Human mobility data and machine learning reveal geographic differences in alcohol sales and alcohol outlet visits across U.S. states during COVID-19

As many U.S. states implemented stay-at-home orders beginning in March 2020, anecdotes reported a surge in alcohol sales, raising concerns about increased alcohol use and associated ills. The surveillance report from the National Institute on Alcohol Abuse and Alcoholism provides monthly U.S. alcohol sales data from a subset of states, allowing an investigation of this potential increase in alcohol use. Meanwhile, anonymized human mobility data released by companies such as SafeGraph enables an examination of the visiting behavior of people to various alcohol outlets such as bars and liquor stores. This study examines changes to alcohol sales and alcohol outlet visits during COVID-19 and their geographic differences across states. We find major increases in the sales of spirits and wine since March 2020, while the sales of beer decreased. We also find moderate increases in people’s visits to liquor stores, while their visits to bars and pubs substantially decreased. Noticing a significant correlation between alcohol sales and outlet visits, we use machine learning models to examine their relationship and find evidence in some states for likely panic buying of spirits and wine. Large geographic differences exist across states, with both major increases and decreases in alcohol sales and alcohol outlet visits.

Trends in Deaths Fully Attributable to Alcohol in Minnesota, 2000-2018

Objective Overall trends in rates of fully alcohol-attributable mortality may mask disparities among demographic groups. We investigated overall, demographic, and geographic trends in fully alcohol-attributable mortality rates in Minnesota. Methods We obtained mortality data from Minnesota death certificates and defined fully alcohol-attributable deaths as deaths that would not occur in the absence of alcohol. We calculated age-adjusted death rates during 2000-2018 using 5-year moving averages stratified by decedents’ characteristics and geographic location. Results Chronic conditions accounted for most of the alcohol-attributable deaths in Minnesota (89% during 2014-2018). Alcohol-attributable mortality rates per 100 000 population increased from an average rate of 8.0 during 2000-2004 to 12.6 during 2014-2018. During 2000-2018, alcohol-attributable mortality rates were highest among males (vs females), adults aged 55-64 (vs other ages), and American Indian/Alaska Native people (vs other racial and ethnic groups) and lowest among people aged ≤24 years and Asian or Pacific Islander people. During 2014-2018, the alcohol-attributable mortality rate among American Indian/Alaska Native people was more than 5 times higher than the overall mortality rate in Minnesota. Conclusions Results from this study may increase awareness of racial and ethnic disparities and continuing health inequities and inform public health prevention efforts, such as those recommended by the Community Preventive Services Task Force, including regulating alcohol outlet density and increasing alcohol taxes.

SPATIAL ASSOCIATION PATTERNS OF BINGE DRINKING, ALCOHOL OUTLET DENSITY, AND EARLY STARTED DRINKING IN THAILAND

Thailand had a higher prevalence of binge drinking (BD) behaviors which put them at risks of morbidity and mortality. This study aimed to determine the spatial association of BD and its associated factors among the population of Thailand. This study was conducted using a data set of the National Statistical Office of Thailand and another data set of the Center for Alcohol Studies, Thailand, in 2017. A Moran's I, Local Indicators of Spatial Association (LISA), and Spatial regression were used to identify the spatial autocorrelation between alcohol outlet density, started drinking before 20 years of age, and BD among Thai people. According to the results, among 61,708 participants, the prevalence of BD was 11.47 per 1,000 population. There was global spatial autocorrelation between alcohol outlet density, start drinking before 20 years, and BD with the Moran's I values of 0.10 and 0.54, respectively. The LISA analysis identified significant positive spatial local autocorrelation of BD in the form of two high-high clusters for density of alcohol outlets and seven high-high clusters of started drinking before the age of 20. Started drinking before 20 years of age could predict binge drinking behaviors by 62.8 percent. There were spatial associations between alcohol outlet density and problems with alcoholic beverage control law enforcement that let the youngsters start drinking before 20. It is a general recommendation to strictly enforce the law in prohibited the underage from consuming alcohol, especially in the high density of alcohol outlets.

Alcohol Accessibility and Family Violence-related Ambulance Attendances

There is substantial evidence supporting the association between alcohol license density and violent crime. However, the impact of different types of alcohol licenses on intimate partner and family violence is sparse. We explored the associations between access to alcohol outlets, and family and intimate partner violence using paramedic clinical records, given this service is often the first to respond to acute crises. Coded ambulance attendance data from 694 postcodes in Victoria, Australia, from July 1, 2016 to June 30, 2018 where alcohol or another drug, mental health or self-harm associated with family or intimate partner violence was indicated were examined. A hybrid model of spatial autoregressive and negative binomial zero-inflated Poisson-based count regression models was used to examine associations with alcohol outlet density and socioeconomic factors. We found that access to a liquor license outlet was significantly associated with family violence-related attendances across all types of outlets, including on-premise (late night) licenses ( β = 1.73, SE: 0.18), restaurant licenses ( β = 0.83, SE: 0.28), and packaged liquor licenses ( β = 0.62, SE: 0.06). Our results demonstrate a significant relationship between alcohol-related harms in the context of family violence and provides evidence of the relationship between alcohol-related family violence in both victims and perpetrators. The findings of this study highlight the need for public health interventions such as licensing policy and town planning changes to reduce these harms by restricting alcohol availability.

The Cannabis Retail Environment for Young Adults in Los Angeles: Which Metrics Matter

Background: Currently, there is no consensus on how to measure cannabis retailer density. Researchers and policy makers need clear measures to support policies that mitigate unintended harms of legalization. To address this gap, our unique study leverages cannabis retailer location data in Los Angeles County (LA), California, and home addresses from an LA-based cohort of young adults (21-25 years) to develop a series of cannabis retailer density metrics and assess their relationship with cannabis use. Methods: Drawing from GIS-based measures of alcohol outlet density, we developed a series of cannabis retailer density metrics: proximity, counts within 5- 10- 15-, and 30-minute driving distances, and considered retail licensure. Retailer addresses were compiled by webscraping cannabis registries (e.g. Weedmaps) and conducting field visits (March 2019). Home addresses were geocoded for participants who completed a 2019 survey (n 1097). A series of retailer metrics was created for each person. We fit a series of multi-level logistic regression models with a random intercept by census tract (CT) (models adjusted for age, gender, race/ethnicity, college student, and CT median household income) to assess which retailer metrics were associated with any past month cannabis use. Results: Thirty percent of participants used cannabis in the past month, and 430 retailers were operating in LA in 2019. Thirty-nine percent of participants had a retailer within a mile from home and an average of 14 retailers within a 10-minute drive. Licensed retailers were less prevalent; the nearest licensed retailer was on average 2.4 miles from home. The odds of past month cannabis use significantly increased by 3% (OR:1.03, 95% CI:1.00–1.07) for every additional licensed retailer within a 10-minute drive in adjusted model; use was also significantly associated with licensed retailers within a 30-minute drive (OR:1.01, 95% CI:1.00–1.01). Proximity metrics were not significantly associated with past month cannabis use.