Are Persistent Somatic Symptoms a Trans-Diagnostic Concept? A Cluster Analysis

Abstract Background Psychiatry and other medical specialties have developed various diagnostic constructs related to persistent somatic symptoms (PSS). The risk factors and mechanisms underlying these diagnoses may share trans-diagnostic commonalities. Aims & Objectives This study aimed to consider multiple diagnoses associated with PSS to perform a cluster analysis on patients in a psychosomatic clinic and compare the demographics, psychopathologies, personality traits, and quality of life (QOL) characteristics between the resulting clusters. Method This study, conducted in Taiwan, included 277 adult patients attending a psychosomatic clinic. Diagnostic data were obtained using the Semi-Structured Interview for Somatic Symptoms. The following diagnoses were used for a two-step cluster analysis: DSM-5 somatic symptom disorder, DSM-IV undifferentiated somatoform disorder, ICD-11 bodily distress disorder, ICD-10 undifferentiated somatoform disorder, bodily distress syndrome, irritable bowel syndrome, functional dyspepsia, fibromyalgia, and chronic fatigue syndrome. The two resulting clusters were compared in terms of demographics, psychopathologies, personality traits, and QOL using independent t-tests and Chi-square tests. Results The two-step cluster analysis identified two clusters: cluster 1 (n = 153), with a higher proportion of diagnoses associated with PSS, and cluster 2 (n = 124), with a lower proportion. Diagnoses with higher influence on the clustering included DSM-IV undifferentiated somatoform disorder, ICD-10 undifferentiated somatoform disorder, bodily distress syndrome, and DSM-5 somatic symptom disorder. Intergroup comparisons revealed that cluster 1 had older patients and higher levels of somatic distress (Patient Health Questionnaire-15 scores), general aspect of illness-related anxiety (Health Anxiety Questionnaire scores), behavioral aspect of illness-related anxiety (Scale for the Assessment of Illness Behavior scores), and somatic anxiety (Beck Anxiety Inventory scores). Cluster 1 also exhibited significantly lower overall and physical domain QOL scores. Other demographics, psychopathologies, personality traits, and QOL measures showed no significant intergroup differences. Discussion & Conclusions These findings support the notion of PSS as a trans-diagnostic and trans-symptomatic concept. Patients in psychosomatic clinics can be categorized into two clusters: one characterized by more prominent PSS features and the other with less prominent features. The cluster with more pronounced PSS features tended to exhibit higher somatic distress and illness-related anxiety, and lower QOL, particularly in the physical domain.