The pattern of subcutaneous fat distribution in middle-aged men and the risk of coronary heart disease: the Paris Prospective Study

Abstract

The relationship of subcutaneous fat repartition in a population of 6718 men aged 42-53 employed by the city of Paris with the incidence of coronary heart disease has been studied in the Paris Prospective Study after a follow-up duration of 6.6 years. Fat distribution was described from the measurement of 13 skinfold thicknesses (five on the trunk, four at the triceps level, four at mid-thigh). Principal component analysis of the data disclosed two independent factors which explained most of the total variance of the log measurements: the first one (F1) might be interpreted as a general index of adiposity, the second one (F2) opposed the skinfolds on the trunk to those on the thigh. When men who developed coronary heart disease (CHD) were compared to those who remained free of it, both F1 and F2 differed significantly between the two groups. The combination of F1 and F2 which was the best predictor of CHD in this population (G1) might be interpreted as reflecting trunk adiposity mainly abdominal. Subsequent analysis showed that the combination of F1 and F2 which was statistically independent of G1 (G2) was nearly as correlated with body mass index (BMI) as was G1 (0.49 and 0.59 respectively) but G2 was not a significant predictor of angina pectoris nor myocardial infarction. Moreover mean G2 did not differ between treated diabetics and non-diabetics at entry in the study whereas mean G1 was significantly increased in the former group. Linear correlation coefficients of G2 with systolic blood pressure total cholesterol and triglycerides measured at entry were low (0.08, 0.08, and 0.15 respectively) in comparison with those obtained with G1 (0.26, 0.24 and 0.35 respectively). It is concluded that trunk fat deposit as defined by the G1 index represents the pattern of subcutaneous fat repartition the most tightly associated with CHD complications and their main risk factors in this population of middle-aged men and that it remains a significant CHD predictor after controlling for these factors.