Show Summary Details
Page of

Lung function and cardiopulmonary exercise testing 

Lung function and cardiopulmonary exercise testing

Stephen Chapman

, Grace Robinson

, John Stradling

, Sophie West

, and John Wrightson

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2021. 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: 26 October 2021

Flow–volume loop 1

A good start to understanding lung function tests is the flow–volume loop (see Fig. A6.1). This plots inspiratory and expiratory flow against lung volume during a maximal expiratory and maximal inspiratory manoeuvre.

At the beginning of expiration from a full breath in, the expiratory muscles are at their strongest, the lungs at their largest, and hence the airways are at their most open (A). Because the lungs are at their largest, the radial attachments to the airways, effectively the alveolar/capillary membranes and their connective tissue, are pulling the hardest and supporting the airways against dynamic compression during the exhalation manoeuvre.

This means that the highest flow rates are possible at the beginning of the blow, hence the sudden rise to a PEFR in the first 100ms or so of the forced breath out (B). This is the peak flow and is essentially what a peak flow meter measures (see Fig A6.5).

As the lung empties and the lung volume drops, the dilatory pull on the airways from the radial attachments of the surrounding lung tissue reduces (C). Hence, the airways narrow and become less supported and are less able to resist dynamic compression. This means that the maximal airflow obtainable, regardless of effort, falls too.

Eventually, the expiratory muscles come to the end of their ‘travel’ and cannot squeeze the chest anymore. Also, increasingly with age, the small airways may actually close off, preventing any more emptying (D). The volume at which this begins to happen is called the closing volume.

As maximal inspiration starts, although the inspiratory muscles are at their strongest, the airways are at their smallest. Thus, flow rates start low and increase as the airways open up. However, as the lung expands, the inspiratory muscles are approaching the end of their ‘travel’ and are weakening; this means the flow rates fall again, hence the different rounded appearance of the inspiratory limb of the flow–volume curve.

Thus, normally, the inspiratory and expiratory flow rates depend on lung volume and are termed ‘volume-dependent’. If there is a fixed upper airway narrowing, such as from a solid hard tumour partially blocking the trachea, then the size of the airway at this point may become so narrow that it now limits maximal flows. However, its diameter will vary very little with lung volumes, and hence flow will become ‘volume-independent’. Fig. A6.2 shows this.

At (A), the rise in flow will initially be normal, but, at some point, the maximum flow imposed by the upper airway narrowing will cut in (B). From that point onwards, the flow rate will be fixed at this maximum (C) until, at much lower lung volumes, the lower recoil and narrowing of the small airways again determine the maximum flow (D). The flow–volume curve has been severely ‘clipped’, with a square-ish appearance. The same clipped appearance will be present on the inspiratory limb (E), giving rise to the so-called ‘square box’ appearance.

Sometimes, such upper airways restriction may be variable, rather than fixed, and only obstruct during inspiration (e.g. paralysed and collapsing vocal cords), due to the obstructing elements being sucked in and then blown open again on expiration.

  • Thus, a square inspiratory limb, but normal expiratory limb, provides evidence of a mobile extrathoracic upper airways obstruction.

Conversely, a mobile intrathoracic upper airway obstruction (e.g. soft fleshy tumour at the carina or retrosternal thyroid) may obstruct more during expiration (when the expiratory effort is compressing the lung), compared with inspiration when the chest is being expanded.

  • Thus, a square expiratory limb, but normal inspiratory limb, is evidence of a variable intrathoracic upper airways obstruction.

Sometimes, ratios of maximal inspiratory to maximal expiratory flows are used to characterize the intra- or extrathoracic airway obstruction.

Flow–volume loop 2

The other, more common, causes of airway obstruction are due to narrowing of the lower airways (asthma, COPD). In these conditions, the airway calibre (and thus flow rates) still remains dependent on lung volume. Hence, the flow rates decrease, as the lung volume decreases, but particularly decrease at low lung volumes. This is because resistance to flow is proportional to the airway radius raised to the power of 4 (r4) and therefore most significant when airways are already small. Hence, increasing airflow obstruction produces expiratory flow–volume curves like those in Fig. A6.3. This greater effect of small airways narrowing at low lung volumes has led some to report flow rates at, for example, 25% expired lung volume or averaged between 25% and 75% of the total expired lung volume.

Fig. A6.3 Expiratory flow–volume curves—lower airways obstruction.

Fig. A6.3 Expiratory flow–volume curves—lower airways obstruction.

Airways can be so small that, during expiration, they begin closing off earlier than normal (the closing volume); hence, a full breath out is not possible, producing air trapping and a raised RV (A). Sensitive tests of small airways narrowing have to concentrate on flows at low lung volumes, and peak flow measurements are relatively insensitive.

  • Note, however, that peak flow measurements are the most sensitive to upper airway narrowing and a good way to follow changes in upper airway narrowing during, for example, radiotherapy for a central airway obstructing lung cancer.

Spirometry, peak flow measurements, and CO transfer

  • The ordinary spirometer (mechanical or electronic) records volume against time, rather than flow against volume

  • The two essential measures are FEV1 and VC.

The VC is the maximum amount of air that can be blown out completely. This will be reduced if the lungs are stiff (preventing a full breath in), the inspiratory muscles are weak (preventing a full breath in), or the airways are narrowed such that the small airways collapse during expiration (preventing a full breath out).

The FEV1 is the amount (forced expiratory volume) that can be blown out in 1s. Because the value is taken over a second, a much longer period of flow is being captured during the breath out than the PEFR, but, despite this, the measurement is still being made when the airways are larger. It is less dependent on effort and generally more robust. The ratio of the two figures (FEV1/VC) tells us about the degree of airflow obstruction.

  • A ratio of FEV1/VC of less than about 70–75% indicates airflow obstruction.

This ratio is very useful because it is hardly affected by age, sex, height, ethnic origin, etc.—it is self-normalizing. The individual measures of FEV1 and VC do need corrections for the above factors and are usually quoted as % predicted. The range of normality is considerable, and it may not be clear if results are simply at the bottom end of normal or considerably reduced from the patient’s normally much higher figures. Serial measurements indicating continuing deterioration may be the first clue. Only if the FEV1/VC ratio is normal, can the VC be confidently used to infer whether there is a reduced total lung volume such as from ILD. A low VC with normal FEV1/VC ratio is called a restrictive pattern. A low FEV1/VC ratio is called an obstructive pattern, and a reduced VC cannot then be confidently used to infer that the total lung volume is also reduced and indeed may even be increased because of air trapping. A small print fact is that the FEV1/VC ratio may actually be raised in ILD, as the airways are better supported by the fibrosed radial attachments, which reduces dynamic compression, thus increasing expiratory flow, compared to that expected for the lung volume.

The slope of the volume–time plot from a spirometer is effectively the flow at any particular point; because flow is dropping during expiration, the slope progressively flattens off. However, if there is any fixed upper airways obstruction (as previously discussed), the expiratory flow rate will be constant for a while, and hence the spirometer line will be straighter than usual. An interesting index to detect possible upper airway obstruction the Empey index has been described:

Empey Index ≈ [FEV1 (mL) / PEFR (L/min)]

Because PEFR is clipped first by the presence of upper airflow obstruction, relative to the FEV1, the above index gets larger with such a problem. A figure over 10 is suggestive of upper airflow obstruction, but it is only a pointer, and there will be false positives and negatives.

Although one-off measures of lung function can be made, more interesting information comes from serial measurements, e.g. in asthma, PEFR will fluctuate with characteristic morning ‘dips’.

Spirometry—how to do it

Everyone has their own way to do spirometry (see Table A6.1), but this is a way that works for the authors. Say to the patient, ‘This is a test of how big your lungs are and how fast you can empty them. What I would like you to do is take an enormous breath in, the biggest you can manage, then seal your lips around the tube, and blow as hard and as fast and as long as you possibly can’. Then demonstrate the manoeuvre yourself with a spare tube (not necessarily connected to the spirometer) so that they can then mimic it. Whilst they are blowing, say ‘excellent, well done, keep blowing, come on, come on, come on, keep blowing’. There are various recommendations as to numbers of blow, etc. these are the arguments.

Table A6.1 Spirometer tips

Standing or sitting?

High intrathoracic pressure generated may cause the patient to pass out. Therefore, sitting down is safer, but better and more consistent figures are obtained standing. Have a chair behind patient to sit on, if dizzy.

Nose clip?

Prevents escape through the nose which would give falsely low figures but is uncomfortable. Vast majority of patients do not need it, but, if the line appears to fall off towards the end of the blow or inconsistent volumes, try a nose clip.

Best of three blows?

Needed to demonstrate that maximal blow has been consistently achieved, usually by seeing two identical tracings. A device showing the actual spirometer tracing is extremely helpful. If two tracings are identical, this is probably enough but may need more if blowing is erratic, until satisfied it is maximal.

Keep going to end of page (6s) or not?

Needed to establish correct VC. Therefore, if line still rising, then VC not reached, but have to stop somewhere. In restrictive disease or normals, usually maximal by 3s; in obstructive disease, may need to do a slow VC to establish ‘real’ value, which can be much larger than the forced VC (FVC) due to dynamic compression.

Repeat with a submaximal effort?

Similar to above, it is sometimes useful to ask patient to repeat the expiratory manoeuvre with slightly less effort, when emphysema suspected. This will often give better expiratory volume in 1s and the VC, which would suggest major dynamic airways compression.

CO transfer

Usually done in the lung function laboratory, and it essentially measures the amount of gas-exchanging surface area available. A gas mixture containing CO is inhaled, the breath held for 10s, and then exhaled. The amount of CO that has disappeared (by crossing the alveolar capillary membrane and being taken up by red cells) is calculated. A correction for the Hb concentration is required, as the amount of CO transferred will fall as the available Hb is reduced. The total amount of CO transferred is the TLCO (total lung, TL). When divided by the total lung volume during the breath-hold, it is called the kCO, gas transfer per unit lung volume. The total lung volume ‘reached’ by the CO is the amount breathed in plus the amount of air already in the lung at the start of the breath in. This is measured by including helium in the inhaled gas mixture that is diluted by the air already in the lung; by comparing inspired helium concentration with expired, this total lung volume can be calculated.

  • The TLCO and kCO are reduced most in emphysema when alveoli have been destroyed

  • The TLCO and kCO are reduced in ILD where the alveolar capillary membrane may be thick enough to reduce CO passage

  • The kCO may also be raised (and TLCO normal) when lungs are poorly expanded by, say, weak respiratory muscles, because the lung is ‘more concentrated’ and transfers CO better when quoted per unit volume

  • The kCO may also be raised for a few days when there has been profuse lung haemorrhage, as can occur in, e.g. SLE, Wegener’s, and Goodpasture’s. This is because the free red cells lining the alveoli take up CO directly and ‘falsely’ elevate the figure. As the Hb is broken down, the kCO returns to normal, unless there is another bleed. This helps to distinguish re-bleeding from other causes of lung infiltrates such as infection.

This test requires more cooperation than simple spirometry, as well as a minimum inspired volume, and therefore cannot always be obtained.

Respiratory muscle function, body plethysmography, and lung volumes

Respiratory muscle function

Respiratory failure and small VCs may be due to weak respiratory muscles. It is therefore useful to be able to assess inspiratory and expiratory muscle power. There may be global weakness or specific inspiratory weakness, usually due to diaphragm paralysis. In the clinic, the simplest test is a lying and standing VC. If the diaphragm is paralysed, then, on lying down, the abdominal contents will push up the diaphragm and limit inspiration. On standing, the abdominal contents drop and aid inspiration.

  • A fall in VC of <10% on lying down is probably normal

  • A fall of 10–20% is suspicious of diaphragm paralysis

  • A fall of >20% is abnormal and suggests significant, usually bilateral, diaphragm paralysis.

In the laboratory, there are various ways to test respiratory muscle function. The patient can blow against a pressure meter after a maximum inspiration and inspire against the meter after a full expiration. This is, of course, highly effort-dependent. A manoeuvre, such as a sniff, is very stereotypic, and patients can reproduce this. Measuring the inspiratory pressure produced at the nose during this manoeuvre is a rough and ready way of screening for inspiratory muscle weakness. More accurate assessments of inspiratory muscle function, particularly the diaphragm, can be obtained using two semi-inflated balloons, introduced via the mouth and oesophagus, placed above and below the diaphragm, and connected to pressure transducers. The transdiaphragmatic pressures during maximal inspiratory efforts, sniffing, and breathing to TLC all provide reproducible measures of diaphragm function but depend on good cooperation and effort by the patient. Activating the phrenic nerve directly with a superficial electrical stimulator, or by using high-intensity magnetic stimulation over the nerve roots of C3–C5, whilst measuring transdiaphragmatic pressures, provides a non-effort-dependent way to test diaphragm function.

Body plethysmography

requires the patient to climb into an airtight cabinet and breathe through a shuttered mouthpiece connected to the outside world. It has two particular advantages over simpler lung function tests. It is able to measure the total lung volume or capacity (TLC) in the thorax, and it provides a measure of airways obstruction, involving little or no effort by the patient.

The other main method of measuring TLC involves helium dilution, as described during the TLCO measurement above. However, in the presence of lower airways obstruction, the helium may not ‘reach’ all parts of the lung during the 10s breath-hold, and the volume calculated from this dilution will therefore be lower than the real TLC. The body plethysmograph relies on the pressure changes that occur when all the air in the chest is alternately compressed and expanded by the patient making breathing efforts against an airway closed by a shutter at the mouth. The pressure changes produced in the oral cavity vs those in the box are then proportional to the volume of air being compressed and rarefied, thus allowing calculation of the volume in the chest at the time. Note that this volume will include any bullae or pneumothorax, and the difference between the plethysmographic lung volume and the helium dilution volume will reflect the bullae/pneumothorax volumes, as well as areas not reached by the helium due to increased airways resistance.

Measurement of airways resistance with the body box relies on a similar principle. If there were no airways resistance, then breathing in and out would not compress or rarefy the air in the chest. With increasing resistance, the air in the chest will be compressed during expiration and rarefied on inspiration. It is this phenomenon that allows calculation of the airways resistance during quiet breathing or panting (the latter ensures the vocal cords are fully open and not contributing to the measured resistance). See Fig. A6.4 and Table A6.2.

Fig. A6.4 Lung volumes in normal, obstructive, and restrictive lung conditions. TLC, total lung capacity (not always increased when obstructive pattern); VC, vital capacity; RV, residual volume; FRC, functional residual capacity; FEV1, forced expiratory volume in 1s.

Fig. A6.4 Lung volumes in normal, obstructive, and restrictive lung conditions. TLC, total lung capacity (not always increased when obstructive pattern); VC, vital capacity; RV, residual volume; FRC, functional residual capacity; FEV1, forced expiratory volume in 1s.

Table A6.2 Lung volume patterns




FEV1 (% predicted)


VC (% predicted)

↓ or →

FEV1/VC ratio

→ or ↑ (increased recoil)

TLC (% predicted)

↑ or →

RV (% predicted)

FRC (% predicted)

RV/TLC ratio

↑ (gas trapping)

→ or ↓

Further information

Gibson GJ. Clinical tests of respiratory function, 3rd edn. Hodder Arnold, 2008.Find this resource:

Peak flow reference ranges

See Fig. A6.5.

Fig. A6.5 Normal values for peak flow, based on original Gregg and Nunn values (BMJ 1973) but corrected for new EU scale peak flow meters. Normal range extends about ± 15% or roughly 100L/min in men and 50L/min in women.

Fig. A6.5 Normal values for peak flow, based on original Gregg and Nunn values (BMJ 1973) but corrected for new EU scale peak flow meters. Normal range extends about ± 15% or roughly 100L/min in men and 50L/min in women.

Note: old Wright peak flow meters over-read in the middle of the scale (e.g. reading about 400 when actual value was 350L/min) and were replaced from October 2004 by a corrected scale.

Cardiopulmonary exercise testing (CPET)

General points

  • An exercise test, with additional measurement of ventilatory gases

  • Useful for assessing:

    • Cardiorespiratory fitness

    • Relative contribution of cardiovascular (CV) and respiratory (RS) disease to exercise limitation

    • Disease severity assessment and prognostication

    • Risk stratification pre-lung resection, lung and cardiac transplantation, cardiac medical device therapy, other surgical evaluations

    • Response to an intervention.

Undertaking CPET

  • Baseline spirometry and maximum voluntary ventilation (MVV)

  • Measurements include serial ECGs, SaO2, BP, HR, expired concentrations of O2 and CO2, tidal volume, and breathing frequency (via face mask or mouthpiece with nose clips)

  • Calculation of VE (minute ventilation), VO2 (O2 consumption), VCO2 (CO2 production) at rest and throughout exercise

  • Ventilatory threshold represents a point where several ventilatory parameters show a threshold-like behaviour, related to the onset of anaerobic respiration and lactic acidosis. Many ways of estimating; look for the inflection point on plots of VCO2 vs VO2 or VE vs VCO2

  • ABGs (via an arterial line) are occasionally taken

  • Usually performed on a cycle ergometer (alternatively, a treadmill)

  • Cycle is initially unloaded, and then work is ramped (based on usual activity level; pre-test FEV1 and MVV results can also be helpful)

  • Doctor and physiologist monitor patient during test and ensure mask/mouthpiece closely fitting. Patient encouragement during testing enhances performance and can make results more meaningful

  • Test usually stopped due to exhaustion (e.g. tired legs or too dyspnoeic). Stop immediately with significant arrhythmias, ST depression ≥2mm, heart block, significantly falling BP, ischaemic-sounding chest pain, severe symptomatic hypoxaemia, or near syncope

  • Unloaded pedalling at end of test.

Interpreting variables

  • See Table A6.3.

Table A6.3 Interpreting CPET variables



Normal values

Peak VO2 (mL/kg/min) (or VO2 max)

Maximum O2 utilization Global prognostic marker/severity assessment. Influenced by CV, RS, and muscular function

Influenced by age and sex (15–80mL/kg/min)Reported as % predicted

VO2 at ventilatory threshold (VT)(mL/kg/min)

Associated with anaerobic threshold. Limit of workload sustainable for prolonged periods

± 50–65% peak VO2 Influenced by training and genetic predisposition

Peak respiratory exchange ratio (RER)

RER = VCO2/VO2 ratio ↑work → ↑VO2 but ↑↑VCO2 → ↑RER Marker of effort during exercise

Good effort suggested by RER ≥1.1

VE/VCO2 slope (VE on y-axis; VCO2 on x-axis)

Determined by V/Q matching Marker of disease severity

<30 normal. Particularly high in PHT

VE/VO2 at peak exercise

Ventilatory cost of O2 uptake at peak exercise

≤40 normal

End-tidal CO2 partial pressure PETCO2

Determined by V/Q matching and cardiac function Marker of disease severity

Rest: 4.8–5.6kPa At VT: ↑ 0.4–1.1kPa Above VT: ↓ due to increased ventilatory response to metabolic acidosis

VE/MVV VE measured at peak exercise, MVV at rest

Helps determine if dyspnoea is related to a pulmonary cause

≤0.8 normal >0.8 suggests pulmonary limitation

O2 pulse (mL O2/beat)

O2 pulse = VO2/HR Surrogate for stroke volume response to exercise Helpful for assessing possible myocardial ischaemia

Normally rises during exercise, reduced rise in chronic heart failure

Change in VO2/change in workload (∆VO2/∆W)

Helpful for assessing possible myocardial ischaemia

Normally linear rise of VO2 with work Average 10mL/min/W

FEV1 and PEF (L/min)

Compare pre- and post-exercise Changes suggest respiratory cause of dyspnoea (not asthma-specific)

<15% reduction with exercise

Heart rate recovery (HRR)

Compares max HR with HR after 1min recovery. Related to parasympathetic activation

Normally HRR >12 beats lower at 1min. If <12, suggests cardiac cause

Exercise BP

CV response to exercise

Usual to have modest increase in SBP with exercise. DBP usually static or decreases due to vasodilatation


Useful for assessing for respiratory causes of exertional dyspnoea

Should not fall >5%. Greater fall in RS disease and PHT

Patterns in unexplained exertional dyspnoea

  • Decreased VO2 peak/max—defines degree of impairment, independent of mechanism

  • VE/VCO2 slope increase and PETCO2 decrease—consider causes of exercise-related PHT

  • SaO2 fall—suggests V/Q mismatching

  • FEV1 or PEF decreases, VE/MVV increases—suggests respiratory cause

  • Lack of ventilatory threshold—suggests cause related to ventilation.

Further information

Guazzi M et al. Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Eur Heart J 2012;33:2917–27.Find this resource: