LONGITUDINAL CHANGES IN SERUM BETA-CAROTENE AND C-REACTIVE PROTEIN

Abstract

Elevated high sensitivity c-reactive protein (hs-CRP) and decreased serum beta-carotene are associated with increased risk of cardiovascular disease. The acute phase response, characterized by a rapid increase in hs-CRP, decreases serum beta-carotene levels, and hs-CRP and beta-carotene are inversely correlated in cross-sectional studies. However, the effect of chronic low-level inflammation on serum beta-carotene is unknown and the value of a single time point measurement of hs-CRP is questionable. PURPOSE To longitudinally examine the association between serum beta-carotene and hs-CRP over one year. METHODS Serum beta-carotene and hs-CRP were measured at baseline and quarterly (5 times) over one year in 109 subjects (62 M, 47 F) aged 20–70 yr. Dietary beta-carotene intake was assessed 12–15 times during the study (3 times per quarter) via 24-hour recall. RESULTS Average hs-CRP was not related to serum beta-carotene or to natural changes in serum beta-carotene across the 5 quarters. When subjects were grouped by BMI (n=34 <25 kg/m2, n=44 25–29.9 kg/m2, n=31 >30 kg/m2), average hs-CRP was significantly greater in obese (3.2 ± 1.9 mg/L) and overweight (2.1 ± 1.7 mg/L) than normal weight (1.1 ± 1.0 mg/L) subjects, and serum beta-carotene was significantly lower in obese subjects at each time point (obese <normal weight; p<0.05). In a regression model that combined within- (longitudinal effect) and between-individual effects (cross-sectional effect) the natural change in serum beta-carotene across the 5 visits was inversely related to hs-CRP in normal weight subjects only (p<0.05). When physical activity, age, gender, smoking status, and dietary beta-carotene were included in the model, serum beta-carotene remained significantly associated with hs-CRP (p<0.05) in normal weight subjects. CONCLUSION These data show a significant association between natural changes in serum beta-carotene and hs-CRP in normal weight subjects. This association was not evident in overweight and obese subjects. Supported by NHLBI Grant R01-HL52745 (to I.S. Ockene)