Cushing's syndrome is characterized by endogenous glucocorticoid excess, either adrenocorticotropic hormone (ACTH)-dependent or ACTH-independent, both with a variety of underlying causes (1). The vast majority of patients has Cushing's disease caused by an ACTH-secreting pituitary adenoma, with an estimated incidence of 1.2–1.7 per million each year (2). First-choice treatment for Cushing's disease is transsphenoidal pituitary surgery, selectively removing the corticotroph adenoma (3). Ectopic Cushing's syndrome is a rare condition resulting from a non-pituitary ACTH-producing source, and is generally approached by removing the ACTH-producing tumor, if identified and resectable. Excess of glucocorticoids causes osteoporosis, central obesity, insulin resistance, dyslipidemia, hypertension, hypercoagulability, and neuropsychiatric disorders (4, 5). Despite biochemical cure, mortality risk remains increased in patients with Cushing's disease (6).

Sex distribution differs markedly between different etiologies. There is consensus that Cushing's disease occurs up to five times more often in females, and that male patients with ACTH-dependent Cushing's syndrome have a relatively higher risk of an ectopic ACTH-secreting tumor. However, exact figures are lacking and a pathophysiological explanation for the sex distribution is absent (7, 8). Interestingly, the female preponderance is not yet present in prepubertal cases, suggesting that males are diagnosed with Cushing's syndrome at a younger age (9, 10). This is confirmed by some studies in adults with Cushing's syndrome (11, 12), but rejected by others (13, 14). Some cohort studies have reported males to have more severe clinical presentation (higher body mass index and waist circumference, reduced libido and sexual dysfunction, more striae, myopathy, and hypokalemia), biochemical parameters (higher ACTH, serum cortisol, and urinary free cortisol [UFC] concentrations), complications at diagnosis (higher HbA1c concentrations, more often hypertension, anemia, spine osteoporosis with vertebral fractures, rib fractures, and hypercoagulable state), and worse outcome after surgery (more often anemia, lower cortisol normalization rate, and higher recurrence rate), than females (8, 9, 11–16). Pituitary tumors were less easily visualized by pituitary magnetic resonance imaging (MRI) in males (12, 13). One study reported more macroadenomas, with higher invasion and apoplexy rate (11), whereas another study reported equal percentages of macroadenomas in males and females (12). No differences were found in quality of life between males and females on multiple validated questionnaires (CushingQoL, EQ-VAS, and SCL-90-R) (8, 11, 17). Table 1 summarizes the reported literature regarding sex differences (also reported as gender differences) in Cushing's syndrome (8–18).

Potential differences between males and females could for example be based on: (1) Different concentrations of corticosteroid binding globulin (CBG) (19), (2) Different ACTH secretion and cortisol production rate (20, 21), (3) Different interaction between corticosteroids and the gonadotroph axis, and (4) Different disease entity of Cushing's syndrome. A large cohort study focusing on a broad range of potential differences between both sexes is needed to confirm the sometimes conflicting results from previous small studies focusing on a limited number of aspects, to enable a more accurate understanding of the clinical picture of sex-specific ACTH-dependent Cushing's syndrome.

In this cohort study, we compared clinical presentation, biochemical parameters, diagnostic test results, surgical outcome, and morbidity in male vs. female patients with ACTH-dependent Cushing's syndrome. With similar serum cortisol concentrations and UFC, males had higher ACTH concentrations at time of diagnosis than females, with no difference in etiology of Cushing's syndrome and pituitary tumor size between the sexes. Cushing's disease was the most common cause of ACTH-dependent Cushing's syndrome in both male and female patients. At diagnosis of Cushing's syndrome, we found a higher prevalence of osteoporosis with (vertebral) fractures in males than females. After surgery, the higher prevalence of osteoporosis with (vertebral) fractures persisted, and a higher prevalence of anemia in males than in females was found. It is important to note that there were no differences in surgical outcome, recurrence, or mortality between sexes. Thus, male patients with ACTH-dependent Cushing's syndrome seem to constitute a different clinical pattern regarding symptoms and biochemistry, which does, however, not affect the further diagnostic strategy, therapy, or surgical outcome in this cohort study, although comorbidities did differ between sexes. Therefore, no different diagnostic or therapeutic strategy is indicated based on sex. However, based on the differences in comorbidities, extra attention should be given to male patients for diagnosing and treating osteoporosis (with fractures).

A major strength of this study is that it includes a large cohort from two centers and focuses on a broad spectrum of potential differences between sexes, thereby allowing a more accurate description and clear understanding of the clinical picture of ACTH-dependent Cushing's syndrome based on sex, than previous cohort studies. The differences in biochemistry and morbidity found in this study are largely in line with previous studies: higher plasma ACTH (9, 12–14), more often anemia (16) and osteoporosis (with fractures) (8, 11–13). The lack of difference in etiology and pituitary tumor size was in agreement with some studies (9, 12), but not with others (8, 11). Similarly, the lack of difference in surgical outcome was in agreement with one study (9), but not with two others (11, 12). Thus, this study adds evidence to the existence of a sex difference with respect to ACTH, anemia, and osteoporosis (with fractures), and to the lack of a sex difference in etiology and pituitary tumor size. Due to the low number of fractures found in our study, no extensive analyses could be performed to compare the incidence of various types of fractures between men and women. We suggest future studies to focus on the occurrence of fractures in general in men and women with ACTH-dependent Cushing's syndrome, as well as vertebral and femoral fractures separately. The existence of a sex difference regarding surgical outcome is doubtful, as now two studies show a difference, whereas two others do not, indicating that more research is needed into the existence of a sex difference in surgical outcome. As the associations in our study are restricted to ACTH-dependent Cushing's syndrome as exposure, and not as outcome, no explanation can be provided as to why Cushing's disease is more prevalent in women than in men. This question could be answered by future studies examining potential causes of Cushing's disease.

When interpreting the results, the following study limitations should be taken into account. As this study compared a broad spectrum of potential differences between sexes, the Bonferroni method was used to control for multiple testing, leading to tests considered significant only if p < 0.001. Consequently, the cohort size was insufficient to detect any differences that were smaller than 40% between the study groups. In order to find a 20% difference between the study groups with a significance level of p < 0.001, we would have needed at least 450 patients in total, given the sex distribution in this study, which would have taken many more study centers or multiple extra decades to assemble. However, none of the previously published studies on potential sex differences controlled for multiple testing, without exception using a significance level of p < 0.05 despite using many statistical tests per study. This should be taken into account when comparing our study results to the literature. Furthermore, as there were only six patients with ectopic Cushing's syndrome, no analyses could be performed for patients with ectopic Cushing's syndrome separately.

Loss to follow-up could have led to selection bias, since more female than male patients were lost to follow-up. However, as no patient was lost to follow-up within 6 months postoperatively, and most study endpoints were measured within 6 months postoperatively, selective loss to follow-up is unlikely to have influenced study results.

As this study was performed retrospectively, using information from patient files, information could not always be collected from similar time points for each patient, leading to missing data in some patients (e.g., anemia at diagnosis), and information bias is likely to have occurred due to selective questioning of patients (e.g., sex-related disturbances in the Cushing's syndrome Severity Index score). This presumably led to the improbable low percentage of males reporting sex-related disturbances, whereas other studies found males to have sex-related disturbances more often than females (8, 11). Thus, it is important to inquire patients at disease presentation thoroughly about this subject and ideally also interview the patient's partner in this regard.

As should be discussed according to the STROBE guideline (24), this study is theoretically generalizable to all patients with ACTH-dependent Cushing's syndrome. However, the small number of patients with ectopic Cushing's syndrome precludes generalizability for this etiological ultrarare subgroup.

ACTH has been reported to be higher in males than in females with ACTH-dependent Cushing's syndrome consistently across multiple studies (9, 12–14). Some studies, that also reported higher concentrations of UFC in males than in females, suggested that ACTH-dependent Cushing's syndrome is a more aggressive disease in males (12, 13). The increased ACTH concentrations could not be related to larger tumor size of the pituitary adenoma (14). Likewise, as the percentage of patients with ectopic Cushing's syndrome was similar for both sexes, etiological differences are no satisfying explanation for the observed variation in ACTH concentrations. Until now, no study has found a convincing pathophysiological explanation why ACTH is higher in males than in females. However, also in healthy humans, ACTH secretion and cortisol production rate have been found increased in males compared with females (20, 21). Interestingly, higher concentrations of ACTH in males do not consistently seem to increase concentrations of cortisol (serum cortisol as well as UFC). As higher cortisol concentrations are assumed to represent aggressiveness of disease, the lack of increased cortisol concentrations reduces the probability that ACTH-dependent Cushing's syndrome in males actually is a more aggressive disease than in females. Future research with the aim to discover why ACTH is higher in males than in females could focus first on which factors other than pituitary tumor size or etiology might be related to a higher ACTH per sex.

Osteoporosis (with fractures) and anemia were more prevalent in males than in females, which may be explained, at least in part, by patient delay during the diagnostic process, as males may have less pronounced symptoms than females in the earlier stages of the disease (i.e., menstrual disturbances), leading to a greater delay in diagnosis and therefore more, and more severe complications at time of diagnosis in males. Furthermore, it has been suggested that interaction between corticosteroids and the gonadotropic axis leads to hypogonadism in males more often than in females, which may markedly influence bone damage in Cushing's syndrome (8, 13). This is in accordance with our study results, which showed postoperative hypogonadism more often in males than females. This hypogonadism may also cause an endocrine anemia due to the insufficient drive of testosterone on erythropoiesis (25). However, a meta-analysis of prior corticosteroid use and (osteoporotic) fracture risk found no difference between men and women (26). Increased fracture risk and decreased bone mineral density in hypercortisolism is mainly due to suppression of bone formation, which is negatively regulated by an osteocyte-produced factor: sclerostin. Sclerostin is a protein that acts as a key inhibitor of the Wnt signaling pathway in osteoblasts (27). Interestingly, in mice, glucocorticoids stimulate the production of sclerostin (28), whereas circulating sclerostin was decreased in the only available study in patients with Cushing's syndrome to date during hypercortisolism, and increased after correction of glucocorticoid excess (29). Differences in circulating sclerostin between the sexes were not reported in the limited number of patients studied. Both estradiol and testosterone regulate bone turnover, but circulating sclerostin appears to be differentially regulated by sex steroids in women and men (30). Specifically, estradiol, but not testosterone, prevented increases in sclerostin levels following induction of sex steroid deficiency in older men and women, indicating that changes in sclerostin production are likely affected by a complex interplay between estrogens and testosterone and glucocorticoids on bone turnover.

In conclusion, male patients with ACTH-dependent Cushing's syndrome seem to show a different clinical picture than females. However, no different diagnostic strategy or treatment is indicated based on sex, in view of the similar surgical outcome. Clinicians should pay more attention to male patients with ACTH-dependent Cushing's syndrome in particular regarding diagnosis and treatment of accompanying osteoporosis (with fractures).

The datasets supporting the conclusions of this study are available upon reasonable request from the corresponding author.

Patients gave written informed consent for use of their data for scientific research in accordance with the Declaration of Helsinki. Permission from the ethical committees in the LUMC and Charité Universitätsmedizin was granted.

LB, TK, NB, CS, OD, and AP contributed to the conception and design of the study. LB and FH collected the data and organized the database. LB performed the statistical analyses and wrote the first draft of the manuscript. All authors contributed to the manuscript revision, read and approved the submitted version.