The political picture of a cannabis user in Britain is still a teenager in a park. The American data no longer supports that picture. Neither, quietly, does our own.
On 28 September 2026 the journal Addiction published a nationally representative study of people aged 65 and over in the United States. Benjamin Han at the University of California San Diego, with colleagues at Duke, NYU and Rutgers, analysed four years of the National Survey on Drug Use and Health. The sample was 21,189 older adults.
An estimated 8.9 per cent had used cannabis in the previous year. Among those who had, 11.4 per cent met DSM-5 criteria for cannabis use disorder — one in nine. Most of that disorder was mild (76.0 per cent). 20.1 per cent was moderate. 3.9 per cent was severe.
That is not a story about children. It is a story about pensioners. And it is the latest point on a line Han has been drawing for a decade.
What the American numbers actually say
Han’s 2016 paper in the same journal found that past-year cannabis use among Americans aged 50–64 rose 57.8 per cent between 2006/07 and 2012/13. Among those aged 65 and over it rose 250 per cent, from a very small base. By 2015–18, past-year use in the 65-plus group had reached 4.2 per cent, up from 2.4 per cent. In June 2025, in JAMA Internal Medicine, Han reported that current use — the past month — hit 7.0 per cent in 2023, up from 4.8 per cent in 2021. That is a 46 per cent rise in two years. In 2006/07, fewer than 1 per cent of older Americans had used cannabis in the previous year.
The new Addiction paper is the first to describe, at national scale, not just whether older Americans use cannabis but how, and who among them is in trouble.
Smoking was still the most common method (65.8 per cent of past-year users). Eating or drinking came next (40.5 per cent). Vaping was 15.4 per cent. Only 19.9 per cent said a doctor had recommended any of it. Four in five older users are not on a formal medical script. They are buying it, growing it, or being given it, in a country where most adults now live in a state that has legalised some form of cannabis.
Among those who met criteria for a use disorder, the two most common signs were spending a lot of time obtaining, using or recovering from the drug (74.4 per cent) and craving (67.5 per cent). That is not the cartoon of a collapsed life. It is closer to a habit that has become sticky — the same pattern GPs already know from alcohol, sleeping tablets and painkillers.
The risk was not evenly spread. Men were more likely to meet disorder criteria than women (relative risk 1.69). So were people with past-year mental illness (2.93), tobacco use (1.52), other illicit drug use (3.22), or a history of being arrested (1.59). Smoking cannabis, as opposed to other methods, carried a relative risk of 4.94. Using on 300 or more days a year carried a relative risk of 3.70. The people most likely to get into difficulty look less like a Saga brochure and more like patients already known to the NHS: men, smokers, drinkers, people with depression or anxiety, people who use a lot.
Han’s conclusion is blunt. Older adults who use cannabis need screening, and some of them need treatment. Clinicians who assume that a 70-year-old can be left to it because they are near the end of life are making a clinical error, not a kindness.
The paper is American. The legal context is American. Copying the percentages onto Britain would be lazy. Ignoring the direction of travel would be lazier.
Britain is already ageing into the same question
The Crime Survey for England and Wales still treats cannabis as a young person’s drug in the way it publishes the figures. Last-year cannabis use in the year ending March 2025 was 6.5 per cent among 16- to 59-year-olds and 12.5 per cent among 16- to 24-year-olds. Youth use is down on a decade ago: 16.4 per cent of 16- to 24-year-olds reported last-year cannabis use in the year ending March 2015.
The more interesting sentence in the latest bulletin is the one about the age gap. It has been narrowing since the 1990s, ONS says, “a result of drug use decreasing among younger age groups and increasing among older age groups.” Any-drug use fell among 16- to 19-year-olds compared with 2015. It rose among 45- to 54-year-olds. The year before, ONS had already noted rises among 45- to 54 and 55- to 59-year-olds compared with 2014.
That is a cohort moving through the tables. People who were in their late teens and twenties when last-year cannabis use in England and Wales was at its 1990s peak are now in their forties and fifties. They did not all stop. Some of them never started again. Some of them kept a quiet habit. Some of them came back for sleep, pain, or the same reason they first tried it, which was that they liked it.
What the survey does not tell us is what happens after 59. Questions on drug use are still capped there. ONS is explicit about why: “the low prevalence of drug use among those aged 60 years and over.” The last time the office published a 60–74 figure, for the year ending March 2020, about 1 per cent of that age group had taken any drug in the last year.
One per cent is not zero. It is also, by 2026, an old number. The United States was under 1 per cent past-year use among over-65s less than twenty years ago. It is now approaching one in eleven. Britain cannot know whether it is on the same curve because it has chosen not to count.
That choice is getting harder to defend. Around one in five people in the UK is already 65 or over — about 12.7 million in 2022, on Commons Library figures, with ONS projecting that share toward 27 per cent by 2072. The leading edge of the British baby boom is already in that group. The rave generation is in the Crime Survey’s rising 45–54 band. If reform ever arrives, it will not arrive into a country of teenagers. It will arrive into a greying consumer market that official statistics still refuse to describe.
The British picture is not a teenager with a spliff
Cannabis in British public argument is still a youth problem, a stop-and-search problem, a “skunk” problem. That is not invented. Last-year use is still highest at 16–24. Most possession files are not opened on people drawing a state pension. County lines do not run on Saga tours.
It is incomplete.
The person the new American paper describes has a British cousin who is already in the waiting room. Chronic pain, bad sleep, a few prescriptions, a cautious interest in whether “the medical stuff” is real. They are more likely to be talking about oil than a three-skinner. They are not posting about it. They may not tell the GP, because the GP has never asked.
UK medical cannabis, legal since November 2018, already shows that split. In Drug Science’s Project Twenty21 registry, people aged 65 and over were only 4.7 per cent of patients at entry — 198 of 4,228. They did not look like the rest of the list. They were more likely to be women (52.5 per cent against 47.0 per cent). 76.3 per cent were there for pain, against 45.6 per cent of younger patients. Only 20.2 per cent reported current daily use, against 60.3 per cent. They were more likely to be prescribed a CBD-dominant oil (17.5 per cent against 5.7 per cent) and much less likely to be prescribed THC-dominant flower (32.5 per cent against 75.2 per cent). Quality of life, mood and general health improved. Sleep improved less than it did for younger patients.
That is a different product, a different patient, and a different harm profile from the 22-year-old buying high-THC flower on a burner phone. A reform debate that only models the second person will mis-build the shops, the labels, the tax and the NHS response for the first.
There is a class wrinkle too. In the Crime Survey, last-year cannabis use is still higher in lower-income households. Han’s 2025 letter found the recent American surge among older adults was most pronounced among people who are college-educated, married, female, and on higher incomes. Legalisation in the US did not just add volume. It changed who felt entitled to use. A licensed British market would likely do something similar: pull in older, more affluent, more female buyers who will not currently meet a dealer, while the existing younger and poorer market does not disappear. Those are two customer bases. They want different things from a counter.
If Britain regulated, the grey pound would show up
A legal UK market would not create older cannabis users from nowhere. It would make them visible, easier to supply, and harder for the health system to ignore.
The customer would not only be 18. North American legalisation debates were sold, and opposed, as a youth story. The fastest growth after the laws passed was often among people who already had mortgages. Canada’s over-65 past-three-month use went from tens of thousands in 2012 to more than 400,000 by 2019, according to Statistics Canada figures reported after legalisation. Ontario emergency departments then saw cannabis-poisoning visits among over-65s roughly triple, from 5.8 per 100,000 person-years before legalisation to 21.1 once edibles were on sale — 2,322 visits over eight years, in Nathan Stall’s 2024 JAMA Internal Medicine letter. The absolute numbers were small against three million Ontario seniors. The direction was not. Edibles were the second step of the spike. Britain has a sweet tooth and a gummy market already waiting in the illegal channel. A legal edible without ruthless dosing, delay warnings and child-resistant packaging is how grandparents end up in A&E, and how grandchildren eat something from a biscuit tin.
The product mix would skew. Younger British medical patients already lean to flower. Older ones lean to oils. A regulated adult-use shelf built as a head shop will underserve the people with the money and the time. Lower-THC, clearer milligram labelling, capsules, and a pharmacist-adjacent counter are not a vibe. They are how you stop smoking — the method Han found almost five times more associated with disorder — becoming the default for people whose lungs and balance are not what they were.
The NHS would have to ask. Older adults metabolise THC more slowly, take more other medicines, fall over more easily, and drive. Polypharmacy, postural hypotension, confusion and delayed edible onset are geriatrics, not subculture. Han’s American users mostly did not have a doctor in the loop. Stall’s Ontario poisonings included people with cancer, alcohol problems and dementia. A UK law that legalises the shop and leaves GPs untrained is how you get a rise in “unexplained” falls and a coroner who has to learn a new sentence. Screening is not a moral project. It is medicines reconciliation.
The politics would change. Pensioners vote. A consumer who is 68, in pain, and already buying CBD in a health shop is a different leaflet from a consumer who is 19. Opponents will still lead with children, psychosis and high-THC flower, and those arguments have to be answered on the evidence. They should not be allowed to pretend that the median legal customer in a mature market looks like a sixth-former. If anything, a well-designed market has a reason to look boring: age checks at the door, no cartoon packaging, hours like a pharmacy, advertising rules that would make a lager brand weep. Boring is how Ontario got people through a licensed door. Youth-coded is how you lose the public and keep the WhatsApp groups.
The medical market would need a landing. Britain already has a private cannabis-based medicines sector. Older patients in it are a minority, but they are the ones on oils for pain. Adult-use reform that undercuts those prescriptions on price without keeping a clinical route will dump some of them into shops with no follow-up, and leave others paying clinic fees for a product the high street now sells. That is a design choice, not a footnote.
The tax would not all come from the young. Older households have, on average, more assets and more time. They are also more price-sensitive about being seen. A legal product that is tested, labelled and sold without a teenager behind the till is the product more of them will actually buy. Capture rates in the models — Transform’s 80 per cent legal share in a commercial scenario, against 45 per cent for clubs and home-grow — depend on people like that walking in. They will not walk into a shop that feels like a joke at their expense.
None of this means use among British over-65s is about to hit 8.9 per cent. The US had state legalisation, medical programmes, and a different 1960s. Britain had the Misuse of Drugs Act, a later youth peak, and a medical regime that is still mostly private and expensive. The honest UK statement is narrower. Use is falling among the youngest adults and rising in middle age. The generation that made cannabis ordinary in the 1990s is not 19 anymore. The generation that made it a moral panic in the 1960s is 65. We have almost no published data on the second group, and a health system that still does not ask.
The question is not whether they are getting older
They are. In the United States that is now a measured fact, including a one-in-nine rate of use disorder among older people who consume. In Britain it is a Crime Survey footnote about a narrowing age gap, a medical registry in which the over-65s look like pain patients rather than stoners, and an official decision that over-60s are too rare to report.
Reform, if it comes, will not mainly be a youth policy with a geriatrics appendix. It will be a consumer-protection and NHS-capacity problem for people who already take five other tablets, drive to Waitrose, and vote. The American paper is the warning about screening, smoking, frequency and mental illness. The Ontario paper is the warning about edibles. The British contribution, so far, is that we have not looked.
The first useful reform would not even require a Bill. It would be for ONS to publish last-year cannabis use for people aged 60 and over, with the same honesty it already applies to 16-year-olds. You cannot plan a market, or a harm-reduction service, for a customer you have declined to count.
Sources
Han, Eschliman, Palamar, Triguero Roura and Mauro, “Cannabis use and cannabis use disorder among older adults in the USA”, Addiction, 28 September 2026 (NSDUH 2021–24, N = 21,189 aged ≥65); Han, Yang, Cleland and Palamar, “Trends in Past-Month Cannabis Use Among Older Adults”, JAMA Internal Medicine, 2025; Han and Palamar, “Trends in Cannabis Use Among Older Adults in the United States, 2015–2018”, JAMA Internal Medicine, 2020; Han et al., “Demographic trends among older cannabis users in the United States, 2006–13”, Addiction, 2016. ONS, Drug misuse in England and Wales, year ending March 2025 and year ending March 2024; ONS, year ending March 2020 (60–74 any-drug figure). Lynskey, Thurgur, Athanasiou-Fragkouli, Schlag and Nutt, Project Twenty21, “Prescribed medical cannabis use among older individuals” (June 2024). Stall et al., “Edible Cannabis Legalization and Cannabis Poisonings in Older Adults”, *JAMA Internal Medicine*, 2024. House of Commons Library, “The UK’s changing population”; ONS mid-2024 population estimates (65+ still rising in every UK country).
