Accelerometer compared with questionnaire measures of physical activity in relation to body size and composition: a large cross-sectional analysis of UK Biobank

OBJECTIVES: Previous studies of the association between physical activity and adiposity are largely based on physical activity and body mass index (BMI) from questionnaires, which are prone to inaccurate and biased reporting. We assessed the associations of accelerometer-measured and questionnaire-measured physical activity with BMI, waist circumference and body fat per cent measured by bioelectrical impedance and dual-energy X-ray absorptiometry (DXA). DESIGN: Cross-sectional analysis of UK Biobank participants. SETTING: UK Biobank assessment centres. PARTICIPANTS: 78 947 UK Biobank participants (35 955 men and 42 992 women) aged 40–70 at recruitment, who had physical activity measured by both questionnaire and accelerometer. MAIN OUTCOME MEASURES: BMI, waist circumference and body fat per cent measured by bioelectrical impedance. RESULTS: Greater physical activity was associated with lower adiposity. Women in the top 10th of accelerometer-measured physical activity had a 4.8 (95% CI 4.6 to 5.0) kg/m(2) lower BMI, 8.1% (95% CI 7.8% to 8.3%) lower body fat per cent and 11.9 (95% CI 11.4 to 12.4) cm lower waist circumference. Women in the top 10th of questionnaire-measured physical activity had a 2.5 (95% CI 2.3 to 2.7) kg/m(2) lower BMI, 4.3% (95% CI 4.0% to 4.5%) lower body fat per cent and 6.4 (95% CI 5.9 to 6.9) cm lower waist circumference, compared with women in the bottom 10th. The patterns were similar in men and also similar to body fat per cent measured by DXA compared with impedance. CONCLUSION: Our findings of approximately twofold stronger associations between physical activity and adiposity with objectively measured than with self-reported physical activity emphasise the need to incorporate objective measures in future studies.