The Association of Sugar Sweetened Beverages with Body Weight and Adiposity

Abstract

Dietary intake plays a crucial role in maintaining health and mitigating the risk of chronic diseases. However, accurately assessing food consumption for nutritional research is challenging due the use of self-reported methods, which may be affected by misreporting. Nutritional research is key in determining the impact of the diet on many aspects of health. Among dietary components, the amount of sugar consumed is of particular concern, as it contributes to higher total energy intake and has been consistently associated with increased body weight and obesity-related health risks. This thesis comprises of three studies all adding to the field of nutritional research. The first study developed a new algorithmic tool to evaluate the accuracy of self-reported dietary intake by screening large datasets using physical parameters such as age, sex, height, weight, and ethnicity. The tool predicted an individual’s daily energy expenditure (DEE) which, when weight stable, should match the daily energy intake (DEI). The resulting data was compared to previous methods for determining energy intake misreporting (the Goldberg and Black cutoffs and the McCrory approach). The tool itself was an equation developed by Professor John Speakman based on doubly labelled water (DLW) data (n = 6,497; ages 4–96), the study predicted DEE in the UK National Diet and Nutrition Survey (NDNS, 2008–2019; n = 12,694) and the US National Health and Nutrition Examination Survey (NHANES, 2017–2018; n = 5,873). Predicted total energy expenditure was compared with reported daily energy intake, allowing classification of participants as under-, over-, or acceptable reporters. The results showed that approximately 27% of participants in both datasets, were outside the acceptable range for accurate self-reporting of their DEI. Under reporting of DEI was significantly greater in individuals with a higher body mass index (BMI) (p value <0.005). In both datasets, protein intake was negatively associated, and fat intake positively associated ii with the discrepancy between DEE and DEI, whereas carbohydrate intake showed no association in the NDNS but a negative association in the NHANES. The second cross-sectional study also utilised a large dataset the UK Biobank (n =175444; aged 30-69), to explore the genetic variation in sweet taste receptors and its effects on sugary drink consumption and body fatness. Initially associations between sugary beverage consumption and single nucleotide polymorphisms (SNPs) in the sweet taste receptor gene TAS1R2 were examined. Beverage intake was assessed using a using four self-administered 24-hour dietary recalls. The results identified six SNPs in the TAS1R2 gene that cause amino acid changes; these were significantly associated with higher intakes of sugary beverages and intake of sugars/sweeteners (p value < 0.002). The study also investigated the associations between sugary beverage consumption and adiposity measures (BMI, waist-to-hip ratio (WHR), body fat percentage (BF %), abdominal subcutaneous adipose tissue (ASAT) and visceral adipose tissue (VAT)). Consumption of fizzy drinks was associated with higher levels of all adiposity measures (p value < 0.0001). While low-calorie drinks were associated with increased BMI, BF%, ASAT, and VAT, they were not associated with WHR. Pure fruit and vegetable beverages as well as intake of sugar to coffee or tea were linked to higher BMI and WHR, whereas intake of sweeteners in coffee or tea were associated with increased BMI, WHR, and body BF % (p < 0.0001). The third study generate a new dataset and was an observational study including 128 healthy adults living in Aberdeenshire, UK. Body composition was assessed using air displacement plethysmography, bioelectrical impedance analysis, and deuterium dilution. Sugar-sweetened beverage intake was assessed using two questionnaires, and sweet taste sensitivity was measured. The results showed that females had a higher body fat percentage than males (p value < 0.0001), although BMI did not differ by sex (p value > 0.05). Body fat percentage measured by all three methods was strongly correlated, with the strongest agreement between iii deuterium dilution and air displacement plethysmography. Sweet taste sensitivity did not differ by sex or smoking status and was not associated with BMI, body fat percentage, or sugar intake from sweetened beverages (p value > 0.05). In conclusion, improving the accuracy of dietary intake assessment is an important step before implementing nutrition-related health policies as reliable data are essential for food recommendations and monitoring national dietary habits. The identification of six TAS1R2 variants significantly associated with sweetened beverage consumption, suggests a biological influence on taste preferences. The strong association between intake of sweet beverages which include SSBs and ASBs and multiple adiposity markers underscores the need for targeted health recommendations and policies. These policies may include adding warning labels on caloric and non-caloric sweetener beverages to enhance informed consumption, as well as reducing the availability of sweetened beverages on public locations as vending machines in hospitals and universities. However, the lack of association between taste sensitivity and sugar-sweetened beverages intake in the Aberdeen study sample, suggests that environmental or behavioural factors may also play a significant role. 4