Your body clears alcohol in two steps, and the chemical in the middle is the one that breaks DNA

The rest of the day

59 more stories on this beat.

Each with its own sources. None of these is a link to the story above.

1. 02

   Genes tested the cancer link, and most of it did not survive

   Researchers used Mendelian randomisation, a method that compares people by the drinking-related genes they were born with rather than by what they say they drink. Genes are dealt at conception, so they cannot have been caused by illness, income or habit. Across roughly 1.49 million people in four research databases, genetically higher drinking raised head and neck cancer odds by 51%, with weaker signals for bowel and gullet cancer. Breast cancer showed nothing. Kidney and womb cancer went the other way. [14]

   Why it matters — The World Health Organization names seven cancers caused by alcohol and breast cancer is the one that reaches the most people. The strongest design available short of a trial could not confirm it.

2. 03

   843 studies re-graded, and one link came out strong

   A team inside the Global Burden of Disease project re-analysed alcohol against 20 health outcomes, pooling 843 cohort and case-control studies published between 1961 and 2023. Each link was graded from zero to five stars, where five means the evidence holds up even after allowing for how much the studies disagree with each other. Exactly one outcome reached five stars: cancer of the pharynx, the passage behind the nose and mouth, at 105% higher risk. Nothing scored four. [13]

   Why it matters — Grading the evidence rather than the headline shows how thin much of this literature is. The team also found publication bias, where positive results get printed more often, for eight of the twenty outcomes.

3. 04

   The dementia U-shape vanished under the genes

   In 559,559 adults in two research databases, one in the United States and one in Britain, dementia risk formed a U: both non-drinkers and heavy drinkers looked worse than light drinkers. 14,540 people developed dementia during follow-up. The team then repeated the question using genetics drawn from studies of 2.4 million people. The U disappeared and risk rose steadily with intake. Tracking individuals showed why: people who went on to develop dementia had been cutting down for years beforehand. [15]

   Why it matters — The apparent protection of a nightly glass was partly the disease reaching backwards into the data. The authors estimate that cutting rates of alcohol use disorder could lower dementia cases by up to 16%.

4. 05

   Most people with the slow enzyme drink anyway

   A team at the University of California read the ALDH2 gene in 311,290 people enrolled in the All of Us research programme in the United States. Among participants who identified as Asian, 23.5% carried at least one faulty copy, against under 2.5% in every other group. Of those with one faulty copy, 61.2% had drunk in the past year and 30.3% reported binge drinking. Even among those with two faulty copies, 24.4% had drunk. [7]

   Why it matters — The warning is built into the body and most people override it. That is the gap between knowing a risk and feeling one.

5. 06

   Flushers who drank heavily had strokes seven years earlier

   Doctors in Taiwan enrolled 930 people admitted with a first-ever ischemic stroke, the kind caused by a blocked vessel, and matched their ALDH2 genotype against their drinking. Among men, those carrying the faulty enzyme who drank heavily, meaning more than 45 grams of alcohol a day, had their stroke at an average age of 56.1. Carriers who did not drink heavily had theirs at 63.6, later than men with the normal enzyme. [25]

   Why it matters — The same gene shifted the age of a stroke seven years in either direction, depending only on drinking. It is one of the cleanest gene-and-behaviour interactions in the day's evidence.

6. 07

   A drug designed to block the clearing enzyme

   Disulfiram has been used for decades to help people stay off alcohol, and it works by blocking ALDH2 so that any drink floods the body with acetaldehyde and feels awful. It is unreliable, has to be swallowed daily and can harm the liver. Researchers have now put a related compound, SOPH-110S, through preclinical testing as a long-acting injection, reporting high potency, no off-target hits across 84 common targets and enough evidence for US regulators to allow human trials. [26]

   Why it matters — The treatment deliberately recreates the flush that half a billion people are born with. It only works because acetaldehyde is genuinely unpleasant to carry.

7. 08

   The gut, not just the liver, clears the poison

   The liver has always been treated as the place where acetaldehyde is destroyed. New work in mice shows a good share of it is instead pushed out of the liver in bile, into the gut, where gut ALDH2 finishes the job. Changing how much bile flowed changed both the acetaldehyde level in blood and how much the animals chose to drink. [27]

   Why it matters — It suggests the clearing system is two organs working together rather than one. It also opens a second place a drug could act.

8. 09

   What the flush is, and what it is not

   Alcohol intolerance is a genetic condition in which the body cannot break alcohol down efficiently, and its usual signs are a blocked nose and reddened skin within minutes. A true allergy to alcohol itself is rare. What people often react to is something else in the drink: sulphites used as preservatives, which up to 13% of people with asthma react to, or histamine and salicylates found in wine, beer, cider and rum. In Britain, drinks makers are not required to list ingredients. [11][12]

   Why it matters — Two different problems get the same name, and only one of them is about the alcohol. Someone reacting to sulphites in wine may be fine with a spirit.

9. 10

   Mouth and lip cancer from heavy drinking has nearly doubled since 1990

   Using Global Burden of Disease data for 2021, researchers counted the health lost worldwide to lip and mouth cancer attributable to heavy drinking. The measure is disability-adjusted life years, which adds years lost to early death to years lived in poor health. The total reached 1,191,084, up 90% from 628,484 in 1990. Growth was fastest in low-middle income regions, at 279%, and Southeast Asia saw deaths rise 357%. [29]

   Why it matters — The burden is moving toward places with the least cancer treatment. Central Europe still carries the highest rate per head.

10. 11

    Alcohol pushed into the cervical cancer picture

    Cervical cancer is the fourth most common cancer in women worldwide and is caused by persistent infection with human papillomavirus, or HPV. A review pulling together studies from several regions found moderate and heavy drinking tracked with higher risk, and that the effect was strongest in women already carrying a high viral load. Heavy drinking and second-hand smoke were both linked to infections that fail to clear. [30]

    Why it matters — Alcohol here is not acting alone. It appears to change whether an infection the body would normally shake off becomes permanent.

11. 12

    Normal liver tests do not mean a safe liver

    Korean researchers followed 19,035 adults over 40 who had kept normal liver enzyme readings across several health checks between 2002 and 2008, then tracked them to 2019. Heavy drinkers, defined as drinking three or more times a week, had a 73% higher risk of liver disease than abstainers despite the clean bloodwork. For alcohol-related liver disease specifically, moderate drinkers were 29% higher and heavy drinkers 186% higher. [31]

    Why it matters — A blood test that comes back normal is often read as a clearance. In this cohort it was silent about damage that was already accumulating.

12. 13

    The nightcap gets you to sleep and then wrecks the night

    A review of the past decade's evidence on alcohol and sleep, giving priority to systematic reviews, found the same pattern across every study design. Alcohol shortens the time it takes to fall asleep. It then suppresses rapid eye movement sleep, the stage tied to memory and emotional processing, breaks sleep into fragments and worsens breathing, with most of the damage in the second half of the night. Habitual drinking tracked with worse sleep and more insomnia. [32]

    Why it matters — The one effect people notice is the only one that helps. Everything the drink does afterwards happens while they are asleep.

13. 14

    Wine's melatonin is too small to count

    Wine contains melatonin, the hormone that signals night to the body, and polyphenols, plant compounds often credited with health effects. Both are regularly used to argue that wine is a gentler drink before bed. The same review measured what is actually present and concluded the amounts are far too low to produce any meaningful effect on sleep. Timing mattered more than beverage: drinking earlier in the evening reduced some of the disruption. [32]

    Why it matters — A real ingredient at an irrelevant dose is one of the most common ways a health claim gets built. The melatonin is genuinely in there.

14. 15

    Alcohol makes a stopped-breathing disorder worse

    Obstructive sleep apnoea is a condition where the throat closes during sleep and breathing repeatedly stops and restarts, causing loud snoring and daytime exhaustion. Alcohol relaxes the muscles that hold the airway open. It also lengthens the delay before the body rouses itself to breathe again, so oxygen falls further before the sleeper wakes. Studies show moderate to heavy drinking can produce apnoea episodes in people who do not otherwise have the condition. [33]

    Why it matters — The mechanism explains why heavy drinkers snore and why a drinker's partner notices before they do. People with alcohol use disorder have a higher risk of developing the condition.

15. 16

    Why the anxiety arrives the morning after

    Alcohol boosts GABA, the brain's main calming signal, and dampens glutamate, the main firing signal. The brain notices the imbalance and pushes back, ramping up its excitatory systems and stress pathways to compensate. Those compensations do not switch off the moment the alcohol clears. What is left is a temporary state of over-excitement, with cortisol, the main stress hormone, still raised and sleep already broken. [35][36]

    Why it matters — The dread is not a mood, it is the correction overshooting. It is also why the second night after heavy drinking is often calmer than the first.

16. 17

    Alcohol hits more brain systems at once than any other common drug

    Cocaine works mainly on dopamine and noradrenaline. MDMA works mainly through serotonin and oxytocin. Heroin works on the endorphin system. Alcohol touches all of those and the two most common signals in the nervous system as well. Researchers call this being pharmacologically promiscuous. As blood levels rise it moves down the brain, taking out judgement first, then the balance and coordination centre, and at critical levels the brainstem that drives breathing. [35]

    Why it matters — One drug covering that many uses is why it has stayed in every culture that had it. It is also why nothing simple explains what it does.

17. 18

    The heavy-drinking brain learns to burn a different fuel

    Alcohol raises acetate in the blood, and the brain can burn acetate instead of its usual glucose. Researchers infused labelled acetate into light drinkers, at-risk heavy drinkers, people in long-term recovery and people entering supervised detox, then measured the brain's uptake with magnetic resonance spectroscopy. The heavy-drinking brain had built extra capacity to use it. In the detox group the rate was lowest at one week of abstinence, and they were scanned again a month in. [37]

    Why it matters — The brain reorganised its energy supply around a by-product of drinking. The paper's title calls the change reversible.

18. 19

    Alcohol reclassified as a changeable Alzheimer's risk

    Alzheimer's disease researchers have begun listing alcohol among the risk factors a person's circumstances can alter. Reviews collected this year set out where the two overlap in the body: oxidative stress, activation of the brain's immune cells, tangling of the tau protein, altered handling of amyloid, and damage to the hippocampus and prefrontal cortex. Heavy drinking is the clear case. What light and moderate drinking do to the same pathways is still described as uncertain. [38][39]

    Why it matters — Drinking rates are rising in middle-aged and older groups at the same time as dementia rates. Whether one is feeding the other is exactly what the genetic studies were built to test.

19. 20

    Nerve damage from years of heavy drinking

    Alcohol-related neuropathy is nerve damage caused by long-term heavy drinking together with the poor nutrition that often goes with it. Alcohol is directly toxic to nerve tissue, and heavy drinkers frequently run short of the B vitamins nerves need. Symptoms start in the feet and hands as pain, numbness and pins and needles, and can progress to weakness and unsteadiness. Stopping early gives the best chance of recovery. [41] Reviews of long-term exposure trace the damage to alcohol's own by-products, acetaldehyde among them, hitting the parts of the cell that make energy. [40]

    Why it matters — It is one of the alcohol harms that can partly reverse. That depends on catching it before the nerve fibres die, which is a question for a doctor.

20. 21

    Blackouts and the slower kind of memory loss

    Two different things get called alcohol memory loss. One is the blackout, where events during heavy drinking are never recorded at all. The other is a persistent decline in memory and thinking that follows years of heavy use and can amount to dementia. The US National Institute on Alcohol Abuse and Alcoholism lists the brain changes behind the second. Alcohol use disorder is defined as being unable to stop or control drinking. [42]

    Why it matters — People often reassure themselves with the first while carrying the second. The blackout is dramatic and temporary; the other is quiet and is not.

21. 22

    Who wakes with the headache, and what it feels like

    Thirty-two healthy volunteers drank alcohol and filled in questionnaires across three separate drinking and hangover episodes. A headache featured in 55 of the 96 episodes, 57.3% of them. It was almost always over the whole head, pressing rather than throbbing in two thirds of cases, and rated 6 out of 10 for intensity. It worsened with movement in every single case and changed when people stood up in 89%. Anyone with a history of headaches was three and a half times more likely to get one. [43]

    Why it matters — The standing-up pattern points at pressure around the brain rather than at dehydration alone. Those whose headache came on when they stood also got over it faster.

22. 23

    Twenty-one hangover trials, and the evidence is weak

    Researchers at King's College London reviewed every randomised controlled trial claiming to prevent or ease a hangover. They found 21, testing curcumin from turmeric, clove extract, B vitamins, the amino acid L-cysteine, red ginseng and Korean pear juice. Seven reported an improvement against a dummy treatment. The studies varied enormously in how much participants drank and whether they ate, very few included women, and not one intervention had been tested a second time. [44]

    Why it matters — A finding nobody has repeated is a lead, not a result. The hangover market is built almost entirely on findings nobody has repeated.

23. 24

    A turmeric drink lowered blood alcohol in 27 people

    A randomised, double-blind, crossover trial gave 27 healthy adults one of two turmeric-based products or a dummy 30 minutes before a set alcohol dose, then measured blood and breath alcohol for 15 hours. The combined product, containing Theracurmin along with an extract of the oriental raisin tree and a yeast-derived glutathione, cut the total alcohol exposure over time compared with placebo. Turmeric alone did less. Every participant took all versions, so each acted as their own comparison. [45]

    Why it matters — The trial measured alcohol in the blood, not how bad anyone felt the next morning. Products like this get sold on the first measurement and bought for the second.

24. 25

    Dead bacteria cut the middle chemical in a whiskey test

    Twenty-eight people drank whiskey equivalent to 90 grams of alcohol, roughly six standard drinks, having taken either a heat-killed bacterial preparation or a dummy. Researchers drew blood eight times over 15 hours. Both alcohol and acetaldehyde came out lower in the treated group, and acetaldehyde peaked lower and later. Scores on two standard hangover questionnaires were also lower. Each person did both arms of the trial. [46]

    Why it matters — It is the only trial in the day's pool that measured the actual chemical rather than just how people felt. It has not been repeated.

25. 26

    Silicon and hydrogen tried on hangovers in mice

    Japanese researchers tested a silicon-based agent that reacts with water in the gut to release hydrogen gas continuously, on the basis that hydrogen has been reported to ease acute alcohol damage to the liver. Mice given an alcohol overdose showed less of the movement impairment that follows, and less liver damage. The work is at the animal stage and the mechanism is an inference from earlier hydrogen studies rather than a demonstration. [47]

    Why it matters — It is a reminder of where most hangover science sits. A mouse that walks better is several long steps from a person who feels better.

26. 27

    Nothing you eat speeds the liver up

    Two medical centres put the same question to their specialists and got the same answer. No food, before, during or after drinking, changes how fast the liver processes alcohol. What food can do is ease the symptoms around it: salt helps hold onto fluid, fat slows the stomach emptying, protein steadies blood sugar, which dips after drinking. Eggs are often mentioned because they contain cysteine, an amino acid involved in clearing the by-product. [4][48]

    Why it matters — The distinction is between shortening a hangover and making it tolerable. Only the second is on the table.

27. 28

    Alcohol dehydrates by switching off a hormone

    The body holds onto water using vasopressin, also called antidiuretic hormone, which tells the kidneys to reabsorb water rather than send it to the bladder. Alcohol reduces its release, and starts doing so within about 15 minutes of the first sip. Three things follow: the kidneys reabsorb less, more water ends up in urine, and the bladder fills faster. The net result can be losing more water than the drink supplied. [49]

    Why it matters — It explains why drinking water alongside does not fully cancel the effect. Alcohol has already switched off the hormone that would have told the kidneys to hold on to that water.

28. 29

    A drug that removes the point of the drink

    Naltrexone blocks opioid receptors and was developed for opioid addiction. It also blunts the euphoric kick of alcohol. The Sinclair Method uses this differently from standard treatment: rather than taking it to stay away from alcohol, a person takes it before drinking, so that each drink delivers less reward and the learned craving gradually fades. It is prescription-only and does not stop someone becoming impaired. [50][51]

    Why it matters — It treats the habit as something learned rather than as a failure of will. Whether it suits a given person is a matter for a doctor.

29. 30

    Two similar drugs, two different sets of responders

    Nalmefene and naltrexone work through nearly the same brain receptors and differ mainly at one of them. Trials of the two have produced conflicting results for years, with no head-to-head comparison. Researchers gave 56 mice all conditions in turn: nalmefene, naltrexone, a comparison compound and a dummy. On average the two drugs performed identically. Individually they did not: the animals helped by one got no benefit at all from the other. [52]

    Why it matters — An average that hides two opposite groups looks like a tie. A blood test predicted which group a mouse belonged to, which is where a human version would have to start.

30. 31

    Weight-loss drugs slowed alcohol into the blood

    Weight-loss drugs of the GLP-1 class, such as semaglutide, have repeatedly been observed to reduce how much people drink, and nobody is sure why. Twenty people with obesity, 14 already taking one of the drugs and the rest not, drank a set dose of alcohol while researchers measured breath alcohol and how intoxicated they felt. The treated group's readings rose more slowly at 10, 15 and 20 minutes, and nausea did not explain the gap. [53]

    Why it matters — It suggests part of the effect happens in the stomach rather than the brain, which is how disulfiram works and not how naltrexone works. The sample is 20 people and the authors call it preliminary.

31. 32

    Heavy drinking and obesity turn out to share circuits

    About 42% of adults in the United States have obesity and about 10% meet the criteria for alcohol use disorder, with roughly 8 million adults in both groups. A review in Molecular Psychiatry argues the two are not simply parallel cases of consuming too much. It sets out overlapping biology in the brain's reward and appetite systems, which is one explanation for why a drug aimed at appetite keeps turning up in drinking data. [54]

    Why it matters — It connects two conditions usually treated by different specialists. It also predicts that a treatment for one will keep leaking into the other.

32. 33

    Dry January's volunteers are not a random sample

    A scoping review screened 90 publications on Dry January and examined 16. Compared with drinkers in the general population who did not take part, participants tended to have higher incomes, university degrees, and to be younger, female and heavier drinkers at the start. Lighter drinkers were more likely to finish the month than heavier ones. Among those who did finish, drinking stayed lower for months and biological measures and wellbeing improved. [24]

    Why it matters — Every result about Dry January is a result about the people who volunteer for it. That is not a criticism of the month, it is a limit on what its numbers can be read to mean.

33. 34

    What people say when asked why they drink less

    Researchers in Germany followed 540 low-to-moderate drinkers over three years, sorted their drinking paths into three patterns, then interviewed 22 of them in depth. Social reasons dominated every group: drinking with others, and not drinking when there was nobody to drink with. Reasons to cut down were mostly practical and immediate rather than about long-term disease. The interviewees averaged 36 years old and 55% were women. [55]

    Why it matters — Public messages are built around cancer risk and the people in this study were not thinking about cancer. The gap between the message and the motive is where campaigns fail.

34. 35

    Perimenopause, low mood and drinking to cope

    A survey of 936 women aged 40 to 65 measured menopause symptoms, drinking, mental health and wellbeing together. Women in perimenopause, the years of hormonal change before periods stop, reported the most symptoms, the lowest wellbeing and the strongest drinking-to-cope motives. Statistically, those coping motives partly accounted for the link between symptoms and harmful drinking. It is a cross-sectional survey, so it captures one moment rather than a sequence. [56]

    Why it matters — It identifies a life stage where drinking and distress rise together. Which one moves first is exactly what this design cannot say.

35. 36

    The same drink hits harder after 65

    The chemistry of alcohol processing does not change with age but the body's capacity to do it does. Liver enzymes shift and work more slowly. Lean muscle falls, so there is less tissue for alcohol to spread into and more of it stays in the blood. Medications interact, and conditions such as diabetes affect liver function. Older adults also carry less of the clearing enzyme, which is part of why hangovers get worse with age. [12][57]

    Why it matters — The same two glasses become a larger dose without anyone drinking more. This is one of the clearest cases where a fixed guideline fits some people badly.

36. 37

    Wine looked better than spirits in 340,924 people

    An analysis of UK Biobank data presented at the American College of Cardiology's 2026 meeting compared wine, beer and cider, and spirits against death rates over an average 13 years. High consumption of any type raised the risk of death: 24% for all causes, 36% for cancer. At low to moderate levels the types parted company. Moderate wine drinking tracked with 21% lower cardiovascular death, while even low intakes of beer, cider or spirits tracked with 9% higher. [58]

    Why it matters — Wine drinkers differ from spirits drinkers in income, diet and when they drink, and none of that is the wine. The work was presented at a conference and has not yet been peer reviewed.

37. 38

    The US dropped its daily drinking numbers

    The previous US guidance was no more than two drinks a day for men and one for women. The 2025-2030 Dietary Guidelines for Americans replaced that with a sentence: consume less alcohol for better overall health. No figure is attached. Mehmet Oz, who runs the US Medicare and Medicaid agency, summarised it in a January press conference as meaning a small amount. The guidelines do not define small. [22][23]

    Why it matters — A number can be checked against a glass and a sentence cannot. Health analysts warned the change may make it harder for doctors to spot when drinking warrants attention.

38. 39

    Two commissioned reports, one dropped

    The Biden administration commissioned two reviews to inform the guidelines. One, led by Columbia University epidemiologist Katherine Keyes, found risk rising from low levels and put the chance of an alcohol-related death above 1 in 1,000 at two drinks a day. The alcohol industry called it ideologically driven and asked for it to be disregarded. In August her team was told it would not go into the guidelines packet. The Department of Health and Human Services gave no explanation. [18][20]

    Why it matters — The second review, from the National Academies, reached the opposite conclusion and did make it in. The two teams were reading overlapping evidence with different exclusion rules.

39. 40

    Alcohol deaths in the US are still above where they were

    There were 46,756 alcohol-induced deaths in the United States in 2024, meaning deaths from conditions caused directly by drinking. That is about 50% higher than ten years earlier and about 20% above the level before the Covid-19 pandemic, having risen sharply during the pandemic years and eased since. Separately, alcohol contributed to more than 4.2 million emergency department visits in 2022, and about 178,000 deaths a year across 2020 and 2021. [22][59]

    Why it matters — The numbers moved in the opposite direction to the guidance. Polling finds 54% of US adults say a family member has struggled with alcohol, while only 7.6% of people with alcohol use disorder received any treatment in 2022.

40. 41

    Education changed what alcohol did to the heart

    Researchers followed 387,914 people aged 40 to 69 in the UK Biobank who were drinking at the start, and tracked deaths from ischemic heart disease. Among men, drinking between 1 and 30 grams a day without heavy sessions was linked to 39% lower risk than occasional drinking. But that apparent benefit differed by education: men with low or medium education showed a worse pattern than highly educated occasional drinkers. No such difference appeared among women. [60]

    Why it matters — If a drink protected the heart chemically, a degree would not change the result. The difference points at what else travels with education.

41. 42

    The genes said red wine, and even that is contested

    A Mendelian randomisation study looked at bleeding inside the brain, a stroke type with an annual fatality rate around 50%. Genetically predicted alcohol intake overall raised the risk by 21%. Genetically predicted red wine intake specifically lowered it, and about 13% of that ran through lower blood pressure. Survey data from older Chinese adults pointed the same way. The study sits against a separate re-analysis that downgraded alcohol's heart evidence to one star. [61][13]

    Why it matters — The same genetic method that removed alcohol's protective glow elsewhere produced one here. That is what an unsettled question looks like from the inside.

42. 43

    The old wine evidence had a hole in the middle of it

    PREDIMED was a Spanish trial of 7,447 people at high heart risk, whose diet score included one item on wine. People scoring well on the Mediterranean diet including wine had a 45% lower rate of cardiovascular events than poor compliers; excluding wine, the reduction was 16% and not statistically clear. A second cohort of 23,133 people was followed for 22 years. But the analyses least vulnerable to abstainer bias were not significant. [16]

    Why it matters — Abstainer bias is what happens when people who quit through illness are counted as non-drinkers. Strip it out and the wine result thinned.

43. 44

    The researcher who changed his mind in public

    Tim Stockwell, of the Canadian Institute for Substance Use Research, says he published work twenty-five years ago arguing that doubting alcohol's protective effect on the heart was unreasonable. He now says the more rigorous studies have moved him the other way, and that bodies including the World Health Organization and the European Heart Network have shifted toward there being no safe level and no protective effect. Canadian guidance now puts low risk at two drinks a week. [18]

    Why it matters — Canada's number and the new US non-number were set from overlapping evidence in the same few years. The gap between them is judgement, not data.

44. 45

    The case for not throwing out the moderate-drinking evidence

    A public health scientist writing in STAT worked through the same reports and reached a different position. A meta-analysis of 60 prospective studies found no link between low-level drinking and overall cancer risk apart from a weak breast cancer signal. The National Academies found small breast cancer associations, weaker bowel evidence, and no consistent link to mouth or throat cancer. And absolute risks are small: among women, fewer than seven drinks a week carried a 0.56% ten-year breast cancer risk against 0.64% at seven or more. [17]

    Why it matters — A 14% relative increase and a 0.08 percentage point absolute increase are the same number described two ways. Which one gets printed usually decides how the finding lands.

45. 46

    Why nobody has simply run the trial

    A review titled The Unfinished Debate lays out how the field got here. The Global Burden of Disease study said in 2018 that zero was safest, then in 2022 added age-specific adjustments allowing net benefit for some groups. The 2025 US surgeon general report discouraged any intake; the National Academies report the same year argued for tailoring to individual risk. Mendelian randomisation studies found no benefit. No large trial comparing cessation against moderation has been run. [19]

    Why it matters — Every position in the argument is drawn from observation, and observation cannot settle it. The trial that would settle it is expensive, long, and hard to keep people in.

46. 47

    Alcohol, the gut wall and ageing pull in the same direction

    A review this year traced how alcohol, the ALDH2 variant and ageing interact in the gut. Slow acetaldehyde clearance raises oxidative stress and makes the intestinal lining more permeable, letting bacterial fragments reach the bloodstream and trigger inflammation, even at low intakes. Heavy drinking erodes the lining further and shifts which microbes live there. Ageing adds its own low-grade inflammation, so the same drink lands on a body less able to absorb it. [62]

    Why it matters — It ties the gene, the gut and the calendar into one account. It is also the route by which drinking affects mood through inflammation rather than through the brain directly.

47. 48

    When you drink may matter as much as how much

    A review of alcohol and the body clock collected evidence that circadian disruption and alcohol damage amplify each other. In animal and human studies, disturbing the clock genetically or by behaviour increased alcohol-driven gut permeability, systemic inflammation and liver disease. Shift work, poor diet and smoking made it worse. The liver, intestine, heart, blood vessels and skeletal muscle all run on daily rhythms that alcohol pushes out of phase. [63]

    Why it matters — It gives a mechanism for why the same amount of drinking harms a night-shift worker more. It also means timing is a variable most alcohol studies never recorded.

48. 49

    Intelligence at 18 and later drinking problems

    Researchers used the Swedish military conscription register, which tested nearly every 18-year-old man for decades, and followed 573,855 men born between 1950 and 1962 for an average of 60 years. Lower measured IQ at 18 was linked to higher risk of alcohol use disorder after adjusting for parental substance use, psychiatric diagnoses, family finances and birth year. Genetic analyses using data from over a million people pointed the same way. [64]

    Why it matters — It puts a difference measured decades before the first drink into the causal chain. That makes alcohol use disorder harder to read as a series of individual decisions.

49. 50

    Red wine's reputation as a migraine trigger is shakier than it sounds

    Migraine affects about 14% of the world and red wine is among the triggers patients name most often. A review of alcohol and headache found the few prospective studies, which ask people to record drinks and attacks as they happen rather than recall them afterwards, do not consistently confirm it. Cluster headache, an intensely painful and rarer type, is reliably triggered by alcohol, yet many patients keep drinking, sometimes during an attack. [65]

    Why it matters — Recalling what preceded an attack is a different measurement from recording it in advance. Migraine can start hours before anyone notices, which is long enough to reach for a drink.

50. 51

    What repeated binges and hangovers do to social behaviour

    A review pulled together human studies and animal experiments on binge drinking, meaning a large amount in a short period, and the withdrawal that follows. A single binge tends to increase sociability and interest in new company. Repeated cycles of binge and hangover run the other way, producing lasting low mood and social withdrawal. The pattern is most studied in adolescents, where binge drinking is most common. [66]

    Why it matters — The reason people drink and the reason they later avoid company come from the same cycle. It is one of the few places where alcohol's short and long effects are exactly opposite.

51. 52

    The full list of glands alcohol reaches

    The endocrine system is the network of glands that release hormones into the blood to coordinate growth, metabolism, reproduction, stress and daily rhythms. A critical review this year set out where alcohol interferes across it, covering the hypothalamus and pituitary in the brain, the thyroid, the pancreas, the adrenal glands and the gonads. Disruption at one gland propagates, because these organs regulate one another through feedback loops. [67]

    Why it matters — It is the reason alcohol shows up in stories about fertility, thyroid tests, blood sugar and stress hormones at once. One input, many outputs.

52. 53

    The kidneys take a direct and an indirect hit

    A Cleveland Clinic nephrologist sets out two routes. Directly, heavy drinking changes how the kidneys handle fluid and electrolytes and can injure the tissue itself. Indirectly, alcohol raises blood pressure and damages the liver, and both of those are leading causes of kidney disease in their own right. Kidney damage typically produces no symptoms until it is well advanced. [68]

    Why it matters — The indirect route is the larger one and it is the one nobody attributes to drinking. A blood pressure problem rarely gets recorded as an alcohol problem.

53. 54

    Inflammation in the liver that can start after you stop

    Alcohol-induced hepatitis is inflammation of the liver caused by drinking. The liver treats alcohol as a toxin and breaks it into by-products that damage liver cells; the inflammation is the body's repair response, and sustained inflammation destroys healthy tissue and leaves scarring. It can appear as a short-term reaction to heavy drinking, and it can also develop after someone has stopped, because the damage accumulated silently. [69]

    Why it matters — The delay is what makes it dangerous. Feeling fine while drinking is not evidence about the liver, and neither is stopping.

54. 55

    The cannabis-for-sleep claim is thinning too

    Eighteen studies were identified and nine pooled in a review of what cannabis does to sleep measured in a laboratory rather than by self-report. Cannabis did not consistently change how long people slept, how quickly they fell asleep, how often they woke, or the balance of sleep stages. Early work suggesting it suppresses rapid eye movement sleep used small samples and high THC doses. Newer studies at therapeutic doses mostly found nothing. [70]

    Why it matters — It is the same shape as the alcohol nightcap: a substance taken for sleep whose effect on measured sleep is not what users report. Withdrawal disrupts sleep in both.

55. 56

    Sleep trouble runs through addiction treatment and is rarely treated

    A systematic review and meta-analysis of sleep in substance use disorders notes that disturbed sleep persists into abstinence and is linked to relapse, yet is rarely addressed in treatment. The loop is specific with alcohol: people with insomnia use it as a sedative, the sedative effect fades quickly, sleep gets worse, and drinking rises. Relapse rates within weeks or months remain high across existing therapies. [34][52]

    Why it matters — A modifiable factor sitting inside a treatment gap is unusual. The same review notes cannabis users cite sleep as a reason to use and then relapse during withdrawal insomnia.

56. 57

    Alcohol lands in the same body as tobacco, and they compound

    The National Academies held a two-day workshop in March 2025 on tobacco, alcohol and cancer. It set the two side by side deliberately: tobacco remains the leading changeable cancer risk factor, alcohol ranks third, and people who smoke are more likely to drink and the reverse. The 1964 US surgeon general report on smoking led within a year to warning labels and a broadcast advertising ban. [72]

    Why it matters — Alcohol in 2026 sits roughly where tobacco sat in 1964: mechanism established, labels contested. The overlap between the two groups means their harms are hard to count separately.

57. 58

    Where the flush stops being a joke

    Among 244 Asian American college students followed over four years, researchers separated those who flush when they drink from those who do not, then tested how drinking motives related to how much they drank. Among flushers, drinking for enjoyment and drinking to socialise were more strongly tied to higher consumption. Among non-flushers, fitting in mattered more. Coping motives behaved the same in both groups. [28]

    Why it matters — The people with the built-in warning were not drinking less for the reasons everyone else drinks. Guidelines that recommend lower limits for flushers have to reach past that.

58. 59

    Alcohol as a driver of the metabolic damage nobody attributes to it

    A review of heavy drinking and metabolism describes a cascade: reactive oxygen species damage fats in cell membranes, the cell's supply of NAD+ is depleted, and mitochondria, the parts of the cell that make energy, start to fail. The review traces this through the liver, heart, pancreas and brain, and links it to alcohol-related liver disease, type 2 diabetes and cardiovascular disease. [71]

    Why it matters — Fatty liver, diabetes and heart disease are usually filed under diet. This is the route by which drinking arrives at the same diseases from a different direction.

59. 60

    Blood vessel lining as the shared route of injury

    A review of acetaldehyde and organ damage focuses on the endothelium, the single-cell lining of every blood vessel. Acetaldehyde damages its DNA, raises oxidative stress and disturbs calcium handling and insulin and oestrogen receptor signalling inside those cells. The review notes that one DNA repair route, the NEIL3 pathway, is used ahead of another, which can leave rearranged genetic material behind. [6][5]

    Why it matters — The lining of the blood vessels is common to every organ, which is one explanation for why a single chemical shows up in liver, brain, heart and kidney stories at once.