Ref erence Values

Cole Bowerman, MSc, MDa,b,*, Sanja Stanojevic, PhDb

KEYWORDS
• Pulmonary function test • Spirometry • Lung function • Reference values • Normal range

KEY POINTS
• Reference equations act as growth charts for the lungs in that they facilitate comparison to a
healthy standard population.
• The Global Lung Function initiative has produced the currently recommended equations for spirom-
etry, transfer capacity of the lung for carbon monoxide, and lung volumes.
• Reference equations do not consider the social determinants of health, which may introduce bias in
interpretation, especially for marginalized populations.

INTRODUCTION Defining a Healthy Reference Population

Measures of lung function, most commonly from Lung size and function are determined by both
spirometry, can provide valuable information to biological and environmental factors. Demo-
support clinical decision-making. The values graphic factors such as height, biological sex,
measured by the device, in liters, are difficult to and age are key determinants of lung size and
interpret alone. For instance, a measured value function and are routinely incorporated into refer-
of 2.5 L can reflect what is expected for a 70- ence equations, which are essentially growth
year-old female (170 cm) but would reflect charts for lung function. The characteristics of
impaired lung function for a 20-year-old male of the healthy population used to derive reference
the same height. It is therefore necessary to equations can have an important influence of the
compare measured values to what is expected in range of expected values.
otherwise healthy individuals of similar height, bio- To be generalizable, the healthy population
logical sex, and age. Pediatric growth charts should be representative of individuals across all
are an example of how a child’s measured value ages and heights and represent individuals without
(eg, weight or height) are compared to what can an exposure history that could impair lung func-
be expected of other healthy children of the tion. An obvious example is tobacco cigarette
same age. The growth chart provides a range of exposure.1 The healthy population selected to
expected values to facilitate interpretation and create a reference equation would not include indi-
clinical decision-making. In respiratory medicine, viduals with tobacco cigarette exposure; however,
measured values are similarly compared to a refer- individuals with secondhand tobacco exposure
ence range derived from healthy individuals. If an and other environmental exposures may be
individual’s measured values are far below (or in included, which can influence interpretation. The
some cases higher) than what is expected for a factors that influence lung growth and develop-
healthy individual of similar height, biological sex, ment, and lung function itself are complex and
and age, there is a greater probability of underlying multifactorial, and in some cases intergenera-
pathophysiology and need for further testing and/ tional,2 which makes it challenging to define a
or early intervention. healthy population that represents ideal lung
[Link]

a
Department of Medicine, Queen’s University, Kingston, Canada; b Department of Community Health and
Epidemiology, Dalhousie University, Centre for Clinical Research, 5790 University Avenue, Halifax, NSB3H
1V7, Canada
* Corresponding author. Core Internal Medicine Program, 94 Stuart Street, Room 3022, Kingston, ON K7L 2V6,
Canada.
E-mail address: [Link]@[Link]

Clin Chest Med 46 (2025) 569–578

[Link]
0272-5231/25/© 2025 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training,
and similar technologies.
Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on
September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
570 Bowerman & Stanojevic

Abbreviations in utero and in childhood,7 the size of the central

airways,8 and the number of alveoli and bronchi-
ATS American Thoracic Society oles even after accounting for lung size.9 Function-
FEV1 forced expiratory volume in 1 second ally, sex hormones such as progesterone and
FVC forced vital capacity testosterone influence bronchial contractility and
GLI global lung function initiative relaxation10 as well as inspiratory and expiratory
NHANES national health and nutrition pressures, which can alter lung function between
examination study
the sexes.11 The role of the social concept of
gender is yet to be fully understood and is not
routinely incorporated into interpretation at this
health. Furthermore, the determinants of lung
time. Interpretation of lung function in transgender
health can change over time, and an individual
individuals, particularly those who have had
identified as healthy today may later go on to
gender reaffirming surgery or hormone therapy,
develop lung disease. The concept of health has
is an evolving area of respiratory medicine. It is
historically been considered as the absence of dis-
important for clinicians to understand and respect
ease, whereas more contemporary approaches
an individual’s gender identity and to explain the
acknowledge the importance of the social deter-
role of biological sex in determining lung size and
minants of health. Therefore, unlike other clinical
consequently how lung function measurements
decision tools, reference equations for lung func-
are interpreted.
tion represent a summary of an otherwise healthy
Lung function also follows a unique pattern with
population at 1 point in time and the inclusion of
age (Fig. 1). During childhood, the lungs grow
people at risk of lung disease, or those with lifetime
rapidly and often synchronously with somatic
exposures that are not excluded, may be included
growth until lung growth and development peaks
in the healthy population and influence the accu-
in the mid 20’s. Male lungs typically continue to
racy of interpretation.
grow until age 25 years, whereas female lung
growth peaks at age 20 years.12,13 During periods
Known Determinants of Lung Volumes/
of rapid growth and development (eg, puberty),
Function
parenchymal lung growth typically precedes
Intuitively, people with larger thoraxes will have airway growth in a process known as dysanapsis.
larger lungs, and thus would be expected to This is particularly evident in males where the air-
have larger lung volumes. Chest dimensions are ways develop slower than the lung paren-
difficult to measure; therefore, standing height is chyma.13,14 Sex-specific hormones play an
used as a proxy of chest size. Height is the single important role in lung development during puberty.
most important predictor of lung volume.3,4 The Rapid somatic growth (height) as well as matura-
general approximation is that lung volume is a tion and growth of the lung parenchyma itself influ-
function of height cubed, or more precisely ence final lung size. Following early adulthood,
height2.4.5 Depending on a range of genetic, envi- lung function is understood to gradually decline
ronmental, and epigenetic factors, 2 individuals of in later adulthood.15 To a certain extent, loss of
the same standing height can have quite different elasticity of the lung parenchyma, due to aging,
chest sizes and thus lung volumes. Height is only leads to lung function decline.16 An accumulation
a proxy for chest size, which may not have the of exposures over the life course may also
same relationship (ie, correlation) between individ- contribute. While the degree of lung function
uals or groups of individuals. In individuals for decline due to aging is not fully understood, there
whom standing height does not reflect the ex- is consensus that both genetic and environmental
pected chest size (ie, individuals with shorter factors can influence the rate of lung function
legs and longer thoraxes) there will be greater decline in adulthood.17,18
measurement error and thus reduced accuracy in The combination of standing height, biological
estimating lung function. sex, and age are recognized as key determinants
During childhood, differential influence of sex of lung size and lung volume. Thus, these determi-
hormones, as well as anatomic differences in nants are used to create reference equations/
chest shape between males and females (biolog- growth charts lung function.
ical sex at birth) influence lung volumes. For the For an individual of a measured height, biolog-
same standing height, males and females have ical sex, and age, the measured lung function
anatomic differences in the shape and size of the value is compared against what is expected (the
chest, diaphragm, and ribcage.6 Differences average of all healthy individuals of the same
have also been observed between males and fe- age, biological sex, and age) and the range of
males in terms of the timing of lung development values that can be expected in the healthy

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
 Reference Values 571

Fig. 1. Predicted values for (A) Male FEV1, (B) Female FEV1, (C) Male FVC, (D) Female FVC, (E) Male FEV1/FVC (F)
Female FEV1/FVC across age. (Reproduced with permission of the © ERS 2025. Quanjer PH, Stanojevic S, Cole TJ, et
al. Multi-ethnic reference values for spirometry for the 3–95-yr age range: the global lung function 2012 equa-
tions. Eur Respir J 2012; 40: 1324-1342 [[Link]

population (the range of values between the lower were Black or of African ancestry on average had
and upper limit of normal). longer legs for a given standing height (ie, smaller
Cormic Index (sitting height: standing height ra-
tio)), meaning that on average, thorax size, and
Race and Ethnicity
thus lung size, was smaller compared to groups
For many decades, observed differences in lung of White individuals of the same standing
function between populations of different races height.19–21 In other words, this assumed that
or ethnicities were widely considered to be innate lung volumes were smaller for a Black individual
biological differences. The combined influence of compared with that of a white individual of the
genetic and environmental factors on lung growth same standing height, biological sex, and age.
and development across childhood are closely These observations led to the assumption that
related to the observed differences in lung function reduced lung function in Black individuals was
between populations around the world. For innate and that race-specific reference population
example, people of African ancestry on average was necessary to avoid misinterpreting smaller
have lower lung function compared with individ- lung volume as impaired lung function or disease.
uals of white European ancestry of the same The original intent was to improve precision of lung
standing height, biological sex, and age. Several function interpretation; however, invertedly this led
observational studies found that individuals who to the underestimation of lung function impairment

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
572 Bowerman & Stanojevic

in Black individuals. It is now recognized that most Exposures such as secondhand tobacco, biomass
of the previous research did not consider the influ- fuel, air pollution, and noxious chemicals, espe-
ence of the social determinants of health, including cially during lung growth and development, can
environmental and nutritional exposures during all impair lung function. Inclusion of individuals
pregnancy and early childhood on lung function.22 with these exposures can decrease the accuracy
Today, we better understand that there is as of reference equations. However, many social de-
much, if not more, genetic heterogeneity within a terminants of health are not easily captured,
racial or ethnic group as there is between groups despite influencing populations across a large
of individuals,23 and that social and environmental geographic area. Deriving reference equations for
determinants of health have important influences lung function using a crude definition of lung health
on lung function across the life course.24 The (ie, no respiratory diagnosis and nonsmoker) may
assumption that observed differences between mask pathology and normalize impairment. For
populations are entirely due to innate biological dif- instance, current criteria to define healthy refer-
ferences ignores the potential influence of social ence populations may include individual with early
and environmental factors on lung health. This childhood exposures to biomass fuels, which
notion, perpetuated through race-specific reference would widen the range of healthy values and may
equations, may have further widened disparities in misclassify reductions in lung function as normal.31
lung health for the most marginalized populations. Another example is the influence of altitude.32
In 2023, The American Thoracic Society (ATS) One approach is to derive a reference equation
published a statement acknowledging the potential from high altitude communities to control for un-
harms of the continued use of race and ethnicity- measured factors. However, many communities
specific equations.25 Other respiratory societies that live at high altitudes are also socioeconomi-
have subsequently followed.26 The ATS statement cally disadvantaged with higher rates of food inse-
acknowledges that the use of race and ethnicity- curity.33 Comparing measured values against a
specific equations perpetuates the belief that racial healthy population living at altitude may account
differences in lung function are innate, which is not for the physiologic effects of altitude and would
the case. There is not a universally accepted defini- also normalize the negative social determinants
tion of race or ethnicity, which practically makes of health and thus, may misclassify impairment
selection of reference equations difficult or impos- as normal.
sible. Genetic differences are known to be greater Given the complex influence of exposures on
within racial and ethnic groups than between lung health across the life span, and even across
them, and even lung function itself is highly variable generations, it may be futile to define a reference
within racial and ethnic groups.23,27 Further, self- population, which reflects the circumstances
identified race is now known to be a poor proxy across the lifespan that promote ideal growth
for genetic ancestry.28 Even if there was an accu- and development of the lungs. Lung health at the
rate proxy for genetic ancestry, environmental ex- population level has been found to improve over
posures within the same ethnic group likely time and cohort effects may also be present,
contribute more to differences in lung health than which further limits the accuracy of a given refer-
genetics alone.22 Many clinical decision aids rely ence range.34 These limitations highlight the
on cut-offs to define disease, eligibility for treat- need to acknowledge the challenges in applying
ment, and employment or insurance coverage, reference equations for lung function and the un-
therefore it is important to recognize the impact us- certainty that is implicit in the interpretation of
ing a race-neutral reference equation/growth chart measured values. Measured values should be
can have on who is classified as eligible/ineligible interpreted in the context of the medical history
for a range of outcomes.29 of an individual patient.

Other Considerations HOW IS THE NORMAL RANGE DEFINED?

An individual’s weight is not included in reference The range of values that are defined by a healthy
equations for lung function. Nonetheless, weight, population is based on the distribution of
specifically obesity, can impact lung function.30 measured values in a population of otherwise
Obesity may increase the likelihood of extrathora- healthy individuals. The limits of normal are set to
cic restrictive lung disease (eg, obesity hypoventi- reflect the range of values within which we can
lation syndrome) and obstructive sleep apnea expect 95% of the observations to lie. Values
through mechanical and inflammatory processes. outside of this range can be considered atypical,
There is also increasing interest in the nonan- with values further away from the range more likely
thropomorphic determinants of lung function. to represent pathophysiology. In the case of

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
 Reference Values 573

spirometry where only low values are of concern lower limit of normal can lead to systematic misinter-
(eg, in forced expiratory volume in 1 second pretation of values in women, children, and older
[FEV1] or forced vital capacity [FVC]), the lower limit adults.38
is set at 5% of a healthy population, such that 5%
of healthy individuals would be expected to have PREVIOUSLY RECOMMENDED REFERENCE
values below the limit of normal (Fig. 2). Practically, EQUATIONS
the range of expected values is now incorporated
in pulmonary function software, which uses Historically, many geographic areas used a sample
z-scores and visual analog scales depicting red, of the healthy population to create a reference equa-
yellow, and green z-score zones that show how tion and reference range. However, there was very
close an individual’s pulmonary function test little standardization of the equipment, methods
(PFT) result is to the lower limit of normal (Fig. 3). and methodological approaches, which made appli-
PFT results in the yellow zone require more careful cation difficult and the false impression that there
interpretation given their proximity to the cut-off as were large differences in lung function between pop-
compared to results in the green or red z-score ulations. An important study that led to the standard-
zones, which are more likely to represent health ization of how lung function reference equations
or pathology respectively, given how far the result were made and interpreted was that by Hankinson
falls from the expected value and the properties of and colleagues, which derived reference equations
the normal distribution. for spirometry using the third iteration of National
Lung function is often interpreted as a percentage Health and Nutrition Examination Study (NHANES
of the predicted value where the average lung func- III). Spirometry data from 7429 individuals between
tion expected for an individual of similar height, bio- the ages of 8 and 80 years who were asymptomatic,
logical sex, and age is set as 100%. A fixed threshold lifelong nonsmokers residing in the United States
of 80% is often set to reflect the lower limit of were summarized to define the reference range.39
normal.35,36 Although intuitive and easy for most Spirometry was performed according to the ATS
people to understand, this approach does not incor- standards for spirometry40,41 and evaluated meticu-
porate the wide range of values that can be lously for quality control. The NHANES III study was
observed in people of different heights and ages, purposefully designed to improve equity and diver-
and between males and females. Indeed, 80% of sity in PFT interpretation. Differences in lung function
predicted reflects the 5th centile of the normal distri- between racial and ethnic differences groups
bution for young adult white males, whereas the (Caucasian, African American, and Mexican Amer-
lower limit of normal in terms of percent predicted ican) were investigated to better understand why
is much lower for young children and older adults.37 people of different racial and ethnic groups were
The use of 80% percent-predicted to define the different in terms of lung function.39 However, the

Fig. 2. The normal distribution, highlighting how the farther away from the mean a result is, the less likely it is
from the same population. (Adapted from Levy ML, Quanjer PH, Booker R, et al. Diagnostic spirometry in primary
care. Proposed standards for general practice compliant with American Thoracic Society and European Respira-
tory Society recommendations. Prim Care Respir J 2009; 18: 130–147.)

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
574 Bowerman & Stanojevic

Fig. 3. A prebronchodilator and postbronchodilator spirometry result with visual analog scale showing how close
the individual’s result is to the lower limit of normal. Note the red, yellow, and green zones.

study population used to derive the NHANES equa- recommendations from major respiratory societies
tions was entirely from the United States and in North America recommend the use of race-
included only 3 options for race and ethnicity, which neutral approaches to interpret lung function.25,38
limited its application globally. Race neutral equations have been consistently
In recognition of the limitations of the NHANES III shown to better correlate with survival,42 incident
and other reference equations, the Global Lung diagnosis of lung disease,43 and respiratory symp-
Function Initiative (GLI) network was established in toms44 compared to race-specific alternatives. In
2008. The GLI aimed to combine existing data 2022 the GLI-Global equation was created using
from healthy individuals around the world to create data from the original GLI dataset. The data were
a single reference equation that could be applied weighted, such that the original 4 racial and ethnic
globally. Researchers from across the world were groups contribute an equal percentage to the pre-
invited to contribute individual level spirometry dicted lung function values.45 When compared to
data to the data repository.12 The measured spirom- GLI-Other, GLI-Global prediction equations pro-
etry values, participant demographic characteris- duce similar predicted lung function values, with a
tics, including age, sex, height, and race/ethnicity, wider limit of normal (ie, a lower 5th percentile).
as well as details regarding the protocol were used GLI Global serves as a single race-neutral equation
to harmonize the data sources. Data from 97,759 in- (race-neutral in the sense that the patient does not
dividuals were collated and analyzed for systematic need to identify their race or ethnic background)
differences.12 The GLI 2012 multiethnic reference with which to compare an individual’s lung function
equations for spirometry were created, which and does not imply that the range represents
spanned the age range of 3 years to 95 years of optimal lung health. It is not intended to be the
age. Four ethnic-specific equations were derived best lung function reference equation, but instead
from the useable data: Caucasian (White European), serves as step toward a truly race-neutral
African American, North East Asian, and South East approach. The most recent recommendations
Asian. These groupings were created based on also highlight the importance of incorporating bio-
analysis of the reported ethnicity of participants. A logical variability and uncertainty in a clinician’s
fifth composite equation, based on an average of interpretation of a PFT, especially for populations
these 4 racial and ethnic groups, called GLI-Other, that are underrepresented in the GLI dataset.
was also developed for individuals who did not iden- For lung function tests other than spirometry, the
tify with one of these groups. The authors acknowl- GLI network has also derived reference equations
edged that over 80% of the original GLI sample was for carbon monoxide transfer factor (DLCO),46 static
derived from a reportedly White population, which lung volumes,47 and multiple breath washout.48
would disproportionately alter lung function predic- Importantly, these equations were derived primarily
tions toward the White population. The equations from White European populations (ie, DLCO, static
were derived using an advanced statistical tech- lung volumes) or were purposefully created to be
nique that allowed for a single equation across the race neutral (eg, multiple breath washout).
life course and incorporate the biological variability
at different ages to accurately define the lower limit Limitations of the Global Lung Function
of normal (5th percentile) (see Fig. 1). Initiative Equations
The standardization of lung function measurement
Currently Recommended Reference Equations
and interpretation has helped further our under-
Given the evidence of the potential harms of using standing of lung health; however, there are impor-
race or ethnic-specific reference equations, current tant limitations. Large geographic areas of the

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
 Reference Values 575

world, where lung function data collection has his- used as binary decision aids in diagnosis, insur-
torically been limited, were not included in the GLI ance decisions, and even employment, our cur-
equations, most notably continental Africa and rent understanding of the limitations of reference
South Asia.12 Even with the current race-neutral equations requires careful reconsideration of
GLI equations, it must be acknowledged that a these approaches.
large proportion of the original sample was classi-
fied as White European, and the Black subpopula- Case Study
tion in the GLI spirometry database came entirely
To illustrate the potential differences in interpreta-
from the United States. It is unclear whether GLI
tion of lung function with race-specific and race-
equations are appropriate for populations that
neutral equations and the pitfalls of using PFTs
were excluded, including indigenous commu-
as binary decision aids, take the example of a
nities,49 Southern Asia,50 and individuals who live
75-year-old male, who is 180 cm tall. He is a life-
at altitude.51 Our understanding of lung health in
long cigarette smoker with a 60-pack year history,
these communities, which have historically been
who has been experiencing worsening dyspnea
underrepresented is limited and must be further
for the past 12 months. His initial spirometry
investigated to avoid perpetuating disparities.
testing has shown a FEV1 of 1.8 L, and an FVC
A second major limitation of reference equa-
of 2.9 L, with an FEV1/FVC of 0.62. Under the
tions is in the definition of a healthy reference pop-
GLI 2012 African American equations, this individ-
ulation. A very limited definition of health, defined
ual is above or at the lower limit of normal (LLN) for
as lifelong nonsmoking individuals without respi-
all 3 indices (5th percentiles: 1.79, 2.51, and 0.62
ratory symptoms, or physician-diagnosed respi-
respectively). However, under the GLI White equa-
ratory disease was used for inclusion in the GLI
tions, this individual would be below the LLN for
reference equations.12 We now know that there
FEV1 and FVC (2.21 and 3.06 respectively), and
are many factors across the life course that can
above the LLN for FEV1/FVC (0.61). Under race-
negatively impact lung health long before symp-
neutral equations, this individual would be at or
toms occur. Individuals with these exposures
below the LLN for all 3 indices (2.18, 3.0, and
may have inadvertently been included in refer-
0.62 respectively). These results demonstrate
ence equations as healthy, which would be ex-
that for individuals with results at the border of
pected to lower the mean predicted lung
the LLN, changes in reference equations can shift
function for individuals and thus, bias interpreta-
the PFT diagnosis. Thus, it is critical that clinicians
tion. The unmeasured contribution of the social
consider the clinical context and the uncertainty
determinants of health is difficult to predict and
around the upper and lower limits of normal to
is likely multifactorial.
make informed diagnostic and clinical decisions.
A third consideration is that lung function mea-
surements are inherently variable. Two healthy in-
Future Developments and Considerations
dividuals of the same standing height, age, and
biological sex can have very different measured There is an urgent need to better understand the
values. In some cases, the differences can simply factors that influence lung health as well as to
reflect the natural biological variability of the mea- explore alternative interpretation strategies for
surement, whereas in cases where individuals lung function measurements. The solution is not
have been exposed to potentially harmful expo- necessarily to collect more data but rather to
sures, the differences can reflect an underlying consider novel approaches that will improve the
pathophysiology. Even in repeated measurements accuracy, precision and equity of pulmonary func-
of the same individual over time, there is a degree tion test interpretation.
of variability given the effort-dependent nature of One such example for the interpretation of
spirometry testing. While quality-control criteria spirometry results is FEV1 quotient, in which lung
are in place to prevent this, this can be difficult function is compared to the lowest lung function
to fully eliminate and thus must be considered in needed for survival (being 0.4 L and 0.5 L for
interpreting results. men and women, respectively). This approach cir-
Given this uncertainty, PFT results should be cumvents the need for reference equations as it
interpreted in the context of the individual’s clin- compares an individual’s lung function to the
ical history, comorbidities, exposures, and symp- endpoint of respiratory failure or death.52 Howev-
toms. In cases where results lie near the lower er, this approach is not applicable in children,
limit of normal, serial measurements, and other nor is it applicable or useful in situations where
adjunct diagnostic tools, such as chest imaging mortality is not the outcome of interest. In many
should be used to augment clinical decision- clinical scenarios, such as in individuals with
making. While PFT results have historically been early-stage disease, symptom management and

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
576 Bowerman & Stanojevic

hospitalization avoidance are likely to take prece- REFERENCES

dence over mortality prediction.
Another example is the Z-score for change, 1. ATS statement–Snowbird workshop on standardiza-
which compares changes in lung function over tion of spirometry. Am Rev Respir Dis 1979;119(5):
time to the individual’s baseline while accounting 831–8.
for the time between measurements and expected 2. Mahon GM, Koppelman GH, Vonk JM. Grandmater-
change in FEV1 with age. In other words, the nal smoking, asthma and lung function in the
change score identifies changes in lung function offspring: the lifelines cohort study. Thorax 2021;
between 2 timepoints that are outside the limits 76(5):441–7.
of biological variability of the test and indicate a 3. Clausen JL. Prediction of normal values in pulmo-
significant change.53 With this, individuals serve nary function testing. Clin Chest Med 1989;10(2):
as their own controls, which avoids some of the 135–43.
pitfalls associated with reference equations, but 4. Whittaker AL, Sutton AJ, Beardsmore CS. Are ethnic
still requires the standardization of measured differences in lung function explained by chest size?
values for height, biological sex, and age using a Arch Dis Child Fetal Neonatal Ed 2005;90(5):
reference equation. F423–8.
5. DeMuth GR, Howatt WF, Hill BMI. Lung volumes. Pe-
SUMMARY diatrics 1965;35(1):162–76.
6. Garcı́a-Martı́nez D, Torres-Tamayo N, Torres-
Reference equations are an important tool to facil- Sanchez I, et al. Morphological and functional impli-
itate comparison of an individual’s measured lung cations of sexual dimorphism in the human skeletal
function values to a healthy population. The thorax. Am J Phys Anthropol 2016;161(3):467–77.
inherent biological variability of lung function mea- 7. Thurlbeck WM. Postnatal human lung growth. Tho-
surements and uncertainty of interpretation need rax 1982;37(8):564–71.
to be considered to ensure clinical decisions are 8. Dominelli PB, Ripoll JG, Cross TJ, et al. Sex differ-
appropriate for each patient. ences in large conducting airway anatomy. J Appl
Physiol 2018;125(3):960–5.
CLINICS CARE POINTS 9. Bouwsema MM, Tedjasaputra V, Stickland MK. Are
there sex differences in the capillary blood volume
• Reference equations for lung function repre- and diffusing capacity response to exercise?
sent the range of values that can be expected J Appl Physiol 2017;122(3):460–9.
in an otherwise healthy population. 10. Verma MK, Miki Y, Sasano H. Sex steroid receptors
• The influence of the social determinants of in human lung diseases. J Steroid Biochem Mol
health on lung growth and development Biol 2011;127(3–5):216–22.
have not been considered to date, and exist- 11. Schrader PC, Quanjer Ph H, Olievier ICW. Respira-
ing approaches to interpret pulmonary func- tory muscle force and ventilatory function in adoles-
tion measurements may mask impairment, cents. Eur Respir J 1988;1(4):368–75.
especially in marginalized populations.
12. Quanjer PH, Stanojevic S, Cole TJ, et al. Multi-ethnic
• Interpretation of PFTs requires consideration reference values for spirometry for the 3-95-yr age
of an individual’s medical and family history, range: the global lung function 2012 equations.
not just whether the numbers reported are Eur Respir J 2012;40(6):1324–43.
below a threshold.
13. Hibbert M, Lannigan A, Raven J, et al. Gender differ-
• Given the limitations with PFT reference ences in lung growth. Pediatr Pulmonol 1995;19(2):
equations, PFTs should not be used as binary 129–34.
decision aids for diagnosis, insurance, or 14. Pagtakhan RD, Bjelland JC, Landau LI, et al. Sex dif-
employment, but rather as 1 piece of infor-
ferences in growth patterns of the airways and lung
mation in the complete clinical picture.
parenchyma in children 1984;56(5):1204–10.
15. Bush A. Lung development and aging. Ann Am
Thorac Soc 2016;13:S438–46.
16. Miller MR. Structural and physiological age-associated
DISCLOSURES changes in aging lungs. Semin Respir Crit Care Med
C. Bowerman is a Junior Executive on the Global 2010;31(5):521–7.
Lung Function Initiative Network. S. Stanojevic is 17. Green F, Pinkerton KE. Environmental determinants
a co-chair of the Global Lung Function Initiative of lung aging. In: The lung. 2004. Available at:
Network. Payments received from ndd medical [Link]
advisory board and receives speaker fees from B9780123247513500620. Accessed August 5, 2024
Vyaire medical. 2004.

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
 Reference Values 577

18. Frasca D, Blomberg BB. Inflammaging decreases residents of European ancestry. Am J Phys Anthro-
adaptive and innate immune responses in mice pol 1988;75(1):77–85.
and humans. Biogerontology 2016;17(1):7–19. 33. Tremblay JC, Ainslie PN. Global and country-level es-
19. Lum S, Bountziouka V, Sonnappa S, et al. Lung func- timates of human population at high altitude. Proc
tion in children in relation to ethnicity, physique and Natl Acad Sci U S A 2021;118(18):e2102463118.
socioeconomic factors. Eur Respir J 2015;46(6): 34. Allinson JP, Afzal S, Çolak Y, et al. Changes in lung
1662–71. function in European adults born between 1884
20. Harik-Khan RI, Muller DC, Wise RA. Racial differ- and 1996 and implications for the diagnosis of
ence in lung function in African-American and white lung disease: a cross-sectional analysis of ten
children: effect of anthropometric, socioeconomic, population-based studies. Lancet Respir Med
nutritional, and environmental factors. Am J Epide- 2022;10(1):83–94.
miol 2004;160(9):893–900. 35. Sobol BJ, Sobol PG. Per cent of predicted as the
21. Harik-Khan RI, Fleg JL, Muller DC, et al. The effect of limit of normal in pulmonary function testing: a statis-
anthropometric and socioeconomic factors on the tically valid approach. Thorax 1979;34(1):1–3.
racial difference in lung function. Am J Respir Crit 36. Global Initiative for Chronic Obstructive Lung Dis-
Care Med 2001;164(9):1647–54. ease. Pocket guide to COPD diagnosis, manage-
22. Holland R, Bowerman C, Stanojevic S. The contribu- ment, and prevention: a guide for health care
tion of anthropometry and socioeconomic status to professionals. 2020. Available at: [Link].
racial differences in measures of lung function: a org. Accessed January 4, 2022.
systematic review. Chest 2022;162(3). [Link] 37. Swanney MP, Ruppel G, Enright PL, et al. Using the
org/10.1016/[Link].2022.04.017. lower limit of normal for the FEV1/FVC ratio reduces
23. Witherspoon DJ, Wooding S, Rogers AR, et al. Ge- the misclassification of airway obstruction. Thorax
netic similarities within and between human popula- 2008;63(12):1046–51.
tions. Genetics 2007;176(1):351–9. 38. Stanojevic S, Kaminsky DA, Miller MR, et al. ERS/
24. Quanjer PH. Lung function, genetics and socioeco- ATS technical standard on interpretive strategies
nomic conditions. Eur Respir J 2015;45(6):1529–33. for routine lung function tests. Eur Respir J 2022;
25. Bhakta NR, Bime C, Kaminsky DA, et al. Race and 60(1):2101499.
ethnicity in pulmonary function test interpretation: 39. Hankinson JL, Odencrantz JR, Fedan KB. Spiro-
an official American Thoracic Society Statement. metric reference values from a sample of the general
Am J Respir Crit Care Med 2023. [Link] U.S. population. Am J Respir Crit Care Med 1999;
10.1164/RCCM.202302-0310ST. 159(1):179–87.
26. Marciniuk DD, Becker EA, Kaminsky DA, et al. Effect 40. Gardner RM, Hankinson JL. Standardization of
of race and ethnicity on pulmonary function testing Spirometry—1987 update. Am Rev Respir Dis
interpretation: an American college of chest physi- 1987;136(5):1285–98.
cians (CHEST), American Association for Respira- 41. Crapo O, Hankinson JL, Irvin C, et al. Standardiza-
tory Care (AARC), American Thoracic Society tion of spirometry: 1994 update. Am J Respir Crit
(ATS), and Canadian Thoracic Society (CTS) evi- Care Med 1995;152(3):1107–36.
dence review and research statement. Chest 2023; 42. McCormack MC, Balasubramanian A, Matsui EC,
164(2):461–75. et al. Race, lung function, and long-term mortality
27. Kumar R, Seibold MA, Aldrich MC, et al. Genetic in the National Health and Nutrition Examination Sur-
ancestry in lung-function predictions. N Engl J vey III. Am J Respir Crit Care Med 2022;205(6):
Med 2010;363(4):321–30. 723–6.
28. Witonsky J, Elhawary JR, Eng C, et al. Race- and 43. Elmaleh-Sachs A, Balte P, Oelsner EC, et al. Race/
ethnicity-based spirometry reference equations: ethnicity, spirometry reference equations, and pre-
are they accurate for genetically admixed children? diction of incident clinical events :the multi-ethnic
Chest 2022;162(1):184–95. study of atherosclerosis (MESA) lung study. Am J
29. Diao JA, He Y, Khazanchi R, et al. Implications of Respir Crit Care Med 2022;205(6):700–10.
race adjustment in lung-function equations. N Engl 44. Baugh AD, Shiboski S, Hansel NN, et al. Reconsi-
J Med 2024;390(22):2083. dering the utility of race-specific lung function pre-
30. Dixon AE, Peters U. The effect of obesity on lung diction equations. Am J Respir Crit Care Med
function. Expet Rev Respir Med 2018;12(9):755–67. 2022;205(7):819–29.
31. Vaz Fragoso CA, McAvay G, Van Ness PH, et al. 45. Bowerman C, Bhakta NR, Brazzale D, et al. A race-
Phenotype of normal spirometry in an aging popula- neutral approach to the interpretation of lung func-
tion. Am J Respir Crit Care Med 2015;192(7): tion measurements. Am J Respir Crit Care Med
817–25. 2023;207(6).
32. Greksa LP, Spielvogel H, Paz-Zamora M, et al. Effect 46. Stanojevic S, Graham BL, Cooper BG, et al. Official
of altitude on the lung function of high altitude ERS technical standards: Global lung function

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.
578 Bowerman & Stanojevic

initiative reference values for the carbon monoxide ethnic norms comparative study. Expet Rev Respir
transfer factor for Caucasians. Eur Respir J 2017; Med 2021;15(6):833–43.
50(3):1700010. 50. Abdullah N, Borhanuddin B, Shah SA, et al. Global
47. Hall GL, Filipow N, Ruppel G, et al. Official ERS tech- lung initiative 2012 spirometry reference values in
nical standard: global lung function initiative refer- a large Asian cohort of Malay, Chinese and Indian
ence values for static lung volumes in individuals ancestry. Respirology 2018;23(12):1173–9.
of european ancestry. Eur Respir J 2021;57(3). 51. Rangel LGG, Martı́nez-Briseño D, Bowerman C, et al.
48. Ramsey KA, Stanojevic S, Chavez L, et al. ERS tech- Effect of altitude on spirometry values in Latin Amer-
nical standard: Global lung function initiative refer- ican: a GLI network ERS clinical research collabora-
ence values for multiple breath washout indices. tion. Eur Respir J 2023;62(suppl 67):OA872.
Eur Respir J 20242400524. [Link] 52. Miller MR, Pedersen OF. New concepts for express-
13993003.00524-2024. ing forced expiratory volume in 1 s arising from sur-
49. Heraganahally SS, Howarth T, White E, et al. Lung vival analysis. Eur Respir J 2010;35(4):873–82.
function parameters among Australian Aboriginal 53. Stanojevic S, Filipow N, Ratjen F. Paediatric repro-
‘apparently healthy’ adults: an Australian Caucasian ducibility limits for the forced expiratory volume in
and global lung function initiative (GLI-2012) various 1 s. Thorax 2020;75(10):891–6.

Downloaded for Dr Rios D Nicaragua (drriosdubois@[Link]) at [Link] from [Link] by Elsevier on

September 03, 2025. For personal use only. No other uses without permission. Copyright ©2025. Elsevier Inc. All rights reserved.