Introduction
Impulsivity, defined as “a predisposition toward rapid, unplanned responses to internal or external stimuli with disregard for the negative consequences of these reactions,” is a core feature of bipolar disorder (BD).Reference Swann, Dougherty and Pazzaglia1 Individuals with BD exhibit greater impulsivity than those with other affective disorders.Reference Dervic, Garcia-Amador and Sudol2 This heightened impulsivity persists not only during manic episodes but also during remission in BD, suggesting it is both a state feature and a trait feature of BD.Reference Swann, Pazzaglia and Nicholls3 Emerging research indicates that BD may be more accurately conceptualized as a multidimensional condition rather than a categorical diagnosis, characterized by distinct neurobiological underpinnings and a range of comorbidities.Reference Fornaro, Orsolini and Marini4, Reference Sepede, Chiacchiaretta and Gambi5 Elevated impulsivity in patients with BD is linked to a spectrum of adverse outcomes, including cognitive dysfunction, increased risk of substance abuse and suicidal behaviors, reduced medication adherence, and sustained aggression post-remission.Reference Strakowski, Fleck and DelBello6, Reference Muhtadie, Johnson and Carver7 Furthermore, our recent research has identified higher trait impulsivity, associated with altered prefrontal cortex (PFC) structure, in patients with BD with a history of suicide attempts compared to those without such history and healthy individuals.Reference Huang, Kuan and Tu8 Sleep disturbance, a prevalent feature of BD, may be related to heightened impulsivity.Reference Harvey, Schmidt and Scarna9 Both impulsivity and sleep disturbance are associated with symptomatic relapse in BD and can predict the trajectory of BD.Reference Carr, Saunders and Bilderbeck10, Reference Alloy, Urosevic and Abramson11 Poor sleep has been connected to several maladaptive neurocognitive processes that regulate mood and impulsive behavior,Reference Yoo, Gujar and Hu12, Reference Telzer, Fuligni and Lieberman13 underscoring the importance of sleep as a potential mediator and therapeutic target for impulsivity and its detrimental effects. Nonetheless, research into the interplay between these two domains in patients with BD remains scarce.
Impulsivity consists of state impulsivity and trait impulsivity. State impulsivity refers to momentary responses to contextual triggers and is generally assessed with behavioral tasks, while trait impulsivity denotes an individual’s personal characteristics. Studies investigating impulsivity and sleep have mostly enrolled healthy individuals or patients with attentional deficits, with the majority focusing on state impulsivity.Reference Gillett, Watson and Saunders14 Experimental studies have found that sleep deprivation is associated with more impulsive responses in behavioral tasks related to decision-making or response inhibition.Reference Gillett, Watson and Saunders14 McGowan and Coogan observed inverse correlations between sleep duration/efficiency and trait impulsivity, as measured by the Barratt Impulsiveness Scale Version 11 (BIS-11), within an undergraduate student population.Reference McGowan and Coogan15 In summary, extensive research suggests that sleep duration or efficiency negatively correlates with impulsivity in healthy individuals.Reference Gillett, Watson and Saunders14 Despite these significant connections, the study of BD has largely examined sleep and impulsivity as separate entities, and the association between the two features in BD has been scarcely examined. Russo et al. demonstrated associations between disturbed sleep/daytime dysfunction and trait impulsivity among euthymic patients with BD.Reference Russo, Mahon and Shanahan16 However, the data were limited by the exploratory design, which described only bivariate associations between variables without adjusting for important covariates such as age, sex, educational level, and psychotropic medication use.
Recognizing that sleep and impulsivity are critical elements of mood vulnerability in BD, and considering their significant interactions in nonclinical subjects, a deeper comprehension of trait impulsivity and its correlation with sleep disturbance could provide valuable insights into the prognosis and potential therapeutic target of BD. Our objective was to explore the relationship between subjective sleep quality and trait impulsivity in patients with BD. We hypothesized that poor sleep quality would be associated with increased trait impulsivity in euthymic patients with BD.
Methods
Forty-seven consecutive outpatients, aged 20–50 years, diagnosed with BD based on the Mini-International Neuropsychiatric Interview (MINI) and the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), were enrolled from the psychiatric outpatient department of Taipei Veterans General Hospital. Fifty-two patients were screened; 5 were excluded for being over 50 years old. According to DSM-5, bipolar I disorder (BDI) requires at least 1 manic episode for diagnosis, while bipolar II disorder (BDII) requires at least 1 hypomanic episode and 1 major depressive episode. All participants were in a stable mood state, indicated by scores below 7 on both the Montgomery–Asberg Depression Rating Scale (MADRS) and the Young Mania Rating Scale (YMRS).Reference Montgomery and Asberg17, Reference Young, Biggs and Ziegler18 Exclusion criteria included major physical diseases (e.g., head injury with sustained loss of consciousness or cognitive aftermath, neurological disorders) and a lifetime history of schizophrenia, other psychoses, or intellectual disability as defined by the DSM-5. Patients were receiving treatment with various antipsychotics, mood stabilizers, and hypnotics according to the Taiwan consensus on the biological treatment of BD.Reference Cheng, Chang and Lin19 A cohort of 58 healthy participants without any DSM-5 diagnoses, matched for age and sex, was recruited through posters within the clinic and community. These participants underwent thorough clinical evaluations by a psychiatrist to confirm the absence of psychiatric illness. The Institutional Review Board of Taipei Veterans General Hospital approved the study, which adhered to the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants before study inclusion.
Clinical assessment
Demographic characteristics, including age, sex, education, duration of illness, history of substance or alcohol use disorder (according to DSM-5 criteria), and use of psychotropic medications, were collected from all participants. The severity of depression and manic symptoms was evaluated using the MADRS and the YMRS, respectively. The BIS-11 was used to gather self-reported assessments of trait impulsivity through 30 queries, graded on a four-point Likert scale from 1 (rarely/never) to 4 (almost always/always),Reference Patton, Stanford and Barratt20 assessing three impulsivity categories: attentional, motor, and non-planning. Higher scores indicate increased trait impulsivity. Subjective sleep quality over the previous month was assessed using the Pittsburgh Sleep Quality Index (PSQI).Reference Buysse, Reynolds and Monk21 This index includes 19 items covering seven components: subjective sleep quality (1 item), sleep latency (2 items), sleep duration (1 item), habitual sleep efficiency (3 items), sleep disturbance (9 items), use of sleep medication (1 item), and daytime dysfunction (2 items). The sleep disturbance component is derived from the sum of eight item scores reflecting the frequency of sleep-related difficulties. Each component is scored from 0 to 3, indicating the degree of difficulty, with a total score ranging from 0 to 21 for overall sleep quality; higher scores suggest poorer sleep quality. A PSQI score > 5 was used to identify sleep disorders, and a score > 10 was used to determine clinically significant insomnia.Reference Buysse, Reynolds and Monk21
Statistical analyses
Continuous variables were compared using analysis of variance (ANOVA), and categorical data were evaluated with the chi-squared test. The Levene test assessed variance homogeneity among groups. When the variance distribution was uneven, Welch’s 1-way ANOVA was used, followed by Games–Howell post hoc pairwise comparisons. Participants were stratified by their PSQI ratings (≤5, 5–10, > 10), and general linear models (GLMs) assessed impulsivity levels across groups, adjusting for age, sex, education, mood symptoms, substance/alcohol use disorder history, psychotropic medication use, and disease group. Linear regression models examined the association between BIS-11 and PSQI scores, controlling for the same covariates. The main effects of demographic data and clinical symptoms on BIS-11 and PSQI scores were tested, as well as interactions between diagnosis and PSQI scores in relation to BIS-11 scores. The Benjamini–Hochberg false discovery method managed multiple comparisons. A p-value of <0.05 was considered significant. Statistical analyses were conducted using Statistical Package for the Social Sciences (SPSS) 11.5 (SPSS Inc., Chicago, IL, USA).
Results
Compared with controls, patients with BDI or BDII exhibited higher rates of substance or alcohol use disorder history and were more frequently receiving treatment with antipsychotics, mood stabilizers, and hypnotics (all p < 0.005; Table 1). Patients with BD had significantly higher BIS-11 total scores and exhibited greater attention and non-planning impulsiveness than healthy subjects. Additionally, patients with BD experienced poorer sleep quality across global PSQI measures, including sleep latency, sleep disturbance, use of sleeping medications, and daytime dysfunction, relative to healthy controls (all p < 0.005; Table 1). The percentage of patients dissatisfied with their sleep quality (PSQI total score > 5) was 95.5% for BDI, 72% for BDII, and 17.2% for controls.
Table 1. Demographic Data and Clinical Variables between Patients with BD and Healthy Controls
* Independent t-test comparing the BD and HC groups.
+ ANOVA comparing the BDI, BDII, and HC groups.
BDI, bipolar I disorder; BDII, bipolar II disorder; BIS, Barratt Impulsiveness Scale; MADRS = Montgomery–Asberg Depression Rating Scale; PSQI, Pittsburgh Sleep Quality Index; YMRS, Young Mania Rating Scale.
Patients with BD who took hypnotics (n = 18) had higher PSQI total scores and greater daytime dysfunction than those who did not use hypnotics (n = 29) (PSQI total: 11.39 versus 8.72, p = 0.04; daytime dysfunction: 2.00 versus 1.24, p = 0.11). Those who experienced interrupted sleep (question 5b in PSQI) more than once per week (n = 31) exhibited higher attentional impulsiveness compared to patients with less frequent interrupted sleep (less than once per week, n = 16) (18.77 versus 16.20; p = 0.013). Subgroup analyses revealed this association only in patients with BDI (18.73 versus 14.00; p = 0.002), not in those with BDII. Among the eight items assessing sleep disturbance (5b–5i), only interrupted sleep (5b) was linked to attentional impulsiveness.
Linear regression analyses revealed a positive association between the PSQI total score and the BIS-11 total score after adjusting for covariates (Table 2). The second-order attention factor in the BIS-11 was positively correlated with the PSQI total score, sleep disturbances, and daytime dysfunction. However, subanalyses stratified by disease groups did not demonstrate the aforementioned associations between the PSQI and BIS-11 scores. In the total sample, tests for an interaction effect between the PSQI and disease groups revealed no significant interactions (Table 3).
Table 2. Associations between the BIS and PSQI, with the Adjustment of Age, Sex, Education, Substance/Alcohol Use Disorder History, Mood Symptom Severity, Psychotropic Medication Use, and Disease Group
BIS, Barratt Impulsiveness Scale; CI, confidence interval; PSQI, Pittsburgh Sleep Quality Index.
Bold font indicates statistical significance.
Table 3. Correlation of the BIS and PSQI from GLMs with the Adjustment of Age, Sex, Education, Substance/Alcohol Use Disorder History, Mood Symptom Severity, and Psychotropic Medication Use
BDI, bipolar I disorder; BDII, bipolar II disorder; BIS, Barratt Impulsiveness Scale; PSQI, Pittsburgh Sleep Quality Index.
The GLM, adjusted for covariates, revealed that participants with higher PSQI total scores (>10) exhibited elevated BIS-11 total and non-planning scores compared to those with low and medium PSQI scores (≤5, 5–10) (all p < 0.05; Figure 1). The GLM detected no differences in the second factor (motor) between groups.
Figure 1. Comparison of BIS-11 scores among participants of different PSQI score ranges, adjusting for age, sex, education, psychotropic medication use, substance use disorder, mood symptoms, and disease group.
Discussion
The current study found that patients with BD scored higher on the BIS-11 and the PSQI, and subjective sleep quality was negatively associated with trait impulsivity. Given the multifaceted nature of impulsivity, we further explored the contribution of the three BIS-11 subscales to subjective sleep quality. Our findings demonstrated significant correlations between sleep disturbance and daytime dysfunction with attentional impulsiveness, highlighting that the connections between sleep problems and impulsivity primarily stemmed from a specific relationship with the subscale covering difficulties concentrating. We also found that a higher frequency of interrupted sleep (difficulty maintaining sleep) was the only component of the 8 sleep disturbance items to be associated with higher attentional impulsiveness, which includes impaired task focus, intrusive thoughts, and racing thoughts. Despite the non-significant linear association between PSQI scores and the non-planning factor of the BIS-11 after multiple testing corrections, participants with PSQI total scores higher than 10 exhibited greater non-planning impulsiveness than those with low or medium PSQI scores. As patients with BD also scored high on non-planning impulsiveness items, it could be that other aspects of trait impulsivity, as opposed to attentional impulsiveness, are less likely to be directly related to sleep. Additionally, we found that total scores of the BIS-11 and PSQI were positively associated, while subgroup analyses based on different disease categories showed no such association. Our findings suggest the possibility of a shared and transdiagnostic role of sleep-related mechanisms in influencing inhibitory control, which could be independent of the specific pathological underpinnings of distinct mood disorders. The absence of statistical significance after employing multiple testing corrections or in the subgroup analyses might have resulted from the limited sample size within each subgroup. Future studies with larger cohorts are warranted to confirm our findings.
The relationship between sleep and impulsivity continues to be a subject of debate, with the bidirectional hypothesis gaining traction. Experimental studies have demonstrated that sleep deprivation significantly increases neural and autonomic responses to negative emotional stimuli,Reference Yoo, Gujar and Hu12 and disturbed sleep may lead to increased impulsive action.Reference Demos, Hart and Sweet22 Conversely, evidence suggests that poor impulse control or a propensity for risk-taking may contribute to the development of sleep problems.Reference Dahl and Lewin23 Individuals with deficits in self-regulation or higher levels of trait impulsivity and reward-seeking behavior are thought to be inclined to delay bedtime by engaging in activities that provide immediate gratification, such as media consumption or playing video games.Reference Reynolds, Gradisar and Kar24, Reference Kroese, Evers and Adriaanse25 An alternative view suggests that sleep and impulsivity may not have a causal relationship but may instead share a common psychopathological basis. Future research should aim to clarify the relationship between poor sleep quality and impulsivity. Prospective studies could investigate whether improving sleep quality effectively reduces impulsivity. There have been a few longitudinal studies: Serena V. Bauducco found bidirectional links between sleep duration/insomnia and trait impulsivity in a cohort of 2767 young adolescents.Reference Bauducco, Salihovic and Boersma26 In contrast, Van Veen et al. found no longitudinal correlations between sleep quality and impulsivity, whether self-reported or clinician-rated, in 83 male forensic psychiatric patients.Reference Van Veen, Rutters and Spreen27 The interaction between sleep and impulsivity may be particularly significant in individuals with abnormal prefrontal function, which could explain the inconsistencies in previous research findings. Sleep deprivation disrupts the functional connectivity of the PFC, a critical region for inhibitory control, with this effect being more pronounced in those with prefrontal dysfunction.Reference Chai, Whitfield-Gabrieli and Shinn28 Poor sleep quality is correlated with higher impulsivity specifically in individuals with reduced PFC connectivity.Reference Tashjian, Goldenberg and Galvan29 From an evolutionary standpoint, the reduction in inhibitory control during sleep loss may represent an adaptation for increased responsiveness to negative stimuli.Reference Gujar, McDonald and Nishida30
A potential neural explanation exists for the impact of poor sleep on trait impulsivity. The top-down control exerted by the PFC, which mediates inhibitory capabilities, is vulnerable to sleep loss,Reference Chee and Choo31 and functional connectivities between prefrontal networks, the insula, and the limbic system undergo alterations during inadequate sleep.Reference Yoo, Gujar and Hu12, Reference Telzer, Fuligni and Lieberman13 Sleep deprivation-induced impairments in attention and vigilance coincide with reduced metabolic activity in the prefrontal and posterior parietal cortices. These brain regions are also pivotal in the regulation of trait impulsivity.Reference Thomas, Sing and Belenky32–Reference Chuah, Venkatraman and Dinges34
Our results have clinical significance for psychiatric cohorts, as concurrent insomnia may exacerbate trait impulsivity. This is noteworthy because impulsivity is linked to adverse outcomes in BD, such as suicidality and risky behavior.Reference Costanza, Vasileios and Ambrosetti35 Therefore, reducing impulsivity is a critical therapeutic objective, and improving sleep quality could serve as a potential intervention. Chronotherapeutic strategies aimed at increasing the regularity of an individual’s routines, including interpersonal and social rhythm therapy and cognitive behavioral therapy for insomnia, have proven effective in treating BD-related insomnia. These therapies have been shown to delay relapse and decrease suicidal ideation.Reference Sylvia, Janos and Pegg36, Reference Soreca37 However, it remains unclear whether these behavioral interventions can also reduce impulsivity, necessitating further research. Additionally, the potential pathophysiological mechanisms connecting sleep and impulsivity warrant investigation. Specifically, poor sleep quality is associated with maladaptive changes in the hypothalamus–pituitary–adrenal (HPA) axis, which has been implicated in impulsivity.Reference van Dalfsen and Markus38, Reference Stanley, Michel and Galfalvy39 A better understanding of HPA axis responsiveness and its moderating effect on the sleep–impulsivity relationship could inform psychiatric classification and the development of targeted interventions for impulsivity-related undesirable behaviors.Reference Berardelli, Serafini and Cortese40
This study had several limitations. First, sleep and impulsivity indices were evaluated using self-report measures, which are susceptible to memory bias. Incorporating objective measures of sleep and a more comprehensive assessment of behavioral impulsivity would have been beneficial. Actigraphy, with its extended sampling intervals, could have provided estimates of sleep phase, duration, interdaily stability, and amplitude of the circadian rest–activity rhythm. Second, the cross-sectional design limits our ability to determine directionality. Future studies should identify which actigraphic parameters most strongly correlate with psychopathologies in affective disorders, and longitudinal designs could clarify potential intraindividual relationships between actigraphic metrics and trait impulsivity. Additionally, this study only examined the sleep–impulsivity relationship in euthymic patients with BD. The mediating role of affective disturbance in this relationship warrants further investigation. Third, the patients were on psychotropic medications during clinical assessments, which could have influenced the results, as antipsychotics and mood stabilizers can have sedative effects and modify impulsive behaviors. Nevertheless, allowing patients to continue their medications was ethically sound and likely prevented relapse and recurrence; we have accounted for the influence of psychotropic medications in our statistical analyses. Fourth, we used the PSQI, a comprehensive measure of subjective sleep quality that assesses various facets of sleep and associated symptoms, including insomnia, snoring, breathing difficulties, and daytime fatigue. However, we could not determine the specific impact of sleep disorders such as sleep apnea on impulsivity trajectories. Symptoms of particular sleep disorders could inflate cumulative PSQI scores and act as a predisposing factor for increased impulsivity. Finally, chronotype may influence the effects of sleep dysregulation on impulsivity. Previous research has linked later chronotype and eveningness preference with multiple aspects of impulsivityReference Hasler, Wallace and Graves41; eveningness may adversely affect PFC and striatal function, which are implicated in impulsivity.Reference Hasler, Dahl and Holm42, Reference Zhang, Tomasi and Shokri-Kojori43 Future research should explore the relationship between impulsivity and chronotype in BD.
The findings corroborate the hypothesis that poor sleep quality correlates with increased trait impulsivity in individuals with BD, bolstering the growing body of evidence that enhancing sleep quality could be beneficial for those suffering from BD.
Author contribution
Validation: Y.E.C.; Writing – review & editing: Y.E.C., W-C.M., T-P.S.; Data curation: Y-H.K., T-P.S., M-H.H.; Project administration: Y-H.K., T-P.S.; Resources: Y-H.K., T-P.S.; Methodology: W-C.M., M-H.H.; Supervision: W-C.M., T-P.S.; Funding acquisition: T-P.S.; Conceptualization: M-H.H.; Formal analysis: M-H.H.; Investigation: M-H.H.; Writing – original draft: M-H.H.
Funding
The study was supported by grants from the Taiwan Ministry of Science and Technology (NSC 103-2314-B-075-065-MY2, NSC 102-2420-H-010-003, and NSC 103-2420-H-010-001) and Taipei Veterans General Hospital (V104C-128). The funding source had no role in any process of our study.
Competing interest declaration
All authors have no financial relationships relevant to this article to disclose.
Credit authorship contribution statement
Mao-Hsuan Huang conceptualized the study, curated the data, involved in formal analysis, investigated the data, designed the methodology, and wrote the original draft. Yi-Hsuan Kuan curated the data, administered the project, and provided resources. Yee-Lam E Chan wrote, reviewed, and edited the manuscript and validated the data. Wei-Chung Mao designed the methodology; supervised the data; and wrote, reviewed, and edited the manuscript. Tung-Ping Su curated the data; acquired funding; administered the project; provided resources; supervised the data; and wrote, reviewed, and edited the manuscript.
