heart disease (RR = 1.96; p = 0.022) and stroke (RR = 4.17; p = 0.002) (Eichholzer et al., 1992; Gey et al., 1993b). Based upon these and other data, Gey et al. (1993a) proposed that more than 0.4 to 0.5 µmol/L (21 to 27 µg/dL) α-plus β-carotene or 0.3 to 0.4 µmol/L (16 to 21 µg/dL) β-carotene is needed to reduce the risk of ischemic heart disease.
Total serum carotenoids, measured at baseline in the placebo group of the Lipid Research Clinics Coronary Primary Prevention Trial, were inversely related to subsequent coronary heart disease events (Morris et al., 1994). Men in the highest quartile of total serum carotenoids (more than 3.16 µmol/L [172 µg/dL]) had an adjusted relative risk of 0.64 (95 percent CI = 0.44−0.92); among those who never smoked, the relative risk was 0.28 (95 percent CI = 0.11−0.73). Riemersma e t al. (1991) reported that persons with plasma carotene concentrations in the lowest quintile (less than 0.26 µmol/L [14 µg/dL]) had 2.64 times the risk of angina pectoris. Adjustment for smoking reduced the magnitude of risk. However, because smoking may be part of the causal path, adjustment may not be appropriate.
The U.S. Health Professionals Follow-up Study of over 39,000 men reported a relative risk for coronary heart disease of 0.71 (95 percent CI = 0.55−0.92) for those at the top quintile of total carotene intake relative to the lowest quintile of intake (Rimm et al., 1993). The effect of β-carotene varied by smoking status: among current smokers, the relative risk was 0.30 (95 percent CI = 0.11−0.82); among former smokers, the risk was 0.60 (95 percent CI = 0.38−0.94), and among nonsmokers, the risk was 1.09 (95 percent CI = 0.66−1.79). A prospective cohort study of postmenopausal women found that the lowest risk of coronary heart disease was found for dietary carotenoid intakes greater than 8,857 IU/day (RR = 0.77; p = NS) (Kushi et al., 1996). A case-control study in 10 European countries found that lycopene concentrations, but not other carotenoid concentrations, in adipose tissue were inversely associated with the risk of myocardial infarction (Kohlmeier et al., 1997).
Cardiovascular epidemiology studies are now pursuing the use of intermediate endpoints, such as intima-media thickness, which can be estimated via ultrasonography as a measure of atherosclerosis. Bonithon-Kopp et al. (1997) reported a decrease in the intima-media thickness of the common carotid arteries with increasing concentrations of total plasma carotenoids in both men and women. Plasma carotenoid concentrations in excess of 2.07 µmol/L (111 µg/dL) were optimal for men; concentrations in excess of 3.73 µmol/L (200 µg/dL) were optimal for women. Salonen et al. (1993)