Fig. 2
Forest plot showing odds ratios for the prevalence of walking for leisure amongst women in comparison to men among (a) all participants and (b) by age group. Clear markers indicate bivariate analysis; filled markers indicate multivariate analysis, and footnotes detail control variables included in each model. Marker size is proportional to the sample size of the study. * Results plotted here from one of several models in original study; reported ORs (95% CI) ranged from 1.47 to 1.49 (1.16 to 1.19, 1.81 to 1.85). 1 Age, income, marital status, chronic conditions, obesity, neighbourhood income, land use. 2 Age, ethnicity, household composition, education, household income and wealth, neighborhood population density. 3 Age, income, working status, self-rated health status, smoking status, leisure screen time, city density, population. 4 Age, family income, marital status, neighborhood SES and walkability. 5 Age, gender, education, country of origin. 6 Age, education, employment, household economic status, presence of children in household, self-rated health, social environment, aesthetics, and presence of pain due to orthopedic disorders. 7 Age, suburb SES, other recreational MVPA, education, retirement status, functional limitations, urban vs rural, park proximity and quality