Construct validity and descriptive validity of somatoform disorders in light of proposed changes for the DSM-5

Abstract

Objective

Current diagnostic criteria for somatoform disorders demand revisions due to their insufficient clinical as well as scientific usability. Various psychological and behavioral characteristics have been considered for the proposed new category Somatic Symptom Disorder (SSD). With this study, we were able to jointly assess the validity of these variables in an inpatient sample.

Methods

Using a cross-sectional design, we investigated N = 456 patients suffering from somatoform disorder, anxiety, or depression. Within one week after admission to the hospital, informed consent was obtained and afterwards, a diagnostic interview and a battery of self-report questionnaires were administered. Logistic regression analyses were performed to determine which variables significantly add to construct and descriptive validity.

Results

Several features, such as somatic symptom severity, health worries, health habits, a self-concept of being weak, and symptom attribution, predicted physical health status in somatization. Overall, our model explained about 50% of the total variance. Furthermore, in comparison with anxious and depressed patients, health anxiety, body scanning, and a self-concept of bodily weakness were specific for DSM-IV somatoform disorders and the proposed SSD.

Conclusions

The present study supports the inclusion of psychological and behavioral characteristics in the DSM-5 diagnostic criteria for somatoform disorders. Based on our results, we make suggestions for a slight modification of criterion B to enhance construct validity of the Somatic Symptom Disorder.

Introduction

For a long time, revisions for somatoform disorders have been demanded since current diagnostic criteria are insufficient for therapeutic as well as scientific use [1], [2], [3], [4], [5]. Different changes have been proposed, such as less restrictive time criteria or the addition of psychosocial symptoms, such as health care utilization, catastrophizing, or a self-concept of being weak [6], [7], [8], [9]. Issues concerning diagnostic validity and clinical utility should be considered in these revisions [10], [11], [12]. The current diagnosis proposed is the Somatic Symptom Disorder (SSD) (updated April 27, 2012; http://www.dsm5.org/ProposedRevision/Pages/proposedrevision.aspx?rid=368). For this new category, medically unexplained symptoms (MUS) are no longer required and somatic complaints have had to be present for at least 6 months. A major revision is the addition of positive criteria (criterion B): concerns about the symptoms' medical seriousness, elevated health anxiety as well as excessive time or energy devoted to the complaints. In the following we give a brief overview of the current status concerning construct validity and descriptive validity of somatization.

Construct validity is given when a diagnosis has an empirically supported theoretical framework that ideally explains its development and maintenance. In the current DSM-IV diagnostic criteria, no existing biopsychosocial model of somatization is reflected [13] even though cognitive-behavioral models of somatization provide useful tools for research and therapy [13], [14], [15]. All models postulate the interaction of various cognitive, behavioral, and physiological factors for the perpetuation of symptoms. In brief, physiological disturbances and emotional arousal produce symptoms that draw bodily attention. If these are attributed to an organic origin, several psychological processes, such as illness worry or catastrophizing may follow, leading to increased illness behavior (help seeking, avoidance, expression of symptoms) and a difference in how symptoms are presented towards the environment (disability, distress, social response).

Patients suffering from somatization show substantial functional impairment [16], [17]. Abnormal illness behavior, such as increased disability days and health care utilization [18], [19], [20], [21], [22], [23], [24], pronounced expression of symptoms [19], [21], or body scanning [19], is present even though versatile [8]. Furthermore, somatization is associated with more catastrophic interpretations of somatic complaints [8], [25], [26]. Patients exhibit a stronger self-concept of physical weakness, higher scores in illness anxiety, and consider themselves more vulnerable to somatic diseases [7], [27], [28]. Greater somatic symptom attribution, which is related to greater impairment than attributing symptoms to psychological causes, has often been found [21], [29], [30], [31], [32]. On the other hand, Hiller and colleagues [33] showed that somatoform patients equally attributed their symptoms to mental/emotional problems and somatic disease.

Descriptive validity is given when a disorder's diagnostic criteria are specific and can be clearly differentiated from others. The well documented high comorbidity of somatoform disorders with depression and anxiety [34], [35], [36] raises the question of diagnostic overlap and increases the demand for more specific criteria for somatization [23]. In the following, findings for psychosocial characteristics included in the proposed revisions for DSM-5, namely SSD's criterion B, are briefly outlined.

Illness worry and catastrophizing have been shown to be increased in somatoform patients compared to patients suffering from other mental disorders [7], [26], [27], [37], [38]. They appear to be especially relevant for immediate and long-term outcome and therefore be valid diagnostic criteria. On the other hand, often no differences were found [7], [39], [40], [41]. Whether a self-concept of bodily weakness is specific for somatization or it is more a result of impairment in general still needs to be investigated [7], [27]. Two promising aspects of illness behavior, body scanning and health care utilization, yield contradictory results. Some indicate specificity [6], [24], [36], [42], [43], [44], [45], others found no difference in illness behavior between somatoform and other patients [19], [24], [46]. Organic attributions as part of previous suggestions for DSM-5 appears to be a promising construct as well. Subjects with somatic complaints exhibited pronounced somatic illness attribution in numerous studies [6], [20], [21], [31] where as in depression and anxiety more psychological attributions can be found [20], [21], [31].

The current literature indicates that there is a deep need to add specific symptoms to the pure existence of physical complaints in order to increase construct and descriptive validity of somatoform diagnoses. As Rief and colleagues [6] pointed out, psychological features should be evaluated together in a study to assess their validity. Aiming at an empirical evaluation of the proposed changes, we conducted a large prospective, diagnostic study to investigate psychological features that might enhance diagnostic validity of somatoform disorders. The current paper focuses on construct validity and descriptive validity; results regarding predictive validity and clinical utility will be reported elsewhere. Based on the current evidence, we tested the following hypotheses:

• Compared to the current diagnostic criteria, additional psychosocial variables (e. g., health anxiety, self-concept of being weak, somatic attribution, illness behavior) will add significantly to the explained variance of self-rated physical health status in somatoform patients (construct validity).

• Psychosocial features will be specific for somatoform disorders when compared with anxiety and depressive patients (descriptive validity).