Measurement error from assessing use of vitamin supplements at one point in time.

Although many epidemiologic studies ask about current use of vitamin supplements, long-term use is usually the exposure of etiologic interest. We conducted a mailed survey to investigate the relation between current and long-term (10-year) supplement use (N = 325 adults). Estimates of current daily intake for supplemental micronutrients were roughly twice that of average daily intake over the past 10 years. Correlations between current intake and long-term intake from supplements alone were 0.77, 0.75, and 0.65 for vitamin C, vitamin E, and calcium, respectively. A measure of supplement use at one point in time incorporates measurement error that will attenuate measures of association.