Most smokers start in adolescence. Therefore, policies that restrict the sale of tobacco products by age have the potential to prevent smoking and related harm. We undertook a scoping review to identify evidence for two potential policies to restrict the sale and supply of tobacco to younger people:

• Raising the legal smoking age each year to create a tobacco-free generation (TFG).
• Increasing the minimum legal age of sale (MLA) to 21 (MLA21) or 25 (MLA25).

We undertook searches of eight bibliographic databases and an advanced Google search for peer-reviewed and grey literature to July 2023. We included English-language studies on the effect of age-restriction interventions for tobacco on any health and non-health outcome among young people up to the age of 25. We only looked at evidence for combustible cigarettes and nicotine vapour products (NVPs, an umbrella term that includes e-cigarettes and refill containers (e-liquids) intended for nicotine vaping). We did not conduct critical appraisal of the evidence due to time constraints and the nature of the review, but we have summarised the types of evidence available and high-level findings. The scoping report should be used as a precursor to an evidence review or other evidence outputs.

What we found
Introducing a tobacco-free generation (TFG) We identified six studies that looked at TFG as an endgame tobacco control measure, either alone or in combination with other strategies.

We found:

• No primary studies investigated the short-term effectiveness of TFG on health or non-health outcomes.
• Evidence from four modelling studies indicates that a TFG policy could reduce smoking prevalence, increase health-adjusted life years and offer cost savings. However, these benefits may take several decades to be realised.
• There is evidence from three modelling studies that combined tobacco-control strategies which include TFG could be more effective at reducing smoking prevalence than TFG alone.
• There is evidence from two modelling studies that TFG could reduce inequalities by sex and ethnicity.

Increasing the minimum legal age of sale to 21 (MLA21)
We found 39 studies that looked at MLA21, of which 34 relate to the United States (US), where Tobacco 21 (T21) legislation has been implemented variably over time.

We found:

• Most of the observational evidence and all modelling evidence found MLA21 reduced, or was projected to reduce, combustible tobacco cigarette prevalence. US observational evidence for the impact of MLA21 on NVP prevalence was mixed. There is some evidence to suggest variation in effect of MLA21 on smoking prevalence rates by ethnicity, sex, rurality, socioeconomic status and sexual orientation.
• US observational evidence is inconsistent for combustible cigarettes and NVP initiation among children and young people but modelling data from US and Australia suggest MLA21 reduced or could reduce initiation rates.
  There is some US evidence that MLA21 reduces smoking among pregnant young mothers and improves neonatal outcomes.
• There is limited evidence that MLA21 could prevent future cardiovascular events.
• Some modelling evidence forecast reductions in US healthcare costs due to reductions in smoking and burden of cardiovascular diseases, with differential effects by ethnicity and sex.
• Most US evidence found that MLA21 successfully reduced purchasing of tobacco products and NVPs. However, there was some evidence of retailer non-compliance, cross-border purchasing and under 21s accessing products through older people in their social networks. Young people obtained tobacco and NVPs from sources other than directly from retailers, which could reduce the effectiveness of MLA21.
• Evidence suggests that young people living in places with MLA21 may be more likely to perceive combustible cigarettes as riskier than those in regions with a lower minimum sales age. There is evidence that inclusion of NVPs in MLA21 in some US states led to a greater increase in young people's perceptions of the risks associated with NVPs than with combustible tobacco cigarettes.
• US evidence suggests that young people were largely aware of, and supportive of, MLA21 policies, apart from some doubts about its effect on consumption.
• Current smokers were less optimistic about effectiveness than non-smokers. Retailers were highly aware of the intervention, predominantly supportive, and typically thought it was easy to comply with.
  We found limited evidence that knowledge of MLA21 was associated with reduced intention to use all tobacco products.

Key points for consideration

• Our scoping review does not make overall conclusions about the effectiveness of TFG and MLA21 policies. It would therefore be useful to consider whether further, more robust and specific, evidence synthesis would be required.
• Modelling studies are not a substitute for evaluation of policy implementation but provide useful evidence for policies that have not yet been tried (i.e. TFG); for the potential impact of a policy in isolation or in combination with other tobacco control policies; and for long-term outcomes.
• Quality appraisal of included studies was not undertaken in this scoping review; many of the eligible studies used designs that were not well suited to policy evaluation. Consideration should be given to what quasi-experimental designs would be suitable and feasible to evaluate key health outcomes.
• Collecting evidence for non-health outcomes (e.g. purchasing behaviour, knowledge and attitudes, compliance) should be considered as part of a programme of policy evaluation.