From iPod to iGod Are 12step Groups Hip Enough for Adolescents

1
From iPod to iGod? Are 12-step Groups Hip
Enough for Adolescents?

• John F. Kelly, Ph.D.
• Associate Professor in Psychiatry
• Harvard Medical School
• Associate Director
• Massachusetts General Hospital
• Center for Addiction Medicine

2
Outline

• Developmentally-related barriers to participation
  in 12-step mutual-help groups among adolescents
• Review evidence regarding 12-step participation
  and its influence on relapse
• 8-year course of recovery for young people in
  relation to AA/NA
• The implications of the available evidence for
  clinical practice guidelines

3
Acknowledgments

• UCSD/VA San Diego Healthcare System
• Sandra A. Brown, Ph.D.
• Mark G. Myers, Ph.D.
• Brown University
• Christopher Kahler, Ph.D.
• Ana Abrantes, Ph.D.
• NIAAA (R01AA007033-09 R01 5AA015526-03)

4
Treatment and Continuing Care

• Alcohol/drug disorders viewed as chronic -
  frequently require ongoing management/monitoring,
  re-intervention for lasting remission (McLellan,
  2002 Dennis et al, 2005 White et al, 2006)
• To aid recovery efforts informal continuing care
  resources (e.g., AA and NA) are often utilized
  have been shown to be cost-effective adjuncts to
  formal treatment (Bond et al., 2003 Gossop et
  al, 2007 Humphreys, 2004 Humphreys and Moos,
  2001 2007 Kaskutas et al., 2002 Kelly et al.,
  2006 Kissin et al., 2003 Magura et al., 2005
  McKay, 2001 Tonigan et al., 2002 Weiss et al.,
  2005 White, 2009).
• Thus, newer interventions incorporate standard
  referral (Longabaugh et al., 2005 Pettinati et
  al., 2005) and addiction and mental health
  organizations advocate in adult practice
  guidelines (e.g., APA, 2006 IOM, 1998 NIDA,
  2000 VHA, 2001)

5
Limited Youth-specific Knowledge

• Comparatively little known about AA/NA as youth
  continuing care resources
• Yet widespread use of 12-step treatment
  approaches and referrals to AA/NA by youth
  providers (Brown, 2004 Drug Strategies, 2003
  Kelly et al., 2008 Knudsen et al, 2008).

6
AA/NA as continuing care resource for adolescents

• More than three quarters of treatment programs
  espouse a 12-step orientation (Roman et al, 2004)
  yet it remains one of the least evaluated (Miller
  Willbourne, 2002).
• 66 or more of youth facilities incorporate
  12-step elements to some degree (Drug Strategies,
  2003) only 9 base treatment solely on 12-step
  model but 47 require AA/NA participation during
  treatment, and 85 of programs link adolescents
  to AA/NA groups (Knudsen et al, 2008).
• From 12-step treatment perspective, ongoing
  community AA/NA participation is seen as key
  mediator of successful, long-term remission of
  the disorder (McElrath, 1997)
• Other treatment orientations may see community
  AA/NA groups as a useful continuing care resource
  irrespective of theoretical orientation and most
  pts appear to be referred (Humphreys, 1999
  Laudet White, 2005 Kelly et al, 2008)

7
Clinical Staff Perceptions, Expectations, and
Referrals to AA/NA
Kelly, Yeterian Myers, 2008, Alcoholism
Treatment Quarterly.
8
In theory 12-step groups promising treatment
adjunct

• Lots of meetings, in most communities, at
  high-risk times (evenings/ weekends) when
  professional support often not available
• Access to a supportive abstinent social network
• Free of charge, pts can attend as intensively for
  as long as desired
• Provide an adaptive community based
  recovery-oriented system, responsive to
  undulating relapse risk (Kelly Yeterian, in
  press)

9
In actuality, youth face a number of
developmentally-related Barriers

• AA Membership 11.3 ltage 30 and 2.3 lt21 (AA,
  2008)
• Clinical differences
• Addiction severity (substance use
  profile/withdrawal/consequences)
• Problem recognition/motivation for abstinence
• Greater psychological distress
• Barriers to identification with pertinent
  life-stage issues
• marriage, employment, NIH funding problems
• 12-step Specific barriers
• Disinterest or opposition to spiritual/religious
  content
• May signify poor fit with 12-step fellowships
  emphases on abstinence and spiritual growth

10
Youth-Specific AA/NA Knowledge Limited
Existing evidence suggests participants better
outcomes at least in the short-term
11
Youth-Specific AA/NA Knowledge Further steps
needed(2)

• Given small number of studies, we still we know
  very little regarding
• Which youth attend and at what rate?
• Whether youth benefit over the short and
  longer-term as they transition into young
  adulthood
• What levels of utilization are associated with
  benefits

12
What we do know lacks rigor

• In most studies 12-step participation and
  outcomes examined concurrently obfuscating causal
  attributions related to temporal precedence
  (Nock, 2007)
• In many studies relevant time-invariant confounds
  related to better prognosis (e.g., baseline
  substance involvement) not controlled
• No existing youth studies have taken into account
  influential time-varying factors (e.g.,
  professional txs, levels of substance use in
  preceding follow-up periods)
• Thus, a major aim of current investigation was to
  use rigorous longitudinal methods (GEE), lagged
  approach, controlling for static and time-varying
  covariates

13
Which youth attend? Predictors of AA/NA
Attendance (1)

• Based on theory and prior evidence we examined a
  number of predictor variables
• Demographic (e.g., gender)
• Psychiatric (e.g. conduct disorder)
• Severity of substance involvement (e.g.
  dependence sxs, lifetime intoxication,
  withdrawal)
• Psychological (ability to moderate alcohol/drug
  use, abstinence self-efficacy)
• Prior-treatment
• Religious (affiliation, practices)

14
Do youth benefit in the short and
longer-term?Influence of Early Ongoing AA/NA

• Critical juncture - inpatients discharged back to
  community first 6m requires demanding
  adjustments - remains highest risk relapse period
  (Brown, 1993 Brown et al., 1989 Godley et al.,
  2002 Hunt et al., 1971 Vik et al., 1992).
• Findings from adult studies suggest also that pts
  who achieve immediate, rather than delayed, AA/NA
  may experience better long-term outcomes (Moos
  and Moos, 2004).
• Given pervasive use of 12-step treatment (Drug
  Strategies, 2003) and pressure for evidence-based
  practice (IOM, 1998), we aimed to provide
  knowledge regarding extent treated youth meet the
  key proximal outcome and whether its associated
  with better long-term outcome (Finney, 1995
  Suchman, 1967).

15
What levels of utilization are associated with
benefits?

• Prior studies support notion linear relationship
  between AA/NA and outcome
• Stress 90 in 90? Or would 1 or 2 meetings a
  week help my patient? What would be optimal?
  Little empirical basis for recommendations
• Hence, further aim here to analyze relationship
  between levels of AA/NA participation and
  abstinence using non-parametric regression (i.e.,
  locally-weighted scatterplot smoothing LOWESS
  Cleveland Devlin, 1988 Cleveland, 1993,
  chapter 3)

16
Study Aims

• Describe rates and predictors of participation in
  AA/NA over 8-years following inpatient treatment
• Examine the relationship between early, and
  ongoing AA/NA participation, and substance use
  outcome using lagged, longitudinal models
  (Generalized Estimating Equations GEE)
  controlling for static and time-varying
  covariates
• Explore dose-response thresholds of AA/NA
  attendance and outcome (LOWESS)

17
Methods 1

• Participants (N166 40 female M age 16 75
  Caucasian) consecutive admits to adolescent
  12-step oriented, inpatient, SUD programs in San
  Diego (M stay 4 wks)
• Followed at 6m and 1, 2, 4, 6, and 8yrs
  (follow-up rates gt 84)
• Neither demographic nor tx/clinical vars found
  associated with follow-up (psgt.27).

18
Methods 2

• Prior outpatient 47
• Prior inpatient 8
• School Attendance
• 73.8 was enrolled in public or private school
• 19.2 reported quitting school
• 6.2 were expelled or suspended.
• Alcohol -most commonly used drug.
• Reported drug of choice was
• meth/amphetamines (53)
• cannabis (32)
• hallucinogens (7)
• alcohol (4)
• cocaine (3)
• opiates (lt1)

19
Results Rates of Attendance
20
AA/NA Predictors Results
21
Relationship to Outcome Zero-Order Spearman rs
AA/NA x PDA Across 8 Years
Correlation is significant at the 0.01 level
(2-tailed). Correlation is significant at the
0.05 level (2-tailed).
22
Relationship to Outcome Generalized Estimation
Equations (GEE) Models (1) AA/NA in relation to
Outcome

• Preliminary Analyses Static control variables
  (sig. on 2 or more follow-ups)
• Examined 5 sets of variables
• Demographic (age, gender, ethnicity, religion)
• Substance-related (pre-treatment frequency of
  use, lifetime alcohol/drug intoxication, DSM-IV
  alcohol/drug dependence sxs, withdrawal sxs,
  beliefs about ability to use alcohol/drugs in
  moderation)
• Psychological (conduct disorder diagnosis,
  self-efficacy)
• Treatment (prior outpatient or inpatient)
• Family (family history of alcohol/drug
  dependence)

23
Generalized Estimation Equations (GEE) Models
(2) AA/NA in relation to Outcome

• Preliminary Analyses
• Two sig. results for (static) control variables
• Gender predicted outcome at 4-, 6-, and 8-years
  (girls/young women having more abstinent days
  than boys (M r -.30, pslt.001).
• Greater perceived ability to use alcohol/drugs in
  moderation consistent predictor of less
  abstinence at every follow-up time point through
  8 years (M r -.21 range -.18 to -.22, ps
  .008-.03).
• Thus, in addition to baseline use (past 90 days),
  gender, perceptions of ability to moderate use
  included
• Time-varying covariates
• Substance use (prior period)
• Aftercare meetings (6m)
• Any outpatient/inpatient SUD treatment (all
  periods)

24
Lagged GEE Model of Youth Treatment Outcome over
8 Years in Relation to AA/NA Attendance in the
first 6 months post-treatment
1. Square root transformed N 150
Kelly, Brown, Abrantes, Kahler, Myers (2008)
Alcoholism Clinical Experimental Research, 32, 8
1468-1478.
25
Lagged GEE Model of Youth Treatment Outcome in
relation to AA/NA attendance over 8 Years
1 Sq root transformed 2 Time varying covariate
26
Levels of participation in relation to
outcomeDose-Response Thresholds for Attendance
(LOWESS)

• LOWESS Locally weighted fitting and smoothing
  scatterplots Cleveland, 1979 1991)
• Instead of forcing the data to fit some a priori
  function of x on y, this method that lets the
  data speak for themselves (thus, no single
  parameter estimate of relationship)
• Data is modeled locally by WLS polynomial
  regression- weighted more for local points
  (robust to outliers)

27
(No Transcript)
28
(No Transcript)
29
Results Summary

• AA/NA attendance early post-treatment was common
  and relatively intensive, but declined steadily
  over time
• Objective and subjective indices related to
  addiction severity predicted youth AA/NA
  attendance following treatment
• Despite declining attendance rates over the 8
  years, early post-treatment attendance predicted
  long-term salutary outcomes
• Effects associated with ongoing attendance
  remained significant and consistent
• Exploratory LOWESS analyses suggested that even
  relatively small amounts of AA/NA associated with
  benefits, with a strong linear association
  evident up through 3 meetings per week

30
Discussion (1)AA/NA Participation over 8 years

• Participation very high early post residential
  tx, but quickly declines in proportion
  attending/freq AA/NA
• Clinical staff survey (n114 5 programs Kelly
  et al, 2008) suggest desire is for long-term
  AA/NA participation (especially if treated in
  12-step oriented program) and staff refer most
  patients
• Suggests youth may encounter other developmental
  barriers to AA/NA participation (intrinsic,
  age-barriers)

31
Relation between Age Composition of Attended
Meetings and Percent Days Abstinent
Kelly, Myers Brown, (2005) Journal of Child and
Adolescent Substance Abuse
32
Predictors of AA/NA Attendance

• Similar to other studies youth in the current
  study were more likely to participate if they had
  
• greater hx of alcohol/drug involvement
• thought they had a problem with alcohol
• considered themselves unable to control their
  substance use (possible screening question)
• Results consistent with HBM (Rosenstock, 1974),
  construct of perceived severity (Finney Moos,
  1995)
• Religious involvement did not predict attendance
  and girls/young women as likely as boys/young men
  to attend

33
Outcome in Relation to Early AA/NA Attendance

• Effects from early attendance (0-6m) suggests
  what is accomplished therapeutically early
  post-treatment has diminishing, but longer-term,
  recovery ramifications (Moos and Moos, 2004)
• Specific mediators of effects remain to be
  clarified (e.g., avoidance of high risk social
  environments or cognitive-emotional shifts)

34
How might AA/NA reduce relapse risk?
35
Conceptualization of How AA May Reduce Relapse
Risk

Decreased
Coping
Self-efficacy
probability
Response
of
relapse





High Risk Situation
AA

Increased
Initial use
No coping
Decreased
Abstinence
probability
of
response

self
-
efficacy

violation
of relapse

substance

effect

Dissonance
Positive
Outcome
conflict and
Expectancies

self
-
for initial use
attribution

of the
(guilt
substance

loss control)


Source Marlatt Gordon, 1985

36
plt.001 plt.01 plt .05 plt.10 N99
Kelly, Myers, Brown (2000) Psychology of
Addictive Behaviors
37
plt.001 plt.01 plt .05 plt.10 N99
Kelly, Myers, Brown (2000) Psychology of
Addictive Behaviors
38
Ongoing AA/NA Participation in Relation to Outcome

• Although attendance declines substantially,
  consistent effects of AA/NA over time, after
  accounting for formal care utilization and other
  confounds
• for every meeting attended there was a subsequent
  gain of about 2 days of abstinence, independent
  of other factors also associated with good
  outcome
• Findings suggest AA/NA may be beneficial for some
  youth in their recovery in the short- and
  long-term, particularly those more severely
  affected and who believe it unlikely they can
  moderate substance use
• Ongoing recovery-specific support from AA/NA may
  help reduce relapse risk, in part, by providing a
  social context where role-models are available
  and friendships can develop that provide
  alternative socially rewarding activities (Brown,
  2001 Moos, 2007 Longabaugh et al., 1998).

39
Substance Use and Problem Onset and Offset
NSDUH Age Groups
100
Severity Category
90
No Alcohol or Drug Use
80
70
Light Alcohol Use Only
60
Any Infrequent Drug Use
50
40
Regular AOD Use
30
Abuse
20
10
Dependence
0
65
12-13
14-15
16-17
18-20
21-29
30-34
35-49
50-64
NSDUH and Dennis Scott under review
40
Prevalence of DSM-IV Alcohol Dependence across
the Lifespan
Source Grant, Dawson et al, 2004
41
Serious Psychological Distress (NSDUH, 2007)
42
Dose-Response

• 3 or more AA/NA meetings per week, on average,
  appears associated with complete abstinence early
  post-treatment
• However, even 1 or 2 meetings per week was
  associated with sharp increases in abstinence
• Suggests even though no 12-step-specific actions
  completed general group-therapeutic processes may
  nevertheless be helpful (e.g. Kelly, Myers
  Rodolico, 2008).

43
What do you like best about AA/NA meetings? (6m
follow-up n53)
Source Kelly, Myers Rodolico (2008) Journal of
Substance Abuse
44
Limitations

• Although rigorous test of AA/NA participation
  over time cause-effect conclusions limited
• Aggregated data over long time periods (up to 2
  years at a time) do not lend themselves to
  interpreting the more complex relationship of
  attendance topography in relation to outcome
• Our measure of AA/NA participation was frequency
  only, did not measure degree of involvement in
  AA/NA or age composition of attended meetings
• Although all met for AUD, many identified main
  substance as meth/amphetamine (53)
• Inpatient, 12-step oriented, treatment (Drug
  Strategies, 2003)

45
Conclusions

• AA/NA attendance is likely to be high initially,
  when patients are treated in 12-step-oriented
  residential programs encouraged/facilitated to
  attend
• Even small amounts of participation appear
  related to improvement strong linear
  relationship up through 3 meetings per week
• Similar to adults, more severe youth attend, and
  youth who continue attending appear to experience
  better outcomes into young adulthood (M range
  16-24yrs)
• Social contexts facilitated by AA/NA may help
  buffer youth relapse during stage of life when
  rates of use/disorders peak in the general
  population and where support for recovery may be
  minimal

46
Provisional Clinical Guidelines

• Substance dependent youth should be encouraged to
  try NA or AA meetings
• Locate and direct youth to meetings where some
  other youth may be present
• Prepare youth for what to expect
• Emphasize potential benefits reported by other
  youth that go (e.g., feeling connected, less
  alone that somebody cares)
• Emphasize social component