Social ecological factors associated with physical activity and screen time amongst mothers from disadvantaged neighbourhoods over three years

Background Mothers from socioeconomically disadvantaged neighbourhoods are at elevated risk of physical inactivity and high levels of screen time. Yet, little is known regarding the social ecological factors that are longitudinally associated with physical activity and screen time in this target group, and whether the age of their children impacts these relationships. This study aimed to longitudinally examine the social ecological factors associated with physical activity and screen time amongst mothers living in socioeconomically disadvantaged neighbourhoods, and whether these differed according to their child's age. Methods Data were from 895 mothers living in socioeconomically disadvantaged neighbourhoods (mean age 36.7 years) at baseline and three-year follow-up. Mothers self-reported weekly discretionary physical activity (leisure-time, LTPA; transport-related, TRPA) and screen time durations. Linear regression models assessed associations between five intrapersonal, three social and five physical environmental factors and LTPA, TRPA and screen time, adjusting for confounding factors, clustering by neighbourhood and baseline variables. Interaction analysis was conducted for age of children (younger and older children (n = 442) and mothers with older children (aged 5-12 years) only (n = 453). Results In adjusted models, all intrapersonal factors (self-efficacy, enjoyment, outcome expectations, behavioural intentions and behavioural skill), social support from friends, neighbourhood cohesion and number of televisions were longitudinally associated with LTPA amongst all mothers. Interaction models showed that findings were generally consistent across groups (i.e., those with both younger and older children compared to those with older children only), with three exceptions. Physical activity enjoyment and social support from family were associated with LTPA only among mothers with older children. Neighbourhood cohesion was associated with screen time only amongst mothers with both younger and older children. No associations were detected for TRPA. Conclusion Intrapersonal, social and physical environmental factors were longitudinally associated with mother's LTPA, whilst neighbourhood cohesion was longitudinally associated with screen time behaviours amongst mothers. Interventions aimed at increasing LTPA amongst mothers (particularly those from socioeconomically disadvantaged neighbourhoods) may need to target all domains of the social ecological model and may require some tailoring according to the age of children. Further work is needed to identify longitudinal associations with screen time and TRPA in this population group.