The sentence is three claims
"Young men are most at risk from gambling." Some version of that sentence is everywhere, and it reads like a single finding. It is three, and they do not all hold up.
Against young women, it holds
This is the part that is solid. Every study I found that splits young people by sex finds young men higher. How much higher depends on the study, and so does what is being counted.
Three studies, three countries, three different quantities:
- Canada. 3.3% of young men aged 15 to 24 were moderate-risk or problem gamblers, against 1.1% of young women. A snapshot of how many are in trouble, on a deliberately wide definition.
- Sweden. 3.32% of males aged 16 to 24 developed a first gambling problem during one year, against 1.14% of females. Not how many have one. How many newly got one.
- United States. The poll from the last entry found 10% of men aged 18 to 30 scoring as problem gamblers, against 7% of women the same age. A snapshot again, but on a narrow definition, and from a small subgroup of 801 voters.
Line the ratios up and the agreement is thinner than it looks. Canada and Sweden both put young men at about three times young women. The US poll puts them at about one and a half times. Same direction, different size, and the American gap is the smallest of the three even though its headline number is the biggest.
Two cautions. The Canadian study was published in 2007, which makes it nineteen years old, and it includes 15-to-17-year-olds.
The second is those near-identical digits. Canada reports 3.3 and 1.1; Sweden reports 3.32 and 1.14. Side by side that reads as two studies confirming each other. One is a standing total and the other is an arrival rate, so the resemblance is a coincidence between two different measures. It is exactly the kind of thing that gets quoted as corroboration.
Even so, the direction is consistent across three countries and two different instruments. Against young women, young men are at higher risk. I would call that answered.
Against older men, nobody has published it
This is where it stops.
To settle "most at risk" you need problem gambling broken down by sex and age at the same time: young men against middle-aged men against older men. That cell is not in the published tables. Not in the British national survey, not in anything I opened this week.
I went looking for a counter-example, because this is the claim the entry rests on. The strongest candidate is the largest study of its kind: a 2024 review pooling 366 representative samples across 68 countries. It reports problem gambling by sex, at 2.8% of adult men who gambled against 1.2% of women. It reports age too, as adults against adolescents. Separately. Its subgroups are sex, broad age, region, country and how people were recruited, and the two never cross. The largest pooled estimate we have cannot tell you whether a man of twenty is at more risk than a man of fifty.
One paper comes closer. The meta-analysis behind the circulating "3.4 times" does carry a young-men figure, in its written section rather than its pooled results: three of the studies it collected put problem gambling at somewhere between 4% and 8.1% of young men. A range that wide is three studies disagreeing, and there is no older-men figure to set beside it. It tells you young men have been counted. It does not tell you they are the most at risk.
That is the pattern everywhere. Men against women, all ages. Young adults against middle-aged adults, both sexes. Different questions, and you cannot multiply them together to get the one you wanted. That invents a number nobody can check.
Britain's 2023 official survey shows why the separation matters. Across all adults, the highest problem-gambling rate was not in the youngest group. It was in 25-to-34-year-olds, at 5.2%, declining with age to 0.2% among those 75 and over. Men were at 3.3% and women at 1.8%. Both true, neither combined.
That figure counts all adults, including everyone who never gambles. The same report also gives the rate among people who did gamble in the past year, and on that base the order changes: 18-to-24-year-olds are highest, at 9.1%, with 25-to-34-year-olds just behind at 8.4%.
So the survey supports both answers, depending which denominator you pick. Across everybody, the peak is 25 to 34. Across people who gamble, the peak is 18 to 24. My reading of why: fewer young adults gamble at all, so dividing by all of them drags the rate down, while dividing by only the ones who do pushes it up.
It is worth knowing where those second numbers live. They are values in a chart. The report's text never states them, and says only that "similar patterns were observed" among past-year gamblers, which is a generous way to describe a peak that moves to a different age group.
So the missing cell is the finding.
To be clear about what I am and am not saying. I am not saying young men are fine. I am not saying the claim is wrong. I am saying the measurement that would settle it has not been published, so anyone stating it as established has gone further than the evidence goes.
That is an unsatisfying thing to write and a worse thing to put in a headline. It is also the only position the published data supports.
And the measured numbers are smaller
The third claim is a rate, and this is where the distance between circulation and measurement is widest.
Two large Swiss studies measured young men directly. The first surveyed 2,349 army conscripts aged 18 to 24, between August 2023 and January 2024. Among those who had gambled in the past year, 4.4% scored as problem gamblers, which is 2.1% of all the young men surveyed. The second followed 5,352 young Swiss men to around age 28, with its data collected between April 2019 and November 2020: 28.5% had gambled in the past year, and 0.8% met the clinical criteria for gambling disorder.
Those dates matter, and they matter most to American readers. The United States began opening up sports betting in 2018, and the market here has moved quickly since. The older Swiss figure comes from a market that looked different, and neither study is a forecast for this one. They are the best direct measurements of young men I could find, from somewhere else.
Set them against the figures that circulate about young men and the direction is consistent: the numbers measured on young men themselves are smaller than the numbers attached to them.
The obvious question is why I am quoting Switzerland at an American audience. Because there is no American version of these numbers. The United States has run the largest gambling experiment of any rich country in the last decade, and what it has measured is the damage rather than the prevalence.
There is good work on what legalisation did. One study tracked consumer credit records across states as sports betting rolled out and found bankruptcies, collections and delinquencies rising. Another, published this year, used a national health survey covering millions of responses and found that in states which legalised online sports betting, young men who already binge drink did so about 10% more often.
Both are about consequences. Neither tells you what share of young American men have a gambling problem. For that, the best thing anyone has is the 801-person poll from the last entry, whose young-male subgroup is small enough that its size was never published.
So the honest summary for an American reader is this. We can see the wreckage reasonably well. We cannot see how many people are in it.
The conscript survey covers the French-speaking region only, which its authors do not flag as a limit and I will. The second study recruited across three centres in both language regions. Both are cross-sectional, and they use different instruments at different ages, so they are not a trend and I will not line them up as one.
What happened when Switzerland opened the market
This one earns a section rather than a footnote, because it is the closest thing in the week's reading to the question an American reader actually has.
Switzerland legalised online casinos in 2019. The same research team surveyed the same population of conscripts before and after. Among those who gambled, problem gambling went from 3.6% to 4.4%, a change the authors themselves call a slight, non-significant increase. Participation rose and moved online. The problem rate did not clearly follow.
Two things stop that from being reassurance. The later survey covered 18-to-24-year-olds where the earlier one covered 18-to-22, so the two are not quite like for like. And one country's experience with online casinos is not a prediction about another country's experience with sports betting. It is one data point, from one market, measured carefully.
A smaller note from the same body of work: among young Swiss men who gambled, the ones with the most symptoms were not the ones who gambled online. They were the ones doing both, online and offline, and that held after adjusting for how much time and money they spent. The authors are explicit that a cross-sectional study cannot tell you which way that runs.
What I retired
One claim came out of the library this week. An earlier paper of mine described a prospective Swiss cohort in which online gambling and impulsivity predicted later problems. Going back to it, that study does not exist as described: the Swiss gambling paper is cross-sectional and never measures impulsivity, and the impulsivity research is on adolescent boys in Montreal and Baltimore. The sentence had fused two separate literatures into one citation.
Nothing downstream was built on it. It is gone, and it is logged.
Where this leaves the question
Against young women: yes, consistently, across three countries.
Against older men: unknown, because the comparison is not published, and that absence is a finding about the evidence rather than a fact about young men.
At what rate: lower than what circulates, wherever young men were measured directly.
One more thing, underneath all of it
Britain's survey is the best single source above, and it has trouble reaching the people this entry is about.
In the 2023 annual report, 4% of the men who responded were aged 18 to 24. That age group is 11% of the male adult population. Young women were much closer to their share, at 7% of female respondents against 10%. The overall response rate was 19% against a target of 22%, and the report says plainly that those in the youngest age group were the least likely to take part. The same pattern repeats across six published waves, with young men running between 4% and 6%.
The data is weighted to correct for that, and weighting does fix the share. It cannot manufacture precision the responses never contained. So every figure above that comes from a national survey rests on fewer young men than you would assume.
What is next
The queue says the next question is why. If young men are being pulled toward this harder than other people, what is doing the pulling, and why now? Phones, always-on apps, and the fact that a betting ad is now part of watching sport at all.
That is the next entry. The survey-reach problem goes on the list behind it, because a series that keeps saying the data is thin eventually has to ask why it is thin.
Sources and how I checked
Every number in this entry was opened at its original source. The table lines up what the AI search summary said against what the sources actually say. Where a figure was not opened, it is listed as unverified rather than used.
| What the search said | What the source says | How I checked |
|---|---|---|
| “In Canada, 3.3% of young men aged 15 to 24 are moderate-risk or problem gamblers, against 1.1% of young women” | Holds up. But the study was published in 2007 and the summary presented it undated, linked to an unrelated 2025 paper. It also includes 15-to-17-year-olds. | Opened at source (abstract-level (CRDCN record)), 2026-10-15 |
| “In Sweden, 3.3% of young men aged 16 to 24 against 1.1% of young women” | Holds up, but it measures something else: the incidence of a FIRST episode in a year, not how many people have a problem. The near-identical digits to the Canadian figure are a coincidence between two different measures, not corroboration. | Opened at source (abstract), 2026-10-15 |
| “Young men are most at risk” | Not shown by the one large survey that comes closest. In Britain's 2023 official survey, across all adults the highest rate was in 25-to-34-year-olds (5.2%), not the youngest group. Sex and age are reported separately and cannot be combined. | Opened at source (full section text plus Figure 16 chart values (web report)), 2026-10-15 |
| “Youth problem gambling prevalence in Europe ranges from 0.2% to 12.3%” | Traced to Calado, Alexandre & Griffiths (2017), a review of 44 studies. It is about adolescents, not young men, it covers studies from 2000 onward, and the authors attribute the width of the range to differing instruments and cut-offs. Not opened at origin. | Not verified |
| “10% of men aged 16 to 34 gambled in more than seven different ways in the past year” | Does not hold up. The 10% traces to an advocacy article rather than the primary report; the Gambling Commission's own summary of the same survey gives 4% of men for seven or more activities; and the threshold wording differs between sources. It is also a breadth-of-gambling measure from 2012, not problem gambling. | Not verified |
| “In one wave of the British survey, 18-to-24-year-olds were 5% of the male sample against 11% of the male adult population” | Holds up, with the figure corrected. The 2023 annual report says 4% of responding males were 18 to 24, against 11% of the male adult population; later waves run 5-6%. Young women were closer to their share, at 7% against 10%. | Opened at source (full section text (web report)), 2026-10-15 |
| “Men are 3.4 times more likely than women to gamble at a problem level” | The largest meta-analysis (366 samples, 68 countries) puts it at 2.8% of adult men who gambled against 1.2% of adult women, conditional on gambling. Like every other source here it reports sex and age separately and never together, so it carries no young-men figure either. | Opened at source (full text), 2026-10-15 |
Sources
- Tomei, A., Bieri, M., Porchet, C. & Simon, O. (2026). Gambling progression in young adults following online casino-gambling legalization in Switzerland. J. Gambling Studies 42(1):249-263. doi 10.1007/s10899-025-10435-6
Measured on: Swiss Army conscripts at the mandatory recruitment programme, French-speaking region; all male; n = 2,349 analysed (98.2% response) · 18-24 (mean 20.6) · Switzerland · What I read: Opened at source (full text), 2026-10-15 · Funding: Programme intercantonal de lutte contre la dependance au jeu (PILDJ); authors declare no conflict of interest; no operator funding reported - Marmet, S., Studer, J., Wicki, M., Khazaal, Y. & Gmel, G. (2021). Online gambling's associations with gambling disorder and related problems in a representative sample of young Swiss men. Frontiers in Psychiatry 12:703118. doi 10.3389/fpsyt.2021.703118
Measured on: C-SURF cohort, young Swiss men enrolled at army recruitment centres (Lausanne, Windisch, Mels); n = 5,352 · mean 28.3 (SD 1.3) · Switzerland · What I read: Opened at source (full text), 2026-10-10 · Funding: Swiss National Science Foundation; authors declare no commercial or financial relationships - Gambling Commission (2024). Gambling Survey for Great Britain: annual report 2023 (official statistics), section 'Problem Gambling Severity Index'. Published 25 July 2024
Measured on: Adults 18+ in Great Britain; 9,633 participants (5,790 past-year gamblers), unweighted · 18+, in 10-year bands from 18-24 · Great Britain · What I read: Opened at source (full section text plus Figure 16 chart values (web report)), 2026-10-15 · Funding: Gambling Commission (regulator); funding mechanism not stated on the page - Huang, J.-H. & Boyer, R. (2007). Epidemiology of youth gambling problems in Canada: a national prevalence study. Canadian Journal of Psychiatry 52(10):657-665
Measured on: Canadian youth in the Canadian Community Health Survey: Mental Health and Well-Being; n = 5,666 · 15-24 · Canada · What I read: Opened at source (abstract-level (CRDCN record)), 2026-10-15 · Funding: not reported - Froberg, F., Rosendahl, I.K., Abbott, M., Romild, U., Tengstrom, A. & Hallqvist, J. (2015). The incidence of problem gambling in a representative cohort of Swedish female and male 16-24 year-olds, in comparison with 25-44 year-olds. J. Gambling Studies 31(3). doi 10.1007/s10899-014-9450-9
Measured on: Swedish Longitudinal Gambling Study (Swelogs), people with no history of problem gambling at baseline; 4,358 selected aged 16-44 · 16-24 vs 25-44 · Sweden · What I read: Opened at source (abstract), 2026-10-15 · Funding: not reported (full text paywalled) - Gambling Commission (2024). Gambling Survey for Great Britain: annual report 2023, Response to the survey
Measured on: Adults 18+ in Great Britain responding to the GSGB; unweighted sample composition · 18-24 vs all adults · Great Britain · What I read: Opened at source (full section text (web report)), 2026-10-15 · Funding: Gambling Commission (statutory regulator); official statistics - Tran, L.T., Wardle, H., Colledge-Frisby, S., Taylor, S., Lynch, M., Rehm, J., Volberg, R., Marionneau, V., Saxena, S., Bunn, C., Farrell, M. & Degenhardt, L. (2024). The prevalence of gambling and problematic gambling: a systematic review and meta-analysis. The Lancet Public Health 9(8), e594-e613
Measured on: Adults and adolescents in 366 representative samples across 68 countries; mean age 29.7 · adults vs adolescents only · 68 countries and territories · What I read: Opened at source (full text), 2026-10-15 · Funding: not recorded on the open-access copy - Hollenbeck, Larsen & Proserpio (2024). The Financial Consequences of Legalized Sports Gambling. SSRN 4903302
Measured on: US consumers in the University of California Consumer Credit Panel · all adults; subgroups include men under 45 by county income · USA · What I read: Opened at source (full text (July 2024 version)), 2026-10-18 - American Institute for Boys and Men (2025). How sports betting can harm young men
Measured on: US consumers in the University of California Consumer Credit Panel · all adults; subgroups include men under 45 by county income · USA · What I read: Opened at source (full text (July 2024 version)), 2026-10-18 · Funding: Advocacy nonprofit; funders not checked - Dasgupta, K. & Ghimire, K. (2026). Gambling and Substance Use: Early Evidence From Sports Betting Laws. Health Economics Letters. doi 10.1002/hec.70088
Measured on: US adults in the Behavioral Risk Factor Surveillance System; young-men subsample aged 35 and under · 35 and under · USA · What I read: Opened at source (full text), 2026-10-18 · Funding: not stated on the article; authors at the Federal Reserve Board and University of Cincinnati - Dellosa & Browne (2024). The influence of age on gambling problems worldwide: a systematic review and meta-analysis of risk among younger, middle-aged, and older adults. J. Behavioral Addictions
Measured on: Problem-gambling prevalence samples in 21 studies; 156,249 participants (47.6% male) · Grouped 18-35 / 30-55 / 45-65 (overlapping) · 18 countries (Europe 52%, Asia 19%, Oceania 19%, North America 10%) · What I read: Opened at source (abstract), 2026-10-19
What I couldn’t verify
- “Youth problem gambling prevalence in Europe ranges from 0.2% to 12.3%”: Traced to Calado, Alexandre & Griffiths (2017), a review of 44 studies. It is about adolescents, not young men, it covers studies from 2000 onward, and the authors attribute the width of the range to differing instruments and cut-offs. Not opened at origin.
- “10% of men aged 16 to 34 gambled in more than seven different ways in the past year”: Does not hold up. The 10% traces to an advocacy article rather than the primary report; the Gambling Commission's own summary of the same survey gives 4% of men for seven or more activities; and the threshold wording differs between sources. It is also a breadth-of-gambling measure from 2012, not problem gambling.
Corrections
None yet. If something here changes, the change is logged here with its date.
How this was made
This series is researched with an AI desk: it searches, reads, pulls the numbers and drafts, including this entry. I edit and decide what is published.
How I check sources · The research library
If gambling is costing you or someone close to you more than it should, the US National Problem Gambling Helpline is free and confidential: call or text 1-800-MY-RESET (1-800-697-3738), or chat at ncpgambling.org.