Depression is a highly prevalent mental disorder and a significant cause of global disability (WHO, 2017). Epidemiological studies on depression generally indicate that females are about twice as likely to be affected as males (Salk et al., 2017). Furthermore, female diagnosed with depression tend to have more symptoms with greater severity as well as more atypical symptoms, and a higher rate of co-morbid anxiety disorder than male depressed patients (Marcus et al., 2008, Schuch et al., 2014). Consequently, it is mandatory to consider sex-specific aspects of depression for understanding the characteristics and optimization of diagnostics and treatment for female and male patients diagnosed with depression.

The hypothalamic-pituitary-adrenal (HPA; Stratakis and Chrousos, 1995) axis is one of the major neuroendocrine systems. It is cascadically activated in response to circadian rhythmic signals or stressful situations: The hypothalamus perceives stressful stimuli and releases corticotropin-releasing hormone (CRH), triggering the secretion of adrenocorticotropic hormone (ACTH) from the anterior pituitary, which ultimately leads to the production of glucocorticoids by the adrenal glands (see review: Herman et al., 2012, Spencer and Deak, 2017). Further, these neuroendocrine signals act upon virtually all tissues and promote whole-body responses to internal and external stimuli by regulating psychophysiological responses such as mood, immune system, energy storage and expenditure. Therefore, a properly functioning HPA axis is essential for restoring homeostasis and maintaining mental and physical health. Conversely, prolonged exposure to stress can alter the adaptive response of the HPA axis to environmental threats, leading to changes in brain function eventually contributing to the subsequent development of depression or other stress-related mental disorders (see reviews: McEwen, 2017, McEwen, 2007). Cortisol, as the primary glucocorticoid end product of the HPA axis, serves as a crucial indicator of HPA axis function and is a widely used peripheral measure in mental health research. Various biological samples can be employed to measure cortisol levels. Commonly, hair samples provide insights into cortisol levels over an extended time period, reflecting chronic stress levels. Blood samples measure cortisol bound to cortisol-binding globulin, saliva and urine samples capture unbound, free cortisol (El-Farhan et al., 2017). Samples of blood, saliva and urine allow for the evaluation of immediate or shorter-term changes in cortisol levels, thereby reflecting more recent cortisol fluctuations. Overall, these biological samples all offer a relatively simple and even non-invasive (for hair, saliva and urine) sampling procedure of measuring cortisol levels.

In adult patients suffering from depression, the HPA axis seems to be hyperactive, as evidenced by excessive secretion of cortisol (see meta-analyses: Belvederi Murri et al., 2014, Stetler and Miller, 2011). Differences in HPA axis activity between patients with depression and healthy individuals have been widely investigated across different cortisol parameters: For morning and evening basal salivary cortisol (see meta-analysis: Knorr et al., 2010), the cortisol awakening response (CAR, Bauduin et al., 2018, Schmidt et al., 2013), which manifests as a significant increase in cortisol levels across the first 30-45 minutes post-awakening in the morning (see review: Stalder et al., 2016), cortisol stress reactivity (see meta-analysis: Ciufolini et al., 2014) all indicate increased levels in depressed patients compared to healthy individuals. Contrarily, a more recent meta-analysis on hair cortisol as an indicator for long-term HPA axis activity, showed no significant differences in patients with depression and healthy individuals (see meta-analysis: Psarraki et al., 2021). Thus, it seems that patients with depression have higher basal cortisol, CAR and stress reactivity but no differences in long-term hair cortisol appeared in comparison to healthy participants.

In the context of neuroendocrinology, it is necessary to consider the influence of the hypothalamic-pituitary-gonadal (HPG) axis when discussing the activity of the HPA axis. The HPG axis is a neuroendocrine system that primarily regulates and controls reproductive function. Sex hormones influence activity of the HPA axis, e.g., while estradiol generally enhances the activity of the HPA axis, testosterone suppresses the HPA axis activity (see reviews: Heck and Handa, 2019, Kudielka and Kirschbaum, 2005, Viau, 2002). These effects are also observed in studies involving female and male participants. For example, girls compared to boys (aged<18 years) tended to have greater cortisol variability throughout the day, higher CAR, and stronger cortisol reactivity to social stress (see review: Hollanders et al., 2017). Adult healthy males showed higher salivary cortisol levels at peak and recovery phases following the Trier Social Stress Test (TSST, Kirschbaum et al., 1993) than females (see meta-analysis: Liu et al., 2017). In the case of basal cortisol in healthy individuals, there was no quantitative result on sex difference in healthy individuals, however, a study based on a large sample (n=1671, 838 females) reported that females had significantly higher morning basal cortisol than males (Larsson et al., 2009). In summary, these findings suggest that sex highly affects HPA axis activity and is reflected in the differences in cortisol levels.

Despite this evidence that females and males differ in their cortisol levels, in the prevalence rates of depression, and in sex hormones, up to now, the assessment of sex differences in cortisol in patients with depression remains scarce. To the best of our knowledge, the meta-analysis of Zorn et al. (2017) on cortisol reactivity to psychosocial stress is yet the only one that examines cortisol levels of patients diagnosed with depression in females and males separately. The meta-analytical data indicates that currently depressed females demonstrated a blunted cortisol reactivity to psychosocial stress compared to female healthy controls, whereas currently depressed males showed an increased cortisol stress reactivity compared to male healthy controls (see meta-analysis: Zorn et al., 2017). However, a direct comparison of cortisol levels between depressed females and males has yet to be conducted. Additionally, findings from experimental studies are conflicting: whereas some studies showed differences in cortisol levels between female and male depressed patients, e.g., in 24h urinary cortisol with higher levels in males (Bos et al., 2005, Grant et al., 2007), others did not, e.g., in CAR (Aubry et al., 2010), and cortisol reactivity to TSST (Lange et al., 2013, Young et al., 2004). Hence, the results across and within different cortisol parameters results on sex differences in depressed patients are mixed, and a quantitative evaluation of basal cortisol, long-term hair cortisol and CAR between females and males diagnosed with depression and also in comparison with sex-matched healthy individuals are still missing.

Therefore, based on existing research, the present systematic review and meta-analysis reviewed studies assessing cortisol levels in females and males diagnosed with depression, including the following cortisol parameters: basal cortisol, hair cortisol, CAR and cortisol stress reactivity. By doing so, we aimed at providing comprehensive and quantitative evidence on the effect of sex on HPA axis activity in patients with depression.