Statistical analyses
Based on theirMVPAlevel at 15 weeks of gestation, women were
divided into a group with lowMVPA(n12) and highMVPA(n
12)usingthemedianof182minutesperweekasacutpoint.Fortesting
differences between these 2 groups, we performed unadjusted Student
t tests. The relationship between time spent in MVPA at 15 weeks of
gestationwithmetabolicoutcomesat24and32weeksofgestationwas
assessed in linear regression models (model 1). All models were controlled
for objectively measured maternal BMI at week 15, age, and
sedentary behavior. In model 2, effects of changes in MVPA from 15
weeks of gestation to 24 or 32 weeks were assessed by adding the in
MVPA (in h) to model 1. The s for (change in) MVPA indicate the
change inoutcomeperhourperweekincrease ofMVPA.Associations
were considered statistically significant when