Tomohiro Saito1, Seiichiro Nanri, Ikuo Saito. 1. Division of Epidemiology and Department of Health Policy, National Institute for Child Health and Development, Keio University, Tokyo, Japan. tmsaito@nch.go.jp
Abstract
BACKGROUND: Family history of lifestyle-related diseases is an important risk factor and is widely used in epidemiologic studies. To justify its use the validity of the family history needs to be assessed. The aim of the present study was therefore to examine reliability, that is, consistency in repeated measurements, of the family history obtained on questionnaire. METHODS: A questionnaire survey on the family history of lifestyle-related diseases was conducted for 438 high school students about students' parents, grandparents, uncles and aunts, twice with a 1 year interval. The questionnaire was filled in by their parents. Discordance between the two questionnaires in age, age at death and a positive history of lifestyle-related diseases expressed by age at onset by decade age among the family members was examined. Kappa was used as a measure of agreement. RESULTS: The kappas for relatives were high, mostly >0.7 for hypertension and >0.5 for myocardial infarction, angina pectoris, stroke, diabetes mellitus and hypercholesterolemia. The discordance tended to be higher among grandparents than parents, uncles and aunts. The discordance with regard to relatives was mostly <4% for angina pectoris, myocardial infarction, stroke and diabetes mellitus; for hypertension and hypercholesterolemia the discordance was <6%. CONCLUSIONS: The questionnaire provides a reliable and valid method of collecting family history on relatives of high school students. Thus, family history obtained through questionnaires can be used justifiably as a risk factor for lifestyle-related diseases among children and youths.
BACKGROUND: Family history of lifestyle-related diseases is an important risk factor and is widely used in epidemiologic studies. To justify its use the validity of the family history needs to be assessed. The aim of the present study was therefore to examine reliability, that is, consistency in repeated measurements, of the family history obtained on questionnaire. METHODS: A questionnaire survey on the family history of lifestyle-related diseases was conducted for 438 high school students about students' parents, grandparents, uncles and aunts, twice with a 1 year interval. The questionnaire was filled in by their parents. Discordance between the two questionnaires in age, age at death and a positive history of lifestyle-related diseases expressed by age at onset by decade age among the family members was examined. Kappa was used as a measure of agreement. RESULTS: The kappas for relatives were high, mostly >0.7 for hypertension and >0.5 for myocardial infarction, angina pectoris, stroke, diabetes mellitus and hypercholesterolemia. The discordance tended to be higher among grandparents than parents, uncles and aunts. The discordance with regard to relatives was mostly <4% for angina pectoris, myocardial infarction, stroke and diabetes mellitus; for hypertension and hypercholesterolemia the discordance was <6%. CONCLUSIONS: The questionnaire provides a reliable and valid method of collecting family history on relatives of high school students. Thus, family history obtained through questionnaires can be used justifiably as a risk factor for lifestyle-related diseases among children and youths.