Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 11.
Published in final edited form as: Am J Respir Crit Care Med. 2026 Jan 1;212(1):117–128. doi: 10.1164/rccm.202412-2393OC

Inflammation and obesity correlate in pulmonary hypertension but associate with diverging outcomes

Eckart De Bie 1,2,3, Priscilla Correa-Jaque 4, Rowena Jones 1,2, Harm J Bogaard 5, Justine Chan 1, Colin Church 6, John G Coghlan 7, Akshay Gaur 8, Stefano Ghio 9, Hossein-Ardeschir Ghofrani 10, Ze Ming Goh 8, Luke S Howard 11, Marc Humbert 12, Gabor Kovacs 13, Allan Lawrie 11, James Lordan 14, Wei-Yu Lin 15, Dharshan Neelam-Naganathan 8, Joseph Newman 1,2, Christopher J Rhodes 11, Karen Sheares 2, Olivier Sitbon 12, Thomas W Willis 3,15, Stephen J Wort 11; ASPIRE Registry, UK PAH COHORT CONSORTIUM, UNIPHY CLINICAL TRIALS NETWORK, Stefan Gräf 1, David G Kiely 8, Raymond L Benza 4, Alex Rothman 8, Chris Wallace 3,15,*, Mark Toshner 1,2,*
PMCID: PMC12888913  NIHMSID: NIHMS2137156  PMID: 40788756

Abstract

Rationale & Objectives

Inflammation is associated with all types of Pulmonary Hypertension (PH) both as a known cause and/or a putative confounder. The most common marker of inflammation, C-reactive protein (CRP), has not been widely studied in PH. This study set out to clarify if CRP informed clinical endotyping and outcomes.

Methods & Measurements

Time-series clustering of longitudinal CRP levels was employed. Clinical differences between clusters were validated in three independent UK/international cohorts using clinical cut off values (n=10,301; UK-cohort, ASPIRE and FDA cohort). Associations were analysed with functional and mortality outcomes by linear and Cox regression models including all-causes of PH (groups 1-5). To add mechanistic insight, multi-omics were interrogated from associated previously published arrays.

Main Results

Patients segregated into two stable CRP clusters (median CRP 2 versus 6.5 mg/l), with the high cluster exhibiting significantly higher BMI (difference between medians DBM=5.4 kg/m2), higher RAP (DBM=2mmHg) and reduced 6-minute walk distance (6MWD (DBM=55m)).

Inflammation was associated with worse survival, comorbidities, higher pulmonary vascular resistance (PVR), and smoking status. CRP and BMI were associated with differing inflammatory profiles in proteomic and transcriptomic analyses. Despite the relationship with CRP, higher BMI associated with improved survival, lower PVR and did not negatively affect 6MWD treatment-related functional responses.

Conclusions

We establish a relationship between CRP and BMI across all-cause PH, though CRP and BMI associate with diverging clinical outcomes. Inflammation and obesity are relevant phenotypes for consideration in clinical trial design. Understanding their impacts on outcomes is important for clinical practice.

MeSH terms: Hypertension, Pulmonary [C08.381.423], C-Reactive Protein [D12.776.034.145], Body Mass Index [E01.370.600.115.100.125]

Introduction

Pulmonary Hypertension (PH) is a group of severe diseases characterised by increased pulmonary arterial pressure associated with right ventricular dysfunction and significant mortality. PH is estimated to affect one percent of the world’s population (1). Whilst PH is common, the most studied subtype of PH, Group 1 Pulmonary Arterial Hypertension (PAH), is rare (2, 3). Like many other cardiovascular diseases, PAH has been linked with inflammation and aberrant immune regulation (3). (Auto-)inflammatory disease endotypes, defined by autoantibodies and raised immunoglobulin G transcription which associate with distinct phenotypes and survival, have also been described in PAH (4, 5). Similarly, inflammation is associated with all other groups of PH, either as measured by humoral immune markers or associated inflammatory conditions (e.g. sarcoidosis) (6-9). For example, IL-6 is raised in group 3 PH and associates with haemodynamics (7).

Whilst some reports suggest that immunomodulatory drugs may be beneficial in PH, clinical trials, such as one targeting IL-6, have thus far failed to show a clear beneficial effect (10, 11). However, these trials did not stratify patients by disease endotype or inflammatory status. Identification and phenotypic characterisation of patients with an inflammatory disease endotype and stratification of trials based on this may help in personalising treatment.

C-reactive protein (CRP), the clinically most used inflammatory marker, may help in identifying patients with inflammatory disease endotypes including autoimmunity. CRP has known associations with cardiovascular disease, autoimmune diseases, and obesity amongst others (12-14). This is of interest as ~30% of PAH patients are obese and this may associate with outcomes (15). Treatment of obese patients with heart failure with preserved ejection fraction (HFpEF) with GLP-1 agonists reduces CRP in addition to improving weight and functional outcomes, demonstrating the inter-related nature of inflammation/CRP and BMI and their relevance for outcomes in the context of cardiovascular diseases (16). The link between obesity and inflammation in PH, however, has not been clearly delineated.

A few small studies have identified associations between CRP levels and PH outcomes, restricted mostly to PAH, including survival and haemodynamics (17, 18). Moreover, the trajectory of CRP levels in PH over the course of disease has never been studied. We hypothesised that systemic inflammation, informed by raised CRP, would associate with previously published autoinflammatory endotypes (4).

Here, we investigate inflammatory phenotypes/endotypes in PH based on CRP level and validate findings using multicentre data from cohorts in Idiopathic/Heritable PAH (UK cohort), all-cause PH (ASPIRE registry) and the U.S. Food and Drug Administration (FDA) trials.

Methods

Included patients and ethical consent

Details of the methods used are available in the supplemental methods.

Patients were included from three different cohorts (Figure 1); A) 1308 patients from the National Cohort Study of Idiopathic and Heritable PAH (UK cohort; NCT01907295, Ethics: 13/EE/0203), which was used as ‘discovery’ cohort; B) 5036 patients from the Assessing the Spectrum of Pulmonary hypertension Identified at a REferral centre registry (ASPIRE; Ethics: 16/YH/0352 & 22/EE/0011), used as validation cohort in all-cause PH (19); and C) 3957 PAH patients recruited into randomised controlled trials from the FDA cohort. Patients participated in the AMBITION (20), ARIES-1 & ARIES-2 (21, 22), FREEDOM-EV (23), GRIPHON (24), PATENT (25), PHIRST (26), and SERAPHIN (27) Phase-3 trials. All patients included in the current study had a PH diagnosis and provided written consent if applicable. For the UK cohort and ASPIRE, patients were ≥18 years or older at diagnosis with a CRP level recorded at diagnosis, or the first study visit and/or a BMI. For the patients in the FDA registry, only a BMI and survival data needed to be recorded. Baseline demographics for each cohort can be found in Supplemental Table E1.

Figure 1 – Overview of the study design and employed methods.

Figure 1 –

Open in a new tab

UK PAH cohort patients (n=236) were clustered based on longitudinal CRP levels after which clinical differences between clusters were compared. Subsequently, clinical characteristics, omics and outcomes between patients with normal and mildly elevated CRP were compared in the UK cohort (n=968) and ASPIRE (n=5036). Finally, differences between WHO BMI groups were compared in the UK cohort (=1308), ASPIRE (n=4365) and the FDA cohort (n=3957). Cox-proportional hazard models and linear models were constructed adjusting for synergistic effects of CRP and BMI and other key covariates. Abbreviations: BMI: body mass index; CRP: C-reactive protein; PAH: pulmonary arterial hypertension; PH: pulmonary hypertension; WHO: World Health Organisation. Created in https://BioRender.com.

Statistical analyses

All statistical analyses were performed in R version 4.3.2. Non-parametric tests were used unless otherwise specified. P-values in the ‘discovery’ UK cohort were corrected with FDR using the Benjamini-Hochberg procedure (significance threshold: <0.05). Clinical differences reported are all at diagnosis unless otherwise specified.

To assess survival differences, right-censored Kaplan-Meier curves, for transplant free survival, were constructed using the survival (version 3.5-8) and Survminer (version 0.4.9) packages. Log-rank tests for survival times were performed. Similarly, the survival package was used to fit Cox-proportional Hazard models. For these models, numeric variables were Z-scaled and log-rank tests for survival differences were used. For the UK cohort, survival was assessed at 10 years post diagnosis, for ASPIRE at 5 years post diagnosis and for the FDA cohort at 2 years after study inclusion, due to the differing survival times and follow-up.

When comparing high with low CRP groups a threshold for raised CRP of >5mg/l was used, apart from the clustering analyses, based on previous literature and common cutoff in UK hospitals (17, 28).

Time series clustering

A subset of 236 patients from the discovery (UK) cohort with longitudinal CRP levels from diagnosis to the fourth study visit was selected for unsupervised clustering. Using the dtwclust package (version 5.5.12), Dynamic Time Warping distances were used to assess differences between individual time series. After assessment of stability indexes, Partitioning Around Medoids with two clusters was identified as optimal clustering algorithm/number of clusters (Supplemental Figure E1). Clinical differences between clusters were compared.

Linear models of the effect of CRP and BMI on outcome variables

To assess the effects of (diagnostic/first study visit) CRP and BMI on outcome variables, whilst correcting for covariates, Bayesian information criterion (BIC) scores were used to find the optimal model in the UK cohort (Supplemental Table E2).

Cox-proportional hazards model of BMI group and CRP level

To understand the association of BMI and CRP with mortality in all-cause PH, a Cox-proportional hazard model was used to estimate the effects of BMI and CRP (up to 50mg/l to prevent overfitting due to outliers) on 5-year mortality within the ASPIRE cohort, because of its larger sample size. This model adjusts for the associations of CRP and BMI so that their combined survival association can be estimated.

Multi-omics

Associations between BMI and CRP groups and previously published whole blood transcriptomics (n=310 for CRP groups and 374 for BMI groups) (29) and plasma proteomics (n=387 for CRP groups and 441 for BMI groups) of inflammatory genes/proteins were assessed using nonparametric tests in the UK cohort (30).

Code availability

Available via: https://github.com/EckartDeBie/Inflammation_and_obesity_in_PH

Results

Time series clustering of CRP levels in PAH identifies two phenotypically distinct clusters

Using unsupervised time series clustering of patients in the UK PAH cohort with consecutive longitudinal CRP over annual outpatient study visits (n=236, Figure 2), two distinct clusters were identified in the 'discovery' UK cohort, a ‘normal CRP’ (n=105) and a ‘high CRP’ (n=131) cluster with mildly raised CRP. Both clusters had stable CRP levels over five years post-diagnosis (median diagnostic CRP level for the high and normal clusters: 6.5 mg/l and 2 mg/l respectively, (Figure 2 & Table 1). This stability of CRP levels over time suggests diagnostic CRP levels can be used to stratify patients. The clusters differed in body mass index (BMI; p<0.0001) with higher BMI in the high CRP cluster (30.8 versus 25.4 kg/m2) and clinical markers strongly associated with or mediated by BMI, including right atrial pressure (RAP; p=0.00099), were all raised in the high CRP cluster (10 versus 8 mmHg Figure 2 & Table 1). Notably, the high CRP cluster had a significantly reduced 6-minute walk distance (6MWD; 300 versus 356 metres, p=0.00045) compared to the normal cluster.

Figure 2 -. Clustering of CRP levels reveals two phenotypically distinct clusters and increased CRP levels at diagnosis associate with worse haemodynamics and functional outcomes.

Figure 2 -

Open in a new tab

(A) Geometric means with standard deviations of serial CRP levels based on clustering of longitudinal log-transformed CRP levels in 236 PAH patients from the discovery UK cohort at diagnosis (visit 0 and the subsequent annual study visits for five years) 131 of whom were in the 'high' CRP cluster and 105 in the normal CRP cluster. (B) Box Plots (presented as median +/− IQR) of clinical characteristics between normal and high CRP clusters (n=236): body mass index (BMI), right atrial pressure (RAP), pulmonary arterial wedge pressure (PAWP), estimated glomerular filtration rates (eGFR), and 6-minute walk distance (6MWD).

Table 1 -. Differences between CRP clusters.

Abbreviations: CTD PAH: connective tissue disease PAH, PVOD: pulmonary veno-occlusive disease, PCH: pulmonary capillary haemangiomatosis, TSH: thyroid stimulating hormone, fT4: free T4, eGFR: estimated glomerular filtration rate, 6MWD: 6-minute walk distance, Hb: haemoglobin, PVR: pulmonary vascular resistance, RAP: right atrial pressure, dPAP: diastolic pulmonary artery pressure, mPAP: mean pulmonary artery pressure, PAWP: pulmonary arterial wedge pressure, FEV1: forced expiratory volume in 1 second, FVC: forced vital capacity, KCO: carbon monoxide transfer coefficient, DLCO: Diffusing Capacity of the Lung for Carbon Monoxide.

Variable Level Normal CRP
(n=105)		 High CRP (
n=131)		 Total (n=236) P-
value		 FDR
corrected p-
value
CRP (mg/l) Median [iqr] 2 [1.2, 3.5] 6.5 [3.2, 10.6] 3.9 [2, 8] NA
Baseline characteristics
Sex female 79 (75.2%) 93 (71.0%) 172 (72.9%) 0.56 0.76
male 26 (24.8%) 38 (29.0%) 64 (27.1%)
Age at diagnosis median [iqr] 40 [32.0, 54.5] 45.6 [34.2, 58.4] 43.3 [33.4, 57.2] 0.037 0.20
BMI (kg/m2) median [iqr] 25.4 [21.9, 29.6] 30.8 [26.6, 36.1] 28.2 [23.8, 33.0] < 0.0001 < 0.0001
PAH subtype Congenital heart disease PAH 3 (2.9%) 4 (3.1%) 7 (3.0%) 0.71 0.87
CTD PAH 0 (0.0%) 1 (0.8%) 1 (0.4%)
PVOD 3 (2.9%) 2 (1.5%) 5 (2.1%)
HPAH 8 (7.6%) 11 (8.4%) 19 (8.1%)
IPAH 91 (86.7%) 109 (83.2%) 200 (84.7%)
WHO group 2 PH 0 (0.0%) 2 (1.5%) 2 (0.8%)
WHO group 3 PH 0 (0.0%) 1 (0.8%) 1 (0.4%)
WHO group 4 PH 0 (0.0%) 0 (0.0%) 0 (0.0%)
WHO group 5 PH 0 (0.0%) 1 (0.8%) 1 (0.4%)
TSH (mu/l) median [iqr] 1.9 [1.1, 2.5] 2.1 [1.3, 3.5] 2 [1.2, 3.0] 0.16 0.39
fT4 (pmol/l) median [iqr] 14.2 [12.8, 16.6] 14.1 [12.9, 16.1] 14.2 [12.8, 16.3] 0.74 0.87
missing 38 44 82
eGFR (ml/min/1.73m2) median [iqr] 73.7 [62.0, 86.7] 70.3 [58.3, 84.1] 71.4 [59.3, 86.0] 0.062 0.24
Platelets (*109/l) median [iqr] 218 [183.8, 272.2] 228 [183.8, 269.8] 225.5 [183.8, 272.0] 0.87 0.98
Hb (g/l) median [iqr] 155 [140.0, 163.2] 151 [137.0, 169.5] 153 [138, 167] 0.97 1
NT-pro-BNP (ng/l) median [iqr] 462 [106, 965] 2,000 [ 304, 2,477] 1,220.5 [ 269.2, 2,403.2] 0.070 0.24
BNP (ng/l) median [iqr] 126.8 [ 42.4, 468.8] 252.9 [ 73, 456] 205.7 [ 47.9, 462.4] 0.37 0.76
Functional markers
Resting SpO2 (%) median [iqr] 97 [95, 98] 95 [92, 97] 96 [94, 98] 0.0044 0.039
Heart rate (BPM) median [iqr] 80.5 [73.2, 94.0] 83 [75, 92] 82 [74.0, 92.5] 0.73 0.87
6MWD distance (metres) median [iqr] 355.5 [263.2, 441.5] 300 [125.5, 363.0] 313 [196, 414] 0.00045 0.0061
WHO functional class 1 2 (2.0%) 0 (0.0%) 2 (0.9%) 0.23 0.51
2 14 (13.7%) 11 (8.6%) 25 (10.9%)
3 71 (69.6%) 99 (77.3%) 170 (73.9%)
4 15 (14.7%) 18 (14.1%) 33 (14.3%)
Haemodynamics
PVR (dynes/sec/cm−5) median [iqr] 1,000 [ 643.9, 1,328.7] 962.1 [ 615.4, 1,263.2] 978.9 [ 626.5, 1,277.8] 0.94 1
Cardiac Output (ml/min) median [iqr] 3.6 [3.1, 4.6] 3.9 [3.0, 5.1] 3.8 [3.0, 4.7] 0.46 0.76
Cardiac Index (l/min/m2) median [iqr] 2 [1.7, 2.4] 1.9 [1.6, 2.5] 2 [1.6, 2.5] 0.41 0.76
RAP (mmHg) median [iqr] 8 [ 6, 11] 10 [ 7, 14] 9 [ 6, 13] 0.0099 0.067
dPAP (mmHg) median [iqr] 34.5 [29, 42] 36.5 [28.2, 42.0] 36 [29, 42] 0.45 0.76
mPAP (mmHg) median [iqr] 54 [45.5, 62.5] 56 [46, 65] 55.5 [46, 65] 0.50 0.76
PAWP (mmHg) median [iqr] 10 [ 7, 12] 10 [ 8, 13] 10 [ 8, 13] 0.045 0.20
Vasodilator responder no 30 (71.4%) 36 (73.5%) 66 (72.5%) 1 1
yes 12 (28.6%) 13 (26.5%) 25 (27.5%)
Pulmonary function tests
FEV1 (% predicted) median [iqr] 84.5 [74, 99] 81.5 [70.0, 94.2] 83.5 [73, 96] 0.12 0.33
FVC (% predicted) median [iqr] 94 [ 82.0, 107.2] 90.4 [ 80, 103] 92 [ 80, 104] 0.11 0.33
KCO (% predicted) median [iqr] 74.5 [64.0, 83.8] 75 [61.0, 90.5] 75 [64.0, 89.5] 0.51 0.76
DLCO (% predicted) median [iqr] 71.7 [60.7, 81.6] 69.8 [51.7, 83.0] 70.3 [58.2, 81.8] 0.54 0.76
Missing variables (n) BMI: 10, TSH: 41, fT4: 82, eGFR: 44, Plaetelets: 8 Hb: 9, NT-proBNP: 202, BNP: 129, Resting SpO2: 18, Heart rate: 25, 6MWD: 92, WHO FC: 6, PVR: 28, Cardiac output: 18, Cardiac index: 111, RAP: 25, dPAP: 14, mPAP: 10, PAWP: 39, Vasodilator responder: 145, FEV1: 27, FVC: 31, KCO: 42, DLCO: 91
Open in a new tab

Diagnostic CRP levels associate with clinical characteristics in PH

As the CRP levels in the clusters remained stable over time, we investigated whether the identified associations could be replicated using diagnostic (or first study visit) CRP in both the UK PAH cohort (n=936) and the all-cause PH ASPIRE cohort (n=5036). We applied the clinical threshold for CRP of >5 mg/l. We found again that predominantly BMI-attributable clinical differences, such as increased BMI and RAP and decreased forced vital capacity (FVC) and walk distance were present in those with CRP above the clinical threshold. Renal function (eGFR) was significantly lower (62.6 versus 68.6 ml/min/1.73m2) and pulmonary arterial wedge pressure (PAWP) was significantly higher in patients with mildly raised CRP in the UK cohort (10 versus 9 mmHg) and pulmonary vascular resistance (PVR) was significantly worse in these patients in the ASPIRE cohort (524 versus 484 dynes/sec/cm−5; Supplemental Figure E2 and Supplemental Tables E3/E4). Longitudinal CRP trajectories of these two groups also remained stable in all cohorts (Supplemental Figure E3).

CRP and BMI are correlated in PH when it comes to clinical characteristics and comorbidities, but have opposing effects on key haemodynamics

To better examine the relationship between CRP and BMI and clinical characteristics, linear models were constructed in both the UK cohort and ASPIRE. The models showed that CRP and BMI positively correlate with each other (p<0.0001) and that, even after correction for synergistic effects, to ensure the CRP association was not marking an association with BMI, and relevant covariates (including age, sex, and comorbidities), they significantly influence clinical outcomes and physiological markers. Raised CRP and BMI associated with higher RAP, worse walking distance, eGFR, and FVC. However, CRP and BMI have opposing effects on PVR and cardiac output (CO). Raised CRP is associated with higher PVR, but high BMI associates with lower PVR, possibly related to their differing associations with CO (Figure 3 & Supplemental Figure E4A/B). This remains true when adjusting for connective tissue disease (CTD) status in group 1 PAH, in which CRP is significantly raised (Supplemental Figure E5).

Figure 3 -. CRP and BMI are interrelated in PH and have opposing effects on haemodynamics.

Figure 3 -

Open in a new tab

Linear models describing the effect of CRP (A) and BMI (B) on Walk distance, PVR: pulmonary vascular resistance, CRP, and BMI. Summarised effects are presented as standardised coefficients with 95% confidence interval, of CRP level (A) and BMI level (B) on clinical parameters in both the discovery UK cohort (n=936)) and validation ASPIRE cohort (n=5036). Models were adjusted for age at diagnosis, sex, BMI (excluding the model for BMI), CRP level (excluding the model or CRP), and Charlson Comorbidity Index (CCI) score. Walk distance was measured as 6-minute walk distance tests in the UK cohort and Incremental Shuttle Walk Distance tests in ASPIRE.

The findings of these linear models relating to BMI were reconfirmed when comparing clinical differences between BMI groups in the UK cohort and validation ASPIRE and FDA cohorts. For example, increased BMI correlated with reduced PVR (Supplemental Tables E5/E6/E7; average difference between population and of obesity-class III median in all groups: −136 dynes/sec/cm−5) and high baseline BMI significantly associated with reduced walk distance in all three cohorts. Despite correlating with worse baseline walk distance (average 6MWD difference between population and obesity-class III median: −102 metres), BMI group does not associate with functional treatment response (measured by improved walk distance in all three cohorts; Supplemental Figure E6).

To assess associations between comorbidities and CRP levels as well as BMI levels, we calculated odds ratios for comorbid diseases for patients with raised CRP and obese patients in the UK cohort (supplemental figure E4C/D). In PAH, raised CRP levels associate with obesity (OR 2.7) and comorbidities that also associate with increased BMI, such as hypertension (OR 1.4), type-II diabetes (OR 1.5), and smoking history (OR 1.7). Increased CRP levels further associate with COPD (OR 1.7) and myocardial infarctions (OR 1.9) (Supplemental Figure E4C/D) and increased BMI associates with a diagnosis of asthma (OR 2.1). Only six patients in the UK cohort were being treated with SGLT-2 inhibitors and eight with GLP-1 analogues respectively, reducing the risk of these drug effects on associations. Overall, patients with raised CRP levels had more (severe) comorbidities when measured with a Charlson Comorbidity Index (CCI) score (p=0.017; Supplemental Table E3).

Raised CRP levels associate with poor survival whilst raised BMI is associated with improved survival in PH

As CRP levels and BMI associated with clinical characteristics and comorbidities, we investigated the survival associations between raised CRP (>5 mg/l) and BMI survival in both the UK, ASPIRE and FDA cohorts. Increased CRP levels at diagnosis, in both PAH and all-cause PH respectively, were significantly associated with worse survival (HR 1.24-1.56; Figure 4). This association remains when correcting for associated covariates, age at diagnosis, sex, haemodynamics, BMI, and CCI score (latter only available in the UK cohort; Figure 4), or when including REVEAL Lite 2 risk scores in solely PAH patients (Supplemental Figure E7). To account for the possibility that the significant association between raised CRP levels and mortality was due to intercurrent illness/infections, patients with an initial CRP level over four interquartile ranges from the median were excluded from survival analysis , yielding similar results (Supplemental Figure E8), proving the robust association between CRP levels and survival. Raised CRP levels significantly associated with increased mortality in each PH subgroup when analysed separately (HR 1.45-1.95), except in group 3 PH (HR 1.16, 95%-CI 0.95-1.37), which was notable for its poor prognosis (Figure 4 and Supplemental Figure E9).

Figure 4 -. CRP is an independent predictor of mortality in both PAH and all-cause PH.

Figure 4 -

Open in a new tab

A & B: Kaplan-Meier survival curves, stratified based on diagnostic (or first study visit) CRP (threshold of >5mg/l), for 10 year survival in the discovery UK cohort (n=823) (A) and 5 year survival in the validation ASPIRE cohort (n=5036) (B). C: Cox-proportional hazard model for raised versus normal CRP level adjusting for age at diagnosis, sex, BMI, right atrial pressure (RAP), pulmonary vascular resistance (PVR), Charlson Comorbidity Index (CCI) score in the UK cohort (n=736), and additionally PH subtype in ASPIRE (n=5036).

A significant association between increased BMI with improved outcomes for all-cause PH was observed in the ASPIRE cohort, even after adjustment for CRP level, (p<0.0001, Figure 5), whilst non-significant directionality towards better outcomes was seen in the UK cohort (Supplemental Figure E10). A Cox-proportional hazard model showed increased BMI had clear opposing associations with survival compared to increases in CRP level (Figure 5). This finding was further validated in PAH in the FDA cohort (n=3957) (Supplemental Figure E10).

Figure 5 -. Increased BMI has a protective association with survival in PAH and all-cause PH.

Figure 5 -

Open in a new tab

(A) Kaplan-Meier curves for BMI classes in ASPIRE (n=4365). (B) A Cox-proportional hazard model, estimating survival probability at 5 years post diagnosis, using WHO BMI group and CRP level (up to 50 mg/l to prevent overfitting).

Smoking can cause reverse causation bias (31). Rates were low and not different between groups in the FDA cohort (0.9% current and 0.4% past smokers). Additionally, a between groups Cox-proportional hazards model was constructed for all three cohorts, accounting for age, sex, haemodynamics, CRP level, smoking status, as well as CCI score. Increased BMI remained a significant predictor of improved survival (Supplemental Figure E10). To address the possibility of reverse causation bias due to early mortality because of cachexia in the underweight group, survival analysis was performed after excluding patients with survival less than one year post diagnosis which confirmed the association of raised BMI with reduced mortality (Supplemental Figure E11). Similarly, substituting PVR with indexed PVR in our Cox-proportional hazards models resulted in similar findings (Supplemental Figure E5).

CRP and BMI associate with partially overlapping molecular endotypes with notable differences in multi-omics analyses

To gain mechanistic insight in the observed clinical differences, multi-omics datasets in the UK cohort were interrogated, including autoantibodies (n=374 in the whole cohort and 164 for the patients in the time series clustering), transcriptomics (n up to 374), and proteomics (n up to 441).

CRP levels in 374 UK cohort patients, and between the CRP clusters, did not correlate with 19 clinically relevant autoantibody levels, suggesting systemic inflammation in PAH is not always associated with autoreactivity (Supplemental Figure E12).

Analysis of transcriptomics of 85 pre-defined inflammatory genes showed that differences in CRP group associated with an inflammatory signature notable for increased TGF-beta and IL-1 signalling which was not present in the BMI groups (Supplemental Figure E13). There were no significantly altered genes in the BMI analyses after false discovery correction. Eighty inter-related inflammatory proteins in the proteomics dataset were assessed between CRP and BMI groups. There remained significant differences in TGF-beta and IL-1 signalling with shared significant expression restricted to chiefly CRP itself, HCC-1, MIP-1a and IL-1RA (Supplemental Figure E13 and Supplemental Table E8). Reassuringly, CRP was the most significant protein in both CRP and BMI analyses (p=6.9*10−13 and p=1.4*10−13 respectively).

Discussion

This study, the largest to date on the association between inflammation and outcomes in PH, showed unexpected prominence of BMI and obesity-related phenotypic features in the clustering analyses. Whilst we were unable to support our original hypothesis that inflammation would associate with autoreactivity, we report a strong association between BMI and CRP in PH. Whilst CRP is correlated with BMI in all-causes of PH, raised CRP and BMI have diverging associations with haemodynamics and survival. These observations are consistent in some of the largest studies ever undertaken in PH and validated in diverse datasets representing both UK and international patient populations.

It is well known that low grade inflammation, as measured by mildly increased levels of CRP associates with cardiovascular diseases and mortality (12). In HFpEF haemodynamics and CRP level also associate (32). Moreover, the association between PH, inflammation and associated adverse functional outcomes, mortality and haemodynamics has been described before albeit in modestly sized studies (17, 18, 33). We robustly confirm mildly raised CRP levels associate with increased mortality, worse haemodynamics, and worse functional outcomes in all cause PH and uniquely in the field the size of our studies has allowed us to adjust for key covariates including BMI, haemodynamics, and CTD status. The decision to use a 5 mg/l cut-point for CRP was informed by clinical practice and previous literature (17, 28). However, we note that our linear and Cox-proportional hazard models for BMI, still show significant associations with clinical variables and outcomes.

Other markers of inflammation, such as IL-6, which associates with CRP, also associate with mortality in PAH (12, 34). The association of CRP with mortality, however, is unlikely to be a direct consequence of raised CRP. Inflammation acts as surrogate of a potentially large number of measured and unmeasured confounders including increased prevalence of comorbidities (Figure 6). Indeed, causal inference using Mendelian randomisation has thus far failed to identify causal associations between CRP and, for example, cardiovascular mortality (35). Raised CRP levels significantly associated with age, raising the question whether it is a mediator or confounder of the survival association, in effect a marker of PH sub-phenotypes, some of which associate with age (36, 37).

Figure 6 -. Proposed summary of associations related to obesity and inflammation influencing survival in PH patients.

Figure 6 -

Open in a new tab

Summarised effects of associations between obesity and inflammation. Stimulating associations are visualised with a green arrow, inhibitions are visualised with a red inhibitor and when causal pathways are hypothesised but not supported by data in this current paper, lines are dotted. Created in https://BioRender.com.

Obesity is an increasing problem in PH, affecting 24.1%-39.8% of patients in our studies. Like CRP, increased baseline BMI associated, after corrections for key covariates including CRP, with worse baseline functional outcomes including PAWP, eGFR, and 6MWD; a major regulatory endpoint for trials. The effect size of increased BMI on baseline 6MWD was greater than historical treatment effects and the minimal important difference for 6MWD (38, 39) and BMI did not influence treatment effects in our three studies, validating previous studies (40-42),

Despite associating with worse baseline characteristics, increased BMI associated with improved survival and PVR in our studies. This ‘obesity paradox’ has been contentious in other cardiorespiratory diseases and has conflicting evidence in PH (31, 41, 43, 44). Previous study sample sizes were either limited or case composition was less broad and inclusive compared to our studies.

Our dataset is unusual for its size and variables available to address confounding or collider bias, such as smoking status and concomitant medication (31, 45). We were also able to address reverse causation and a Cox-proportional hazards model showed the protective association of raised BMI was not simply due to increased mortality in the underweight group.

Despite the fact that BMI and CRP generally both associate with worse baseline characteristics, the association between BMI and PVR and survival diverges from that of CRP. There are multiple potential explanations for this.

One possibility is that there is an endotype of PH with better survival that is enriched for obese patients, in which BMI would only indirectly be associated with survival.. Indeed, molecular signatures in BMI and CRP groups, whilst partly overlapping, differed in our study. The CRP groups had a distinct transcriptomic and proteomic signature notable for TGF-beta signalling. The observational nature and smaller sample size of our multi-omics sub-studies limit the ability to explore or comment on causality. Moreover, social determinants of health are associated with both BMI and outcomes (46, 47) but were not measured in the current study. Broadly, however, our findings are in line with inflammatory pathways known to relate to both CRP and BMI.

Increased BMI in all three cohorts in this study was associated with improved cardiac output and index as well as lower PVR. There are a number of potential explanations for this phenomenon including changes in peripheral vascular bed volume, changes in thoracic pressures and altered cardiac filling (48). An obvious question is whether this is the dominant cause of improved long-term outcomes. Additional obesity-related factors such as adipokines and oestrogens are thought to be involved in the pathophysiology of PH and further work is required to understand their associations with clinical outcomes in PH (40, 49).

Our study shows that the PH population is heterogeneous and inflammation is a complicated concept associating with more than just autoreactivity or a simple PH focussed inflammatory response. Confounding bias in modern PH populations is high from diverse demographic features including increasing co-morbidity. The desire to stratify patients by biomarkers, both for endotyping and for consideration in clinical trials will potentially miss the contribution of confounding variables. For example, a patient with raised CRP driven by BMI may have a different disease course than one with raised CRP related to other factors and may not be amenable to immunomodulatory drugs as the inflammation does not contribute directly or indirectly to PH and the disease course is different in this context. Despite trials frequently excluding obese patients, for example the recent Sotatercept study (NCT03496207), stratification on obesity in trials may be of importance due to its association with outcomes.

Limitations of our work include the observational nature of our studies, which precludes causal inference and unmeasured confounders may have been present. Additionally, we lacked alternative markers of adiposity such as waist height ratio, which, for example, in HFpEF associates with adverse outcomes (50).

In summary, we present the largest study to date showing a previously unreported, strong association between CRP and BMI in PH populations. Despite the interrelatedness of CRP and BMI and their comparable effects on functional outcomes, they have opposing effects on haemodynamics and survival. This has implications for our understanding of PH, the impact of comorbidity on disease and the design of clinical trials.

Supplementary Material

supplemental material

At a Glance Commentary.

What is known:

Inflammation is a key feature of pulmonary hypertension (PH) and known to associate with diverse disease endotypes and outcomes. In the general population raised body mass index (BMI) and inflammation are correlated and both associate with increased morbidity and mortality. Roughly one third of pulmonary hypertension patients are obese and the association of BMI and survival in PH is contentious. No definitive studies have explored the relationship and importance of BMI and inflammation in PH.

What this study adds:

This work demonstrates that CRP and PH are associated across three international cohorts in over 10,000 PH patients. C-reactive protein (CRP) levels remain stable over time in PH patients and associate strongly with BMI and BMI-related traits. Despite being strongly associated, raised CRP was associated with worse outcomes across all cause PH whilst raised BMI associated with improved survival. This could potentially be explained by differing associations with pulmonary haemodynamics, underlying disease phenotypes/endotypes and confounding via comorbid diseases. This has important implications for clinical trial design and clinical practice.

Acknowledgements

Dr. Guillermo Reales (Cambridge Institute for Therapeutic Immunology and Infectious Disease) for his advice regarding data analysis/software setup, Dr. Garnett and Dr. Stockbridge (U.S. Food and Drug Administration) for providing the FDA dataset to the team of Prof. Benza and the National Institutes of Health for funding the research on the FDA dataset (R01 HL164906-05). UK PAH COHORT CONSORTIUM, UNIPHY CLINICAL TRIALS NETWORK, ASPIRE registry.

Sources of Funding

National Institute for Health Research BioResource and National Institute for Health Research Cardiorespiratory Biomedical Research Centre, Gates Cambridge Trust (grant number OPP1144), MRC (MC_UU_00040/01), Wellcome Trust (WT220788), NIHR Sheffield Biomedical Research Centre (NIHR203321), BHF Basic Science Research fellowship (FS/SBSRF/21/31025), Wellcome Trust Clinical Research Development Fellowship (206632/Z/17/Z), British Heart Foundation Clinical Research Training Fellowship (FS/CRTF/22/24390). Royal Netherlands Academy of Sciences (CVON-2017-10 Dolphin-Genesis), the Dutch Heart Foundation, Dutch Federation of University Medical Centres, Netherlands Organisation for Health Research and Development, The Netherlands Organization for Scientific Research (NWO-VICI: 918.16.610, NWO-VIDI: 917.18.338), Royal Netherlands Academy of Sciences (CVON-2012-08 PHAEDRA & CVON-2018-29 PHAEDRA-IMPACT), National Institutes of Health (R01 HL164906-05). The UK National Cohort of Idiopathic and Heritable PAH was supported by: the NIHR BioResource; the British Heart Foundation (BHF SP/12/12/29836) and the UK Medical Research Council (MR/K020919/1).

The funders had no role in study design, data collection, data analysis, data interpretation, or writing of this article.

National Cohort Study of IPAH/HPAH Consortium

Marta Bleda, Charaka Hadinnapola, Matthias Haimel, Kate Auckland, Tobias Tilly, Emilia Swietlik, Margaret Day, Alan Greenhalgh, Debbie Shipley, Val Irvine, Fiona Kennedy, Shahin Moledina, Lynsay MacDonald, Eleni Tamvaki, Anabelle Barnes, Victoria Cookson, Latifa Chentouf, Souad Ali, Shokri Othman, Lavanya Ranganathan, J. Simon, R. Gibbs, Mahitha Gummadi, Rosa DaCosta, Joy Pinguel, Natalie Dormand, Alice Parker, Della Stokes, Dipa Ghedia, Yvonne Tan, Tanaka Ngcozana, Ivy Wanjiku, Gary Polwarth, John Cannon, Karen K. Sheares, Dolores Taboda, Rob V. Mackenzie Ross, Jay Suntharalingam, Mark Grover, Ali Kirby, Richard Trembath, Nicholas Morrell, Martin Wilkins, Colin Church, John G Coghlan, Luke Howard, Stephen J. Wort, Christopher J. Rhodes, Allan Lawrie, Joanna Pepke-Zaba, Stefan Gräf, Mark Toshner.

ASPIRE Consortium author list

David G Kiely*#, Lisa Watson*#, Iain Armstrong*#, Catherine Billings*#, Athanasios Charalampopoulos*#, Robin Condliffe*#, Charlie Elliot*#, Abdul Hameed*#, Neil Hamilton*#, Judith Hurdman*#, Allan Lawrie^#, Robert A Lewis*#, Smitha Rajaram*#, Alex Rothman^#, Andy J. Swift^#, Steven Wood*#, AA Roger Thompson^#, Jim Wild^#.

* Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK

^ University of Sheffield, Sheffield, UK.

# National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)

Acknowledgements for ASPIRE:

This ASPIRE registry is supported by the National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care

Data sharing

De-identified data for the UK cohort will be shared upon reasonable request (based on local protocols and law). Requests for access to the UK cohort data can be made to the corresponding author Dr Mark Toshner (mrt34@cam.ac.uk). Requests for ASPIRE can be made via the following link: https://bit.ly/aspire-registry.

References

  • 1.Mocumbi A, Humbert M, Saxena A, Jing Z-C, Sliwa K, Thienemann F, Archer SL, Stewart S. Pulmonary hypertension. Nature Reviews Disease Primers 2024; 10: 1. [Google Scholar]
  • 2.Humbert M, Sitbon O, Guignabert C, Savale L, Boucly A, Gallant-Dewavrin M, McLaughlin V, Hoeper MM, Weatherald J. Treatment of pulmonary arterial hypertension: recent progress and a look to the future. The Lancet Respiratory Medicine 2023; 11: 804–819. [DOI] [PubMed] [Google Scholar]
  • 3.Simon B, Rafael Sobrano F, Sue G, Andrea F, Tim L. Pathophysiology and new advances in pulmonary hypertension. BMJ Medicine 2023; 2: e000137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Jones RJ, De Bie EMDD, Groves E, Zalewska KI, Swietlik EM, Treacy CM, Martin JM, Polwarth G, Li W, Guo J, Baxendale HE, Coleman S, Savinykh N, Coghlan JG, Corris PA, Howard LS, Johnson MK, Church C, Kiely DG, Lawrie A, Lordan JL, Mackenzie Ross RV, Pepke Zaba J, Wilkins MR, Wort SJ, Fiorillo E, Orrù V, Cucca F, Rhodes CJ, Gräf S, Morrell NW, McKinney EF, Wallace C, Toshner M. Autoimmunity is a Significant Feature of Idiopathic Pulmonary Arterial Hypertension. Am J Respir Crit Care Med 2022. [Google Scholar]
  • 5.Kariotis S, Jammeh E, Swietlik EM, Pickworth JA, Rhodes CJ, Otero P, Wharton J, Iremonger J, Dunning MJ, Pandya D, Mascarenhas TS, Errington N, Thompson AAR, Romanoski CE, Rischard F, Garcia JGN, Yuan JXJ, An T-HS, Desai AA, Coghlan G, Lordan J, Corris PA, Howard LS, Condliffe R, Kiely DG, Church C, Pepke-Zaba J, Toshner M, Wort S, Gräf S, Morrell NW, Wilkins MR, Lawrie A, Wang D, Bleda M, Hadinnapola C, Haimel M, Auckland K, Tilly T, Martin JM, Yates K, Treacy CM, Day M, Greenhalgh A, Shipley D, Peacock AJ, Irvine V, Kennedy F, Moledina S, MacDonald L, Tamvaki E, Barnes A, Cookson V, Chentouf L, Ali S, Othman S, Ranganathan L, Gibbs JSR, DaCosta R, Pinguel J, Dormand N, Parker A, Stokes D, Ghedia D, Tan Y, Ngcozana T, Wanjiku I, Polwarth G, Mackenzie Ross RV, Suntharalingam J, Grover M, Kirby A, Grove A, White K, Seatter A, Creaser-Myers A, Walker S, Roney S, Elliot CA, Charalampopoulos A, Sabroe I, Hameed A, Armstrong I, Hamilton N, Rothman AMK, Swift AJ, Wild JM, Soubrier F, Eyries M, Humbert M, Montani D, Girerd B, Scelsi L, Ghio S, Gall H, Ghofrani A, Bogaard HJ, Vonk Noordegraaf A, Houweling AC, Veld AHit, Schotte G, Trembath RC, Consortium UKNPCS. Biological heterogeneity in idiopathic pulmonary arterial hypertension identified through unsupervised transcriptomic profiling of whole blood. Nature Communications 2021; 12: 7104. [Google Scholar]
  • 6.Breitling S, Ravindran K, Goldenberg NM, Kuebler WM. The pathophysiology of pulmonary hypertension in left heart disease. American Journal of Physiology-Lung Cellular and Molecular Physiology 2015; 309: L924–L941. [DOI] [PubMed] [Google Scholar]
  • 7.Chaouat A, Savale L, Chouaid C, Tu L, Sztrymf B, Canuet M, Maitre B, Housset B, Brandt C, Le Corvoisier P, Weitzenblum E, Eddahibi S, Adnot S. Role for Interleukin-6 in COPD-Related Pulmonary Hypertension. Chest 2009; 136: 678–687. [DOI] [PubMed] [Google Scholar]
  • 8.Quarck R, Wynants M, Verbeken E, Meyns B, Delcroix M. Contribution of inflammation and impaired angiogenesis to the pathobiology of chronic thromboembolic pulmonary hypertension. European Respiratory Journal 2015; 46: 431–443. [DOI] [PubMed] [Google Scholar]
  • 9.Kalantari S, Gomberg-Maitland M. Group 5 Pulmonary Hypertension: The Orphan’s Orphan Disease. Cardiology Clinics 2016; 34: 443–449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Matsumoto K, Miyawaki Y, Katsuyama T, Nakadoi T, Shidahara K, Hirose K, Nawachi S, Asano Y, Katayama Y, Katsuyama E, Takano-Narazaki M, Matsumoto Y, Mori A, Akagi S, Sada K-E, Wada J. Immunosuppressive Treatment for an anti-U1 Ribonucleoprotein Antibody-positive Patient with Pulmonary Arterial Hypertension: A Case Report. Internal Medicine 2023; advpub. [Google Scholar]
  • 11.Toshner M, Church C, Harbaum L, Rhodes C, Villar Moreschi SS, Liley J, Jones R, Arora A, Batai K, Desai AA, Coghlan JG, Gibbs JSR, Gor D, Gräf S, Harlow L, Hernandez-Sanchez J, Howard LS, Humbert M, Karnes J, Kiely DG, Kittles R, Knightbridge E, Lam B, Lutz KA, Nichols WC, Pauciulo MW, Pepke-Zaba J, Suntharalingam J, Soubrier F, Trembath RC, Schwantes-An T-HL, Wort SJ, Wilkins MR, Gaine S, Morrell NW, Corris PA. Mendelian randomisation and experimental medicine approaches to interleukin-6 as a drug target in pulmonary arterial hypertension. European Respiratory Journal 2022; 59: 2002463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Collaboration ERF. C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. The Lancet 2010; 375: 132–140. [Google Scholar]
  • 13.Rizo-Téllez SA, Sekheri M, Filep JG. C-reactive protein: a target for therapy to reduce inflammation. Front Immunol 2023; 14: 1237729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-Reactive Protein Levels in Overweight and Obese Adults. JAMA 1999; 282: 2131–2135. [DOI] [PubMed] [Google Scholar]
  • 15.Poms AD, Turner M, Farber HW, Meltzer LA, McGoon MD. Comorbid Conditions and Outcomes in Patients With Pulmonary Arterial Hypertension: A REVEAL Registry Analysis. Chest 2013; 144: 169–176. [DOI] [PubMed] [Google Scholar]
  • 16.Kosiborod MN, Abildstrøm SZ, Borlaug BA, Butler J, Rasmussen S, Davies M, Hovingh GK, Kitzman DW, Lindegaard ML, Møller DV, Shah SJ, Treppendahl MB, Verma S, Abhayaratna W, Ahmed FZ, Chopra V, Ezekowitz J, Fu M, Ito H, Lelonek M, Melenovsky V, Merkely B, Núñez J, Perna E, Schou M, Senni M, Sharma K, Van der Meer P, von Lewinski D, Wolf D, Petrie MC. Semaglutide in Patients with Heart Failure with Preserved Ejection Fraction and Obesity. New England Journal of Medicine 2023; 389: 1069–1084. [DOI] [PubMed] [Google Scholar]
  • 17.Quarck R, Nawrot T, Meyns B, Delcroix M. C-Reactive Protein: A New Predictor of Adverse Outcome in Pulmonary Arterial Hypertension. J Am Coll Cardiol 2009; 53: 1211–1218. [DOI] [PubMed] [Google Scholar]
  • 18.Joppa P, Petrasova D, Stancak B, Tkacova R. Systemic inflammation in patients with COPD and pulmonary hypertension. Chest 2006; 130: 326–333. [DOI] [PubMed] [Google Scholar]
  • 19.Hurdman J, Condliffe R, Elliot CA, Davies C, Hill C, Wild JM, Capener D, Sephton P, Hamilton N, Armstrong IJ, Billings C, Lawrie A, Sabroe I, Akil M, O’Toole L, Kiely DG. ASPIRE registry: Assessing the Spectrum of Pulmonary hypertension Identified at a REferral centre. European Respiratory Journal 2012; 39: 945–955. [DOI] [PubMed] [Google Scholar]
  • 20.Galiè N, Barberà JA, Frost AE, Ghofrani H-A, Hoeper MM, McLaughlin VV, Peacock AJ, Simonneau G, Vachiery J-L, Grünig E. Initial use of ambrisentan plus tadalafil in pulmonary arterial hypertension. New England Journal of Medicine 2015; 373: 834–844. [DOI] [PubMed] [Google Scholar]
  • 21.Oudiz RJ, Torres F, Frost AE, Badesch DB, Olschewski H, Galie N, McGoon MD, McLaughlin V, Rubin LJ. ARIES-1: a placebo-controlled, efficacy and safety study of ambrisentan in patients with pulmonary arterial hypertension. Chest 2006; 130: 121S. [Google Scholar]
  • 22.Galiè N, Olschewski H, Oudiz RJ, Torres F, Frost A, Ghofrani HA, Badesch DB, McGoon MD, McLaughlin VV, Roecker EB. Ambrisentan for the treatment of pulmonary arterial hypertension: results of the ambrisentan in pulmonary arterial hypertension, randomized, double-blind, placebo-controlled, multicenter, efficacy (ARIES) study 1 and 2. Circulation 2008; 117: 3010–3019. [DOI] [PubMed] [Google Scholar]
  • 23.White RJ, Jerjes-Sanchez C, Bohns Meyer GM, Pulido T, Sepulveda P, Wang KY, Grünig E, Hiremath S, Yu Z, Gangcheng Z, Yip WLJ, Zhang S, Khan A, Deng CQ, Grover R, Tapson VF, Svetliza GN, Lescano AJ, Bortman GR, Diez FA, Botta CE, Fitzgerald J, Feenstra E, Kermeen FD, Keogh AM, Williams TJ, Yousseff PP, Ng BJ-H, Smallwood DM, Dwyer NB, Brown MR, Lang IM, Steringer-Mascherbauer R, Arakaki JSO, Campos FTAF, de Amorim Correa R, de Souza R, Bohns Meyer GM, Moreira MAC, Yoo HHB, Lapa MS, Swiston J, Hirani N, Mehta S, Michelakis E, Sepulveda PA, Blancaire MMZ, Liu J, Shuyang Z, Pan L, Chunde B, Qun Y, Xiaoshu C, Zaixin Y, Li X, Hua Y, Gangcheng Z, Zhu X, Chen Y, Zhaozhong C, Yang Y, Daxin Z, Jieyan S, Nielsen-Kudsk JE, Carlsen J, Bourdin A, Hachulla E, Dromer C, Chaouat A, Reynaud-Gauber M, Seronde M-F, Klose H, Halank M, Hoffken G, Ewert R, Rosenkranz S, Grunig E, Kruger U, Kronsbein J, Hauptmeier BM, Koch A, Held M, Lange TJ, Neurohr C, Wilkens H, Wilhelm Wirtz HR, Konstantinides S, Argyropoulou-Pataka P, Orfanos S, Hiremath S, Kerkar PG, Suresh PV, Baxi HA, Oomman A, Abhaichand RK, Arjun PKE, Chopra V, Mehrotra R, Rajput RK, Sawhney JPS, Bimalendu S, Sharma KH, Sastry BKS, Kramer MR, Segel MJ, Ben-Dov I, Berkman N, Yigla M, Adir Y, D’Alto M, Vizza CD, Scelsi L, Vitulo P, Pulido TR, Jerjes-Sanchez C, Boonstra A, Vonk MC, Sobkowicz B, Mularek-Kubzdela T, Torbicki A, Podolec P, Teik LS, Yip WLJ, Chang H-J, Kim H-K, Park J-B, Chang S-A, Kim D-K, Chang S-A, Chung W-J, Song J-M, Nissell M, Hjalmarsson C, Rundqvist B, Huang W-C, Cheng C-C, Hsu C-H, Hsu H-H, Wang K-Y, Coghlan JG, Kiely DG, Pepke-Zaba JW, Lordan JL, Corris PA, Cadaret L, Hansdottir S, Oudiz RJ, Badesch DB, Mathier M, Schilz R, Hill N, Waxman A, Markin CJ, Zwicke DL, Fisher M, Franco V, Sood N, Park MH, Allen R, Feldman JP, Balasubramanian V, Seeram VK, Bajwa A, Thompson AB, Migliore C, Elwing J, McConnell JW, Mehta JP, Rahaghi FF, Rame JE, Khan A, Patel B, Oren RM, Klinger JR, Alnuaimat H, Allen S, Harvey W, Eggert MS, Hage A, Miller CE, Awdish RLA, Cajigas H, Grinnan D, Trichon BH, McDonough C, White RJ, Rischard F. Combination Therapy with Oral Treprostinil for Pulmonary Arterial Hypertension. A Double-Blind Placebo-controlled Clinical Trial. Am J Respir Crit Care Med 2019; 201: 707–717. [Google Scholar]
  • 24.Sitbon O, Channick R, Chin KM, Frey A, Gaine S, Galiè N, Ghofrani H-A, Hoeper MM, Lang IM, Preiss R. Selexipag for the treatment of pulmonary arterial hypertension. New England Journal of Medicine 2015; 373: 2522–2533. [DOI] [PubMed] [Google Scholar]
  • 25.Ghofrani H-A, Galiè N, Grimminger F, Grünig E, Humbert M, Jing Z-C, Keogh AM, Langleben D, Kilama MO, Fritsch A. Riociguat for the treatment of pulmonary arterial hypertension. New England Journal of Medicine 2013; 369: 330–340. [DOI] [PubMed] [Google Scholar]
  • 26.Galiè N, Brundage BH, Ghofrani HA, Oudiz RJ, Simonneau G, Safdar Z, Shapiro S, White RJ, Chan M, Beardsworth A. Tadalafil therapy for pulmonary arterial hypertension. Circulation 2009; 119: 2894–2903. [DOI] [PubMed] [Google Scholar]
  • 27.Pulido T, Adzerikho I, Channick RN, Delcroix M, Galiè N, Ghofrani H-A, Jansa P, Jing Z-C, Le Brun F-O, Mehta S. Macitentan and morbidity and mortality in pulmonary arterial hypertension. New England Journal of Medicine 2013; 369: 809–818. [DOI] [PubMed] [Google Scholar]
  • 28.Guler S, Sarbu A-C, Stalder O, Allanore Y, Bernardino V, Distler J, Gabrielli A, Hoffmann-Vold A-M, Matucci-Cerinic M, Müller-Ladner U, Ortiz-Santamaria V, Rednic S, Riccieri V, Smith V, Ullman S, Walker UA, Geiser TK, Distler O, Maurer B, Kollert F. Phenotyping by persistent inflammation in systemic sclerosis associated interstitial lung disease: a EUSTAR database analysis. Thorax 2023; 78: 1188. [DOI] [PubMed] [Google Scholar]
  • 29.Rhodes CJ, Otero-Núñez P, Wharton J, Swietlik EM, Kariotis S, Harbaum L, Dunning MJ, Elinoff JM, Errington N, Thomson AAR, Iremonger J, Coghlan JG, Corris PA, Howard LS, Kiely DG, Church C, Pepke-Zaba J, Toshner M, Wort SJ, Desai AA, Humbert M, Nichols WC, Southgate L, Trégouët D-A, Trembath RC, Prokopenko I, Gräf S, Morrell NW, Wang D, Lawrie A, Wilkins MR. Whole Blood RNA Profiles Associated with Pulmonary Arterial Hypertension and Clinical Outcome. Am J Respir Crit Care Med 2020. [Google Scholar]
  • 30.Rhodes CJ, Wharton J, Swietlik EM, Harbaum L, Girerd B, Coghlan JG, Lordan J, Church C, Pepke-Zaba J, Toshner M, Wort SJ, Kiely DG, Condliffe R, Lawrie A, Gräf S, Montani D, Boucly A, Sitbon O, Humbert M, Howard LS, Morrell NW, Wilkins MR. Using the Plasma Proteome for Risk Stratifying Patients with Pulmonary Arterial Hypertension. Am J Respir Crit Care Med 2022; 205: 1102–1111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Banack HR, Stokes A. The ‘obesity paradox’ may not be a paradox at all. International Journal of Obesity 2017; 41: 1162–1163. [DOI] [PubMed] [Google Scholar]
  • 32.DuBrock HM, AbouEzzeddine OF, Redfield MM. High-sensitivity C-reactive protein in heart failure with preserved ejection fraction. PLOS ONE 2018; 13: e0201836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Scognamiglio G, Kempny A, Price LC, Alonso-Gonzalez R, Marino P, Swan L, D’ Alto M, Hooper J, Gatzoulis MA, Dimopoulos K, Wort SJ. C-reactive protein in adults with pulmonary arterial hypertension associated with congenital heart disease and its prognostic value. Heart 2014; 100: 1335. [DOI] [PubMed] [Google Scholar]
  • 34.Schwiening M, Swietlik EM, Pandya D, Burling K, Barker P, Feng OY, Treacy CM, Abreu S, Wort SJ, Pepke-Zaba J. Different Cytokine Patterns in BMPR2-Mutation-Positive Patients and Patients With Pulmonary Arterial Hypertension Without Mutations and Their Influence on Survival. Chest 2022; 161: 1651–1656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Association between C reactive protein and coronary heart disease: mendelian randomisation analysis based on individual participant data. BMJ 2011; 342: d548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Skowasch D, Klose H, Ewert R, Wilkens H, Richter M, Rosenkranz S, Setzer G, Grünig E, Halank M. Phenotypes and treatment outcomes in idiopathic pulmonary arterial hypertension patients with comorbidities. ERJ Open Research 2024; 10: 00668–02023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Nossent EJ, Smits JA, Seegers C, Meijboom LJ, Boonstra A, Aman J, De Man FS, Bogaard HJ, Radonic T, Dorfmüller P, Vonk Noordegraaf A. Clinical Correlates of a Nonplexiform Vasculopathy in Patients With a Diagnosis of Idiopathic Pulmonary Arterial Hypertension. CHEST 2024; 166: 190–200. [DOI] [PubMed] [Google Scholar]
  • 38.Mathai SC, Puhan MA, Lam D, Wise RA. The Minimal Important Difference in the 6-Minute Walk Test for Patients with Pulmonary Arterial Hypertension. Am J Respir Crit Care Med 2012; 186: 428–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Pan H-M, McClelland RL, Moutchia J, Appleby DH, Fritz JS, Holmes JH, Minhas J, Palevsky HI, Urbanowicz RJ, Kawut SM, Al-Naamani N. Heterogeneity of treatment effects by risk in pulmonary arterial hypertension. European Respiratory Journal 2023; 62: 2300190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.MacLean MR, Pandya D, Swietlik EM, Denver N, Mair K, Morrell NW, Gräf S, National Cohort Study for I, Heritable Pulmonary Arterial Hypertension C. A pilot study to examine association of BMI with functional class and 6 min walk distance in idiopathic and heritable PAH: Possible association with estrogen metabolism. Pulm Circ 2022; 12: e12139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Weatherald J, Huertas A, Boucly A, Guignabert C, Taniguchi Y, Adir Y, Jevnikar M, Savale L, Jaïs X, Peng M, Simonneau G, Montani D, Humbert M, Sitbon O. Association Between BMI and Obesity With Survival in Pulmonary Arterial Hypertension. Chest 2018; 154: 872–881. [DOI] [PubMed] [Google Scholar]
  • 42.McCarthy BE, McClelland RL, Appleby DH, Moutchia JS, Minhas JK, Min J, Mazurek JA, Smith KA, Fritz JS, Pugliese SC, Urbanowicz RJ, Holmes JH, Palevsky HI, Kawut SM, Al-Naamani N. BMI and Treatment Response in Patients With Pulmonary Arterial Hypertension: A Meta-analysis. CHEST 2022; 162: 436–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Frank RC, Min J, Abdelghany M, Paniagua S, Bhattacharya R, Bhambhani V, Pomerantsev E, Ho JE. Obesity Is Associated With Pulmonary Hypertension and Modifies Outcomes. Journal of the American Heart Association 2020; 9: e014195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.McLean LL, Pellino K, Brewis M, Peacock A, Johnson M, Church AC. The obesity paradox in pulmonary arterial hypertension: the Scottish perspective. ERJ Open Research 2019; 5: 00241–02019. [Google Scholar]
  • 45.King NE, Brittain E. Emerging therapies: The potential roles SGLT2 inhibitors, GLP1 agonists, and ARNI therapy for ARNI pulmonary hypertension. Pulm Circ 2022; 12: e12028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Bann D, Johnson W, Li L, Kuh D, Hardy R. Socioeconomic Inequalities in Body Mass Index across Adulthood: Coordinated Analyses of Individual Participant Data from Three British Birth Cohort Studies Initiated in 1946, 1958 and 1970. PLOS Medicine 2017; 14: e1002214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Booth CM, Li G, Zhang-Salomons J, Mackillop WJ. The impact of socioeconomic status on stage of cancer at diagnosis and survival: a population-based study in Ontario, Canada. Cancer 2010; 116: 4160–4167. [DOI] [PubMed] [Google Scholar]
  • 48.Packer M. The conundrum of patients with obesity, exercise intolerance, elevated ventricular filling pressures and a measured ejection fraction in the normal range. European Journal of Heart Failure 2019; 21: 156–162. [DOI] [PubMed] [Google Scholar]
  • 49.Jia Q, Ouyang Y, Yang Y, Yao S, Chen X, Hu Z. Adipokines in pulmonary hypertension: angels or demons? Heliyon 2023; 9: e22482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Borlaug BA, Jensen MD, Kitzman DW, Lam CSP, Obokata M, Rider OJ. Obesity and heart failure with preserved ejection fraction: new insights and pathophysiological targets. Cardiovascular Research 2022; 118: 3434–3450. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental material
NIHMS2137156-supplement-supplemental_material.pdf (4.2MB, pdf)

Data Availability Statement

De-identified data for the UK cohort will be shared upon reasonable request (based on local protocols and law). Requests for access to the UK cohort data can be made to the corresponding author Dr Mark Toshner (mrt34@cam.ac.uk). Requests for ASPIRE can be made via the following link: https://bit.ly/aspire-registry.

RESOURCES