Title: Assessing predictors of smoking behavior among students in Senegal – evidence from the 2020 Senegal Global Youth Tobacco Survey (GYTS)

‌Abstract

Background: There is a growing trend in cigarette use among students in Africa in general, and in Senegal in particular. While the association between parental smoking and children's smoking has been described elsewhere, differences in family structures and dynamics in Africa, such as the extended nature of most families, may influence the impact of parental smoking on students' smoking. This study aims to examine the impact of parents' smoking behavior on students’ smoking habits, using the most recent school-based survey in Senegal.

Methods: We used the publicly available 2020 Senegal Global Youth Tobacco Survey (GYTS), a cross-sectional survey conducted between January and December 2020 by the National Tobacco Control Program of Senegal. Our outcome of interest was students' current cigarette smoking habits as a binary variable, and the predictor was parents’ current smoking habits. Additional covariates include friends' smoking habits, student sex, grade, weekly pocket money, and parents' employment status. We utilized descriptive statistics and weighted logistic regression for the association between parents’ smoking and students’ smoking.

Results: Among the total sample of 3,723 students, 11.04% (unweighted n = 402) reported being smokers, while 88.96% (unweighted n = 3,321) reported not smoking. In the crude model,

students with at least one smoking parent had nearly twice the odds of smoking compared to those with no smoking parent (OR = 1.93, p-value = 0.0001). This association remained significant in the multivariate model after adjusting for whether the student had at least one friend who smokes, students’ sex, grade, pocket money, and parental work status (adjusted OR = 1.59, p- value = 0.0073).

Conclusions: In this study, we found that despite structural differences in family composition between Western and African contexts, parental smoking remains a significant predictor of students’ smoking behavior, emphasizing the need for including parents in smoking-prevention programs targeted at students.

‌Introduction

The link between cigarette smoking and adverse health outcomes is well demonstrated. Indeed, smoking is associated with cancers (Li & Hecht, 2022), cardiovascular diseases (Kondo, Nakano, Adachi, & Murohara, 2019), stroke (Shah & Cole, 2010), lung diseases, including chronic obstructive pulmonary disease (COPD)(Lu et al., 2024), and diabetes (Zubizarreta, Mezquita, García, & Ferrero, 2017) among others. Smoking weakens the immune system (Stämpfli & Anderson, 2009), impairs blood circulation, and increases the risk of infections (Stämpfli & Anderson, 2009). Consequently, smoking is a leading cause of morbidity and mortality worldwide (Goldenberg, Danovitch, & IsHak, 2014). Interestingly, the risk associated with the occurrence of smoking-related morbidities and mortality is increased with early smoking (Choi & Stommel, 2017). This multifaceted adverse impact of smoking on health emphasizes the need for preventative approaches, especially among youth.

The link between parents and children smoking is already established. Children of current or former smokers are significantly more likely to become smokers themselves (Doherty & Allen, 1994; Otten, Engels, van de Ven, & Bricker, 2007). Some studies suggested that having two smoking parents was associated with early experimentation with smoking (Alves & Perelman, 2022). Other studies suggested that the influence of the father's smoking behavior was stronger than the mother's on offspring smoking (Yu, Qin, & Li, 2022).

However, studies in African countries are scarce, as most research has been conducted in Western countries (Vuolo & Staff, 2013). The link between culture and cigarette smoking is multifaceted, encompassing social norms, identity, values, and historical context (Nichter, 2003). Culture shapes both the initiation and persistence of smoking behaviors, as well as attitudes toward cessation and public health interventions (Egbe, Petersen, Meyer-Weitz, & Oppong Asante, 2014). Unlike in many Western societies where families typically consist of only parents and children, African youth—particularly in rural areas—are more likely to grow up within extended family structures that include uncles, aunts, and grandparents (Wilson & Ngige, 2006). Consequently, the influence of parental smoking habits on students may not mirror the patterns commonly observed in Western contexts. Furthermore, differences in gender roles and educational norms may also influence smoking behaviors among youth, as societal expectations around masculinity and femininity shape the acceptability of smoking (Weber et al., 2019).

Some studies attempted to tackle the issue in Africa but either remained mostly descriptive (James, Bah, Kabba, Kassim, & Dalinjong, 2022) or focused on other types of exposure, such as shisha (Cham, Weaver, Jones, Popova, & Jacques, 2024). In the context of rising non- communicable diseases across the developing African continent (Barry et al., 2025; Bigna &

Noubiap, 2019), understanding the complexities of parental smoking behaviors and their influence on youth smoking experiences can contribute to the development or refinement of effective tobacco prevention policies.

In this study, using data from the 2020 Global Youth Tobacco Survey 2020 conducted in Senegal, we aim to assess the link between parents' smoking and children's smoking, considering the confounding effect of parent's employment status and the interaction between the sex of student, the availability of money for them to use in a week, and the smoking behavior among their friends.

‌Materials and Methods

‌Study population

Data used came from extracted from the Senegal Global Youth Tobacco Survey (GYTS) 2020, a cross-sectional study conducted between January and December 2020 by the National Tobacco Control Program of Senegal, under the coordination of the Ministry of Health and Social Action (Senegal), the World Health Organization (WHO) and the US Centers for Disease Control and Prevention (CDC)(Centers for Disease Control and Prevention (CDC), 2020). A population- representative sample of students aged 13 to 15 was obtained through a two-stage sampling cluster design. First, schools were selected with a probability proportional to student enrollment, in Dakar – the capital city of the Country – and the rest of the country. Second, classes were randomly selected, and all students of selected classes – 4,320 in total – self-completed a standard core questionnaire, complemented with optional questions. The survey generated a response rate of 93.9%. The collected data was processed by the CDC, and made publicly

accessible on the WHO website. Ethical approval and informed consent were not required for this paper, as it involved secondary analysis of de-identified data. While the specific Senegal 2020 GYTS documentation does not detail the consent form process in the available summaries, the standardized GYTS protocol requires that informed consent (and often parental consent for minors) be obtained prior to participation in the survey (Ebraima Manneh, 2008).

‌Measures

Outcome

Our outcome of interest was cigarette smoking among surveyed participants, assessed by the response to the question, "Have you ever tried or experimented with cigarette smoking, even one or two puffs?" with available responses as “yes” or “no”.

Exposure

Our exposure of interest was cigarette smoking among parents of surveyed participants, assessed by the question "Do your parents smoke tobacco?'. Parent smoking was coded as 1) None, 2) Both, 3) Father only, 4) Mother only, and 5) Don't know. (note: all these response options are taken verbatim from the code book). Because some categories had very few observations, we decided to recode parental smoking into a dichotomous variable: "None" when no parent smoked, and "At least one" when either or both parents smoked.

Confounder

From the set of possible covariates, we identified the employment of parents as the only one that fulfills the criteria for a confounder – a cause of both the exposure and the outcome, and

not caused by either the exposure or the outcome of interest. Indeed, we hypothesized that unemployment leads to financial hardship, which in turn causes both parents and children to smoke, but smoking of either parents or students may not lead to financial hardship. Employment status was assessed with the question, "Do your parents work?". Parents' work status was classified as 1) Father (stepfather or mother's partner), 2) Mother (stepmother or father's partner), 3) Both, 4) Neither, and 5) Don't know. We also created a dichotomous variable for parents’ work status, with “Yes” indicating that the student reported at least one working parent, and “No” otherwise.

Effect measure modifiers

We hypothesized that the smoking habits of parents will induce smoking habits in children differently, based on the sex of the children, the amount of money the student has for free spending per week, and whether the student has friends who smoke as well. As such, these variables were considered as effect measure modifiers. With regard to sex, participants were classified as male or female based on response to the question, "What is your sex?". Money was assessed with this question: "During an average week, how much money do you have that you can spend on yourself, however you want?", with the following possible responses: “I usually don't have any spending money”, “ Less than 2500 F”, “ 2500-5000 F”, “ 5001-7500 F”, “ 7501- 10000 F”, and “More than 10000 F”. For the purpose of this manuscript, we recoded participants into 3 groups because of fewer observations for participants reporting at least 2500F: “usually don't have any spending money”, “less than $ 4.38 ”, and “at least $ 4.38 ”. The amount of $4.38 corresponds to 2,500 F as of April 2025, which is the local currency reported in the survey data. We assumed that students may have additional sources of income from their own activities,

enabling them to earn money independently of their parents. Friends’ smoking behavior was assessed with this question “Do any of your closest friends smoke tobacco?”, with 4 possible answers: “None of them”, “Some of them”, “Most of them” and “All of them”. We also recoded the friends smoking habit into a dichotomous variables with “None” if the student’s response is “none of them”, or “at least one” if the student provided any of the 3 remaining answers, “some of them”, “most of them” or “all of them”.

We finally decided to control for grade as a proxy for age, which also contains information on educational attainment. Roughly, 6eme, 5eme, 4eme, and 3eme correspond to 6th , 7th , 8th , and 9th grade, respectively, in the US school system.

For all responses, participants who replied "don't know" were classified as missing data.

‌Statistical analysis

We described the characteristics of the study population overall in terms of student smoking habit, friends' smoking habits, sex, grade, pocket money, and parent working status, by parent smoking status. Differences in the distribution of variables by parent smoking status were assessed using the Rao-Scott Chi-square test.

We then proceeded to assess the relationship between parental smoking behavior and students’ smoking behavior using univariate and multivariate logistic regression, adjusted for the selected covariates. We also assessed effect measure modifications (EMM) on the multiplicative scale using logistic regression, and made the determination to proceed or not, with the assessment of EMM on the additive scale using log-linear regression, based on the significance of the p-values of the interaction terms.

Analyses were adjusted for the complex survey design by including sampling weights, stratification, and clustering variables. All statistical analysis was conducted using SAS® Studio online (SAS Institute Inc., Cary, NC, USA), and statistical significance was set at α = 0.05, except for the interaction terms, where α was set to 0.1.

‌Results

Among the total sample of 3,723 students, 11.04% (unweighted n = 402) reported being smokers, while 88.96% (unweighted n = 3,321) reported not smoking. Most students (unweighted n = 3,153) reported that neither parent smoked, among which 9.95% (unweighted n = 305) were smokers. Among students who reported to have at least one parent who smokes (unweighted n

= 570), 17.56% (unweighted n = 97) reported smoking. The Rao-Scott chi-square p-value was < 0.0001, suggesting that student smoking behavior differed by parent smoking behavior.

Most students had no friends with smoking behavior, with 90.79% (unweighted n = 3,419) in the total sample, 93.14% (unweighted n = 2,958) among students whose parents were non-smokers and 77.21% (unweighted n = 461) among those with at least one smoking parent. This difference was statistically significant (p < 0.0001), suggesting that students’ exposure to peers who smoke also varied by parental smoking status.

Males represented 46.38% of the total participants (unweighted n = 1,727), 45.41% (unweighted n = 1,418) among students with non-smoking parents, and 51.96% (unweighted n = 309) among those with at least one smoking parent. This difference was statistically significant (p = 0.0048), suggesting that the distribution of participants’ sex varied by parental smoking status.

In the total sample, 26.84% (unweighted n = 1039) were in 6th grade, 24.68% (unweighted n = 1221) in 7th grade, 25.87% (unweighted n = 865) in 8th grade, and 22.60% (unweighted n = 726) in 9th grade. Among students with non-smoking parents, 26.60% (unweighted n = 867) were in 6th grade, 24.74% (unweighted n = 1,038) in 7th grade, 26.00% (unweighted n = 723) in 8th grade, and 22.66% (unweighted n = 613) in 9th grade. For those with at least one smoking parent, 28.26% (unweighted n = 172) were in 6th grade, 24.33% (unweighted n = 183) in 7th grade, 25.15% (unweighted n = 142) in 8th grade, and 22.26% (unweighted n = 113) in 9th grade. This difference was not statistically significant (p = 0.9447), indicating that participants’ grade distribution did not differ by parental smoking behavior.

With regard to access to pocket money, in the total sample, the majority of students, 47.69% (unweighted n = 1,914), reported having less than $4.38 per week, free to use. This observation remained true among those who reported at least one parent who smoked, with 48.16% (unweighted n = 1,622) and 45.00% (unweighted n = 292) among those who reported at least one parent who smoked. This distribution did not significantly differ by parental smoking status (p = 0.1201).

Finally, most students reported having at least one working parent, with a total of 92.78% (unweighted n = 3,500) for the entire sample, 92.80% (unweighted n = 2,947) among students with non-smoking parents, and 92.67% (unweighted n = 553) among students with smoking parents. This distribution did not significantly differ by parental smoking status (p = 0.8858).

A full description of the study population by parental smoking status is presented in Table 1.

We then assessed the association between having at least one parent who smokes and students’ smoking behavior (Table 2). In the univariate model, students with at least one smoking parent had nearly twice the odds of smoking compared to those with no smoking parent (OR = 1.93, p- value = 0.0001). This association remained significant in the multivariate model after adjusting for whether the student had at least one friend who smokes, students’ sex, grade, pocket money, and parental work status (adjusted OR = 1.59, p-value = 0.0073).

Similarly, students with at least one friend who smokes had significantly higher odds of smoking (crude OR = 3.43, p-value <.0001, adjusted OR = 2.56, p-value = 0.0002) compared to students who had no smoking friends. The odds of smoking were significantly higher among males (crude OR = 3.33, p-value = <.0001, adjusted OR = 3.28, p-value <0.0001) compared to females, and student with at least $4.38 per week, (crude OR = 1.58, p-value = 0.0258, adjusted OR = 2.07, p- value = 0.0019), compared to students who reported not having weekly money to spend. We did not find any association between grade and student smoking behavior, or between parents' working status and students' working behavior.

We then moved forward to assess whether the association between parents’ smoking behavior and students’ smoking behavior was modified by students’ sex, students' friends’ smoking behavior, and their access to money. Interestingly, the interaction terms between sex, money and friends smoking behavior with parental smoking status remained non-significant (p-value level of significance set for < 0.1), suggesting no effect measure modification on the multiplicative scale (p-value for interaction parent smoking and student sex was 0.6925, parent smoking and having some money but less than $4.38 was 0.6985, parent smoking and having at

least $4.38 was 0.8092, and parent smoking and at least one friend who smokes was 0.3303) (Table 3). Consequently, we did not proceed with the assessment for EMM on the additive scale.

Table 1: Characteristics of students by parental smoking status from the 2020 Senegal Global Youth Tobacco Survey (GYTS)

Table 2: Association between parent smoking and student smoking from the 2020 Senegal Global Youth Tobacco Survey (GYTS)

Table 3: Interaction between parent smoking behavior, and sex, access to money and friends’ smoking behavior

‌Discussion

In our study, we found a significant association between parental smoking and students’ smoking behavior in Senegal, using nationally representative data from the 2020 Global Youth Tobacco Survey of 3372 participants. Students with at least one smoking parent were significantly more likely to report current smoking, after adjusting for friends smoking behavior, sex, access to money, grade, and parents work status. Different explanations have been formulated with regard to the intergenerational continuity observed in smoking behavior between parents and their children. One explanation relates to the role of socialization, in which children absorb parental attitudes and norms around tobacco use (Yu et al., 2022). Indeed, youth may be more inclined to view smoking as an acceptable behavior in families where smoking is normalized or implicitly condoned (Albers, Biener, Siegel, Cheng, & Rigotti, 2008; Yu et al., 2022). A second explanation correlates with behavioral modeling, where, during formative years, children often emulate the actions of adults who hold authority or emotional significance in their lives—most notably, their parents (Yu et al., 2022). A third explanation relates to easy access to cigarettes in families where parents smoke. Indeed, when tobacco products are easily found at home and boundaries regarding their use are unclear or inconsistently enforced, adolescents may be more likely to experiment with smoking (Yu et al., 2022). Finally, genetic factors may also contribute to the intergenerational persistence of smoking behavior (Loukola, Hällfors, Korhonen, & Kaprio, 2014; Yu et al., 2022).

Likewise, students who had at least one friend who smokes had significantly higher odds of smoking themselves. Peer influence plays a significant role in both the initiation and maintenance of cigarette smoking among adolescents (Robalino & Macy, 2018). Much like parental influence,

the mechanisms through which peers shape smoking behavior are complex and operate through multiple pathways. Adolescents may encounter direct or indirect pressure to smoke, ranging from explicit encouragement to more subtle forms of social influence, such as observing peers who smoke (Robalino & Macy, 2018). Through social modeling, the visibility of smoking within peer groups increases the likelihood that adolescents will perceive the behavior as typical or aspirational. Social reinforcement further amplifies this effect, as peer groups may reward smoking with approval or inclusion, while non-smoking may be met with exclusion or ridicule (Liu, Zhao, Chen, Falk, & Albarracín, 2017). Additionally, peer networks often provide the first access to cigarettes, lowering the barriers to experimentation and creating environments where smoking becomes a shared group activity (Liu et al., 2017). We also demonstrated that boys and students with greater weekly financial resources were significantly more likely to report current tobacco use. These findings suggest that both gender norms and economic access may play key roles in facilitating smoking behavior among youth.

In this study, we found that despite structural differences in family composition between Western and African contexts—where nuclear families are more common in Western countries and extended family arrangements prevail in many African settings, parental smoking remains a significant predictor of students’ smoking behavior. Additionally, being male, having close friends who smoke, and having access to discretionary money were all associated with an increased likelihood of smoking among students. Family-centered tobacco control interventions should be broadened to include extended family members, emphasizing the role of adult behavior in shaping adolescent choices. Schools should implement peer-focused prevention programs that address group norms and equip students with strategies to resist peer pressure, especially

targeting male students who appear at greater risk. Further studies could be conducted, including qualitative or mixed-methods research, to assess the impact of advertising and anti- tobacco education in relation to the influence of parental smoking on student smoking behavior.

This study has some limitations. As a school-based, self-administered questionnaire, the responses to the questions may be subject to selection bias, as it excludes adolescents who are not enrolled in school or absent on the day of data collection — groups that may differ systematically in tobacco use behaviors. Moreover, the reliance on self-reported data introduces the possibility of information bias, particularly social desirability bias, where students may underreport their smoking behavior. This might be particularly true for girls, since smoking is generally not considered a desirable trait for women in general. The cross-sectional design of the GYTS limits causal inference, as it is not possible to establish temporality or directionality of associations between variables.

SAS Synthax

*============== Importing data ==============; libname paper "/home/nanawandre0/AppliedRDMA/paper"; run;

filename refile '/home/nanawandre0/AppliedRDMA/paper/SENEGAL_DATA_20.csv'; proc import datafile = refile

dbms=CSV out=WORK.sen; getnames=YES;

run;

*============== Data formatting ==============; data sen1;

set sen;

*---------------- Outcome ;

* CR5 Have you ever tried or experimented with cigarette smoking, even one or two puffs?;

if CR5 = 1 then smoking = "Yes"; else if CR5 = 2 then smoking = "No"; else smoking = ""; * Missing values;

*---------------- Exposure ;

** Parent_smoke for logistic regression model; if OR45 = 1 then parent_smoke = "None";

else if OR45 in (2,3,4) then parent_smoke = "Any"; *At least one parent smokes;

else if OR45 = 5 then parent_smoke = ""; * responded "Don't know"; else parent_smoke = ""; * Missing values;

*---------------- Other covariates ;

• CR2 sex;

  if CR2 = 1 then sex = "M";

  else if CR2 = 2 then sex = "F"; else sex = ""; * Missing values;

• Grade;

  if SNR3 = 1 then grade = "6th grade";

  else if SNR3 = 2 then grade = "7th grade"; else if SNR3 = 3 then grade = "8th grade"; else if SNR3 = 4 then grade = "9th grade"; else grade = ""; * Missing values;

  
  

• SNR4 During an average week, how much money do you have that you can spend on yourself, however you want?;

  if SNR4 = 1 then money = "Usually no money";

  else if SNR4 = 2 then money = "Less than $4.38 CFA"; else if SNR4 => 3 then money = "At least $4.38 CFA"; else money = ""; * Missing values;

• OR1 parent_work;

  if OR1 in (1,2,3) then parent_work = "Yes"; * at least one working parent;

  else if OR1 = 4 then parent_work = "No"; else if OR1 = 5 then parent_work = ""; else parent_work = ""; * Missing values;

• OR46 friends_smoke;

  if OR46 = 1 then friends_smoke = "None";

  else if OR46 in (2, 3, 4) then friends_smoke = "Any"; *Some friends to all friends;

  else friends_smoke = ""; * Missing values;

  rename

  FinalWgt = weight;

  
  

  label

  smoking = "Student smoking behavior" parent_smoke = "Parent smoking behavior" sex = "Student sex"

  grade = "Grade"

  money = "Pocket money per week" parent_work = "Parent work status" friends_smoke = "Friends smoking behavior"

  ;

  run;

  
  

• checking new variables against old variables for correctness; proc freq data = sen1;

  table CR5 smoking

  OR45 parent_smoke CR2 sex

  SNR3 grade SNR4 money

  OR1 parent_work OR46 friends_smoke

  ;

  run;

  *============== Table 1 ==============;

• Computing all levels of parental smoking status (None, Father only Mother only, Both) accross variables of interest;

proc surveyfreq data=sen1; tables smoking*parent_smoke

sex*parent_smoke grade*parent_smoke money*parent_smoke parent_work*parent_smoke friends_smoke*parent_smoke

/ col chisq; strata Stratum; cluster PSU;

weight weight;

run;

*================== Table 2: Logistic regression ===============;

*1. crude association between smoking and selected covariates;

*1.a parent smoking behavior at least one; proc surveylogistic data = sen1;

strata Stratum; cluster PSU; weight weight;

class parent_smoke(ref= "None")/ param=ref; model smoking(event='Yes') = parent_smoke;

run;

*1.b sex;

proc surveylogistic data = sen1; strata Stratum;

cluster PSU; weight weight;

class sex(ref = "F")/ param=ref; model smoking(event='Yes') = sex;

run;

*1.c grade;

proc surveylogistic data = sen1; strata Stratum;

cluster PSU; weight weight;

class grade(ref= "6th grade")/ param=ref; model smoking(event='Yes') = grade;

run;

*1.d pocket money;

proc surveylogistic data = sen1; strata Stratum;

cluster PSU; weight weight;

class money(ref= "Usually no money")/ param=ref; model smoking(event='Yes') = money;

run;

*1.e Parent Work Status;

proc surveylogistic data = sen1; strata Stratum;

cluster PSU; weight weight;

class parent_work(ref= "No")/ param=ref;

model smoking(event='Yes') = parent_work; run;

*1.f Friends smoking;

proc surveylogistic data = sen1; strata Stratum;

cluster PSU; weight weight;

class friends_smoke(ref= "None")/ param=ref; model smoking(event='Yes') = friends_smoke;

run;

*2. adjusted model with selected covariates; proc surveylogistic data = sen1;

strata Stratum; cluster PSU; weight weight;

class parent_smoke(ref= "None") sex(ref= "F")

grade(ref= "6th grade") money(ref= "Usually no money") parent_work(ref= "No")

friends_smoke(ref= "None")/ param=ref;

model smoking(event='Yes') = parent_smoke sex grade money parent_work friends_smoke;

run;

*3. Assessing effect measure modification on the multiplicative scale; proc surveylogistic data = sen1;

strata Stratum; cluster PSU; weight weight;

class parent_smoke(ref= "None") sex(ref= "F")

grade(ref= "6th grade") money(ref= "Usually no money") parent_work(ref= "No")

friends_smoke(ref= "None")/ param=ref;

model smoking(event='Yes') = parent_smoke sex grade money parent_work friends_smoke parent_smoke*sex parent_smoke*money parent_smoke*friends_smoke;

run;

References

‌Albers, A. B., Biener, L., Siegel, M., Cheng, D. M., & Rigotti, N. (2008). Household smoking bans and adolescent antismoking attitudes and smoking initiation: findings from a longitudinal study of a Massachusetts youth cohort. Am J Public Health, 98(10), 1886-1893. doi: 10.2105/ajph.2007.129320

‌Alves, J., & Perelman, J. (2022). Intergenerational transmission of parental smoking: when are offspring most vulnerable? , 32(5), 741-746. doi: 10.1093/eurpub/ckac065

‌Barry, A., Impouma, B., Wolfe, C. M., Campos, A., Richards, N. C., Kalu, A., . . . Farham, B. (2025). Non- communicable diseases in the WHO African region: analysis of risk factors, mortality, and responses based on WHO data. Scientific Reports, 15(1), 12288. doi: 10.1038/s41598-025- 97180-3‌

‌Bigna, J. J., & Noubiap, J. J. (2019). The rising burden of non-communicable diseases in sub-Saharan Africa. The Lancet Global Health, 7(10), e1295-e1296. doi: 10.1016/S2214-109X(19)30370-5

Centers for Disease Control and Prevention (CDC), M. o. H. a. S. A. S., World Health Organization (WHO),. (2020). Senegal Global Youth Tobacco Survey 2020 [Survey]. Retrieved from: https://extranet.who.int/ncdsmicrodata/index.php/catalog/913/get-microdata

‌Cham, B., Weaver, S. R., Jones, C. K., Popova, L., & Jacques, N. (2024). Prevalence and associated factors of shisha smoking among students in Senegal: Global Youth Tobacco Survey 2020. Tob Induc Dis,

22. doi: 10.18332/tid/186656

‌Choi, S. H., & Stommel, M. (2017). Impact of Age at Smoking Initiation on Smoking-Related Morbidity and All-Cause Mortality. American Journal of Preventive Medicine, 53(1), 33-41. doi: https://doi.org/10.1016/j.amepre.2016.12.009

‌Doherty, W. J., & Allen, W. (1994). Family functioning and parental smoking as predictors of adolescent cigarette use: A six-year prospective study. Journal of Family Psychology, 8(3), 347.

‌Ebraima Manneh. (2008). A Global Youth Tobacco Survey (GYTS) Country Report (2008). Gambia.‌

Egbe, C. O., Petersen, I., Meyer-Weitz, A., & Oppong Asante, K. (2014). An exploratory study of the socio- cultural risk influences for cigarette smoking among Southern Nigerian youth. BMC Public Health, 14(1), 1204. doi: 10.1186/1471-2458-14-1204

‌Goldenberg, M., Danovitch, I., & IsHak, W. W. (2014). Quality of life and smoking. The American journal on addictions, 23(6), 540-562.

‌James, P. B., Bah, A. J., Kabba, J. A., Kassim, S. A., & Dalinjong, P. A. (2022). Prevalence and correlates of current tobacco use and non-user susceptibility to using tobacco products among school-going adolescents in 22 African countries: a secondary analysis of the 2013-2018 global youth tobacco surveys. Archives of Public Health, 80(1), 121. doi: 10.1186/s13690-022-00881-8

‌Kondo, T., Nakano, Y., Adachi, S., & Murohara, T. (2019). Effects of tobacco smoking on cardiovascular disease. Circulation Journal, 83(10), 1980-1985.

‌Li, Y., & Hecht, S. S. (2022). Carcinogenic components of tobacco and tobacco smoke: A 2022 update.

Food and Chemical Toxicology, 165, 113179. doi: https://doi.org/10.1016/j.fct.2022.113179

‌Liu, J., Zhao, S., Chen, X., Falk, E., & Albarracín, D. (2017). The influence of peer behavior as a function of social and cultural closeness: A meta-analysis of normative influence on adolescent smoking initiation and continuation. Psychol Bull, 143(10), 1082-1115. doi: 10.1037/bul0000113

‌Loukola, A., Hällfors, J., Korhonen, T., & Kaprio, J. (2014). Genetics and smoking. Curr Addict Rep, 1(1), 75-82. doi: 10.1007/s40429-013-0006-3

‌Lu, W., Aarsand, R., Schotte, K., Han, J., Lebedeva, E., Tsoy, E., . . . Halpin, D. M. (2024). Tobacco and COPD: presenting the World Health Organization (WHO) tobacco knowledge summary.

Respiratory Research, 25(1), 338.

‌Nichter, M. (2003). Smoking: what does culture have to do with it? Addiction, 98(s1), 139-145. doi: https://doi.org/10.1046/j.1360-0443.98.s1.9.x

‌Otten, R., Engels, R. C., van de Ven, M. O., & Bricker, J. B. (2007). Parental smoking and adolescent smoking stages: the role of parents' current and former smoking, and family structure. J Behav Med, 30(2), 143-154. doi: 10.1007/s10865-006-9090-3

‌Robalino, J. D., & Macy, M. (2018). Peer effects on adolescent smoking: Are popular teens more influential? PLoS One, 13(7), e0189360. doi: 10.1371/journal.pone.0189360

‌Shah, R. S., & Cole, J. W. (2010). Smoking and stroke: the more you smoke the more you stroke. Expert review of cardiovascular therapy, 8(7), 917-932.

‌Stämpfli, M. R., & Anderson, G. P. (2009). How cigarette smoke skews immune responses to promote infection, lung disease and cancer. Nature Reviews Immunology, 9(5), 377-384.

‌Vuolo, M., & Staff, J. (2013). Parent and child cigarette use: a longitudinal, multigenerational study.

Pediatrics, 132(3), e568-577. doi: 10.1542/peds.2013-0067

‌Weber, A. M., Cislaghi, B., Meausoone, V., Abdalla, S., Mejía-Guevara, I., Loftus, P., . . . Rao Gupta, G. (2019). Gender norms and health: insights from global survey data. The Lancet, 393(10189), 2455-2468. doi: 10.1016/S0140-6736(19)30765-2

‌Wilson, S. M., & Ngige, L. W. (2006). Families in Sub-Saharan Africa. Families in global and multicultural perspective, 247-273.

‌Yu, Z., Qin, W., & Li, J. (2022). Intergenerational transmission of parental risky health behaviors in Chinese children: Are there socioeconomic status differences? Front Med (Lausanne), 9, 842817. doi: 10.3389/fmed.2022.842817

‌Zubizarreta, M. L., Mezquita, M. Á. H., García, J. M. M., & Ferrero, M. B. (2017). Tobacco and diabetes: clinical relevance and approach to smoking cessation in diabetic smokers. Endocrinología, Diabetes y Nutrición (English ed.), 64(4), 221-231.