A methodological and substantive review of the evidence that schools cause pupils to smoke

https://doi.org/10.1016/j.socscimed.2003.08.012Get rights and content

Abstract

The objectives of this review were to examine whether smoking prevalence varies between schools independently of health promotion programmes and pupil composition, to show which school characteristics are responsible for this variation, and to examine the methodological adequacy of such studies. Searches for published studies were performed on medical, educational and social science databases, relevant articles’ reference lists, and citation searches. Any study was included that described inter-school variation in smoking prevalence, or related such variation to school characteristics. A model relating pupil smoking to school, neighbourhood, and pupil characteristics unlikely and likely to be influenced by school was used to examine the adequacy of control of confounding by pupil composition. Data from studies were combined qualitatively considering methodological adequacy to examine the relation of smoking prevalence to school characteristics. Theoretical frameworks underpinning the choice of school characteristics and postulated relationships between these characteristics and smoking prevalence were described. There were large variations in smoking prevalence between ostensibly similar schools. Evidence that pupil composition did not cause this was weak, because all studies had methodological problems, including under control of relevant pupil compositional factors and over control of factors likely to represent the mechanism through which schools influence pupils’ smoking. There was little evidence that elements of tobacco control policy other than bans and enforcement deterred smoking. Academic practice and school ethos were related to smoking. Academically selective schools did not influence smoking, once pupil composition was controlled. There was one study on neighbourhood influences, which were unrelated to smoking. Studies frequently offered little or no theoretical justification for associating school characteristics with smoking. Some aspects of school influence pupils’ smoking, probably independently of pupil composition. However, under-control and over-control of confounding and lack of theoretical underpinning precludes definitive conclusions on how particular school characteristics influence pupils’ smoking.

Introduction

This review summarises evidence from observational studies that some schools not operating specific interventions appear to inhibit smoking whilst other schools with equivalent pupils appear to promote smoking. Two types of observational studies are included. Ecological studies measure smoking prevalence at the school-level and relate this to school characteristics or aggregated pupil characteristics. Multilevel studies validly relate school-level characteristics to the pupil-level outcome of smoking.

An epidemiological framework for assessing causality from observational studies is described by Hennekens and Buring (1987). In this framework, there should be evidence that association is not due to chance, bias, or confounding. Associations that are strong, credible, consistent across studies, where cause precedes effect, and a where a dose–response relationship exists support causality. In particular, we examine confounding by pupil characteristics as the main threat to the validity of observed associations. If pupil characteristics explain inter-school smoking variation, this implies adolescent risk factors for smoking that are not of the schools’ making, such as parental smoking, are common amongst pupils in some schools but not others. School-level characteristics include, for example, tobacco control policies and school culture. Macintyre, Ellaway, and Cummins (2002) describe differences in pupil characteristics between schools as compositional characteristics, and school-level characteristics as contextual and collective characteristics.

To date, school-based anti-smoking interventions have had very limited success (Peterson, Kealey, Mann, Marek, & Sarason, 2000) If school-level contextual/collective factors influence pupils’ smoking, this could generate new school smoking prevention programmes. The search for the attributes of schools and schooling that influence pupils’ smoking is the first stage of this development.

Section snippets

Data sources

To conduct this review, we searched Medline, Embase, Psychlit, the British Education Index (BEI), ERIC (an educational database), and the Science and Social Science Citation databases using index terms and text words, and searched our own library of papers. In Medline, we used the index terms ‘culture’ (exploded), ‘institutional management teams’ (exploded), ‘smoking’ (exploded) and text words either ‘college’ or ‘school’. In Embase, we used index terms organisation (exploded), sociology

Does smoking prevalence vary by school?

Five UK studies reported on variation between schools (Table 1) Smoking prevalence increases rapidly as adolescents move through the secondary school (ages 11–16 years), which influences measures of variation between schools. To provide a common index of variation, we calculated a ‘school effect’ odds ratio (OR). This OR is the odds of smoking in a school one standard deviation above the average school divided by the odds of smoking in a school with average smoking prevalence, calculated from

Conclusion

This review has documented that smoking prevalence varies markedly between schools, but what is the evidence that this is caused by school factors? For the most part, where associations have been significant, chance does not provide an adequate explanation of inter-school smoking prevalence variation. Selection bias is unlikely because most studies used appropriate sampling and it cannot explain variation per se. Information bias is more possible. Particularly, for example, in schools with

References (41)

  • G Davey-Smith et al.

    Confounding in epidemiological studiesWhy “independent” effects may not be all they seem

    British Medical Journal

    (1992)
  • C de Moor et al.

    The association between teacher attitudes, behavioral intentions, and smoking and the prevalence of smoking among seventh-grade students

    Adolescence

    (1992)
  • S.T Ennett et al.

    The contribution of influence and selection to adolescent peer group homogeneityThe case of adolescent cigarette smoking

    Journal of Personality & Social Psychology

    (1994)
  • S.T Ennett et al.

    School and neighborhood characteristics associated with school rates of alcohol, cigarette, and marijuana use

    Journal of Health & Social Behavior

    (1997)
  • M Galt et al.

    Smoking prevalence among 15–16 year olds in Doncaster, England

    Journal of Public Health Medicine

    (1994)
  • C Hagquist

    Variations in adolescents’ smoking and alcohol behaviour between Swedish schools—an ecological analysis

    Drugs: Education, Prevention and Policy

    (1997)
  • C.H Hennekens et al.

    Epidemiology in medicine.

    (1987)
  • W.W Holland et al.

    Indications for measures to be taken in childhood to prevent chronic respiratory disease

    Millbank Memorial Quarterly

    (1969)
  • M.J Jarvis et al.

    Children's purchases of single cigarettesEvidence for drug pushing?

    British Journal of Addiction

    (1990)
  • M.R Johnson et al.

    Schools and smokingSchool features and variations in cigarette smoking by children and teachers

    British Journal of Educational Psychology

    (1985)
  • Cited by (117)

    • The association between smoke-free school policies and adolescents’ anti-smoking beliefs: Moderation by family smoking norms

      2019, Drug and Alcohol Dependence
      Citation Excerpt :

      The injunctive and descriptive norms were combined into an overall family smoking norms score that ranged from 0 to 8 and was dichotomised by using a median split (0–6 and 7–8), representing individuals with relatively permissive and prohibitive family smoking norms, respectively. Age (in years), gender (male vs female), city, parents’ migration background (0, 1, or 2 parents not from country of residence), mother’s educational level and father’s educational level (low, middle, high or unknown) were measured as potential confounders (Aveyard et al., 2004). The education level of parents was assigned country-specific categories that were later standardized into low-, middle- and high-education.

    • Role theory of schools and adolescent health

      2019, The Lancet Child and Adolescent Health
      Citation Excerpt :

      Mental health and behavioural problems have a profound effect on student engagement and educational attainment.2 Conversely, schools have an equally substantial effect on students' health, with great variation existing between schools in terms of rates of violence, and use of tobacco, alcohol, and other drugs: differences that reflect school characteristics rather than merely differences between the students entering schools.3–5 A range of theories have been suggested for the health benefits observed in students, including avoidance of health risks associated with early entry into labour markets, sexual activity, or criminality;6 education of young people about how to promote their physical and mental health;7 and enhanced provision of nurturing social environments, with growing evidence that interventions that modify the school environment also improve student mental and physical health across a range of outcomes.8–10

    View all citing articles on Scopus
    View full text