OPTIMI: Early Prediction and Prevention of Depression

Institute for Response-Genetics, Departement of Psychiatry (KPPP)

Psychiatric Hospital, University of Zurich


Zurich Standard Texts

Major Depressive Disorders

Major depressive disorders (MDD) affect about 4-8% percent of the general population, causing the loss of the ability to work, to have close relationships, and to have a fulfilling life. Available treatments, though effective, have only modest response rates and, most disturbingly, the question of predicting treatment response in the individual patient is not answerable for any of the currently available therapeutic approaches. A considerable number of patients suffer from recurrent episodes of depression so that the early detection of relapse is key for prevention and the development of effective longterm therapies.

Speaking Behavior and Voice Sound Characteristics

Depression significantly reduces the dynamic expressiveness of human voices, thus greatly reducing inter-individual differences. As a direct consequence, the patients' voices become more similar to each other ("depressive voice"): The patients speak in a low voice, slowly, hesitatingly, monotonously, sometimes stuttering, whispering, try several times before they bring out a word, become mute in the middle of a sentence (Kraepelin 1921). During recovery, however, the patients' speaking behavior and voice sound characteristics return to "normal" values. In a series of clinical studies we found a close correlation in two thirds of patients between speech parameters on the one hand, and psychopathology scores on the other. Consequently, speech analyses allow one to monitor the time course of improvement (and possible deterioration) in these patients.

Refining Predictor Performance

OPTIMI, an EU-funded collaborative project run by a consortium of 11 academic and industrial partners, aims at refining the overall performance of the approach through a combination of several predictor variables. Goal is the reliable prediction of relapse in the individual patient at a high sensitivity and specificity. Predictors under investigation include sleep disturbances, cortisol levels, physical activity, and brain wave patterns, complemented by side effects and medical comorbidity. Predictors are constructed by means of advanced Neural Networks technologies with focus on the individual patient.


Braun S, Botella C, Bridler R, Chmetz F, Delfino JP, Herzig D, Kluckner VJ, Mohr C, Moragrega I, Schrag Y, Seifritz E, Soler C, Stassen HH: Affective State and Voice: Cross-Cultural Assessment of Speaking Behavior and Voice Sound Characteristics. A Normative Multi-Center Study of 577+36 Healthy Subjects. Psychopathology 2014; 47(5): 327-340
Delfino JP, Barragán E, Botella C, Braun S, Bridler R, Camussi E, Chafrat V, Lott P, Mohr C, Moragrega I, Papagno C, Sanchez S, Seifritz E, Soler C, Stassen HH: Quantifying Insufficient Coping Behavior under Chronic Stress. A cross-cultural study of 1,303 students from Italy, Spain, and Argentina. Psychopathology 2015; 48: 230-239
Mohr C, Braun S, Bridler R, Chmetz F, Delfino JP, Kluckner VJ, Lott P, Schrag Y, Seifritz E, Stassen HH: Insufficient Coping Behavior under Chronic Stress and Vulnerability to Psychiatric Disorders. Psychopathology 2014; 47: 235-243


Voice recordings are carried out as self-assessments in the test person's home environment by means of a 2nd-generation microphone with integrated A/D-converter that plugs into any standard USB port of a laptop or netbook. An easy-to-use "voice app" is currently available for Laptops, Tablets, and Smartphones under Windows and Android. The iPhone version is in preparation.

Everis, Spain
ETH, Switzerland
UZH, Switzerland
Freiburg, Germany
MA Systems, UK
Bristol, UK
Xiwrite, Italy
Ultrasis, UK
Jaume, Spain
Valencia, Spain
Lanzhou, China


EU-Grant (FP7):

[ Mail to Webmaster ] k454910@bli.uzh.ch