Table 2. Hazard ratios (95% CI) for total stroke according to sleep duration.

	Sleep duration
	<6 h	6–8 h	>8 h
Cases, (%)	265 (3.98)	2800 (3.23)	70 (4.39)
Model 1†	0.96 (0.85–1.10)	reference	1.37 (1.08–1.74)
Model 2‡	0.96 (0.84–1.09)	reference	1.37 (1.08–1.74)
Model 3§	0.92 (0.81–1.05)	reference	1.29 (1.01–1.64)
Model 4※	0.92 (0.81–1.05)	reference	1.29 (1.01–1.64)
Sensitivity analysis*	0.92 (0.80–1.05)	reference	1.29 (1.01–1.65)
women
Cases, (%)	32 (2.89)	292 (1.62)	9 (2.40)
Model 1†	1.16 (0.80–1.68)	reference	2.27 (1.16–4.43)
Model 2‡	1.14 (0.79–1.66)	reference	2.33 (1.19–4.55)
Model3§	1.09 (0.75–1.59)	reference	1.91 (0.98–3.74)
Model 4※	1.09 (0.75–1.60)	reference	1.91 (0.98–3.74)
Sensitivity analysis*	1.03 (0.69–1.53)	reference	2.00 (1.02–3.92)
men
Cases (%)	233 (4.20)	2508 (3.65)	61 (5.00)
Model 1†	0.93 (0.82–1.07)	reference	1.31 (1.02–1.69)
Model 2‡	0.93 (0.81–1.07)	reference	1.31 (1.02–1.69)
Model 3§	0.90 (0.79–1.04)	reference	1.24 (0.96–1.61)
Model 4※	0.90 (0.78–1.03)	reference	1.24 (0.96–1.60)
Sensitivity analysis*	0.91 (0.79–1.04)	reference	1.24 (0.95–1.61)

CI, confidence interval.

^†Model 1 was stratified by hospitals, and adjusted for age and sex.

^‡Model 2 was stratified by hospitals, and adjusted for as model 1 plus marital status, family per member monthly income, education level, smoking status, drinking status, physical activity, and family history of stroke.

^§Model 3 was stratified by hospitals, and adjusted for the variables in Model 2 plus body mass index, systolic blood pressure, diastolic blood pressure, fasting blood glucose, total cholesterol, hypotensive drug use, lipid-lowering drug use, hypoglycemic drug use, history of myocardial infarction, and snoring status.

^※Model 4 was stratified by hospitals, and adjusted for the variables in Model 3 plus high sensitive C-reactive protein, and atrial fibrillation.

^*Adjusted for model 3 and further excluded individuals with myocardial infraction and cancer.