Actigraphic assessment of delayed sleep–wake rhythm could predict functional impairment in patients with bipolar disorder

Authors: Kazuhiro Kurihara, Futenma Kunihiro, Yoshikazu Takaesu

Abstract

Aim

A delayed sleep–wake rhythm is a common but often under‐recognized characteristic of bipolar disorder (BD). This study aimed to examine the association between delayed sleep–wake rhythm and functional impairment using actigraphy.

Methods

Participants with BD in clinical remission (N = 47) were recruited. Sleep parameters, including the midpoint of sleep (o'clock), total sleep time (minutes), sleep efficiency (%), and wake after sleep onset (minutes), were averaged over a 14‐day period using continuous actigraphy. Functional impairment was assessed using the 12‐item World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0). Correlation analyses were conducted to examine associations between total WHODAS 2.0 scores and sleep parameters. Multiple regression analyses were performed with WHODAS 2.0 total scores as the dependent variable and sleep parameters as independent variables, controlling for mood symptoms and insomnia severity.

Results

Correlation analysis revealed a significant positive relationship between the midpoint of sleep and total WHODAS 2.0 scores (r = 0.424, p = 0.003). Multiple regression analysis identified the midpoint of sleep as a significant predictor of total WHODAS 2.0 scores (β = 0.488, p = 0.005).

Conclusion

Delayed sleep–wake rhythms may contribute to functional impairment in euthymic individuals with BD. These findings suggest that advancing sleep–wake rhythm may contribute to improvements in social functioning. Actigraphy holds potential as a digital biomarker for assessing functional outcomes in this population.

INTRODUCTION

Bipolar disorder (BD) is a recurrent mood disorder characterized by repeated episodes of mood elevation and depression. ^1^ , ^2^ , ^3^ In the management of BD, achieving not only symptomatic improvement in mood symptoms but also recovery from functional impairment, which includes the restoration of daily life and social roles, is considered a critical therapeutic goal. ^4^ , ^5^ , ^6^ However, functional impairment often persists during the euthymic periods, ^7^ , ^8^ , ^9^ and this residual dysfunction may increase the risk of relapse. ^10^ Therefore, residual functional impairment represents a major obstacle to long‐term recovery.

Recently, delayed sleep–wake rhythms have gained attention as potential barriers to functional recovery in patients with BD. These circadian disturbances are commonly observed in patients with BD ^11^ , ^12^ and have been associated with an increased risk of relapse. ^13^ Moreover, disruptions in sleep–wake rhythms play a key role in mood regulation and are closely associated with the core pathophysiology of BD. ^14^ Given these findings, clinicians should carefully consider the impact of delayed sleep–wake rhythms when managing patients with BD as these disturbances may significantly affect long‐term recovery.

However, delayed sleep–wake rhythms are challenging to monitor and frequently under‐recognized in clinical practice. This is partly because more prominent sleep disturbances, such as reduced sleep need during mania and both insomnia and hypersomnia during depression, are more readily recognized in BD. ^15^ , ^16^ Additionally, clinical assessments, such as the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM‐5), tend to prioritize mood symptoms and insomnia, which are clearly defined in diagnostic criteria and frequently self‐reported by patients. ^17^ As a result, delayed sleep–wake rhythms are often unrecognized, and their potential contribution to functional impairment remains insufficiently studied.

In recent years, actigraphy has gained attention as a practical tool for the continuous and objective monitoring of rhythm disturbances. ^18^ , ^19^ Its validity has been well‐established in sleep research on BD, supporting its application in both clinical and research settings. ^20^ , ^21^ , ^22^ , ^23^ Given its capacity to generate continuous, objective data, actigraphy may be particularly valuable for detecting under‐recognized delayed sleep–wake rhythms and providing novel insights into their clinical relevance in BD.

Accordingly, this study aimed to investigate the association between delayed sleep–wake rhythms and functional impairment using actigraphy. By providing objective, continuous data, this approach may help clarify the clinical significance of delayed sleep–wake patterns that are often overlooked in standard assessments.

METHODS

Participants and procedures

The participants in this study were patients with BD (aged 20–75 years) who were recruited from the outpatient clinic of Kyorin University Hospital between April 2019 and March 2020. BD was diagnosed according to the DSM‐5. ^17^ Patients were eligible if they had been in clinical remission for at least 8 weeks before enrollment.

The exclusion criteria were the (1) comorbid severe physical diseases; (2) dementia; (3) alcohol or substance use disorder; (4) shift workers; (5) active suicidal ideation; and (6) other sleep disorders, including sleep apnea syndrome, restless legs syndrome, periodic limb movement disorder, narcolepsy, and related disorders. A total of 47 patients meeting the inclusion criteria were enrolled.

On the first day of the study, participants provided demographic information and completed clinical assessment questionnaires. Each participant wore a wrist actigraphy device continuously for 14 days to collect objective sleep–wake data.

Assessments

Demographic and clinical information collected included age, sex, education, employment, marital status, living situation, family history of psychiatric disorders, age at onset of BD, comorbid physical illnesses, and baseline medications. Medication data included the diazepam equivalent dose of hypnotics (mg), lithium dose (mg), valproate dose (mg), lamotrigine dose (mg), and the chlorpromazine (CP) equivalent dose of antipsychotics (mg).

Functional impairment was assessed using the 12‐item World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0). ^24^ This scale is a standardized self‐report instrument developed by the WHO to evaluate functional difficulties related to physical or mental health conditions. It covers the following six cognition, mobility, self‐care, getting along with others, life activities, and participation in society. Each item is rated on a 5‐point Likert scale ranging from 1 (none) to 5 (extreme or cannot do). Higher total scores reflect greater functional impairment.

Subjective sleep disturbances were assessed using the Japanese version of the Insomnia Severity Index (ISI), a validated 7‐item screening tool commonly used to evaluate insomnia severity. ^25^ , ^26^ Each item is rated on a 5‐point Likert scale ranging from 0 to 4 and assesses domains such as sleep‐onset difficulties, sleep maintenance, satisfaction with sleep, and interference with daily functioning. Total scores range from 0 to 28, with higher scores indicating greater insomnia severity.

Depression severity was evaluated using the Montgomery–Asberg Depression Rating Scale (MADRS). ^27^ The MADRS consists of 10 items, each rated on a 7‐point scale from 0 (no symptoms) to 6 (severe symptoms), yielding a total score ranging from 0 to 60. Higher scores indicate more severe depressive symptoms. The MADRS consists of four subscales. The sadness subscale comprises Item 1 (apparent sadness) and Item 2 (reported sadness); the neurovegetative symptoms subscale includes Item 3 (inner tension), Item 4 (reduced sleep), and Item 5 (reduced appetite); the detachment subscale consists of Item 6 (concentration difficulties), Item 7 (lassitude), and Item 8 (inability to feel); and the negative thoughts subscale comprises Item 9 (pessimistic thoughts) and Item 10 (suicidal thoughts).

The severity of manic symptoms was assessed using Young's Mania Rating Scale (YMRS). ^28^ The YMRS comprises 11 items, with four and seven rated on a scale of 0–8 and 0–4, respectively. Total score ranged from 0 to 60, with higher scores reflecting greater severity of manic symptoms.

Objective sleep parameters were measured using wrist actigraphy (Actiwatch Spectrum; Philips, Best, the Netherlands). Participants were instructed to wear the device continuously for 14 days, except during water‐related activities such as bathing. Sleep parameters included midpoint of sleep (o'clock), total sleep time (min), sleep efficiency (%), and wake after sleep onset (min). Mean values over the 14‐day period were used for analysis.

Statistical analyses

Demographic characteristics, baseline medications, psychological questionnaire responses, and actigraphy‐derived sleep parameters were summarized for the entire sample. Categorical variables are presented as frequencies and percentages, whereas continuous variables are reported as means and standard deviations (SDs). Pearson correlation coefficients were computed to examine associations between total WHODAS 2.0 scores and actigraphy‐based sleep parameters, including midpoint of sleep (o'clock), total sleep time (min), sleep efficiency (%), and wake after sleep onset (min).

Multiple regression analyses were performed with WHODAS 2.0, total scores as the dependent variable. To minimize multicollinearity, predictors with a variance inflation factor (VIF) < 5 were included in the final models. The final models included age, sex, total MADRS score, total YMRS score, total ISI score, sleep midpoint, total sleep time, and sleep efficiency, diazepam equivalent dose of hypnotics, lithium dose, valproate dose, lamotrigine dose, and the CP equivalent dose of antipsychotics. Additionally, we conducted an alternative analysis using the four MADRS subscales instead of the total MADRS score.

RESULTS

Participant characteristics and baseline measures

The clinical characteristics of the participants are summarized in Table 1. The mean age of participants was 45.7 ± 15.1 years, with 21 participants (44.7%) being men. The mean onset age of BD was 33.4 ± 14.2 years. Hypnotic medications included benzodiazepines (n = 8, 17.0%), non‐benzodiazepines (n = 11, 23.4%), ramelteon (n = 10, 21.3%), and suvorexant (n = 4, 8.5%). Mood stabilizers included lithium (n = 25, 53.2%), valproate (n = 7, 14.9%), and lamotrigine (n = 22, 46.8%). Antipsychotic medications included aripiprazole (n = 7, 14.9%), quetiapine (n = 10, 21.3%), extended‐release quetiapine (n = 3, 6.4%), olanzapine (n = 1, 2.1%), perospirone hydrochloride (n = 7, 14.9%), and levomepromazine (n = 1, 2.1%). Additionally, antidepressants (n = 5, 10.6%) and anxiolytics (n = 14, 29.8%) were also used. The mean baseline doses were as diazepam equivalent dose of hypnotics, 15.3 ± 59.4; lithium, 378.7 ± 407.0; valproate, 74.5 ± 205.9; lamotrigine, 79.8 ± 103.0; and CP equivalent of antipsychotics, 107.9 ± 147.8 (Table 1). Regarding comorbid physical conditions, the following were observed among the 47 hypertension (n = 4, 8.5%), dyslipidemia (n = 3, 6.4%), asthma (n = 2, 4.3%), diabetes mellitus (n = 1, 2.1%), essential tremor (n = 1, 2.1%), osteoporosis (n = 1, 2.1%), uterine fibroids (n = 1, 2.1%), glaucoma (n = 1, 2.1%), menopausal disorder (n = 1, 2.1%), temporomandibular disorder (n = 1, 2.1%), atopic dermatitis (n = 1, 2.1%), and allergic rhinitis (n = 1, 2.1%).

Table 2 presents psychological measures and actigraphy‐derived sleep parameters. The mean scores for WHODAS 2.0, MADRS, YMRS, and ISI scores were 21.1 ± 8.1, 5.8 ± 6.2, 1.2 ± 1.8, and 9.6 ± 6.1, respectively. The midpoint of sleep and average total sleep time, sleep efficiency, and wake after sleep onset were 29 ± 3:20, 506.5 ± 108.9 min, 84.4% ± 6.5%, and 54.5 ± 27.6 min, respectively (Table 2).

Correlations between actigraphy‐derived sleep parameters and total WHODAS 2.0 scores

Figure 1 illustrates the correlations between objective measured actigraphy‐derived sleep parameters and functional impairment, as assessed by total WHODAS 2.0 scores. A significant positive correlation was observed between the total WHODAS 2.0 scores and the midpoint of sleep (r = 0.424, p = 0.003). In contrast, total sleep time (r = 0.182, p = 0.221), sleep efficiency (r = –0.043, p = 0.775), and wake after sleep onset (r = 0.116, p = 0.437) were not significantly correlated with WHODAS 2.0 scores (Figure 1).

Multivariable regression analyses of functional impairment

Table 3 presents the results of a multiple regression analysis examining factors associated with total WHODAS 2.0 scores. The predictors included demographic variables (age and sex), clinical scale scores (total MADRS, YMRS, and ISI), actigraphy‐derived sleep parameters (midpoint of sleep, total sleep time, and sleep efficiency), and medication‐related variables (diazepam equivalent dose of hypnotics, lithium dose, valproate dose, lamotrigine dose, and CP equivalent dose of antipsychotics). Among these variables, the total MADRS score (β = 0.482, p = 0.007) and midpoint of sleep (β = 0.488, p = 0.005) were significantly associated with total WHODAS 2.0 scores. The overall model was statistically significant (F[13, 33] = 3.714, p = 0.011), with an R² of 0.515 and an adjusted R² of 0.323 (Table 3).

In contrast, when the four MADRS subscales were entered into the model instead of the total MADRS score, none of the subscales showed significant associations with WHODAS 2.0 sadness (β = 0.084, p = 0.738), neurovegetative symptoms (β = 0.246, p = 0.306), detachment (β = 0.368, p = 0.116), and negative thoughts (β = –0.127, p = 0.585).

DISCUSSION

This is the first study to investigate the association between delayed sleep–wake rhythms, objectively measured using actigraphy, and functional impairment in patients with BD, while controlling for subjective mood and insomnia symptoms. Delayed sleep–wake rhythms were significantly associated with both functional impairment and depressive symptom severity (Table 3). Although relatively few studies have examined the relationship between actigraphy‐derived parameters and functional outcomes, several have reported relevant associations. For instance, one study involving patients with mood disorders reported that a decline in activity levels from day to night, as measured using actigraphy, was associated with functional impairment. ^29^ Other studies have shown that sleep–wake rhythm parameters correlate with brain functions related to working memory in patients with BD. ^30^ However, unlike the current study, these studies either included patients with major depressive disorder (MDD), did not specifically examine rhythm indicators, or focused only on limited aspects of cognitive function. In addition, previous research on functional impairment in BD has primarily concentrated on common subjective symptoms such as depression, ^31^ , ^32^ , ^33^ mania, ^34^ and insomnia. ^35^ , ^36^ , ^37^ , ^38^ In contrast, our study uniquely highlights the impact of objectively measured delayed sleep–wake rhythms on functional outcomes in BD. In summary, the novelty of the present study lies in its specific focus on delayed sleep–wake rhythms and the use of actigraphy as an objective assessment tool.

In recent years, actigraphy has emerged as a valuable tool for the clinical assessment of circadian rhythm sleep–wake disorders. ^39^ Building on this foundation, the present study investigated whether actigraphy‐based measures provide clinically meaningful information beyond subjective symptom reports. Our findings demonstrated that a delayed sleep midpoint, objectively measured using actigraphy, was significantly associated with functional impairment in patients with BD even after adjusting for subjective symptoms (Table 3). These results suggest that actigraphy can provide clinically relevant information. Furthermore, actigraphy offers several practical advantages. It is well suited for both clinical and research settings due to its simplicity and the ability to monitor sleep–wake patterns over extended periods without interfering with daily life activities. ^18^ In addition, actigraphy is a cost‐effective alternative, particularly valuable in contexts requiring long‐term monitoring and strong patient compliance. ^39^ These advantages, along with our findings, underscore the potential of actigraphy not only as a clinically meaningful assessment tool but also as a practical and patient‐friendly option for routine clinical use.

Circadian rhythm dysfunction is increasingly recognized as a core feature of BD. ^40^ However, clinical attention has traditionally focused on sleep quantity and quality, such as reduced sleep needs, insomnia, and hypersomnia. ^15^ , ^16^ In the present study, among the actigraphy‐derived sleep parameters, including total sleep time and sleep efficiency, only the midpoint of sleep was significantly associated with functional impairment. No significant association was found between subjective insomnia symptoms and functional impairment (Figure 1 and Table 3). These findings suggest that sleep–wake rhythm (midpoint of sleep) may play a more clinically important role in BD than sleep quantity or quality. A delayed sleep–wake phase is a commonly observed circadian disturbance in patients with BD, ^12^ and often persists even during euthymic periods. ^41^ Furthermore, a delayed sleep–wake phase is more frequently observed in BD than in MDD. ^11^ Importantly, the presence of a comorbid delayed sleep–wake phase disorder has been shown to predict relapse in patients with euthymic patients with BD. ^13^ Taken together, these findings suggest that the delayed sleep–wake phase may serve as both a trait and a state marker of BD. The present study further reinforces the clinical relevance of delayed sleep–wake rhythms in BD, as emphasized in previous studies.

The severity of depressive symptoms is strongly associated with functional impairment in patients with BD. ^42^ Depressive symptoms have also been significantly linked to impairments across occupational, social, and familial functioning. ^31^ , ^32^ , ^33^ Furthermore, patients with BD often exhibit residual depressive symptoms even during the euthymic phase, which have been associated with reduced resilience, higher risk of relapse, and persistent functional impairment. ^43^ , ^44^ In the present study, subjective depressive symptoms remained significant predictors of functional impairment, even after accounting for actigraphy‐derived sleep parameters (Table 3). These findings underscore the possibility that even mild depressive symptoms may have an impact on functional outcomes, even during the remission phase. To the best of our knowledge, no previous study has simultaneously analyzed both depressive symptoms and actigraphy‐derived sleep parameters in patients with BD. Our findings align with existing literature and further strengthen the evidence supporting a robust association between depressive symptoms and functional impairment in this population. These results highlight the importance of achieving full syndromal remission of depressive symptoms in patients with BD to improve functional outcomes and overall well‐being.

This study has several limitations. First, the relatively small sample size may have limited the statistical power to detect subtle associations. Second, the cross‐sectional design precludes any conclusions about the causal relationships between delayed sleep–wake rhythms, insomnia, and functional impairment. Third, the absence of a healthy control group limits the ability to contextualize findings, as the analysis was restricted to relative differences among patients with BD. Fourth, although basic information on comorbid physical conditions was collected, this study did not assess their severity, duration, or treatment status in detail. These factors may have influenced the findings and should be considered in future research.

CONCLUSION

Delayed sleep–wake rhythms may contribute to functional impairment in patients with BD even during the euthymic phase. These findings suggest that advancing the sleep–wake rhythm may contribute to improvements in social functioning. Actigraphy may serve as a promising digital biomarker for assessing functional impairment in this population.

ANIMAL STUDIES

N/A.

AUTHOR CONTRIBUTIONS

Kazuhiro Kurihara: Writing—original draft; methodology; formal analysis. Futenma Kunihiro: Writing—review and editing; methodology. Yoshikazu Takaesu: Writing—original draft; writing—review and editing; formal analysis; investigation; project administration; funding acquisition.

CONFLICT OF INTEREST STATEMENT

Yoshikazu Takaesu received lecture fees from Takeda Pharmaceutical, Otsuka Pharmaceutical, Daiichi Sankyo Company, Shionogi, Mochida Pharmaceutical, Lundbeck Japan, Eisai, and MSD, and Viatris Pharmaceuticals outside of the submitted work. The other authors declare no competing financial interests or personal relationships that could influence the work reported in this study.

ETHICS APPROVAL STATEMENT

Ethical guidelines were strictly followed throughout this study. All participants provided informed consent before participation. To ensure confidentiality, all data were anonymized and analyzed in an aggregate form. The participants were fully informed about the study's purpose, data protection procedures, and their right to withdraw at any time without penalty. This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Kyorin University (approval 1203).

PATIENT CONSENT STATEMENT

All participants provided written informed consent before participating in the study.

CLINICAL TRIAL REGISTRATION

N/A.