Which muscles hold their strength through weight loss? A gene hunt
A transcriptomic sub-study of the CALERIE calorie-restriction trial links 151 genes to preserved leg strength during weight loss, independent of how much body mass was lost.
Does stevia mess with your blood sugar or hormones? A crossover trial says no
In 69 people across normal weight, overweight and type 2 diabetes, a stevia drink behaved like water — no shift in glucose, insulin, gut hormones, appetite or how much they ate next.
Fathers' B12 appears in a birth-defects cohort, and folate's curve flattens early
A Shanghai study of 3,032 couples measured nutrients before conception. Predicted birth-defect prevalence fell across maternal B12 levels — and, more weakly, across paternal ones.
Picky eating is usually a phase: most toddlers outgrow it within a few years
A cohort of 4,018 children found nearly half were picky at some point, but two-thirds remitted within three years and prevalence fell to 13.2% by age 6. About 4% stayed persistently picky.
Eating late does change your metabolism. Whether it makes you fatter is less clear.
A tightly controlled crossover trial found late eating increased hunger, cut waking energy expenditure and shifted fat-tissue gene expression. The population data show a 17% higher odds of obesity.
The fibre evidence points at 25 to 29 grams a day. A supplement is not the same thing.
A Lancet series covering 135 million person-years found the largest risk reductions in that range. Trials of isolated fibre supplements produce much smaller effects on much narrower outcomes.
Coffee is consistently linked to lower liver cirrhosis and cancer risk — in weak studies
Two meta-analyses tie an extra two cups a day to roughly a third lower risk of both liver cancer and cirrhosis, with decaf showing a smaller effect. It is all observational, and the evidence quality is graded very low.
The world eats more than twice the sodium WHO advises. Cutting it works in trials.
Global mean intake was 4,278 mg a day in 2021 against a recommended limit under 2,000. A week of low-sodium eating moved blood pressure in most people; a five-year trial of salt substitute cut strokes.
Meta-analysis: protein supplements nudge cholesterol down in overweight adults, modestly
Pooling 32 trials, researchers found milk, whey, casein, and soy protein lowered LDL and total cholesterol slightly — but the effect shrank once weaker studies were excluded.
How much caffeine is too much? 400 mg is the number, and it is a soft one.
The FDA's threshold traces to a Health Canada assessment and a 2017 systematic review of 381 studies. It describes a population, not a person, and the variation between people is large.
Drinking more fluid roughly halves kidney-stone recurrence, trials show
Low fluid intake is a genuine, well-evidenced stone risk factor. Increasing fluid to produce plenty of urine is the single best-supported way to prevent another stone — better established than any diet tweak.
How common is iron deficiency in women? It depends entirely on the cutoff you use
In one study of 62,685 women, iron deficiency prevalence ranged from 3% to 15% depending on the ferritin threshold. On treatment, a recent trial found alternate-day iron caused fewer side effects than daily dosing.
Higher protein does not harm healthy kidneys, meta-analysis of 28 trials finds
The fear that a high-protein diet damages normal kidneys is not supported by the trial evidence. Protein restriction is a therapy studied in people who already have kidney disease — a different question entirely.
The alcohol J-curve did not survive its own control group
Across 107 cohort studies and 4.8 million people, the protective effect of moderate drinking disappeared once the comparison group stopped including people who had quit because they were ill.
Dairy prevented fractures in one large trial and nothing in the cohort data. Both hold.
Adding milk, yoghurt and cheese cut hip fractures 46% in Australian aged-care residents. In the general over-50 population, dietary calcium intake shows no association with fracture risk at all.
Sugar does not make children hyperactive. The trials settled this in the 1990s.
A meta-analysis of 23 blinded studies found no effect on any of 14 behavioural and cognitive measures. A separate trial fed children extreme doses and found nothing. The belief survives anyway.
The fermented-food boom rests on one 17-week trial with 18 people per arm
It found rising microbiome diversity and falling inflammatory markers in the fermented-food arm — and no change in its own primary outcome in the high-fibre arm it was compared against.
WHO called aspartame a possible carcinogen and left the safe intake unchanged
Two WHO bodies reported on the same day in July 2023 and reached findings that sound contradictory. Both were right, because they were answering different questions.
Fructose harms the liver mainly when it adds calories, the trials suggest — not by itself
Feeding studies find that when fructose replaces other carbohydrate at equal calories, liver fat barely moves. The liver damage tracks the extra energy, not the sugar's molecular identity.
Breakfast is not the most important meal for weight. The trials point the other way.
Pooling 13 randomised trials, people assigned to eat breakfast ended up 0.44 kg heavier and ate 260 more calories a day than people assigned to skip it. The trials were small, short and mostly low quality.
Two 2026 papers test the protein advice older adults keep getting. Neither result is clean
A NHANES analysis of 5,736 US adults over 60 found the benefit of higher protein directionally present and statistically imprecise. A meta-analysis of protein at breakfast found nothing at all.
The MIND diet slowed decline in cohort studies. The randomised trial found no difference.
A diet built to protect the brain looked powerful when researchers watched who chose to eat it. When they assigned it in a three-year trial, both groups improved by almost exactly the same amount.
Processed meat is a Group 1 carcinogen. That is a statement about evidence, not risk.
IARC put bacon in the same category as tobacco in 2015, which says how sure scientists are, not how dangerous it is. The attributable-death figures are three orders of magnitude apart.
Fatty liver disease affects roughly a third of adults. Weight loss is what reverses it.
MASLD is excess fat in the liver unrelated to alcohol, and about 30% of adults have it. The best-evidenced way to reverse the inflamed form is weight loss — and how much you lose tracks how much the liver recovers.
No trial shows seed oils cause disease. No trial shows they prevent it either.
Pooled biomarker data from 30 studies in 13 countries link higher linoleic acid to lower cardiovascular risk. The randomised evidence, 19 trials and 6,461 people, is null and rated low quality.
Low-FODMAP ranked first in every IBS comparison. No trial tested the part you do alone.
A network meta-analysis of 13 randomised trials put the diet ahead of every alternative for IBS symptoms. Its authors note the trials skipped reintroduction and personalisation entirely.
Moderate drinking is not safe for the liver, and the risk rises steeply with dose
The two dominant causes of chronic liver disease are alcohol and metabolic fat. For alcohol, cirrhosis risk climbs exponentially with intake — and in women it was already raised at about one drink a day.
Are eggs bad for cholesterol? The cohorts disagree and the trial found nothing.
A pooled analysis of 1.72 million people found no link between an egg a day and cardiovascular disease. A pooled analysis of six US cohorts found one. A randomised trial in 140 high-risk adults found neither.
The eight-glasses-a-day rule has no study behind it. Here is what does.
A 2002 review went hunting for the origin of "8 x 8" and found nothing. Two decades later, isotope measurements in 5,604 people put daily water use anywhere from about 1 litre to about 6.
104 companies sell DNA-based diet advice. Nearly half won't say which genes they read
A scoping review mapped 204 consumer nutrigenetic panels at a mean price of $234. Across them, 3,309 different genes were linked to dietary recommendations.
Medicaid paid for meals as treatment. Hospitalisations fell 31% and costs nearly balanced
The first large-scale policy evaluation of medically tailored meals covers 1,866 recipients across 11 health systems. It is quasi-experimental, not randomised — and that is the finding's main limit.
Is coffee good or bad for your heart? Across 2.86 million people, genuinely both.
Pooled data across 38 studies found higher coffee intake linked to a higher risk of heart attack — but a lower risk of stroke and irregular heartbeat, with a separate benefit for people with high blood pressure.
A Canadian cohort links toddlers' ultraprocessed intake to later behavior problems
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.
A French cohort links several preservatives to cancer risk. Eleven of 17 showed nothing
In 105,260 people followed for 7.6 years, higher intakes of sorbates, sulfites, nitrates and acetates tracked with modestly higher cancer incidence. This is an observational study of self-reported diets.
New US dietary guidelines raise protein, cap added sugar, and leave a fat problem
The 2025–2030 edition was released January 7. It is the first to name highly processed foods, and the first whose saturated-fat arithmetic does not obviously close.
They ate 500 fewer calories on unprocessed food. A reanalysis asks what they chose.
On the unprocessed diet, participants built plates 57% heavier and 110 calories lighter. The authors argue that fruit and vegetables were crowding out calories on purpose.
Diabetes care's annual rulebook lands, and obesity medicine moves toward the centre of it
The ADA's 2026 Standards of Care add guidance on individualising obesity-drug dosing, name two eating patterns for type 2 diabetes prevention, and recommend continuous glucose monitoring from diagnosis.
The Lancet's ultra-processed food series makes a displacement argument
Three papers published on 18 November argue the harm is not one ingredient but the replacement of whole-food diets. The evidence they assemble is strongest for diet quality and weakest where causation is hardest.
A Mediterranean diet beat standard first-line dietary advice for IBS in a UK trial
The trial was designed to test non-inferiority against traditional dietary advice. It found superiority instead, with 62% versus 42% achieving a clinical response over six weeks.
A review of low energy availability in team sport found prevalence between 0% and 80%
Fourteen studies covering 265 elite athletes used six different methods to identify relative energy deficiency. The spread in the answers is mostly a spread in the definitions.
In 77,100 Norwegian women, processed food risk landed on one side of the colon
Over 17.4 years, high ultra-processed food intake was linked to a 24% higher colorectal cancer risk overall — and a 58% higher risk for right-sided colon cancer specifically.
Ultra-processed food and diabetes risk, tested outside a Western diet
A pooled analysis of 72,776 Korean adults found an 11% higher type 2 diabetes risk in the top intake quartile. An updated meta-analysis of 17 cohorts put the dose-response at 10% per 10% of intake.