Reviews examining sugar-sweetened beverages and body weight: correlates of their quality and conclusions

Abstract

Background: The role of sugar-sweetened beverages (SSBs) in increasing obesity is of great scientific, clinical, and public health interest. Many reviews have been published on this topic in recent years with very different conclusions.

Objective: We sought to assess the scientific quality and other characteristics that may be associated with the conclusions of reviews regarding the causal relation between SSB consumption and body weight.

Design: A systematic search of reviews in English language-published peer-reviewed journals in 2006-2013 was performed. Their methodologic quality was assessed by 2 judges using 2 scoring systems: the Assessment of Multiple Systematic Reviews and the American Dietetic Association Quality Criteria Checklist. The conclusions were blindly assessed by 11 independent readers using a Likert scale ranging from a position score of 0 = no evidence of a causal relation to 5 = strong evidence of a causal relation.

Results: Twenty reviews were identified: 5 meta-analyses, 3 qualitative systematic reviews, and 12 qualitative nonsystematic reviews. Four received funding from the food industry. Quality scores were neither correlated with the readers' perception of conclusions nor with the source of funding. However, industry-funded reviews were more likely to suggest that evidence supporting a causal relation between SSB consumption and weight gain was weak (mean position score = 1.78), whereas evidence was generally considered well-founded in other reviews (mean position score = 3.39; P ≤ 0.01).

Conclusions: For a complex and controversial scientific issue, it is important to minimize perceived or actual threats to scientific objectivity and methodologic quality. More refined tools are needed to better assess their scientific quality and to identify factors and mechanisms that may influence authors' conclusions.