Distance to parks and non-residential destinations influences physical activity of older people, but crime doesn’t: a cross-sectional study in a southern European city

Table 3 Association between time spent in leisure-time physical activity of active participants and neighborhood characteristics. Association between daily minutes spent in leisure-time physical activity (log-transformed) of active participants and neighborhood characteristics, stratified by sex (Porto, 2005–2008)

	Model 1^a				Model 2^b
	Women		Men		Women		Men
	Coefficient	p-value	Coefficient	p-value	Coefficient	p-value	Coefficient	p-value
Distance to the nearest park (hm)	−0.0275	0.017	−0.0063	0.573	−0.0262	0.029
Distance to the nearest sport space (hm)	−0.0297	0.068	0.0471	0.017			0.0462	0.032
Distance to the nearest non-residential destination (hm)	−0.0750	0.014	0.0125	0.680	−0.0735	0.019
Distance to the sea/riverside (hm)	−0.0031	0.669	−0.0011	0.852
Intersection density^c (nodes/ha)	−0.0073	0.549	−0.0070	0.471
Bus/metropolitan stops (no.)^c	0.0093	0.828	0.0089	0.823
Land gradient (%)^c	−0.0254	0.221	−0.0102	0.628
Population density (inhab./ha)^c	0.0006	0.495	−0.0005	0.596
Neighborhood SES^d
1 – least deprived	Ref		Ref
2 – medium deprived	−0.0394	0.832	0.0242	0.879
3 – most deprived	−0.1358	0.612	0.1921	0.393
Neighborhood crime (crimes/1000 inhab.)
Incivilities	−0.0008	0.996	−0.0008	0.995
Crime without violence	−0.0015	0.615	0.0029	0.423
Crime with violence	−0.0081	0.593	0.0012	0.991
Traffic crime	0.0045	0.422	0.0020	0.669
Overall crime	−0.0156	0.689	0.00038	0.883

^aUnivariable regression
^bMultivariable regression adjusted for age, educational attainment, marital status, retirement status, residence in Porto for 20 years or more, comorbidities, BMI and smoking habits
^cWithin 200 m circular buffer
^dSES neighborhood socioeconomic status