Table 2.

Multivariate cox proportional-hazards regression models examining the relation between interarm differences in systolic blood pressure ≥ 10mmHg and the risk of dementia, across the entire sample and stratified by Apolipoprotein ε4 status.

		Whole sample			Apoe ε4−			Apoe ε4+
Outcome	Model	n cases / subjects	HR (95% CI)	P	n cases / subjects	HR (95% CI)	P	n cases / subjects	HR (95% CI)	P
Any dementia
	1	224/2018	1.05 (0.76, 1.45)	0.78	162/1579	0.83 (0.55, 1.24)	0.35	59/416	1.92 (1.09, 3.40)	0.02
	2	218/1979	1.00 (0.72, 1.40)	0.99	158/1554	0.80 (0.53, 1.21)	0.29	58/404	1.81 (1.00, 3.28)	0.05
	3	217/1926	0.97 (0.70, 1.35)	0.86	157/1510	0.75 (0.50, 1.12)	0.16	58/396	1.97 (1.12, 3.47)	0.02
Alzheimer’s disease
	1	184/2018	1.07 (0.75, 1.52)	0.72	129/1579	0.75 (0.47, 1.19)	0.22	52/416	2.32 (1.29, 4.18)	0.005
	2	179/1979	1.02 (0.71, 1.47)	0.92	125/1554	0.72 (0.44, 1.16)	0.18	52/404	2.28 (1.24, 4.20)	0.008
	3	178/1926	0.97 (0.68, 1.39)	0.87	124/1510	0.66 (0.41, 1.06)	0.09	52/396	2.30 (1.29, 4.11)	0.005

Model 1 adjusts for age, sex, education and systolic blood pressure in the left arm; model 2 adjusts for the stroke risk factors outlined in the Framingham Stroke Risk Profile [5]; model 3 adjusts for a positive Dementia Screening Indicator score (score ≥ 22). APOE = Apolipoprotein ε4 status.