Chronic fatigue syndrome: Identification of distinct subgroups on the basis of allergy and psychologic variables

Abstract

Background: We investigated a role for allergic inflammation and psychologic parameters in the development of chronic fatigue syndrome (CFS). Methods: The design was a comparison between subjects with CFS and age- and sex-matched control cohorts. Studies were performed on CFS subjects (n = 18) and control cohorts consisting of normal subjects (n = 11); allergic subjects (n = 14), and individuals with primary depression (n = 12). We quantified cytokines at baseline as cell-associated immunoreactive peptides and as transcripts evaluated by means of semiquantitative RNA-based polymerase chain reactions. Psychologic evaluations included administration of the Diagnostic Interview Schedule, the Structured Clinical Interview, and the Symptom Checklist 90-Revised. Results: Increases in tumor necrosis factor (TNF)-α were identified in individual subjects with CFS (50.1 ± 14.4 pg TNF-α per 107 peripheral blood mononuclear cells [PBMCs]; mean ± SEM) and allergic subjects (41.6 ± 7.6) in comparison with normal subjects (13.1 ± 8.8) (P < .01 and P < .05) but no further modulation of TNF-α or IL-10. Similarly, self-reported exacerbations of CFS were associated with a further increase in IFN-α (from 2.5 ± 1.0 to 21.9 ± 7.8; P < .05) and occurred at times of seasonal exposures to allergens. This linkage does not permit making any definitive conclusions regarding a causative influence of either seasonal allergies or the increase in cellular IFN-α with the increase in CFS symptoms. The close association between atopy and CFS led us to speculate that CFS may arise from an abnormal psychologic response to the disordered expression of these proinflammatory and antiinflammatory cytokines. Psychologic variables were predictive of immune status within the CFS sample (65.9% of the variance in immune status; F (3,10) = 6.44, P < .05). Specifically, the absence of a personality disorder but greater endorsement of global psychiatric symptoms was predictive of immune activation. Conclusions: Most of our subjects with CFS were allergic, and the CFS and allergy cohorts were similar in terms of their immune status. However, the CFS subjects could be discriminated by the distinct psychologic profiles among subjects with and without immune activation. We propose that in at least a large subgroup of subjects with CFS who had allergies, the concomitant influences of immune activation brought on by allergic inflammation in an individual with the appropriate psychologic profile may interact to produce the symptoms of CFS. In a psychologically predisposed individual, symptoms associated with allergic inflammation are recognized as illness.