Inflammatory biomarkers, hormone replacement therapy, and incident coronary heart disease: prospective analysis from the Women's Health Initiative observational study

Abstract

Context: Postmenopausal hormone replacement therapy (HRT) has been shown to elevate C-reactive protein (CRP) levels. Several inflammatory biomarkers, including CRP, are associated with increased cardiovascular risk. However, whether the effect of HRT on CRP represents a clinical hazard is unknown.

Objectives: To assess the association between baseline levels of CRP and interleukin 6 (IL-6) and incident coronary heart disease (CHD) and to examine the relationship between baseline use of HRT, CRP, and IL-6 levels as they relate to subsequent vascular risk.

Design, setting, and participants: Prospective, nested case-control study of postmenopausal women, forming part of the Women's Health Initiative, a large, nationwide, observational study. Among 75 343 women with no history of cardiovascular disease or cancer, 304 women who developed incident CHD were defined as cases and matched by age, smoking status, ethnicity, and follow-up time with 304 study participants who remained event free during a median observation period of 2.9 years.

Main outcome measure: Incidence of first myocardial infarction or death from CHD.

Results: Median baseline levels of CRP (0.33 vs 0.25 mg/dL; interquartile range [IQR], 0.14-0.71 vs 0.10-0.47; P<.001) and IL-6 (1.81 vs 1.47 pg/mL; IQR, 1.30-2.75 vs 1.05-2.15; P<.001) were significantly higher among cases compared with controls. In matched analyses, the odds ratio (OR) for incident CHD in the highest vs lowest quartile was 2.3 for CRP (95% confidence interval [CI], 1.4-3.7; P for trend =.002) and 3.3 for IL-6 (95% CI, 2.0-5.5; P for trend <.001). After additional adjustment for lipid and nonlipid risk factors, both inflammatory markers were significantly associated with a 2-fold increase in odds for CHD events. As anticipated, current use of HRT was associated with significantly elevated median CRP levels. However, there was no association between HRT and IL-6. In analyses comparing individuals with comparable baseline levels of either CRP or IL-6, those taking or not taking HRT had similar CHD ORs. In analyses stratified by HRT, we observed a positively graded relationship between plasma CRP levels and the OR for CHD among both users and nonusers of HRT across the full spectrum of baseline CRP.

Conclusions: These prospective findings indicate that CRP and IL-6 independently predict vascular events among apparently healthy postmenopausal women and that HRT increases CRP. However, use or nonuse of HRT had less importance as a predictor of cardiovascular risk than did baseline levels of either CRP or IL-6.