Supplement use, other dietary and demographic variables, and serum vitamin C in NHANES II.

Abstract

Our objective was to evaluate the effect of regular use of nutritional supplements on serum vitamin C levels in a multivariable regression model, taking into account other dietary and demographic variables which may affect nutritional status. We analyzed NHANES II data for subjects age 3 to 74. Analysis was limited to regular supplement users and nonusers, excluding irregular users. Multivariable regression analysis was performed with SUDAAN, incorporating sample weights and accounting for the complex survey design. Regular supplement users had substantially higher serum vitamin C levels than nonusers (p < 0.001). The magnitude of the effect of supplement use on serum vitamin C was 0.23-0.33 mg/dL in children and teens, and 0.36-0.46 mg/dL in adults. In adults who smoked, bottom quartile vitamin C levels were 0.3 mg/dL in men and 0.4 mg/dL in women who did not use supplements, compared to 0.9 and 1.1 mg/dL in regular supplement users. There was a significant interaction of smoking and supplement use in men (p < 0.001). Regular supplement use has a strong impact on serum vitamin C levels, independent of other dietary and demographic characteristics of supplement users which may favor improved nutritional status.