Original article
Adolescence and Later Life Disease Burden: Quantifying the Contribution of Adolescent Tobacco Initiation From Longitudinal Cohorts

https://doi.org/10.1016/j.jadohealth.2017.02.011Get rights and content

Abstract

Purpose

Adolescence is a time of initiation of behaviors leading to noncommunicable diseases (NCDs). We use tobacco to illustrate a novel method for assessing the contribution of adolescence to later burden.

Methods

Data on initiation of regular smoking during adolescence (10–19 years) and current adult smoking were obtained from the 1958 British Birth Cohort, the U.S. National Longitudinal Study of Adolescent Health (Add Health), the Pelotas 1982 Birth Cohort, and the Victorian Adolescent Health Cohort Study. We estimated an “adolescent attributable fraction” (AAF) by calculating the proportion of persisting adult daily smoking initiated < age 20 years. We used findings to estimate AAFs for >155 countries using contemporary surveillance data.

Results

In the 1958 British Birth Cohort, 81.6% of daily smokers at age 50 years initiated < age 20 years, with a risk ratio of 6.1 for adult smoking related to adolescent initiation. The adjusted AAF was 69.1. Proportions of smokers initiating <20 years, risk ratio, and AAFs were 83.3%, 7.0%, and 70.4% for Add Health; 75.5%, 3.7%, and 50.2% in Victorian Adolescent Health Cohort Study; and 70.9%, 5.8%, and 56.9% in Pelotas males and 89.9%, 6.4%, and 75.9% in females. Initiation <16 years resulted in the highest AAFs. Estimated AAFs globally ranged from 35% in China to 76% in Argentina.

Conclusions

The contribution of adolescent smoking initiation to adult smoking burden is high, suggesting a need to formulate and implement effective actions to reduce smoking initiation in adolescents. Similar trends in other NCD risks suggest that adolescents will be central to future efforts to control NCDs.

Section snippets

Methods

Three sets of analysis were undertaken. We first used data from four national or regional longitudinal cohorts which provided data on age of initiation of smoking from adolescent waves and data on adult smoking beyond age 25 years.

The British National Childhood Development Study (NCDS) [11] is a nationally representative birth cohort which followed all UK children born in 1 week in 1958 (n = 17,638) and followed them up frequently till age 50 years (2008).

The U.S. National Longitudinal Study of

Results

Data on smoking adulthood by smoking status in adolescence (10–19 years) are shown for each cohort (the NCDS at 50 years and persistent adult smoking, Add Health, VACHS, and Pelotas cohorts) in Table 1. For each cohort, the table shows the proportion of adult smokers who initiated regular smoking in adolescence <20 years together with the associated risk ratio (RR) for being an adult smoker related to adolescent smoking, the AAF for adult smoking, and the adjusted AAF (adjusted for

Discussion

The contribution of adolescence to adult smoking burden is high, with 50%–72% of adult smoking related to adolescent initiation in the longitudinal cohorts studied here. Initiation in young people ≤ 16 years was more strongly associated with adult smoking burden than initiation later in adolescence. We found effects to be similar across older and more contemporary cohorts in high- and middle-income countries. Our findings in these cohorts took into account recent secular trends toward fewer

Acknowledgments

The first draft of this article was written by R.V. Authors’ contributions: R.V. and D.H. had the idea for the study. R.V. led the writing of the study and undertook the analyses of all cohorts except the Add Health cohort. D.H. undertook the analyses in the Add Health cohort and contributed to writing the article. J.M. and B.H. provided the Pelotas data, assisted in analyses of the Pelotas data and contributed to writing the article. A.M. and G.P. contributed to the ideas underlying the

References (40)

  • R. Campbell et al.

    An informal school-based peer-led intervention for smoking prevention in adolescence (ASSIST): A cluster randomised trial

    Lancet

    (2008)
  • G.C. Patton et al.

    Pubertal transitions in health

    Lancet

    (2007)
  • S. Greenland

    Concepts and pitfalls in measuring and interpreting attributable fractions, prevented fractions, and causation probabilities

    Ann Epidemiol

    (2015)
  • L. Degenhardt et al.

    Toward a global view of alcohol, tobacco, cannabis, and cocaine use: Findings from the WHO World Mental Health Surveys

    Plos Med

    (2008)
  • Global health risks: Mortality and burden of disease attributable to selected major risks

    (2009)
  • M.C. Seto et al.

    Sexual coercion experience and sexually coercive behavior: A population study of Swedish and Norwegian male youth

    Child Maltreat

    (2010)
  • T. Norstrom et al.

    Associations between adolescent heavy drinking and problem drinking in early adulthood: Implications for prevention

    J Stud Alcohol Drugs

    (2012)
  • J. Elliot et al.

    Now we are 50: Key findings from the National Child Development Study

    (2008)
  • Add Health Study: About Public-Use Data

  • F.C. Barros et al.

    [Methodology of the Pelotas birth cohort study from 1982 to 2004-5, Southern Brazil]

    Revista de saude publica

    (2008)
  • Cited by (8)

    View all citing articles on Scopus

    Conflicts of Interest: The authors have no conflicts of interest to disclose.

    Ethical review: No new ethical review permissions were required for these secondary data analyses of existing historical cohorts.

    View full text