E-cigarette use and smoking-related disparities in England: an observational study using small area estimation and nonparametric regression - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Harm Reduct J . 2026 Mar 7;23:76. doi: 10.1186/s12954-026-01435-9 Search in PMC Search in PubMed View in NLM Catalog Add to search E-cigarette use and smoking-related disparities in England: an observational study using small area estimation and nonparametric regression James E Prieger James E Prieger 1 School of Public Policy, Pepperdine University, Malibu, CA USA Find articles by James E Prieger 1, ✉ , Samuel C Hampsher-Monk Samuel C Hampsher-Monk 2 BOTEC Analysis, Santa Clarita, CA USA Find articles by Samuel C Hampsher-Monk 2 , Nima Shahidinia Nima Shahidinia 2 BOTEC Analysis, Santa Clarita, CA USA Find articles by Nima Shahidinia 2 , Eliza R W Hunt Eliza R W Hunt 2 BOTEC Analysis, Santa Clarita, CA USA Find articles by Eliza R W Hunt 2 Author information Article notes Copyright and License information 1 School of Public Policy, Pepperdine University, Malibu, CA USA 2 BOTEC Analysis, Santa Clarita, CA USA ✉ Corresponding author. Received 2024 May 15; Accepted 2026 Feb 26; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13081496 PMID: 41794738 Abstract Background The uptake of e-cigarettes has been associated with reduced smoking in the UK, and British public health policy has sought to leverage e-cigarettes’ potential to encourage cessation. Smoking harms are concentrated in disadvantaged communities in the UK. If e-cigarette use were similarly concentrated in these groups, that could help redress the inequity. On the other hand, disadvantaged communities may be slower to adopt e-cigarettes, and this could perpetuate or exacerbate smoking-related disparities. Methods This study examines the association between community-level e-cigarette prevalence and disparities in smoking between those employed in routine and manual occupations (R&M) and those employed in professional, managerial, intermediate, and other (PMI) occupations in England between 2013 and 2019. Using observational data, a small area estimation model is employed to form synthetic estimates of e-cigarette prevalence at the local level. The local socioeconomic gap in smoking (or smoking prevalence by occupation group) is then nonparametrically regressed on the estimated local e-cigarette prevalence. Results The uptake of e-cigarettes is negatively associated with smoking rates in both occupational groups, affirming the role of e-cigarettes as an opportunity for harm reduction. However, the decrease in smoking is smaller for R&M workers. Thus, the uptake of e-cigarettes may have slightly increased the disparity between the two groups. Conclusions Small area estimation provides a useful way to synthesize local estimates of e-cigarette use when direct estimation is impossible. The analysis suggests that the uptake of e-cigarettes, while associated with cessation from smoking for both types of workers, may not immediately reduce smoking-related disparities. Additional investigation should examine why the uptake of e-cigarettes may have a less marked effect on smoking cessation among R&M workers and whether these results change as the e-cigarette market matures. The analysis highlights the need to ensure that cessation interventions delivering net benefits reach the communities whose need is greatest. Supplementary Information The online version contains supplementary material available at 10.1186/s12954-026-01435-9. Keywords: E-cigarettes, Vaping, Electronic nicotine delivery systems (ENDS), Safer nicotine products (SNPs), Tobacco harm reduction, Public health disparities, Tobacco, Smoking, Tobacco control Introduction Although the prevalence of smoking in England among adults has been cut in half since 2009 (see Fig. 1 ), there are still 4.3 million British adults who smoke [ 40 ] and an average of 64,000 deaths per year were attributed to smoking in England during 2017 to 2019 (OHID, 2026). Smoking and its associated health risks disproportionately affect historically disadvantaged communities. Amid these persistent challenges, electronic cigarettes (also known as e-cigarettes, vapes, or electronic nicotine delivery systems [ENDS]) have emerged as a possible tool in smoking cessation efforts, since the uptake of electronic cigarettes is correlated with smoking cessation [ 5 , 29 , 56 ]. The potential of e-cigarettes to address disparities in smoking prevalence among different socioeconomic groups is an open question, however. Fig. 1. Open in a new tab Prevalence (%) of adult (16+) current smoking and current e-cigarette use in England, 2009–2024 Despite the great decline in British smoking, which has been attributed to robust tobacco controls and high tobacco taxes [ 7 , 28 ], smoking and its related harms remain concentrated among historically disadvantaged groups in England [ 33 , 38 , 46 ]. In 2023, more than 27% of Britons without formal educational qualifications used cigarettes, whereas just 5.8% of those with a university degree did so, and the smoking rate among unemployed Britons is about twice that of those with paid employment [ 39 ]. Among workers, those in routine and manual (R&M) occupations smoke at more than 2.5 times the rate of those in managerial and professional occupations [ 39 ]. The intersection of socioeconomic and other disadvantages may compound the effect on smoking behaviors [ 44 ]. Furthermore, higher levels of cigarette use in a local community may reinforce smoking behaviors by creating environmental, social and psychological cues. Accordingly, there are wide variations in smoking rates across the UK. Some northern towns and cities in England have smoking rates that are more than three times those of affluent localities in the south. While British policy towards cigarettes aims to reduce their use, policy toward e-cigarettes has sought to balance their apparent benefits – most importantly, facilitating cessation from smoking – and the risks they may create. While not risk-free, e-cigarettes represent a significantly less harmful alternative to combustible tobacco [ 4 , 33 , 42 , 43 , 48 , 52 , 58 ] and appear to out-perform other methods for supporting smoking cessation, both in efficacy [ 23 ] and popularity [ 47 ]. Nevertheless, concerns about the use of e-cigarettes by people who do not smoke, particularly youth, remain; such concerns were exacerbated worldwide by the so-called e-cigarette, or vaping, associated lung injuries (EVALI) crisis in the US [ 10 ]. Sales of e-cigarettes to minors are banned, nicotine concentrations are capped at 20 mg/mL, and e-cigarettes carry warning labels. Still, public agencies in the UK such as the National Health Service [ 33 ] promote e-cigarettes as safer alternatives to combustible cigarettes and encourage those who are otherwise unable or unwilling to quit smoking to switch to e-cigarettes. In 2023, the “Swap to Stop” campaign [ 57 ] announced that it would provide free e-cigarette starter kits to almost one in five adults who smoke to support smoking cessation. Unlike many other countries, the UK permits flavors in e-cigarettes to support cessation [ 19 ]. Whereas combustible tobacco products in the UK carry some of the highest tax rates in the world [ 3 ], e-cigarettes carry no excise taxes other than the standard VAT, providing a price incentive for substitutions. Although a new excise duty on vaping products is scheduled to be implemented in October 2026, concurrent increases in taxes on traditional tobacco products are expected to maintain a relative price advantage for e-cigarettes and thus preserve the substitution incentive. Disparities in smoking-related health outcomes have become a recent focus of British public health agencies [ 12 , 13 , 45 ]. In 2022, the UK launched an independent review to address such disparities as part of the government’s strategy to make England smoke-free by 2030 [ 35 ] The subsequent Kahn Review explicitly promotes the use of e-cigarettes “as an effective tool to help people to quit smoking tobacco” [ 26 ]. However, the effect of e-cigarette use on smoking-related health disparities remains uncertain. E-cigarettes have been shown to be economic substitutes for combustible tobacco [ 1 , 9 , 16 , 41 ], and use of e-cigarettes in England has risen as smoking has declined (Fig. 1 ). Plausibly, therefore, e-cigarette use will be concentrated among the same communities where smoking is—or was—most prevalent. If there were greater uptake of e-cigarettes among individuals with lower socioeconomic status (SES) who smoke compared to those in higher SES groups, and if the effect of e-cigarette use on smoking cessation were similar among groups, then the cessation-facilitating effects may be concentrated in the communities where smoking harms are greatest, reducing smoking-related disparities. However, the diffusion of innovation theory [ 50 ] suggests that disadvantaged communities may be slower than wealthier communities to adopt innovative products [ 51 ]. That could concentrate any smoking-cessation benefits among more affluent groups, perpetuating or even exacerbating smoking disparities, at least for a time [ 17 , 36 ]. To date, the literature contains mixed reports on how e-cigarette use varies across socioeconomic groups. In England, lower socioeconomic position was associated with higher odds of e-cigarette use between 2014 and 2019 [ 27 ]. The opposite holds among those who smoke: e-cigarette use in Great Britain among those who smoke has been associated with higher socioeconomic status [ 6 , 27 ]. Recently, Green et al., [ 18 ] found more nuanced results: socioeconomic disadvantage was associated with vaping among some groups in the UK (youth who have never smoked and adults who formerly did so) but not others (adults who never smoked and adults who currently do so). Regarding cessation, socioeconomic disadvantage is associated with reduced odds of having quit among adults, but this association is moderately weaker among those who vaped (OR: 0.88; CI: 0.82–0.95) than among those who do not (OR: 0.82; CI: 0.80–0.84; p-value for difference = 0.081; [ 18 ]). Studies from other countries also suggest that e-cigarette use is less likely to lead to cessation among lower SES groups [ 53 ], which may exacerbate socioeconomic disparities in smoking-related harms [ 22 , 32 ], even if the overall burden of disease from tobacco use is reduced [ 15 ]. Given the imperative to redress smoking-related health disparities, and the fact that more adults now use e-cigarettes than smoke in England (Fig. 1 ), it is vital to gain a better understanding of how the uptake of e-cigarettes interacts with smoking-related disparities. This is particularly important considering the significant role that e-cigarettes appear to have had in driving smoking cessation at the national level. Methodology The aim of this study is to assess the relationship between the local prevalence of e-cigarette use and the socioeconomic gap in smoking. For the latter, we follow other observational studies that use employment type as a proxy for SES [ 24 ]. 1 Data on the socioeconomic gap in smoking, disaggregated to the local authority/district level, were publicly available from Public Health England (PHE). 2 The smoking gap is between employment classifications: workers in routine and manual occupations (R&M) and those in other roles, mainly professional, managerial, and intermediate occupations (PMI). 3 However, there are no publicly available estimates for the prevalence of e-cigarette use at the local geographic level, and so those must be estimated. The empirical exploration therefore comprises three steps, which are described below and in Table 1 . Full details on the data and methods are in the appendix. Table 1. Summary of the steps in the analysis Estimation step Goal Unit of observation Data 1 Estimate a model to predict the probability of smoking based on observed characteristics An individual in England, ages 16–64 years Health Survey England, 2013–2019 2 Predict the prevalence of current use of e-cigarettes A local authority district (LAD) Mainly from the Office of National Statistics, Annual Population Survey. See the appendix for sources. 3 Assess the relationship between use of e-cigarettes and the socioeconomic gap in smoking A local authority district (LAD) Outcome variable: Public Health England. Use of e-cigarettes: the results from Step 1 Open in a new tab Step 1: modeling an individual’s decision to use e-cigarettes Steps 1 and 2 are for a small area estimation (SAE) to predict local prevalence of e-cigarette of use. SAE has been used in the public health literature to estimate smoking prevalence [ 14 , 21 , 31 ] but apparently not yet for estimating the prevalence of e-cigarette use. In the first step, the parameters of a person-level model predicting the probability of using e-cigarettes based on demographic and socioeconomic data are estimated using microdata from Health Survey England (HSE), an annual household survey. The dependent variable is a binary indicator, ecig , for whether the individual states that they currently use e-cigarettes. The estimation in step 1 is for individuals aged 16 to 64 years. Although the socioeconomic gap variable used in step 3 is defined for workers between the ages of 18 and 64, it is not possible in the HSE data to exclude 16- and 17-year-olds. A linear probability model, estimated via weighted least squares regression using the survey weights, is used to predict usage of e-cigarettes as a function of age, gender, ethnicity (Asian, Black, mixed, White, and other ethnicity), household income, employment and occupation, the region of residence in England, the quintile of the index of multiple deprivation ( qIMD ) of the neighborhood of residence (1 = most deprived, 5 = least deprived), 4 and the survey year. Smoking status is not used to predict e-cigarette usage to avoid simultaneity, since the predictions from this model are used in step 3 as the regressor in regressions of smoking-related variables. The choice of variables in the survey data that can be used as regressors in the linear probability model is also constrained by the need for data on the average values of the same variables to be available from census data for local areas (for the prediction in step 2). A linear probability model is used, instead of a model for limited dependent variables such as logit regression, because the next step requires the predictions to be linear in the predictors. Step 2: predict local area prevalence of current e-cigarette use The publicly available survey data do not identify the local area of residence of the respondent, and even if they did, there would not be enough people for each area and year to estimate the local prevalence of e-cigarette use accurately. Instead, the fitted model from step 1 is used to create a synthetic estimate of the prevalence of e-cigarette use from a prediction based on the demographic and socioeconomic characteristics of the locality. This procedure is an example of SAE, which allows us to “borrow strength” from observations on e-cigarette usage from outside the focal small area to increase the effective sample size [ 49 ]. In the parlance of SAE, our estimator is domain- and time-indirect, since the prediction for e-cigarette use in a specific local area and year borrows strength from observations from other geographical areas and time periods. The appendix shows that the results from the linear probability model from step 1 can be used to estimate the prevalence of current e-cigarette use in English Local Authority Districts (LADs) by plugging LAD-level average values of the regressors into the model fitted to individuals. Accordingly, a panel dataset was gathered on the LAD average values of the same variables used as regressors in step 1. Data on age, gender, ethnicity, quintile of the multiple deprivation index, and average annual household income were obtained from official sources (see the appendix). Step 3: regression of the socioeconomic gap in smoking (or smoking rates) on prevalence of e-cigarette use The final step is to perform regression analyses of the socioeconomic gap between the smoking rates of the R&M and PMI groups to determine the relationship between e-cigarette prevalence and disparities in smoking. The first dependent variable in the analysis is the socioeconomic gap in a local authority district among adults who currently smoke (ages 18 to 64), based on responses to the annual population survey conducted by ONS and made available by PHE. This variable is defined by PHE as an odds ratio, comparing the smoking prevalence among adults who work in R&M occupations—a proxy for lower socioeconomic status 5 —and those in PMI occupations. 6 Given the skewness in the distribution of the smoking gap, the log odds ratio is taken for the dependent variable in the regression. Log odds ratios greater than zero indicate that the odds of adult smoking are higher in the area for R&M workers than PMI workers. The independent variable in the regression is the synthetic estimate of the prevalence of e-cigarette use at the local (i.e., LAD) level from step 2. The regression function and its derivative (the marginal effect of e-cigarette prevalence on the socioeconomic gap) are computed by nonparametric local-linear regression, a technique chosen to minimize the need for parametric assumptions about the nature of the relationship between the two variables [ 30 ]. 7 Pointwise bootstrap percentile confidence intervals [ 8 ] are computed from 799 replications. Results Direct estimates of e-cigarette use National estimates produced from the HSE data of the current use of e-cigarettes, broken out by socioeconomic group, are in Fig. 2 . The third category, labeled neither , is reserved for people without occupations (i.e., those who have never worked or the long-term unemployed) or for those who did not answer the relevant survey question. The survey weights are employed so that each prevalence shown is an estimate for the population aged 16–64; the p -value from the chi-squared statistic for differences in prevalence among groups is 0.002 or smaller for each year. As with smoking, the use of e-cigarettes is much more common among R&M workers than other groups (see Fig. 2 ). Fig. 2. Open in a new tab Estimated population prevalence of current e-cigarette use by employment group Model for predicted e-cigarette use The first-step estimates for the SAE used to predict current usage of e-cigarettes by an individual are shown in Table 2 . The results show that individuals aged 25–34 and 45–54 are the most likely to use e-cigarettes, followed closely by those aged 35–44. Men are more likely to use e-cigarettes than women. E-cigarette usage rises with household income and falls with the deprivation of the area. The latter result may stem from environmental effects, by which individuals in less deprived areas are more likely to use e-cigarettes because they see others in their neighborhood doing so. 8 Individuals of white ethnicity are most likely to use e-cigarettes, which has also been shown for the smoking population in the US [ 22 ]. After controlling for the other factors, the propensity to vape is lowest in London and highest in the East Midlands region (an area that includes Northampton, Leicester, Nottingham, and Derby). The coefficients for the years roughly match the time trend for e-cigarette usage shown in Fig. 1 . Table 2. First-step linear probability model estimation for the SAE: Predicting current e-cigarette usage Y = current e-cig use Coefficient S.E. Age 16 to 24 -0.023 *** 0.006 Age 35 to 44 -0.005 0.004 Age 45 to 54 -0.002 0.005 Age 55 to 64 -0.024 *** 0.005 Male 0.019 *** 0.003 Household income (100,000s) 0.023 *** 0.003 IMD quintile = 2 0.000 0.004 IMD quintile = 3 0.012 ** 0.005 IMD quintile = 4 0.015 *** 0.005 IMD quintile = 5 (least deprived) 0.020 *** 0.005 Black ethnicity -0.054 *** 0.007 Asian ethnicity -0.052 *** 0.004 Mixed ethnicity -0.033 *** 0.008 Other ethnicity -0.048 *** 0.010 Not employed 0.037 *** 0.005 Occupation: Managers/senior officials 0.015 *** 0.006 Occupation: Assoc. prof. & tech./admin. 0.013 *** 0.004 Occupation: Skilled trades 0.015 ** 0.006 Occupation: sales/service/manual/elementary 0.038 *** 0.004 Region: North East -0.003 0.007 Region: North West 0.010 0.007 Region: East Midlands 0.014 * 0.008 Region: West Midlands -0.001 0.007 Region: East of England 0.001 0.007 Region: London -0.016 ** 0.006 Region: South East -0.003 0.006 Region: South West -0.011 0.007 Year: 2014 0.015 *** 0.004 Year: 2015 0.023 *** 0.005 Year: 2016 0.044 *** 0.006 Year: 2017 0.042 *** 0.005 Year: 2018 0.050 *** 0.005 Year: 2019 0.045 *** 0.005 Constant 0.018 *** 0.008 R-squared 0.022 N (obs.) 34,145 F-test statistic 16.97 Prob > F 0.000 Open in a new tab *** p < 0.01, ** p < 0.05, * p < 0.1 Note: Estimation is pooled OLS on the repeated cross-sectional survey data. Standard errors account for the complex survey design of the HSE (stratification, clustering, and weights). The excluded region is Yorkshire and The Humber . IMD is the index of multiple deprivation Small area estimates of e-cigarette use The synthetic estimates from step 2 are shown at the regional level in Fig. 3 , which also shows direct estimates of the prevalence of e-cigarette use. The former estimates are the weighted average predicted probability of use for individuals in the region, using the model estimated in step 1, while the latter are the mean regional estimated prevalences computed directly from the HSE. The figure shows that the two regional estimates are generally close. The prevalence of current e-cigarette use grew rapidly from 2013 to 2016, after which it largely appears to have stabilized. However, the prevalence of use differs greatly among regions: in the latest years, about 10% of adults in the East Midlands used e-cigarettes, while only about 5% did so in London. The complete list of small area estimates is contained in the appendix. 9 Fig. 3. Open in a new tab Direct and synthetic estimates of the prevalence of current e-cigarette use, by region Regression results for the socioeconomic gap The socioeconomic gap in smoking varies widely by area, per the data from Public Health England. While the mean odds ratio across local areas is 2.35 (or 2.5 when weighted by population), the s.d. of the statistic is 1.2, the minimum value is 0.179 (Melton in 2016), and the maximum value is 9.824 (St. Albans in 2014). All but 3.9% of the estimated gaps exceed unity, and none of the remaining estimates differ significantly from 1.0. 10 Thus, the estimated odds of smoking are higher among R&M workers than among PMI workers in nearly all areas, and the hypothesis of uniformly higher odds across areas and years cannot be rejected. A nonparametric regression of the socioeconomic gap in smoking on local authority-level synthetic estimates of e-cigarette use prevalence is shown in Fig. 4 . Note that, as is typical with nonparametric regression, there is only a single regressor. The predicted prevalence of current e-cigarette usage is a linear combination of the first-step covariates (see the appendix), and thus the influence of all these variables on the smoking gap is expressed through the e-cigarette regressor. Fig. 4. Open in a new tab Socioeconomic gap in prevalence of smoking, workers age 18–64, as a function of current use of e-cigarettes The scatter plot in Fig. 4 shows the data color coded by year. Since the use of e-cigarettes has increased over time in England, more recent observations tend to be toward the right side of the graph, although demographic differences among areas imply that the predicted prevalence of e-cigarette use is not simply ordered by year. The nonparametric regression function is shown in Fig. 4 . 11 The graph shows that there is a mildly positive relationship between e-cigarette usage and the socioeconomic gap, but only when e-cigarette usage is low. In the range where most of the predicted e-cigarette prevalences lie, there is no apparent relationship with the smoking gap. Figure 5 shows the estimated marginal effects of e-cigarette use on the socioeconomic gap, which is the derivative of the regression function plotted in Fig. 4 . The marginal effect and its confidence band allow us to see where there is a significant association between the two variables by noting where the band does not span zero. The marginal effect is significantly positive, indicating that an increase in the prevalence of e-cigarette use is positively associated with increases in the socioeconomic gap, for intermediate levels of e-cigarette use, but only up to a prevalence around 6.3%. There is no significant association above that level, where the top 60% of local areas (ranked by e-cigarette prevalence of use) lie. Fig. 5. Open in a new tab Marginal effect of e-cigarette prevalence on the socioeconomic gap in prevalence of adult smoking Given these equivocal results, the smoking gap is next broken down by the numerator and denominator in the odds ratio to understand why the gap appears to be insensitive to e-cigarette use over the largest range of the data. Figure 6 shows results from a nonparametric regression of the log odds of smoking for both occupational groups. The green curve in the graph for the R&M group shows that, at least in the region where most of the data lie, there is a mild negative relationship between the odds of smoking and e-cigarette use. Figure 7 shows that when the prevalence of current e-cigarette use is above 4%, which covers 90% of the distribution, the negative marginal effect (in green) is statistically significant for R&M workers. Thus, at all but the lowest prevalence levels, e-cigarettes are associated with reduced odds of smoking for R&M workers. Fig. 6. Open in a new tab Log odds of smoking as a function of current adult use of e-cigarettes, by worker group Fig. 7. Open in a new tab Marginal effect of e-cigarette prevalence on the log odds of adult smoking, by worker group The exercise is repeated for PMI workers, with results also shown in Figs. 6 and 7 . 12 The regression curve for these workers (in red) shows starker results: except for the extremes of the distribution of e-cigarette prevalence, there is a clear negative relationship between e-cigarette use and the odds of smoking. The marginal effect, graphed in red in Fig. 7 , shows that for PMI workers, the estimated derivative of the log odds of smoking with respect to e-cigarette use in the local area is negative everywhere. The estimated marginal effects are statistically significantly negative for all but the 0.05% lowest predicted e-cigarette prevalence. These results show that the lack of a relationship found between the socioeconomic gap and e-cigarette use masks the phenomenon that use of e-cigarettes is associated with lower odds of smoking for both occupational groups (at least when e-cigarette prevalence is high enough, in the case of R&M workers). With the numerator and the denominator of the odds ratio (i.e., the socioeconomic gap) both falling as e-cigarette prevalence rises, the socioeconomic gap itself at first rises and then remains largely unchanged as use of e-cigarettes increases. Notwithstanding, e-cigarette use is associated with less smoking for both groups (apart from the lowest decile of prevalence for R&M workers). Further exploration The associations found between the odds of smoking and the predicted prevalence of e-cigarette use may not be causal. Recall that the nonparametric estimations control for the demographic and socioeconomic characteristics of the local communities, the region effects, and the year effects only through their impact on the single regressor, predicted e-cigarette prevalence. One may wonder whether variation in these predictors that is orthogonal to the e-cigarette prediction is confounding the regressions. Since the e-cigarette prediction is a linear combination of the predictors, they cannot all be added to the final-step estimations due to multicollinearity. Instead, to verify that the association found in Fig. 6 for R&M workers is not merely due to, say, a confounding time trend, we (parametrically) regress their log odds of smoking on a quadratic function of predicted e-cigarette prevalence, year and region fixed effects, and a random effect for the LAD. Modeling the confounding factors specific to the communities as a random effect is a compromise between not accounting for any demographic or socioeconomic confounders on the one hand and either including all of them (which is not possible) or using LAD-level fixed effects (which leaves too little variation in the predicted e-cigarette prevalence to identify its impact on smoking). The regression results, in Table 3 , show that the log odds of smoking for R&M workers are still significantly negatively associated with the local e-cigarette prevalence after controlling for community, region, and year confounding factors. The implied marginal effects are shown in Fig. 8 ; the marginal effects are negative everywhere and significant when e-cigarette prevalence is greater than approximately 0.041, which covers the upper 87% of the range. Thus, the results confirm those from Fig. 7 . Table 3. Local-area regression of blue-collar smoking odds on predicted e-cigarette prevalence Y = log odds of smoking (blue collar) Coefficient Standard Error E-cigarette prevalence -0.537 † 2.261 Squared e-cigarette prevalence -43.064 *† 13.794 Region fixed effects Included Year fixed effects Included Number of observations 2,157 Number of clusters 315 R-squared (between) 0.092 R-squared (within) 0.080 R-squared (overall) 0.085 Chi-squared test statistic 186.670 Prob > Chi-square 0.000 Open in a new tab * p < 0.01 † Coefficients are jointly significant at the 1% level Estimation is random effects regression. The unit of observation is an LAD. Standard errors are robust to heteroskedasticity and clustering on the LAD Fig. 8. Open in a new tab Marginal effect of e-cigarette prevalence on the log odds of adult smoking, R&M workers (from quadratic regression) Discussion The negative relationship found between the odds of smoking and e-cigarette use for both worker groups supports the claim that e-cigarettes are contributing to declining smoking rates in the UK [ 2 ]. The relatively stronger decline in smoking among the PMI group implies that the uptake of e-cigarettes is associated with a mild widening of the gap in smoking rates between the R&M and PMI groups. The increase in the gap is most pronounced when e-cigarette use is low. This may be because uptake of e-cigarettes (and subsequent smoking cessation) is initially concentrated among groups with higher SES, in accord with the diffusion of innovation theory popularized by Rogers [ 50 ], the conjectures of some public health researchers [ 17 , 36 ], and the evidence of the history of the impact of life-saving innovations [ 11 ]. 13 To the extent that completely substituting combustible tobacco with e-cigarettes reduces risk exposure and leads to improved health outcomes [ 4 ], the present evidence suggests that those in R&M roles are not benefitting to the same degree as those in PMI occupations, at least for the study period. The results do not explain why the association between e-cigarette use and reduced demand for cigarettes is weaker among the R&M group. There are several possibilities. Even after controlling for access to cessation supports, people in lower SES groups who smoke are less likely to be successful in quitting smoking relative to others [ 25 ]. That perpetuates the smoking-related disparity between higher and lower SES communities. Second, group-specific psychosocial and environmental factors may reinforce smoking among some groups [ 54 ]. Should such factors persist after an individual begins to use e-cigarettes, this circumstance could impede full substitutions and increase the likelihood of dual or poly-product use, particularly within low socioeconomic status (SES) communities where these factors predominate. Dual use may be a transitory phase between smoking and cessation rather than an end-point; nevertheless, while it persists, dual or poly-product use denies the consumer the health benefits of complete substitution. Third, misinformation may lead some people who smoke to miss out on the benefits of e-cigarette-assisted smoking cessation. Those who currently smoke, particularly those “from disadvantaged groups, incorrectly and increasingly believe that vaping is as harmful as smoking” [ 34 ]. This belief – almost certainly false given current evidence – is an important impediment to switching to e-cigarettes and thus perpetuates smoking-related morbidity and mortality. Finally, if the uptake of e-cigarettes by people who do not smoke were concentrated among disadvantaged groups, that could explain why the association between e-cigarette use and smoking prevalence is weaker in the R&M group. That possibility emphasizes the importance of understanding the motivations around e-cigarette use and how these differ across communities. For example, Hiscock et al., [ 24 ] found that higher-SES clients of Stop Smoking Service (SSS) in England were more likely to use e-cigarettes as a cessation aid, whereas lower SES clients were more likely to use e-cigarettes without quitting smoking. 14 These possibilities deserve further exploration. However, if – as the evidence suggests – the uptake of e-cigarettes is associated with reductions in smoking in both groups, then the differential effect should not be interpreted as evidence that e-cigarettes are an undesirable intervention. Rather, the implication is that policy messaging and cessation services should be tailored to specific communities so that benefits attendant to substituting combustible tobacco with safer alternatives might be shared among the disadvantaged as well as other communities. Limitations As with all SAE methodologies, the present study has several limitations. First, the synthetic estimates assume that “demography is destiny”. Given the lack of any local identifiers in the HSE data, methods of SAE that include random effects to allow for unobserved heterogeneity in local e-cigarette use cannot be employed. Second, it is not clear from the present analysis whether the associations found are specific to the immature e-cigarette market or if they will evolve as consumption patterns change as the market matures. Further analysis may help shed light on these possibilities. Third, the current-use variable fails to account for the possibility that some people have transitioned from smoking to e-cigarettes and thence to abstaining from all nicotine. Such individuals are not included in the estimate of e-cigarette prevalence, but their cessation from smoking may affect the socioeconomic gap. Additional studies could utilize data on former use of e-cigarettes as well as current use. Fourth, the results found in the study are observational and associational. Quasi-experimental methods designed to elicit causal effects could be employed, a task for future research that awaits the availability of different and better data. Fifth, defining the socioeconomic gap in smoking based on occupational class may be less applicable outside the UK, where class and social status have historically been closely tied to occupation, and where official statistics are routinely produced using such classifications [ 20 ]. In less developed countries, for example, one could follow Sen [ 59 ] and focus on relative deprivation and capability shortfalls rather than occupation. One might group individuals by access to key resources or opportunities—such as educational attainment, housing quality, or employment stability—that reflect the capacity to lead healthy lives. These groupings would then proxy for socioeconomic status, capturing both material disadvantage and the broader social constraints that influence smoking behavior. Finally, recall that a linear probability model is employed in step 1, instead of a nonlinear estimator tailored to binary dependent variables such as logit, because the synthetic estimator in the next step requires linearity. However, an analogous logit regression produces estimates that are highly similar to the linear model in terms of the signs, significance, and magnitudes of the effects of the regressors on usage of e-cigarettes (i.e., the OLS coefficients vs. the logit average marginal effects; see the appendix). Thus, the main difference between the estimations would be the logit model’s restriction of predicted values to the unit interval and the distance between the predicted values. 15 Indeed, there are a few estimated prevalences for local areas that are below zero (but see footnote 9). However, there are few differences in the ranks of the predicted values between the two models; the Spearman correlation coefficient is 0.988, indicating nearly perfect correlation of the ranks. The estimations in step 3, in which these predicted values are the regressor, are nonparametric. Thus, transforming the regressor with an essentially rank-preserving function is likely to have only minor impact on the estimated substantive relationship between e-cigarette use and local-area smoking rates. In particular, a rank-preserving transformation of the true local e-cigarette prevalences, if they are indeed so distorted by the linear model, and if there is any impact on the results, would mainly affect the level of the nonparametric marginal effects shown in Fig. 7 , not their sign or significance. Conclusion This is, to the best of our knowledge, the first use of small area estimation for the local prevalence of use of e-cigarettes, much less to investigate the effect of e-cigarette uptake on smoking-related disparities. While results based on observational data should be interpreted with caution, the findings nonetheless underscore some important observations. First, the uptake of e-cigarettes is associated with reductions in smoking in both study groups (except for R&M workers in the case of very low levels of e-cigarette prevalence). Second, the uptake of e-cigarettes may have had a less significant impact on smoking among those in the R&M group compared with those in the PMI group. As such, the smoking-related disparity between the two groups appears to have been perpetuated and perhaps even exacerbated during the first decade of the e-cigarette market in the UK. Further studies will be needed to assess whether the associations found are indeed causal and to explore the mechanisms for this effect. The suggestive evidence highlights the need for effective risk-communication in all groups, and for cessation interventions, including those involving e-cigarettes, to be tailored to the specific communities and localities in which smoking harms are concentrated. This could facilitate the acceleration in smoking cessation afforded by transitions to e-cigarettes to benefit the communities most in need. Supplementary Information Supplementary Material 1. (873.9KB, pdf) Acknowledgements Not applicable. Abbreviations ENDS Electronic nicotine delivery systems HSE Health survey England IMD Index of multiple deprivation LA Local authority LAD Local authority district LSOA Lower-level super output area MHRA Medicines and health-care products regulatory agency NHS National health service (of England) NRT Nicotine replacement therapy OHID Office for health improvement and disparities ONS Office of national statistics PHE Public health England PMI Professional, managerial, and intermediate (workers) QIMD Quintile of index of multiple deprivation R&M Routine and manual (workers) SAE Small area estimation SES Socio-economic status SSS Stop smoking service VAT Value added tax Author contributions J.P. designed the empirical approach, procured data, analyzed and interpreted the data and estimations, and was a major contributor in writing and editing the manuscript. S.H.-M. conceived the scope of this manuscript and was a major contributor in writing and editing the manuscript. N.S. procured data and assisted with the analysis and interpretation of the data and estimations. E.H. contributed to the writing and editing of the manuscript. Funding This report was funded with a grant from the Foundation for a Smoke-Free World, a US nonprofit 501(c)(3) private foundation with a mission to end smoking in this generation, to BOTEC Analysis. The Foundation accepted charitable gifts from PMI Global Services Inc. (PMI); under the Foundation’s Bylaws and Pledge Agreement with PMI, the Foundation was independent from PMI and the tobacco industry. The contents, selection, and presentation of facts, as well as any opinions expressed herein are the sole responsibility of the authors and under no circumstances shall be regarded as reflecting the positions of the Foundation for a Smoke-Free World, Inc. Data availability The data pertaining to the first step of the small area estimation (the Health Surveys for England) that support the findings of this study are available from the UK Data Archive but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. The datasets generated or analyzed during the current study pertaining to the subsequent local-area analysis are available in the OSF repository, [ https://osf.io/w6ch5/overview ]. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests E.H. and N.S.: Nothing to declare apart from the funding statement.J.P. and S.H.-M.: We have previously received compensation from a third-party for research and consulting services performed for a tobacco company on matters of illicit trade. Footnotes 1 The authors acknowledge that employment type is an imperfect proxy with which to investigate socio-economic disparities related to smoking. Our choice is prompted by the Office of National Statistics’ decision to use the type of occupation to construct the National Statistics Socio-Economic Classification (NS-SEC) in England and ONS’s subsequent decision to use the NS-SEC to define two groups of workers (R&M and PMI, in the language of this article) to compute a variable the ONS calls the ‘socio-economic gap in current smokers’ ( https://tinyurl.com/ONS-NSSEC ). The authors recognize that categorizing R&M workers as ‘lower class’ may create a stigmatizing label and that the stigmatization of people who smoke confers additional harms on them and reinforces smoking behaviors [ 55 ]. However, the ONS’s socio-economic gap variable is the only one that allows investigation of disparities in at the local level. 2 PHE has since been dissolved and this indicator has been discontinued. PHE’s health protection and improvement functions were transferred to the UK Health Security Agency (UKHSA) and the Office for Health Improvement and Disparities (OHID) [ 37 ]. 3 The PMI category also includes small business owners and the self-employed. See footnote 8. 4 The qIMD variable in the HSE data are for the LSOA (lower level super output area), which is the smallest unit of census geography in England. LSOAs comprise about 400 to 1,200 households. 5 R&M occupations refer to those in major groups 5 (Lower supervisory and technical), 6 (Semi-routine occupations) and 7 (Routine occupations) of the National Statistics Socio-economic classification (NS-SEC) schema, 8-group version. 6 The PMI occupations include those in NS-SEC groups 1 (Higher managerial, administrative and professional occupations, 2 (Lower managerial, administrative and professional occupations), 3 (Intermediate occupations, i.e., those involving clerical, sales, or service roles), and 4 (Small employers and own-account workers). 7 Estimates are computed with the npregress kernel command in Stata 17.0 with the Epanechnikov kernel. The bandwidths for the estimation of the conditional means and derivatives are chosen by cross-validation. 8 A similar regression for current cigarette use shows that smoking is also more likely in less deprived areas, other things equal. Thus, the negative relationship between e-cigarette use and area deprivation may merely reflect that there are more people who smoke than would be predicted by the other regressors, and hence more dual users, in the least deprived areas. 9 Nothing in the linear predictor constrains the predicted prevalences to be non-negative, and a few are below zero. However, fewer than 1% of the prevalence estimates are negative, and 13 of 15 are within 0.01 of zero. 10 The estimates of the socioeconomic gap in smoking from Public Health England for local areas are provided with 95% confidence intervals. 11 The 95% confidence band includes error from the mean smoothing only, not the first-step estimation error in the synthetic estimates of prevalence. 12 One change is made to the estimation procedure, since the cross-validated bandwidth did not converge. Instead of cross-validation, the bandwidths for the mean and derivative estimations were chosen to equal those from the R&M worker estimations. 13 Based on historical investigation, Deaton ([ 11 ], p.25) contends that “most innovations do appear to initially favor the better-off…”. 14 A reviewer suggested another possible explanation: if lower-income people who smoke purchase less expensive cigarettes, then the price incentive to switching to e-cigarettes may be attenuated, causing fewer R&M workers to switch. 15 For example, logit predictions near the lower end are necessarily compressed toward zero, while the lowest linear predictions, having no lower bound, may be more widely spread. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Abouk R, Courtemanche C, Dave D, Feng B, Friedman AS, Maclean JC, Pesko MF, Sabia JJ, Safford S. Intended and unintended effects of e-cigarette taxes on youth tobacco use. J Health Econ. 2023;87:102720. 10.1016/j.jhealeco.2022.102720. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Action on Smoking and Health (ASH) (UK). Use of e-cigarettes (vapes) among young people in Great Britain. 2023. https://ash.org.uk/uploads/Use-of-vapes-among-young-people-GB-2023.pdf?v=1686042690 3. Asen E. Cigarette Taxes in Europe. Tax Foundation. 2019. https://taxfoundation.org/cigarette-tax-europe-2019 4. Balfour DJ, Benowitz NL, Colby SM, Hatsukami DK, Lando HA, Leischow SJ, West R. Balancing consideration of the risks and benefits of e-cigarettes. Am J Public Health. 2021;111(9):1661–72. 10.2105/AJPH.2021.306416. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Beard E, West R, Michie S, Brown J. Association of prevalence of electronic cigarette use with smoking cessation and cigarette consumption in England: a time-series analysis between 2006 and 2017. Addiction. 2020;115(5):961–74. 10.1111/add.14851. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Brown J, West R, Beard E, Michie S, Shahab L, McNeill A. Prevalence and characteristics of e-cigarette users in Great Britain: findings from a general population survey of smokers. Addict Behav. 2014;39(6):1120–5. 10.1016/j.addbeh.2014.03.009. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Cancer Research UK (CRUK). British smokers down by 1.9 million since the ban. 2017 https://news.cancerresearchuk.org/2017/07/01/british-smokers-down-by-19million-since-the-ban 8. Cattaneo MD, Jansson M. Kernel-based semiparametric estimators: Small bandwidth asymptotics and bootstrap consistency. Econometrica. 2018;86:955–95. 10.3982/ECTA12701. [ Google Scholar ] 9. Cotti C, Courtemanche C, Maclean JC, Nesson E, Pesko MF, Tefft NW. The effects of e-cigarette taxes on e-cigarette prices and tobacco product sales: evidence from retail panel data. J Health Econ. 2022;86:102676. 10.1016/j.jhealeco.2022.102676. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Dave D, Dench D, Kenkel D, Mathios A, Wang H. News that takes your breath away: risk perceptions during an outbreak of vaping-related lung injuries. J Risk Uncertain. 2020;60:281–307. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Deaton A. Policy implications of the gradient of health and wealth. Health Aff. 2002;21(2):13–30. [ DOI ] [ PubMed ] [ Google Scholar ] 12. Department of Health (DH). Towards a Smokefree Generation – A Tobacco Control Plan for England. Department of Health, Global and Public Health/Population Health/Health/Cost Centre. 2017. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/630217/Towards_a_Smoke_free_Generation_-_A_Tobacco_Control_Plan_for_England_2017-2022__2_.pdf 13. Department of Health and Social Care (DHSC). Government launches landmark reviews to tackle health disparities. 2022. https://www.gov.uk/government/news/government-launches-landmark-reviews-to-tackle-health-disparities 14. Eberth JM, McLain AC, Hong Y, Sercy E, Diedhiou A, Kilpatrick DJ. Estimating county-level tobacco use and exposure in South Carolina: A spatial model-based small area estimation approach. Ann Epidemiol. 2018;28(7):481–8. 10.1016/j.annepidem.2018.03.015. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Friedman AS, Horn SJ. Socioeconomic disparities in electronic cigarette use and transitions from smoking. Nicotine Tob Res. 2019;21(10):1363–70. 10.1093/ntr/nty120. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Friedman AS, Pesko MF. Young adult responses to taxes on cigarettes and electronic nicotine delivery systems. Addiction. 2022;117(12):3121–8. 10.1111/add.16002. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Giovenco DP. Different smokes for different folks? E-cigarettes and tobacco disparities. Am J Public Health. 2019;109(9):1162. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Green MJ, Gray L, Sweeting H, Benzeval M. Socioeconomic patterning of vaping by smoking status among UK adults and youth. BMC Public Health. 2020;20(1). 10.1186/s12889-020-8270-3. [ DOI ] [ PMC free article ] [ PubMed ] 19. Doward J, Fraser T . UK attacked for defence of flavoured e-cigarettes. The Guardian. 2019. https://tinyurl.com/37n6af4r 20. Goldthorpe JH, Hope K. The social grading of occupations: A new approach and scale. Clarendon; 1974. 21. Ha NS, Lahiri P, Parsons V. Methods and results for small area estimation using smoking data from the 2008 National Health Interview Survey. Stat Med. 2014;33(22):3932–45.10.1002/sim.6219 [ DOI ] [ PubMed ] [ Google Scholar ] 22. Harlow AF, Stokes A, Brooks DR. Socioeconomic and racial/ethnic differences in e-cigarette uptake among cigarette smokers: Longitudinal analysis of the population assessment of tobacco and health (PATH) study. Nicotine Tob Res. 2019;21(10):1385–93. 10.1093/ntr/nty141. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Lindson N, Butler AR, McRobbie H, Bullen C, Hajek P, Begh R, Hartmann-Boyce J. Electronic cigarettes for smoking cessation. Cochrane Database Syst Reviews. 2024. 10.1002/14651858.CD010216.pub8., 1, Article CD0102161. [ Google Scholar ] 24. Hiscock R, Arnott D, Dockrell M, Ross L, McEwen A. Stop Smoking practitioners’ understanding of e-cigarettes’ use and efficacy with particular reference to vapers’ socioeconomic status. J Smok Cessat. 2019;14(1):21–31. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Hiscock R, Dobbie F, Bauld L. Smoking cessation and socioeconomic status: An update of existing evidence from a national evaluation of English stop smoking services. BioMed Research International, 2015, Article 274056. 10.1155/2015/274056 [ DOI ] [ PMC free article ] [ PubMed ] 26. Kahn J. The Khan review: Making smoking obsolete. Office for Health Improvement and Disparities. 2022. https://www.gov.uk/government/publications/the-khan-review-making-smoking-obsolete 27. Kock L, Brown J, Shahab L. Association of socioeconomic position with e-cigarette use among individuals who quit smoking in England, 2014 to 2019. JAMA Netw open. 2020;3(6):e204207–204207. 10.1001/jamanetworkopen.2020.4207. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Kuipers MA, Beard E, West R, Brown J. Associations between tobacco control mass media campaign expenditure and smoking prevalence and quitting in England: A time series analysis. Tob Control. 2018;27(4):455–62. 10.1136/tobaccocontrol-2017-053662. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Levy DT, Sánchez-Romero LM, Li Y, Yuan Z, Travis N, Jarvis MJ, McNeill A. England SimSmoke: The impact of nicotine vaping on smoking prevalence and smoking‐attributable deaths in England. Addiction. 2021;116(5):1196–211. 10.1111/add.15269. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Li Q, Racine J. Cross-validated local linear nonparametric regression. Statistica Sinica. 2004;14(2):485–512. [ Google Scholar ] 31. Li W, Land T, Zhang Z, Keithly L, Kelsey JL. Small-area estimation and prioritizing communities for tobacco control efforts in Massachusetts. Am J Public Health. 2009;99(3):470–9. 10.2105/AJPH.2007.130112. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Lucherini M, Hill S, Smith K. Potential for non-combustible nicotine products to reduce socioeconomic inequalities in smoking: a systematic review and synthesis of best available evidence. BMC Public Health. 2019;19(1469):1–12. 10.1186/s12889-019-7836-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. McNeill A, Brose LS, Calder R, Bauld L, Robson D. Evidence review of e-cigarettes and heated tobacco products 2018: A report commissioned by Public Health England. Public Health England. 2018. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/684963/Evidence_review_of_e-cigarettes_and_heated_tobacco_products_2018.pdf 34. McNeill A, Brose LS, Calder R, Bauld L, Robson D. Vaping in England: An evidence update including mental health and pregnancy, March 2020: A report commissioned by Public Health England. Public Health England. 2020. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/869401/Vaping_in_England_evidence_update_March_2020.pdf 35. Medicines and Healthcare Products Regulatory Agency (MHRA). E-cigarettes: regulations for consumer products. 2016. https://www.gov.uk/guidance/e-cigarettes-regulations-for-consumer-products 36. Nayak P, Pechacek TF, Weaver SR, Eriksen MP. Electronic nicotine delivery system dual use and intention to quit smoking: Will the socioeconomic gap in smoking get greater? Addict Behav. 2016;61:112–6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Office for Health Improvement and Disparities. Smoking profile: Data. Public Health England Fingertips. 2026. https://tinyurl.com/48yetk62 38. Office for National Statistics (ONS). Likelihood of smoking four times higher in England’s most deprived areas than least deprived. 2018. https://tinyurl.com/2k5xjx6a 39. Office for National Statistics (ONS). Adult smoking habits in the UK: 2024. https://tinyurl.com/yr4kzjx7 40. Office for National Statistics (ONS). Adult smoking habits in England, 2023 edition. 2025a. https://tinyurl.com/5bmyh296 41. Pesko MF, Courtemanche CJ, Maclean JC. The effects of traditional cigarette and e-cigarette tax rates on adult tobacco product use. J Risk Uncertain. 2020;60(3):229–58. 10.1007/s11166-020-09330-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Polosa R, Morjaria JB, Prosperini U, Busà B, Pennisi A, Malerba M, Caponnetto P. COPD smokers who switched to e-cigarettes: Health outcomes at 5-year follow up. Therapeutic Adv Chronic Disease. 2020;11:2040622320961617. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Polosa R, Morjaria J, Caponnetto P, Caruso M, Strano S, Battaglia E, Russo C. Effect of smoking abstinence and reduction in asthmatic smokers switching to electronic cigarettes: Evidence for harm reversal. Int J Environ Res Public Health. 2014;11(5):4965–77. 10.3390/ijerph110504965. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Potter LN, Lam CY, Cinciripini PM, Wetter DW. Intersectionality and smoking cessation: Exploring various approaches for understanding health inequities. Nicotine Tob Res. 2021;23(1):115–23. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Public Health England (PHE). PHE Strategy 2020-25. 2019 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/831562/PHE_Strategy_2020-25.pdf 46. Public Health England (PHE). Health matters: Smoking and mental health. 2020. https://www.gov.uk/government/publications/health-matters-smoking-and-mental-health/health-matters-smoking-and-mental-health 47. Public Health England (PHE). Vaping better than nicotine replacement therapy for stopping smoking, evidence suggests. 2021 https://www.gov.uk/government/news/vaping-better-than-nicotine-replacement-therapy-for-stopping-smoking-evidence-suggests 48. Pulvers K, Nollen NL, Rice M, Schmid CH, Qu K, Benowitz NL, Ahluwalia JS. Effect of pod e-cigarettes vs cigarettes on carcinogen exposure among African American and Latinx smokers: A randomized clinical trial. JAMA Netw Open. 2020;3(11):e2026324–2026324. 10.1001/jamanetworkopen.2020.26324. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Rao JNK, Molina I. Small area estimation (Second). Wiley; 2015. 50. Rogers E. Diffusion of Innovations, 5th Edition. Simon and Schuster. 2003 51. Sage D, Vitry C, Dainty A. Exploring the organizational proliferation of new technologies: An affective actor-network theory. Organ Stud. 2020;41(3):345–63. 10.1177/0170840618815524. [ Google Scholar ] 52. Shahab L, Goniewicz ML, Blount BC, Brown J, McNeill A, Alwis KU, West R. Nicotine, carcinogen, and toxin exposure in long-term e-cigarette and nicotine replacement therapy users: a cross-sectional study. Ann Intern Med. 2017;166(6):390–400. 10.7326/m16-1107. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Sharma A, Kasza KA, O’Connor RJ, Felicione NJ. Personal characteristics associated with switching from cigarettes to noncombustible tobacco and nicotine products among US adults: Findings From PATH study waves 1–5. Nicotine Tob Res. 2025;27(1):80–90. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 54. Thirlway F. Explaining the social gradient in smoking and cessation: the peril and promise of social mobility. Sociol Health Illn. 2020;42(3):565–78. 10.1111/1467-9566.13039. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 55. Volkow ND. Stigma and the toll of addiction. N Engl J Med. 2020;382(14):1289–90. 10.1056/NEJMp1917360. [ DOI ] [ PubMed ] [ Google Scholar ] 56. West R, Shahab L, Brown J. Estimating the population impact of e-cigarettes on smoking cessation in England. Addiction. 2016;111(6):1118–9. 10.1111/add.13343. [ DOI ] [ PubMed ] [ Google Scholar ] 57. Wise J. Vaping: Government announces swap to stop scheme to cut smoking rates. BMJ. 2023;381:815. 10.1136/bmj.p815. [ DOI ] [ PubMed ] [ Google Scholar ] 58. Yayan J, Franke KJ, Biancosino C, Rasche K. Comparative systematic review on the safety of e-cigarettes and conventional cigarettes. Food Chem Toxicol. 2024;185:114507. [ DOI ] [ PubMed ] [ Google Scholar ] 59. Sen A. Liberty, unanimity and rights. Economica. 1976;43(171):217-245. 10.2307/2553122 Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1. (873.9KB, pdf) Data Availability Statement The data pertaining to the first step of the small area estimation (the Health Surveys for England) that support the findings of this study are available from the UK Data Archive but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. The datasets generated or analyzed during the current study pertaining to the subsequent local-area analysis are available in the OSF repository, [ https://osf.io/w6ch5/overview ]. Articles from Harm Reduction Journal are provided here courtesy of BMC ACTIONS View on publisher site PDF (3.3 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top