This study used data from the UK Millennium Cohort Study to examine associations between SSB and ASB consumption and adiposity changes in UK children between 7 and 11 years. This age period is important as it is associated with accelerated weight gain [24]. We found that regular consumption of SSBs and ASBs was associated with greater increases in BMI and percentage body fat. These results remained statistically significant after adjusting for a variety of socio-demographic and behavioral factors associated with an increased risk of adiposity. ASB and SSB consumption was lower among children of mothers with higher educational qualifications and SSB consumption was higher among south Asian children. Children who did not eat breakfast every day and those watching more than two hours of TV per day were more likely to drink both ASBs and SSBs.

While SSB consumption is an established risk factor for adiposity [1], few studies have examined associations with ASBs. A regional study of 1200 children in England found a positive association between ASB consumption and changes in adiposity between 5 and 9 years [25]. However, the authors suggest that this finding may reflect reverse causality due to overweight children switching to ASBs as part of an unsuccessful weight loss strategy. Associations between ASB consumption and adiposity remained statistically significant in our sensitivity analysis which excluded children who were obese at age 11 years. This suggests that our findings are not fully explained by substitution to ASBs among heavier children. Our findings are also consistent with those from an intervention trial which found that substitution of SSBs with water resulted in substantially lower calorie intake when compared with substitution with ASBs [26] while an intervention directed at adolescents encouraging substitution to ASBs was not associated with lower BMI than controls [27]. These latter findings support the conclusion that in the absence of energy restrictions, switching to ASBs is not an effective strategy for weight loss [28].

A variety of possible mechanisms by which artificial sweeteners may be linked to weight gain have been proposed: such sweeteners may increase appetite and lead to increased consumption, or they may train the palate to enjoy similar sweet foods, which may or may not be low calorie [28, 29]. Another possible explanation is that people may consciously overcompensate for these low calorie options and overindulge in other intakes [28, 29]. There is at present no conclusive evidence on these possible mechanisms, and the area remains an open area of research [10, 28]. In addition to adiposity, ASBs have also been linked to other health outcomes, such as incident diabetes in a recent meta-analysis [30].

This study is the first national longitudinal examination of associations between SSBs and ASBs and adiposity among pre-adolescent children. We used two objective measures of adiposity in children and found similar associations for both which strengthens the findings. The MCS is designed to be representative of the population of the UK and the percentage of our sample which was overweight or obese is similar to officially published estimates. 29.9 % of our sample was overweight or obese, which is similar to data from the National Child Measurement Programme for England which found 33.3 % of age 10 and 11 year old children to be overweight or obese in 2012/13 [31]. Nonetheless, there are a number of study limitations that should be considered. Measurement of beverage consumption was based on caregiver reporting, as were potential confounding variables. The MCS is not specifically designed with a nutrition focus and it remains possible that there are other important elements of diet which we have not been able to control for. In particular, inclusion of data on overall diet quality would have been useful in exploration of these issues. However, there was no data available on overall diet quality, and the data we use on snacking behavior comes from age 7 and may not accurately reflect snacking at age 11.This means that residual confounding remains a possible explanation for these findings, especially given that adjustment reduced effect sizes. Nonetheless we did control for a wide range of factors related to adiposity in previous work.Our predictor variable was based on sweetened beverage consumption at age 11 years, rather than at both time points in the study. Inclusion of data on consumption from both time points would have allowed less cautious interpretation of the results presented here and without this further research is needed to determine potential associations between these beverages and adiposity in children. Although parents were asked about beverage consumption at age 7 years, the question referred only to intake between meals and did not discriminate between artificial and sugar sweetened beverages. We fitted a categorical variable for exposure to SSBs or ASBs which meant that there was some sweetened beverage consumption in the reference group (at most once a month). Exposure to ASBs and SSBs was not mutually exclusive in the sample as a large number of children (40.7 %) were consuming both. However, findings from sensitivity analysis which involved fitting a four way categorization where weekly SSB/ASB consumption was mutually exclusive were consistent with our main findings. Although changes in adiposity seen were modest and may be accounted for by measurement error at the individual level, given our sample size it is unlikely that our findings areentirely due to artefact. Our findings may be partly explained by reverse causality e.g. more adipose children switching to ASBs as part of an unsuccessful weight loss strategy. However, associations between ASB and adiposity were statistically significant in sensitivity analyses which excluded obese children.

The World Health Organization recently published draft guidance on sugar intake; that adults should keep their intake below 10 % of total calories. The guidance also states that reducing sugar intake to below 5 % of total calories would yield additional population health benefits [32]. The WHO guidance is underpinned by robust epidemiological evidence indicating sugar consumption as an important cause of excess weight gain in both adults and children globally [1, 33]. Reducing SSB consumption offers considerable scope to achieve the WHO recommendations as SSBs account for up to 10 % of children’s energy intake in the UK [25]. Policy options to reduce intakes include upstream interventions such as cap and trade policies on adding sugar to the food chain, similar to the international carbon trading system [34]. Other proposed interventions include warning labels on SSBs [35], removing such drinks from schools [36] or additional taxes on such drinks (recently passed in Berkeley California). Findings from this and previous research indicate that encouraging substitution of SSBs with ASBs may not yield the health benefits originally intended in terms of adiposity or diabetes and substitution with water may be preferable Although evidence on policy impacts is emerging, the ubiquity of both SSBs and ASBs suggests that a variety of interventions are likely to be required in tandem, and that national governments will need support, co-ordination and guidance from international organizations [37]. Additionally, synergies identified between behaviors such as screen viewing and consumption [38], as well as the importance of physical activity [21] suggest that multi-faceted interventions tackling a variety of risk factors are likely to be be needed. Given our findings that consumption of sweetened beverages is higher in lower socio-economic and ethnic minority groups, it is important that any impacts on inequalities in childhood obesity are carefully monitored.