Intake of Added Sugar from Different Sources and Risk of All-Cause Mortality and Cardiovascular Diseases: The Role of Body Mass Index

Background: Recent studies have shown an ascending trend in global added sugar consumption. Evidence for the long-term consequences of added sugar from different sources on all-cause mortality and cardiovascular diseases (CVDs) remains limited. Objectives: This study aimed to examine the associations between added sugar from various sources and the risk of all-cause mortality and CVDs, and to determine whether genetic predisposition and body mass index (BMI, in kg/m(2)) influence or mediate these associations. Methods: We included 196,414 UK Biobank participants who completed a 24-h food survey between 2009 and 2012. Sugar contents were collected from the Composition of Foods Integrated Data set (CoFID). The National Death Registries and hospital records provided data on death and the disease diagnosis. We employed a polygenic risk score (PRS) to assess the genetic predisposition. Cox proportional hazards regression was used to analyze the associations. Results: Totally, 10,081 deaths, 38,563 hypertension cases, 12,306 ischemic heart diseases (IHD), and 5491 cerebrovascular diseases were documented. Compared with the lowest quartile group of added sugar intake, the hazard ratios for all-cause mortality in the highest quartile were 1.21 (95% CI: 1.14, 1.30) for total added sugar, 1.03 (95% CI: 0.97, 1.10) for solids, and 1.16 (95% CI: 1.10, 1.23) for beverages. For CVDs, significant associations were observed in total added sugar and beverage sources. These associations were not altered by PRS, and individuals at greatest risk showed higher PRS along with excessive added sugar consumption ( P (trend) < 0.001). BMI was found to mediate the highest proportion of the association between added sugar and hypertension (19.10% for total; 36.95% for beverages). Conclusions: Higher intake of added sugar, especially from beverages, is associated with an increased risk of all-cause mortality and CVDs. BMI mediates a proportion of these associations.