Daily Physical Activity in Pulmonary Arterial Hypertension: Insights from a Multicenter Longitudinal Trial Using Accelerometry

Abstract

Background:

Pulmonary arterial hypertension (PAH) is a progressive disease associated with high morbidity and mortality. Traditional assessments such as the six-minute walk distance (6MWD) may not adequately capture daily physical activity or patient experience in real-world settings.

Research Question:

Among patients with PAH, do accelerometer-derived measures of daily physical activity reflect disease severity, functional status, and outcomes over time?

Study Design and Methods:

PHANTOM, a multicenter randomized controlled trial of anastrozole in PAH enrolled males and post-menopausal females with PAH at seven centers in the US. Patients completed 4 in-person study visits over 12 months. Participants wore hip-based accelerometers for 7 days at each visit. Mixed-effects models assessed associations between daily physical activity and demographic, clinical, echocardiographic, and environmental variables. Clinical outcomes included 6MWD, health-related quality of life (HRQoL; emPHasis-10 and SF-36), bone mineral density (BMD), and time to clinical worsening (hospitalization, treatment escalation, or death). Cox models examined associations between activity and time to clinical worsening.

Results:

Eighty-four participants were enrolled (mean age 61 years; 67% female). Device adherence was high across visits. Activity declined over time and was lower in patients with WHO functional class III, worse RV function, and connective tissue disease–associated PAH. Lower daily activity was associated with worse HRQoL, worse baseline 6MWD, and worse baseline femoral/hip BMD. Lower baseline activity was associated with increased risk of clinical worsening (HR 1.06 per hour/week less active; p<0.01).

Interpretation:

Accelerometer-derived physical activity reflects disease severity and is associated with HRQoL, BMD, and clinical worsening in PAH. These metrics may offer a dynamic, patient-centered complement to traditional assessments.

Pulmonary arterial hypertension (PAH) is a rare and life-threatening disease characterized by elevated pulmonary vascular resistance and pulmonary arterial pressure culminating in right ventricular (RV) failure. Despite therapeutic advancements, outcomes remain poor, with an age-adjusted mortality rate at 4.5–12.3 per 100,000 adults.^1,2 Clinical trials and treatment decisions in PAH are largely guided by changes in exercise capacity, often assessed using the six-minute walk distance (6MWD), an established regulatory endpoint.^1,3,4 While 6MWD reflects sub-maximal exercise capacity, it is a cross-sectional clinic-based test that may not fully capture patients experiences.⁵ Many patients cite day-to-day activity limitations as a primary concern, yet standard tools often fail to capture these.⁶ Although efforts towards remote adaptations of 6MWD exist,^7,8 most assessments remain in-clinic and overlook “real-world” fluctuations, underscoring the need for patient-centered metrics.

Physical activity encompasses all energy-expending movement, extending beyond structured exercise.^9,10 In PAH, activity limitation occurs early and contributes to reduced health-related quality of life (HRQoL). While disease severity contributes, environmental, demographic and behavioral factors also influence activity.¹¹ Some argue improved cardiopulmonary hemodynamics may not translate to greater activity due to sedentary habits. However, meaningful physiologic improvement should promote increased activity and better HRQoL. Wearable devices enable continuous, real-world monitoring of physical activity, offering a patient-centered assessment of functional status beyond clinic-based measures.

This study leveraged longitudinal data from PHANTOM, a multicenter randomized clinical trial (RCT) of anastrozole in PAH,^12,13 to evaluate associations between physical activity and disease burden. Using accelerometer data collected at baseline, 3, 6, and 12-months, we assessed predictors of physical activity and examined associations of daily physical activity with bone mineral density (BMD), HRQoL, 6MWD, and clinical worsening over time. This analysis builds on our prior baseline-work¹³ by including longitudinal data from the full cohort and evaluating additional clinical and behavioral domains.

Methods

Study population and Study Sample

PHANTOM¹² was a multicenter RCT of anastrozole (1 mg daily) conducted at the University of Pennsylvania, Stanford University, University of Colorado, Johns Hopkins University, Washington University, Rhode Island Hospital and Vanderbilt University. The study was approved by the Penn IRB. Eligible participants included adult males and postmenopausal females on stable PAH therapy ≥3-months. Key exclusions included hospitalization or illness, World Health Organization (WHO) functional class IV symptoms, significant musculoskeletal or coordination impairments, and contraindications to anastrozole (osteoporosis, breast cancer history, hormone replacement therapy, and pre-menopausal status as anastrozole by itself does not suppress estrogen in this group).

Between 2018 and 2022, 84 participants were enrolled, including during the COVID-19 pandemic. Study visits occurred at baseline, 3, 6 and 12-months, during which demographic and clinical data were collected. Participants underwent 6MWT per guidelines.⁵ Transthoracic echocardiograms (TTE) were performed at baseline, 6 and 12-months following a research-grade protocol and analyzed centrally using American Society of Echocardiography guidelines.¹⁴ Blinded reviewers assessed RV function, including tricuspid annular plane excursion (TAPSE) and qualitative RV metrics.

Participants completed the Short Form-36 (SF-36) and the emPHasis-10 (E-10) at each visit.^15–17 The SF-36 evaluates general HRQoL across mental and physical domains (on a scale of 0 – 100, lower scores worse),¹⁷ while the E-10 is PH-specific (0 and 50, higher scores worse).^15,16 Both instruments are extensively used in PAH.^15,16,18,19

Accelerometry

Participants wore the ActiGraph GT9X hip-based accelerometer during waking hours for 7-days at each visit. Nonwear time was defined using standardized algorithms.²⁰ Sixteen hours/day was considered 100% adherence, allowing 8-hours of sleep. A valid wear day was defined as ≥3 hours of device wear (~20% adherence), and median values were calculated across valid days within each 7-day period to provide robust daily activity estimates. Devices recorded triaxial movement at 30Hz. Data were aggregated into 1-minute epochs and summarized as vector magnitude (VM) counts, the square root of the sum of squares of movement along the three axes [(x²+y²+z²)^1/2].²¹ Epochs with counts ≥1853 were classified as active minutes based on standard thresholds. Total active minutes/day were calculated by summing all qualifying epochs.

The primary activity metric was total time spent active/day. We also evaluated VM counts and activity intensities categories: sedentary (1.0–1.5 metabolic equivalent of tasks [METs]), light (1.6–3.0 METs), moderate (3.1–6.0 METs), and vigorous activity (>6.1 METs), defined using standard cutoffs.²² Median values from valid wear days were used for all final analyses.¹³

Statistical Analysis

Descriptive statistics were presented as means with standard deviations (SDs) or medians with interquartile ranges (IQRs). To examine factors associated with device wear time, we used linear mixed-effects models adjusting for visit number, age, sex, body mass index (BMI), WHO functional class, and PAH etiology, with random intercepts for subject-ID. Covariates were selected a priori based on clinical relevance and prior literature.

Two sets of mixed-effects models were constructed. The first evaluated longitudinal associations between patient characteristics, disease severity, and time spent active (dependent variable), adjusting for the same covariates and device weartime, with random intercepts for subject-ID. All models were adjusted for wear time to avoid confounding by differences in adherence. Two-way interactions (e.g., age × sex, age × BMI, age × PAH etiology, age × wear-time etc.) were tested in separate models. Similar models assessed echocardiographic variables (e.g., TAPSE, qualitative RV size) and WHO functional class as predictors of daily activity.

Centers were grouped into four geographic regions - Northeast, Midwest/Central, Mountain West, and West Coast - based on proximity. Study visits were categorized as spring/summer or fall/winter based on calendar month. Separate mixed-effects models evaluated associations between region, season and time spent active, including region-season interactions. All models used the same covariates and random intercepts as above.

For clinical outcomes, we used mixed-effects models to examine the associations between time spent active (independent variable) and 6MWD, E-10 and SF-36 scores (mental and physical components). Models included same covariates and random intercepts as before.

Cox proportional hazards models assessed the association between activity and time to clinical worsening (composite of death, PAH-related hospitalization, or treatment escalation). Physical activity was modeled both as a baseline (fixed) variable and as a time-dependent covariate in separate models, with the latter updated at each follow-up visit (3, 6, and 12 months) using the median value from the associated 7-day wear period. Censoring occurred at the end of follow-up (12 months after enrollment) or at the time of clinical worsening, whichever occurred first. Events were prospectively captured at each study visit and verified by independent adjudication in the parent trial. Results were reported as hazard ratios (HR) with 95% CIs.

Linear regression models assessed cross-sectional associations between baseline physical activity and BMD, measured by dual-energy X-ray absorptiometry (DEXA). These models included the same covariates. Due to potential anastrozole effects on BMD, longitudinal analyses were limited to the placebo group, but only 33 participants had both DEXA and accelerometry at 12 months, precluding analysis.

To evaluate pandemic-related effects, visits were categorized as pre-COVID, COVID-pre-vaccine, and COVID-post-vaccine. Mixed-effects models assessed the association between pandemic period and device wear time or time spent active, adjusting for standard covariates and random intercepts.

The PHANTOM trial found no differences in physical activity by treatment group (anastrozole vs. placebo).¹² Including treatment group assignment as a covariate did not alter model estimates and was omitted from final analyses.

Results

Eight-four participants enrolled in PHANTOM (Table 1). The mean age was 61-years, and 67% were female. Idiopathic PAH was the most common etiology, 22% received parenteral prostacyclin analogues at baseline, and most were WHO functional class II. Accelerometer data were available for 83 participants at baseline, with excellent retention: 81 participants completed assessments at both 3 and 6-months, and 78 at 12-months. Reasons for missing assessments are in Supplementary Table 1.

Table 1:

Patient characteristics at baseline

	N = 84
Age, years	59.7 ± 10.7
Female sex	51 (61%)
BMI, kg/m2	29 [25, 34]
Race
White	64 (76%)
Black	8 (10%)
Asian	5 (6%)
Other/Unknown	7 (8%)
Ethnicity
Hispanic/Latino	7 (8%)
Non-Hispanic	77 (92%)
PAH etiology
Idiopathic/heritable	40 (47%)
Connective tissue disease	19 (23%)
Drug or toxin induced	10 (12%)
Congenital heart disease	7 (8%)
HIV	4 (5%)
Portopulmonary Hypertension	4 (5%)
Treatment
Oral only	52 (62%)
Inhaled ± oral	7 (8%)
Subcutaneous/IV ± oral	25 (30%)
Echo parameters
RV enlargement
None	10 (12%)
Mild	33 (39%)
Moderate	35 (42%)
Severe	6 (7%)
TAPSE, cm	2.1 [1.7, 2.4]
NT-proBNP	204 [92, 505]
WHO functional class
I	9 (11%)
II	54 (64%)
III	21 (25%)
6MWD, meters	426 [355, 517]
HRQOL scores
SF-36 PCS	41 [35, 47]
SF-36 MCS	55 [49, 59]
E-10	17 [10, 25]
ActiGraph parameters at baseline
Active time, hours	4.8 [3.6, 6.0]
Time spent in light activity, minutes	196 [141, 251]
Time spent in moderate - vigorous activity, minutes	2 [1, 8]
Device wear time, hours	14 [12.4, 15.3]

Values are presented as mean ± SD, median [IQR]. IV = intravenous. WHO = World Health Organization. TAPSE: Tricuspid annular plane systolic excursion. 6MWD = six-minute walk distance. SF-36 = short form 36. PCS = physical component score. MCS = mental component score.

Wear time behaviors

Participants consistently engaged with accelerometer use, wearing the device for a median of 5 days per visit, with a median of 13 hours and 36 minutes daily (Supplementary Figure 1). Across all visits, 2,182 patient-days of accelerometer data were collected (Supplementary Table 2).

Wear time did not differ by age, sex, or BMI, but varied by disease severity and etiology. Participants with WHO functional class III wore the device for 0.8 hours less per day compared to those with class I/II (95%CI:0.33–1.30 hours/day, p=0.002). Those with connective tissue disease-associated PAH wore it 1.5 hours less per day than those with idiopathic or heritable PAH (95%CI:0.3–2.7 hours/day, p=0.023). Wear time declined by 1 hour/day on follow-up compared to baseline (95% CI:1.16–0.42 hours/day, p<0.001), after adjusting for covariates (Figure 1). Wear time was not significantly associated with any clinical outcomes.

Figure 1:

Linear mixed effects models examining the longitudinal association of (A) time since enrollment, (B) WHO functional class and (C) PAH etiology with device wear time after adjusting for age, sex, BMI with random effects for subject ID.

Time spent active

Participants spent most of their time in sedentary behaviors, with limited light and minimal moderate-to-vigorous activity. This pattern did not change over study visits (Figure 2A, B). Compared to baseline, activity declined by 15 minutes/day (1.75 hours/week) at follow-up visits (95%CI:4–15 minutes/day, p=0.009) (Figure 3).

Figure 2A:

Time spent in different activity intensities across all visits

Figure 2B:

Time spent (hours/day) in different activity intensities – sedentary, light activity and moderate-vigorous activity. Individual participant trajectories (thin gray lines) and the overall smoothed trend (thick blue line) are displayed. The overall smoothed trend was estimated using restricted cubic splines with three knots, adjusted for wear time.

Figure 3:

Linear mixed effects models examining the longitudinal association of (A) age, (B)time since enrollment, and (C) interaction between age and BMI with time spent active per day after adjusting for sex, PAH etiology and device wear time, with random effects for subject ID.

Daily time active decreased by 24 minutes/day for every 10-year increase in age (95%CI:10–38 minutes/day, p=0.038). Overweight patients were 20 minutes/day more active (1.25 hours/week) than those with normal weight or obesity (95%CI:3–37 minutes/day, p=0.022; Figure 3A, B). An age-BMI interaction showed that younger overweight individuals had substantially more active time than older ones, a pattern not observed in other BMI categories (p for interaction=0.01; Figure 4). For example, a 40-year-old overweight patient averaged 75 minutes/day of activity, versus 15 minutes/day in a 60-year-old overweight patient. No significant differences in activity were found by sex or PAH etiology, and no other interactions were detected.

Figure 4:

Linear mixed effects models examining the longitudinal interaction between age and BMI with time spent active per day after adjusting for sex, PAH etiology and device wear time, with random effects for subject ID.

RV function and WHO functional class

No significant longitudinal associations were identified between echocardiographic measures of disease severity and time spent active (Supplement Table 3), including TAPSE and RV size. However, reduced RV function was associated with lower vector magnitude (VM) counts. For every 0.4 cm (1SD) decrease in TAPSE, VM counts declined by 19000 (95% CI:6140–32000, p=0.004). Similarly, greater RV dilation was associated with lower VM counts (p=0.01; Figure 5). Patients with WHO functional class III were 50 minutes/day (5.8 hours/week) less active than those in class I or II (95%CI:30–58 minutes/day, p<0.001; Supplement figure 2).

Figure 5:

Linear mixed effects models examining the longitudinal association of (A) tricuspid annular plane systolic excursion (TAPSE) and (B) right ventricular dilation with vector magnitude counts after adjusting for age, sex, BMI, PAH etiology and device wear time, with random effects for subject ID. (ß coefficients reported per 0.4cm increase in TAPSE)

Geographic region and seasonal variation

Across 307 study visits, 163 occurred in spring/summer months and 144 visits in fall/winter. Patients were 15 minutes/day (1.75 hours/week) less active during the fall/winter compared to spring/summer (95%CI:10–27 minutes/day, p=0.02). These seasonal effects were not modified by age, sex, BMI, PAH etiology, or visit number. No differences in activity were observed between geographic regions overall. However, a significant region-season interaction was identified: West Coast patients were 32 minutes/ day (3.5 hours/week) more active during the fall/winter compared to other regions (95%CI:14–53 minutes/day, p for interaction=0.001).

Time spent active and clinical outcomes

Less time spent active was significantly associated with worse HRQOL. At baseline, each additional hour of daily activity corresponded to a 1.5 lower E-10 score (95%CI:−2.9 to −0.50, p=0.009). Across all visits, a one hour higher daily active time was associated with 3.5 point lower E-10 score (95%CI:−7.91 to −1.1, p=0.04) for the 7-day wear period. One-hour higher daily active time was also associated with a higher SF-36 physical (ß:0.98, 95%CI:0.14–1.82, p=0.02) and mental component scores (ß:2.94, 95%CI:−0.56 to 6.44, p=0.09) although the latter was not statistically significant (Figure 6). These increments in HRQOL scores did not meet the threshold for minimal clinically important differences.^15,17

Figure 6:

Linear mixed effects models examining the longitudinal association of time spent active with (A) emPHasis-10 and (B) SF-36 physical component scores, after adjusting for age, sex, BMI, PAH etiology, device wear time and visit number, with random effects for subject ID.

At baseline, higher physical activity was associated with a 22-meter greater 6MWD per hour of activity (95%CI:10 meters–35 meters, p=0.002). However, no significant longitudinal association was observed (ß:3 meters/additional hour of activity, p=0.19).

During 364 median follow-up days (IQR:361, 377), 20 clinical worsening events (3 deaths, 10 hospitalizations, and 7 treatment escalations) occurred. Five of these were recurrent events. The event rate for clinical worsening was 17.9 per 100 person-years. Thirteen of the 20 events occurred in patients below the cohort median for daily activity.

Lower baseline activity was associated with higher risk of clinical worsening (HR:1.06 per hour/week, 95% CI:1.03–1.10, p<0.01) after adjusting for adjusting for age, sex, BMI, PAH etiology and device wear time. In time-dependent models, each one-hour reduction in daily activity was linked to a shorter time to clinical worsening (HR:1.02, 95%CI:1.01–1.05, p=0.03), adjusted for adjusting for age, sex, BMI, PAH etiology and device wear time (Figure 7).

Figure 7:

Adjusted Time to Clinical Worsening by Baseline Physical Activity: Survival estimates from a Cox proportional hazards model adjusted for age, sex, BMI, PAH etiology, and device wear time, comparing participants below the 25th (2 hours/day) and above the 75th (7 hours/day) percentiles of baseline daily physical activity (p < 0.01).

Bone mineral density

At baseline, more physical activity was associated with a higher BMD at the hip and femoral neck. For every 1-hour-per-day increase in physical activity sustained across the week, BMD at the hip was higher by 0.14 g/cm² (95%CI:0.03–0.25 g/cm², p=0.01), while femoral BMD was higher by 0.13 g/cm² (95%CI:0.03–0.22 g/cm², p=0.015) after adjusting for age, sex, BMI, PAH etiology and device wear time (Figure 8). No significant associations were observed for lumbar spine BMD.

Figure 8:

Linear regression of association of active hours per day with bone mineral density at (A) Hip and (B) Femur at baseline adjusted for age, sex, BMI, etiology of PAH and device wear time.

Impact of the COVID-19 pandemic

Of 323 total study visits, 183 occurred pre-COVID, 78 during the pre-vaccine phase, and 62 post-vaccine. The pandemic did not significantly impact study enrollment, visit completion, device wear time or physical activity. No significant interactions were found with age, sex, BMI, PAH etiology, or visit number.

Discussion

This study leveraged longitudinal data from the multicenter PHANTOM RCT to examine the relationship between daily physical activity and clinical characteristics, functional status, and outcomes in PAH. Younger age and overweight status were associated with higher activity, while worse RV function and advanced WHO functional class were associated with lower activity. Higher activity was associated with better HRQoL, greater baseline hip and femoral BMD, and lower risk of clinical worsening. Differences in device wear time by PAH etiology and functional class highlighted behavioral and disease-related factors influencing engagement.

Our findings reinforce prior reports of profound sedentary behavior in PAH.^13,23,24 PHANTOM participants averaged 11.9 sedentary hours/day, a pattern stable across 12-months. In contrast, U.S. adults aged ≥ 60-years average 8.5 sedentary hours/day, despite being older (median age:71.6 vs 59.7 years in PHANTOM).²⁵ This underscores disproportionate inactivity in PAH – even among highly motivated trial participants. Reduced daily activity was associated with worse RV function and WHO functional class, reinforcing the functional impact of disease.

Physical activity also varied by age, BMI and environmental factors (geography and season), suggesting influences beyond cardiopulmonary impairment. A significant age-BMI interaction emerged: younger overweight individuals had higher daily activity than older counterparts - a pattern not seen in normal weight or obese groups. This may reflect body composition differences, as BMI does not distinguish lean from fat mass. In younger adults, higher BMI may indicate greater muscle mass, while in older adults, it may reflect adiposity and sarcopenia.^26–31 In PAH, normal BMI may not imply preserved muscle mass, and overweight status may sometimes indicate better musculoskeletal reserve. In contrast, the lack activity in young obese individuals may reflect the additive burden of adiposity, including impaired ventilatory efficiency, increased RV afterload, and reduced exercise capacity.^32,33 Obesity is also linked to higher rates of sarcopenia and frailty,^34–37 limiting mobility.^34–37 These findings highlight BMI’s limitations as a proxy for body composition and support future studies using more granular phenotyping of lean and fat mass. The observed decline in activity over time may reflect behavioral adaptation or waning novelty of monitoring - a reverse Hawthorn effect. Identifying demographic, clinical, and environmental influences was a key objective, acknowledging that such factors can complicate interpretation in broader populations. These findings inform the ongoing ACTiPH study (R01-HL159997), a large real-world prospective cohort designed to validate wearable metrics and develop interpretable models for heterogeneous settings.

Daily physical activity was associated with HRQoL, with stronger associations for the PAH specific E-10 than the general SF-36.^16,38,39 While statistically significant, the effect sizes did not exceed minimal clinically important differences.^15,17 As this was an observational analysis within an RCT, whether activity changes yield meaningful HRQoL improvements or define thresholds for clinical benefit. Future RCTs should test whether increasing daily activity improves patient-reported outcomes.

Despite a strong cross-sectional association between activity and 6MWD at enrollment, no longitudinal association was observed, likely reflecting differences in what these measures capture:6MWD is a standardized submaximal exercise test⁵, whereas daily activity reflects real-world behavior shaped by disease, environment and behavior. In our clinically stable RCT cohort, 6MWD changed little over time (mean baseline: 428 meters, mean 12-months: 429 meters).¹² Activity may change more gradually, or in a different time scale than what the 6MWD can detect, supporting its complementary value to clinic-based assessments.

Higher baseline activity was associated with greater femoral neck and total hip BMD, the first such finding in PAH, where osteoporosis risk is elevated.^40–42 This mirrors general population studies linking physical activity to improved skeletal health.^43,44 A recent longitudinal study in older adults participating in a yearlong multicomponent exercise intervention demonstrated that increased moderate-vigorous physical activity, measured by accelerometry, attenuated age-related BMD decline.⁴⁵ While our BMD analyses were limited by cross-sectional design and small sample size, they raise the hypothesis that activity may help preserve skeletal integrity in PAH - warranting further investigation in larger, longitudinal cohorts.

Lower baseline daily physical activity was associated with an increased risk of clinical worsening over 12-months. Time-varying analyses showed that declining activity further increased this risk. While these associations may reflect disease severity, they provide confirmatory evidence in a rigorously followed PAH cohort. The 7-day wear periods offered snapshots rather than continuous monitoring, potentially missing early changes. Given the modest number of events, these findings are hypothesis-generating and support future studies with denser longitudinal data to assess predictive utility. A smaller single study using step counts similarly found that lower daily activity was associated with hospitalization risk, whereas 6MWD was not.⁴⁶ These results align with European Respiratory Society recommendations¹ supporting physical activity in PAH, largely based on rehabilitation studies. A recent pilot study demonstrated that even unsupervised, remote activity interventions increased step counts and improved HRQoL in PAH.⁴⁷ These preliminary data highlight that daily physical activity is modifiable and merits further interventional research.

This study offers several strengths. We collected high-quality accelerometer data over 12-months with excellent adherence, even during the COVID-19 pandemic, demonstrating the feasibility of remote, real-world monitoring in PAH. By integrating activity data with echocardiography, PROs, BMD, and outcomes, we provide a comprehensive assessment of disease burden. Granular data enabled analyses of both common (e.g., time spent active), and nuanced (e.g., vector magnitude) metrics. We also identified factors associated with device adherence offering insights into behavioral and disease-specific contributors to engagement to inform future studies.

Limitations include the use of hip-based accelerometers, which while widely used and validated in large population studies like NHANES⁴⁸, may underestimate total body activity by missing upper body or non-ambulatory movement. Wrist-based devices capture these but introduce noise and systematically higher activity counts, complicating data interpretation.⁴⁹ Most validated metrics, including activity intensity thresholds were developed using hip-based devices, supporting their continued use despite limitations.⁵⁰ The high adherence and engagement likely reflect trial-related support and motivation, limiting generalizability to broader populations; however, these data establish feasibility and inform future real-world studies. Small BMD sample size limited longitudinal analysis, and as an observational analysis, causality cannot be inferred. Finally, regional variation in pandemic-related activity may not be fully captured.

This study offers novel insights into the relationship between physical activity, disease severity, and outcomes in PAH. Accelerometer-derived daily activity emerged as a dynamic, patient-centered metric complementing traditional assessments. High adherence and retention, support the feasibility, and by capturing real-world behavior, activity monitoring may help align care with patients lived experiences. These findings support integrating physical activity metrics into PAH assessments and management, ultimately advancing our ability to improve outcomes.

Abbreviations:

6MWD

six-minute walk distance

BMD

bone mineral density

HRQoL

health-related quality of life

PAH

pulmonary arterial hypertension

RV

right ventricular

TTE

transthoracic echocardiogram