202
Hikmet Yazıcı
Psychological Symptoms and Smoking Status
A logistic regression analysis was performed with smoking status as a dependent variable
and gender, age, SES, maternal smoking, paternal smoking, depression, anxiety, hostility,
self-negativity and somatization as predictor variables. A total of 779 cases were analyz-
ed and the full model was significantly reliable (chi-square =146.40, df = 16, p<.001). This
model accounted for 17.1% to 25.0% of the variance in smoking status, with 67% of the
current smokers successfully predicted. The model correctly predicted the outcome, with
76% (Table 2).
Bivariate analyses identified demographic and psychopathological variables correlated
with smoking. As can be seen in Table 2, smoking was more likely among male (OR=6.39,
95% CI = 4.34-9.39, p<.001). Smoking was also correlated with maternal smoking
(OR=1.64, 95% CI = 1.09-2.45, <.05), self-negativity (OR=0.95, 95% CI = 0.91-0.99,
p<.05) and hostility (OR=1.12, 95% CI = 1.07-1.17, p<.01). There was no correlation be-
tween smoking and other psychological symptoms, paternal smoking, and SES.
Table 2 Logistic Regression Analyses for Correlates of Smoking.
Predictor Variables
OR
CI
Gender
Female*
1.00
Male
6.39
4.34-9.39
SES
1 (Very Low)*
1.00
2
1.09
0.29-4.11
3
1.17
0.43-3.19
4
1.33
0.37-4.81
5 (Very High)
3.44
0.54-22.11
Maternal smoking
No*
1.00
Yes
1.64
1.09-2.45
Paternal Smoking
No*
1.00
Yes
1.13
0.72-1.77
Depression
1.02
0.98-1.06
Anxiety
0.99
0.95-1.04
Self-negativity
0.95
0.91-0.99
Somatization
1.03
0.98-1.07
Hostility
1.12
1.07-1.17
* reference group; OR = odds ratio; CI = 95% confidence interval.
