Prevalence and determinants of e-cigarette use in South Asia: a systematic review

Abstract

Introduction

E-cigarette use has become a significant global public health concern. Recently, several systematic reviews have synthesised evidence on e-cigarette use in Europe, America and Southeast Asia. However, there is no known systematic review of such on South Asia (SA). This systematic review aimed to systematically synthesise empirical evidence on the prevalence and determinants of e-cigarette use in SA.

Methods

This systematic review adhered to Preferred Reporting Items for Systematic Review and Meta-Analysis guidelines and Cochrane Handbook for Systematic Reviews of Interventions. Nine databases, as well as additional sources, were searched to retrieve relevant articles on the prevalence and determinants of e-cigarette use in SA. Only those peer-reviewed articles which met the eligibility criteria were included in this systematic review. Data were extracted from the included articles and synthesised using a narrative synthesis approach.

Results

Seventeen articles were included in this systematic review and all were graded as above average after quality appraisal. Lifetime prevalence of e-cigarette use was reported in three SA countries only: India (0.1–23.8%), Bangladesh (0.4–0.5%) and Nepal (21.2%). The prevalence of current e-cigarette use was reported in four SA countries only: Pakistan (6.20–10.1%), India (0.01–5.39%), Nepal (5.9%) and Bangladesh (0.5%). The following factors were the reported determinants of e-cigarette use that were significant (p<0.05) among SA populations: self-perception of e-cigarette use as addictive, less harmful, an alternative to tobacco smoking and as a safe product; leisure use; individual appeal to the flavours of e-cigarette; individual stress; presence of mental depression; peer influence; social acceptance of e-cigarette use and prices of e-cigarettes.

Conclusion

Empirical evidence on e-cigarette use is limited to specific sociodemographic, socioeconomic and geographical regions in SA. Additional empirical investigations are required to outline the prevalence and determinants of e-cigarette use in SA countries lacking adequate data. Also, there is a need for tailored interventions targeting e-cigarette use in SA.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• E-cigarette use in South Asia (SA) is a public health issue of concern. No known systematic review has synthesised evidence on the prevalence and determinants of e-cigarette use in SA.

WHAT THIS STUDY ADDS

• This study provides a comprehensive overview of the prevalence of e-cigarette use in SA and identifies key determinants influencing its use.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The synthesised region-specific findings on e-cigarette use in SA offer valuable insights for tailored public health interventions and policies needed for the effective prevention and control of e-cigarette use in SA.

Introduction

Tobacco ranks as the second most commonly used psychoactive substances globally, with around 1.18 billion regular tobacco smokers and 7 million tobacco-related deaths reported in 2020.^{1 2} The widespread use of both smokeless and smoked tobacco products poses a significant global public health challenge due to their well-documented adverse health consequences—cardiovascular diseases, cancer and type 2 diabetes.³ In response to the growing global burden, tobacco harm reduction programmes have been implemented as a part of public health approach to reduce the impact of smoking-related diseases.⁴ In addition, electronic cigarettes (e-cigarettes) or electronic nicotine delivery systems (ENDS) have emerged as a popular, less harmful alternative to tobacco smoking.^{5 6}

E-cigarettes are battery-powered devices that simulate smoking by vaporising a liquid solution—nicotine, glycerol, propylene glycol–based solutions and flavourings using a heating component.^{6 7} Their combustion-free mechanism and the perception that they expose users to fewer toxicants compared with conventional tobacco products have contributed to their rapid uptake, particularly among adolescents and young adults.^{6 8} Globally, the prevalence of e-cigarette use varies across regions, and based on the global regional statistics, the current prevalence was reported as 7.1% in Africa,⁹ 6.5% in America¹⁰ and 2.4% in Europe.¹¹ The prevalence of e-cigarette usage among a few high-income countries was reported as 25.46% in France¹² and 12.52% in Australia.¹³ Since 2015, the prevalence of e-cigarette use among adolescents increased in several Asian countries, such as China, Indonesia, Malaysia, the Philippines and Thailand.¹⁴,¹⁷

Despite their perceived harm reduction potential, the growing body of evidence highlights several health risks and unintended consequences associated with e-cigarette use.^{18 19} These include cardiovascular effects,²⁰,²³ respiratory dysfunction,²⁴,²⁶ adverse neurological effects,²⁷,³⁰ the potential for nicotine addiction^{16 31} and dual use with conventional cigarettes. Moreover, the long-term health effects of e-cigarettes remain inadequately understood due to their relatively recent emergence.

In the context of the South Asian region, it is already burdened with high rates of tobacco use, weak regulatory infrastructures and significant health disparities, and the rise of e-cigarettes poses additional challenges.^{32 33} The aggressive marketing of flavoured e-liquids, increasing online accessibility and gaps in policy and enforcement contribute to the risk of widespread uptake, especially among the youth.³⁴ Furthermore, the potential normalisation of nicotine use through e-cigarettes could undermine existing tobacco control efforts, complicate cessation initiatives and introduce a new generation to nicotine dependence.³⁵ Given these concerns, a critical examination of the public health implications of e-cigarette use in South Asia (SA) is both urgent and essential for guiding future policy and intervention strategies. Therefore, this review aims to systematically synthesise empirical evidence on the prevalence and determinants of e-cigarette use in SA to uncover usage trends, identify vulnerable groups and inform targeted, evidence-based regulatory and public health strategies.

Methods

Protocol registration and reportage

This systematic review was registered in the International Prospective Register of Systematic Reviews (PROSPERO) registry (Ref: CRD42024541855) on 13 May 2024, and it was reported based on the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines³⁶ and the Cochrane Handbook for Systematic Reviews of Interventions.³⁷

Research questions

The following are the research questions which were addressed in this systematic review:

1. What is the prevalence of e-cigarette use in SA?

2. What are the determinants of e-cigarette use in SA?

Eligibility

The inclusion or exclusion of retrieved literature was determined based on predefined eligibility criteria. Specific inclusion and exclusion criteria applied in this systematic review are outlined below:

Inclusion criteria

Original research articles were included if they were of any study design—qualitative, quantitative or mixed methods published in peer-reviewed journals in English and had accessible full texts. Studies were eligible if they investigated populations in South Asian countries—Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan or Sri Lanka—and examined the prevalence and/or determinants of e-cigarette use, such as sociodemographic, behavioural or environmental factors. These criteria were established to ensure the inclusion of methodologically robust and contextually relevant studies suitable for comprehensive synthesis.

Exclusion criteria

On the other hand, studies were excluded if they were classified as grey literature (eg, edited books, monographs, technical reports, blogs, vlogs and book chapters) or non-original peer-reviewed publications (eg, commentaries, letters, reviews and editorials). Additionally, original peer-reviewed research articles were excluded if the full text was not accessible, if the study population was located outside the South Asian region or if the focus was unrelated to the prevalence or determinants of e-cigarette use. These criteria were applied to maintain the relevance, quality and consistency of the evidence synthesised in this review.

Search strategy

To ensure a robust and comprehensive literature search, two major sources were used for the search process. These sources include a database and reference list of relevant literature.

Database search

A bibliometric search of all relevant literature, published from inception till December 2023, was carried out on nine major databases, including PubMed, SCOPUS, Child Development & Adolescent Studies, CINAHL Ultimate, APA PsycInfo, Dentistry and Oral Sciences Source, AMED – The Allied and Complementary Medicine Database, SPORTDiscus with Full Text, and Psychology and Behavioural Sciences Collection databases. This search was conducted on 2 June 2024. The search strategy followed the PEO (P – Population, E – Exposure and O – Outcome) framework, where the population of interest was e-cigarette users residing in SA, the exposure of interest was e-cigarette and the outcomes of interest were the prevalence and determinants of e-cigarette use. Multiple search terms, aided by Boolean operators (“OR” and “AND”) and truncation (“*”), were used for the database search. The choice of these search terms was informed by this review’s research question and its inclusion criteria; these terms were obtained through a thorough review of existing review articles on e-cigarettes,³⁸,⁴⁰ the World Bank list of South Asian countries,⁴¹ Thesaurus, and the PubMed’s Medical Subject Heading (MeSH) Dictionary. The obtained search terms are as follows: “South Asia”, “India”, “Pakistan”, “Sri Lanka”, “Bangladesh”, “Nepal”, “Bhutan”, “Maldives”, “Afghanistan” “E-cigarette”, “Electronic cigarette”, “E-cig”, “Vape*”, “Vaping” “Electronic nicotine delivery system”, “E-Hookah”, “Electronic hookah”, “Epidemiolog*, “Determinants”, “Cause”, “Causa*”, “Frequen*”, “Inciden*”, “Prevalen*”, “Use”, “Usage”, “Barrier”, “Facilitat*” and “Influenc*”. The search strings generated from the database search are depicted in online supplemental tables 1–4.

Reference lists of relevant literature

Manual searches of the reference lists of included articles were conducted to identify any additional relevant studies that may not have been retrieved through the initial database searches. This manual screening process was carried out after completing the identification of eligible articles from the database search.

Literature selection

All the literature retrieved from the searched databases were imported into the Rayyan web application to remove duplicate entries.⁴² Following deduplication, each literature was screened by at least two blinded independent reviewers (YAJ, PP, GN or SR) in a two-stage process. In the first stage, title and abstracts of the deduplicated literature were screened at face value to exclude non-relevant literature. Thereafter, the literature that were not excluded in the first stage were subjected to second-stage screening, where full text screening was done. However, in cases where conflicts arose in the screening process, a senior reviewer (KKK or RDJ) was consulted to resolve such conflicts.

All screenings were informed by the review’s inclusion and exclusion criteria, and only those literature which met the review’s inclusion criteria were included.

Data extraction

Data extraction was done using a data extraction sheet which was developed based on the Joanna Briggs Institute’s guidelines for data extraction in systematic reviews.⁴³ The sheet was used to extract the following data from the included literature: author names, publication year, study design, study objectives, study population characteristics (including sociodemographic features and sample size), study instruments, results (findings) and conclusions.

Data synthesis

The extracted data were synthesised using a systematic narrative synthesis approach (online supplemental tables 9 and 10). The extracted data on study characteristics were categorised into publication trend by year, country location and participants’ characteristics (sample size, age, gender, socioeconomic status and urban rural setting) (online supplemental table 9). The extracted data on study outcomes (findings/results) were mainly divided into prevalence and determinants of e-cigarette use. The extracted data on the prevalence of e-cigarette use was categorised into lifetime prevalence and current prevalence. Also, the extracted data on the determinants of e-cigarette use were categorised using the Dahlgren–Whitehead Rainbow Model of classification of social determinants of health⁴⁴ (online supplemental table 10). Where applicable, all p values, ORs and CIs were presented in the synthesised findings and in accordance with existing clinical guidelines for reporting statistics in clinical research.⁴⁵

Notably, meta-analysis was intended to be conducted in this systematic review; however, based on the outcome measures and methodological heterogeneities of the study designs of the included articles, conducting a meta-analysis was not feasible.⁴⁶

Evaluation of risk of bias

All included articles were independently and jointly appraised by two reviewers using the Mixed Methods Appraisal Tool (MMAT),⁴⁷ selected due to the inclusion of studies with diverse research designs. The MMAT facilitates the evaluation of qualitative studies, quantitative randomised controlled trials, quantitative non-randomised studies, quantitative descriptive studies and mixed methods research. Each study was assessed using two generic screening questions applicable to all study types, followed by five design-specific criteria: appropriateness of the sampling strategy, adequacy of data collection methods, alignment between data sources and interpretation, quality of measurement tools and data/statistical analysis and consideration of potential confounders—resulting in a total of seven appraisal questions. Each criterion was rated using a three-point scale where “Yes” receiving a score of one (1), “Can’t tell” receiving a score of zero (0), and “No” receiving a score of zero (0), obtaining a cumulative score ranging from 0 to 7 for each study. Based on this cumulative score, the quality of each appraised article was graded as ‘above average’ if the score was ≥4 or ‘below average’ if the score was ≤3. The discrepancies were resolved through discussions among the reviewers, with a third reviewer (KKK) consulted to provide consensus and confirm the accuracy of the quality appraisal of the included studies. The detailed risk of bias assessments of each study are summarised in online supplemental tables 5–8.

Results

A total of 305 literature were retrieved from the database search (PubMed=78, SCOPUS=119, Child Development & Adolescent Studies=4, CINAHL Ultimate=52, APA PsycInfo=34, Dentistry and Oral Sciences Source=4, AMED – The Allied and Complementary Medicine Database=0, SPORTDiscus with Full Text=7, and Psychology and Behavioural Sciences Collection=7). After removal of duplicate copies (n=149), only 156 literatures remain for title/abstract screening. Twenty-five pieces of literature were subjected to full-text evaluation, after which only 15 articles were included, after meeting the inclusion criteria of this systematic review. Additionally, two other articles were manually retrieved from the reference lists of these 15 included articles, after undergoing a thorough full-text screening process. Thereby, a total of 17 articles were finally included in this systematic review (online supplemental table 4, figure 1).

Figure 1. Preferred Reporting Items for Systematic Review and Meta-Analysis flowchart diagram.

Study designs

The majority (n=13; 76.5%) of the included articles adopted quantitative non-randomised study designs,⁴⁸,⁶⁰ two (11.8%) each adopted quantitative descriptive study designs^{61 62} and qualitative study designs.^{63 64} Among those articles that adopted quantitative non-randomised study designs (n=13), only one (7.7% [1/13]) of them adopted a cohort study design,⁴⁹ while the rest (n=12/13; 92.3%) adopted a cross-sectional analytical study design.⁴⁸⁵⁰,⁶⁰

Risk of bias assessment outcomes

The quality of all the included articles (n=17) was above average, with their quality appraisal scores ranging between 4 and 7. One of these two articles adopting quantitative descriptive study scored 7⁶¹ while the other one obtained a score of 4.⁶² Among those articles, adopting quantitative non-randomised design (n=13), only one article (7.7% [1/13]) obtained a score of 4,⁵⁷ three (23.1% [3/13]) obtained a score of 5,^{48 51 56} two (15.4% [2/13]) obtained a score of 7^{50 53} and the seven remaining articles obtained a score of 6.4952 54 55 58,⁶⁰ The only two articles adopting qualitative study design received a score of 7^{63 64} (figure 2) (online supplemental tables 5–8).

Figure 2. Quality appraisal of the included articles.

Publication trend by year

All the included articles were published between the year 2017 and 2024. Among them, only six (35.3%) were published in the year 2023,⁴⁸,⁵⁰⁵⁵ only one (17.7%) each was published in the years 2017,⁶⁰ 2019⁶¹ and 2024,⁶³ three (17.6%) each were published in the years 2018^{53 54 62} and 2022,^{51 52 64} and only two (11.8%) were published in the year 2020.^{58 59} However, there was no article published in the year 2021 (online supplemental table 9).

Country location and participants’ characteristics

Geographical distribution

The included articles were reports on only four South-Asian countries, of which eight (47.1%) of them were reports on India.⁴⁸,5055 56 59 62 64 However, less than one-quarter each of these articles were reports on Pakistan (n=4; 23.5%)^{53 54 60 63} and Bangladesh (n=3; 17.6%),^{52 58 61} while only one article was a report on Nepal (5.9%; n=1).⁵⁷ Furthermore, only one article, by Pan et al,⁵¹ was a report on Bangladesh and India combined. However, none of the included article was a report on Sri Lanka, Maldives or Bhutan (online supplemental table 9).

Sample size

A total of 122 716 participants were studied in all the included articles, with the sample size in each article ranging from 24⁶⁴ to 74 037 persons.^{51 56} In most of the included articles, the age range of the participants was between 11 to 65 years of age (online supplemental table 9).

Age (adolescent vs. adult)

The two major categories of study cohorts that were investigated in the included articles were adolescents^{48 50 51 56 60 64}and adults.⁴⁸,⁵⁹⁶¹ The majority (n=11; 64.7%) of the articles were on adults⁴⁹⁵²,^{55 57} and only two articles (11.8%) were on adolescents.^{60 64} The remaining four (23.5%) articles investigated a combined cohort of adolescents and adults^{48 50 51 56} (online supplemental table 9).

Gender

Based on the overall cumulation of the gender distribution of the study participants in the included articles, sixteen articles (94.1%) investigated both genders. However, only one article, by Gravely et al,⁶¹ investigated the male population alone (online supplemental table 9).

Socioeconomic status

In the majority (n=12; 70.59%) of the included articles, the socioeconomic status of their study participants was reported.⁴⁹,⁵¹⁵⁴ However, the remaining articles (n=5; 29.41%) did not provide such information concerning their participants.^{48 52 53 59 61} Among those articles which reported the socioeconomic status of their participants (n=12), only one-third (n=4) of them reported the status as low, middle and high socio-economic groups,^{49 51 55 56} two (16.7%) reported the status as middle and high-socioeconomic groups only,^{56 57} and only one article (8.3%) reported the status as low-socioeconomic groups only⁶⁴ (online supplemental table 9).

Setting (urban vs. rural)

Additionally, data on the urban/rural setting of the study location of the participants were provided in 70.6% (n=12) of the included articles.⁴⁸⁵⁰,^{52 54 56} Among these articles (n=12), only half of them investigated population residing in urban areas alone,⁵⁰,⁵²⁵⁶ while the remaining half investigated population residing in both urban and rural areas.^{48 54 59 60 63 64} However, no article exclusively investigated populations in rural areas alone (online supplemental table 9).

Prevalence of e-cigarette use

More than three-quarters (n=14; 82.4%) of the included articles reported the prevalence rates of e-cigarette use among their study participants.⁴⁹⁵¹,^{62 64} Among them, eight articles (57.1% [8/14]) exclusively reported the current prevalence of e-cigarette use among their participants.⁵²,^{5456 58} Only six articles (42.9% [6/14]) reported on both lifetime and current prevalences of e-cigarette use among their participants^{49 51 55 57 61 62} (online supplemental table 10).

Lifetime prevalence

The prevalence rates of lifetime e-cigarette use were reported only in India,^{49 55 62} Bangladesh⁶¹ and Nepal.⁵⁷ Additionally, the article by Pan et al⁵¹ reported lifetime prevalence of e-cigarette use in both Bangladesh and India. The reported lifetime prevalence of e-cigarette use ranged between 0.1 and 23.8% among Indian populations^{49 51 55 62} and between 0.4 and 0.5% among Bangladeshi populations,^{51 61} whereas the lifetime prevalence of e-cigarette use was 21.2% among the Nepali population.⁵⁷ Only one article reported the lifetime prevalence of e-cigarette use based on age categories of their study participants,⁴⁹ and in the study, the majority of the lifetime e-cigarette users were between 25 and 30 years of age (online supplemental table 10).

Current prevalence

The prevalence of current e-cigarette use ranged between 0.01 and 5.39% among Indian populations,^{49 51 55 56 62} between 6.20 and 10.1% among Pakistani populations,^{53 54 60} whereas the prevalences were 0.5% in a Bangladeshi population^{51 61} and 5.9% in a Nepali population.⁵⁷ Based on the gender distribution, only 30.8% of the included studies (n=5/13) reported the prevalence of current e-cigarette use^{49 51 56 57 64} (online supplemental table 10).

Determinants of e-cigarette use

The determinants (predictors) of e-cigarette use were only reported in 35.3% (n=6/17) of the articles included in this systematic review.^{49 51 52 57 60 64} The reported determinants were categorised into three levels: individual lifestyle factors; social and community networks; living and working conditions and general socioeconomic, cultural and environmental conditions⁴⁴ (online supplemental table 10).

Individual lifestyle factors

Three articles reported individual lifestyle factors that were significant determinants of e-cigarette use among South Asian populations. These factors included self-perception of e-cigarette use as addictive (p=0.035; OR=0.64; [95% CI=0.43–0.97]),⁴⁹ self-perception of e-cigarette use as less harmful (p=0.030; OR=0.56; [95% CI=0.34–0.94]),⁴⁹ self-perception of e-cigarette as a safe product (p<0.001; OR and CI were unspecified), leisure use (p<0.001; OR and CI were unspecified), individual appeal to the flavours of e-cigarette (p=0.001; OR and CI were unspecified), individual stress (p<0.001; OR and CI were unspecified), presence of mental depression (p<0.001; OR and CI were unspecified)⁶⁰ and to use it as an alternative to tobacco smoking (p<0.001; OR and CI were unspecified).^{52 60} Additionally, age brackets between 20 and 26 years were found to be significant determinants of lifetime (p<0.05; COR=4.56; [95% CI=0.98–6.24)) and current (p=0.036; COR=5.07; [95% CI=0.93–8.19]) e-cigarette use.⁵⁷ Considering gender and e-cigarette use, being a man is a significant determinant of lifetime e-cigarette use among Indian population (p=0.024; OR=1.77; [95% CI=1.07–2.91]).⁴⁹ On the other hand, depending on the study population, being a woman was found to be a significant determinant of current e-cigarette use in Indian population (p<0.05; 95% CI=0.002–0.02; OR was unspecified)⁵¹ while being a man was a significant determinant of its current use among Nepali population (p=0.002; COR=6.48 [adjusted data: p=0.008; AOR=2.88])⁵⁷ (online supplemental table 10).

Social and community networks

Peer influence and social acceptance of e-cigarette use were the main contributing factors in e-cigarette use.^{49 60} Peer influence by family and friends towards e-cigarette use^{49 60} and acceptability of e-cigarette use within the family member(s) of e-cigarette users⁶⁰ were significant determinant of e-cigarette use (p<0.001; OR and CI were unspecified). Also, social appreciation of e-cigarette use was another significant determinant of e-cigarette use (p=0.002; OR and CI were unspecified)⁵² (online supplemental table 10).

Living and working conditions

Inexpensive prices of e-cigarettes compared with cigarettes were the only significant determinant of e-cigarette use (p<0.001; OR and CI were unspecified),⁵² while the ability to use e-cigarettes in smoke-free areas was not a significant determinant of e-cigarette use (p=0.904; OR and CI were unspecified).⁵² Based on the urban–rural setting, lifetime e-cigarette use was the only significant factor among urban residents in Nepal of the crude data (p<0.05; COR=5.19; [95% CI=0.83, 8.02]),⁵⁷ whereas the current e-cigarette use was a significant determinant among the rural residents in India (p<0.05; 95% CI=0.002–0.02; OR was unspecified)⁵¹ (online supplemental table 10).

General socioeconomic, cultural and environmental conditions

In all the included articles, only one of them investigated the role of general socioeconomic, cultural and environmental conditions in determining the use of e-cigarettes.⁴⁹ The only reported factor under this category was social advertising.⁴⁹ However, this factor (ie, social advertising) was not found to be a significant determinant of e-cigarette use (p=0.080; OR=1.59; 95% CI=0.95–2.66)⁴⁹ (online supplemental table 10).

Discussion

E-cigarette use has emerged as a prominent alternative to traditional smoking, particularly among the youth population, and currently, it has raised significant public health concerns.⁶⁵ Therefore, it is essential to investigate prevalence and assess determinants of e-cigarette use to strategise global health policies and regulations for a high life expectancy. This review systematically synthesised data on available evidence on both current and lifetime prevalence of e-cigarette use and categorises its determinants using the Dahlgren–Whitehead Rainbow Model, focusing on adolescent and adult populations in SA. To the best of our knowledge, this is the first systematic review in the region addressing these aspects, offering valuable insights for future research and policy development.

According to the systematically synthesised empirical evidence of this review, lifetime and current prevalence rates across SA showed substantial variability. Thereby, the lifetime e-cigarette use ranged from 0.1%⁵¹ to 23.8%⁶² in India, while the current e-cigarette use ranged from 0.01% in India⁵⁶ to 10.1% in Pakistan.⁵⁴ It is believed that this variability among lifetime e-cigarette use in the same geographic location is due to the variations in study designs, population of interest and sampling methods—such as Global Adult Tobacco Survey (GATS)⁵¹ used in each study. The wide variations in the prevalence of current e-cigarette use across different geographical locations in SA may be largely influenced by the diversity in regulatory frameworks governing these products. For instance, countries with more stringent regulations tend to report lower usage rates, while those with limited or poorly enforced policies often exhibit higher prevalence.^{66 67} This observation is supported by empirical evidence from a systematic review conducted in Southeast Asia, which similarly found that variations in national regulatory environments were closely associated with differences in e-cigarette use across populations.¹⁴ These findings highlight the significant role of policy in shaping e-cigarette use patterns within and across regions.

The findings of this review highlight that the determinants of e-cigarette use in SA are multifaceted and context-specific. In line with the published systematic review by Ling et al,¹⁴ this review proves that both lifetime and current e-cigarette use are particularly prevalent among young adults and males in countries such as Nepal and India.^{49 57} However, contradictory observations were reported in Pan et al,⁵¹ which found that current e-cigarette use was more significant among females in India, suggesting potential gender-based differences that may vary by study context or population subgroup. Furthermore, the significance of several determinants—such as personal beliefs, peer pressure, social normalisation of e-cigarette use and the cost of these products—varied among South Asian countries. These findings indicate that e-cigarette use is influenced by a complex interplay of cultural, social and economic factors that are specific to each country’s context.⁶⁸

Regulatory context of e-cigarettes in SA

The regulatory landscape for e-cigarettes in SA is highly fragmented and evolving, reflecting varying national priorities and capacities. As per reports by the WHO, only three SA countries in the region—India, Sri Lanka and Nepal have officially banned e-cigarettes.⁶⁹ India implemented a comprehensive ban on the sale, marketing and production of e-cigarettes in 2019.^{48 64} However, enforcement remains weak, and reports indicate that adolescents and young adults can still access these products through shops and social networks.⁶⁴ In contrast, both Pakistan and Bangladesh lack comprehensive regulatory frameworks for e-cigarettes.^{61 63} However, Pakistan’s tobacco control efforts focus largely on conventional cessation approaches and do not address Reduced Risk Products (RRPs) like e-cigarettes,⁶³ while there was no regulation of nicotine vaping products (NVP) in Bangladesh.⁶¹ Additionally, lack of literature on regulatory policies of e-cigarettes in Sri Lanka, Nepal, Bhutan, Maldives and Afghanistan highlights significant gaps in tobacco policy monitoring and regulation across the region.

Potential harm reduction strategies and public health implications

Although e-cigarettes were promoted as a potential harm reduction strategy to tobacco, their role in public health remains highly debatable and underexplored, particularly due to cultural relevance in SA. To mitigate the risk of the rising of non-communicable diseases (NCD) due to e-cigarette use, comprehensive harm reduction strategies and public health implications are necessary.⁷⁰ A key priority is the establishment of standardised regulations across the region, as current policies on e-cigarettes are inconsistent and fragmented.⁷¹ Measures such as enforcing age restrictions, banning flavoured products, implementing taxation and mandating health warnings can help limit youth access while supporting adult smokers in transitioning away from more harmful tobacco products.⁷²,⁷⁴

Additionally, culturally relevant public health campaigns are critical to oust the common misconception that e-cigarettes are completely safe. Strengthening surveillance systems and cessation support services will be vital to track usage trends, dual use and long-term health impacts. In countries like India, where e-cigarettes are banned, policymakers should reconsider existing regulations in response to emerging scientific evidence.⁴⁸,⁵⁰⁶⁴ Meanwhile, in countries with minimal regulation, such as Pakistan and Bangladesh, proactive and comprehensive policy development is urgently needed.^{61 63 71} Regional collaboration and evidence-based policymaking will be central to balancing the harm reduction benefits of e-cigarettes with the need to protect youth and advance broader tobacco control efforts.^{71 75 76}

Strengths and limitations

This systematic review has several notable strengths and limitations. As the first, it synthesises evidence on the prevalence and determinants of e-cigarette use in SA, and it addresses a significant gap in the regional literature. The use of a comprehensive search strategy across multiple databases, adherence to PRISMA guidelines and the inclusion of both prevalence estimates and determinants enhance the methodological rigour and relevance of the review. The findings have direct implications for public health policy and the development of targeted interventions in a region where tobacco control remains a pressing concern.

However, the review is not without limitations. The included studies exhibit considerable heterogeneity in terms of design, population groups and definitions of e-cigarette use, limiting comparability and the possibility of meta-analysis. The evidence base is relatively limited, with many studies being cross-sectional⁴⁸,⁶² and of varying quality, which constrains the ability to draw assumptions. Finally, given the rapidly evolving landscape of e-cigarette use and regulation, some findings may already be outdated, affecting the review’s timelines.

Conclusion

This systematic review highlights the emerging, yet under-researched issue of e-cigarette use in SA. The variation in reported prevalence and determinants reflects the complexity of factors influencing vaping behaviours in the region. Therefore, to understand and address these trends, future research should prioritise longitudinal designs, include countries and populations currently underrepresented—particularly adolescents—and investigate broader social, economic and policy-related factors. The findings of this review should inform policymakers in developing culturally appropriate regulations and public health interventions to curb the growing trend of vaping across SA.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.