Can family-based interventions prevent young people from starting to smoke?


Early uptake of cigarette smoking is associated with heavier smoking, greater nicotine dependency, increased mortality and reduced likelihood of quitting (Royal College of Physicians, 2010). Evidence indicates that the majority of adult smokers (two-thirds) report that they took up smoking before the age of 18 (Robinson and Bugler, 2010).

Currently it is estimated that 207,000 children in the UK start smoking each year (Hopkinson et al., 2013). It is therefore important that interventions to prevent smoking target children and adolescents at this critical time for smoking initiation. Interventions aimed at young people are likely to impact positively on the pattern of tobacco-related illness and deaths in coming decades.

Research has shown that a wide range of factors influence smoking uptake in children and adolescents, including gender, concerns with body weight, attitudes toward smoking, peer group and parental/sibling smoking (Mayhew et al., 2000). In particular children with parents or siblings that smoke are three times more likely to become a smoker themselves (Leonardi-Bee et al., 2011). Parenting style and restrictions on smoking at home have also been shown to impact the likelihood of experimenting with smoking (Jackson et al., 1998; Proescholdbell et al., 2000).

Given the evidence for the likely impact of childhood initiation of smoking on several smoking outcomes and the documented role of parental and sibling factors on early uptake, family interventions may represent an opportunity to prevent children and adolescents from starting to smoke. However, evidence for such interventions is yet to be evaluated.

The recent Cochrane systematic review discussed here (Thomas et al., 2015) aimed to examine the effectiveness of family interventions designed to strengthen non-smoking attitudes and promote non-smoking to influence children and adolescents to not smoke.

It's estimated that 207,000 children in the UK start smoking each year.

It’s estimated that 207,000 children in the UK start smoking each year.


The authors searched the Cochrane Tobacco Addiction Group Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, PsycINFO, CINAHL, Web of Science and ERIC databases for studies.

Studies were deemed suitable for inclusion if they:

  • Reported randomising student/families to either an intervention or a no intervention control group
  • Included appropriate methods of group allocation and outcome measure (cross-sectional studies were excluded)
  • Included children (aged 5-12 years) or adolescents (aged 13-18 years)
  • Examined family-based interventions designed to deter the start of tobacco use
  • Examined school/community-based interventions in addition to family interventions
  • Examined other substance use, providing outcome measures of tobacco use were included
  • Had a primary outcome measure of the effect of the intervention on the smoking status of children or adolescents who reported a non-smoking status at baseline

Studies were excluded if they:

  • Did not conduct a baseline measure of smoking status
  • Did not measure smoking behaviour
  • Did not distinguish between the effects of family-based and other interventions
  • Focused on cessation rather than prevention
  • Had a follow-up period of less than 6 months


Using these criteria, 27 studies were selected for the review; 12 were Randomised Controlled Trials (RCTs) and 15 were cluster RCTs (C-RCTs).

Study characteristics

23 of these studies were conducted in the USA, and 1 in each of the following countries: India, Australia, Netherlands and Norway. Interventions in each of the studies were classified as low, medium or high intensity, according to the level of the family component in the intervention.

The focus of the 23 studies was varied:

  • 15 studies examined substance use prevention:
    • 6 on tobacco
    • 1 on alcohol
    • 1 on general substance use
    • 3 on alcohol, tobacco and marijuana
    • 2 on tobacco and cardiovascular health
  • 2 studies assessed HIV and unsafe sex
  • 10 studies looked at family functioning, child development and adolescent behaviour change

Despite the varied focus of included studies, interventions were similar in their components of the family-based intervention used.

Family-based intervention vs no intervention

9 studies with 4,810 participants examined the effects of family-based interventions versus no intervention on smoking behaviour. Data from 8 studies could not be included due to insufficient information for analysis.

  • Family-based interventions showed a significant benefit in preventing experimentation and uptake of regular smoking compared with a no intervention control
  • Findings suggest that family interventions could reduce the number of adolescents who try smoking by 16-32%
  • Most of these studies were classified as high intensity interventions; however, subgroups of low and medium intensity interventions yielded similar results
  • 2 studies with 4487 participants, where some participants had experience of smoking at baseline, did not detect an effect of intervention.
The review found that, compared with no intervention, family-based intervention could reduce the number of adolescents who try smoking by 16-32%.

The review found that, compared with no intervention, family-based intervention could reduce the number of adolescents who try smoking by 16-32%.

Family-based intervention vs Family-based and School-based intervention

Of 8 studies that assessed combined family-based and school-based interventions versus family-based interventions alone. 2 studies with 2,301 participants included only never smokers at baseline.

  • These studies indicated that the addition of a family-based interventions to school-based interventions may reduce the onset of smoking by 4-25%
  • One study with 1,096 participants, which included children who had smoked at baseline also detected a benefit of including family-based interventions
  • Of the five remaining studies, insufficient data was available for analysis
  • One study (included in the family versus no intervention analysis above), also examined a combined school/family intervention compared with a “good behaviour” school intervention, indicating no difference between these interventions

The authors observed that a common feature of effective interventions was encouraging authoritative parenting style (showing strong interest and care for children and adolescents, with rule setting).

The review found that adding family-based interventions to school-based interventions may reduce the onset of smoking by 4-25%.

The review found that adding family-based interventions to school-based interventions may reduce the onset of smoking by 4-25%.


These data add further evidence to the notion that families influence smoking initiation in children and adolescents. The results examined here suggest a benefit of family-based interventions for preventing children and adolescents from starting to smoke. The evidence in this review is strongest for high intensity, family-based interventions that are independent of school-based programmes. Additionally, encouraging authoritative parenting was a common feature amongst effective interventions.

However, the authors do caution that data could not be included from all studies initially included in the review. Data from 13 studies were not sufficient for analysis. Additionally, given that many of the RCTs included in the review did not provide detailed study methods, the authors further caution that they cannot be confident that results were not biased. As a consequence the quality of the evidence should be viewed as moderate.

Of the 14 studies that were eligible for analysis, the majority varied considerably with regards to follow-up period (6 months to 15 years), focus of the programme (HIV, general substance use, tobacco only) and the intensity of the intervention offered.

In addition, variation in estimates of reduction in smoking experimentation and onset of smoking were extremely broad (e.g. 16-32% for family interventions versus no intervention and 4-25% for combined family/school based interventions respectively). Therefore, it is difficult to draw strong conclusions across a relatively small and highly varied group of studies.

Further research is required to examine whether family-based interventions can prevent smoking in children and adolescents, in order to determine the effectiveness and cost benefit of these programmes.

It's difficult to draw any hard and fast conclusions from systematic reviews that contain small and highly varied group of studies.

It’s difficult to draw any hard and fast conclusions from systematic reviews like this one, which contain a small and highly varied group of studies.


Primary study

Thomas RE, Baker PRA, Thomas BC, Lorenzetti DL. Family-based programmes for preventing smoking by children and adolescents. Cochrane Database of Systematic Reviews 2015, Issue 2. Art. No.: CD004493. DOI: 10.1002/14651858.CD004493.pub3.

Other references

Hopkinson et al (2013) Child uptake of smoking by area across the UK. Thorax. Thorax thoraxjnl-2013-204379

Action on Smoking and Health (2014) Young People and Smoking. Young People and Smoking (PDF).

Robinson and Bugler (2010) Smoking and drinking among adults (PDF). General Lifestyle Survey 2008.

Ferrence R (2010) Passive smoking and children (Editorial). BMJ 2010;340:c1680

Mayhew KP et al (2000) Stages in the development of adolescent smoking. Drug and Alcohol Dependence, Volume 59 , 61 – 81 [abstract]

Leonardi-Bee et al (2011) Exposure to parental and sibling smoking and the risk of smoking uptake in childhood and adolsecence: a systematic review and meta-analysis. Thorax 2011;66:847-855 doi:10.1136/thx.2010.153379

Jackson et al (1998) The authoritative parenting index: predicting health risk behaviours among children and adolescents. Health Education and Behaviour. [abstract]

Proescholdbell et al (2000) Home smoking restrictions and adolescent smoking. Nicotine & Tobacco Research. Vol 2 Issue 2  159-167 [abstract]

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+