Exercise Induced Changes of Vital Signs in Adults with Sickle Cell Disease

Abstract

The six-minute walk test (6MWT) has been used in patients with sickle cell disease (SCD), in conjunction with tricuspid regurgitant velocity (TRV) and plasma N-terminal pro-brain natriuretic peptide (NT-pro BNP), to assess risk of having pulmonary hypertension. Exercise induced vital signs change (VSC) are predictors of clinical outcomes in other diseases. In this study, we assess the predictors and prognostic value of 6MWT VSC in adult SCD patients.

Data from a multinational study of SCD patients (Treatment of Pulmonary Hypertension with Sildenafil: walk-PHaSST) was used to calculate the 6MWT VSC. Predictors of VSC were identified by multivariable analysis and survival analysis was conducted by the Cox proportional hazard method.

An increase in heart rate was observed in 90% of the 630 SCD adults, 77% of patients had an increase in systolic blood pressure (SBP) and 50% a decrease in oxygen saturation. TRV (OR = 1.82, P = 0.020), absolute reticulocyte count (OR = 1.03, P < 0.001) and hemoglobin (OR = 0.99, P = 0.035) predicted oxygen desaturation ≥ 3% during the 6MWT. In adjusted analysis, SBP increase during the 6MWT was associated with improved survival (HR=0.3, 95%CI: 0.1–0.8).

Increases in heart rate and blood pressure, as well as oxygen desaturation, are common in adults with SCD during the 6MWT. VSC is associated with markers of anemia and TRV and can be used for risk stratification. Any increase in SBP during the 6MWT was associated with improved survival and may be indicative of a patient’s ability to increase stroke volume.

Introduction

Cardiovascular dysfunction is a frequent complication of sickle cell disease (SCD) that is associated with worse outcomes^1,2. The effects of chronic intravascular hemolysis and anemia, compounded by vaso-occlusion and ischemia, adversely affect the cardiovascular system of people with sickle disease as they age^3,4. A combined study of two cohorts of adults with SCD found that 36% of participants were members of a cardiovascular high-risk group defined by elevated pulmonary arterial pressure measured by tricuspid regurgitation velocity (TRV), increased plasma N-terminal pro-brain natriuretic peptide (NT-pro BNP), and chronic kidney disease¹. The newly characterized cardiovascular high-risk group was found to be at more than five times the risk of mortality compared to other patients in the cohort¹. Each measure of cardiovascular function used to define the high-risk group has also been independently associated with increased risk of mortality^5–7. TRV measured by Doppler-echocardiography is used to detect elevated systolic pulmonary arterial pressure which is predictive of pulmonary hypertension. Research has shown that the development and severity of pulmonary hypertension increases the risk of mortality in the adult SCD population^7–9. Concordantly, increased TRV ≥ 2.5 estimated by Doppler-echocardiography has been associated with hazard ratios of mortality ranging from 2.3 to 5.1 and a TRV ≥ 3.0 was associated with a hazard ratio for mortality of 11.1 in a large international cohort^6,7,10–12. Plasma NT-pro BNP is a marker of ventricular stress that is also an established predictor of increased mortality for adults with SCD⁷. Additionally, cardiovascular dysfunction has been cited in large scale investigations of autopsy and death certificates as a leading a cause of mortality in sickle cell disease patients¹³.

Identification of SCD patients at risk for the development of cardiovascular sequelae is necessary to improve clinical outcomes, and the gold standard for the diagnosis of pulmonary hypertension is right heart catheterization. However, right heart catheterization is an invasive procedure, relatively expensive and not available in resource poor settings internationally. Exercise testing can identify cardiovascular dysfunction that may not be evident at rest^14,15. The six-minute walk test (6MWT) is an established exercise assessment used in the SCD population that can increase the specificity of cardiovascular diagnostic testing and can also be indicative of disease manifestations¹⁶. To reduce the number of invasive right heart catheterizations needed to screen for pulmonary hypertension, it has been recommended that patients with a high estimated pulmonary pressure measured by TRV assessment be further stratified using plasma NT-pro BNP and six-minute walk testing^16–18. Although TRV is not as specific as right heart catheterization for diagnosis of pulmonary hypertension, studies have confirmed that sensitivity for identification of high-risk patients increases as TRV values increase and can be augmented by inclusion of plasma NT-pro BNP level analysis^9,19,20. It should also be noted that a TRV value of ≥ 2.5 is approximately 2-standard deviation above the normal age matched control values, while a TRV value of ≥ 3.0 is approximately 3-standard deviations above the age matched control values²¹. A mean pulmonary artery pressure ≥ 20 mm Hg is 2-SD above normal and ≥ 25 mm Hg is 3-SD above normal, and new consensus definitions now identify a mean pulmonary artery pressure ≥ 20 mm as abnormal based on associated risk of mortality²². It is therefore appropriate to consider a TRV value of ≥ 2.5 as abnormal and indicative of higher risk of death and higher risk of having pulmonary hypertension. A recent large population screening study evaluated TRV and risk of death using unbiased receiver operator curve analysis and found that a TRV value of ≥ 2.5 had the highest sensitivity and specificity characteristics for identifying the SCD population at highest risk of death²³. Indeed, adult SCD patients with a TRV ≥ 2.5 m/sec, a six-minute walk distance (6MWD) less than 333 meters or NT-pro BNP plasma level above 164 pg/mL, have an increased probability of having a mean pulmonary artery pressure above 25 mm Hg measured by right heart catheterization, which is indicative of pulmonary hypertension^9,13,24. Additionally, 6MWD is known to be associated with several measures of cardiological dysfunction in the adult SCD population. Markers of anemia and hemolysis are associated with six-minute walk induced oxygen desaturation in the pediatric SCD population²⁵. Yet to date, no one has analyzed the other physiological parameters measured during this simple inexpensive test^25,26.

Exercise induced vital sign change (VSC) has been shown to be a predictor of clinical outcomes in other diseases such as pulmonary hypertension, pulmonary fibrosis and heart failure^27–29. The prognostic significance of 6MWT VSC for the adult SCD population has yet to be determined. To address the knowledge gap regarding the importance of exercise induced VSC in adults with SCD, we aim to assess the distribution and predictors of VSC, as well as identify associations with patient survival in a large cohort of adults with SCD.

Methods

We analyzed data of adult SCD patients from the multicenter international trial Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy (walk-PHaSST). The screening stage of walk-PHaSST investigated the relationship between echocardiographic markers of pulmonary hypertension and left ventricular diastolic dysfunction with markers of SCD severity and patients’ survival⁷. The methodology of original data collection from this cohort is described in greater detail in^7,30. The cohort consisted of 720 SCD patients recruited in the United States and the United Kingdom. At baseline, study participants were evaluated for history of clinical events, lifetime treatments, and by physical examination. Patients also underwent laboratory screening, transthoracic doppler echocardiography, and completed a six-minute walk test with vital sign monitoring⁷. Measurement of hemoglobin oxygen saturation was measured by pulse oximetry at room temperature and atmospheric oxygen pressure³⁰. Laboratory testing including serum chemistry profile, complete blood count, and lactate dehydrogenase was performed from samples collected at screening visits⁷. Local review boards and ethics committees were consulted prior and during the study. Written consent was obtained from all participants³⁰.

The 6MWT were performed to American Thoracic Society standards, using a hard-flat surface in an indoor hallway longer than 30 meters. Subjects were instructed to cover the greatest possible distance by walking at a self -determined pace for six minutes. The distance walked was recorded in meters for all subjects. Transcutaneous pulse oximetry, hemoglobin oxygen saturation measurements were obtained before and immediately after the six-minute walk test²⁵.

Statistical Analysis

Vital sign change was calculated as post 6MWT value – pre 6MWT value. For subsequent analysis, we used the ratio of vital sign change divided by the distance walked during the 6MWT, multiplied by 100, to ascertain the ratio of VSC per 100 meters walked (Supplementary Figure 1). Patient characteristics and vital signs are presented using median (interquartile ranges) or n (%). Pearson correlation was used to measure association between continuous variables and vital sign changes. We used Student’s t-test or one-way ANOVA to test the vital signs changes by categorical variables. We applied two modelling strategies to test the multivariable association between clinical variables and change in vital signs. In the first model, we used a backward variable selection using all variables in univariate analysis. Variables with P < 0.05 were kept in the model. Alternatively, we used the best generalized modelling. This approach keeps variables regardless of their P value in a model which provides the best prediction ability using the Bayesian Information Criteria. All coefficients in multivariable models were standardized using the ratio of SD (X)/ SD (Y). Crude and adjusted (for NT-pro BNP and tricuspid regurgitation velocity) Hazard ratios were reported for all vital sign changes. Proportional Hazard was tested in models. All analyses were performed by STATA 16.2 (StataCorp, College Station, TX) and R 4.0.3 (R Core Team, 2020).

Results

Patients Characteristics

The median age of the 630 adult SCD patient in this study was 37 years, 47% were male, 77% had the HbSS phenotype and 19% were HBSC (Supplementary Table 1). During the 6MWT, 90% of our cohort had an increase (defined as any increase from baseline) in heart rate and 77% exhibited increases in systolic blood pressure. Fifty eight percent had an increase in diastolic blood pressure and 68% had an increase in pulse pressure. Fifty percent of patients exhibited a decrease in oxygen saturation of at least 1%. The median values for the change of heart rate, O2 saturation, systolic blood pressure, and diastolic blood pressure were 16 beats per minute, −1%, 8 and 2 mmHg, respectively. The median change of pulse pressure was 6 mmHg (Supplementary Table 2).

Bivariate Predictors of the Ratio of Vital Sign Change to Distance

There was a strong correlation of 6MWD with change of heart rate, systolic and pulse blood pressure ranging from 0.40 to 0.22 (P <0.001). The correlation 6MWD with oxygen saturation and diastolic blood pressure was −0.13 (P = 0.002) and 0.07 (P = 0.070), respectively. So, we used the ratio of VSC to distance to control the potential effect of 6MWD in analysis. Bivariate analysis found positive correlations between the change in heart rate during the 6MWT and age (r=0.09), TRV (r=0.08), ALT (r=0.10), and BUN (r=0.11). Bivariate analysis also identified negative correlations between the change in heart rate during the 6MWT and hemoglobin (r=−0.11), red blood cell (RBC) counts (r=−0.10) (Supplementary table 3, Figure 1a). Heart rate increased more in patients with history of pneumonia/acute chest syndrome (p=0.007), history of pulmonary hypertension (p=0.039), and history of hip replacement procedure (p=0.018). Heart rate increased more in groups of patients who were on hydroxyurea treatment (p=0.004), with a past history of smoking (p=0.005), and in patients meeting the criteria of NYHA/WHO III and IV classes (p=0.021) (Table 1).

Figure 1a:

Correlogram of univariate association between continuous variables and change of vital sign during 6MW test. The color in each cell identifies the direction and magnitude of correlation between variables in row and column.

Table 1:

Medical History Correlation with Change of Vital Signs

	Δ heart rate per 100m		Δ O2 saturation % per 100m		Δ systolic blood pressure per 100m		Δ diastolic blood pressure per 100m		Δ pulse pressure per 100m
Characteristic	Median (p25-p75)	p value	Median (p25-p75)	p value	Median (p25-p75)	p value	Median (p25-p75)	p value	Median (p25-p75)	p value
Gender: Female	4.4 (1.6–7.6)	0.3	0.0 (−0.5–0.00)	0.3	1.6 (0.0–3.8)	0.1	0.3 (−0.7–1.7)	0.3	1.3 (−0.5–3.1)	0.3
Male	3.9 (1.4–7.4)		−0.2 (−0.5–0.00)		2.2 (0.5–3.8		0.5 (−0.2–1.6)		1.4 (0.0–3.0)
SS, Genotype	4.2 (1.5–7.5)	0.1	−0.2 (−0.6–0.00)	0	2.1 (0.3–3.9)	0.6	0.4 (−0.5–1.6)	0.5	1.4 (−0.2–3.0)	0.2
Other Genotype	3.6 (1.0–7.2)		0.0 (−0.2–0.00)		1.5 (0.0–3.6)		0.5 (−0.5–1.8)		1.3 (−0.6–2.9)
No α-gene deletion	4.2 (1.4–7.7)	0.8	−0.2 (−0.5–0.00)	0.6	2.1 (0.4–3.9)	0.3	0.5 (−0.5–1.6)	0.5	1.5 (−0.2–3.2)	0.6
α-gene deletion	4.0 (1.5–7.2)		0.0 (−0.5–0.00)		2.0 (0.0–3.6)		0.4 (−0.7–1.6)		1.2 (−0.3–2.9)
Hydroxyurea use: Current	4.8 (1.4–8.5)	0.004	−0.2 (−0.4–0.00)	0.8	2.3 (0.4–3.9)	0.2	0.4 (−0.5– 1.7)	0.7	1.4 (−0.2–3.2)	0.4
None	3.4 (1.3–6.9)		0.0 (−0.5–0.00)		1.6 (0.0–3.4		0.4 (−0.5–1.5)		1.3 (−0.4–2.9)
Past	4.4 (2.0–8.6)		−0.2 (−0.7–0.00)		2.0 (0.2–4.3)		0.5 (−0.7–1.8)		1.6 (−0.4)
Smoking status: Current	3.0 (1.4–6.0)	0.005	−0.2 (−0.4–0.00)	0.6	1.6 (−0.1–3.9)	0.6	0.3 (−0.9–1.6)	0.7	1.1 (−0.4–3.1)	0.9
Never	4.2 (1.4–7.7)		−0.2 (−0.6–0.00)		2.2 (0.3–4.0)		0.5 (−0.5–1.6)		1.5 (−0.3–3.1)
Past	4.9 (1.7–8.5)		0.0 (−0.5–0.00)		2.0 (0.4–3.1)		0.5 (−0.3–1.6)		1.2 (−0.2–2.9)
History of pneumonia/ACS	4.5 (1.7–8.1)	0.001	−0.2 (−0.5–0.0)	0.3	2.0 (0.2–4.0)	0.2	0.4 (−0.5–1.6)	0.5	1.4 (−0.4–3.1)	0.4
No	3.0 (1.2–6.5)		0.0 (−0.5–0.0)		1.9 (0.3–3.6)		0.5 (−0.5–1.7)		1.3 (−0.2–2.9)
History of PH	3.4 (1.3–6.5)	0.039	−0.2 (−0.6–0.0)	0.5	2.2 (0.7–3.9)	0.1	0.2 (−0.7–1.6)	0.4	1.6 (0.0–3.5)	0.039
No	4.3 (1.5–7.9)		0.0 (−0.5–0.0)		1.8 (0.0–3.6)		0.5 (−0.5–1.6)		1.3 (−0.5–2.9)
History hip replacement	6.1 (2.5–9.2)	0.018	−0.2 (−0.6–0.0)	0.1	2.2 (1.2–4.3)	0.1	0.5 (−1.0–1.7)	0.5	1.8 (−0.2–3.7)	0.2
No	3.8 (1.4–7.3)		0.0 (−0.5–0.0)		1.9 (0.2–3.6)		0.5 (−0.5–1.6)		1.3 (−0.3–3.0)
NYHA/WHO		0.021		0.1		0.1		0.4		0.1
Class I	3.4 (1.3–7.3)		0.0 (−0.5–0.0)		1.8 (0.2–3.6)		0.5 (−0.5–1.6)		1.2 (−0.3–2.7)
Class II	4.7 (2.2–8.4)		−0.2 (−0.6–0.0)		2.1 (0.3–4.1)		0.2 (−0.7–1.4)		1.6 (−0.5–3.7)
Class III & IV	5.6 (1.7–7.3)		−0.1 (−0.6–0.0)		2.4 (0.5–5.0)		0.6 (−0.6–2.9		2.3 (0.0–4.2)

PH: pulmonary hypertension, NYHA: New York Heart Association, ACS: Acute chest syndrome

Negative correlations were identified between the change in O₂ saturation during the 6MWT and measures of cardiac function such as TRV (r=−0.09), left atrial volume index (r=−0.18), left ventricular mass index (r=−0.14), left ventricular diastolic dimension (r=−0.13) and left lateral E/è ratio (r=−0.11) by bivariate analysis (Supplementary table 3, Figure 1a). Certain hematologic characteristics of complete blood counts and hemolysis also correlated negatively with O₂ desaturation such as the white blood cell count (r=−0.10), LDH (r=−0.11), and total bilirubin (r=−0.09) (Figure 1a). O₂ desaturation correlated positively with ejection fraction (r=0.01), hemoglobin levels (r=0.15), RBC count (r=0.15) and ejection fraction (r=0.01) (Supplementary table 3, Figure 1a).

Bivariate analysis identified positive correlations between the change in systolic blood pressure and TRV (r=0.13) and BUN (r=0.08) (Supplementary table 3, Figure 1a). Diastolic blood pressure was found to correlate negatively with left atrial volume index (r=−0.12) There were positive correlations between the change in pulse pressure and TRV (r=0.14), left ventricular mass index (r=0.01), MCV (r=0.09) and BUN (r=0.10). The change in pulse pressure correlated negatively with levels of hemoglobin (r=−0.09) and red blood cells (r=−0.11) (Supplementary table 3, Figure 1a). The mean of change of pulse pressure was higher in patients with self-reported previously diagnosed pulmonary hypertension (p=0.039) (Table 1).

Multivariable Predictors of the Ratio of Vital Sign Change to Distance

In most parsimonious model, absolute reticulocyte count (β=−0.18) and ejection fraction (β=0.10) were associated with the change in oxygen saturation. The change in systolic blood pressure was associated with LDH (β=−0.10), TRV (β=0.14) and ejection fraction index (β=−0.09). Diastolic blood pressure change was associated with age (β=−0.12), and SS genotype (β=−0.09). The change in pulse pressure was observed to be associated with RBC count (β=−0.13), lactate dehydrogenase (β=−0.10) and blood urea nitrogen levels (β=0.13) (Table 2).

Table 2:

Multivariable Association of Clinical Variables and Change of Vital Signs (Adjusted for Walk Distance) During 6MW Test^*.

	Change in O2 saturation	Change in Diastolic blood pressure	Change in Systolic blood pressure	Change in Pulse pressure
Age	NA	−0.12 (−0.19– −0.03)	NA	NA
SS genotype	NA	−0.09 (−0.18– −0.001)	NA	NA
LDH	NA	NA	−0.10 (−0.21– −0.04)	−0.10 (−0.20 – −0.002)
RBC	NA	NA	NA	−0.13 (−0.24– −0.02)
TRV	NA	NA	0.14 (0.04– 0.24)	NA
Ejection fraction	0.10 (0.01–0.19)	NA	−0.09 (−0.20 – −0.001)	NA
Absolute reticulocyte count	−0.18 (−0.29– −0.08)	NA	NA	NA
BUN	NA	NA	NA	0.13 (0.03–0.24)
Adjusted R2	0.04	0.02	0.02	0.03

^*Cell represent standardized beta (mean and 95%CI) for change of vital sign per 100 m walk. A negative beta indicates that variable predicts more decrease in vital sign from before to after 6MWT.

For each outcome, predictors that were selected from backward linear regression with a P<0.05.

NA: Not in the model.

Alternative modelling strategy indicated that higher TRV and lower RBC predicted more oxygen saturation decline. LA volume index predicted less increase in systolic blood pressure (Supplementary table 4).

Change of heart rate was positively associated with blood urea nitrogen (β=0.13), BMI (β=0.12), ALT (β=0.11), history of acute chest syndrome (β=0.14) and current hydroxyurea use (β=0.15). Change of heart rate was negatively associated with RBC count (β=−0.22) and MCV (β=−0.17), left atrial volume as a marker for diastolic heart dysfunction (β=−0.14), and pain crises in the last 12 months (β=−0.10) (Figure 1b).

Figure 1b:

Beta (P Values) of Factors Impacting Change of Heart Rate, Adjusted R-squared = 0.06

Survival Analysis

During a median time of follow up of 29 (IQR: 25–33) months, 22 (3.7% of 592 with follow-up data) participants died. Survival analysis revealed that any increase in systolic blood pressure was associated with reduced risk of mortality when adjusted for TRV and NT-pro BNP (HR=0.3; p=0.019) (Table 3 and Figure 1c).

Table 3:

Vital Sign Change During 6MW (Adjusted for Walk Distance) and Survival in Adult Patients with SCD.

	n (%)	Crude Hazard ratio (95%CI, P value)	Hazard ratio adjusted for TRV and NT-pro BNP (95%CI, P value)
SBP increase*	480 (77)	0.68 (0.28–1.65, 0.39)	0.28 (0.10–0.81, 0.019)
DBP increase*	364 (58)	0.85 (0.37–1.96, 0.70)	0.72 (0.26–2.02, 0.54)
Pulse BP increase*	428 (68)	1.28 (0.50–3.26, 0.61)	0.91 (0.33–2.49, 0.85)
Oxygen % decrease**	309 (50)	1.42 (0.60–3.35, 0.43)	1.06 (0.40–2.82, 0.92)
Heart rate increase*	565 (90)	2.40 (0.33–17.61, 0.39)	1.15 (0.16–8.54, 0.89)

^*Was defined as any increase in vital sign from before to after 6MWT

^**Was defined as any decrease in O2 saturation from before to after 6MWT

Figure 1c:

Cox Regression Analysis of Change in Systolic Blood Pressure

Sensitivity Analysis

One hundred thirty-eight (22.0%) of patients had ≥3% decrease in oxygen saturation. This frequency was 25.5% in HbSS vs. 12.5% in other genotypes (P <0.001). TRV (OR = 1.82, P = 0.020), absolute reticulocyte count (OR = 1.03, P < 0.001) and hemoglobin (OR = 0.99, P = 0.035) were significant predictors of oxygen saturation ≥ 3% during 6MW test. In a sensitivity analysis including 467 patients with HbSS, the frequency of systolic blood pressure rate increase was 76.8%. The predictors of systolic blood pressure change were ejection fraction, TRV and LDH and any increase in systolic blood pressure was associated with lower mortality (HR = 0.28, 95%CI = 0.09–0.83, P = 0.022).

Discussion

The results of this study identify novel associations between the patterns of vital sign change of adult SCD patients completing the 6MWT and clinical characteristics, morbidity and mortality. Increase in heart rate was the most frequently observed VSC during 6MWT, followed by systolic and diastolic blood pressure. Indicators of anemia and hemolysis were associated with greater oxygen desaturation and increase in heart rate. Elevated TRV correlated with more exercise induced blood pressure increase and markers associated with left ventricular diastolic dysfunction were associated with less increase of blood pressure. Patients with clinical characteristics associated with severe SCD, such as a history of acute chest syndrome, were observed to have higher increase in their heart rate while completing the 6MWT.

For all of our analysis we used the ratio of VSC divided by the 6MWD. The distance walked during a 6MWT is known to be influenced by many factors³¹. Dividing VSC by the completed 6MWD in a well characterized patient cohort removed confounders that are associated with the completed distance and strengthened the association between 6MW VSC and SCD pathology. This method of analysis should better account for variance of patient age, height, lung size and sex. Similar analysis has been shown to be effective in other conditions, but our study is the first to provide evidence supporting its use in the SCD population³². This method of analysis is also better suited for longitudinal clinical monitoring as it should facilitate better analysis of lifetime 6MW VSC progression and change.

We found that any increase in systolic blood pressure during the 6MWT was associated with better survival. This finding’s independence from the effect of TRV and NT-pro BNP is important as both are independently associated with mortality in the adult SCD population⁷. The protective effect associated with increased 6MWT systolic blood pressure could be indicative of a patient’s ability to increase stroke volume and therefore maintain cardiac output during exercise. This concept was supported by our finding that the other component of cardiac output, heart rate, was not a predictor of mortality. It is possible that inability to increase systolic blood pressure during exercise that we observed is related to the development of heart failure with preserved ejection fraction (HFpEF), or right heart dysfunction in the setting of high pulmonary pressures, or a combination of both pathologies^33,34. Decreased risk of mortality associated with exercise induced increases of systolic blood pressure has also been identified in other conditions. Increased peak exercise systolic blood pressure in primary pulmonary hypertension patients, and increased 6MWT pulmonary systolic blood pressure in a combined cohort of patients with pulmonary arterial or thromboembolic hypertension was associated with increased survival rates^35,36. Additionally, greater post 6MWT systolic pressure is associated with increased functional capacity in adolescents with chronic kidney disease³⁷. Clinical characteristics of SCD severity and increased functional ability have also been shown to correlate with hemodynamic response to exercise in the setting of cardiopulmonary function. Previous work showed that adult SCD patients with normal pulmonary spirometry exhibited a higher average change of systolic blood pressure, significantly different post 6MWT systolic blood pressure levels, decreased history of acute chest syndrome and increased functional capacity when compared to patients with abnormal spirometry³⁸.

Our analysis identified associations between the 6MW VSC and markers of diastolic dysfunction that may have contributed to risk of premature death. Diastolic heart dysfunction with HFpEF is common in the adult sickle cell disease population and associated with increased risk of mortality^9,12,13. The myocardial fibrosis that often coincides with HFpEF is known to decrease cardiac compliance and contractility, which impairs a patient’s ability to increase stroke volume^9,13. Previous investigation has shown that patients with HFpEF exhibit smaller increases in stroke volume during exercise³⁹. Increased left atrial filling pressures are also associated with the pathophysiology of the HFpEF with diastolic dysfunction that commonly affects the SCD community^9,34. In our study, patients with higher left atrial volume experienced lower increases in systolic blood pressure during the 6MWT. Patients in our study with increased left atrial volume were also observed to experience smaller changes of heart rate. Above average ejection fraction is another characteristic of diastolic heart dysfunction. Interestingly, increased ejection fraction, along with elevated LDH levels, also coincided with smaller increases of 6MWT systolic blood pressure. These associations support the possibility that the progression of diastolic dysfunction may have contributed to the increased risk of death observed in patients who could not increase their systolic blood pressure during the 6MWT. Previous work has already established that exercise related hemodynamic analysis is effective in detecting HFpEF that may not be apparent in a resting patient¹⁵. It may also be possible that a pattern of deviation of 6MWT VSC from baseline values exists that is associated with adult SCD patients at increased risk for the development of HFpEF with diastolic dysfunction.

Oxygen desaturation was the least frequent vital sign change that we observed, however, it was still present in half of our study participants. A previous study found exercise induced oxygen desaturation (as measured by at least 3% change) to be present in eight percent of pediatric SCD patients in a cohort of mixed genotypes²⁵. Another study in a cohort of pediatric patients observed the prevalence to be more frequent in HbSS participants in⁴⁰. Our data suggests that the prevalence of exercise induced oxygen desaturation may be higher in adults with SCD. Participants with sickle cell anemia in our cohort exhibited decreased instance of 3% oxygen desaturation; however, the HbSS genotype was not found to be a predictor of 3% oxygen desaturation by multivariable analysis. Of note, predictors of 6MWT oxygen desaturation of at least 3% in our study were found to be TRV, and markers of both hemolysis and anemia as has been found in previous investigations^25,41. It is common for adult SCD patients to be afflicted by pulmonary dysfunction with a restrictive pattern that worsens with aging⁴². The increased prevalence of exercise associated oxygen desaturation that we observed may be associated with the increased incidence of restrictive lung disease among adult SCD patients or compounded cardiovascular effects of the disease that impair cardiac output.

This study also identified associations between 6MW VSC and characteristics of severe sickle cell disease. Participants with increased frequency of acute chest syndrome/pneumonia or those currently using hydroxyurea, were associated with increased heart rate change. Both the mean change of heart rate and pulse pressure were found to be different in groups with and without pulmonary hypertension by bivariate analysis. Our investigation did not however identify any associations between aspects of clinical history that directly impact 6MWT test performance such as history of hip replacement, avascular necrosis of hip, leg ulcers, or stroke by multivariate analysis. These findings suggest that 6MWT VSC in adults with SCD is significantly influenced by SCD severity.

Longitudinal studies of 6MWT VSC analysis may be able to identify patterns of VSC derangement that precede the development of cardiovascular and cardiopulmonary dysfunction in the SCD population. Such data could become significant in the implementation of preventative clinical measures. The prognostic significance of vital sign recovery rates and mortality and morbidity in the sickle cell disease population remain unknown. Decreased post 6MWT heart rate recovery rate is a predictor of mortality in both heart disease and pulmonary fibrosis patients^28,29. Investigation of 6MWT vital sign recovery rates may also prove to have prognostic significance for patients with sickle cell disease.

Several limitations exist in the interpretation of our study data. Right heart catheterization to confirm pulmonary hypertension diagnosis was not performed consistently, so it was only possible to make correlations between TRV values and self-reported data. Blood oxygen saturation was not measured by arterial blood gas with CO-oximetry, so dyshemoglobinemia was not accounted for. Pulse oximetry as a measure of oxygen saturation can be questionable for SCD patients, most notably during periods of severe anemia and or vaso-occlusion which could have affected our results^43–45.

In conclusion, we found that increase in heart rate was the most frequently observed VSC, followed by systolic and diastolic blood pressure. Oxygen desaturation was the least frequent VSC observed. 6MWT VSC in the adults with SCD is associated with anemia and pulmonary hypertension, as assessed by TRV values and can be used for risk stratification. Any increase in systolic blood pressure during exercise testing was associated with a reduced risk of mortality after adjustment for TRV and NT-proBNP, suggesting systolic blood pressure increase during the 6MWT may correlate with retention of the ability to increase stroke volume and cardiac output. Our work indicates that exercise induced VSC analysis can be used to identify adult SCD patients at risk for worse clinical outcomes.