Brain–Lung and Brain–Heart Physiological Interactions

The brain, lungs, and heart share a circulation and influence one another through pressure, gases, neural signals, and inflammation. An intervention can improve one organ’s measurement while worsening another organ’s function. Understanding the separate pathways makes such responses interpretable: oxygenation, intracranial pressure, perfusion pressure, cardiac output, and tissue oxygen availability are related observations, not interchangeable endpoints.

63.1 Connected organs communicate through several pathways

The lungs determine arterial oxygenation and carbon dioxide elimination. The heart provides the flow carrying that blood toward the brain. Cerebral vessels regulate its distribution, while venous drainage returns blood through pathways affected by head position and thoracic pressure. An injury anywhere along this sequence can change cerebral supply without a new structural brain lesion.

Communication also runs outward from the brain. Brainstem networks organize breathing and airway reflexes. Cortical, hypothalamic, and brainstem autonomic networks influence sympathetic and parasympathetic activity. Acute injury can therefore alter ventilation, heart rate, vascular tone, and myocardial function. Inflammatory signals and changes in immune function provide additional routes to extracranial injury.

These routes operate on different timescales. A cough changes thoracic and intracranial pressures within seconds. CO₂ accumulation changes cerebrovascular tone over a different trajectory. Myocardial dysfunction, inflammatory lung injury, pneumonia, and evolving fluid balance can persist after the initial neural event has passed. A single simultaneous set of measurements cannot reconstruct all these processes.

Cause and consequence can also form a loop. Brain injury can impair airway protection and promote aspiration. Lung injury then worsens gas exchange, increases respiratory effort, and changes the ventilatory support required. The resulting hypoxemia, CO₂ disturbance, or circulatory deterioration can further threaten the brain. Likewise, neurocardiac dysfunction can reduce cerebral supply and promote pulmonary edema, making both oxygenation and systemic perfusion harder to sustain.

The first interpretive step is to name the pathway being proposed. Direct transmission of pressure is different from a fall in cardiac output, and both differ from CO₂-mediated cerebral vasodilation. Several can occur together, but one should not be inferred merely because another was measured. Chapters 58–62 establish intracranial mechanics, cerebral perfusion, metabolism, CSF transport, and edema; this chapter connects those processes to the chest and circulation.

63.2 Airway pressure is partitioned between lung and chest wall

Positive airway pressure expands the respiratory system, which includes both lung and chest wall. Part raises the pressure distending the lung, and part changes pleural pressure. The pleural component influences intrathoracic vessels and cardiac chambers. Consequently, the same airway pressure can have different circulatory consequences in patients with different lung and chest-wall mechanics.

During a passive, no-flow measurement, assume airway pressure represents alveolar pressure in communicating units. Transpulmonary pressure, Pᴸ, is alveolar pressure minus surrounding pleural pressure, Ppl. Expressing both pressures in cmH₂O:

Consider constructed end-expiratory measurements. Total PEEP rises from 8 to 14 cmH₂O while an appropriate esophageal-pressure estimate rises from 5 to 8 cmH₂O. Estimated transpulmonary pressure rises from 8 − 5 = 3 to 14 − 8 = 6 cmH₂O. Thus, the 6 cmH₂O airway-pressure increase is partitioned into approximately 3 cmH₂O of pleural-pressure increase and 3 cmH₂O of additional lung-distending pressure.

Using 1 cmH₂O ≈ 0.736 mmHg, the estimated pleural increment is approximately 2.21 mmHg. This conversion aligns units; it does not predict a 2.21 mmHg increase in CVP or ICP. Pressure transmission, vascular geometry, venous collapse, and intracranial accommodation still intervene.

Esophageal pressure estimates a regional pleural condition and requires attention to balloon position, calibration, body position, and surrounding anatomy. It is not uniform pressure throughout the thorax. Regional alveolar pressure can also differ with airway closure or trapped gas. These constructed partitions therefore explain the mechanism without supplying a universal PEEP correction.

Spontaneous effort adds another force. Inspiratory muscle contraction can lower pleural pressure even as positive airway pressure is applied, increasing transpulmonary pressure beyond what the airway display alone suggests. Airway pressure must therefore be interpreted with effort and measurement conditions before its cerebral or cardiac consequences are predicted.

63.3 Venous drainage and arterial supply determine the intracranial response

Rising thoracic venous pressure can impede cerebral venous drainage. Increased upstream venous pressure and intracranial blood volume may then raise ICP, particularly when remaining volume accommodation is limited. However, cerebral veins are collapsible pathways, not rigid extensions of the right atrium.

When an outflow segment is narrowed by surrounding pressure, its upstream flow can become partly separated from a lower downstream pressure. Experiments in anesthetized dogs demonstrated a Starling-resistor behavior between cerebral veins and the sagittal sinus during PEEP and venous-pressure changes. The findings explain conditional transmission rather than a fixed human bedside formula. 1

By raising right atrial pressure relative to upstream systemic veins, a thoracic-pressure rise can also reduce the gradient for venous return. Changes in venous capacitance and abdominal pressure can modify this response. Altered filling and right ventricular loading can then reduce cardiac output. If arterial pressure at the brain falls, cerebral perfusion can deteriorate even when ICP barely changes. Under the conventional assumptions discussed in Chapter 59, with both pressures referenced appropriately and expressed in mmHg:

For illustration, MAPbrain 85 and ICP 18 give CPP 67 mmHg. If a later state has MAPbrain 75 and ICP 21, CPP is 54 mmHg. Only 3 mmHg of the 13 mmHg fall comes from increasing ICP; the larger contribution is falling arterial pressure. Looking only for an ICP increase would miss most of the change.

In a prospective multicenter observational study of 109 patients, modest PEEP changes produced small average ICP changes without a significant average CPP change. The relatively low starting ICP and selected clinical changes limit extrapolation to refractory intracranial hypertension or other ventilatory conditions. Group averages do not establish individual tolerance. 2

Externally transmitted CVP still represents intravascular pressure affecting connected veins. It should neither be dismissed as meaningless nor equated automatically with local cerebral venous pressure. Hydrostatic references, neck pathways, collateral drainage, and thoracic conditions determine the connection.

63.4 Recruitment and overdistension create different brain–lung responses

Recruitment opens previously poorly aerated units and can reduce shunt. If ventilation becomes more evenly distributed, effective gas exchange may improve without concentrating the tidal breath in a small remaining lung. Overdistension instead expands already open units excessively and can compress alveolar vessels, increase wasted ventilation, and increase right ventricular afterload.

These responses can coexist in different regions. A rise in end-expiratory lung volume alone does not establish how much represents recruitment rather than inflation of already open tissue. Neither improved oxygenation nor a favorable compliance change independently describes regional lung stress or cardiac tolerance.

In twelve severely brain-injured patients with acute lung injury, Mascia and colleagues compared PEEP responses in recruitable and non-recruitable lungs. Recruitment accompanied improved oxygenation without a significant ICP increase. In non-recruiters, increasing elastance and PaCO₂ accompanied rising ICP. This small physiological experiment links lung behavior to a CO₂ pathway; it does not establish a universal safe PEEP level or prove that all intracranial responses arise from CO₂. 3

A 2025 ten-patient proof-of-concept study illustrates a different combination. Lowering PEEP increased MAP, CPP, and ICP together, while oxygenation worsened. Patients were deeply sedated and paralyzed, head-elevated, and studied after a recruitment procedure. Arterial CO₂ showed no significant paired change in available samples, but some gas measurements were missing. The pattern is compatible with interacting systemic pressure and cerebrovascular responses; it does not show that higher PEEP uniformly improves brain perfusion. 4

Right ventricular effects add another route. Recruitment may lessen hypoxic vasoconstriction in newly ventilated regions. Overdistension compresses alveolar vessels, while hypoxemia, hypercapnia, and acidemia can further increase pulmonary vascular tone and loading. Falling right ventricular output then reduces left ventricular filling and potentially cerebral supply. The lung, heart, and brain responses must therefore be read together rather than attributed to airway pressure alone.

63.5 Carbon dioxide control depends on effective alveolar ventilation

Minute ventilation is the total gas moved each minute. Alveolar ventilation is the portion participating effectively in CO₂ exchange. Reducing tidal volume while keeping minute ventilation constant does not necessarily preserve alveolar ventilation, because dead-space ventilation may occupy a larger fraction of each breath.

Let tidal volume VT and effective physiological dead-space volume VD be expressed in liters, and frequency f in breaths/min. In a simplified steady-state model:

For constructed values, VT 0.50 L, VD 0.15 L, and f 12 give minute ventilation 6.0 L/min and alveolar ventilation 4.2 L/min. Changing to VT 0.40 L and f 15 preserves minute ventilation at 6.0, but gives alveolar ventilation 3.75 L/min. If CO₂ production, dead space, and other conditions stayed fixed, PaCO₂ would increase by the ratio 4.2/3.75 = 1.12; an initial 40 mmHg would become approximately 44.8 mmHg.

Actual dead space changes with perfusion, recruitment, overdistension, and equipment. CO₂ production changes with temperature, activity, and metabolism. Increasing frequency also shortens expiration and can produce dynamic hyperinflation when emptying is incomplete. The calculation is therefore an explanation of a possible response, not a ventilator prescription.

CO₂ enters brain tissue readily and changes the local acid–base environment governing vascular tone. Rising PaCO₂ generally favors dilation, potentially increasing both flow and vascular volume. Falling PaCO₂ favors constriction, which can lower ICP while reducing flow. Classic human experiments demonstrated substantial gas-related flow changes without corresponding increases in cerebral oxygen consumption. 5

End-tidal CO₂ is useful for trends but is not always PaCO₂. Changing dead space and pulmonary perfusion can change their difference. A ventilation change should consequently be interpreted through actual gas exchange and the cerebral response, rather than assuming that unchanged minute ventilation or end-tidal values guarantee unchanged arterial CO₂.

63.6 Effort, airway protection, and position alter the connection

Acute brain injury can disturb respiratory rhythm, reduce airway protection, impair swallowing, and weaken cough. Pulmonary complications may therefore begin through aspiration, retained secretions, atelectasis, or infection rather than a primary neurogenic vascular event. These mechanisms can overlap with autonomic and inflammatory lung injury.

Respiratory effort has its own circulatory effects. Strong inspiration makes pleural pressure more negative, increasing venous inflow and the pressure the left ventricle must generate relative to its surroundings. In susceptible patients this can increase pulmonary vascular filling and favor edema. Large transpulmonary swings can also increase regional lung stress despite modest displayed airway pressure.

Coughing, straining, and some episodes of ventilator asynchrony produce abrupt thoracic-pressure and ICP changes. Pain and agitation can simultaneously increase sympathetic activity, CO₂ production, and cerebral demand. A transient spike must be interpreted with its duration, accompanying perfusion, and the underlying intracranial state; it does not by itself measure new structural injury.

Position changes several boundaries at once. Neck rotation or external compression can obstruct venous pathways. Abdominal compression may transmit pressure toward the chest and alter venous return. Head elevation changes arterial hydrostatics at brain level as well as venous drainage. Transducer leveling must remain appropriate if before-and-after pressures are to be compared.

Prone positioning illustrates the potential tradeoff. It may improve pulmonary gas distribution while changing thoracic, abdominal, and cranial venous conditions. In a retrospective series of 27 patients with brain injury and moderate-to-severe ARDS, oxygenation improved, but intracranial hypertension occurred in some patients and proning was stopped for sustained ICP increases in five. Technique was incompletely characterized, and the analyzed records sampled ICP at specific hourly times. The series does not provide a universal screening threshold or establish the outcome effect of proning in this population. 6

63.7 Brain injury can prime the lung for a second insult

An acute neural injury can activate sympathetic pathways and initiate systemic inflammatory signaling. Endothelial activation and leukocyte recruitment can alter the lung before a major gas-exchange abnormality appears. The lung may then respond more severely to a later exposure such as aspiration, infection, or injurious mechanical stretch.

This “second insult” concept describes changed susceptibility rather than an inevitable sequence. An initially normal oxygenation ratio does not exclude biological priming, but neither does every subsequent pneumonia or ARDS episode prove that the brain initiated the lung injury.

In an experimental rabbit study, lungs obtained after induced massive brain injury developed greater edema-related changes and alveolar hemorrhage during subsequent ex vivo injurious ventilation than control lungs. The design demonstrates increased susceptibility in that preparation; it does not reproduce every exposure or time course of a living patient. 7

A mouse concussion model provided more specific inflammatory evidence. Brain injury produced pulmonary interstitial neutrophil accumulation without initial alveolar injury. A later small acid aspiration produced augmented alveolar injury, which was reduced by neutrophil depletion. This supports a neutrophil-dependent mechanism in that model, not a clinical recommendation to deplete neutrophils or assume the same response magnitude in humans. 8

Human lung dysfunction after brain injury remains multifactorial. Aspiration, chest trauma, transfusion, infection, fluid accumulation, cardiac dysfunction, and ventilation can all contribute. Injury severity also affects exposure to these events and the duration of mechanical support. An association between neurological severity and pulmonary complications can therefore contain both direct biological effects and shared clinical causes.

The reverse inflammatory route is also plausible: pulmonary injury and systemic inflammation can affect vascular and cellular function in an already vulnerable brain. Its contribution in an individual patient cannot be calculated from a circulating cytokine concentration. The mechanistic value is to recognize that protecting gas exchange alone does not fully describe the brain–lung relationship.

63.8 Neurogenic pulmonary edema combines vascular pressure and barrier effects

Neurogenic pulmonary edema describes pulmonary edema developing in relation to an acute neurological insult. The label proposes a causal connection; it should not terminate examination for aspiration, acute cardiac disease, infection, or other contributors. Several mechanisms can produce a similar radiographic and oxygenation pattern.

An intense sympathetic response can constrict systemic vessels and redistribute blood centrally. Increased left ventricular afterload, impaired ventricular relaxation or contraction, and altered pulmonary vascular tone can raise pulmonary microvascular pressure. Fluid filtration then increases. A sufficiently abrupt pressure surge may also damage the alveolar–capillary barrier, so a hydrostatic initiating event and subsequent permeability abnormality are not mutually exclusive.

Animal experiments support a substantial pressure-mediated pathway under particular conditions. In a rat cerebral-compression model, reducing the pressor response attenuated pulmonary hemorrhagic edema, supporting a hemodynamic contribution in that preparation. It does not identify a clinical drug strategy or establish that every neurogenic edema episode requires the same systemic pressure response. 9

Other experiments demonstrate that altered barrier function cannot always be explained by the measured hemodynamic changes alone. McClellan and colleagues found increased pulmonary protein leak after experimentally raised ICP in dogs; reproducing the modest pressure and flow changes through other maneuvers did not reproduce the leak. This supports a neurological influence on permeability in that model, while leaving its exact human contribution uncertain. 10

The observed state also depends on clearance. Lymphatic drainage and epithelial fluid transport oppose accumulation, while sustained venous congestion can hinder removal. A transient initial vascular surge can therefore leave pulmonary edema after the surge has resolved. A later normal wedge pressure does not reconstruct the earlier pulmonary capillary pressure or exclude prior cardiac involvement.

The resulting edema worsens shunt and respiratory mechanics, potentially increasing oxygen requirements, ventilatory pressures, and cardiac strain. These changes close the loop by threatening cerebral oxygen supply and venous drainage. “Neurogenic” therefore describes an interaction requiring explanation, not a purely noncardiac category.

63.9 Sympathetic activation can injure the myocardium

Acute injury to central autonomic networks can abruptly alter sympathetic outflow and adrenal catecholamine release. Adrenergic stimulation initially increases heart rate, contractility, vascular tone, and myocardial energy requirements. Excessive stimulation can promote calcium-handling disturbances, metabolic stress, and impaired contraction. The later circulatory state need not resemble the initial surge.

In an experimental canine subarachnoid-hemorrhage model, catecholamines, arterial pressure, wall motion, and cardiac output initially increased, followed by declining cardiac performance. This temporal sequence supports a neural–cardiac mechanism in the studied model, rather than a prediction that every patient passes through identical phases. 11

Experimental mouse subarachnoid hemorrhage also produced reduced cardiac output and altered peripheral vascular reactivity through distinguishable adrenergic pathways. The study included receptor-directed interventions, but its model, timing, and pretreatment conditions prevent direct conversion into a human prophylactic strategy. It also showed why preserved ejection fraction need not establish preserved output: the fraction ejected and the absolute volume ejected are different quantities. 12

Human observations support, without proving every causal step, a related neurocardiac phenotype. Banki and colleagues studied myocardial perfusion and sympathetic innervation after subarachnoid hemorrhage. Interpretable perfusion scans were normal, while abnormal sympathetic-innervation imaging was associated with wall-motion abnormalities and troponin elevation. The selected cohort supports a neural hypothesis; it does not establish that coronary disease is absent in every patient with neurological injury. 13

Regional or global systolic dysfunction, impaired relaxation, and stress-cardiomyopathy patterns can reduce forward output or raise pulmonary venous pressure. Increased systemic vascular resistance can partly preserve arterial pressure while output falls. Thus, hypertension, a preserved ejection fraction, or a later recovery of contraction cannot independently establish that cerebral flow and oxygen delivery were adequate throughout the episode.

63.10 Arrhythmia and troponin changes require a cardiac explanation

Autonomic disturbances can change heart rate and repolarization. Electrolyte abnormalities, drugs, hypoxemia, acid–base changes, and pre-existing heart disease add further influences. A new rhythm disturbance after brain injury may therefore have several simultaneous causes.

Tachyarrhythmia can shorten diastolic filling and coronary perfusion time while increasing myocardial oxygen demand. Bradyarrhythmia can reduce cardiac output when stroke volume cannot compensate. The resulting hypotension can threaten cerebral supply, while elevated filling pressures or myocardial dysfunction can worsen pulmonary gas exchange. Whether an observed rhythm change produces these consequences depends on duration, ventricular function, and the rest of the circulation.

In a prospective cohort of 580 patients with spontaneous subarachnoid hemorrhage, clinically significant arrhythmias occurred in approximately 4%, most commonly atrial fibrillation or flutter. Arrhythmias were associated with worse outcomes after adjustment, but this observational relationship does not establish how much risk would be prevented by a particular rhythm-management strategy. 14

Troponin addresses injury rather than cause. The 2026 Fifth Universal Definition identifies acute myocardial injury through a rise or fall in troponin with at least one value above the appropriate sex-specific 99th-percentile upper reference limit. Infarction additionally requires an ischemic explanation and the appropriate clinical, electrocardiographic, or imaging evidence. Its primary, secondary, and procedure-related categories describe different contexts; the presence of a sympathetic stressor alone does not establish secondary infarction. 15

Neurocardiogenic injury, primary coronary pathology, ischemia from supply–demand imbalance, and other cardiac diseases can overlap. Neurological impairment or sedation may prevent symptom reporting, increasing the importance of serial ECG, biomarker trajectories, and appropriate imaging. A wall-motion pattern or the temporal relation to brain injury can support an explanation without excluding alternatives. Calling an abnormality “neurogenic” must not become a substitute for establishing what the heart is doing and why.

63.11 A plausible protective component does not prove a beneficial strategy

Lower tidal volume can reduce some forms of lung stretch, while PEEP can support recruitment. Each also changes other conditions: alveolar ventilation, expiration time, pleural pressure, venous return, and right ventricular loading. A combined strategy must therefore be judged by its complete effects rather than by its intended physiological label.

PROLABI randomized 190 adults with severe acute brain injury without ARDS to a lower-tidal-volume/higher-PEEP strategy or conventional ventilation. The 28-day composite of death, ventilator dependence, or ARDS occurred in 61.5% versus 45.3%, respectively; mortality was also higher in the combined-strategy group. The trial stopped early because funding ended, and its authors emphasized the need for confirmation. 16

These findings challenge automatic adoption of that tested bundle in its selected population. They do not isolate tidal volume, PEEP, CO₂, or a circulatory pathway as the cause. They also do not determine the treatment of established ARDS or establish that every lower-volume approach is harmful. Component effects cannot be extracted from randomization to a whole bundle simply by choosing a plausible mechanism afterward.

Observational evidence answers another question. VENTIBRAIN documented ventilation practices in more than two thousand patients with acute brain injury and found associations between ventilatory parameters and mortality. These comparisons are influenced by injury severity, respiratory mechanics, treatment selection, and evolving complications; they do not identify a setting that would improve survival if imposed on every patient. 17

The physiological assessment remains valuable alongside these limits. It asks whether oxygenation, CO₂, respiratory mechanics, arterial pressure, output, and cerebral measurements change coherently after a clinical intervention. Outcome evidence then determines whether a specified strategy produces a favorable balance in a specified population. Neither reasoning process replaces the other, and a favorable short-term measurement is not itself proof of neurological recovery.

63.12 An integrated case separates oxygenation from delivery

Consider constructed observations before and after a clinically chosen ventilatory change. Initially, hemoglobin is 10 g/dL, arterial saturation 0.93, PaO₂ 70 mmHg, MAPbrain 85 mmHg, and ICP 18 mmHg. Afterward, saturation is 0.98 and PaO₂ 100, but MAPbrain is 75 and ICP 21. The oxygenation improvement accompanies a CPP decrease from 67 to 54 mmHg.

To examine oxygen transport, use hemoglobin in g/dL, saturation as a fraction, PaO₂ in mmHg, and CBF in mL blood/100 g/min. Using the same teaching coefficients as Chapter 60:

Initial content is 1.34 × 10 × 0.93 + 0.003 × 70 = 12.672 mL/dL. Later content is 13.432 mL/dL. Suppose comparable flow measurements, rather than an inference from CPP, show average CBF falling from 50 to 40. Delivery then changes from 50 × 12.672/100 = 6.336 to 40 × 13.432/100 = 5.3728 mL O₂/100 g/min. Better arterial oxygenation has accompanied lower calculated cerebral delivery.

This does not identify the cause. Increased pleural pressure, impaired right ventricular output, altered arterial CO₂, or concurrent myocardial dysfunction could contribute. A rise in CVP might reflect pressure transmission, altered ventricular loading, or both. A troponin rise would require its own cardiac evaluation. The timing, respiratory mechanics, gas measurements, echocardiographic findings, and regional cerebral observations help distinguish these possibilities.

Nor does the average delivery calculation establish local consumption or tissue viability. Extraction, demand, regional perfusion, and probe location remain relevant. Rechecking pressure references and measurement comparability precedes a causal interpretation.

The purpose of combining these observations is to explain the patient’s response across organs. A successful brain–lung–heart interpretation accounts for how oxygen reaches the brain, how blood leaves it, what the heart can deliver, and what the current metabolic activity requires.

References

1. Luce JM, Huseby JS, Kirk W, Butler J. A Starling resistor regulates cerebral venous outflow in dogs. J Appl Physiol Respir Environ Exerc Physiol. 1982;53:1496–1503. Original experiment.

2. Barea-Mendoza JA, Molina-Collado Z, Ballesteros-Sanz MÁ, et al. Effects of PEEP on intracranial pressure in patients with acute brain injury: an observational, prospective and multicenter study. Med Intensiva (Engl Ed). 2024;48:594–601. Original study.

3. Mascia L, Grasso S, Fiore T, Bruno F, Berardino M, Ducati A. Cerebro-pulmonary interactions during the application of low levels of positive end-expiratory pressure. Intensive Care Med. 2005;31:373–379. Original study.

4. Bencze R, Kawati R, Hånell A, et al. Intracranial response to positive end-expiratory pressure is influenced by lung recruitability and gas distribution during mechanical ventilation in acute brain injury patients: a proof-of-concept physiological study. Intensive Care Med Exp. 2025;13:43. Original study.

5. Kety SS, Schmidt CF. The effects of altered arterial tensions of carbon dioxide and oxygen on cerebral blood flow and cerebral oxygen consumption of normal young men. J Clin Invest. 1948;27:484–492. Original experiment.

6. Bernon P, Mrozek S, Dupont G, Dailler F, Lukaszewicz AC, Balança B. Can prone positioning be a safe procedure in patients with acute brain injury and moderate-to-severe acute respiratory distress syndrome? Crit Care. 2021;25:30. Original series.

7. López-Aguilar J, Villagrá A, Bernabé F, et al. Massive brain injury enhances lung damage in an isolated lung model of ventilator-induced lung injury. Crit Care Med. 2005;33:1077–1083. Original experiment.

8. Humphries DC, O’Neill S, Scholefield E, et al. Cerebral concussion primes the lungs for subsequent neutrophil-mediated injury. Crit Care Med. 2018;46:e937–e944. Original experiment.

9. Chen HI, Huang HS, Yang JG, Wang D. Vasodilator and oxidant scavenger in the neurogenic pulmonary edema induced by cerebral compression. Chin J Physiol. 1992;35:123–131. Original indexed experiment.

10. McClellan MD, Dauber IM, Weil JV. Elevated intracranial pressure increases pulmonary vascular permeability to protein. J Appl Physiol. 1989;67:1185–1191. Original experiment.

11. Masuda T, Sato K, Yamamoto S, et al. Sympathetic nervous activity and myocardial damage immediately after subarachnoid hemorrhage in a unique animal model. Stroke. 2002;33:1671–1676. Original experiment.

12. Dinh DD, Lidington D, Kroetsch JT, et al. Experimental subarachnoid hemorrhage drives catecholamine-dependent cardiac and peripheral microvascular dysfunction. Front Physiol. 2020;11:402. Original experiment.

13. Banki NM, Kopelnik A, Dae MW, et al. Acute neurocardiogenic injury after subarachnoid hemorrhage. Circulation. 2005;112:3314–3319. Original study.

14. Frontera JA, Parra A, Shimbo D, et al. Cardiac arrhythmias after subarachnoid hemorrhage: risk factors and impact on outcome. Cerebrovasc Dis. 2008;26:71–78. Original cohort.

15. Mills NL, Newby LK, Zaman S, et al. Fifth Universal Definition of Myocardial Infarction (2026). J Am Coll Cardiol. 2026;88:1314–1368. Original joint consensus document.

16. Mascia L, Fanelli V, Mistretta A, et al. Lung-protective mechanical ventilation in patients with severe acute brain injury: a multicenter randomized clinical trial (PROLABI). Am J Respir Crit Care Med. 2024;210:1123–1131. Original trial.

17. Robba C, Giardiello D, Almondo C, et al. Ventilation practices in acute brain injured patients and association with outcomes: the VENTIBRAIN multicenter observational study. Intensive Care Med. 2025;51:318–331. Original cohort.