Comparison of two physical activity questionnaires in obese subjects: the NUGENOB study.

PURPOSE Simple instruments are needed to assess habitual physical activity (PA) in obese subjects. In a multicenter European obesity project, we tested whether PA assessments by two questionnaires were correlated and similarly associated to selected obesity-related variables. METHODS A total of 757 obese subjects (75% female; age 37.1 [7.9] yr, BMI 35.5 [4.9] kg.m(-2), mean [SD]) completed the Baecke questionnaire (assessing work, sport, and nonsport leisure activity) and the short last 7-d version of the International Physical Activity Questionnaire (IPAQ; assessing vigorous, moderate-intensity, walking activity, and sitting). We assessed percent body fat (bioimpedance), waist circumference, and fasting plasma concentrations of glucose, insulin, leptin, and FFA. Insulin sensitivity was assessed by the HOMA index for insulin resistance (HOMAIR). RESULTS Using the IPAQ, only about one third of men and women were classified as insufficiently active. Total habitual PA assessments by the Baecke and IPAQ were significantly related (Spearman rho = 0.51 in total sample, P < or = 0.0001, with adjustment for age, gender, and center). Using principal component analysis, we built two uncorrelated indices corresponding to general obesity (determined by high body fat and leptin) and abdominal obesity (determined by high waist circumference and HOMAIR). PA scores from both questionnaires were negatively related to general and abdominal obesity indices, except for abdominal obesity with the IPAQ in men. CONCLUSIONS Total PA assessments by the two questionnaires were found to correlate significantly, and the general pattern of associations of PA with general obesity was similar for the two questionnaires. However, the IPAQ may capture less of the relationships between PA and abdominal obesity than the Baecke, especially in men. Reporting of habitual PA in obese subjects with the IPAQ warrants further evaluation against objective assessment methods.