Table 3.

Weighted multivariate regression models of sleep duration predicting patient-reported outcomes and healthcare use among adults with asthma (n = 1389)

	Multivariate Analysesa,b
	≤ 5 h of sleep vs. 6–8 h of sleep			≥ 9 h of sleep vs. 6–8 h of sleep
	β or ORc (95% CI)	P value		β or ORc (95% CI)	P value
Patient-reported outcomes
Asthma attack in past yr	1.58 (1.13–2.21)	0.008		1.27 (0.83–1.95)	0.26
Wheezing in past yr	1.31 (0.91–1.90)	0.14		0.88 (0.55–1.43)	0.61
Dry cough in past yr	1.95 (1.32–2.87)	0.001		1.15 (0.56–2.36)	0.70
Activity limitation due to wheezing in past yr
None	1.00			1.00
A little	1.09 (0.71–1.69)	0.64		1.82 (1.13–2.91)	0.01
Frequent	0.67 (0.30–1.48)	0.31		1.25 (0.70–2.24)	0.44
Health-related quality of life
Days physical health was not good	2.70 (0.90–4.51)	0.004		1.70 (−0.60–3.99)	0.15
Days mental health was not good	3.43 (2.07–4.79)	<0.001		1.62 (−0.71–3.97)	0.16
Inactive days due to health	2.53 (0.74–4.32)	0.007		2.23 (0.00–4.46)	0.05
Healthcare use
Emergency room visits for asthma in past yr	1.19 (0.79–1.79)	0.39		1.16 (0.62–2.17)	0.49
Number of times received healthcare for wheezing in past yr	0.09 (−0.12–0.31)	0.40		0.21 (−0.17–0.59)	0.27
Number of times received general healthcare in past yr	0.34 (0.11–0.57)	0.005		0.43 (−0.01–0.86)	0.06
Overnight hospitalizations in past yr	2.14 (1.37–3.36)	0.001		1.17 (0.62–2.22)	0.63

^aAnalyses used linear regression (for continuous dependent variables: days physical health was not good, days mental health was not good, inactive days due to health, number of times received healthcare for wheezing, number of times received healthcare), logistic regression (for dichotomous dependent variables), or multinomial logistic regression (for categorical dependent variable).

^bAdjusted for age, sex, race/ethnicity, socioeconomic status, and smoking status.

^cβ = standardized beta coefficient from linear regression; OR = odds ratio from logistic regression