Table 3.

Relative Risk Ratios from Multinomial Logistic Regression Models Predicting Sleep Duration stratified by Respondent Race/Ethnicity, Sample Adults living with Romantic Partner, aged 25–84 Years in the National Health Interview Survey, 2004–2018.

	Short Sleep (vs. Normal)									Long Sleep (vs. Normal)
	Own/Race/Ethnicity									Own/Race/Ethnicity
	NH White		NH Black		Hispanic		NH Other			NH White		NH Black		Hispanic		NH Other
	RRR		RRR		RRR		RRR			RRR		RRR		RRR		RRR
Model 1
Romantic Partner not same Race/Ethnicity (Ref = Partner Same Race/Ethnicity)	1.21	***	1.12	***	1.23	***	1.13	**	0.96			1.01		0.94		1.12
Model 2
Partner Race/Ethnicity
Non-Hispanic White	Ref		1.08		1.05		0.97			Ref		1.06		1.00			1.17
Non-Hispanic Black	1.01		Ref		1.40	**	1.37			0.98		Ref		0.93			1.20
Hispanic	1.07		0.99		Ref		1.18			0.89		1.02		Ref			1.20
Non-Hispanic Other	1.16	**	1.01		1.21		Ref			1.09		0.83		0.99			Ref
Constant	0.33	***	0.27	***	0.21	***	0.24	***		0.14	***	0.16	***	0.19	***		0.16	***
N = 243,552	155,667		29,085		42,814		15,986			155,667		29,085		42,814			15,986

Notes: RRR = Relative Risk Ratio. For each model, the reference group of the partner is the group studied (e.g., the reference group for partner race/ethnicity for the Non-Hispanic White model is non-Hispanic White partners). All models account for age group, gender, number of offspring in the home, cohabitation status, region of residence, nativity status, household income, respondent educational attainment, partner educational attainment, hours worked per week, home ownership status, smoking status, alcohol consumption, self-reported health, kessler-6 scale, health insurance status and survey year.

Source: National Health Interview Survey, 2004–2018. *p < .05 **p < .01 ***p < .001.