RESUMO
BACKGROUND: Epidemiological studies aimed at correlating coronary heart disease (CHD) with serum ferritin levels have thus far yielded inconsistent results. We hypothesized that a labile iron component associated with non-transferrin-bound iron (NTBI) that appears in individuals with overt or cryptic iron overload might be more suitable for establishing correlations with CHD. METHODS AND RESULTS: We investigated the relation of NTBI, serum iron, transferrin saturation, and serum ferritin with risk of CHD and acute myocardial infarction (AMI). The cohort used comprised a population-based sample of 11 471 postmenopausal women aged 49 to 70 years at enrollment in 1993 to 1997. During a median follow-up of 4.3 years (quartile limits Q1 to Q3: 3.3 to 5.4), 185 CHD events were identified, including 66 AMI events. We conducted a case-cohort study using all CHD cases and a random sample from the baseline cohort (n=1134). A weighted Cox proportional hazards model was used to estimate hazard ratios for tertiles of iron variables in relation to CHD and AMI. Adjusted hazard ratios of women in the highest NTBI tertile (range 0.38 to 3.51) compared with the lowest (range -2.06 to -0.32) were 0.84 (95% confidence interval 0.61 to 1.16) for CHD and 0.47 (95% confidence interval 0.31 to 0.71) for AMI. The results were similar for serum iron, transferrin saturation, and serum ferritin. CONCLUSIONS: Our results show no excess risk of CHD or AMI within the highest NTBI tertile compared with the lowest but rather seem to demonstrate a decreased risk. Additional studies are warranted to confirm our findings.