Table 3

Univariate, partially adjusted and fully adjusted models representing the mean differences in accelerometer-assessed steps/day across quartiles of neighbourhood walkability (n=2949)*,†,‡

	Quartile 2			Quartile 3			Quartile 4
	Steps/day difference	95% CI		Steps/day difference	95% CI		Steps/day difference	95% CI
	Steps/day difference	Lower bound	Upper bound	Steps/day difference	Lower bound	Upper bound	Steps/day difference	Lower bound	Upper bound	R²
GIS-derived walkability
Model 1	−357	−744	30	−389	−776	−2	−471	−858	−84	0.0023
Model 2	−412	−787	−37	−538	−913	−162	−744	−1121	−367	0.0681
Model 3	−395	−768	−21	−448	−823	−73	−530	−916	−144	0.0865
Model 4	−397	−766	−28	−343	−717	31	−234	−630	163	0.1093
Walk Score
Model 1	−322	−713	70	−582	−975	−189	−485	−870	−99	0.0033
Model 2	−393	−772	−14	−757	−1137	−376	−772	−1147	−397	0.0698
Model 3	−418	−795	−40	−623	−1005	−242	−555	−941	−169	0.0875
Model 4	−390	−763	−18	−538	−917	−158	−232	−631	167	0.1104

*Quartile 1 (least walkable) served as the reference; GIS-derived walkability index quartiles: <−1.5, ≥1.5, <−0.3, ≥−0.3, <1.1, ≥1.1; Walk Score quartiles: <22, ≥22, <48, ≥48, <68, ≥68.
†Model 1: unadjusted; model 2: adjusted for age, sex and body mass index; model 3: adjusted for age, sex, body mass index, married/common law, income, children, immigrant and mood disorder; model 4: adjusted for age, sex, body mass index, married/common law, income, children, immigrant, mood disorder, perceived health, ever smoker and season.
‡Rural location was not included in the final multivariate models as it was correlated with both GIS-derived walkability and the Walk Score.
GIS, Geographic Information System.