Background Processed meat is classified as a carcinogen and unprocessed red meat a probable carcinogen, but their associations with stomach cancer are not established. We evaluated intakes of these food groups and poultry, seafood, and eggs in relation to stomach cancer risk. Methods Individual-level data for 1 100 370 females and 891 919 males from 20 cohorts were analysed. Diet was assessed at baseline by using food-frequency questionnaires or diet histories generally covering the past year and harmonized across studies. We used Cox regression to estimate multivariable-adjusted study-and sex-specific hazard ratios (HRs) and 95% confidence intervals (CIs) for associations with each food group and combined study-specific estimates by using random-effects meta-analysis. We repeated analyses using an aggregated dataset combining all studies. Results During follow-up (mean = 8-23 years across studies), 8561 invasive gastric adenocarcinomas were diagnosed. Processed meat intake showed weak evidence of an overall association with stomach cancer risk (HR, 1.15 for >= 45 vs <3 g/day; 95% CI, 1.00-1.32 and HR, 1.03 per 25-g/day increment; 95% CI, 0.99-1.07). This association was limited to current smokers (P-interaction = .01) and differed by geographic region, with evidence for an association in Australian and European studies and no association observed in Asian studies (P-interaction = .02). Positive associations with stomach cancer were observed for egg intake; no association was found for unprocessed red meat, poultry, or seafood. Conclusion Higher processed meat intake was associated with a modestly increased risk of stomach cancer among current smokers, although residual confounding by smoking dose remains possible. Findings for eggs warrant further investigation.
Background Loneliness is increasingly recognized as a major public health challenge. Yet, its impact on cognitive function remains poorly understood. A complex interplay of life-course risk factors makes it challenging to disentangle the effect of loneliness from other factors and to assess underlying mechanisms when using traditional approaches.Methods We applied the parametric g-formula with the Health and Retirement Study (1996-2018; n = 34 774; aged 50-80 years) to simulate an intervention estimating the effect of preventing loneliness on cognitive function. We assessed whether this effect varies by gender. Formal mediation analysis was also performed to assess depression as a potential pathway.Results Preventing loneliness was protective against cognitive decline at all ages, with the magnitude increasing with age. Averaged across all ages, an intervention to prevent loneliness would slow cognitive decline by 0.075 points at the population level [95% confidence interval (CI): 0.063-0.089] and by 0.157 points among individuals experiencing loneliness (95% CI: 0.138-0.178). Overall, 17.3% of the total effect of preventing loneliness was mediated via depression, with the proportion mediated being 19.4% among men and 16.9% among women.Conclusion Employing an innovative counterfactual framework, this study shows that preventing loneliness could provide an insulative effect against cognitive decline, even at the population level. Importantly, with advancing age, addressing loneliness yields progressively greater cognitive benefits. The effects partly operate through reductions in depression, especially in men. Interventions against loneliness may help maintain cognitive function, with benefits extending to mental health, highlighting a promising avenue for public health strategies.
Abstract Background Tobacco smoking is a leading preventable cause of cancer. We estimated its attributable cancer burden by smoking status (current vs. former), further stratified by cancer type, sex, country, and Human Development Index (HDI). Methods The Population Attributable Fraction (PAF) for cancer incidence and mortality due to tobacco were estimated across 175 countries by using 2000 Global Burden of Disease smoking prevalence and 2022 Global Cancer Observatory cancer data. Relative risks (RRs) were derived from large pooled and meta-analyses, and population data from the World Bank. Countries were categorized by HDI. Results The global tobacco PAF was 17.2% for cancer incidence and 22.4% for mortality. Lung cancer had the highest RR for current smokers (8.96) and laryngeal cancer for former smokers (4.65). Males had a higher cancer incidence (26.1%) and mortality (32.1%) burden than females (8.0% and 10.2%, respectively). Cancers with the highest smoking PAF were laryngeal (70.1%), lung (66.9%), pharyngeal (63.5%), oral cavity (38.5%), and bladder (39.4%). PAF estimates for incidence were 5.4% in low-HDI (49 countries), 15.0% in medium-HDI [41], 19.3% in high-HDI [32], and 17.4% in very-high-HDI [32] countries; corresponding mortality estimates were 6.4% (low-HDI), 17.4% (medium-HDI), 26.0% (high-HDI), and 23.6% (very-high-HDI), respectively. Conclusion Approximately one in six cancer cases and one in four cancer deaths globally are attributable to tobacco smoking. Despite declining smoking prevalence, former smokers account for a substantial burden long after cessation. Males and higher-HDI countries bear a greater burden.
Abstract Background Menopause marks a key reproductive milestone with significant implications for women’s long-term cardiometabolic health. However, the influence of cardiometabolic health during the reproductive years on the timing of menopause remains poorly understood. We prospectively examined the association between cardiometabolic health profiles and menopause onset in South Asian women, a population at high risk of early cardiometabolic dysfunction. Methods We analyzed longitudinal data from menstruating women aged 30–50 years in the CARRS cohort from 2010 to 2024. Baseline cardiometabolic measures included overweight/obesity (body mass index ≥25 kg/m2), hypertension (systolic blood pressure ≥140 or diastolic blood pressure ≥90 mmHg, prior history, or medication use), diabetes (hemoglobin A1c ≥6.5%, physician diagnosis, or use of glucose-lowering medication), and non-high-density lipoprotein cholesterol (non-HDL-C ≥ 130 mg/dl). Using Cox proportional hazards models, we examined associations between baseline cardiometabolic status and time to natural menopause, adjusting for sociodemographic and reproductive factors. Results Among 2225 women (median age 37 years; IQR: 33, 40), 44% reached natural menopause at a median age of 46 (IQR: 43, 49) years. Women with hypertension reached natural menopause approximately 1 year earlier than those without hypertension. In adjusted models, women with hypertension experienced relatively earlier natural menopause over follow-up (HR 1.34, 95% CI: 1.19–1.52). In contrast, the timing of natural menopause did not differ by baseline diabetes, weight, or non-HDL-C status. Conclusion Hypertension was associated with earlier timing of menopause in South Asian women. Further research is needed to explore underlying mechanisms and long-term health implications.
Abstract Background Preclinical studies suggest perinatal antidepressant exposure may alter pubertal timing, but human evidence is lacking. Objective To investigate whether maternal antidepressant exposure is associated with pubertal timing in human offspring. Study design We used data from the Norwegian Mother, Father, and Child Cohort Study linked with the Medical Birth Registry of Norway, including 1396 singleton births to women with depression/anxiety before pregnancy. Group-based trajectory modeling was used to identify four antidepressant exposure trajectories: continuers (n = 49, 3.5%), early discontinuers (n = 140, 10.0%), late discontinuers (n = 51, 3.7%), and non-exposed (n = 1156, 82.8%). The Pubertal Development Scale with five items assessed at 14 years allowed us to classify the outcome variable as early, normal, and delayed puberty. Modified Poisson regression with generalized overlap weighting was used to estimate weighted risk ratios (wRR) and their 95% confidence interval (CI). Results In total, delayed puberty occurred in 17 children (1.2%), mainly in males. Early puberty occurred in 631 (45.2%), more frequently among females. Compared with unexposed children, children born to continuers, early discontinuers, and late discontinuers had lower point estimates of risk of early puberty (wRR range 0.77–0.82). When early discontinuers were the comparator, estimates of early puberty risk were attenuated closer to the null (wRR range 1.00–1.07), with differences observed between males and females. Crude estimates, however, suggested higher risk of delayed puberty. Conclusion Children born to mothers who continued antidepressants or discontinued late in pregnancy did not have increased risk of early puberty, but the risk of delayed puberty might be higher.
