For how we live.
In 947 Chinese children followed for ten years, those with insulin resistance despite normal blood sugar had a 4.3-fold higher risk of later dysglycaemia, and six amino acids helped predict who would develop it.
In nearly 40,000 UK women, abnormal blood pressure six weeks after a hypertensive pregnancy roughly doubled later cardiovascular disease. A Dutch cohort shows pregnancy history sharpens risk prediction.
A cohort of 102 children born very preterm found epigenetic age acceleration linked to higher cognitive and motor scores at three — even, unexpectedly, in those with severe brain injury.
In 7,186 Korean adults followed up to 20 years, people flagged by fasting glucose, glucose tolerance and HbA1c all at once had a 15.3-fold higher risk of type 2 diabetes than normal controls.
Among 287 people with REM sleep behaviour disorder followed about four years, 15.7% developed a synucleinopathy. Higher coffee intake was linked to a lower risk of progression.
An Israeli records study of 181,000 adults found people now diagnosed are older and sicker, but start with better glucose and cholesterol, and far more leave the clinic on SGLT2 or GLP-1 drugs.
An observational study of 8,127 UK women on GLP-1 drugs, almost all tirzepatide, reported 16.1 kg average loss at six months and 91.4% retention. There was no control arm.
Among 93,406 children in Japan's national birth cohort, preterm birth was linked to higher odds of thinness at ages 4 to 6 and persistently lower BMI from 3 to 6, with no sign of catch-up.
In 4,299 people with type 1 diabetes and 17,196 matched controls, the death rate was 131 vs 35 per 100,000 person-years. The excess risk was far larger in women, though absolute mortality was higher in men.
Among 7,579 pregnancies in people with migraine, anti-CGRP antibody use early in pregnancy carried a 42% higher adjusted risk of spontaneous abortion; propranolol showed no increase. The data are observational.
A prospective UK surveillance study followed 210 pregnancies in 209 women with a pre-pregnancy cardiomyopathy diagnosis over three years. There were no maternal deaths; 10% had heart failure and 8% an arrhythmia.
In 221 male childhood-cancer survivors, FSH levels during puberty tracked adult testicular function. Radiotherapy damage persisted into adulthood, while alkylating-chemotherapy effects often recovered.
A nationwide study of 624,713 first births found quadrivalent HPV vaccination before pregnancy associated with modestly lower odds of preterm birth and growth restriction, strongest among women vaccinated youngest.
A 37-hospital registry tracked 17,602 adults whose hearts stopped while admitted. CPR usually started within a minute, yet only 9.1% survived to discharge — and after-care lagged far behind the initial response.
In 1,448 pregnancies, lower early placental growth factor was linked to more preterm births — yet adding it to a prediction model lifted accuracy only from 0.59 to 0.68.
Across 10 US children's hospitals and 697 patients, reoperation for recurrent twisting was rare either way — but laparoscopic repair carried lower odds of surgery for later bowel obstruction.
In the MONEAD cohort, six-year-olds whose mothers breastfed while taking antiseizure medicines scored no worse on language or adaptive behaviour — and longer breastfeeding tracked with slightly better scores.
A nationwide Brazilian cohort of 4.1 million children found that babies exposed to Zika in the womb but without congenital Zika syndrome had the same pneumonia-hospitalisation risk as unexposed children.
A Korean cohort of 1.37 million women found more dementia after hypertensive pregnancies. The association did not survive adjustment for the vascular and metabolic conditions that followed.
A population-linked Australian cohort followed children to school entry after antenatal metformin. The adjusted relative risk of developmental vulnerability was 0.97 — no different from unexposed children.
The PURE study measured airflow obstruction in adults across 25 countries and found most cases had never been labelled, with the largest undiagnosed share in lower-income settings.
A US analysis of the 2025-2026 vaccines targeting the LP.8.1 lineage found both an mRNA-1283 and a BNT162b2 shot reduced COVID hospitalisation in older adults, with the strongest effect in those aged 75 and over.
For people without coeliac disease, a large cohort found gluten intake was not linked to heart disease — and cutting it often means fewer whole grains, the foods tied to lower cardiovascular and mortality risk.
"Superfood" is a marketing word, not a scientific one. Blueberries carry real evidence for the heart and, tentatively, the ageing brain — but most of it is observational or short-term markers, not hard outcomes.
Cohort data link green tea drinking to lower cardiovascular death, but trials show only surrogate changes and concentrated extracts have caused liver injury at everyday catechin doses.
A banana is mostly sugar and starch, but it comes with fibre and potassium, and large cohort studies link whole fruit — bananas included — to slightly lower type 2 diabetes risk, not higher.
A French study of 11,558 patients without heart failure linked atenolol to fewer cardiovascular events than bisoprolol. It is observational, so it shows an association, not proof one is better.
Physically acting out dreams is one of the strongest known early signs of Parkinson's and related diseases. In a 1,280-patient cohort, most converted within 12 years — but it is a risk marker, not a verdict.
A pooled study of 1.44 million adults links activity to lower risk of 13 cancers — but not all, and not melanoma. A 2025 randomised trial now shows exercise after colon cancer improves survival.
The sugar alcohol in many 'keto' and sugar-free foods is generally recognised as safe. But newer studies tie high blood levels to cardiovascular events and show it makes platelets more prone to clotting.
Large cohorts link higher optimism and a stronger sense of purpose to longer life and lower mortality, in a dose-dependent way. But the studies are observational, and no trial has proved the causal arrow.
The WHO sets clear screen-time limits by age: none under two, an hour or less from two. The evidence behind them is real but modest — and the direction of cause and effect is not as simple as it sounds.
A landmark US cohort found frequent hot flushes lasted a median of 7.4 years, and 4.5 years after the last period. When they start earlier in the transition, they tend to last far longer.
Whole fruit and fruit juice behave oppositely for type 2 diabetes risk, and a trial that told newly diagnosed patients to eat less fruit found no benefit at all.
Guidelines settle on a limit near 200 mg a day — roughly two mugs of coffee. But the underlying studies are observational, and a vocal minority argues there is no clearly safe level.
A retrospective analysis of 54,114 women on a digital weight-loss service found those with PCOS lost as much as those without. But only 40 PCOS users reached the 10-month mark, and the data are observational.
Organic produce carries far fewer pesticide residues and modestly more of some antioxidants. But there is no strong evidence that eating organic makes you healthier, and the cancer signal is observational.
The fear that soy fuels breast cancer is not supported by the human data; if anything, soy food intake tracks with lower recurrence and death after diagnosis. Its other health claims are weaker.
People who eat nuts have lower rates of heart disease and early death in large cohorts, and a Mediterranean diet with nuts cut cardiovascular events in one trial. Most of the evidence is observational.
Whole fruit is linked to lower type 2 diabetes risk, sugary drinks to clearly higher risk, and fruit juice sits closer to the drinks than to the fruit it comes from.
For heart disease and death, dairy looks roughly neutral to modestly favourable in large cohorts, and the long-standing warning against whole-fat dairy is not well supported by outcome data.
Across nearly 1.1 million people, each extra daily cup tracks with lower diabetes risk in a clear dose-response. It is consistent and large — and still only observational.
Health authorities converge on a ceiling — under 10% of daily calories, and lower is better still. Large cohorts single out sugary drinks as the form most clearly tied to dying earlier.
A cohort of 5.24 million UK adults linked higher BMI to 17 of 22 cancers — strongly for the uterus, weakly or not at all for others, and inversely for a couple. The pattern is specific, not uniform.
The federal program enrols nearly 140,000 responders and survivors, most certified for a 9/11-related illness. Cohort data show a heavy, still-growing burden — and hint the program itself lowers mortality.
Twenty-five years of cohort data show the alkaline dust caused an acute cough syndrome and a real, largely permanent drop in lung function, graded by how much each person breathed.
Registries show a real but narrow signal: a few cancers, led by prostate and thyroid, run high in 9/11 responders, while overall cancer incidence sits at or below the rate expected.
In half a million Chinese adults, higher exhaled carbon monoxide tracked with lower Parkinson's risk even among people who never smoked. It is an association, not a therapy.
A British Columbia study emulated a trial across 41,000 patients. Those started on take-home methadone within weeks to months of induction stayed in treatment longer and died less than those on daily witnessed dosing.
Analysing 64.8 million births from 2005 to 2022, researchers found policies such as physician-only requirements and 20-week bans tied to higher odds of preterm birth or low Apgar scores.
A 411,000-patient claims study found adults on GLP-1 drugs who started oxycodone had more severe constipation and bowel obstruction at 30 days than those starting hydrocodone or tramadol.
A Norwegian study of 204,663 children found no raised risk of autism or most neurodevelopmental disorders after maternal COVID in pregnancy — with one exception in the early pandemic.
Across 16,893 high school athletes screened over 15 years, 0.36% had T-wave inversions. Four were later found to have hypertrophic cardiomyopathy, a leading cause of sudden cardiac death in the young.
A Swedish study of 10,693 ACL reconstructions in registered football players found no link between graft type and revision surgery. Age and playing level mattered far more.
In 5,520 children across 34 states, prenatal exposure to regulated public-water fluoride showed no association with emotional or behavioural problems overall — with one caveat in a secondary model.
Patients whose carotid plaque contained plastics had a higher rate of heart attack, stroke or death. It is an association in 257 people, not proof of cause.
In 2.5 million Swedish births followed to adulthood, mothers' early-pregnancy obesity was linked to a stepwise rise in offspring cardiovascular disease, though sibling analyses temper the size.
In 53 patients, oral fluconazole — but not a mouth-only antifungal — cut gut Candida and coincided with improved disease activity. The study was observational, brief and unblinded.
Among 3,032 couples sampled before conception, higher maternal and paternal vitamin B12 both tracked with lower predicted birth-defect prevalence. The paternal pattern was weaker, and this is observational.
In 1,238 people from two community cohorts, apnea severity showed no link to depressive symptoms. Sleep fragmentation did, and that association disappeared once insomnia symptoms entered the model.
A cohort with serial MRI and autopsy data found earlier menopause tracked faster cognitive decline and white matter damage. The per-year coefficients are very small.
A Nurses' Health Study II analysis estimated 37% of incident reflux symptoms were attributable to five modifiable factors. Body weight was the largest single one, at 19%.
A consortium reduced overnight sleep studies to nine composites. Two tracked with better cognition. None tracked with whether a person later developed dementia.
In 53,100 California women, higher trihalomethanes in drinking water tracked with modestly higher uterine cancer risk. Nitrate, arsenic and uranium showed no link.
Three cohorts, 147,374 people, 30 years and 35,798 deaths. The dose-response curve flattened above 120 minutes per week, and the cancer signal appeared at the low end, not the high end.
Moderate-to-severe sleep apnea alone roughly doubled all-cause mortality risk. Adding comorbid insomnia to the same patients raised it further — evidence for treating the combination as its own risk category.
A six-year cohort study using Japan's long-term care records found a roughly 25% lower risk. The design cannot separate cooking from the ability to cook.
In 3,873 middle-aged adults followed with repeated testing, waist-based categories of overweight and obesity showed larger effect sizes than BMI-based ones. The link to actual decline appeared only in men.
In 2,077 children, higher processed-food intake at age 3 tracked with worse behavioral and emotional scores at age 5. It's an association from a birth cohort, not a trial, and can't establish that the food caused it.