Nonlinear analysis of heart rhythm in preeclampsia: a route for translational clinical applications in neuroinflammation

According to some studies, the immune profile of women with preeclampsia shifts from a steady state of mild inflammation; from an equilibrium of both pro-inflammatory and anti-inflammatory mechanisms to a downregulated state characterized by increased inflammatory cytokines and effector immune cells with a concomitant decrease in anti-inflammatory factors and regulatory cells [32]. Other recent findings have documented an increased ratio of pro- to anti-inflammatory cytokines in preeclampsia, suggesting that the decreased levels of cytokines such as interleukin 4 (IL-4) and interleukin 10 (IL-10) accelerate the production of pro-inflammatory cytokines resulting in excessive inflammation [33]. Interestingly, further experimental studies showed that by infusing IL-10 to restore the immune system’s balance, the blood pressure associated with placental ischemia in pregnant rats is reduced [13]. These results support a relevant role of the anti-inflammatory cytokines in preeclampsia. In line with such findings, novel studies report an association between the pro-inflammatory cytokine profile and lower plasma concentrations of soluble CD163 molecule and IL-10 of women with severe preeclampsia [34]. Authors have thus endorsed that assessing the cytokine profile of women with preeclampsia is convenient [35].

Preclinical evidence has indicated that interleukin 6 (IL-6) may play a role in mediating hypertension and reducing renal hemodynamics observed in uterine perfusion reductions in pregnant rats [36]. According to Conrad et al., elevated levels of tumor necrosis factor-alpha (TNF-α) and IL-6 may contribute to preeclampsia’s putative endothelial dysfunction [37]. Novel evidence confirms that high levels of TNF-α have been associated with an increased risk of preeclampsia [38]. Additionally, new findings support a non-bacterial maternal immune response role in the development of preeclampsia and eclampsia [39]. For example, a sterile inflammation may be initiated by molecules in the host organism known as damage associated molecular patterns (DAMPs), Fig. 1. In preeclampsia, various DAMPs may be involved in this disorder’s etiology or exacerbation [39].

Yang et al. [40] reported reduced vagus nerve function in preeclamptic women. Consequently, in these women, evidence suggests that diminished acetylcholine (ACh) synthesis may contribute to excessive inflammation and hypertension [41]. ACh serves as a neuromodulator in the central nervous system, affecting neuronal excitability, synaptic transmission, and plasticity [41]. Findings show that the mRNA expression levels of the alpha 7 nicotinic acetylcholine receptor (α7nAChR) in preeclamptic women’s monocytes were lower than control healthy women’s monocytes. Thus, the authors concluded that the downregulation of α7nAChR may be linked to the development of preeclampsia by boosting pro-inflammatory and decreasing anti-inflammatory cytokine release via the nuclear factor kappa β (NF-κβ) pathway [42]. Considering its ability to modulate anti-inflammatory processes by reducing the expression of pro-inflammatory effectors and cytokines, α7nAChR has received much attention in the last few decades [43]. Therefore, preeclampsia can be considered an altered neuroimmune condition in pregnant women, becoming a frontier topic for multiple disciplines, Fig. 1.

Novel findings suggest that the vagal-driven activation of the α7nAChR attenuates preeclampsia-like symptoms, and this protective effect is likely result of the inhibition of inflammation via the NF-κβ pathway [44].

McAllen et al., have provided compelling evidence in preclinical studies that indicates that splanchnic sympathetic outflow acts to suppress immune-mediated inflammation [45]. However, recent evidence shows that preeclampsia is not associated with elevated sympathetic reactivity evaluated by sympathetic nervous system activity [46]. Consistently, current evidence suggests that an autonomic dysfunction is highly prevalent in preeclamptic women [12].

There is still a lack of knowledge in how preeclampsia modifies the neuroimmune pathways; for example, some studies have reported an association among maternal immune activation, the manifestation of preeclampsia, and an altered neonatal neurodevelopment [47]. A possible pathway for this association involves the dysfunction of the microglial cells, which are the brain’s immune cells [47]. In this sense, studies are necessary to provide more apprehensible knowledge about the neuroinflammatory mechanisms in preeclampsia. Biomedical research is pushing forward to developing potential non-invasive and fast diagnostic biomarkers for clinical translational applications.

Nonlinear HRV analysis is based on chaos theory, in which the heart rhythm generation, besides being sensitive to initial conditions and evolving very fast over time, is considered nonlinear [48, 49]. On the other hand, when using linear analysis techniques of HRV data, several assumptions must be made to ensure that the data is meaningfully interpreted. For example, the HRV data must be stationary, i.e., the mean and variance must be stable. However, it is recognized that the cardiac system is dynamic, nonlinear, and nonstationary [49]. Thus, linear methods of HRV may not account for all aspects of cardiac performance, especially the complex interactions between the control mechanisms that regulate heart function, due to assumptions and conditioning needed for linear HRV analysis [49]. In any case, both linear and nonlinear processing algorithms of HRV have provided useful computational tools for diagnosing a wide range of inflammatory conditions and pathologies [26‐29]. According to a review of 2020, the most measurable nonlinear features of HRV are representation features (e.g., Poincaré plot representation, recurrence plot analysis, asymmetry); fractal (e.g., detrended fluctuation analysis, Hurst exponent); entropy or complexity-based features (e.g., Shannon entropy, sample entropy, multiscale entropy) and symbolic dynamics [50].

Evidence has also shown negative associations between linear HRV indexes such as the standard deviation of the normal-to-normal interval with TNF-α in heart failure patients, suggesting that the immune system’s activation is related to autonomic imbalance [51]. More specifically, our previous research indicates that nonlinear features of HRV such as entropy-based methods, symbolic dynamics, and fractal methods have successfully characterized systemic acute inflammatory conditions such as the lipopolysaccharide (LPS) induced endotoxemia [52‐54] and low-risk human parturition at term [55].

Specifically, it is known that preeclampsia affects the cardiovascular health of both mother and offspring [56]. Studies also indicated that different hypertensive disorders such as chronic hypertension, preeclampsia, and pregnancy-induced hypertension (PIH) are different in HRV, indicating that different regulatory mechanisms are involved [57]. In Mexican women, PIH’s progression into preeclampsia has been appreciated in one of each four patients [35]; thus, the evaluation of HRV may be considered an additional biomarker for monitoring PIH patients who may develop preeclampsia. A recent review study indicates that, in general, a decrease in overall HRV is found in preeclampsia compared to normotensive pregnant controls [58]. Among different autonomic alterations in preeclampsia, a reduction in the autonomic vagal modulation and increased sympathetic autonomic modulation assessed by HRV analysis have been reported [9, 40].

Novel nonlinear measures have been introduced to describe HRV, which present the advantage of not being affected by nonstationary effects [59]. Table 1 summarizes the primary outcomes of clinical and preclinical studies of HRV and preeclampsia [9, 40, 57, 60‐66]. Table 1 specifies the type of HRV measure (linear or nonlinear), and if the analysis was performed in fetal or maternal HRV data. Interestingly, only one (1/10) of the consulted studies have evaluated nonlinear analysis of maternal or fetal HRV in women with preeclampsia.

Nonlinear HRV measures or features have shown potential as biomarkers of internalizing psychopathology, often associating lower information-based complexity with internalizing psychopathology [67]. The relationships between physiological variability, interorgan communication, and disease suggest that monitoring complexity may reveal disease state; thus, decreased complexity would correspond with disease progression, and increased complexity (towards homeostasis) would correspond with recovery [68]. The complexity analysis of HRV data, which also quantifies nonlinear interactions among frequencies reflecting underlying ANS dynamics, represents a relevant topic in assessing neuroimmune disorders. However, this complexity also means that there is potentially a variety of information from a wide range of physiological systems embedded in the HRV, indicating an opportunity to utilize computational approaches to reveal the physiological significance of HRV.

Nonlinear measures of HRV have already allowed the discrimination of depressive patients from healthy subjects, consistently showing a significant decrease of HRV complexity in neuroimmune pathologies [69]. Mainly, multiscale entropy, a complexity measure of HRV, was more sensitive than current clinical markers for evaluating inflammation in a preclinical model of LPS-induced inflammation [70]. Authors suggest that multiscale entropy may prove a fruitful component of a bedside monitor to detect inflammatory insults.

According to the consulted literature, no clinical studies have been conducted to simultaneously evaluate the nonlinear features of HRV and multiple inflammation markers in women diagnosed with preeclampsia. However, results indicate that preeclamptic women reveal a reduced complexity by incrementing periodicity assessed by multiscale asymmetry indexes of HRV [71]. The authors concluded that the mathematical indexes obtained with the nonlinear methods of HRV could differentiate between normal and preeclamptic pregnancies. At present, there is a clinical need to diagnose preeclampsia accurately and promptly, which usually progresses to an adverse fetal or maternal outcome.

The CAP activity can be monitored easily and non-invasively via the linear analysis of HRV data derived from either maternal or fetal electrocardiogram, Fig. 2. However, the linear analysis of HRV is not entirely appropriate to offer information about the complex dynamics of heartbeat fluctuations, and consequently, from the CAP activity. This is because the mechanisms involved in cardiovascular physiology interact with each other in a nonlinear way [72]. The activation of the CAP could rather be crucial for controlling preeclampsia symptoms. Besides, nonlinear methods of HRV not only present clinical relevance but also offer an improved interpretation of pathological conditions. Additionally, the main advantage of HRV signals is that they can be calculated in real-time and non-invasively, while all current biomarkers used in clinical practice are discrete and imply blood sample analyses which delay results.

The information derived from chaos theory, fractal mathematics, and the dynamic complexity of HRV has not yet been fully applied in the obstetrics field. It is a productive area for research in both healthy and pathological scenarios. Also, the complex characterization of the HRV time series during preeclampsia has rarely been studied. Theoretically, the maternal and fetal nonlinear analysis of HRV offers unique opportunities to explore the neuroimmune interactions in the mother-child binomial.