Abstract :
[en] The heterogeneity of depression (i.e., symptomatology profiles, treatment responsiveness) is more and more evidenced. From a clinical perspective, having a clearer overview of the symptoms’ heterogeneity of depression will help (1) to deeper understand its underlined psychological processes and (2) to tailor clinical treatments (i.e., psychological interventions and/or antidepressant medications). In order to so, an effort to determine subtypes of depression has been developed through a cluster-analysis approach. Andreasen et al. (1980) identified three clusters of depressed patients based on the Schedule for Affective Disorders and Schizophrenia; Schacht et al. (2014) evidenced five clusters based on the Hamilton Depression Rating Scale (i.e., lack of insight, sleep/sexual/somatic, typical MDD, gastrointestinal/weight loss, mild MDD); Hybels et al. (2012) found three clusters of depressive patients based on the Center for Epidemiologic Studies-Depression Scale in community seniors; and Guidi et al. (2011) found two clusters of depressives in medically ill patients based on the Structured Clinical Interview for DSM-IV (i.e., depressed somatizers, irritable/anxious depression).
The current study aims at replicating and extending these previous findings in adults presenting depressive symptoms through a cluster-analysis approach. Unlike previous studies, the highlighting of the clusters will be based on the most frequently used assessment of depression, the Beck Depression Inventory-Second Edition (BDI-II). Further, the definition of the clusters will be based on the presence of the depressive symptoms rather than on their severity because symptoms’ severity on the BDI-II can be misrepresentative of the phenomenology of depressive symptoms’ clusters.
A sample of 619 adults from community and mental healthcare centers has been recruited. Inclusion criteria were as follows: being a French-speaking adult aged between 18 and 60 years and presenting at least five symptoms on the BDI-II, irrespective of their severity. Data grouping was achieved through a combination of hierarchical (Ward’s method with squared Euclidian distance measurement) and nonhierarchical procedures (K-means cluster analysis), as recommended by recent theoretical trends.
A six-cluster solution was evidenced: heavy sleepers (high levels of tiredness, loss of energy and increase of sleep), cognitive depressives (high levels of pessimism, past failures, guilty feelings, self-dislike and worthlessness), affective-somatic depressives (high levels of affective symptoms - loss of interest and pleasure - and somatic symptoms - increase of appetite and sleep, irritability), mild depressives (mild levels of all symptoms), sleepless depressives (high levels of decrease of sleep and tiredness), typical depressives (high levels of all symptoms).
Results evidenced the heterogeneity of depressive symptoms, as six different clusters of depressive adults have been found. Two of the clusters identified in the current study are similar to two clusters identified by Schacht et al. (2014) (sleepless depressives are similar to the cluster “sleep/sexual/somatic” because their profile are highly characterized by symptoms related to insomnia and tiredness; mild depressives are similar to the cluster “mild MDD/symptoms” because their profile are characterized by low to average levels on all symptoms).
The take-home message of this study is that depression is a heterogeneous condition. Consequently, it is necessary to consider this heterogeneity in order to tailor the psychological intervention.