1. Introduction
Adolescence is marked as a period of rapid developmental changes and often perceived as a time of changing behaviour and health across the life course [
1,
2]. As a transition from the childhood to adulthood, adolescence is a time of opportunity and vulnerability to health risk behaviour with lifelong consequences for health and well-being [
3,
4]. During this period smoking is most commonly initiated and addiction is likely to happen [
5]. Adolescent girls and boys who start smoking believe that they will be able to stop soon and easily, but the addictive nature of nicotine causes most of them to develop a dependence on this substance and to continue smoking for many years [
6]. In addition to the problems of tobacco addiction, smoking in adolescence has immediate consequences on physical health [
7], it is linked with depressive symptoms [
8,
9], suicidal ideation [
10] and with other addictive behaviours such as intensive alcohol consumption [
11], cannabis use, or gambling [
12]. Faced with this reality, the prevention of smoking at this age should be considered of high priority [
13].
Family support, which can be defined as positive parent–child interactions grounded in open communication and high parental sensitivity and responsibility to their child’s needs, has a beneficial impact on the psychological well-being of adolescents as well as on protecting against poor health outcomes and health-risk behaviour [
14,
15,
16]. Love, support, trust, and optimism from their family make adolescents feel safe and secure, and are powerful weapons against peer pressure, life’s challenges, and disappointments [
17]. However, the role of parents and parent–adolescent relations undergo a process of change through transitions. Even though family support decreases from early to late adolescence [
18] (pp. 31–34), parents continue to play a fundamental role in adolescent development, socialization, health, and well-being, and this role may be as important as it is in the early developmental stages, even though it is different and less noticeable [
19].
Several studies have found a positive relationship between perceived family support (especially parental support) and adolescents’ mental and physical health and prevalence of engagement in health-risk behaviours [
19]. A high level of parental support is associated with better emotional well-being, fewer internalizing and externalizing problems [
20,
21,
22], and better educational outcomes [
23]. A supportive family environment can also play a crucial role in health promotion, for example, assisting healthy changes of lifestyle. Family support is a protective factor against health-risk behaviour in ordinary samples [
21,
24,
25] and against maladjustment in at-risk populations [
26]. High parental support buffers against the negative consequences of adverse life events and peer-victimisation, especially among young female adolescents [
27].
Family environment and health behaviours during adolescence is one of the foci of the cross-national Health Behaviour in School-aged Children (HBSC) study [
28]. The previous reports of the study [
18,
29] have highlighted that over two thirds of adolescents reported high levels of support from their family, but wide cross-national variations were observed. Across most countries, younger girls and boys were more likely to report high family support. Significant gender differences were observed in less than half of countries/regions, with boys more likely to report higher levels of family support in most of these. In more than half of countries/regions, adolescent boys and girls from more affluent families reported higher levels of family support [
29]. However, despite the evidence of developmental changes in perceived family support there is still a paucity of research about the possible changes in the impact of family support on smoking prevention during adolescence period [
18]. The association between family factors and adolescent smoking habits may vary depending on the social and cultural context, so it is important to examine the relationships between adolescents from different countries. The HBSC study involves a wide network of researchers from more than 50 countries and regions, so its data allows us to successfully address such challenges [
28].
The present article is aimed to contribute more specifically to the current evidence based on the role of family support during adolescent transitions. Consistent with recent research [
18], the analysis has an objective to explore whether higher family support is associated with less risk of smoking behaviour, whilst controlling for demographic variables, focusing especially on gender, age, and country differences on the impact of family support on adolescent smoking. In line with this objective, the first hypothesis was that higher parental support is associated with a lower adolescent smoking risk. The second specific objective was to identify whether in having higher family support, adolescents can more easily pass the challenges during their transitions from early to late adolescence. Noting that the strength of the relationship between family support and adolescent smoking varied across countries, a second hypothesis was formulated. It claims that differences in smoking prevalence between 11- and 15-year-olds (e.g., during a period of transition from early to late adolescence) in each country appear to be related to the strength of the relationship between family support and adolescent smoking.
4. Discussion
The current study examined similarities and differences in perceived family support and its impact on overcoming the risk of smoking during adolescents’ transition from early to late adolescence. Data of the survey among 11–15-year-old adolescents from 42 countries that participated in the cross-national HBSC study in 2017/2018 were employed [
29]. The results of analysis confirmed both hypotheses that were raised for the objectives of the study. First, using the individual data records, it was found that higher family support is associated with a lower adolescent smoking risk. Second, using aggregated country data, it was revealed that the countries with a stronger mean effect of family support in reducing smoking risk have indicated lower rates of adolescent smoking prevalence as well as a lower increase in the current cigarette smoking prevalence during the age period from 11 to 15 years. In summary, these findings allow us to conclude that family support is an important asset helping adolescents to overcome the risk of smoking traversing adolescence.
The findings suggest that most young people feel family, generally their parents, supporting them. Proportion of adolescents who reported high family support in this HBSC wave remained almost the same as in the previous HBSC wave in 2013/2014 [
18] (pp. 31–34). In line with previous research [
41], our results indicate that perceived family support decreases with age. Traditionally older adolescents report having less family support compared to their younger counterparts. This result clearly reflects the beginning of the individualization process where the relationships with family members move from asymmetrical to a more symmetrical interaction being the adolescents treated as more autonomous individuals [
4]. According to our findings, boys reported having higher level of family support compared to girls, which is in line with recently manifested research that girls, more often than boys, start problems of connectedness with parents in early adolescence, at the age of eleven [
42]. Current research considers the parents–boys relationship to be more centred on independence and a higher need of psychological separation, while the parents–girls relationship is simultaneously based on independence but also connectedness, intimacy and reciprocity at the same time [
43].
The associations of adolescent smoking behaviour with familial or parental variables have been extensively examined [
44,
45,
46,
47]. Although various variables to describe family functioning and parenting have received a great deal of attention [
15,
48], in this study we focussed on the associations of current smoking with family support. Only a few articles were found in the literature to analyse such an association [
18] (pp. 31–34). Moreover, parenting includes several dimensions, including support and control. Parental support has been described as the variation in the amount of parental responsiveness and warmth, such as responding to the child needs, while parental control is a continuum that ranges from restrictiveness to permissiveness [
49]. The effects of parental support and parental control on early adolescence smoking was analysed in a longitudinal study in the Netherlands [
50]. Logistic regressions demonstrated that low parental control predicted adolescent smoking initiation but neither support nor control predicted adolescent smoking increase or continuation. Parental smoking status was important in adolescent smoking continuation and cessation. Unfortunately, it is beyond the scope of the present study to further analyse different components of family support.
The results of the study showed a significant association between family support and smoking prevalence, indicating a positive effect of family support on prevention of smoking among adolescents. This finding was in line with the previous research supporting the idea that family support can help to form the most important basic values, attitudes and patterns of behaviour making adolescent transitions easier [
51,
52,
53]. Consequently, family support has an overall protective effect on the risk of smoking among 11–15 years old adolescents. However, the strength of the association and the protective effect may vary, depending on other family characteristics, such as the material status of the family, parental monitoring and control, parental communication, and parenting styles. Research conducted by Mahabee-Gittens et al. (2012) revealed that higher parental monitoring and the attitudes towards smoking are significantly associated with recent smoking and ever smoking among US adolescents [
54]. Aho et al. (2018) also found a significant protective effect of parental involvement (tested as a composite measure of parent–child relationship, family connectedness, and parental monitoring) on Finish adolescents’ risk of smoking [
55]. Research conducted by Moore and Littlecott (2015) shows that higher family SES was associated with significantly lower likelihood of smoking and other health risk behaviours among Welsh adolescents [
56]. These findings suggest that family socioeconomic status and family support may have independent and combined effect on young people’s risk behaviours, particularly on the risk of smoking, while family support may have stronger protective effect on the risk of smoking among adolescent from lower affluent families [
56]. We did not find such associations in the current research. Instead, we found that family structure or living with both biological parents is a more important factor than family affluence. This factor may have an interaction with other familial factors in prevention of adolescent smoking [
44]. These combined effects of family support and other determinants of family functioning could explain the non-significant association between the increase in current cigarette smoking prevalence from 11 to 15 years and the percentage of boys and girls who reported high family support.
This study is exceptional because it involved adolescents from 42 countries. This allowed the analysis of variable associations at the country-level. One such analysis revealed that the prevalence of adolescent smoking was lower in countries where the greater proportion of adolescents felt high family support. This association was slightly stronger among girls than among boys, but it was significant only among the youngest adolescent groups. The prevalence of adolescent smoking had also a negative correlation with the strength of association between family support and adolescent smoking at country-level; however, in contrast with the previous association, it was significant in older adolescent groups only. Another country-level analysis found a significant correlation between the increase in adolescent smoking prevalence from 11 to 15 years and the strength of association between family support and adolescent smoking at country-level. The analysis of age-related developmental trajectory in adolescent health behaviours is important as such trajectories may also track into adulthood [
57,
58].
These results imply that the age-related increase in smoking prevalence during adolescence is less pronounced in countries where family support can be regarded as stronger protective factor against smoking than in countries where its impact is weaker. These findings are in line with previous research indicating the important role of family support during adolescents’ transition from early to late adolescence, especially as close family relationships can ameliorate the impact that adversity has on lifespan physical health [
16]. Moreover, a high level of perceived family support is related to lower levels of risk behaviours reducing their risk behaviours [
21,
25] and it is a protective factor for children in adverse environments [
24]. The mentioned increase in smoking prevalence at country level was not significantly correlated with the proportion of adolescents who reported high family support. The findings from examination of these associations may be generalized to support an epidemiologically based inference that the preventive effect or delay in the onset of many life-threatening conditions in the country does not depend on the extent of preventive measures but depends on their effectiveness. This means that the relationship between family support and smoking at the country level was moderated by the strength at the country-level of association between family support and adolescent smoking. Due to the unique cross-national design of our study, we did not find confirmation of this assumption among other studies. However, a preliminary analysis of HBSC data shows that this assumption is also valid for other outcome variables (e.g., alcohol consumption, and low life satisfaction), as well as observed in previous HBSC waves (e.g., in 2013/2014). Thus, the observed regularity deserves further investigation. Family can have important protective and preventive role, but also detrimental, such as being a supplier of alcohol to under-aged people [
59]. The study implies that the well-documented age-related increase in smoking prevalence during adolescence is less pronounced in countries where family support can be regarded as stronger protective factor against this behaviour than in countries where its impact is weaker.
Strengths and Limitations
The study is the product of an international network of researchers who work in topic-focused groups that collaborate to researching adolescent health. The research protocol includes scientific rational, international mandatory questions, and required procedures for sampling, data collection, and preparation of data set for ensuring high quality data. The measures of family life were based on valid scales. The use of large, nationally representative sample and the inclusion of 42 countries increases the generalizability of our finding. The analytical procedure facilitates the analysis of the relationships between family support and multiple self-reported aspects of adolescence transformation.
There are several limitations to this research. The measure of outcome variable in this study was cigarette smoking in the past 30 days. This may misclassify some respondents who smoked cigarettes occasionally but did not smoke in the past 30 days or used e-cigarettes to smoke. There is also a likelihood of recall bias with a question covering the past 30 days; such a time frame applies to many population-based studies of youth lifestyles, so the bias would be consistent across studies. Moreover, using sensitive questions can also be affected by the possibility for social fear bias in adolescent responses. However, every effort was made to minimize that possibility by ensuring strict anonymity of respondents. This study relied only on self-reported data, although these data are considered to hold the most valid information when studying subjective measures such as relations with parents. The proportion of current cigarette smoking among 11-year-old adolescents, especially among girls, was relatively small (1–2%); therefore, the estimations of associations should be considered with caution. This study did not measure peer influence that may play significant role in preventing or promoting adolescent smoking [
60]. Cultural factors may also contribute to family supports. Instead, we relied on prior studies’ findings of these cultural contexts [
61,
62,
63]. Future studies should attempt to study country-level factors (e.g., the tobacco control legislation and policy) that contribute to cultural differences in the association between family support and development of behaviour in adolescence transition. Our study was cross-sectional in nature; therefore, the findings of such a study can only suggest associations but not causation [
64]. Finally, we assessed the change in smoking prevalence during adolescence by comparing reports of 11- and 15-year-old adolescents in a cross-sectional survey, and further analysis of the associations were performed at the country level. Future research should continue to study the long-term associations between individual adolescent transitions such as family support and health behaviour trajectories. Particular attention should be paid to adolescents who are just trying to smoke (smoke 1–2 cigarettes per month) as they are most in need of family help and can benefit more than regular smokers.