Introduction
Bipolar disorder (BD) is a chronic and severe illness stemming from the complex interaction between genetic, neurobiological, and environmental factors [Reference Fico1–Reference Oliva3]. Among the neurobiological systems altered in BD, sleep/wake and circadian rhythms are affected, showing a strong overlap with disturbances in energy levels that are central to the disorder [Reference McCarthy4]. Sleep disturbances are part of the diagnostic criteria for BD [5] and structurally contribute to its clinical course and outcome, as they often persist in inter-episodic phases, thus contributing to relapses [Reference Gold and Sylvia6]. They also typically precede the onset of BD [Reference Pancheri7]. Regrettably, insomnia is often neglected as a symptom target in the management of affective disorders [Reference Murru and Sommerhoff8].
People with BD present a substantially increased risk of death by suicide, reportedly 10- to 30-fold higher than in the general population [Reference Schaffer9]. The rate of attempted suicide is also very high, with a lifetime risk of approximately 30%–50% for people with BD [Reference Dong10]. In addition, suicide attempts (SA) are often more lethal in patients with BD than in those with other psychiatric disorders [Reference Baldessarini and Tondo11]. Suicidal thoughts and behaviors are strongly associated with depressive or mixed mood episodes, and with depressive illness onset [Reference Schaffer9, Reference Plans12, Reference Tondo, Vazquez and Baldessarini13]. Other established correlates of suicidality include male gender, younger age, age at illness onset, family history of suicide, previous SA, comorbid personality disorders, anxiety disorders, alcohol and substance use, and worse quality of life [Reference Schaffer9, Reference Liu14–Reference Oliva16].
Notably, sleep disturbances, which are strongly linked to suicidal ideation and behavior in the general population [Reference Pigeon, Titus and Bishop17], also play the aforementioned, significant role in BD exacerbating mood instability, leading to an increased suicide risk.
Until now, the existing literature on this topic appears unclear due to the often heterogeneous definitions of both sleep disturbances and suicidal outcomes, which may encompass a wide and heterogeneous range of phenomena. Also, a lack of control for relevant confounders complicates the overall understanding of the possible relationship between SA, sleep disturbances, and BD.
This systematic review and meta-analysis addresses this gap by evaluating and quantifying the association between sleep disturbances and suicidality in individuals with BD.
Methods
The current SRMA followed the “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” (PRISMA) 2020 guidelines [Reference Page18] and a registered protocol (PROSPERO-ID: CRD42023421381). The PRISMA checklist, the original protocol, and detailed deviations from the original protocol are reported in the Supplementary Materials – Appendices 1 and 2.
Search strategy
The PsycINFO, PubMed, and Scopus databases were systematically searched from inception until May 13, 2024 (search strings are available in Supplementary Materials, Appendix 3). The references of the included articles, books, and other pertinent materials were manually searched and inspected to identify additional original studies that were not captured by the search strings.
Eligibility criteria
The inclusion criteria were original articles that: (a) were published in peer-reviewed journals; (b) included individuals diagnosed with BD according to any edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) [Reference Association19] or the International Statistical Classification of Diseases and Related Health Problems (ICD) [20]; (c) evaluated sleep disturbances in people with suicidality or suicidality in people with sleep disturbances; and (d) reported quantitative data about these association. Studies were eligible for inclusion if they examined sleep disturbances (e.g., insomnia, nightmares, extended sleep onset latency, wakefulness after sleep onset, reduced total sleep time, decreased sleep efficiency, or poor sleep quality) and suicidality (e.g., ideation, planning, attempts, or completed suicide), regardless of the specific definitions used. Only studies including BD patients as both cases and controls (e.g., BD patients with vs. without sleep disturbances or suicidality) were eligible. Both observational (cross-sectional and longitudinal) and interventional studies were eligible for inclusion, but only baseline data were considered in longitudinal and interventional studies in order to minimize the potential confounding effects of treatment, follow-up duration, and time-varying exposures. No language or age restrictions were applied. Studies were excluded if they were as follows: (a) reviews, clinical cases, abstracts, letters to the editor, conference proceedings, or study protocols; and (b) only included non-human samples.
Study selection and data extraction
After excluding irrelevant articles by title and abstract based on previously defined inclusion and exclusion criteria, potentially eligible articles were examined by reading their full texts. Data extraction included, when available: first author, year of publication, geographical region and country, study design, diagnostic criteria employed and (semi)structured interview used, study setting, age group of participants (categorized as children/adolescents, adults, elderly, or mixed), definition of cases (people with sleep disturbances or people with suicidality) and definition of controls (people without sleep disturbances or people with suicidality), mean age of cases and controls, number and percentage of females in both groups, the number and percentage of individuals diagnosed with BD-I among cases and controls, and the percentage of euthymic, depressed, or (hypo)manic, or mixed patients for cases and controls, pharmacological treatment for both cases and controls, the mean duration of illness and SD for cases and controls, and the percentage of BD familiarity among cases and controls, as well as the percentage of suicide familiarity among cases and controls, broad outcome (suicidality in people with sleep disturbances and sleep disturbances in people with suicidality), specific definition of the outcome (as detailed above), information regarding the time frame of suicidality when available, distinguishing between lifetime suicidality and current suicidality, details regarding the assessment method for suicidality (clinical diagnosis, standardized scale, clinical records), details regarding the assessment method for sleep disturbance (self-report, validated scale, or objective measures such as polysomnography (PSG) or actigraphy), number of individuals with and without sleep disturbances or number of individuals with or without suicidality, mean scores and standard deviation (SD) obtained on severity scale for the outcome of interest for cases and controls, statistics that quantify association between outcomes and predictors (correlation coefficients, odds ratios (ORs), or standardized mean differences (SMDs)). When information was not available, we contacted the authors to request relevant data.
Two investigators (CP and MB) conducted all steps described independently. Discrepancies were resolved by consensus with the third author (VO or MDP).
Quality control
Two investigators (CP and MB) independently assessed the Risk of Bias using the “Newcastle-Ottawa Scale” (NOS) [Reference Stang21]. Discrepancies were resolved by consensus with the third author (VO or MDP). The obtained scores were converted according to the standards set by the “Agency for Healthcare Research and Quality” (AHRQ) as previously described [Reference De Prisco22].
Statistical analysis
Statistical analyses were conducted using R version 4.3.1 [23], and separate random-effect meta-analyses (restricted maximum-likelihood estimator) [Reference Harville24] were performed using the metaphor R-package [Reference Viechtbauer25]. SMDs, ORs, and Pearson’s r coefficients with 95% confidence intervals (CI) were used to calculate effect sizes for continuous and dichotomous outcomes and correlations, respectively.
The results were visualized using jungle plots, which display SMDs, ORs, correlation coefficients, and 95% CI for each outcome [Reference De Prisco and Oliva26].
Heterogeneity was evaluated using Cochran’s Q test and I 2 statistic. If heterogeneity was detected (Cochran’s Q p-value < 0.10 or I 2 > 50%), meta-regressions were conducted according to predefined predictors (i.e., mean age, percentage of females, percentage of BD-I, percentage of euthymic patients, percentage of patients with depressive episode, and percentage of patients with (hypo-)manic episode).
Leave-one-out sensitivity analysis, excluding one study at a time from the main analysis, was used to investigate the influence of each study on the overall effect size estimation. Publication bias was examined using funnel plots and Egger’s test [Reference Egger27] when at least 10 studies were available.
Results
Study characteristics
The overall study selection process is shown in the PRISMA flowchart in Figure 1. The literature search identified 828 records, which became 714 after supervised removal of duplicates. Of these, 652 were excluded from the title and abstract screening, and 46 were excluded after reading the full text. Sixteen studies [Reference Aubert28–Reference Sylvia43] fulfilled our inclusion criteria and were included in the quantitative synthesis. A comprehensive list of the excluded studies, with the respective reasons for their exclusion, is provided in the Supplementary Materials – Appendix 4.
Figure 1. PRISMA flowchart, 2020 edition, adapted.
The included studies were published between 2012 and 2022, with six studies from Europe, five from North America, two from Asia, two from Africa, and one multicenter study (i.e., from multiple sites in Europe, Africa, and the Middle East). Fourteen studies used a cross-sectional design, and two studies were longitudinal [Reference Bertrand31, Reference Sylvia43]. The sample sizes across the studies varied from 8 to 16,411, encompassing a total of 19,084 individuals with BD. A comprehensive overview of the included studies is provided in Table 1.
Table 1. Characteristics of the studies included in the systematic review and meta-analysis
Abbreviations: AIS, abbreviated injury scale; BD, bipolar disorder; CDRS-R, Children’s Depression Rating Scale-Revised; C-SSRS, Columbia suicide severity rating scale; DIGS, diagnostic interview for genetic studies; DSM, diagnostic and statistical manual of mental disorders; ESS, Epworth Sleepiness Scale; HAM-D, Hamilton Depression Rating Scale; ICD, International Statistical Classification of Diseases and Related Health Problems; ISI, Insomnia Severity Index; MADRS, Montgomery Asberg Depression Rating Scale; NOS, Newcastle–Ottawa scale; PSG-TST, polisomnography-total sleep time; PSQI - Pittsburgh Sleep Quality Index; SCID, Structured Clinical Interview for DSM Disorders; SSI, Scale for Suicide Ideation; WASH-U-KSADS, Washington University in St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia.
a Results only with the point estimate of the odds ratio.
For a thorough understanding of the scales used to assess suicidality and sleep disturbances included in the quantitative analysis, and additional information such as illness duration, family history of suicide and psychiatric disorders, and current treatment, please refer to Supplementary Materials – Appendices 5 and 6.
Suicidality in patients with BD and sleep disturbances
A total of 11 studies [Reference Aubert28, Reference Bishop32, Reference Esan and Fela-Thomas33, Reference Keskin, Tamam and Ozpoyraz36–Reference Sylvia43] on patients with sleep disturbances were included, comprising a total of 18,420 participants (3,692 cases and 14,728 controls). The overall mean age was 36.25 years (SD = 18.88), and 56.73% of the participants were female.
Three studies evaluated sleep quality using validated scales [Reference Aubert28, Reference Esan and Fela-Thomas33, Reference Keskin, Tamam and Ozpoyraz36]. Three studies assessed general sleep disturbances, with one using a validated scale [Reference Sylvia43] and two relying on clinical assessment [Reference Bishop32, Reference Stubbs42]. Regarding insomnia, two studies used validated scales [Reference Palagini39, Reference Palagini40], while one study assessed insomnia and hypersomnia clinically [Reference Murru38]. Two studies focused exclusively on nightmares, one with a clinical assessment and the other with a validated scale [Reference Marinova37, Reference Stanley41]. Six studies clinically evaluated suicidal ideation or attempts [Reference Aubert28, Reference Esan and Fela-Thomas33, Reference Keskin, Tamam and Ozpoyraz36, Reference Murru38, Reference Stubbs42, Reference Sylvia43], considering both lifetime and current instances. Four studies assessed current suicidality using validated scales [Reference Marinova37, Reference Palagini39–Reference Stanley41]. One study used medical records to assess completed suicides [Reference Bishop32].
According to the NOS scale, four studies were rated as “Good,” five as “Fair,” and two as “Poor.” For more information, please refer to Supplementary Material – Appendix 7.
Sleep disturbances in patients with BD and suicidality
Three studies [Reference Benard29, Reference Bishop32, Reference Fekih-Romdhane34] focusing on BD patients with suicidality (either current or lifetime) were included. The total number of participants was 16,666 (13,412 cases and 3,254 controls). The overall mean age was 48 years (SD = 11.2), with 57% female participants.
Two out of the three studies clinically assessed lifetime suicidality [Reference Benard29, Reference Fekih-Romdhane34]. One study evaluated current suicidal ideation using a validated scale [Reference Fekih-Romdhane34]. Another study consulted medical records for suicide attempts [Reference Bishop32]. Two studies assessed sleep using validated scales [Reference Benard29, Reference Fekih-Romdhane34], one of them used actigraphy data [Reference Benard29], and the third one used clinical assessment without scales [Reference Bishop32].
Two studies were rated as “Good” quality and one as “Fair.” For more information, please refer to Supplementary Material – Appendix 7.
Correlations between sleep disturbances and suicidality
Three studies reported correlation data between sleep disturbances and suicidality [Reference Bernert30, Reference Bertrand31, Reference Hashmi35], with a total sample size of 322 participants. The mean age was 35.98 years (SD = 12.83), and 42.37% were female. Two studies provided objective sleep data using actigraphy and polysomnography [Reference Bernert30, Reference Bertrand31], while suicidality was assessed using validated clinical scales. The quality of the three studies was rated as “Poor.”
Meta-analyses results
The main meta-analytic results are presented in Table 2. Jungle plots of the main results are presented in Figure 2. Forest plots and further details of the leave-one-out sensitivity analyses and meta-regressions are reported in Supplementary Materials – Appendix 8.
Table 2. Results of the meta-analyses in detail
Abbreviations: CIs, confidence intervals; I 2, Higgin and Thompson’s I 2 estimating of the total heterogeneity; PIs, prediction intervals; Q p, p-value for the Cochran’s Q-test of (residual) heterogeneity; OR, odds ratio; SMD, standardized mean difference; τ2, between-study variance.
Significant results are depicted in bold.
Figure 2. Jungle plots summarizing the main effect sizes: (A) Odds ratios (OR), (B) Standardized mean differences (SMD), (C) Correlation coefficients. Black-filled dots represent statistically significant comparisons, while white-filled dots indicate non-significant comparisons. Dot size corresponds to the sample size of each comparison.
Suicidality in patients with BD and sleep disturbances
Lifetime suicidality
Sleep disturbances were associated with significantly higher lifetime suicidality (OR = 1.51, 95% CI = 1.23, 1.86, p-value <0.001, I 2 = 20.1%). The leave-one-out sensitivity analysis did not show a significant influence of single studies on the overall results.
Current suicidality
No significant results were observed (OR = 2.52, 95% CI = 0.93, 6.81, p-value = 0.07, I 2 = 36.1%). The meta-analysis became significant when two studies were excluded in the leave-one-out sensitivity analysis.
Suicidality scores on assessment scales
Patients with sleep disturbances presented significantly higher scores at suicidality assessment scales (SMD = 0.79, 95% CI = 0.53, 1.05, p-value <0.01, I 2 = 0%).
Sleep disturbances in patients with BD and suicidality
Current sleep disturbances
SA were significantly associated with the presence of sleep disturbances (OR = 1.37, 95% CI = 1.2, 1.55, p-value <0.01, I 2 = 0%).
Daytime sleepiness
No significant association was observed between suicide attempts and daytime sleepiness measured by the Epworth scale (ESS; SMD = −0.12, 95% CI = −0.44, 0.21, p-value = 0.49, I 2 = 39.8%).
Sleep quality
Poorer sleep quality, measured by the Pittsburgh Sleep Quality Index (PSQI), was significantly associated with SAs (SMD = 0.42, 95% CI = 0.12, 0.73, p-value <0.01, I 2 = 30.9%).
Correlations between sleep disturbances and suicidality
Association between sleep quality and suicidal ideation
A significant positive correlation was found (r = 0.24, 95% CI = 0.10, 0.36, p-value <0.01, I 2 = 66.67%). Leave-one-out sensitivity analysis did not show a significant influence of single studies on the overall results. Meta-regression analyses indicated a significant dependence on age (β = 0.014, 95% CI = 0.001, 0.028, p-value = 0.05). No significant differences were observed in terms of gender, bipolar disorder diagnosis, or affective state.
Association between total sleep time and suicidal ideation
No significant correlation between suicidal ideation and total hours of sleep in BD was identified (r = 0.06, 95% CI: −0.29, 0.41, p-value = 0.73, I 2 = 58.6%).
Publication bias
Since none of the meta-analyses included a minimum of 10 studies, publication bias could not be assessed [Reference Higgins44].
Discussion
The present SRMA aimed to assess the association between sleep disturbances and suicidality in individuals diagnosed with BD. Overall, we found a significant association between sleep disturbances and suicidality in BD in both directions: individuals with BD who report sleep disturbances (i.e., insomnia, hypersomnia, nightmares) have increased odds of suicidal behavior (lifetime suicidal attempts or suicidal ideation), while those with a history of suicide attempts are more likely to experience sleep disturbances. Additionally, poor sleep quality, as measured with PSQI, positively correlated with suicidal ideation and SAs.
Our findings align with previous SRMAs conducted in populations with unipolar major depression [Reference Wang, Cheng and Xu45], as well as in other psychiatric diagnoses such as schizophrenia, anxiety, panic disorder [Reference Malik46, Reference Rogers, Gresswell and Durrant47], and in the general population [Reference Harris48, Reference Baldini49], where sleep disturbances were consistently linked to suicidality.
In BD, both insomnia and hypersomnia are clinically relevant. While insomnia is often the focus of suicide risk, hypersomnia, a common feature of atypical depression, is frequently underestimated [Reference Barateau50]. Atypical depressive features are more prevalent in BD and have been associated with increased suicide risk [Reference Sánchez-Gistau51]. Additionally, hypersomnia in BD is also linked with increased illness severity, a higher frequency of mood episodes, prolonged depressive or (hypo)manic phases, and psychiatric comorbidities [Reference Murru38]. Furthermore, genetic studies suggest that both suicidality and hypersomnia may share underlying proinflammatory pathways [Reference Kappelmann52]. Thus, despite hypersomnia being perceived as a less worrisome symptom when assessing suicide risk, it still should be routinely evaluated [Reference Yan, Xu and Sun53].
Interestingly, individuals with atypical depression may also exhibit elevated levels of emotion dysregulation [Reference Fornaro54]. Emotion dysregulation is a transdiagnostic construct characterized by difficulty in understanding, accepting, and regulating emotions [Reference De Prisco55, Reference De Prisco56]. It is highly prevalent in BD and correlates with both depressive and (hypo)manic symptoms [Reference Oliva57]. Emotion dysregulation worsens when sleep disturbances are present, increasing impulsivity and elevating the risk of suicide [Reference Harvey58]. More precisely, rumination is the emotion regulation strategy most strongly associated with symptoms of BD [Reference Oliva59], mediating between sleep and suicidality [Reference Holdaway, Luebbe and Becker60]. In line with this, neuroimaging studies support that sleep deprivation disrupts emotion-regulating neural circuits, while heightened emotional arousal negatively affects sleep [Reference Zohar61]. So, emotion dysregulation would act as a state-dependent and sleep-dependent factor increasing the vulnerability to suicidal behavior, particularly in BD, where impulsivity and risk-taking are core and prevalent features [Reference Palagini39, Reference Jiménez62]. In addition to these state-dependent factors, individuals with BD also present stable, trait-like factors (i.e., affective temperaments) contributing to the likelihood of suicide. Temperaments such as irritable, cyclothymic, depressive, and anxious have been associated with impulsivity and mood instability [Reference Konopka and Benzer63–Reference Schrader65]. These characteristics may interact with environmental stressors such as sleep disturbances, exacerbating emotion dysregulation and promoting both self- and hetero-aggressive behaviors.
Along with sleep disturbances, broader circadian rhythm disruption represents another key trait-related framework in BD.
Circadian rhythms are regulated by feedback loops involving the commonly defined “clock genes” [Reference Konopka and Benzer63], which orchestrate not only the sleep–wake cycle but also metabolic and neurophysiological processes [Reference Ozburn64, Reference Schrader65]. Polymorphic variations in clock genes are significantly associated with BD [Reference Chung66, Reference Oliveira67], and linked to greater severity and recurrence of mood episodes [Reference Benedetti68–Reference Murray and Harvey70]. Importantly, several clock genes relate to key biomarkers for the prediction of suicidality [Reference Levey71], such as CLOCK and ARNTL (also known as BMAL1). Animal studies reinforce this connection: in CLOCK mutant rodents, circadian disruption is associated with increased activity in the ventral tegmental area, leading to manic-like behaviors [Reference Beyer and Freund72]. This circadian misalignment alters dopamine dysregulation, representing a possible neurobiological substrate for suicidality in BD [Reference Alloy, Nusslock and Boland73].
Although our SRMA did not reveal significant differences between BD-I and BD-II in terms of sleep disturbances and suicidality, prior literature suggests that they may differ in relevant clinical features, including patterns of sleep dysregulation and risk profiles for suicidal behavior. In BD-I, suicide risk has been more consistently associated with insomnia, particularly difficulties initiating sleep and associated daytime impairments such as anhedonia [Reference Goldschmied74]. Some studies also report higher rates of hypersomnia in BD-I, which may reflect a real clinical feature, measurement limitations in detecting hypersomnia, or treatment-related aspects [Reference Kaplan and Williams75]. Conversely, in BD-II, suicide risk appears more strongly linked to evening chronotype, emotional dysregulation, childhood trauma, and low resilience [Reference Romo-Nava76]. This pattern aligns with the circadian vulnerability model, in which eveningness constitutes a transdiagnostic marker of emotional instability [Reference McClung77]. Finally, the presence of mixed features across BD subtypes is consistently associated with more severe clinical outcomes, including rapid cycling, substance use comorbidity, and elevated suicide risk [Reference Tondo, Vazquez and Baldessarini13].
Another aspect to take into account in the relationship between sleep and suicidality in BD is the potential moderating role of age. Our meta-regression analysis indicated that the association between poor sleep quality and suicidality becomes stronger with increasing age. This pattern aligns with prior longitudinal and clinical studies emphasizing the relevance of sleep disturbances in suicide risk among older adults [Reference Bernert78], possibly due to the cumulative burden of chronic sleep disruption, comorbid medical conditions, or reduced resilience.
According to our results, sleep contributes to shaping the clinical cascade leading to suicidal behavior in BD, so that their early detection is crucial. Clinicians must not underestimate any complaints and should always thoroughly explore sleep patterns, ideally with validated scales. Prospective, real-time monitoring of sleep, circadian rhythms has the potential to enhance standard clinical care in the near future. The use of technology and wearable devices alongside ecological momentary assessment tools (e.g., mobile-based self-report systems) offers a promising approach to accurately catch sleep, energy, and mood fluctuations [Reference Anmella79].
Although treatment considerations fall outside the scope of this review, the clinical relevance of managing sleep disturbances in the context of suicidality in BD warrants some consideration. Lithium remains the most robust pharmacological agent with dual effects on sleep regulation and suicide prevention [Reference Yan, Xu and Sun53, Reference Lewitzka80]. A syndrome-based pharmacological approach, tackling both sleep and mood fluctuations, seems the most rational approach to sleep alterations in BD, so that the same considerations in the general management of BD are due, such as caution in the use of antidepressants and the unwarranted use of benzodiazepines for chronic insomnia [Reference Vieta2, Reference De Crescenzo81, Reference Riemann82]. Non-pharmacological interventions such as sleep hygiene education, light therapy, and cognitive-behavioral therapy for insomnia (CBT-I) may complement standard treatment in BD [Reference Hertenstein83, Reference Harvey84].
Limitations
The present SRMA is the first to examine the relation between sleep disturbances and suicidality in individuals with BD. Our results must be considered in the light of some limitations.
The most important limitation is the overall scarcity of prospective studies on the topic, and only two longitudinal studies were included, so that while this SRMA univocally supports an association between sleep disturbances and suicidality in BD, the directionality of this relationship remains unclear. Undoubtedly, sleep disturbances act as proximal triggers anticipating suicidal behaviors, but they might also arise as downstream effects of depressive symptoms or other clinical aspects.
These limitations reflect a technological barrier that digital tools for ecological and continuous monitoring will hopefully help to bypass [Reference Lipschitz85]. Furthermore, the evidence reviewed in this study could not differentiate between acute and euthymic phases of the disorder or finer aspects in the clinical course, such as the predominant polarity of relapses. Also, most of the studies lack the control for comorbid somatic, psychiatric, and substance use confounding conditions, which are all well-known factors associated with sleep disturbances and suicidality in general populations and in individuals with BD [Reference McIntyre86–Reference Oliva88]. Comorbid sleep disorders are especially relevant, unconsidered, and underdiagnosed in psychiatric populations [Reference Abad and Guilleminault89]. Additionally, the majority of the patients in the included studies were on medication, and its effects on sleep and suicidality may be a confounder [Reference Ilzarbe and Vieta90]. Lastly, the lack of ethnic data restricts the generalizability of our findings, as the relationship between sleep disturbances and suicidal behavior in BD may vary across different ethnic and environmental contexts, as suggestive evidence exists on the association of high temperature and suicidal behaviors [Reference Radua91].
Future research should aim to improve methodological rigor – particularly through the use of more prospective designs – and incorporate ecological, continuous assessments of sleep and activity patterns, which might allow for differentiating the broad definition of sleep disturbances into a meaningful and clinically relevant stratification [Reference Anmella92–Reference Oliva94]. This temporal distinction is clinically meaningful, as it informs whether sleep interventions may serve a preventive versus palliative function in suicide risk management. Integrating this bidirectional model into clinical frameworks could help both refine and redefine screening and treatment priorities.
Conclusions
This systematic review and meta-analysis found a significant association between sleep disturbances and suicidality in individuals with bipolar disorder, with the most consistent relationship observed between poor sleep quality and suicidal ideation. Future studies using standardized, multidimensional assessments of sleep and prospective designs are needed to clarify the temporal and physiopathological links between sleep alterations and suicidality in bipolar disorder, and digital innovation will likely allow us to fill this gap both in research and clinical practice. Nonetheless, integrating structured sleep assessment into routine care may offer clinicians a practical and accessible opportunity to improve monitoring, guide timely interventions, and ultimately enhance the safety and well-being of individuals living with bipolar disorder.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1192/j.eurpsy.2025.10102.
Data availability statement
Requests to see any data that are not included in the article or the appendix should be directed to the corresponding author.
Acknowledgments
MB thanks the support of a Marató-TV3 Foundation grant 202230-31.
VO is supported by a Rio Hortega 2024 grant (CM24/00143) from the Spanish Ministry of Science, Innovation and Universities, financed by the Instituto de Salud Carlos III (ISCIII) and co-financed by the Fondo Social Europeo Plus (FSE+).
MDP is supported by the Translational Research Programme for Brain Disorders, IDIBAPS.
AM thanks the support of the Spanish Ministry of Science and Innovation (PI19/00672, PI22/00840), integrated into the Plan Nacional de I + D + I and co-financed by the ISCIII-Subdirección General de Evaluación and the Fondo Europeo de Desarrollo Regional (FEDER) and from La Marató-TV3 Foundation grants 202230-31.
EV thanks the support of the Spanish Ministry of Science, Innovation and Universities (PI21/00787, PI24/00432) integrated into the Plan Nacional de I + D + I and co-financed by the Instituto de Salud Carlos III -Subdirección General de Evaluación and the Fondo Europeo de Desarrollo Regional (FEDER); the Generalitat de Catalunya and Secretaria d’Universitats i Recerca del Departament d’Economia i Coneixement (2021 SGR 01128), CERCA Programme, Generalitat de Catalunya; La Marató-TV3 Foundation grants 202234-30; the European Union Horizon 2020 research and innovation program (H2020-EU.3.1.1. - Understanding health, wellbeing and disease, H2020-EU.3.1.3. Treating and managing disease: Grant 945151, HORIZON.2.1.1 - Health throughout the Life Course: Grant 101057454 and EIT Health (EDIT-B project).
Financial support
This research received grants from the Spanish Ministry of Science and Innovation (PI22/00840), integrated into the Plan Nacional de I + D + I and co-financed by the ISCIII-Subdirección General de Evaluación and the Fondo Europeo de Desarrollo Regional (FEDER).
Competing interests
GF has received CME-related honoraria or consulting fees from Angelini, Janssen-Cilag, and Lundbeck.
AM has received grants and served as a consultant, advisor, or CME speaker for the following entities: Angelini, Idorsia, Lundbeck, Pfizer, Takeda, outside of the submitted work.
EV has received grants and served as a consultant, advisor, or CME speaker for the following entities: AB-Biotics, AbbVie, Angelini, Biogen, Biohaven, Boehringer-Ingelheim, Celon Pharma, Compass, Dainippon Sumitomo Pharma, Ethypharm, Ferrer, Gedeon Richter, GH Research, Glaxo-Smith Kline, Idorsia, Janssen, Lundbeck, Medincell, Neuraxpharm, Newron, Novartis, Orion Corporation, Organon, Otsuka, Rovi, Sage, Sanofi-Aventis, Sunovion, Takeda, Teva, and Viatris, outside the submitted work.
All the other authors have no conflicts to declare.
Open access
Comments
No Comments have been published for this article.