Phenotyping of type 2 diabetes mellitus by 'genetic field work': are data sufficient that are based only on a questionnaire?

摘要

BACKGROUND AND OBJECTIVE: Genetic studies of complex diseases such as diabetes mellitus (DM) require the recruitment of patients and acquisition of patient data in a time- and cost-effective manner. An effective method to do so may be genetic field work (GFW) that recruits the relatives of index cases, these relatives then replying to appropriate questionnaires. This study investigated the validity of GFW data compared with patient interviews by a physician who had examined those persons. METHODS: 122 relatives were identified through GFW among a cohort of 35 families with DM. Questionnaires with self-reported information on past medical history, cardiovascular risk factors and life style were obtained. In a second step these famiy members were interviewed and examined by a physician in an outpatient setting. RESULTS: Comparison of the two data sets of the 122 individuals yielded clear differences which favored interviews by a physician. The number of persons declaring themselves as having DM was significantly lower by GFW than direct examination (18 vs. 27). Diabetic nephropathy was reported by four individuals by GFW, while examination resulted in the identification of 16 cases. Body weight was also clearly too low when self-reported, so that significantly lower values were recorded for BMI in the questionnaire. Hypertension was reported by 25 vs. 54 participants. The amount of cigarettes smoked was significant lower in the GFW than in physicians' examination. But the numbers recorded for myocardial infarction, peripheral vascular disease and stroke were not significantly different. CONCLUSION: These data show that information obtained by GFW is less accurate than that collected in interview and examination by physicians. The latter are necessary to gain valid data for accurate phenotyping, while GFW is useful in the initial recruitment of relatives.