Skip to main content
. Author manuscript; available in PMC: 2015 Apr 1.
Published in final edited form as: J Adolesc Health. 2013 Nov 19;54(4):460–466. doi: 10.1016/j.jadohealth.2013.09.014

Table 3.

Multinomial logistic regression models for overall frequency of alcohol use and heaviness of cigarette use.

Frequent vs. no alcohol consumption Infrequent vs. no alcohol consumption Heavy cigarette use vs. no cigarette use Light cigarette use vs. no cigarette use

Variable O.R. 95% C.I. for O.R. p-value O.R. 95% C.I for O.R. p-value O.R. 95% C.I. for O.R. p-value O.R. 95% C.I for O.R. p-value
Age 1.53 1.38-1.70 < .001 1.24 1.14-1.34 < .001 1.55 1.36-1.78 < .001 1.29 1.16-1.44 < .001
Gender 0.86 0.68-1.07 .170 0.80 0.67-0.96 .014 0.87 0.64-1.18 .376 0.55 0.43-0.70 < .001
Race 1.94 1.40-2.68 < .001 1.83 1.43-2.36 < .001 1.25 0.81-1.92 .314 1.38 0.98-1.94 .069
Ethnicity 1.94 1.31-2.88 .001 1.72 1.25-2.36 .001 1.99 1.24-3.19 .004 1.07 0.70-1.64 .755
Regular participation in activities 0.54 0.43-0.67 < .001 0.71 0.60-0.85 < .001 0.45 0.33-0.61 < .001 0.67 0.54-0.85 .001

Impulsivity 1.15 1.08-1.22 < .001 1.03 0.98-1.08 .207 1.22 1.12-1.33 < .001 1.11 1.04-1.18 .001
Sensation-seeking 1.25 1.19-1.32 < .001 1.20 1.15-1.25 < .001 1.05 0.97-1.15 .227 1.21 1.12-1.30 < .001

Have a part-time job 2.00 1.57-2.54 < .001 1.50 1.24-1.82 < .001 0.82 0.57-1.20 .313 1.38 1.04-1.81 .024

Sensation-seeking-by-part-time-job interaction ns 1.25 1.11-1.42 < .001 1.03 0.94-1.13 .537

Note. Gender: male coded 1, female coded 0; race: white coded 1, non-white coded 0; ethnicity: Hispanic/Latino(a) coded 1, non-Hispanic/Latino(a) coded 0; activity participation: at least one activity coded 1, no activities coded 0; part-time job coded 1, no job coded 0; O.R. = odds ratio; C.I. = 95% confidence interval. Statistically significant results at p ≤ .01 indicated in bold. Two-way interactions of part-time job status with impulsivity and with sensation seeking were tested initially in each model, however these interactions were dropped when not statistically significant. Thus the results presented here include only statistically significant interaction terms.