Background: Recreational cannabis legalization has expanded rapidly across US states. The regulatory approaches states adopt vary widely, with varying implications for public health. This study aimed to characterize heterogeneity in recreational cannabis laws (RCLs) across US states and to identify state–level characteristics associated with these regulatory models. Methods: We conducted Latent Class Analysis (LCA) of state–year RCL provisions from 2013 to 2024 (n=612) to identify distinct RCL approaches. Descriptive analyses and exploratory multinomial regression analyses were used to examine correlations between state characteristics and RCL approaches from 2020 to 2024, when sufficient cross–state variation in RCL adoption was available. Eleven recreational cannabis policy provisions spanning governance, potency limits, consumption restrictions, access controls, taxation, marketing regulations, and driving prohibitions are primarily from the Alcohol Policy Information System. State–level characteristics included cannabis use prevalence, market conditions, medical cannabis history, political factors, demographic, and socioeconomic covariates obtained from multiple secondary data sources. Results: We identified four latent classes of state–year RCL provisions representing different regulatory approaches: No RCL, Pre–commercial, Full Access, and Dispensary Access. The No RCL corresponded to state–years without RCL. The Pre–commercial class represented state–years in early–stage legalization with a minimal regulated approach in terms of commercial infrastructure. The Full Access class was characterized by permitting on–site retail consumption and home delivery and restricting (but not prohibiting) public use. In contrast, the Dispensary Access class limited retail sales to off–site consumption only, prohibited public use, and imposed stricter market controls. Higher past–month cannabis use prevalence was associated with a greater likelihood of membership in the Full Access class (RRR = 1.78; 95% CI: 1.21–2.62), relative to No RCL. A longer duration since medical cannabis legalization was associated with a higher likelihood of membership in the Dispensary access class (RRR = 1.47; 95% CI: 1.02–2.12). Higher beer excise taxes were associated with a lower likelihood of membership in any RCL class relative to No RCL. Conclusions: From 2013 to 2024, US recreational cannabis regulations clustered into four distinct regulatory approaches, with two distinct commercial models: one permitting on–site retail consumption and home delivery, the other restricting sales to off–premises only and prohibiting public use. Higher cannabis use prevalence and longer medical cannabis history were associated with more access–oriented and more restrictive commercial approaches, respectively. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement Dr. Ellicott C Matthay was supported in part by the National Institute for Alcohol Abuse and Alcoholism (grant number R00AA028256) ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All data produced in the present work are contained in the manuscript

Socioenvironmental disasters such as wildfires, floods, extreme heat events, and zoonotic epidemics have been increasing in magnitude and frequency. These events are sometimes are followed by increases in overdose mortality. This paper applies an adaptation of the Big Events framework, originally developed to understand macrosocial changes and their effects on HIV, to guide research into the pathways through which socioenvironmental disasters may cause increases in overdose mortality. These pathways include economic disruption, health and social service disruption, population displacement, changes in social networks, and social movements. They also include processes and conditions more causally proximal to overdose events such as housing insecurity and changes in drug markets and drug-use behaviors. The paper also discusses relevant existing measures and variables that must be developed to conduct research on the effects of socioenvironmental disasters on overdose The end goal of this application of the Big Events framework is to encourage scholarship that identifies potential interventions for people who use drugs, public and private service agencies and governments to mitigate the harms that socioenvironmental disasters can cause for people who use drugs.

Given intensifying political polarization and sweeping legal actions targeting marginalized populations across the U.S., there are increasing calls to examine how these policy changes are shaping population health and health inequities. To meet these calls, researchers have developed and applied various approaches to support rigorous, theoretically-informed quantitative research on policies and health. To date, much of the literature has examined single policies; however, there is growing interest in examining alternative ways of conceptualizing policy exposures, namely as policy clusters or policy climates, to better capture how policies are enacted (and experienced) in "real-world" contexts. To advance this work, greater clarity is needed regarding how different approaches to policy conceptualization, measurement, and analysis align with distinct research questions and goals, ranging from identifying specific, manipulable policy levers to informing ways of extending the "policy space" beyond already existing laws to support broader social change. In this essay, we help fill this gap by outlining key issues related to policy exposures, including conceptualization, methods for measure development, and analytic approaches for understanding the relationship between policies and health. We end by discussing future directions for population health research on single policies, policy clusters, and policy climates, with an eye towards advancing health equity.

BACKGROUND:Cannabis use and alcohol use are associated with self-harm injuries, but little research has assessed links between recreational cannabis outlet openings on rates of self-harm within communities or the interactions of cannabis outlets with the density of alcohol outlets. We estimated the associations of recreational cannabis outlets, alcohol outlets, and their interaction on rates of fatal and nonfatal self-harm injuries in California, 2017-2019. METHODS:Using California statewide data on recreational cannabis outlets, alcohol outlets, and hospital discharges and deaths due to self-harm injuries, we conducted Bayesian spatiotemporal analyses of quarterly ZIP code-level data over 3 years, accounting for confounders and spatial autocorrelation. Using the model posteriors, we estimated parameters corresponding to hypothetical shifts in outlet densities. RESULTS:If recreational cannabis outlets had never opened, we estimated that nonfatal self-harm injuries would have been -0.35 per 100,000 lower (95% credible interval [CI]: -1.25, 0.51), while fatal self-harm injuries would have been -0.004 per 100,000 lower (95% CI: -0.26, 0.25). These associations did not depend on alcohol outlet density, but a hypothetical 20% reduction in alcohol outlet densities was associated with fewer self-harm injuries (risk difference per 100,000, nonfatal: -1.59; 95% CI: -2.60, -0.59; fatal: -0.10; 95% CI: -0.37, 0.16). Associations for nonfatal incidents were strongest for people aged 15-34 years, and White and Hispanic people. CONCLUSION:We did not find evidence that the introduction of recreational cannabis outlets was associated with self-harm injuries or that cannabis and alcohol outlet densities interact, but alcohol outlet density had a strong association with nonfatal self-harm injuries.

Growing evidence suggests exposure to high temperatures may result in increased urban crime, a known driver of health and health inequity. Theoretical explanations have been developed to describe the heat-crime relationship without consensus yet achieved among experts. This scoping review aims to summarize evidence of heat-crime associations in U.S. cities. Further examination of empirical and translational inconsistencies in this literature will ensure future studies of urban heat-crime relationships in the U.S., and their policy impacts are informed by a thorough understanding of existing evidence. We performed a comprehensive literature search of empirical studies on heat-crime relationships in U.S. cities published between January 2000 and August 2023. The included studies were qualitatively synthesized based on operationalized exposures, outcomes, covariates, methodologies, theoretical framing, and policy implications. In total, 46 studies were included in this review. Most studies (93

INTRODUCTION:In 2016, California legalized cannabis for adult recreational use; after recreational sales began in 2018, it became the largest retail market worldwide. This study profiled specific risks and prevention opportunities across age groups and examined changes in medically significant child cannabis exposures before and after legalization. METHODS:Researchers conducted analyses, including interrupted time series, to examine 1,695 California Poison Control System reports of cannabis exposure in children aged 0-17 years from 2010 to 2020. Analyses were confined to moderate and severe exposures, identified by California Poison Control System toxicologists as requiring medical attention. RESULTS:Monthly rates of moderate/severe cannabis exposure per million children increased after legalization (β=0.06; 95% CI=0.05, 0.08), especially in children aged <5 years. Fourteen percent required critical care admission. Exposures were primarily unintentional in younger children (87.7%-99.2%) and intentional in adolescents (85.5%). Across all ages, most exposures occurred in the home (94.0%) and involved edible products (83.5%). An analysis of packaging on edible brands frequently cited in health records found that most could be easily mistaken for popular candies and snack foods. CONCLUSIONS:After cannabis legalization in California, reports of child cannabis exposures requiring medical attention increased significantly. Most reported cannabis exposures occurred in the home through the ingestion of edible products, often packaged to look like popular candy and snack food brands. To prevent these harms, cannabis legalization should be accompanied by robust marketing and packaging regulations (e.g., plain labels, larger warning labels). Secondary prevention should focus on educating parents and caregivers on safe cannabis storage in the home.

Policy Points We can leverage data science and artificial intelligence to inform state and local resource allocation for overdose prevention. Data science and artificial intelligence can help us answer four questions: (1) What is the impact of laws on access to interventions and overdose risk? (2) Where should interventions be targeted? (3) Which types of demographic subgroups benefit the most and the least from interventions? and (4) Which types of interventions should they invest in for each setting and population? Advances in data science and artificial intelligence can accelerate the pace at which we can answer these critical questions and help inform an effective overdose prevention response.

Between May 2020 and December 2021, there were 159,872 drug overdose deaths in the US. Higher eviction rates have been associated with higher overdose mortality. Amid the economic turmoil caused by the COVID-19 pandemic, 43 states and Washington, DC, implemented eviction moratoria of varying durations. These moratoria reduced eviction filing rates, but their impact on fatal drug overdoses remains unexplored. We evaluated the effect of these policies on county-level overdose death rates by focusing on the dates the state eviction moratoria were lifted. We obtained mortality data from NCHS and eviction moratoria dates from the COVID-19 US State Policy Database. We employed a longitudinal targeted minimum-loss-based estimation with Super Learner to flexibly estimate the average treatment effect (ATE) of never lifting the moratoria. Lifting state eviction moratoria was associated with a 0.14 per 100,000 higher rate of monthly overdose mortality (95%CI: -0.03, 0.32), although confidence intervals were wide and included zero. Eviction moratoria may not be sufficient to prevent overdose mortality during crises such as the COVID-19 pandemic.

In 1992, Wacholder et al. developed a theoretical framework for case-control studies to minimize bias in control selection. They described 3 comparability principles (study base, deconfounding, and comparable accuracy) to reduce the potential for selection bias, confounding, and information bias in case-control studies. Wacholder et al. explained how these principles apply to traditional sources of control participants for case-control studies, including population controls, hospital controls, controls from a medical practice, friend or relative controls, and deceased controls. The goal of the present article is to extend this seminal work on case-control studies by providing a modern perspective on sources of control participants. Today, there are many more potential sources of control participants s for case-control studies than there were in the 1990s. This is due to technological advances in computing power, internet access, and availability of "big data" resources. These advances have vastly expanded the quantity and diversity of data available for case-control studies. We discuss control selection from electronic health records, health insurance claims databases, publicly available online data sources, and social media-based data. We focus on practical considerations for unbiased control selection, emphasizing the strengths and weaknesses of each modern source of controls for case-control studies.

Recent advances in Artificial Intelligence (AI) present new and not widely recognized opportunities to advance the rigor, scope, efficiency, and impact of epidemiologic research aiming to make causal inferences or causal decisions. We describe recent developments, challenges, and examples for integrating varied AI tools into the steps of Petersen and van der Laan’s causal inference roadmap and causal decision-making tasks. AI tools relevant to causal research in epidemiology include predictive models, unsupervised learning, causal structure learning, causal estimation, and generative models. Opportunities exist to integrate AI at each stage of the causal roadmap. This includes the use of generative models to synthesize scientific literature and identify knowledge gaps; causal structure learning to discover or hypothesize causal structures from data; unsupervised learning from unstructured text to generate quantitative variables for analysis; predictive models to drive clinical or policy interventions; generative or causal models to assess or establish identifiability; causal models for estimating statistical parameters; and generative models to create text, tables, and figures to interpret and disseminate findings. Researchers must be mindful of potential pitfalls of AI tools such as insufficient training data, poor accuracy, biases, and ethical and legal concerns. Diverse AI tools are available to support causal research in epidemiology. Steps of the causal inference roadmap cannot yet be fully automated, but thoughtful “collaboration” between investigators and AI tools may accelerate or deepen the research at each step.

Recreational cannabis outlets may influence rates of interpersonal violence, but research has yielded inconsistent findings. Modification by alcohol outlet density may help explain inconsistencies. We estimated the impacts of recreational cannabis outlets on neighborhood-level assault injury rates in California and evaluated whether alcohol outlet density moderated these associations. We applied Bayesian spatiotemporal analyses to ZIP code-level statewide data on alcohol outlets, recreational cannabis outlets, and injuries and deaths due to firearm and nonfirearm assault, from 2017 to 2019, accounting for confounders and spatial autocorrelation. Using the model posteriors, we estimated parameters corresponding to hypothetical shifts in outlet densities, overall and by age, sex, and race/ethnicity. If recreational cannabis outlets were never introduced, we estimated that nonfirearm assault injuries would have been 1.63 per 100 000 lower (95% CI, -3.08 to 0.01), but we observed no association with firearm assault injuries (risk difference [RD] per 100 000: -0.07; 95% CI, -0.34 to 0.21). These associations did not depend on alcohol outlet density, but a hypothetical 20% reduction in alcohol outlet densities was associated with fewer firearm (RD per 100 000: -1.89; 95% CI, -0.46 to 0.09) and nonfirearm (RD per 100 000: -5.67; 95% CI, -7.44 to -3.95) assault injuries. The introduction of recreational cannabis outlets may have contributed to a small increase in nonfirearm assault injuries.

This cross-sectional study describes self-harm rates from 2005 to 2021 among US youth by age group, sex, and race and ethnicity.

Research into clinical interventions rarely translates to improved mental health at the population level. Adequately powered studies leveraging advances in statistical methods to assess and translate multicomponent interventions will be better positioned to yield improvements in population mental health.

Despite well-studied associations of state firearm laws with lower state- and county-level firearm homicide, there is a shortage of studies investigating differences in the effects of distinct state firearm law categories on various cities within the same state using identical methods. We examined associations of 5 categories of state firearm laws—pertaining to buyers, dealers, domestic violence, gun type/trafficking, and possession—with city-level firearm homicide, and then tested differential associations by city characteristics. City-level panel data on firearm homicide cases of 78 major cities from 2010 to 2020 was assessed from the Centers for Disease Control and Prevention’s National Vital Statistics System. We modeled log-transformed firearm homicide rates as a function of firearm law scores, city, state, and year fixed effects, along with time-varying city-level confounders. We considered effect measure modification by poverty, unemployment, vacant housing, and income inequality. A one z-score increase in state gun type/trafficking, possession, and dealer law scores was associated with 25

The target trial framework is a well-known tool for estimating causal effects from observational data. The target trial approach can be used with data from any type of observational study, but it has most often been used to emulate a hypothetical target trial using data from a prospective cohort study. In this manuscript, we present the target cohort framework for estimating causal effects from case-control studies. The target cohort approach extends the existing target trial framework for estimating causal effects using observational data but has an explicit case-control perspective. There are clear conceptual links from randomized trials to cohort studies and from cohort studies to case control studies. The target cohort framework uses a nested case control study to emulate a cohort study that has been designed to emulate a hypothetical pragmatic randomized controlled trial. Both target trial and target cohort frameworks require clear specification of eligibility criteria, treatment strategies, treatment assignment (randomization), follow-up period, outcomes of interest, causal contrast, and analysis plan. We demonstrate the target cohort approach using an example of an observational study to estimate the causal effect of semaglutide, a type of GLP-1 medication sold under the brand name Ozempic, on adverse gastrointestinal events.

Observational data provide invaluable real-world information in medicine, but certain methodological considerations are required to derive causal estimates. In this systematic review, we evaluated the methodology and reporting quality of individual-level patient data meta-analyses (IPD-MAs) conducted with non-randomized exposures, published in 2009, 2014, and 2019 that sought to estimate a causal relationship in medicine. We screened over 16,000 titles and abstracts, reviewed 45 full-text articles out of the 167 deemed potentially eligible, and included 29 into the analysis. Unfortunately, we found that causal methodologies were rarely implemented, and reporting was generally poor across studies. Specifically, only three of the 29 articles used quasi-experimental methods, and no study used G-methods to adjust for time-varying confounding. To address these issues, we propose stronger collaborations between physicians and methodologists to ensure that causal methodologies are properly implemented in IPD-MAs. In addition, we put forward a suggested checklist of reporting guidelines for IPD-MAs that utilize causal methods. This checklist could improve reporting thereby potentially enhancing the quality and trustworthiness of IPD-MAs, which can be considered one of the most valuable sources of evidence for health policy.

BACKGROUND:Cannabis exposures reported to the California Poison Control System increased following the initiation of recreational cannabis sales on 1 January 2018 (i.e., "commercialization"). We evaluated whether local cannabis control policies adopted by 2021 were associated with shifts in harmful cannabis exposures. METHODS:Using cannabis control policies collected for all 539 California cities and counties in 2020-2021, we applied a differences-in-differences design with negative binomial regression to test the association of policies with harmful cannabis exposures reported to California Poison Control System (2011-2020), before and after commercialization. We considered three policy categories: bans on storefront recreational retail cannabis businesses, overall restrictiveness, and specific recommended provisions (restricting product types or potency, packaging and labeling restrictions, and server training requirements). RESULTS:Localities that ultimately banned storefront recreational retail cannabis businesses had fewer harmful cannabis exposures for children aged <13 years (rate ratio = 0.82; 95% confidence interval = 0.65, 1.02), but not for people aged >13 years (rate ratio = 0.97; 95% confidence interval = 0.85, 1.11). Of 167 localities ultimately permitting recreational cannabis sales, overall restrictiveness was not associated with harmful cannabis exposures among children aged <13 years, but for people aged >13 years, a 1-standard deviation increase in ultimate restrictiveness was associated with fewer harmful cannabis exposures (rate ratio = 0.93; 95% confidence interval = 0.86, 1.01). For recommended provisions, estimates were generally too imprecise to detect associations with harmful cannabis exposures. CONCLUSION:Bans on storefront retail and other restrictive approaches to regulating recreational cannabis may be associated with fewer harmful cannabis exposures for some age groups following statewide commercialization.

ImportancePsychosis is a hypothesized consequence of cannabis use. Legalization of cannabis could therefore be associated with an increase in rates of health care utilization for psychosis.ObjectiveTo evaluate the association of state medical and recreational cannabis laws and commercialization with rates of psychosis-related health care utilization.Design, Setting, and ParticipantsRetrospective cohort design using state-level panel fixed effects to model within-state changes in monthly rates of psychosis-related health care claims as a function of state cannabis policy level, adjusting for time-varying state-level characteristics and state, year, and month fixed effects. Commercial and Medicare Advantage claims data for beneficiaries aged 16 years and older in all 50 US states and the District of Columbia, 2003 to 2017 were used. Data were analyzed from April 2021 to October 2022.ExposureState cannabis legalization policies were measured for each state and month based on law type (medical or recreational) and degree of commercialization (presence or absence of retail outlets).Main Outcomes and MeasuresOutcomes were rates of psychosis-related diagnoses and prescribed antipsychotics.ResultsThis study included 63 680 589 beneficiaries followed for 2 015 189 706 person-months. Women accounted for 51.8% of follow-up time with the majority of person-months recorded for those aged 65 years and older (77.3%) and among White beneficiaries (64.6%). Results from fully-adjusted models showed that, compared with no legalization policy, states with legalization policies experienced no statistically significant increase in rates of psychosis-related diagnoses (medical, no retail outlets: rate ratio [RR], 1.13; 95% CI, 0.97-1.36; medical, retail outlets: RR, 1.24; 95% CI, 0.96-1.61; recreational, no retail outlets: RR, 1.38; 95% CI, 0.93-2.04; recreational, retail outlets: RR, 1.39; 95% CI, 0.98-1.97) or prescribed antipsychotics (medical, no retail outlets RR, 1.00; 95% CI, 0.88-1.13; medical, retail outlets: RR, 1.01; 95% CI, 0.87-1.19; recreational, no retail outlets: RR, 1.13; 95% CI, 0.84-1.51; recreational, retail outlets: RR, 1.14; 95% CI, 0.89-1.45). In exploratory secondary analyses, rates of psychosis-related diagnoses increased significantly among men, people aged 55 to 64 years, and Asian beneficiaries in states with recreational policies compared with no policy.Conclusions and RelevanceIn this retrospective cohort study of commercial and Medicare Advantage claims data, state medical and recreational cannabis policies were not associated with a statistically significant increase in rates of psychosis-related health outcomes. As states continue to introduce new cannabis policies, continued evaluation of psychosis as a potential consequence of state cannabis legalization may be informative.

Observational data provide invaluable real-world information in medicine, but certain methodological considerations are required to derive causal estimates. In this systematic review, we evaluated the methodology and reporting quality of individual-level patient data meta-analyses (IPD-MAs) published in 2009, 2014, and 2019 that sought to estimate a causal relationship in medicine. We screened over 16,000 titles and abstracts, reviewed 45 full-text articles out of the 167 deemed potentially eligible, and included 29 into the analysis. Unfortunately, we found that causal methodologies were rarely implemented, and reporting was generally poor across studies. Specifically, only three of the 29 articles used quasi-experimental methods, and no study used G-methods to adjust for time-varying confounding. To address these issues, we propose stronger collaborations between physicians and methodologists to ensure that causal methodologies are properly implemented in IPD-MAs. In addition, we put forward a suggested checklist of reporting guidelines for IPD-MAs that utilize causal methods. This checklist could improve reporting thereby potentially enhancing the quality and trustworthiness of IPD-MAs, which can be considered one of the most valuable sources of evidence for health policy.

Objective: Growing availability of cannabis products through home delivery services may affect cannabis-related health outcomes. However, research is impeded by a lack of data measuring its scale. Prior research demonstrated that crowdsourced websites can be used to validly enumerate brick-and-mortar cannabis outlets. We piloted an extension of this method to explore the feasibility of measuring availability of cannabis home delivery. Method: We tested implementation of an automated algorithm designed to webscrape data from Weedmaps, the largest crowdsourced website for cannabis retail, to count the number of legal cannabis retailers offering home delivery to the geographic centroid of each Census block group in California. We compared these estimates to the number of brick-and-mortar outlets within each block group. To assess data quality, we conducted follow-up telephone interviews with a subsample of cannabis delivery retailers. Results: We successfully implemented the webscraping. Of the 23,212 block groups assessed, 22,542 (97%) were served by at least one cannabis delivery business. Only 461 block groups (2%) contained one or more brick-and-mortar outlets. In interviews, availability varied dynamically as a function of staffing levels, order sizes, time of day, competition, and demand. Conclusions: Webscraping crowdsourced websites could be a viable method for quantifying rapidly evolving availability of cannabis home delivery. However, key practical and conceptual challenges must be overcome to conduct a full-scale validation and develop methodological standards. Acknowledging data limitations, cannabis home delivery appears to be nearly universal in California, whereas availability of brick-and-mortar outlets is limited, underscoring the need for research on home delivery.