Background: Four common factors-cigarette smoking, risky drinking of alcoholic beverages, physical inactivity, and overweight-contribute substantially to chronic disease prevalence. Methods: We used data from the 2001 National Health Interview Survey to provide an up-to-date picture of multiple risk factor prevalence and clustering in the U.S. population. We conducted a multinomial logit analysis to examine the independent association between each covariate and the dependent ordinal risk factor variable with three levels (none or one risk factor, two risk factors, and three or four risk factors). Results: Seventeen percent of the sample of 29,183 subjects had three or more risk factors. For the entire sample, the mean number of risk factors was 1.68 (95% confidence interval [CI] = 1.66-1.70). Many demographic and health factors were significantly associated with the mean number of risk factors including gender, age, ethnic/racial categories, education, martial status, presence of chronic diseases, level of mental distress, country of birth, and presence and type of health insurance. Using the risk factor score as the ordinal dependent variable, adjusted odds for having a risk score of three or four versus zero or one were as follows: men aged <65, 2.49 (95% CI=2.29-2.72); education attainment of high school graduate or less, 3.24 (95 % CI=2.86-3.67); and individuals with high levels of mental distress, 2.06 (95% CI=1.65-2.58). Conclusions: Our analyses confirm earlier reports of the high prevalence of multiple, clustered behavioral risk factors and underline the challenge this presents for primary care and public health systems. (C) 2004 American journal of Preventive Medicine.
