CHOLESTEROLEMIA AND TOTAL, CARDIOVASCULAR AND CANCER MORTALITY - A COHORT OF 220,000 PEOPLE

The distribution of cholesterolemia and their relationship to total mo rtality, cardiovascular mortality and cancer mortality were studied in a population of 125,513 men and 96,301 women, aged 16-90. Cholesterol emia in the population increases with age, up to age 50-60 for men and 60-70 for women. Beyond these ages, cholesterolemia decreases very si gnificantly for men and moderately for women. Upon global examination of the entire population, the relationship between cholesterolemia and total mortality is observed as a U-shaped curve. The relationship wit h cancer mortality is slightly negative in men and appears as a U-shap ed curve in women. The relationship,vith cardiovascular mortality in m en is strongly positively correlated when adjusted for body mass index , blood pressure, tobacco consumption, gamma GT, and age. This relatio nship is not significant in women. There is a significant interaction in the relationship between cholesterolemia and mortality in men for w eight, body mass index, vital capacity, maximum expiration volume per second, tobacco consumption, urea, serum albumin, hematocrit, hemoglob in, alkaline phosphatases, gamma GT, red cell volume, and sedimentatio n rate. Age is fundamental variable to take into account. Thus, in men under 50 years of age at the time of inclusion, the relationship betw een cholesterolemia and total mortality shows a positive tendency, and the relationship with cardiovascular mortality is strongly positive. In men over 65 years of age, these relationships are negative. The sam e tendencies exist for cancer mortality. The reasons for these negativ e relationships may be related to the decreased cholesterol levels in subjects with malnutrition or diseases (essentially hepatic diseases); this is especially true in older subjects. These data push for additi onal in-depth analysis of these relationships and interactions, accord ing to age categories and detailed causes of mortality. They also rein force the idea that, in some patients, low cholesterolemia appears to be a marker for predisposition or the result, but not a cause of cance r.