Table 4. Association Between Caregiver Strain and Composite Diabetes Self-Care Score Among REGARDS Participants With Diabetes Who Were Caregivers^a.

Caregiver strain	Prevalence ratio (95% CI)
	Unadjusted	Model 1b	Model 2c	Model 3d	Model 4e	Model 5f	Model 6g
No strain	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Some strain	0.92 (0.71-1.20)	1.02 (0.78-1.33)	1.07 (0.82-1.38)	1.07 (0.83-1.39)	1.08 (0.84-1.40)	1.05 (0.80-1.37)	1.05 (0.80-1.37)
High strain	0.65 (0.44-0.98)	0.73 (0.48-1.10)	0.79 (0.53-1.17)	0.81 (0.55-1.21)	0.87 (0.58-1.29)	0.81 (0.53-1.22)	0.82 (0.53-1.27)

Abbreviation: REGARDS, Reasons for Geographic and Racial Differences in Stroke.

^aA sample size of 510 was used only for the composite diabetes self-care score outcome (1 of the 4 outcomes in this study); this was less than 795 (the main sample size) owing to missing data, primarily from the Mediterranean diet (1 of the components of the score). To preserve power for the individual self-care outcomes, a sample size of 795 was used.

^bModel 1 was adjusted for sociodemographic data (age, sex, race, education, income, and geographic region).

^cModel 2 adjusted for the variables in model 1 plus medical conditions (history of heart disease, hypertension, obesity, and dyslipidemia) and physical health.

^dModel 3 adjusted for the variables in model 2 plus social support (marital status, social network).

^eModel 4 adjusted for the variables in model 3 plus caregiver strain data (living with care recipient, relationship to care recipient).

^fModel 5 adjusted for variables in model 4 plus hours spent caregiving.

^gModel 6 adjusted for variables in model 5 plus mental health (12-item Short Form Health Survey, Mental Composite Summary Score), depressive symptoms (the Center for Epidemiological Studies-Depression scale), and perceived stress (4-item Perceived Stress Scale).