Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
N Am J Psychol. Author manuscript; available in PMC 2013 January 23.
Published in final edited form as:
N Am J Psychol. 2011 June; 13(2): 175–186.
PMCID: PMC3552303

How Type of Treatment and Presence of PTSD affect Employment, Self-regulation, and Abstinence


The present study examined self-regulation, unemployment, and substance use outcomes for individuals with and without posttraumatic stress disorder (PTSD) who had transitioned from substance use treatment centers to the community. Participants, recruited from substance abuse treatment centers, were randomly assigned to an Oxford House self-help communal living environment (n = 75) or received usual aftercare (n = 75). Among these 150 individuals, 32 participants (27 women, 5 men) were diagnosed with lifetime PTSD. At a two year follow-up, individuals with PTSD in the usual aftercare condition showed significantly lower levels of self-regulation than those in the Oxford House condition with or without PTSD. These findings highlight the importance of abstinence supportive settings following substance use treatment, especially for individuals with PTSD.

Posttraumatic stress disorder (PTSD) is associated with increased risk for substance use disorders (SUDs; Najavits, Weiss, & Shaw, 1997). Studies have found rates of PTSD and SUD comorbidity as high as 25–59% (Brown, Recupero, & Stout, 1995; Najavits, et al., 1997; Stewart et al., 2000). Read, Brown and Kahler (2004) found that having PTSD and increased psychiatric distress associated with comorbid disorders was associated with poorer substance use outcomes. Additionally, Ritsher et al. (2002) found that dually diagnosed patients were not only less likely to be in remission when compared to an SUD-only group, but that they did have more severe levels of distress. However, other studies suggest that there are no significant differences for treatment outcomes between those with comorbid PTSD and SUD, and SUD-only groups (e.g., Norman, Tate, Anderson, & Brown, 2007).

Several theorists believe that using substances for extended periods of time may be a causal factor in mental health symptomatology, or that it exacerbates existing psychiatric symptoms (e.g. Blume, Schmaling, & Marlatt, 2000; Miller, Eriksen, & Owley, 1994). Alternatively, self-medication theorists assert that individuals use substances as a coping mechanism for negative emotions (Khantzian, 1997). Ouimette et al. (2007) found that patients with PTSD reported that their substance abuse relapse was in response to depression more often than people without PTSD. Meanwhile, Waldrop, Back, Verduin, and Brady (2007) proposed that alcohol may have dampening effects that help regulate the anxiety of patients with PTSD and that cocaine may increase hypervigilance and self-confidence to help individuals with PTSD feel more in control in social situations. There is some evidence for both theoretical points of view. For example, Gil-Rivas, Prause, and Grella (2009) found that individuals with co-occurring disorders reported that they experienced depressive and anxiety symptoms before relapse, which supports the self-medication theory. However, these individuals reported that those symptoms did not diminish, and in fact, were exacerbated after drug use.

Symptoms of PTSD include intense feelings of fear and anxiety, which may lead individuals with this disorder towards avoidance of people, places, or situations that could cause them to re-experience the trauma (American Psychiatric Association, 2000). Individuals with PTSD also experience self-regulation impairments – a reduction in their ability to logically regulate responses to goals, priorities and environmental demands (Tangney, Baumeister, & Boone, 2004). Impairment of self-regulation causes individuals to experience increased emotional distress, periods of dissociation, loss of trust in relationships and meaning in life, and chronic health problems that cannot be medically explained (Roth, Newman, Pelcovitz, van der Kolk, & Mandel, 1997). Cognitive structures responsible for managing emotional responses become impaired for individuals who have been exposed to both extreme stress and who are chronically dependent on substances (Phan, Wager, Taylor, & Liberzon, 2002). Inhibition of cognitive control processes may affect decision making and impulse control that impacts substance abuse or relapse (Matto, 2007). Conversely, people with high self-control show less impulse-related problems, such as alcohol problems. High self-regulation is also associated with better psychological adjustment (Tangney et al., 2004).

Another factor that may increase positive outcomes for those with PTSD and SUDs is employment. People with psychological disorders find work as a meaningful and satisfying way to expand the broader social and economic networks in their lives (Bluestein, 2008). For instance, in one study, veterans who were able to form social bonds were also more likely to be in remission of PTSD. In contrast, veterans’ feelings of isolation and weakening social bonds were more likely to predict chronic PTSD (Koenen, Stellman, Stellman, & Sommer, 2003). People with psychological disorders who are able to work derive internalized values and satisfaction from their work-related experiences. They value their independence and perceive themselves as able to influence their environment, a factor that is positively associated with mental health (Auerbach & Richardson, 2005). Although being employed has economic and mental health advantages, having PTSD makes it more difficult to find employment. Savoca and Rosenheck (2000) found that PTSD status greatly decreased the probability of current employment, and that PTSD status had a greater effect on unemployment than years of education. Similarly, Smith, Schnurr and Rosenheck (2005) found a positive correlation between PTSD symptom levels and the probability of unemployment. Employment is beneficial for substance users. Chronically dependent individuals have been found to increase their likelihood of abstinence when employment was a component in their substance use intervention (Silverman et al., 2002). Employment might act as a strong influence in increasing mental health and maintaining abstinence.

One specific type of setting that has been found to increase levels of abstinence and promote employment is Oxford House. Oxford House is a financially self-supporting recovery model where residents learn skills and behaviors that will allow them to return to independent settings once abstinence is sustained (Jason, Davis, Ferrari, & Bishop, 2001). Oxford Houses are democratically-run, self-help recovery environments where residents develop long-term abstinence skills without the involvement of professional staff. Members of Oxford House share responsibilities such as house chores and paying the bills, and therefore most of the members are employed. There is no limit on the length of stay; however, residents are evicted if they use drugs or alcohol. A person can seek re-admission after sobriety is re-established, but because this process involves house members voting on the person’s re-application, the time period of sobriety varies with each person seeking re-admission. Many Oxford House members also attend twelve-step programs to assist them in their abstinence (Oxford House, Inc., 2008). Research has shown not only the effectiveness of the Oxford House model in maintaining sobriety among its members, but also the practicality of having self-supporting recovery environments, especially in financially stressful times (Jason et al., 2001). For example, when compared to a group that received usual aftercare, those in the Oxford House condition had significantly lower substance use (31.3% compared to 64.8%) at the two-year follow-up. Furthermore, the Oxford House group earned approximately $550 more per month than the usual aftercare group (Jason, Olson, Ferrari, & Lo Sasso, 2006).

The present study examined whether individuals in two treatment groups (Oxford House versus usual aftercare) with comorbid PTSD and SUD had different self-regulation scores, unemployment rates and substance use outcomes than individuals with SUD but not PTSD. We hypothesized that the individuals with PTSD in usual aftercare would have the worst substance use outcomes, higher unemployment rates, and lower self-regulation scores than those in the usual aftercare without PTSD and those in the Oxford House condition with and without PTSD. We also hypothesized that those in the Oxford House condition with PTSD would have similar substance use outcomes, employment and self-regulation levels to those without PTSD in the Oxford House condition.


Participant Recruitment

Data collection occurred from 2002 to 2005. A total of 150 participants were recruited from residential substance abuse treatment centers in northern Illinois for the present study (for a more detailed discussion of these recruitment methods, see Jason, et al., 2007). Of those persons approached to be in the study, only four individuals indicated that they were not interested in being involved in the project. Participants were approached at treatment centers. There was a complete discussion of the study with participants and written informed consent was obtained after this discussion. The study was conducted in accordance with the Declaration of Helsinski, and the DePaul IRB who approved and monitored the study.

After signing the consent forms, 75 adults (46 women, 29 men) were randomly assigned to an Oxford House and 75 adults (47 women, 28 men) to a usual after-care condition and baseline interviews were given on site. A computer-generated randomization sequence was created and printed out in hard copy. A member of the research team who was not involved in interactions with, or care given to, participants maintained this printout and was the only person with knowledge of the actual randomization sequence. When an eligible participant required randomization, this research team member was contacted and revealed the next assignment in the sequence. With the assistance of the research staff, participants in the Oxford House condition visited one of Illinois’ Oxford Houses to complete an application for entry and to be interviewed by House residents. Participants moved into an Oxford House on their planned date of release from their treatment facility. Participants in the usual aftercare condition were referred by their case-workers to different outpatient treatment facilities, self-help groups or other resources in the community. Following discharge, the usual aftercare individuals went to the following sites: a relative’s home (32%), a staffed recovery home (18%), a partner’s or spouse’s home (16%), their own home or apartment (16%), a homeless shelter (10%), a substance abuse treatment program (4%) or a friend’s home (3%).

Incentives of $40 were given for participation in the baseline interview and equal amounts for subsequent interviews. Chi square and independent sample t-tests indicated no significant differences between conditions on socio-demographic variables. At baseline, over the past 90 days, 30% of the sample indicated using alcohol and 92% indicated that they had used drugs. The completion rate across the 2-year study was comparable for the Oxford House (89%) and usual after-care (86%) conditions. Over the 2-year follow-up, Oxford House participants spent an average of 256.2 days (range 8 730, SD = 247.1) in this setting. Of the 75 Oxford House participants, 5% stayed in Oxford House for the entire 24 months of the study, 35% moved into their own home or apartment after leaving the Oxford House, 20% went to relatives’ homes, 15% moved into a partner’s or spouse’s home; 9% went to a friend’s home; 5% went to a treatment program; 4% went to jail; 4% went to another staffed recovery home; and 3% went to a homeless shelter.


Diagnostic Interview Schedule-IV (DIS-IV)

The DIS-IV is a structured psychiatric instrument that was administered at baseline (Robins & Helzer, 1994). Lifetime and current prevalence of DSM-IV diagnoses were assessed for each participant. Lifetime PTSD status was determined by this instrument. Only 32 of the 150 participants had a lifetime diagnosis of PTSD (19 in the Oxford House condition and 13 in the usual aftercare condition). Of the other 118 participants without a PTSD diagnosis, 56 were in the Oxford House condition and 62 were in the usual aftercare condition. The no-PTSD variable was defined by “negative,” “not determined” and “indeterminate” cases, whereas comorbid PTSD variable was defined by the “positive, but not current” and “positive and current” cases. Horton, Compton, and Cottler (1998) tested the validity of DIS-IV by comparing it to the diagnoses obtained with the WHO Schedules for Clinical Assessment in Neuropsychiatry (SCAN; Wing, Babor, Brugha, et al., 1990). The overall comparison of DIS and SCAN indicated moderate agreement for PTSD diagnoses (kappa .46) (Horton et al).


Self-regulation was measured in this study by an instrument developed by Tangney et al. (2004). The self-report measure consists of 36 items (e.g., “I get carried away by my feelings,” “I wish I had more self-discipline”) which were answered using a five-point Likert scale (1 = not at all like me, 5 = very much like me). An average total summary score of the 36 items was used to arrive at a self-regulation score. Tangney et al. (2004) reported that this measure had good internal consistency (alpha r = 0.83 to 0.85), and with the present sample Cronbach’s alpha was 0.82 (M = 44.1; SD = 8.2). A lower mean score signifies better self-regulation (Tangney, et al., 2004).

Addiction Severity Index (ASI)

All participants completed the 5th edition of the ASI-Lite (McLellan, et al., 1992). This measure assesses substance use and areas of life that may be affected by it (e.g., legal, medical and psychological status). Results of the interviews on these sensitive issues were kept confidential for all participants in the study. Employment status was also determined by this measure. For purposes of this study, at the two year follow-up, participants were asked about their employment status and were coded as either “unemployed” or “employed” (employed group consisted of participants who were currently employed, students or had a disability status). Those who had a disability were placed in the employed group as they had a steady source of income from their disability. For items assessing substance use, participants were asked whether they used any illegal drugs or consumed any alcohol. Using the ASI self-reports of alcohol/drug use, the primary dependent variable measured at the 24 month follow-up whether the participants relapsed into any substance use at any point during the last 6 months. This instrument has been used extensively over the last 15 years and has demonstrated excellent test retest reliability (≥ .83; McLellan et al., 1992).

Data Analysis

In order to test the study’s hypotheses, we examined differences in self-regulation, unemployment rates and substance use outcomes among people with and without PTSD who were either assigned to the Oxford House or the usual aftercare conditions. We created four distinct groups to test the differences: individuals in the usual aftercare group who did not report having PTSD in their lifetime, individuals in usual aftercare with PTSD, individuals in the Oxford House condition without PTSD, and individuals in the Oxford House condition who reported having PTSD at some point in their lifetime.



As we did have self-regulation scores at the baseline, we first conducted a one-way analysis of variance (ANOVA) to compare the four groups at baseline. We then used the baseline self-regulation scores as covariates and performed an analysis of covariance (ANCOVA) with the dependent measure being the 24 month self-regulation scores. At baseline, the ANOVA indicated there were no significant differences in self-regulation scores among the four groups. Controlling for baseline self-regulation scores, the ANCOVA revealed significant differences among the four groups at the 2 year follow-up [F (3,130) = 2.65, p = .05]. Lower scores indicate better self-regulation. Significantly worse self-regulation scores were found among those in the usual aftercare condition with PTSD (M = 2.97, SE =.21; 95% CI: 2.56–3.38) than those individuals in the Oxford House condition with PTSD (M = 2.39, SE =.17; CI: 2.05–2.73) and without PTSD (M = 2.41, SE = .10; CI: 2.21–2.62). Those in the usual aftercare condition with PTSD had directionally worse self-regulation scores than those in the usual aftercare condition without PTSD (M = 2.65, SE = .09; CI: 2.46–2.83).


At baseline, all individuals were finishing treatment, were abstinent, and unemployed, so we conducted chi-square tests to compare the four groups on the dependent measures (whether the individual relapsed or was abstinent; whether the person was employed or not) at the two year follow-up. At the 24 month follow-up, there were significant differences between the four groups for unemployment [χ2 (3, N = 125) = 10.00, p < .05]. Unemployment rates were 54% for individuals with PTSD in usual aftercare, 47% for individuals without PTSD in usual aftercare, 29% for individuals with PTSD in the Oxford House condition, and 20% for individuals without PTSD in the Oxford House condition. Individuals without PTSD in Oxford House had significantly lower unemployment than those without PTSD in the usual aftercare condition (20% versus 47%) [χ2 (1, N=98) = 7.93, p < .01]. There was a directional but not significant difference when we compared those with PTSD in the Oxford House condition to those with PTSD in the usual aftercare condition (29% versus 54%) [χ2 (1, N=27) = 1.78, p = .18].

Substance Use Outcomes

There was a significant difference in the relapse rates [χ2 (3, N = 138) = 16.44, p < .01] among the four groups. Forty-one percent of those in the Oxford House condition with PTSD and 28% of those in the Oxford House condition without PTSD relapsed at the 24 month follow-up. However, relapse rates were higher in the usual aftercare condition, with 69% of those with PTSD and 64% of those without PTSD relapsing. When we compared relapse rates for those with PTSD in the Oxford House condition to those with PTSD in the usual aftercare condition, there was a directional but not significant difference (41% versus 69%) [χ2 (1, N=30) = 2.33, p =.13]. Individuals without PTSD in Oxford House had significantly lower relapse rates than those without PTSD in the usual aftercare condition (28% versus 64%) [χ2 (1, N=108) = 13.80, p < .001].


The present study found that there was a significant difference in self-regulation scores among the four groups. The PTSD group in the usual aftercare condition had the lowest self-regulation levels when compared to the two Oxford House groups. These findings are in accordance with research that indicates that individuals who are chronically dependent on substances and have PTSD have impairments in their ability to manage emotional responses (Phan, Wager, Taylor, & Liberzon, 2002). These types of impairments may affect decision making and impulse control, and this could have an effect on substance abuse or relapse (Matto, 2007). In contrast, participants with PTSD who lived in Oxford House after their substance use treatment showed higher self-regulation, which is associated with better psychological adjustment (Tangney et al., 2004). This is consistent with research by Tangney et al. (2004) who found people with high self-regulation showed less impulse-related problems, such as alcohol problems. These findings suggest that abstinence supportive settings such as Oxford Houses might be particularly important for individuals with PTSD.

Oxford Houses might represent an environment that increases self-regulation for individuals due to its structured model. Individuals living in Oxford House are responsible for paying rent, completing chores, participating in house governance. They also provide the individuals with a stable place to live that does not have restrictions on length of stay and promotes abstinence. Perhaps the stability and the responsibilities that are associated with Oxford House are what help individuals manage their behavior and emotions. This is particularly useful for individuals with comorbid SUD and PTSD, since those with PTSD were found to have lower levels of self-regulation. Oxford House is an environment that supports sobriety and is associated with higher levels of self-regulation, which together produce good outcomes (Jason, et al., 2007). It is very likely that individuals with co-morbid conditions such as PTSD are at higher risk following completion of substance abuse treatment, and finding placements that provide support for abstinence might be particularly important for such high risk individuals.

Those in the Oxford House condition without PTSD had significantly lower rates of unemployment and lower substance use than those in the usual aftercare condition without PTSD. A person living in an Oxford House is required to be employed and people in Oxford Houses may find out about employment opportunities through other Oxford House residents. The group with the directionally highest rates of unemployment and substance use was the PTSD group in the usual aftercare condition. This finding is important because previous research shows that PTSD status greatly decreases the probability of employment (Savoca & Rosenheck, 2000). Therefore, these findings suggest that Oxford House might be a positive environment not only for people in recovery from substance abuse, but also for people with comorbid PTSD and SUD.

This study had a limited sample size of people with PTSD, with only 32 of the 150 participants having a lifetime diagnosis of PTSD. Future research should include a larger PTSD group as well as evaluate the generalizeability of our findings across other samples. In addition, future studies should also consider assessing variables such as mental health treatment that occurs for individuals following discharge from substance abuse treatment settings.

In summary, the present study found that individuals with SUD and PTSD in the usual aftercare condition showed significantly lower levels of self-regulation than those with SUD and PTSD who were randomized to the Oxford House condition. Other findings also point to possible benefits of providing an abstinence environment following substance abuse treatment. Those without this type of supportive environment appear to have problematic outcomes, and those with other co-morbid factors like PTSD, are probably at even higher risk. These findings highlight the importance of abstinence supportive settings following substance use treatment, especially for individuals with PTSD.


The authors appreciate the financial support from the National Institute on Alcohol Abuse and Alcoholism (NIAAA grant numbers AA12218 and AA16973), the National Institute on Drug Abuse (NIDA grant numbers DA13231 and DA19935), and the National Center on Minority Health and Health Disparities (grant MD002748).


  • American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-IV-TR. 4. Washington, DC: American Psychiatric Association; 2000.
  • Auerbach ES, Richardson P. The long-term work experiences of persons with severe and persistent mental illness. Psychiatric Rehabilitation Journal. 2005;28:267–273. [PubMed]
  • Blume AW, Schmaling KB, Marlatt GA. Revisiting the self-medication hypothesis from a behavioral perspective. Cognitive and Behavioral Practice. 2000;7:379–384.
  • Blustein DL. The role of work in psychological health and well-being. American Psychologist. 2008;63:228–240. [PubMed]
  • Brown PJ, Recupero P, Stout R. PTSD substance abuse comorbidity and treatment utilization. Addictive Behaviors. 1995;20:251–254. [PubMed]
  • Gil-Rivas V, Prause J, Grella CE. Substance use after residential treatment among individuals with co-occurring disorders: The role of anxiety/depressive symptoms and trauma exposure. Psychology of Addictive Behaviors. 2009;23:303–314. [PMC free article] [PubMed]
  • Horton J, Compton WM, Cottler LB. Assessing psychiatric disorders among drug users: Reliability of the revised DIS-IV. In: Harris L, editor. NIDA Research Monograph-Problems of Drug Dependence. Washington, DC: 1998. NIH Publication No. 99–4395.
  • Jason LA, Davis MI, Ferrari JR, Bishop PD. Oxford house: A review of research and implications for substance abuse recovery and community research. Journal of Drug Education. 2001;31:1–27. [PubMed]
  • Jason LA, Olson BD, Ferrari JR, Lo Sasso AT. Communal housing settings enhance substance abuse recovery. American Journal of Public Health. 2006;96:1727–1729. [PubMed]
  • Jason LA, Olson BD, Ferrari JR, Majer JM, Alvarez J, Stout J. An examination of main and interactive effects of substance abuse recovery housing on multiple indicators of adjustment. Addiction. 2007;102:1114–1121. [PMC free article] [PubMed]
  • Khantzian EJ. The self-medication hypothesis of substance use disorders: A reconsideration and recent applications. Harvard Review of Psychiatry. 1997;4:231–244. [PubMed]
  • Koenen KC, Stellman JM, Stellman SD, Sommer JFJ. Risk factors for course of posttraumatic stress disorder among Vietnam veterans: A 14-year follow-up of American Legionnaires. Journal of Consulting and Clinical Psychology. 2003;71:980–986. [PubMed]
  • Matto HC. An innovative treatment: stage I evaluation of a dual processing treatment protocol on self-regulation capacities of chronic substance dependent adults. Best Practices in Mental Health. 2007;3:41–51.
  • McLellan AT, Kusher H, Metzger D, Peters R, Smith I, Grissom G, et al. The fifth edition of the addiction severity index. Journal of Substance Abuse Treatment. 1992;9:199–213. [PubMed]
  • Miller NS, Eriksen A, Owley T. Psychosis and schizophrenia in alcohol and drug dependence. Psychiatric Annals. 1994;24:418–423.
  • Najavits LM, Weiss RD, Shaw SR. The link between substance abuse and posttraumatic stress disorder in women: a research review. American Journal on Addictions. 1997;6:273–283. [PubMed]
  • Norman SB, Tate SR, Anderson KG, Brown SA. Do trauma history and PTSD symptoms influence addiction relapse context? Drug and Alcohol Dependence. 2007;90:89–96. [PubMed]
  • Ouimette P, Coolhart D, Funderburk JS, Wade M, Brown PJ. Precipitants of first substance use in recently abstinent substance use disorder patients with PTSD. Addictive Behaviors. 2007;32:1719–1727. [PubMed]
  • Oxford House, Inc. Oxford House Manual. Silver Springs, MD: Oxford House, Inc; 2008.
  • Phan KL, Wager T, Taylor SF, Liberzon I. Functional neuroanatomy of emotion: A meta-analysis of emotionactivation studies in PET and fMRI. Neuroimage. 2002;16:331–348. [PubMed]
  • Read JP, Brown PJ, Kahler CW. Substance use and posttraumatic stress disorders: Symptom interplay and effects on outcome. Addictive Behaviors. 2004;29:1665–1672. [PubMed]
  • Ritsher JB, McKellar JD, Finney JW, Otlingam PG, Moos RH. Psychiatric comorbidity, continuing care and mutual help as predictors of five-year remission from substance use disorders. Journal of Studies on Alcohol. 2002;63:709–715. [PubMed]
  • Robins L, Helzer JE. The half-life of a structured interview: The NIMH Diagnostic Interview Schedule (DIS) International Journal of Methods in Psychiatric Research. 1994;4:95–102.
  • Roth S, Newman E, Pelcovitz D, van der Kolk B, Mandel FS. Complex PTSD in victims exposed to sexual and physical abuse: Results from the DSM-IV Field trial for posttraumatic stress disorder. Journal of Traumatic Stress. 1997;10:539–555. [PubMed]
  • Savoca E, Rosenheck R. The civilian labor market experiences of Vietnam-era veterans: The influence of psychicatric disorders. The Journal of Mental Health Policy and Economics. 2000;3:199–207. [PubMed]
  • Silverman K, Svikis D, Wong CJ, Hampton J, Stitzer ML, Bigelow GE. A reinforcement-based therapeutic workplace for the treatment of drug abuse: Three-year abstinence outcomes. Experimental and Clinical Psychopharmacology. 2002;10:228–240. [PubMed]
  • Smith MW, Schnurr PP, Rosenheck RA. Employment outcomes and PTSD symptom severity. Mental Health Services Research. 2005;7:89–101. [PubMed]
  • Stewart S, Conrod PJ, Samoluk SB, Pihl RO, Donigier M. Posttraumatic stress disorder symptoms and situation-specific drinking in women substance abusers. Alcoholism Treatment Quarterly. 2000;18:31–47.
  • Tangney JP, Baumeister RF, Boone AL. High self-control predicts good adjustment, less pathology, better grades, and interpersonal success. Journal of Personality. 2004;72:271–322. [PubMed]
  • Waldrop AE, Back SE, Verduin ML, Brady KT. Triggers for cocaine and alcohol use in the presence and absence of posttraumatic stress disorder. Addictive Behaviors. 2007;32:634–639. [PubMed]
  • Wing JK, Babor T, Brugha T, et al. SCAN-Schedules for Clinical Assessment in Neuropsychiatry. Archives of General Psychiatry. 1990;47:589–593. [PubMed]