When anorexia nervosa conceals Cushing syndrome: a diagnostic pitfall in hypercortisolemia Authors: Miao Yunn Lim, Serena Chong, Kerry-Lee Milner Published: 2026 ABSTRACT AbstractA 37-year-old woman presented with severe hypertension, mood changes, and weight loss (body mass index 14.6 kg/m2). She exhibited restrictive eating behaviors consistent with anorexia nervosa, alongside biochemical hypercortisolemia with serum cortisol 43.6 µg/dL (1204 nmol/L; reference range: 6.2-18.1 µg/dL; 170-500 nmol/L), 24-hour urinary free cortisol 430.3 µg/24 hours (1187 nmol/24 hours; reference range: <60.2 µg/24 hours; <166 nmol/24 hours), and unsuppressed adrenocorticotropic hormone of 19.1 pg/mL (4.2 pmol/L; reference range: 7.3-63.1 pg/mL; 1.6-13.9 pmol/L). The cortisol level after the 1 mg dexamethasone suppression test was 23.9 µg/dL (659 nmol/L; reference range: <1.8 µg/dL; <50 nmol/L). Cortisol remained markedly elevated after nasogastric feeding and weight restoration. Pituitary magnetic resonance imaging revealed a 4 mm left anterior pituitary lesion. High-dose dexamethasone suppression, corticotropin-releasing hormone stimulation, and inferior petrosal sinus sampling confirmed a pituitary source of adrenocorticotropic hormone. The patient underwent transsphenoidal resection of a corticotroph adenoma. Eleven months later, her hypertension and restrictive eating behaviors had resolved, and her mood significantly improved. Cushing syndrome can coexist with anorexia nervosa, masking typical cushingoid features and presenting with psychiatric manifestations and weight loss. Persistent hypercortisolemia despite nutritional rehabilitation should prompt dynamic endocrine testing and pituitary evaluation to ensure timely diagnosis and treatment.Keywords: Cushing syndrome, anorexia nervosa, hypercortisolemia Introduction Cushing syndrome (CS), characterized by pathological hypercortisolemia, is associated with significant morbidity and mortality. Classical features such as weight gain, central adiposity, and skin changes may be absent when comorbid conditions obscure the phenotype. Anorexia nervosa (AN) can independently elevate cortisol through hypothalamic–pituitary–adrenal (HPA) axis activation, leading to non-neoplastic hypercortisolemia (NNH) that may normalize with nutritional rehabilitation. The coexistence of AN and CS is rare and may present with prominent psychiatric symptoms and weight loss, creating a diagnostic challenge. We report a patient with AN whose evaluation for hypertension revealed higher-than-expected hypercortisolemia, with suspicion of CS leading to a diagnosis of Cushing disease, emphasizing the importance of clinical suspicion and dynamic endocrine testing in distinguishing true CS from functional hypercortisolemia. Case presentation A 37-year-old woman was referred to the endocrinology clinic after investigation of severe hypertension revealed a serum cortisol of 43.6 µg/dL (1204 nmol/L; reference range: 6.2-18.1 µg/dL; 170-500 nmol/L). She had been diagnosed with hypertension 18 months prior and started on ramipril 2.5 mg daily. She reported easy bruising, fatigue, recurrent vaginal candidiasis, and headaches. Approximately 12 months earlier, restrictive eating behaviors (present for at least 4 years) worsened, and she lost 5 kg with resulting secondary amenorrhea. Her husband described frequent self-weighing, calorie counting, and restrictive intake. Despite her low body weight, she perceived herself as “chubby” and “fat.” She had a history of depression and self-discontinued antidepressant treatment 18 months prior. Her medical history was notable for previous renal calculi. She was a nonsmoker, consumed alcohol socially, and denied any recreational drug use. She had 2 biological children and worked for an airline company. Diagnostic assessment Blood pressure was 187/127 mmHg. Height was 147 cm, weight was 32.3 kg (body mass index [BMI] 14.9 kg/m2). Visual fields and extraocular movements were normal. She had rounded facies but no dorsocervical or supraclavicular fat pads. Apart from thin skin and a few bruises, there were no other cutaneous features of CS, including purpura, acne, or violaceous striae. Proximal myopathy was absent. Laboratory evaluation revealed an elevated serum cortisol level of 43.6 µg/dL (1204 nmol/L), unsuppressed adrenocorticotropic hormone (ACTH) of 19.1 pg/mL (4.2 pmol/L; reference range: 7.3-63.1 pg/mL; 1.6-13.9 pmol/L), elevated late-night salivary cortisol of 1.1 µg/dL (31.2 nmol/L; reference range: <0.1 µg/dL; <3.2 nmol/L) and increased 24-hour urinary free cortisol of 430.3 µg/24 hours (1187 nmol/24 hours; reference range: <60.2 µg/24 hours; <166 nmol/24 hours). The cortisol level after the 1 mg dexamethasone suppression test was 23.9 µg/dL (659 nmol/L; reference range: <1.8 µg/dL; <50 nmol/L). She had a low gonadotropin axis with follicle-stimulating hormone 6.2 mIU/mL (6.2 IU/L; reference range: 3.5-12.5 mIU/mL; 3.5-12.5 IU/L), luteinizing hormone 12.6 mIU/mL (12.6 IU/L; reference range: 2.4-12.6 mIU/mL; 2.4-12.6 IU/L), estradiol 15.8 pg/mL (58 pmol/L; early follicular-phase reference range: 30-100 pg/mL; 110-368 pmol/L). The remainder of anterior pituitary function was unremarkable apart from a mildly elevated prolactin (Table 1). Table 1Investigations before and after weight restorationHormoneReference rangeBaseline laboratory valuesaLaboratory values after weight restorationAdrenocorticotropic hormone1.6-13.9 pmol/L(7.3-63.1 pg/mL)5.2 pmol/L(23.6 pg/mL)6.0 pmol/L(27.2 pg/mL)Cortisol170-500 nmol/L(6.2-18.1 µg/dL) 707 nmol/L (25.6 µg/dL) 499 nmol/L(18.1 µg/dL)24-hour urinary free cortisol<166 nmol/24 hours(<60.2 µg/24 hours) 1187 nmol/24 hours (430.3 µg/24 hours) 1195 nmol/24 hours (433.2 µg/24 hours) Late-night salivary cortisol<3.2 nmol/L(<0.1 µg/dL) 31.2 nmol/L (1.1 µg/dL) 33.5 nmol/L (1.2 µg/dL) 1 mg dexamethasone suppression test<50 nmol/L(<1.8 µg/dL) 659 nmol/L (23.9 µg/dL) 559 nmol/L (20.3 µg/dL) Thyroid-stimulating hormone0.27-4.2 mIU/L(0.27-4.2 µIU/mL)0.4 mIU/L(0.4 µIU/mL)—Free thyroxine12-22 pmol/L (0.9-1.7 ng/dL)12.9 pmol/L (1.0 ng/dL)—Insulin-like growth factor 111.2-36.2 nmol/L (85.7-276.9 ng/mL)14 nmol/L (107.1 ng/mL)—Growth hormone<5 µg/L(<5 ng/mL)1.3 µg/L(1.3 ng/mL)—Prolactin100-500 mIU/L (4.7-23.5 µg/L) 569 mIU/L (26.7 µg/L) —Abnormal values are shown in bold. Laboratory values were measured in SI units and converted to conventional units for manuscript data presentation. a Baseline values were drawn at different time points compared with the initial presentation to the clinic. Investigations before and after weight restoration Abnormal values are shown in bold. Laboratory values were measured in SI units and converted to conventional units for manuscript data presentation. a Baseline values were drawn at different time points compared with the initial presentation to the clinic. An outpatient pituitary magnetic resonance imaging (T1-weighted with dynamic and delayed postcontrast sequences) performed 1 month prior to presentation demonstrated a focal hypointense lesion in the left anterior pituitary gland measuring 4 × 3 × 2.5 mm (Fig. 1). Dual-energy X-ray absorptiometry demonstrated low bone mineral density for age (right total hip Z-score −1.6; lumbar spine Z-score −1.0). Gadolinium-enhanced T1-weighted magnetic resonance imaging of the brain demonstrating focal, hypointense lesion in the left anterior pituitary gland measuring 4 × 3 × 2.5 mm in the coronal view. Given the association between elevated cortisol levels and AN, the patient was admitted for voluntary nasogastric feeding for 4 weeks, during which her weight increased from 32.3 kg (BMI 14.6 kg/m2) to 38.7 kg (BMI 17.9 kg/m2). Despite sustained weight restoration and menstrual cycle restoration, screening tests for hypercortisolemia remained persistently elevated for the next 8 months (Table 1). Subsequent endocrine investigations (Table 2) included a high-dose dexamethasone suppression test demonstrating partial cortisol suppression, a corticotropin-releasing hormone (CRH) stimulation test suggestive of a pituitary source of ACTH secretion, and inferior petrosal sinus sampling (IPSS; Table 3) confirming a central-to-peripheral ACTH gradient at baseline and post-CRH stimulation, collectively supporting a diagnosis of Cushing disease. Table 2Results of further dynamic testingInvestigationHormoneReference rangeBaseline valuePosttest value8 mg dexamethasone suppression testACTH1.6-13.9 pmol/L(7.3-63.1 pg/mL)0.6 pmol/L (2.7 pg/mL)—Cortisol170-500 nmol/L(6.2-18.1 µg/dL)709 nmol/L (25.7 µg/dL)266 nmol/La (9.6 µg/dL)Corticotropin-releasing hormone testACTH1.6-13.9 pmol/L(7.3-63.1 pg/mL)6.2 pmol/L (28.2 pg/mL)11 pmol/Lb (50.1 pg/mL)Cortisol170-500 nmol/L(6.2-18.1 µg/dL)587 nmol/L (21.3 µg/dL)730 nmol/Lc (26.5 µg/dL)Laboratory values were measured in SI units and converted to conventional units for manuscript data presentation.Abbreviation: ACTH, adrenocorticotropic hormone. a 63% suppression from baseline cortisol; >50% suppression of cortisol post high-dose dexamethasone is consistent with a central source of ACTH. b Peak value during test; peak ACTH >50% from basal value favors Cushing disease over ectopic source. c Peak value during test; peak cortisol >30% from basal value favors Cushing disease over ectopic source. Results of further dynamic testing Laboratory values were measured in SI units and converted to conventional units for manuscript data presentation. Abbreviation: ACTH, adrenocorticotropic hormone. a 63% suppression from baseline cortisol; >50% suppression of cortisol post high-dose dexamethasone is consistent with a central source of ACTH. b Peak value during test; peak ACTH >50% from basal value favors Cushing disease over ectopic source. c Peak value during test; peak cortisol >30% from basal value favors Cushing disease over ectopic source. Table 3Inferior petrosal sinus sampling results with CRH stimulationSampleLeft IPS ACTHRight IPS ACTHPeripheral ACTHIPS:P ratioBasal5.9 pmol/L (26.7 pg/mL)61.6 pmol/L (279.7 pg/mL)3.9 pmol/L (17.7 pg/mL)15.8Basal5.8 pmol/L (26.3 pg/mL)53.4 pmol/L (242.5 pg/mL)3.7 pmol/L (16.8 pg/mL)14.4Basal7.6 pmol/L (34.5 pg/mL)148.0 pmol/L (672.0 pg/mL)2.7 pmol/L (12.3 pg/mL)54.8CRH, 100 μg IV +2 minutes89.8 pmol/L (407.8 pg/mL)85.4 pmol/L (387.8 pg/mL)3.0 pmol/L (13.6 pg/mL)29.9 +5 minutes604.0 pmol/L (2742.8 pg/mL)682.0 pmol/L (3097.0 pg/mL)7.3 pmol/L (33.1 pg/mL)93.4 +10 minutes367.0 pmol/L (1666.5 pg/mL)270.0 pmol/L (1226.1 pg/mL)10.6 pmol/L (48.1 pg/mL)34.6The maximal basal IPS:P ratio was 54.8, and the maximal post-CRH IPS:P ratio was 93.4, both exceeding established diagnostic thresholds for a pituitary source of ACTH excess (basal ≥2; post-CRH ≥3). Laboratory values were measured in SI units and converted to conventional units for manuscript data presentation.Abbreviations: ACTH, adrenocorticotropic hormone; CRH, corticotropin-releasing hormone; IPS, inferior petrosal sinus; IPS:P, inferior petrosal sinus-to-peripheral ratio. Inferior petrosal sinus sampling results with CRH stimulation The maximal basal IPS:P ratio was 54.8, and the maximal post-CRH IPS:P ratio was 93.4, both exceeding established diagnostic thresholds for a pituitary source of ACTH excess (basal ≥2; post-CRH ≥3). Laboratory values were measured in SI units and converted to conventional units for manuscript data presentation. Abbreviations: ACTH, adrenocorticotropic hormone; CRH, corticotropin-releasing hormone; IPS, inferior petrosal sinus; IPS:P, inferior petrosal sinus-to-peripheral ratio. Secondary hypertension workup including renal artery Doppler ultrasonography, autoimmune and vasculitis screens, serum normetanephrine and metanephrine levels, and aldosterone-to-renin ratio were unremarkable. Treatment The patient underwent transsphenoidal resection 1 year after presentation (4 months after IPSS). Intraoperatively, a small, soft left-sided pituitary tumor was identified and excised along with a segment of involved dura mater. Histopathology confirmed a corticotroph pituitary adenoma: T-box transcription factor positive; pituitary-specific transcription factor 1 and steroidogenic factor 1 negative; weak ACTH immunoreactivity in ∼20% of cells; Ki-67 <3%; no mitoses; and dural invasion present. This corresponded to a corticotroph pituitary adenoma without high-risk features (clinicopathological score 1a). Outcome and follow-up Postoperative day 1 cortisol was 2.5 µg/dL (68 nmol/L). Hydrocortisone replacement (40 mg in the morning, 20 mg at midday) was initiated. Postoperative magnetic resonance imaging showed no residual tumor. Blood pressure medication was discontinued. Hydrocortisone tapering was interrupted by recurrent respiratory and genitourinary infections, worsening fatigue and mood, requiring initiation of mirtazapine 30 mg in conjunction with the mental health team. At last review (11 months postsurgery), she had successfully remained off hydrocortisone for the past 4 months. Morning cortisol was 9.3 µg/dL (256 nmol/L) with ACTH 8.6 pg/mL (1.9 pmol/L). Blood pressure was 130/80 mmHg, and weight was 40 kg with a BMI of 18.5 kg/m2. She continued to experience oligomenorrhea. Gonadotropins had largely normalized with follicle-stimulating hormone 1.7 mIU/mL (1.7 IU/L), luteinizing hormone 7.1 mIU/mL (7.1 IU/L), estradiol 354 pg/mL (1300 pmol/L; late follicular-phase reference range: 100-400 pg/mL; 367-1469 pmol/L). She reported an improved mood and less restrictive eating and had returned to work. Discussion This case demonstrates the diagnostic complexity when low body weight, with compensatory endogenous hypercortisolemia, blunts the typical clinical features of CS, resulting in a diagnostic challenge differentiating true CS from NNH, formerly known as pseudo-CS. Both conditions can present with biochemical hypercortisolemia and overlapping clinical features, yet their underlying pathophysiology and management differ substantially [1, 2]. Once CS was suspected, further investigations confirmed Cushing disease, the most common cause of CS caused by an ACTH-secreting pituitary adenoma [3, 4]. Non-neoplastic hypercortisolemia reflects physiologic activation of the HPA axis in medical conditions such as psychiatric disorders, alcoholism, obesity, or polycystic ovarian syndrome [2, 4, 5]. Anorexia nervosa is a recognized cause of NNH, with elevated cortisol attributed to increased CRH secretion, reduced cortisol clearance, and an adaptive response to low energy availability that lowers basal metabolic rate to maintain euglycemia [1, 6-8]. Individuals with AN typically exhibit serum morning cortisol levels and 24-hour urinary free cortisol 2- to 3-fold higher than those of controls, along with an unsuppressed 1 mg dexamethasone suppression test. Cortisol secretion generally normalizes after weight restoration [9-12]. Dynamic endocrine testing remains essential, although no single test reliably distinguishes CS from NNH [4]. The Endocrine Society guidelines recommend biochemical confirmation of hypercortisolemia, followed by dexamethasone-CRH (dex-CRH) or midnight serum cortisol testing when diagnostic uncertainty remains [9]. The dex-CRH test may be particularly useful in AN because NNH often shows preserved suppression after prolonged dexamethasone exposure and a blunted CRH response, possibly due to CRH receptor downregulation from chronic HPA activation [1, 10]. Late-night salivary or midnight serum cortisol may also assist, as loss of circadian rhythm is characteristic of CS but is usually preserved in NNH [11, 12]. However, these tests remain limited by nonstandardized protocols and assay variability [1]. In our patient, serum and urinary cortisol concentrations were higher than expected in NNH, and persistent hypercortisolemia despite weight restoration and return of menses raised concern for endogenous cortisol excess. A dex-CRH test could not be performed due to patient factors; however, serial late-night salivary cortisol remained ∼10-fold the upper limit of normal before and after weight restoration. Additionally, a CRH test demonstrated a peak ACTH increase >50% and cortisol rise >30%, rather than the blunted response typical of AN. Together with psychiatric symptoms and hypertension, these findings supported true CS. Although IPSS confirmed a pituitary source, interpretation was challenging, as both CS and NNH can involve central ACTH secretion. The concurrence of both AN and CS contributed to an atypical phenotype. Weight gain and central adiposity, the hallmark features of CS, can be attenuated or entirely masked in individuals engaging in restrictive eating behaviors and excessive dietary control [13]. Sustained caloric restriction and vigorous exercise may counteract even the anabolic and adipogenic effects of cortisol excess, resulting in a net phenotype of weight loss. Complicating this diagnostic dilemma, CS itself may present with prominent psychiatric manifestations and weight loss [14, 15]. In their conference abstract, McDonnell et al [14]. described a 22-year-old woman presenting with psychotic depression, food-related delusions, and weight loss as the initial manifestation of Cushing disease, with improvement after transsphenoidal surgery. Neuropsychiatric symptoms are common in hypercortisolemia, with depression reported in up to 81% of cases and psychosis in ∼8% [14, 16]. The mechanism is unclear but may involve cortisol-induced changes in dopamine and glutamate signaling, neuroinflammation, and oxidative stress [16, 17]. Given that our patient's mental health and eating behaviors worsened in the lead-up to her presentation and significantly improved on follow-up, we hypothesize that hypercortisolemia may have played a role in the development of AN, or behaviors characteristic of AN. Few reports describe CS presenting concurrently with AN. These include a 26-year-old woman with longstanding AN who developed overt CS from an adrenal adenoma [18]; a 32-year-old woman with a 2-year history of an eating disorder and BMI of 12.8 kg/m2 whose depression and low weight persisted through standard psychiatric care until an adrenal tumor was identified [19]; and a 17-year-old who developed CS 2 years after AN onset who underwent bilateral adrenalectomy for adrenal hyperplasia before an ACTH-secreting pituitary adenoma was found [20]. These suggest AN and CS may coexist more frequently than currently recognized. It remains undetermined whether the sustained corticotroph drive of starvation-associated hypercortisolemia furnished the milieu from which the adenoma emerged, whether hypercortisolemia from the adenoma generated the eating behaviors and weight loss recognized as AN, or whether the 2 arose coincidentally with one masking the other. What this case establishes is that persistent hypercortisolemia after weight restoration should not be attributed to AN alone. Learning points Anorexia nervosa can mask classical features of CS, including weight gain and central adiposity. Anorexia nervosa can mask classical features of CS, including weight gain and central adiposity. Hypercortisolemia exceeding the 2- to 3-fold elevation typical of AN, or persisting after weight restoration, should prompt evaluation for CS with dynamic testing and assessment of cortisol circadian rhythm. Hypercortisolemia exceeding the 2- to 3-fold elevation typical of AN, or persisting after weight restoration, should prompt evaluation for CS with dynamic testing and assessment of cortisol circadian rhythm. Early recognition and definitive treatment of CS are essential to reduce morbidity and improve psychiatric, metabolic, and reproductive outcomes. Early recognition and definitive treatment of CS are essential to reduce morbidity and improve psychiatric, metabolic, and reproductive outcomes. Anorexia nervosa and CS may coexist more often than recognized, and refractory psychiatric or eating-disorder symptoms despite appropriate treatment warrant endocrine evaluation. Anorexia nervosa and CS may coexist more often than recognized, and refractory psychiatric or eating-disorder symptoms despite appropriate treatment warrant endocrine evaluation. Acknowledgments The authors acknowledge Prof. Ann McCormack for her input and care of the patient, as well as Dr Ben Jonker, the neurosurgeon who performed the transsphenoidal surgery. Abbreviations adrenocorticotropic hormone anorexia nervosa corticotropin-releasing hormone Cushing syndrome dexamethasone-CRH hypothalamic–pituitary–adrenal non-neoplastic hypercortisolemia Contributor Information Miao Yunn Lim, Department of Endocrinology, The Prince of Wales Hospital, Randwick, NSW 2031, Australia; Faculty of Medicine and Health, The University of New South Wales, Randwick, NSW 2031, Australia. Serena Chong, Department of Endocrinology, The Royal Prince Alfred Hospital, Camperdown, NSW 2050, Australia. Kerry-Lee Milner, Department of Endocrinology, The Prince of Wales Hospital, Randwick, NSW 2031, Australia; Faculty of Medicine and Health, The University of New South Wales, Randwick, NSW 2031, Australia. Contributors All authors made individual contributions to authorship. K.-L.M. and S.C. were involved in the initial diagnosis and management of the patient. K.-L.M., S.C., and M.Y.L. were involved in the postoperative care and follow-up of the patient. All authors reviewed and approved the final draft. Funding No public or commercial funding. Disclosures None declared. Informed patient consent for publication Signed informed consent obtained directly from the patient. Data availability Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study. References Associated Data Data Availability Statement Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.