Table 3 .

OLS regression analysis: predicting days of physical health not good at time 2

	Unstandardized coefficient	Standard error	P value
Intercept	−0.32	1.10	0.77
Days of physical health not good at time 1	0.37	0.03	<0.01
Male	−0.08	0.58	0.90
Race (ref.: non-Hispanic white)
Non-Hispanic black	−0.02	0.65	0.98
Others	−1.39	1.12	0.21
Injury severity (ref.: ambulatory)
Non-ambulatory C1–C4	1.49	0.89	0.10
Non-ambulatory C5–C8	−0.37	0.69	0.59
Non-ambulatory non-cervical	−0.26	0.63	0.68
Age at injury	0.07	0.02	<0.01
Years since injury at time 1	0.07	0.03	0.01
Policies subscale at time 1	−0.05	0.29	0.87
Physical/structural subscale at time 1	0.42	0.20	0.03
Attitudes and support subscale at time 1	−0.09	0.27	0.74
Services and assistance subscale at time 1	1.07	0.31	<0.01
Adjusted R2	0.18