Movement patterns of youth and their parents: cross-sectional and longitudinal associations.

BACKGROUND: Establishing early physical activity (PA) habits is vital for long-term health, with parents considered as key influencers on children's PA. Yet, most previous parent-offspring dyads examining PA associations were cross-sectional, rarely used device-based measures, and often overlooked movement composition. The aim of this study was to determine whether mother's and father's waking movement composition is cross-sectionally or longitudinally associated with those of their children. METHODS: The SOPHYA cohort recruited families from a nation-wide population-based random sample stratified by child's sex, birth year, and language. All youth aged 6-16 years and their parents officially residing in Switzerland, were eligible. Baseline and follow-up assessment occurred in 2013-2015 and 2019-2020, respectively. Questionnaire information and accelerometer measurements were collected remotely. The main predictor was parental movement composition at baseline. The associations between parental and child movement compositions were examined using Dirichlet regression models, adjusting for child's age and sex, parental education, and language region. The endpoints were children's movement composition at baseline (cross-sectional) and follow-up (longitudinal), respectively. RESULTS: Baseline assessment provided accelerometer and self-reported covariate data for the same measurement week in 686 mother-child and 373 father-child pairs. Follow-up assessment provided accelerometer data for 263 children with maternal and 149 with paternal baseline data. Cross-sectionally, replacing parental sedentary behaviour (SB) with moderate-to-vigorous activity (MVPA) (mothers: 0.10, p < 0.001; fathers: 0.09, p = 0.002) or replacing SB with light physical activity (LPA) (mothers: 0.13; < 0.001; fathers: 0.09; p < 0.005) was associated with similar, but smaller shifts in children. Longitudinally, replacing parental SB with LPA was associated with similar, but smaller shifts in children five years later (mothers: coefficient: 0.12, p = 0.021; fathers: coefficient: 0.10, p = 0.108). The cross-sectional change in children's LPA/SB ratio predicted from a parent's 20% decrease in SB and corresponding 20% increase in LPA was about 18-fold smaller than the observed maternal shift and about 29-fold smaller than the paternal shift from which it was predicted. CONCLUSIONS: Dirichlet regression results suggest that parental movement composition may predict children's movement composition, highlighting parental movement patterns, particularly SB, as potentially effective targets for short- and long-term interventions to increase PA in both parents and children.