Table 3.

Cross-sectional association between dual sensory impairment and low perceived social support at Year 1 and Year 5 on study.

Low perceived social support	Hearing Impairment Only		Vision Impairment Only		Dual Sensory Impairment
	Relative Risk	95 % Confidence Interval	Relative Risk	95 % Confidence Interval	Relative Risk	95 % Confidence Interval
Year 1
None of the components	REF	REF	REF	REF	REF	REF
At least one component is present	1.10	(0.85, 1.44)	1.41	(1.03, 1.92)	1.25	(0.93, 1.67)
Two components are present	0.76	(0.47, 1.24)	0.86	(0.49, 1.53)	0.87	(0.52, 1.47)
All three components are present	0.38	(0.13, 1.14)	0.14	(0.02, 1.14)	0.52	(0.17, 1.63)
Year 5
None of the components	REF	REF	REF	REF	REF	REF
At least one component is present	1.07	(0.82, 1.39)	1.01	(0.74, 1.39)	1.24	(0.93, 1.66)
Two components are present	0.65	(0.40, 1.06)	0.97	(0.56, 1.67)	1.02	(0.61, 1.70)
All three components are present	0.67	(0.17, 2.62)	2.48	(0.70, 8.82)	0.19	(0.02, 1.78)

Note: Multivariable multinomial logistic models were adjusted by sex, race, postsecondary education, age, body mass index, history of Type II diabetes and hypertension, and marital status (married vs. not). All continuous variables were mean-centered. Low perceived social support is defined by the following: (i) dissatisfaction defined as the lower 10th percentile of the sample distribution on both questions (score < 6), (ii) low social contact defined as one time per week/less than once per week for friends and/or relatives, and (iii) low emotional support defined as needing more emotional support. Bolded estimates and their respective 95 % confidence intervals indicate that the 95 % confidence intervals did not overlap with 1.