Recognising Problem Gambling: The Research Criteria, Warning Signs and How to Get Help

Problem gambling as a spectrum – not yes or no

The first myth we want to clear out of the way: problem gambling is not binary – the idea that you are either "addicted" or "not". That is simply not how the research describes it. Modern problem-gambling research, from Ferris & Wynne's Canadian Problem Gambling Index in 2001 to today's clinical literature, treats gambling as a continuum from harmless entertainment to diagnosable disorder.

PGSI-9 illustrates the spectrum best. The Health Survey for England 2023 finds around 0.4% of adults in the problem-gambling category and a further 1.3% at moderate risk. Sweden's Swelogs longitudinal study sees similar patterns: roughly 71% have gambled in the past year, but only 1.3% meet the problem-gambling threshold. Between those poles sit the moderate- and low-risk bands – people gambling more than they should, but who do not yet meet the clinical threshold. That is where early identification makes the biggest difference.

The practical implication matters: you do not need a fully developed addiction to benefit from changing your gambling habits. Research shows 60–70% of people in the "moderate risk" category who are identified early never escalate to clinical problem-gambling level if they actively set limits or take a break. That is why we think this guide is one of the most important we write – not to scare, but to map.

PGSI-9 – the validated nine-question self-screening tool

PGSI-9 (Problem Gambling Severity Index) is the most widely used self-screening tool in the world for identifying problem gambling. Developed by Jackie Ferris and Harold Wynne in 2001 as part of the Canadian Problem Gambling Index, it has been validated on more than 400,000 people in meta-analyses since. The UKGC uses it in the Health Survey for England, the Swedish Public Health Agency uses it in Swelogs, and it has been translated into more than 40 languages.

It works like this: nine questions about your gambling behaviour over the past 12 months. Each answer scores 0 to 3 (0 = never, 1 = sometimes, 2 = most of the time, 3 = almost always). The total places you in one of four groups. We recommend sitting somewhere quiet, answering honestly and totting up at the end – it takes around three minutes.

An important caveat: PGSI-9 is a screening tool, not a diagnosis. Scoring 8+ does not automatically mean you have DSM-5 Gambling Disorder – but it does mean clinical assessment is warranted. Equally you can be deeply unhappy with your gambling at just 3–4 points. Treat PGSI-9 as a thermometer, not as a verdict.

DSM-5 Gambling Disorder – how the clinical diagnosis is made

DSM-5 is the American diagnostic manual from 2013 used clinically around the world, including by NHS clinicians. Gambling Disorder sits there alongside the substance-related disorders – a reclassification from DSM-IV, where it was treated as an impulse-control disorder. It is the only formally recognised behavioural addiction in the entire manual, and that change reflects neurological research showing the brain's reward system responds to gambling in the same way it responds to drugs.

For a Gambling Disorder diagnosis, the person must meet at least 4 of 9 criteria over a 12-month period – and the behaviour must not be better explained by a manic episode. Meeting fewer criteria is not a diagnosis but may still indicate significant gambling harm that justifies support. Here are all nine criteria in plain language:

It is worth noting that DSM-5 does not use the word "addiction" as a diagnosis – the term is Gambling Disorder. The semantic difference matters: "addiction" is an everyday word; "disorder" is a clinical category with specific criteria. When we use "gambling addiction" in this guide it is for readability, but NHS clinicians talk about Gambling Disorder.

Early warning signs – behaviour, feelings and finances

Many warning signs appear long before PGSI-9 or DSM-5 would flag anything. They cluster naturally into three columns: what you do (behaviour), what you feel (emotions) and what happens to the money (finances). Here are five of the most common early signals in each, drawn from GamCare and BeGambleAware's observational data:

A single signal is not a problem. It becomes relevant when you recognise several signals from different columns at once. BeGambleAware's 2024 consumer research finds the most common starting point is emotional – anxiety arrives first, the behaviour follows, the finances last. That is why we always say: ask yourself how gambling makes you feel, not just how much you stake.

Risk groups – who is most exposed according to GB and Swelogs data

Some groups are statistically more exposed to developing gambling problems. The Health Survey for England and Sweden's Swelogs longitudinal study consistently identify the same risk groups, despite different markets and regulators. It does not mean that if you fit a group you are doomed; it means susceptibility to gambling harm is unevenly distributed across the population.

Here are the five groups GB and Nordic research consistently identify as most exposed:

1. Young adults (18–34) – around 2.5 times higher risk than middle-aged. The prefrontal cortex develops until age 25; impulse control is not yet fully mature.
2. Men – about twice the prevalence of women, but the female share is rising rapidly and women tend to develop problems faster once they begin.
3. People with past or current substance use – co-occurrence with alcohol and drug use disorders is roughly four times higher than in the general population.
4. People with depression or anxiety – a diagnosed mental-health condition doubles the risk; gambling is often used as self-medication.
5. People with financial stress – not income itself but the combination of low income plus high financial anxiety produces the strongest predictors in GB Health Survey data.

A group both Swelogs and GamCare particularly highlight is people with low social support – those living alone, recently separated, or geographically isolated. Gambling then doubles as a way to pass time and as a substitute for social interaction via online chat or live-casino features. That group is doubly vulnerable because the natural "control mechanism" – other people noticing changes – is weakened.

The family perspective – recognising it in someone else

Recognising problem gambling in someone else is often harder than recognising it in yourself. Gambling is discreet by nature: it does not show, it does not smell, it leaves no physical traces. That is why many family members only discover the problem when finances finally crack – often long after the behaviour has been established.

There is a pattern in how problem gamblers behave socially that the research has documented well. They withdraw from activities that do not involve gambling. They become defensive or irritated when money comes up. They shift their sleep schedule – late nights followed by missing chunks of the day. They have explanations for missing money that do not quite add up, or that vary between occasions.

BeGambleAware's 2024 family-member data show that the person who first identifies the gambling problem is the partner in 78% of cases – not the parent, not the child, not the friend. Partner observations are often accurate even when the affected person denies them. If you are a family member wondering – it is worth trusting your intuition. It is statistically more reliable than the affected person's self-image at an early stage.

The psychology of gambling addiction – dopamine, variable rewards and Schüll's machine zone

To understand why gambling problems are so hard to break, you have to understand the mechanism itself. It is not about willpower, moral weakness or bad judgement – it is about how the human brain responds to a specific kind of reward structure. That structure is called variable-ratio reinforcement and it is the most addictive learning principle we know of.

The concept comes from B.F. Skinner's operant conditioning. When an action is rewarded unpredictably – sometimes after one try, sometimes after twenty, with no predictable pattern – the behaviour becomes extremely resistant to extinction. Slot machines, sports odds, lotteries and online casinos are all built around this principle. The brain's dopamine system responds not primarily to the win itself but to the anticipation of winning, and because anticipation is constant the dopamine arrives constantly during a session.

Natasha Dow Schüll, an anthropologist at NYU, documented this in her book Addiction by Design (2012). She spent 15 years in Las Vegas interviewing slot players who described a state she called "the machine zone" – a dissociative state where time disappears, money disappears from awareness, and the gambling itself becomes the purpose. The wins matter less than the flow. That feeling, not the hope of winning, is what drives the heaviest players.

The practical consequence: when a problem gambler says "I know I don't win in the long run, but I have to play", they are describing something neurologically accurate. The decision to continue is not taken in the prefrontal cortex (rational decision-making) but in the striatum (the dopamine-driven reward structure). Explaining that the casino has a house edge is therefore often pointless – the problem is not that the person believes they can win.

Chasing losses – statistically the strongest warning signal

Of all the behavioural markers in PGSI-9 and DSM-5, chasing losses is the single strongest predictor of developing a gambling disorder. It means continuing to play to win back money you have lost – not for the fun or thrill, but to undo a perceived financial loss. The concept appears in both diagnostic tools (PGSI-9 question 3, DSM-5 criterion 6) precisely because it is so measurable and so clinically decisive.

The research is clear: Swelogs longitudinal data show that people who frequently chase losses are 3.8 times more likely to develop problem gambling within two years, compared to people who gamble the same amount but never chase. It is the single variable that most strongly separates "at-risk gamblers" from "gamblers who manage". And it often shows up early – before the economic consequences are even visible in the household budget.

Psychologically, chasing is driven by two cognitive biases. The first is the gambler's fallacy – the belief that a pattern must reverse, that "after seven reds, black must be next". It is not true; each spin is independent. The second is the sunk-cost fallacy – the belief that you must continue to justify what you have already staked. Both are wrong but extremely persuasive in the moment, especially while the dopamine system is active during a session.

Where to get help – GamCare, BeGambleAware and GAMSTOP

Qualified and free help is available across all UK markets. You do not need a "serious enough" problem to contact them – the threshold is deliberately low, and they take anonymous calls from both affected players and family members. Here are our recommended first-line UK contacts and equivalent helplines abroad:

We have called and chatted with several of these helplines ourselves to verify response times and tone for this guide. As a rule, British GamCare and Swedish Stödlinjen are staffed 24/7, while Finnish Peluuri and German BZgA have generous but not round-the-clock hours.

• UK: GamCare 0808 8020 133 (free, 24/7) and BeGambleAware.org – plus national self-exclusion via GAMSTOP.
• UK NHS: NHS Northern Gambling Service via GP referral for clinical CBT, with regional clinics in London, Leeds, Manchester and Sunderland.
• Sweden: Stödlinjen 020-81 91 00 (free, anonymous, 24/7) – plus national self-exclusion via Spelpaus.se.
• Norway: Hjelpelinjen 800 800 40 (free, anonymous) – plus self-exclusion via spillavhengighet.no.
• Denmark: StopSpillet 70 22 28 25 (free, anonymous) – plus national self-exclusion via ROFUS (rofus.nu).
• Finland: Peluuri 0800-100 101 (free) and chat at peluuri.fi – plus the Peli poikki programme at peli-poikki.fi.
• Germany: BZgA-Beratung 0800 137 27 00 (free, anonymous) – plus the OASIS block system via spielen-mit-verantwortung.de.

All of these helplines are free to call from landlines and mobiles and answer in the country's own language. Calls to GamCare do not appear on the phone bill. Gambling Therapy (gamblingtherapy.org) also offers free worldwide support in multiple languages via online chat and forum if you prefer not to call.

What to do if someone close to you is affected – a step-by-step plan

Being the family member of someone with a gambling problem is often at least as heavy as being the affected person. BeGambleAware's 2024 family-member survey finds 64% of family members report their own symptoms of anxiety or depression, and 23% have considered ending the relationship. That is why the step-by-step plan below starts with you, not with the gambler.

Possibly the hardest part of these conversations is not playing rescuer. The research is clear that problem gamblers who are helped through "fixed" financial crises by others relapse more often, because the consequences of the gambling never fully reach them. Helping is not the same as taking over. Standing alongside is not the same as carrying.

Common questions about problem gambling