AA Effectiveness – Faith Meets Science Lee Ann Kaskutas, Dr.P.H. Alcohol Research Group, Emeryville, CA School of Public Health, UC Berkeley Presented at the Annual Meeting and Conference of the Federation of State Physician Health Programs (FSPHP) April 25, 2012 Fort Worth, TX Perspectives on AA Effectiveness • AA members: I know AA works and there is no need for research to prove that Perspectives on AA Effectiveness • AA members: I know AA works and there is no need for research to prove that • AA critics: AA is a cult that relies on God as mechanism of action Perspectives on AA Effectiveness • AA members: I know AA works and there is no need for research to prove that • AA critics: AA is a cult that relies on God as mechanism of action • Cochrane Review: no experimental evidence of effectiveness Perspectives on AA Effectiveness • AA members: I know AA works and there is no need for research to prove that • AA critics: AA is a cult that relies on God as mechanism of action • Cochrane Review: no experimental evidence of effectiveness • Moos: first send people to AA, not treatment Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 % abstinent Abstinence & AA exposure 50 45 40 35 30 25 20 15 10 5 0 at 1 yr. at 18 mos. AA male VA inpatients 1 yr n = 3018; 18 mo n = 91 no AA Ouimette et al., J Stud Alcohol 1998 Thurstin et al., Int J Addict 1987 Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Abstinence & meeting amount % abstinent mos. 9-12 70 60 50 40 30 20 10 0 0 1-19 20-49 50+ # of meetings mos. 9-12 Male VA residential patients n = 2376 Moos et al., J Clin Psychol 2001 % abstinent at 2 yrs Abstinence & meeting frequency 80 70 60 50 40 30 20 10 0 never less than weekly weekly meeting frequency, mos. 19-24 LA Target Cities, outpatients n = 262 Fiorentine, Am J Drug Alcohol Ab 1999 % abstinent 16 years Abstinence & sustained attendance 80 70 60 meetings 50 40 30 20 10 0 none 1-8 wks 9-26 wks 27+ wks year 1 years 2-3 years 4-8 duration☼ of AA attendance Previously untreated problem drinkers n = 461 Moos & Moos, J Clin Psychol 2006 Also see Moos & Moos, JSAT 2004 AA meeting trajectories # of AA meetings, pst yr 250 200 declining AA high AA medium AA low AA 150 100 50 0 TxEntry Dependent treatment seekers n = 349 1 year 3 years 5 years Kaskutas et al., ACER 2005 % abstinent Abstinence and meeting trajectories 100 90 80 70 60 50 40 30 20 10 0 declining AA high AA medium AA low AA no AA 1 year Dependent treatment seekers n = 349 3 years 5 years Kaskutas et al., ACER 2005 Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Consistency across samples & time 70 % abstinent 60 50 nothing outpatient only AA only AA + outpatient 40 30 20 10 0 1 year aMale VA inpatients n = 3018 a 3 years bPreviously b untx prob drnkrs n = 466 8 years b a Ouimette et al., J Stud Alcohol 1998 b Timko et al., J Stud Alcohol 2000 Consistency across samples hazards of abstinence (vs. nothing) 6 5 4 3 2 1 0 formal tx only NESARC; dependent n = 4422 Tx n=239 12-step only 12-step n=138 both n=829 12-step + tx nothing n=3217 (Dawson, Addiction 2006) Consistency across time AA involvement over 10 years Inpatients n=158 significant Abstinence at 10 years Cross et al., ACER 1990 Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Substance use following AA/NA 12-step meetings mos. 1-3 No Alcohol and drug use Yes mos. 4-6 12-step activities mos. 1-3 Cocaine-dependent outpatients n = 336 Weiss et al., Drug Alcohol Depen 2005 Alcohol abstinence following AA AA involvement Outpatient β = .29 mos. 1-6 Aftercare β = .34 Project MATCH n = 480 outpatients, n = 434 aftercare Percent days abstinent mos. 7-12 Connors et al., J Stud Alcohol 2001 Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Randomizing to AA to remove selection effect % abstinent at 2 yrs 40 35 30 25 20 15 10 5 0 Hospital inpatient Alcohol abusers + EAP referred n = 227: n=73 hospital; n=83 AA; n=71 choice AA meetings Choice Walsh et al., New Engl J Med 1991 % abstinent, past 3 mos. Randomize to TSF to remove selection effect: Outpatient sample 40 35 30 25 20 15 10 5 0 12-step Motiv Cog Beh 1-yr follow-up p = .0024 Project MATCH n = 806 outpatients at yr 3 3-yr follow-up p < .007 PMRG, J Stud Alcohol 1997 PMRG, ACER 1998 % abstinent mos 13-15 Randomize to TSF to remove selection effect: Aftercare sample 50 45 40 35 30 25 20 15 10 5 0 12-step Project MATCH aftercare n = 714 at 1-year follow-up Motiv Cog Beh PMRG, J Stud Alcohol 1997 AA Meeting Attendance by Project MATCH Sample And Treatment Assignment % Days AA Mtgs .5 .4 .5 Green = 12-step Yellow = Motiv Grey = Cog Beh .4 .3 .3 .2 .2 .1 .1 0.0 0.0 Intake Outpatient* Project MATCH *n = 952 Outpatients, **n = 774 Aftercare 15 Month Green = 12-step Yellow = Motiv Grey = Cog Beh Intake 15 Month Aftercare** Tonigan et al., Tx Match Alcohol 2003 Specificity: Randomizing to Intensive AA/NA Referral % abstinent at 6, 12 mos 60 50 40 30 20 10 51.4% 40.7% Intensive Referral Standard Referral 0 VA outpatients n = 207: n=161 intensive; n=146 standard referral Timko & DeBendetti, Drug Alc Dep, 2007 Timko’s results for 12-step involvement at 6 months Standard Intensive Referral Referral # of meetings P Value 57 64 ns Did service 45% 56% < .05 Had awakening 39% 52% < .01 Had a sponsor 38% 48% <.05 Timko, Addiction, 2006 Groups where Intensive referral mattered in that study • The low prior AA group attended significantly more meetings in IR (vs. SR) at 6 months. Differences in meeting attendance were nonsignificant in the high prior AA group. Specificity: Randomizing to MAAEZ (Making AA Easier) 12-month Outcomes Usual MAAEZ care % Clean 71 79 % Sober 83 92 % Clean & Sober 72 82 Groups where MAAEZ dose seems to really matter • • • • • • • Unmarried Had prior treatment Attended 91+ prior AA meetings in lifetime Has high ASI psychiatric problem severity Drug dependent Atheist, agnostic, unsure No heavy drinkers or drug users in social network Statistical models to study selection effect Baseline 1 year 2 years Motivation AA involvement negative Alcohol problems Psychopathology Male VA inpatients N=2,319 McKellar et al., J Consult Clin Psych 2003 Example Using Propensity Score Method to Remove Selection Effect: Odds of Abstinence before adjustment N Non-AA 228 AA 334 * p<0.05, ** p<0.01, *** p<0.01 Odds Ratios associated with AA attendance 3.60 Odds of Abstinence after propensity score stratification N Stratum 1 (lowest PS) Stratum 2 Stratum 3 Stratum 4 Stratum 5 (highest PS) Total Adjusted * p<0.05, *** p<0.01 Non-AA AA Non-AA AA Non-AA AA Non-AA AA Non-AA AA 88 23 68 45 44 68 18 96 10 102 Non-AA 228 AA 334 Odds Ratios associated with AA attendance 2.72* 6.85*** 4.10*** 2.39 1.33 2.93* Criteria to establish causation • • • • • • Strength of association Dose-response relationship Consistency of association Temporally-correct association Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Consistency with theory: MEETINGS • What you do – Place to go instead of bar – Talk about your problems • What you hear – Others had similar experiences – Ways people coped instead of drinking • What happens – Your mood changes – Don’t drink a day at a time Cog Beh Soc Lrn alternative cue Psy need succeed skills mood alternative efficacy Consistency with theory: FELLOWSHIP Cog Beh • Friendship – Adds sober people supportive of your abstinence – Role models of new behavior – Helps learn how to have fun sober alternative • Sponsorship – Someone to call emerg plan – Someone to provide emotional support Soc Lrn Psy cue model cue need need Consistency with theory: STEPS Cog Beh Soc Lrn • Change how you treat others, or you will drink – Personal inventory; amends • Key to sobriety is helping others – Helping gets you to relinquish negative self focus Psy need mood Evidence of mechanism: cognitive behavioral AA involvement Abstinence Self-efficacy* Coping skills** *1 Resi or IOP n = 100 *2 asolescent inpatients n=74 **1 initially untx PDs n=466 *1Morgenstern et al., J Consult Clin Psych 1997 *2Kelly et al., J Stud Alcohol 2002 **1Timko et al., ACER 2005 **2 male VA inpatients 2 ** Humphreys et al., Ann Behav Med 1999 n=2,337 Evidence of mechanism: social learning AA involvement Abstinence Fewer pro-drinking influences* Enhanced friendship networks** More friends† # who support abstinence from AA‡ * treated ** male VA inpat. † init. untx. PDs ‡ treated n = 722 n=2,337 n=466 n=655 *Kaskutas et al., Addiction 2002 **Humphreys et al., Ann Behav Med 1999 † Timko et al., ACER 2005 ‡ Bond et al., J Stud Alcohol 2003 Evidence of mechanism: social learning? MAAEZ Abstinence Service 6+12mos; high psych Sponsor Comfort speak; be @ meetings # who support abstinence Kaskutas et al., JSAT 2009 Evidence of mechanism: psychodynamic AA involvement Abstinence Life meaning* Motivation for abstinence** *In recovery n = 354 **adolescent inpatients n = 74 *White & Laudet, CPDD 2006 **Kelly et al., J Stud Alcohol 2002 Evidence of mechanism: spirituality AA involvement Abstinence Δ religious beliefs & behaviors Spiritual awakening Day Hosp & Residential, managed care n = 537 Zemore, ACER in press Criteria to establish causation Strength of association Dose-response relationship Consistency of association Temporally-correct association – Specificity of the association Coherence with existing information Mausner & Kramer, Epidemiology -- text 1985 Getting People to AA Making AA Easier (MAAEZ) • Six 90-minute manual-guided GROUP sessions led by recovering counselor • Goals: To reduce patients’ immediate rejection of AA, and to help them connect with the AA fellowship • Manual available now (lkaskutas@arg.org) Kaskutas & Oberste, MAAEZ 2002 Why care about AA facilitation? • AA is the most widely-used source of help for alcoholism • Over time, those who attend AA have higher rates of abstinence than those who drop-out • Drop-out from AA is very high – Over half drop-out within their first year Goal of MAA*EZ To prepare clients to engage in the culture of AA/NA/CA – Making it easier to connect with AA members • Changing social networks is a mechanism of AA’s effect on abstinence 5,6 – Helping clients fill a purposeful role early on • Helping helps the helper (Reissman’s Helper Therapy Principle) 7,8 and is consistent with AA’s stated goal 9 5 Kaskutas et al, Addiction 2002;97(7):891-900 et al, JSA 2003;64(4):579-588 7 Riessman, Social Policy 1976;7:41-45 8 Zemore, Southern Medical Journal, forthcoming 9 Alcoholics Amonymous, 1939 6 Bond Structure of MAA*EZ • 6-week intervention – Six sessions, attended 1-week apart • Need time for doing weekly homework • Intro session – First and Last session attended • For newcomers and graduates • 4 core sessions – Spirituality, sponsorship, principles not personalities, living sober • Attended in any order Introduction • Graduate vignettes of their AA and MAA*EZ experience • How do you pick a meeting? – Meeting directories are passed-out • Rules of the road at meetings • Homework: go to a meeting – Pick meeting now using directory Spirituality • AA is spiritual, not religious program – Things to think about when you hear ‘God as we understood Him” • Spirituality in AA – It’s about your behavior and taking responsibility for it • “Act your way into good thinking” • Get outside yourself; do service • Homework – Talk to someone at a meeting who you don’t know, who has more sobriety than you Principles, not Personalities • Common objections to AA – It is a cult • No single person speaks for AA • Homework – Ask someone you don’t know at meeting, for their phone # • Telephone them (and talk to them) before next session Sponsorship • Who should you ask? – You’re not imposing • Role playing to ask someone to be temporary sponsor; 4 vignettes: – – – – Asking someone you went to coffee with Ask someone whose phone # you got Ask speaker at meeting whose talk you liked Ask someone who said something you connected with, but they said ‘no, they’re too busy’ • Homework: ask somebody to be your temporary sponsor Living Sober • Sober fun: or, Is there life after sobriety? – What are some things you are worried about being able to do without the help of alcohol or drugs? • Homework: do something with one or more people from AA who have more sobriety than you Acknowledgements: Funding • NIAAA grants – – – – – R01 AA 11279 (Kaskutas, PI) R21 AA 13066 (Kaskutas, PI) R01 AA 14688 (Kaskutas, PI) R01 AA 9750 (Weisner, PI) P50 AA 5595 (Greenfield, PI) • NIDA grant – R01 DA 12297 (Kaskutas, PI) • CSAT contract – #270-94-0001 (Kaskutas, PI) Acknowledgements: People • Eddie Oberste – co-developed MAAEZ • Thomasina Borkman, Constance Weisner, David Pating – expert panel on MAAEZ development • Ann Munoz, Sarah Zemore, Jane Witbrodt, Mina Subbaraman – fieldwork and analysis for MAAEZ study