Health Survey for England 2018 reports on adult gambling behaviours

NHS Digital (formerly the NHS Information Centre) has published the 28th Health Survey for England (“HSE”). Contrary to what might be imagined from a number of popular media reports, the survey data contains some encouraging signs that gambling industry efforts to address gambling-related harms have been taking effect.

You can access here a Quick Guide to the HSE 2018. The complete resources are available here.

The HSE is a series of annual surveys that provide regular information that cannot be obtained from other sources about the public’s health and health-related behaviour. It includes a report on adults’ health-related behaviours, including gambling behaviours.

Key findings of the HSE in relation to gambling are:

  • 54% of adults had participated in some form of gambling activity during the previous 12 months
  • 40% of adults had participated in some form of gambling activity during the previous 12 months (excluding National Lottery)
  • 15% of men had participated in online gambling in the previous 12 months, compared with 4% of women.

The section of the HSE dealing with gambling behaviours (that you can download below) states as follows:

Gambling behaviour


Great Britain has one of the most accessible gambling markets in the world. Opportunities to gamble exist on most high streets and, with the spread of the internet, in virtually every home.

The gambling industry is increasingly being called upon to do more to protect participants and prevent problem gambling from occurring, and the National Responsible Gambling Strategy emphasises the need for joint action between industry, government, healthcare providers and other public bodies to tackle gambling-related harm.

This chapter looks at levels of participation in gambling, and whether these vary by age and sex. It covers overall participation in any form of gambling. For all gambling activities, participation was defined as having spent money on the activity over the past year. Participants were shown a list of gambling activities and were asked to think about any gambling they had done over the past 12 months. The activities included in the list were intended to cover all types of gambling available. However, to allow for the possibility that an activity was missed or that participants may have misunderstood an activity description, an option was provided for participants to mention another form of gambling.

This chapter also presents information about the prevalence of problem gambling among adults aged 16 and over. ‘Problem gambling’ is typically defined as gambling to a degree that compromises, disrupts or damages family, personal or recreational pursuits.


Problem gambling

The HSE uses the following screening tools to identify at risk or problematic gambling: the DSM-IV criteria and the Problem Gambling Severity Index (PGSI). A score of 3 or more for DSM-IV or a score of 8 or more for PGSI is indicative of problem gambling.

At-risk gambling

A score of 1 or more for PGSI is indicative of at-risk gambling. A score of 1 to 2 is considered low risk and a score of 3 to 7 is considered moderate risk.

Problem gambling screening

Diagnostic and Statistical Manual of Mental Disorders, fourth version (DSM-IV

The DSM-IV screening instrument is based on criteria from the fourth edition of the Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-IV). It was created as a clinical diagnostic tool and was not intended for use as a screening instrument among the general population. An adapted version of the DSM-IV for use in a survey setting was developed for the British Gambling Prevalence Survey (BGPS) series and was subject to a rigorous development and testing process, including cognitive testing and piloting.

The DSM-IV contains ten diagnostic criteria ranging from ‘chasing losses’ to ‘committing a crime to fund gambling’. Each item is assessed on a four-point scale, ranging from ‘never’ to ‘very often’. This report follows the scoring method used by the BGPS; each item is coded according to whether the respondent had a positive score, resulting in a total score between 0 and 10.

Among clinicians, a diagnosis of pathological gambling is made if a person meets five out of the ten criteria. Many surveys, when adapting the DSM-IV criteria into a screening instrument for use within a general population survey, have included a further category of problem gambler for those who meet at least three of the DSM-IV criteria.

Problem Gambling Severity Index (PGSI)

The PGSI was designed for use among the general population rather than within a clinical context. It was developed, tested and validated within a general population survey of over 3,000 Canadian residents76. The instrument itself has been subject to critical evaluation and was revised in 2003.

The PGSI consists of nine items ranging from ‘chasing losses’ to ‘gambling causing health problems’ to ‘feeling guilty about gambling’. Each item is assessed on a four- point scale: never, sometimes, most of the time, almost always. Responses to each item are given the following scores: never = 0; sometimes = 1; most of the time = 2; almost always = 3. When scores to each item are summed, a total score ranging from 0 to 27 is possible. A PGSI score of 8 or more represents a problem gambler. This is the threshold recommended by the developers of the PGSI and the threshold used in this report. The PGSI was also developed to give further information on sub-threshold problem gamblers.

PGSI scores between 3 and 7 are indicative of moderate risk gambling and a score of 1 or 2 is indicative of low risk gambling.

Participation in gambling activities in the last 12 months, by age and sex

In 2018, 54% of adults had participated in some form of gambling activity during the previous 12 months. This proportion falls to 40% when gambling on the National Lottery is excluded.

The proportion of adults who had participated in any gambling activities was highest in the 45 to 54 age group at 60% compared with 39% in the 16 to 24 age group.

Among men, the highest proportions were among the 25 to 34 age group (65%) with the 16 to 24 age group being the lowest with 45%. Note: most forms of gambling (excluding lotteries or scratch cards) are illegal for the under 18s and this will affect gambling prevalence in the 16 to 24 age group.

For women, the highest proportions participating in any gambling activity were the 55 to 64 age group, with 59% having participated in gambling activities. When gambling on the national lottery is excluded, women aged between 25 and 64 are most likely to have participated in the last 12 months. As with men, the lowest prevalence of any gambling activity was among the 16 to 24 age group, with 33% having participated.

The most pronounced differences across both age and sex were for online gambling other than the National Lottery.

15% of men had participated in online gambling in the previous 12 months, compared with 4% of women.

After reaching a peak in the 25 to 34 age group, participation in online gambling then gradually declines with age for both sexes. 28% of the 25-34 age group for men had participated in online gambling, compared with less than 5% of those aged 65 and over. Similarly, 9% of women in the 25 to 34 age group had participated in online gambling.

Prevalence of at-risk and problem gambling (DSM-IV and PGSI scores), by sex and age

Using PGSI scores, 0.4% of adults were identified as problem gamblers (score 8+) and 3.6% as problem or at-risk gamblers (score 1+). Using the DSM-IV scores, 0.5% of adults were identified as problem gamblers, defined as DSM-IV score 3+.

Using the PGSI scores, the proportion of men identified as problem or at-risk gamblers is substantially higher than women, with 6% of men and 2% of women identified. The proportion of problem or at-risk gamblers decreases with age from between 5% and 7% in those age 16 to 44 to 1% of those age 75+.

By downloading the document itself, you will be able to gain access to footnotes in relation to the above extract and diagrams illustrating:

  • Participation in gambling activities in the last 12 months, by age
  • Participation in any online gambling (excluding National Lottery) in the last 12 months, by age and sex.

You can access here a supplementary analysis on gambling published by NHS Digital in conjunction with the HSE.

Predictably, certain media commentators sought to find negative aspects within the survey data to justify headlines such as the Mail Online‘s “A deadly gamble on our health: NHS boss Simon Stevens slams gaming giants after millions more turn to betting online and new figures reveal more than HALF of us like a bet”.

That article goes on to quote Stevens saying that: “the new statistics are a stark reminder of how common gambling is in our society, and how easy it is to become addicted”. Had the Mail Online (and others carrying similar headlines) – and, one is tempted to say, Simon Stevens himself – investigated the position more perceptively themselves, they would have seen that a rather more balanced picture could have been painted.

For example:

  • as indicated above, the proportion of adults who had participated in some form of gambling activity during the previous 12 months (i.e. 54%) reduces to quite considerably less than half (i.e. 40%) when the government-sponsored National Lottery gambling activity is excluded,
  • that 54% adult gambling participation statistic has reduced materially (from 62% in 2015 and 56% in 2016),
  • the above-mentioned 0.5% of adults identified as problem gamblers has also reduced (from 0.9% in 2015 and 0.7% in 2016) which can be interpreted as a welcome sign that UK licensed gambling operators have indeed turned a significant corner in embracing the Gambling Commission’s call for safer gambling provision, and
  • in terms of the claim that “millions more” have turned to betting online, the survey data shows that the overall rate of participation in online gambling as a whole has remained the same as recorded in 2016 (i.e. 9.4%), a small upward shift in numbers betting online being counter-balanced by a reduction in numbers gaming online.

SBC News has studied the statistics within that supplementary analysis of the HSE 2018 and (in an article entitled “New NHS figures shed more light on problem gambling levels” that you can also download below) has reported as follows:

Figures from the latest Health Survey for England show that around 13% of the country’s population engaged in online betting with a bookmaker in 2018, with the percentage of people who had gambled overall dropping to 51%.

The Health Survey also measured problem gambling rates using two different measures – DSM-IV and PGSI – both of which recorded rates below the generally accepted rate of 0.7%. Under DSM-IV, the problem gambling rate for 2018 was recorded as 0.5% of the population, with men more likely to have problems than women (0.7% compared to 0.3%). Using the PGSI standard, the problem gambling rate is 0.4%, with men once again higher at 0.6% compared to women at 0.1%.

The PGSI also suggested a 0.8% of the population could be classified as ‘moderate risk’ gamblers. In 2016, the figure for the whole of Great Britain (not just England) was much higher at 1.1%.

The NHS has been incorporating questions around gambling into its Health Surveys ever since the funding was pulled for the National Gambling Prevalence Studies, the last of which was in 2010. Since then, the Gambling Commission has been reliant on the Health Survey to monitor the effect of gambling in society.

Given the 2010 survey had much higher DSM-IV problem gambling prevalence rates for the whole population were 1.5% for men, 0.3% for women and 0.9% overall and PGSI reported rates of 1.3% for men, 0.2% for women and 0.7% overall, this could be construed as good news.

However the chief executive of NHS England Simon Stevens still used the data to call for more money from the gambling industry. He told the media: “These new stats are a stark reminder of how common gambling is in our society and how easy it is to become addicted, particularly with the aggressive push into online gambling,” said Stevens on Tuesday.

“The NHS never stands still as health needs change, which is why we’re rolling out new specialist services to tackle mental ill health linked to gambling addiction, as part of our long-term plan.

“But it is high time that all these firms who spend many millions on marketing and advertising step up to the plate and take their responsibilities seriously.”

Other figures from the report show a range of gambling participation depending on product:

  • Online betting with a bookmaker – 13% of population
  • Betting exchange – 2%
  • Horse races (not online) – 10%
  • Dog races (not online) – 3%
  • Sports events (not online) – 7%
  • Other events (not online) – 3%
  • Spread betting – 1%
  • Private betting – 6%
  • Football pools – 5%
  • Bingo (not online) – 3%
  • Slot machines – 8%
  • Machines in a bookmakers – 4%
  • Casino table games (not online) – 4%
  • Poker played in pubs or clubs – 1%
  • Online gambling on slots, casino or bingo games – 4%