The scale of the findings is striking. At their peak, annual prevalence estimates reached 347 cases per 100,000 population in primary care settings in 2017 and 314 per 100,000 in hospitals in 2016, figures that consistently exceed the threshold of more than 100 cases per 100,000 person-years used to define high-burden settings. Even after years of gradual decline, the most recent estimates from 2023 stood at 143 per 100,000 in primary care and 167 per 100,000 in hospitals, still well above that benchmark. Hospital admissions accounted for 47.62 percent of all typhoid-related visits, and cumulative deaths over the nine-year period included 11,991 recorded in hospitals, corresponding to an average annual hospital mortality rate of 0.57 per 100,000 population.

Methodologically, the study is a serial cross-sectional analysis built on an unweighted sample of 2,499,681 unique individual records, which, after applying BPJS Kesehatan’s stratified sampling weights, represents a national population of approximately 290 million people. Cases were identified using physician-assigned ICD-10 codes, primarily A01.0 for typhoid fever, recorded as either primary or secondary diagnoses. To avoid inflating prevalence with repeat consultations for the same illness, the researchers applied an episode-based filtering approach: encounters occurring within 30 days of a preceding record for the same individual were treated as part of a single clinical episode, a boundary chosen because typhoid’s incubation period typically ranges from 6 to 30 days and relapses most commonly occur two to three weeks after treatment ends. Only the final record within each episode was retained for analysis.

The choice of data source matters because diagnosing typhoid fever in endemic settings is notoriously difficult. The Widal agglutination test, long used in low-resource settings, has poor sensitivity and specificity because of high background seroprevalence, and rapid diagnostic tests offer only modest improvements. Blood culture remains the reference standard, but its sensitivity drops after prior antibiotic use, and laboratory capacity and cost constraints mean microbiological confirmation is rarely performed in routine Indonesian practice. In primary care, where reimbursement is capitation-based, laboratory confirmation is not routinely required at all; in hospitals, supporting investigations such as Tubex, Typhidot, or Widal testing are generally undertaken for reimbursement claims, though these do not replace culture confirmation. The study’s estimates therefore describe the burden of physician-coded typhoid requiring healthcare utilisation rather than microbiologically confirmed disease, a distinction the authors acknowledge carefully.

Demographically, the burden fell unevenly. Adults aged 19 to 64 accounted for roughly half of all typhoid-related visits, yet the highest prevalence was observed among children under five years of age in both primary care and hospital settings. Adjusted analyses using Poisson regression with robust variance estimation revealed that children aged 0 to 5 years had a modestly higher likelihood of hospitalisation compared with working-age adults, with an adjusted prevalence ratio of 1.02, while adolescents aged 6 to 18 showed a ratio of 1.03. The most pronounced elevation appeared among people aged 65 and older, whose adjusted prevalence ratio of 1.07 was the highest of any age group, despite this group having the lowest overall prevalence of typhoid-related visits. The authors suggest this may reflect comorbidities, frailty, and a lower clinical threshold for admitting older patients rather than greater disease severity per se.

Socioeconomic vulnerability emerged as another powerful determinant. Individuals enrolled in the subsidised insurance scheme, which serves low-income households, had a 7 percent higher likelihood of hospitalisation than those in the formal sector, with an adjusted prevalence ratio of 1.07. Members of Class III, the most affordable tier of care with the simplest inpatient facilities, contributed the largest share of visits across all settings. These patterns, the researchers argue, reflect underlying socioeconomic disadvantage and potential barriers to timely access to care, meaning that by the time subsidised patients reach hospital, their illness may be more advanced. The intersection of age-related vulnerability and social disadvantage, they conclude, shapes severe disease trajectories in ways that targeted interventions must address.

Geographically, the study uncovered pronounced subnational clustering that national-level statistics had long obscured. Primary care prevalence was highest in Central Java province at a mean annual 302 per 100,000, followed by Bengkulu at 289, while hospital prevalence peaked in Aceh at 496 per 100,000, followed by Gorontalo and North Sumatra. When provinces were aggregated into the INA-CBGs regional classification, the highest annual prevalence reached 2,367 visits per 100,000 population in region 1, covering Java, Indonesia’s most populous island, for primary care, and 2,213 per 100,000 in region 3, covering Sulawesi and parts of Sumatra and Kalimantan, for hospitals. Urban areas consistently showed higher prevalence than rural ones across both care settings throughout the study period.

The concentration of cases in densely populated Java is particularly revealing because it runs counter to a simple sanitation narrative. Several provinces in Java exhibited both relatively high percentages of households with access to improved sanitation and high typhoid prevalence, while provinces outside Java with better sanitation access generally showed lower prevalence. Previous research has documented widespread use of shallow, contaminated groundwater as a household drinking water source and poor sanitation management in dense and informal urban settlements on Java, and parts of Sumatra and Sulawesi where cases clustered also show WASH infrastructure access below the national figure. The eastern provinces of Papua and Papua Barat recorded the lowest prevalence, at 24 and 10 per 100,000 respectively in primary care, but the authors caution this likely reflects lower healthcare utilisation and under-ascertainment rather than genuinely low transmission.

The temporal trends carry their own lessons. A marked decline in cases occurred during 2020 and 2021, coinciding with the COVID-19 pandemic, likely reflecting the reallocation of 20 to 40 percent of general ward beds to COVID-19 care, reduced healthcare-seeking, and non-pharmaceutical interventions such as social distancing and improved hand hygiene that may have limited faecal-oral transmission. When cases rebounded in 2022, the increase was more pronounced among inpatients than outpatients, which the authors attribute to the restoration of routine inpatient services rather than a true rise in disease severity. Earlier fluctuations during 2016 and 2017 may relate to the rapid expansion of insurance membership, evolving referral policies, and maturation of coding practices during the early phase of the scheme’s implementation.

The implications extend well beyond Indonesian borders. The Global Burden of Disease 2023 study estimates that Indonesia has the highest typhoid incidence, 99 per 100,000, and mortality rate, 1.29 per 100,000, in Southeast Asia, yet those figures rely substantially on modelled extrapolation in a country lacking nationwide culture-confirmed surveillance. The new claims-based estimates, which exceed even the GBD figures, arrive as neighbouring countries move forward with typhoid conjugate vaccine introduction: Nepal and Pakistan have incorporated the vaccine into routine immunisation, and Bangladesh introduced it in 2025, following the World Health Organization’s 2018 recommendation of a single dose for children from six months of age. Indonesia, however, continues to face financing constraints and competing priorities, with policymakers citing a lack of robust contemporary epidemiological data. The authors also point to complementary environmental surveillance showing Salmonella Typhi in 13 percent of 406 wastewater samples collected from treatment plants, manholes, rivers, and public spaces, evidence of persistent community transmission. Combined with rising multidrug and extensively drug-resistant strains documented across endemic settings, and imported cases linked to travel from Indonesia reported abroad, the converging evidence strengthens the case for introducing the vaccine alongside targeted water and sanitation improvements, prioritising high-burden regions and vulnerable populations, and building serosurveillance and wastewater monitoring into the country’s surveillance architecture.

Subject of Research: Population-based epidemiology of typhoid fever in Indonesia using national health insurance claims data

Article Title: Burden of typhoid fever in primary and hospital care in Indonesia: a serial cross-sectional study using national health insurance (BPJS Kesehatan) data, 2015–2023

Article References: Oktaria, V., Ramadhani, A. D., Maula, A. W., Murni, I. K., Watts, E., Baehaqi, I. S., Dinari, R., Wiryawan, D. A., Daniel, D., Donato, C. M., & Bines, J. E. (2026). Burden of typhoid fever in primary and hospital care in Indonesia: a serial cross-sectional study using national health insurance (BPJS Kesehatan) data, 2015–2023. The Lancet Regional Health – Western Pacific, 75, Article 101994. https://doi.org/10.1016/j.lanwpc.2026.101994

Image Credits: AI Generated

DOI: 10.1016/j.lanwpc.2026.101994

Keywords: typhoid fever, Indonesia, Salmonella Typhi, BPJS Kesehatan, epidemiology, hospitalisation, typhoid conjugate vaccine, water sanitation and hygiene, antimicrobial resistance, health insurance claims, subnational disparities, wastewater surveillance

The research, published in the Journal of Neurology by a team at Karolinska Institutet, drew on the Environmental Investigation in Multiple Sclerosis, or EIMS, a population-based case-control study that has been running in Sweden since 2005. The investigators analyzed data from 3,008 people newly diagnosed with multiple sclerosis and 5,773 matched controls drawn from the national population register. Crucially, the team restricted their analysis to participants whose disease onset, or corresponding reference year among controls, occurred within five years before study enrollment, a design choice intended to ensure that reported exercise habits genuinely preceded the disease rather than being shaped by its early, invisible stages.

Participants reported their leisure-time physical activity five years before inclusion, categorized into four levels ranging from largely sedentary living to sweat-inducing exercise at least three times per week. The lowest level corresponded to less than two hours per week of light activity such as walking or cycling, while the highest required at least thirty minutes of vigorous exercise three or more times weekly. Sun exposure over the same period was captured through a composite index combining sunbathing in Sweden, travel to sunnier countries, and use of sunbeds, with scores summed into a scale ranging from three to twelve.

At first glance, the results seemed to confirm what many earlier studies had suggested. In a model adjusting only for age, sex, and residential area, the most active participants had a 22 percent lower risk of multiple sclerosis than the least active, with an odds ratio of 0.78. A clear downward trend emerged across the activity levels, with each step up the activity ladder associated with roughly 8 percent lower odds of disease. For anyone who has followed the exercise-prevention literature, this is exactly the pattern one would expect to see.

But the picture changed dramatically once the researchers accounted for the full web of lifestyle and environmental factors that travel alongside physical activity. After adjusting for ancestry, educational level, smoking status, alcohol consumption, body mass index at age twenty, and sun exposure, the odds ratio for high versus low activity drifted to 0.94, a figure statistically indistinguishable from no effect at all. The per-level trend estimate similarly weakened from 0.92 to 0.97. In other words, much of what looked like protection from exercise appeared to be protection from something else that active people tend to have more of, most notably sunshine.

The stratified analyses revealed the study’s most intriguing finding. Among participants with low sun exposure, defined as a composite index score below six, high physical activity remained associated with a 24 percent reduction in multiple sclerosis risk, with an adjusted odds ratio of 0.76 and a statistically significant inverse trend across activity levels. Among those with high sun exposure, no such association existed; the odds ratio was a flat 0.97. Formal statistical testing confirmed that the interaction between physical activity and sun exposure was unlikely to be a fluke, with a p-value of 0.01, and model-predicted probabilities visualized the gradient clearly: the protective slope of exercise was steepest in the sun-deprived and vanished as sun exposure rose.

Skeptics might argue that physical activity is simply a proxy for time spent outdoors, and that sunshine, not exercise, does the real work. To test this, the team conducted a sensitivity analysis in a subgroup of participants recruited after 2015 who had provided detailed information on time spent outdoors during summer leisure and at work. Remarkably, the inverse association between physical activity and multiple sclerosis risk persisted essentially unchanged after additional adjustment for outdoor time, with the odds ratio per activity level moving only from 0.87 to 0.88. This suggests that exercise itself, or something closely tied to it beyond mere outdoor exposure, may genuinely matter for people whose sun exposure is otherwise low.

What biological mechanisms could underlie such a pattern? Regular exercise is known to exert systemic anti-inflammatory effects, influencing cytokine profiles, metabolic regulation, adiposity-related inflammation, and the signaling of myokines, molecules released by contracting muscle. These pathways intersect with the immune processes implicated in the autoimmune attack on the central nervous system that defines multiple sclerosis. Sun exposure, meanwhile, is the principal driver of vitamin D production and has well-documented immunomodulatory effects, and low sun exposure has previously been shown in the same Swedish cohort to raise multiple sclerosis risk both directly and indirectly. One plausible interpretation is that sunlight and exercise act on overlapping biological pathways, so that the marginal benefit of exercise becomes visible only when the sun-related pathway is underactive.

The authors are careful to emphasize the caveats. All exposures were self-reported retrospectively, opening the door to recall bias, and the interval between the reported activity period and clinical onset varied among participants. Reverse causation remains a serious concern, because multiple sclerosis is increasingly recognized to have a prodromal phase, sometimes years before diagnosis, in which fatigue, reduced exercise tolerance, pain, and subtle neurological changes may quietly reshape a person’s activity habits. The broad four-level activity categories also cannot capture every nuance of duration, intensity, and seasonal variation, and the sun exposure index is an imperfect stand-in for total ultraviolet radiation dose. The sun-exposure interaction, the team notes, should be considered exploratory until replicated.

Still, the study’s strengths are considerable: a population-based design, nearly 3,000 incident cases, high participation rates, and unusually comprehensive confounder data, including a check against residual confounding by population structure using genetic principal components. The takeaway for the public is not that exercise is useless against multiple sclerosis, but that its potential preventive role is context-dependent, most apparent in people with limited sun exposure and robust even after accounting for time outdoors. As the authors conclude, physical activity should be understood in relation to sun exposure, outdoor behavior, and the broader lifestyle patterns in which it is embedded, a reminder that in disease-prevention research, no single behavior operates in isolation.

Subject of Research: The association between leisure-time physical activity, sun exposure, and risk of developing multiple sclerosis

Article Title: Leisure-time physical activity, sun exposure, and multiple sclerosis risk: a population-based case–control study

Article References: Guo, Q., Olsson, T., Alfredsson, L., & Hedström, A. K. (2026). Leisure-time physical activity, sun exposure, and multiple sclerosis risk: a population-based case–control study. Journal of Neurology, 273(10), Article 623. https://doi.org/10.1007/s00415-026-14170-9

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14170-9

Keywords: multiple sclerosis, physical activity, sun exposure, case-control study, epidemiology, autoimmune disease, vitamin D, outdoor activity, risk factors, Karolinska Institutet, disease prevention, Journal of Neurology

The research, led by Anna Karin Hedström, Tomas Olsson, and Lars Alfredsson of Karolinska Institutet and published in the Journal of Neurology, drew on the Epidemiologic Investigation of Multiple Sclerosis, a population-based study that recruited newly diagnosed patients from 40 neurology units across Sweden, including every university hospital, between April 2005 and December 2021. Of 3,792 incident cases enrolled with a participation rate of 91 percent, the final analytic sample comprised 3,094 individuals after excluding those with uncertain family history or missing follow-up disability data. Each participant completed a structured questionnaire at inclusion covering demographics, environmental exposures, lifestyle factors, and family history of the disease.

Participants were sorted into three groups based on their closest affected relative: those with no known family history, those with a first-degree relative with multiple sclerosis, and those with a second- or third-degree relative with the disease. In total, 239 participants, or 7.7 percent, reported an affected first-degree relative, while 468, or 15.1 percent, reported an affected second- or third-degree relative. This hierarchical classification by degree of relatedness is a methodological strength, because many earlier studies lumped all familial cases together or relied on small clinic-based samples that could not disentangle the effects of relatedness from other prognostic factors.

The study then linked participants to the Swedish MS Registry, a nationwide resource integrated into routine clinical practice in which neurologists prospectively record disease activity, neurological function, and treatment exposure at every visit. Baseline was defined as the first available Expanded Disability Status Scale assessment, and the primary outcome was confirmed disability worsening, defined as a sustained increase of at least 1.0 EDSS point confirmed at a subsequent visit at least six months later, with adjustments for lower baseline scores. The researchers also tracked time to EDSS milestones 3 and 4, new or enlarging T2 lesions on MRI, cognitive decline on the Symbol Digit Modalities Test, and worsening on the patient-reported Multiple Sclerosis Impact Scale-29, capturing both physical and psychological dimensions of the disease.

Using Cox proportional hazards regression adjusted for age, sex, ancestry, education, calendar year of diagnosis, disease duration, baseline disability, smoking, body mass index, and initial disease-modifying therapy, the team found no significant associations between family history status and any of these outcomes. The hazard ratio for confirmed disability worsening was 1.06 among people with a first-degree relative with the disease and 1.04 among those with a second- or third-degree relative, with confidence intervals spanning unity in both cases. Separate analyses of relapse-associated worsening and progression independent of relapse activity, the two components of confirmed worsening, likewise showed no clear differences, and sensitivity analyses that modeled treatment as a time-varying covariate or reclassified participants with uncertain family history left the conclusions unchanged.

The negative findings extended across every domain examined. MRI disease activity, measured as the first appearance of new or enlarging T2 lesions, showed hazard ratios of 1.11 and 0.95 for the first-degree and second-/third-degree groups respectively. Cognitive worsening, defined as a drop of eight or more points on the Symbol Digit Modalities Test, yielded hazard ratios of 1.05 and 0.93, and a stricter four-point threshold produced similar results. Patient-reported worsening on both the physical and psychological subscales of the MSIS-29 was also unrelated to family history. Linear mixed-effects models tracking longitudinal change over time found no evidence that trajectories differed among the family history groups for cognition or for either patient-reported subscale, with interaction p-values ranging from 0.267 to 0.941.

One baseline difference did emerge: individuals with a first-degree relative with multiple sclerosis were significantly more likely to carry the HLA–DRB1*15:01 allele, the strongest genetic risk factor for the disease, with 68.3 percent positivity compared with 52.8 and 53.4 percent in the other groups. Beyond this, however, the groups were remarkably similar at disease onset. Age at onset, sex distribution, disease phenotype, smoking habits, body mass index, sun exposure, physical activity, Epstein-Barr virus antibody levels measured against the main MS-associated EBNA-1 peptide segment, and vitamin D levels showed no meaningful differences, suggesting that people with and without familial disease constitute clinically and epidemiologically comparable populations once the disease has declared itself.

The authors argue that this overall similarity points to a fundamental biological distinction between disease initiation and disease progression. Most of the more than 200 genetic susceptibility loci identified through genome-wide association studies relate to immune regulation and the risk of developing multiple sclerosis in the first place, not to how severe the disease becomes afterward. Familial clustering, on this view, primarily reflects shared predisposition to disease onset, while later progression is driven by other forces, including treatment exposure, aging, comorbidities, and lifestyle factors. A recent genome-wide association study published in Nature identified a severity locus implicating central nervous system resilience, underscoring that the genetics of progression may be largely separate from the genetics of susceptibility.

The researchers also acknowledge that family history is a blunt instrument. It does not capture the underlying burden or composition of genetic risk variants, and it conflates inherited factors with shared familial environment. Individuals classified as having familial disease may carry very different genetic risk profiles, while some people without a known affected relative may nonetheless harbor substantial inherited susceptibility. Such heterogeneity would tend to dilute any true association toward the null. Self-reported family history may also be incomplete, particularly for more distant relatives, introducing nondifferential misclassification. Yet the absence of any consistent gradient across degrees of relatedness in the sensitivity analyses argues against a hidden signal of clinically meaningful magnitude.

For patients and clinicians, the practical message is considerable. A family history of multiple sclerosis, which often looms large in the minds of newly diagnosed patients and their relatives, should not by itself be treated as a marker of worse prognosis or a reason to escalate therapy more aggressively. The study does not rule out modest effects, particularly among the comparatively small first-degree group, and the observational design leaves open the possibility of residual confounding by treatment decisions or unmeasured factors such as depression. But within the limits of a nationwide cohort with long-term follow-up, the conclusion is clear: inherited susceptibility shapes who develops multiple sclerosis, not how the disease unfolds once it has begun. The search for the determinants of progression must therefore look elsewhere, toward the emerging genetics of severity and the modifiable exposures that follow diagnosis.

Subject of Research: The prognostic impact of family history of multiple sclerosis on long-term disease outcomes

Article Title: Family history of multiple sclerosis and long-term outcomes

Article References: Hedström, A. K., Olsson, T., & Alfredsson, L. (2026). Family history of multiple sclerosis and long-term outcomes. Journal of Neurology, 273(10), Article 622. https://doi.org/10.1007/s00415-026-14162-9

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14162-9

Keywords: multiple sclerosis, family history, disability progression, HLA-DRB1*15:01, genetic susceptibility, prognosis, MRI disease activity, cognitive decline, Swedish MS Registry, EIMS, relapsing-onset MS, epidemiology

The review team searched four major databases—PubMed, Web of Science, MEDLINE, and Scopus—for both observational and interventional studies that assessed whether fluoride interventions reduced social inequalities in tooth decay. They deliberately kept the focus tight. Studies conducted in institutional populations were excluded, as were studies that combined fluoride with other preventive measures, since bundling interventions makes it impossible to isolate what fluoride itself contributes to equity. The methodological rigor of the exercise was considerable: the team assessed the risk of bias in each included study using the ROBINS-E tool, designed specifically for non-randomized studies of exposures, and graded the certainty of the evidence with an adapted GRADE approach. Because the studies were too heterogeneous in design, populations, and outcome measures to pool statistically, no meta-analysis was possible—a limitation that itself says something about the state of this field.

What emerged from the search was a body of evidence dominated by a single intervention. Of the twenty-eight studies that met the inclusion criteria, twenty-six examined community water fluoridation, the practice of adjusting the fluoride concentration of public drinking water to a level known to protect teeth against decay. Only one study looked at fluoride toothpaste and one at fluoride mouth rinse. That imbalance matters, because water fluoridation is the flagship population-level fluoride policy in many countries, and it is precisely the kind of universal intervention that equity theorists argue should disproportionately benefit the most disadvantaged—reaching everyone regardless of income, dental insurance status, or access to a dentist.

The theoretical case for fluoridation as an equalizer is straightforward. Dental caries is driven by frequent sugar exposure, inadequate oral hygiene, and limited access to preventive care, all of which cluster along socioeconomic lines. A fluoride intervention delivered through the water supply requires no behavior change, no appointment, and no out-of-pocket payment, so it should, in principle, lift the oral health of those with the fewest resources the most. The review’s findings offered partial support for that logic. Some studies reported narrower social inequalities in dental caries in the presence of community water fluoridation, particularly when inequality was measured in absolute terms—for example, the raw difference in decayed, missing, and filled teeth or surfaces between socioeconomic groups.

But the picture darkened when the reviewers examined relative measures of inequality, such as ratios or indices that express the gap as a proportion of overall disease levels, and when they looked at formal effect-modification analyses, which test whether the fluoride effect differs by socioeconomic position. Across those analyses, the findings were inconsistent. This distinction between absolute and relative inequality is not a statistical nicety. Because tooth decay is generally more common in disadvantaged groups, an intervention that reduces decay everywhere can narrow the absolute gap while leaving the relative gap unchanged, or even widening it, simply because the baseline difference is so large. Which measure matters most is a value judgment as much as a scientific one, and the review’s inability to find consistent results across both types of measurement underscores how sensitive equity conclusions are to methodological choices.

There was also a striking demographic blind spot. Every one of the included studies focused on children and adolescents; not a single study examined adults or older populations. That is a significant gap given that dental caries is a lifelong, cumulative disease, and given that older adults—particularly those in aged care or on fixed incomes—face some of the highest rates of decay and the worst access to dental services. If fluoridation and other fluoride interventions are to be justified as equity-promoting policies, the evidence base will eventually need to extend beyond childhood. For now, claims about fluoride and inequality in adults rest on inference rather than direct evidence.

The quality assessment was, frankly, sobering. Twenty-seven of the twenty-eight included studies were rated as being at high or very high risk of bias, and only one achieved a moderate rating. None were judged low risk. This is not necessarily an indictment of the individual research teams; studying the population-level effects of water fluoridation is inherently difficult. Randomized trials are practically and ethically fraught when the exposure is a public water supply, so researchers rely on natural experiments, cross-sectional surveys, and before-and-after comparisons, all of which are vulnerable to confounding. Communities that fluoridate their water may differ systematically from those that do not, in wealth, in health infrastructure, in other preventive programs, and disentangling the fluoride signal from that background noise is genuinely hard.

Reflecting those limitations, the certainty of the evidence was graded as very low across the board: for community water fluoridation among children and adolescents, for topical fluoride interventions such as toothpaste and mouth rinse, and for evidence emerging from low- and middle-income country settings. In GRADE terminology, very low certainty means that the true effect could plausibly be substantially different from what the available studies suggest. The reviewers were careful in their conclusion: the findings suggested that community water fluoridation may contribute to narrower inequalities in dental caries among children and adolescents, but the certainty underpinning that suggestion was very low, and high-quality research is needed to establish both the causal relationship and the magnitude of the effect across different populations and settings.

Why does this matter beyond academic circles? Community water fluoridation remains politically contested in many countries, with periodic referendums and campaigns questioning its safety, its ethics, and its effectiveness. Proponents have long cited its equity benefits as a core justification, and the World Health Organization has endorsed it as a cornerstone population strategy against dental decay. This review does not overturn that case, but it does reveal how thin the evidentiary foundation for the equity argument actually is. Policymakers weighing fluoridation decisions are being asked to act on evidence that is suggestive rather than definitive, and on studies whose risk-of-bias ratings would make any clinical guideline committee wince.

The review also carries a message for researchers. The dominance of water fluoridation in the literature leaves topical fluoride—varnishes, toothpastes, rinses—almost unstudied from an equity perspective, despite the fact that these interventions require adherence and access, the very resources that disadvantaged populations lack. And the near-total absence of evidence from low- and middle-income countries means that global oral health policy is being guided by findings generated almost entirely in high-income settings, where water infrastructure, sugar consumption patterns, and dental care systems differ profoundly. The reviewers’ call for high-quality research is therefore not a routine academic plea; it is a roadmap. If fluoride is to serve as a genuine instrument of health equity, the next generation of studies will need to measure inequality explicitly, in both absolute and relative terms, across the full life course, and in the populations where the dental divide is widest.

Subject of Research: The effectiveness of fluoride interventions in reducing social inequalities in dental caries

Article Title: The effectiveness of fluoride interventions in reducing social inequalities in dental caries: a systematic review

Article References: Ge, Z., Guarnizo-Herreño, C. C., Ye, L., Kaur, G., Chen, Z., Hopcraft, M., & Singh, A. (2026). The effectiveness of fluoride interventions in reducing social inequalities in dental caries: a systematic review. International Journal for Equity in Health. https://doi.org/10.1186/s12939-026-03045-0

Image Credits: AI Generated

DOI: 10.1186/s12939-026-03045-0

Keywords: fluoride, community water fluoridation, dental caries, health inequalities, systematic review, socioeconomic position, oral health, public health, epidemiology, health equity, topical fluoride, children and adolescents

The research, published in eClinicalMedicine, drew on the Clinical Practice Research Datalink Aurum database, an anonymized repository of medical records covering more than 50 million patients registered with over 1800 general practices across England. The investigators linked these primary care records to Hospital Episode Statistics, which capture every admission to publicly funded hospitals in the country, allowing them to track both routine diagnoses and serious exacerbations of lung disease. From this vast data resource, they assembled a cohort of 444,668 adults aged 50 and over with a physician diagnosis of COPD, who together contributed nearly 2.3 million person-years of follow-up. Within this group, 37,645 people, or 8.5 percent, were newly diagnosed with atrial fibrillation during the study window.

The scale of the comorbidity problem is difficult to overstate. COPD affects more than 390 million people worldwide and is the third leading cause of death globally, while atrial fibrillation touches over 52 million lives. Previous meta-analyses have estimated that patients with COPD face roughly twice the risk of developing atrial fibrillation compared with those without the lung disease, and that risk climbs as lung function deteriorates. The new study reinforces this dose-response relationship with unusual granularity. Incidence rates rose consistently with age, were higher in men than women across every age band, and were markedly elevated among patients with more severe COPD, whether severity was measured by forced expiratory volume in one second, breathlessness scores, or recent exacerbations requiring hospitalization. Patients hospitalized for a severe COPD flare in the previous year showed incidence rates above 55 per 1000 person-years by 2022, several times higher than those with well-preserved lung function.

The temporal pattern also tells a story about modern healthcare. After a gradual year-on-year climb, incidence plateaued between 2016 and 2019, then dropped 22 percent in 2020 compared with 2019, a dip the authors attribute to reduced access to primary care during the COVID-19 pandemic. The rebound afterward suggests the underlying trend was never interrupted, merely masked. The researchers note that the rising trajectory mirrors patterns seen in the general UK population, implying that improved detection and better survival among people with atrial fibrillation risk factors, rather than something unique to COPD, may be driving much of the increase. Still, the consistency of the elevation across every marker of COPD severity points to a genuine biological contribution, likely involving chronic inflammation, hypoxia, and the structural cardiac changes that accompany long-standing lung disease.

Why does this matter so much? Because atrial fibrillation in a COPD patient is more dangerous than atrial fibrillation alone. Recent cohort evidence indicates that people with both conditions face approximately 1.6 times the risk of ischemic stroke compared with those who have atrial fibrillation without COPD. Oral anticoagulants are the cornerstone of stroke prevention in atrial fibrillation, and current guidelines recommend that treatment decisions be made independently of COPD status. Yet the real-world evidence on whether this actually happens has been sparse and contradictory, with some studies suggesting higher anticoagulation rates in COPD patients and others suggesting lower or equivalent rates.

The new findings go further than any previous work by tracking prescribing trends over thirteen years. Among the 34,851 patients newly diagnosed with atrial fibrillation during follow-up, 66.3 percent received an oral anticoagulant within a year of diagnosis. The rate of initiation nearly tripled over the study period, from 10.3 prescriptions per 100 person-months in 2010 to 30.1 per 100 person-months in 2022, a 2.9-fold increase. The most dramatic rise occurred among patients aged 90 and older, whose initiation rate was more than nine times higher in 2022 than in 2010, albeit from a very low baseline. This transformation was driven almost entirely by the direct oral anticoagulants, or DOACs, which received European approval for stroke prevention in 2011. Warfarin and other vitamin K antagonists, once the only option, collapsed to a 96 percent lower initiation rate by 2022, while DOAC use surged 64-fold between 2012 and 2022.

The fine-grained prescribing data reveal how completely the therapeutic landscape has changed. Apixaban emerged as the dominant agent, peaking at 208.1 prescriptions per 1000 person-months in 2021 and accounting for 61.6 percent of all oral anticoagulants started between 2020 and 2022. Edoxaban, introduced in 2016, overtook rivaroxaban by 2020 and then jumped from 49.4 to 95.4 prescriptions per 1000 person-months between 2021 and 2022. Rivaroxaban, an early favorite, peaked in 2016 and declined steadily thereafter. Dabigatran, the first DOAC approved, never gained significant traction. Notably, these patterns closely match prescribing trends reported in atrial fibrillation patients generally, suggesting that the presence of COPD did not materially alter which drug clinicians chose, only, in some cases, whether they prescribed one at all.

And that is where the treatment gap emerges. Despite the overall tripling of anticoagulation, several high-risk groups were persistently less likely to receive treatment. Patients aged 90 and older had the lowest initiation rates of any age group throughout the study, even though stroke risk climbs steeply with age and guidelines recommend anticoagulation for most patients with atrial fibrillation over 75. Non-White patients were prescribed anticoagulants less often than White patients, echoing documented disparities in atrial fibrillation care more broadly. Most strikingly, people with the most severe COPD, identified by low percent-predicted lung function, severe breathlessness, or recent hospitalization for an exacerbation, consistently received fewer prescriptions than those with milder disease, a pattern that held across all years, ages, and sexes. Patients with chronic kidney disease, liver disease, a history of bleeding, cancer, or dementia were also less likely to be treated.

The authors caution that these patterns may not reflect inappropriate care alone. Contraindications, frailty, polypharmacy, and patient preferences all legitimately shape prescribing decisions, and clinicians may reasonably hesitate before adding a blood thinner to a regimen already burdened by multiple respiratory medications. Yet the concern about physician hesitancy is hard to dismiss, because evidence on the safety and efficacy of anticoagulants specifically in patients with coexisting COPD and atrial fibrillation remains thin. Without dedicated trials or robust observational outcome data in this population, uncertainty itself becomes a barrier to treatment, and the patients facing the highest thromboembolic risk are left in a zone of clinical ambiguity.

The study has limitations worth noting. The database does not capture prescriptions issued by specialists or during hospitalizations, which may lead to modest underestimation of prescribing rates, though the central role of general practitioners in the UK system likely mitigates this. Restricting the cohort to adults over 50 reduced but did not eliminate the possibility of misclassifying some asthma patients as having COPD. The ethnicity-stratified analyses aggregated small numbers of non-White patients into a single category, limiting insight into specific minority subgroups, and the cohort included only individuals with COPD, so comparisons with the general population are indirect. Even so, the sheer size of the cohort, the representativeness of the data, and the consistency of results across sensitivity analyses using multiple imputation lend considerable weight to the conclusions. The message for clinicians and health systems is clear: as atrial fibrillation becomes increasingly common among people with COPD, stroke prevention must keep pace, and the oldest, sickest, and most ethnically marginalized patients, precisely those at greatest risk, should not be left behind. Integrated care pathways and routine bidirectional screening for both conditions, identified as research priorities by the American Thoracic Society, may offer a route forward, but the first step is recognizing that a growing treatment gap exists at all.

Subject of Research: Temporal trends in atrial fibrillation incidence and oral anticoagulant prescribing among adults with COPD in UK primary care

Article Title: Trends in the incidence of atrial fibrillation and the prescription of oral anticoagulants in adults with chronic obstructive pulmonary disease in UK primary care (2010–2022): a cohort study

Article References: Trends in the incidence of atrial fibrillation and the prescription of oral anticoagulants in adults with chronic obstructive pulmonary disease in UK primary care (2010–2022): a cohort study. (n.d.). https://doi.org/10.1016/j.eclinm.2026.104245

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104245

Keywords: atrial fibrillation, COPD, oral anticoagulants, DOACs, stroke prevention, UK primary care, CPRD, cohort study, health disparities, warfarin, apixaban, epidemiology

The scale of the underlying dataset is what sets this analysis apart. The investigators assembled a cohort of every individual registered with the Lombardy Regional Health Service as of January 2020, aggregating monthly data by sex, age, country of origin, municipality or ZIP code area, and pre-existing comorbidities. The region’s eight local public health authorities, which administer surveillance and prevention across distinct catchment areas, contributed records that allowed the researchers to track three separate outcomes: confirmed COVID-19 incidence, hospital admissions, and mortality. Crucially, the team also had access to SARS-CoV-2 lineage data from the regional genomic repository, which enabled them to divide the study period into four sub-periods defined by the predominant viral variants circulating at the time. This variant-based segmentation meant the researchers could ask not only how the epidemic changed over time, but how the relationship between risk factors and outcomes itself evolved as the virus mutated.

The demographic profile of the population under study helps explain why Lombardy was hit so hard. Of the more than ten million residents included in the cohort, 45 percent were aged 50 or older, and 22 percent carried a history of cardiac disease. These figures describe a region with a substantial reservoir of older adults and cardiovascular vulnerability, precisely the characteristics epidemiologists associate with severe COVID-19 outcomes. The study’s multivariable regression models, run separately for each of the four variant-defined sub-periods, confirmed that advancing age remained the dominant risk factor for both infection and death throughout the entire pandemic. Neurologic and renal comorbidities also emerged as consistent predictors of mortality, alongside cardiac conditions, underscoring that the burden of severe disease fell disproportionately on people with chronic organ dysfunction.

One of the most striking findings concerns the geography of the early epidemic. Contrary to the intuitive expectation that a respiratory virus would ignite first in the largest urban centers, the outbreak in Lombardy was initially concentrated in a handful of Local Health Districts, none of which contained large cities. This observation aligns with the now well-understood story of the virus’s silent spread through smaller provincial towns in the Po Valley before the first cases were even recognized. The geographic heterogeneity persisted throughout the study period: incidence, hospitalization rates, and mortality varied substantially across districts, and the researchers found that geographical location itself remained an independent predictor of outcomes even after accounting for age, sex, comorbidities, and other individual-level characteristics.

That persistent geographic signal carries an important interpretive message. Because the models adjusted for measurable individual risk factors, the residual effect of location hints at the influence of local epidemic dynamics, such as the timing and intensity of transmission waves in each community, and of health system factors, including differences in testing capacity, care pathways, and hospital load between districts. The authors suggest that areas impacted early in the pandemic experienced distinct trajectories compared with those hit later, and that these local dynamics shaped outcomes in ways that individual-level data alone cannot capture. In practical terms, the same person’s risk of severe outcomes depended not only on their age and health status but also on where and when they lived through the pandemic, a finding with clear implications for how regional health systems should allocate resources during future epidemic emergencies.

The temporal trends documented in the study reveal a pattern that public health officials observed qualitatively but had rarely been able to quantify so systematically. As overall SARS-CoV-2 transmission increased in the population, driven by successive waves of more transmissible variants and eventually by widespread immunity from infection and vaccination, hospital admissions and mortality declined over time. The decline was especially pronounced in areas that had been impacted early, such as the districts at the epicenter of the first wave, where the combination of prior infection, vaccination coverage, and adapted clinical management appears to have progressively decoupled case numbers from severe outcomes. The researchers also tracked test positivity rates and vaccination coverage across the local public health authorities and districts, providing a parallel account of how diagnostic intensity and immunization campaigns varied geographically and interacted with the epidemic’s course.

Methodologically, the study exemplifies the power of routinely collected administrative health data when linked at population scale. Rather than relying on sampled surveys or voluntary reporting, the team exploited the complete registry of Regional Health Service beneficiaries, pseudonymized according to the ISO 25237:2017 standard by each health protection agency before analysis. Only fully anonymized, aggregated data were made available to the study group, a design that satisfied Italian data protection regulations and the EU General Data Protection Regulation while preserving the statistical power of a true total-population cohort. The multivariable regression framework, applied within each variant-defined sub-period, allowed the investigators to estimate how the hazard of infection and death associated with each risk factor changed as the virus and the population’s immunity evolved, a level of temporal resolution that few previous studies have achieved.

The findings also carry lessons that extend well beyond Lombardy. The confirmation that neurologic and renal comorbidities predict mortality, in addition to the more widely recognized cardiac and pulmonary conditions, refines the clinical risk stratification that guided vaccination prioritization and protective measures during the pandemic. The demonstration that geographic location independently shapes outcomes even in a region with a single, relatively uniform health system suggests that epidemic management cannot rely on one-size-fits-all policies; sub-regional surveillance and locally tailored interventions matter. And the observation that severity declined as transmission rose, reflecting the transition from a naive population facing a lethal novel virus to an immunized population facing endemic circulation, provides a quantitative template for anticipating how future respiratory pathogens might behave as they make the same transition.

Lombardy’s experience was in many respects the harshest test any European health system faced in 2020, when hospitals in Bergamo, Brescia, and Cremona became global symbols of the pandemic’s toll. This new analysis converts that traumatic experience into structured evidence, mapping exactly where and when the virus struck, who was most vulnerable, and how the determinants of severe disease shifted across the ancestral, Alpha, Delta, and Omicron eras. The work was funded by Fondazione Cariplo as part of a broader initiative on post-COVID syndrome research, and it forms part of the PASCNET network’s effort to build lasting analytical capacity from the pandemic’s data legacy. As the authors conclude, the findings contribute essential insights into the dynamics of COVID-19 in one of Europe’s earliest and most severely affected regions, evidence that will inform pandemic preparedness, health system planning, and epidemiological modeling for years to come. For a region that served as the pandemic’s European ground zero, the three-year map now drawn represents both a record of what was endured and a guide for what may come.

Subject of Research: Population-based epidemiological mapping of COVID-19 incidence, hospitalization, and mortality in Lombardy, Italy, from 2020 to 2022

Article Title: Mapping the COVID-19 epidemic in Lombardy, Italy, years 2020–2022: a population-based study of temporal trends, geographic distribution, and risk factors for incidence, hospitalization, and mortality

Article References: Villa, S., Magnoni, P., Mazzali, C., Zucchi, A., Maifredi, G., Cavalieri d’Oro, L., Gambino, M. L., Fanetti, A. C., Perotti, P. G., Villa, M., Valsecchi, M. G., Vigani, D., Lucifora, C., Pregliasco, F. E., Cereda, D., Leoni, O., Russo, A. G., the PASCNET study group, Franzoni, F., … Zuccaro, V. (2026). Mapping the COVID-19 epidemic in Lombardy, Italy, years 2020–2022: a population-based study of temporal trends, geographic distribution, and risk factors for incidence, hospitalization, and mortality. BMC Public Health. https://doi.org/10.1186/s12889-026-29682-2

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29682-2

Keywords: COVID-19, Lombardy, SARS-CoV-2, epidemiology, spatial analysis, risk factors, hospitalization, mortality, variants of concern, vaccination coverage, public health, population-based study

NAFLD has become one of the most common chronic liver conditions in the world, affecting roughly a third of adults globally, with prevalence climbing steeply over the past two decades. The disease, characterized by the accumulation of fat in liver cells in people who drink little or no alcohol, is classically tied to obesity, insulin resistance, and metabolic syndrome. But those familiar culprits explain only part of the picture, and scientists have increasingly looked toward environmental exposures for missing pieces. Benzene was a natural candidate. Earlier work had linked low-level ambient benzene to elevated mortality in a national English cohort, and laboratory studies had shown that the chemical disrupts mitochondrial function, impairs fatty acid oxidation, and reprograms lipid metabolism, disturbances that closely mirror the core pathology of fatty liver disease.

The new investigation, published in iScience, harnessed the UK Biobank, a prospective cohort of roughly half a million community-dwelling adults recruited across 22 assessment centers since 2006. After exclusions, 401,472 participants entered the primary analysis, followed for a median of 12.4 years, during which 5,848 developed incident NAFLD. To estimate each participant’s exposure, the team linked geocoded home addresses to annual benzene concentrations generated at a one-kilometer resolution by the UK Department for Environment, Food and Rural Affairs atmospheric dispersion and chemical transport model, an exposure surface validated against ground monitoring with a coefficient of determination of about 0.82. For people who moved during follow-up, the researchers calculated time-weighted average concentrations across each address, a refinement that reduces exposure misclassification.

The headline result was striking. For each interquartile-range increment in benzene concentration, just 0.151 micrograms per cubic meter, the risk of developing NAFLD rose by 26 percent in fully adjusted Cox proportional hazards models, an association with a p-value on the order of 10 to the minus 40. When participants were sorted into exposure quartiles, those in the highest quartile faced a 54 percent higher risk than those in the lowest. The relationship was not strictly linear. A two-piecewise threshold analysis identified an inflection point near 0.3 micrograms per cubic meter, below which the association was modest and above which it grew markedly steeper, hinting that the liver’s defenses may be progressively overwhelmed as exposure climbs past a certain level.

Robustness was tested from multiple angles. The association held when death was treated as a competing event, when NAFLD was redefined using a combination of the hepatic steatosis index and hospital coding, and when the analysis was restricted to participants who had lived at the same address for at least ten years. Excluding early follow-up cases to guard against reverse causation changed little. In multipollutant models adjusting for nitrogen dioxide and fine and coarse particulate matter, the benzene effect was attenuated but remained significant, suggesting the chemical carries an independent signal even within the complex cocktail of urban air pollution. Subgroup analyses across ethnicity, smoking status, physical activity, and body mass index showed consistent directions of effect, with slightly stronger associations among current drinkers and people with diabetes.

Genetics entered the picture through a polygenic risk score built from 31 independent variants identified in large genome-wide association studies, including well-known NAFLD loci such as PNPLA3, TM6SF2, GCKR, and MBOAT7. Each standard-deviation increase in the score was associated with a 36 percent higher risk of NAFLD, and participants in the highest genetic-risk tertile faced a 44 percent greater risk than those in the lowest. Most provocatively, the joint analysis revealed that people carrying both high genetic risk and high benzene exposure had the highest incidence of all, a 73 percent elevation compared with the low-risk, low-exposure reference group. Formal tests showed the interaction departed significantly from additivity, with a relative excess risk due to interaction of 0.27, indicating that benzene and genetic susceptibility act synergistically rather than simply stacking their effects.

The mechanistic centerpiece of the study, however, was biological aging. The researchers computed four complementary aging indices from baseline blood biomarkers: PhenoAge, which integrates nine clinical markers through a Gompertz regression model; homeostatic dysregulation, which measures how far multiple physiological parameters deviate from a healthy reference distribution using Mahalanobis distance; allostatic load, a composite of twelve biomarkers reflecting cumulative physiological stress; and accelerated biological aging, the residual of biological age regressed on chronological age. Mediation analysis with 5,000 bootstrap resamples showed that accelerated biological aging mediated about 6.5 percent of the benzene-NAFLD association, with homeostatic dysregulation and allostatic load contributing smaller but directionally consistent shares of roughly 3.1 and 2.5 percent. PhenoAge alone showed no stable mediating effect.

That aging pathway is biologically plausible, and the surrounding literature helps explain why. Benzene is metabolized in the liver by the enzyme CYP2E1 into reactive intermediates such as benzoquinone and hydroquinone, which generate reactive oxygen species and erode mitochondrial antioxidant defenses. The resulting oxidative injury activates inflammatory signaling through NF-kappa-B and p38-MAPK, promotes lipid peroxidation, and disturbs hepatic energy metabolism. Occupational studies have shown that highly exposed workers carry epigenetic clocks averaging about 2.1 years ahead of their chronological age, along with shortened leukocyte telomeres and aberrant methylation at immune and metabolic gene sites. In mice, twelve weeks of low-dose benzene inhalation raised hepatic inflammatory cytokines more than twofold, cut mitochondrial membrane potential by 35 percent, and nearly doubled malondialdehyde, a marker of lipid peroxidation. Accelerated hepatocyte aging, in turn, is accompanied by impaired fatty acid oxidation and compensatory upregulation of fat-synthesis pathways, a recipe for lipid accumulation.

The team also pushed the findings toward prediction and external consistency. Among ten machine-learning algorithms, a gradient boosting machine performed best, achieving an area under the receiver-operating-characteristic curve of 0.753 in a held-out validation set, with Shapley additive explanations ranking PhenoAge acceleration as the single most important predictive feature, ahead of sex, body mass index, and benzene exposure itself. A nomogram derived from the model offers individualized risk estimation. In an independent analysis of 1,681 participants from the US National Health and Nutrition Examination Survey, higher urinary trans,trans-muconic acid, a benzene exposure biomarker, was associated with greater odds of hepatic steatosis, though the authors caution that this supportive result should not be regarded as definitive replication given differing exposure metrics and outcome definitions.

The study’s limitations are worth keeping in view. Benzene in ambient air travels with other volatile organic compounds such as toluene and xylene, whose contributions could not be fully disentangled, and NAFLD was ascertained primarily from hospital coding, which may miss early or non-hospitalized cases. Alcohol intake was captured through self-report, and the UK Biobank’s famously healthy volunteers may temper generalizability. Even so, the scale of the cohort, the rigor of the sensitivity analyses, and the convergence of genetic, aging, and exposure data make a compelling case that benzene deserves a place among the environmental risk factors for metabolic liver disease. The findings also carry a broader message for environmental health: aging-related pathways, not just traditional metabolic measures, may be essential to understanding how everyday chemical exposures inscribe themselves on long-term health, and air quality standards may need to consider the liver alongside the lungs and the heart.

Subject of Research: The association between long-term ambient benzene exposure and incident non-alcoholic fatty liver disease, mediated by accelerated biological aging.

Article Title: Long-term ambient benzene exposure increases incident NAFLD risk through accelerated biological aging

Article References: Li, Y., Peng, Z., Zhang, J., Cao, H., & Zhang, J. (2026). Long-term ambient benzene exposure increases incident NAFLD risk through accelerated biological aging. iScience, 29(10), Article 117035. https://doi.org/10.1016/j.isci.2026.117035

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117035

Keywords: benzene, NAFLD, fatty liver disease, air pollution, biological aging, UK Biobank, polygenic risk score, epigenetics, oxidative stress, mitochondrial dysfunction, environmental health, epidemiology

The study was led by PhD student Maria Östman of the University of Oslo and colleagues, who exploited one of the fastest and most complete vaccine introductions in Europe as a natural experiment. When Norway added the monovalent Rotarix vaccine to its national childhood immunisation programme in September 2014, uptake among infants reached 92.8 percent within a short period. That near-total and abrupt shift created a clean dividing line in the population: children born before the introduction had essentially no rotavirus vaccination in infancy, while those born after it were overwhelmingly vaccinated. Such a sharp exposure boundary is rare in epidemiology, and it allowed the researchers to compare diabetes incidence across birth cohorts with unusual clarity.

Rotavirus is a highly common virus transmitted by the faecal–oral route and a leading cause of severe gastroenteritis in young children, particularly those under five years of age. Immunity builds with each infection, so repeat infections tend to be milder, but first infections can be dangerous. Rotavirus vaccines, which are oral live-attenuated products made from a laboratory-weakened virus, have dramatically reduced hospitalisation and mortality from severe rotavirus disease worldwide. Because viral infections have long been suspected of influencing the risk of type 1 diabetes — an autoimmune disease in which the immune system destroys the insulin-producing beta cells of the pancreas — researchers have debated whether rotavirus vaccination might either raise or lower diabetes risk. Previous studies in other countries have produced conflicting results, with some suggesting a protective effect and others finding no association.

To resolve the question in the Norwegian setting, the researchers followed all children born in Norway between 2007 and 2019, a total of 740,744 children, from the age of six months to five years. Crucially, they ensured that the length of follow-up was identical for children born before and after the vaccine’s introduction, eliminating a common source of bias in cohort comparisons. Data collection continued until 31 December 2024. Among the full cohort, 846 children were diagnosed with type 1 diabetes before their fifth birthday. The researchers then applied an interrupted time series analysis, a statistical technique that uses the observed pre-vaccination trend to predict what incidence would have been without the intervention, and compares that prediction with the actual observed trend afterwards.

The main analysis showed a relative increase of 11 percent in type 1 diabetes incidence among children born after the vaccine’s introduction compared with the predicted trend. In a separate adjusted analysis, the relative hazard ratio for children fully vaccinated with two doses of the monovalent vaccine, compared with unvaccinated children, was 1.17 — an apparent elevation that, however, did not reach statistical significance. The authors summarised their central finding bluntly: the study found no evidence for a reduced risk of type 1 diabetes after the introduction of rotavirus vaccination in infancy, but instead found that incidence before age five increased after vaccination was introduced. Taken at face value, that combination of results could have fuelled concerns about the vaccine’s safety.

It is here that the study’s secondary analyses become decisive. To test whether the observed increase was genuinely attributable to the rotavirus vaccine, the researchers examined what happened to diabetes incidence after other, unrelated changes to the Norwegian immunisation programme. These served as negative controls: there is no biological or epidemiological evidence that the introduction of the hepatitis B vaccine in 2016, or the switch from the 7-valent to the 13-valent pneumococcal vaccine in 2011, would affect type 1 diabetes risk. If the association seen after the rotavirus rollout reflected a true vaccine effect, similar associations should not appear around these unrelated programme changes. Yet they did. The incidence of type 1 diabetes among children born after the hepatitis B vaccine’s introduction rose by a magnitude similar to that seen after the rotavirus rollout, and the same pattern followed the pneumococcal vaccine switch.

The authors interpret this convergence as strong evidence that the initial 11 percent finding is probably not related to the rotavirus vaccine at all, but rather to other, as yet unknown changes in the environment affecting Norwegian children. Type 1 diabetes incidence has generally been increasing over recent decades, with year-to-year fluctuation, and the researchers emphasise that they do not believe any of the changes in the national vaccination programme, individually or in combination, explain the rise. As they put it, despite their best efforts to design the study to remove the influence of other unknown factors, the identical pattern appearing around every programme change — including changes hypothesised to have opposite or null effects — points to unidentified environmental factors operating across the period.

The team also systematically considered alternative explanations for the observed increase. The COVID-19 pandemic, which was followed in many countries by a temporary surge in paediatric type 1 diabetes diagnoses, was one candidate; Norway itself experienced a temporary increase in incidence in 2022 followed by a decrease. However, a recent study of Swedish and Norwegian adolescents found no increase in type 1 diabetes incidence after COVID-19 infection, weakening that hypothesis. The researchers also examined a change in Norwegian prenatal care guidelines in June 2018, which began recommending measurement of serum ferritin concentration in the first trimester of pregnancy to determine the need for iron supplementation, replacing the less sensitive measurement of haemoglobin at week 30. Because high iron levels have been suspected of association with type 1 diabetes, the change warranted scrutiny. The authors concluded it was unlikely to be relevant, noting that iron levels are usually low in pregnancy and that only children born in 2019 within the study window were affected by the new recommendations.

The authors are careful about the limits of their conclusions. They state that they do not believe rotavirus vaccination increases the risk of type 1 diabetes, but they also argue that additional studies of rotavirus vaccination and type 1 diabetes are needed in different settings and with different vaccines before any definitive conclusion is drawn. This matters because vaccine formulations differ across countries: Norway and the United Kingdom use the monovalent Rotarix vaccine alone, Germany and Italy use both Rotarix and the pentavalent RotaTeq vaccine, while Finland and Iceland implemented RotaTeq alone. Replication across these different exposure profiles would help determine whether the Norwegian findings hold elsewhere or are artefacts of local circumstances. The authors also stress the broader need for research into other environmental factors that could be associated with type 1 diabetes, given that the disease’s incidence continues to climb in many populations without a fully established cause.

For the public, the practical message is one of reassurance grounded in methodological rigour. Rotavirus vaccination has efficiently reduced hospitalisation and mortality from severe rotavirus infections in children, and the Norwegian data, once subjected to negative-control scrutiny, do not support the idea that the vaccine drives diabetes risk in early childhood. The study instead illustrates a subtle pitfall of interrupted time series designs: when a long-running upward trend in a disease coincides with a policy change, the change can appear causal even when it is not. By testing the same pattern against interventions with no plausible link to diabetes, the Oslo team showed that the 11 percent rise was more plausibly a marker of some other environmental shift affecting children born from September 2014 onward. The authors declare no conflict of interest, and their work stands as a model of how vaccine safety signals should be interrogated — thoroughly, transparently, and with controls designed to catch spurious associations before they harden into public fear.

Subject of Research: Association between rotavirus vaccination and type 1 diabetes incidence in Norwegian children

Article Title: Nationwide study in Norway shows a relative 11% increase in incidence of type 1 diabetes in children aged 0.5 to 5 years after nationwide introduction of rotavirus vaccine, but secondary analyses suggest vaccine probably not the cause

Article References: Nationwide study in Norway shows a relative 11% increase in incidence of type 1 diabetes in children aged 0.5 to 5 years after nationwide introduction of rotavirus vaccine, but secondary analyses suggest vaccine probably not the cause. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: type 1 diabetes, rotavirus vaccine, Rotarix, Norway, EASD, Diabetologia, interrupted time series, childhood immunisation, epidemiology, negative controls, autoimmune disease, vaccine safety

The scientific literature documenting this field has grown explosively. Over the past two decades, publications on epigenetic epidemiology have multiplied in both number and variety, and the scope of the research now extends well beyond the DNA methylation markers that once dominated the field. Histone modifications and non-coding RNAs have joined the toolbox, and the study designs on display range from epigenome-wide association studies, commonly known as EWAS, to candidate gene studies and clinical trials. Epigenetic markers appear in the literature in two distinct roles. Sometimes they are investigated as risk factors, as in studies of DNA methylation patterns associated with the incidence of type 2 diabetes. Sometimes they are investigated as outcomes, as in research documenting methylation changes in response to air pollution. Keeping track of this expanding and heterogeneous body of work has become a genuine challenge, and that challenge motivated a team of researchers affiliated with the Centers for Disease Control and Prevention and Emory University to build a purpose-made solution.

The result is the Epigenetic Epidemiology Publications Database, or EEPD, a publicly available, web-based application described in a recent correspondence published in Epigenetics Communications. The database is designed to offer a user-friendly way to explore the literature at the intersection of epigenetics, epidemiology, and public health. The scope of the collection is carefully defined. The developers aimed to include population-based studies of epigenetics in association with an exposure or outcome of interest, systematic reviews related to epigenetic associations, and published descriptions of epigenetic resources such as cohorts, protocols, and methods. Just as important is what the database excludes: articles focused on gene expression, including transcriptomics, non-population-based laboratory studies, non-human studies, and non-systematic reviews are all filtered out. That boundary keeps the collection focused on the population perspective that defines the field.

The technical machinery behind the database is where the project becomes genuinely interesting for anyone following the application of machine learning to scientific curation. Manually screening the biomedical literature is notoriously labor-intensive, and the volume of relevant publications makes hand-curation impractical at scale. The team therefore developed an automatic screening process that begins with a PubMed query and then applies a machine learning method, a support vector machine, combined with text mining techniques. Support vector machines are supervised learning algorithms that classify items into categories after being trained on labeled examples, and they have a long track record in text classification tasks. To evaluate how well their pipeline performed, the researchers assessed it against a random sample of 1,999 manually annotated abstracts. Of these, 285 positive records and 114 negative records were reserved for testing. The results were striking: the screening process achieved an estimated area under the curve of 96.4 percent, with 90 percent sensitivity and 90 percent specificity. In practical terms, the automated system caught the overwhelming majority of relevant articles while correctly discarding most irrelevant ones, a level of performance that rivals careful human screening at a fraction of the effort.

Once the screening pipeline is in place, the database essentially runs itself. Records are automatically selected and uploaded on a daily basis, which means the collection stays current without a team of curators reading every new abstract. Each record is indexed with genes, diseases, and other factors, creating a structured layer of metadata on top of the raw publication records. The system also automatically subsets the data according to 14 special topics, ranging from cancer and diabetes to environmental health, health equity, infectious diseases, neurological disorders, pharmacogenomics, rare diseases, and reproductive and child health. These subsets are generated by searching the database with topic-relevant terms, so a researcher interested specifically in, say, heart, lung, blood and sleep disorders can work within a curated slice of the literature rather than the full collection.

The payoff from all this automation is best illustrated by a direct comparison the authors conducted. On September 28, 2023, a search for cancer-related epigenetic epidemiology on EEPD returned 8,942 records. A search on PubMed using the terms cancer, epigenetics, and epidemiology returned only 1,198 records, a dramatic undercount that reflects how often relevant studies fail to use all three terms in their indexing. Removing the epidemiology term from the PubMed search pushed the result to 17,927 records, but at the cost of flooding the searcher with irrelevant material. The EEPD sits in the sweet spot between those two failures, capturing relevant work that conventional keyword searches miss while excluding the noise that broader searches admit. For researchers conducting literature reviews, that difference is not a convenience; it can determine whether a review is comprehensive or fundamentally incomplete.

Using the database is deliberately simple. Users type a query of interest into a search box at the top of the page, and results are returned as a list of related publications. From there, filtering functions allow progressive refinement using indexed factors including gene, disease, publication year, publication journal, and study design. Filters can be applied continuously until the desired result set is achieved, and the final results can be downloaded as an Excel spreadsheet with a single click on the download button. Alternatively, users can subset the entire database to one of the 14 special topics through a dropdown menu in the search bar, and all of the navigation functions remain available within that subset. The design philosophy is clear: the database should spare investigators from having to construct complicated PubMed queries while still offering the precision that expert searchers demand.

The scale of the collection as of August 18, 2023, underscores how much material the automated pipeline has already accumulated. The database contained 21,699 PubMed records, indexed against 9,378 genes and 1,019 disease terms. The annual number of publications has increased markedly since 2000, reflecting the field’s rapid maturation. Cancer dominates the collection as the most studied topic, followed by rare diseases. The five most frequently indexed disease terms are breast neoplasm, colorectal neoplasms, squamous cell carcinoma, lung neoplasm, and stomach neoplasms, a roster that mirrors the prominence of oncology in epigenetic research more broadly. On the gene side, the five most frequently indexed entries are CDKN2A, RASSF1, DNMT1, MGMT, and CDH1, names that will be immediately familiar to anyone who has followed methylation research in tumor biology, since several of these genes are classic targets of epigenetic silencing in cancer.

The authors are candid about the limitations of their tool. The sensitivity of EEPD is not perfect, and the search algorithms and indexing processes of both PubMed and the database have inherent constraints in a field that is rapidly evolving and not fully defined. Because EEPD relies exclusively on PubMed as its data source, potentially relevant articles not indexed there are simply not included, which means the database cannot claim exhaustive coverage of the global literature. These caveats matter, and the authors do not shy away from them. Even so, they argue that EEPD can serve as a valuable complement to traditional search methods for literature review, offering investigators an efficient way to gain an overview of available research on a particular topic as well as to identify articles relevant to their own projects without labor-intensive screening.

The roadmap for the future is equally pragmatic. The authors hope to expand the data sources feeding the database, drawing on resources such as Scopus and Embase to capture material that PubMed misses. Further tuning of the PubMed search query and of the database’s own algorithms could enhance both the sensitivity and specificity of data collection and retrieval, pushing the already strong performance metrics even higher. For a field that sits at the crossroads of epigenetics, epidemiology, and public health, and that produces new publications at an accelerating pace, tools like EEPD may prove essential infrastructure. The promise of epigenetic epidemiology, linking the environments and experiences of populations to the molecular marks that shape disease risk, depends on researchers being able to find, synthesize, and build upon the work of their predecessors. By automating the discovery of that work, the Epigenetic Epidemiology Publications Database aims to make sure that no important study is lost in the sheer volume of an exploding literature.

Subject of Research: A machine learning-curated public database for navigating epigenetic epidemiology literature

Article Title: Navigating epigenetic epidemiology publications

Article References: Yu, W., Drzymalla, E., Gyorfy, M. F., Khoury, M. J., Sun, Y. V., & Gwinn, M. (2023). Navigating epigenetic epidemiology publications. Epigenetics Communications, 3(1), Article 8. https://doi.org/10.1186/s43682-023-00023-3

Image Credits: AI Generated

DOI: 10.1186/s43682-023-00023-3

Keywords: epigenetics, epidemiology, DNA methylation, machine learning, support vector machine, PubMed, database, EWAS, cancer, public health, text mining, literature curation

Social determinants of health encompass the conditions in which people are born, live, work and age, including income, education, housing, neighborhood environment, social connection and psychosocial stress. These factors are increasingly recognized in epidemiology as powerful modifiers of disease outcomes, yet they are often examined one at a time, which can obscure the reality that disadvantages tend to cluster within the same individuals and communities. To address this, the Tulane team used an elastic-net weighting approach, a statistical technique that can handle many correlated variables simultaneously and assign each social indicator a weight reflecting its contribution to the overall score. Participants were then sorted into favorable, medium and unfavorable categories, allowing the investigators to test whether graded levels of cumulative social burden corresponded to graded differences in dementia incidence over time.

The primary analysis drew on the UK Biobank, one of the world’s largest prospective biomedical cohorts, and included 10,071 participants with type 2 diabetes and 11,780 euglycemic participants. Over approximately fourteen years of follow-up, 247 participants with diabetes and 149 euglycemic participants developed dementia. Among people with type 2 diabetes, unfavorable social determinants, compared with favorable ones, were associated with roughly a 2.4-fold higher hazard of dementia after adjustment for a broad set of covariables, including age, race, sex, smoking status, alcohol consumption, physical activity, total cholesterol, body mass index, hemoglobin A1C, depression and an Alzheimer’s disease polygenic risk score. A similar overall association emerged among euglycemic participants, indicating that cumulative social disadvantage matters for cognitive outcomes even in the absence of diabetes.

The most arresting result, however, concerned sex differences. Among men with type 2 diabetes, unfavorable versus favorable social determinants were associated with a 1.7-fold higher dementia risk. Among women with the same disease, the corresponding figure was 5.35-fold, and the statistical test for interaction confirmed that the association between social burden and dementia differed significantly between men and women. This pattern suggests that women with type 2 diabetes may bear a disproportionate cognitive cost when social circumstances are adverse. The authors note that their findings support the evidence that women with type 2 diabetes may be especially vulnerable to the cognitive consequences of social determinants of health, a conclusion with potential implications for how cognitive risk is assessed and how diabetes care is tailored.

The interplay between diabetes and social environment also showed up when the researchers compared people with and without diabetes within the same social strata. Among participants in the medium or unfavorable social determinant categories, those with type 2 diabetes had nearly three times the hazard of dementia compared with euglycemic participants in the same strata. Among women in these categories, diabetes was associated with approximately a 3.1-fold higher risk compared with euglycemic women. In other words, the metabolic disease carried a steeper cognitive penalty in socially disadvantaged settings than it did overall, a pattern consistent with the idea that social adversity and metabolic dysfunction can interact rather than merely add together.

To guard against the possibility that the UK Biobank findings were an artifact of a single population or health system, the team replicated key analyses in the United States All of Us Research Program, a large and diverse national cohort. Similar patterns for both dementia and mild cognitive impairment were observed, lending weight to the conclusion that the association between cumulative social disadvantage and cognitive risk in diabetes is not confined to one country or one set of measurement conventions. Replication across two independently constructed cohorts with different demographic compositions is an important strength in observational research of this kind, where confounding and measurement error are persistent concerns.

Beneath the aggregate score, several individual social indicators stood out. Among UK Biobank participants with type 2 diabetes, dementia was associated with household income, area-level education deprivation, home ownership, loneliness and perceived stress. Household income and loneliness were linked to dementia risk in both men and women, with stronger associations among women, while perceived stress was additionally associated with dementia risk in women with type 2 diabetes. These component-level findings hint at mechanisms: loneliness and chronic stress are known to influence hypothalamic-pituitary-adrenal axis activity, inflammation and health behaviors, while income and housing stability shape access to care, nutritious food and safe opportunities for physical activity.

The authors outline several plausible biological and behavioral pathways that could connect social disadvantage to cognitive decline in diabetes. Social adversity can restrict access to healthcare, healthy food, physical activity and medication adherence, and it can undermine diabetes self-management while amplifying psychosocial stress. These pressures may interact with the vascular and metabolic abnormalities that accompany type 2 diabetes, including microvascular disease, endothelial dysfunction, chronic inflammation and insulin resistance, all of which have been implicated in neurodegeneration and cerebrovascular injury. The brain’s vulnerability to metabolic disease may therefore be compounded when the social environment fails to buffer the demands of managing a chronic condition. The researchers are careful to emphasize, however, that these are observational findings and do not establish that adverse social determinants directly cause dementia.

The study also carries important limitations that temper interpretation. Dementia and mild cognitive impairment were identified using diagnostic codes, a method that can produce misclassification and can miss undiagnosed cases, particularly among people with limited healthcare access, which could itself bias estimates. The social determinant measures differed between the UK Biobank and All of Us and required harmonization, neither cohort fully represents the general population, and residual confounding cannot be excluded despite extensive statistical adjustment. In addition, the neuroimaging and cognitive measures available to the researchers were cross-sectional, which limits the ability to draw mechanistic conclusions about how social conditions translate into brain changes over time.

Looking forward, the authors propose that future studies incorporating longitudinal brain imaging, repeated cognitive assessments, inflammatory markers, proteomics and metabolomics could help clarify the biological routes by which social disadvantage relates to cognitive decline. For now, the findings carry a clear practical message: cumulative social conditions are associated with dementia risk and may be especially consequential for people living with type 2 diabetes, and the markedly stronger association observed among women suggests that both social circumstances and sex deserve a place in cognitive risk assessment and diabetes care. Whether interventions that address adverse social determinants, from income support to loneliness reduction, can actually lower dementia risk remains an open question that only targeted trials and long-term follow-up studies can answer.

Subject of Research: The association between adverse social determinants of health and dementia risk in people with and without type 2 diabetes

Article Title: Adverse social determinants linked to higher dementia risk in people with type 2 diabetes

Article References: Adverse social determinants linked to higher dementia risk in people with type 2 diabetes. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: dementia, type 2 diabetes, social determinants of health, UK Biobank, All of Us Research Program, sex differences, cognitive decline, loneliness, household income, epidemiology, aging, Tulane University