Is a perceived supportive physical environment important for self-reported leisure time physical activity among socioeconomically disadvantaged women with poor psychosocial characteristics? An observational study

Table 4 OR ^a for LTPA according to environment score, stratified by psychosocial score, among women in the READI study

Environment score by psychosocial score	LTPA
	≥1 min/wk (vs. 0 min/wk)	≥120 mins/wk (vs. <120 mins/wk)	≥280 mins/wk (vs. <280 mins/wk)
	OR (95% CI)	OR (95% CI)	OR (95% CI)
Low psychosocial score
Low environment score (n = 402)	1.0 (reference)	1.0 (reference)	1.0 (reference)
Med environment score (n = 350)	1.21 (0.88, 1.65)	1.76 (1.22, 2.54)*	1.34 (0.75, 2.37)
High environment score (n = 323)	1.19 (0.87, 1.65)	1.58 (1.07, 2.34)*	1.38 (0.80, 2.40)
Med psychosocial score
Low environment score (n = 316)	1.0 (reference)	1.0 (reference)	1.0 (reference)
Med environment score (n = 342)	1.14 (0.80, 1.64)	0.92 (0.70, 1.22)	0.78 (0.54, 1.14)
High environment score (n = 430)	0.94 (0.66, 1.32)	0.91 (0.67, 1.23)	0.79 (0.54, 1.16)
High psychosocial score
Low environment score (n = 291)	1.0 (reference)	1.0 (reference)	1.0 (reference)
Med environment score (n = 420)	1.29 (0.76, 2.20)	1.08 (0.76, 1.54)	1.11 (0.80, 1.54)
High environment score (n = 742)	1.24 (0.73, 2.10)	1.14 (0.81, 1.60)	1.10 (0.81, 1.47)

LTPA leisure time physical activity, READI resilience for eating and inactivity despite inequality; * p < 0.05.
^a Odds ratios and 95% confidence intervals from generalised ordered logistic regression model adjusted for age, number of children, country of birth, employment status, marital status, smoking status, and injury/illness/disability; robust standard errors adjusting for clustering by neighbourhood (the unit of recruitment). Generalised ordered logistic regression is appropriate for the ordered but not evenly spaced outcome data, and compares each increasing category to all data below the cut-point for that category.

ISSN: 1471-2458