Predictors of Probiotic Use in the United States (P21-023-19)

Abstract

Abstract Objectives To identify predictors associated with probiotic use in the US using National Health and Nutrition Examination Survey (NHANES) data. Methods The study examined a large cross-sectional US population using NHANES data from 2005–2014. We utilized 24-hour dietary recall (DR) interviews and Dietary Supplement Use 30-Day recall (DSQ) to extract the information of Probiotic Food (PF) or Probiotic Supplements (PS) consumption, respectively, using text search terms. Adults (≥18 years), who participated in DR interviews and DSQ were included, while pregnant women were excluded from the study. Differences in probiotic consumers and non-consumers were analyzed by χ2/Wilcoxon tests. Potential predictors of probiotic use, such as gender, race, Healthy Eating Index (HEI), antibiotic use, Kcal Consumption, Body Mass Index (BMI), age and physical activity were evaluated using multivariate survey-weighted logistic regression. Predictors of PF and PS use were analyzed separately and weighted to reflect national estimates. Results Our study included 15,104,111 (13.79%) PF consumers and 94,421,839 (86.21%) non-consumers. The adjusted analysis suggested that, females were more likely to be PF consumers than males [Weighted Odds Ratio (OR) = 1.75, 95% confidence interval (CI) = 1.56–1.97]. Non-Hispanic white were more likely to consume PF than other races. Each unit increase in HEI was associated with 3% higher odds of PF consumers [1.03, 1.02–1.03] than non-consumers. Similarly, we identified 1108,619 (1.08%) PS users and 101,698,442 (98.92%) non-users. The odds of PS use were 1.75 times higher in females compared to males [1.75, 1.14–2.70]. Mexican American [0.20, 0.09–0.45] and Non Hispanic Black [0.52, 0.32–0.83] were less likely to be PS consumers compared to Non-Hispanic white population. Each unit increase in HEI was associated with 4% higher odds of PS use [1.04, 1.03–1.05] than non-consumers. Antibiotic use (9.37% vs. 4.52%) was significantly associated with PS use [2.30, 1.11–4.74]. Kcal consumption, BMI, age and physical activity were not associated with probiotic consumption. Conclusions This large population based nationally representative study identified females, HEI and Non-Hispanic white population as predictors of probiotic use, while antibiotic use was additional predictor for PS use. Funding Sources None.