Association of lifestyle and socioeconomic factors with duration and intensity of physical activity assessed using 7-day 24 h-accelerometry

Abstract

Background Duration and intensity of physical activity (PA) are independent protective factors for chronic diseases. Knowing factors associated with duration and intensity of PA might help tailor preventive measures aiming to reduce chronic disease risk. However, previous studies mainly relied on self-reported PA, which is prone to bias, and provides only limited information on PA intensity. The aim of the present study was to examine factors associated with objectively assessed duration and intensity of PA by accelerometry. Methods In a multicentric, nationwide pretest of the German National Cohort, PA of 262 men and women aged 20 to 69 years was assessed by 7-day 24 h-accelerometry using hip-worn triaxial accelerometers (ActiGraph GT3X+). Time in overall PA (cut point: ≥ 79 counts per minute, cpm) and proportion of time in overall PA spent in low, moderate, and vigorous activity were determined based on the triaxial vector magnitude cpm (cut points: 79–2690, 2691–6166, and ≥ 6167 cpm, respectively). Further, participants underwent a personal interview and anthropometric measurement following standardized protocols. In cross-sectional analyses, factors potentially associated with overall PA and PA intensity were studied in a multivariable linear regression model including sex, age, body mass index (BMI), waist circumference (residually adjusted for BMI), smoking status, alcohol consumption, university entrance qualification, employment status, net household income, marital status, diabetes, dyslipidemia, and study center. It was also tested for sex differences. Results On average, participants were 417 min/d overall active, with 79%, 19%, and 1% of their active time spent in low, moderate, or vigorous activity, respectively. Regarding overall PA, persons having versus not having a university entrance qualification were 36 min/d (95% confidence interval; 13, 59) less active. Similarly, persons reporting to be unemployed versus full time employed were 66 min/d (35, 98) less active. Regarding PA intensities, per 5-years higher age, persons spent 0.2% (0.1, 0.4) less of their active time in vigorous and 0.6% (0.2, 1.0) more time in low activity. Furthermore, persons reporting to be current or former smokers spent 1.0% (0.4, 1.5) or 0.6% (0.0, 1.1) less time in vigorous activity compared to persons reporting to have never smoked. Finally, there were sex differences in factors associated with PA intensities, and per 1–unit higher waist circumference (residually adjusted for BMI), men spent 0.03% (0.1, 0.5) less of their active time in moderate in favor of low activity [0.4% (0.2, 0.6)], while in women, there were no associations [moderate activity, 0.0% (−0.2, 0.2), sex interaction: P = 0.03; low activity, −0.1% (−0.3, 0.1), P = 0.01]. In contrast, women reporting to be current smokers spent 4.4% (2.2, 6.6) less time in moderate and 5.5% (3.2, 7.9) more time in low activity than women reporting to have never smoked, whereas in men, no associations emerged [moderate activity, 0.6% (−2.2, 3.4), sex interaction: P = 0.02; low activity, 0.0% (−3.1, 0.3), P = 0.02]. Conclusion This data suggests that having a university entrance qualification and unemployment are associated with less time in overall PA, while higher age, higher waist circumference in men, and current smoking in women are related to more time spent in low activity and less time spent in moderate or vigorous activity. Thus, biological, behavioral, and economic factors should be taken into account for planning and analyses of epidemiological studies.