, Gmel G, Probst C, Shield KD. Lifetime-risk of alcohol-attributable mortality based on different levels of alcohol consumption in seven European countries. Implications for low-risk drinking guidelines. Toronto: Centre for Addiction and Mental Health; 2015. [PubMed]
  • Reinbach HC, Martinussen T, Møller P. Effects of hot spices on energy intake, appetite and sensory specific desires in humans. Food Quality and Preference. 2010;21:655–661.
  • Robinson E, Nolan S, Tudur-Smith C, Boyland EJ, Harrold JA, Hardman CA, et al. Will smaller plates lead to smaller waists? A systematic review and meta-analysis of the effect that experimental manipulation of dishware size has on energy consumption. Obesity Reviews. 2014;15:812–21. [PubMed]
  • Rodin J. Current status of the internal-external hypothesis for obesity. What went wrong? American Psychologist. 1981;36:361–72. [PubMed]
  • Rolls BJ, Hetherington M, Burley VJ. The specificity of satiety: the influence of foods of different macronutrient content on the development of satiety. Physiology & Behavior. 1988;43(2):145–53. [PubMed]
  • Rolls BJ, Bell EA, Castellanos VH, Chow M, Pelkman CL, Thorwart ML. Energy density but not fat content of foods affected energy intake in lean and obese women. American Journal of Clinical Nutrition. 1999;69:863–71. [PubMed]
  • Rolls BJ. The relationship between dietary energy density and energy intake. Physiology and Behavior. 2009;14(5):609–15. [PMC free article] [PubMed]
  • Rolls BJ. What is the role of portion control in weight management? International Journal of Obesity. 2014;38:S1–8. [PMC free article] [PubMed]
  • Russell MAH, Sutton SR, Feyerabend C, Saloojee Y. Smokers' response to shortened cigarettes: dose reduction without dilution of tobacco smoke. Clinical Pharmacology and Therapeutics. 1980;27(2):210–8. [PubMed]
  • Rychetnik L, Frommer M, Hawe P, Shiell A. Criteria for evaluating evidence on public health interventions. Journal of Epidemiology and Community Health. 2002;56(2):119–27. [PMC free article] [PubMed]
  • Schünemann HJ, Oxman AD, Vist GE, Higgins JPT, Deeks JJ, Glasziou P, Guyatt GH. Chapter 12: Interpreting results and drawing conclusions. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from http://www.cochrane-handbook.org.
  • Scisco JL, Blades C, Zielinski MJ, Muth ER. Dividing a fixed portion into more pieces leads to larger portion size estimates of JELL-O® squares. Perception. 2012;41:988–90. [PMC free article] [PubMed]
  • Shah M, Schroeder R, Winn W, Adams-Huet B. A pilot study to investigate the effect of plate size on meal energy intake in normal weight and overweight/obese women. Journal of Human Nutrition and Dietetics. 2011;24(6):612–5. [PubMed]
  • Skidelsky R, Skidelsky E. How Much is Enough?: Money and the Good Life. New York, NY: Other Press; 2013.
  • Small L, Lane H, Vaughan L, Melnyk B, McBurnett D. A systematic review of the evidence: the effects of portion size manipulation with children and portion education/training interventions on dietary intake with adults. Worldviews on Evidence-Based Nursing. 2013;10(2):69–81. [PubMed]
  • Smith LP, Ng S-W, Popkin BM. Trends in US home food preparation and consumption: analysis of national nutrition surveys and time use studies from 1965–1966 to 2007–2008. Nutrition Journal. 2013;12:45. [PMC free article] [PubMed]
  • Spanos S, Kenda AS, Vartanian LR. Can serving-size labels reduce the portion-size effect? A pilot study. Eating Behaviors. 2015;16(0):40–2. [PubMed]
  • Spears D. Economic decision-making in poverty depletes behavioral control. 2010. CEPS Working Paper.
  • Steenhuis I, Vermeer W. Portion size: review and framework for interventions. International Journal of Behavioral Nutrition and Physical Activity. 2009;6(1):58–67. [PMC free article] [PubMed]
  • Stewart LA, Tierney JF, Clarke M. Chapter 19: Reviews of individual patient data. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from http://www.cochrane-handbook.org.
  • Stunkard AJ, Messick S. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. Journal of Psychosomatic Research. 1985;29(1):71–83. [PubMed]
  • Tedstone A, Anderson S, Allen R. Sugar reduction: responding to the challenge. 2014. London: Public Health England.
  • Thomas J, Brunton J, Graziosi S. EPPI-Reviewer 4.0: software for research synthesis. 2010. EPPI-Centre Software. London: Social Science Research Unit, Institute of Education.
  • United Nations. Outcome document of the high-level meeting of the General Assembly on the comprehensive review and assessment of the progress achieved in the prevention and control of non-communicable diseases. New York: United Nations; 2014. p. 2.
  • Us Food, Drug Administration. Chapter I: Food and Drug Administration, Department of Health and Human Services; Subchapter B - Food for Human Consumption (Title 21, Volume 2, 21CFR101.12, Revised April 1, 2014) 2014;2 Code of Federal Regulations - Title 21.
  • Van Strien T, Frijters JER, Bergers GPA, Defares PB. The Dutch Eating Behavior Questionnaire (DEBQ) for assessment of restrained, emotional and external eating behavior. International Journal of Eating Disorders. 1986;5(2):295–315.
  • Versluis I, Papies EK, Marchiori D. Preventing the pack size effect: exploring the effectiveness of pictorial and non-pictorial serving size recommendations. Appetite. 2015;87(0):116–26. [PubMed]
  • Wagenaar AC, Salois MJ, Komro KA. Effects of beverage alcohol price and tax levels on drinking: a meta-analysis of 1003 estimates from 112 studies. Addiction. 2009;104(2):179–90. [PubMed]
  • Wallis DJ, Hetherington MM. Emotions and eating. Self-reported and experimentally induced changes in food intake under stress. Appetite. 2009;52(2):355–62. [PubMed]
  • Wang Y-C, Ludwig DS, Sonneville K, Gortmaker SL. Impact of change in sweetened caloric beverage consumption on energy intake among children and adolescents. Archives of Pediatric and Adolescent Medicine. 1009;163(4):336–43. [PubMed]
  • Wansink B, van Ittersum K. Shape of glass and amount of alcohol poured: comparative study of effect of practice and concentration. BMJ. 2005;331(7531):1512–4. [PMC free article] [PubMed]
  • Wansink B, Payne CR, Chandon P. Internal and external cues of meal cessation: the French paradox redux? Obesity. 2007;15(12):2920–4. [PubMed]
  • Welch V, Petticrew M, Tugwell P, Moher D, O'Neill J, Waters E, et al. and the PRISMA-Equity Bellagio Group. PRISMA-Equity 2012 Extension: Reporting Guidelines for Systematic Reviews with a Focus on Health Equity. PLoS Medicine. 2012;9(10):e1001333. [PMC free article] [PubMed]
  • Welsh JA, Sharma AJ, Grellinger L, Vos MB. Consumption of added sugars is decreasing in the United States. American Journal of Clinical Nutrition. 2011;94(3):726–34. [PubMed]
  • White IR. Multivariate random-effects meta-regression: updates to mvmeta. Stata Journal. 2011;11(2):255–70.
  • Williams J, Clemens S, Oleinikova K, Tarvin K. The skills for life survey: a national needs and impact survey of literacy, numeracy and ICT skills. 2003. Norwich: Department of Education and Skills.
  • World Health Organization. WHO Framework Convention on Tobacco Control. Geneva: World Health Organization; 2003.
  • World Health Organization. World Health Statistics 2014. 2014. p. 88. Geneva: World Health Organization.
  • World Health Organization. World Health Statistics 2014. 2014. p. 46. Geneva: World Health Organization.
  • Yeomans MR, Lee MD, Gray RW, French SJ. Effects of test-meal palatability on compensatory eating following disguised fat and carbohydrate preloads. International Journal of Obesity and Related Metabolic Disorders. 2001;25(8):1215–24. [PubMed]
  • Young L, Nestle M. The contribution of expanding portion sizes to the US obesity epidemic. American Journal of Public Health. 2002;92:246–9. [PubMed]
  • Young LR, Nestle M. Reducing portion sizes to prevent obesity. American Journal of Preventive Medicine. 2012;43(5):565–8. [PubMed]
  • Zlatevska N, Dubelaar C, Holden SS. Sizing up the effect of portion size on consumption: a meta-analytic review. Journal of Marketing. 2014;78:140–54.
  • Zung WWK. Zung Self-Rating Depression Scale and Depression Status Inventory. In: Sartorius N, Ban TA, editors. Assessment of Depression. Berlin: Springer-Verlag; 1986. pp. 221–31.

Articles from Wiley-Blackwell Online Open are provided here courtesy of Wiley-Blackwell, John Wiley & Sons