Table 2.
Associations between density of physical activity facilities and adiposity outcomes: multilevel regression results
| Model 0 | Model 1 | Model 2 | Model 3 | ||
|---|---|---|---|---|---|
| Waist circumference, cm (n=401 917) | |||||
| Number of facilities | |||||
| 0 | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | |
| 1 | 0·15 (−0·02 to 0·32) | −0·15 (−0·30 to −0·01) | −0·13 (−0·26 to 0·00) | −0·19 (−0·32 to −0·06) | |
| 2–3 | 0·08 (−0·20 to 0·35) | −0·43 (−0·70 to −0·17) | −0·29 (−0·51 to −0·08) | −0·40 (−0·62 to −0·17) | |
| 4–5 | −0·14 (−0·55 to 0·27) | −0·80 (−1·19 to −0·42) | −0·51 (−0·82 to −0·19) | −0·65 (−0·96 to −0·33) | |
| ≥6 | −0·67 (−1·25 to −0·09) | −1·51 (−2·04 to −0·98) | −1·03 (−1·45 to −0·62) | −1·22 (−1·64 to −0·80) | |
| Body-mass index, kg/m2 (n=401 435) | |||||
| Number of facilities | |||||
| 0 | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | |
| 1 | 0·04 (−0·04 to 0·12) | −0·08 (−0·15 to 0·00) | −0·06 (−0·13 to 0·01) | −0·07 (−0·14 to 0·00) | |
| 2–3 | −0·03 (−0·15 to 0·10) | −0·22 (−0·34 to −0·09) | −0·16 (−0·26 to −0·05) | −0·18 (−0·28 to −0·08) | |
| 4–5 | −0·18 (−0·36 to 0·00) | −0·43 (−0·59 to −0·26) | −0·30 (−0·43 to −0·17) | −0·33 (−0·46 to −0·20) | |
| ≥6 | −0·42 (−0·67 to −0·17) | −0·73 (−0·96 to −0·50) | −0·52 (−0·69 to −0·34) | −0·57 (−0·74 to −0·39) | |
| Body fat, % (n=395 640) | |||||
| Number of facilities | |||||
| 0 | 0 (ref) | 0 (ref) | 0 (ref) | 0 (ref) | |
| 1 | 0·03 (−0·07 to 0·14) | −0·11 (−0·20 to −0·01) | −0·08 (−0·16 to 0·00) | −0·11 (−0·20 to −0·03) | |
| 2–3 | −0·07 (−0·24 to 0·11) | −0·30 (−0·46 to −0·13) | −0·21 (−0·34 to −0·07) | −0·27 (−0·40 to −0·13) | |
| 4–5 | −0·28 (−0·53 to −0·02) | −0·58 (−0·81 to −0·35) | −0·40 (−0·58 to −0·22) | −0·48 (−0·67 to −0·29) | |
| ≥6 | −0·63 (−0·96 to −0·30) | −1·00 (−1·29 to −0·70) | −0·71 (−0·92 to −0·49) | −0·81 (−1·03 to −0·59) | |
Density is defined as number of physical activity facilities in a 1000 m street-network buffer. Data are mean difference (95% CI). Model 0: adjusted for age and sex. Model 1: model 0 plus adjustment for ethnicity, urban or non-urban status, and area deprivation. Model 2: model 1 plus adjustment for individual socioeconomic characteristics (income, education, and employment status). Model 3: model 2 plus adjustment for residential density and distance to nearest fast-food outlet.