Validity of cerebrovascular mortality rates.

Cerebrovascular disease (CVD) mortality rates are being increasingly used to establish a relationship with coronary heart disease risk factors. Cerebrovascular mortality rates suffer from the same general defects as coronary heart disease mortality rates but with greater diagnostic error in regard to subdivisions of CVD or stroke. These vital statistics, whether national or regional, are too inaccurate for scientific use and cannot be regarded as representing true mortality rates for any country. Moreover, clinical CVD is an inappropriate surrogate for severity of atherosclerosis, providing inexact data and allowing indeterminate overlap in epidemiologic studies.