Relationship between serum progesterone concentrations and cardiovascular disease, diabetes, and mortality in elderly Swedish men and women: An 8-Year prospective study

Abstract

The use of synthetic progesterone-like substances in hormone replacement therapy and birth control pills has been associated with increases in cardiovascular morbidity and the prevalence of diabetes. The primary aims of this study were to investigate whether physiologic concentrations of progesterone might also be associated with cardiovascular disease and diabetes, and to explore potential gender differences in these associations in elderly Swedish men and women. This prospective, longitudinal study was performed in a Swedish population-based sample of opposite-sex twins aged between 71 and 80 years who were not receiving sex hormone therapy. Serum concentrations of progesterone, estradiol, C-reactive protein (CRP), and urea were measured at baseline (1996) and at 8-year follow-up (2004), and serum concentrations of cystatin and insulin were measured only at follow-up. The outcomes of interest were cardiovascular morbidity (myocardial infarction, angina pectoris, peripheral arterial disease, stroke, congestive heart failure [CHF], cardiac arrhythmia, hypertension, and thromboembolism), diabetes, and mortality throughout the observation period. At baseline, the study sample included 230 men and 195 women (mean [SD] age, 74.6 [2.6] years). At follow-up, 132 men and 145 women (mean age, 82.4 [2.5] years) met the inclusion criteria, of whom 128 men and 112 women did so at both baseline and follow-up. Serum progesterone concentrations, which did not differ significantly from reported concentrations for the age group, were significantly associated with mortality across the observation period (P < 0.001). At follow-up, higher serum progesterone was significantly associated with the occurrence of CHF (P < 0.01); this association remained robust after adjustment for CRP, cystatin, and insulin levels. In these elderly Swedish men and women, increased physiologic concentrations of progesterone were found to be associated with an increased prevalence of CHF, independent of inflammatory factors, markers of renal function, and insulin metabolism.