In a nationally representative survey of smoking and alcohol use in Great Britain, our pilot gambling questions collected during one month of data collection returned similar estimates for overall gambling participation, and for at-risk and disordered gambling as other national population surveys [1, 6]. Approximately half of adults reported some gambling activity in the past year, and descriptively the prevalence of smoking (16%) and increasing and higher risk drinking (41%) was higher in those who gambled compared with those who did not (13% and 28%, respectively).
One in 14 people who gambled were classified as being at any risk of harm from gambling, but our data were likely insensitive to detect associations between any-risk of harm from gambling, smoking and higher risk drinking, if true associations ranging from OR = 0.5 to 1.9 existed. Although gambling at any level of harm is consistently associated with alcohol consumption [2, 26], the aetiology of this relationship is likely multidimensional. Observed associations in the wider literature may reflect common underlying genetic, social and environmental determinants [27, 28], but also involve bi-directional causality whereby frequency of gambling is higher under the influence of alcohol [27]. Alcohol may be used as an avoidant coping mechanism following losses [29] and as a cued response following wins [30]. The mechanism through which alcohol consumption might lead to disordered gambling requires more research. For example, a recent review examining the salient hypothesis that acute alcohol consumption leads to harm from gambling by encouraging greater risk taking concluded that there was no reliable effect [31].
In 2009 a review into the comorbidity of smoking and gambling concluded that comorbidity was highly prevalent [32]. However, an evidence review published in 2023 conducted by Public Health England concluded that cigarette smoking was not associated with gambling among adults [2]. While this may not hold true in certain priority subgroups, further data examining this issue in the STS/ATS could provide important information at the population-level.
The paucity of data in our survey on individuals experiencing disordered gambling reflects the relatively small proportion of the population falling into this category, but also that population surveys cannot comprehensively capture relatively rare behaviours - like disordered gambling [33], or injecting drug use [34] – which are more common in subgroups of the population who fall outside of traditional sampling frames [35]. Nonetheless, because they are more numerous, the majority of harms from gambling at the population-level is driven by those classified as low, and moderate risk of harm [36]. Understanding the relationships between gambling behavior and other licit and commercially influenced addictive behaviors like smoking and alcohol use, and gambling at all levels of risk remains an important endeavour. While pilot data were collected in one survey month, extending data collection in a longer monthly time-series would allow these intersections to be interrogated with greater accuracy and reliability.
Finally, while there was wide uncertainty in our estimates for expenditure on gambling among those at any risk of harm from gambling, the outliers driving this uncertainty remain plausible given the extreme spending that can occur in those experiencing disordered gambling [36]. Indeed, due to the rise in online gambling in recent years, in their recent white paper the UK government has proposed introducing financial risk-checks for moderate to high spending [37]. While our estimates should be interpreted with caution, there was a signal for higher expenditure on gambling among those catagorised as drinking at an increasing and higher risk level. If true, this pattern of spending would conform to studies highlighting a positive relationship between increasing alcohol consumption and gambling spend [7].
In conclusion, the collection of pilot data on gambling in a population-level survey on smoking and alcohol use yielded estimates of gambling participation and at-risk and disordered gambling that are similar to other population-level surveys. Further data collection would help elucidate the intersections more reliably between gambling, smoking and alcohol use, and inform population-level approaches to reduce the harms to public health conferred by these behaviors.