Table 4

 Logistic regression of absence in last week (smoking status by health)

Matched health sample (n = 71824)
nORp Value
Data are from the US Department of Commerce, Census Bureau, National Cancer Institute Sponsored Tobacco Use Supplement to the Current Population Survey, 1992–1999.
The matched health sample refers to a roughly 1 of 6 random sub-sample for whom extended survey data on health status were available.
Four smoking status indicators were interacted with three categories of health status to generate results.
All models also control for age, education, race, ethnicity, marital status, number of children, occupation, industry, metropolitan statistical area, state, and month/year.
Column (a) displays the p value for the test of significance of the odds ratio (OR) relative to the referent “never smoker”.
Column (b) displays the p value for the test of equality between the indicated OR and “current smoker”.
Health main effects
Good/very good426520.4770.000
n OR (a) (b)
Smoking effects by health
Fair/poor health
    Quit <1 year ago583.0790.0010.002
    Quit 1 year5541.0210.9080.841
    Never smoked16621.000
Good/very good health
    Quit <1 year ago7041.9780.0000.008
    Quit 1 year62351.3250.0000.606
    Never smoked226161.000
Excellent health
    Quit <1 year ago4002.0200.0120.880
    Quit 1 year36671.3200.0350.007
    Never smoked158731.000