Table 4. Association of Each Form of Disadvantage With Smoking Status Among 278 048 Respondents, 2008 to 2017.

Disadvantage	Association With Smoking Status Collapsed Across Survey Years, OR (95% CI)a			Time Trend Analysis for Association With Current vs Never Smoking Statusb
	Current vs Never Smoker	Current vs Former Smoker	Former vs Never Smoker	Disadvantage × Survey Year Interaction Variable, P Valuec,d	Survey Year Variable Stratified by Disadvantage Status, OR (95% CI)c
Unemployed
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	0.96 (0.95-0.96)e
Yes	1.83 (1.73-1.93)e	1.44 (1.36-1.52)e	1.27 (1.21-1.33)e	.002e	0.98 (0.96-0.99)e
Poverty
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	0.96 (0.95-0.96)e
Yes	1.81 (1.74-1.88)e	1.89 (1.81-1.97)e	0.96 (0.92-1.00)	<.001e	0.98 (0.97-0.99)e
Low education
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	0.96 (0.95-0.96)e
Yes	1.69 (1.63-1.76)e	1.60 (1.50-1.64)e	1.08 (1.04-1.12)e	<.001e	0.99 (0.98-1.00)
Disability/limitation
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	0.95 (0.95-0.96)e
Yes	1.42 (1.37-1.47)e	1.13 (1.08-1.18)e	1.26 (1.22-1.30)e	<.001e	0.98 (0.97-0.99)e
Serious psychological distress
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	0.96 (0.95-0.96)e
Yes	2.25 (2.12-2.38)e	1.83 (1.71-1.95)e	1.23 (1.15-1.32)e	.02e	0.98 (0.96-1.00)e
Heavy drinking
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]	0.96 (0.95-0.96)e
Yes	4.83 (4.48-5.21)e	2.47 (2.29-2.66)e	1.96 (1.80-2.13)e	.42	0.97 (0.94-0.99)e

Abbreviation: OR, odds ratio.

^aAssociation between regressor and respective smoking status pairwise contrast in multivariable polytomous regression models that included unemployment, poverty, low education, disability/limitation, serious psychological distress, and heavy drinking as simultaneous regressors adjusted for sex, race/ethnicity, age, survey year, and region weighted to the US population. The Methods section gives the operational definition of each form of disadvantage. The eMethods in the Supplement describes disability/limitation.

^bTime trend analysis of footnote a.

^cYear treated as a 10-level continuous variable (2008-2017).

^dInteraction term added in subsequent model; all other results are from models that excluded interaction term.

^eStatistically significant after Benjamini-Hochberg correction to maintain studywise false discovery rate at 0.05.