Further evidence of self-medication: personality factors influencing drug choice in substance use disorders.


According to Khantzian's (2003) self-medication hypothesis (SMH), substance dependence is a compensatory means to modulate affects and self-soothe in response to distressing psychological states. Khantzian asserts: (1) Drugs become addicting because they have the power to alleviate, remove, or change human psychological suffering, and (2) There is a considerable degree of specificity in a person's choice of drugs because of unique psychological and physiological effects. The SMH has received criticism for its variable empirical support, particularly in terms of the drug-specificity aspect of Khantzian's hypothesis. We posit that previous empirical examinations of the SMH have been compromised by methodological limitations. Also, more recent findings supporting the SMH have yet to be replicated. Addressing previous limitations to the research, this project tested this theory in a treatment sample of treatment-seeking individuals with substance dependence (N = 304), using more heterogeneous, personality-driven measures that are theory-congruent. Using an algorithm based on medical records, individuals were reliably classified as being addicted to a depressant, stimulant, or opiate by two independent raters. Theory-based a priori predictions were that the three groups would exhibit differences in personality characteristics and emotional-regulation strategies. Specifically, our hypotheses entailed that when compared against each other: (1) Individuals with a central nervous system (CNS) depressant as drug of choice (DOC) will exhibit defenses of repression, over-controlling anger, and emotional inhibition to avoid acknowledging their depression; (2) Individuals with an opiate as DOC will exhibit higher levels of aggression, hostility, depression, and trauma, greater deficits in ego functioning, and externalizing/antisocial behavior connected to their use; and (3) Individuals with a stimulant as DOC will experience anhedonia, paranoia, have a propensity to mania, and display lower levels of emotional inhibition. MANOVAs were used to test three hypotheses regarding drug group differences on the personality variables that were in keeping with the SMH. The MANOVAs for Hypothesis I (Depressant group) and Hypothesis II (Opiate group) were statistically significant. Findings partially support the SMH, particularly in its characterization of personality functioning in those addicted to depressants and opiates.