Added Sugar Consumption Is Not Relevantly Associated With an Increased Odds of Prediabetes in U.S. Adults

Abstract Objectives To examine whether self-reported added sugar consumption is associated with an increased risk for prediabetes in U.S. adults. Methods A secondary data analysis was performed using National Health and Nutrition Examination Survey data collected between 2013–2018 in a subsample of adults (≥20 years) with prediabetes (HgbA1c 5.7% to 6.4%) and normoglycemia (HgbA1c < 5.7%) who did not have a diagnosis of type 2 diabetes (HgbA1c ≥ 6.5%). Mean usual intake of added sugar (g/day) was estimated from two 24-h dietary recalls. Survey-weighted logistic regression was used to test whether (1) total mean intake (g/day) of added sugar or (2) tertiles of added sugar as an overall percentage of added sugar intake (<10%,10–15%, >15% g/day) were associated with an increased odds of prediabetes. Results In the sample of 10,671 adults, 34% were identified as having prediabetes, consuming an estimated average of 49.4 g/day of added sugar. In unadjusted models, added sugar was not relevantly associated with an increased odds of prediabetes [total: (OR = 1.001, P = 0.625); tertiles (reference: <10%): 10–15% (OR = 1.119, P = 0.206), >15% (OR = 1.057, P = 0.531)]. Similarly, in adjusted models that controlled for age, gender, race/ethnicity, total energy intake, physical activity status, smoking status, BMI, and socioeconomic covariates, added sugar was not relevantly associated with an increased odds of prediabetes [total: (OR = 1.001, P = 0.601); tertiles (reference: < 10%): 10–15% (OR = 1.032, P = 0.763), >15% (OR = 1.053, P = 0.665)]. Conclusions Results indicate that self-reported added sugar consumption does not appear to relevantly increase the odds of prediabetes in adults. Because prediabetes is an early indicator of type 2 diabetes risk, it is possible that exposure to added sugar over time contributes to the significant associations observed in individuals with type 2 diabetes and not prediabetes. Funding Sources University of Alabama at Birmingham Graduate School and School of Nursing.