"Natural histories" of mitral valve prolapse. Influence of patient selection on cardiovascular event rates.

Background: In previous studies the reported incidence of cardiovascular events among mitral valve prolapse patients has differed more than 10 fold. We endeavored to determine the relation between the clinical features and mode of ascertainment of mitral valve prolapse and the resulting event rate.

Methods: Between January 1979 and August 1996, 275 patients (129-47% men, 146-53% women, mean age 43 +/- 19 years), were followed for a mean of 98 months after evaluation in a referral center for valvular heart disease. Comparative data were obtained from a separate, less selected population consisting of 316 patients.

Results: A total of 65 events occurred (2.9/100 patient-years): 46 (2.0/100 patient-years) mitral surgery, 12 cardiac deaths (0.5/100 patient-years), 6 neurologic ischemia (0.26/100 patient-years), and 1 infective endocarditis (0.04/100 patient-years). The overall event rate varied significantly according to demographic, clinical and echocardiographic variables (all p < 0.0001). It was higher among males (odds ratio-OR 2.1), subjects > or = 45 years of age (OR 14.7), those with a holosystolic murmur (OR 25.9), an enlarged left ventricle (OR 13.5) or left atrium (OR 34.9) and those with 3-4+ mitral regurgitation at color Doppler echocardiography (OR 40.0). It was lower in those with an audible mid-systolic click (OR 0.05). These ORs closely resembled those we reported previously in a less selected population. At multivariate analysis, male gender (p = 0.013), severe Doppler mitral regurgitation (p = 0.0048), and left atrial enlargement (p = 0.046) were all independent predictors of events.

Conclusions: In a population of mitral valve prolapse patients, including many with significant mitral regurgitation at baseline, we identified similar predictors of events but an overall event rate nearly 3 times higher than that we previously reported for relatively unselected patients or family members in New York City (1/100 patient-years). Therefore, the impact of patient selection on the prevalence of mitral regurgitation, older age and male gender strongly affects the adversity of the "natural history" of mitral valve prolapse.