Hypoadrenalism in patients with fatigue and rheumatic disease

Many patients with rheumatic disease complain of fatigue. Clinicians may interpret this as part of the disease process in the absence of anaemia or hypothyroidism, and sometimes respond with the empiric addition of steroids to therapy. The possibility of true hypoadrenalism is only occasionally considered, and little data exists on the frequency with which it coexists with rheumatic disease. Random serum cortisol may be requested by clinicians to help exclude hypoadrenalism as a factor in fatigue. We undertook a survey to assess how frequently this test was of clinical relevance, what was done in patients with low results, and which patients were most likely to have true adrenal failure. All random cortisol assays requested by the members of a rheumatological team over one year were identified and the notes examined retrospectively. The indication for the request, the result, the ultimate clinical diagnosis and all prior diagnoses were recorded. Where further investigations were undertaken, these too were noted. The results were compared to those in an age and gender matched population of patients with general medical conditions (excluding endocrine disorders) for whom cortisol assays had also been requested. Random cortisol was requested by a team of four consultants in 74 patients with a variety of rheumatic disorders over 12 months, usually because of unexplained fatigue. Among the control group of 75 medical patients, the commonest reasons for requesting cortisol assay were fatigue, low sodium and unexplained anaemia. Mean cortisol levels were significantly higher in medical patients (512 nmol/L) than those with rheumatic disease (351 nmol/L) [P=0.04]. The results were low (<200 nmol/L) in 14 rheumatic patients and 7 medical patients. Among these 21 individuals, synacthen tests were performed in 16 and were indicative of hypoadrenalism in 6 cases. Further investigations revealed primary hypoadrenalism in 3 patients, with tertiary adrenal suppression from oral steroids in three others. All six of these patients had underlying rheumatic disorders, usually RA (3) or systemic lupus erythematosus (SLE) (2). None of the medical patients had an abnormal synacthen test. Fatigue in patients with rheumatic disease may be a presenting feature of hypoadrenalism. Although adrenal failure is rare in medical patients with anaemia, hyponatraemia or fatigue, patients with rheumatic disease and unexplained fatigue merit a random cortisol. If this is low, synacthen testing may be appropriate. Steroids should not be commenced empirically in such patients until hypoadrenalism has been excluded, as a significant minority will have genuine primary adrenal failure, usually due to autoimmune disease.