Cross-sectional association between soda consumption and body mass index in a community-based sample of twins

BACKGROUND Consumption of sugar-sweetened beverages, such as soda, have been shown to play an important role in weight gain. Although soda consumption has been associated with body mass index (BMI) in many studies, it has been difficult to ascertain a true causal relationship between soda consumption and BMI for two reasons. First, findings have been based largely on observational and cross-sectional studies, with much less evidence from randomized controlled trials. Second, the reported relationships may be confounded by genetic and shared environmental factors that affect both soda consumption and BMI. In the present study, we used the twin design to better understand the relationship between soda consumption and BMI by accounting for measured and unmeasured confounds in non-experimental data. Associations from genetically informed tests in twins are considered "quasi-causal," suggesting that our confidence in the causal underpinning of the association between soda consumption and BMI has been strengthened. We hypothesized that the association between soda consumption and BMI would be significant both between and within twins. METHODS This was a cross sectional study of 5787 same sex adult twin pairs (18-97 years, 66% female) from the community based Washington State Twin Registry. Structural equation modeling (SEM) was employed to investigate associations between soda consumption and BMI in the population (the phenotypic association between exposure and outcome among all twins treated as individuals) and within pairs of identical and fraternal twins (the quasi-causal association controlling for between pair genetic and environmental confounds). RESULTS Among all twins, there was a significant phenotypic association between soda consumption and BMI that held when controlling for age, sex, race, annual household income, and education level (P < 0.05). In the quasi-causal model, however, the effect of soda consumption on BMI was greatly reduced and no longer significant, with a large genetic confound in both men and women (P < 0.05). CONCLUSION Among a large group of adult twin pairs, increased soda consumption was associated with increased BMI; however, the observed association was mediated by a genetic background common to both.