Methamphetamine dependence rose to epidemic levels during the 1990s, creating staggering effects on the healthcare system.1,2 For example, in California, methamphetamine-related admissions to acute care facilities tripled from 3,437 in 1990 to 10,167 in 1994.3 Furthermore, 22,000 of the 67,000 individuals (32%) who recently sought treatment in publicly funded drug abuse centers in California identified methamphetamine as their primary drug of use.1 Despite the increased dependence on methamphetamine, treatment researchers have yet to create specific treatments demonstrated to promote abstinence or reduce relapse. For example, in a study of 57 treatment-seeking methamphetamine-dependent individuals, 56 had received treatment twice prior to their current treatment episode.1
Although there has been a marked increase in methamphetamine dependence, the psychiatric sequelae of methamphetamine dependence have not been well characterized. Syndromes such as amphetamine psychosis4—6 have received descriptive attention over the past four decades, but the prevalence of this and other more frequently occurring disorders or sequelae (e.g., depression) in methamphetamine users has yet to be reported. Prevalence estimates of psychiatric disorders in abusers of other stimulants, such as cocaine, vary, but at maximum they suggest that 3 of 4 cocaine abusers seeking treatment met lifetime criteria for a psychiatric disorder.7,8
In substance dependence disorders other than methamphetamine dependence, increased psychiatric comorbidity is associated with poorer treatment outcome. Substance-dependent individuals with a comorbid psychiatric disorder are more likely to relapse9,10 than individuals without a comorbid psychiatric disorder.
In order to generate basic data on methamphetamine dependence, we obtained estimates of self-reported psychiatric disturbance over the past year in 1,580 arrestees.
The sample included male and female arrestees in city and county jails located in the 14 most populous counties in California. A total of 2,485 men and 978 women were approached. Ninety percent of the men (n=2,236) and 81% of the women (n=790) consented to participate in the study. For this study, 1,446 participants were excluded because data were missing for at least one of the variables of interest, leaving a final sample size of 1,580.
Los Angeles County facilities were oversampled because of greater access to the jails in this region through the national Drug Use Forecasting (DUF) program and because Los Angeles County represents 29% of the total state population and 28% of all arrests. DUF, now called Arrestee Drug Abuse Monitoring (ADAM), was established by the National Institute of Justice (NIJ) for the early detection of metropolitan drug epidemics, allocation of law enforcement, treatment and prevention resources, evaluation of drug abuse and crime reduction programs, and the tracking and forecasting of national trends in drug use. We utilized this sample because of its size and the diversity of the study participants.
Each of the 1,580 arrestees underwent a four-hour interview by a trained research assistant within 48 hours of arrest. Prior to commencing the interview, participants were provided with a complete description of the study and written informed consent was obtained. Subjects were interviewed by using a standardized semistructured instrument that assessed a wide range of domains. A subset of the interview questions focused on participants' demographic profile (see t1), substance use history (see t2), and psychiatric history. A structured diagnostic instrument was not used because the study focused primarily on issues aside from psychiatric status, such as legal history, human immunodeficiency virus—type 1 (HIV) risk behaviors, and treatment history. For this study, pertinent demographic data included age, education, ethnicity, gender, and HIV serostatus as determined by urinalysis. Participants were classified as substance dependent if, during the 12 months prior to the interview, they had 1) used the drug, 2) unsuccessfully tried to decrease their use of the drug, and 3) felt addicted. Psychiatric history was characterized by asking arrestees if they 1) now needed help for psychological or emotional problems or their nerves and 2) had ever been treated as an outpatient in a clinic for emotional or psychological problems. They also were asked if, in the 12 months prior to the survey, they had 1) ever had thoughts of suicide, 2) experienced hallucinations or delusions (did you see or hear things that no one else could or did you think that someone else controlled your thoughts?), 3) had trouble controlling their temper or violent behavior, and/or 4) felt seriously depressed, and lost interest or pleasure in things they usually cared about.
Participants' responses to questions assessing their psychiatric and substance use histories were dichotomously coded (1=yes; 0=no) and entered into two sets of logistic regression equations. In the first set of equations, self-reported methamphetamine dependence was used to predict prevalence of self-reported psychiatric disturbance. In the second set, in order to control for the influence of demographic profile and polysubstance abuse, a three-step hierarchical logistic regression model was employed. In the first step, demographic variables (age, education, gender, ethnicity, and HIV serostatus) were simultaneously entered. In the second step, variables indicating whether participants reported dependence on other drugs (alcohol, powder cocaine, crack cocaine, heroin, and/or marijuana, each coded separately) were simultaneously entered. In the third step, methamphetamine dependence was used as a predictor. By statistically controlling for the association between psychiatric symptoms and demographic profile and substance dependence variables, we could determine more precisely the contribution of reported methamphetamine dependence to psychiatric symptomatology.
t3 shows reported psychiatric symptom occurrence in methamphetamine-dependent individuals compared with those denying methamphetamine dependence. Both the percentages and the logistic regression showed that individuals reporting methamphetamine dependence were significantly more likely to report needing and having received psychiatric treatment; having experienced suicidal ideation; having had difficulty with anger or violence; or having experienced depressive symptoms.
t4 reveals the results of a three-step hierarchical logistic regression analysis in which presence of psychiatric symptoms was regressed on methamphetamine dependence after controlling for demographic profile (age, education, gender, ethnicity, HIV serostatus) and dependence on other substances (alcohol, powder cocaine, crack cocaine, heroin, marijuana). Methamphetamine-dependent individuals were significantly more likely to say they needed psychiatric assistance than individuals denying methamphetamine dependence. Moreover, arrestees reporting a history of methamphetamine dependence, compared with arrestees denying such history, were more likely to indicate that they had experienced suicidal ideation and depressive symptoms in the 12 months prior to the interview.
Because the methamphetamine-dependent participants reported significantly greater levels of depressive symptomatology than other participants and because the prevalence of syndromal depression is two to three times greater for females than for males, the three-step hierarchical logistic regression was repeated while controlling for gender. Females reporting methamphetamine dependence were more likely to endorse a need for psychiatric assistance (χ2=4.25, df=1, P=0.04; odds ratio [OR]=2.04, confidence interval [CI] 1.04—4.47) and to say that they have experienced suicidal ideation (χ2=4.80, df=1, P=0.03; OR=2.13, CI 1.08—4.19) and depressive symptomatology (χ2=6.17, df=1, P=0.02; OR=2.38, CI 1.18—4.78) in the 12 months prior to the interview in comparison to females denying methamphetamine dependence. In contrast, males reporting methamphetamine dependence were more likely than males denying such dependence to report having experienced depressive symptomatology in the 12 months prior to the interview (χ2=7.50, df=1, P=0.006; OR=1.83, CI 1.19—2.82) but were not more likely to report having experienced suicidal ideation in the 12 months prior to the interview or to report needing psychiatric assistance.
Our study revealed that methamphetamine-dependent individuals, particularly women, were more likely to report a syndrome consisting of depression and suicidal ideation. These findings remained statistically significant after controlling for demographic profile, HIV serostatus, and history of other substance dependence. Although difficulties with temper control and inhibition of violent behavior were initially associated with methamphetamine dependence, more controlled statistical analyses revealed that these behaviors were not specific to methamphetamine dependence. Increased risk for depressive symptoms was observed for both women and men reporting methamphetamine dependence. Women, but not men, reporting methamphetamine dependence were more likely to report suicidal ideation and a need for psychiatric assistance at the time of the interview.
These findings are concordant with anecdotal reports from Scandinavia and Japan that describe a constellation of depressive symptoms in abstinent methamphetamine users that did not resolve, even 10 years after the cessation of use. They also are consistent with the findings from studies of abstinent cocaine-dependent individuals, a cohort in which the incidence of depressive disorder far exceeds that of the general population.6,11,12
Both neurobiological and psychosocial mechanisms likely account for increased incidence of depressive symptoms in this sample and the differences in symptom presentation for men and women. Methamphetamine dependence is associated with adverse psychosocial outcomes, such as unemployment and lack of medical insurance.13 It also is associated with chronic alterations in the dopaminergic system.14 Both of these factors most probably contribute to the distress observed in this group of subjects.
We recognize that the survey instruments employed in this type of epidemiological study do not provide the in-depth data that would be available in a study focused exclusively on psychiatric comorbidity. For example, a more focused study could have used standardized measures (e.g., SCID15) to assess for the presence of psychiatric disturbance. Furthermore, the sample included only those arrestees agreeing to participate. Moreover, the study did not assess premorbid psychopathology; thus, for example, we could not distinguish whether depressive symptoms preceded onset of drug use or drug use resulted in the emergence of depressive symptoms.
These limitations notwithstanding, the association between methamphetamine dependence and specific patterns of psychiatric symptoms underscores the need to further investigate and identify consequences of methamphetamine dependence.
Future studies using more precise assessment instruments could determine the validity of the self-reported symptoms in this context. If this syndrome is replicated in a more focused study, then future studies might examine the efficacy of antidepressant medications in the treatment of methamphetamine dependence. Moreover, because a recent study has shown that increased severity of depressive symptoms is associated with the euphorigenic effects of cocaine,16 it would be useful to determine if similar outcomes are observed in recently abstinent methamphetamine-dependent individuals.
This research was supported in part by a grant from the National Institute on Drug Abuse (DA07272). Parts of this paper were presented at the 49th Annual Meeting of the College on Problems of Drug Dependence, Nashville, TN, July 1997.