Table 5.

Logistic regression analysis of variables associated with odds of being a current smoker in relation to amount of tobacco marketing

Variable (reference category)	No	Odds ratio (95% CI)	P value
Total number of tobacco marketing techniques of which aware		1.13 (1.06 to 1.20)	<0.0001
Number of friends who smoke:
All of them (none of them)	43	29.38 (7.56 to 114.12)	<0.0001
Most of them (none of them)	205	16.05 (4.82 to 53.50)	<0.0001
About half of them (none of them)	146	6.72 (1.95 to 23.07)	0.002
A few of them (none of them)	126	1.92 (0.51 to 7.21)	0.336
Not sure (none of them)	19	7.21 (1.37 to 37.92)	0.020
Any siblings who smoke:
Yes (no)	185	3.01 (1.94 to 4.66)	<0.0001
Not sure (no)	47	2.93 (1.26 to 6.82)	0.013
Mother smokes:
Yes (no)	231	1.97 (1.29 to 2.98)	0.002
No mother (no)	9	3.26 (0.77 to 13.82)	0.109
Not sure (no)	20	0.21 (0.04 to 1.13)	0.069
Constant		0.01	0.010

Variables were included in the analyses in a forward stepwise procedure. A probability of 0.05 was used for a variable to enter the equation and a probability of 0.10 to be removed. 

The dependent current smoker (1) was compared against non-smokers and tried smokers (0). 

The analysis included the following controls: gender, age, friends' smoking, sibling's smoking, mother's smoking, father's smoking, socioeconomic group, marital status of parents, future education intentions, and parental presence during interview.