Gestational age and device-measured physical activity in adolescents and adults: A linkage of the HUNT Study and Medical Birth Registry.

There is limited knowledge of the relationship between gestational age and physical activity assessed by accelerometers in population-based cohort studies. We examined how gestational age predicts physical activity among adolescents and adults. Data from 12 896 participants (13-52 years) with physical activity assessed by two tri-axial accelerometers for seven days in the population-based Trøndelag Health Study (HUNT Study) were linked to the Medical Birth Registry of Norway. Metabolic equivalent of task (MET) min/day and min/day in total, moderate to vigorous, and light physical activity were calculated. Effects of gestational age (in weeks) and birth weight (in grams) on physical activity were modelled using fractional polynomials and linear regression. Gestational age predicted MET min/day in light physical activity (B = 1.53, 95% CI: 0.02 to 3.07) in the total sample, especially among females (B = 2.14, 95% CI: 0.33 to 3.93), and in total physical activity among adolescents (B = 2.44, 95% CI: 0.16 to 4.88). Gestational age predicted min/day in total physical activity in the total sample (B = 0.99, 95% CI: 0.07 to 1.90), among females (B = 1.25, 95% CI: 0.11 to 2.40) and adolescents (B = 1.06, 95% CI: 0.02 to 2.12). There were no non-linear or linear effects of birth weight on physical activity. Shorter gestation predicted less physical activity in the total sample, among females, and in adolescents. The associations with total physical activity are by and large consistent with previous observations that individuals born preterm with very low birth weight have lower rates of physical activity.