Table 3.

Multiple linear regression analysis with several independent baseline variables of biopsychosocial health predicting PCS 3 years later, separated by gender

Baseline variables	Women				Men
	Ba	SEb	βc	p-value	Ba	SEb	βc	p-value
Age (years)
55–64	1.26	.49	.06	.011	2.25	.55	.10	< .001
65–74 (ref.d)
75–84	− 2.02	.58	− .08	.001	− 1.33	.59	− .05	.024
Education (years)
0–8	− .42	2.13	< -.01	.834	− 2.06	1.95	− .02	.291
9–10 (ref.)
≥ 11	− 1.15	1.00	− .02	.249	− .73	.66	− .03	.267
11–12	.41	.89	.01	.643	.76	.73	.02	.300
Marital status
Single	− .34	1.15	− .01	.765	− 1.16	1.26	− .02	.356
Married (ref.)
Divorced/widowed	− .16	.46	− .01	.731	.25	.75	.01	.743
Physical health
No chronic disease	1.07	.53	.05	.044	2.12	.55	.10	< .001
1 chronic disease	1.14	.54	.05	.037	.48	.58	.02	.404
≥ 2 chronic diseases (ref.)
Mental health
Somatic symptomst0	− 3.38	.89	− .09	< .001	− 1.81	1.54	− .03	.239
Depression symptomst0	.91	.88	.03	.304	− 1.34	1.25	− .03	.289
GAD symptomst0	.74	1.11	.02	.503	1.61	1.86	.02	.388
Social health
Lonelinesscontinuously	− .33	.13	− .06	.008	.22	.17	.03	.190
Social networklarge	− .98	.67	− .03	.146	− 1.68	.65	− .06	.010
HRQOL
MCSt0	.06	.03	.07	.010	.10	.03	.08	.001
PCSt0	.63	.02	.63	.010	.65	.03	.62	< .001

Significant associations are printed in bold, variables with gender differences are framed

^aNon-standardized regression coefficient

^bStandard error

^cStandardized regression coefficient

^dReferent