Population-representative analysis of dietary supplementation among Americans with diabetes mellitus

Few population-representative studies have examined dietary supplement use among Americans with diabetes mellitus (DM). This investigation analyzed dietary supplementation and DM data from the 2013-2014 National Health and Nutrition Examination Survey (NHANES).

Data from 5040 adult participants were analyzed, including 639 participants reporting a diagnosis of DM (Method 1), which increased to 1036 after including participants reporting "borderline" DM or prediabetes (Method 2). Participants reported dietary supplements used over the past month. The prevalence of supplementation was compared among participants with and without DM using the Rao-Scott likelihood Chi-squared test; multivariate logistic regression was used to examine whether DM was an independent predictor of supplementation.

Regardless of whether Method 1 or Method 2 was used, dietary supplementation was more prevalent among participants with DM (Method 1, 61.4% vs. 54.4%, P = 0.024; Method 2, 62.7% vs. 53.5%, P < 0.001). After adjusting for a variety of demographic and health-related variables, the presence of DM was no longer associated with supplementation for Method 1 (odds ratio [OR] 1.01; 95% confidence interval [CI] 0.78-1.32; P = 0.925) or Method 2 (OR 1.13; 95% CI 0.91-1.41; P = 0.238). Approximately four of every five products taken for the purpose of managing DM were self-prescribed, the most common being multivitamins-minerals, cinnamon-containing supplements, and chromium-containing supplements.

Slightly more than six out of 10 Americans with DM use dietary supplements each month and, although this is higher than Americans without DM, it does not appear to be due to DM per se.