A total of 361 students responded, giving a response rate of 2.3%. Of these respondents, 256 were females and 105 were males. Majority of the participants (60.4%, n = 218) aged 22 years and above. 36.6% (n = 132) were aged 20–21 years and 3.0% (n = 11) aged 18–19 years. 215 students (Females, 144; males, 71) identified as smokers. Female students (43.9%, n = 112) were significantly (p = 0.04) more likely to report a negative smoking status compared to male students (32.4%, n = 34). Table 1 shows the range of variable responses students who reported positive smoking status by gender.
Table 1
Student (smokers) responses by gender
Variables Responses
|
Total (smokers) N = 215
No. (%)
|
Females
N = 144
No. (%)
|
Males
N = 71
No. (%)
|
P-values
|
*Smoking frequency
▪ At least, once a day
▪ At least, once a week
▪ At least, once a month
▪ Less often than once a month
|
41 (18.1)
16 (7.0)
23 (10.1)
147 (64.8)
|
22 (14.4)
8 (5.2)
17 (11.1)
106 (69.3)
|
19 (25.7)
8 (10.8)
6 (8.1)
41 (55.4)
|
0.05
|
**Age Initiated smoking
▪ Below 18 years
▪ Above 18 years
|
73 (38.6)
116 (61.4)
|
50 (39.7)
76 (60.3)
|
23 (36.5)
40 (63.5)
|
0.673
|
**Cigarette source
▪ Self
▪ Parents/Siblings
▪ Friends/some else
|
77 (37.6)
4 (2.0)
124 (60.5)
|
47 (34.3)
2 (1.5)
88 (64.2)
|
30 (44.1)
2 (2.9)
36 (52.9)
|
0.268
|
**Lifetime number of Cigarette smoked
▪ 1–2 puffs
▪ 2–10 puffs
▪ 10–50 puffs
▪ More than 50 puffs
|
29 (13.2)
36 (16.4)
41 (18.6)
114 (51.8)
|
20 (13.5)
30 (20.3)
32 (21.6)
66 (44.6)
|
9 (12.5)
6 (8.3)
9 (12.5)
40 (66.7)
|
0.012
|
**Parental smoking status
▪ Both
▪ Father
▪ Mother
▪ None
|
32 (15.2)
29 (13.8)
27 (12.9)
122 (58.1)
|
21 (14.8)
21 (14.8)
20 (14.7)
80 (56.3)
|
16 (16.2)
8 (11.8)
7 (10.3)
42 (61.8)
|
0.775
|
** Parents smoke at home
▪ No
▪ Sometimes
▪ Yes
|
143 (68.4)
13 (6.2)
53 (25.4)
|
91 (64.1)
8 (5.6)
43 (30.3)
|
5 2(77.6)
5 (7.5)
10 (14.9)
|
0.058
|
* Variables added up above the total N because smoking experimenters may have regarded themselves as non-smokers |
**Variables did not add up to the total N because of missing values |
A three-class model was considered the best fit based on the model fit indices (AIC and BIC) and application of the principle of latent class separation and homogeneity. Table 2 shows the details of the model fit indices of the classes |
Table 2
Latent class analysis of the smoking typology of students
Model Class
|
AIC*
|
BIC**
|
(X^2) Chi-square goodness of fit
|
Maximum log-likelihood
|
2 classes
|
1903.643
|
1849.114
|
1280.424
|
-909.8217
|
3 classes
|
1849.114
|
2054.463
|
584.7915
|
-860.5568
|
4 classes
|
1969.812
|
2244.406
|
614.2618
|
-898.9059
|
5 classes
|
2182.732
|
2527.572
|
1829.542
|
-983.3661
|
*AIC, Akaike Information Criterion; |
**BIC, Bayesian Information Criterion. The best-fit model is indicated in bold. |
The 2-class model has the lowest BIC model fit index while the 3-class model has the lowest AIC model fit index. Both the 2-class model and 3-class model has similar latent class homogeneity. Though the 2-class model has a higher latent class separation than the 3-class, the parsimonious 3-class model was selected over the 2-class model due to a lower AIC and better maximum log-likelihood. Using item-response probabilities, we interpreted the 3 classes distinct (Table 3) as ‘late-onset, social smokers’ (class 1), ‘late-onset, light smokers’ (class 2) and early-onset, nicotine addicts (class 3). Table 4 represents the table of prediction of latent class membership with gender as covariates.
Table 3
Item response probabilities for a 3-class model of students’ smoking typology
Variables
|
Late-onset,
Social smokers
|
Late-onset, light
smokers
|
Early-onset,
nicotine addicts
|
Smoking Frequency
• Less often than once a month
• At least, once a month
• At least, once a week
• Daily
Life-time Number of Cigarettes Smoked
• 1–2 puffs
• 2–10 puffs
• 10–50 puffs
• more than 50 puffs
Age Initiated Smoking
• Before 18 years
• After 18 years
n (% of samples)
|
0.80
0.17
0.03
0.00
0.20
0.22
0.28
0.31
0.16
0.84
88 (41.3)
|
0.81
0.13
0.06
0.00
0.05
0.18
0.34
0.42
0.43
0.57
52 (24.3)
|
0.15
0.05
0.17
0.63
0.00
0.00
0.00
1.00
0.60
0.40
74 (34.4)
|
Table 4
Latent class membership prediction with gender as covariates
2/1
|
Coefficient
|
Standard error
|
t-value
|
Pr(>|t|)
|
Intercept
Gender (female)
|
0.2357
-0.61815
|
0.71401
0.56636
|
0.330
-1.091
|
0.742
0.277
|
3/1
|
Coefficient
|
Standard error
|
t-value
|
Pr(>|t|)
|
Intercept
Gender (female)
|
-0.5542
-0.0337
|
1.3003
0.9612
|
-0.426
-0.035
|
0.084
0.116
|
Overall, females are more likely to be ‘late-onset, social smokers’ (class 1) while the males are more likely to be ‘late-onset, light smokers’ (class 2) and ‘early-onset, addicted smokers’ (class 3). Females have 46.0% lesser odds of being in class 2 and 3.4% lesser odds of being in class 3 compared to the males.
Class 1: Late-onset, Social smokers
This class (n = 88, 41.3%) was considered social smokers because of the low probability of regular use of tobacco and a low number of lifetime cigarettes smoked. More than 83% of students that belong to this class had initiated smoking in their adulthood. Nearly 100% of them had neither of the parents as smokers. Regarding the source of cigarettes, 90% get from friends and peers and 10% purchase the cigarettes themselves.
Class 2: late-onset, light smokers
This class (n = 74, 34.4%) was considered moderate smokers because of the low probability of regular tobacco use and a higher probability of smoking more cigarettes in their lifetime compared to class 1. A majority (54.3%) of those in this group had initiated smoking in their adulthood. 33.8% had both parents as smokers; 42.0% had mothers who smoke and 33.8% reported their father was a smoker. 64.6% of this class reported their parents to smoke in the home. Regarding the source of cigarettes, 73.8% get from friends and peers; 24.0% buy themselves and 2% from parents/elder siblings.
Class 3: Early-onset, Nicotine addicts
This class (n = 52, 24.3%) were considered addicted smokers because of a higher probability of regular tobacco use and has smoked the largest number of cigarette in their lifetime (100% of them) compared to classes 1 and 2. The majority (60.0%) of those in this group had initiated smoking as a teenager. 14.9% had both parents as smokers; 10.6% had mothers who smoke and 20.6% reported their father was a smoker. 23.7% of this class reported their parents to smoke in the home. Regarding the source of cigarettes, 10.9% get from friends and peers; 87.5% buy themselves and 2% from parents/elder siblings.