Change in health and social factors in mid-adulthood and corresponding changes in leisure-time physical inactivity in a prospective cohort

Table 2 Associations (RRRs (95% CIs)) between baseline health and social factors (33y) and physical inactivity 33y-to-50y^a

33y factor	Persistent vs. never inactive	Deteriorating vs. never inactive	Improving vs. persistently inactive
Poor self-rated health	2.01 (1.70,2.38)	1.38 (1.16,1.64)	0.77 (0.63,0.94)
Obesity	1.56 (1.30,1.86)	1.40 (1.18,1.67)	0.72 (0.57,0.90)
Depression	1.28 (1.05,1.57)	1.10 (0.89,1.36)	1.09 (0.88,1.35)
Low self-efficacy	1.41 (1.22,1.62)	1.19 (1.04,1.36)	0.86 (0.73,1.01)
Manual SEP	1.41 (1.24,1.61)	1.44 (1.28,1.63)	0.89 (0.76,1.04)
Not married/co-habiting	1.06 (0.91,1.24)	1.15 (0.98,1.35)	0.84 (0.69,1.02)
Parenthood (≥2 children)	1.11 (0.97,1.27)	1.03 (0.90,1.16)	1.15 (0.98,1.35)

Results from two multinomial logistic regression models: comparing (i) persistently vs never inactive and deteriorating to inactivity vs never inactive and (ii) improving from inactivity vs persistent inactivity
Models adjust for gender and for all factors in the table
p for gender interaction (in univariable analysis) ≥ 0.33 for all health and social factors (except p = 0.03, for parenthood)
^aPhysical inactivity N (%) at 33y: 3426 (31.3); at 50y: 2955 (30.4); % inactive 33–50y (averaged over 20 imputed datasets): Never inactive: 51.4; persistently inactive: 13.6; deteriorating: 17.2; improving: 17.9
SEP Socio-economic position

ISSN: 1479-5868