Article Text
PDF
Abstract
Background Japan’s ‘Health Japan 21’ initiative targets a reduction in adult smoking prevalence to 12% by 2032. This study evaluates the probability of meeting this target at both national and prefectural levels by estimating and comparing prevalence trends.
Method Using crude smoking prevalence data from 2001 to 2022 for the whole nation and across Japan’s 47 prefectures, we used Bayesian linear regression to project future smoking trends up to 2100. We calculated the posterior probabilities of each prefecture achieving the target by 2032 and projected the timeline for meeting this target. We defined ‘meeting the target’ as having a 60% or higher probability.
Results Nationally, 2022 smoking prevalence was 16.09%, accompanied by an annual reduction rate of 3.75%. There is a 64.3% probability of achieving the 12% smoking prevalence target by 2032, with considerable prefectural variation. Out of 47 prefectures, 19 are on track to meet the target by 2032, whereas 28 are predicted to lag this deadline. For example, Tokyo (99.3% probability) and Nara (98.0% probability) could potentially reach this target as early as 2026 whereas Fukushima, Iwate and Aomori have <1% probability of achieving the target by 2032 and are not projected to reach the target until 2046 or later.
Conclusions While Japan may achieve its national smoking reduction target by 2032, there are large regional variations. This variability underscores the need for tailored public health strategies that address the unique challenges faced by each prefecture to ensure a cohesive and effective tobacco control approach across the nation.
- Disparities
- Surveillance and monitoring
- Global health
- Public policy
Data availability statement
Data are available in a public, open access repository. The data used in the project are open-access data from the Comprehensive Survey of Living Conditions, published by the Ministry of Health, Labour and Welfare of Japan (https://ganjoho.jp/reg_stat/statistics/data/dl/index.html%23smoking).
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Japan’s progress in implementing comprehensive tobacco control measures has lagged behind international standards set by the WHO Framework Convention on Tobacco Control (FCTC) and MPOWER guidelines (a policy package developed by WHO consisting of six evidence-based strategies: Monitor tobacco use and prevention policies; Protect people from tobacco smoke; Offer help to quit tobacco use; Warn about the dangers of tobacco; Enforce bans on tobacco advertising, promotion, and sponsorship; and Raise taxes on tobacco). Despite some progress, regional variations in tobacco control efforts persist, and smoking prevalence remains high, particularly among men. The ‘Health Japan 21’ initiative aims to reduce adult smoking prevalence to 12% by 2032, but the trajectory to achieving the target is unclear.
WHAT THIS STUDY ADDS
This study provides the first prefecture-level projections of smoking prevalence in Japan up to 2100. While Japan overall has a 64.3% probability of achieving the 12% target by 2032, 28 out of 47 prefectures are predicted to miss this deadline, with some not reaching the target until the 2050s. These findings may indicate significant regional disparities in tobacco control effectiveness across Japan, resulting from the gap between national policy and local implementation.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These projections can inform more targeted and effective tobacco control strategies in Japan. The findings underscore the need for improved coordination between national policies and local implementation, as well as tailored, region-specific approaches. This may involve efforts to strengthen local government capacity to implement and enforce national tobacco control laws and to align local policies more closely with the FCTC and MPOWER guidelines, which could ultimately reduce geographical disparities in smoking prevalence.
Introduction
Japan’s battle against tobacco consumption is marked by significant contention, especially due to the influential lobbying by major tobacco companies like Japan Tobacco.1 Over the past two decades, the government has introduced multiple phases of its ‘National Health Promotion Movement in the 21st Century (Health Japan 21),’ each outlining ambitious targets for reducing smoking prevalence. The second term of this initiative (2013–2022) aimed to lower adult smoking rates from 19.5% in 2010 to 12% by 2022; however, an interim evaluation in 2018 found that only 21 out of 53 health objectives were likely to be met, indicating that the smoking-reduction goal would almost certainly be missed.2 Subsequent projections suggested that Japan might not achieve a 12% smoking rate until 2041 under existing policies, prompting a third term of Health Japan 21 with an extended timeline.3 In 2023, despite ongoing industry-related challenges, the Japanese government reaffirmed its commitment to reduce overall smoking prevalence to 12%—this time by 2032.4 However, the effectiveness of national policies often hinges on local implementation strategies, with prefectural and municipal governments playing crucial roles in implementing public health measures. Discrepancies in local implementation of tobacco control have been shown to lead to significant regional variations in smoking prevalence.5 In 2001, Japan’s overall adult smoking prevalence was estimated to be about 30.5%, with men’s smoking rates often exceeding 48.4% and women’s hovering around 14%. While the national average has steadily declined over the past two decades—reaching approximately 16.9% by 2022—significant regional disparities persist. For instance, Tokyo now reports one of the lowest prefectural prevalences at just over 13%, whereas Aomori and Fukushima continue to exceed 20%.6
Japan’s 12% reduction target also appears relatively ‘soft’ relative to endgame strategies set by other high-income nations—such as New Zealand, Ireland and England—many of which aim to bring smoking prevalence below 5%.7–9 Against these global benchmarks, Japan’s target may reflect both historical policy constraints and strong commercial interests that have limited the speed and scope of tobacco-control reforms.
This study aims to evaluate progress towards meeting the Health Japan 21 target by estimating the probability of reaching the target and time needed to meet the target at both national and prefectural levels. By quantifying regional disparities and timelines, we provide crucial data to inform policy-makers, enabling them to tailor and potentially intensify interventions in regions that are behind.
Methodology
Data preparation
This study used point estimate prevalence of people who currently smoke spanning from 2001 to 2022 for the entire nation and each of Japan’s 47 prefectures, accessed through e-Stat, the Japanese government’s official statistics portal.10 These point estimates were precalculated based on data from the Comprehensive Survey of Living Conditions, and people who currently smoke were defined as individuals who responded ‘yes’ to either of the following statements: ‘I smoked daily’ or ‘I smoked sometimes’.11 The dataset includes distinct records of smoking prevalence yet relies on a single survey question that does not distinguish between cigarettes and other tobacco products, every 3 years from 2001 to 2022 nationally and for each prefecture. For the purposes of statistical analysis, these prevalence figures were converted to the logit scale to suit the constraints of the data. A prefecture is the first-level jurisdiction in Japan, roughly comparable to a state or province in other countries. Japan has 47 prefectures, which vary widely in geography, population density and socioeconomic status (SES).
Bayesian model specification
We employed a Bayesian linear regression model for the national trend and each prefecture separately to assess changes in smoking prevalence over time. The model is specified as follows:
In these expressions, Yi represents the logit-transformed smoking prevalence for each data point i, that is,
Where Pi is the proportion of people reporting that they smoked tobacco at time point t.
In the linear term, μi denotes the linear predictor, α is the intercept, β indicates the slope or the rate of change per year, and τ stands for the precision of the normal distribution. The term timei is the year.
The model includes prior distributions for the parameters, reflecting weak prior knowledge:
We calculated the annual reduction rate (ARR) in smoking prevalence (%) using the coefficient obtained from the linear regression model on the logit scale. Specifically, the ARR is computed as:
Where β indicates the slope. The 95% credible intervals (95% CrI) for the beta coefficient were estimated.
Statistical inference and forecasting
The model was implemented using Just another Gibbs Sampler (JAGS). It was employed to simulate the Bayesian model, initialising three Markov chains and conducting a burn-in phase of 5000 iterations to ensure convergence, followed by 10 000 iterations for sampling and forecasting of smoking prevalence until 2100.12 Model convergence was assessed visually by trace plots.
Target prevalence forecasting
We estimated the likelihood of reaching the Japanese government smoking prevalence target of a reduction to 12% prevalence nationally by 2032. The posterior probabilities of achieving the target at the national and prefectural levels were calculated based on the distributions generated from our model simulations. These probabilities were categorised into:
Highly Likely (0.8–1.0 probability).
Likely (0.6–0.79).
Possible (0.4–0.59).
Unlikely (<0.4).
We then visually assessed geospatial patterns by plotting these probabilities on a map.
Additionally, we projected the years by which each prefecture is expected to reach the target. We also estimated a secondary target, the year of reaching the endgame smoking reduction goal (<5% smoking prevalence). To conduct a post hoc investigation into potential socioeconomic influences on achieving the tobacco reduction target, we examined minimum wage levels across prefectures in 2024 as a proxy indicator for regional economic conditions and population density in 2019 (population per 1 km2 of total land area).13 14 We defined reaching the target as having a probability of 60% or higher. The analysis was done using R V.4.3.3 and JAGS V.4.3.1.
Results
Our national analysis predicted that Japan was projected to meet its target of reducing smoking prevalence to 12% by 2032, with a likelihood of 64.3%. The national smoking prevalence in 2022 was 16.09%, accompanied by an ARR of 3.75% (95% CrI 2.76% to 4.73%). This national trend, however, masks significant variation at the prefecture level, both in current smoking prevalence and in the rates of annual reduction. In 2022, Tokyo exhibited the lowest smoking prevalence at 13.54%, with an ARR of 4.84% (95% CrI 3.76% to 5.9%), leading to the highest probability of 99.3% of meeting the target by 2026. Similarly, Kanagawa and Gifu, with 2022 prevalences of 14.67% and 15.56%, respectively, and high ARRs of 4.5% (95% CrI 3.51% to 5.46%) and 3.72% (95% CrI 2.55% to 4.78%), are also likely to meet the target ahead of 2032, with probabilities exceeding 80%. Conversely, prefectures like Fukushima, Iwate and Aomori present a stark contrast. Fukushima has the highest 2022 smoking prevalence at 21.38% and a low ARR of 2.27% (95% CrI 1.14% to 3.37%), resulting in a negligible probability of reaching the target by 2032 and a projected achievement year of 2053. Similarly, Iwate and Aomori, with prevalences of 18.98% and 20.44%, and reduction rates of 2.41% (95% CrI 1.45% to 3.38%) and 2.81% (95% CrI 1.78% to 3.84%), show probabilities under 1%, with projected target years extending to 2046 and beyond. The estimated probabilities and projected years by which each prefecture is expected to achieve the 12% smoking prevalence target are detailed in table 1.
Probabilities of reaching the Japanese government target of 12% smoking prevalence or less in 2032 with the year of achieving the target
The achievement levels are shown on a choropleth map in figure 1. Urban regions such as Tokyo, Kanagawa and Osaka display higher probabilities (green and blue). In contrast, rural areas, especially in the northern (Hokkaido and Tohoku) and southwestern regions (Kyushu and Okinawa), are less likely to meet the target (red). There is a notable clustering of higher likelihoods in central prefectures around the Tokyo metropolitan area and other urban areas such as Osaka, while lower probabilities are concentrated in the northern and southwestern regions. Other high performers, such as Ehime and Hiroshima, are also evident. As illustrated in figure 2A,B, prefectures occupying the highest quintiles of both minimum wage and population density generally have higher probabilities of achieving the 12% smoking target (figure 2A, red scale) and faster ARRs (figure 2B, blue scale). This pattern indicates that wealthier urban areas are experiencing faster declines in prevalence and have a higher probability of reaching the stated policy goal.
Geospatial distribution of the likelihood of reaching the ‘Health Japan 21’ smoking reduction target by 2032 across Japan’s 47 prefectures.
(A) Heatmap depicting prefectures’ probability of reaching a 12% smoking-prevalence target by 2032 (red scale) based on minimum wage (x-axis) and population density (y-axis) and (B) heatmap showing annual reduction rates in smoking (blue scale) under the same socioeconomic axes. Deeper colours in each panel indicate higher probabilities or faster reductions, while lighter shades represent lower probabilities or slower reductions.
The projected year of reaching the endgame smoking reduction target is presented in online supplemental material 1. A comprehensive set of trend lines for both national and regional analyses is presented in online supplemental material 2, alongside observed versus predicted regression lines for each prefecture to evaluate the model fit.
Supplemental material
Supplemental material
Discussion
This study aimed to evaluate the progress towards the ‘Health Japan 21’ target of 12% smoking prevalence at the national and prefectural level. The analysis indicates a promising national trajectory towards achieving Japan’s 2032 tobacco reduction target, with a greater probability of success (0.643) than failure. However, we uncovered significant regional discrepancies in our prefecture-level forecasts. Of Japan’s 47 prefectures, 22 are deemed unlikely to meet the 2032 target for reducing smoking prevalence. Furthermore, approximately nine prefectures may require an additional decade beyond 2032 to achieve this goal. Regional variations in meeting the 2032 target can be attributed to disparities in ARR that likely reflect differences in tobacco control effectiveness. A compelling example of this is the contrast between Fukushima and Hokkaido prefectures. Hokkaido, despite starting with the highest baseline prevalence in 2001 (38%), is projected to reach the target by 2039—just 7 years after the deadline—due to its robust ARR of 3.98% (95% CrI 2.59% to 5.28%). In contrast, Fukushima, which started with a lower 2001 baseline prevalence (30%), is not projected to reach the target until 2053, primarily due to its significantly lower ARR of 2.27% (95% CrI 1.14% to 3.37%). This stark difference in progress, despite Hokkaido’s higher starting prevalence, suggests that the effectiveness of local tobacco control measures may be more crucial than initial smoking rates in determining success in meeting the national target.
Despite Japan ratifying the WHO Framework Convention on Tobacco Control (FCTC) in 2005, its national tobacco control policies have not fully complied with WHO FCTC standards, particularly in the areas of smoke-free regulations and tobacco advertising.15 16 Comprehensive smoke-free laws were only introduced in 2020 under revision of the Health Promotion Act, and still included numerous exceptions such as designated smoking rooms and small-sized restaurants. The varying levels of smoking prevalence across prefectures may hint at the influence of local legislation, such as the early ordinances. For example, Kanagawa Prefecture’s ordinance from 2010, which bans smoking in public spaces including schools and hospitals, has significantly advanced its progress.17 We project that Kanagawa prefecture will meet the 12% prevalence target by 2028. Similarly, robust indoor smoking bans in Hyogo and Tokyo may have set these regions on a faster track to reach their targets, as our analysis showed that Tokyo will reach the target in 2026 followed by Hyogo in 2028.18 19
Another significant challenge to local tobacco control efforts is the tobacco industry’s influence, which includes active lobbying of government and parliament candidates. Candidates primarily from the ruling party (Liberal Democratic Party) receive substantial donations from the tobacco industry during election campaigns, which is thought to have hindered the progress of tobacco control in Japan.20 An ecological study by Yorifuji et al has also demonstrated a connection between tobacco leaf production and adherence to tobacco control regulations, with poorer compliance with tobacco control laws in regions that are major producers of tobacco leaves.5 21 Our results are in line with the study by Yorifuji et al indicating that prefectures—often rural areas—housing tobacco-related industries tend to lag in reducing smoking prevalence. For instance, Iwate and Aomori, which are key tobacco-producing rural prefectures, are projected to lag by approximately 15 years in meeting the Japanese government’s smoking reduction target. Socioeconomic factors appear to play an additional role: prefectures with lower minimum wages exhibit lower probabilities of achieving the 12% threshold and slower annual reductions. This economic disparity aligns with previous research showing that individuals with lower incomes and greater economic difficulties are more likely to smoke.22 Taken together, these dynamics suggest that rural location and limited economic resources may exacerbate regional disparities, highlighting the need for tailored, locally adaptive policies to accelerate tobacco control. Although the present study suggests that Japan overall is on track to meet its own target, historical data and current results demonstrate that further efforts are needed at the regional levels. Due to the suboptimal national tobacco control, it is recommended to implement more stringent tobacco control measures at the regional level, following the WHO MPOWER guidelines, which have been shown to accelerate the reduction of smoking prevalence in Japan, as well as other novel strategies, such as reducing the number of tobacco retailers.3 Furthermore, the current target of 12% prevalence is much higher than the target set in other high-income countries such as New Zealand, Ireland and England, where they set tobacco endgame as a national target close to 5%.7–9 To align with the global movement towards a tobacco endgame (ie, setting a smoking prevalence target below 5%), setting a more ambitious target for the nation and regions that are on track to meet the current target is crucial.23 Lastly, given the rapidly changing landscape of tobacco and alternative nicotine products, notably the growing popularity of heated tobacco products, comprehensive monitoring of these products and updating laws to reflect their growing role is also important.5 24
Since the 12% target was set based on national smoking prevalence—an average that naturally balances higher prevalence in some areas against lower rates in others—significant regional disparities can remain obscured. In particular, large metropolitan areas like Tokyo, with comparatively progressive legislation and a substantial population, may drive the national average down sufficiently to meet headline targets, even as other regions continue to lag behind. This approach, set in 2023, risks under-representing the need for more robust interventions in less-supported prefectures and may create a lack of transparency about the true scope of tobacco use in Japan. To address these concerns, it is critical to establish targeted strategies that account for geographic, demographic and product-specific factors—distinguishing, for instance, between smoked cigarettes, heated tobacco products and nicotine pouches. Implementing area-specific and gender-specific goals, alongside systematic monitoring of progress for each tobacco product category, would offer a more nuanced understanding of where the most urgent efforts are needed. Such an approach fosters equitable resource allocation and ensures that high-prevalence prefectures are neither overlooked nor masked by the successes of more urbanised regions.
Furthermore, the observed regional disparities highlight the potential value of simulation studies to examine how implementing MPOWER measures at the prefectural level may influence smoking prevalence forecasts and help prioritise interventions. Collecting data on local tobacco control policies and MPOWER scores across prefectures, and subsequently applying simulation models similar to those used by Yang et al in a national-level study for Japan, could offer critical, granular insights.3
This study has limitations. The analysis primarily relied on temporal trends as the predictive variable, omitting other influential factors such as age, SES and tobacco outlet density. Second, the model did not disaggregate smoking prevalence by sex, despite clear evidence of higher smoking rates among men in Japan. Due to using the Japanese government’s sex-aggregated target, this might lead to overestimation or underestimation of smoking rates. These elements, along with the consumption patterns of other tobacco and nicotine products like heated tobacco products, waterpipes and nicotine pouches, could significantly impact smoking prevalence and should be integrated into future research to enhance forecast accuracy and relevance. Additionally, because the questionnaire did not differentiate between cigarette smoking and exclusive heated tobacco use, the resulting estimates may conflate distinct usage patterns. Future surveys should adopt product-specific measures to capture the growing diversity of tobacco consumption. Furthermore, the indicators used in the post hoc investigation were at the prefecture level, which might increase the tendency of ecological fallacy, and future analysis should consider using smaller administrative units like municipalities.
Conclusions
This study demonstrates Japan’s mixed outcomes towards achieving the ‘Health Japan 21’ target of reducing smoking prevalence to 12% by 2032. While national-level data predict a probable achievement of this goal, significant discrepancies exist at the prefectural level, highlighting the complex interplay of factors influencing smoking rates across regions. This underscores the critical need for region-specific strategies that address local challenges and disparities in tobacco control efforts, ensuring equitable health outcomes across all prefectures. Additionally, having a more ambitious target for the nation and regions that are on track could motivate policymakers and stakeholders to implement more stringent tobacco control measures, leading to greater advancements towards achieving the ultimate goal of eliminating tobacco use and its associated health risks.
Data availability statement
Data are available in a public, open access repository. The data used in the project are open-access data from the Comprehensive Survey of Living Conditions, published by the Ministry of Health, Labour and Welfare of Japan (https://ganjoho.jp/reg_stat/statistics/data/dl/index.html%23smoking).
Ethics statements
Patient consent for publication
Acknowledgments
This work was supported by the Health Labour Sciences Research Grant of the Ministry of Health, Labour and Welfare of Japan (Grant Numbers: 22FA1002, 22FA2001).
Footnotes
Contributors HJ and KT conceived the study idea. HJ and SG performed the data analyses. HJ created all visualisations and drafted the initial manuscript. KT, SG and KK reviewed and revised the manuscript. KT supervised the project. All authors reviewed the final draft and approved it for submission. KT is the guarantor of this work.
Funding This work was supported by the Ministry of Health, Labour and Welfare of Japan (grant numbers 22FA1002 and 22FA2001).
Map disclaimer The depiction of boundaries on this map does not imply the expression of any opinion whatsoever on the part of BMJ (or any member of its group) concerning the legal status of any country, territory, jurisdiction or area or of its authorities. This map is provided without any warranty of any kind, either express or implied.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.