Environmental and psychosocial contexts of device-measured physical activity and sedentary behaviour in semi-urban Nepal: a multilevel analysis

Insufficient physical activity and prolonged sedentary behaviour are growing public health concerns in low- and middle-income countries (LMICs). However, evidence on their environmental and psychosocial determinants remains limited, especially in South Asia. Most existing studies rely on self-reported measures and focus on higher-income settings, limiting their relevance to LMIC populations. This study investigated associations between recreational/built and productive/utilitarian environmental opportunities and device-measured physical activity and sedentary behaviour among adults in semi-urban Nepal, adjusting for selected psychosocial factors. We conducted a cross-sectional analysis of 421 adults who participated in the Community-Based Promotion of Physical Activity (COBIN-PA) trial in Pokhara Metropolitan City, Nepal. We assessed physical activity and sedentary behaviour using wrist-worn accelerometers (Axivity AX3) worn continuously for seven days. Participants reported environmental characteristics reflecting recreational or built opportunities and productive or utilitarian opportunities, and selected psychosocial factors were assessed using structured questionnaires. Multilevel linear mixed-effects models with random intercepts for clusters were used to estimate associations with daily moderate-to-vigorous physical activity (MVPA) and sedentary time. Interaction terms were assessed to determine whether perceived behavioural control modified the associations with the environment. Productive/utilitarian environmental opportunity was associated with higher MVPA in the environment-only model, but these associations were attenuated in the combined model. Perceived behavioural control remained positively associated with MVPA (β = 34.35, 95% CI: 6.98 to 61.72), whereas intention showed no clear association. For sedentary time, an interaction was observed between the highest productive/utilitarian environment tertile and perceived behavioural control (β = −35.26, 95% CI: −64.92 to − 5.61). In semi-urban Nepal, productive and utilitarian environmental opportunities showed stronger associations with device-measured MVPA than recreational or built features, although these associations were attenuated after accounting for psychosocial factors. Perceived behavioural control remained positively associated with MVPA. Sedentary time showed weaker, less consistent environmental associations. These results suggest that in rapidly urbanising LMICs, physical activity promotion should consider both everyday movement opportunities and individual-level behavioural factors.