Table 3.

Separate models of QOL for patients and partners

Effect	Patients (df = 186)*			Partners (df = 195)*
	Estimate	SE	Pr > |t|	Estimate	SE	Pr > |t|
Intercept	79.89	2.55	<.0001	72.71	3.49	<.0001
Age	0.18	0.05	0.0008	0.21	0.07	0.0019
Education	−0.31	0.12	0.0119	−0.47	0.25	0.0627
Family income	0.80	0.48	0.0971	1.68	0.65	0.0108
Uncertainty2	−0.21	0.03	<.0001	−0.15	0.03	<.0001
Social support1	0.21	0.03	<.0001	0.15	0.03	<.0001
Communication1	3.65	0.57	<.0001	4.64	0.68	<.0001
Time since diagnosis	−0.006	0.04	0.9064	0.002	0.04	0.9607
Time since diagnosis_sq	−0.0004	0.0004	0.3951	0.00005	0.0004	0.9020
Phase of illness (referent: advanced)
Localized	1.81	1.43	0.2083	3.42	1.87	0.0683
Recurrent	5.59	2.22	0.0128	4.80	2.62	0.0684
General symptoms2	−1.02	0.12	<.0001	−1.33	0.11	<.0001
Pca3 symptoms_urine1	0.11	0.40	0.7775	0.44	0.40	0.2742
Pca3 symptoms_bowel1	−0.05	0.35	0.8836	−0.61	0.41	0.1416
Pca3 symptoms_sexual1	0.52	0.19	0.0067	−0.31	0.34	0.3677
Pca3 symptoms_hormonal1	1.98	0.44	<.0001	0.48	0.41	0.2457

Bold numbers are significant findings

Separate models of QOL for patients and partners were fitted as a function of measurements of their own predictors

^*Sample sizes of patients and partners differ due to different attrition rates, 9 partners of patients who dropped out of the study remained in the study

¹Higher scores indicate more positive results: more support from others, more communication, and less prostate cancer-specific symptoms

²Higher scores indicate more negative results: more uncertainty about the illness and more symptom distress such as pain, fatigue, and insomnia

³Prostate cancer-specific