Is low serum chloride level a risk factor for cardiovascular mortality?

Background Serum chloride level is routinely assayed in clinical laboratories in the management of patients with kidney disorders and with metabolic diseases. It is a biological parameter that is easily, precisely and relatively cheaply measured. The epidemiological features of serum chloride levels have not been studied before. Methods For the random sample of men and women from the Belgian Interuniversity Research on Nutrition and Health aged 25-74 years, free of symptomatic coronary heart disease at baseline, serum chloride concentrations were measured, among those of other electrolytes. The cohort was followed up for 10 years with respect to subsequent cause-specific mortality. Results The results are based on observations of 4793 men and 4313 women. According to Cox regression analysis serum chloride level was one of the strongest predictors of total, cardiovascular disease (CVD) and non-CVD mortalities independently of age, body mass index, sex, smoking, systolic blood pressure, levels of total and high-density lipoprotein cholesterol, uric acid, serum creatinine and serum total proteins and intake of diuretics. This relation was proved to be independent of levels of other serum electrolytes and similar for men and women. The estimated adjusted risk ratio for CVD death for subjects with a serum chloride level ≤ 100 mmol/l compared with those with levels above that limit was 1.65 (95% confidence interval 1.06-2.57) for men and 2.16 (95% confidence interval 1.11-4.22) for women. The study of adjusted risk ratios for four groups of subjects defined on the basis of their baseline serum chloride levels revealed a decreasing log-linear ‘dose-response’ relation to total and cardiovascular mortalities. Conclusion This is the first report from a population-based study to indicate that there is an association between serum chloride level and the incidence of total, CVD and non-CVD mortalities. The risk ratio for CVD mortality associated with a low serum chloride level was comparable to or higher than those observed for well-established CVD risk factors