© 1973 Oxford University Press
other |
Serum Cholesterol in a National Sample of U. S. Adults: A Study from Prevalence Data of its Relationship to Physique, Blood Pressure, Blood Glucose and Other Variables
1 Department of Epidemiology and Public Health, Yale University School of Medicine New Haven, Connecticut 06510
2 National Center for Health Statistics, U. S. Public Health Service Rockville, Maryland 20852
Requests for reprints may be addressed to Dr. R. M. Acheson.
Present address: London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, England.
Present address: Office of the Assistant Commissioner for Planning and Evaluation, Food and Drug Administration, U. S. Public Health Service, Rockville, Maryland 20852.
The data were collected during the U. S. National Health Examination Survey between 1960–1962 from 5,545 of the probability sample of 7,710 adults of both sexes; 4,854 respondents were white and 691 black. A stepwise multiple regression analysis was used to examine, on a sex and colour specific basis, the inter-relationships between serum cholesterol on the one hand (which was treated as the dependent variable) and hematocrit, blood glucose, height, weight, body bulk, blood pressure, age, family income, occupation, population density at place of residence and degree of urbanization at place of residence on the other (which were treated as independent variables).
It was found that in general the ability of the independent variables to predict serum cholesterol was poor, though it was better in females than in males. R2 values were as follows: white males 0.198; black males 0.169; white females 0.295; black females 0.230. The independent variables which contributed consistently and significantly to the predictions in all four groups were age and body bulk; in whites of both sexes hematocrit also contributed significantly. It is emphasized that inter-relationships of this kind cannot necessarily be taken as cause and effect.