Katherine Keenan

Antimicrobial resistance is shaped by biosocial and microbiological processes, yet sex and gender have not been systematically framed across the three domains (infection, antibiotic selection and transmission) that lead to drug-resistant infection. Synthesising available evidence, we find no consistent sex or gender bias in resistance prevalence; where differences are reported, prevalence is more often higher in men. Context-specific analyses are needed to determine how gendered interventions should be, if at all.

Volume 55 - Article 18 | Pages 533–572

Background: Global estimates have established antimicrobial resistance (AMR) as a major public health threat, but translating this burden into locally actionable estimates remains challenging where integrated epidemiological and microbiological data are scarce. We developed a bottom-up probabilistic framework combining routinely collected hospitalisation data with pathogen attribution and antimicrobial non-susceptibility estimates to quantify AMR burden at the pathogen, syndrome, and resistance level.Methods: We integrated statewide bacterial community-acquired pneumonia (bCAP) hospitalisations in São Paulo, Brazil, during 2022–25 with pneumococcal attribution estimates from a structured evidence review and age-specific penicillin and ceftriaxone non-susceptibility from statewide surveillance. Monte Carlo simulations propagated parameter uncertainty, and global sensitivity analysis identified its main drivers.Findings: Among 444,198 bCAP hospitalisations, 192,026 were estimated as pneumococcal CAP (pCAP). Of these, 74,985 were attributable to penicillin-non-susceptible and 45,224 to ceftriaxone-non-susceptible pneumococci. Estimated pCAP burden was 49,789 (95% uncertainty interval [UI] 11,103–87,186) among children younger than 5 years and 142,237 (95%UI 39,147–262,706) among individuals aged 5 years or older. Across all scenarios, pneumococcal attribution was the dominant driver of uncertainty.Interpretation: This bottom-up framework converts routinely available but fragmented data into locally relevant AMR burden estimates while quantifying and attributing uncertainty. It revealed a substantial burden of non-susceptible pneumococcal pneumonia and identified pathogen attribution as the principal evidence gap limiting precision. The approach could complement global AMR estimates by providing actionable subnational estimates and is potentially scalable to other pathogen, syndrome, and resistance combinations.

Antimicrobial resistance (AMR) prevalence in bacterial carriage and infection varies across age and gender, reflecting a complex interplay of underlying biological, social, structural, and healthcare-related determinants rather than demographic characteristics alone. As the importance of incorporating demographic perspectives into AMR surveillance and research is increasingly recognised, guidance is needed on assessing data quality and analytical approaches to allow for responsible use and interpretation of demographic differences in AMR. We convened an interdisciplinary workshop to identify priorities for integrating demographic perspectives into AMR research, surveillance, and policy. The workshop developed eight recommendations to support a more contextualised understanding of AMR burden and promote more equitable and responsible research: (1) strengthen demographic variables in AMR surveillance; (2) adopt a life-course approach; (3) address bias in AMR surveillance; (4) transparently report data limitations; (5) account for local context; (6) apply analytical frameworks beyond description; (7) measure multiple types of burden beyond resistance prevalence and mortality; and (8) interpret demographic differences in AMR responsibly. These recommendations provide a structured framework to support the analysis and interpretation of demographic AMR data in research and identify priorities for further quantitative methodological development and surveillance data collection. They also offer guidance for policymakers, surveillance programmes, and researchers to strengthen AMR data systems, improve interpretation of population level demographic differences in AMR, and considerations for more responsible targeting of interventions to support equitable AMR prevention and care.

We noticed that there was a coding mistake in our Stata do-file. This affected 298 participants, who had taken part in sweep 6. For these participants, the possible imputed time range should have started in year 2000, which was the year of data collection for sweep 6, but instead it started in 1986 (age 16), although this value should have only been used for those who skipped wave 6. We corrected the mistake and ran the analyses again. The corrected analyses did not change the main conclusions of the published paper. However, the results are somewhat different for the 30-33 age group, which was the most affected by the coding mistake. Thus, below we show the corrected tables and figures.

Little is known regarding the relationship between the local social environment and multiple long-term conditions (MLTC, also referred to as multimorbidity). We investigated the association between social gifting, the neighbourhood-level latent willingness to gift time for community reciprocity, and four measures of MLTC presence (‘2+ long-term conditions (LTCs)’, mental-physical MLTC, ‘3+ LTCs’ and complex MLTC). We further explored variations in these relationships across types of urban–rural settlement. We linked participants of the Scottish Longitudinal Study who participated in Census 2011, aged 40–75, with no MLTC before 2010 (n = 98 296), to their hospitalisation records (2010–19) and an established neighbourhood-level index reflecting social gifting. Two-level logistic regression was used to model the onset of MLTC (2010–19), accounting for the clustered data structure of individuals nested within neighbourhoods. Lower social gifting was associated with increased odds of MLTC in all measures, except for ‘2+ LTCs’, with the strongest association observed for mental-physical MLTC. There was a statistically significant interaction between social gifting and types of urban–rural settlement for mental-physical MLTC but not for other measures of MLTC, suggesting that social gifting was more strongly associated with mental-physical MLTC in urban than other areas. The findings highlight the important role of the local social environment in the development of MLTC. Policies targeted at supporting neighbourhood-level social cohesion and social participation may benefit population health, particularly for mental-physical MLTC in urban areas where observed associations were strongest.

The rise in divorce, cohabitation, non-marital childbearing and multi-partner fertility means that today’s children are more likely to experience less common or less stable family settings compared to previous generations. This may lead to increasing inequalities across the life course. Unlike most existing studies on family change, we investigate family trajectories in the United Kingdom from children’s perspective. We map the family trajectories characterising children’s first ten years of life using multi-channel sequence analysis on data from the UK Household Longitudinal Study, jointly capturing the dynamics of maternal partnership histories and paternal co-residence patterns from the children’s perspective. Multinomial logistic regression is applied to understand the characteristics associated with experiencing different childhood family trajectories. Children experience six typical family trajectories: continuously married; early separation; continuously cohabiting; later separation; early solo motherhood; and a new father. From birth to age ten, over a quarter of children do not continuously live with their two biological parents. Children with lower-educated mothers, mothers in the youngest or oldest groups, who live in urban areas, and belong to certain ethnic groups (White British, Mixed, Caribbean, Black African) tend to experience less common or less stable trajectories. Our elucidation of factors associated with more/less stable childhood family pathways can inform policy decision-making around support for families to mitigate growing short- and long-term inequalities giving rise to children’s diverging destinies.

Objectives: This paper estimates rates and age-trajectories of disease accumulation across South Korean birth cohorts and assesses whether observed cohort differences persist after accounting for early-life exposures and adult characteristics. Methods: Data were from eight waves of the Korean Longitudinal Study of Aging (2006-2020) and included 8,202 participants aged 50-74 years. Birth cohorts were defined by historical periods: Japanese annexation (1932-1944), Korean liberation (1945-1949), Korean War (1950-1953), and Post-war (1954-1961). Poisson mixed-effects models were used to estimate disease accumulation using counts of self-reported chronic diseases. Models were built stepwise from a baseline model with cohort fixed effects. Subsequent models added a cohort/age interaction, early-life exposures (parental death and education), and adult characteristics (own education, residence, smoking, obesity). Results: The Post-war (1954-1961) cohort has lower rates of disease accumulation than the older cohorts, which do not differ substantially from each other. Controlling for early-life exposures slightly reduces cohort differences, but controlling for adult characteristics leads to a larger reduction, leaving only the Korean liberation (1945-1949) cohort significantly different from the Post-war (1954-1961) cohort (IRR 1.17, 95% CI 1.07-1.29). This suggests adult characteristics explain most of the observed differences between the Post-war (1954-1961) and older cohorts, except the Korean liberation (1945-1949) cohort, which remains uniquely different due to unobserved factors. Discussion: These findings highlight how historical context and life course experiences are jointly associated with disease accumulation across cohorts. Adult characteristics play an important role, especially for older cohorts, and could be considered as important targets for disease prevention strategies.