Total cholesterol and mortality in China, Poland, Russia, and the US.

Purpose: To examine the relationships of total and cause-specific mortality to serum cholesterol in four diverse populations.
Methods: Chinese, Polish, Russian, and US population-based samples were studied. The relationship between cholesterol levels and mortality was assessed by Cox proportional hazard regression with restricted piecewise cubic splines.
Results: The cholesterol and total mortality relationship was statistically significantly J-shaped for all men combined. In country-specific relationships, cholesterol was significantly, linearly, and positively related to total mortality in Russian and US men. For women, the relationship was non-linear, but not statistically significant, and became statistically significant upon adjustment for other risk factors. For Polish women, a statistically significant inverse relationship existed. CHD mortality and cardiovascular disease (CVD) mortality increased linearly with cholesterol in Polish, Russian, and US men and the aggregate of men, but there was no relationship for women. Cancer mortality was not related to cholesterol except for the Polish cohort and Russian women, where there was an inverse relationship.
Conclusions: Serum cholesterol was a strong, consistent predictor of CHD and CVD mortality in Polish, Russian, and US men despite their social diversity. In contrast to CHD mortality, the relation of cholesterol to total mortality and non-CVD mortality varied by country and gender.