This study uses an algorithm to determine whether patients dually diagnosed
with a mood disorder and cocaine dependence have either (a) an autonomous
mood disorder (onset of mood disorder prior to substance use disorder (SUD)
or mood disorder symptoms persist during periods of abstinence or (b) a no
nautonomous mood disorder (onset of SUD preceded mood disorder and mood dis
order symptoms remit during periods of abstinence). The relationship among
autonomy, patient characteristics, and treatment completion is examined. Th
e sample included 67 methadone patients with a mood disorder (87% major dep
ression, 13% bipolar) who were enrolling in a 6-month psychosocial treatmen
t for cocaine use. Of these subjects, 27% were rated as having an autonomou
s mood disorder and 73% a nonautonomous mood disorder. Mean age was 37 year
s; 55% were female and 82% were Hispanic or African-American. All subjects
had been stabilized on methadone (mean = 70 mg). During the 30 days prior t
o study intake, subjects with an autonomous mood disorder, compared to subj
ects with a nonautonomous mood disorder, reported fewer days using cocaine
(12.5 versus 21.1) and fewer days drinking four or more drinks of alcohol (
1.1 versus 6.1). Treatment completion was associated with less cocaine use,
autonomy, and African-American ethnicity. However, when these variables we
re controlled using logistic regression, only autonomous mood disorder and
ethnicity predicted treatment completion. These results suggest that autono
my may be a useful construct to measure, and that subjects with nonautonomo
us mood disorders may need special efforts to ensure treatment retention.