#32 Clustering of chronic disease risk factors in children in low-income multiethnic neighborhoods in montreal

Abstract

PURPOSE: Risk factors (RF) for chronic disease are prevalent in youth, track to adulthood, and are associated with adult morbidity. We studied clustering (i.e. any combination) of 9 behavioral RF in 6841 children aged 9–12 in low-income multiethnic urban neighborhoods. METHODS: Subjects from 16 control schools in a prevention trial, completed questionnaires each spring from 1993-7, collecting data on smoking, unhealthy diet, obesity, sedentary behavior, and physical inactivity; height and weight were measured in a standardized protocol. Family origin (FO) was based on language(s) spoken and country of birth of subjects and parents. We examined the 5 years of data cross-sectionally, removing repeat observations of the same subject. We ranked 13 FO groupings (mean n/grouping = 449 (range 116–1403) according to mean number of RF. We then tested FO as an independent correlate of clustering controlling for age, sex, parental employment, single-parent family status, number of persons/household, and % lifetime in Canada. RESULTS: The mean age of subjects was 11.1 years; 60% of subjects and 80% of parents were born outside Canada. Clustering was high among children of Portuguese and Haitian FO, moderate among those of Canadian, Italian, Eastern European, Chinese, Cambodian, South American, Salvadorian, and other Central American FO, and low among children of Vietnamese, South Asian, and Arabic FO. Control for confounding substantiated univariate findings. When obesity was removed from the list of RF, children of Chinese, Vietnamese, and Cambodian FO were among those at highest risk (odds ratios 1.71–2.21) showing that these children, while unlikely to be overweight, had a high prevalence of other RF. CONCLUSION: Differential clustering of behavioral RF is observable by FO as early as pre-puberty. Inclusion of obesity in the list of RF masked clustering of other RF in children of Asian FO.