Uncovering the different subtypes of Problem Gamblers: An empirical testing of the Pathways Model and its clinical implications

1
Uncovering the different subtypes of Problem
Gamblers An empirical testing of the Pathways
Model and its clinical implications

• Gupta, Blaszczynski, Nower, Derevensky, Faregh

March 11, 2010 Sacramento, California
2
Context

• After many years of researching and working with
  teens, it became evident that not all were
  similar in their presentation and etiology.
• Their gambling behaviors may have been similar,
  but distinct sub-types of people were emerging.

3
In order to better understand

• Apply a theoretical framework to represent a
  complex phenomenon
• Measure this phenomenon with identifiable and
  measurable units of information
• A scientific approach will never reflect the true
  depth of human behavior, but it may reflect
  enough so that we may advance the current state
  of knowledge

4
Background

• Jacobs General Theory of Addictions (1989)
  postulates that all addictions are a dependant
  state acquired over time by a predisposed person
  in an attempt to relieve a chronic stress
  condition
• Under/ Overactive physiological resting state
• Psychological characteristics such as low
  self-esteem

5
Pathways model

• Blaszczynskis model is based on the premise that
  there exist different typologies of gamblers and
  that it is counter productive to merge distinctly
  different types of people into a single
  heterogeneous group.

6
Nower Blaszczynski (2004)

• Proposed that the Pathways model can also serve
  as
• an effective template for the development of
  early intervention, prevention, targeted
  clinical management strategies for adolescent and
  young adult gamblers.

7

• The Pathways model proposes that there are a
  minimum of three subgroups of problem and
  pathological gamblers with distinct clinical
  features and etiological processes.
• This theoretical model is based on the premise
  that there exist different typologies of gamblers
  and that it is counter-productive to merge
  distinctly different types of people into a
  single heterogeneous group

8
Pathway 1- Behaviorally conditioned youth
problem gamblers.

• These youth develop gambling problems as a result
  of conditioning rather than impaired control.
  They are characterized by an absence of premorbid
  psychopathology. Anxiety can surface as a
  consequence of gambling problems.

9
Pathway 2- Emotionally-Vulnerable Youth Problem
Gamblers

• These young gamblers present with premorbid
  depression and/or anxiety, low self-esteem, poor
  coping, a history of familial neglect or abuse,
  and present with other adverse behaviors.

10
Pathway 3- Antisocial Impulsivist Youth Problem
Gamblers

• These young gamblers present with premorbid
  psychopathology suggestive of neurochemical
  dysfunctions. They share many things in common
  with Pathways 2 individuals, but they are
  distinct in their features of impulsivity and
  antisocial personality disorder which impair
  their overall psychosocial functioning.

11
Science may be described as the art of
systematic over-simplification. -Sir Karl
Raymund Popper

• Testing a model empirically is a very complex
  undertaking.
• This study is the first to attempt to do so, and
  should be considered exploratory in nature.
• This study is limited by its cross-sectional
  nature and needs replication.

12
The sample

• 1133 students from Ontario and Quebec schools
• All English speaking
• 49.5 male, 50.5 female
• Grades 7 through 12

13
Gambling severity

• Non-gamblers 19.3 (n 214)
• Social gamblers 70.4 (n 779)
• At-Risk gamblers 7.2 (n 80)
• PPG 3 (n 33)

14
Instruments

• DSM-IV-MR-J
• GAQ
• Millon Adolescent Clinical Inventory (MACI)
• Conners-Wells self report scale
• Erroneous cognitions questions

15
Erroneous cognitions questions

• When I gamble, I believe that I win most of the
  time.
• When gambling, I use several tricks and
  strategies that give me an advantage over other
  players.
• I have figured out a system that will allow me to
  win frequently when gambling.
• The more I gamble, the better I become.
• At times, I have considered becoming a
  professional gambler because I am skilled enough
  to make a good living.
• When deciding to gamble, I usually do so because
  I am convinced that I will win.
• I dont hesitate to place large bets when I
  gamble because I know that I will eventually win
  it back and more.
• I am usually a luckier person than most others,
  and I believe that it makes me a more successful
  gambler.
• Most gambling games involved high elements of
  skill, and if you play wisely, you will usually
  come out ahead.
• There are things that I can do to increase my
  chances of winning when gambling.
• I have lucky clothing, personal objects, or
  rituals that increase my chances of winning when
  gambling.

16
MACI description

• 160 item self-report inventory designed
  specifically for assessing adolescent personality
  characteristics and clinical syndromes.
• It is useful in the evaluation of troubled
  adolescents, and it may be used for developing
  treatment plans and as an outcome measure.
• -p. 1 manual

17

• The MACI measures
• 12 personality patterns
• 8 expressed concerns
• 7 clinical syndromes
• Reliability
• Internal consistency alpha coefficients range
  from 0.73 to 0.91
• Test-retest obtained a median stability
  coefficient of 0.82

18
Plan of Analysis
19

• An empirical classification of homogeneous
  subgroups by distinctive traits through 8 MACI
  variables, making use of Latent Class Analysis
  (LCA)
• 9 variables selected based on Pathways Model
  postulates

20

• Problem gamblers were composed of at-risk and ppg
  youth.
• Step-wise mixture modelling technique in Mplus
  5.21.
• Global fit indices for model parsimony and fit
  (BIC and AIC), overall classification soundness
  (entropy), the Vuong-Lo-Mendell-Rubin Likelihood
  ratio, the Lo-Mendell-Rubin Adjusted LRT test,
  and the bootstrapped parametric likelihood ratio
  tests for the K-1 models.

21
Results
22
A 4 class solution optimal
Classes Log Likelihood Bayesian criterion (BIC) Entropy
2 -3708.846 7454.827 0.925
3 -3623.88 7298.26 0.951
4 -3557.01 7177.88 0.963
5 -3522.99 7123.22 0.947
6 -3496.483 7083.574 0.954
Sample size adjusted
23
Classes and control group for 8 MACI variables
(mean scores)

• Variable C1 C2 C3 C4 Control group (N924)
• Self-demeaning 19.3 29.5 90.0 56.1 45.623
• Family discord 36.8 76.5 77.8 61.3 59.15
• Childhood abuse 9.9 25.9 70.6 32.6 28.51
• SAP 19.5 84.7 73.2 39.5 43.53
• Impulsivity 31.5 73.7 72.3 46.9 47.67
• Anxious feeling 63.9 30.4 55.8 64.5 59.76
• Depressive affect 35.4 30.7 99.1 73.1 59.87
• Suicidal tendency 8.4 13.9 78.5 30.7 26.19

A mean MACI score of gt 60 reflects clinical
range Substance abuse propensity
24
Class descriptions
Class 1 (n) Class 2 (n) Class 3 (n) Class 4 (n)
Depressive
Anxious
SAP
Impulse
Suicidal
Self-Demeaning
Family Discord
Childhood Abuse
25

• Class 1 Anxious feelings (minimally significant)
• Class 2 Impulsive, prone to abusing substances
  (poor coping/ impulsive nature), reported family
  discord
• Class 3 Impulsive, prone to abusing substances,
  reported family discord, childhood abuse,
  self-demeaning (very significant), depressive
  affect (very significant), suicidal tendency
• Class 4 Anxious feelings, depressive affect,
  family discord (minimally significant)
• Self-demeaning, suicidal tendencies, and high
  depressive affect linked to childhood abuse only

26
Additional findings

• Problem gamblers are more likely to fall into
  class 3 (distinguished by childhood abuse, and
  combination of impulsivity and mood disorder)
  than any other class
• No significance in endorsement rates of
  DSM-IV-MR-J items between classes
• No significant differences in types of gambling
  activities and rates of play between classes

27
How do our 4 classes match up with the pathways
model?
28

• Three of our classes align very nicely with the 3
  proposed pathways in the PM
• Pathway 1 latent class 1
• Pathway 2 latent class 4. However, unlike those
  in PM, those experiencing childhood abuse are in
  a class of their own.
• Pathway 3 latent class 3

29
There is one meaningful difference

• our data points to a 4th class of people who are
  impulsive, prone to other addictions, but do not
  experience depressive/anxious symptoms (our
  latent class 2).
• Based on our clinical experience with youth, we
  often treated gamblers who were either primarily
  impulsive OR depressive/anxious, so the addition
  of the 4th class makes a lot of sense, as the
  Pathways \model does not account for a unique
  subset of problem gamblers who are primarily
  impulsive but not depressed.

30
What does this tell us?

• Blaszczynski was onto something.
• Nower and Blaszczynski were also warranted in
  their belief that this pathways model could apply
  to youth.

31
Should the theory be modified?

• Replication studies are necessary, with different
  methodologies and larger subject pools, before
  any modifications or additions to the Pathways
  Model are warranted.
• Having said that, further exploration into the
  validity of a 4th pathway is warranted at this
  time.

32
Conclusions

• Strong support for the premise that not all
  problem gamblers are the same

33
Implications

• Assessment on intake of gamblers in treatment
  should be extensive (clinical syndromes, history
  of abuse, impulsivity, personality concerns).
• Treatment should be tailored to these etiological
  factors and co-morbid disorders.

34
Implications

• Treatment providers must be proficient in
  treating mood impulse control disorders, and
  co-morbid addictions. Furthermore, it is
  important to be comfortable working with a family
  systems approach as well as in dealing with
  underlying contributing issues such as childhood
  abuse.
• Therefore, an eclectic training is ideal so that
  treatment can be properly tailored to each of the
  4 subgroups of problem gamblers.

35
Therapy Class 1

• Cognitive restructuring of erroneous beliefs
  pertaining to gambling participation.
• Financial counselling. Outcome excellent.

36
Therapy Class 2

• Impulse control management is primary focus.
  Environmental and emotional triggers need to be
  identified and client is taught more effective
  ways of coping. Family systems approach would be
  helpful as well. Cognitive restructuring,
  financial counselling and management of budget
  skills. Anticipated outcome good, but relapse
  risk is high due to impulsive style.

37
Therapy Class 3

• Primary focus on mood disorder and underlying
  cause of childhood abuse. Focus also on making
  them feel better about themselves and be less
  self-demeaning, self-destructive. Secondary focus
  on triggers and impulse control issues. Minimal
  focus on cognitive restructuring until primary
  mood and underlying issues dealt with.
  Anticipated outcome guarded. Risk of relapse
  high.

38

• Those experiencing childhood abuse are going to
  require a more comprehensive approach
  psycho-therapeutic approach and lengthy treatment
  protocol. Abuse leads to self-demeaning thought
  patterns, depression, and self-harming behaviours.

39
Therapy Class 4

• Treatment of mood disorders is primary focus.
  Family systems approach, and financial
  counselling also useful. Cognitive restructuring
  may be necessary but minimal. Outcome good, lower
  risk of relapse if depression/anxiety under
  control.

40

• Pharmacotherapy may be indicated for those
  experiencing underlying impulsivity, depression,
  and/or anxiety disorder. Grant et al. have
  proposed the benefits of SRIs, mood stabilizers,
  and Naltrexone for the treatment of problem
  gamblers.

41
Common therapeutic threads across subtypes

• Focus on gradually cutting down gambling
  involvement (time, money, frequency). Requesting
  immediate abstinence will result in elevated
  treatment dropout rates. (Best achieved by
  establishing a baseline rate and working from
  there.)

42

• Restructuring free time Most temptation occurs
  when faced with free time. Help them plan and
  organize their time on a weekly basis.

43

• Replace the gambling with other activities.
  Encourage a balanced lifestyle (i.e., exercise,
  taking on a hobby)

44

• Involve others as supports, such as significant
  other, friends, relatives. (Support groups, such
  as GA can be useful in conjunction with therapy,
  especially when other supports are unavailable.)