Nineteen female adolescent inpatients diagnosed with anorexia nervosa, restricting type (AN-R) and 16 non-eating disordered (ED) controls were assessed for plasma dehydroepiandrosterone (DHEA), dehydroepiandrosterone-sulphate (DHEA-S), and cortisol levels, and for eating-related and non-eating-related psychopathology. AN-R patients were assessed at admission, 1 month and 4 months following hospitalization. The non-ED controls were assessed once. No baseline between-group differences were found in plasma cortisol, DHEA, and DHEA-S levels, whereas the patient group had a significantly lower Cortisol/DHEA-S ratio and elevated scores on most psychopathological parameters. A significant increase was found in the body mass index of the AN-R patients at 4 months post-hospitalization, accompanied by a decrease in plasma cortisol levels and a trend towards decreased Cortisol/DHEA and Cortisol/DHEA-S ratios, whereas no change occurred in psychopathology. The difference in Cortisol/DHEA-S ratio between AN-R patients and non-ED controls, and the different patterns of change in cortisol vs. DHEA(-S) levels following weight restoration, may in part account for the feeding difficulties in AN, particularly during refeeding.
Neurosteroids are steroids synthesized either directly in the central nervous system (CNS) or in the periphery that have definite effects on the brain (Monteleone et al., 2001 and Patchev et al., 1994). The most important of these neuroactive steroids include allopregnanolone (3α5α-tetrahydroprogesterone), synthesized within the CNS glial cells from pregnanolone, dehydroepiandrosterone (DHEA), and its sulphated metabolite dehydroepiandrosterone-sulphate (DHEA-S). The synthesis of DHEA and DHEA-S has not been demonstrated in the brain, apparently because the enzyme cytochrome P45017α, which converts pregnanolone to DHEA, is not found within the CNS ( Monteleone et al., 2001). Nevertheless, these two neurosteroids may persist in the brain unchanged for over 2 weeks following adrenalectomy and gonadectomy, demonstrating that central DHEA and DHEA-S levels are likely independent of their peripheral biosynthesis in the adrenals or gonads ( Corpechot et al., 1981 and Baulieu, 1991).
Neurosteroids can alter neuronal excitability by binding to ion channel-coupled neurotransmitter receptors at the cell membrane. They may act at inhibitory γ-amino-butyric acid type-A (GABA-A) and glycine receptors, at excitatory glutamate receptors, and at serotonin, N-methyl-d aspartate, sigma type 1, and nicotinic acetylcholine receptors ( Engel and Grant, 2001).
Converging evidence suggests that neurosteriods may be relevant to the pathophysiology of several psychiatric disorders such as depression, schizophrenia post-traumatic stress disorder and attention deficit hyperactivity disorder (Spivak et al., 2000, Michael et al., 2000, Strous et al., 2001, Strous et al., 2004 and Ritsner et al., 2004). Specifically, they may modulate the expression of mood, anxiety, aggression, cognition, general well-being and activity (Van Goozen et al., 1998, Wolf and Kirschbaum, 1999, Monteleone et al., 2001 and Michael et al., 2000).
There is some evidence for a putative role of neurosteroids in the pathophysiology of feeding behaviors. Allopregnanolone, which is a positive modulator of inhibitory GABA-A and glycine receptors, may increase feeding behavior and weight, whereas DHEA and DHEA-S, which are negative modulators of GABA-A and glycine receptors and positive modulators of excitatory glutamate receptors and serotonin receptors, may decrease food intake and weight (Engel and Grant, 2001 and Kaur and Kulkarni, 2001). The interaction between these neurosteroids and serotonin may be of relevance in the regulation of feeding behavior because of the down-regulation of serotonin found in underweight individuals with anorexia nervosa (AN), and its elevated activity in long-term recovered AN patients (Kaye et al., 2004). Additionally, elevated serotonin activity may increase several AN-related behavioral constraint and dysphoric traits, including obsessionality, rigidity, perfectionism, harm-avoidance, depression and anxiety (Casper et al., 1992, Pollice et al., 1997 and Kaye et al., 2004).
Corticotropin Releasing Factor (CRF) stimulates the release of adrenocorticotropic hormone (ACTH) from the anterior pituitary, which in turn stimulates the adrenal cortex to secrete DHEA and DHEA-S, in addition to cortisol (Leowattana, 2004). Underweight individuals with AN show elevated plasma cortisol secretion, reflecting hypersecretion of endogenous CRF and an overdrive of the CRF/ ACTH system (Hotta et al., 1986). The secretion of CRF and ACTH in AN normalizes after weight restoration, leading to normalization of cortisol levels (Kaye et al., 2004) and theoretically also of DHEA and DHEA-S levels, although this has not been investigated yet. As the secretion of DHEA and DHEA-S by the adrenal gland is driven by the CRF/ACTH system, it seems plausible that alterations in the production of these neurosteroids may occur in patients diagnosed with AN (Monteleone et al., 2001).
Only a few studies have previously assessed neurosteroid function in AN. Zumoff et al. (1983) have found decreased mean daily values of plasma DHEA and DHEA-S in 14 AN patients with a concomitant increase in plasma cortisol concentrations, whereas Winterer et al. (1985) reported normal cortisol levels and reduced baseline and ACTH-stimulated concentrations of DHEA and DHEA-S in six underweight AN women. Other studies have also found reduced levels of DHEA and DHEA-S in acutely ill AN patients (Devesa et al., 1987 and Gordon et al., 2000). In the largest and most rigorous study performed yet, Monteleone et al. (2001) have found increased plasma levels of allopregnanolone, DHEA, DHEA-S, and cortisol in 30 female AN patients, compared with 30 age-matched healthy controls.
The aim of the present study was to assess whether a change would occur in circulatory neurosteroid levels in AN patients following weight restoration compared to the acute underweight stage of the disorder, and whether these changes will be associated with changes in core eating-related and non-eating-related psychopathology. We hypothesized that plasma DHEA, DHEA-S and cortisol levels will be elevated in underweight AN patients compared with non-eating disordered (non-ED) controls, whereas weight restoration will be associated with normalization in the level of these neurosteroids.
