Circadian Phase Instability in Bipolar Disorder: A Neglected Essence

Abstract

Circadian system disruption is an essential but poorly understood feature of bipolar disorder (BD) and the associated comorbidities. Here, current evidence is summarized regarding the emerging concept of circadian phase instability (CPI) as a neglected phenomenon with possibly unique features in BD that could be harnessed to develop individually tailored chronobiological interventions.

Bipolar disorder in context: a multi-system illness

Bipolar disorder (BD) is a multi-system chronic illness characterized by recurrent depressive, manic/hypomanic, or mixed mood episodes [1]. Because of its clinical features and associated comorbidities, it is a leading cause of morbidity and early mortality worldwide. Even with current therapeutics, relapse and subsyndromal symptoms are common [1]. A poor understanding of its neurobiology hampers the development of novel interventions for BD and the associated multi-system comorbidities.

The circadian system is a robust brain-body interaction pathway that is key to organizing most physiological processes [2] (Figure 1–A). Different forms of circadian system disruption are linked to the neurobiology of BD and the associated adverse health outcomes but remain poorly understood [3, 4] (Figure 1–B). Circadian phase instability (CPI) in BD can be defined as an intrinsic susceptibility or inability to maintain a stable circadian phase, which translates into a trait and state mistiming of the circadian system. Arguably, while circadian phase disturbances have been documented in other psychiatric and non-psychiatric illnesses, no other medical condition is currently known to display (spontaneously or triggered) such extreme and recurrent phase shifts (both advances and delays) as a key feature of a symptomatic state. Here, the evidence on CPI in BD is summarized, along with its clinical relevance and rationale for its exploration as a target to develop individually tailored chronobiological interventions.

Figure 1. Expected organization of the circadian system in healthy individuals and individuals with a disrupted circadian system (e.g., circadian phase delay, extreme phase shifting).

In healthy individuals (A), the suprachiasmatic nucleus (SCN) is entrained by temporal cues (e.g., light, nightly melatonin release or administration) and interacts with peripheral oscillators through non-neural (e.g., hormonal axes) and neural pathways. The latter include efferents to other hypothalamic nuclei and brain areas, as well as to the pre-autonomic neurons in the paraventricular nucleus (PVN) and its output to peripheral oscillators via the sympathetic nervous system (SNS) and parasympathetic nervous system (PNS). Likewise, input from the periphery (e.g., hormonal, autonomic and spinal afferent pathways) contributes to entrain the SCN and sustain an optimal organization which translates in circadian rhythms of physiological processes including cardiovascular, metabolic, and immune function. Circadian phase delay or extreme phase shifts are expected to impact this organization through conflict with environmental cues and internal components also known as circadian misalignment. This interferes with SCN output and/or input from the brain and periphery resulting in multi-system desynchrony (decoupling) and dysregulation of physiological processes (B). Figure generated using Biorender (www.biorender.com).

Delineating the concept of CPI in BD

Diagnostic criteria for BD mood episodes focus on “polarity-dependent” sleep disturbances, which are highly prevalent. Mainly insomnia/hypersomnia for depressive episodes, and a decreased need for sleep for mania/hypomanic episodes, which become a target for therapeutic interventions. Remarkably, pronounced shifts in sleep onset time are strong predictors of upcoming mood symptoms and states in BD [5]. This is important because sleep/wake cycles are driven by the circadian system, and emerging evidence suggests that an inability to maintain a stable circadian phase may play a key role in BD with implications beyond sleep disturbances [6, 7].

The characteristics of circadian rhythms in physiological parameters (e.g., phase, MESOR, amplitude) inform on functional aspects of the circadian system. A circadian phase is defined as a time point (e.g., onset, offset, peak) that provides temporal information of a biological rhythm and alignment with others. To date, the most reliable and specific marker of circadian phase (internal time) in humans is called dim light melatonin onset (DLMO) [8]. It consists of sequential measurements of melatonin levels in saliva or blood samples in dim light conditions during the evening/night period. The procedure aims to capture the moment at which the SCN initiates the signal for the release of melatonin from the pineal gland and marks the biological transition from the day into the night period. The timing of this phenomenon is mediated by endogenous melatonin action on melatonin receptors in the SCN.

Circadian phase (i.e., DLMO) and self-reported chronotype (diurnal preference) are highly correlated and fairly stable traits during adulthood [9]. In contrast, CPI represents a critical neurobiological component with recognizable trait and state-dependent features during the course of BD. In support of a CPI trait feature, DLMO has been reported to be delayed among euthymic patients with BD as compared to control individuals [6]. This is consistent with multiple reports showing a fivefold higher prevalence of an evening chronotype, a proxy marker of a trait tendency towards a phase delay, in adults with BD compared to the general population [4]. This is clinically relevant because individuals with BD and evening chronotype show worse illness features including more mood episodes, higher rates of a past suicide attempt, and increased comorbidity with anxiety and substance use disorders. Notably evening chronotype in BD is also associated with increased medical comorbidities like hypertension, migraine, asthma, obstructive sleep apnea, as well as higher rates of binge eating behavior, bulimia nervosa, night binge eating, a higher body mass index, and poor dietary quality [4]. These multi-system adverse health outcomes resemble those observed in models of circadian misalignment [2, 10]. (Figure 2–A)

Figure 2. Circadian phase instability (CPI) in bipolar disorder (BD) as a therapeutic target.

CPI trait and state dependent mistiming of the circadian system due to an intrinsic inability to maintain a stable phase (Top) and putative physiological consequences (Bottom) are illustrated in (A). A higher prevalence of an evening chronotype as proxy measure of a tendency towards phase delay in BD is an example of a CPI trait feature, which is non-specific to BD and has been observed in other psychiatric and medical conditions. CPI also presents as a neurobiological feature that is possibly unique to BD in the form of state-dependent extreme phase shifts, including phase advances (in mania) or delays (prior or during depressive or mixed episodes) illustrated as clocks with a counter clock (advance) and clockwise (delay) arrow direction. Through time, CPI trait and state dependent features could then contribute to chronic circadian misalignment and the gradual appearance or worsening of the associated multi-system adverse health outcomes, including a worse BD course, and increased psychiatric and non-psychiatric comorbidities. Hypothetically, CPI could be explored as a therapeutic target in BD to develop precision medicine chronobiotic interventions as illustrated in (B). For example, dim light melatonin onset (DLMO) could be assessed as a marker of circadian phase during euthymia among individuals with BD and a circadian-related clinical phenotype (e.g., evening chronotype) to induce and maintain (phase lock) a controlled circadian phase shift and evaluate its effect in relapse prevention through an individually timed chronobiotic intervention administered according to phase response curves (Top). It is hypothesized that addressing CPI could contribute to mitigate/prevent circadian misalignment and the associated multi-system adverse health outcomes (Bottom). Figure generated using Biorender (www.biorender.com).

Remarkably, sleep/wake cycles and circadian phase show extreme time shifts (advances and delays) in anticipation to and during a mood episode, a CPI state-dependent feature that is possibly unique to BD [5, 11] (Figure 2–A). To exemplify, one study followed adults with BD across mood episodes and compared measures of circadian phase to a control group [11]. Cortisol and gene expression circadian rhythms were determined via saliva and buccal cells during two consecutive days. During manic episodes, circadian phases were advanced approximately 7 hours. In depressive episodes and mixed episodes, 4 to 5 hour and 6-to-7-hour phase delays were observed, respectively. The circadian phase tended to normalize during euthymia.

Collectively, these observations support CPI as a key feature in BD with clinical implications including worse illness features and increased multi-system comorbidities (Figure 2–A). An intrinsic inability of the SCN to maintain a stable phase and an enhanced susceptibility to light-induced phase delay are putative contributors to phase instability trait and state features, circadian misalignment and associated health outcomes in BD [6]. However, the underlying mechanisms are unknown.

CPI in BD and the development of precision medicine chronobiological interventions

Because of CPI, circadian phase is highly variable across mood states and a moving target for chronotherapy in BD [11]. Neglecting CPI contributes to the suboptimal utilization of chronobiotic (phase shifting) interventions (e.g., melatonergic agents, light therapy), which continue to be administered according to sleep/wake times. This approach may be inappropriate because phase response curve (PRC) studies demonstrate that the chronobiotic effect of such interventions depends on the time of administration relative to the internal time or circadian phase (i.e., DLMO) [12]. For example, exogenous melatonin administered 3 hrs. prior to DLMO may induce a 1.5 hr phase advance (to occur earlier) but would induce a 1.5 hr. delay if administered 12 hr. after DLMO. In this complex scenario, the time of administration of a chronobiotic intervention would require the assessment of circadian phase (e.g. DLMO) for each patient to maximize the intended effect and minimize adverse effects (e.g., spillover effects). However, such an approach has not been applied to any of the studies using chronobiological interventions in BD, nor is it customary to do so in the clinical setting. This key limitation may account for past negative and inconsistent outcomes for acute episodes and relapse prevention and may warrant a reappraisal of chronobiotic interventions in BD [13].

An important step forward will be to characterize CPI during BD. This could be achieved by using DLMO as marker of circadian phase already in use for chronobiotic interventions PRCs. Although challenging, it is technically feasible to assess DLMO across mood states in outpatient and inpatient settings. A baseline DLMO could inform an individually timed chronobiotic intervention as a precision medicine approach to induce and hypothetically maintain (phase lock) a controlled circadian phase shift and assess its therapeutic effect in BD. Preliminary data suggests that exogenous melatonin can advance sleep onset in adults with BD and delayed sleep phase disorder. While limited by a lack of phase assessment, these findings support the exploration of personalized interventions that target CPI guided by clinical phenotyping and ideally with DLMO monitoring [14]. Unfortunately, DLMO assessment continues to be laborious, costly, and result processing may take days or weeks, posing an additional challenge during acute episodes or rapid cycling. Until rapid DLMO determination or better markers of circadian phase are available, CPI could be initially evaluated for efficacy and safety among vulnerable individuals with BD and an evening chronotype sub-phenotype during euthymia. A chronobiotic intervention that follows such personalized approach would inform on its potential to induce and maintain controlled phase shifts aiming to stabilize circadian phase, prevent relapse and mitigate the medical comorbidities (e.g., cardiometabolic) associated with circadian misalignment [15] (Figure 2–B).

Concluding remarks and future perspectives

CPI in BD remains an understudied and poorly characterized phenomenon not targeted by available interventions. A growing body of evidence already supports CPI as a key and possibly unique feature in BD with clinical relevance that may be associated with multi-system comorbidities and warrants immediate attention. Increased knowledge regarding circadian-related sub-phenotypes in BD (e.g., evening chronotype) and the availability of simpler and cost-effective home-based methods to assess DLMO could be harnessed to develop individually tailored chronobiological interventions to induce and maintain controlled phase shifts targeting CPI.

Upon characterization of CPI in BD and its evaluation as a potential therapeutic target, other knowledge gaps will remain. It will be necessary to determine if CPI has a genetic component and/or environmental triggers, if it precedes or follows the onset of mood symptoms and whether it constitutes a “bipolarity” marker or is shared with other psychiatric and medical conditions. In addition, mood stabilizers like lithium have unexplored chronobiotic properties in CPI. Moreover, the study of CPI in BD offspring and at-risk groups is warranted and whether personalized chronobiotic interventions implemented early could mitigate BD course is unknown. A large effort will be required to delineate CPI in BD and its potential as a target for precision medicine interventions.