Currently, the diagnosis of BD is based on a comprehensive clinical evaluation, there are no specific biomarkers that can indicate the diagnosis, prognosis, or treatment outcome of BD. Coupled with the similarity of the clinical manifestations of BD and MDD, BD is often misdiagnosed, resulting in insufficient or even inappropriate treatment[15]. There are studies suggesting a difference in single nucleotide polymorphisms of biological clock genes associated with BD and MDD[16], as well as biomarker studies suggesting that patients with BD of the evening time phenotype have lower serum IL-6 and TNFα levels than patients with MDD[17], however, the specific differences in circadian rhythm biomarkers between the BD studied above and controls were relatively small. About the genetics of circadian rhythms, assessment of polygenic risk score (PRS) using GWAS demonstrated that circadian rhythm disorders show considerable overlap in psychiatric disorders, particularly among BD, MDD and schizophrenia[12]. Because the phenotypes of psychiatric disorders vary widely and it is unclear how alterations in circadian biomarkers affect changes in gene expression associated with mood, there may still be as-yet-undiscovered circadian biomarkers specific to BD.
Cortisol release in healthy individuals follows a circadian pattern, peaking shortly after waking and minimizing at night. Although dysregulation of the hypothalamic-pituitary-adrenal system has been shown to play a role in the pathogenesis of BPD and mood disorders such as MDD, it is unclear what role circadian rhythms play in this dysregulation. Research suggests that patients with BPD exhibit hypersensitivity of melatonin secretion to light compared to healthy controls, i.e., melatonin secretion in patients with BPD receives stronger inhibition from light and is positively correlated with light intensity[18]. In this regard, it has been found that the Dim Light Melatonin Onset (DLMO) measured in BPD patients under narrow-band blue light irradiation is significantly delayed compared to healthy controls[19]. In addition, the fact that lithium salts can act on the retino-hypothalamic pineal pathway to affect light sensitivity and melatonin secretion also support these observations[20]. Our study distinguishes itself from previous studies that estimated circadian rhythms by measuring the time of initial release of dark light melatonin[21], by monitoring plasma melatonin and cortisol levels at different time point throughout the day, combined with polysomnography to understand the secretion pattern of melatonin and cortisol in different mood states, which can provide clues for the diagnosis and differentiation of bipolar depressive episodes.
In studies of patients with BD, it is generally recognized that the age of onset is early, with more than 70% of these patients presenting with the corresponding clinical manifestations of depression before the age of 25 years[3]. The onset age is 4–5 years earlier in patients with BD than patients with MDD[22]. The demographic characteristics of the present study showed that the age of the patients in the BD group was younger than that of the MDD group, which is consistent with the above findings.
Sleep rhythm disorder is one of the characteristics of circadian rhythm disorders in BPD and also as an independent risk factor for the development of BPD. Sleep rhythm disorders can predict relapse of BPD, so it should be used as an indicator for preventing relapse in patients with BPD. Some studies have shown that BPD patients have abnormalities in total sleep time, sleep latency, and REM sleep latency, whereas other studies have not found similar results[23]. In this study, PSG results showed that the overall distribution of sleep efficiency was higher in the healthy control group than the other three groups, suggesting that patients in RBD still lack continuity of sleep even if clinical symptoms are reduced or even disappeared, and that some of them still need to use medication to ensure the quality of their sleep, and that sleep disturbances may remain as residual symptoms of remission as a circadian rhythm disorder. Previous studies have shown prolonged sleep latency and reduced REM latency in patients with bipolar depressive episodes, and prolonged sleep latency in patients with bipolar remission[24], increased REM phase duration, decreased REM latency and decreased slow wave sleep[25] were shown in patients with MDD. In contrast, our results showed no statistically significant differences in sleep latency, REM sleep latency, and sleep staging among the four groups, which is inconsistent with previous study, that may be related to the small number of study subjects we enrolled. However, there is no consensus on the results of current studies on objective sleeping in patients monitored by PSG due to high variability.
In this study, we measured melatonin and cortisol levels, as one of the characteristics of circadian rhythm disorders, at different time points. A previous meta-analysis found that morning cortisol levels were elevated in patients with schizophrenia and BD compared to healthy controls[26]. This study found that patients with BD had higher levels of melatonin and cortisol at 8:00, cortisol levels at 12:00 and 20:00 than group with MDD, suggesting that disturbances in the secretion of melatonin and cortisol levels may provide clues for differentiating BD and MDD. The above findings also confirm that there is a more pronounced circadian dysregulation in BD than MDD. Of course, there are different results. A recent study that monitored the circadian pattern of cortisol in patients with BPD and healthy controls throughout the day did not find any difference in cortisol levels, and the study did not further clarify whether cortisol level is associated with mood, clinical symptoms, and severity of illness[27]. This may be related to temporal differences in sampling of cortisol and the fact that some of the studies measured salivary cortisol. Previous results have shown that melatonin levels are higher and produced earlier in patients with bipolar manic episodes compared to bipolar depressive episodes[10]. These results provide us some clues: First, the timing and magnitude of melatonin secretion can distinguish the different seizure phases of bipolar. Second, whether the difference in the timing of melatonin secretion can be used as a bioindicator of MDD to distinguish BD requires further research.
Elevated salivary cortisol levels at bedtime are associated with increased suicide attempts[28], a meta-analysis found that cortisol levels were elevated in patients with schizophrenia and BD compared to healthy controls, and the elevation was more pronounced in the schizophrenia group[26]. In our findings, correlation analysis suggested that plasma cortisol levels at 12:00 and 20:00 were positively correlated with suicidal ideation and suicidal ideation, i.e., elevated plasma cortisol was associated with a high risk of suicide, which is in line with the results of previous studies, suggesting that disturbances in the levels of cortisol and abnormalities in circadian rhythms are high risk factors for suicide in patients with BD, and they need to be taken into account in the clinical diagnosis and management of BD.
Limitations
There are some limitations in this study. Firstly, our study was conducted on patients with MDD and BD who had developed mania/hypomania, if a long-term longitudinal study could be conducted before or after an appearance of developed manic/hypomanic episodes in patients with BD, plasma melatonin and cortisol could be collected, and the assessment of the PSG and scale could be improved and compared between the groups, which would provide a more reliable clinical basis for the pathogenesis of MDD and BPD pathogenesis, and enable the identification of validated biomarkers to predict BPD. In addition, subsequent studies that include patients in the manic phase may provide more information.
This is a single-center cohort study with a relatively small sample size. Future studies should further expand the sample size, be multicenter, and control for influencing factors. Even though drug discontinuation was initiated one day before dynamic monitoring of plasma melatonin and cortisol levels and refinement of polysomnography, the patients' blood concentrations were still at high levels, the drug still had an effect on the study results.
Currently, our study only monitored plasma melatonin and cortisol levels at four time points, and we will add a multi-temporal study to further explore the secretion profiles and secretion peaks of melatonin and cortisol in patients with bipolar disorder.
The method of this study is invasive, which will bring a certain degree of psychological burden to the study subjects, and may even affect the results of the patients themselves, their condition and the treatment effect. We will use non-invasive methods such as salivary cortisol collection in the subsequent studies, but this method needs to minimize the interference caused by eating and drinking before collecting the samples. If the subjects can cooperate, we should increase the days number of melatonin and cortisol collection to avoid the results bias caused by the inaccuracy of the test level on a certain day, and collect more dynamic changes of melatonin and cortisol at different time points to obtain the secretion curves and the peak of secretion in patients with bipolar depression, to further compare the differences of BD and MDD in the circadian rhythms adequately.