Comparison of exhaustion symptoms in patients with stress-related and other psychiatric and somatic diagnoses

Stress-related health problems seem to be increasing in several counties [1‐3]. Data from various sources indicate that stress-related health problems also constitute a significant problem in Denmark. For example, the number of mental disorders reported to the Danish Working Environment Authority almost doubled from 2011 to 2014, making mental disorders the second most frequently reported occupational disorder after musculoskeletal diseases [4]. These data, as well as other data [3, 5, 6], suggest that stress-related health problems are prevalent and have increased in recent years in Denmark. Successful prevention and treatment of stress-related health problems are urgently needed as well as instruments for assessment and monitoring of symptoms. To serve this purpose, a rating scale should be based on the core symptoms of stress-related disorders. However, the mental health problems that result from chronic stress are not recognized as a unitary psychiatric disease in Denmark, and therefore no common diagnostic criteria for stress-related disorders exist. This means that there is no agreement on the core symptoms that such an instrument should address. In contrast to this situation in Denmark, the National Board of Health and Welfare in Sweden has provided tentative diagnostic criteria for stress-related mental health problems and has suggested that the term exhaustion disorder and the ICD-10 diagnostic code F43.8A be used [7]. The core symptom of exhaustion disorder is physical and mental exhaustion, which manifests itself as fatigue, an increased need for recovery after mental exertion, and reduced mental and physical stamina. The symptoms of exhaustion are typically accompanied by other symptoms such as sleeping problems, cognitive difficulties, emotional lability, increased sensitivity to stimuli (sounds, light, etc.), and increased sensitivity to further stress [7].

A number of rating scales have been developed to conform to the diagnostic criteria of exhaustion disorder [8‐10]. KEDS [11] is a brief, symptom-focused rating scale with excellent sensitivity and specificity when comparing patients with exhaustion disorder to healthy control individuals. However, little is known about the performance of KEDS in other patient groups. Because the symptoms addressed in KEDS cover several domains and may occur in many psychiatric and somatic disorders, it can be presumed that they may also be reported to some extent by patients with other diagnoses.

The purpose of the present study was to test a newly-translated Danish version of KEDS in patients from three DOMs. We aimed to answer four questions. First, how are symptoms of exhaustion distributed in different patient groups, including patients with different diagnoses, men and women, and young and old patients? Second, are the symptom levels for exhaustion different in patients diagnosed with stress-related disorders compared to patients with other mental disorders and patients diagnosed with somatic disorders? Third, is KEDS useful for distinguishing between patients diagnosed with stress-related disorders and patients with other disorders, and may it as such be used for diagnostic purposes? Finally, is KEDS free of item bias, i.e. does the response to individual questions in the rating scale differs systematically between different patient groups?

The Danish version of KEDS was tested in 698 patients from three Danish DOMs. The findings of this study are, firstly, that the level of the symptoms of exhaustion as assessed by KEDS is similar in patients with stress-related diagnoses and patients with major depressive disorder and secondly, that both of these groups report more symptoms of exhaustion than patients with somatic diagnoses and patients with ‘problems with employment and unemployment’ (diagnosis code Z56). A third finding is that the Z56 group also reports more symptoms of exhaustion than patients with somatic diagnoses. This finding corroborates the general experience that Z56 patients often display mild signs of a mental disorder but do not meet the criteria for the disorder. The examination of the distribution of the nine specific symptoms addressed by KEDS showed that patients with stress-related diagnoses and patients with major depressive disorder had higher levels of all symptoms than patients with somatic diagnoses. This result underscores that the symptoms of exhaustion addressed in KEDS are indeed characteristics of psychiatric disorders, although they are not very specific. The examination also revealed that the item addressing sleep problems was least able to differentiate between patients with psychiatric and somatic diagnoses. This is in accordance with the observation by Besèr et al. (2014), who found that sleep problems (together with anger and irritability) were associated with the weakest loading in the exploratory factor analysis when investigating the factor structure of the KEDS items.

With regard to the question of KEDS’ ability to differentiate between patients with stress-related diagnoses and patients with somatic diagnoses, the answer is negative, i.e. KEDS cannot be used for diagnostic purposes. Although the mean KEDS sum score of patients with stress-related diagnoses was significantly higher than that of patients with somatic diagnoses, the specificity was deemed to be too low at the cut-off score of 19 suggested by Besèr et al. [11] for the KEDS sum score to be useful for diagnostic purposes. Lastly, the statistical analysis established that none of the nine items on the KEDS scale showed DIF with regard to the patient group (stress-related disorder compared to somatic disorder), gender, or age. This means that the differences in the KEDS sum scores between these groups are not driven by single items to any significant extent, thus providing the answer to the fourth and final research question.

KEDS is a relatively new rating scale and there is only a limited amount of data available to which we can compare our results. In the validation study of KEDS, Besèr et al. [11] found a mean KEDS score of 29.9 (SD = 6.6) in 200 patients with exhaustion disorder and a mean KEDS score of 6.2 (SD = 5.3) in 117 healthy individuals. The mean sum scores of patients with stress-related diagnoses (29.3; SD = 8.0) and major depressive disorder (31.4, SD = 8.1) in this study are similar to the mean sum score of patients with exhaustion disorder in their study. However, it is noteworthy that the mean KEDS sum score of patients with somatic diagnoses was relatively high compared to that of the healthy individuals in the study by Besèr et al. [11]. This is not surprising, considering that it is well-known that various non-psychiatric diseases are associated with increased levels of fatigue [15‐21]. Moreover, it has been shown that symptoms of fatigue are prevalent in primary care patients [22, 23] and that somatic disease is associated with increased levels of mental fatigue among the general population [24]. Nevertheless, the results in this study suggest that pronounced exhaustion is a characteristic of patients with psychiatric disorders, including stress-related disorders.

The earlier phases of exhaustion disorder are amenable to preventive measures [25]. The high level of symptoms of exhaustion in the Z56 patients observed in this study suggests that KEDS might be used to identify an early prodromal phase of the exhaustion disorder. A high level of symptoms of exhaustion could serve as a warning sign and may lead the doctor to inquire about possible underlying causes. This may lead to a treatment that could help the patient and alleviate the symptoms.

Moreover, symptoms of exhaustion also persist for a long time in patients with stress-related disorders [26]. Since mental and physical exhaustion will affect the patient’s daily functioning, it is plausible that the manifestation of symptoms of exhaustion may interfere with the patient’s compliance with treatment or may complicate the return to work from a sickness-related absence [27]. Hence, KEDS could also be useful in assessing symptoms of exhaustion in secondary prevention as well as in determining when patients are ready to return to work.

This study has strengths as well as limitations. A strength is that it investigates symptoms of exhaustion in a clinical population [28]. Furthermore, it is also a strength that the patients were recruited from three different DOMs, which increases the generalizability of the findings to patients in other DOMs. However, the low overall response rate must be mentioned as one of the limitations of the study. This might have influenced the mean score of the different patient subgroups to some extent, because highly exhausted patients are more likely to refuse to participate in the study. The mean KEDS sum score of patients with stress was, however, almost identical to the mean score of patients with exhaustion disorder in the study by Besèr et al. [11], which we interpret to indicate that this effect was not very strong. Another limitation is that information about the patients who chose not to participate was not collected in the study. Thus, a drop-out analysis could not be made. It is also a limitation that the classification of patients was based solely on the diagnosis assigned at the DOM. Hence, patients diagnosed with a somatic disorder may also have a stress-related disorder, depression, or another psychiatric disorder. However, double diagnoses were recorded and constituted only a negligible fraction. Therefore, we believe that the possible contribution from undetected severe psychiatric disorders does not explain the level of symptoms of exhaustion in patients with somatic diagnoses. Finally, it is a limitation that we did not include a non-patient comparison group in the current study.

Patients with stress-related diagnoses (ICD-10 diagnostic codes F43) and patients diagnosed with major depression (F32 + F33) reported stronger symptoms of exhaustion as assessed by KEDS than a diverse group of patients with various somatic diagnoses. However, the specificity of the KEDS sum scores was too low for KEDS to be useful for differential diagnostics. On average, and regarding both the total KEDS score and the scores on individual KEDS items, women tended to select the more severe response alternatives (i.e., score higher). Patients without a disorder diagnosis but with problems due to employment or unemployment (Z56) reported a level of symptoms between that of patients with psychiatric diagnoses and patients with somatic diagnoses. This might indicate a mild exhaustion state that warrants attention from the doctor. The ability of KEDS to detect mild states of exhaustion, which are possible pre-stages to severe stress-related disorders, needs to be further investigated.