Atypical depression and emotion dysregulation: Clinical and psychopathological features.

BACKGROUND

Most atypical depression (AD) cases endorse prominent mood reactivity, anxiety, and interpersonal sensitivity, resembling some of the characteristics of emotional dysregulation (ED). The present study assesses the frequency and clinical features of different levels of ED in ADyes vs. non-AD(ADno) cases.

METHODS

The present cross-sectional study discriminated depressed outpatients screened with the Hamilton Depression rating scale with the Atypical Depression Supplement (SIGH-ADS), Symptom Checklist-90-Revised, Temperament Evaluation of Memphis, Pisa, Paris, and San Diego Auto-questionnaire, 110-item version, 36-item Difficulties in Emotion Regulation Scale (DERS), and Young Mania Rating Scale into people with high (EDhigh) vs. low (EDlow) for a broad range of clinical and psychopathological features. Descriptive statistics were followed by random forest analysis with "out-of-bag"[OOB] computation.

RESULTS

We included 326 patients (MDD = 204[62.60 %], BD-II = 105[32.20 %], and BD-I = 17[5.20 %]). ADyesEDhigh cases had the earliest age at the onset of depression and overall clinical burden. Higher scores at interpersonal sensitivity, somatization, early age at onset of depression, anxious features, non-atypical core of depression, cyclothymic and depressive temperament, DERS total, and strategies scores predicted higher odds of atypical depression (OOB = 0.25). Among other predictors, age at onset of depression somatization and cyclothymic temperament predicted EDhigh group membership (OOB = 0.23). Hyperthymic temperament, the SIGH-ADS atypical balance percentage score, and somatization emerged as top predictors of treatment-resistant-depression (OOB = 0.12) in contrast to the SIGH-ADS-8-item atypical balance, psychotic features, and age at onset for treatment-resistant-bipolar-depression (OOB = 0.16).

LIMITATIONS: Cross-sectional design; treatment-seeking outpatients.

CONCLUSIONS

AD and ED represent intertwined clinical entities potentially relevant to enhanced treatment outcomes, warranting more accurate random-forest models.