Abnormal hypothalamic functional connectivity and serum arousal-promoting neurotransmitters in insomnia disorder patients: a pilot study

Introduction

Insomnia disorder (ID) is a common sleep disorder characterized by challenges in falling asleep or staying asleep, even with sufficient time and proper conditions for sleep, resulting in decreased daytime performance and quality of life (Krystal, Ashbrook & Prather, 2021; Perlis et al., 2022). It impacts a substantial portion of the global population, with prevalence estimates varying from 10% to 30% in the general population (Chung et al., 2015). However, the exact causes of ID are complex and not fully understood. Various hypotheses have been proposed by scholars to explain ID, with particular attention given to the hyperarousal hypothesis (Riemann et al., 2010).

Recent advancements in neuroimaging techniques have offered valuable insights into the neural correlates of ID. Specifically, functional magnetic resonance imaging (fMRI) studies have revealed changes in brain activity and functional connectivity (FC) linked to ID, emphasizing abnormal functioning in regions important for sleep-wake regulation (Kay & Buysse, 2017; Khazaie et al., 2017). Previous studies have utilized regions of interest (ROI) including the amygdala (Huang et al., 2012), the anterior cingulate cortex (ACC) (Wang et al., 2023), the thalamus (Li et al., 2019), and the default mode network (DMN) (Aquino et al., 2024) to investigate connectivity patterns in ID, revealing altered FC in various brain regions of ID patients; however, conclusions were inconsistent (Fasiello et al., 2022; Aquino et al., 2024).

The hypothalamus plays a critical role in regulating the sleep-wake cycle (Saper, Chou & Scammell, 2001; Saper, Scammell & Lu, 2005; Szymusiak & McGinty, 2008). Human observational studies on the anatomical basis of sleep-wake regulation indicate that the posterior hypothalamus (PH) promotes arousal while the anterior hypothalamus (AH) facilitates sleep (Boes et al., 2018). Early animal studies on sleep-wake regulation pathways have identified the PH nucleus as a major component of the ascending arousal system promoting wakefulness. Meanwhile, the ventrolateral preoptic nucleus within the AH has emerged as a key player in promoting sleep through inhibitory actions (Köhler et al., 1986; Ko et al., 2003; Takahashi, Lin & Sakai, 2009). Lesions in the PH can trigger narcolepsy episodes, whereas lesions in the AH can result in insomnia symptoms, thus supporting the role of the hypothalamus in sleep-wake regulation (Swett & Hobson, 1968; Lu et al., 2000). Additionally, AH and PH exhibit negatively correlated state-dependent discharge patterns, with the PH predominantly discharging during wakefulness, while the AH shows specific activation during non-rapid eye movement (NREM) sleep (Steininger et al., 1999; De Luca et al., 2022).

The precise localization of the human ROI within the hypothalamus remains unclear. Boes et al. (2018) utilized fMRI to examine 98 healthy participants, delineating a negative correlation between the anterior and posterior regions of the hypothalamus. The ROIs within the hypothalamus were selected by two anatomists with expertise in hypothalamic anatomy, utilizing a standard Montreal Neurological Institute (MNI) space high-resolution T1-weighted MRI template (Fonov et al., 2011). Furthermore, a recent study refined the coordinates of anterior and posterior regions of the hypothalamus during wakefulness and NREM sleep by simultaneously collecting electroencephalography (EEG) and fMRI data from healthy participants to investigate brain functionality (Jiang et al., 2021). Currently, there are relatively few studies using MRI to examine structure and function of hypothalamus in ID patients. Ding et al. (2021) found higher FC between the left hypothalamus and other brain regions, including the medial prefrontal cortex and pallidum in ID patients compared with healthy control (HC). Additionally, ID patients showed increased volumes in the specific subunits of the hypothalamus when compared to HC (Luo et al., 2024). Therefore, it is currently crucial to investigate the FC in the AH and PH of ID patients, as these regions are associated with sleep-promoting and arousal-promoting functions, respectively.

Orexin-A and histamine are neurotransmitters known to promote wakefulness. Previous studies have explored the role of these neurotransmitters together in relation to the pathophysiology of diseases such as Alzheimer’s disease and narcolepsy (Shan, Dauvilliers & Siegel, 2015; Gabelle et al., 2017). Orexin-A primarily exerts excitatory effects on sleep-wake cycles, typically coupling with intracellular Gq signaling proteins (Hoang et al., 2003; Karteris et al., 2005). Histamine, another key arousal-promoting neurotransmitter, acts via H1 and/or H3 receptors and crucially regulates sleep-wake cycles (Thakkar, 2011). Administration of histamine or H1 receptor agonists induces wakefulness, while H1 receptor antagonists promote sleep. Anatomically, histamine neurons and orexin neurons partially overlap in the PH (Anaclet et al., 2009). Histamine neurons located in the tuberomammillary nucleus (TMN) and surrounding areas exhibit increased firing during wakefulness (Steininger et al., 1999), whereas orexin neurons in the perifornical area discharge during both wakefulness and rapid eye movement (REM) sleep (Mileykovskiy, Kiyashchenko & Siegel, 2005).

Based on the hyperarousal theory and the role of the hypothalamus in sleep regulation, we hypothesized that ID patients would exhibit altered connectivity patterns between the hypothalamus and other sleep-related regions, accompanied by elevated concentrations of arousal-promoting neurotransmitters. Therefore, using seed-based FC analysis, this study aimed to investigate the resting-state FC of the AH and PH within the whole brain in ID patients. Additionally, we measured serum levels of orexin-A and histamine to explore their relationship with the FC of the PH within the whole brain in ID, as these arousal-promoting neurotransmitters are primarily secreted by neurons in the PH.

Materials and Methods

Results

Demographic and clinical characteristics

In this study, 30 ID patients and 30 HCs were recruited (Fig. 1). There were no significant differences between the ID and HC groups in terms of age, gender, BMI, or some objective sleep parameters derived from PSG (SOL, REM_L, N1, N2, N3 duration, N1_perc, N2_perc, N3_perc, R_perc, ArI_NREM, ArI_total) (P > 0.05). However, significant differences were observed between the ID and HC groups in other PSG parameters (TST, SE, SOL, Wake duration, R duration, ArI_REM) (P < 0.05). Significant differences were also found between the ID and HC groups in sleep scales (ISI, PSQI), emotional scales (HAMA, HAMD), and cognitive scale MOCA (P < 0.05) (Table 1). These findings align closely with the clinical features of sleep onset and maintenance difficulties, as well as emotional and cognitive aspects in ID patients reported in previous studies (Baglioni et al., 2014; Künstler et al., 2023).

Table 1:

Demographic and clinical characteristics of participants with ID and HC groups.

Characteristics	ID (N = 30)	HC (N = 30)	χ2/Z/t	P
Age (year)	32.00 (19.00)	27.70 (6.55)	−1.591a	0.112
Gender, male/female	10/20	10/20	0.000b	1.000
BMI (kg/m2)	20.13 (3.89)	21.58 (3.93)	−1.567a	0.117
ISI	15.00 (7.00)	2.00 (4.75)	−6.104a	<0.001
PSQI	11.00 (5.00)	4.00 (2.00)	−5.925a	<0.001
HAMA	4.00 (4.00)	1.00 (2.00)	−4.710a	<0.001
HAMD	5.00 (4.00)	0.50 (3.00)	−5.161a	<0.001
MOCA	27.00 (2.00)	28.00 (3.00)	−2.394a	0.017
TST (min)	355.48 ± 91.08	413.85 ± 70.23	2.78c	0.007
SE (%)	0.78 (0.17)	0.88 (0.12)	−2.602a	0.009
WASO_time (min)	68.50 (80.00)	42.75 (52.40)	−2.499a	0.012
WASO_number	26.00 (11.00)	21.50 (19.00)	−2.153a	0.031
SOL (min)	16.50 (19.50)	9.50 (20.10)	−1.863a	0.062
REM_L (min)	115.00 (84.50)	94.75 (53.50)	−0.577a	0.564
Wake duration (min)	107 (113.00)	57.75 (67.10)	−2.476a	0.013
N1 duration (min)	44.00 (34.50)	37.75 (20.00)	−0.458a	0.647
N1_perc (%)	11.80 (9.50)	9.05 (8.40)	−1.582a	0.114
N2 duration (min)	167.97 ± 50.78	189.83 ± 40.47	1.844c	0.070
N2_perc (%)	45.33 ± 9.73	46.19 ± 7.88	0.376c	0.708
N3 duration (min)	68.11 ± 33.43	84.82 ± 39.33	1.773c	0.081
N3_perc (%)	19.07 ± 9.01	20.44 ± 8.23	0.615c	0.541
R duration (min)	75.00 ± 27.43	94.57 ± 36.48	2.348c	0.022
R_perc (%)	20.22 ± 5.88	22.25 ± 6.68	1.251c	0.216
ArI_REM	1.20 (3.20)	0.70 (1.50)	−1.972a	0.049
ArI_NREM	5.90 (8.60)	3.75 (5.30)	−0.904a	0.366
ArI_total	9.30 (5.40)	6.40 (7.50)	−1.390a	0.165

DOI: 10.7717/peerj.18540/table-1

Notes:

^aMann Whitney U-test;^bChi-square test; ^cStudent t-test. Numbers in bold represent P < 0.05. Normally distributed data are presented as mean ± standard deviation, non-normally distributed data as median (interquartile range).

ID, insomnia disorder; HC, healthy control; BMI, body mass index; ISI, Insomnia Severity index; PSQI, Pittsburgh Sleep Quality Index; HAMA, Hamilton Anxiety Scale; HAMD, Hamilton Depression Rating Scale; REM, Rapid Eye Movement; NREM, Non-Rapid Eye Movement; TST, Total Sleep Time; SE, Sleep Efficiency; WASO, Wake After Sleep Onset; SOL, Sleep Onset Latency; REM_L, REM Latency; N1, NREM sleep stage 1; N2, NREM sleep stage 2; N3, NREM sleep stage 3; N1_perc, N2_perc, N3_perc, R_perc, Percentage of stage N1, N2, N3, REM sleep; ArI, Arousal index.

Alterations of FC in ID patients

Compared to the HCs, PH-based FC analysis revealed decreased FC between the PH and several brain regions, including the bilateral orbital superior frontal gyrus, the bilateral angular gyrus, the right ACC, the left precuneus, and the right medial superior frontal gyrus in ID patients (Table 2 and Fig. 2). Additionally, the AH-based FC analysis showed decreased FC between the AH and the right ACC in ID patients (Table 3 and Fig. 3). The significance threshold was set at a voxel-level P < 0.001 and a cluster-level P < 0.05 (Bonferroni corrected).

Table 2:

Brain regions showing reduced functional connectivity with the PH in ID patients compared to HCs.

Brain regions	MNI coordinates			Cluster size	T-value
	X	Y	Z
Frontal_Sup_Orb_R	18	60	−6	59	4.32
Frontal_Sup_Orb_L	−12	60	9	85	4.38
Cingulum_Ant_R	9	36	6	58	4.55
Precuneus_L	−6	−48	21	440	4.87
Angular_L	−36	−72	39	148	5.18
Angular_R	42	−57	30	137	4.75
Frontal_Sup_Medial_R	15	27	48	91	4.76

DOI: 10.7717/peerj.18540/table-2

Notes:

R represents right and L represents left side of the brain.

ID, insomnia disorder; HC, healthy control; PH, posterior hypothalamus; Frontal_Sup_Orb, Orbital superior frontal gyrus; Cingulum_Ant, Anterior cingulate cortex; Angular, Angular gyrus; Frontal_Sup_Medial, Medial superior frontal gyrus.

Table 3:

Brain regions showing reduced functional connectivity of the AH in ID patients compared to HCs.

Brain regions	MNI coordinates			Cluster size	T-value
	X	Y	Z
Cingulum_Ant_R	15	27	48	91	4.79

DOI: 10.7717/peerj.18540/table-3

Notes:

R represents right side of the brain.

ID, insomnia disorder; HC, healthy control; AH, anterior hypothalamus; Cingulum_Ant, Anterior cingulate cortex.

Correlation between clinical characteristics and FC values in ID patients

Correlation analyses were performed between clinical characteristics and FC values of AH and PH with brain regions exhibiting significant differences in ID patients. The results revealed that in ID patients, the FC values of the PH with left precuneus were positively correlated with R duration (P < 0.05), N3 duration (P < 0.01), N3_perc (P < 0.05), while negatively correlated with N1 duration (P < 0.05). The FC values of PH with the left angular gyrus showed positive correlation with R duration (P < 0.01), R_perc (P < 0.01) in ID patients. The FC values of the PH with the left orbital superior frontal gyrus showed a negative correlation with N1 duration (P < 0.05) in ID patients. The FC values of the PH with the right ACC demonstrated positive correlations with N1 duration, N1_perc, N3 duration, and N3_perc (P < 0.01), while a negative correlation with N2_perc (P < 0.05) in ID patients. However, the FC value of AH was not observed to correlate with clinical characteristics in ID patients. The absolute values of the correlation coefficients (|r|) for the above associations were all greater than 0.48 (Fig. 4).

Changes in the levels of serum arousal-promoting neurotransmitters in ID patients

Among the recruited participants, 25 ID patients and 26 HCs provided eligible blood samples for analysis (demographic information was shown in Table S1). We measured serum orexin-A and histamine concentrations using ELISA. Serum orexin-A (12.30 [5.59] pg/mL in ID, 10.87 [1.21] pg/mL in HC, Z = 4.315, P < 0.001) and histamine (19.59 [2.05] µg/L in ID, 16.52 [1.80] µg/L in HC, Z = 5.634, P < 0.001) concentrations in ID patients were significantly elevated compared to HCs (Fig. 5).

Correlation between FC of PH and arousal-promoting neurotransmitters

Given that arousal-promoting neurotransmitters (orexin-A and histamine) are predominantly secreted by neurons in the PH, we only conducted correlation analysis between FC values of PH with brain regions exhibiting significant differences and the levels of arousal-promoting neurotransmitters in ID patients. The results revealed a positive correlation between the FC of the PH and right medial superior frontal gyrus and histamine in ID patients (R = 0.41, P = 0.04) (Fig. 6).

Discussion

In the present study, seed-based FC analysis was applied to investigate the resting-state FC between the AH/PH and the whole brain in ID patients. Additionally, we measured serum concentrations of orexin-A and histamine to explore the relationship between FC and arousal-promoting neurotransmitters in ID patients. The results revealed altered FC of the hypothalamus in ID patients, which correlated with the levels of arousal-promoting neurotransmitters. To our knowledge, this is the first study to investigate the FC of AH/PH with the whole brain and to examine the correlation between PH and arousal-promoting neurotransmitters in ID patients, which indicated shared alterations in hypothalamic function and humoral factors in ID.

Our study found decreased FC between the PH and brain regions associated with the DMN and ACC in ID patients. The PH is a primary brain region where arousal-promoting neurons are located, and the changes of connectivity within PH may reflect alterations in arousal mechanisms. The DMN is an interconnected brain system that is most active during undisturbed solitary thinking, involving a wide range of advanced cognitive functions (Regen et al., 2016). ID patients show significantly decreased FC between the anterior DMN and posterior DMN (Huang et al., 2017; Dong et al., 2018; Yu et al., 2018); however, several studies have also reported increased FC between subregions of the DMN (Ma et al., 2018; Zhou et al., 2018; Leerssen et al., 2019). Changes in FC within the DMN may explain the sensitivity and alterations in self-awareness observed in ID patients, which could potentially affect the process of falling asleep (Morin et al., 2015). This study identified reduced FC between the PH and specific DMN brain regions (the bilateral angular gyrus, the left precuneus). We speculate that ID may be linked to a failure in arousal mechanisms, hindering the transition from wakefulness to sleep states. Additionally, this study found significantly reduced FC between the ACC and both the anterior and posterior hypothalamus, suggesting that ACC may be a key brain region involved in the onset and development of ID. Two previous studies revealed alterations in FC between ACC and brain regions such as the thalamus and orbitofrontal cortex in ID patients (Yan et al., 2018; Li et al., 2018). Additionally, correlation analysis in our study showed that the FC of PH and the precuneus, angular gyrus, orbital superior frontal gyrus, ACC were correlated with sleep stages and their latency periods, suggesting the function of the hypothalamus is associated with sleep quality and architecture. Similarly, Guo et al. (2023) also investigated the relationship between FC and sleep stages in ID patients, finding a positive correlation between transition probabilities from N2 to wakefulness and the FC of the ACC with the anterior cerebellum during N2. However, research in this area is still limited, and further studies are required.

Elevated serum levels of orexin-A and histamine were found in ID patients in this study. Our findings are consistent with that of Gabelle et al. (2017) who reported significantly higher levels of orexin-A in the peripheral blood of ID patients compared to HCs. However, the comparison of peripheral blood histamine levels in ID patients has not been previously reported. Both histamine and orexin-A are involved in regulating brain functions, including the sleep-wake cycle, energy regulation, and endocrine homeostasis (Shan, Dauvilliers & Siegel, 2015). Histaminergic neurons are innervated by orexinergic neurons, both located in the PH, playing crucial roles in regulating the sleep-wake cycle (Yamanaka et al., 2002; Saito et al., 2018). The levels of histamine in the brain are related to alertness and the circadian rhythm, with histamine release increasing during wakefulness (Yoshikawa, Nakamura & Yanai, 2021). The activation of orexin neurons promotes arousal and maintains wakefulness, whereas decreased orexin activity induces sleep (Taheri, Zeitzer & Mignot, 2002). While measuring neurotransmitter concentrations in the CSF is ideal, obtaining CSF from individuals can be challenging (Chien et al., 2015). Strawn et al. (2010) demonstrated a strong correlation between levels of orexin-A in CSF and peripheral blood samples from individual subjects. Therefore, this study collected serum samples rather than CSF samples. The correlation between neurotransmitters (histamine and orexin-A) in the brain and peripheral blood has not been fully clarified. In normal conditions, histamine is typically unable to pass through the blood-brain barrier (BBB) (Thakkar, 2011). However, histamine can increase BBB permeability by opening the interendothelial cell tight junctions (Sakurai et al., 2009; Sedeyn et al., 2015). Moreover, histamine in the blood can affect hypothalamic activity through circumventricular organs that lack the BBB (Alstadhaug, 2014). Orexin-A efficiently cross the BBB through simple diffusion to reach the brain tissue (Kastin & Akerstrom, 1999). Previous studies have indicated a positive correlation between wakefulness and extracellular histamine concentration in the hypothalamus (Rozov et al., 2014). Therefore, the elevation of peripheral arousal-promoting neurotransmitters may be one of the neural mechanisms underlying the occurrence of ID.

Furthermore, we conducted a correlation analysis between arousal-promoting neurotransmitters and FC values of the PH, revealing a positive correlation between the FC of the PH and right medial superior frontal gyrus and histamine in ID patients. Stimuli from peripheral tissues can be transmitted to the brain, and changes in brain function are reflected in peripheral immune responses (Cacabelos et al., 2016). In this neuroimmune regulation, histamine serves as a significant pleiotropic factor (Smolinska et al., 2014). Additionally, histamine neurons are found solely within the TMN of the PH (Scammell et al., 2018), which may explain the changes in FC of PH linked to histamine. This result indicated a combined influence of arousal-promoting neurotransmitters and brain function, providing additional evidence for the hyperarousal theory of ID. However, previous studies have not examined the FC between the hypothalamus and arousal-promoting neurotransmitters, and further research is needed.

However, our study had several limitations. First, as an observational study, it lacked interventional research, thereby limiting the depth of mechanistic understanding and the strength of evidence guiding targeted therapies for the ID. Second, the studies analyzed neurotransmitters measured in the blood rather than in the CNS, including direct measurements of concentrations in the brain or CSF. Third, ID encompasses various subtypes, yet this study did not extensively investigate the differences between these subtypes among the included patients, leading to a restricted comprehensive interpretation of the results and limiting the generalizability of the research findings. Fourth, integrating real-time EEG-fMRI could further explore the FC changes in the anterior and posterior hypothalamus between ID patients and HCs across wakefulness and different sleep stages. Finally, this is a small sample exploratory study, and therefore the results are preliminary. Future research should aim for larger sample sizes and increased replication to establish more robust conclusions.

Conclusion

The ID patients exhibited aberrant FC between the AH/PH and the brain regions related to sleep-wake regulation, particularly involving the DMN and ACC, which may correlate with elevated peripheral levels of arousal-promoting neurotransmitters (orexin-A and histamine). These findings enhance our understanding of the potential neuroimaging and neurohumoral mechanisms underlying ID, suggesting a possible collaborative role of the hypothalamus and arousal-promoting neurotransmitters in the pathogenesis of ID. However, further research with larger sample sizes is needed to validate the findings of this pilot study.

Supplemental Information