Legal blood alcohol concentration limits and alcohol-attributable traffic mortality rates: an analysis across 165 countries

Summary

Background

Alcohol-impaired driving remains a leading cause of road traffic deaths worldwide, yet despite widespread adoption, the effectiveness of national blood alcohol concentration (BAC) limits across diverse structural conditions and population risks remains poorly understood.

Methods

We conducted a cross-national multilevel analysis of 165 countries using 2019 data on alcohol-attributable traffic mortality rates (ATMRs), national legal BAC limits, and structural country-level indicators. The primary outcome was ATMRs, defined as age-standardized death rates per 100,000 population among individuals aged 15 and older. Key predictors included national BAC limits, sex, and their interaction. Models were adjusted for national income, healthcare system infrastructure, gender inequality, and per capita alcohol consumption.

Findings

Lower national BAC limits were associated with lower ATMRs, with significantly stronger effects observed among males versus females (β = −2.58, p < 0.01). Structural factors, including lower national income (β = −0.31, p < 0.01), greater gender inequality (β = 1.61, p < 0.01), and higher alcohol consumption (β = 0.16, p < 0.001) each predicted higher ATMRs. Our final model explained 71% of the variance in ATMRs.

Interpretation

ATMRs represent a preventable global burden, with World Health Organization (WHO) BAC guidelines being exceeded in ∼30% of countries examined. While lower BAC limits reduce ATMRs overall for both women and men, men experienced a disproportionately higher share of the mortality burden. Further reduction in ATMRs would be maximized by considering country-level structural factors.

Funding

NIH T32AA029259(CLJC), U54AA027989(SAM), K01AA029706(YZ), R01AA030971(TLV), Office of the Assistant Secretary of Defense for Health Affairs W81XWH-22-2-0081, PASA3(MRP/AAM) and Veterans Affairs VISN1-CDA(MRP).

Research in context.

Evidence before this study

Prior studies and policy reports have examined the effects of lower blood alcohol concentration (BAC) limits on alcohol-related traffic fatalities; however, most analyses have been limited to single-country studies and have generally not accounted for sex differences or broader country-level structural factors. This pattern reflects the lack of comparable cross-national data on legal BAC limits. To assess the availability of such data, we reviewed policy documents and legal sources identified through bibliographic databases, government and road safety agency websites, international legal repositories, and reports from global agencies. Searches were conducted between December 2023 and March 2025, without language restrictions, using terms related to BAC legislation, alcohol-impaired driving, and road safety. We identified no global dataset compiling comparable national BAC limits for most countries included in the 2019 WHO estimates of alcohol-attributable traffic mortality rates (ATMRs) and found no cross-national studies examining the association between legal BAC limits and ATMRs.

Added value of this study

This study is the first to use sex-disaggregated, cross-national data to examine associations between legal BAC limits and ATMRs across 165 countries. By integrating national BAC policy data with structural indicators, including income level, healthcare system infrastructure, gender inequality, and per capita alcohol consumption, this investigation offers a multilevel, sex-informed understanding of how impaired driving policies perform within diverse national contexts. Findings demonstrate that more permissive BAC thresholds are associated with higher mortality rates, particularly among males, and that this association is strongly influenced by broader country-level conditions, such as lower national income, greater gender inequality, and higher per capita alcohol consumption.

Implications of all the available evidence

Together with existing evidence on the effectiveness of BAC legislation, these findings reinforce the notion that national impaired driving laws are embedded within broader social systems that influence their effect. Structural factors, including high alcohol availability, greater gender inequality and lower national income, amplify mortality risks, particularly among males, while strong healthcare systems are associated with lower mortality overall. Effective policy responses will require both stricter BAC thresholds in tandem with broader public health investments in infrastructure, gender equity, and alcohol harm reduction, with particular attention to vulnerable populations. Future research should consider enforcement practices and cultural norms to better understand the conditions under which BAC policies can most effectively reduce ATMRs.

Introduction

Alcohol-impaired driving is a persistent and preventable cause of road traffic mortality worldwide. Despite decades of road safety initiatives, alcohol use continues to play a major role in traffic fatalities, accounting for an estimated 27% of all road traffic deaths.¹ This burden disproportionately affects low- and middle-income countries (LMICs), where alcohol involvement in fatal crashes substantially higher than in high-income countries.² As countries continue to revise legislation on blood alcohol concentration (BAC) limits for drivers, understanding how these policies perform across diverse socioeconomic contexts remains a global health priority.

Legal limits on BAC are among the most widely implemented policy measures to reduce alcohol-attributable traffic mortality rates (ATMRs). The World Health Organization (WHO) recommends a maximum BAC of 0.05 g/dL for the general population and 0.02 g/dL for novice drivers.² To put this into context, a 68 kg man will reach a BAC of approximately 0.05 g/dL after consuming three standard drinks in 2 h, and a BAC of 0.02 g/dL after a single drink consumed in 30 min. Women reach these limits after consuming significantly less alcohol due to sex-based differences in alcohol metabolism.³ Evidence from high-income countries suggests that stricter BAC limits are associated with a reduction in ATMRs.⁴ However, the impact of this legislation not only depends on the numerical threshold, but also on the consistency and severity of enforcement, which varies widely across countries.⁵ Beyond enforcement, whether BAC policies yield comparable outcomes across diverse global regions remains unclear. Much of the existing literature relies on single-country analyses that overlook sex-specific outcomes and rarely consider country-level structural factors such as income, alcohol consumption levels, healthcare system infrastructure, or gender inequality, factors known to influence policy effectiveness.⁶

Sex differences are an important and understudied factor of ATMRs. Globally, males are more likely than females to operate a motor vehicle under the influence of alcohol and be victims of alcohol-related accidents (Fig. 1).⁷ Despite these disparities, few national BAC policies explicitly consider sex differences in their development or evaluation, limiting the ability to assess whether such measures are equally protective across populations. In addition, country-level structural factors, including income level, healthcare system infrastructure, alcohol consumption patterns, and gender inequality, may also influence the effectiveness of alcohol-related traffic policies.8, 9, 10 In many LMICs, underregulated enforcement systems and limited resources often undermine the intended impact of BAC laws.^8,9 Gender inequality, characterized by disparities in reproductive health outcomes, political participation, and employment, further compounds these risks, especially in countries where women's mobility, access to care, and representation in policymaking are limited. Evidently, country-level structural factors play an important role in shaping policy decisions, influencing both the strength of protections in place and where those protections remain incomplete. Yet, few cross-national studies account for these structural conditions, limiting the relevance of existing evidence for global policy guidance.

Fig. 1.

Sex differences in global alcohol-attributable traffic mortality rates (ATMRs). Note: ATMR data were obtained from the 2019 World Health Organization Global Health Observatory Data Repository. Values shown are untransformed and expressed per 100,000 population for interpretability.

Thus, this study examined the association between national legal BAC limits and ATMRs across 165 countries. Specifically, we assessed whether more restrictive BAC limits were associated with lower ATMRs and whether this relationship differed between sexes. Furthermore, we evaluated whether these relationships persisted after accounting for country-level structural factors, including income level, healthcare system infrastructure, gender inequality and per capita alcohol consumption. Based on the existing literature, we hypothesized that more restrictive BAC thresholds would be associated with lower ATMRs, with stronger effects among males due to higher rates of alcohol-impaired driving. We also hypothesized that the protective effect of stricter BAC limits would be weakest in LMICs, where systemic barriers limit policy impact.

Methods

Study design and data sources

We conducted a multilevel, cross-national study using country-level data from 2019 to examine associations between national legal BAC limits and ATMRs. The initial sampling dataset included all 193 United Nations (UN) member states. We first obtained 2019 data on ATMRs from the WHO Global Health Observatory Data Repository, which represented the most recent year with sex-disaggregated data available as of August 2024. National legal BAC thresholds were obtained through an extensive review of government sources, international road safety reports, including the WHO Global Status Report on Road Safety, and other public health sources. The final sample included a total of 165 countries. Additional country-level covariates including, income classification (based on gross national income), per capita alcohol consumption, healthcare system infrastructure, and gender inequality, were obtained from publicly available data from the World Bank Group, Global Health Security Index, and UN Development Programme, respectively. Ethical approval and informed consent were not required, as this study involved secondary analyses of publicly available, de-identified data.

Primary outcome

The outcome of interest was the ATMR, defined as the age-standardized death rate per 100,000 individuals aged 15 years or older, disaggregated by sex. For each country, the dataset included six combinations of sex at birth (male, female, both) and two age groups (all ages, ages 15+); we selected the sex-specific estimates for individuals aged 15 years and older to reflect the legal driving age in most countries. A total of fourteen countries were excluded due to missing ATMR data.

ATMRs were obtained from the WHO Global Health Estimates (GHE) database, which reports modeled, age-standardized death rates for causes of death statistically attributable to alcohol, including road-traffic injuries. These rates are modeled by the WHO to enable global comparability. In the WHO framework, alcohol-attributable mortality rates are estimated through alcohol-attributable fractions (AAFs), the proportion of deaths that would not have occurred in the absence of alcohol use. AAFs are calculated from national data on alcohol consumption and the established risk relationships between different levels of exposure and specific injury or disease categories.^11,12 For road-traffic mortality rates, these modeled fractions are applied to each country's cause-specific mortality data to estimate the proportion of deaths attributable to alcohol. The resulting alcohol-attributable mortality rates are then age-standardized to the WHO world standard population to ensure comparability across populations with differing age distributions.¹² ATMRs analyzed in this study represent WHO-standardized, age-adjusted estimates of deaths from road traffic injuries statistically attributable to alcohol consumption, expressed per 100,000 population.

Independent variable

The national legal BAC limit was defined as the maximum permissible BAC limit for drivers in each country and measured as a continuous variable in grams per deciliter (g/dL). BAC limits were recorded as of 2019 to align with the mortality data. Values were compiled through an extensive review of publicly available sources on government websites, road safety reports, international legal databases, and, when necessary, direct correspondence with national ministries of health. Countries with zero-tolerance policies or alcohol prohibition were coded as 0.00 g/dL. In countries where BAC limits varied across regions, the most commonly applied threshold (i.e., mode) was used to represent the national value. A total of fourteen countries were excluded due to insufficient information on legal BAC limits, including cases where limits were either unknown or not specified (Supplemental Table S1).

Covariates

Country income classification was obtained from the World Bank Group's 2019 gross national income per capita-based income categories and coded as ordinal variable between 1 and 4 (1 = low income, 2 = lower-middle income, 3 = upper-middle income, 4 = high income). Gross National Healthcare system infrastructure was assessed using the 2019 Global Health Security Index (GHSI) (Category 4), with scores ranging from 0 to 100 entered as a continuous variable (0 = poor health system capacity, 100 = fully prepared system). Gender inequality was assessed using the 2019 Gender Inequality Index (GII) from the UN Development Programme, with published fractional values ranging from 0 (equality) to 1 (maximum inequality) entered as a continuous variable. Per capita alcohol consumption was obtained from the World Bank Group and measured in liters of pure alcohol consumed annually per person aged 15 years or older, using 2019 projected estimates. All covariates were measured at the national level; while subnational variation exists, nationally standardized indicators were necessary to facilitate cross-national comparability.

Statistical analyses

To examine the relationship between national BAC limits and sex-specific ATMRs, we conducted a series of multilevel regression models, with male and female ATMRs nested within countries using a random intercept for country and fixed slopes for all predictors. Model 1 included national legal BAC limits as the primary predictor. Model 2 added sex and a BAC limit × sex interaction term. Model 3 further incorporated additional country-level covariates including, income classification, healthcare system infrastructure, gender inequality, and per capita alcohol consumption. All covariates were entered as fixed effects. The dependent variable, ATMR, was log-transformed using ln(ATMR + 0.1) to account for non-normality; corresponding analyses using untransformed values are presented in Supplemental Table S2 and yielded consistent results. Exploratory models testing three-way interactions between sex, national BAC legal limits, and each additional country-level predictor are also reported in Supplemental Table S3 (untransformed data) and Supplemental Table S4 (transformed data). Extreme outliers, defined as values greater than 3 times the interquartile range, were removed using the 3 x IQR method.¹³ Analyses and figure generation were conducted using SPSS (v.28) and JMP Pro (v.17), with statistical significance set at p < 0.05.

Role of the funding source

The funding sources reported in this study were not involved in study design, data collection, statistical analyses, interpretation of the data, or writing of this publication.

Results

Table 1 summarizes the descriptive statistics for the study variables among the 165 countries included in the analysis. Male ATMRs were notably higher than female ATMRs across countries, with mean rates approximately five times greater for males. Legal BAC limits ranged from 0.00 g/dL (zero tolerance) to 0.15 g/dL, reflecting considerable variation in national drink-driving policies. Country-level covariates, including healthcare system infrastructure, gender inequality, and per-capita alcohol consumption, also differed across countries. These descriptives provide context for the multilevel analyses presented below.

Table 1.

Mean rates of alcohol-attributable traffic mortality rates (ATMRs; per 100,000), policy, and health metrics across countries.

	Mean [95% CI]	N (observations)
Male ATMR	9.77 [8.40, 11.14]	164
Female ATMR	2.05 [1.75, 2.36]	164
Legal BAC	0.05 [0.04, 0.05]	330
Gender inequality index	0.33 [0.31, 0.36]	308
Per capital alcohol intake	5.61 [5.18, 6.04]	330
Healthcare index	31.57 [29.61, 33.53]	330
Low income group	–	42
Lower-middle income group	–	90
Upper-middle income group	–	86
High income group	–	110

Note: Values represent mean estimates with 95% confidence intervals in brackets. ATMRs are age-adjusted, expressed per 100,000 population, and reported separately for males and females. All other variables are country-level measures and not age-adjusted. Sample sizes vary due to data availability across predictors. Descriptive statistics are presented using untransformed ATMR data.

In the initial mixed-effects model (Model 1; Table 2), higher legal BAC limits were significantly associated with increased ATMRs (β = 19.39, 95% CI: [14.75, 24.03]). Although the association was positive overall, considerable cross-national variability was evident (Fig. 2). The model explained 20% of the variance in mortality (R² = 0.20), suggesting that BAC limits alone do not fully account for cross-national differences in ATMRs.

Table 2.

Comparison of mixed-effects models for log-transformed alcohol-attributable traffic mortality rates (ATMRs).

	Model 1 Mixed model (ATMR ∼ Legal BAC)	Model 2 Mixed model (ATMR ∼ Legal BAC x Sex)	Model 3 Mixed model (ATMR ∼ Legal BAC x Sex + covariates)
Intercept	β = 0.15, SE = 0.13 95% CI: [−0.10, 0.41]	β = 0.83∗∗∗, SE = 0.13 95% CI: [0.57, 1.09]	β = 0.70, SE = 0.47 95% CI: [−0.23, 1.62]
Legal BAC	β = 19.39∗∗∗, SE = 2.35 95% CI: [14.75, 24.03]	β = 20.85∗∗∗, SE = 2.40 95% CI: [16.12, 25.59]	β = 13.34∗∗∗, SE = 2.00 95% CI: [9.39, 17.29]
Sex (Ref: Male)	–	β = −1.34∗∗∗, SE = 0.05 95% CI: [−1.43, −1.24]	β = −1.35∗∗∗, SE = 0.05 95% CI: [−1.46, −1.25]
Income Level	–	–	β = −0.31∗∗, SE = 0.10 95% CI: [-0.50, −0.11]
Healthcare Index	–	–	β = −0.00, SE = 0.00 95% CI: [−0.01, 0.00]
Gender Inequality	–	–	β = 1.61∗∗, SE = 0.56 95% CI: [0.51, 2.70]
Per Capita Intake	–	–	β = 0.16∗∗∗, SE = 0.02 95% CI: [0.13, 0.20]
Legal BAC x Sex	–	β = −2.92∗∗, SE = 0.89 95% CI: [−4.68, −1.16]	β = −2.58∗∗, SE = 0.95 95% CI: [−4.45, −0.70]
Marginal R2	0.20	0.51	0.71
N (Total Observations)	328	328	304

Note: N reflects sex-disaggregated observations across 165 countries. One outlier per sex was removed. ATMRs were log-transformed using ln (ATMR + 0.1) prior to analysis. Sample size varies in Model 3 due to missing data from covariates. Significance levels are indicated as follows: p < 0.001∗∗∗ |p < 0.01∗∗ |p < 0.05∗.

Fig. 2.

Legal BAC limit and alcohol-attributable traffic mortality rates (ATMRs). Note: Shaded region around the line represents the 95% confidence interval. Values shown are untransformed and expressed per 100,000 population for interpretability.

The inclusion of sex in the second model (Model 2; Table 2) revealed that at a BAC limit of 0.00 g/dL, females had significantly lower mortality rates than males (β = −1.34, 95% CI: [−1.43, −1.24]). Moreover, a significant interaction between sex and BAC limit (β = −2.92, 95% CI: [−4.68, −1.16]) indicated that the increase in mortality associated with higher BAC thresholds was disproportionately greater among males. As shown in Fig. 3, the relationship between legal BAC limits and ATMRs was steeper among males. The inclusion of sex and its interaction with BAC improved the explanatory power of the model, increasing the variance explained to 51% (R² = 0.51).

Fig. 3.

Sex Differences in alcohol-attributable traffic mortality rates (ATMRs). Note: Shaded region around the line represents the 95% confidence interval. Values shown are untransformed and expressed per 100,000 population for interpretability.

Building on Model 2, the third model (Model 3; Table 2) included the additional country-level covariates of income level, healthcare system capacity, gender inequality and per capita alcohol consumption, to account for structural differences across countries. The interaction between legal BAC limit and sex remained significant (β = −2.58, 95% CI: [−4.45, −0.70]), indicating that the association between BAC limits and ATMRs continued to differ by sex even after adjusting for the covariates. Lower national income (β = −0.31, 95% CI: [−0.50, −0.11]), higher gender inequality (β = 1.61, 95% CI: [0.51, 2.70]), and greater alcohol consumption (β = 0.16, 95% CI: [0.13, 0.20]) were associated with increased mortality. No significant main effect of healthcare system infrastructure was observed (p = 0.36); however, a significant three-way interaction with legal BAC and sex emerged (Fig. 4); these results are detailed in the Supplementary Materials (Model S2, Supplemental Table S4). Model 3 explained 71% of the variance in ATMRs (R² = 0.71).

Fig. 4.

Legal BAC limit and sex predict alcohol-attributable traffic mortality rates (ATMRs) by Healthcare Infrastructure. Note: Graph displays a median split on the Healthcare System Index. Higher index values reflect stronger healthcare infrastructure. Shaded region around the line represents the 95% confidence interval. Values shown are untransformed and expressed per 100,000 population for interpretability.

Additionally, we also examined three-way interactions between legal BAC limits, sex, and the remaining country-level covariates including income group, gender inequality, and per capita alcohol intake. No significant three-way interactions emerged (all p > 0.05). However, significant two-way interactions were observed between legal BAC limits and each covariate. The protective association of stricter BAC policies was amplified in countries with a lower national income (β = −5.05, 95% CI: [−8.43, −1.67]), higher gender inequality (β = 39.10, 95% CI: [19.34, 58.87]), and lower per-capita alcohol consumption (β = −1.99, 95% CI: [−2.96, −1.02]). For a complete summary of all two-way and three-way interactions, see Supplemental Table S4 and Supplemental Figures S1–S3.

Discussion

In this cross-national study, we examined the relationship between national legal BAC limits and ATMRs across 165 countries. Our results support global evidence indicating that stricter BAC thresholds are associated with lower ATMRs. This pattern persisted across sexes, although the association was significantly stronger among males. In addition to legal thresholds, we found that several broader country-level factors, such as lower national income, greater gender inequality, and higher alcohol consumption, were each associated with higher mortality rates. While healthcare system infrastructure was not directly related to ATMRs, it moderated the BAC-sex relationship, indicating that a stronger healthcare system may reduce sex disparities. Controlling for these structural factors improved model performance by reducing the unexplained variance and provided a better understanding of the contexts in which ATMRs are highest. Together, these findings highlight the need to consider BAC limits within the broader sociopolitical contexts in which they operate.

In our initial model, higher national legal BAC limits were associated with increased ATMRs. However, when sex and its interaction with BAC were included, a clearer pattern emerged; although the association between BAC limits and mortality was evident in both sexes, the effect was significantly greater among males. This suggests that under more permissive BAC policies, males bear a disproportionate share of the mortality burden. This finding is consistent with evidence indicating that men account for approximately three-quarters of road traffic deaths worldwide,² reflecting both higher driving exposure and greater engagement in high-risk behaviors, including hazardous drinking and alcohol-impaired driving.^7,14 When BAC laws are more lenient, men, who are more likely to drive under the influence, are disproportionately affected, resulting in sharper mortality increases.¹⁵

Although legal thresholds apply equally across sexes, females generally reach these limits with fewer drinks due to well established biological differences in alcohol metabolism.^7,3 Some evidence also suggests that females may experience greater cognitive and psychomotor impairment at equivalent BAC levels,^14,16 although findings are mixed with respect to driving-related performance.¹⁷ While our results did not show a steeper increase in mortality among females, these physiological differences remain important for prevention. Public education efforts should emphasize that legal BAC limits represent a regulatory threshold rather than a physiological safety level, and that measurable impairment can occur at lower concentrations, particularly among female drivers. Recent global patterns in alcohol use also show rising rates of risky drinking behaviors among women, narrowing the historical gender gap in alcohol-related harm.⁷ Recognizing these shifts is paramount for ensuring that BAC policies remain effective in addressing evolving patterns of alcohol use and sex-based risk. However, our findings indicate that national BAC standards can be applied consistently across sexes.

Building on the observed sex differences, we next examined how broader national conditions, including economic, infrastructural, social, and behavioral factors, influenced ATMRs. Several of these country-level factors were significantly associated with ATMRs. Specifically, lower-income countries generally exhibited substantially higher mortality rates, consistent with evidence indicating that over 90% of road traffic deaths occur in LMICs despite these regions having fewer vehicles per capita.¹ These disparities likely reflect structural challenges, including under-resourced enforcement systems, inadequate roadside testing equipment, and limited access to emergency medical care, all contributing to higher ATMRs.⁹ As noted in the Supplementary Materials (Supplemental Figure S1), the association between legal BAC limits and mortality was also steeper in these low- and lower-middle income countries. By contrast, high-income countries experienced lower mortality and flatter associations, suggesting that countries with greater financial resources may offer safer road conditions and more access to effective post-crash treatment.

However, economic conditions alone do not fully explain variation in outcomes. Healthcare system infrastructure also shaped patterns of ATMRs. Countries with stronger healthcare systems, characterized by greater emergency care capacity, trauma coordination, and access to medical services, showed smaller sex disparities and flatter BAC-ATMR slopes, as reflected in the significant three-way interaction with legal BAC limits and sex (Fig. 4). This pattern likely reflects not only better emergency response and trauma care systems, but also stronger public health system infrastructure capable of supporting education, prevention and outreach around alcohol-related risks.¹⁸ Conversely, countries with weaker healthcare systems often face limited access to emergency medical services, delayed transport, and inadequate trauma care, increasing the likelihood that alcohol-related traffic injuries result in death. Together, these patterns suggest that the effectiveness of BAC legislation may depend on the broader system context in which it is implemented.

This broader system-level approach has shown real-world impact. In Lithuania, for example, national reforms combining higher taxes, restricted sales hours, advertising bans, and expanded public health programming were linked to significant reductions in alcohol-attributable traffic injuries and deaths between 2004 and 2019.¹⁹ These coordinated policies not only strengthen enforcement and healthcare response capacity but can also help shift social norms around drinking and driving. Lower legal BAC thresholds may help cultivate social norms that impaired driving is socially unacceptable, reinforcing preventive behaviors such as using public transportation, designated drivers, or rideshare services. Taken together, these findings suggest that the effectiveness of BAC limits depends not only on legislation, but also on the broader health, enforcement, and social systems that support their implementation and impact.

Beyond healthcare systems, social inequities, such as gender inequality, contribute to determining who is most at risk of elevated ATMRs. Specifically, countries with greater gender inequality experienced a stronger protective effect of BAC policies on ATMRs (Supplemental Figure S2). However, these differences in gender inequality likely reflect cultural and structural constraints that limit women's mobility, such as restricted access to vehicles and alcohol, and gendered norms that reduce women's driving exposure, which may appear protective but instead reflect underlying inequities.^20,21 Gender inequality is also most prominent in many LMICs,²² compounding other challenges such as weak enforcement and under-resourced medical care. Together, these factors reinforce the importance of evaluating legal BAC thresholds within the broader sociopolitical environments in which they are enforced.

National alcohol consumption levels were also strongly associated with ATMRs, with higher per-capita intake predicting higher mortality rates. However, countries with lower consumption levels exhibited a more pronounced protective association between stricter BAC limits and ATMRs (Supplemental Figure S3). This finding indicated that low average consumption does not reliably reflect reduced alcohol-related traffic risk. Consistent with the alcohol-harm paradox, infrequent but high-intensity drinking can yield disproportionate harm even in low-intake settings.²³ Moreover, alcohol-related risk in these settings is often concentrated within a smaller population of individuals who engage in high-risk drinking rather than distributed across the population.²³ In line with this pattern, our findings also demonstrated that males in low-intake countries had elevated ATMRs, mirroring global evidence that alcohol-related injury and mortality disproportionately affect men.²⁴ Taken together, these observations suggest that average national consumption levels alone may be insufficient markers of traffic safety risk and that the effectiveness of BAC laws is shaped by the distribution and severity of drinking behaviors within populations.

Policy measures that address these systemic conditions can enhance the effectiveness of BAC legislation. For example, Scotland introduced a minimum unit pricing policy in 2018, setting a floor of £0.50 per unit of alcohol. The policy aimed to reduce access to the cheapest and most harmful forms of alcohol. Over 32 months, alcohol-attributable deaths fell by 13.4%, with the largest reductions among men and those living in the most socioeconomically deprived areas.²⁵ This example highlights the potential for pricing policies to reduce alcohol-related harms, particularly when paired with broader efforts to address structural drivers of excessive consumption.

While these findings offer important policy implications, several limitations must also be addressed. First, although this study aimed to include all 193 countries, data were incomplete or inconsistent for several countries, limiting full global representation of all UN member states. Additionally, mortality estimates are based on the most recent available data from 2019 and may not reflect the impact of more recent policy reforms. Moreover, these estimates were not disaggregated by road user type, limiting the ability to quantify how alcohol-attributable traffic fatalities were distributed across drivers, passengers, and bystanders. Data representing urbanization levels across countries were also not available, though urbanization may influence exposure risk, enforcement and access to emergency services. Similarly, covariates were measured at the national level and may not reflect subnational variation, particularly in decentralized systems, though nationally standardized indicators were necessary for cross-national comparability. Data on alcohol dependence were often also unavailable, particularly in low- and middle-income countries, and their inclusion would have likely introduced bias from inconsistent definitions and reporting years. Lastly, our analysis of ATMRs was based on sex-disaggregated data and reflects differences by sex assigned at birth, not gender identity. While we were limited to sex-based data, we included the Gender Inequality Index as a country-level proxy to acknowledge broader societal gender disparities.

In summary, this study provides valuable policy-relevant evidence regarding the effectiveness of national legal BAC limits in reducing ATMRs. Using the most recent comparable global data, we determined that 51 of the 165 countries examined maintained national BAC limits exceeding the WHO-recommended maximum of 0.05 g/dL for driving under the influence.2 Across countries, less permissive BAC laws were significantly associated with lower ATMRs. However, substantial variation in ATMRs was observed even among countries with highly restrictive or zero-tolerance BAC policies. While lower legal thresholds reduce fatalities, further reduction in ATMRs depend on the broader structural and social contexts in which it is implemented. In countries with stronger healthcare systems, BAC legislation was more effective in lowering ATMRs, particularly among males. Similarly, less permissive BAC laws were more effective in reducing ATMRs in LMICs, as well as countries with higher gender inequality and lower per capita alcohol intake. Across all countries, males consistently faced a higher burden of ATMRs, with mortality increasing as BAC limits became more permissive. To maximize impact, legal BAC limits should be implemented alongside infrastructural investments in systems that enable reliable enforcement, ensure access to timely emergency medical care, reduce higher per capita alcohol intake, and deliver targeted prevention strategies. Without these foundational systems in place, even the most well-designed BAC policies are unlikely to achieve maximal impact.

Contributors

CLJC, MRP, TLV, YZ, RK, BAB, NRR, and SAM conceptualized the study. CLJC, MRP, AAM, ABM contributed to data curation. CLJC cleaned and analyzed the dataset with statistical guidance from BP and SAM. CLJC and MRP accessed and verified the data. SAM acquired funding. CLJC drafted the manuscript, and all authors reviewed and edited.

Data sharing statement

The data analyzed in this study are available from the corresponding author upon reasonable request.

Editor note

The Lancet Group takes a neutral position with respect to territorial claims in published maps and institutional affiliations.

Declaration of interests

All authors have submitted the ICMJE COI forms. CLJC, RK, SAM, TLV, YZ, and RK have received grants or contracts from the National Institute on Alcohol Abuse and Alcoholism. TLV has received support from the Brain & Behavior Research Foundation Young Investigator Grant. AAM and MRP have received grants or contracts from the Office of the Assistant Secretary of Defense for Health Affairs through the Alcohol and Substance Use Research Program. SAM has received payment from role associate editor in an Elsevier journal. BAB has received payment regarding consulting fees from Novo Nordisk and Boehringer Ingelheim. Authors (ABM, BP, NRR, VGR) declare no further conflicts of interest.