Around two-thirds of Australians are overweight or obese and, in 2008, the total cost of obesity (excluding overweight) in Australia was $ 8.3 billion [1
]. Primary prevention of childhood overweight is a high priority given 20-25% of Australian 2-8 year olds are already overweight [2
] and at substantially increased risk of becoming overweight adults, with attendant increased risk of morbidity and mortality [4
]. There is also correlational evidence linking parenting style and early feeding practices to child weight status [6
] but prospective longitudinal or intervention studies are rare.
Why target early feeding practices?
Parents, particularly mothers, are the 'gate keepers' of children's eating environments [7
]. Parent early feeding practices (i) determine infant exposure to food (type, amount, frequency) and (ii) include responses (e.g. coercion) to infant feeding behaviour (e.g. food refusal). These feeding practices strongly influence children's eating patterns, which are firmly established by five years of age and lay the foundation of adult eating habits [8
]. The degree of parental control over early feeding (restriction, monitoring and pressure) has been associated with child eating behaviour (preferences and intake) and weight status [6
]. Rapid early weight gain before two years of age is associated with a 2-3 fold increase in risk of later overweight [10
]. Most excess weight gained before puberty is gained by the age of five years (91% girls, 70% boys) [12
]. Contemporary feeding practices are seen to stem from culture, tradition and family experience. They evolved in the context of relative food scarcity (less than 2-3 generations ago) and have not adapted to western environments, where excess food can pose a major health risk. Therefore, new approaches that reflect key contemporary determinants of child eating behaviour are required [7
Given that poor eating patterns emerge early in life, early-life interventions are required. Recent US, nationally representative, cross-sectional data from the Feeding Infants and Toddlers Study (FITS) [13
] (n = 3022, 4-24 months), report poor intakes of fruit and vegetables and frequent use of non-core foods [14
]. Webb et al. [15
] report similar dietary quality issues in 429 Australian children aged 16-24 months who were enrolled in an asthma prevention trial. Approximately half the children drank cordial daily and two-thirds consumed fried potato, confectionary and non-milk sweetened beverages at least once over the three day record period. The mean consumption of 'extras' foods (energy-dense, nutrient poor) was 157 g per day and contributed 27% of daily energy intake.
Approaches to improving eating patterns in preschoolers: the evidence gap
A 2008 review [6
] examining the role of parenting and feeding practices in child eating behaviour and weight status highlights the explosion of research interest in this area. It concluded that (i) the vast majority of evidence is cross-sectional or experimental from a 'quasi laboratory' setting (only 7 of 67 studies were longitudinal, none of which included children under 5 years); (ii) very few studies examine parent feeding practices, child eating behaviour, intake and weight as a multidirectional mediation model; (iii) only two studies (both in preschools) examined whether parent feeding practices can be modified; and (iv) most studies failed to evaluate covariates, particularly maternal weight status and family socioeconomic status [6
Only two intervention studies have evaluated the influence of parenting and feeding practices on child eating behaviour and weight status. NEAT [16
] used a quasi-experimental design to evaluate a 6-month home-visit program to enhance feeding practices in toddlers (n = 135; mean age 19 months at baseline) and found minimal effects. This study was limited by a non-randomised design, use of a convenience sample, a short time-frame and absence of direct outcome data on child weight status, food preference or intake. Further, the intervention may have been too late to change already-established eating patterns. The second was a pilot study [17
] that involved a 16-week home-visit intervention for 40 Native American families (children aged 9-36 months). At the end of the program there was a reduction in weight-height z-score in active versus control groups (mean -0.27 ± 0.31; P
The NOURISH randomised controlled trial (RCT) is designed to promote feeding practices that will support healthy weight and growth. It will provide impact evaluation with respect to improving feeding practices and infant food preferences and intake up to age two years, as potentially modifiable determinants of weight status. It is intended that weight status at five years of age will be the primary outcome in longer follow up of the cohort, subject to further funding.
Rationale for the proposed NOURISH intervention
While parents and infants share a common genetic propensity for weight gain, the early feeding environment is critical for establishing eating habits [18
]. Figure summarises key factors that influence the reciprocal relationships between parent feeding practices and infant feeding behaviour, child food preferences and early food intake patterns. These, in turn, lay the foundation for later eating habits [7
]. The NOURISH intervention reflects the key determinants of healthy eating behaviour in infants and children.
Key factors that influence the reciprocal relationships between parent feeding practices and infant feeding behaviour.
Exposure and acceptance
Neophobia, the rejection of novel foods, is a normal adaptive response, but is readily modified by experience, particularly familiarity arising from repeated exposure [8
]. Repeated (≥ 10) neutral exposures within a short time frame enhance acceptance of new foods; both healthy (eg fruit and vegetables) [8
] and unhealthy (high fat and/or sugar, low nutrient) foods [20
]. Unfortunately, the wide availability of the latter energy dense, low nutrient foods ('non-core' or 'extras') in family diets means even very young children have high levels of exposure, potentially enhancing their access to and preference for such foods [7
]. There is a dearth of food intake data from Australian children under two years. Our pilot study of a random sample of 361 mothers of toddlers (aged 12-36 months) found evidence of poor dietary quality. On the day of survey 30% consumed ≥ 2 non-core foods and 39% had sweetened drinks.
Self-regulation of intake
Self regulation of intake in response to internal hunger and satiety cues is innate in infancy, but easily overridden by social and emotional cues from adults [21
]. Parental feeding practices such as explicit encouragement and praise, coercion, coaxing and the use of alternatives or rewards (food or otherwise) have been shown to be ineffective in improving food intake and variety [8
]. Satter argues for a 'parent provide child decide'
approach where the parent is responsible for providing safe, nutritious, developmentally-appropriate food and the child decides if, and how much to eat [22
]. However, data from our pilot study (see above) showed that such an approach is uncommon: 75% of mothers self-reported coaxing or coercing their child to eat more; only 56% interpreted general food refusal as satiety and 40% at least sometimes used food as a reward. More mothers were concerned about their child being underweight (22%) than overweight (9%) [23
]. These data are consistent with results from FITS [24
] and indicate a concerning prevalence of maternal anxiety about feeding, use of non-neutral approaches to food refusal, emotional use of food and failure to appropriately respond to internal hunger and satiety cues.
Attachment and parenting skills
Attachment refers to the enduring emotional tie between an infant and their primary caregiver [26
] who share repeated, characteristic interactions that shape each others' behaviour. Secure attachment develops when care is consistent, warm and sensitive [27
]. Enhancing attachment is a common goal of early intervention and prevention programs to promote parenting competence and skills and child health and well being [27
]. A meta analysis [28
] of 88 interventions (n = 1503), concluded that brief behavioural interventions (with 5-16 versus more than 16 sessions), that start in mid-infancy rather than perinatally, are most effective in enhancing maternal sensitivity (appropriate and prompt emotional and verbal responses to infant signals). Sample characteristics (SES, multiple social risk factors, adolescent mother, prematurity) were not effect modifiers.
While the attachment paradigm has not been used directly in the nutrition promotion context, attachment interventions commonly use video taping of feeding sessions as an intervention strategy and/or outcome measure [28
]. However, given that maternal sensitivity to infant cues of hunger and satiety are central to positive feeding practices, attachment provides a highly plausible and novel framework within which to develop behavioural strategies to enhance parental competence and skills in early feeding.
Parenting styles can be defined on the dimensions of behavioural control and responsiveness (warmth) and are related to parenting behaviours and feeding practices [6
]. In a cross-sectional study of 4-year olds (n = 231), authoritative parenting and feeding styles (high control, high warmth) were independently associated with higher intakes of dairy foods and vegetables, whilst authoritarian styles (high control, low warmth) were associated with lower intake of vegetables [30
]. A prospective study of 5-year olds (n = 872), reported that those exposed to authoritarian parenting were five times more likely to be overweight two years later than those exposed to authoritative parenting practices (after adjustment for a range of covariates, including child weight) [31
]. Several authors recommend targeting parenting and feeding styles, specifically encouraging authoritative feeding, in interventions to prevent child overweight [6
is a proactive and preventive approach. It provides parents with information about behaviours they can expect and positive ways to manage these, rather than waiting until parents seek advice once problems have become established. This approach has been shown to be effective in improving family and child outcomes across a range of domains [32
Overall, the following problems appear to be prevalent for Australian infants and toddlers: high exposure to non-nutritive, energy dense foods; maternal concern about feeding; use of non-neutral approaches to food refusal; emotional use of food; and coercive maternal feeding practices that fail to respond appropriately to infant hunger and satiety cues. These practices are linked to increased obesity risk [6
] and their reduction is the focus of the NOURISH intervention modules.
Aims and hypotheses
The NOURISH study aims to implement and undertake impact evaluation of a community-based intervention for first-time mothers of infants aged 4-7 months at enrolment that will
(i) foster healthy food preferences, dietary intakes and eating behaviours in very young children;
(ii) initiate and maintain positive maternal feeding practices in very young children; and
(iii) enhance maternal efficacy (knowledge, skills, confidence) with respect to child feeding.
A RCT will compare self-directed access to 'usual child health services' (control) with participation in a structured, comprehensive, maternal education and peer support program delivered when the infants are 4-7 months and 13-16 months of age and which will provide anticipatory guidance to improve early feeding practices (intervention). Follow up will be at two years of age. It is anticipated the intervention will result in:
H1: increased infant/child preferences for, and intake of, fruit and vegetables (frequency and variety);
H2: reduced infant/child preferences for, and intake of, non-core (low nutrient, energy dense) foods;
H3: increased frequency of maternal feeding practices that recognise and respond appropriately to infant cues of hunger and satiety and that support infant/child self-regulation of intake; and
H4: improved maternal efficacy and confidence with respect to child feeding.