Impact of smoke-free legislation on stroke risk: A systematic review and meta-analysis

Abstract

Purpose:

Secondhand smoke significantly increases the risk of cerebrovascular diseases, prompting recent public smoking bans. We aimed to ascertain the effects of smoke-free legislation on stroke incidence and mortality.

Methods:

We systematically searched Medline, Embase, Cochrane Library, and Scopus up to August 13, 2023, for studies reporting changes in stroke incidence following partial or comprehensive smoking bans. A random-effects meta-analysis was conducted on hospital admissions and mortality for stroke, stratified based on comprehensiveness of the ban ((i) workplaces-only, (ii) workplaces and restaurants, (iii) workplaces, restaurants and bars). The effect of post-ban follow-up duration was assessed visually by a forest plot, while meta-regression was employed to evaluate for any dose-response relationship between ban comprehensiveness and stroke risk.

Findings:

Of 3987 records identified, 15 studies analysing bans across a median follow-up time of 24 months (range: 3–67) were included. WRB bans were associated with reductions in the rates of hospital admissions for stroke (nine studies; RR, 0.918; 95% CI, 0.872–0.967) and stroke mortality (three studies; RR, 0.987; 95% CI, 0.952–1.022), although the latter did not reach statistical significance. There was no significant difference in the risk of stroke admissions for studies with increased ban comprehensiveness and no minimum duration for significant post-ban effects to be observed.

Discussion and conclusion:

Legislative smoking bans were associated with significant reductions in stroke-related hospital admissions, providing evidence for its utility as a public health intervention.

Graphical abstract

Keywords

Smoke-free policy secondhand smoke stroke acute cerebrovascular disease

Introduction

Stroke is the second leading cause of death globally.¹ According to the World Stroke Organisation, the incidence of stroke exceeds 12.2 million patients annually and over 143 million years of healthy life are lost every year due to stroke-related deaths and disability.² Smoking is a strong independent risk factor of stroke. Studies performed across various populations have demonstrated that current smokers have at least a two- to fourfold increased risk of stroke compared with lifelong nonsmokers or ex-smokers who had quit smoking more than 10 years prior.³ Moreover, secondhand smoke (SHS) exposure has been associated with adverse health effects. SHS, also known as environmental tobacco smoke or passive smoke, consists of smoke that is emitted during burning of tobacco-containing products and directly exhaled by the smoker.⁴ SHS exposure has been reported to increase the risk and mortality of stroke by up to 30% and 100% respectively.^5,6

Numerous countries have implemented smoke-free legislations in public areas to mitigate the deleterious effects of SHS exposure. However, smoke-free legislations in different regions vary in their degree of comprehensiveness. Comprehensive smoking bans have been implemented in workplaces, bars, restaurants, public transport, schools and hospitals, whilst partial smoking bans still allow for smoking in certain establishments. This has direct implications on the efficacy of legislative policies on the general population.⁷ Additionally, there is a large variation in the scope and level of enforcement of such smoking bans among different cities and countries. Despite this, certain EU countries such as Switzerland,^8,9 Ireland,¹⁰ Scotland¹¹ and Sweden¹² have reported a clear reduction in SHS. There is also a notable positive effect on other health outcomes such as reduction in incidence of acute coronary events and respiratory diseases.¹³

While the protective effects of smoke-free legislation on the incidence and mortality rates of cardiovascular and respiratory conditions are well defined,^14–16 its effects on stroke are less clear. Several studies have reported a decrease in hospital admissions for stroke after the implementation of a smoking ban^17–20 while others showed no significant change.^21–24 One study¹⁸ showed that Florida’s 2003 statewide law, which banned smoking in all workplaces and restaurants, was associated with a 18.1% (95% CI, 9.3–30.0) reduction in stroke hospitalisations over the period of 3 years. Conversely, a similar study in Canada²² examined a 2003 province-wide smoking ban, encompassing public places and workplaces, and found a non-significant change of −3.04 (95% CI, −13.14, 7.06) admissions per 100,000 person-months in mean monthly stroke hospitalisation rates over 67 months.

In this systematic review and meta-analysis, we aimed to elucidate the effects of public smoking bans on the incidence and mortality rates of stroke, stratified by comprehensiveness of bans, length of follow-up and methodological quality.

Methods

This systematic review and meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 Statement²⁵ and took reference from the Cochrane Handbook for Systematic Review of Interventions (version 6.3).²⁶ The study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO, Record ID: CRD42024524342).

Search strategy

A thorough literature search was performed in Medline, Embase, Scopus and Central Cochrane databases from inception to 13th August 2023. A search was conducted for studies that reported associations of population level smoke free control policies in relation to stroke risk. Key search terms like ‘smoke-free policy’, ‘legislation bans’ and ‘stroke’ were included in our search strategy developed alongside the expertise of two medical information specialists (Medical Library, National University of Singapore and Medical Library, Nanyang Technological University). Reference lists of primary studies and systematic reviews were also reviewed for additional eligible studies. The detailed search strategy is available in the Supplemental Materials (eMethods).

Study selection

The research question was formulated using the PICOS (population, intervention, comparator, outcome, study type)²⁷ framework. The intervention group would comprise a general population situated within a geographical region in which a smoke-free ban was implemented, either partial or complete. Complete bans were defined as comprehensive smoking bans encompassing workplaces, restaurants and bars while partial bans referred to restrictions that do not fulfil the criteria for comprehensive bans. This group would be compared against control regions with no legislative smoking bans and control conditions without smoking as a risk factor. The primary outcomes were chosen to be any stroke-related population health outcomes including disease incidence, mortality rates and hospital admissions and discharges due to stroke. Study designs included observational studies (cross-sectional studies and cohort studies) and quasi-experimental studies (interrupted time series, uncontrolled before-and-after and controlled before-and-after).

Studies that met the following criteria for smoking bans were eligible: (1) the smoking ban must have been legislated at a local, regional or national level, (2) smoking was banned either completely in all settings including the hospitality sector or in specific areas, (3) smoking ban applied to both smokers and non-smokers, (4) outcomes reported were exclusively related to stroke or stroke with transient ischaemic attack (TIA) and were not combined with other unrelated medical diagnoses into a composite outcome. The following types of studies were excluded: (1) Systematic reviews; (2) Commentaries; (3) Case reports; (4) not in English; (5) studies that only investigated indirect measures to prohibit smoking, such as increasing the tobacco tax, advertising on the harmful effects of smoking, etc. (i.e. no form of comprehensive or partial ban established); (6) studies for which the full text of the article could not be retrieved for full evaluation; (7) studies that did not provide sufficient data about population size, the number of stroke mortality/incidence/deaths; (8) studies that did not report or present data that permitted estimating relative risk and confidence intervals; (9) studies on regions which had existing smoking bans in the pre-ban phase; (10) studies which reported only on transient ischaemic attack.

For each study, two authors independently reviewed the titles and abstracts. After excluding studies that did not meet the inclusion criteria, the full texts of the remaining studies were then independently screened by the same two authors to further assess eligibility. Discrepancies at each stage were resolved via discussion with all authors.

Data extraction

The following data was extracted from each study: proportion of population smoking, comprehensiveness of smoking ban, duration of study pre- and post-smoking ban, region that ban was implemented in, age, gender, population size, mean stroke admission rates, mean stroke discharge rates, mean stroke mortality rates, relative risk (RR) and confidence intervals (CI) comparing pre-ban versus post-ban stroke incidence, stroke incidence in control regions as well as change in incidence of non-smoking related control conditions. Two reviewers independently extracted data for each study and cross-checking was done by a third reviewer. The data extraction process was blinded, and discrepancies were resolved via consensus among all authors.

Study quality assessment

A comprehensive evaluation of bias in population-level observational studies was conducted using the Risk of Bias In Non-randomised Studies - of Exposures (ROBINS-E)²⁸ tool. The tool assessed the risk of bias in exposure effect estimates from individual observational studies that examine the effect of implementation of smoking bans on stroke incidence.^29–31 Each included study was assessed in seven key domains– confounding factors, selection of participants, classification of exposures, departure from intended exposure, missing data, measurement of outcomes and selection of the reported result– with the risk of bias for each domain expressed as low, some concerns, or high. Two reviewers independently assessed the risk of bias for each study and all conflicts were resolved as detailed above.

Statistical analysis

Data analysis was conducted using R version 4.3.2 (The R foundation)³² with the meta package version 7.0-0.³³ The Hartung Knapp Sidik-Jonkman estimator³⁴ served as our between-study variance estimator, and random-effects models were employed to account for anticipated inter-study heterogeneity. A p-value of 0.05 was chosen as the cutoff for statistical significance in this study.

Statistical heterogeneity was assessed using Cochran’s Q test and the I² statistic. Cochran’s Q test was utilised to detect significant heterogeneity of intervention, with I² > 50% indicating significant heterogeneity.

The RR was reported in only three studies.^20,35,36 For these three studies, the gradual effects model was chosen to better represent the change over the entire follow-up period. The remaining 10 studies did not report RR and CI directly, thus various methods had to be employed to estimate them from the reported outcome. In five studies,^19,37–40 the reported Incident Rate Ratio (IRR) and their respective CI served as an approximation of the RR. In two studies that applied a Poisson regression model,^23,41 the results were reported as a percentage change which reflected the difference between the RR and the null value of 1.0. As such, we could convert the percentage changes back to RR by adding them to the null value (1.0). Likewise, the CI was obtained by adding the upper and lower limit to the null value. In one study,²² RR and CI was estimated using the exponential of the change in trend of monthly admission rate obtained from the ARIMA model used in the study. In two studies,^17,42 RR and CI was determined using a Poisson regression model accounting for temporal trends. For one study,⁴² the effects of each phase of the smoking ban were isolated, and the final phase (workplaces, restaurants and bars bans) was directly compared with the pre-ban period. In all cases, the risk change associated with the law was always directly compared with the pre-ban period.

Included studies were categorised based on the outcomes measured and the comprehensiveness of the ban, with separate analyses conducted for every category with at least three or more studies. Meta-analysis was conducted on each study category to determine the significance of relative risk reduction.

Random-effects meta-regression utilising dummy variables coding for comprehensiveness of the law was performed to assess for its association with reductions in hospital admissions.

Visual analysis of the main forest plot with studies arranged in order of post-legislation follow-up length was used to assess for (i) any significant association in magnitude of risk reduction in stroke with length of follow-up and (ii) any minimum duration of follow-up for a significant post-ban effect to be observed. If a significant association was found, further regression analysis based on duration of follow-up would be performed.

Subgroup analysis was performed on studies which incorporated control regions (no smoking bans) and non-tobacco related conditions (e.g. cholecystitis, intestinal obstruction) in their study methodology as they adjusted for potential confounders such as secular trends and minimised internal validity threats to their study design.^43,44 Further subgroup analysis was also performed on studies based on their risk of bias assessment using the ROBINS-E tool.

Finally, the risk of publication bias was assessed using funnel plots, Egger’s and Begg’s tests on our primary outcome of change in hospital admissions due to stroke.

Results

Study selection

A total of 3987 articles were retrieved from four databases, including 665 articles from PubMed, 2940 articles from Embase, 299 articles from Cochrane Library, and 83 articles from Scopus. 702 duplicates were removed, with a further 3258 articles excluded following the title and abstract sieve. Full text evaluation of the remaining 27 articles yielded 15 texts that met the inclusion criteria, of which 13 studies were eligible for meta-analysis. The number of references included and excluded at each screening stage are shown in a PRISMA flowchart (Figure 1).

Figure 1.

PRISMA flow diagram for review of effect of smoke-free legislation on stroke risks.

One paper¹⁸ examined the smoke-free legislation in three US states (Florida, New York and Oregon) and only Florida had no county-level comprehensive smoke-free air laws in place before the statewide law was enacted. However, the paper did not provide sufficient data to evaluate the effect estimate. Another paper⁴⁵ did not provide the incidence of stroke events before and after the smoke-free ban was implemented necessary to calculate the RR. These two papers were hence not included in the meta-analysis.

Baseline characteristics

The characteristics of included studies (published between 2010 and 2022) are shown in Table 1 and eTable 1. Six studies^{23,35,37,38,40,45} were based in Europe, five^{17–19,22,42} in North America and four studies^20,36,39,41 in East Asia. The scale of legislation differed between studies, with six studies^{35,37–40,45} analysing country-wide legislation, five^{20,23,36,41,42} city-wide, three^17–19 state-wide and one²² at the provincial level. Smoking legislation included in these studies did not encompass e-cigarette and vaping products. In terms of the statistical models used, nine studies^{18,20,22,23,35,36,40,41,45} were interrupted time series, four studies^17,19,39,42 were controlled before-and-after and two studies^37,38 were uncontrolled before-and-after.

Table 1.

Summary of study design and characteristics of included studies.

Study	Study design	Location	Population size	Proportion of population smoking	Age of patients (years old)	Gender of patients	Comprehensiveness of smoking ban	Pre-ban study duration	Post-ban study duration
Naiman 2010	Controlled before-and-after	Canada (Toronto)	2,503,281	18% (in Toronto)	⩾45	All (unspecified)	Phase 1 (October 1999) – workplaces-only	Jan 1996–Oct 1999 (45 months)	Post phase 1: Oct 1999–June 2001 (21 months)
				26.6% (109,000 in Thunder Bay – control city)			Phase 2 (June 2001) – WR^c		Post phase 2: June 2001–June 2004 (36 months)
				24.8% (561,000 in Durham – control city)			Phase 3 (June 2004) – WRB^d		Post-phase 3: June 2004–March 2006 (21 months)
Herman 2011	Controlled before-and-after	USA (Arizona)	NA^a	NA^a	NA^a	All (unspecified)	WRB^d	1 Jan 2004–1 May 2007 (40 months)	1 May 2007–31 May 2008 (12 months)
Kent 2012	Before-and-after	Ireland	NA^a	NA^a	20–69	All (unspecified)	WRB^d	2002–2003 (24 months)	2005–2006 (24 months)
Loomis 2012	Interrupted time series	USA (Florida, New York and Oregon)	NA^a	NA^a	>35	All (unspecified)	Florida: WR^c	Florida: Jan 1990–July 2003 (163 months)	July 2003–Dec 2006 (41 months)
Loomis 2012	Interrupted time series	USA (Florida, New York and Oregon)	NA^a	NA^a	>35	All (unspecified)	New York and Oregon: excluded as there were existing state/county-level laws	Florida: Jan 1990–July 2003 (163 months)	July 2003–Dec 2006 (41 months)
Head 2012	Controlled before-and-after	USA (Beaumont, Texas)	Total: 107 876 (2005), 115 053 (2008)	NA^a	All ages	All (unspecified)	WRB^d	July 2004–June 2006 (24 months)	July 2006–June 2008 (24 months)
			Non-Hispanic blacks: 53 829 (2005), 58 217 (2008)
			Non-Hispanic whites: 36 288 (2005), 42 180(2008)
Gaudreau 2013	Interrupted time series	Canada (Prince Edward Island)	143,000	NA^a	⩾35	M & F^b	WRB^d	1 April 1995–31 May 2003 (98 months)	1 June 2003–31 December 2008 (67 months)
Stallings-Smith 2013	Interrupted time series	Ireland	1,900,000	NA^a	⩾35	M & F^b	WRB^d	Jan 2000–Mar 2004 (51 months)	Apr 2004–Dec 2007 (45 months)
Mackay 2013	Interrupted time series	Scotland	5,100,000	NA^a	All ages	M & F^b	WRB^d	1 Jan 2000–25 Mar 2006 (75 months)	26 Mar 2006–Dec 2010 (57 months)
Humair 2014	Before-and-after	Switzerland (Geneva)	450,000	NA^a	⩾16	M & F^b	WRB^d	Pre-ban: 30 June 2006–30 June 2008 (24 months)	31 October 2009–31 December 2010 (14 months)
								Phase 1 ban: 1 July 2008 − 30 September 2008 (3 months)
								Ban suspension period: 1 October 2008–30 October 2009 (13 months)
Galan 2015	Interrupted time series	Spain (Madrid, Barcelona)	Madrid − 3,124,090	NA^a	⩾18	All (unspecified)	WRB^d	Jan 2003–Dec 2005 (36 months)	Jan 2006–Dec 2006 (12 months)
			Barcelona − 1,591,529
			Total − 4,715,619
Saijo 2019	Controlled before-and-after	Japan (Bibai City, Hokkaido)	23,013	NA^a	NA	All (unspecified)	Workplaces-only	July 2014–July 2016 (24 months)	July 2016–June 2018 (24 months)
Xiao 2020	Interrupted time series	China (Tianjin)	15,600,000	300,000,000	⩾35	M & F^b	WRB^d	January 2007–31 May 2012 (65 months)	1 June 2012–December 2015 (43 months)
Xiao 2020	Interrupted time series	China (Qingdao)	4,700,000	NA^a	⩾35	M & F^b	WRB^d	January 2007–31 May 2012 (65 months)	1 September 2013–31 December 2015 (28 months)
Chu 2022	Interrupted time series	China (Xi’an)	NA^a	NA^a	⩾15	M & F^b	WRB^d	9 February 2017–31 October 2018 (21 months)	1 November 2018–25 December 2019 (14 months)
Jarlstrup 2023	Interrupted time series	Denmark	NA^a	NA^a	⩾30	All (unspecified)	WRB^d	1 January 2002–31 August 2007 (68 months)	1 June 2012–December 2015 (43 months)

Not applicable.

Males and females.

Workplaces and restaurants.

Workplaces, restaurants and bar.

Among the 15 studies included in the analysis, 12 studies^{17,19,20,22,23,35–38,40,41,45} specifically investigated the impact of WRB bans on stroke risk. There was a single study¹⁸ in the categories of WR and workplaces-only bans each. The remaining study⁴² analysed three separate phases on a city-wide ban. Seven out of fifteen studies^{17,19,22,35,36,39,42} had accounted for non-tobacco related conditions and/or control regions.

Hospital admissions were the most commonly reported,^{17,18,22,23,36–39,41,42} followed by mortality rate,^20,35,36 disease incidence,^40,45 hospital discharge¹⁹ and emergency transportations.³⁹ The median length of follow-up was 24 months (range: 3–67).

Three papers^23,38,41 were evaluated to be of low risk, eight papers^{17,18,20,35,36,40,42,45} had some concerns and four papers^19,22,37,39 were deemed to be of high risk (eTable 2) based on the ROBINS-E tool.²⁸

Meta analysis

Several studies examined more than a single region,²³ outcome³⁶ or stroke subtype⁴⁰; as such, 13 eligible studies yielded 16 effect estimates which were included in the final meta-analysis.

Nine studies assessed the association of comprehensive smoking bans involving WRB with the rate of hospital admission due to stroke. There was a significant reduction in the risk, with a RR of 0.918 (95% CI, 0.872–0.967) (Figure 2).

Figure 2.

Meta-analysis of smoke-free legislation in workplaces, restaurants and bars (WRB) on stroke hospital admissions.

We observed a non-significant pooled reduction in stroke mortality with comprehensive smoking bans involving WRB in three studies (RR, 0.987; 95% CI, 0.952–1.022) (Figure 3).

Figure 3.

Meta-analysis of smoke-free legislation in workplaces, restaurants and bars (WRB) on stroke mortality.

Sensitivity analysis of studies accounting for non-smoking related conditions and/or control regions with no smoking bans (n = 4) revealed a similar effect estimate with reduced inter-study heterogeneity (RR, 0.906; 0.806–1.019; I² = 55%) (Figure 4).

Figure 4.

Meta-analysis of smoke-free legislation in workplaces, restaurants and bars (WRB) on stroke hospital admissions, stratified by presence or absence of control region and/or condition.

The subgroup meta-analysis by high-moderate (n = 6) versus low (n = 3) risk of bias showed no significant difference in pooled risk ratio between the two groups (p = 0.25) (eFigure 1).

In order to ascertain whether studies with longer follow-up duration produced larger effect sizes and to identify any minimum duration for a significant effect to be observed, we produced a forest plot with the studies arranged in ascending order of follow-up duration. There was no visual correlation between follow-up duration and effect sizes, hence we did not proceed with further analysis to interrogate this (eFigure 2).

The effect of more comprehensive bans on hospital admissions for stroke was determined using meta-regression with dummy coded variables representing the three main categories (W, WR, WRB) (eTable 3). Regression coefficients were found to be statistically insignificant.

The funnel plot appears asymmetrical, suggesting possible publication bias for studies examining hospital admissions due to stroke and comprehensive smoke-free bans in WRB (eFigure 3). However, Egger’s and Begg’s tests for regression intercept did not indicate a statistically significant publication bias (p = 0.516 and 0.835 respectively).

Discussion

This is, to our knowledge, the largest systematic review and meta-analysis focussed on ascertaining the effects of public smoke-free legislation on stroke risk specifically. After examining 15 studies of smoke-free legislations around the world, the main finding was that the implementation of comprehensive smoking bans involving WRB were associated with a significant reduction in stroke hospital admissions. We did not find evidence that this conclusion was affected by systemic bias in the included studies, follow-up duration or presence of controls. There were insufficient data points to draw conclusions about workplaces-only and WR bans.

While the effectiveness of public smoke-free legislations in reducing the risk of acute myocardial infarction¹⁵ is well studied, the evidence regarding stroke is less conclusive. Findings from different regions vary, with some studies reporting reductions in both stroke-related hospital admissions^17–19 and stroke mortality rates²⁰ following smoking bans, while others^21–24 show no such effect. Factors including differences in the way the bans were designed, implemented, enforced and evaluated may explain these inconsistent findings. Discrepancies in findings may be influenced by the varying prevalence of strokes across different regions. For instance, Loomis and Juster reported that areas with a greater burden of stroke present greater potential for significant reduction in stroke-related hospital admissions.¹⁸ Our meta-analysis, which pooled data from 13 eligible studies conducted across 15 diverse geographical areas, revealed a substantial and statistically significant decrease of 7.4% on average in the risk of stroke admissions. Our findings provide assurance that policy interventions effectively mitigate secondhand smoke exposure at a population level, consequently reducing the incidence of strokes. This makes sense in the context that a body of epidemiological research has identified secondhand smoke exposure as an important risk factor for stroke.⁴⁶

The inclusion of control regions and/or non-tobacco related conditions as non-equivalent dependent variables, which are not expected to respond to the intervention, but exposed to the same historical validity threats,⁴⁷ enhances the ability of studies to effectively control for confounding factors.⁴⁴ This approach taken by some studies enabled a more precise delineation of the impact of legislative policies on stroke risk. The subgroup analysis of these higher quality studies revealed a RR of 0.906 which is materially similar to that of our main analysis of 0.918, demonstrating that our effect estimate from the main analysis is robust. Moreover, the subgroup analysis of these studies yielded a lower interstudy heterogeneity, suggesting that the impact of smoking bans may have been underestimated in certain studies without adequate controls. Any future studies on smoke-free legislation health outcomes should include control conditions and control regions for greater internal validity to further substantiate their findings.

Although we did not find evidence amongst the included studies to show that a longer follow-up period post-smoking ban leads to greater effects on stroke risk, current evidence supports a temporal association between the introduction of national smoke-free bans and subsequent reductions in smoking-related morbidity and mortality, specifically cardiovascular risk.¹³ Bans do require time for enforcement and compliance for their effects to show.⁴⁸ Similarly, effects may take time to accumulate; this explains the lack of significant change in stroke mortality seen amongst the pooled studies in which the median follow up length was only 2 years. Longitudinal studies examining more data points over longer follow-up periods are warranted to confirm this association and ascertain the true impact on stroke mortality.

Whilst originally intended, data limitations posed a significant challenge in uncovering a dose-response relationship to elucidate a possible link between comprehensiveness of smoke-free legislation and magnitude of stroke risk reduction. Regression analysis did not reveal any demonstrable relationship between the comprehensiveness of the smoking ban (W, WR and WRB) with the risk of hospital admission for stroke in the present study, but this is presumably due to the aforementioned data limitations (only one dataset each for workplaces-only and WR). However, we note that there is evidence from prior studies that the risk of SHS is dependent on the comprehensiveness of the smoke-free law implemented.⁴⁶ Future studies on upcoming laws implementing partial bans can help to augment this analysis for greater statistical power.

While smoke-free legislations are thought to reduce stroke risk through reduced SHS exposure for non-smokers, it is possible that another mechanism is through the reduction in the prevalence of smoking itself. While not specifically studied in our study, the contributing role of this possible mechanism is supported by multiple studies from Anglo-America,⁴⁹ Europe⁵⁰ and Asia⁵¹ provide strong evidence that such legislation can lead to a modest decline in smoking rates. Notably, comprehensive bans have a greater effect on smoking cessation and prevalence compared to partial bans. However, this effect is often delayed due to the challenges of overcoming addiction and behavioural change. A study in Switzerland found it took a year for comprehensive bans to significantly reduce smoking rates, highlighting the importance of early implementation for maximum impact.⁵²

Various mechanisms through which smoking can increase the risk of stroke have been proposed. Smoking can lead to carotid atherosclerosis,^53,54 endothelial dysfunction^55–57 and inflammation,⁵⁸ and expose smokers to oxidative damage from highly reactive oxygen species generated by cigarette combustion.⁵⁹ A meta-analysis also demonstrated that smoking was associated with an increased risk of atrial fibrillation in a dose-dependent manner.⁶⁰ Studies have also shown that secondhand smoke can impact non-smokers to a significant and deleterious extent.^61,62 Therefore, smoke-free bans are likely to confer benefits to the whole population.

Successful implementation of smoke-free legislation is multifactorial and is in part dependent on stringent enforcement of the ban by local law enforcers and public compliance of the ban.^63,64 Mackay et al.⁴⁵ reported an initial benefit in ischaemic stroke reduction which partially reverted back to pre-legislative levels, a trend attributed to waning enforcement over time. In this regard, studies have demonstrated that the compliance of smoking bans can be improved through anti-tobacco educational campaigns to enhance smokers’ knowledge on health risks,⁶⁵ and signages and penalties for violations.⁶⁶ General attitudes towards smoking culture and tobacco industry play a huge role in the success of smoking bans, especially in regions where tobacco holds historical and economical importance.^67–69 Furthermore, secondhand smoke exposure frequently occurs within homes, which falls outside public smoke-free legislation.^70,71 This may have limited the effect of public smoking bans on reduction in stroke risk.

Recently, there has also been a dramatic increase in the use of alternative tobacco products and electronic nicotine delivery systems (ENDS) worldwide, such as vaping and e-cigarettes, which produce deleterious effects and are not necessarily covered within public smoking bans.^72–74 More studies investigating the health impact of ENDS are needed in the future.

There has also been a growing trend of increasing ischaemic stroke incidence among younger individuals,⁷⁵ in which smoking prevalence likely plays a contributory role.⁷⁶ Existing studies have reported that the dose-response relationship between cigarette smoking and ischaemic stroke risk in both young men⁷⁷ and women⁷⁸ is the strongest compared to other age groups. Future studies should focus on the effect of smoking bans on the prevalence of ischaemic stroke rates in young people.

The present study demonstrated a significant decrease in hospital admissions for stroke with implementation of smoke-free laws and this potentially translates to lower healthcare expenditure and more efficient utilisation of healthcare resources. A similar effect has been observed in studies examining AMI-related hospitalisations and the cost savings of the smoking bans.^79,80 This economic effect can be considered by policy makers when developing smoking-related legislation.

Limitations

Our study has several limitations. Firstly, differentiation of the impact of smoking bans on the risks of ischaemic and hemorrhagic stroke was not possible due to the lack of separation within most included studies regarding these two stroke types. Ischaemic and hemorrhagic strokes have different pathophysiologies and there may be value to disentangle them as separate outcomes. Likewise, some studies did not separate between transient ischaemic attack and stroke, and this may have diluted the actual effect of smoke-free legislation on stroke risk. Secondly, a majority of primary studies included in our review reported outcomes that were limited mainly to hospital admissions and mortality, while other adverse outcomes of stroke such as duration of hospitalisation or disability-adjusted life years (DALY) were not as commonly measured. Thirdly, studies that reported smoke free legislation other than in WRB and studies within administrative regions that already had a partial ban in place prior to comprehensive smoking bans were excluded to minimise inconsistencies in pre- versus post-ban analysis. As a result, the full effect of smoke free legislation implemented may be underestimated. Fourthly, we could not ascertain the effects on stroke risk in subpopulations of interest (e.g. age and gender) due to lack of data from the primary studies. Finally, the difference in timescales (data collected weekly vs monthly vs annually) between studies could have contributed in part to the heterogeneity observed in this study. Future research, in collaboration with epidemiologists, policymakers, and relevant stakeholders, would benefit from a standardised methodology to enhance utility for policymaking and ease of assessment, considering seasonal variations, autocorrelation, and other non-linear trends. The time scale should be made consistent across studies to allow for easier comparison (e.g. standardising to a monthly rate of hospital admission). Standardisation across studies will contribute to a more comprehensive understanding of the true effects of smoke-free laws on stroke, though we recognise the potential for collecting data at a finer time resolution will differ for resource-challenged settings.

Conclusion

This systematic review and meta-analysis found that legislative smoking bans in public spaces across different regions were associated with a reduced risk of hospital admission for stroke, supporting their role in public health policymaking.

Supplemental Material

sj-pdf-1-eso-10.1177_23969873241293566 – Supplemental material for Impact of smoke-free legislation on stroke risk: A systematic review and meta-analysis

Supplemental material, sj-pdf-1-eso-10.1177_23969873241293566 for Impact of smoke-free legislation on stroke risk: A systematic review and meta-analysis by Zhuo Xun Chua, Chern Yeh Lai Amanda, Timothy Jia Rong Lam, Jamie Si Pin Ong, Shermane Yun Wei Lim, Shivaram Kumar, Mervyn Jun Rui Lim, Benjamin Yong Qiang Tan, Joel Aik and Andrew Fu Wah Ho in European Stroke Journal

Footnotes

Acknowledgements

We would like to express gratitude to Madam Rebecca Lavanie David, Senior Medical Librarian at Nanyang Technological University, and Madam Toh Kim Kee, Senior Medical Librarian at National University of Singapore, for their guidance in the search strategy design and literature search.

Correction (May 2025):

Article Type updated from Review Article to Systematic Review

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project is funded by Singapore Heart Foundation. T.J.R.L. is supported by the SingHealth Medical Student Talent Development Award.

Ethical approval

Not applicable as it is a systematic review.

Informed consent

Not applicable as it is a systematic review.

Contributorship

All authors reviewed and edited the manuscript and approved the final version of the manuscript. Study concept and design: AFWH. Acquisition, analysis, or interpretation of data: ZXC, CYLA, TJRL, JSPO, SYWL, SK. Drafting of the manuscript: ZXC, CYLA, TJRL, SYWL, SK. Critical revision of the manuscript for important intellectual content: ZXC, TJRL, MJRL, BYQT, JA, AFWH. Statistical analysis: ZXC, TJRL, JA, AFWH. Administrative, technical, or material support: ZXC, TJRL, MJRL, BYQT, JA, AFWH. Study supervision: AFWH.

Trial registration

Not applicable as it is a systematic review.

ORCID iD

Zhuo Xun Chua

Supplemental material

Supplemental material for this article is available online.

References

World Health Organization. The top 10 causes of death [Internet], https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death (2024, accessed 16 March 2024).

Feigin

Brainin

Norrving

, et al. World stroke organization (WSO): global stroke fact sheet 2022. Int J Stroke 2022; 17: 18–29.

Shah

Cole

JW.

Smoking and stroke: the more you smoke the more you stroke. Expert Rev Cardiovasc Ther 2010; 8: 917–932.

National Cancer Institute. Definition of secondhand smoke - NCI dictionary of cancer terms - NCI [Internet], https://www.cancer.gov/publications/dictionaries/cancer-terms/def/secondhand-smoke (2011, accessed 16 March 2024).

CDCTobaccoFree. SGR: the health consequences of smoking—50 years of progress [Internet]. Centers for Disease Control and Prevention; 2014. https://archive.cdc.gov/www_cdc_gov/tobacco/sgr/50th-anniversary/index.htm (accessed 16 March 2024).

Lin

Ovbiagele

Markovic

, et al. Association of secondhand smoke with stroke outcomes. Stroke 2016; 47: 2828–2835.

Jones

Barnoya

Stranges

, et al. Cardiovascular events following smoke-free legislations: an updated systematic review and meta-analysis. Curr Environ Health Rep 2014; 1: 239–249.

Durham

A-D

Bergier

Morisod

, et al. Improved health of hospitality workers after a Swiss cantonal smoking ban. Swiss Med Wkly 2011; 141: w13317.

Rajkumar

Schmidt-Trucksäss

Wellenius

, et al. The effect of workplace smoking bans on heart rate variability and pulse wave velocity of non-smoking hospitality workers. Int J Public Health 2014; 59: 577–585.

10.

Goodman

Agnew

McCaffrey

, et al. Effects of the Irish smoking ban on respiratory health of bar workers and air quality in Dublin pubs. Am J Respir Crit Care Med 2007; 175: 840–845.

11.

Pell

Haw

Cobbe

, et al. Smoke-free legislation and hospitalizations for acute coronary syndrome. New Engl J Med 2008; 359: 482–491.

12.

Larsson

Boëthius

Axelsson

, et al. Exposure to environmental tobacco smoke and health effects among hospitality workers in Sweden–before and after the implementation of a smoke-free law. Scand J Work Environ Health 2008; 34: 267–277.

13.

Frazer

Callinan

McHugh

, et al. Legislative smoking bans for reducing harms from secondhand smoke exposure, smoking prevalence and tobacco consumption. Cochrane Database Syst Rev 2016; 2: CD005992.

14.

Tan

Glantz

SA.

Association between smoke-free legislation and hospitalizations for cardiac, cerebrovascular, and respiratory diseases: a meta-analysis. Circulation 2012; 126: 2177–2183.

15.

Gao

Wang

, et al. The effect of smoke-free legislation on the mortality rate of acute myocardial infarction: a meta-analysis. BMC Public Health 2019; 19: 1269.

16.

Akter

Islam

Rahman

, et al. Evaluation of population-level tobacco control interventions and health outcomes: a systematic review and meta-analysis. JAMA Netw Open 2023; 6: e2322341.

17.

Herman

Walsh

ME.

Hospital admissions for acute myocardial infarction, angina, stroke, and asthma after implementation of Arizona’s comprehensive statewide smoking ban. Am J Public Health 2011; 101: 491–496.

18.

Loomis

Juster

HR.

Association of indoor smoke-free air laws with hospital admissions for acute myocardial infarction and stroke in three states. J Environ Public Health 2012; 2012: 589018.

19.

Head

Jackson

Bae

, et al. Hospital discharge rates before and after implementation of a city-wide smoking ban in a Texas city, 2004–2008. Prev Chronic Dis 2012; 9: E179.

20.

Xiao

Zhang

Wang

, et al. Impact of smoke-free legislation on acute myocardial infarction and stroke mortality: Tianjin, China, 2007–2015. Tob Control 2020; 29: 61–67.

21.

Juster

Loomis

Hinman

, et al. Declines in hospital admissions for acute myocardial infarction in New York State after implementation of a comprehensive smoking ban. Am J Public Health 2007; 97: 2035–2039.

22.

Gaudreau

Sanford

Cheverie

, et al. The effect of a smoking ban on hospitalization rates for cardiovascular and respiratory conditions in Prince Edward Island, Canada. PLoS One 2013; 8: e56102.

23.

Galán

Simón

Flores

, et al. Assessing the effects of the Spanish partial smoking ban on cardiovascular and respiratory diseases: methodological issues. BMJ Open 2015; 5: e008892.

24.

Nazzal

Harris

JE.

Lower incidence of myocardial infarction after smoke-free legislation enforcement in Chile. Bull World Health Organ 2017; 95: 674–682.

25.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021; 372: n71.

26.

Cochrane Handbook for Systematic Reviews of Interventions [Internet]. [date unknown], https://training.cochrane.org/handbook (2023, accessed 16 March 2024).

27.

The well-built clinical question: a key to evidence-based decisions - PubMed [Internet] [date unknown], https://pubmed.ncbi.nlm.nih.gov/7582737/ (1995, accessed 16 March 2024).

28.

ROBINS-E Development Group ( Higgins

, et al.). Risk of bias tools - ROBINS-E tool [Internet] [date unknown], https://www.riskofbias.info/welcome/robins-e-tool (2024, accessed 16 March 2024).

29.

Higgins

JPT

Morgan

Rooney

, et al. A tool to assess risk of bias in non-randomized follow-up studies of exposure effects (ROBINS-E). Environ Int 2024; 186: 108602.

30.

Zeraatkar

Kohut

Bhasin

, et al. Assessments of risk of bias in systematic reviews of observational nutritional epidemiologic studies are often not appropriate or comprehensive: a methodological study. BMJ Nutr Prev Health 2021; 4: e000248.

31.

Jeyaraman

Rabbani

Copstein

, et al. Methodologically rigorous risk of bias tools for nonrandomized studies had low reliability and high evaluator burden. J Clin Epidemiol 2020; 128: 140–147.

32.

R Core Team. _R: a language and environment for statistical computing_. Vienna: R Foundation for Statistical Computing [Internet]. https://www.R-project.org/ (2023, accessed 16 March 2024).

33.

Schwarzer

Meta: general package for meta-analysis [Internet], https://cran.r-project.org/web/packages/meta/index.html (2024, accessed 16 March 2024).

34.

IntHout

Ioannidis

Borm

GF.

The hartung-Knapp-Sidik-Jonkman method for random effects meta-analysis is straightforward and considerably outperforms the standard DerSimonian-Laird method. BMC Med Res Methodol 2014; 14: 25.

35.

Stallings-Smith

Zeka

Goodman

, et al. Reductions in cardiovascular, cerebrovascular, and respiratory mortality following the national Irish smoking ban: interrupted time-series analysis. PLoS One 2013; 8: e62063.

36.

Xiao

Jia

, et al. Impact of Qingdao’s smoke-free legislation on hospitalizations and mortality from acute myocardial infarction and stroke: an interrupted time–series analysis. Addiction 2020; 115: 1561–1570.

37.

Kent

Sulaiman

Nicholson

, et al. Acute pulmonary admissions following implementation of a national workplace smoking ban. Chest 2012; 142: 673–679.

38.

Humair

J-P

Garin

Gerstel

, et al. Acute respiratory and cardiovascular admissions after a public smoking ban in Geneva, Switzerland. PLoS One 2014; 9: e90417.

39.

Saijo

Ido

Sato

, et al. Acute myocardial infarction and stoke after the enactment of smoke-free legislation in public places in Bibai city: data analysis of hospital admissions and ambulance transports. Hypertens Res 2019; 42: 1801–1807.

40.

Jarlstrup

Thygesen

Grønbæk

, et al. The impact of the Danish national smoking ban from 2007 on the incidence of eight smoking-related diseases: a nationwide register-based interrupted time series analysis. Nicotine Tob Res 2023; 25: 648–656.

41.

Chu

Liu

Fang

, et al. Effects of a smoke-free policy in xi’an, China: impact on hospital admissions for acute ischemic heart disease and stroke. Front Public Health 2022; 10: 898461.

42.

Naiman

Glazier

Moineddin

Association of anti-smoking legislation with rates of hospital admission for cardiovascular and respiratory conditions. Can Med Assoc J 2010; 182: 761–767.

43.

Wagenaar

Sherr

Fernandes

, et al. Using routine health information systems for well-designed health evaluations in low- and middle-income countries. Health Policy Plan 2016; 31: 129–135.

44.

Coryn

CLS

Hobson

. Using nonequivalent dependent variables to reduce internal validity threats in quasi-experiments: rationale, history, and examples from practice. New Dir Eval 2011; 2011: 31–39.

45.

Mackay

Haw

Newby

, et al. Impact of Scotland’s comprehensive, smoke-free legislation on stroke. PLoS One 2013; 8: e62597.

46.

Oono

Mackay

Pell

JP.

Meta-analysis of the association between secondhand smoke exposure and stroke. J Public Health 2011; 33: 496–502.

47.

Mercer

DeVinney

Fine

, et al. Study designs for effectiveness and translation research: identifying trade-offs. Am J Prev Med 2007; 33: 139–154.

48.

Trinidad

Gilpin

Pierce

JP.

Compliance and support for smoke-free school policies. Health Educ Res 2005; 20: 466–475.

49.

Bajoga

Lewis

McNeill

, et al. Does the introduction of comprehensive smoke-free legislation lead to a decrease in population smoking prevalence? Addiction 2011; 106: 1346–1354.

50.

Nagelhout

de Vries

Boudreau

, et al. Comparative impact of smoke-free legislation on smoking cessation in three European countries. Eur J Public Health 2012; 22 Suppl 1: 4–9.

51.

Wan

Protecting people from tobacco smoke in China: current status and challenges. China CDC Wkly 2022; 4: 460–464.

52.

Boes

Marti

Maclean

JC.

The impact of smoking bans on smoking and consumer behavior: quasi-experimental evidence from Switzerland. Health Econ 2015; 24: 1502–1516.

53.

Rehill

Beck

Yeo

, et al. The effect of chronic tobacco smoking on arterial stiffness. Br J Clin Pharmacol 2006; 61: 767–773.

54.

Wilson

Hoeg

D’Agostino

, et al. Cumulative effects of high cholesterol levels, high blood pressure, and cigarette smoking on carotid stenosis. New Engl J Med 1997; 337: 516–522.

55.

Abbruscato

Lopez

Mark

, et al. Nicotine and cotinine modulate cerebral microvascular permeability and protein expression of ZO-1 through nicotinic acetylcholine receptors expressed on brain endothelial cells. J Pharm Sci 2002; 91: 2525–2538.

56.

Abbruscato

Lopez

Roder

, et al. Regulation of blood-brain barrier Na,K,2Cl-cotransporter through phosphorylation during in vitro stroke conditions and nicotine exposure. J Pharmacol Exp Ther 2004; 310: 459–468.

57.

Hossain

Sathe

Fazio

, et al. Tobacco smoke: A critical etiological factor for vascular impairment at the blood–brain barrier. Brain Res 2009; 1287: 192–205.

58.

Mazzone

Tierney

Hossain

, et al. Pathophysiological impact of cigarette smoke exposure on the cerebrovascular system with a focus on the blood-brain barrier: expanding the awareness of smoking toxicity in an underappreciated area. Int J Environ Res Public Health 2010; 7: 4111–4126.

59.

Tsuchiya

Asada

Kasahara

, et al. Smoking a single cigarette rapidly reduces combined concentrations of nitrate and nitrite and concentrations of antioxidants in plasma. Circulation 2002; 105: 1155–1157.

60.

Aune

Schlesinger

Norat

, et al. Tobacco smoking and the risk of atrial fibrillation: a systematic review and meta-analysis of prospective studies. Eur J Prev Cardiol 2018; 25: 1437–1451.

61.

Fischer

Kraemer

Meta-analysis of the association between second-hand smoke exposure and ischaemic heart diseases, COPD and stroke. BMC Public Health 2015; 15: 1202.

62.

Office on Smoking and Health. Toxicology of secondhand smoke [Internet]. In: The health consequences of involuntary exposure to tobacco smoke: a report of the surgeon general. USA: Centers for Disease Control and Prevention; 2006. https://www.ncbi.nlm.nih.gov/books/NBK44321/ (accessed 23 March 2024).

63.

Septiono

Kuipers

MAG

, et al. The impact of local smoke-free policies on smoking behaviour among adults in Indonesia: a quasi-experimental national study. Addiction 2020; 115: 2382–2392.

64.

Bird

Kashaniamin

Nwankwo

, et al. Impact and effectiveness of legislative smoking bans and anti-tobacco media campaigns in reducing smoking among women in the US: a systematic review and meta-analysis. Healthcare 2020; 8: 20.

65.

Huisman

Kunst

Mackenbach

JP.

Inequalities in the prevalence of smoking in the European Union: comparing education and income. Prev Med 2005; 40: 756–764.

66.

Wynne

Guillaumier

Twyman

, et al. Signs, fines and compliance officers: a systematic review of strategies for enforcing smoke-free policy. Int J Environ Res Public Health 2018; 15: 1386.

67.

Weinehall

Ohman

“If I don’t smoke, I’m not a real man’ Indonesian teenage boys” views about smoking. Health Educ Res 2006; 22: 794–804.

68.

Julianto

Killing me softly with your corporate social irresponsibility: tobacco industry and its so-called corporate social responsibility [Internet], https://www.academia.edu/66902011/Killing_Me_Softly_With_Your_Corporate_Social_Irresponsibility_Tobacco_Industry_and_Its_So_Called_Corporate_Social_Responsibility (2015, accessed 23 March 2024).

69.

Donahoe

Titus

Fleischer

NL.

Key factors inhibiting legislative progress toward smoke-free coverage in Appalachia. Am J Public Health 2018; 108: 372–378.

70.

Wang

Lam

TH.

Parental smoking, exposure to secondhand smoke at home, and smoking initiation among young children. Nicotine Tob Res 2011; 13: 827–832.

71.

Wang

, et al. Smoking family, secondhand smoke exposure at home, and quitting in adolescent smokers. Nicotine Tob Res 2013; 15: 185–191.

72.

Feeney

Rossetti

Terrien

E-cigarettes-a review of the evidence-harm versus harm reduction. Tob Use Insights 2022; 15: 1179173X221087524.

73.

Banks

Yazidjoglou

Brown

, et al. Electronic cigarettes and health outcomes: umbrella and systematic review of the global evidence. Med J Aust 2023; 218: 267–275.

74.

Honeycutt

Huerne

Miller

, et al. A systematic review of the effects of e-cigarette use on lung function. NPJ Prim Care Respir Med 2022; 32: 45–47.

75.

Putaala

. Ischemic stroke in the young: current perspectives on incidence, risk factors, and cardiovascular prognosis. Eur Stroke J 2016; 1: 28–40.

76.

George

MG.

Risk factors for ischemic stroke in younger adults: a focused update. Stroke 2020; 51: 729–735.

77.

Markidan

Cole

Cronin

, et al. Smoking and risk of ischemic stroke in young men. Stroke 2018; 49: 1276–1278.

78.

Bhat

Cole

Sorkin

, et al. Dose-response relationship between cigarette smoking and risk of ischemic stroke in young women. Stroke 2008; 39: 2439–2443.

79.

Sargent

Demidenko

Malenka

, et al. Smoking restrictions and hospitalization for acute coronary events in Germany. Clin Res Cardiol 2012; 101: 227–235.

80.

Matheos

Liew

Zomer

, et al. Cost-effectiveness analysis of tobacco control strategies in Indonesia. Value Health Reg Issues 2023; 33: 65–75.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.02 MB