Article Text


Ear acupuncture for co-occurring substance abuse and borderline personality disorder: an aid to encourage treatment retention and tobacco cessation
  1. Elizabeth B Stuyt1,2
  1. 1Department of Psychiatry, University of Colorado, Denver, Colorado, USA
  2. 2Circle Program, Colorado Mental Health Institute at Pueblo, Pueblo, Colorado, USA
  1. Correspondence to Dr Elizabeth B Stuyt, Department of Psychiatry, University of Colorado, 1600 W. 24th Street, Pueblo, Denver, CO 81003, USA; Elizabeth.stuyt{at}


Objectives Retention of individuals with co-occurring borderline personality disorder (BPD) and substance use disorders in treatment is known to be difficult. An outcome study of a tobacco-free 90-day inpatient dual-diagnosis treatment programme that uses several evidenced-based treatments in addition to ear acupuncture (acudetox) was undertaken to determine overall treatment effectiveness.

Methods Between January 2009 and December 2011, 231 patients were treated in the programme, 88% with nicotine dependence and 79% with personality disorder diagnoses. All patients completing the programme were invited to enrol in a 1-year follow-up study in which they responded to monthly questionnaires to assess outcomes.

Results 185 patients (80%) successfully completed the programme. There was no correlation between successful programme completion and gender, race, age, primary drug dependence diagnosis or primary psychiatric diagnosis. The use of acudetox was positively correlated with successful completion (p=0.006). Of the 78 patients with BPD, 100% of men and 83% of women successfully completed the programme. Their use of acudetox was positively correlated with successful completion (p=0.026). At the end of the year, 140 questionnaires were returned: 51 patients with BPD reported outcomes similar to the group as a whole, with 55% sober and doing well.

Conclusions The use of acudetox was positively correlated with both successful completion of the programme for those with BPD as well as successful tobacco cessation, which ultimately improves the ability to maintain sobriety.

  • Substance Abuse Treatment
  • Tobacco Cessation
  • Borderline Personality Disorder
  • Auricular Acupuncture
  • 5-point NADA protocol
  • Treatment Retention

Statistics from


The prevalence of personality disorders in those with substance use disorders treated in inpatient or residential settings has been shown to be as high as 70–80%.1 Verheul2 found the median prevalence rate of personality disorders among a population of treated addicts to be 56.5%. Ross et al3 reported 53% personality disorders in another population, with borderline personality disorder (BPD; 74%) and antisocial personality disorder (ASPD; 66%) being the most prevalent. Patients with ASPD and BPD have consistently been associated with risk for early dropout from all types of substance abuse treatment.4 BPD is thought to be a serious mental illness, beginning in early adulthood and marked by a pervasive pattern of instability of interpersonal relationships, self-image, mood and affects, marked impulsivity and a marked fear of abandonment.5 Individuals with BPD often display recurrent suicidal and self-mutilation behaviours. Samuel et al6 evaluated the impact of 10 personality disorders on early attrition (within the first 30 days) as well as time to dropout during a 9-month therapeutic community residential treatment programme and found that BPD was the only personality disorder negatively related to overall programme retention. While therapeutic community treatment does have an established record of effectiveness in reducing substance use and criminality, the most important outcomes have been shown to be retention in treatment and successful programme completion, as longer time in treatment predicts better long-term outcomes.7–9 Tull and Gratz10 studied the impact of BPD on dropout from residential substance abuse treatment among men. They found that men with BPD had significantly higher treatment dropout rates than those without BPD, even when accounting for other covariates relevant to treatment dropout such as court-ordered treatment status, major depressive disorder and contract duration. They suggested that, based on the literature, programmes should incorporate dialectical behavioural therapy (DBT) to teach emotion regulation and distress tolerance skills to help with treatment dropout.

DBT has been found to be an effective addition to substance abuse treatment. In a theoretical paper on which components of DBT may be useful in retaining clients in substance abuse treatment, Bornovalova and Daughters11 identified the mechanisms that can predict treatment dropout among those with comorbid BPD and substance use disorders. These include lack of motivation to engage in treatment, difficulty developing therapeutic alliance and deficits in distress tolerance. Although DBT is useful in promoting all three, a systematic review of studies using DBT in the treatment of comorbid BPD and substance use disorders still found high attrition rates with difficulty retaining participants in treatment.12 Treatment retention is essential for achieving positive outcomes.

The five-point ear acupuncture protocol known as NADA (National Acupuncture Detoxification Association) or acudetox has been shown to help with treatment retention.13–15 The NADA protocol has also been shown to be an effective treatment to significantly reduce self-injurious behaviour in depressed adolescents,16 a behaviour common among those with BPD. In a review of the literature on the use of acupuncture for drug dependence, White17 points out that, while early trials produced some astonishing results, subsequent large randomised trials reported disappointing results. White reports that, despite the lack of support from systematic reviews which summarise the evidence for acupuncture as ‘equivocal’ and ‘inconclusive’, the use of acupuncture in treatment programmes has held steady and is still widely used, which suggests a ‘mismatch between practice and research’. White suggests that the benefits of acupuncture may have been missed due to the choice of outcome measures and future studies should use outcome measures suggested by clinical experience.

The Circle Program at the Colorado Mental Health Institute at Pueblo is a 90-day inpatient modified therapeutic community treatment programme funded by the State of Colorado to treat people with co-occurring substance abuse and mental health disorders who have failed previous treatment. Most have been in numerous previous treatments and have dropped out or been unsuccessful. Personality disorders are present in 70–80% of those admitted. The programme incorporates cognitive behavioural treatment including Strategies for Self-Improvement and Change (SSC)18 and DBT,19 and all patients participate in both. A significant number of patients have experienced abuse and trauma both as a child and as an adult, contributing to the symptoms of post-traumatic stress disorder (PTSD) and BPD. All patients participate in trauma therapy in the form of group and individual psychotherapy on a weekly basis.

The five-point NADA ear acupuncture protocol was incorporated in 2000 when the programme went completely tobacco-free, initially to help patients with nicotine withdrawal and smoking cessation.20 It is offered 4–5 days per week on a voluntary basis and is used by most patients. Staff members who are trained in the protocol place five sterile needles in each ear using the points called Sympathetic, Shenmen, Kidney, Liver and Lung. Patients sit quietly in a group for 40–45 min, often listening to soft music. Patients who do not want to receive the needles are also allowed to sit in the group and they frequently do but do not receive needles.

Anecdotally, treatment staff noticed that patients with BPD tended to participate in the acudetox more than others and frequently asked for treatments more often. Because the programme has had good success at retaining people with BPD for the full 90 days of treatment in spite of being a modified therapeutic community programme, it was hypothesised that the NADA protocol could aid patients with BPD to remain in treatment and, as a result of longer time in treatment, possibly contribute to successful outcomes after treatment.


An outcome study was undertaken from January 2009 to December 2011, during which time 231 patients were admitted to and discharged from the programme. The goal of this study was to follow patients who successfully completed the programme monthly for up to 1 year after discharge to determine outcomes. Contact was made by telephoning the patient, a family member, their mental health provider, their substance abuse counsellor and/or their probation/parole officer, if applicable, to have them complete a questionnaire regarding the patients’ recovery efforts. Psychiatric and personality disorder diagnoses (table 1) were made by clinical assessment and continual re-evaluation during the patients’ stay by the programme psychiatrist using the Diagnostic and Statistical Manual of Mental Disorders IV criteria.21

One hundred and eighty-five patients (80%) successfully completed the 90-day programme and 179 were eligible to enrol in the one-year follow-up outcome study (six had to return to another secure environment making them ineligible). One hundred and fifty-four patients (86%) enrolled (25 declined) and 140 patients (91%) completed the 1-year follow-up. Fourteen patients (9%) were lost to follow-up. Information was obtained using a standardised questionnaire. Responses provided data on patients’ use of drugs, alcohol and tobacco; employment; participation in school or training programmes; participation in support groups; participation in individual or group therapy; maintaining contact with probation officer; frequency of tissue testing for drugs or alcohol; new legal charges; quality of physical health; quality of mental health; hospitalisations; compliance with prescribed medications; living situation; financial stability; and use of their support system.

All statistical calculations were performed using the StatView program 5 (SAS Institute, Cary, North Carolina, USA). Statistical methods included χ2 analysis of contingency tables, the Fisher exact test for categorical variables and unpaired t tests for continuous variables. In all analyses, p values <0.05 were considered evidence of a statistical difference not attributable by chance.


Results of 90-day inpatient treatment

For all patients there was no correlation between programme completion and the following factors: primary substance dependence diagnosis; primary psychiatric diagnosis; tobacco use on admission; gender; race; age; or number of previous treatment programmes. Tobacco use during treatment (against the rules) and the patient's plan to use tobacco as soon as possible after treatment were both significant predictors of a less successful programme completion (table 1).

Table 1

Patient characteristics and completion of 90-day treatment program

The programme was successfully completed by 98% of patients with no personality disorder diagnosis, 87% of those with BPD and 59% of those with ASPD (p<0.0001). Of the 78 patients with BPD, 19 were men and 59 were women (p<0.0001). All 19 of the men with BPD (100%) and 49 of the women (83%) successfully completed the programme. This difference was not statistically significant.

There was no difference between those with BPD (n = 78) and those with no personality disorder diagnosis (n = 49) with regard to legal status; race; primary substance dependence diagnosis; tobacco use on admission; use of tobacco in treatment; or attitude regarding tobacco use after treatment. Those with BPD were more likely to have a primary psychiatric diagnosis of PTSD (47%) than those with no personality disorder diagnosis (18%) (p=0.004). This is not surprising given that all the patients with PTSD and BPD had a history of significant abuse and trauma. While both groups initially in treatment were resistant to open up about their trauma history, those with BPD tended to be much more behaviourally disruptive, engaging in ongoing ‘chaos’ in the living environment. They complained of constant mood swings, going from happy to sad to angry within minutes. They reported an inability to slow their mind, sit still and initially found it difficult to learn and practice the stress reduction skills taught to all the patients including mindfulness, progressive relaxation and biofeedback.

Thirty-one (84%) of those with ASPD were men (p<0.0001). The most significant difference between those with ASPD and those with BPD and no personality disorder diagnosis was their attitude about tobacco use after treatment. While 59% of those without a personality disorder diagnosis reported a desire to stay away from tobacco and 62% of those with BPD (56% of women and 79% of men) reported a desire to stay away from tobacco after discharge, the majority of those with ASPD (52%) reported a plan to smoke as soon as possible once discharged (p=0.0013).

One hundred and ninety-four patients (84%) participated in the NADA acudetox. The use of NADA acudetox was positively correlated with successful completion of the programme, with 84% of those using it successfully completing the programme compared with only 62% of those not using it (p=0.006). The length of stay for those successfully completing the programme was 87±7 days compared with 40±22 days for those not completing the programme. Tobacco users who used acudetox (n = 170) stayed an average of 80±19 days compared with 33 tobacco users who did not use acudetox who stayed 63±33 days (p<0.0001). The use of the NADA protocol appeared to help those who reported high stress levels due to nicotine cravings. For those reporting a plan to use tobacco as soon as possible after discharge, a greater number of acudetox sessions was associated with a longer stay in treatment (p=0.001), even though they returned to tobacco use after discharge (figure 1). There was no difference in length of stay for those who reported a desire to try to stay away from tobacco and the number of acudetox sessions (p=0.2714).

Figure 1

Length of stay (LOS) in the programme in days by number of acudetox sessions compared with attitude about tobacco use after discharge.

The number of acudetox sessions attended by patients with BPD was positively correlated with successful completion of the programme. Sixty-eight patients with BPD who successfully completed the programme participated in 13±9 sessions compared with 7±6 sessions for those with BPD who did not complete the programme (p=0.026) (figure 2). Men with BPD participated in 14±10 sessions while women with BPD participated in 13±8 sessions. Using this treatment, those with BPD found that they were able to sit still quietly without disruption and feel relaxed, which then enabled them to learn and practice more easily the other relaxation techniques being taught. Those with no personality disorder diagnosis participated in 12±10 acudetox sessions while those with ASPD participated in only 6±6 sessions (p=0.003).

Figure 2

Programme completion by acudetox sessions. PD, personality disorder.

Results of 1-year follow-up

At the end of the 1-year follow-up period, 54% of the 140 patients reported that they were sober and doing well. Of these, 30% reported that they were continuously abstinent from drugs and alcohol the entire year, 18% reported a relapse (>2 episodes of use or persistent use) but were back on track, and 6% reported one or more slips (1 or 2 isolated uses) and were back on track; 19% were in relapse at the end of the year, 3% had died (three of the four had relapsed at the time of death) and 24% had re-offended and were incarcerated. There was no significant difference between the status at the end of the 1-year follow-up period and gender; race; primary substance dependence diagnosis; primary psychiatric diagnosis; tobacco use on admission; presence of a personality disorder diagnosis; or legal status. Age was a factor, with those who were incarcerated (32±9 years) or relapsed (34±9 years) being younger than those who were sober (35±9 years), those who were continuously abstinent (38±12 years) and those who had died (46±10 years). The difference in age between those who were incarcerated and those who were continuously sober was significant (p=0.04).

Fifty-one (36%) of the 140 patients had BPD (14 men and 37 women). They reported outcomes similar to the group as a whole, with 55% sober and doing well and 29% continuously abstinent from drugs and alcohol for the entire year and no significant difference in the variables identified compared with the group as a whole (figure 3). Recidivism was somewhat lower in the BPD group, with only 18% incarcerated at the end of the year compared with 29% for the 17 in the ASPD group and 23% for the 35 with no personality disorder diagnosis.

Figure 3

Status at end of 1-year follow-up. PD, personality disorder.

For the entire group of 140, tobacco use after treatment was highly correlated with relapse to drugs or alcohol. Those still using tobacco were much more likely to relapse, with 69% of tobacco users relapsing compared with only 45% of non-tobacco users (p=0.01). Patients who were using tobacco on admission but were willing to remain off tobacco after discharge were even more successful at staying sober than those not using tobacco at the time of admission (60% of non-tobacco users on admission relapsed vs 28% of those who refrained from using tobacco vs 69% of those continuing to use tobacco, p=0.0115). The time to first relapse after discharge was significantly longer for those not using tobacco than for those using tobacco (9±5 months vs 6±5 months; p=0.008). Those who maintained continuous abstinence from drugs and alcohol throughout the year were more likely not to be using tobacco (p=0.03). The number of people refraining from using tobacco at the end of the follow-up period nearly doubled from 20 (14%) on admission to 38 (27%). Those with BPD were more likely to use the time in treatment to quit tobacco than those with ASPD or those with no personality disorder diagnosis. Six out of 51 with BPD (12%) were not using tobacco on admission and 17 (33%) reported not using tobacco at the end of the follow-up period. This compares with one of 17 (6%) with ASPD at the start and two (12%) at the end of the follow-up period, and eight of 35 (23%) with no personality disorder diagnosis at the start and 10 (29%) at the end. Interestingly, for all 140 people, those not using tobacco at the end of the follow-up period used significantly more acudetox sessions when they were in treatment (15±9) than those who were still using tobacco (12±8) (p=0.04).


Patients with BPD in this study did remarkably well, both in remaining for the entire 90 days of treatment and then in maintaining sobriety in the year after treatment. They were also more successful at tobacco cessation and avoiding recidivism than those with ASPD or even those with no personality disorder diagnosis. While all patients are required to participate in cognitive behavioural therapy (SSC, DBT) groups, gender-specific trauma groups and individual trauma psychotherapy weekly throughout the programme, the NADA ear acupuncture is totally voluntary. There are four acudetox groups per week. Patients are encouraged to attend the groups as often as they like during their first month, but at the start of their second month they are required to attend other groups offered at the same time such as cue exposure and anger management.

The limitations of this study include the naturalistic design and retrospective nature of the analysis. There was no randomisation, placebo treatment or true control group for the acupuncture. While the strengths of this study include the assistance by probation officers who reported results of tissue testing including urine, breath and hair to document use of drugs or alcohol during the 1-year follow-up, limitations include the fact that none of these tests included nicotine, nicotine metabolites or carbon monoxide testing. Data on tobacco use were therefore obtained by self-report and observation by others. However, research indicates that self-report correlates with serum cotinine levels and can be a good indicator of actual smoking status.2,2 Another limitation was that 22% of the patients were not followed by a probation officer and therefore did not have tissue testing results. Slips, relapses and tobacco use in these individuals were obtained by self-report and observation by others only. Nevertheless, empirical studies have indicated that carefully collected self-report alcohol and drug use data are accurate.2,3 ,2,4

In our experience, NADA acudetox is a safe, effective, low-cost, non-verbal intervention that can be used to help patients to be more present and mindful, allowing them to be more open to other therapeutic interventions. Patients with BPD and substance use disorders often have never experienced what it is like to sit quietly for 40–45 min, and it is therefore difficult to teach them mindfulness and other stress tolerance skills. When they sit in a group with five needles in each ear, they experience an immediate calming effect and realise that they can sit still. The calming effect they experience increases their motivation to engage in other aspects of the treatment programme. The protocol definitely fosters the therapeutic alliance, in that the patient demonstrates trust in the treatment providers by allowing them to put needles in their ears and the positive effect they experience reinforces the alliance.

Summary points

  • Drug use is commonly associated with personality disorder.

  • In a residential drug treatment programme, we offered acupuncture for detoxification including smoking cessation.

  • The patients who used acupuncture were more likely to complete the programme and be drug-free at 1 year.

  • Those persisting with tobacco use were more likely to relapse.


The author would like to thank Spencer Troy Beck, Sandra M Cordova, Kimberly Dionysus, Jessica Espinoza, Karen A Farr, Rosie Freeman, Jeanette Grant, Peggy Hicks, Brenda McBride, Jose Vega and Stacey Waring for their work in making numerous telephone calls in order to complete the questionnaires and obtain the data.


View Abstract


  • Correction notice This article has been corrected since it was published Online First. Figure 1 legend has been amended to read ‘Cell: Plans to use tobacco as soon as possible N=67’ and figures 2 and 3 have been updated.

  • Contributors The author conceived the paper, analysed the data and wrote the report.

  • Competing interests None.

  • Patient consent Patients were provided with informed consent and signed releases of information for all persons to be contacted.

  • Ethics approval The local Institutional Review Board of Parkview Hospital in Pueblo, Colorado approved the study.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.