Characterization of fatigue states in medicine and psychiatry by structured interview.

CONTEXT: Unexplained fatigue states are prevalent, with uncertain diagnostic boundaries.

OBJECTIVE: Patients with fatigue-related illnesses were investigated by questionnaire and a novel semistructured interview to identify discriminatory features.

METHODS: Cross-sectional samples of women from specialist practices with chronic fatigue syndrome (n = 20), postcancer fatigue (PCF; n = 20), or major depression (n = 16) were recruited. Additionally, two longitudinal samples were studied: women with fatigue associated with acute infection who subsequently developed postinfective fatigue syndrome (n = 20) or recovered uneventfully (n = 21), and women undergoing adjuvant therapy for breast cancer experiencing treatment-related fatigue who subsequently developed PCF (n = 16) or recovered uneventfully (n = 16). Patients completed self-report questionnaires, and trained interviewers applied the Semi-structured Clinical Interview for Neurasthenia. The receiver operating characteristics curves of the interview were measured against clinician-designated diagnoses. Cluster analyses were performed to empirically partition participants by symptom characteristics.

RESULTS: The interview had good internal consistency (Cronbach alpha "fatigue" = .83), and diagnostic sensitivity and specificity for chronic fatigue syndrome (100% and 83%) and major depression (100% and 72%), with reasonable parameters for PCF (72% and 58%). Empirical clustering by "fatigue" or "neurocognitive difficulties" items allocated most patients to one group, whereas "mood disturbance" items correctly classified patients with depression only.

CONCLUSIONS: The Semi-structured Clinical Interview for Neurasthenia offers reliable diagnostic use in assessing fatigue-related conditions. The symptom domains of fatigue and neurocognitive difficulties are shared across medical and psychiatric boundaries, whereas symptoms of depression such as anhedonia are distinguishing.