Send to

Choose Destination
Addict Behav. 2016 Jun;57:13-20. doi: 10.1016/j.addbeh.2016.01.010. Epub 2016 Jan 22.

Gambling transitions among adult gamblers: A multi-state model using a Markovian approach applied to the JEU cohort.

Author information

BALANCED Clinical Investigation Unit ("BehaviorAL AddictioNs and ComplEx mood Disorders"), Department of Addictology and Psychiatry, University Hospital of Nantes, France; EA 4275 SPHERE "MethodS for Patients-centered outcomes and HEalth REsearch", University of Nantes, France.
EA 4430 CLIPSYD "CLInique PSYchanalyse Développement", University of Paris Ouest Nanterre La Défense, France; Louis Mourier Hospital of Colombes, Assistance Publique - Hôpitaux de Paris (APHP), France.
Marmottan Medical Center, GPS Perray-Vaucluse, Paris, France.
Department of Adult Psychiatry, Sainte-Marguerite University Hospital of Marseille, France.
Psychiatry Laboratory, Sanpsy CNRS USR 3413, University of Bordeaux and Charles Perrens Hospital, Bordeaux, France.
Psychiatry Department, University Hospital of Clermont-Ferrand, France.
Psychiatry and Addictology Department, Paul Brousse University Hospital of Villejuif, Assistance Publique - Hôpitaux de Paris (APHP), France.
EA 4275 SPHERE "MethodS for Patients-centered outcomes and HEalth REsearch", University of Nantes, France; Unit of Methodology and Biostatistics, University Hospital of Nantes, France.



The aim of this paper is to study transitions between two states of gambling in adulthood (problem gambling and non-problem gambling) and to identify factors that might influence these transitions.


Data for this 2-year long longitudinal study were collected in a French Outpatient Addiction Treatment Center, in gambling establishments and through the press. Both problem gamblers and non-problem gamblers were evaluated using a structured interview and self-report questionnaires. The statistical analysis was carried out using a Markovian approach.


The analyzed cohort consisted of 304 gamblers with 519 observed transitions. Participants with no past-year gambling problems (based on the DSM-IV) had a probability of about 90% of also having no past-year gambling problems at the following assessment, whereas the observed percentage of problem gamblers transitioning to non-problem gambling was of 48%. We reported (i) vulnerability factors of transitioning to problem gambling (such as an anxiety disorder or an Attention Deficit Hyperactivity Disorder (ADHD) during the childhood), (ii) protective factors for non-problem gamblers, (iii) recovery factors (such as ongoing treatment and younger age) and (iv) persistence factors of a gambling problem (such as a persistent ADHD).


The status of problem gambler is unstable over time, whereas we found stability among non-problem gamblers. Our findings suggest the existence of vulnerability and protective factors in gambling. These results lead to think about preventive actions and adaptive care, such as cognitive-behavioral therapy or researching gambling problems in people with an anxiety disorder or ADHD.


Gambling; Longitudinal study; Markov process; Prevalence; Risk factors; Transitions

[Indexed for MEDLINE]
Free full text

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center