This systematic review and meta-analysis presented strong evidence concerning the effectiveness of smoke-free legislation on reducing the AMI mortality rate. Previous meta-analyses have estimated reductions in the incidence rate or hospital admission rate of the AMI after the implementation of smoke-free legislations [28,29,30]. To the best of our knowledge, this may be the first study to examine the relationship between AMI mortality rate and the corresponding smoke-free legislation.
The results of this study demonstrated that smoke-free legislation were associated with an 8% decrease in the AMI mortality rate in general. Furthermore, the degree of the decline in risk of AMI varied across different subgroups (2–23%). The AMI mortality rate decreased greatly among females (RR, 0.90; 95% CI: 0.87–0.94) and people aged ≤65 (RR, 0.90; 95% CI: 0.84–0.98). The different beneficial effects of legislation may be attributed to the diversity of study designs, including differences in the target populations, statistical analyses, types of smoke-free legislations, and differences in compliance with the law. . There were 2 studies that did not find that smoke-free legislation was associated with a lower AMI mortality rate [15, 32]. Different methodology concerning the coverage of local smoke-free legislations and the time during which the bans were in effect may result in contradicting results for AMI mortality.
The largest effect of smoke-free legislation on the AMI mortality rate was found in.
places with more comprehensive bans [14, 16, 21,22,23, 25]. Allowing smoking in designated smoking areas or in ventilated smoking rooms would not effectively prevent secondhand smoke exposure . More comprehensive smoke-free legislation could significantly reduce the number of active smokers , raise people’s awareness about the side effects of smoking, and more importantly, change social norms about the perception of smoking. Moritsugu noted that enacting more comprehensive tobacco-control legislation could effectively prevent exposure to secondhand smoke and reduce the number of smokers . Based on European studies, Ward found that the indoor PM2.5 concentration generally decreased by 68.4%, while for areas with partial bans on smoking, indoor PM2.5 concentration reduced by 40%, after introducing comprehensive smoke-free legislation . This provided sufficient evidence that enacting comprehensive smoke-free legislation was associated with lower level of exposure to secondhand smoke.
In this study, larger estimates of the effect were found in the smaller samples rather than in the larger samples, which was consistent with the results of a previous meta-analysis . In studies of small samples, the RR values may be more sensitive to random factors; in our study, the standard deviation of the mean RR values was 0.11, which was higher than that of the larger samples (SD = 0.05).
Although some studies (with a post-ban duration ≤2 years) indicated that smoke-free legislation could result immediate effects on the AMI mortality rate [14, 17, 22, 25], the present study showed that the impacts of smoke-free legislation could increase over time. Four studies suggested that the effect of smoke-free legislation on the AMI mortality rate increased with a longer follow up period [14, 17, 20, 24]. Dove believes that the smaller reduction in AMI deaths after the legislation maybe because we examined AMI mortality rates were what we examined rather than hospitalization rates; therefore, it may take a longer time to reveal an immediate downward trend . Figure 4 shows that studies with a longer follow-up period have smaller heterogeneity in the risk estimates than those in studies with a shorter follow up period.
The present study also showed that the decline in the AMI mortality rate was also associated with the situation where there were no smoking control measures before the smoke-free legislation [14, 16, 23]. The effect of smoke-free legislation may have been weakened when people had been protected by pre-existing local regulations. For example, California had developed local smoke-free measures before introducing nationwide smoke-free legislation, and almost 70% of the state’s population was protected . Enacting smoking control measures could reduce secondhand exposure to some extent, which may minimize the effect size of implementing smoke-free legislation on the AMI mortality rate .
Subgroup analyses were performed to explore the effect of smoke-free legislation on AMI mortality by sex and age (as shown in Figs. 10 and 11). The decline in the AMI mortality rate in the female group was 10%, which was 1% higher than that in the male group (Fig. 10). In addition, there was also a decrease of 10% in the group with aged less than 65, which was larger than that in the group with aged greater than 65 (Fig. 11). Furthermore, a sensitivity analysis was performed, after excluding studies with a high risk of bias (as shown in Fig. 12), to ensure the robustness of the analysis, and the overall RR of AMI mortality was 0.93 (95% CI: 0.90–0.96), there was no appreciable change.
It is worth noting that 2 studies, lacking estimates for the whole target population, could not be directly used for meta-analysis of the whole target population. However, in our sensitivity analyses, we included in the analysis from a study in Flanders stratified by sex and age, and from another study in Rome stratified by age [32, 39]. The results of the sensitivity analysis showed that smoke-free legislation resulted in larger effects on AMI mortality rate within the female group [14, 20] and the younger people group (aged less than 65) [14, 17, 32]. One plausible explanation is that prohibiting smoking in the workplace could reduce the smoking prevalence in work areas among young working populations . In addition, the older people are less likely to go to bars and similar venues, which could lead to smaller relatives risks associated with secondhand smoke exposure in this age group people . 35% of females and 33% of males were exposed to secondhand smoke , but the smoking prevalence among females was relatively lower compared to that among males. Therefore, after implementing smoke-free legislation, more females would be protected from secondhand smoke exposure, which would lead to fewer women dying of AMI.
Other smoking control measures after smoke-free legislation also influence the AMI mortality rate. Thach found that in Hong Kong, the increase in the tobacco tax strengthened the effectiveness of smoke-free legislation . McAlister also estimated that the influence of raising the tobacco tax would be reflected in the smoking prevalence .
In this study, several limitations should be noted. First, we analyzed the change in the AMI mortality rate before and after implementing smoke-free legislation, but the causal relationship between them could not be explored. It was not possible to find a location that was identical and that did not have smoke-free legislation to include as a control group when we assessed the effects of smoke-free legislation on the AMI mortality rate. The data included in this study were all extracted from time series data to evaluate the potential effect of smoke-free legislation.
Second, during subgroup analysis, the comprehensiveness of the smoke-free legislation was entered into the model as an ordinal variable (0 for workplaces only; 1 for workplaces, restaurants and bars) to test whether comprehensive laws were more beneficial to the considerable decline in the AMI mortality rate. Hence, the expected decrease in risk per 1% or per 1-standard deviation (SD) decrease in the AMI mortality rate, in the meta-analysis, could not be calculated.
Third, studies included in this meta-analysis did not account for the nonlinear trend in the AMI mortality rate, which might be concern in our estimates. The nonlinear secular trend could be explained by the concomitant effect of other time-varying factors , only 2 studies compared models with different specifications of secular trend and showed that the estimated effect was attenuated under the condition of nonlinearity in the secular trend of declining AMI [22, 41] .
Fourth, inaccuracies might have existed due to the lack of control for other confounding factors. Some studies included in this meta-analysis only considered individual factors, such as sex and age, and few studies have considered air quality. However, other studies have documented that environmental factors such as air temperature, air pressure, and air quality were related to AMI mortality rates [42,43,44]. Furthermore, technological advances in medicine  and increases in tobacco taxes and prices would also influence the AMI mortality rate . In this meta-analysis, 2 studies did not consider confounding factors [16, 24].
Despite the methodological limitations of the individual studies included, the present study still provided evidence-based assessments of the effect of smoke-free legislation on AMI mortality rates around the world. These results may provide evidence for promoting smoke-free legislation in areas that allow indoor and outdoor smoking.