Purpose: The aim of this study was to examine the prevalence of self-reported multivitamin use in the Physicians' Health Study (PHS) cohort and its association with various lifestyle, clinical, and dietary factors to improve our understanding of who tends to use multivitamins.
Methods: Among 18,040 middle-aged and older men, information on lifestyle and clinical factors was collected from a baseline enrollment questionnaire, and supplement use and dietary factors were assessed through a food-frequency questionnaire. Four categories of multivitamin use were considered: (1) no supplement use, (2) use of multivitamins only, (3) use of multivitamins with other individual vitamin/mineral supplements, and (4) use of other supplements only. We used logistic regression to calculate multivariate odds ratios and 95% confidence intervals of taking multivitamin supplements for various lifestyle, clinical and dietary factors.
Results: Overall, 36% of men reported current multivitamin use. Men who were older, current smokers, and currently using aspirin were 143, 43, and 74% more likely to use multivitamins only. Men having a history of hypercholesterolemia were 16% more likely to use multivitamins only. A 14, 24, and 26% greater likelihood of using multivitamins was also observed among men consuming more fruits and vegetables, whole grains, and tea, respectively. Similar associations were observed for the likelihood of using multivitamins with other supplements; however, men with higher physical activity, history of cancer, hypertension, higher consumption of nuts, and lower consumption of red meat and coffee were also more likely to use multivitamins with other supplements (all P < 0.05).
Conclusion: Self-reported multivitamin use associated with lifestyle, clinical and dietary factors may be an indicator of healthy behaviors. These results provide important information for the interpretation of the recent findings from the PHS II trial and consideration of results from observational studies of multivitamin use and chronic disease.