INTRODUCTION
The pulmonary function test (PFT) is an important tool used to identify and monitor patients with respiratory disease. This test reveals the statuses of the small and large airways and the integrity of the pulmonary parenchyma [
1]. Although the PFT itself is not diagnostic, an abnormal result can help establish confirmative diagnoses of many respiratory diseases when combined with the patient’s history. The PFT also helps to predict the risk and prognosis associated with lung resection [
2]. Chronic obstructive pulmonary disease can be diagnosed using the ratio of forced expiratory volume in 1 second (FEV1) and the forced vital capacity (FVC) [
3]. Furthermore, improvement of FEV
1 is an important target of asthma treatments [
4]. The PFT results therefore provide important information about these major chronic respiratory diseases.
Abnormal spirometry is usually defined as an FEV
1, FVC, or FEV
1/FVC value below the normal range. Although unusually high values may also be encountered during PFT, the clinical meaning of this type of abnormality has not been evaluated and is not usually mentioned in the guidelines. Only one guideline briefly comments on the unusual nature of a percent predicted value > 140% and instructs clinicians to ensure that the employee’s age or height has been measured and entered correctly in such cases. If the unusually high percent predicted value cannot be explained by an error, the guideline recommends recalibration of the spirometer [
5].
One might assume that such unusually high spirometry values merely represent a few exceptional cases, mistakes in the input of demographic data, or calibration errors. The reference equations were developed using data collected from representative healthy populations. In other words, the references for minority populations, such as people with unusual heights and weights or ages > 80 years, are based on small numbers of subjects or extrapolations [
6,
7]. We therefore hypothesized that the reference equations might not be sufficiently accurate, and this limitation could lead to the underestimation of predicted values in some minority populations. Therefore, an unusually high predictive value may be possible in specific groups and should not be considered merely an error of demographics or calibration. This study aimed to characterize the demographic features of subjects with a high percent predicted pulmonary function value.
DISCUSSION
Our study found that 0.08% of the total subjects had an unusually high pulmonary function, defined as a FEV1 or FVC ≥ 140% of the predicted value. This pulmonary function pattern was more common in older (> 80 years) female participants with a lower body weight (< 50 kg) and shorter height (< 140 cm). An unusually high pulmonary function was not rare in this specific population (6.1%). Furthermore, an older age and shorter height remained significant predictors of a high pulmonary function in multivariate analysis.
The PFT plays an important role in the diagnosis and management of respiratory diseases [
9,
10]. Because the PFT results depend on height, age, sex, and ethnicity, standardization and development of correct reference equations using “local healthy controls” are critical issues [
8,
11]. As noted above, the reference equations were derived using cross-sectional data from representative populations [
12-
16], indicating a requirement for sufficient numbers of participants in each age, sex, height, and weight category. However, it is not possible to collect sufficient data from all populations, particularly individuals with unusual heights or body weights. Therefore, the values generated using prediction equations may differ from the observed values in some minority populations [
8], and this limitation should be considered when interpreting such data. The Korean reference equation was developed using data from 1,212 nonsmokers to provide a better local standard [
7]. However, although data from more than 1,000 people were used to derive the reference equation, the subject group included only nine subjects older than 70 years [
7]. This situation is consistent with that associated with the development of other reference equations; in a British study, only 6% of male and 9% of female subjects were older than 75 years [
13], whereas several other previous studies included no subjects older than 65 years [
14,
17,
18]. Our data suggest that reference values for elderly populations may be inaccurate and that a wider range of populations is needed to avoid under- or over-estimation of spirometry results.
Only a few previous studies have analyzed high PFT results. In swimmers, a relatively higher pulmonary function was associated with increases in the chest wall width and muscle strength [
19,
20]. Cochet et al. [
21] found that 5.5% of all patients had either a FEV
1 or FVC > 110% of the predicted value, whereas 1.5% of the patients had an FEV
1 or FVC > 120% of the predicted value; furthermore, the proportion of patients with supranormal lung function did not differ significantly between active duty military personnel and a non-active duty population of the same age range. However, even these studies did not mention patients with unusually high PFT results (≥ 140% of the predicted value), and only one suggestion was made to re-check the calibration and demographic data in such cases [
5]. This guidance is partly reasonable because the cases with values ≥ 140% of the predicted value are very rare in the general population. Only one (0.004%) of 20,115 male subjects with a height > 170 cm had a value ≥ 140% of the predicted value. Similarly, only one (0.01%) of 12,821 male subjects aged younger than 50 years had an abnormally high value (
Table 5). However, we note that such values were not rare in some specific demographic groups, such as short elderly women.
In our study, old age was associated with a high predicted pulmonary function. Increasing age corresponds with decreases in the static elastic recoil of the lung, chest wall compliance and respiratory muscle strength [
22]. In humans, the FEV
1 and FVC generally increase up to the middle of the third decade of life, then decrease thereafter [
23,
24]. However, physiologic measurements vary more widely among older populations, compared with younger populations [
25]. Indeed, age was not normally distributed (Kolmogorov-Smirnov test,
p = 0.032; histogram in
Fig. 2) among 55 subjects with high pulmonary function. Although subjects with a high pulmonary function tended to be older, many previous reference equations were based on extrapolations about minority populations, especially those older than 70 years [
7,
13-
15,
26]. Given the expected global increase in life expectancy [
27], a reference equation that has been validated in a sufficiently large population of older subjects is essential.
Although female sex was a significant predictor of a high pulmonary function in the univariate analysis, male sex was associated with a high pulmonary function after adjusting for age, height, and weight. This conflicting result might be attributable to the small number of subjects with high pulmonary function and an imbalance of anthropometric values between female and male subjects. Indeed, within the high pulmonary function group, female subjects had a lower weight (47.6 kg vs. 61.8 kg, p < 0.001) and shorter height (143.9 cm vs. 157.0 cm, p < 0.001) than did the men, and this imbalance might have affected our results. In addition, although height was a significant predictor of a high pulmonary function in both the univariate and multivariate analyses, body weight was only significant in the former, which suggests that height might be a stronger predictor of pulmonary function. As the body weight correlated positively with height (r = 0.641, p < 0.001), the former parameter might lose significance in a multivariate analysis including the latter.
This study has several limitations of note. First, the design was retrospective in nature. However, we included a large number of patients with little missing data. Furthermore, we included all subjects who underwent PFT during a single calendar year to minimize confounding factors. Second, our analysis included only patients at a single referral center in South Korea. Third, only 55 subjects (0.08% of all included subjects) were found to have high pulmonary function and thus it is difficult to generalize the results of our current study. External validation studies in other populations of subjects may be necessary.
In summary, an unusually high PFT was associated with the female sex, an older age, lower body weight and shorter height, and this outcome was not unusual among subjects who exhibited all of these characteristics. Therefore, we should consider the demographics of the examinees and the limitations of reference equations when interpreting PFT results. The meticulous inclusion of various healthy controls when generating reference equations may reduce this type of overestimation.