Homocysteine Not Framingham Predicts Cardiovascular Mortality in Very Elderly

By Peggy Peck

LEIDEN, The Netherlands, 12 jan2009-- Homocysteine, a biomarker that has a rocky record in cardiovascular disease prevention trials, may prove to be useful as a predictor of cardiovascular mortality in elderly patients, researchers here reported. And while homocysteine demonstrated prognostic power in 302 octogenarians, Framingham risk score -- a widely validated and widely used measure -- was not predictive, according to Wouter de Ruijter, M.D., of Leiden University Medical Center, and colleagues who reported the findings online in BMJ. Other biomarkers -- C-reactive protein, folic acid, and interleukin 6 -- were not predictive of five-year mortality in these elderly patients either. When the researchers plugged the biomarkers and the Framingham risk score into different models, only "the model based on homocysteine resulted in significant differences between the risk categories (log rank test P=0.002). The high-risk category had a 3.4-fold increased risk of cardiovascular mortality compared with the low-risk category (95% confidence interval 1.4 to 8.1), they wrote. For many years homocysteine has appeared to be a player in the pathogenesis of cardiovascular disease, but therapies designed to lower homocysteine concentration have repeatedly failed to demonstrate a benefit. As the homocysteine star has been in descent, C-reactive protein has been in the ascent, most recently with the release of results of the JUPITER trial. That study found that using a potent statin in apparently healthy patients who had elevated highly sensitive C-reactive protein was associated with a lower rate of myocardial infarction, stroke, and cardiovascular death. The Leiden 85-plus Study followed 215 women and 87 men 85 years and older for five years. Plasma concentrations of homocysteine, C-reactive protein, folic acid, and interleukin 6 were measured at baseline and Framingham risk score assessed based on gender, systolic blood pressure, total cholesterol, HDL, diabetes mellitus, smoking, and electrocardiogram. During follow-up 108 participants died and 32% of those deaths were from cardiovascular causes. When the elderly participants were classified into standard Framingham high-, moderate-, and low-risk rankings, a high-risk score did not predict cardiovascular mortality. When the Framingham risk factors were entered as separate variables in a Cox proportional hazard model specifically calibrated to predict risk in very old people, "three new risk categories were obtained, but these categories also did not predict the observed five-year cardiovascular mortality," they wrote. When the Framingham risk score and the homocysteine risk model "were compared by receiver operating characteristic curves, the area under the curve for the Framingham risk score was 0.53 (95% confidence interval 0.42 to 0.63) and that for the homocysteine-based model was 0.65 (0.55 to 0.75)." And when Framingham models and models using other biomarkers were compared, the areas under the curve were the same. But adding homocysteine values into a model using Framingham risk scores, did not increase the predictive power. Dr. de Ruijer and colleagues cautioned that their study has a number of limitations including the limited number of biomarkers selected for inclusion, the lack of coronary artery calcium measurements, and the small sample size. They concluded that the findings were preliminary and therefore require confirmation in a separate cohort. But if confirmed, the findings "could eventually lead to a revision of current guidelines and corresponding indicators of quality of care." The study was funded by an unrestricted grant from the Dutch Ministry of Health, Welfare, and Sports. Dr. de Ruijer reported no conflicts.

Primary source: BMJ Source reference: de Ruijter W, et al "Use of Framingham risk score and new biomarkers to predict cardiovascular mortality in older people: population based observational cohort study" BMJ 2008: 337: a3083.