A total of 835 questionnaires were collected in the five schools. Only a few students, 2–4 students from each school returned their parents’ signed refusal forms and did not complete the questionnaires. A total of 30 cases were excluded due to incompleteness; as a result, 805 cases were included in the analysis for this study.
shows the demographic characteristics of the respondents. There were more boys than girls in this study (73.4% vs. 26.6%). A majority of the adolescents had never smoked (91.1%), 4.5% had tried smoking, 1.7% smoke occasionally, and 2.7% were regular smokers. More adolescents had drunk alcohol (25.3%), with 30.7% of them drinking occasionally. Less than 1% of adolescents (n = 8) reported ever having used illegal drugs. Nearly 80% of the adolescents lived with both parents and had siblings. Adolescents from single parent families comprised 17.4% of the respondents.
Demographic Characteristics of the Adolescents in the Study (n = 805).
Nearly half of the respondents had a father who smoked (44.6%), while a respective 5.6% and 7.8% of the adolescents had a mother and sibling(s) who smoked. About a quarter (22.9%) of the adolescents had at least one friend who smoked. More adolescents were satisfied with their mother’s role fulfillment (51.7%) than with their father’s (43.2%) and their own role fulfillment as a child (45.1%).
The calculation of the composite scores for family support and control showed that most of the parents (43.2%) were moderate in their support and control of their children, 38.5% were permissive, 10.9% were “neglectful”, 6.5% were “authoritarian”, and only 0.6% were “warm and directive” in their style of parenting ().
shows the comparison between adolescents who do and do not use substances, smoke, and drink, with reference to the smoking behaviors of family members, family structure, satisfaction with the roles of one’s parents and oneself, and the influence of friends. Those defined as smokers are those who had ever tried smoking or who smoke occasionally or regularly, and those defined as drinkers are those who had ever tried drinking. No comparisons were made for drug use since less than 1% of adolescents admitted to having tried illicit drugs.
Comparison of smoking behaviors of family members, family structure, satisfaction of roles, and friends’ factors between adolescents with and without risk behaviors (n = 805).
The comparison between smoking and non-smoking adolescents showed that more smokers than non-smokers have a father who smokes (62.5% vs. 42.9%, p = 0.002), a mother who smokes (22.2% vs. 4.0%, p ≤ 0.001), and siblings who smoke (25.0% vs. 6.1%, p = 0.001). Fewer smokers than non-smokers were living with both parents (62.5% vs. 81.0%, p ≤ 0.001) and perceived that their family objected to their smoking (72.2% vs. 84.9%, p ≤ 0.003). The smokers were less likely than the non-smokers to be satisfied with the role fulfillment of their father (29.1% vs. 53.2%, p ≤ 0.001), mother (47.2% vs. 64.2%, p = 0.003), and their own self as a child (23.6% vs. 46.6%, p ≤ 0.001). More of the smokers than the non-smokers had friends who smoke (66.6% vs. 16.4%), and would accept a friend’s offer to smoke (48.6% vs. 4.0%), but had fewer friends who object to smoking (33.3% vs. 63.9%), all with p ≤ 0.001 ().
The comparison between drinkers and non-drinking adolescents showed that more of the drinking than non-drinking and smoking adolescents have a father (48.6% vs. 38.1%, p = 0.005), mother (7.1% vs. 3.0%, p ≤ 0.013), and siblings (11.2% vs. 3.0%, p ≤ 0.001) who drink. More of the drinkers than non-drinkers were living with a single parent (20.2% vs. 13.3%, p = 0.012). Fewer drinkers than non-drinkers were satisfied with the role fulfillment of their father (48.8% vs. 54.2%), mother (61.3% vs. 64.8%), and their own self as a child (42.1% vs. 48.1%), but there were no statistically significant differences. More of the drinkers than non-drinkers had friends who drink (28.0% vs. 10.9%, p ≤ 0.001), and would accept a friend’s offer to drink (72.7% vs. 17.3%, p ≤ 0.001), but fewer had friends who object to drinking (59.0% vs. 64.9%, p = 0.019) ().
A comparison was made of the family process (family activities, conflicts with parents, and parental support and control) of adolescents who smoke/drink and those who do not (). The results show that when compared with non-smokers, smokers have a lower mean score for “time spent with parents” (mean = 9.06 vs. 11.78, p ≤ 0.001) and a higher mean score for “conflict with parents” (mean = 7.24 vs. 5.77, p = 0.002). The smokers, in comparison with those who do not smoke, also tended to think that their parents were providing them with less support (mean = 6.81 vs. 7.99, p ≤ 0.001) and exerting more control (mean = 4.32 vs. 3.45, p = 0.002). With regard to the comparison between drinkers and non-drinkers, the former have a higher mean score for “conflict with parents” (mean = 6.30 vs. 5.29, p ≤ 0.001). The drinkers, when compared with those who do not drink, also tended to think that their parents provide less support (mean = 7.58 vs. 8.33, p ≤ 0.001) and exert more control (mean = 3.83 vs. 3.09, p ≤ 0.001).
Comparison of family process between adolescents with and without health risk behaviors (n = 805).
Factors Contributing to Substance Use by Adolescents
Factors that contribute to substance use among the adolescents in this survey were identified using logistic regression. All statistically significantly differences between the users and non-users of substances were included in the analysis. The logistic regression analyses were carried out in two steps for each type of substance use (smoking or drinking). The first set of analyses was conducted by inputting all family-related factors. The second step in the analyses was then done by adding the variables relating to friends.
shows that when only family-related factors were analyzed, the factors contributing to adolescent smoking were: having a mother who smokes (OR = 4.633, 95% CI = 1.87–11.49, p ≤ 0.001), having a sibling who smokes (OR = 3.16, 95% CI = 1.51–6.64, p = 0.0012), and having parents who are “authoritarian” (OR = 1.856, 95% CI = 1.185–2.905, p = 0.007). Adolescents were less likely to smoke if they were satisfied with their father’s fulfillment of his role (OR = 0.478, 95% CI = 0.23–0.99, p = 0.048), and satisfied with their own self (OR = 0.410, 95% CI = 0.185–0.905, p ≤ 0.027). When friends’ factors were added in the second step, only the factors relating to siblings who smoke (OR = 2.61, 95% CI = 1.118–6.09, p = 0.027) and satisfaction with oneself (OR = 0.274, 95% CI = 0.121–0.624, p = 0.002) remained. Having friends who smoked (OR = 5.446, 95% CI = 2.608–11.374, p ≤ 0.001), and friends who invite one to smoke appeared to be the dominant contributors to adolescent smoking with OR = 10.455, 95% CI = 4.434–24.649, p ≤ 0.001).
Logistic regression analysis # of risk behaviors of adolescents (smoking cigarettes & drinking).
When only family-related factors were analyzed, the factors contributing to adolescent drinking were: having a sibling who drinks (OR = 4.53, 95% CI = 2.174–9.444, p ≤ 0.001) and having conflicts with one’s parents (OR = 1.423, 95% CI = 1.004–2.018, p = 0.048). Those who have parents who are “permissive” were less likely to drink (OR = 0.885, 95% CI = 0.786–0.996, p = 0.042). When friends’ factors were added in the second step, both the factors of having siblings who drink (OR = 3.607, 95% CI = 1.51–8.62, p = 0.004) and conflicts with parents (OR = 1.529, 95% CI = 1.01–2.31, p = 0.045) remained significant. Having friends who drank (OR = 1.894, 95% CI = 1.083–3.311, p = 0.025), and having been invited by friends to drink appeared to be the dominant contributors to adolescent drinking with OR = 11.825, 95% CI = 7.715–18.126, p ≤ 0.001) ().