If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. University of Maryland School of Medicine, Division of Pulmonary and Critical Care Medicine, 110 South Paca Street, 2nd Floor, Baltimore, MD, 21201, USA.
Division of Pulmonary and Critical Care Medicine, University of Maryland School of Medicine, Baltimore, MD, USADivision of Occupational and Environmental Medicine, University of Maryland School of Medicine, Baltimore, MD, USA
Currently available reference values for adult impedance measures are limited.
•
Most current adult reference values are based on European Caucasians and Chinese.
•
There is marked variability between studies in measurement technique.
•
Forced oscillation and impulse oscillometry values may not be interchangeable.
•
Normal values obtained from large populations using standardized method are needed.
Abstract
Objective
To provide an evidence-based review of published data regarding normal range reference values and prediction equations for measurements of respiratory impedance using forced oscillation technique (FOT) and impulse oscillometry (IOs) in adults.
Methods
A non-language-restricted search was performed using forced oscillation technique and impulse oscillometry as primary terms. Original research studies reporting respiratory system impedance reference values or prediction equations based on cohorts of ≥100 healthy adults were included. Publications cited in identified studies were also considered for inclusion.
Results
Of 882 publications identified, 34 studies were included: 14 studies of FOT, 19 studies of IOs, and one study of both techniques. Nineteen studies provided prediction equations. Most reports were from Europe (n = 20) and Asia (n = 12) and included relatively small cohorts (median = 264 subjects). Across publications, there was marked variability in performance and technique of impedance measurements. Height and sex emerged as major contributors to available prediction equations. The contribution of weight was more pronounced at the obese end of the weight spectrum. The contribution of age was less clear, and elderly were largely under-represented. Ethnicity likely plays a role, but was under-reported in currently available literature. Inclusion of current and former smokers in some studies further confound the results.
Conclusions
Currently available literature providing reference values and prediction equations for respiratory impedance measurements in adults is limited. Until larger-scale standardized studies are available, the choice of prediction equations should be based on datasets that best represent the target patient population and modality in use within each pulmonary physiology laboratory.
], the measurement of lung impedance using pressure sound waves, termed forced oscillation technique (FOT) and its younger derivative, impulse oscillometry (IOs), have yet attained an established position in the arsenal of lung function testing tools that are routinely employed in the adult pulmonary function laboratory. Indeed, in 2003 the American Thoracic Society (ATS) officially adopted FOT as a modality only with respect to pulmonary evaluation of children under six years of age without providing clear guidance concerning technical aspects of study performance and quality control [
An official American Thoracic Society workshop report: optimal lung function tests for monitoring cystic fibrosis, bronchopulmonary dysplasia, and recurrent wheezing in children less than 6 years of age.
]. The same year, an European Respiratory Society (ERS)-designated Task Force acknowledged the need for further standardization of FOT in children and adults [
]. No similar statements regarding the use of IOs have been published by prominent pulmonary medicine societies to date.
The lack of guidance recommendations regarding standardization of performance and quality control of impedance testing parallels other gaps. First, guidance on appropriate use and clinical application of impedance testing in the adult population is somewhat limited [
]. Second, large-scale adult population-based datasets of normal values are also limited, culminating in lack of well-agreed upon prediction equations that will allow appropriate in-context interpretation of these measurements. Among more conventional testing such as spirometry, guidance from the ATS and ERS suggests that clinicians should select reference values for testing that are most similar in demographic and anthropometric characteristics to the clinical population being tested [
]. Similar strategies would be appropriate for selection of reference values in studies of respiratory impedance.
In this review, we summarize major studies published to date reporting normal reference values or prediction equations for both FOT and IOs in adults. The purpose of this review is to present options for selection of reference values in performance of respiratory impedance measurements in adult clinical practice. In this regard, we elaborate on descriptive characteristics of the various studies' populations of healthy adult volunteers as well as the specific modalities and settings used to derive reference values and prediction equations.
2. Background
2.1 Forced oscillation technique and impulse oscillometry
Unlike spirometry, both FOT and IOs measurements are effort-independent and superimposed on tidal breathing, requiring only a good mouth seal and application of pressure on the cheeks and the floor of the mouth in order to reduce the upper airway shunt effect [
], but may also provide a rationale for their more routine use in adults with both physical and cognitive limitations, institutionalized patients, and the elderly [
]. The use of this modality gained much of its support based on the finding that the application of a range of forced oscillation frequencies to the airways by means of a loudspeaker allows to distinguish between the amount of resistance to flow contributed by the large and the small airways. Low oscillations in the range of 5–15 Hz transmit more distally in the airways, thus representing more of the whole respiratory system, while higher oscillations of ≥20 Hz tend to travel as far as the intermediate size airways, thus representing the respiratory system resistance from the mouth and up to that point [
]. This prospect was utilized for characterization of abnormalities localized to the more peripheral airways, for which spirometry was shown to be is less reliable [
FOT was originally based on multiple single-frequencies of sinusoidal oscillations in the range of 2–30 Hz, providing excellent temporal resolution. The incorporation of the fast Fourier transform of pseudorandom noise allowed the breakdown of different wave frequencies into unique single-frequency sine waves, later derived by spectral analysis, thus greatly simplifying the process of performing measurements over a range of frequencies [
]. Over the years, laboratories worldwide employed different FOT apparati, different standards of measurement, and different processing software, at least in part explaining the variations in results between major contributors to the bulk of the FOT literature. This problem was at least partially resolved following the publication of the ERS Task Force recommendations for standardization, quality control measures, and implementation guidelines in 2003 [
], however, currently available commercial FOT systems, such as Resmon Pro Diary® (Restech srl, Milan, Italy), termoFlo C-100® (Thorasys Medical Systems, Montreal, Canada), and MostGraph-01® (CHEST M.I., Tokyo, Japan), have largely overcome this limitation [
]. IOs uses short impulses of fixed frequency-rectangular pressure waves from which all other frequencies can be derived using spectral analysis. This technique, which somewhat compromises the temporal resolution of the measurements, greatly simplifies study performance, allowing it to be more easily implemented in the laboratory [
Analogous to an electric circuit in series, measurements of the mechanical properties of the airways, lungs, and chest wall obtained by FOT and IOs are described as a function of series-resistive elements. When a wave of airflow is superimposed on tidal breathing and applied to the air column in the respiratory system, the resultant transthoracic pressure, flow amplitude, and phase differences, as obtained by flow and pressure transducers, reflect the total impedance of the respiratory system (Zrs). This information can be broken down to provide the following measurements:
Resistance of the respiratory system (Rrs): derived from analyzing the pressure oscillations that are the real part of impedance and in-phase with the pressure-flow relationship across the range of oscillation frequencies. Resistance measures reflect the sum of three respiratory system resistance components in series: the extrathoracic and central airways, the peripheral small airways, and the chest wall. Since low frequency oscillations transit more distally than high frequency oscillations, resistance measures at low and high frequencies allow differential characterization of the frequency dependence (Rlow – Rhigh) of the respiratory system, providing information which is more inclusive of the smaller airways [
Reactance of the respiratory system (Xrs): the reflection of the imaginary part of impedance or the out-of-phase component of the pressure-flow relationship, and attributable to two opposing mechanical factors: capacitance and inertance. Small airways, lung parenchyma, and chest capacitative pressure loss is dominant at low frequencies while the large airway inertive pressure loss is dominant at high frequencies. The resonant frequency (Fres) is the frequency at which the relative contribution of both vectors is roughly equal providing an Xrs value of zero. Integration of the low frequency reactance between 5 Hz (X5) and Fres provides reactance area (AX), which describes the reactant properties of the more peripheral airways [
Coherence (CO) describes the temporal variability of the data sample at different frequencies, usually reported at 5, 10, and 20 Hz. CO is decreased by improper technique such as irregular breathing or flow obstruction by the tongue, resulting in its use as a technical quality measure, but it is also asserted that abnormal CO, representing the non-uniformity of respiratory mechanics, may be a property of obstructive lung disease such as chronic obstructive lung disease [
Current literature provides sufficient evidence for a role of impedance measurements as complimentary to history, physical examination, spirometry, and imaging in the evaluation of patients with large and small airways disease as well as parenchymal lung diseases. Indeed, several studies have demonstrated the additive value of FOT in disease states characterized by respiratory abnormalities not yet reflected by spirometry [
Advantage of impulse oscillometry over spirometry to diagnose chronic obstructive pulmonary disease and monitor pulmonary responses to bronchodilators: an observational study.
Aside from its relevance in elderly and debilitated patient populations due to the above mentioned technical advantages, patterns of respiratory impedance have been described in patients with obstructive lung disease [
], amongst others. In occupational setting, impedance measurements were shown to be useful as an adjunct to spirometry in evaluation of subjects exposed to environmental irritants [
]. Following the terrorist attack on the World Trade Center in September of 2001, several studies have explored the role of FOT and IOs in early detection and characterization of respiratory dysfunction secondary to massive exposure to dust and smoke [
]. Nonetheless, the ability of oscillometric studies to distinguish between the various pathophysiologic mechanisms of respiratory abnormalities as a stand-alone modality in symptomatic patients is less clear [
E. Oostveen, ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine (US). Identifier NCT01868607, Reliability of different FOT devices to measure the human respiratory impedance, (n.d.). https://clinicaltrials.gov/ct2/show/record/NCT01868607 (accessed January 1, 2017).
A structured search of the literature was performed to identify all studies that reported reference values and/or prediction equations for FOT or IOs, based on cohorts of healthy subjects 18 years of age and above (Fig. 1). We performed a non-language-restricted PubMed search for the period from inception of the database to April of 2017 using forced oscillation technique and impulse oscillometry as primary search terms. A similar search was performed in the Cochrane Library (including Cochrane Database of Systemic Reviews, Database of Abstracts of Review of Effects, and Cochrane Central Register of Controlled Trials) and ClinicalTrials.gov registry. The search was updated in June of 2017.
Fig. 1Flow chart for study search and selection process.
The initial PubMed search identified 882 potential papers. No relevant Cochrane Library reviews were identified and one relevant ongoing clinical trial was found in ClinicalTrials.gov (NCT01868607). Titles and abstracts were initially screened, and selected articles underwent full review. We excluded animal studies, studies reporting results only in pediatric subjects, studies not reporting results for healthy subjects, and review articles. Each retained publication's reference list was then reviewed for additional potentially relevant citations. Citations not found in PubMed were further searched online using the Google Scholar tool. Publications reporting data for less than 100 subjects were excluded from this review. This number was chosen based on a survey of the initial volume of publications following the completion of literature review. Short-listed publications were then scrutinized for data including modality (whether FOT, IOs, or both), waveform, frequency range, number of healthy subjects included, sex distribution, anthropometric variables including age, height, weight, and ethnicity, smoking status, and provision of reference values and/or prediction equations. For studies comparing an exposed or intervention group with a control group, only data related to the control group were extracted, given that this group was comprised of healthy “unexposed” adults. All retained publications were also explored for the differential contribution of sex, age, height, and weight to reported prediction equations.
4. Results
A total of 34 studies providing impedance reference values or prediction equations met inclusion and exclusion criteria and were short-listed for this review [
]. Table 1 provides a descriptive summary of these studies. Table 2 provides a summary of individual study pertinent data. All publications listed in Table 2 provide reference values for Rrs, Xrs, or both measures of impedance. Whereas a majority of early studies employed FOT, a shift towards IOs occurred with the turn of the new millennium. Of the 19 studies providing prediction equations, 8 were for FOT [
] was documented in 27/34 studies. While most studies included in this review were performed in Western Europe, a fair amount of data were derived from studies performed in China. In general, cohorts were relatively small, reporting results for no more than 1000 subjects (Table 1, Table 2). Prominent among short-listed publications were Fischer and Matthys who in 1977 reported FOT prediction equations derived from a cohort of 2350 German subjects, 51.1% of which were female; however, these results were reported in the form of an abstract only [
]. Additionally, in his Master's degree thesis published in 2014, Xue provided IOs reference values from a cohort of 6945 pediatric and adult subjects from over 25 Chinese provinces, but failed to disclose the proportion of adults or the male to female ratio in his cohort [
]. Of the 31 studies reporting sex distribution of participants, approximately two thirds included relative equal representation of men and women. The remaining studies were mostly male-predominant. Only 23% of studies reported the ethnic background of the participating subjects. Of these, a majority included predominantly Caucasian subjects [
]. Approximately 60% of studies included active cigarette smokers or failed to clearly report the smoking status in their cohorts. Moreover, among studies that included current non-smokers, there was considerable variability in the classification of former smokers as non-smokers. Table 3 summarizes studies providing meaningful data regarding subjects ≥60 years of age. Of those studies, three focused almost exclusively on this age group [
Table 2Major publications reporting normal reference values and prediction equations for forced oscillation technique and impulse oscillometry in healthy adults.
Accounting for upper airway shunt using cheek support was also considered documented if impedance measurements were reported to have been performed in accordance with the 2003 ERS Task Force recommendations [3].
b Accounting for upper airway shunt using cheek support was also considered documented if impedance measurements were reported to have been performed in accordance with the 2003 ERS Task Force recommendations [
Table 4 summarizes the relative contributions of anthropometric variables of height, weight, and age to Rrs and Xrs measurements in male and female subjects. Many studies demonstrated a negative association between height and Rrs [
]. This association was less consistent among females in comparison to males. Concomitantly, a majority of studies across both sexes showed positive associations between height and Xrs [
]. This association was more notable in low frequencies in comparison to high frequencies. Finally, associations between age and impedance measures were generally inconsistent. When comparing between the sexes, most authors reported higher Rrs measurements in women when compared with men across all age groups and cohorts [
]. Sex-related differences seemed less pronounced in Xrs measures. Sixteen of the 19 studies reporting prediction equations provided separate sets for men and women.
Table 4Relative contribution of anthropometric covariates to prediction of respiratory system impedance measurements across studies reporting impedance prediction equations.
In this review, we characterize and summarize major publications providing reference values and prediction equations for measurements of respiratory system impedance. As impedance measures gain more acceptance as an adjunct tool in the arsenal of pulmonary function tests, understanding the origins and validity of currently used data sets is crucial for accurate interpretation of the results within a given clinical context. A discussion regarding the definition of “normal” and “abnormal” in adult impedance testing is beyond the scope of this review article. It is important to note, however, that much of our current practice in terms of definitions of the normal range for both FOT and IOs in adults, stems from the seminal 2005 publication by Smith, Reinhold, and Goldman [
]. In this context, measurements obtained by FOT are not necessarily exchangeable with those obtained by IOs due to differences in sound wave generation and method of analysis. Comparison studies have demonstrated discrepancies between FOT devices from different manufacturers [
]. In that respect, IOs was shown to generate higher Rrs measurements when compared with FOT. These data raise concern over the accuracy of prediction equations derived from cohorts in which both modalities have been used interchangeably [
], can impact impedance values. Several studies reviewed here did not address hands-on-cheek technique to control for this potential confounder (Table 2). Optimally, reference values should be selected from studies where upper airway shunt was minimized as described by proper hands-on-cheek technique.
When choosing a set of prediction equations for a patient population, prior guidance suggests considering two fundamental questions: (a) does the cohort from which the equations were derived represents the target patient population? (b) does the modality by which the reference equations were derived resemble the technical approach adopted by the local pulmonary physiology laboratory both in general terms (FOT versus IOs) and in the specifications (frequency range, waveform, etc.)?
To illustrate, among sets of reference equations provided by manufacturers of IOs wave analysis software there is a tendency to default to the very first IOs reference set published by Johannes Vogel and Udo Smidt in 1994 [
]. This set of values, however, was obtained from a cohort of a mostly Caucasian population recruited in the industrial city of Erfurt in former Eastern Germany, which suffered significant air pollution during the second half of the 20th century [
]. In addition, this reference population included a relatively high proportion of smokers, the actual number of which was not reported. These data together raise some concern as for the generalizability of the German Erfurt prediction equations over other populations, specifically non-Caucasians or Caucasians of non-European descent, even though it remains one of the most comprehensive reference sets for IOs in terms of size, age, and sex distribution.
The relative contribution of anthropometric variables to impedance is closely related to the population studied. As can be inferred from Table 4, anthropometric variables have different and sometimes opposing effects on impedance measures in different study populations. In this context, the increase in Rrs and decrease in Xrs with increasing height is speculated to be a result of larger diameter of the airways [
]. Interestingly, in an early FOT study this effect was no longer significant when spirometry measurements were added into regression equations, possibly suggesting that the relation between Rrs and height may in fact reflect the relation between height and lung volume [
] in women. In that respect, none of the currently available prediction equations account the measurements of impedance for lung volume.
The increase in Rrs and decrease in Xrs with increasing weight suggest that weight plays a role in determination of impedance. This role appears more pronounced in the morbidly obese end of the weight spectrum with body mass index of ≥40 kg/m2 [
], supporting the hypothesis that these findings may in fact reflect obesity-related pulmonary inhomogeneity, possibly as a consequence of mechanical airway compression and microatelectasis [
]. In a study exploring resistance measurements before and after bariatric surgery in morbidly obese subjects without known airway disease, a relatively high prevalence of airway hyper-responsiveness was also postulated to be related to excessive airway compression and obstruction [
]. Since weight appears to be important in impedance measures, reference sets and prediction equations should therefore account for weight variations, including in the obese range.
Reports of reference values in the elderly population deserve special attention. The healthy elderly is relatively underrepresented in studies of respiratory impedance and other modalities of lung function testing. Predictive values for this age group are often extrapolated from younger patient populations. As illustrated in Table 4, three studies to date provide relatively robust data for this age group. Guo et al. found no correlation between age and FOT measures of Rrs or Xrs in 223 Caucasian geriatric hospital residents in Switzerland, all of which were 65 years and above, both within their cohort [
]. Importantly, Rrs values reported in this study tended to be lower than those predicted by extrapolation from reference values provided by Vogel and Smidt [
]. In contrast, Schulz et al. studied a cohort of 397 Caucasian German Bavarians, 80% of which were 55 years of age and above, and reported a significant age-dependency of IOs measurements of R5-R20, AX, and Fres in both sexes and X5 in females only [
], Rrs measures tended to be higher than those reported by Guo et al., probably reflecting higher mean height and weight in the former cohort and possibly differences in technique. Age-related increase in resistance was also demonstrated in Japanese subjects over 70 years of age when compared with younger subjects [
]. While these inconsistent data make it difficult to draw a conclusion regarding the magnitude of how age contributes to impedance, they also reinforce the need for larger and more inclusive studies encompassing a wide range of age groups.
The smoking status of participants in most reference cohorts confounds the data among those thought to be normal. In their seminal report from 1994, Vogel and Smidt maintained that due to the high prevalence of cigarette smoking in the studied population, smokers can be regarded as “healthy” [
]. This argument is in conflict with large population-based studies used for the derivation of spirometry prediction equations, such as the third National Health and Nutrition Examination Survey (NHANES III), which excluded both current and former smokers from their analysis [
Finally, the impact of race and ethnicity on impedance measures is difficult to characterize based on current literature. Ethnicity is significantly under-reported and this may limit the applicability of current equations to heterogenous patient populations. Moreover, most currently available impedance prediction equations were derived from studies performed in Western Europe and China and to a lesser degree North America, Japan, and Australia. South Asian, African, and Central and South American populations remain under- or non-represented. Two studies to date reported prediction equations based on multinational cohorts [
], however, most participating countries in both studies can be classified as developed, thus not representing the developing world. The studies by Newbury et al. and Brown et al. are noteworthy for deriving prediction equations for IOs and FOT, respectively, based on cohorts of non-smoking Caucasian Australians [
]. Interestingly, prediction equations derived from these studies provided values that were different from those provided by studies based on seemingly similar Caucasian European populations [
]. Similarly, Schulz et al. observed a discrepancy between prediction equations derived from their Caucasian Bavarian Germans and those derived from the East German cohort of Vogel and Smidt [
]. Currently, there is no accepted sets of reference values or established correction factors for populations such as blacks or Hispanics.
Possibly adding more confounders into the overall picture, in his cohort of Chinese subjects, Xue demonstrated an association between impedance measures and climate, geography, geology, and other environmental factors.
To summarize, current data suggests that height and to a somewhat lesser degree weight influence impedance. Sex may be important for Rrs, but likely less important for Xrs. The contribution of age, while probable, appears not as pronounced as in spirometry measures [
]. Smoking may have an effect even in smokers without respiratory symptoms; hence, future reference sets should ideally exclude current and preferably former smokers. The impact of race and ethnicity is not well-defined; however, differences in reference sets across different nations and their established role in spirometry prediction equations imply that they may influence impedance as well. At the same time, currently used prediction equations are difficult to generalize over other, even seemingly similar, populations due to (a) non-standardized hardware setup and measurement settings, which are not necessarily interchangeable with different sets of apparati and specifications; (b) non-standardized approach to the definition of the “healthy volunteer”; and (c) the potential influence by yet poorly-defined genetic and environmental factors.
6. Conclusions
Impedance measurements gain more and more favor as complimentary to routine pulmonary function testing for better characterization of respiratory abnormalities, and their role in specific disease-states in adults is constantly explored and expanded. Review of currently available literature yielded a surprisingly large number of publications reporting reference values and prediction equations for adults; however, these studies are characterized by relatively small cohorts, major differences in subjects' inclusion and exclusion criteria, and non-standardized study performance and technique. These findings underscore the need for large population-based studies that will provide reliable adult reference value datasets and prediction equations that are based on standardized methodology. These must account for both sexes and a wide range of heights, weights, and ages, as well as include a wide diversity of ethnic origins. Such data will allow the clinician to make an informed and conscious decision when basing the individual patient's results on a reference set of normal values that reflects standard technical practice as well as an appropriate patient population.
Funding source
None.
Conflicts of interest
None.
Acknowledgments
The authors would like to thank Xianglan Yao, MD, PhD, Cuilian Dai, MD, and Haitau Xu, BSc for their assistance with acquisition and translation of publications in the Chinese language and Montserrat Diaz-Abad, MD for her critical review of the manuscript.
An official American Thoracic Society workshop report: optimal lung function tests for monitoring cystic fibrosis, bronchopulmonary dysplasia, and recurrent wheezing in children less than 6 years of age.
Advantage of impulse oscillometry over spirometry to diagnose chronic obstructive pulmonary disease and monitor pulmonary responses to bronchodilators: an observational study.
E. Oostveen, ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine (US). Identifier NCT01868607, Reliability of different FOT devices to measure the human respiratory impedance, (n.d.). https://clinicaltrials.gov/ct2/show/record/NCT01868607 (accessed January 1, 2017).
30024-6&id=gr1.jpg){kind=link}