Across the globe, rates of depression and anxiety have risen substantially since the COVID pandemic. Consequently, poor mental health is now a top health policy priority in many countries and more people than ever are seeking treatment. While the segment of people with poor mental health is large and growing, there is a dearth of data about their demographics and health needs and their use of and experience in the health system. Health systems require this information to effectively organize and provide services. We investigated population prevalence of fair or poor mental health and compared health system experience and quality of care among adults with poor versus good mental health in 18 high-, middle-, and low-income countries using data from the People's Voice Survey (n = 32,419). Data were collected in 2022 and 2023 through a combination of nationally representative telephone, online, and in-person surveys. Prevalence of self-reported poor mental health ranged from 4.7% in Nigeria to 39.6% in China and was unrelated to national income per capita. More women than men reported poor mental health in most countries. Across all countries, people with poor mental health had worse self-rated overall health and more chronic illness. Between 0.9% (Lao PDR) and 52.4% (UK) of those with poor mental health had received mental healthcare in the past year. People with poor mental health reported lower patient activation, worse care quality, and lower confidence in the health system. A study limitation is that results are based on self-reported mental health rather than clinical diagnoses. People with poor mental health have markedly different health profiles and health system experience. These findings should prompt health systems to re-assess their services to better serve this growing patient group. Comparison of user experience and quality over time and across countries with similar health systems may assist in benchmarking performance.

Magnesium sulphate halves the risk of eclampsia. There is no consensus on who to give magnesium sulphate prophylaxis because clinical tools are poor at identifying those at risk. Known prodromal symptoms such as headache, visual disturbance, or epigastric pain have modest associations with eclampsia. We set out to identify new prodromal symptoms of eclampsia. This case-control study prospectively recruited participants in South Africa and Pakistan who had eclampsia, preeclampsia, or normotensive pregnancies. We asked whether they experienced 20 neurological symptoms, within 7 days of the seizure for those who had eclampsia. The primary analysis was the likelihood of symptoms occurring before eclampsia, compared to being present with preeclampsia. 341 participants were recruited with eclampsia, 1,355 with preeclampsia and 389 with normotensive pregnancies. When comparing symptoms among those who had eclampsia versus preeclampsia, the odds ratios (OR) were 2.56 (95% confidence interval (CI) [1.81,3.62]; p < 0.001) for headache, 5.73 (95% CI [4.44,7.39]; p < 0.001) for visual disturbances and 2.25 (95% CI [1.76,2.89]; p < 0.001) for epigastric pain. We identified 10 symptoms with odds ratios over 10 for eclampsia. Odds ratios were 42.03 (95% CI [23.66,74.68]; p < 0.001) for twitching/jerking limbs (30.5% eclampsia versus 1% preeclampsia); 36.00 (95% CI [18.34,70.65]; p < 0.001) for affected hearing (21.1% versus 0.7%)' 33.60 (95% CI [21.39,52.78]; p < 0.001) for affected mind state (38.7% versus 1.8%); 33.12 (95% CI [19.46, 54.37]; p < 0.001) for impaired speech; 23.71 (95% CI [16.49,34.10]; p < 0.001) for feelings of doom; 26.59 (95% CI [7.82,90.41]; p < 0.001) for severe vertigo; 20.52 (95% CI [14.22,29.63]; p < 0.001) for confusion; 18.16 (95% CI [10.76,30.66]; p < 0.001) for jitters; 15.18 (95% CI [11.34,20.33]; p < 0.001) for difficulty concentrating; and 10.49 (95% CI [6.76,16.27]; p < 0.001) for weakness/paralysis. These symptoms were rare among normotensive pregnancies. Only 2.4% of women with eclampsia did not experience any prodromal symptoms. This study is limited by the fact that we asked about prodromal symptoms after the seizure happened, and the potential for recall bias. Ten prodromal symptoms exhibit far stronger associations with eclampsia than headache, visual disturbances, or epigastric pain. Eclampsia is uncommon without any prodromal symptoms. It may be useful to screen these symptoms among women with preeclampsia as part of clinical history taking to guide management. They could help direct magnesium sulphate prophylaxis to those with a higher risk of eclampsia.

Postnatal care (PNC) plays a crucial role in averting newborn mortality, yet its use remains low, particularly in regions with the highest mortality. While demographic and social determinants of PNC use have been well studied and have informed current strategies focused on changing care-seeking behaviors, the stagnating decline in neonatal mortality highlights the need for upstream and system-wide approaches to increase PNC uptake. Limited evidence exists to guide system-level reforms; therefore, we investigated whether health systems that ensure high-quality perinatal care are associated with increased use of PNC. We performed a cross-sectional observational study using Demographic and Health Survey data from 38 countries that had not met SDG neonatal mortality targets by 2020. The study population comprised women aged 15-49 years whose most recent live birth occurred within five years preceding the survey. We employed logistic regression models with country fixed effects to examine associations between: (1) perinatal service utilization, (2) service quality; and their relationship with postnatal checkups for newborns within 28 days. We analyzed utilization-quality interactions to determine how the effects of service coverage on PNC use varied by care quality and conducted wealth-stratified analyses to assess how quality effects on PNC use differed across socioeconomic groups. High-quality perinatal care was associated with a 2-fold increase in the probability of postnatal checkups within 28 days (0.406 in the highest quality tertile versus 0.221 in the lowest quality tertile). For mothers who received the lowest tertile of service quality, full utilization of perinatal services yielded negligible changes in postnatal checkup probability (0.217 to 0.216). Conversely, for mothers who received the highest quality of antenatal and perinatal care, full access to perinatal care improved the probability of postnatal check-ups (0.392 to 0.428). Notably, women in the lowest wealth quintile experienced a substantial increase in postnatal checkup probability from 0.224 to 0.481 between low and high-quality care cohorts, while this differential was less pronounced in the highest wealth quintile (0.236 to 0.450). Key limitations include restricted quality indicators, potential recall bias from self-reported measures, limited information on follow-up care, and the cross-sectional nature of the data, which limits causal interpretation. Our interaction analysis reveals a critical insight: high-quality care substantially enhanced the magnitude of the association between expanded perinatal service use and PNC use, whereas increased utilization alone was not linked to higher PNC uptake. Notably, the impact of improved care quality was most pronounced among the lowest wealth groups, highlighting its potential as a mechanism for promoting equity. By demonstrating the central role of care quality in promoting PNC utilization, particularly among disadvantaged populations, our findings suggest that improving health system quality could be a more effective strategy for achieving universal maternal healthcare coverage than traditional access-focused approaches.

Long-acting injectable pre-exposure prophylaxis (LAI-PrEP) is a highly effective biomedical intervention for the prevention of HIV acquisition. There is a strong interest among communities and policymakers for LAI-PrEP scale-up, accelerating the demand for clear guidance on testing approaches that balance accuracy with scalability. Unlike oral pre-exposure prophylaxis, LAI-PrEP may overcome adherence challenges, such as difficulty with frequent clinic visits. However, LAI-PrEP results in prolonged subtherapeutic drug levels after discontinuation, which can increase the risk of drug resistance among those who have an undetected HIV infection. This systematic review evaluates how different HIV testing strategies, including rapid diagnostic tests (RDTs), laboratory-based immunoassays and nucleic acid testing (NAT), affect clinical utility and programme delivery of LAI-PrEP. We searched databases and retrieved studies up to April 8, 2025, and supplemented findings with data collected through a World Health Organization (WHO) survey among ongoing and completed LAI-PrEP implementation studies. We included publications reporting original or primary data on clinical, diagnostic and resource-use outcomes of HIV testing for LAI-PrEP. Meta-analyses were conducted using random-effects models. Chi-square tests were used to examine differences between related outcomes. Certainty of evidence was determined using the GRADE methodology (Prospero: CRD42024605562). Risk Of Bias In Non-randomised Studies of Interventions, Version 2 (ROBINS-I V2) assessment tool was used to assess bias for non-randomised comparative studies. Of 7,698 records identified, 38 reports representing 22 studies (cabotegravir: 20, lenacapavir: 2) across 15 countries were included. The overall certainty of evidence was low. Most were observational cohorts (n = 13) or non-randomised comparator studies (n = 7). Among 8,171 LAI-PrEP users in four randomised controlled trials, HIV detection rates were similar across strategies (9/8171 (RDT) versus 14/8171 (NAT) (Odds ratio (OR) 0.66 (95% confidence interval: 0.29-1.50; P = 0.87)), with no difference in adverse events. Compared with laboratory-based tests, RDTs enabled faster turnaround (same-day versus up to 7 days), more rapid treatment initiation (1 day versus 6-9 days), and lower test costs (US$4 versus US$22). All tests had similar negative predictive value (~100%) at LAI-PrEP initiation and comparable positive predictive value (~55%) at continuation. There was little difference in delayed HIV detection (11/8171 (RDT) versus 0/8171 (NAT)). In the HPTN 083 trial, NAT use was occasionally associated with false-positive results, leading to unnecessary PrEP holds or discontinuation (7/2483). NAT might have detected HIV before resistance emerged, though no prospective or modelling evidence showed clinical benefit at a population level. There was limited evidence of HIV self-testing for LAI-PrEP delivery. We noted that our assessment of performance accuracy in different testing strategies may introduce selection bias. RDT-based testing strategies have comparable accuracy to laboratory-based strategies and are more accessible and scalable, which can ensure that testing does not become a barrier to accessing or continuing LAI-PrEP. As countries expand access to LAI-PrEP amid increasingly constrained resources, adoption of new WHO guidance supporting the use of RDTs can enable simpler, more affordable, and user-centred HIV testing approaches.

Dementia, cardiovascular disease (CVD), and functional impairment (FI) often co-occur in aging populations, with abdominal obesity as a shared modifiable risk factor. The long-term impact of abdominal obesity on these comorbidities is unclear. We projected the 30-year burden of dementia, FI, and CVD in China under different trajectories of abdominal obesity prevalence. We modeled three trajectories of abdominal obesity prevalence from 2020 to 2050 using data from the China Health and Nutrition Survey (2000-2015): continuation of the observed growth prevalence trend (persistent), stabilization at 2015 levels (optimal), and a 50% reduction in the growth rate (improved). Abdominal obesity was defined as a waist circumference of ≥90 cm for men and ≥85 cm for women. A Markov model was used to estimate occurrence of dementia, CVD, FI, and mortality among adults aged ≥65 years by sex and year. Under the persistent scenario, dementia cases were projected to rise to 37.8 million (95% uncertainty interval (UI) [36.7, 38.9]) by 2050, alongside 68.1 million (95% UI [66.7, 69.5]) cases of FI, 198.3 million (95% UI [196.5, 200.4]) CVD cases, and 17.6 million (95% UI [16.5, 18.7]) deaths. Compared with the persistent scenario, dementia and FI burdens increased under the optimal and improved scenarios by 2050, by 1396.0 thousand (95% UI [589.1, 2293.9]) and 711.4 thousand (95% UI [294.4, 1150.3]) for dementia, and by 2570.6 thousand (95% UI [1081.0, 4024.9]) and 1289.5 thousand (95% UI [569.5, 2034.2]) for FI, mainly due to reduced CVD mortality expanding the population at risk. These shifts are most pronounced among adults aged ≥80 years and women. For CVD, reductions in the number of cases were projected in the short term (by 2030), but these changes remain uncertain by 2050. Main limitations include the assumption that other risk factors remain unchanged, and the lack of modeling of multiple co-occurring dementia's risk factors. Abdominal obesity control may reduce CVD incidence and mortality, thereby shifting the disease burden toward dementia and FI due to increased longevity, highlighting the need for integrated, life-course public health strategies responsive to the patterns of dementia and its comorbidity in older people.

Clinical guidance in Canada and the United States recommends faster methadone dose titration for individuals using fentanyl. An initial dose increase is recommended between days 4 and 6, but the impact on patient outcomes remains unclear, particularly when provided in an outpatient setting. Therefore, we evaluated the association between early methadone dose titration (i.e., within treatment days 4-6) and treatment discontinuation and opioid toxicity. We conducted a retrospective propensity-score weighted cohort study of Ontario residents who initiated methadone in an outpatient setting between January 2017 and December 2022. Exposure was defined as provision of a dose titration between treatment days 4-6, with the index date defined as the first date of dose increase. For unexposed individuals, the index date was randomly assigned and followed the same distribution as those exposed. Individuals were followed for up to 181 days following index date for study outcomes. Primary outcomes included methadone discontinuation and opioid toxicity, using an intention-to-treat approach. Secondary outcomes included methadone versus non-methadone-related toxicity while on treatment. We fit a propensity-score model to estimate the probability of dose titration, conditional on baseline demographic and clinical (e.g., comorbidities, past medication use) variables. Propensity score weighted Cox proportional hazard models were then estimated to determine the association between early dose titration and study outcomes. Among 13,560 incident methadone recipients (61.4% [N = 8,322] provided early dose titration), the mean age was 36.5 years old, and approximately one-third were female. Individuals who received an early dose titration had a lower weighted hazard of discontinuation, which was most pronounced during the first seven days of follow-up (weighted hazard ratio [wHR]: 0.55; 95% confidence interval [CI]: 0.51, 0.60), with attenuation towards the null over time. The observed weighted rate of opioid toxicity was lower among those exposed versus unexposed (10.1 versus 12.3 per 100 person-years; wHR: 0.82; [95% CI: 0.69, 0.97]). Early dose titration was not associated with opioid toxicity while on treatment. Most toxicities were attributed to non-methadone opioids (exposed: 3.89 per 100 person-years versus unexposed: 4.06 per 100 person-years; wHR: 1.00; [95% CI: 0.70, 1.44]). Methadone-related toxicity while on treatment remained low and comparable between exposed and unexposed groups (2.02 versus 2.65 per 100 person-years; wHR: 0.80; [95% CI: 0.46, 1.41]). The main study limitation is the lack of information on drug use history and time-varying clinical factors that may influence both dose titration and subsequent outcomes, introducing the potential for residual confounding. This population-based analysis demonstrated that early methadone dose titration was associated with improved treatment retention and lower hazard of opioid toxicity. Logistical barriers that hinder timely provision of dose increases must be addressed to improve methadone retention among people who use fentanyl.

The explosion of Mendelian Randomization (MR) submissions of dubious quality to journals globally is well recognized. Contributing to this deluge of publications may be the poor understanding among practitioners, reviewers, and editors of gene-environment equivalence, the fundamental principle of MR.

In this Formal Comment, representatives from PLOS, Nature and JAMA call for action on adopting a principle-based approach for a responsible authorship culture.

Postoperative delirium is the most common postoperative complication in older individuals. Genome-wide association studies (GWAS) can provide insights into how genetic factors influence postoperative risk. We examined the genetic architecture of postoperative delirium after major surgery and its relationship with related cognitive conditions (delirium of any type and Alzheimer's disease, including the APOE ε4 allele). A case-control GWAS was performed in the UK Biobank to identify genetic variants associated with postoperative delirium, adjusted for age, sex, genetic chip, and the first 10 principal components. These results were then used in genetic correlation and polygenic risk score analyses to investigate shared genetic risk between postoperative delirium and a) delirium of all causes, and b) Alzheimer's disease. The GWAS (1,016 cases, 139,148 controls) identified seven Single Nucleotide Polymorphisms (SNPs) that mapped to four genes (APOE, TOMM40, APOC1, and PVRL2); p < 5 x 10-8. Five SNPs remained significant after excluding pre-existing dementia, and two after excluding subsequent dementia. The lead SNP was rs429358, a missense variant of APOE. Genetic correlation and polygenic risk score analyses revealed evidence of shared genetic architecture and risk between postoperative delirium and Alzheimer's disease (rho 0.68, 95% CI [0.46, 0.81]; p < 0.001). After adjustment for age and sex, the APOE ε4 isoform had a dose-response effect on risk (odds ratios for one and two copies: 1.75, 95% CI [1.53, 2.0], and 4.19, 95% CI [3.25, 5.41], respectively; p < 0.001). The main limitations of the study include the reliance upon clinical coding for outcome definition and limited statistical power to detect small or modest genetic effects. We identified genetic variants associated with increased risk of postoperative delirium. We also found evidence of shared genetic liability with Alzheimer's disease via APOE, complementing recent large-scale studies in all-cause delirium. If validated, the findings have potential clinical applications, including preoperative risk stratification and early identification of pre-clinical Alzheimer's disease risk.

Antibody-drug conjugates are redefining treatment options in advanced breast cancer, demonstrating efficacy across all breast cancer subtypes. However, their rapid clinical expansion has resulted in several unresolved challenges, including the need for rational sequencing strategies, appropriate and ethical trial design, drug tolerability, and the limitation of mono-national development programs.