Water is arguably the most critical nutrient, as its absence can be lethal within a few days [1
]. To promote adequate water intake at the population level, several countries and transnational authorities have developed water intake recommendations based on national estimates of water intake [4
]. In the UK, data on total water intake (TWI) is not normally published in national survey reports and there is currently no Dietary Reference Value. The main purpose of this study was to quantify TWI and its relation to patterns of beverage consumption and then to explore associations between types of beverage consumed and the intake of water and energy. It has been estimated from studies in Europe and the USA that around 70% -80% of TWI comes from beverages of various types (including water, tea and coffee, milk, soft drinks, juice and alcoholic drinks), with the remainder contributed by water in food [4
], and we expected the British population would have a similar pattern. Beverages are widely available in developed societies, where tap water is essentially free, but there are nevertheless concerns that some people may not be consuming sufficient fluid for optimal health and some people may be over-consuming. Children, elderly or infirm people, and those working in hot environments, are some of those most vulnerable to the effects of dehydration, but many adults may also be inadequately hydrated at some time. Unfortunately, hydration status is rarely measured in epidemiological studies and this hampers attempts to assess the adequacy of water intakes at a population level. However, guidelines have been established to determine how much water humans require (on average) to avoid dehydration and to optimise physical and psychological function.
In 2005, the Food and Nutrition board of the Institute of Medicine (IOM) published adequate intake (AI) values for TWI in temperate climates [5
]. The AI for total water (from a combination of drinking water, beverages, and food) was set based on the median total water intake from US survey data. For young men and women (19–30
years) this is 3.7
L and 2.7
L per day, respectively. These recommended intakes are based on median intakes of generally healthy individuals who are adequately hydrated, but the report pointed out that individuals can be adequately hydrated at levels below, as well as above, the AIs provided. The American AIs exceed those of other authorities, while the recent recommendations produced by the European Food Safety Authority (EFSA) in 2010 are the most conservative to date, at 2.0
L per day for adult females and 2.5
L per day for adult males [4
Limited data exist on daily water intake in other countries, and comparison is sometimes hampered by different methods of definition and data collection [8
]. In the UK, TWI is not currently quoted in the published reports but the mean value has been calculated as 2494
g/day among adults aged 19–64 y in 2008/9 [9
]. Average water intakes in other European countries appear to be broadly similar to those in the UK (e.g. mean 2461
ml/day in Sweden [10
]), or else lower than the UK, (e.g. mean intake 1984
ml/d in France [11
ml/day in Germany [12
ml/d in the Netherlands [13
]). Reported intakes of total water in North America are considerably higher than in Britain and Europe. In 2005–2006, American adults participating in the National Health and Nutrition Examination Surveys (NHANES) reported consuming 3.18
L of total water within the previous 24
], slightly less than the 3.35
L reported in 1999–2004.
While all beverages support hydration by virtue of their high water content, many also supply calories. Excessive consumption of caloric beverages has been widely viewed as contributing to the obesity epidemic [14
], although systematic reviews have highlighted the need for better randomised controlled trials, in order to demonstrate a causal effect [16
]. The main putative mechanism involves reduced satiety and incomplete compensation for calories ingested in liquid form [18
]. According to this theory, so-called “liquid calories” are more likely than solid calories to result in passive overconsumption and excess energy intake [19
]. There are, however, few studies that have examined relationships between beverage consumption patterns and energy intake in the British population, apart from the recent paper by Ng et al. [9
], which reported that the proportion of energy from beverages changed very little between 1986 and 2008/9, although there were some shifts between sources. Our paper differs in scope, focusing on consumption over 24
hours and 7
days of the week and comparing water intakes, in men and women with reference values.
The National Diet and Nutrition Survey database provides what is probably the best source of detailed information on the diets of normal individuals in Britain [20
]. In 2008 the survey adopted a new method, collecting data via a (non-weighed) diet record over 4
days, in place of the former weighed record over 7
days, in 2000/2001. The latter, slightly older, data provide the opportunity to study variation over days of the week and may give a better indication of participants’ usual intake. The food records list the weight in grams of each item of food and drink consumed (including tap water) for each of the 1724 participants (12,068 person-days of data), while the nutrient database includes the water, energy and nutrient content of each item. Time of consumption is also recorded, providing a rich resource for exploring patterns of consumption (timing, frequency, variety etc.) and linking this with personal data. Using raw data from the NDNS we attempted to address the following questions:
1) Is the UK adult population consuming adequate amounts of total water?
2) How does beverage consumption vary by age and gender, day of the week and time of day and is this related to total water intake?
3) Is the variety of beverages consumed a positive predictor of total water intake?
4) How much energy do beverages contribute to the UK diet and in what proportion?
5) Is energy in liquid form (i.e. from beverages) positively associated with total energy intake?