Dietary supplement use and medical conditions: the VITAL study

Am J Prev Med. 2003 Jan;24(1):43-51. doi: 10.1016/s0749-3797(02)00571-8.


Background: Over half of U.S. adults use vitamin or mineral supplements, and some are likely using supplements to treat chronic diseases or risk factors for disease. Information on the relationship between supplement use and medical conditions is useful to health professionals to understand the self-medication behavior of their patients, and important for researchers because medical conditions may be potential confounding factors in observational studies of supplement use and disease risk.

Methods: The cross-sectional data in this report are from 45,748 participants, aged 50 to 75 years, who completed a self-administered, mailed questionnaire on current dietary supplement use (multivitamins plus 16 individual vitamins or minerals), medical history (cancer, cardiovascular-related diseases, and other self-reported medical conditions), and demographic characteristics.

Results: Supplement use (mean number used at least once a week) was higher among respondents who were older, female, highly educated, Caucasian, and of normal body mass index (all p<0.001). After controlling for these covariates, supplement use was higher among those with the condition for 13 of the 21 conditions examined (p<0.01); only having diabetes or high stress was associated with using fewer supplements. For specific supplements, the strongest associations were for cardiovascular disease and its risk factors with vitamin E, niacin, and folate, and for calcium with indigestion and acid reflux disease. For several conditions, the relative odds of using specific supplements were consistently higher for men than for women.

Conclusions: Supplement use was associated with many medical conditions in this cohort. However, these cross-sectional data do not permit inferences about the temporal sequence. Some associations appeared to be based on evidence for efficacy (e.g., folate with coronary artery disease), and others could be based on misinformation (e.g., selenium with benign prostatic hyperplasia).

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Age Distribution
  • Aged
  • Body Mass Index
  • Cohort Studies
  • Dietary Supplements / statistics & numerical data*
  • Educational Status
  • Female
  • Health Status*
  • Humans
  • Male
  • Middle Aged
  • Obesity
  • Sex Distribution
  • Surveys and Questionnaires
  • Vitamins / administration & dosage*
  • Washington


  • Vitamins