Journal of Epidemiology and Community Health最新文献

What is a good society and how might we build one? This essay, an edited version of the Cochrane lecture delivered at the Society for Social Medicine and Population Health in Bradford 2025, sets out my personal vision of what we should be aspiring to and how we might achieve it. I open with evidence connecting socioeconomic inequalities to well-being and highlight that component parts of a utopian place exist, but not all in the same place. Building upon public health lessons around prevention and upstream intervention, I then introduce what I believe are the two most promising policies we can implement: a universal basic income and a wealth tax, underpinned by citizen's assemblies, participatory budgeting and institutional structures to support evidence-based social policymaking. I make the case that we have the evidence and tools needed to collectively create a good society and that it is possible to change course and bequeath a better world on future generations. A good society is one where everyone's physical and mental health is as good as it could be, because prevention is prioritised, and health inequalities are levelled out by addressing the wider determinants of health. It is one where those who need care-whether children, those with disabilities or the elderly-are looked after without incurring financial stress, in settings where their emotional and social well-being are as important as their physical needs. Our children and young people should flourish in an education system that engages their imaginations, inspires their creativity, equips them with skills for life and leaves no one behind.

Background: We investigated whether ghost gun recovery rates are significantly associated with firearm mortality rates in the following year across California's 58 counties from 2014 to 2023.

Methods: We obtained yearly county-level data on ghost guns recovered in California from The Trace's Gun Violence Data Hub. County-level firearm death counts (total, suicide and homicide) were pulled from the Centers for Disease Control and Prevention's Restricted-Use Vital Statistics Data. Spatiotemporal models quantified the covariate-adjusted associations between ghost gun recoveries per capita and firearm death rates (total, suicide and homicide) in the following year. Secondary analyses examined suicide and homicide models stratified by sex and race/ethnicity. RESULTS : For every 20 ghost guns recovered per 100 000 population, there was an associated 6.4% increase in firearm suicide rate (adjusted incidence rate ratio (aIRR): 1.064, 95% credible interval (CrI) 1.019 to 1.111) in the following year. We found no evidence of a significant ghost gun recovery association with total firearm death rate (aIRR: 1.036, 95% CrI 0.999 to 1.075) and firearm homicide rates (aIRR: 1.002, 95% CrI 0.946 to 1.064). Stratified models for firearm suicide rates suggested variations across sex and racial/ethnic groups, with significant positive associations observed for male (6.5% increase; aIRR: 1.065, 95% CrI 1.017 to 1.115), non-Hispanic white (6.2% increase; aIRR: 1.062, 95% CrI 1.005 to 1.122) and Hispanic (12.6% increase; aIRR: 1.126, 95% CrI 1.031 to 1.230) individuals. A different pattern emerged for firearm homicide death rates, where associations across demographic groups were not statistically significant.

Conclusions: Practitioners concentrating on suicide prevention efforts should be advised about the threat that ghost guns may present.

Background: Although voting is recognised as a social determinant of health, the association between electoral participation and subsequent mortality at an individual level has not been established.

Objective: To assess whether voters and non-voters differ in mortality risk.

Methods: We used register-based information on electoral participation in the 1999 parliamentary elections from the full electorate of at least 30-year-old Finnish citizens living in mainland Finland linked to registers containing sociodemographic and mortality information by Statistics Finland. Mortality was assessed with Cox proportional hazards regression models, with follow-up until the end of 2020 (n=3 185 572 individuals; 58 133 493 person-years; 1 053 483 deaths).

Results: The age-adjusted HR of non-voters compared with voters for all-cause mortality was 1.73 (95% CI 1.72 to 1.74) for men and 1.63 (95% CI 1.62 to 1.64) for women. These differences were strongest for external causes of death and for younger age groups. Among the older (age ≥75 years) population, voting men had lower mortality than non-voting women. The difference in mortality between non-voters and voters was stronger than between those with basic and higher education.

Conclusion: Due to a strong relationship, information on voting contributes to a more comprehensive analysis of social differentials in mortality.

Introduction: Accounting for missing data by imputing or weighting conditional on covariates relies on the variable with missingness being observed at least some of the time for all unique covariate values. This requirement is referred to as positivity, and positivity violations can result in bias. Here, we review a novel approach to addressing positivity violations in the context of systolic blood pressure.

Methods: To illustrate the proposed approach, we estimate the mean systolic blood pressure among children and adolescents aged 2-17 years old in the USA using data from the 2017-2018 National Health and Nutrition Examination Survey (NHANES). As blood pressure was not measured for those aged 2-7, there exists a positivity violation by design. Using a recently proposed synthesis of statistical and mathematical models, we integrate external information with NHANES to address our motivating question.

Results: With the synthesis model, the estimated mean systolic blood pressure was 100.5 (95% CI 99.9 to 101.0), which is notably lower than either a complete-case analysis or extrapolation from a statistical model. The synthesis results were supported by a diagnostic comparing the performance of the mathematical model in the positive region.

Discussion: Positivity violations pose a threat to quantitative medical research, and standard approaches to addressing non-positivity rely on restrictive untestable assumptions. Using a synthesis model, like the one detailed here, offers a viable alternative.

Background: Deprived communities, migrants and ethnic minorities were disproportionately affected by COVID-19 and may, therefore, be at a higher risk of post-COVID condition (PCC). This analysis, using data from the Virus Watch study, investigates how deprivation, migration status and ethnic minority status influence PCC risk in both the full cohort (all regardless of infection status) and those with a confirmed COVID-19 infection.

Methods: A subset of participants from Virus Watch, a prospective community cohort study in England, were included. We used logistic regression to compare the predicted probability of developing PCC in both full and infected cohorts among different deprivation levels, migration and ethnic minority status categories by sex-at-birth during pre-Omicron and Omicron periods, adjusting for sociodemographic covariates.

Results: During the pre-Omicron period, PCC probability increased with deprivation levels, especially in females (most deprived: 7.8%, 95% CI 4.6% to 11.0%; least deprived: 3.5%, 2.5%-4.5%). Migrant and ethnic minority males had a higher likelihood of PCC than their respective counterparts, particularly in the full cohort for migrants (6.3%, 1.8%-10.8%) and the previously infected cohort for ethnic minorities (38.8%, 21.2%-56.4%). However, these disparities were less pronounced in females. In the Omicron period, these differential probabilities were also less evident.

Conclusion: Our findings suggest that greater PCC probability among these populations is driven by increased infection risk and postinfection determinants. This underscores the need for policies and interventions to reduce infection risk and affordable and easily available healthcare services for those with PCC.

The reduction of health inequalities has been a priority of researchers, decision-makers and practitioners for many years. Advances in causal mediation analysis offer great promise for identifying intervention targets and inferring how policy actions might alter health inequalities. However, these methods are sometimes presented in a manner that is not accessible to the wider community of health researchers. Causal mediation methods also have a range of limitations and assumptions that have implications for their application and the interpretation of results. In this paper, we consider three types of questions that can be used to guide policy actions to reduce health inequalities, addressed using causal mediation methods: (1) which mediating pathways offer most promise for the reduction of health inequalities and should be the focus of further, more indepth analysis? 2) In the face of two competing pathways, which one is most likely to lead to a narrowing of health inequalities? 3) What would be the impact of a hypothetical intervention on one specific mediating pathway when implemented under different scenarios? Focusing on early years' health, we use real life examples of the application of causal mediation methods to address these three types of question. In doing so, we discuss the relative strengths and limitations of these methods and introduce key mediation concepts relevant to health inequalities researchers.