Area-level Effects Research Articles

The Causal Effect of Increasing Area-Level Income on Birth Outcomes and Pregnancy-Related Health: Estimates From the Marcellus Shale Boom Economy.

Income is positively correlated with pregnancy health and birth outcomes, but the causal evidence for this association is limited. Leveraging a natural experiment based on the Pennsylvania boom economy created by the extraction of natural gas from the Marcellus Shale geological formation, we test whether area-level income gains impact birth outcomes (birth weight, gestational length, and preterm birth) and pregnancy health (prepregnancy and prenatal smoking, prepregnancy weight status, gestational weight gain, and the timing and adequacy of prenatal care). We append tax data to birth certificate data and compare health outcomes before and after the boom for births occurring in school districts above the Marcellus Shale. We also explore income effects with a subsample of siblings and test for nonlinear income effects by considering preboom district poverty rates. Using instrumented difference-in-differences models, we find that plausibly exogenous income gains increase the likelihood of having adequate prenatal care in the full sample. In the sibling sample, income gains decrease the likelihood of low birth weight but increase the likelihood of prepregnancy underweight among birthing parents. Results are statistically significant in initially high-poverty districts. We thus affirm prior findings of a causal effect of income on birth weight and prenatal care use but find minimal area-level income effects on other pregnancy-related health behaviors and birth outcomes.

Global-Local Priors for Spatial Small Area Estimation

Small area estimation is gaining increasing popularity among survey statisticians. Since the direct estimates of small areas usually have large standard errors, model-based approaches are often adopted to borrow strength across areas. The models often use covariates to link different areas and random effects to account for the additional variation. In the classic Fay-Herriot model, the random effects are assumed to have independent normal distributions with a shared variance. Recent studies showed that random effects are not necessary for all areas, so global-local priors have been introduced in Tang et al.[ 26 ] to effectively characterize the sparsity in random effects. This article introduces global-local priors in the context of small area estimation where the area level random effects exhibit a spatial structure. This generalizes the findings of Tang et al.[ 26 ] where independence of the area level effects is assumed. Our findings are illustrated via both simulation and real data examples. AMS Subject Classification: 62D05, 62M30

Determinants of rural household poverty in Pakistan with multilevel approach

Rural poverty is often discussed in combination with spatial inequality where rural poor face worse problems with diverse remedies in comparison to urban poor. The present study aims to find out area-level effects and possible determinants of rural poverty based on data from the Rural Household Panel Survey of Pakistan Round-1 (2012). The multilevel binary logistic model is used to integrate concurrent individual and district-level variables to explore area-level effects on poverty. A two-level multilevel model is used for the analysis of the poverty status of 4804 respondents nested in 19 districts. Individual-level variables included in the model are education, socioeconomic status, and spending behaviour. It is found that approximately 60% of individuals of the overall population were found poor, and district-level effect accounts for 7% variations in poverty. A particularly significant ratio of poor is found for people who attained deeni madrassa/religious education or no education at all. Poverty existence is significantly associated with spending behavior and socioeconomic status. The persons with relatively middle/high socioeconomic status and successful educational achievement but a lacking attitude toward spending money are considered poor.

The Swiss neighbourhood index of socioeconomic position: update and re-validation

The widely used Swiss neighbourhood index of socioeconomic position (Swiss-SEP 1) was based on data from the 2000 national census on rent, household head education and occupation, and crowding. It may now be out of date. We created a new index (Swiss-SEP 2) based on the 2012-2015 yearly micro censuses that have replaced the decennial house-to-house census in Switzerland since 2010. We used principal component analysis on neighbourhood-aggregated variables and standardised the index. We also created a hybrid version (Swiss-SEP 3), with updated values for neighbourhoods centred on buildings constructed after the year 2000 and original values for the remaining neighbourhoods. A total of 1.54 million neighbourhoods were included. With all three indices, the mean yearly equivalised household income increased from around 52,000 to 90,000 CHF from the lowest to the highest index decile. Analyses of mortality were based on 33.6 million person-years of follow-up. The age- and sex-adjusted hazard ratios of all-cause mortality comparing areas in the lowest Swiss-SEP decile with areas of the highest decile were 1.39 (95% confidence interval [CI] 1.36-1.41), 1.31 (1.29-1.33) and 1.34 (1.32-1.37) using the old, new and hybrid indices, respectively. The Swiss-SEP indices capture area-based SEP at a high resolution and allow the study of SEP when individual-level SEP data are missing or area-level effects are of interest. The hybrid version (Swiss-SEP 3) maintains high spatial resolution while adding information on new neighbourhoods. The index will continue to be useful for Switzerland's epidemiological and public health research.

Long-term trends in urban-neighbourhood inequalities in cause-specific mortality and hospitalisation – multilevel analyses among individuals nested in Finnish post-code areas, 1991–2018

BackgroundHigh-income countries yield mixed evidence concerning the long-term trends of neighbourhood inequalities in health outcomes. The reasons why these inequalities persist and the factors driving any changes over time remain unclear. We analysed trends in general neighbourhood differences in mortality and hospitalisation, compared specific area-level and individual-level income effects, and assessed whether area-level effects were attributable to the neighbourhood population composition. MethodsThis prospective cohort study used individual-level register-linked information on sociodemographic factors covering the total population of 20–64-year-olds living in Finnish cities at the beginning of seven four-year periods in 1991–2018 (N = 952,493–1,200,431). We used random-effects Poisson models to assess all-cause and external mortality and hospitalisations among individuals nested in postal-code areas. ResultsThe general contextual effect of the neighbourhood on all-cause mortality and hospitalisation was stable across time, with a median incidence-rate ratio of around 1.20–1.30, and it was mainly attributable to the population's composition. The association between area-level income and both mortality and hospitalisation was also robust and increased slightly even after accounting for population composition. The lowest neighbourhood income quintile in 2015–2018 had 15% (95% CI:5–26%) and 30% (95% CI:15–47%) excess mortality among men and women, respectively. These differentials were particularly large for external causes, but all area-level income associations were much smaller than the corresponding individual-level associations. ConclusionThe overall relevance of the neighbourhood context to mortality and hospitalisation was stable across time, and generally attributable to population composition. However, there were substantial relative area-level income disparities between neighbourhoods, which had grown over time.

Association between socioeconomic status and chronic obstructive pulmonary disease in Jiangsu province, China: a population-based study.

Background:Chronic obstructive pulmonary disease (COPD) is a common public health problem worldwide. Recent studies have reported that socioeconomic status (SES) is related to the incidence of COPD. This study aimed to investigate the association between SES and COPD among adults in Jiangsu province, China, and to determine the possible direct and indirect effects of SES on the morbidity of COPD.Methods:A cross-sectional study was conducted among adults aged 40 years and above between May and December of 2015 in Jiangsu province, China. Participants were selected using a multistage sampling approach. COPD, the outcome variable, was diagnosed by physicians based on spirometry, respiratory symptoms, and risk factors. Education, occupation, and monthly family average income (FAI) were used to separately indicate SES as the explanatory variable. Mixed-effects logistic regression models were introduced to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for examining the SES-COPD relationship. A pathway analysis was conducted to further explore the pulmonary function impairment of patients with different SES.Results:The mean age of the 2421 participants was 56.63 ± 9.62 years. The prevalence of COPD was 11.8% (95% CI: 10.5%–13.1%) among the overall sample population. After adjustment for age, gender, residence, outdoor and indoor air pollution, body weight status, cigarette smoking, and potential study area-level clustering effects, educational attainment was negatively associated with COPD prevalence in men; white collars were at lower risk (OR: 0.60, 95% CI: 0.43–0.83) of experiencing COPD than blue collars; compared with those within the lower FAI subgroup, participants in the upper (OR: 0.68, 95% CI: 0.49–0.97) tertiles were less likely to experience COPD. Such negative associations between all these three SES indicators and COPD were significant among men only. Education, FAI, and occupation had direct or indirect effects on pulmonary function including post-bronchodilator forced expiratory volume in 1 s/forced vital capacity (FEV1/FVC), FEV1, FVC, and FEV1 percentage of predicted. Education, FAI, and occupation had indirect effects on pulmonary function indices of all participants mainly through smoking status, indoor air pollution, and outdoor air pollution. We also found that occupation could affect post-bronchodilator FEV1/FVC through body mass index.Conclusions:Education, occupation, and FAI had an adverse relationship with COPD prevalence in Jiangsu province, China. SES has both direct and indirect associations with pulmonary function impairment. SES is of great significance for COPD morbidity. It is important that population-based COPD prevention strategies should be tailored for people with different SES.

Association between the regional variation in premature mortality and immigration in Ontario, Canada

ObjectivesHealth region differences in immigration patterns and premature mortality rates exist in Ontario, Canada. This study used linked population-based databases to describe the regional proportion of immigrants in the context of provincial health region variation in premature mortality.MethodsWe analyzed all adult premature deaths in Ontario from 1992 to 2012 using linked population files, Canadian census, and death registry databases. Geographic boundaries were analyzed according to 14 health service regions, known as Local Health Integration Networks (LHINs). We assessed the role of immigrant status and regional proportion of immigrants in the context of these health region variations and assessed the contribution using sex-specific multilevel negative binomial models, accounting for age, individual- and area-level immigration, and area-level material deprivation.ResultsWe observed significant premature mortality variation among health service regions in Ontario between 1992 and 2012. Average annual rates ranged across LHINs from 3.03 to 6.40 per 1000 among males and 2.04 to 3.98 per 1000 among females. The median rate ratio (RR) decreased for men from 1.14 (95% CI 1.06, 1.19) to 1.07 (95% CI 1.00, 1.11) after adjusting for year, age, area-based material deprivation, and individual- and area-level immigration, and among females reduced from 1.13 (95% CI 1.05, 1.18) to 1.04 (95% CI 1.00, 1.05). These adjustments explained 84.1% and 94.4% of the LHIN-level variation in males and females respectively. Reduced premature mortality rates were associated with immigrants compared with those for long-term residents in the fully adjusted models for both males 0.43 (95% CI 0.42, 0.44) and females 0.45 (0.44, 0.46).ConclusionThe findings demonstrate that health region differences in premature mortality in Ontario are in part explained by individual-level effects associated with the health advantage of immigrants, as well as contextual area-level effects that are associated with regional differences in the immigrant population. These factors should be considered in addition to health system factors when looking at health region variation in premature deaths.

Ethnic disparities in psychotic experiences explained by area-level syndemic effects.

Ethnic inequalities in health outcomes are often explained by socioeconomic status and concentrated poverty. However, ethnic disparities in psychotic experiences are not completely attenuated by these factors. We investigated whether disparities are better explained by interactions between individual risk factors and place-based clustering of disadvantage, termed a syndemic. We performed a cross-sectional survey of 3750 UK men, aged 18-34 years, oversampling Black and minority ethnic (BME) men nationally, together with men residing in London Borough of Hackney. Participants completed questionnaires covering psychiatric symptoms, substance misuse, crime and violence, and risky sexual health behaviours. We included five psychotic experiences and a categorical measure of psychosis based on the Psychosis Screening Questionnaire. At national level, more Black men reported psychotic experiences but disparities disappeared following statistical adjustment for social position. However, large disparities for psychotic experiences in Hackney were not attenuated by adjustment for social factors in Black men (adjusted odds ratio, 3.24; 95% CI 2.14-4.91; P < 0.002), but were for South Asian men. A syndemic model of joint effects, adducing a four-component latent variable (psychotic experiences and anxiety, substance dependence, high-risk sexual behaviour and violence and criminality) showed synergy between components and explained persistent disparities in psychotic experiences. A further interaction confirmed area-level effects (Black ethnicity × Hackney residence, 0.834; P < 0.001). Syndemic effects result in higher rates of non-affective psychosis among BME persons in certain inner-urban settings. Further research should investigate how syndemics raise levels of psychotic experiences and related health conditions in Black men in specific places with multiple deprivations.

Parametric Modelling of M-Quantile Regression Coefficient Functions with Application to Small Area Estimation

Summary Small area estimation methods can be used to obtain reliable estimates of a parameter of interest within an unplanned domain or subgroup of the population for which only a limited sample size is available. A standard approach to small area estimation is to use a linear mixed model in which the heterogeneity between areas is accounted for by area level effects. An alternative solution, which has gained popularity in recent years, is to use M-quantile regression models. This approach requires much weaker assumptions than the standard linear mixed model and enables computing outlier robust estimators of the area means. We introduce a new framework for M-quantile regression, in which the model coefficients, β(τ), are described by (flexible) parametric functions of τ. We illustrate the advantages of this approach and its application to small area estimation. Using the European Union Survey on Income and Living Conditions data, we estimate the average equivalized household income in three Italian regions. The paper is accompanied by an R package Mqrcm that implements the necessary procedures for estimation, inference and prediction.

The impact of the 2016 flood event in Anhui Province, China on infectious diarrhea disease: An interrupted time-series study

Climate change may bring more frequent and severe floods which will heighten public health problems, including an increased risk of infectious diarrhea in susceptible populations. Affected by heavy rainfall and an El Niño event, a destructive flood occurred in Anhui province, China on 18th June 2016. This study investigates the impact of this severe flood on infectious diarrhea at both city-level and provincial level, and further to identify modifying factor. We obtained information on infectious diarrheal cases during 2013–2017 from the National Disease Surveillance System. An interrupted time-series design was used to estimate effects of the flood event on diarrhea in 16 cities. Then we applied a meta-analysis to estimate the area-level pooled effects of the flood in both flooded areas and non-flooded areas. Finally, a meta-regression was applied to determine whether proximity to flood was a predictor of city-level risks. Stratified analyses by gender and age group were also conducted for flooded areas. A significant increase in infectious diarrhea risk (RR = 1.11, 95% CI: 1.01, 1.23) after the flood event was found in flooded area with variation in risks across cities, while there was no increase in non-flooded areas. Diarrheal risks post-flood was progressively higher in cities with greater proximity to the Yangtze River. Children aged 5–14 were at highest risk of diarrhea post-flood in the flooded areas. Our study provides strong evidence that the 2016 severe flood significantly increased infectious diarrheal risk in exposed populations. Local public health agencies are advised to develop intervention programs to prevent and control infectious diarrhea risk when a major flood occurs, especially in areas close to water bodies and among vulnerable populations.

Area-Level Deprivation and Adverse Consequences in People With Substance Use Disorders: Findings From the Psychiatric and Addictive Dual Disorder in Italy (PADDI) Study

ABSTRACTBackground: Environmental factors may operate with individual ones to influence the risk of substance use. Research has focused on severe adverse consequences influenced by contextual variables. However, the literature on community level factors influencing substance use behaviors is relatively limited across Europe so far. Objective: We capitalized on data from a National survey, exploring individual and contextual characteristics, to study adverse consequences among people with substance use disorders. Methods: The impact of area-level deprivation on nonfatal overdose, hepatitis C or B infections, and major involvement with the criminal justice system, was explored. Logistic regression models with cluster-robust errors, modeling subject-level and area-level effects, were used. Results: Living in deprived and intermediate areas, as compared with affluent ones, was associated with greater likelihood of both nonfatal overdose and jail sentences longer than 6 months, though not of active viral hepatitis. Conclusions: Area-level deprivation may play an important role in determining adverse consequences in people with substance use disorders, also after controlling for individual-level characteristics. More research is needed to understand the aspects of social and physical environments that matter for drug outcomes before effective policy and research interventions can be developed.

A Longitudinal Mixed Logit Model for Estimation of Push and Pull Effects in Residential Location Choice

We develop a random effects discrete choice model for the analysis of households’ choice of neighborhood over time. The model is parameterized in a way that exploits longitudinal data to separate the influence of neighborhood characteristics on the decision to move out of the current area (“push” effects) and on the choice of one destination over another (“pull” effects). Random effects are included to allow for unobserved heterogeneity between households in their propensity to move, and in the importance placed on area characteristics. The model also includes area-level random effects. The combination of a large choice set, large sample size, and repeated observations mean that existing estimation approaches are often infeasible. We, therefore, propose an efficient MCMC algorithm for the analysis of large-scale datasets. The model is applied in an analysis of residential choice in England using data from the British Household Panel Survey linked to neighborhood-level census data. We consider how effects of area deprivation and distance from the current area depend on household characteristics and life course transitions in the previous year. We find substantial differences between households in the effects of deprivation on out-mobility and selection of destination, with evidence of severely constrained choices among less-advantaged households. Supplementary materials for this article are available online.

Are area-level effects just a proxy for school-level effects? Socioeconomic differences in alcohol consumption patterns among Swedish adolescents

AimsAlthough recent studies have found significant variations in adolescent alcohol consumption across neighbourhoods, these investigations did not address another important context in adolescents’ lives: schools. The purpose of this study was to not only simultaneously assess variations in adolescent alcohol use and binge drinking at the city district level and the school level but also analyse whether any such variations could be ascribed to the socioeconomic characteristics of the examined city districts, schools, and students. DesignCross-sectional study. SettingStockholm, Sweden. ParticipantsNinth-grade students (n=4349) attending schools (n=75) located in the city districts of the Stockholm municipality (n=14). MeasurementsTwo measures based on information regarding alcohol consumption were constructed: alcohol use (no or yes) and binge drinking among alcohol users (ranging from “very seldom” to “a few times a week”). A wide range of socioeconomic characteristics was included at the city district, school, and student levels. Alcohol use was analysed using mixed-effects logistic regression, whereas binge drinking among users was modelled using mixed-effects ordered logistic regression. FindingsThe results indicated that the school was more important than the city district in assessments of contextual variations in adolescent alcohol use in general and binge drinking in particular. Moreover, proportions of well-educated parents and high-performing students accounted for part of the school-level variation in alcohol use but not binge drinking. ConclusionsFailure to account for the school context may have caused past research to overestimate city district differences in alcohol consumption among adolescents.

Relevant social and spatial contexts for elementary school children: An examination of multiple scales

This study examined direct area-level effects of 4 common child well-being outcomes across multiple scales of residential and school neighborhoods to identify relevant contexts for measuring neighborhood effects on elementary school children. Results from growth curve models indicate that neighborhoods operationalized as residential and school census tracts exerted similar effect sizes while neighborhoods operationalized as school attendance zones showed attenuated effects. These results suggest that it may be reasonable to interchange residential and school census tracts when examining contextual effects on child well-being. In addition, results suggest that school attendance zones represent conservative, theoretically sound neighborhoods for elementary school children.

Relationships Between Deprivation and the Self-Reported Health of Older People in Northern Ireland.

There are few studies on relationships between deprivation and the self-reported health of people aged above 64 years, and no studies fully representative of Northern Ireland's older population. This article addresses this gap. Deprivation of older people as reported in the 2001 and 2011 Censuses and the relationship with self-reported health are analyzed over a 10-year span using multilevel modeling. The data are from the Northern Ireland Longitudinal Study (NILS) linked to 2001-2011 Census returns. Deprivation measures include housing tenure; property value; access to a car; and educational, employment, and area-level income deprivation. Older people suffering deprivation face a significant health disadvantage over a 10-year time span. This health disadvantage is stronger in men than in women, likely due to conservative gender roles that are prevalent among Northern Ireland's older population, leading to psychological distress especially among deprived men. The analysis found strongly significant area-level effects, aggravating the health impact of deprivation.

Exploring Spatial Variability in the Relationship between Long Term Limiting Illness and Area Level Deprivation at the City Level Using Geographically Weighted Regression.

Ecological influences on health outcomes are associated with the spatial stratification of health. However, the majority of studies that seek to understand these ecological influences utilise aspatial methods. Geographically weighted regression (GWR) is a spatial statistics tool that expands standard regression by allowing for spatial variance in parameters. This study contributes to the urban health literature, by employing GWR to uncover geographic variation in Limiting Long Term Illness (LLTI) and area level effects at the small area level in a relatively small, urban environment. Using GWR it was found that each of the three contextual covariates, area level deprivation scores, the percentage of the population aged 75 years plus and the percentage of residences of white ethnicity for each LSOA exhibited a non-stationary relationship with LLTI across space. Multicollinearity among the predictor variables was found not to be a problem. Within an international policy context, this research indicates that even at the city level, a “one-size fits all” policy strategy is not the most appropriate approach to address health outcomes. City “wide” health polices need to be spatially adaptive, based on the contextual characteristics of each area.

Adjusting for geographic variation in observational comparative effectiveness studies: a case study of antipsychotics using state Medicaid data.

BackgroundArea-level variation in treatment and outcomes may be a potential source of confounding bias in observational comparative effectiveness studies. This paper demonstrates how to use exploratory spatial data analysis (ESDA) and spatial statistical methods to investigate and control for these potential biases. The case presented compares the effectiveness of two antipsychotic treatment strategies: oral second-generation antipsychotics (SGAs) vs. long-acting paliperiodone palmitate (PP).MethodsA new-start cohort study was conducted analyzing patient-level administrative claims data (8/1/2008–4/30/2011) from Missouri Medicaid. ESDA techniques were used to examine spatial patterns of antipsychotic prescriptions and outcomes (hospitalization and emergency department (ED) visits). Likelihood of mental health-related outcomes were compared between patients starting PP (N = 295) and oral SGAs (N = 8,626) using multilevel logistic regression models adjusting for patient composition (demographic and clinical factors) and geographic region.ResultsESDA indicated significant spatial variation in antipsychotic prescription patterns and moderate variation in hospitalization and ED visits thereby indicating possible confounding by geography. In the multilevel models for this antipsychotic case example, patient composition represented a stronger source of confounding than geographic context.ConclusionBecause geographic variation in health care delivery is ubiquitous, it could be a comparative effectiveness research (CER) best practice to test for possible geographic confounding in observational data. Though the magnitude of the area-level geography effects were small in this case, they were still statistically significant and should therefore be examined as part of this observational CER study. More research is needed to better estimate the range of confounding due to geography across different types of observational comparative effectiveness studies and healthcare utilization outcomes.Electronic supplementary materialThe online version of this article (doi:10.1186/1472-6963-14-355) contains supplementary material, which is available to authorized users.

Longitudinal Effects on Mental Health of Moving to Greener and Less Green Urban Areas

Despite growing evidence of public health benefits from urban green space there has been little longitudinal analysis. This study used panel data to explore three different hypotheses about how moving to greener or less green areas may affect mental health over time. The samples were participants in the British Household Panel Survey with mental health data (General Health Questionnaire scores) for five consecutive years, and who relocated to a different residential area between the second and third years (n = 1064; observations = 5320). Fixed-effects analyses controlled for time-invariant individual level heterogeneity and other area and individual level effects. Compared to premove mental health scores, individuals who moved to greener areas (n = 594) had significantly better mental health in all three postmove years (P = .015; P = .016; P = .008), supporting a "shifting baseline" hypothesis. Individuals who moved to less green areas (n = 470) showed significantly worse mental health in the year preceding the move (P = .031) but returned to baseline in the postmove years. Moving to greener urban areas was associated with sustained mental health improvements, suggesting that environmental policies to increase urban green space may have sustainable public health benefits.

Individual- and area-level effects on mortality risk in Germany, both East and West, among male Germans aged 65+

This study investigates whether mortality inequalities based on individual- and area-level deprivation exist at older ages in Germany, and whether there are differences between eastern and western Germany. Data on population and death counts according to the individual-level socioeconomic status of male German pensioners aged 65+ years in Germany in 2002-2004 were obtained from the German Federal Pension Fund. Area-level characteristics for the 439 German districts were incorporated. Multilevel Poisson models were fitted. Individual-level socioeconomic mortality inequalities exist among elderly men in Germany. After controlling for differential population composition in the districts, we found that district-level factors contribute to the explanation of mortality inequalities in (western) Germany. The analysis further indicated that mortality and mortality inequalities tend to be higher in more economically deprived districts, and that minor mortality differences attributable to regional conditions exist in eastern Germany. The results showed that regional conditions have moderate effects on health inequalities at older ages in (western) Germany, when the differential population composition in the districts is controlled for.

Do flexible opening hours reduce violence? An assessment of a natural experiment in alcohol policy

BackgroundThroughout history, regulatory approaches to prevent alcohol-related problems have restricted the times and places at which alcohol can be sold. In 2005, the English and Welsh Government took a unique and controversial approach to violence prevention by removing restrictions on opening hours for alcohol outlets. Contrary to traditional restrictive controls, the Licensing Act (2003) proposed to reduce violence by removing restraints on regulated opening hours for licensed premises; thus potentially increasing the physical availability of alcohol. In this study, we treated the removal of trading hour restrictions as a natural experiment in alcohol policy. With a quasi-experimental pattern matching approach, we analysed the Act's effect on violence. MethodsThe study took place in the City of Manchester, UK (population 464 200, 33 electoral wards) during 2004–08. Data for violent incidents were obtained from Greater Manchester Police Service (GMP). We obtained addresses and trading hours for more than 600 licensed premises from magistrates courts (before intervention records) and local government licensing authorities (after intervention records). All data were cleaned and merged into one database with a geographical information system (GIS). We used three strategies to examine the Act's effect on violence. First, we used an interrupted autoregressive integrated moving average (ARIMA) time-series design to estimate the Act's immediate and delayed effect on violence across the city. Sensitivity analysis was done to examine effects of confounding events (ie, police enforcement campaigns, the smoking ban, etc). Second, we did a process assessment to examine how changes to exposure (ie, alcohol availability and staggered closing hours) were affected by the Act, and whether expected changes were achieved. Third, we did a dose-response analysis to examine whether changes to area-level exposures were related to changes in area-level outcomes (ward-level recorded violence). Ordinary least squares regression models adjusted for changes to ward-level outlet density, and spatial autocorrelation was examined to monitor potential geographical spill-over effects. FindingsThe results of our analyses identified little evidence that the deregulation of alcohol opening hours affected citywide violence rates. Analysis of total violence showed no evidence of any immediate, temporary, or delayed intervention effects (ω=0·781, 95% CI −0·06 to 0·17). However, analysis of violence at different times of the day showed a significant 36% increase (gradual and permanent) in weekend violence between 0300 h and 0600 h (δ=−0·897, 95% CI −1·02 to −0·77). The process assessment showed much variation in how areas implemented extra trading hours under the new regulations. Most premises extended trading hours (n=405, 67%), some did not change (n=103, 16%), and some restricted trading hours (n=18, 3%). Only one premise acquired a license to sell alcohol 24 h a day. This variation could have affected the extent to which areas staggered closing times, and increased (or decreased) the availability of alcohol. With a dose-response analysis, we examined whether changes to trading times were associated with changes to area-level violence rates. We noted no relation between changes to staggered closing times, alcohol availability, and area-level violence. InterpretationWe identified no evidence to indicate that the Licensing Act (2003) caused changes to rates of violence in the 2 years after its implementation. Our findings offer little support for fears that increased alcohol availability would lead to substantial increases in violence rates. However, explanations for this counterintuitive finding remain somewhat uncertain because of several limitations, including the short timeframe (2 years), insufficient statistical power to detect small area-level effects, possible under-reporting of violent incidents, and unmeasured and unknown confounding. Many of the limitations encountered in this study could have been averted with improved planning and collaboration between policy makers and scientists to ensure that effective monitoring and assessment was prospectively designed. FundingThe Economic and Social Research Council funded this research with a 1+3 doctoral studentship grant.