Sir

A recent paper in your journal (Stewart et al, 2002), about DSM-IV atypical depression (AD), showed that the typical better response to MAOI than to TCA was seen only in the early-onset and chronic subtype, suggesting the need to better study this important subtype. The study aim was to find support for this subtyping of AD. A large database by the author, collected during the last 5 years for other studies (so eliminating any interviewer bias), was scanned for such evidence. Study methods, briefly reported below, are fully explained in previous reports (Benazzi, 1999, 2000a, 2000b, 2002a, 2002b, 2002c, 2003a,2003b).

METHODS

Study Setting

An outpatient psychiatry private practice, more representative of mood disorders usually seen in clinical practice in Italy.

Interviewer

A senior clinical and mood disorder research psychiatrist.

Patients

Consecutive 320 bipolar-II (BP-II) and 243 major depressive disorder (MDD) outpatients, presenting voluntarily for MDE treatment. Substance-related and borderline personality disorders were excluded because of confounding diagnosis of BP-II, and rare in the setting. Clinically significant general medical illnesses and cognitive disorders were also excluded.

Interview methods

During the assessment visit (off psychoactive drugs for at least 2 weeks, apart from a few cases on small doses of benzodiazepines), the following instruments were used: (1) Structured Clinical Interview for DSM-IV Axis I Disorders-Clinician Version (SCID-CV) (First et al, 1997). (2) Global Assessment of Functioning scale (GAF, in SCID-CV) for index MDE severity. (3) Family History Screen (Weissman et al, 2000). Often, family members/close friends supplemented clinical information during interview. DSM-IV criteria for MDE with atypical features specifier (AD) were followed. ‘Early onset’ was defined as the onset of first MDE before age 21 years (according to Stewart et al, 2002), ‘chronic’ as index MDE symptoms lasting more than 2 years. Variables often reported to distinguish atypical vs nonatypical depression were assessed (see references by Benazzi, Agosti and Stewart, 2001; Angst et al, 2002; Matza et al, 2003 and Rabkin et al, 1996). Early-onset chronic AD (EO-C-AD) was compared to nonearly-onset, nonchronic AD (non-EO-C-AD).

Statistics

Logistic regression was used to study associations and to control for confounding and interactions (STATA 7). P-values were two-tailed, α level 0.05.

RESULTS

AD was present in 44.5% (251/563). EO-C-AD was present in 26.2% of AD (66/251). In the AD sample, univariate logistic regression of EO-C-AD vs BP-II, index age, female gender, recurrences, axis I comorbidity, GAF, bipolar family history, DSM-IV atypical symptoms, found significantly associated BP-II (odds ratio=2.2, 95% CI=1.2–4.0), age (odds ratio=0.9, 95% CI=0.9–0.9), more recurrences (odds ratio=4.4, 95% CI=2.3–8.3), more axis I comorbidity (odds ratio=2.4, 95% CI=1.4–4.4), more bipolar (I+II) family history (odds ratio=2.9, 95% CI=1.4–5.9), and hypersomnia (odds ratio=1.8, 95% CI=1.0–3.2). Multivariate logistic regression of EO-C-AD vs variables found significant in the univariate analysis (not including age), including all possible interactions, found that recurrences (odds ratio=4.3, 95% CI=1.8–10.0), axis I comorbidity (odds ratio=2.4, 95% CI=1.2–5.1), bipolar family history (odds ratio=2.7, 95% CI=1.2–6.1), and hypersomnia (odds ratio=2.5, 95% CI=1.1–5.3) were still significant and independently associated.

DISCUSSION

Findings support, on clinical and family history grounds, a distinction between EO-C-AD and non-EO-C-AD. The findings support the pharmacological distinction found by Stewart et al (2002).