Table 2.

BMI estimates of the univariate ordered probit model with no assumption of endogeneity between smoking and body weight

Variables	BMI
	Full sample	Males	Females
Smoker (smoking participation)	− 0.058	− 0.089	− 0.022
	(0.045)	(0.060)	(0.068)
Sociodemographic characteristicsa
Gender (women)	−0.282***	–	–
	(0.041)
30–60 years old	0.441***	0.446***	0.437***
	(0.064)	(0.089)	(0.092)
Over 60 years old	0.608***	0.508***	0.696***
	(0.075)	(0.105)	(0.107)
Married	0.167***	0.213***	0.141*
	(0.054)	(0.074)	(0.080)
Divorced/widowed	0.213***	0.157	0.209**
	(0.072)	(0.110)	(0.098)
Big city	−0.154***	−0.128*	−0.175**
	(0.049)	(0.069)	(0.071)
Village/Rural area	0.060	0.047	0.091
	(0.047)	(0.066)	(0.067)
Frequent consumption of:
Whole grain bread	−0.029	−0.088	0.041
	(0.040)	(0.056)	(0.060)
White/toasted bread	−0.053	− 0.134**	0.044
	(0.043)	(0.058)	(0.066)
Fruits	−0.023	−0.044	0.017
	(0.045)	(0.060)	(0.071)
Vegetables	−0.023	0.031	−0.063
	(0.042)	(0.060)	(0.059)
Meat products	0.299***	0.364***	0.254***
	(0.054)	(0.093)	(0.066)
Deep fried foods	−0.016	−0.121	0.160
	(0.074)	(0.091)	(0.132)
Confectionery	−0.159***	−0.191***	− 0.133**
	(0.039)	(0.055)	(0.056)
μ1	−1.742	−1.984	−1.287
μ2	0.359	0.280	0.768
μ3	1.428	1.483	1.690
Log-Likelihood	− 3644.031	− 1835.938	− 1780.192

Standard errors are given in parentheses

^aAge: 18–29 years old (reference category), Marital status: single (reference category)

Area of residence: town (reference category)

*p < 0.1, **p < 0.05, ***p < 0.01 (Statistically significant variables are highlighted in bold)