Neil S. Cherniack

Mathematical models of periodic breathing and their usefulness in understanding cardiovascular and respiratory disorders

Periodic breathing is an unusual form of breathing with oscillations in minute ventilations and with repetitive apnoeas or near apnoeas. Reported initially in patients with heart failure or stroke, it was later recognized to occur especially during sleep. The recurrent hypoxia and surges of sympathetic activity that often occur during the apnoeas have serious health consequences. Mathematical models have helped greatly in the understanding of the causes of recurrent apnoeas. It is unlikely that every instance of periodic breathing has the same cause, but many result from instability in the feedback control involved in the chemical regulation of breathing caused by increased controller and plant gains and delays in information transfer. Even when it is not the main cause of the periodic breathing, unstable control modifies the ventilatory pattern and sometimes intensifies the recurrent apnoeas. The characteristics of disturbances to breathing and their interaction with the control system can be critical in determining ventilation responses and the occurrence of periodic breathing. Large abrupt changes in ventilation produced, for example, in the transition from waking to sleep and vice versa, or in the transition from breathing to apnoea, are potent factors causing periodic breathing. Mathematical models show that periodic breathing is a ‘systems disorder’ produced by the interplay of multiple factors. Multiple factors contribute to the occurrence of periodic breathing in congestive heart failure and cerebrovascular disease, increasing treatment options.

Cytokines across the Night in Chronic Fatigue Syndrome with and without Fibromyalgia

ABSTRACT The symptoms of chronic fatigue syndrome (CFS) are consistent with cytokine dysregulation. This has led to the hypothesis of immune dysregulation as the cause of this illness. To further test this hypothesis, we did repeated blood sampling for cytokines while patients and matched healthy controls slept in the sleep lab. Because no one method for assaying cytokines is acknowledged to be better than another, we assayed for protein in serum, message in peripheral blood lymphocytes (PBLs), and function in resting and stimulated PBLs. We found no evidence of proinflammatory cytokine upregulation. Instead, in line with some of our earlier studies, we did find some evidence to support a role for an increase in interleukin-10, an anti-inflammatory cytokine. Although the changes were small, they may contribute to the common complaint in CFS patients of disrupted sleep.

Cardiovascular Evaluation of Electronic Control Device Exposure in Law Enforcement Trainees: A Multisite Study

Objective: Occupational health risk with regard to training exercises is a relatively under studied domain for law enforcement officers. One potential health risk is exposure to electronic control devices (ECDs). Methods: Seven different training facilities in six states participated. Law enforcement trainees (N = 118) were exposed to Taser Internationals (Scottsdale, AZ) X26® for up to 5 seconds. Results: There was no evidence of cardiac or skeletal muscle breakdown. Exposure did not adversely affect electrocardiogram (ECG) morphology obtained 24 hours after exposure in 99 trainees. For two trainees with preexisting ECG abnormalities, ECG morphology differed in the post-ECD samples. Conclusions: The results from this large, multisite study suggest that, for most trainees, ECD exposure does not represent a significant health risk. Further investigation is warranted for cardiac vulnerability and potential interactions with ECD exposure.

Respiratory and Cardiovascular Response during Electronic Control Device Exposure in Law Enforcement Trainees

Objective: Law enforcement represents a large population of workers who may be exposed to electronic control devices (ECDs). Little is known about the potential effect of exposure to these devices on respiration or cardiovascular response during current discharge. Methods: Participants (N = 23) were trainees exposed to 5 s of an ECD (Taser X26®) as a component of training. Trainees were asked to volitionally inhale during exposure. Respiratory recordings involved a continuous waveform recorded throughout the session including during the exposure period. Heart rate was calculated from a continuous pulse oximetry recording. Results: The exposure period resulted in the cessation of normal breathing patterns in all participants and in particular a decrease in inspiratory activity. No significant changes in heart rate during ECD exposure were found. Conclusion: This is the first study to examine breathing patterns during ECD exposure with the resolution to detect changes over this discrete period of time. In contrast to reports suggesting respiration is unaffected by ECDs, present evidence suggests that voluntary inspiration is severely compromised. There is no evidence of cardiac disruption during ECD exposure.

Influence of arousal threshold and depth of sleep on respiratory stability in man: analysis using a mathematical model.

We examined the effect of arousals (shifts from sleep to wakefulness) on breathing during sleep using a mathematical model. The model consisted of a description of the fluid dynamics and mechanical properties of the upper airways and lungs, as well as a controller sensitive to arterial and brain changes in CO2, changes in arterial oxygen, and a neural input, alertness. The body was divided into multiple gas store compartments connected by the circulation. Cardiac output was constant, and cerebral blood flows were sensitive to changes in O2 and CO2 levels. Arousal was considered to occur instantaneously when afferent respiratory chemical and neural stimulation reached a threshold value, while sleep occurred when stimulation fell below that value. In the case of rigid and nearly incompressible upper airways, lowering arousal threshold decreased the stability of breathing and led to the occurrence of repeated apnoeas. In more compressible upper airways, to maintain stability, increasing arousal thresholds and decreasing elasticity were linked approximately linearly, until at low elastances arousal thresholds had no effect on stability. Increased controller gain promoted instability. The architecture of apnoeas during unstable sleep changed with the arousal threshold and decreases in elasticity. With rigid airways, apnoeas were central. With lower elastances, apnoeas were mixed even with higher arousal thresholds. With very low elastances and still higher arousal thresholds, sleep consisted totally of obstructed apnoeas. Cycle lengths shortened as the sleep architecture changed from mixed apnoeas to total obstruction. Deeper sleep also tended to promote instability by increasing plant gain. These instabilities could be countered by arousal threshold increases which were tied to deeper sleep or accumulated aroused time, or by decreased controller gains.

Paradoxical potentiation of exercise hyperpnea in congestive heart failure contradicts Sherrington chemoreflex model and supports a respiratory optimization model.

Congestive heart failure (CHF) patients suffer decreased exercise tolerance, yet they demonstrate an augmented ventilatory response to exercise such that P(aCO2) remains normal (isocapnic) from rest to maximal exercise in the face of increased pulmonary dead space (Fig. 1). On the other hand, the effect of a large external dead space is hypercapnic instead of isocapnic. This discrepancy suggests that external dead space and pulmonary dead space may exert distinct influences on control of breathing. These paradoxical clinical phenomena are at variance with the conventional chemoreflex model (Johnson 2001), but appear to be consistent with the predictions of the optimization model (Poon 2001; Poon, Tin et al. 2007).

Behavioral State Control and Airway Instability

Sleep is an actively generated state induced by excitation of specific networks that alter the activity of different cell populations along the neuraxis, including serotonin-containing neurons. The main region identified in slow wave sleep (SWS) generation is located within the ventrolateral preoptic area (VLPO), which contains neurons that are specifically activated during sleep20. These cells innervate cell bodies and dendrites of neurons that participate in arousal, such as histamine containing cells13. In addition, the lateral preoptic area projects to midline neurons (Haxhiu, unpublished data), indicating possible pathways by which sleep influences activities of upper airway dilating muscles and cholinergic outflow to the tracheobronchial system.

Excitatory and Inhibitory Influences on the Ventilatory Augmentation Caused by Hypoxia

Maintenance of an adequate supply of O2 to tissues depends on a complex pattern of circulatory and ventilatory adjustments that are triggered by environmental perturbations, alterations in arterial O2 levels, and changes in metabolic rate (16, 17, 66). Ventilatory responses to hypoxia are mediated by discrete neural pathways involving peripheral chemoreceptors, primary afferents, and central neurons (16, 17, 66). Hypoxia increases ventilation entirely by its effects on the peripheral chemoreceptors, mainly the carotid body (17, 18, 51, 63, 66). The characteristics of the response, however, depend on complex interactions at multiple levels of the neuraxis, including the primary afferent neurons and bulbopontine pathways through which the carotid body signals are processed and transduced to ventilation. Neurons in the nucleus of the tractus solitarius and near the ventrolateral surface of the medulla, for example, particularly in the nucleus paragigantocellularis lateralis, play a pivotal role in amplifying respiratory responses to hypercapnia and hypoxia (8, 38, 50, 62). Moreover, hypoxia can affect breathing by direct actions on the brain, including alterations in cerebral blood flow, stimulation or depression of diencephalic and cortical neurons, and possibly changes in cerebral metabolism (16–18, 54, 66).

The Sleep Apnea Number Game: Counting the Apnea-Hypopnea Index

cur with every apnea, arousals can nonetheless predispose to further apneas as well as terminate current apneas [4] . Although not addressed by Aarab et al. [3] , their study surely affects treatment decisions. A single-night polysomnography is not enough to exclude sleep apnea [5] . Hutter et al. [6] recently reported on patients with symptoms of sleep apnea but with normal polysomnograms. Repeat studies measuring esophageal pressure showed that some of them had abnormal polysomnograms while others had the upper airway resistance syndrome. It is still controversial whether patients should be treated for sleep apnea on the basis of the AHI level alone in the absence of sleep-related complaints [7, 8] . In patients with no complaints, treating patients with AHI levels 1 30 per hour with continuous positive airway pressure did not improve the quality of life [9] . At the very least, the study by Aarab et al. [3] showed that no single number can be used to decide on treatment but rather a range. How much reduction in the AHI need occur before treatment is deemed successful? Should the goal of therapy be the reduction in the AHI to ! 5, the supposed normal value [10] ? The linear relationships between cardioand cerebrovascular risks and the AHI suggest a zero AHI might be the ideal treatment goal [11, 12] . However, if sleep apnea is being treated as a preventive measure, is it not at least equally important to treat comorbidities such as obesity? People who had difficulty falling asleep used to count sheep, but now the people who watch others sleep count apneas and hypopneas or apnea-hypopnea indices (AHIs). Disturbed and unrefreshing sleep is an extremely common symptom, reported in some articles to be experienced by as many as 16% of the populace [1] . Hence, these symptoms of obstructive sleep apnea but also of other conditions are no guarantees that sleep apnea is present. Therefore, more objective criteria are needed and, in particular, the AHI measured during whole-night polysomnography [2] . The article by Aarab et al. [3] in this issue of Respiration demonstrates the variability of the AHI. One important conclusion is that apparent improvement in the AHI following various sorts of treatment may not be real but rather a manifestation of AHI variability. In a small group of 15 subjects who were studied 4 times over a 10-week period, Aarab et al. [3] found that the minimal detectable difference in the AHI was 12.8. The greater the AHI, the higher the variability. These findings are not unexpected given the number of different factors that cause apneas and hypopneas. The most important one may be body position, not just because of passive gravitational effects on upper airway tissue but because of effects on the alignment and force-generating capabilities of the various muscles of the upper airways. The degree of sleepiness is also inconstant from night to night and may affect the AHI. Though they need not ocPublished online: November 18, 2008

N-terminal pro-B-type naturetic peptide (NTBNP): so much promise and such a disappointment.

Obstructive sleep apnea (OSA) is well-known for its adverse effects on the cardiovascular system affecting both the blood vessels and the heart. About 15% to 20% of the population of North America has high blood pressure, and about 40% of patients with OSA have hypertension [1]. There are about five million cases of heart failure and about half of them have central or obstructive sleep apneas [2]. Strokes, heart attacks, and arrhythmias are other serious cardiovascular consequences of OSA [1]. OSA wreaks its damaging effects on the cardiovascular system in several ways [2–6] Intermittent hypoxia probably by generating oxygen radicals causes endothelial dysfunction producing tissue injury and interfering with the release of NO [2, 4]. In addition, in mouse studies, intermittent hypoxia increases the susceptibility of the heart to ischemia–reperfusion injury [4] Cycles of apnea and arousal raise sympathetic activity, which can injure both the heart and blood vessels and increase coagulability. Intermittent hypoxia stimulates the release of inflammatory cytokines such as tumor necrosis factor-alpha (TNF-α) [2]. Increased pulmonary vascular resistance as a result of hypoxia leads to cor pulmonale [2]. The increased negative pressures occurring in inspiration during obstructive apneas decrease left ventricular preload and increase left ventricular after load by boosting left ventricular transmural pressures. This slows left ventricular filling and can interfere with normal cardiac function [2, 6]. Inflammatory cytokines and leptin released by adipose tissue contribute to cardiovascular damage in OSA [7]. Adipokines, peptides released from white fat, can modulate the effects of insulin and with the inflammatory cytokines and leptin promote the occurrence of atherosclerosis [7, 8]. Echocardiography and exercise testing can be used to detect cardiac dysfunction. More recently, interest has centered on the measurement of B-type naturetic peptide (BNP) and/or NTBNP in the blood as a tool not only for detecting cardiac dysfunction, but also distinguishing heart failure from respiratory failure and as a guide to the treatment of heart failure [9, 10]. ProBNP (the pro peptide) is formed by ventricular myocytes in response to changes in volume and pressure. It is cleaved by an endopeptidase to BNP, a 32-amino acid molecule [6], which has important biological properties and an N-terminal fragment (NTBNP), which has no known function. They both circulate in the blood but NTBNP has a much longer half-life of several hours than BNP [2, 6]. BNP is destroyed by an endopeptidase in the proximal tubule of the kidney [6]. BNP acts on the cell membrane to increase the intracellular formation of cyclic GMP [6]. It has a cardioprotective effect, lowering cardiac output and reducing afterload. BNP acts on the kidney to produce naturesis and diuresis opposing the actions of the renin–angiotensin system and aldosterone. It sequesters plasma, thereby increasing the hematocrit. It decreases sympathetic activity and by an action on the vagus nerve decreases heart rate. It helps regulate cardiac growth preventing overgrowth and cardiac hypertrophy. BNP levels are increased both in systolic and diastolic dysfunction, left ventricular hypertrophy, after myocardial infarction [10]. BNP levels are also increased by hypoxia [2]. They are higher in the elderly, in women, and in the presence of reduced renal clearance [6, 10, 11]. BNP and NTBNP are equally useful in the detection of cardiac dysfunction [6, 10, 11]. Sleep Breath (2008) 12:3–5 DOI 10.1007/s11325-007-0144-8