Background
Bipolar disorder is generally characterized by alternating mood episodes of opposite polarity, specifically the emotional highs (mania or hypomania) or lows (depression) or the mixed states. These clinical symptoms vary between the two major types of this disease, namely, bipolar disorder type I (BD I) and type II (BD II), as well as between individuals [
1]. In depressive or mixed episodes, and conditions associated with aggression/ impulsivity, the major risk for the patients is suicidal ideations and attempts [
2,
3]. During manic/ hypomanic episodes, patients are usually unable to correctly judge the negative consequences of the excessive involvement in potentially harmful activities in their lives [
4]. Additionally, due to symptom overlap, the two types of bipolar disorder are difficult to separate accurately. Compared to BD I, BD II is characterized by increased frequency of depressive episodes, equal or increased rates of disability, and increased risks of suicidal behavior [
5]. Unfortunately, the medications available to treat these disorders are not optimally effective and their effects vary between BD I and BD II [
6]. All these studies imply that there are affective, cognitive and behavioral differences between the two bipolar disorder types, which might influence their psychological and pharmacological management in the clinic. Therefore, there is a practical need to delineate the two bipolar types, for instance, by typifying them with more neurophysiological endophenotypic markers.
Researchers worldwide have identified some potential differences between BD I and BD II including genetic [
7‐
9], anatomical [
10‐
12], and clinical [
13,
14] aspects. For instance, clinically, BD I patients presented increased sexual thoughts, desire, and activity [
15,
16], while BD II patients often displayed diminished libido and sexual dysfunction [
17]. In neuroimaging studies, during a negative-emotional task, BD I patients displayed a reduced activation of the left insula coupled with a reduced activation of the right supramarginal gyrus [
18]. The diffusion tensor imaging techniques have also shown that the functional connectivity between structures in the fronto-limbic networks was reduced in bipolar disorders, especially in BD I, and the reduction was independent of the serotonin transporter function [
19]. The neurophysiological evidence has also demonstrated higher ratios of P50 cerebral potential in BD I patients with and without a history of psychosis than those in BD II patients [
20].
When processing emotions in bipolar disorder, cerebral dysfunctions were found in interactions between the frontal and temporal lobes, insula and subcortical structures such as amygdala, basal ganglia, hypothalamus and brainstem [
21]. For instance, the abnormal structure and function of cortical and subcortical networks were directly related to the emotional suppression and reappraisal in bipolar disorder [
22]. When facing emotional stimuli of fear and happiness, BD I patients showed increased activation in frontal regions, including the anterior cingulate cortex, dorsolateral prefrontal cortex, and amygdala [
23]. In addition, when processing external negative emotions, BD I manic patients displayed significantly lower amplitude of the cerebral P3 potential compared to healthy volunteers [
24]. However, there was less evidence regarding the cerebral activation when responding to facial or external emotional-stimuli in BD II patients.
On the other hand, the emotional processing involves multi-level interactions between the brainstem and cortex, and emotional expressions of complex, organized somatic and visceromotor activities are generated from brainstem structures [
21]. One question arises – how about the brainstem or brainstem-cortical activities when responding to external emotions in bipolar disorder? A way of illustrating these functions might be through the assessment of brainstem reflexes under cortical drive. In normal men, there is a neurophysiological test which reflects the function of the brainstem inhibitory interneurons, that is, when applying slight painful electrical stimuli to the labial commissure, two successive exteroceptive suppressions (ES1 and ES2) of the jaw-closing muscles (masseter and temporalis) are elicited. ES1 is believed to be mediated by an oligosynaptic circuit in the pontine area, and ES2 by a polysynaptic circuit probably located in the lateral tegmental field [
25]. The ES2 duration was decreased in patients with chronic tension-type headache [
26], dystonia [
27], stroke [
28], generalized anxiety disorder [
29], Alzheimer’s disease [
30], Parkinson’s disease [
31], and in participants with high aggression trait [
32]. Notably, ES2 duration was prolonged and its latency was shortened under relaxation or depressive mood [
33]. The abnormal temporalis ES2 in these neuropsychiatric disorders also supports the notion that the brainstem inhibitory reflex is influenced by the emotional-processing centers in the limbic system, such as amygdala, periaqueductal gray matter, and hypothalamus [
33].
Accordingly, studying the exteroceptive suppression activities in response to external stimuli of positive or negative emotions in different types of bipolar disorder may help to better understand and diagnose the symptoms of this disease. Based on the above-mentioned research, in the current study, we have hypothesized that the external emotional-stimuli reduce the temporalis ES2 duration in both BD I and BD II patients. Therefore, we have asked both BD I and BD II patients and healthy volunteers to undergo the temporalis exteroceptive suppression tests under external stimuli of both positive (e.g., happiness and erotica) and negative (disgust, sadness and fear) emotions, to test whether ES2 durations are different in BD I and BD II under different emotional-stimuli. Additionally, the participants also answered the Mood Disorder Questionnaire [
34], the Hypomania Checklist-32 [
35] and the Plutchik-van Praag Depression Inventory [
36] to characterize their own ongoing affective states.
Discussion
In the current study, after controlling for age, gender and education level factors, we found that patients with BD II scored significantly higher than the healthy volunteers and patients with BD I did on PVP; patients with BD I scored significantly higher than the healthy volunteers and patients with BD II did on MDQ. We also found that patients with BD I and BD II scored significantly higher than the healthy volunteers, while patients with BD I scored higher on HCL-32 than those with BD II. These outcomes are consistent with previous findings [
42‐
44] and support the suggestion that BD I and BD II represent manic and depressive extremes in the bipolar spectrum respectively [
45]. Moreover, we found that ES2 latency was reduced under Erotica in the BD I group, ES2 latencies were prolonged under Disgust and Happiness, and ES2 durations were reduced under Disgust, Happiness and Sadness in the BD II group. Furthermore, there was no relationship between ES1 and ES2 latency and duration and PVP, MDQ or HCL-32 scores in a given group, suggesting that the abnormal brainstem inhibitory reflexes were independent from their clinical, ongoing affective states. To the best of our knowledge, this is the first time that the temporalis ES2 methodology is tried under different stimuli of external emotions to examine the brainstem dysfunctions in the two types of bipolar disorder.
Regarding BD I, patients often display cognitive impairment not only during their own different mood episodes but also during their euthymic phases [
46,
47]. Patients often present their own affective states as mania, depression or mixture of both [
18,
48], they also consistently show increased sexual thoughts, desire, and activity [
15,
16]. The perception, cognition and behavior under erotic stimulation, might be linked to more excitation of cerebral areas such as the prefrontal, hypothalamus and cingulate cortices [
49], which in turn delayed the activation of brainstem inhibitory interneurons, thus delaying the latencies of ES1 and ES2 under erotica. On the other hand, when processing both fear and happiness in patients with BD I, increased activation of the left amygdala/ ventrolateral prefrontal cortex [
50], thalamic, pallidal, and caudate/ putamen regions [
51] were reported, which again might be the reason for the prolonged ES1 latencies under fear and happiness. Indeed, our BD I patients perceived higher intensity under Fear. Apparently, the delayed activation had kept its effect under Blank 2, where both ES1 and ES2 latencies were prolonged. Moreover, BD I participants displayed normal ES1 and ES2 durations, but their perceived happiness and sadness intensities were negatively correlated with the ES2 duration. Possible reasons might be that the cerebral areas involved when processing both positive (such as happiness) and negative (such as sadness) emotions had functioned excessively to inhibit the brainstem inhibitory interneurons. Previous studies did show that bipolar disorders (mainly BD I) often presented elevated emotions such as happiness, disgust and sadness [
52].
Compared to BD II, the ES1 duration under Blank 1 was significantly decreased in BD I patients, which implied that the inhibitory oligosynaptic interneurons were less activated at the baseline (resting) state in the disorder. Although the inhibitory polysynaptic interneurons mediating ES2 duration were less affected in BD I, this finding was in accordance with the baseline manic state in this pathology [
4].
Regarding BD II, the ES2 durations were reduced under Disgust, Happiness and Sadness in the current study. Indeed, fMRI studies have demonstrated the involvement of several cerebral areas when these emotions were processed in BD II. For instance, when responding to facial disgust, there was altered connection between prefrontal-subcortical network, diminished prefrontal and increased limbic areas in both BD I and BD II patients [
53]. When processing happy faces in BD II patients, there was a loose connection between the orbitomedial prefrontal cortex and amygdala, which might result in an abnormal activation of the latter nucleus [
54]. When processing sadness, BD II patients showed a reduced regional orbitofrontal and limbic activations and the magnitude of these activations were correlated with depression severity [
55], and showed an increased amygdala activation in another study [
56]. Interestingly, in our study, BD II patients perceived higher intensity under Disgust. These results pointed to the excessive activation of brain nuclei which might cause more inhibition to the brainstem interneurons mediating ES2 under these emotions.
Moreover, in BD II patients, ES2 latencies were prolonged under Disgust and Happiness. Similarly, the above-mentioned subcortical nuclei under Disgust and Happiness might inhibit the activation onset of the brainstem inhibitory interneurons, thus they might result in delayed ES2 in BD II patients. However, the possibly abnormal activation of the left hippocampal area failed to exert its control effect on the activation-onset of brainstem inhibitory interneurons, thus allowing an intact ES2 latency under Sadness in our BD II group. Although it is still unknown how much the cognition contributes to the ES2 formation, the memory processing function of the hippocampus [
57] might help to facilitate the ES2 activation in this disorder. Indeed, clinically, BD II patients more depression tendency [
1] and might be more adapted to the depressive situation [
45,
58]. Additional support for the reduced ES2 under Sadness in the BD II group might come from the findings that chronic tension-type headache had reduced ES2 duration [
26,
59], and this headache type was often comorbid with depression [
60].
However, it must be noted that our study suffers from at least four design flaws. First, we only enrolled bipolar disorder patients; recruiting other non-bipolar disease-controls, such as unipolar major depression and schizophrenia, which also display depressive or manic episode, would be nice justifications for our findings. Second, we did not record the personality traits or personality disorder functioning styles of our participants; whether their personalities contribute to our current findings remains unknown. Third, we only employed five emotions as external stimuli; additional one, such as surprise, anger, or contempt might also display their effects on the brainstem reflexes. One more concern is that ES1 latencies and durations showed some alterations in both the BD I and BD II groups. Since ES2, rather than ES1, reflects more about the brainstem functions, and provides more meaningful information in this regard [
25], the outcomes in the current study nevertheless provide some hints about the cortico-brainstem dysfunctions under different external emotional-stimuli in different types of bipolar disorder. Regarding the temporalis ES2 reduction, Erotica exerted a pronounced effect on BD I patients, while Disgust, Happiness and Sadness exerted pronounced effects on BD II patients. From a neuropsychophysiological aspect, on the one hand, our results help to delineate the two types of bipolar disorder, and on the other, they offer different therapeutic modalities of emotional control for them. In spite of these clinical implications, our current findings would need further confirmation from other independent laboratories worldwide.