According to an epidemiological study, the estimated prevalence of depressive disorder (DD) in mainland China was 3.02 % (Smith, 2014). Another epidemiological study conducted by Huang et al. in 2017 revealed that the prevalence of DD in China had risen to 5.56 % (Huang et al., 2019). Indeed, the prevalence of DD in China has shown an obvious uptrend since the 1980s. Most DD patients are affected by insomnia to some extent, and the association between depression and insomnia is complex. On the one hand, insomnia may result from depression, which is mainly manifested as early-morning awakening, difficulty in falling asleep, shallow sleep, and decreased alertness in the daytime (Pandi-Perumal et al., 2020); on the other hand, insomnia is also a risk factor for DD, which is even considered to be a manifestation of sub-clinical depression, or a sign of exacerbation and more somatization of symptoms (Carney et al., 2017; Li et al., 2016). Some other studies support the coexistence of DD and insomnia, and suggest that improvement in sleep quality of patients with DD can help improve their prognosis (Gebara et al., 2018). Nonetheless, most psychiatrists still assess the sleep quality of patients through use of scales or based on the patient's self-assessment, which introduces an element of subjective bias, especially in patients with atypical depression. Moreover, insomnia or other sleep disturbances may be more prevalent in patients with mild to moderate DD, in contrast to the typical depressant symptoms in severe major depressive disorder (MDD). Given that Polysomnography (PSG) technique provides a more objective assessment of sleep process and sleep structure, it was used to evaluate the sleep quality of patients with mild to moderate DD in this study.

PSG is one of the most effective methods to determine sleep quality. PSG has helped characterize the following unique sleep characteristics of DD patients. The first characteristic is the short sleep latency of rapid eye movement (REM) which leads to increase in quantity, intensity, and density of REM sleep (Augustinavicius et al., 2014). Some studies have found prolongation of REM in the first sleep cycle in patients with MDD (Hoyos et al., 2020). Another characteristic is the abnormal non-REM (NREM) sleep which is manifested as reduction in slow-wave sleep, slow-wave activities, and slow-wave amplitude (Bernert et al., 2017). The third feature is the impaired sleep continuity including the prolongation of sleep latency, increased frequency of awakening, and early-morning awakening (Mi et al., 2020).

Agomelatine is a new type of antidepressant that improves the mood and sleep quality via stimulating melatonin receptors and inhibiting 5-TH2C receptors simultaneously (Mi et al., 2020). In spite of the development of a large number of antidepressants, their response rates are still not satisfactory (Zhou et al., 2020); thus new therapeutic modalities such as repetitive transcranial magnetic stimulation (rTMS) have gradually been developed over the years. rTMS modulates the physiological activities in the brain via current induced from regular magnetic pulses targeted on cerebral cortex (McClintock et al., 2018). U.S. Food and Drug Administration (FDA) has approved the use of rTMS for treatment of DD with 10 Hz applied to left dorsolateral prefrontal cortex (L-DLPFC) (McClintock et al., 2018). The International Federation of Clinical Neurophysiology (IFCN) has also recommended the use of high-frequency rTMS targeted to L-DLPFC for treatment of DD (Lefaucheur et al., 2020).

Though several studies have explored the sleep quality of DD patients, most of these studies were conducted in patients with severe major depressive disorder owing to the associated risk of self-injury and suicide, while mild to moderate DD has been neglected to some extent. However, mild to moderate depression also has a negative influence on the patients' social function and daily life, accompanied with risk of progression to severe depression (Saccarello et al., 2020; Tayama et al., 2019). Moreover, most previous studies in the field of psychiatry have just used ordinary sleep-related scales to evaluate sleep quality, which may generate some subjective bias, while PSG can provide the most objective evidence of sleep quality (Chen et al., 2018; Rundo and Downey, 2019). Therefore, there is no clear consensus on the effects of rTMS on sleep quality (Nardone et al., 2020).

Thus, in this study, we investigated the effect of combined use of high-frequency rTMS and agomelatine on the sleep quality and some blood biomarkers in adult patients with mild to moderate DD using PSG and scales.