Hypoactivity of the HPA axis, resulting in low cortisol production, has been implicated in the pathogenesis of this disorder. Investigators showed that patients with CFS had subnormal adrenal response to different doses of ACTH, indicating chronic HPA axis underactivity. These findings were supported by a subsequent study by Scott et al who demonstrated that patients with CFS had small adrenal glands compared to normal controls using computed axial tomography, suggesting adrenal atrophy.

Such findings have led many investigators to believe that CFS is a state of "functional" hypocortisolism. Patients with CFS showed a normal cortisol response when "stronger" stimuli of the HPA axis were applied, such as insulin-hypoglycemia. These data indicate that, while patients with CFS suffer from chronic hypocortisolism, they do not suffer from Addison’s disease, which is a nearcomplete destruction of the adrenal glands. The chronic hypocortisolism, though, can have profound effects on several immune and endocrine systems, further complicating the clinical picture of these patients. Specifically, hypocortisolism can lead to overproduction of inflammatory cytokines (such as IL-6), which in turn can further precipitate the symptom complex of CFS.

Overproduction of inflammatory cytokines, such as IL-6 has been reported in states of CFS. However, others have failed to find such elevation of basal production of inflammatory cytokines in CFS. A possible explanation for such a disagreement in the literature may be that CFS is an illness of remission and relapsesand tends to have a fluctuating clinical course.

Moreover, while patients with CFS may be "functioning" at baseline, they clearly have difficulty functioning under physical or mental stress. They typically try to limit their activities and their exposure to stressful stimuli, so as not to trigger a CFS relapse.

Therefore, evaluation of biological parameters, such as IL-6, at basal conditions may not be the best way to approach this illness diagnostically. Rather, a provocative test needs to be applied. Provocative testing is not uncommon in endocrinology; for example, the diagnosis of adrenal insufficiency or that of growth hormone insufficiency cannot be made on the basis of basal sampling, but rather requires the application of a provocative test, such as an ACTH stimulation test for the former and an insulin-tolerance test for the latter. Similarly, a stimulation test would be more appropriate for the diagnosis of CFS. To date such a test does not exist.

The source of circulating IL-6 in humans has been an enigma for some time. Recent studies have shown that adipose (fat) tissue contributes significantly to the circulating IL-6 levels in humans. For example, several groups, including ours, have found that IL-6 is produced by human adipocytes in vitro.

We are investigating the role of endogenous IL-6 in the pathogenesis of all or part of the CFS symptom complex. The major advantage of this approach is that the cytokine under investigation (IL-6) will not be administered exogenously as a pharmacological agent. Instead, corticotropin- releasing hormone (CRH) will be administered to the subject. This will lead to a modest elevation of endogenous IL-6 concentration, which will approximate more closely than exogenous IL-6 administration the cytokine status of patients with fatiguing illnesses.

The stimulation of IL-6 secretion by CRH is intriguing and may be an important link between the HPA axis and the, inflammatory cytokines in the pathogenesis of CFS-like symptoms. Thus patients with CFS may have an exaggerated IL-6 response to peripheral CRH, which would in-turn lead to or exacerbate the symptom complex of this illness.

Our hypothesis is that chronic hypocortisolism leads to increased production of inflammatory cytokines. Such chronic overproduction of inflammatory cytokines can subsequently lead to the development of CFS symptoms.

Recruitment of subjects for the pilot study was restricted to certain groups in order to increase statistical power. To eliminate genderrelated variability and because CFS affects primarily middle-age women, only female subjects (age 30–50) were included in the pilot study. In addition, hypothalamic- pituitary-adrenal (HPA) axis physiology varies among races. To eliminate race-related variability, which would have a negative impact on the statistical power of the study, only Caucasian subjects were recruited to the pilot study.

Caffeine and alcohol ingestion were discontinued 72 hours before being admitted for the inpatient part of the protocol. Only non-smokers were admitted to the study. Because the adipose tissue is one of the major sources of circulating cytokines and to increase the sensitivity of the study we studied overweight, but not obese, subjects (BMI: 27- 30 kg/m2). Obese (BMI > 30 kg/m2) subjects were not studied because  they were not considered to be "healthy" volunteers, as obesity is — by definition — an illness.

CRH administration:
CRH was administered as a 24-hour intravenous infusion instead of an IV bolus injection. The 24- hour infusion was preferred over the bolus administration for the following reasons: i) our preliminary in vitro data using primary cultures of human adipocytes indicated that a prolonged infusion rather that an IV bolus administration would be needed for biological effects to occur; ii) a prolonged infusion would most likely result in a rather prolonged stimulation of IL-6 production, which would better resemble the chronic characteristics of CFS and related illness.

CRH dosing:
The optimal dose of CRH for stimulation of IL-6 production was unknown. Thus a dose-response study design was implemented. The study was conducted in a doubleblind fashion. The subjects received a bolus (over 1 min) injection of placebo (normal saline) or ovine CRH (oCRH) at doses of 0.01, 0.03, 0.1, 0.3 and 1 mcg/kg, reconstituted in normal saline. This was followed immediately by an infusion of placebo or oCRH at doses of 0.01, 0.03, 0.1, 0.3 or 1 mcg/kg/hr, respectively for 24 hours. The maximum dose was 100 mcg/hr regardless of body weight (100 mcg is the maximum FDA-approved dose for this medication). The Baxter Colleague Infusion Pump was used for the infusion. Five subjects were studied per dose (including saline placebo), totaling 30 subjects., The doses of CRH were selected based on earlier studies showing the minimum effective dose for the HPA axis to be 0.03–0.1 mcg/kg, and the maximum 1–3 mcg/kg. Blood was drawn hourly for 48 hours (24 hours before CRH infusion and 24 hours during CRH infusion). Study parameters included: hourly measurements of plasma cortisol, IL-6, tumor necrosis factoralpha (TNF-alpha), and C-reactive protein.

Low doses of CRH failed to raise IL-6 levels significantly, most likely due to the antagonistic effects of cortisol, which was stimulated at the same time. However, CRH exerted its maximum effect on cortisol at the dose of 0.3 mcg/kg/hr, as 1 mcg/kg/hr did not result in a further increase in circulating cortisol levels, nor in a further increase in urine free cortisol. CRH, therefore, had a stimulatory effect on circulating IL-6 and C-reactive protein at 1 mcg/kg/hr.

We believe that at that dose CRH further stimulated IL-6 release by target organs, such as the adipose tissue, whilst it did not stimulate cortisol release any further. The fact that the cortisol response to CRH reached a "plateau" at 0.3 mcg/ kg/hr, whereas the IL-6 response did not, confers an independent role, of CRH in a net increase in IL-6 release at 1 mcg/ kg/hr. The concomitant increase in circulating C-reactive protein further demonstrates the physiological significance of the IL-6 increase. The effects of CRH on IL-6 appear to be specific, since tumor necrosis factor (TNF)-alpha levels (another inflammatorycytokine produced by fat) were not stimulated at any CRH dose. In fact, plasma TNF-alpha levels were decreased at the two highest CRH doses, most likely due to the known suppressive effect of cortisol on TNF-alpha production.

In summary, these preliminary data suggest that peripheral CRH administration resulted in an increase in peripheral IL-6 levels. The CRH effect on systemic IL-6 concentration may have resulted from a direct CRH effect on IL-6 release by adipose cells.Our next study should help clarify the role of CRH-IL-6 interaction in the pathophysiology of CFS. Our long-term objective is to develop a diagnostic tool for CFS and design new treatment strategies, targeting the HPA-cytokine axis.