The aim of this study is to investigate whether the prediction of all-cause mortality from traditional risk factors is improved by adding electrolytes (serum-phosphate (S-P), serum-calcium (S-Ca) and serum-magnesium (S-Mg)) in a Cox regression. The study uses an 18-year follow-up of patients (n=2504) referred by physicians in primary health care and hospitals to the Vindeln Patient Education (VPE) Center, mainly with a diagnosis of hypertension (HT), type 2 diabetes mellitus (DM) and/or obesity.
Cox regression, with the latest registered value and baseline values for risk factors, was used to study all-cause mortality in men and women. 221 out of 1096 men and 157 out of 1408 women died during the 18-year follow-up (20% and 11% respectively). The Cox regression analysis reveals that high blood glucose (B-Glu) and low S-Mg were significantly associated with increased all-cause mortality in the whole patient population as well as in men and women separately. Among women, type 2 DM and systolic blood pressure (SBP) and among men, high S-Ca, S-P, S-urate and body mass index (BMI) were the main predictors of all-cause mortality. There is significantly improved prediction of all-cause mortality with electrolytes added to the traditional risk factors. High B-Glu and low S-Mg in both men and women, and high S-Ca and S-P in men, are significantly associated with all-cause mortality. The metabolic disturbance in this high-risk group of patients can be more fully understood if ionic imbalance is included in the prediction of mortatlity.