Research and analysis

Nicotine vaping in England: 2022 evidence update summary

Published 29 September 2022

Applies to England

Chapter 1. Introduction

Objective of the report

This report is the eighth in a series of independent reports originally commissioned by Public Health England (PHE) and now the Office for Health Improvement and Disparities (OHID) in the Department of Health and Social Care. The series aims to summarise the evidence on vaping products and to inform policies and regulations.

Alternative nicotine delivery devices such as vaping products can play a vital role in reducing the huge health burden caused by cigarette smoking, which remains:

• the largest single risk factor for death and years of life lived in ill-health globally
• a leading cause of health inequalities in England
• the second most important risk factor for death and disability-adjusted life years globally

Issues addressed

This current report focuses predominantly on the potential health risks of vaping. We carried out a systematic literature review of the health risks and health effects of vaping and divided the findings into chapters. These include:

• biomarkers of exposure to nicotine and potential toxicants
• biomarkers of potential harm to health cutting across several diseases, including cancer, respiratory and cardiovascular diseases
• biomarkers specifically associated with cancer, respiratory, cardiovascular or other health outcomes
• poisonings, fires and explosions
• nicotine
• flavours

This report also covers the latest evidence on prevalence and characteristics of vaping in young people and adults in England, with a focus on the data emerging since our last report published in early 2021. It looks at the prevalence of heated tobacco product use in England, incorporating a summary of the latest Cochrane Review on heated tobacco products, and a new systematic review on harm perceptions of vaping products and interventions to affect perceptions.

Our report does not cover 2 important issues. We felt these issues were either being addressed comprehensively elsewhere or had been covered in our previous reports. So, we did not examine:

1. The relationship between vaping and subsequent smoking. This is because a new Cochrane review on electronic cigarettes and subsequent cigarette smoking in young people is examining the existing literature about this, among people under 30 years old.

2. The evidence for the effectiveness of vaping to help people who smoke quit. We have covered this topic in our previous report, and the Cochrane collaboration has an ongoing (updated monthly) systematic review on electronic cigarettes for smoking cessation. This Cochrane review examines the effectiveness of using electronic cigarettes to help people who smoke tobacco achieve long-term smoking abstinence and searches for updates of the evidence monthly.

Throughout our report, we have also tried to reflect on changes in England since our first report in 2015. This may also help to understand underlying trends, given the influence of COVID-19 recently on the availability of data and on smoking and vaping behaviours.

Terminology

As in our 2020 and 2021 reports, we use the term ‘vaping products’ to describe e-cigarettes and refill containers (e-liquids) intended for nicotine vaping. Some vaping products do not always contain nicotine. Where studies explored products without nicotine, we refer to them as non-nicotine vaping or vaping products.

We use ‘vapers’ to refer to people who regularly use vaping products and ‘vaping’ as the act of using a vaping product. These terms do not include cannabis vaping or the vaping of other legal or illegal substances, which are not the subject of this report.

Vaping regulations and guidance

Here we summarise the main regulations in England governing vaping products and their surveillance, the Medicines and Healthcare products Regulatory Agency (MHRA) safety monitoring, relevant government strategies and consultations, recent reports on regulations, the National Institute for Health and Care Excellence (NICE) tobacco guideline, as well as selected international developments.

Main findings

Regulations and licensing

Vaping products containing nicotine are regulated under the Tobacco and Related Products Regulations 2016 (TRPR), and need to be notified to the MHRA and comply to certain standards (for example, nicotine content is limited to 20 milligrams per millilitre (mg/mL)) before they can be legally sold in the UK. An analysis of notifications in 2016 to 2017 found that notified products were unlikely to cause serious harm.

Vaping products that do not contain nicotine come under the General Product Safety Regulations 2005, enforced by local authority trading standards.

Medicinally licensed vaping products are exempt from the TRPR and currently there is no licensed product in the UK. Although, in October 2021, MHRA published updated guidance to provide clearer information on the process and help speed up review times.

Adverse advents

MHRA also collects information on adverse events believed to be associated with vaping products containing nicotine through its Yellow Card scheme. Between 20 May 2016 (implementation of TRPR) and 13 January 2022, MHRA received 257 reports of adverse reactions (26 of those since January 2021). Each report represents an individual for whom more than one adverse reaction could have been reported. A report is not proof that the reaction was caused by a vaping product, just that the reporter thought it might have been.

Since January 2021, the MHRA has considered 14 of the reports as serious and no fatalities were reported.

Adverse reactions to licensed smoking cessation medications are also reported to the MHRA. In 2021 there were 297 reports for nicotine replacement therapy (NRT) and 78 for varenicline. Varenicline has been unavailable since June 2021, further limiting effective pharmaceutical options for smoking cessation.

Age of sale

It is illegal to sell vaping products to anyone aged under 18 and to buy vaping products for anyone under 18. There is a loophole in the legislation allowing free samples of vaping products to be given to people of any age. Surveys by the Chartered Trading Standards Institute to capture tobacco control activities, including enforcement of age of sale vaping and tobacco product laws, have not been carried out since 2020.

A specific project in Scotland between October and December 2021 focused on single use disposable vaping products. It found that most products had not been notified as required with many above the 20mg/mL nicotine content limit. It also identified some violations of age of sale laws. A review of the age of sale legislation in the UK published in January 2021 concluded that overall, it had achieved its original goal of reducing uptake among under 18s.

Smokefree ambition

A government consultation in 2019 – Advancing our health in the 2020s – outlined a new ambition for England to be smokefree by 2030 (meaning only 5% of the population would smoke by then). It included an ‘ultimatum for industry to make smoked tobacco obsolete by 2030, with smokers quitting or moving to reduced risk products like vaping products.’

The All-Party Parliamentary Group on Smoking and Health made recommendations to help achieve the smokefree 2030 ambition. These included reducing the appeal and availability of vaping products and other nicotine products to young people and to update its guidance for medicinal licensing of vaping products.

Advertising and social media

A review of vaping product marketing in the UK between 2016 and 2019 found high compliance with the advertising code in advertisements, but not in social media posts. It found that young people who had never smoked or vaped noticed posts relating to vaping more than adults who smoked. However, compared with the US and Canada, UK regulations were found to have limited exposure to marketing among adults and young people.

Recent and upcoming developments

In March 2022, OHID published the post-implementation review of the TRPR. The review assessed whether the regulations had met their objectives. This review concluded that the evidence indicated the TRPR’s main objectives were being met, and provided a strong argument for retaining the regulations. It also proposed some amendments which could help support the government’s smokefree 2030 ambition.

A new tobacco control plan for England will be published in 2022 and is expected to outline the government’s strategy for England to become smokefree by 2030. The 2017 tobacco control plan, Towards a smoke-free generation, set out ambitions up to 2022 and remains in place, although progress towards meeting the ambitions has been mixed.

The government also commissioned an independent review of tobacco control, which was published in June 2022. It makes recommendations for the best ways to address the health inequalities caused by smoking and to achieve the smokefree 2030 ambition.

Vaping products which do not contain nicotine and are regulated through the General Product Safety Regulations 2005 are less strictly regulated than products that contain nicotine, so their regulation requires further consideration. As other non-tobacco nicotine products (such as nicotine pouches) emerge in the UK, it seems appropriate to review regulations for these products at the same time.

In November 2021, NICE published a new comprehensive guideline on tobacco, Tobacco: preventing uptake, promoting quitting and treating dependence, which includes guidance on:

• preventing uptake of smoking
• promoting quitting
• treating tobacco dependence
• discussing vaping products with patients to help prevent or stop their tobacco use

It also makes recommendations for policy, commissioning and training.

We also summarise recent international developments in vaping product policy, including in the EU, the US, Canada, Australia and New Zealand.

Implications

The smokefree 2030 ambition and developing a new tobacco control plan for England provide an opportunity to review all vaping (and other nicotine and tobacco) regulations. This will ensure that regulations are appropriate and help smokers quit, while managing the risk of uptake for people who have never smoked.

The next tobacco control plan also provides an opportunity to set out the plans needed to achieve the smokefree 2030 ambition and to set intermediate targets for smoking prevalence in different disadvantaged groups.

The continuing lack of a medicinally licensed vaping product is of concern and may require further review of the process involved.

There needs to be consideration of whether some aspects of packaging of vaping products need restricting.

The review of vaping product marketing suggests the UK needs to substantially strengthen its enforcement of marketing regulations on social media.

There is an opportunity to standardise the notification processes using the MHRA database of notified vaping products. This would enable research and help to maximise harm reduction potential.

Local authority trading standards efforts have been scaled down and compliance with regulations is not enough to prevent underage sales and access to illicit products. Also, more frequent surveillance of single-use disposable vaping products is needed. There is a danger that the reduction in local trading standards officers and restructure of the MHRA could result in a lack of surveillance of these products. This could undermine the approach and regulatory framework for vaping products adopted in England.

Lessons should be learned from the mislabelled US ‘e-cigarette, or vaping, use-associated lung injury’ (EVALI) outbreak. These lessons include the impact of miscommunications about nicotine vaping compared to vaping contaminated illicit substances. Communications about EVALI should clearly separate vaping these illicit substances from nicotine vaping. Also, communications about any future cases or outbreaks of poisonings or injuries should be clear about the implicated substances.

Chapter 2: Methods

We used data from 2 surveys for information on smoking and vaping among young people in England. These were the:

• Action on Smoking and Health-Youth (ASH-Y) survey
• International Tobacco Control Policy Evaluation Project (ITC) Youth Tobacco and Vaping survey (hereafter known as ITC Youth survey)

For information on smoking and vaping among adults in England, we used data from 4 surveys, which were:

• ONS Annual Population Survey (APS)
• Smoking Toolkit Study (STS)
• Action on Smoking and Health-Adult (ASH-A) survey
• ONS Opinions and Lifestyle Survey (OPN)

We reported NHS Digital data from stop smoking services on supported stop smoking attempts in England, and from the National Poisons Information Service and London Fire Brigade on suspected poisonings and fires caused by vaping products.

We conducted 2 systematic reviews, one on the health risks of vaping and one on vaping harm perceptions and also summarised findings from a recent Cochrane Review of heated tobacco products.

For chapters on vaping associations with health risks, we first summarised evidence from previous reports by PHE, the National Academies of Sciences Engineering and Medicine (NASEM) and the Committee on Toxicity of Chemicals in Food Consumer Products and the Environment (COT). We then presented findings from our systematic review. To summarise evidence on the health risks of vaping, we developed an algorithm to assess whether to conduct meta-analyses. Details of the algorithm are presented in table 6 in chapter 2 of the full report.

Chapter 3: Vaping among young people

Data collection

Data reported in this chapter were collected in February 2021 (from the ITC Youth survey), in March to April 2021 (from the 2021 ASH-Y survey) and we also report prevalence data from the ASH-Y 2022 survey carried out in February to March 2022.

In response to the COVID-19 pandemic, schools were closed in England between 4 January and 15 March 2021, and there were tight restrictions on social gatherings between 4 January and 19 May 2021. Although no restrictions were in place during 2022 data collection, it is likely that there are ongoing effects of the 2 years of social restrictions on young people. So, conclusions in this chapter may be greatly affected by the COVID-19 regulations.

Main findings

Smoking and vaping prevalence

2022 ASH-Y survey data (11 to 18 year olds) showed:

• smoking prevalence (including occasional and regular smoking) was 6% in 2022 (compared with 4.1% in 2021 and 6.7% in 2020)
• vaping prevalence (including occasional and regular vaping) was 8.6% in 2022 (compared with 4% in 2021 and 4.8% in 2020)

2021 ITC Youth survey data (16 to 19 year olds) showed:

• smoking prevalence (defined as smoking more than 100 cigarettes in their life and having smoked in the past 30 days) was 7.9% in February 2021 (compared with 8.5% in February 2020 and 6.2% in August 2019)
• vaping prevalence (defined as vaping on more than 10 days in their lifetime and having vaped in the past 30 days) was 9.1% in February 2021 (compared with 9.4% in February 2020, and 7.7% in August 2019)

Overall, data from the 2021 ASH-Y and ITC Youth surveys were broadly similar for comparable age categories. Vaping among 19-year-olds has been steadily increasing in the ITC Youth data over recent years.

The 2022 ASH-Y data suggests that overall nicotine use (via smoking or vaping) has increased over the past year, being 11.1% in 2022 compared with 6.2% in 2021. In 2015, the proportion was 7.7%.

Based on the socio-economic grade of 11 to 18 year olds in the 2022 ASH-Y survey, the estimates for smoking and vaping prevalence were similar for the more advantaged groups in social grades A, B and C1 (5.8% for smoking, 8.4% for vaping) to more disadvantaged groups in social grades C2, D and E (5.4% for smoking, 8.1% for vaping). This was a departure from previous years. For example, in 2021, the estimates for smoking and vaping prevalence were higher among the more advantaged groups in social grades A, B and C1 (4.6% for smoking, 4.4% for vaping) than for the more disadvantaged groups in social grades C2, D and E (2.8% for smoking, 3% for vaping), similar to ASH-Y data from previous years.

The 2022 ASH-Y data showed that most young people who had never smoked were also not currently vaping (98.3%). This was consistent with the 2021 ASH-Y and 2021 ITC Youth data, although the proportions were higher (99.2% and 99.1% respectively).

Vaping devices

Disposable models (which are pre-filled with liquid and used only once) were the most popular type of vaping device in the 2022 ASH-Y survey. These were used by 52.8% of 11 to 18 year olds who currently vaped, and 18.7% used tank models (which are reusable and rechargeable kits that users can refill with liquid). This was a stark difference from previous years, where tank models were the most popular type of vaping device. For example, in 2021, only 7.8% of current vapers reported using disposable models, whereas 41% used tank models.

COVID-19

Young people from the 2021 ITC Youth survey reported an effect of COVID-19 on smoking and vaping behaviour, which found:

• 8% of past year vapers reported quitting vaping
• 15% of past year vapers reported cutting down due to the COVID-19 pandemic

However, 15% reported vaping more as an effect of the pandemic. Similar patterns were seen among young people who had smoked in the past year, with:

• 7% reporting quitting
• 20% reporting cutting down
• 18% reporting smoking more

These findings could contribute to the slight increase in former smokers (from 0.8% to 1.7%) and former vapers (from 4.6% to 8.6%) observed in the ITC Youth data between 2019 and 2021.

Reasons for vaping

The main reasons for vaping were to “give it a try” (48.8%, 2021 ASH-Y), and “liking the flavours” (37.2%, ITC Youth). These reasons were most common among young people who have never smoked or only tried smoking. Among young people who smoked, or had smoked, in the ITC Youth survey, harm reduction, and quitting related reasons were common.

In the 2021 ASH-Y survey, most 11 to 18 year olds who had tried vaping had smoked first (38.7%), while 24.7% said they had vaped before they smoked and 29.7% said they had tried a vaping product and never tried smoking.

Flavours

Fruit flavours were the most popular among young people who currently vaped (51.5% in 2021 ASH-Y). This was followed by menthol/mint (13%), then chocolate/dessert/sweet/candy flavours (9.3%), similar to data presented in our 2021 report.

Access to vaping products

Although it is illegal to sell vaping products to under 18 year olds, many young people under the age of 18 bought and owned their own vaping devices. In the 2021 ASH-Y survey, just under a quarter (24.8%) of young people aged 11 to 17 said that they were given products by friends. But others also reported buying them, for example:

• 22.1% said they bought them from newsagents
• 22.1% said they bought them online
• 16.3% said they bought them from a supermarket

Similarly, in the ITC Youth survey, young people aged 16 to 17 who had vaped in the past 30 days commonly reported being given products (37.5%). Many also reported buying products from shops (32.1%) or online (23.3%). Nearly two thirds (64.3%) of 16 to 17 year olds from the ITC Youth survey who had vaped in the past 30 days reported they owned a vaping product.

Nicotine

About a third (34.2%) of 11 to 18 year olds in the 2021 ASH-Y survey who currently vaped or had vaped in the past reported always using vaping products that contained nicotine and 20.4% reported always using nicotine-free products. Just over two-thirds (68.9%) of 16 to 19 year olds who had vaped in the past 30 days and had ever used vaping products with nicotine, reported using nicotine in their current vaping product and 12.3% said their vaping product did not contain nicotine.

In the 2021 ITC Youth survey, the most common nicotine strength used by 16 to 19 year olds who had vaped in the past 30 days was reported to be under 20mg/mL (64%). A total of 17.2% reportedly used a strength between 20mg/mL and 49mg/mL and 5.6% reportedly used 50mg/mL or over. Compared to 2019, fewer participants reported they did not know the strength of their vaping liquid (from 19.6% to 7.3%).

About half (53.1%) of 16 to 19 year olds who vaped in the past 30 days reportedly used nicotine salt e-liquid (a nicotine version which is smoother to inhale, has lower pH and is absorbed faster into the bloodstream than freebase nicotine) similar levels to those seen in 2019 (56.6%). We also found that 40.4% did not use nicotine salts and 6.5% were unsure. This has changed compared to 2019, where 30.6% did not use salts and 12.8% were unsure. Overall, there was higher awareness of the inclusion of nicotine and type of nicotine (freebase or salt) and fewer ‘don’t know’ responses in 2021 compared to 2019.

Perceived addiction and urges to vape

Under half (42.8%) of 16 to 19 year olds in the 2021 ITC Youth survey who currently vaped did not feel addicted to vaping, but half (52.5%) said they felt a little or very addicted. In comparison, 14.5% of 16 to 19 year olds who currently smoked did not feel addicted to smoking, and 83% reported they felt a little or very addicted.

Just under half (44.5%) of 16 to 19 year olds in the 2021 ITC Youth survey who currently vaped reported experiencing urges to vape almost daily or more than daily, and 16.8% reported never experiencing an urge to vape. In comparison, 66.6% of young people who currently smoked reported urges to smoke daily or multiple times a day, and only 4.7% reported never having urges to smoke.

Just over forty per cent of 11 to 18 year olds in the 2021 ASH-Y survey who currently vaped said they did not feel any urges to vape at all (41.5%), and 23.5% reported strong, very strong or extremely strong urges to vape. In comparison, 24.3% of those who currently smoked reported no urge to smoke and 31.4% reported a strong, very strong or extremely strong urge to smoke.

Other tobacco and nicotine products

Just over one-tenth (11%) of 16 to 19 year olds in the ITC Youth survey reported ever using a waterpipe, 4% reported ever using nicotine pouches, and 5% reported ever using smokeless tobacco.

Implications

Further monitoring and research

Dependence on vaping assessed in 2021 appears lower than on smoking for young people. Further research on dependence is needed, including dependence by type of vaping product used, nicotine type and nicotine strength.

Vaping and smoking among young people appear to have decreased between 2020 and 2021 but then increased in 2022. So, it is important that trends continue to be monitored by the government. The differences in estimates between the ASH-Y and ITC Youth surveys in 2021 are likely due to differences in the age groups and a higher prevalence of vaping among 19-year-olds who are included in the ITC Youth but not the ASH-Y. There are also possible lasting effects of the COVID-19 pandemic and its impact on smoking and vaping among young people needs to be monitored.

Enforcement and further regulations

In 2022, higher vaping prevalence was reported across all age categories. So, as recommended in our previous reports, enforcement of age of sale regulations for vaping (and smoking) needs to be improved to reduce young people’s access to vaping products and cigarettes.

The dramatic increase in young people using disposable vaping products should be monitored with improved regulatory oversight. Also, the advertising, packaging and marketing of disposable products to young people should be investigated and, where appropriate, proportionate action taken to reduce appeal to young people.

Chapter 4: Vaping among adults

Data reported in this chapter came from 4 different surveys. Most data were from the Smoking Toolkit Study (STS), collected between January and September 2021, and the 2021 ASH-Adult (ASH-A) survey, collected in February and March 2021. Other data from the ONS Opinions and Lifestyle Survey (OPN) and ONS Annual Population Survey (APS) were collected in 2020. We also report some data from the most recent 2022 ASH-A survey on smoking prevalence, vaping prevalence, the relationship between smoking and vaping and the type of vaping products used.

Main findings

Smoking and vaping prevalence

Smoking prevalence among adults in England in 2021 was between 12.7% and 14.9% depending on the survey and in 2022, based on ASH-A data, 13.2%. These equate to between 5.6 and 6.6 million smokers.

There was variation in smoking prevalence by age, gender, socio-economic status and ethnicity. Most notably, smoking prevalence remained significantly higher among adults from more disadvantaged groups.

Vaping prevalence among adults in England was lower than smoking prevalence across all groups and seemed to have increased by around 1 percentage point from 2020 to 2021, to between 6.9% and 7.1%. This equated to about 3.1 to 3.2 million vapers. In 2022, based on ASH-A data, adult vaping prevalence in England was 8.3%.

There was some variation in vaping prevalence by socio-demographic groups and smoking status. Using 2021 STS data, the highest vaping prevalence was among:

• men (7.8%)
• people from the north of England (8.3%)
• people from social grades C2, D and E (8.8%)
• current smokers (22% compared with 11.6% among former smokers and 0.6% among never smokers)

Among former smokers, 27.9% of short-term former smokers (quit for less than one year) used vaping products, compared with 9.9% of long-term former smokers (quit for longer than one year). This is an increase since 2013 when 1.2% of long-term former smokers vaped. In comparison, a small but steady proportion of long-term former smokers have used NRT (around 2% to 4%) since 2013.

The proportion of vapers who also smoke had been declining since 2012, from 91.9% to 49.8% in 2020 in the STS survey and from 73.7% to 31% in 2021 in the ASH-A survey. However, both STS and ASH-A surveys suggest a recent increase in the proportion of vapers who smoke. The STS survey showed an increase to 51.7% in 2021, and the ASH-A survey showed an increase to 33.4% in 2022. The discrepancy in estimates across surveys is likely due to different definitions of smoking status.

Types of vaping device

In both STS and ASH-A surveys, tank models remained the most popular type of vaping device, used by 59.3% of current vapers in the 2021 STS survey and 64.3% of current vapers in the 2022 ASH-A survey. In the 2021 STS survey, different types of vaping devices reported by current vapers included:

• 20.1% modular vaping products (where people use their own combination of device parts)
• 14.9% cartridge models (a rechargeable vaping device that uses replaceable pre-filled cartridges)
• 4.6% disposables (a non-rechargeable and non-refillable vaping device)

The 2022 ASH-A survey showed higher use of disposable vaping products than in 2021, with 15.2% of current vapers reporting using disposable vaping products in 2022 compared with 2.2% in 2021.

Vaping frequency

Among adults who had ever vaped, daily vaping was associated with their smoking status. Among never smokers who had ever vaped, nearly two-thirds (64.9%) had tried it once or twice and 5% were vaping daily. Among current daily or non-daily smokers who had ever vaped, around 27% vaped daily. Among former smokers who had ever vaped, more than half (57.7%) vaped daily (2021 ASH-A).

Length of time vaping

ASH-A 2021 data suggested an increase in the proportion of current vapers who have vaped for more than 3 years (23.7% in 2018, 29.3% in 2019, 39.2% in 2020 and 43.7% in 2021). People who had vaped in the past mostly stopped after 6 months of use or less (57.2% in 2021).

Reasons for vaping

The most common reasons for vaping reported in the 2021 ASH-A survey were to quit (27.9%) or stay off (17.7%) smoking tobacco or because people enjoyed it (12.6%).

Nicotine

In 2021, strengths of nicotine vaping liquids above those allowed by regulations (more than 20mg/mL) were used by less than 6% of vapers. Just over a third of vapers (34%) reported reducing the strength of the nicotine vaping liquid they use since starting to vape, 31.4% continued using the same strength and 26.2% did not know if they had changed the strength. Just 8.1% of people reported having increased the strength of the nicotine in vaping liquid they use since starting to vape (2021 ASH-A). The proportion of vapers unsure about the strength they are using has increased slightly over the last 2 years.

Flavours

Fruit (35.3%), menthol/mint (22.5%) and tobacco (20.9%) remained the most popular flavours among vapers (2021 ASH-A).

Using vaping to stop smoking

Attempts to stop smoking and success rates for adults who tried to stop smoking increased significantly in the last 2 years. This is most likely due to the COVID-19 pandemic. According to STS data, vaping products remained the most common aid used in a quit attempt.

The ‘Reaching Out’ report from ASH has shown that stop smoking services have greatly improved the provision of vaping products to support a quit attempt. In 2019, 11% of surveyed local authorities offered vaping products to some or all smokers accessing stop smoking services. In 2021, 40% of surveyed local authorities offered vaping products to some or all smokers and a further 15% had plans to do so.

Between April 2020 and March 2021, quit attempts in stop smoking services that involved using a vaping product (alone or in combination with medication) achieved self-reported short-term success rates of 64.9%, compared with 58.6% for attempts not involving a vaping product. Despite this, only 5.2% of quit attempts supported by a stop smoking service involved a vaping product.

Implications

Further monitoring and research

Vaping is more common among disadvantaged adult groups in society. This mirrors smoking prevalence, and research should continue to explore the impact that higher vaping prevalence has on stopping smoking and reducing health inequalities.

The continuing impact of COVID-19 on smoking and vaping among adults needs to be monitored. This should include younger adults who start smoking and vaping and any changing patterns in the data.

There needs to be further research into the increasing proportion of long-term vapers and their motivation to stop vaping, and whether people who want to stop vaping need support. More research is also needed into vaping among:

• never smokers
• younger adults
• people from ethnic minority backgrounds

A recent increase among these groups of using disposable vaping products warrants further monitoring and research.

Implementing NICE guidance

The NICE guidance ‘Tobacco: preventing uptake, promoting quitting and treating dependence’ should encourage more stop smoking services to support smokers who want to stop smoking with the help of a vaping product.

As we recommended in previous reports in this series, and as supported by the new NICE guidance, all smokers should be supported to stop smoking completely, including dual users who smoke and vape.

Chapter 5: Nicotine

In this chapter we discuss the role of nicotine in vaping product use.

Main findings

E-liquids

As discussed in chapters 3 and 4, 2021 survey data from England shows that nicotine would appear to play an important driver of adult vaping, but perhaps less so than for tobacco smoking.

Most adults who vape (about 87%) use vaping products that contain nicotine. The proportion was about 70% for 11 to 18 year olds, with about half of those saying that their vaping products always contained nicotine, and half sometimes contained nicotine. Among 16 to 19 year olds who reported ever using vaping products with nicotine, and who had vaped in the past 30 days, 83% said that their products contained nicotine or that some of their products contained nicotine. Overall, the vast majority were using vaping products with less than 20mg/mL nicotine e-liquids and so complied with current vaping product regulations.

Questions on the use of salt-based nicotine products as opposed to freebase nicotine were not often included in surveys. Among 16 to 19 year olds, there was higher awareness of the inclusion of nicotine and type of nicotine in 2021 compared to 2019. Among adults, uncertainty about whether people who vape were using salt-based vaping products had increased slightly over the last 2 years.

Nicotine intake

Previous reviews showed that nicotine intake from vaping products was variable and dependent on different product characteristics. The updated evidence presented here also provides conclusive evidence of this variability.

The updated evidence from pharmacokinetic studies (studies exploring how nicotine is absorbed, distributed and eliminated from the body) on vaping show that in general, vaping products provide lower peak nicotine levels and lower overall nicotine levels to users than smoking provides. Also, the pharmacokinetic studies show that exposure to nicotine from vaping varies by product characteristics. The studies suggested that exposure to nicotine tends to increase when:

• using e-liquids with higher nicotine concentration
• using e-liquids based on nicotine salts rather than freebase nicotine
• using tank or modular type vaping devices which provide more exposure than cartridge or disposable models
• people with longer vaping experience vape, as they have more effective puffing behaviour

The time taken to reach peak nicotine delivery from vaping products is usually slower compared with smoking a cigarette. But this varies depending on the e-liquid nicotine concentration and the type of vaping device. Flavours may also play a role in nicotine delivery and we review this in chapter 6 on flavours.

The pharmacokinetic studies are consistent with the studies discussed in chapter 7 on biomarkers of exposure to nicotine and potential toxicants which generally showed lower exposure to nicotine when using vaping products over the short term (up to 7 days) compared to smoking. However, there was moderate evidence, in medium to longer term studies (up to 2 years), of similar exposure to nicotine from vaping compared to smoking. For experienced adult vapers, there was substantial evidence of comparable exposure to nicotine from vaping and smoking. There was supportive evidence that over time, people who vape compensate for lower nicotine concentrations by compensatory puffing (such as puffing more frequently, puffing larger volumes of aerosol, or taking longer puffs).

Nicotine dependency

There was substantial evidence from previous reports (from NASEM, COT and our 2018 report) that using vaping products can result in symptoms of nicotine dependency. There was moderate evidence that the risk and severity of nicotine dependency for vaping is lower than for cigarette smoking and would vary by product characteristics. The pharmacokinetic studies reviewed are consistent with this.

Our review showed that there are many scales used to assess nicotine and vaping dependency. But as yet, there is no consensus on which is the best scale to assess vaping dependency. So, this makes assessing the risk and severity of vaping dependency compared to tobacco smoking dependency difficult.

A recent systematic review examining the effects of nicotine concentration and flavours on dependency found that higher nicotine concentrations might increase the abuse potential and appeal of vaping and hence dependency. So, this could help someone completely substitute tobacco cigarettes for vaping products. Also, preliminary evidence suggested that flavours may interact with nicotine concentrations to affect abuse potential.

Health risks

We review the health risks of vaping in chapters 8 to 12.

Isolating the effects of nicotine on health risks in human studies is complex, partly because only a small proportion of people vape non-nicotine products. In general, where studies assessed biomarkers in humans (measurements of potentially harmful smoking or vaping effects in the body) through non-nicotine vaping as well as nicotine vaping, the different methods used in each study made it difficult to compare, and so limited our conclusions.

One biomarker, pulse wave velocity (which measures blood pressure pulse through an artery or arteries), did seem to be affected by nicotine in vaping products, at least in acute exposure studies. Evidence from the reviewed animal and cell studies suggest some adverse effects of nicotine, but the extent to which these findings can be generalised to humans is currently very unclear.

Implications

Improved surveillance and further research

Questions in national surveys sometimes lag behind product developments, such as questions about people using salt-based vaping products or increasing their use of disposable vaping products. Having an appropriately resourced product surveillance system would help to ensure researchers can capture data on product developments.

Exploring how nicotine labelling could be improved could also be useful as there appears to be an increase in adult users not knowing how much nicotine was in their vaping products. It would also be useful to further explore the small proportion of adults who use nicotine-free vaping products. For example, asking them how long and how often they use these products.

Current evidence shows that more experienced vaping product users adjust their puffing behaviour to attain higher levels of nicotine. However, this does not compensate for lower overall nicotine exposure after a single vaping session compared with smoking a cigarette. We found during longer-term vaping sessions or where a person can vape as much or as often as they want to (ad libitum), experienced vapers reach levels of nicotine comparable to those from smoking (as shown by nicotine biomarker data). Vapers’ ability to adjust their puffing behaviour to mirror smoking suggests that vaping enables users to carefully control their nicotine levels. This could be a problem when people using vaping products with lower nicotine concentrations compensate by increasing their puffing and so risk increasing exposure to other constituents, including potentially harmful ones. We explore this issue further in later chapters. A recent systematic review suggested that limiting nicotine concentrations in vaping products might reduce smoking cessation.

Future research should use more longitudinal study designs (studies that assess the same people more than once over a period of time) to explore how, with more experience, vapers change their:

• puffing behaviour
• nicotine intake
• dependency, over time

This is important for people who have smoked as well as those who have never smoked. For people who have never smoked and start using nicotine through vaping, measurements are needed across a range of vaping products and their characteristics. This will help to assess whether higher nicotine limits (more than 20mg/mL) affect a person’s dependency on vaping, and how their vaping behaviour might interact with:

• free-base or salt nicotine levels
• flavours
• other characteristics (for example e-liquid PG/VG ratio)

Research on longer-term vaping behaviour would also allow researchers to clarify how using different nicotine strength e-liquids over time is associated with dependency and potential health risks.

Need for global standards and protocols

Having a global consensus for assessing and measuring nicotine and product dependency would enable comparisons of nicotine and product dependency:

• between vaping and smoking
• across different vaping products
• with different groups of users (such as adults and young people).

In England, it is important for researchers to keep up to date on the ongoing research in this area.

Agreeing a standard protocol for vaping product pharmacokinetic studies would also enable meaningful comparisons across different vaping products and e-liquid characteristics. However, more long-term ad libitum pharmacokinetic studies are also needed to reflect how users’ personal experience and puffing behaviours affect nicotine delivery and dependency.

Isolating the risks of nicotine to health from the risks of other vaping constituents is difficult in human studies compared to animal and cell studies. Having standards, particularly for human cell research, may strengthen how widely or generally applicable such studies are to vapers. Such standards would also be beneficial in helping to examine the effect of nicotine in humans.

Chapter 6: Flavours

This chapter

• describes the use of flavoured vaping products in England
• provides an overview of the role of flavours in vaping product use
• summarises the evidence on potential harm from flavourings in vaping products from studies identified in a systematic review

Main findings

Use of flavours

As we identified in earlier chapters, fruit flavours are the most popular e-liquid among adults and young people who vape in England, followed by menthol/mint. There is some evidence to suggest that non-tobacco flavours, particularly sweet flavours, may play a positive role in helping people switch from smoking to vaping.

A systematic review of the evidence on youth use of e-liquid flavours concluded that existing research does not yet provide a clear understanding of how flavours in vaping products are associated with young people taking up or stopping smoking.

Potential toxicants in flavours

In 3 studies, levels of tobacco specific nitrosamines and volatile organic compounds were significantly reduced in smokers and dual users who switched to vaping products with different flavours. Biomarker levels slightly differed between flavours, but this was not tested for statistical significance. Users of fruit-only flavoured vaping products had significantly higher concentrations of a biomarker for acrylonitrile (CNEMA) compared to users of a single other flavour in one study.

One longitudinal observational study of people who vaped found that:

• flavour preferences changed over time
• 6.9% self-reported an adverse reaction that they associated with the flavour they used
• a third had ever used a cinnamon or cinnamaldehyde containing vaping product

Findings from 13 cell and 9 animal studies suggest there is limited evidence that some flavourings in vaping products, particularly cinnamaldehyde, or buttery or creamy flavours have the potential to alter cellular responses but less than exposure to tobacco smoke. Exposure to propylene glycol or vegetable glycerine (PG/VG) base liquids without added flavourings appeared to have little or no effect. It was not always possible to differentiate the effect of nicotine or solvents from flavourings due to lack of appropriate controls. This was further complicated by variability of e-liquid composition, cell types, dose exposure and duration. Also, there was not a great deal of consistency about whether cells or animals were exposed to e-liquids, aerosol extracts or aerosols.

There was only one study that looked at the stability of e-liquid flavourings over a period of one year (and found they were stable), but no studies conducted assessments to see if this changed how the flavouring tasted and felt over time.

Subjective effects

Two studies assessing acute exposure to flavoured vaping products, under controlled conditions, found that nicotine delivery and ‘positive subjective effects’ (such as liking) for flavoured vaping products were lower than for tobacco cigarettes. The studies also found that positive subjective effects were greater for vaping products and tobacco cigarettes, than for nicotine gum. There were mixed findings on whether or not the subjective effects of flavourings were due to nicotine delivery or increased level of consumption.

A recently published systematic review concluded that flavours affected the abuse potential (for example, liking a product and intending to use it again) of vaping products through increasing product appeal. But it acknowledged that the effect of flavours on smoking cessation needed further research.

Implications

Surveys in England should include detailed questions on the use of flavours (including mixing different flavours) in vaping products annually, to track use over time. Longitudinal data in adults and young people in England would also be helpful in assessing the health effect of flavours in vaping products.

The findings of the systematic review support previous reports, (our 2018 report, the NASEM report and the COT review), which suggest cinnamaldehyde-containing vaping products continue to be a cause of concern. The review also recommends that regulatory bodies should review this flavouring chemical in e-liquids. Although there is less evidence in this systematic review, some in vitro (laboratory cell-based) studies suggest buttery and creamy flavoured e-liquids may also require further review.

A more standardised approach is needed to evaluate the risks associated with flavourings in e-liquids and aerosols in human and cell studies, independent of nicotine and PG/VG. The evaluation framework devised by COT to aid risk assessment of flavouring compounds via inhalation exposure could be considered by regulators at the time of product notification.

COT also suggested that since flavourings can undergo thermal degradation or react with other constituents in e-liquids, research is needed to fill the gap in our knowledge about how heating affects flavours. This included looking at the extent to which thermal degradation may be affected by users customising their vaping devices.

COT also suggested looking at the potential safety of exposure to mixing e-liquid flavours.

Also, further research is also needed about the stability of flavourings over time and whether they degrade or not.

Chapter 7: Biomarkers of exposure

Evidence reviewed

This chapter examined findings from our systematic review on biomarkers of nicotine and potential toxicants (chemicals or their metabolites in a body that show actual human exposure to nicotine or tobacco products) relevant to our 2 review protocol questions.

1. The effect of vaping and secondhand exposure to vaping products that are associated with the risk of health conditions.

2. The effects of vaping among people with existing health conditions on disease outcomes.

However, we did not find a study addressing the second review question. Only one study assessed participants with self-reported respiratory symptoms but did not test for statistical differences across relevant groups. So, our review for this chapter is confined to our first review question.

We assessed both relative (between vapers and smokers) and absolute (between vapers and non-users) vaping risks associated with exposure to nicotine and potential toxicants where the data were available. Where feasible, we included comparisons across different population groups.

The included studies used a range of different designs and had varying quality or risk of bias.

The studies we have included used a range of different definitions of vaping and smoking. For example, findings of some studies were confounded by classifying vapers who smoke, occasional vapers and/or exclusive daily vapers as a uniform group, or comparing occasional vapers with daily smokers. So, findings need to be cautiously interpreted.

Studies looking at participants at more than one time point mostly explored acute exposure to vaping (single use to 7 days) or followed up participants for short to medium term (8 days to 12 months). So, we were unable to summarise findings on longer term (more than 12 months) vaping exposure, with some studies not allowing adequate wash-out periods for biomarkers with longer half-lives.

In line with our algorithm (chapter 2, table 6), we carried out meta-analyses wherever possible. But a lack of consistency in study designs, biomarker reporting, group definitions and exposure periods resulted in only a few studies being included.

Here we summarise our findings for each biomarker for relative and absolute differences in various populations of interest, starting with first-hand vaping exposure.

Main findings

Nicotine

There was substantial variation across the 60 studies included in this section looking at nicotine exposure. Only 5 studies (4 longitudinal and one cross-sectional (measured at a single point in time)) were from the UK. Levels of nicotine and nicotine metabolites in participants using vaping products differed according to:

• study design
• definitions of vaping and smoking
• biomarker and biosample (a biological sample, which could include urine, blood plasma, blood serum and saliva) used
• exposure duration

To assess relative exposures between vaping and smoking, we were able to carry out 5 meta-analyses of nicotine and nicotine metabolites (one longitudinal, 4 cross-sectional) among people who vaped and smoked at least weekly. All found no significant differences between people who vaped and smoked.

From the narrative summaries, evidence suggests that over time and with increased experience of vaping, users can derive similar levels of nicotine as they can from smoking cigarettes. Levels of nicotine metabolites varied with vaping device characteristics (for example, vaping device types, e-liquid nicotine concentrations).

To assess absolute exposures between vapers and non-users, we were able to carry out 4 meta-analyses of nicotine biomarkers which, as expected, showed significantly higher levels among vapers than non-users. In general, findings from the narrative summaries were similar for absolute nicotine exposures.

There were no discernible differences between adults and adolescent exposures to nicotine and its metabolites.

Volatile organic compounds

Twenty-four studies assessed volatile organic compounds (VOCs), with only 5 from the UK. VOCs are potentially harmful gases released into the air, for example while smoking tobacco. Again, there was considerable variation across the studies in:

• design
• definitions of vaping and smoking, biomarker measurements
• exposure duration

To assess relative exposures between vaping and smoking, we were able to carry out 15 meta-analyses of VOCs (4 longitudinal, 11 cross-sectional). Findings varied by biomarker. In general, most findings showed statistically significantly lower levels of VOCs among vapers than smokers, with substantial reductions in some biomarkers, such as the acrolein metabolite 3-HPMA (71%), the acrylonitrile metabolite CNEMA (94%) and 1,3-Butadiene metabolite MHBMA (83%). For a few VOCs, such as formaldehyde and toluene, available evidence was inconclusive on the significant differences between vapers and smokers.

To assess absolute exposures between vapers and non-users, we were able to carry out 10 meta-analyses (all cross-sectional). All showed no significant differences between vapers and non-users, except for the acrylonitrile metabolite CNEMA. One study showed that average levels of acrylonitrile metabolite CNEMA for vapers were over 3 times higher than those among non-users.

In general, findings from the narrative summaries were similar for absolute and relative VOC exposures.

Levels among young people were broadly in the same direction to levels reported among adults, with some differences for individual biomarkers. This may be due to different smoking and vaping patterns.

Tobacco specific nitrosamines

Twenty-eight studies assessed tobacco specific nitrosamines (TSNAs), a group of chemicals found in tobacco and tobacco smoke, some of which are harmful and cause cancer. Only 3 studies were from the UK. As with other biomarkers, there was considerable variation across the studies in:

• design
• definitions of vaping and smoking
• biomarker measurements
• exposure duration

To assess relative exposures between vaping and smoking, we were able to carry out 5 meta-analyses of TSNAs (2 longitudinal, 3 cross-sectional). These all showed significantly lower levels of TSNAs among vapers than smokers, with substantially lower levels for NNAL (58%), NAB (87%), NAT (94%) and NNN (90%). Findings were generally consistent with those reported in the narrative review.

To assess absolute exposures between vapers and non-users, we were able to carry out 3 meta-analyses using cross-sectional data. These all showed significantly higher levels of TSNAs among vapers than non-users. However, the cross-sectional data make it difficult to distinguish exposure from vaping products from previous tobacco use. Also, evidence from a randomised control trial (RCT) and a cross-over study (a study where different products are given to the same participants but in different orders, and participants serve as their own controls) indicates that TSNA metabolite levels among vapers might decrease to a similar level as among non-users.

Levels among young people were in the same direction as among adults, although the magnitude of difference between vapers and smokers was substantially less for young people compared with adults. Again, this may be due to different smoking and vaping patterns among adults and young people.

Other potential toxicants

Nine studies assessed a range of other potential toxicants, such as polyaromatic hydrocarbons, with only one from the UK. We were unable to carry out any meta-analyses. Generally, the very limited findings suggested the levels of these other potential toxicants were lower among vapers than smokers, and higher among vapers than non-users.

Carbon monoxide

Thirty-three studies assessed carbon monoxide (CO) exposure, with 3 studies from the UK. As for other biomarkers, there was considerable differences in methods across the studies and user definitions.

To assess relative exposures between vaping and smoking, we carried out 2 meta-analyses. Both showed significantly lower blood carboxyhaemoglobin levels among vapers than smokers.

We were unable to carry out any meta-analyses of exposures between vapers and non-users. But some interventional studies (such as RCTs, longitudinal and cross-over studies) suggested that exposure to CO in smokers who completely switch to vaping product use might be reduced to levels similar to non-users.

Metals

Ten cross-sectional studies examined a range of metals (arsenic, cadmium, lead, mercury), with none from the UK. No meta-analyses could be carried out.

In general, the studies had mixed findings about relative exposure.

Absolute exposure assessments were also mixed although most studies showed higher levels of exposure among vapers than non-users.

Secondhand exposure

Six studies assessed secondhand exposure to vaping product aerosol, using a variety of biomarkers, none from the UK. The level of exposure varied greatly from people at home to people attending an indoor vaping convention.

Short exposures to secondhand vaping did not result in detectable changes in levels of nicotine, VOCs or TSNAs. However, longer exposures during heavy sustained vaping were associated with significant increases in nicotine or potential toxicants’ metabolites.

Implications

Our systematic review covered a wide range of biomarkers and studies. Our findings are broadly consistent with the few previous reviews in this area, but because of the greater volume of research that has been conducted in recent years, the implications are much clearer.

Vaping reduces toxicant exposure compared with cigarette smoking

The reviewed studies show that compared to smoking, using vaping products leads to a substantial reduction in biomarkers of toxicant exposure associated with cigarette smoking. However, the degree of any residual risk remains unclear, mainly because of the lack of comparisons between long-term former smokers who do and do not vape or comparisons with those who have never smoked or vaped.

Methodological improvements needed

Our quality assessments revealed most studies had some methodological concerns, and these should be addressed in future research as they limit interpretations of our findings. For example, a lack of significant differences between levels of exposure between people who vape and non-users may be due to several reasons. This includes a lack of sensitivity in biomarker measurement methods, background environmental exposures, or because exposure to potential toxicants between people who only vape and non-users is relatively similar.

Improvements in definitions needed

Historical tobacco use can greatly affect many of the biomarkers used to determine exposure to potentially harmful constituents from vaping. So, as most vapers are previous long-term smokers (see chapter 4 on vaping among adults), strict definitions for duration of exclusive vaping (only vaping) should be used consistently in future studies.

Similarly, definitions should not include concurrent smoking, and only include people who exclusively vape. This is particularly important for cross-sectional studies, but longitudinal studies should also use objective measurements to assess concurrent cigarette smoking.

Future studies should always verify biologically participants’ smoking, vaping or non-use status, rather than rely on self-reports. Based on our review findings, measurements of carbon monoxide or NNAL could be used to improve over-reliance on self-reported vaping and smoking.

More research needed on biomarkers of exposure among vapers

More research is needed on biomarkers of exposure among vapers, particularly in the UK, where we identified a lack of studies. We would encourage research with longitudinal and cross-sectional designs. While longitudinal research is more robust, particularly in relation to changes over time, cross-sectional research also offers insight into exposure from realistic and naturalistic use patterns. Longitudinal research would benefit from including longer follow-up periods to be able to assess long-term changes in biomarker exposure among vapers who sustain use over long periods of time (see chapter 4 on vaping among adults). This is also important for biomarkers with longer half-lives.

In our meta-analyses, many findings were from tobacco industry-funded RCTs conducted in confinement (closed settings such as research centres or hospitals in which the participants stay) for periods of up to 7 days. So, future research needs to include more independent research of biomarkers of exposure in people who use vaping products, smoke and do not use tobacco or nicotine outside of confinement (in their own normal settings), and with longer follow-ups.

Need to distinguish between biomarkers of exposure from tobacco and from other sources

Several biomarkers of exposure are not specific to tobacco, and almost all biomarkers are susceptible to the effects of confounders (which can also influence levels of a biomarker). For example, VOCs are prevalent in many household products such as paints and cosmetics and can also be influenced by diet.

Where a person lives can also uniquely influence exposure. There are higher levels of polyaromatic hydrocarbons and other toxicants found in urban environments due to motor vehicle exhaust fumes and other sources of pollution. There are also different toxicant exposures in rural environments, due to pesticide exposure and other agricultural pollutants.

So, strict control for confounders and large sample sizes are needed to reduce the influences of other environmental exposure on findings in cross-sectional research.

Need to identify and study biomarkers which are specific to vaping

Our systematic review used the World Health Organization (WHO) priority toxic contents and emissions list for tobacco products. There are already suggestions to include vaping specific biomarkers in the WHO list when and if these emerge, which will help guide future research. Due to the variety of different metal elements used for vaping product components, there may be exposure to certain metals from vaping that are not present in exposure from tobacco. Future research is needed to identify types of metal exposure that are exclusively from vaping products and how these can be mitigated.

There is a need to address the lack of comparable research on biomarkers of exposure to nicotine and potential toxicants across different groups, such as:

• young people and adults
• different genders
• ethnicity
• socioeconomic status

Given we identified no studies assessing the biomarkers of exposure to vaping among people with existing health conditions on disease outcomes, this is an important gap that should be addressed by funding bodies.

Lower risks of exposure from vaping than smoking

Overall, despite the methodological limitations identified in our systematic review, evidence suggests significantly lower relative exposure from vaping compared to smoking in biomarkers that are associated with the risk of:

• cancer
• respiratory conditions
• cardiovascular conditions
• other health conditions

This is consistent with encouraging people who smoke to switch completely to vaping as a way to stop smoking or as alternative nicotine delivery devices. Also, our findings of higher absolute exposure from vaping compared with not using any nicotine products reinforce the need to discourage people who have never smoked from taking up vaping (or smoking).

Chapter 8: Biomarkers of potential harm to health cutting across several diseases

Evidence reviewed

This chapter examines findings from our systematic review on biomarkers of potential harm to health that are associated with:

• oxidative stress
• inflammation
• endothelial function
• platelet activation

These biomarkers are known to be associated with the development of multiple diseases (see chapter 2, table 6). So, they are relevant to both our review questions:

1. What effect does vaping have on biomarkers that are associated with the risk of cancer, respiratory, cardiovascular and other health conditions?

2. What effect does vaping among people with existing health conditions have on disease outcomes?

Several of the studies we included assessed biomarker changes in participants with existing health conditions (for example, asthma and dental diseases) but did not estimate how these changes affected outcomes of these health conditions. As these studies did not directly address the second review question, we have presented their data alongside findings from participants from the general population.

Main findings

Issues caused by differences between studies

Overall, we identified 41 unique studies in 43 publications that reported biomarkers of potential harm associated with oxidative stress, inflammation, endothelial function and platelet activation biomarkers. There were significant differences in methodologies across the studies we included, which likely resulted in discrepancies and variability of findings. These differences included the following.

1. Studies assessed multiple biomarkers with different sensitivity, speed of onset or offset and reliability of predicting subsequent health risks. These differences obscured overall conclusions.

2. Studies used different definitions for vaping, smoking and non-use groups, usually did not bioverify smoking or vaping status, and used varied methods (for example, different measures, biosamples and follow-up times) to compare a range of biomarkers between these groups. These differences prevented us pooling data from more studies for meta-analyses and made comparisons between studies complicated.

3. Most studies we included assessed acute vaping effects on oxidative stress, inflammation, endothelial and platelet functions. And because the explored biomarkers of potential harm mostly take weeks or months to normalise after people stop smoking, we cannot make clear conclusions about longer-term vaping effects.

4. Tobacco smoking (or vaping) is not the only known risk factor for detrimental changes in many of the explored biomarkers. And conclusions about vaping associations with the explored biomarkers are further limited by potential confounding of other variables and the lack of controlled studies. So, findings need to be cautiously interpreted.

In line with our algorithm (chapter 2, table 6), we carried out meta-analyses wherever possible, but a lack of consistency in study designs, biomarker reporting, group definitions and exposure periods resulted in few studies being included.

Oxidative stress

One RCT, 6 cross-over studies, 5 non-randomised longitudinal studies and 11 cross-sectional studies assessed oxidative stress biomarkers, specifically:

• low-density lipoprotein (LDL)
• high-density lipoprotein (HDL)
• 8‑isoprostane
• soluble Nox2-derived peptide (sNOX2-dp)
• malondialdehyde (MDA)
• 8-hydroxy-2’-deoxyguanosine (8OhdG)
• reactive oxygen species (ROS)

We found no significant differences in LDL levels across studies between vapers, smokers and non-users’ groups after acute and short-to-medium exposure. A meta-analysis of data from 2 cross-sectional studies also confirmed no difference in blood LDL levels between vapers and non-users.

Findings on HDL levels were inconsistent. Smaller studies reported no differences between vapers, smokers and non-users, and larger studies reported lower HDL levels among non-users compared with vapers and smokers. Two meta-analyses of cross-sectional studies found no difference in blood HDL levels between vapers compared with smokers or non-users.

Evidence for 8-isoprostane level changes after vaping product use was mixed. Studies emphasised longer past smoking history, older age and female gender as potential confounders for higher 8-isoprostane levels (these factors are associated with higher 8-isoprostane levels). In general, comparisons were limited by a lack of longer-term controlled exposure studies (considering time for biomarkers’ levels to normalise) and potential confounding in non-randomised longitudinal and cross-sectional studies.

There was limited evidence for other oxidative stress biomarkers. The overall evidence from most of the included studies show no difference in vaping‑associated oxidative stress risks in comparison with smoking or not using tobacco or nicotine products.

Inflammation

Two RCTs, 3 cross-over studies, 3 non-randomised longitudinal studies and 17 cross-sectional studies assessed inflammation biomarkers, specifically:

• white blood cell (WBC) count
• c-reactive protein (CRP)
• interleukin-6 (IL-6)
• interleukin-8 (IL-8)
• tumour necrosis factor alpha (TNF-α)
• soluble intercellular adhesion molecule 1 (sICAM-1)
• fibrinogen
• prostaglandin E2 metabolite (PGE-M)
• monocyte chemoattractant protein-1 (MCP)

Pooled data from 3 cross-sectional studies showed that average blood CRP levels were lower among vapers than smokers and similar between vapers and non-users, and that average blood sICAM-1 levels were significantly lower among vapers than smokers. However, controlled and longitudinal studies did not confirm these cross-sectional findings. Also, due to varied study designs and a lack of studies comparing the same outcome between the same study groups, no definite conclusions could be drawn on the association between vaping and any specific inflammation biomarker.

Endothelial function

One RCT, 4 cross-over studies, 3 non-randomised longitudinal studies, and one cross-sectional study assessed endothelial function biomarkers, specifically:

• flow-mediated dilation (FMD)
• E‑selectin and P-selectin
• nitric oxide
• microvesicles

No studies reporting on these biomarkers could be pooled for a meta-analysis.

While acute exposure studies showed similar short-term reductions in FMD parameters after vaping (with and without nicotine) and smoking sessions, a single RCT showed that switching from smoking to vaping for 4 weeks significantly improved (increased) participants’ FMD function.

Evidence from 2 cross-over studies and one interventional study showed that acute vaping and smoking sessions led to similar reductions in nitric oxide bioavailability (more susceptibility to oxidative damage), but one study also noted that the reduction was directly associated with the length of past smoking history.

Evidence from one cross-over study and one interventional study showed that acute nicotine vaping increased blood endothelial microvesicle levels while acute non‑nicotine vaping did not change this outcome.

There was limited and inconsistent evidence on the other endothelial function biomarkers. Also, we could not make any conclusions about the absolute effect of vaping on endothelial function, as no controlled studies compared vapers and non-users.

Overall, acute vaping might cause endothelial dysfunction as much as acute smoking but switching from smoking to vaping product use might improve endothelial function in the longer-term.

Platelet biomarkers

Only one cross-over study, one longitudinal study and 2 cross-sectional studies assessed platelet activation measures. No data from these studies could be meta‑analysed. So, evidence on the association between vaping and platelet function was limited, and we could not make any conclusions about absolute effects of vaping on platelet activation or effects of vaping relative to smoking.

Implications

Need for methodological improvements and longer term studies

Considering the 2 human studies summarised by the NASEM report and the 41 studies (in 43 publications) included in our systematic review, research on effects that human vaping has on biomarkers that cut across diseases has grown in recent years, though it is still at an early stage.

Our summary of the evidence on associations between vaping and oxidative stress, inflammation, endothelial function and platelet activation came from studies with different methodologies that mostly assessed acute exposure effects. These findings provide important insights allowing us to compare immediate effects between vaping and smoking. However, like smoking, it is the effects of long-term vaping that will be most relevant to public health, and the explored biomarkers of potential harm mostly take weeks or months to normalise after people stop smoking.

Our risk of bias assessments showed that most studies in this chapter had methodological concerns, and these should be addressed in future research as they limit interpretations of our findings. More research is needed, particularly in the UK, where we identified a lack of studies.

There is a need for future research among people who vape and have never smoked. This would allow us to determine long term changes in biomarkers of potential harm exclusively due to vaping and not as a consequence of prior long‑term smoking.

Need to distinguish between biomarkers of potential harm from smoking or vaping from other sources

Also, most biomarkers of potential harm are associated with multiple confounders not related with vaping or smoking (for example diet, physical activity). So, studies that explore acute effects of vaping and/or smoking, but do not include non-users as a comparison group, cannot clearly distinguish between the effects of vaping and/or smoking on these biomarkers. Due to these reasons, most studies that we have summarised in this chapter cannot inform us about the medium or long term vaping-associated risks via effects on the biomarkers we reviewed. This implies that further controlled studies with adequate sample sizes, non-user comparison groups, and longer exposure and follow-up times are needed to clarify how switching from smoking to vaping affects the most reliable biomarkers of harm.

Greater clarity on clinical significance

More research is also needed to develop ranges where biomarkers of potential harm become clinically relevant predictors of disease. This would improve the biomarkers’ ability to estimate the pathways and contributions of vaping and smoking to multiple diseases.

Chapter 9: Cancers

Evidence reviewed

In this chapter we reviewed the existing evidence on how vaping might affect cancer risk. This included summarising previous reports that have addressed this issue, and then presenting findings from our systematic review of health risks and effects of vaping that are relevant to cancer.

Main findings

Toxicants and carcinogens

Our 2018 evidence review of vaping, the report from NASEM and the COT report include some earlier evidence. The 2018 report included one study directly relevant to cancer that suggested people who switched from smoking to vaping were exposed to lower levels of toxicants and carcinogens than in smoking, but also pointed to the need for further research. The NASEM report found no available evidence about whether the chemicals in vaping aerosols or vaping behaviour were associated with cancer risk relative to smoking or non-use. COT also reported that existing evidence was insufficient to draw conclusions about any links between vaping and cancer risk in humans.

Cancer risks

We identified a growing (but still modest) amount of literature on how vaping may affect cancer risks in humans. In our review of human studies, biomarkers of exposure to several human carcinogens in tobacco smoke show lower measured levels in people who vape compared with those who smoke. So, the biomarker of exposure studies compiled in this review (see chapter 7 on biomarkers of exposure) provide conclusive evidence that vaping generally leads to lower exposure to many of the carcinogens responsible for the health risks of smoking.

Inflammation and oxidative stress

Findings from studies of inflammation and oxidative stress do not show any systematic relationship with mixed evidence of differences (or no difference) in levels between vapers and smokers and non-users. So, this evidence is currently insufficient to draw conclusions.

Gene and DNA processes

We identified 2 RCTs, one longitudinal study and 5 cross-sectional studies of gene expression and DNA methylation in humans (none from the UK). Methodological limitations (for example, a lack of smoking comparison groups in some studies) constrain what we can say about these epigenetic studies (the study of how people’s behaviours and environment can cause changes that affect the way our genes work). Even so, methylation and demethylation of specific genes related to smoking and vaping show potential for achieving more clarity in this area.

Existing or previous cancer conditions

There were no studies that assessed how vaping affects people with an existing or previous cancer condition.

Cell and animal studies

It is challenging to interpret the findings from pre-clinical studies using human or animal cells or rodent models to any cancer risks arising from vaping in humans. These pre-clinical studies commonly use acute exposures, sometimes over concentrated periods. So, it is unclear whether the pathways to risk identified would be replicated in vapers. Further challenges arise because of the complex nature of vaping behaviour over time and the wide variety of different aerosols and products used.

Despite these significant limitations, there are suggestions from this literature that vaping aerosols are not benign to people who have never smoked. And that exposure to these aerosols may be implicated in negative outcomes that could affect the viability of cancer treatment for people with pre-existing disease. However, cell and animal studies appear to support the human studies and suggest vaping may trigger alterations in gene expression, but at a lower extent than we see from exposure to tobacco smoke.

Implications

Longer follow up periods are needed

Vaping generally leads to lower exposure to many of the carcinogens responsible for the considerable health risks of smoking. However, studies of biomarkers of exposure that are associated with cancer risk in humans need to have longer follow up periods than has been the case to date, as this will give us better information if vaping reduces cancer risk compared with smoking.

More research is needed

More research is needed on biomarkers of potential harm in humans.

Studies applying potentially important new methods to assess vaping often neglect to include cigarette smoke as a comparator as well as a control (usually filtered air). Even when a tobacco smoke comparison group is included, it is often difficult to compare like with like when the exposure to nicotine and other important parameters are not included in the description of the experiments. Such data are essential when assessing whether human exposure to different forms of nicotine delivery (in this case vaping and smoking) result in different magnitudes of cancer risk.

Further studies are needed to identify the extent to which evidence from pre-clinical studies is directly relevant in humans.

There are a number of gaps in the literature identified in our review, as well as some gaps that came to our attention when preparing the background to this chapter. Although we know a lot about the links between tobacco smoking and cancer, more needs to be done to document the smoking status of cancer survivors. These people will make up an increasing proportion of cancer patients in the future given improvements in survival and an ageing population. This means that the risk of recurrence or a secondary cancer will not be uncommon.

We could also not identify any studies from the UK on vaping prevalence among people diagnosed with cancer or cancer survivors, so this should be a further area of research.

More research is needed with cancer patients and cancer survivors to understand any role for vaping as a smoking cessation aid in improving treatment outcomes or reducing the risk of cancer recurrence.

Studies are also needed that assess the effects of vaping on cancer outcomes in people diagnosed with cancer, both to compare with people not using nicotine or tobacco products and with people smoking.

Support smokers to completely switch from smoking to vaping

For policy makers and practitioners, findings from our review for this chapter suggest that developing and implementing policies and interventions that support smokers to completely switch from smoking to vaping will reduce exposure to toxicants and carcinogens. This may have relevant outcomes for cancer prevention.

Chapter 10: Respiratory diseases

Evidence reviewed

In this chapter we reviewed the existing evidence on how vaping might cause or influence respiratory disease, one of the main causes of premature mortality and morbidity among smokers. This included summarising previous reports that have addressed this issue, and then presenting findings from our systematic review of health risks and effects of vaping that are relevant to respiratory disease. Our systematic review aimed to assess the effects of exposure to vaping on biomarkers associated with the risk of poor health conditions and to assess the effect of vaping on disease outcomes in people with existing health conditions.

Most studies examined healthy participants, which we summarise first. We then summarise the studies that examined participants with respiratory conditions (asthma, chronic obstructive pulmonary disease) and smokers with mental health conditions. We assessed both relative and absolute vaping risks associated with biomarkers of respiratory disease where the data were available (between vapers and smokers, and between vapers and non-users), and where feasible included comparisons across different population groups.

Conclusions for biomarkers of exposure and biomarkers of potential harm cutting across common diseases are presented in chapters 7 and 8.

Main findings

Biomarkers of respiratory diseases

Several biomarkers of exposure are relevant to respiratory diseases. We identified conclusive evidence that under typical use conditions, acute (from single use to 7 days) and short to medium (from 8 days to 12 months) exposure to most potential respiratory toxicants from vaping is significantly lower compared with smoking tobacco cigarettes. And there are substantial reductions in some biomarkers. For the respiratory toxicants that were assessed at long-term exposure (more than 12 months), evidence was moderate that biomarkers of exposure are lower for vaping than smoking

For a few VOCs, such as formaldehyde and toluene, available evidence was inconclusive on the significant differences between vapers and smokers. However, one study suggested formaldehyde exposure might increase during compensatory puffing behaviour with lower nicotine strength e-liquids. In general, there were no significant differences between vapers and non-users, except for acrylonitrile metabolite CNEMA. The evidence suggested that vaping might increase exposure to acrylonitrile in absolute terms.

Biomarkers of potential harm

The evidence was mixed on biomarkers of potential harm relevant to multiple diseases (including respiratory disease), such as 8-isoprostane and inflammation. This would indicate that there was insufficient evidence from these biomarkers of potential harm whether vaping product use is associated with respiratory disease in humans.

We identified 25 studies (3 from the UK) that assessed other biomarkers of potential harm that were specifically related to respiratory disease in humans. Consistent with studies in other chapters, the studies we included used a range of different designs and had varying quality or risk of bias. Studies used a range of different definitions of vaping and smoking. For example, findings of some studies were confounded by categorising vapers who smoke, occasional vapers or exclusive daily vapers as a uniform group or comparing occasional vapers with daily smokers. So, findings need to be cautiously interpreted.

Studies with more than one time point mostly explored acute exposure to vaping or followed-up participants for short to medium term. In line with our algorithm for selecting studies for meta-analyses (chapter 2, table 6), the lack of consistency in study designs, biomarker reporting, group definitions and exposure periods meant we were unable to carry out any meta-analyses.

Of the 25 studies we included:

• 7 were relevant to our second research question about effects of vaping among people with existing health outcomes on disease outcomes
• 4 assessed participants with asthma
• 2 were from the same longitudinal cohort, but with different follow-up rates, assessing participants with chronic obstructive pulmonary disease (COPD)
• 1 assessed participants with mental health disorders

Respiratory tests and imaging

All 25 studies included spirometry measures, which is a breath test used to assess airflow obstruction in the lungs (commonly used to detect respiratory diseases). But the different study designs, groups and duration of exposure limited any conclusions that we can draw.

Overall, the findings showed no acute (from single use to 7 days), short to medium (from 8 days to 12 months) or long-term (more than 12 months) detrimental effects for vapers. Whereas a clear worsening of lung function was seen in one small study of vapers who switched back to smoking for 7 days.

Eight studies assessed FeNO (fractional exhaled nitric oxide, which is measured in the breath and is a marker of airway inflammation and asthma). Again, these studies involved different designs, groups and exposure duration so limited our conclusions. There were mixed findings in the studies, but most reported no significant differences across the user groups.

One study assessed impulse oscillometry (a respiratory diagnostic test), which suggested an effect of acute nicotine exposure on some lung function attributes among healthy occasional smokers but needs repeating.

Five imaging and bronchoscopy studies used a variety of different techniques. These studies either assessed very short-term single-use exposure or were heavily confounded by including smokers (either of tobacco or marijuana) in the vaping groups.

Overall, given the methodological differences, we concluded that there was insufficient evidence from spirometry, FeNO, impulse oscillometer, and bronchoscopy and imaging studies as to whether vaping has any impact on lung function after acute, short to medium and long-term exposure. We also concluded that there was insufficient evidence on whether acute secondhand vaping had any effect on lung function.

Asthma

In relation to our second research question, we first summarise our findings from the 4 studies with participants with asthma. Again, sample sizes were generally very small, and the findings were inconclusive as to whether there are improvements in lung function and respiratory symptoms among adult smokers with asthma who switch to vaping completely.

There was limited evidence that vaping negatively affects lung function among adults with asthma.

Chronic obstructive pulmonary disease

Two longitudinal articles taken from the same group of COPD patients reported that they found statistically significant improvements in some spirometry measurements for the group who used vaping products compared with baseline. But there were no significant differences in the group who smoked. However, only small numbers of participants were involved, and the authors suggested larger studies were needed to confirm these findings.

These findings indicate that there is limited evidence for reduction of COPD exacerbations among adult smokers with COPD who switch to vaping completely and continue vaping for up to 5 years.

Mental health

In one study, smokers with a mental health diagnosis were encouraged to use a vaping product to reduce smoking, and reported no statistically significant changes in one spirometry measure. But since most of them continued smoking, further research is needed with this population.

Cell and animal studies

As previously mentioned, it is challenging to directly translate the findings from pre-clinical studies using human or animal cells or rodent models to any respiratory risks arising from vaping in humans. These pre-clinical studies commonly use acute exposures sometimes over concentrated periods, and it is unclear whether the mechanisms or pathways to risk identified would be replicated in vapers. Further challenges arise because of the complex nature of vaping behaviour over time and the wide variety of different aerosols and products used.

We identified 47 in vitro studies that examined biological impact of exposure to vaping product aerosol or vaping product aerosol extract on various human airway cell types. We also identified 25 animal studies investigating respiratory effects following vaping product exposure.

Taking all the reviewed articles into consideration, the current available data contributes to the evidence that vaping product aerosol, to some extent, may cause airway-related adverse effects in cell and animal models. Although, the evidence is inconclusive as to which constituents of the aerosol play important roles in the observed cellular and physiological effects.

Conclusions

Overall, while the literature has grown considerably since the NASEM report, the conclusions from that report are supported by this review.

The lack of consistency across the studies meant we could not perform meta-analyses of respiratory measures, which limits the conclusions that we can draw.

The limited evidence for improvements in COPD for adult smokers in the NASEM report who switched to vaping has now been reported at the 5 year follow-up by the same study group. This shows that improvements seem mainly to be among participants who switched to exclusive vaping.

More studies have been carried out with people suffering from asthma, but the different designs, diagnoses, and measurements taken prevent us from making any conclusions.

Implications

Improve research methodology

Our quality assessments revealed most studies had some methodological concerns, and these should be addressed in future research as they limit interpretations of our findings. More research is needed, particularly in the UK, where we identified a lack of studies.

As we previously mentioned, all studies we included had used very different methods. This included different designs, definitions of user groups (people who smoke, people who vape, people who smoke and vape, and people who do neither) and biomarkers. This likely resulted in discrepancies and variability in their findings.

Study people who vape over longer periods of time

As discussed in other chapters, most studies exposed participants to brief sessions of vaping, so cannot answer questions on long-term respiratory outcomes. So, studies that assess people who have been vaping over long periods of time are urgently needed. Findings from one long-term group of smokers who had switched to vaping at baseline are promising and should be replicated by other studies with larger numbers of participants.

More studies are needed that compare long-term former smokers who do and do not vape, as well as studies comparing former smokers who vape with people who vape who have never smoked.

Research on vaping needs to measure strength of evidence

As many studies involve small numbers of participants, researchers should use other, less traditional ways to test their findings. This could include using a Bayes factor analysis to measure the strength of evidence. This is relevant to findings from most of the health biomarker studies included in this report.

Switching to vaping likely to slow down respiratory disease development

For policy makers and practitioners, the limited evidence from our review for this chapter suggests that developing and implementing policies and interventions that support smokers to completely stop and switch to vaping is likely to slow down the development of respiratory diseases.

Consider respiratory biomarkers before starting studies

Researchers need to carefully consider their choice of respiratory biomarkers before carrying out their studies. While some found statistically significant changes in spirometry measures, it is not clear whether these changes are too small to be clinically relevant. This raises the question of how useful spirometry measures are in detecting any vaping risks, particularly among healthy smokers. This concern also applies to other biomarkers, such as inflammatory changes. Also, the pathways between these biomarkers and an increased risk of certain respiratory diseases still needs to be clearly mapped out with supportive evidence.

Considerations for human cell studies

For human cell studies, biologically relevant doses of nicotine or flavours that mimic exposure to vaping product aerosol emissions are needed.

Seek a global consensus on measuring changes to the respiratory system

Studying changes to the respiratory system is important as these might be the first signals of potential harms or (relative) benefits from vaping. So, seeking a global consensus on what measures should be studied, as well as over what duration of exposure and follow-up, is urgently needed.

Assess effects of vaping on people with existing respiratory problems

More studies are needed that assess the effects of vaping on people with pre-existing respiratory problems or diseases. This includes both in comparison with no use of nicotine or tobacco and in comparison with smoking.

Chapter 11: Cardiovascular diseases

Evidence reviewed

For cardiovascular diseases, we did not identify any studies on people with existing cardiovascular conditions so we could not address the second aim of the review. We assessed both relative and absolute vaping risks associated with biomarkers of cardiovascular disease where the data were available (between vapers and smokers, and between vapers and non-users). And where feasible, we included comparisons across different population groups.

We present our conclusions for biomarkers of exposure and biomarkers of potential harm across several diseases in chapters 7 and 8.

The studies we reviewed show that compared to smoking, using vaping products leads to a substantial reduction in biomarkers of toxicant exposure. However, the degree of any residual risk (from vaping but also previous smoking and other factors affecting cardiovascular health) remains unclear, mainly because of the lack of studies using appropriate comparators.

Main findings

Cholesterol

Looking at biomarkers of potential harm relevant to multiple diseases, studies of low-density lipoprotein (LDL) cholesterol showed no differences after acute and short-to-medium use of vaping products, smoking or non-use. LDL cholesterol is sometimes described as ‘bad cholesterol’ as it makes heart problems or a stroke more likely. Similar findings were seen for high-density lipoprotein (HDL) cholesterol (or ‘good cholesterol’), except among large-scale samples of non-users where HDL levels were significantly higher than among vapers and smokers.

Oxidative stress

The findings were more mixed for markers of oxidative stress 8-isoprostane and sNOX2-dp. However, as these oxidative stress biomarkers are influenced by other factors, we could not make strong conclusions on their associations with vaping product use.

Inflammation

For inflammation markers, differing study designs prevented us from making strong conclusions. The meta-analyses of cross-sectional studies suggested lower levels of the inflammation biomarkers (blood CRP and sICAM-1) among vapers than smokers, and similar levels between vapers and non-users. But these findings were not confirmed by other interventional studies that largely focused on acute and short-term exposure.

Endothelial function

For endothelial function biomarkers, a single RCT found that switching from smoking to vaping improved FMD after one month. Evidence from the other studies suggested a short-term deterioration in FMD after acute exposure to vaping product use. Evidence from the other endothelial function biomarkers and the 4 studies on platelet activation markers was also difficult to synthesise. This was due to different designs, outcome measures and comparison groups.

Harm specific to cardiovascular disease

We identified 41 studies that assessed biomarkers of potential harm specific to cardiovascular disease in humans. Consistent with studies in other chapters, the studies we included:

• used a range of different designs
• had varying quality or risk of bias
• used a range of different definitions of vaping and smoking

Studies with more than one time point mostly explored acute exposure to vaping or followed-up participants for short to medium term. So, we were unable to summarise findings on longer-term vaping exposure. In line with our algorithm (chapter 2, table 6), we carried out meta-analyses wherever possible, but a lack of consistency in study designs, outcome reporting, group definitions and exposure periods resulted in data from few studies being meta-analysed.

Heart rate

Thirty-one studies assessed heart rate in humans (4 studies from the UK), and 9 of them could be included in meta-analyses. We were able to conduct 2 meta-analyses of findings comparing vaping and smoking (3 cross-over and 2 cross-sectional studies), 2 meta-analyses of findings comparing vaping and non-use (3 cross-over, 2 cross-sectional studies) and one meta-analysis of findings comparing vaping and non-nicotine vaping (4 cross-over studies).

Acutely, immediately after use, vaping increased heart rate less than smoking. Heart rate after short exposure to vaping was similar to heart rate after not using tobacco or nicotine products. There was no difference in heart rate after nicotine and non-nicotine vaping. Any differences may vary with devices, liquids and puffing behaviours influencing the amount of nicotine delivered and this is further explored in chapter 5 on nicotine.

Comparing longer-term changes in heart rate, people who vaped had lower heart rate than people who smoked when the groups were mutually exclusive (people who vaped did not also smoke). Compared with people who did not vape or smoke, heart rate among people who vaped was lower in a meta-analysis of cross-sectional studies but higher in another cross-sectional study. One longer-term study reported the same level of change in heart rate for smokers who started using nicotine or non-nicotine vaping products.

Blood pressure

Thirty studies assessed blood pressure in humans (3 studies from the UK), with 9 studies that could be included in meta-analyses. We conducted 4 meta-analyses of findings comparing blood pressure when vaping and smoking (3 cross-over studies, 2 cross-sectional studies, meta-analysis repeated for systolic (when your heart beats) and diastolic (when your heart rests between beats) blood pressure), 4 meta-analyses of findings comparing vaping and non-use (3 cross-over and 2 cross-sectional studies, again for both systolic and diastolic blood pressure) and 2 meta-analysis comparing nicotine and non-nicotine vaping (4 cross-over studies, again for both systolic and diastolic blood pressure).

Meta-analyses comparing acute effects found no differences in blood pressure after vaping, smoking or doing neither with the exception of a small difference between vaping and non-use for diastolic blood pressure. Studies that could not be meta-analysed found mixed results. A meta-analysis comparing acute effects of nicotine and non-nicotine vaping found no difference as did most other studies that could not be meta-analysed but included non-nicotine vaping.

Meta-analyses of cross-sectional studies where participants had had longer exposure to vaping (at least 3 months or one year) found that people who vaped (presumably mostly former smokers) had lower blood pressure than people who smoked. There was no difference between people who vaped and people who did not vape or smoke. Studies that could not be meta-analysed found mixed results regarding change in blood pressure.

Secondhand exposure

Only 2 small studies at serious risk of bias included secondhand exposure. So, we could not draw conclusions about what effects exposure to secondhand vapour has on heart rate or blood pressure.

Peripheral resistance and arterial stiffness

Nine studies assessed peripheral resistance or arterial stiffness (PWV) in humans (one study from the UK). Results could not be meta-analysed. PWV may decrease (improve) after smokers have switched to vaping for a sustained period. However, the longest follow-up reported was only 4 months.

PWV generally increased after acute exposure to vaping nicotine, but not after non-nicotine vaping, suggesting that any acute effects of vaping on PWV are due to nicotine.

Oxygen saturation

Three studies (all at critical risk of bias, none from the UK) assessed acute effects on oxygen saturation in humans. Results could not be meta-analysed, and we could not draw conclusions based on the available evidence.

Cell and animal studies

Evidence from cell studies was very limited, with only 2 studies identified in our review. Results showed that vaping product aerosol increased damage to cells and that effects varied across different flavours.

Sixteen studies in animals were included. In summary, animal studies showed that exposure to vaping product aerosol increases blood pressure. Some studies found a decrease in heart rate, although most found no effect.

Animal studies also show an increase in biomarkers of arterial stiffness linked to exposure to vaping products. This may be similar to or smaller than increases caused by smoking. Left ventricular mass and vessel wall thickness (in the heart) were increased and left ventricular function reduced after vaping product aerosol exposure. These effects were potentially less than for exposure to cigarette smoke, and there were inconsistencies in findings across studies. These vaping product-induced effects appeared largely to be nicotine-dependent.

Exposure to vaping product aerosol was associated with decreases in animals’ blood vessel health, as well as increases in markers of thrombosis risk, inflammation, oxidative stress, scarring, and cell health. Although, it is inconclusive as to which constituents of the aerosol play important roles in the observed effects.

As previously mentioned, it is challenging to directly translate the findings from pre-clinical studies using human or animal cells or rodent models to any cardiovascular risks arising from vaping in humans. These pre-clinical studies commonly employ acute exposures, sometimes over concentrated periods, and it is unclear whether the mechanisms or pathways to risk identified would be replicated in people who vape.

Role of nicotine

The evidence does not allow us to distinguish pathways to cardiovascular disease. One potential pathway is through nicotine, and the biomarkers of exposure and pharmacokinetic studies show that people who vape can achieve nicotine levels similar to people who smoke. The animal studies suggested that nicotine did play a role in some of the changes seen in cardiovascular biomarkers, specifically:

• blood pressure
• arterial stiffness
• left ventricular mass and function

Some studies included in this chapter assessed cardiovascular biomarkers in humans through non-nicotine vaping as well as nicotine vaping. This could help explain the assumed role of nicotine in any cardiovascular risks of vaping for humans. However, the differences between studies limits our conclusions.

Meta-analyses of cross-over studies from vaping nicotine and non-nicotine products for heart rate and blood pressure found no differences. Studies that we could not meta-analyse did not consistently find this. The findings were more consistent in PWV effects where nicotine did appear to be implicated at least in acute studies.

Comparisons with other reports

Conclusions from the NASEM report are generally supported by this review. As in 2018, to date there is still no available evidence on whether vaping is associated with clinical cardiovascular outcomes (coronary heart disease, stroke, and peripheral arterial disease) and subclinical atherosclerosis (carotid intima-media thickness and coronary artery calcification).

The NASEM report found substantial evidence that heart rate increased shortly after nicotine intake from vaping, which was also seen in this review (whereas evidence was inconsistent for non-nicotine vaping).

NASEM found moderate evidence that diastolic blood pressure increases shortly after nicotine intake from vaping and limited evidence that vaping is associated with a short-term increase in systolic blood pressure. Based on the still limited and mixed evidence, we conclude that there may be reductions in blood pressure after people who smoke switch to vaping and little difference between people who vape and people who do not vape or smoke.

The NASEM report also concluded that there was insufficient evidence that vaping was associated with long-term changes in heart rate, blood pressure, and cardiac geometry and function. In our review, evidence from animal studies suggests that there may be some long-term changes, but we found no evidence from human studies. And, as already discussed, the validity of animal studies for human outcomes has limitations.

Similarly, conclusions by COT are generally supported by this review. COT concluded that exposure to nicotine from vaping was unlikely to be higher than from smoking. This is confirmed by studies included in this review that found no significant difference between people who vaped or smoked at least weekly.

COT also concluded that vaping was associated with some emissions into ambient air, including nicotine, so that pharmacological effects from exposure to nicotine in ambient air may occur in some individuals. In this review, only 2 small studies at serious risk of bias assessed short-term secondhand exposure to nicotine vaping. So, this did not allow us to make any clear conclusions.

Conclusion

Overall, the extent to which vaping presents a risk for cardiovascular health remains uncertain. But based on the toxicant profile in vaping products and aerosols, the risk is expected to be much less than that of cigarette smoking.

Implications

Our quality assessments revealed most studies had some methodological concerns, and these should be addressed in future research as they limit interpretations of our findings. More research is needed, particularly in the UK, where we identified a lack of studies.

Most studies exposed participants to brief sessions of vaping (27 out of 41 included studies were cross-over or acute exposure studies). And although it can address questions about immediate effects of vaping, this study design is not able to answer questions about effects on the cardiovascular health outcomes most relevant to public health.

Studies that compare rates of cardiovascular diseases between non-users, users of tobacco and users of nicotine vaping products are needed (for example, rates of coronary heart disease, peripheral arterial disease and stroke).

Studies should include longer-term follow-ups and more informative measurements. Studies measuring heart rate or blood pressure should try to include 24 hour ambulatory blood pressure and heart rate. This would improve the validity of the measurement rather than rely solely on measurements in single or short sessions. Researchers should consider including heart rate variability (a higher variability can indicate better health) as an outcome measure, for example in people who switch from smoking to vaping. Evidence is also needed on the extent of longer-term changes in other outcomes such as PWV. Alongside longer follow-ups, inclusion of long-term exclusive vapers may also help address this.

Historical tobacco use can greatly affect many of the biomarkers used to determine exposure to potentially harmful constituents from vaping. As most vapers are previous long-term smokers (see chapter 4 on vaping among adults), definitions for vaping should preclude concurrent smoking and a minimum duration of exclusive vaping should be defined. Studies are needed that compare long-term former smokers who do and do not vape, as well as studies comparing former smokers who vape with people who vape who have never smoked.

Compliance with study allocation and definitions of groups should be verified and reported in all studies. For example, the level of CO exhaled by people categorised as not smoking and the level of nicotine in people categorised as vaping or not using any nicotine products.

The existing evidence does not provide insights into the effects of vaping on cardiovascular health in people of different sex, age or ethnicity. So, future research should pay attention to groups with different cardiovascular risk profiles.

Studies are needed that assess the effects of vaping on people with pre-existing cardiovascular conditions, both in comparison with not using nicotine or tobacco and in comparison with smoking.

Cardiovascular health and disease are affected by a wide range of genetic predispositions, behavioural risk factors and environmental exposures. Further research is needed to clarify any unique contributions from vaping while accounting for other factors.

Vaping products vary and any effects on cardiovascular health are likely to differ with device types, nicotine concentration, liquid composition and user behaviours. As one example, most studies in the US used nicotine concentrations above the legal threshold in the UK and EU, but we were unable to run meta-analyses comparing effects of nicotine concentration on outcomes.

For policy makers and practitioners, findings from our review for this chapter suggest that developing and implementing policies and interventions that support smokers to completely switch from smoking to vaping will reduce exposure to toxicants and carcinogens that have links with poorer cardiovascular health.

Chapter 12: Other health outcomes

Evidence reviewed

In this chapter, we address health outcomes not covered in the chapters on the main causes of smoking-related illness and death. From our systematic review, we identified 15 studies in humans that looked at outcomes related to dental health. We also identified 14 studies in humans, 31 in animals and one in cells that investigated other health outcomes.

Studies in humans have assessed associations with a range of health outcomes including oral, ocular and reproductive health, as well as outcomes related to allergies and pre-diabetes. The health outcomes assessed covered a limited range; all were detrimental to health and none of the included studies explored potential positive effects of nicotine or vaping. For instance, no study looked at the effects on Parkinson’s disease, where some have suggested a protective effect of nicotine.

Main findings

Limitations of the evidence

Many studies found that health outcomes for people who vaped were worse than for people who did not vape (or did not smoke) while others found no differences. However, while some studies included large samples, they were almost exclusively cross-sectional in design, making any causal statements impossible.

Studies used a range of different definitions of vaping and smoking. For example, findings of some studies were confounded by categorising vapers who smoke, occasional vapers or exclusive daily vapers as a uniform group or comparing occasional vapers with daily smokers. So, findings need to be cautiously interpreted. Definition of user groups, information on and comparisons with smoking were often lacking or confounded the findings.

Many studies were at risk of bias and other factors (for example, genetic, lifestyle and environment) influencing health outcomes were often not considered, further limiting the validity of findings.

Reproductive health

The evidence base on reproductive health or pregnancy outcomes remains insufficient. Previous reports only found a single study indicating that vaping in pregnancy had little or no effect on birth weight. We were not able to add further evidence to these.

Oral or dental health

Oral or dental health has been researched more extensively than other health areas However, the quality of the studies was often low. Recent reviews concluded that vaping would be detrimental to oral or dental health among people who have never vaped or smoked but would likely be beneficial for smokers switching. We found no studies that would change that conclusion.

Cell and animal studies

The one cell and 31 animal studies provided insights into how vaping products may affect the central nervous, digestive and reproductive systems. They also looked at other areas that exposure to tobacco or no exposure could affect. However, the data are still limited and too inconsistent to evaluate the compounds of vaping product aerosol causing any alterations to systems in the body. Also, variability of animal models, exposure methods and comparators added to the uncertainty.

Implications

Good quality studies in humans are needed that investigate the effects of vaping on a wider range of physical and mental health outcomes. They should also explore the progression of various health disorders in people who vape compared with people who smoke or do not vape nor smoke.

Also, although cancer, respiratory and cardiovascular diseases are the main contributors of tobacco related disease, there is a lack of research on the effects of vaping on other areas, such as renal and hepatic systems, which can be greatly affected by smoking.

Effects of vaping on fetal development and pregnancy outcomes remain in particular need of research, including the effects of switching from smoking to vaping in the perinatal phase.

Chapter 13: Poisonings, fires and explosions

Main findings

Poisonings

In 2021, the National Poisons Information Service (NPIS) reported that they had received 187 vaping product enquiries out of a total of 39,594 telephone enquiries. Of these, 82 involved children aged 5 years or younger. This equates to at least one telephone enquiry every other day involving a healthcare professional managing someone who has apparently been exposed to vaping products.

Two case reports of poisoning from vaping products in the UK were identified, both intentional. In one of the cases, the person died.

In non-UK poisonings, according to data from a 2020 annual report by the American Association of Poison Control Centers’ National Poison Data System, one person died from vaping product use (no details were given of the circumstances). In 20 studies from international poisons and surveillance centres and case reports identified in our systematic review, most participants were young children who accidentally swallowed e-liquids. Almost all children recovered, although there were 2 deaths among the children who were accidentally exposed to e-liquid. Where exposure was intentional or unknown, there were reports of 16 deaths (outside the UK).

Accidental ingestion is the most common cause of poisonings, with fewer incidences of other routes such as ocular (eyes) exposure.

Incidents of poisoning in children are often preventable.

Fires

Between January 2017 and October 2021, the London Fire Brigade reported that there were 5,706 fires caused by cigarettes and cigarette lighters. This compared to 15 fires caused by vaping products. No fire related injuries or deaths were reported from vaping related fires, compared with 676 injuries and 46 deaths from cigarette related fires. These findings are similar to those we discussed in our 2018 report.

Explosions

Exploding vaping products can cause severe burns and injuries that require intensive and prolonged medical treatment, especially when they explode in users’ hands, pockets or mouths.

Incidents appear to be serious but very rare.

We identified 2 case reports involving 4 people in the UK. One involved an explosion in the mouth while vaping, the other 3 involved explosions when the vaping product was being carried in trouser pockets. No deaths were reported.

There were 23 reports identified outside the UK, from case reports and series or data from burn and surveillance of injury centres. Carrying the vaping product in a trouser pocket was again the most common cause of explosions. One death was reported.

Implications

More research on vaping-related poisonings, fires and explosions

There is a lack of UK research or published case reports on poisonings, fires and explosions involving vaping products. The findings reported here are largely from the US and cannot be assumed to be applicable to the UK given the different regulatory frameworks for vaping products.

More research is needed on the type of vaping product resulting in poisoning, fires and explosions. This would then inform future regulations.

Information on poisonings, fires and explosions should be monitored and reported routinely in publicly available reports by relevant authoritative bodies.

Warnings on labelling and devices

Two explosions were identified as caused by mechanical modifiable tank devices, which do not have inbuilt safety features. So, warnings could be highlighted for users of these products by relevant authoritative bodies.

As well as childproof packaging, regulations should require labelling to reinforce safe storage and away from similar looking medicines, such as eye or ear drops and children’s medicine.

Advice on transporting vaping products and batteries

Additional advice by relevant authoritative bodies could be given on transporting vaping products and batteries for example using specialised containers, to avoid thermal runaway incidents (where a battery discharges all its stored energy at once),

Chapter 14: Heated tobacco products

Main findings

Use of heated tobacco products in England

Among young people aged 11 to 18 in the 2021 ASH-Y survey, 0.9% had tried but no longer used heated tobacco products (HTP) and 0.3% reported currently using HTP.

Among young people aged 16 to 19 in the ITC Youth survey, 1.5% had ever tried HTP but not used them in the past week and 0.7% had used HTP in the past week.

Two-thirds (65.7%) of young people aged 16 to 19 who had ever tried HTP had used it once or up to 10 times only.

Among adults in England, 0.3% in the STS and 0.5% in the 2021 ASH-A survey reported currently using HTP.

The proportion of adults who reported having ever used HTP was 1.8%. It was more common among people aged 25 to 34, women and adults who smoked or vaped.

One third of ever or current adult users of HTP had tried HTP once or twice and 16% of current users (less than 0.1% of adults in England) reported daily use.

Among past year smokers who had attempted to stop smoking, 1.6% reported having used HTP to support their attempt.

Cochrane review

The Cochrane review of HTP for smoking cessation and reducing smoking prevalence reported no studies on HTP used to support cessation of cigarette smoking, so the effectiveness of HTP for stopping smoking remains uncertain.

The Cochrane review found moderate certainty evidence that smokers switching to HTPs have lower exposure to toxicants and carcinogens than smokers continuing to smoke. There was moderate to very low certainty evidence of higher exposure than for people attempting abstinence from all tobacco.

There was some evidence for people improving the amount of air they can exhale from the lungs (FEV1) after switching to HTP compared with continuing to smoke. But there was insufficient evidence of any difference for other biomarkers of harm.

There was insufficient evidence for differences in risk of adverse or serious adverse events between people randomised to switch to HTP, smoke cigarettes or attempt tobacco abstinence in the short-term.

The rate of decline in cigarette sales accelerated after Japan made HTP available. However, it is possible that other factors caused this change. A decline in cigarette sales may not translate to declining smoking prevalence, and changes in Japan may not apply elsewhere.

Implications

Monitoring of HTP uptake among young people and adults should continue.

Research independent of manufacturers is needed into whether HTP help people stop smoking, their safety, and their impact on smoking rates.

Chapter 15: Harm perceptions and communications

Evidence reviewed

This chapter drew on surveys carried out in chapters 3 (vaping among young people) and 4 (vaping among adults) and a systematic review that addressed the following questions:

1. What interventions have been effective in changing vaping harm perceptions?
2. To what extent are vaping harm perceptions predictive of any changes in vaping and smoking behaviours?

Main findings

Young people’s harm and other perceptions of vaping in England

Among 11 to 18 year olds, using 2021 ASH-Y data:

• 44.7% accurately perceived that vaping was less harmful than smoking
• 32.4% inaccurately thought that the harms from vaping and smoking were about the same
• 3.6% inaccurately thought that vaping was more harmful than smoking
• 19.3% said they did not know

The proportion of 11 to 18 year olds who accurately thought that vaping was less harmful than smoking declined from 66.7% in 2015 to 43.3% in 2020, and then increased slightly in 2021 to 44.7%. The proportion who did not know has increased from 9.9% in 2015 to 19.3% in 2021.

Among 11 to 18 year olds, inaccurate perceptions that vaping is more or equally as harmful as smoking were similar between young people who currently vaped and those who never vaped. Only half of current smokers aged 11 to 18 years accurately perceived vaping as less harmful than smoking.

Among 16 to 19 year olds (using ITC Youth data), we see slightly different patterns in 2021, with most (62.9%) accurately perceiving vaping is less harmful than smoking. Yet, we also saw:

• 16.8% inaccurately perceived vaping to be equally harmful to smoking
• 10% inaccurately perceived vaping to be more harmful than smoking
• 10% reported that they did not know

In relation to absolute harms, young people (16 to 19 year olds) rated smoking daily higher on the scale of harm than smoking on some days (88% compared with 65.2% rating it ’very’ or ‘extremely’ harmful). However, there was less difference between young people’s perceptions of vaping daily and vaping on some days (31.9% and 22.6% respectively). Slightly greater proportions of young people perceived some day or daily vaping as not at all harmful (6.2% and 2.8% respectively) than they did for smoking (both 0.6%). A greater proportion of young people did not know the harms of vaping (about 11.5%) than did not know the harms of smoking (less than 1%).

Half of 16 to 19 year olds perceived vaping to be ‘slightly’ or ‘somewhat’ addictive (50.7%), one-third perceived vaping to be ‘very’ or ‘extremely’ addictive (31.7%), and few (6.3%) perceived vaping to be ‘not at all’ addictive, with 11.1% saying they did not know.

Over half of 16 to 19 year olds perceived that vaping makes quitting smoking permanently ‘a bit’ or ‘a lot easier’ (60%). Many (14.2%) thought it had ‘no effect’, just under one-tenth (9.6%) perceived that vaping made quitting ‘a bit’ or ‘a lot harder’, with 15.9% saying that they did not know.

Overall, just over half of 16 to 19 year olds reported noticing any education campaign or public health message about vaping in the past 12 months (53%).

Adult smokers’ and vapers’ harm perceptions of vaping in England

Among adult smokers in 2021 STS data, just over a third (34.1%) accurately perceived that vaping was less harmful than smoking. But around a third (32.1%) inaccurately thought that the harms from vaping and smoking were about the same, 11.9% inaccurately thought that vaping was more harmful than smoking, and 22% said they did not know.

The proportion of adult smokers who inaccurately perceived that vaping was more harmful or equally harmful than smoking has declined since 2020 by 2.9 and 5.6 percentage points respectively. The proportion of smokers who accurately perceived that vaping is less harmful than smoking increased by 5 percentage points since 2020 (the first time we have seen an increase in this measure since 2014). However, there seems to be growing confusion about the relative harms of vaping compared with smoking. STS found that the proportion of adult smokers who said that they did not know whether smoking or vaping was more harmful has more than doubled from 9.5% in 2019 to 22% in 2021.

In the ASH-A survey, overall, few (13.9%) current adult smokers and vapers accurately believed that none or a small amount of the risks of smoking were due to nicotine, with:

• 23.9% reporting ‘under half the risk’
• 17.3% reporting ‘around half the risk’
• 26.9% reporting ‘much more than half’ or ‘nearly all’ the risk
• 18.1% reporting that they did not know

There was a notable gradual increase in correct nicotine risk perceptions among adults depending on participants’ experience with vaping. The proportions that correctly reported that ‘none’ or ‘a very small amount’ of the health risks from smoking come from nicotine in tobacco cigarettes included:

• 10.8% of current smokers
• 15.6% of smokers and vapers
• 20.3% of exclusive vapers

Systematic review of vaping harm perceptions

We have included a systematic review of vaping harm perceptions examining interventions to change them, and longitudinal associations with vaping and smoking behaviours.

Interventions to change perceptions

We identified 32 articles (from 29 studies) addressing our first research question:

1. What interventions have been effective in changing harm perceptions?

Studies involved either adults or young people, and addressed:

• relative perceptions of the harms of vaping (compared with smoking)
• absolute perceptions of the harms of vaping or addictiveness (vaping compared to non-use of tobacco or nicotine products), such as the perception that e-cigarettes contain harmful chemicals, cause heart disease or cancer, or that vaping is addictive)
• perceptions of the harms of nicotine (including perceived addictiveness of nicotine)

Of the 32 articles, there were:

• 13 articles (from 10 studies) assessing interventions involving written information about vaping
• 4 studies assessing educational workshops or videos designed to deter vaping
• 5 studies assessing mass media campaigns or advertisements
• 3 studies assessing warning labels and packaging
• 3 studies assessing video games aimed to prevent youth vaping
• 4 studies assessing whether vaping harm perceptions changed after the EVALI outbreak

Our review found that interventions communicating information about the reduced harms of vaping relative to smoking generally increased people’s perceptions that vaping is less harmful than smoking. Most of this evidence came from studies of adults.

We also found that interventions communicating information about the absolute harms of vaping (vaping compared to non-use of tobacco or nicotine products) generally increased the perception that vaping:

• is harmful to health
• can lead to developing diseases or other health issues
• is equally or more harmful than (relative to) smoking

Most of these interventions were aimed at young people or young adults specifically to deter them from vaping by providing information about vaping harms.

EVALI increased people’s harm perceptions of vaping, including inaccurate perceptions relative to smoking.

Warning labels highlighting that vaping is harmful and addictive generally increased people’s perceptions that vaping is harmful to health and is addictive.

Vaping harm perceptions predicting changes in behaviour

We identified 21 studies addressing our second research question:

2. To what extent do vaping harm perceptions predict any changes in vaping and smoking behaviours?

Studies assessed young people, young adults or adults, and assessed associations between vaping harm perceptions and vaping and smoking behaviours.

For vaping among young people and young adults:

• 14 studies assessed associations between vaping harm perceptions and changes in vaping behaviours
• 3 studies assessed associations between vaping harm perceptions and changes in smoking behaviours

For vaping among adults:

• 6 studies assessed associations between vaping harm perceptions and changes in vaping behaviours
• 3 studies assessed associations between vaping harm perceptions and changes in smoking behaviours

Our review found that vaping harm perceptions consistently predicted subsequent changes in vaping behaviours among young people, young adults and adults.

Perceiving vaping as less harmful than smoking predicted subsequent increases in vaping (including starting vaping) among young people and young adults, but also among adults and adult smokers. Conversely, perceiving vaping as harmful was associated with not starting vaping among young people and young adults.

Substantially fewer studies assessed whether people’s vaping harm perceptions predicted subsequent changes in their smoking behaviours. However, the limited evidence suggests that perceiving vaping as equally or more harmful than smoking predicted subsequent relapse to smoking among adult former smokers. Also, perceiving vaping as less harmful than smoking predicted quitting smoking. But among young people and young adults, relative and absolute harm perceptions (sometimes including perceived risk of addiction) were not associated with starting smoking. Absolute harm perceptions were not associated with smoking more.

In general, the findings were broadly consistent with people’s normal expectations for approaching what they perceive to be lower harm and avoiding what they perceive to be greater harm.

Taken together, the findings suggest that messages about the harms of vaping influence vaping perceptions. This in turn affects people’s vaping and smoking behaviours.

Providing information aimed to deter young people from vaping (for example, highlighting the harms of vaping) can increase their perceptions of the harm of vaping to health, which in turn can deter them from trying vaping. Conversely, providing information aimed to increase accurate relative perceptions of vaping compared to smoking can increase accurate relative perceptions of vaping compared with smoking. This could lead adult smokers to try vaping, reduce risk of relapse to smoking among adult former smokers who vape, but it could also lead to young people trying vaping.

The effects of vaping harm perceptions on longer-term vaping, smoking, and vaping as a substitute for smoking, remain unclear.

Risk of bias was high for all included studies for both our research questions.

Implications

The need for carefully designed interventions

Given a substantial proportion of young people and adult smokers and vapers in England still hold inaccurate perceptions of the relative harms of vaping compared with smoking (that vaping is equally or more harmful than smoking), these misperceptions need to be addressed.

Providing accurate information about the relative harms of vaping, and risks of using nicotine, could help to correct misperceptions of vaping and nicotine, respectively, particularly among adults.

Interventions on absolute harms of vaping need to be carefully designed so as not to misinform young people (particularly smokers) about the relative harms of smoking and vaping.

The need for research on effects of warning labels highlighting relative harms of vaping and smoking

Warning labels highlighting that vaping is harmful and addictive generally increased perceptions that vaping is harmful to health and is addictive. No studies assessed the effects of warning labels highlighting the relative harms of smoking and vaping, on relative harm perceptions. So, these studies are needed.

Other research needed

No studies among young people or young adults assessed whether vaping harm perceptions predicted subsequent switching from smoking to vaping, or the other way around. So, studies addressing substituting smoking with vaping in young people, young adults and adults are needed.

More longitudinal randomised studies assessing interventions to change vaping harm perceptions are needed. There is also a need for studies that assess whether changes in vaping harm perceptions (in response to interventions) and vaping and smoking behaviours (associated with harm perceptions) are maintained over time (particularly into adulthood).

Importance of effective communications

Communications about absolute and relative harms of vaping and smoking are likely to reach both young people and adults. From an ethical standpoint, the main aim of these communications must be to ensure that the messages give accurate information about absolute harms of vaping, and the relative harms of vaping compared to smoking, to address the prevalent misperceptions. Messages will need to be carefully developed and nuanced to avoid unintended effects (for example, ‘less harmful’ translating to a perception of ‘safe’) and should be tested on target audiences first. Finally, continued surveillance of perceptions in young people and adults is needed.

Chapter 16. Conclusions

In this chapter, we summarise the findings from each chapter and pull together the above findings in the context of the series of evidence reviews since 2015. We also present the conclusions from the systematic reviews in the form of evidence statements. We then present overall implications for policy, practice and research.

Despite the increase in research on vaping since 2015, weaknesses around the choices of assessments and biomarkers, populations, user groups and exposure, and study designs all limit the conclusions that we can draw.

Overall findings in the context of our series of evidence reviews

We have previously stated, in our 2015 report, vaping poses only a small fraction of the risk of smoking and is at least 95% less harmful than smoking (that is, smoking is at least 20 times more harmful to users than vaping). This was to help the public and health professionals make sense of the difference in the magnitude of risk between vaping and smoking.

We are aware that summarising the relative risks of vaping versus smoking across a range of different products and behaviours and assessed across multiple biomarkers can be simplistic and misinterpreted. Based on the reviewed evidence, we believe that the ‘at least 95% less harmful’ estimate remains broadly accurate, at least over short term and medium term periods. However, it might now be more appropriate and unifying to summarise our findings using our other firm statement: that vaping poses only a small fraction of the risks of smoking. As we have also previously stated and reiterate, this does not mean vaping is risk-free, particularly for people who have never smoked.

This magnitude of relative risk between vaping and smoking is not reflected in current public perceptions which, as our review has shown, can be influenced by interventions.

Evidence statements

In the chapters which reported on our systematic literature reviews of the health risks of vaping and harm perceptions, and the 2022 Cochrane review on heated tobacco products, we listed 61 evidence statements. These statements are based on the strength of evidence, given the quality of the studies we reviewed and their findings. The statements broadly follow the definitions of level of evidence in the NASEM report. As NASEM noted, the framework is a guide, but a great deal of expert judgement, in our case by the co-authors of our report, is also involved.

Recommendations for research

We made a number of recommendations for research. These included:

• involving people who currently smoke or vape to help shape and design research to ensure research questions are relevant, interpret the evidence and support dissemination
• agreeing a common set of biomarkers of exposure and potential harm to be used
• standardising the definitions of who is involved in the research, their exposure to vaping and smoking, and how studies report details of the devices involved
• agreeing protocols for the different designs of studies used
• greater transparency to reduce bias in research, for example pre-registration of study protocols and analytical plans

Overall implications

Evidence from stop smoking services and the Cochrane living review for smoking cessation (not covered in our report) shows that vaping is effective for stopping smoking. These findings, along with our findings that vaping carries a small fraction of the health risks of smoking, suggest that smokers should be encouraged to use vaping products (or medicinally licensed products) for stopping smoking, or as alternative nicotine delivery devices to reduce the health harms of smoking.

Our findings of higher absolute exposure to toxicants from vaping, compared with not using any nicotine products, reinforce the need to discourage people who have never smoked from taking up vaping (or smoking). Cuts to government bodies responsible for overseeing vaping products are concerning. The recent increase in young people using disposable vaping products makes this an even greater concern, because if it continues, it could undermine the approach and regulatory framework for vaping products adopted in England.

As well as educational materials aimed at older smokers on why and how to vape to stop smoking, educational materials are also needed for young people starting vaping who would otherwise not have smoked, and for those who need support in stopping smoking.

It is vital that surveys that assess smoking and vaping are adequately resourced and maintained over time to enable long term trends to be assessed. For example, it would be useful for the Adult Population Survey to include questions about nicotine vaping product use, given the prevalence of vaping

Public perceptions of absolute and relative vaping harm are not in line with the evidence and our findings indicate that these perceptions influence subsequent vaping and smoking behaviours. We also found that interventions can influence perceptions. So, understanding and changing misperceptions is very important.

Systematic reviews are resource intensive, and since our July 2021 cut-off date for searching the relevant literature for the health chapters, new studies have been published. Future evidence reviews of the health harms of vaping should adopt a continual approach to updating the literature, similar to the living systematic review for e-cigarettes for smoking cessation by the Cochrane Tobacco Addiction Group. This would ensure that relevant new evidence would be incorporated as it becomes available, and would help policy makers to use the most up to date evidence.

Authors and citation

Authors

Ann McNeill, Erikas Simonavičius, Leonie Brose, Eve Taylor, Katherine East, Elizabeth Zuikova, Robert Calder, Debbie Robson

King’s College London

Citation

McNeill, A, Simonavičius, E, Brose, LS, Taylor, E, East, K, Zuikova, E, Calder, R and Robson, D (2022). Nicotine vaping in England: an evidence update including health risks and perceptions, September 2022. A report commissioned by the Office for Health Improvement and Disparities. London: Office for Health Improvement and Disparities.