Changes in total cholesterol levels (TC) were studied using data from three epidemiological studies: about 30,000 men and women aged 37-43 were examined between 1974 and 1980 (CB Project), about 80,000 men aged 33-37 between 1981 and 1986 (RIFOH Project) and 42,000 men and women aged 20-59 from 1987 to 1992 (Monitoring Project on CVD Risk Factors). In men a decline in TC of 6.5% was observed between 1974 and 1992. However, the largest decrease took place between 1981 and 1986 in men in a limited age range (33-37 years), and there were indications that this decrease was not generalizable to other age groups. From 1987 to 1992, a decrease of 7% in HDL cholesterol levels (HDL-C) was observed in men, leading to an increase in the non-HDL-C/HDL-C ratio. In women, no changes in TC and HDL-C were observed.Analyses of data from 36,000 men and women aged 20-59 years showed that between ages 30 and 50 about 19-38% of the gender difference in TC was explained by differences in body mass index (BMI) and cigarette smoking between men and women. After age 50, the higher TC in women compared to men was largely due to the effect of the menopause. The TC increase associated with menopause was 0.45 mmol/l in non-smokers and 0.28 mmol/l in smokers. The difference between a low-risk and a high-risk lifestyle was 0.58 mmol/I for TC and 0.38 mmol/l for HDL-C in men, and 0.40 mmol/l for TC and 0.45 mmol/l for HDL-C in women.Twelve year follow-up of 50,000 men and women aged 30-54 (CB Project) showed that the adjusted relative risk for coronary heart disease (CHD) mortality for the highest compared to the lowest cholesterol quintile was 3.0 (95% CI 1.8-5.1) in men and 3.8 (95% Cl 1.1-13.1) in women. It was estimated that a TC reduction of 0.6 mmol/l was associated with a 20% lower CHD mortality. Low TC was not associated with non-cardiovascular mortality. All-cause mortality was positively related to total cholesterol, with a 60% and 46% higher risk in the highest compared to the lowest TC quintile for men and women ...