Table 2.

Predictors of cognitive impairment in multivariable ordinal logistic regression models stratified by ethnicity^*

	MA		NHW
	Minimally adjusted model	Fully adjusted model	Minimally adjusted model	Fully adjusted model
	cOR (95% CI)	cOR (95% CI)	cOR (95% CI)	cOR (95% CI)
History of stroke	1.74 (1.15–2.63)	1.58 (1.04–2.41)	4.09 (2.57–6.52)	3.81 (2.37–6.12)
Age (reference = 65–69)
70–74	1.01 (0.73–1.41)	0.98 (0.70–1.36)	1.09 (0.71–1.67)	1.04 (0.68–1.60)
75–79	2.27 (1.55–3.34)	2.25 (1.53–3.31)	1.57 (1.02–2.41)	1.49 (0.97–2.30)
80–84	3.24 (2.04–5.15)	3.19 (2.00–5.07)	3.67 (2.28–5.91)	3.59 (2.23–5.78)
85+	11.50 (6.50–20.36)	11.76 (6.64–20.85)	9.81 (5.33–14.89)	8.70 (5.18–14.62)
Female sex	0.78 (0.60–1.03)	0.83 (0.63–1.09)	0.84 (0.62–1.14)	0.89 (0.65–1.19)
Education (reference = college degree)
Less than high school	11.08 (7.26–16.91)	10.69 (7.00–16.35)	7.05 (3.45–14.40)	7.10 (3.44–14.67)
High school diploma	2.60 (1.79–3.76)	2.55 (1.75–3.70)	3.21 (2.33–4.41)	3.12 (2.26–4.29)
History of diabetes	-	1.31 (0.99–1.72)	-	1.35 (0.94–1.96)
History of heart disease	-	1.38 (0.99–1.92)	-	1.12 (0.79–1.60)

^*Abbreviations: MA, Mexican American; NHW, non-Hispanic White; cOR, cumulative odds ratio; CI, confidence interval. The minimally adjusted models included the following variables: stroke history, age, sex, and education. The fully adjusted models included the following variables: stroke history, age, sex, education, diabetes, and heart disease.