Show Summary Details
Page of

Cardiovascular interactions in respiratory failure 

Cardiovascular interactions in respiratory failure
Cardiovascular interactions in respiratory failure

Jae Myeong Lee

and Michael R. Pinsky

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 01 July 2022

Key points

  • Spontaneous ventilation is exercise.

  • Acute respiratory failure placed an increased metabolic demand on the cardiovascular system.

  • Spontaneous inspiration increases venous return and impedes left ventricular ejection increasing intrathoracic blood volume.

  • Positive pressure ventilation decreases venous return and augments left ventricular ejection decreasing intrathoracic blood volume.

  • Lung hyperinflation, either by spontaneous ventilation or mechanical ventilation, increases pulmonary vascular resistance and can cause acute cor pulmonale.


Acute respiratory failure (ARF) can directly alter cardiovascular function in a number of seemingly unrelated ways. However, many of these effects are predictable from a knowledge of the determinants of cardiovascular function. The respiratory and the cardiovascular systems are not separate, but tightly integrated. The ultimate cardiovascular response to ARF is dependent on the patient’s baseline cardiovascular state, the type of respiratory dysfunction present, and the ventilatory pattern and ventilatory support being used.

Hypoxaemia causes increased demand on the cardiovascular system to deliver higher cardiac output (CO) to sustain a constant O2 delivery. Spontaneous inspiratory efforts during acute bronchospasm and acute lung injury (ALI) induce marked/prominent negative swings in intrathoracic pressures (ITP) as the muscles of inspiration try to increase lung volume against increased airflow resistance or stiff lungs, respectively. Both of which also increase the work of breathing. Both lung under- and hyperinflation will increase pulmonary vascular resistance directly impeding right ventricular (RV) ejection. Lung hyperinflation also limits diastolic filling. Furthermore, spontaneous inspiratory efforts will decrease ITP, increasing venous return and intrathoracic blood volume, whereas positive-pressure ventilation by increasing ITP will have the opposite effect. These specific processes explain all the relevant determinants of heart–lung interaction.

Thus, heart–lung interactions involve four basic concepts:

  • Inspiration increases lung volume above the end-expiratory volume.

  • Spontaneous inspiration decreases ITP.

  • Positive-pressure ventilation increases ITP.

  • Spontaneous ventilation is exercise.

Haemodynamic effects of changes in lung volume

Lung inflation alters autonomic tone, pulmonary vascular resistance, and at high lung volumes, compresses the heart in the cardiac fossa. The associated diaphragmatic descent also increases intra-abdominal pressure and compresses the liver increasing hepatic vascular resistance. Each of these processes may predominate in determining the final cardiovascular state. Small tidal volumes (<10 mL/kg) increase heart rate (HR) by vagal (parasympathetic) withdrawal, causing an inspiration-associated cardiac acceleration called respiratory sinus arrhythmia. Whereas large tidal volumes (>15 mL/kg) decrease HR, arterial tone, and cardiac contractility by sympathetic withdrawal.

The haemodynamic response to increases in lung volume are mainly mechanical [1]‌. Lung inflation, independent of changes in ITP, primarily affects cardiac function and CO by altering RV preload and afterload, and left ventricular (LV) preload. First, inspiration induces diaphragmatic descent that increases hepatic outflow resistance, while simultaneously increasing intra-abdominal pressure. Systemic venous return to the heart is a function of the pressure difference between the right atrium and the systemic venous reservoirs and the resistance to venous return. Since a large proportion of the venous blood volume is in the abdomen, increases in intra-abdominal pressure will increase the venous pressure in this vascular space augmenting venous blood flow [2]. However, diaphragmatic descent will compress the liver increasing hepatic outflow resistance thus decreasing flow from the splanchnic venous reservoirs to the right heart. Complicating this further, inspiration will shift venous flow from high resistance splanchnic circuits, which must drain through the liver, to low resistance systemic venous circuits, making flow greater for the same driving pressure. Thus, inspiration may increase, decrease, or not alter venous return depending on which of these factors are predominant. Inspiration will increase venous return in volume overloaded states, whereas in hypovolaemic states and with hepatic cirrhosis, the same inspiratory effect will decrease venous return.

RV output is sensitive to changes in pulmonary outflow resistance. Alveolar collapse occurs in ALI and is associated with increased pulmonary vasomotor tone due to hypoxic pulmonary vasoconstriction [3]‌. Alveolar recruitment by restoring end-expiratory lung volume back to functional residual capacity (FRC) and improving oxygenation often decreases hypoxic pulmonary vasoconstriction, thus decreasing RV outflow resistance. Increasing lung volume above FRC in increases RV outflow resistance [4] due to progressive increases in transpulmonary pressure (airway pressure relative to ITP) associated with increasing lung volume. Since the heart and great vessels exist in the thorax, and sense ITP as their surrounding pressure, increases in transpulmonary pressure will induce pulmonary vascular collapse if transpulmonary pressure exceeds pulmonary artery pressure [1]. Thus, hyperinflation increases pulmonary vascular resistance and impedes RV ejection. Using the smallest tidal volumes, least PEEP, and other protective lung ventilation strategies will also improve RV ejection.

LV end-diastolic volume can be altered by ventilatory changes in three ways. First, since the RV and LV outputs are in series, changes in RV preload will eventually alter LV preload. Secondly, by ventricular interdependence changes, RV end-diastolic volume inversely changes LV diastolic compliance [5]‌. Ventricular interdependence is a major factor in altering LV output during spontaneous ventilation when RV end-diastolic volumes may vary widely from expiration (small volumes) to inspiration (large volumes). Thirdly, increasing lung volume restricts absolute cardiac volume by directly compressing the heart [1]. As the lungs expand, the heart is compressed in the cardiac fossa and absolute bi-ventricular volume is limited in a fashion analogous to cardiac tamponade.

Haemodynamic effects of changes in intrathoracic pressure

The heart within the thorax is a pressure chamber within a pressure chamber. Thus, changes in ITP will affect the pressure gradients for both systemic venous return to the RV and systemic outflow from the LV, independent of the heart itself [1,4]. Increases in ITP, by both increasing Pra and decreasing transmural LV systolic pressure, will reduce these pressure gradients, decreasing intrathoracic blood volume. Whereas decreases in ITP, using the same argument, will augment venous return and impede LV ejection, increasing intrathoracic blood volume. Variations in Pra represent the major factor determining the fluctuation in pressure gradient for systemic venous return during ventilation [2]‌. Increases in ITP, as seen with positive-pressure ventilation or hyperinflation, by increasing Pra decrease venous return, whereas decreases in ITP, as seen with usual spontaneous inspiration, by decreasing Pra increase venous return.

Spontaneous inspiratory efforts by decreasing ITP both increases lung volume and decreases right atrial pressure accelerating blood flow into the RV [6]‌ and increasing pulmonary blood flow on the subsequent beat. Thus, normal respiration-associated haemodynamic changes maximize ventilation–perfusion temporal matching because inspiration matches increased alveolar capillary flow. However, this venous flow augmentation is limited because if transmural vascular pressure falls below zero the extrathoracic veins collapse at the thoracic inlet, limiting flow [7]. This ‘flow-limitation’ is useful, because ITP can decrease greatly with obstructive inspiratory efforts. Without this flow-limitation, the RV could overdistend and fail.

Thus, we can see that spontaneous ventilatory efforts performed against a resistive (bronchospasm) or elastic (ALI) load, decrease LV stroke volume via a complex mechanism collectively called pulsus paradoxus that decreases LV end-diastolic volume and LV ejection. Transient intraventricular septal shift into the LV lumen by the dilated right ventricle plus pericardial volume restraint decreases absolute LV end-diastolic volume [1,5]. Furthermore, increases in LV afterload (LV ejection pressure minus ITP) increase LV end-systolic volume [8]‌.

LV afterload is approximated by maximal systolic wall tension, which is proportional to the product of transmural LV pressure and LV volume. Since increasing ITP will mechanically decrease transmural LV pressure, if arterial pressure is constant, increases in ITP will unload the LV, whereas decreases in ITP have the opposite effect [8]‌. Thus, in ventricular failure due to fluid resuscitation, increases in ITP may increase CO by decreasing LV afterload [8,9].

Sudden increases in ITP increase arterial pressure to an amount equal to the increase in ITP without changing aortic blood flow. If the increase in ITP is sustained, however, then the ITP-induced decrease in systemic venous return will eventually decrease LV output, decreasing arterial pressure. In the steady state, changes in ITP that result in altered CO also alter peripheral vasomotor tone through baroreceptor mechanisms. Baroreceptor reflexes tend to keep arterial pressure and CO constant. Thus, if ITP increases arterial pressure without changing transmural arterial pressure, then the periphery would vasodilate to maintain a constant extrathoracic arterial pressure-flow relation [4]‌. Since coronary perfusion pressure is not increased by ITP-induced increases in arterial pressure, whereas mechanical constraint from the expanding lungs may obstruct coronary blood flow, coronary hypoperfusion from a combined coronary compression, and a decrease in coronary perfusion pressure is a potential complication of increased ITP.

Although increases in ITP should augment LV ejection by decreasing LV afterload, this effect has limited therapeutic potential, just as all afterload reducing therapies are limited by both the minimal end-systolic volume and the obligatory decrease in venous return. Thus, the potential augmentation of LV ejection by increasing ITP is limited because increasing ITP, by reducing LV ejection pressure, can only decrease end-systolic volume, which is usually already small and cannot decrease much more except in markedly dilated cardiomyopathies. However, the decrease in venous return associated with the increase in ITP can continue to total circulatory arrest.

From the perspective of the ejecting LV, there is no difference between increasing ITP from a basal end-expiratory level and eliminating negative end-inspiratory ITP swings seen in spontaneous ventilation. Removing negative swings in ITP may be more clinically relevant than increasing ITP for many reasons. First, many pulmonary diseases are associated with exaggerated decreases in ITP during inspiration. In restrictive lung disease states, such as interstitial fibrosis or acute hypoxaemic respiratory failure, ITP must decrease greatly to generate a large enough transpulmonary pressure to ventilate the alveoli. Similarly, in obstructive diseases, such as upper airway obstruction or asthma, large decreases in ITP occur owing to increased resistance to inspiratory airflow. Secondly, exaggerated decreases in ITP require increased respiratory efforts that increase the work of breathing, taxing a potentially stressed circulation. Finally, the exaggerated decreases in ITP can only increase venous blood flow increasing intrathoracic blood volume. The level to which ITP must decrease to induce venous flow-limitation is different in different circulatory conditions but occurs in most patients below an ITP of –10 cmH2O [1]‌. Thus, further decreases in ITP will further increase only LV afterload without increasing venous return. Accordingly, abolishing these markedly negative swings in ITP should disproportionally reduce LV afterload more than venous return (LV preload). These concepts of a differential effect of increasing and decreasing ITP on cardiac function are illustrated for both normal and failing hearts in Figs 87.1 and 87.2 using the LV pressure-volume relationship during one cardiac cycle to interpose venous return (end-diastolic volume) and afterload (end-systolic volume). Using this logic, one would predict that by endotracheal intubation and ventilation in patients requiring markedly negative swings in ITP to breath will abolish the increased LV afterload without impairing systemic venous return. These interactions have important implications in the decision to both institute and withdraw mechanical ventilatory support, as in acute cardiogenic pulmonary oedema.

Fig. 87.1 The effect of increasing (dark shading) and decreasing (no shading) intrathoracic pressure (ITP) on the left ventricular (LV) relation with LV contractility is normal. The slope of the LV end-systolic pressure volume relationship (ESPVR) is proportional to contractility. The slope of the diastolic LV pressure-volume relationship defines diastolic compliance.

Fig. 87.1 The effect of increasing (dark shading) and decreasing (no shading) intrathoracic pressure (ITP) on the left ventricular (LV) relation with LV contractility is normal. The slope of the LV end-systolic pressure volume relationship (ESPVR) is proportional to contractility. The slope of the diastolic LV pressure-volume relationship defines diastolic compliance.

Fig. 87.2 The effect of increasing (dark shading) and decreasing (no shading) intrathoracic pressure (ITP) on the left ventricular (LV) relation in congestive heart failure when LV contractility is reduced and intravascular volume is expanded. The slope of the LV ESPVR is proportional to contractility. The slope of the diastolic LV pressure–volume relationship defines diastolic compliance.

Fig. 87.2 The effect of increasing (dark shading) and decreasing (no shading) intrathoracic pressure (ITP) on the left ventricular (LV) relation in congestive heart failure when LV contractility is reduced and intravascular volume is expanded. The slope of the LV ESPVR is proportional to contractility. The slope of the diastolic LV pressure–volume relationship defines diastolic compliance.

Ventilation as exercise

Spontaneous ventilatory efforts require muscular activity, consume O2 and produce CO2, they represent a metabolic load on the cardiovascular system. Although ventilation normally requires less than 5% of total O2 delivery to meet its demand, in lung disease states where the work of breathing is increased, such as pulmonary oedema or bronchospasm, the requirements for O2 may increase to 25% or more of total O2 delivery [10]. Furthermore, if cardiac output is limited, then spontaneous ventilation may not be possible without additional cardiovascular support. The institution of mechanical ventilation for ventilatory and hypoxaemic respiratory failure may reduce metabolic demand on the stressed cardiovascular system, decrease O2 consumption, and thus for the same cardiac output, mixed venous oxygen saturation (SvO2) will increase. In patients with right-to-left intrapulmonary shunts, this increased SvO2 will increase PaO2 independent of changes in ventilatory status. Intubation and mechanical ventilation, when adjusted to the metabolic demands of the patient, may dramatically decrease the work of breathing, resulting in increased O2 delivery to other vital organs.

Functional haemodynamic monitoring

Positive pressure inspiration by increasing ITP passively increases right atrial pressure, venous return transiently decreases. This dynamic flow variation will alter both RV filling and RV output, if the RV is volume responsive and then after a short pulmonary transient time alter LV filling and LV output if the LV is volume responsive. Numerous studies have documented that the associated LV stroke volume variation (SVV) and arterial pulse pressure variation (PPV) will vary in direct proportion to volume responsiveness and can be used for clinical decision making about fluid therapy in ARF [11,12,13]. A PPV >13% or a SVV >10% on positive-pressure ventilation is highly predictive of volume responsiveness and the accuracy of these variations was excellent in predicting volume responsiveness (PPV 0.94 (receiving operation curve), systolic pressure variation (SPV) 0.86, and SVV 0.84) with PPV significantly better than SPV or SVV (p < 0.001) [14].

If chest wall compliance is normal, however, a minimal tidal volume (VT) of 8 mg/kg is required to cause a large enough swing in ITP to induce theses effects. Since lower VT ventilation is recommended in the management of patients with ALI, the ITP swings may not be large enough to allow PPV and SVV thresholds to be predictive [9]‌. Still, many patients with ALI also have increased intra-abdominal pressure owing to fluid resuscitation and ileus [15]. This increased intra-abdominal pressure decreases chest wall compliance, making the ITP swings during positive-pressure ventilation still predictive of volume responsiveness despite low VT breathing [16,17]. Therefore, PPV or SVV often has a lesser predictive value of identifying volume responsiveness when measured in patients receiving a VT <8 mL/kg [18,19], but if present signifies volume responsiveness.


1. Butler J. (1983). The heart is in ‘good hands’. Circulation, 67(6), 1163–8.Find this resource:

2. Fessler HE, Brower R, Wise RA, and Permutt S. (1992). Effects of positive end-expiratory pressure on the canine venous return curve. American Reviews of Respiratory Diseases, 146(1), 4–10.Find this resource:

3. Mitchell JR, Doig CJ, Whitelaw WA, Tyberg JV, and Belenkie I. (2011). Volume loading reduces pulmonary vascular resistance in ventilated animals with acute lung injury: evaluation of RV afterload. American Journal of Physiology, 300(3), R763–70.Find this resource:

4. Pinsky MR, Matuschak GM, and Klain M. (1985). Determinants of cardiac augmentation by elevations in intrathoracic pressure. Journal of Applied Physiology, 58(4), 1189–98.Find this resource:

5. Taylor RR, Corell JW, Sonnenblick EH, and Ross J, Jr. (1987). Dependence of ventricular distensibility on filling of the opposite ventricle. American Journal of Physiology, 213(3), 711–18.Find this resource:

6. Pinsky MR. (1984). Determinants of pulmonary artery flow variation during respiration. Journal of Applied Physiology, 56(5), 1237–45.Find this resource:

7. Guyton AC, Lindsey AW, Abernathy B, and Richardson T. (1957). Venous return at various right atrial pressures and the normal venous return curve. American Journal of Physiology, 189, 609–15.Find this resource:

8. Buda AJ, Pinsky MR, Ingles NB, Daughters GT, and Alderman EL. (1979). Effect of intrathoracic pressure on left ventricular performance. New England Journal of Medicine, 301(9), 453–9.Find this resource:

9. Renner J, Cavus E, Meybohm P, et al. (2007). Stroke volume variation during hemorrhage and after fluid loading: impact of different tidal volumes. Acta Anaesthesiologica Scandinavica, 51(5), 538–44.Find this resource:

10. Stock MC, David DW, Manning JW, and Rtyan ML. (1999). Lung mechanics and oxygen consumption during spontaneous ventilation and severe heart failure. Chest, 102(1), 279–83.Find this resource:

11. Cannesson M, Le Manach Y, Hofer CK, et al. (2011). Assessing the diagnostic accuracy of pulse pressure variations for the prediction of fluid responsiveness: a ‘gray zone’ approach. Anesthesiology, 115(2), 231–41.Find this resource:

12. Hofer CK, Muller SM, Furrer L, Klaghofer R, Genoni M, and Zollinger A. (2005). Stroke volume and pulse pressure variation for prediction of fluid responsiveness in patients undergoing off-pump coronary artery bypass grafting. Chest, 128(2), 848–54.Find this resource:

13. Perel A, Pizov R, and Cotev S. (1987). Systolic blood pressure variation is a sensitive indicator of hypovolemia in ventilated dogs subjected to graded hemorrhage. Anesthesiology, 67(4), 498–502.Find this resource:

14. Marik PE, Cavallazzi R, Vasu T, and Hirani A. (2009). Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature. Critical Care Medicine, 37(9), 2642–7.Find this resource:

15. Krebs J, Pelosi P, Tsagogiorgas C, Alb M, and Luecke T. (2009). Effects of positive end-expiratory pressure on respiratory function and hemodynamics in patients with acute respiratory failure with and without intra-abdominal hypertension: a pilot study. Critical Care, 13(5), R160.Find this resource:

16. Gunn SR, Kim HK, Harrigan P, and Pinsky MR. (2006). Ability of pulse contour and esophageal Doppler to estimate rapid changes in stroke volume. Intensive Care Medicine, 32(10), 1537–46.Find this resource:

17. da Silva Ramos FJ, de Oliveira EM, Park M, Schettino GP, and Azevedo LC. (2011). Heart–lung interactions with different ventilatory settings during acute lung injury and hypovolaemia: an experimental study. British Journal of Anaesthesia, 106(3), 394–402.Find this resource:

18. Vallee F, Richard JCM, Mari A, et al. (2009). Pulse pressure variations adjusted by alveolar driving pressure to assess fluid responsiveness. Intensive Care Medicine, 35(6), 1004–10.Find this resource:

19. Debacker D, Taccone FS, Holsten R, Ibrahimi F, and Vincent JL. (2009). Influence of respiratory rate on stroke volume variation in mechanically ventilated patients. Anesthesiology, 110(5), 1092–7.Find this resource: