Exercise‐Induced Oxygen Desaturation and Heart Rate Response During 6‐Min Walk Test Predict Pulmonary Hypertension in Exertional Dyspnea: A Retrospective Cohort Study

ABSTRACT

Pulmonary hypertension (PH) is a life‐threatening condition frequently associated with exertional dyspnea. It remains diagnostically challenging due to limitations in current screening modalities. While the 6‐min walk test (6MWT) has been applied for risk stratification in confirmed PH, its potential role in screening remains unexplored. This retrospective cohort study investigated the diagnostic utility of 6MWT‐derived parameters in 180 patients with exertional dyspnea. PH diagnosis was confirmed by right heart catheterization with the definition of mean pulmonary artery pressure > 20 mmHg. Among 79 PH patients (43.9%), a significantly reduced 6‐min walk distance (6MWD) was observed compared to non‐PH patients (469.5 ± 106.4 m vs. 509.8 ± 74.9 m, p = 0.019). Continuous physiological monitoring revealed that the SpO₂ trough and the heart rate (HR) peak occurred at different time points during 6MWT. Propensity score‐matched case‐control analysis further demonstrated greater exercise‐induced desaturation of SpO₂ from the rest to minimal levels (ΔSpO₂_{rest −min}: 9 ± 9% vs. 4 ± 6%, p < 0.001) and exaggerated HR response from the rest to maximal levels (ΔHR_{max −rest}: 51±21bpm vs. 34±14bpm, p < 0.001) in PH patients. Multivariable analysis identified ΔSpO₂_{rest −min} ≥ 5% (AUC = 0.715, 95% CI: 0.640–0.852; p < 0.001) and ΔHR_{max −rest} ≥ 42 bpm (AUC = 0.740, 95% CI: 0.656–0.823; p < 0.001) as independent predictors of PH. The number of these predictors discriminated the risk of PH in dyspneic patients. A risk‐stratification model incorporating these thresholds demonstrated improved predictive value for PH screening, with a C‐statistic of 0.786 (95% CI: 0.710–0.863, p < 0.001). These findings suggest that parameters derived from the 6MWT, particularly exercise‐induced SpO₂ desaturation and HR response, may facilitate noninvasive PH screening in exertional dyspneic patients.

1. Introduction

Pulmonary hypertension (PH), defined by a mean pulmonary arterial pressure > 20 mmHg, arises from progressive pulmonary vascular remodeling associated with diverse cardiopulmonary and systemic disorders [1]. Despite affecting 1% of the global population, current screening relies predominantly on echocardiography with limited sensitivity for early detection [2]. The diagnostic challenge stems from nonspecific symptomatology—particularly exertional dyspnea—and the frequent overlap with comorbid conditions, resulting in delayed diagnosis and poor clinical outcomes.

The 6‐min walk test (6MWT), a validated measure of functional capacity in cardiopulmonary diseases, provides prognostic value in established PH through distance quantification‐the 6‐min walk distance (6MWD) [3], which majorly measures the distance walked in 6 min at a self‐paced speed. However, 6MWD interpretation is confounded by demographic variables including age, gender, and body composition [4]. Emerging evidence highlights the superior diagnostic potential of dynamic physiological parameters during 6MWT: oxygen desaturation patterns predict mortality in chronic obstructive pulmonary disease [5], while heart rate (HR) kinetics correlate with disease severity in heart failure [6]. Notably, exercise‐induced SpO₂ decline and impaired HR recovery demonstrate prognostic significance in sarcoidosis‐associated PH [7], yet their utility for early PH detection remains unexplored.

This study investigates the temporal dynamics of SpO₂ and HR during 6MWT to develop a novel screening model for PH in patients with unexplained exertional dyspnea. By integrating continuous physiological monitoring with propensity‐matched analysis, we aim to establish objective thresholds that enhance screening efficiency beyond conventional 6MWD metrics.

2. Methods

2.1. Study Design and Population

This study was a single‐center, retrospective, observational cohort study. Patients presenting dyspnea on exertion underwent the 6MWT at the Center of Structural Heart Disease, Zhongnan Hospital of Wuhan University, between January 2022 and October 2023 were continuously reviewed. PH was diagnosed based on the criterion of the mean pulmonary artery pressure (mPAP) measured by right heart catheterization exceeding 20 mmHg according to current guidelines [2]. Exclusion criteria were: (1) pre‐existing pulmonary pathologies, such as COPD, interstitial lung disease, and active pneumonia; (2) contraindications to RHC; and (3) incomplete data. Eligible participants were stratified into PH and non‐PH groups for comparative analysis. The study protocol adhered to the Declaration of Helsinki and received approval from the ethics committee of Zhongnan Hospital of Wuhan University (2023046K). Given the retrospective observational design of the study, informed consent was waived.

2.2. Data Collection

Demographic and clinical parameters were systematically extracted from electronic medical records, including age, sex, height, weight, body mass index (BMI), and N‐terminal pro‐B‐type natriuretic peptide (NT‐proBNP). Hemodynamic measurements were obtained from RHC. Transthoracic echocardiography provided left ventricular functional assessment (left ventricular ejection fraction, LVEF), right ventricle dilation (ratio of right ventricle to right atrium and left ventricle separately, RV/RA, RV/LV), and tricuspid regurgitation. The 6MWT protocol followed American Thoracic Society guidelines [3], conducted in a standardized 30‐meter hospital corridor under ambient conditions. Participants refrained from practice walks to prevent learning effects. Continuous physiological monitoring was achieved using wearable devices. Specifically, the wearable device consisted of an oxygen saturation meter and an ECG recorder. The device continuously recorded the SpO₂ and HR of the patients and transmitted the data to the mobile workstation at 1‐min interval. This device offers a dynamic recording of the SpO₂ and HR without significant missing data. The 6MWD was recorded by trained technicians. Predicted 6MWD values were derived from validated equations incorporating age and BMI [8].

2.3. Statistical Analysis

Continuous variables were expressed as mean ± standard deviation (SD) for normally distributed data or median (interquartile range, IQR) for non‐parametric distributions, assessed using the Shapiro‐Wilk test. Categorical variables were reported as frequencies (percentages). Intergroup comparisons utilized Mann–Whitney U tests for non‐normally distributed continuous variables and χ² tests (or Fisher's exact tests for expected frequencies < 5) for categorical variables. Propensity score‐matched analyses was performed including gender, age, BMI, and 6MWD as matching criteria. Comparison between groups employed paired t‐tests (normal distribution) or Wilcoxon signed‐rank tests (non‐normal distribution) for continuous variables, with McNemar's test for categorical variables. Multivariable logistic regression identified independent predictors of PH, incorporating 6MWT‐derived parameters with established clinical covariates. Model discrimination was evaluated through receiver operating characteristic (ROC) curve analysis, reporting the area under the curve (AUC) with 95% confidence intervals. Optimal cut‐off values were determined using Youden's index. All analyses were performed using SPSS Statistics (version 25.0; IBM Corp.), with two‐tailed p‐values < 0.05 considered statistically significant.

3. Results

There were 731 patients presenting with exertional dyspnea as their primary complaint; 442 patients underwent the 6MWT upon admission. After excluding 162 patients who did not undergo RHC and 100 patients with incomplete clinical data, 180 patients were included in the final analysis (Figure 1). Basic clinical characteristics of study population are presented in Table 1. The study population were aged from 32 to 50 years, with a majority being female (63.8%, n = 115). The major comorbidities in the non‐PH group were hypertension (n = 33, 32.7%), stroke (n = 29, 28.7%), and coronary artery disease (n = 23, 22.8%). The PH group had a similar spectrum of comorbidities, but with fewer patients than the non‐PH group. Comorbidities in the PH group included hypertension (n = 9, 11.4%), coronary artery disease (n = 9, 11.4%), and stroke (n = 5, 6.3%). Tricuspid regurgitation and mild dilation of RV were seen more in the PH group than non‐PH (both p < 0.001). The left ventricular systolic function, as assessed by echocardiography, showed no significant intergroup differences (p = 0.741). NT‐proBNP level was significantly elevated in PH group compared to the non‐PH group (p < 0.001). Functional capacity assessment revealed a higher proportion of PH patients classified as WHO functional class III–IV compared to non‐PH patients (p < 0.001). The predominant etiology of PH was pulmonary artery hypertension (PAH, n = 59, 74.7%). Others subtypes included PH associated with left heart disease (PH‐LHD, n = 6, 7.6%), associated with portal hypertension (PoPH, n = 3, 3.8%), and associated with fibrosing mediastinitis (n = 1, 1.3%).

Figure 1.

Study flowchart.

Table 1.

Clinical characteristics of the study population.

	Total (n = 180)	Non‐PH (n = 101)	PH (n = 79)	p value
General
Female, n (%)	115 (63.8)	64 (63.3)	40 (63.4)	0.923
Age (years)	44.5 ± 15.5	43.4 ± 13.9	45.9 ± 17.2	0.241
BMI (kg/m2)	23.2 ± 3.6	23.0 ± 3.3	23.3 ± 3.9	0.892
HB (g/L)	132.9 ± 21.5	130.1 ± 16.3	136.5 ± 26.4	0.151
NT‐pro BNP (pg/mL)	261.0 ± 622.9	64.0 ± 101.3	512.9 ± 873.4	＜0.001
WHO functional class, n (%)				＜0.001
I–II	155 (86.1)	101 (100)	54 (68.4)
III–IV	25 (13.9)	0	25 (31.6)
Echocardiography
LVEF (%)	64 ± 7	65 ± 4	63 ± 9	0.741
RV/RA	0.9 (0.8, 1.0)	0.9 (0.9, 1.0)	0.9 (0.8, 1.1)	0.924
RV/LV	0.7 (0.7, 0.8)	0.7 (0.6, 0.8)	0.8 (0.7, 1.2)	< 0.001
Tricuspid regurgitation, n (%)				< 0.001
Mild	75 (41.7)	35 (34.7)	40 (50.6)
Moderate to severe	24 (13.3)	1 (0.1)	23 (29.1)
PH etiology, n (%)
IPAH	—	—	37 (46.8)
CHD‐PAH	—	—	19 (24.1)
CTEPH	—	—	10 (12.7)
CTD‐PAH	—	—	3 (3.8)
Others	—	—	10 (12.7)
Hemodynamics
mPAP (mmHg)	19.0 (16.0, 29.0)	16.0 (14.5, 18.5)	31.0 (23.0, 49.0)	＜0.001
PVR (WU)	2.4 (1.2, 4.8)	2.1 (1.6, 2.8)	4.4 (3.0, 8.6)	＜0.001
TPR (WU)	3.8 (2.8, 5.9)	3.0 (2.5, 3.7)	6.0 (4.0, 10.0)	＜0.001
CO (L/min)	5.8 (4.9, 6.8)	6.1 (5.1, 6.7)	5.5 (4.4, 6.9)	0.237
CI (L/min × m2)	3.4 (2.8, 4.0)	3.4 (2.8, 3.9)	3.3 (2.7, 4.2)	0.905
PCWP	8.0 (6.0, 8.0)	5.0 (3.0, 8.0)	10.0 (6.0, 14.0)	< 0.001
6MWT
6MWD (m)	492.1 ± 92.0	509.8 ± 74.9	469.5 ± 106.4	0.019
6MWD pred (m)	645.0 ± 108.8	651.8 ± 93.5	636.4 ± 125.8	0.165
6MWD pred%	77.6 ± 16.4	79.2 ± 12.6	75.5 ± 20.2	0.018
Borg scores, n (%)				0.205
0	31 (17.2)	23 (22.7)	8 (10.1)
0.5	37 (20.6)	22 (21.7)	15 (19.0)
1	42 (23.3)	22 (21.7)	20 (25.3)
2	48 (26.7)	28 (27.7)	20 (25.3)
3	17 (9.4)	6 (5.9)	11 (13.9)
4	4 (2.2)	0	4 (5.1)
5–6	1 (0.6)	0	1 (1.3)

Note: The data are expressed as the mean ± standard deviation, quantity (percentage), or median (interquartile range).

Abbreviations: ΔHR_max−rest= HR_max − HR_rest, ΔSpO_2rest−min = SpO_2rest − SpO_2min, 6MWD = 6‐min walking distance, 6MWD pred = predicted 6‐min walking distance, 6MWD pred% = 6MWD/6MWD pred, 6MWT= the 6‐min walking test, BMI = body mass index, CHD = congenital heart disease, CI = cardiac index, CO = cardiac output, CTD = connective tissue disease, HB = hemoglobin, HR_max = maximum heart rate during 6MWT, HR_rest = HR at rest, IPAH = idiopathic pulmonary arterial hypertension, LV=left ventricle, LVEF = left ventricular ejection fraction, mPAP = mean pulmonary arterial pressure, NT‐pro BNP = N‐terminal pro‐B‐type natriuretic peptide, PCWP = pulmonary capillary wedge pressure, PH = pulmonary hypertension, PVR = pulmonary vascular resistance, RV=right ventricle, SpO_2min = the minimum oxygen saturation during 6MWT, SpO_2rest = the oxygen saturation at rest, TPR = total pulmonary vascular resistance.

Comparative analysis of 6MWT‐derived parameters revealed significant intergroup differences. PH patients demonstrated reduced exercise capacity, evidenced by lower absolute 6‐min walk distance (6MWD) (469.5 ± 106.4 m vs. 509.8 ± 74.9 m, p = 0.019) and decreased percentage of predicted 6MWD (6MWD_pred%) (75.5 ± 20.2% vs. 79.2 ± 12.6%, p= 0.018). However, perceived exertion levels, as measured by Borg dyspnea scale, showed no significant difference between groups (p = 0.205). Continuous monitoring during the 6MWT demonstrated that: PH patients exhibited a peak HR of 128 ± 22 bpm concurrent with a SpO₂ trough of 85 ± 12% at 4–5 min into the test. Non‐PH patients reached a peak HR of 119 ± 18 bpm at the end of the test. Their SpO₂ trough of 93 ± 6% occurred at 4 min into the test (Figure 2). During the test, both groups achieved maximal exertion rather than reaching physiologic extremes, as the maximal HR of PH patients was 73.6 ± 12.7% of the predicted maximum HR (174 ± 17 bpm), and that of non‐PH patients was 67.7 ± 10.8% (predicted maximum HR = 176 ± 13 bpm). These HR parameters remained relatively stable after peaking, forming a characteristic plateau phase.

Figure 2.

Continuous monitoring of SpO₂ and HR during the 6MWT. Continuous monitoring of SpO₂ and HR during the 6MWT in non‐PH (A) and PH patients (B). The red dot indicates the change in SpO₂. The black box indicates the change in HR. Data is presented as mean and interquartile. Abbreviations: bpm, beats per minute; HR, heart rate; min, minute; SpO₂, saturation of peripheral oxygen.

To account for potential confounding factors and eliminate bias from differential walking distances, we performed propensity score matching between PH and non‐PH groups. From the original cohort, 70 matched pairs were successfully identified, while 9 PH patients (11.4%) were excluded due to inability to identify suitable matches within the non‐PH group. As shown in Table 2, PH patients demonstrated a slightly lower but bigger variation of SpO₂ as well as lower HR compared to non‐PH patients before the test (both < 0.05). During the 6MWT, PH patients experienced a 4% lower SpO₂ desaturation compared to non‐PH patients (p < 0.001), while the maximal HR increased more than 10% in PH patients than non‐PH patients (p = 0.002). Notably, significant differences were observed in the dynamic changes of SpO₂ and HR during the 6MWT. To further elucidate these findings, we compared the decrease of SpO₂ from rest to minimal levels (ΔSpO_{2rest −min}) and the increase of HR from rest to maximal levels (ΔHR_{max −rest}). The ΔSpO_{2rest −min} in PH patients was approximately threefold greater than that in non‐PH patients (p < 0.001), while the ΔHR_{max −rest} was 1.5‐fold greater in PH patients (p < 0.001). Both ΔSpO_{2rest −min} [HR: 1.084 (1.038, 1.132), p < 0.001] and ΔHR_{max −rest} [HR: 1.106 (1.030, 1.189), p = 0.006] were independently associated with the presence of PH.

Table 2.

SpO₂ and HR in case‐control matched groups.

	Non‐PH (n = 70)	PH (n = 70)	p value
Female, n (%)	44 (62.8%)	44 (62.8%)	1
6MWD (m)	494.1 ± 68.3	495.6 ± 71.5	0.750
SpO2rest (%)	97 (96, 98)	96 (94, 97)	＜0.001
HRrest (bpm)	84.3 ± 11.6	79.10 ± 12.2	0.021
SpO2min (%)	94 (91, 96)	90 (70, 93)	＜0.001
HRmax (bpm)	115 (106, 130)	132 (115, 141)	0.002
ΔSpO2rest −min (%)	4 ± 6	9 ± 9	＜0.001
ΔHRmax −rest (bpm)	34 ± 14	51 ± 21	＜0.001

Note: The data are expressed as the mean ± standard deviation, quantity (percentage), or median (interquartile range).

Abbreviations: ΔHR_max−rest = HR_max − HR_rest, ΔSpO_2rest−min = SpO_2rest − SpO_2min, 6MWD = 6‐min walking distance, HR_max = maximum heart rate during 6MWT, HR_rest = HR at rest, PH = pulmonary hypertension, SpO_2rest = SpO₂ at rest, SpO_2min = the minimum SpO₂ during 6MWT.

ROC curve analysis was performed to evaluate the predictive efficacy of PH based on ΔSpO_{2rest −min} and ΔHR_{max −rest} (Figure 3). The area under the curve (AUC) of ΔSpO_{2rest −min} was 0.715 (95% CI: 0.630–0.800, p < 0.001). The best cut‐off value for ΔSpO_{2rest −min} was established at ≥ 5%, yielding a sensitivity of 58.57% (95% CI: 46.20–70.20) and a specificity of 75.71% (95% CI: 64.00–85.20). The AUC for ΔHR_{max −rest} was 0.740 (95%CI: 0.656–0.823, p < 0.001). The best cut‐off value for ΔHR_{max −rest} was determined to be ≥ 42 bpm, with a sensitivity of 64.29% (95% CI: 51.90–75.40) and a specificity of 80.00% (95% CI: 68.70–88.60). When taking the ΔSpO_{2rest −min} ≥ 5% and ΔHR_{max −rest} ≥ 42bpm as predictors for PH, exertional dyspneic patients with one predictor had threefold increased likelihood of having PH than those without any predictors (p = 0.002). Furthermore, patients who exhibited both predictors faced a eightfold higher risk of PH compared to those without any predictors (p < 0.001), as illustrated in Figure 4. The risk‐stratification model incorporating these thresholds demonstrated incremental predictive value for PH (C‐statistic = 0.786, 95% CI: 0.710–0.863, p < 0.001).

Figure 3.

ROC curve analysis of independent factors for PH. (A) ROC curve analysis of ΔSpO_{2rest −min} for PH. (B) ROC curve analysis of ΔHR_{max −rest} for PH. The blue line presents the ROC curve. Grey lines present the 95% CI. Abbreviations: ΔHR_max−rest, the increase in HR from rest to maximal levels; ΔSpO_2rest−min, the decrease in SpO₂ from rest to minimal levels; AUC, area under the curve; CI, confidence interval; PH, pulmonary hypertension; ROC, receiver operating characteristic.

Figure 4.

Risk assessment of underlying PH. Forrest plot presents the risk of underlying PH in dyspneic patients with different numbers of predictors. Abbreviations: CI, confidence interval; HR, hazard ratio; PH, pulmonary hypertension. *p < 0.01; **p < 0.001.

4. Discussion

The 6MWT is a widely used test for the assessment of cardiopulmonary function among patients with cardiovascular and pulmonary disorders [9]. The walking distance achieved during the test has been established as a prognostic indicator for mortality and morbidity in PH and other conditions [10, 11]. In this study, we explored the parameters derived from 6MWT to identify the predictors of PH in patients presenting with dyspnea on exertion. Our results revealed that PH patients exhibited decreased 6MWD and reduced 6MWD capacity, which are similar as previous study [12]. Furthermore, our study unveiled two key findings: (1) the peak HR and trough of SpO₂ occurred at 4–5 min into the test instead of at the end of the test; (2) exercise‐induced SpO₂ desaturation (ΔSpO_{2rest −min}) and HR response (ΔHR_{max −rest}) during 6MWT were independent predictors of PH in exertional dyspneic patients. Most importantly, our study is the first to use predictors derived from the 6MWT, ΔSpO_{2rest −min} ≥ 5%, and ΔHR_{max −rest} ≥ 42 bpm, for early assessment of underlying PH. The presence of one of these predictors in the 6MWT identified PH in 60% of dyspneic patients, while the positive rate soared to 84% when both predictors were present.

PH stands as a critical global medical challenge and is strongly associated with poor outcomes in various diseases, including heart valvular disease, lung disease, and others. Early diagnosis and treatment of PH are essential for improving prognosis, in addition to addressing the underlying medical conditions. Dyspnea on exertion, while a cardinal symptom of PH, is also prevalent in heart failure and numerous other diseases, posing challenges in accurately predicting PH risk among affected patients [13]. Currently, echocardiography is recommended for PH screening. Chest CT and pulmonary function tests are employed in the strategies to identify PH in patients with interstitial lung disease [14]. However, operator experience and equipment variations impact their prevalence and accuracy. Our study showed that RV/RA did not differ significantly between non‐PH and PH groups, RV/LV exhibited notable disparities. Tricuspid regurgitation severity, though more pronounced in the PH group, showed comparable mild cases between groups. Collectively, these findings underscore the limitations of echocardiography as a screening tool for PH.

Notably, PH is characterized by progressive exercise intolerance, which is often undetectable at early stages. In our study population, 43.9% of whom were PH patients, the majority were in WHO functional class I–II, underscoring the potential clinical relevance of exercise‐induced SpO₂ and HR responses during early disease phases. Oki et al. showed a correlation between SpO₂ < 89% during 6MWT and dilated pulmonary artery in COPD [15]. Desaturation during 6MWT predicted mortality in patients with lung disease [16]. HR response during the 6MWT serves as a critical marker of cardiac effort and physiological reserve, which is reflected not only by absolute HR elevation but also by the rate of HR increase, recovery kinetics, and variability. Ramos et al. reported that patients whose recovery of HR > 18 bpm at the first minute had better outcomes [17]. The variability of HR response during the 6MWT correlated with the distance in PH patients and might reflect their reduced physiological reserve [18]. Our study further supported the HR response in PH patients diverges significantly from non‐PH patients as HR changed more in PH group than that in non‐PH group. The reason might be reduced right ventricular compliance and elevated pulmonary vascular resistance, which may blunt HR augmentation or delay its peak, reflecting impaired cardiac reserve. Our study here not only investigated the relationship between exercise‐induced desaturation and the presence of PH, but also employed it along with HR response for PH screening. These findings will shift the clinical significance of the 6MWT from prognosis assessment to disease prediction, providing a new option for PH screening.

6MWT has been considered as a convenient and safe test to assess cardiopulmonary function, however, it was not suggested to perform the test when the patient's SpO₂ ≤ 85% for safety considerations [9]. However, Afzal et al. observed in a large cohort (n = 549) and reported that SpO₂ < 80% during a 6MWT was not associated with adverse events; therefore, it was unwarranted that 6MWT was terminated for stable patients when their SpO₂ dropped lower than 80% [19]. In this study, all patients had a SpO_2rest ≥ 94%, two of them had minimal SpO₂ that was lower than 80%. Both of the patients had unrepaired cardiac defects. Different from other patients, cardiac defects might lead to right‐to‐left shunt with the increasing of right atrium pressure, which would cause a significant decline in SpO₂. None of our patients claimed to be uncomfortable and asked to stop the test; nonetheless, they achieved no greater effort after the peak HR and lowest SpO₂ occurred during the 6MWT, as their HR and SpO₂ reached a plateau thereafter. The fact that patients stopped to make an effort during the 6MWT after reaching their maximal HR and minimal SpO₂ may also explain the study result that HR_max and SpO_2min occurred at 4 ~ 5 min upon the 6MWT instead of at the end of the test.

Different from pulmonary function test and cardiopulmonary exercise test, 6MWT is performed at low complexity. The walk distance derived from the test is applied as the major predictor for adverse events during follow‐up. A 6MWD of less than 250 m was recognized as a predictor of mortality in patients with COPD [5]. Changes in 6MWD were also considered indicative of worsening cardiac function in patients with heart failure [20]. In PH, 6MWD serves as a significant factor for risk assessment within 1 year [13]. Even in simplified assessment systems, patients are successfully stratified into four categories based on 6MWD, enabling the identification of their condition and prediction of 1‐year mortality risk. However, the limitation of 6MWD lies in its susceptibility to factors such as patients' familiarity with and effort exerted during the 6MWT, which undermines its standalone objective assessment value [21, 22]. Additionally, the current reference ranges for 6MWD categorization are too broad, resulting in imprecise patient stratification, often necessitating calibration with other indicators. In this study, the tests were conducted under conditions where patients could comprehend the instructions, reflecting their true HR, SpO₂, and distance changes. A case‐control matched analysis was further applied to eliminate the effects of age, height, and walking distance. Even so, changes in HR and SpO₂ remained independent predictors of PH, demonstrating their tight correlation with the disease.

Our study has several limitations. First, this was a single‐center, retrospective, observational study in which 180 patients were included in the final analysis, which may have introduced selection bias during patient recruitment. A prospective study with a larger sample size, and more diversity in patients would offer more solid data for better assessment. Second, our study population included patients with dyspnea, and with a narrow spectrum of comorbidities. The lack of more diseases may cause biases in the conclusions. Third and last, follow‐up for PH patients and non‐PH patients whose ΔSpO_{2rest −min} ≥ 5% and ΔHR_{max −rest} ≥ 42 bpm during the 6MWT was not included in our study. The early diagnosis of PH may change the long‐term prognosis, which needs further investigation.

In conclusion, our study expands the usefulness of the 6MWT in clinical practice. The ΔSpO_{2rest −min} ≥ 5% and ΔHR_{max −rest} ≥ 42 bpm during the 6MWT were independent predictors of PH among patients with dyspnea. The 6MWT should be applied not only for risk stratification and cardiopulmonary function assessment, but also to screen the PH patients using the exercise‐induced desaturation and HR response during the test.

Author Contributions

Haojie Zhang and Menghuan Yan enrolled patients, acquired clinical data, analyzed data, drafted the work, and approved the last version of the manuscript. Feng Li, Qi Chen, and Rui Lu enrolled patients, drafted the work, and approved the last version of the manuscript. Ziyu Wang analyzed data and approved the last version of the manuscript. Xuan Zheng and Gangcheng Zhang designed the study, analyzed data, drafted the work, and approved the last version of the manuscript.

Ethics Statement

The study has been approved by the Ethics Committee of Zhongnan hospital of Wuhan University (2023046K). Given the retrospective observational design of the study, informed consent was waived.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.