Behavioural Addiction at the Severe End: 93 Young Patients in a Moscow Case Series
- Retrospective clinical series at the Mental Health Research Centre, Moscow: 93 young patients with pathological gambling, internet addiction or both, examined across 2020-2024, mean age 21.5 ± 3.2 years. This is a help-seeking sample from one academic centre, so it yields no prevalence estimate whatsoever.
- Gambling-predominant group, n = 21 (22.6% of the series): affective disorders in 71.43%, suicidal activity in 38.10%, and the highest impulsivity of the three groups on the Barratt scale, 96 points (94.00-99.00) – a figure the abstract reports without naming either the statistic or the version of the scale.
- Internet-addiction group, n = 43 (46.2%): neurotic disorders of anxiety and obsessive-compulsive type predominated at 72.09%, with eating disorders in 20.93%.
- Mixed group, n = 29 (31.2%), was the most impaired: comorbid chemical dependence in 62.07%, affective pathology in 86.21%, non-suicidal self-injury in 68.97%. The authors state that they identified prognostic markers; the design is retrospective and has no external control group, so what it yields are comparative group profiles, not a validated prediction.
Most of what gets said about gambling and screens in Russia runs on survey arithmetic: a questionnaire, a cut-off, a percentage of the general population. This paper points the telescope the other way. Krylova, Kuleshov and Kaleda at the Mental Health Research Centre in Moscow describe the 93 young people who actually arrived at a specialist psychiatric service over five years with non-chemical addiction as the presenting problem.
Three groups, one clinic
The method is clinical-psychopathological, psychometric and retrospective. Patients were examined between 2020 and 2024, mean age 21.5 ± 3.2 years, all with non-psychotic disorders. They sorted into three groups. Gambling predominant: n = 21, or 22.6%. Internet addiction: n = 43, or 46.2%. Mixed: n = 29, or 31.2%.
The gambling group carried a family history of behavioural and chemical addiction in first- and second-degree relatives. Affective disorders appeared in 71.43% and suicidal activity in 38.10%. Impulsivity on the Barratt scale was reported as 96 points (94.00-99.00), the highest of the three groups; the abstract gives that figure without saying whether it is a median with an interquartile range or a mean with an interval, and without naming the version of the instrument. The internet-addiction group looked different in kind, not merely in degree: anxiety and obsessive-compulsive type neurotic disorders in 72.09%, eating disorders in 20.93%. The mixed group had the latest onset of addictive behaviour, the worst social adaptation, chemical dependence layered on top in 62.07%, affective pathology in 86.21% and non-suicidal self-injury in 68.97%.
Hold those percentages against their denominators before they harden into facts. In a group of 21, 38.10% is eight patients. In a group of 29, 68.97% is twenty. Every proportion in the paper resolves to a whole patient, so the arithmetic is exact rather than inflated. What the two decimal places cannot do is make the ground firmer: one additional or reclassified case moves a group percentage by nearly five points. The figures are precise; the base is small.
What this changes at intake
The practical value here is not epidemiological, it is descriptive triage. When a young patient presents with problem gambling, this cohort says look immediately for mood disorder and for suicidality, and take the impulsivity seriously as a treatment target rather than a personality footnote. When the presentation is problematic internet use, the comorbid neighbourhood is different: anxiety, obsessive-compulsive phenomena, disordered eating. Screening the gambling patient for anxiety alone, or the internet patient for suicide risk alone, misses the pattern this sample describes.
The mixed presentation deserves the most caution. Latest onset, chemical dependence in nearly two thirds, self-injury in over two thirds. If your patient's behavioural addiction sits alongside substance use, the authors read that combination as a prognostic marker of a worse trajectory, and say so in those terms. It is a hypothesis generated by a retrospective series without an external control group, not a validated prediction rule, and it should be held as a reason for closer follow-up rather than as an established risk score.
There is a second lesson, and it cuts against the reflex to dismiss all of this as moral panic. Ninety-three young people in five years is a small number, and it is precisely the point. These are not heavy users mislabelled by a questionnaire. They are patients with suicidal behaviour, self-injury and layered dependence who reached a specialist service. Over-diagnosis in the general population and under-service at the severe end are not competing claims, and this paper speaks only to the second. Your job at intake is to work out which of the two you are looking at, and the comorbidity profile is the most useful signal this paper hands you.
Ninety-three patients in five years is not a prevalence estimate; it is a description of what behavioural addiction looks like once it is severe enough to reach a psychiatrist.
Single-centre retrospective case series with no external control group and no population denominator, so nothing here supports any claim about how common behavioural addiction is. The published abstract gives the Barratt impulsivity figure without naming the statistic or the version of the instrument, and says nothing about the psychometric properties of its Russian adaptation; if the version used is the 30-item BIS-11, whose range runs from 30 to 120, then 96 points with an interval of 94.00 to 99.00 in a group of 21 is both unusually high and unusually narrow, and the full text is behind a paywall.