High-density lipoprotein cholesterol and risk of ischemic stroke mortality. A 21-year follow-up of 8586 men from the Israeli Ischemic Heart Disease Study.

BACKGROUND AND PURPOSE

While there is overwhelming evidence relating low levels of HDL cholesterol (HDL-C) with coronary heart disease, the association with cerebrovascular disease is not clear. The aim of the present report was to assess the association between HDL-C levels and ischemic stroke mortality obtained from a long-term follow-up in the Israeli Ischemic Heart Disease Study.

METHODS

The subjects of this report are 8586 men, tenured civil servants and municipal employees, aged 42 years or older at the time of HDL-C measurements in 1965. They were followed up for mortality for 21 years. Death due to cerebrovascular disease included the International Classification of Disease, 9th Revision, codes 430 to 438, of which presumed ischemic stroke included codes 433 to 438.

RESULTS

During the 21-year follow-up, 295 men died from cerebrovascular events, of which 241 deaths were due to presumed ischemic stroke. Individuals subsequently experiencing a fatal ischemic stroke had a marginally lower age-adjusted mean HDL-C (1.05 mmol/L) and a significantly lower (P < .001) age-adjusted mean percentage of serum cholesterol contained in the HDL fraction (%HDL) (19.3%) than counterparts surviving the follow-up period (1.06 mmol/L and 20.6%, respectively). Decreasing age-adjusted rates of ischemic stroke mortality were observed with increasing %HDL: 14.6, 14.0, and 11.8 per 10,000 person-years in the low, middle, and upper tertiles of %HDL, respectively. In multivariate analysis, a low concentration of HDL-C appeared to be significantly predictive of ischemic stroke mortality. The relative risk associated with a 5% decrease of %HDL was 1.18 (95% confidence interval, 1.03 to 1.34). Men at the lower tertile of HDL-C levels experienced a 1.32-fold increase of covariate-adjusted ischemic stroke mortality risk compared with counterparts at the upper tertile.

CONCLUSIONS

In this prospective study of middle-aged and elderly men from a healthy, working population, we have demonstrated an independent negative association between HDL-C and ischemic stroke mortality during a long-term (21-year) follow-up.