PURPOSE: To investigate environmental determinants of change to sedentary behaviour (SB), light-intensity physical activity (LPA) and moderate-to-vigorous intensity physical activity (MVPA) in those at risk of diabetes. METHODS: Cohort of 808 adults recruited from family practice (age = 63 years, BMI = 32 kg/m2, women = 36%) followed up annually over 3 years. Participants were included based on being above the 90th percentile of the Leicester Diabetes Risk Score, indicating a high risk. SB, LPA and MVPA were measured by accelerometer (Actigraph GT3X) using the Freedson cut-points with at least 4 valid days of data. Geographical and air pollution mapping were used to link environmental data to participant’s home address. Data were analysed using a repeated measures design assessing determinants of change to SB, LPA and MVPA. Results are presented as mean (SD) or mean (95% CI). RESULTS: At baseline, accelerometers were worn for an average of 649 (83) minutes/day, during which an average of 27 (25) minutes/day, 288 (77) minutes/day and 543 (100) minutes/day were spent in MVPA, LPA and SB respectively. Across the study period, wear time did not change, whilst SB increased by 9 (7, 10) minutes /day per year, matched by a decrease in LPA and MVPA of 7 (5, 8) and 2 (1, 2) minutes/day per year respectively. Less green space, higher road connectivity (busier traffic), greater air pollution and a higher density of fast food restaurants were all determinants of increasing SB and decreasing LPA (See Table 1), but none were associated with change to MVPA. Environmental factors were stronger determinants of change to SB and LPA than an overall index of social deprivation. CONCLUSION: In this high risk population, a less healthy environment predicted greater annual increases in SB and decreases in LPA, but not MVPA. Policies and interventions aimed at improving the liveable environment may therefore act to slow the trajectory of increasing SB and decreasing LPA in populations at risk of diabetes.Table 1: Association between environmental factors and annual change to SB, LPA and MVPA.Data as mean (95% CI). Adjusted for age, sex and ethnicity. Bold indicates significance (p < 0.05).
Read full abstract