Global Burden of Pulmonary Arterial Hypertension in Older Adults and Its Causal Association With Body Mass Index

ABSTRACT

Pulmonary arterial hypertension (PAH) is a severe cardiopulmonary disease associated with substantial morbidity and mortality. However, the global burden, future trajectory, and potentially modifiable risk factors for PAH among older adults remain insufficiently characterized. This study aimed to quantify the global burden and projected trends of PAH among adults aged ≥ 60 years and investigate the causal association between body mass index (BMI) and PAH risk using Mendelian randomization (MR). We extracted PAH burden estimates from the Global Burden of Disease (GBD) 2021 study across 204 countries and territories, 21 GBD regions, and five socio‐demographic index (SDI) strata. Temporal trends were assessed using age‐standardized rates and average annual percent changes (AAPCs), while demographic decomposition and Bayesian age‐period‐cohort (BAPC) models were applied to evaluate drivers of burden changes and future trajectories. Two‐sample MR analyses were performed using genome‐wide significant BMI‐associated genetic variants as instrumental variables and PAH genome‐wide association study summary statistics from FinnGen. Between 1990 and 2021, global incident cases of PAH among adults aged ≥ 60 years increased from 6771.01 (95% uncertainty interval [UI]: 4577.81–9656.55) to 15,622.29 (95% UI: 10,546.44–22,296.65), and deaths increased from 6812.93 (95% UI: 5305.00–8473.04) to 15,443.42 (95% UI: 11,746.25–18,388.02). Although the age‐standardized mortality, prevalence, and disability‐adjusted life year (DALY) rates declined, the age‐standardized incidence rate showed a slight increase (AAPC = 0.07%, 95% CI: 0.05–0.09). Marked geographic and SDI‐related disparities were observed, with population growth and aging identified as the predominant contributors to the rising absolute burden. BAPC modeling projected continued increases in absolute PAH cases through 2050 despite declining age‐standardized mortality and DALY rates. In addition, MR analyses provided suggestive evidence for a positive association between genetically predicted higher BMI and increased PAH risk (inverse‐variance weighted odds ratio [IVW OR] = 1.84, 95% CI: 1.11–3.05, p = 0.018). Among older adults, the global absolute burden of PAH has increased substantially over the past three decades and is expected to continue rising, largely driven by population aging and growth. Genetic evidence further suggests that elevated BMI may represent a potentially modifiable risk factor for PAH. These findings highlight the need for enhanced surveillance, improved diagnostic strategies, equitable healthcare access, and targeted metabolic risk management to address the growing burden of PAH in aging populations.

1. Introduction

Pulmonary arterial hypertension (PAH) is a progressive and fatal pulmonary vascular disease characterized by persistently elevated pulmonary vascular resistance, leading to right ventricular overload and heart failure, with a profound impact on patients’ quality of life and survival [1]. According to the World Health Organization classification, PAH belongs to Group 1 pulmonary hypertension. The diagnostic criteria include a mean pulmonary arterial pressure > 20 mmHg, pulmonary vascular resistance > 2 Wood units, and pulmonary arterial wedge pressure < 15 mmHg [2, 3]. PAH has heterogeneous etiologies, encompassing idiopathic and heritable forms, as well as cases associated with connective tissue disease, HIV infection, and exposure to certain drugs or toxins. Against the backdrop of rapid global population aging, older adults (≥ 60 years) have become the principal group affected by PAH incidence and mortality, with increasingly prominent epidemiological trends [4, 5].

The Global Burden of Disease (GBD) study first listed PAH as an independent disease entity in 2021, providing a standardized foundation for long‐term, cross‐regional epidemiological analyses and reporting key indicators‐such as prevalence, incidence, and disability‐adjusted life years (DALYs)‐for 204 countries and territories. The burden of heart failure is attributed to 27 most‐detailed underlying causes, among which PAH is included [6]. However, substantial knowledge gaps remain regarding the global distribution of PAH in older adults, its long‐term temporal trends, and the socio‐economic drivers underlying these patterns, as well as regarding heart failure attributable to PAH [7].

Beyond quantifying disease burden, clarifying causal relationships for modifiable risk factors is equally critical for disease prevention and control. Body mass index (BMI), a key risk factor for metabolic diseases, has been linked to PAH in some observational studies; however, residual confounding limits causal inference, and the directionality of the association remains uncertain [8]. In older adults, obesity, metabolic dysregulation, and cardiovascular aging may further magnify the potential link between BMI and PAH. Mendelian randomization (MR), which leverages genetic variants as instruments to emulate random allocation, helps mitigate confounding and reverse causation, offering a robust approach to interrogate the causal relationship between BMI and PAH [9].

Accordingly, this study integrates GBD 2021 data with a two‐sample MR framework to systematically analyze the spatiotemporal trends in prevalence, incidence, mortality, and DALYs of PAH and heart failure attributable to PAH (PAH‐HF) among individuals aged ≥ 60 years worldwide from 1990 to 2021, and to project disease burden through 2050. In parallel, we assess the potential causal effect of BMI on PAH risk in older adults. Our findings aim to inform targeted prevention strategies and optimize health resource allocation for the global older population affected by PAH.

2. Methods

2.1. Data Sources and Case Definition

This study was based on the GBD 2021 database (URL: http://ghdx.healthdata.org/gbd-2021) [10]. For adults aged ≥ 60 years, we extracted key epidemiological indicators of PAH from 1990 to 2021, including prevalence, incidence, deaths, DALYs, and the corresponding age‐standardized rates (ASRs). The data covered 204 countries and territories, 21 GBD regions, and five socio‐demographic index (SDI) strata [11]. Specifically, the study population was restricted to individuals aged ≥ 60 years, and data were retrieved according to the GBD standard age‐grouping system for the following eight age bands: 60–64, 65–69, 70–74, 75–79, 80–84, 85–89, 90–94, and ≥ 95 years. All indicators were extracted separately by age group using the GBD Results Tool and then aggregated to characterize the overall disease burden among older adults worldwide. This approach was consistent with the GBD age‐stratification framework, thereby ensuring comparability across locations, time periods, and age groups.

PAH cases were defined according to the GBD 2021 cause hierarchy and ICD‐based disease mapping, including International Classification of Diseases, 10th Revision (ICD‐10) codes I27.0, I27.2, and related subcategories [12]. Diagnostic information in the GBD framework integrates available clinical, administrative, and epidemiological data sources and is harmonized through standardized disease modeling procedures. In addition to the overall PAH burden, we also examined the burden of heart failure attributable to PAH, hereafter denoted PAH‐HF. PAH is one of the 27 most‐detailed underlying causes to which the burden of heart failure is attributed in the GBD 2021 causal attribution analysis, a set of causes identified through literature review, quantitative analysis of death certificates and hospital records from 93 country‐years, and expert opinion. This construct, therefore, represents heart failure occurring as a consequence of PAH, which clinically corresponds predominantly to right ventricular failure complicating PAH. Because GBD does not code the anatomical side of heart failure, the construct does not constitute an anatomically or clinically ascertained category. It also does not correspond to WHO Group 2 pulmonary hypertension due to left heart disease, because the underlying cause classification remains PAH. The GBD case definition for heart failure is based on clinical diagnosis using structured criteria such as the Framingham or European Society of Cardiology criteria; since 2016, GBD has used Stage C or above as defined in the Universal Definition of Heart Failure, thereby capturing both currently symptomatic patients and those diagnosed with heart failure who are currently asymptomatic. In the extracted data set (the GBD heart‐failure attribution dimension, combined with PAH as the underlying cause), only prevalence and YLD estimates were available for this construct; incidence, mortality, and DALY estimates were not available.

All epidemiological indicators were age‐standardized to the GBD world standard population to minimize the influence of differences in population age structure across countries and regions [13]. Estimates were reported with 95% uncertainty intervals (UIs), where available, to reflect uncertainty from data sources and model‐based estimation. All data were obtained from publicly available aggregated sources and contained no individual‐level information; therefore, ethics approval and informed consent were not required.

2.2. Disease Estimation and Trend Projection Models

GBD 2021 disease estimation relies on the Bayesian meta‐regression platform DisMod‐MR 2.1, which synthesizes multiple data sources including vital registration, cause‐of‐death modeling, hospital and insurance databases, and systematic reviews [14]. This modeling framework embeds a Bayesian age–period–cohort (BAPC) structure to jointly estimate age, period, and birth cohort effects. For each indicator, 1000 posterior draws were generated, and the 2.5th and 97.5th percentiles were used to report 95% UIs, reflecting estimation uncertainty [15]. We used the GBD‐reported median estimates and corresponding UIs. Based on ASRs for 1990–2021, we further conducted Bayesian projections for 2022–2050 using a BAPC model. Model settings included a second‐order random walk (RW2) prior for smoothing period effects and a linear drift for cohort effects. Posterior predictive distributions were obtained via 10,000 Markov chain Monte Carlo (MCMC) draws, and we report medians and 95% prediction intervals (PIs). Convergence was assessed using the Gelman–Rubin statistic (R̂ < 1.05).

2.3. Trend and Inequality Analyses

Using calendar year as the time unit, we fitted log‐linear regression models to ASRs among adults aged ≥ 60 years to estimate average annual percent change (AAPC) and 95% confidence intervals (CIs). Trends were deemed statistically significant when the 95% CI for AAPC excluded 0. To control the false‐positive rate under multiple comparisons, Monte Carlo permutation tests (n = 4499) were used to identify joinpoints in temporal trends [16].

Following the standard GBD 2021 decomposition approach, we used the Das Gupta method to decompose changes in PAH burden among adults aged ≥ 60 years from 1990 to 2021. Changes in incidence, deaths, prevalence, and DALYs were partitioned into three demographic and epidemiological components: (1) population growth; (2) population aging; and (3) epidemiological change, defined as changes in age‐specific rates independent of demographic shifts. Contributions of each component were presented as absolute changes and relative percentages to assess the extent to which changes in PAH burden were attributable to demographic expansion, population aging, or changes in underlying epidemiological risk.

To evaluate socio‐economic disparities in PAH burden among older adults, we used SDI as the stratification variable [17]. The slope index of inequality (SII) was calculated for age‐standardized PAH indicators in 1990 and 2021 [18]. The SII was derived from weighted least squares regression, using the population distribution across SDI levels as weights, and reflects the absolute difference in disease burden between the lowest and highest positions on the socio‐economic scale. Negative SII values indicate a disproportionate burden among lower‐SDI populations, whereas positive values indicate a greater burden among higher‐SDI populations. Ninety‐five percent CIs were obtained using 1000 bootstrap resamples.

In addition, concentration index (CII) analyses were conducted to quantify relative SDI‐related inequality in PAH burden. Concentration curves were plotted by ranking countries and territories according to SDI and comparing the cumulative distribution of disease burden with the cumulative distribution of the population. A negative CII indicates that the burden is concentrated in lower‐SDI settings, whereas a positive CII indicates concentration in higher‐SDI settings. Frontier analysis was further used to compare observed age‐standardized PAH burden with the theoretically attainable burden at a given SDI level, thereby identifying locations with larger‐than‐expected disease burden relative to their socio‐demographic development.

2.4. MR Analysis

To test the potential causal effect of BMI on PAH, we performed a two‐sample MR analysis. Genetic instruments for the exposure, BMI, were drawn from a genome‐wide association study (GWAS), selecting single‐nucleotide polymorphisms (SNPs) associated with BMI at genome‐wide significance (P < 5 × 10⁻⁸; GWAS ID: ukb‐b‐19953) in 461,460 individuals of European ancestry [19]. Summary statistics for the outcome, PAH, were obtained from the FinnGen consortium (GWAS ID: I9_HYPTENSPUL), comprising 301 PAH cases and 345,634 controls [20]. SNPs were retained as instrumental variables if they showed a strong association with BMI, had an F‐statistic > 10, and were mutually independent after linkage disequilibrium clumping using r ² < 0.001 within a 10 Mb window. Exposure and outcome data sets were harmonized to ensure that the effect alleles were aligned consistently. Palindromic variants with ambiguous allele frequencies were removed when allele orientation could not be reliably inferred. SNPs were also screened to reduce the likelihood of weak‐instrument bias and potential horizontal pleiotropy.

The primary analysis used the inverse‐variance weighted (IVW) method under the assumption that all genetic instruments were valid or that any horizontal pleiotropy was balanced. MR‐Egger regression and the weighted median method were applied as complementary sensitivity analyses because they provide more robust estimates under different assumptions regarding invalid instruments. Simple mode and weighted mode methods were also used as additional sensitivity estimators. Results are reported as odds ratios (ORs) with 95% CIs, representing the change in PAH risk per 1‐standard deviation increase in genetically predicted BMI.

To assess the robustness of MR findings, heterogeneity among SNP‐specific causal estimates was evaluated using Cochran's Q statistic. Directional horizontal pleiotropy was assessed using the MR‐Egger intercept test. Leave‐one‐out analysis was performed to examine whether the overall MR estimate was driven by any single SNP. Scatter plots, forest plots, and funnel plots were generated to visualize the consistency, heterogeneity, and potential asymmetry of SNP‐level estimates. MR findings were interpreted cautiously when the IVW estimate was statistically significant, but sensitivity estimators showed wide CIs or did not reach statistical significance.

2.5. Statistical Software

All analyses were conducted in R (version 4.2.1). MR analyses were performed primarily with the “TwoSampleMR” (MR‐Base platform) and “ieugwasr” packages; trend analyses employed Joinpoint software (version 5.0.2, National Cancer Institute, USA). Statistical significance was set at p < 0.05 [21]. All data and code were based on public resources, and the analytic workflow is reproducible.

3. Results

3.1. Global Trends and Regional Differences in PAH Among Adults Aged ≥ 60 Years (1990–2021)

From 1990 to 2021, the absolute numbers of incident cases and deaths due to PAH among adults aged ≥ 60 years increased globally. Incident cases rose from 6771.01 (95% UI: 4577.81–9656.55) to 15,622.29 (95% UI: 10,546.44–22,296.65), while deaths increased from 6812.93 (95% UI: 5305.00–8473.04) to 15,443.42 (95% UI: 11,746.25–18,388.02) (Supporting Information S1: Table S1). These corresponded to approximate increases of 130.7% and 126.7%, respectively, indicating substantial expansion in absolute burden among older adults.

The fastest‐rising regions for ASIR between 1990 and 2021 were Central Europe (AAPC = 0.72%, 95% CI: 0.64–0.80), Eastern Europe (0.62%, 0.51–0.72), and High‐income North America (0.52%, 0.50–0.54), followed by High‐income Asia Pacific and Australasia (both 0.27%), Oceania (0.24%), and Central Asia and Central Sub‐Saharan Africa (both 0.22%). The largest declines in ASIR occurred in Western Sub‐Saharan Africa (−0.94%, −1.00 to −0.88), Central Latin America (−0.56%, −0.66 to −0.47), and Western Europe (−0.45%, −0.54 to −0.37). Trends in ASMR were more favorable overall: 17 of the 21 GBD regions showed a declining ASMR, most steeply in Eastern Europe (−2.73%, −3.47 to −1.99), the Caribbean (−2.66%, −2.91 to −2.41), and Australasia (−1.73%, −1.98 to −1.49), whereas ASMR rose in High‐income Asia Pacific (0.77%, 0.58–0.96), Central Asia (0.45%, 0.21–0.69), and High‐income North America (0.12%, 0.03–0.20). Several high‐income and central and eastern European populations therefore experienced rising ASIR even as their ASMR fell, whereas the reverse pattern was observed in parts of sub‐Saharan Africa and Latin America.

Geographic patterns showed marked heterogeneity in PAH burden across countries and regions (Figure 1). Moldova showed the lowest reported incidence and mortality rates in the displayed country‐level estimates, whereas Mongolia showed the highest incidence and DALY rates. Several high‐income settings, including Switzerland, Sweden, the Netherlands, Belgium, and Cyprus, showed comparatively high age‐standardized prevalence. In parts of sub‐Saharan Africa, including Zambia, Ethiopia, and Uganda, incidence rates among older adults were higher than in many high‐income countries. These differences should be interpreted as reflecting both underlying epidemiology and variation in diagnostic access, case ascertainment, and health‐system capacity.

Figure 1.

Global disease burden of PAH among adults aged ≥ 60 years in 2021. Geographic heat maps show age‐standardized rates (ASRs) of incidence (A), mortality (C), prevalence (E), and DALYs (G). Bar charts show global and regional AAPCs for incidence (B), mortality (D), prevalence (F), and DALYs (H) from 1990 to 2021.

AAPC estimates indicated heterogeneous temporal trends in ASRs (Figure 1B,D,F,H). Age‐standardized incidence increased slightly at the global level, whereas age‐standardized mortality, prevalence, and DALY rates decreased. Regional AAPCs varied substantially, with some regions showing upward trends and others showing declines, emphasizing that absolute burden and age‐standardized rate trends should be interpreted separately.

3.2. Variations by SDI and Temporal Trends

Across SDI strata, PAH burden among adults aged ≥ 60 years differed by indicator and region. Global mortality rose during the early study period, peaked around 2010, and subsequently declined to 1.51 per 100,000 in 2021. The DALY rate increased from 27.36 to 29.13 per 100,000 between 1990 and 2010 and then declined to 23.54 per 100,000 by 2021. East Asia showed the highest mortality in 1990 and a subsequent decline, whereas several sub‐Saharan African regions showed lower absolute rates but greater variability over time.

Spearman's correlation analyses between SDI and PAH burden showed weak and statistically nonsignificant associations for the indicators explicitly reported in the text. The correlation between SDI and mortality was small (r = 0.0617, p = 0.102), and the correlation between SDI and DALYs was also weak (r = –0.0150, p = 0.6913) (Figure 2). These results indicate that SDI alone did not fully explain cross‐country variation in PAH burden among older adults.

Figure 2.

Global burden of PAH in different Socio‐demographic index (SDI) regions. (A–D) Correlation between SDI and rates of incidence (A), deaths (B), prevalence (C), and DALYs (D) among 21 regions defined by GBD 2021. (E–H) Correlation between SDI and rates of incidence (E), deaths (F), prevalence (G), and DALYs (H) among 204 countries. Spearman's correlation analysis was applied to measure the r indices and p values for the association of ASR with SDI.

In 2021, the distribution of PAH burden showed a clear age gradient within the population aged ≥ 60 years (Figure 3). Incidence, mortality, prevalence, and DALYs were concentrated in older age strata, with the highest burdens generally observed among individuals aged ≥ 70 years. Low‐SDI regions showed higher incidence and prevalence in older age groups, whereas high‐SDI regions contributed a larger share of deaths and DALYs in the oldest groups.

Figure 3.

Temporal and spatial distribution of global burden of PAH among different age groups. (A−D) Composition ratio of incidence (A), death (B), prevalence (C), and DALYs (D) among different age groups globally and regionally in 2021. (E−H) Trends of incidence (E), death (F), prevalence (G), and DALYs (H) of PAH by age groups from 1990 to 2021.

Joinpoint analyses revealed a slight global increase in age‐standardized incidence from 1990 to 2021 (AAPC = 0.07%, 95% CI: 0.05–0.09; p < 0.001). In contrast, age‐standardized mortality decreased (AAPC = −0.26%, 95% CI: −0.36 to −0.16; p < 0.001), DALY rates declined (AAPC = −0.49%, 95% CI: −0.58 to −0.41; p < 0.001), and prevalence decreased slightly (AAPC = −0.08%, 95% CI: −0.10 to −0.06; p < 0.001) (Figure 4A–D).

Figure 4.

Join‐point regression model analysis and age‐period‐cohort analysis to estimate the temporal trend of global burdens of PAH. (A−D) Line plot of average annual percentage change (AAPC) for incidence (A), mortality (B), prevalence (C), and DALYs (D) of PAH from 1990 to 2021. (E−H) Age‐period‐cohort (APC) analysis of PAH in incidence (E), mortality (F), prevalence (G), and DALYs (H). Net and local drifts greater than 0 indicate that the burden of PAH is rising (top left); the impact of age effects on the burden of PAH (top right); the rate ratio of period and cohort effects greater than 1 indicates an increased burden of PAH (bottom panel).

Age‐period‐cohort analyses further indicated increasing age effects for mortality and DALYs among adults aged ≥ 60 years, with the highest burden in the oldest age groups (Figure 4E–H). The net drift for mortality was 0.05% per year (95% CI: −0.05% to 0.16%), suggesting no clear global increase in age‐adjusted mortality over the full period. The cohort effect indicated higher estimated mortality risk in more recent birth cohorts than in the earliest reference cohort, although these results should be interpreted in the context of model assumptions and long‐term changes in case ascertainment.

3.3. Drivers of the Global PAH Burden and Regional Inequalities

From 1990 to 2021, increases in the absolute numbers of PAH incident cases, deaths, prevalent cases, and DALYs were largely attributable to demographic changes (Figure 5). Population growth was the dominant contributor to increased incidence and deaths, accounting for 96.4% and 98.3% of the increases, respectively. Population aging contributed a smaller share in the global decomposition, whereas epidemiological change had a limited net contribution. For prevalence and DALYs, population growth also accounted for most of the absolute increase and was partially offset by favorable epidemiological change in some regions.

Figure 5.

Contributing factors for the disease burden of PAH. (A−D) Decomposition analysis with incidence (A), death (B), prevalence (C), and DALYs (D) of PAH.

Inequality analyses showed persistent SDI‐related differences in PAH burden (Figure 6A–D). For incidence, the SII changed from −1.18 (95% CI: −1.33 to −1.03) in 1990 to –0.81 (95% CI: −0.94 to −0.68) in 2021, indicating attenuated butpersistent absolute incidence inequality. The CII for incidence changed from –0.13 (95% CI: –0.15 to –0.11) in 1990 to –0.10 (95% CI: –0.12 to –0.08) in 2021, indicating that incidence remained somewhat concentrated in lower‐SDI populations. Mortality inequality patterns were less clearly shifted, as shown by the SII and CII panels for age‐standardized mortality.

Figure 6.

Regional disparities for the disease burden of PAH. (A–D) Worldwide health inequality regression curves and concentration curves for the DALYs of PAH in 1990 and 2021, respectively. The relationship between SDI and ASIR (A) and ASMR (B) is depicted by the slope index of inequality. The relative inequalities depicted by the concentration index of ASIR (C) and ASMR (D). Blue represents data in 1990, and red represents data in 2021. (E, F) Frontier analysis exploring the relationship between SDI and ASIR (E) or ASMR (F) for PAH in 204 countries and territories.

Frontier analysis demonstrated that several countries had higher PAH incidence or mortality than expected for their SDI level (Figure 6E,F). Mongolia showed a relatively large gap between observed mortality and the modeled frontier, whereas the United States and Sweden showed lower mortality rates but still remained above the estimated attainable frontier. These findings suggest that both socioeconomic development and health‐system efficiency contribute to cross‐national differences in PAH burden.

3.4. Future Disease Burden Projections (1990–2050)

BAPC projections indicated that absolute numbers of incident cases, deaths, prevalent cases, and DALYs are expected to continue increasing through 2050 (Figure 7, left panels). By contrast, projected ASRs showed different trajectories across indicators (Figure 7, right panels): incidence increased only modestly, prevalence remained broadly stable, and age‐standardized death and DALY rates declined. Thus, the projected increase in future PAH burden is driven mainly by demographic expansion rather than a uniform increase in age‐standardized risk.

Figure 7.

Prediction of the disease burden trend of PAH from 1990 to 2050. Prediction of PAH in all age cases (left) and ASR (right) of incidence (A), death (B), prevalence (C), and DALYs (D) till 2050 by using the Bayesian Age‐Period‐Cohort (BAPC) model.

3.5. Epidemiological Characteristics of Heart Failure Attributable to PAH (PAH‐HF)

Globally, ASPR and ASYR were relatively stable with modest fluctuations over time. High‐SDI strata consistently showed the highest ASPR and ASYR in the displayed data, whereas low‐middle SDI strata showed the lowest rates. Middle‐SDI strata increased during the early and middle study period and then stabilized or declined slightly toward 2021.

Sex‐specific panels showed higher ASPR and ASYR among women than men across most years and SDI strata. This pattern differs from the sex pattern described for some general PAH indicators and therefore should be interpreted as specific to the PAH‐HF metrics shown in Supporting Information S1: Figure S3. The figure does not display incidence; therefore, incidence‐specific statements for PAH‐HF should not be inferred from these panels.

Overall, the PAH‐HF analysis suggests persistent SDI‐and sex‐related heterogeneity in prevalence‐ and YLD‐related rates among older adults. Because the displayed indicators are ASPR and ASYR rather than incidence or mortality, these findings primarily describe non‐fatal burden and prevalent disease patterns rather than new case occurrence.

3.6. MR of the Association Between BMI and PAH Risk

MR analyses provided genetic evidence compatible with a positive association between higher BMI and PAH risk. In the primary IVW analysis, each 1‐standard deviation increase in genetically predicted BMI (approximately 4.7 kg/m²) was associated with higher PAH risk (OR = 1.84, 95% CI: 1.11–3.05; p = 0.018). MR‐Egger (OR = 1.64, 95% CI: 0.43–6.22; p = 0.466) and weighted median (OR = 1.34, 95% CI: 0.61–2.93; p = 0.464) estimates were in the same direction but were statistically nonsignificant. Simple mode and weighted mode estimates were imprecise, with wide CIs. Therefore, the MR results should be interpreted as suggestive evidence rather than definitive proof of causality, particularly given the limited number of PAH cases in the outcome GWAS (Figure 8 and Supporting Information S1: Figures S1–S2).

Figure 8.

Mendeli randomization (MR) results for pulmonary arterial hypertension (PAH) and body mass index (BMI).

4. Discussion

4.1. Main Findings

Using GBD 2021 estimates, this study characterized the global burden, temporal trends, demographic drivers, socioeconomic inequality, and future projections of PAH among adults aged ≥ 60 years, and evaluated genetic evidence linking BMI with PAH risk. The principal finding is that absolute PAH burden increased substantially from 1990 to 2021, whereas age‐standardized trends were indicator‐specific: incidence increased slightly, while mortality, prevalence, and DALY rates declined. Decomposition and projection analyses indicate that population growth and population aging, rather than a uniform increase in age‐standardized risk, are the main drivers of rising absolute burden. The MR analysis supported a positive IVW association between genetically predicted BMI and PAH risk, but the sensitivity estimates were imprecise; therefore, the BMI‐PAH finding should be interpreted as suggestive genetic evidence requiring further validation.

4.2. Global and Regional Trend Differences

Marked geographic and socio‐economic heterogeneity in PAH among adults aged ≥ 60 years was observed worldwide [22]. Moldova showed the lowest burden (incidence and mortality 0.03 per 100,000; DALYs 1.18 per 100,000), potentially reflecting stronger primary care and public‐health measures. In contrast, Mongolia had a much higher incidence (8.34 per 100,000) and DALYs (130.41 per 100,000), suggesting the combined effects of high altitude, chronic hypoxia, and constrained healthcare resources [7]. High‐income countries such as Switzerland and Sweden exhibited higher age‐standardized prevalence (14–22 per 100,000), largely driven by aging populations and higher diagnostic yield. The heaviest burdens in older adults were seen in sub‐Saharan Africa, Southeast Asia, and South Asia; in Zambia, Ethiopia, and Uganda, incidence commonly exceeded 2.5 per 100,000, reflecting the combined influences of altitude, infectious diseases, and limited access to care [23]. By contrast, East Asia showed declining mortality among older adults. These regional contrasts underscore the interplay of economic development, healthcare capacity, and environmental exposures. In low‐resource settings, priorities should include improving diagnostic access, establishing regional PAH surveillance, and strengthening echocardiography and right‐heart catheterization at the primary‐care level to reduce underdiagnosis [24].

4.3. Socio‐Economic Inequality and Health‐System Efficiency

Socio‐economic status is a major determinant of the distribution of PAH burden in older adults. From 1990 to 2021, incidence, mortality, and DALYs remained higher‐and improved more slowly—in low‐SDI regions than in high‐SDI regions [25]. Although mortality in older adults declined globally after 2010, the burden in low‐income countries remained substantial, reflecting limited diagnostic and therapeutic access, diagnostic delay, and suboptimal treatment adherence [4]. In high‐income settings, wider availability of targeted therapies and multidisciplinary care has reduced mortality, yet intensifying population aging continues to push incidence and prevalence upward. Quantitative inequality metrics corroborated these patterns: the SII for incidence changed from −1.18 in 1990 to –0.81 in 2021, indicating attenuated but persistent inequality; however, the SII for DALYs changed little (–5.27 to –5.35), indicating a persistently higher burden among older adults in low‐income regions [26]. Frontier analysis showed Mongolia's mortality among older adults far above its theoretical optimum (efficiency gap = 8.30), whereas Somalia—despite a zero efficiency gap—has limited room for improvement given its low SDI. The United States and Sweden, while high‐income, still exhibited scope for optimization, possibly related to gaps in early diagnosis or uneven care pathways. These findings indicate that imbalances in health‐system efficiency—rather than economic status alone—are a key bottleneck in global PAH control among older adults. Future strategies should focus on basic health‐system capacity in low‐SDI countries and resource equity in high‐SDI countries to promote coordinated, multi‐tier system improvements.

4.4. Age, Demography, and Disease Drivers

Population growth and aging were the central forces behind the rising PAH burden in those aged ≥ 60 years. Between 1990 and 2021, population growth accounted for 96.41%, 98.27%, 102.22%, and 117.25% of the increases in incidence, mortality, prevalence, and DALYs, respectively, with aging playing a more prominent role in high‐income regions [27, 28]. As global aging accelerates, PAH prevalence and mortality continue to increase in older adults, making this group the chief contributor to global PAH trends. Some regions achieved burden reductions through epidemiological improvements—for example, declines in mortality and DALYs in Eastern Europe following advances in public‐health management and treatment access [29]. Sex differences also merit attention: prevalence and mortality in older women generally exceeded those in men, potentially related to estrogen signaling, immune differences, and lifestyle factors [30]. Accordingly, prevention efforts should prioritize older women with early cardiopulmonary assessment and broader access to targeted therapies to reduce morbidity and mortality. Based on projections to 2050, the number of older adults with PAH could exceed 140,000 in the absence of effective interventions. Without substantial improvements in foundational geriatric care, low‐SDI countries are likely to see continued increases, posing sustained pressure on public‐health systems.

4.5. Risk Factors and the BMI‐PAH Causal Link

Our MR analysis further evaluated the potential causal association between BMI and PAH risk in adults aged ≥ 60 years. The IVW estimate indicated that each 1SD (≈4.7 kg/m²) increase in BMI was associated with an 84% higher risk of PAH (OR = 1.84, 95% CI: 1.11–3.05; p = 0.018), suggesting obesity as an important, potentially modifiable risk factor in older adults. Several mechanistic pathways render a BMI‐PAH link biologically plausible. First, expansion and dysfunction of visceral and epicardial adipose tissue may increase the production of pro‐inflammatory mediators, including TNF‐α and IL‐6 [31, 32, 33]. IL‐6 is both necessary and sufficient for experimental pulmonary vascular remodeling: IL‐6‐deficient mice are protected from hypoxia‐induced pulmonary hypertension and lung inflammation, whereas lung‐specific IL‐6 overexpression produces occlusive neointimal lesions and severe pulmonary hypertension, and TNF‐α promotes an apoptosis‐resistant phenotype of pulmonary artery smooth muscle cells (PASMCs) at least partly by inhibiting pyruvate dehydrogenase [31, 32]. Second, adipokine signaling is directly implicated in pulmonary vascular remodeling: leptin is overexpressed in pulmonary endothelial cells from patients with PAH and drives PASMC proliferation and perivascular inflammation through the ObR‐b receptor, whereas adiponectin deficiency causes pulmonary vascular inflammation and elevated pulmonary artery pressure in mice, such that the obesity‐associated shift toward higher leptin and lower adiponectin favors pulmonary vascular remodeling [34, 35, 36]. Third, obesity‐related sleep‐disordered breathing and intermittent hypoxia superimpose sympathetic activation, oxidative stress, and systemic inflammation, and obstructive sleep apnea is a recognized and frequently underdiagnosed cause of pulmonary hypertension [37]. Fourth, and particularly relevant to older adults, obesity is the dominant driver of the HFpEF phenotype: obese patients with HFpEF show greater plasma volume expansion, pericardial constraint, and exercise‐induced increases in pulmonary artery pressure, and approximately one third of HFpEF patients develop combined post‐ and pre‐capillary pulmonary hypertension, which carries a worse prognosis and overlaps phenotypically with PAH [38, 39, 40]. Insulin resistance and dyslipidemia may also enhance vasoconstrictive reactivity via disruption of NO‐PGI₂ signaling. Taken together, these mechanisms provide a coherent biological basis for the positive MR estimate, although they should be regarded as complementary evidence rather than as independent confirmation. Because obesity is also the dominant driver of the HFpEF phenotype, heart failure attributable to PAH, particularly HFpEF‐like phenotypes, may further amplify the BMI‐PAH association; however, our outcome GWAS captured PAH rather than PAH‐HF specifically, so these mechanistic considerations should be regarded as complementary rather than as confirmation of a distinct PAH‐HF entity. Although the exposure and outcome GWASs were derived from independently recruited UK Biobank and FinnGen cohorts, and no sample overlap was expected based on their documented provenance, individual‐level identifiers were unavailable for direct verification. In addition, the small number of PAH cases in the outcome GWAS (n = 301) limited statistical power and precision. Future research should apply stratified and multivariable MR in larger and age‐specific PAH data sets, together with multi‐cohort validation, to determine whether the effect of BMI on PAH is independent of heart failure attributable to PAH. From a policy perspective, our findings support obesity prevention as an entry point for early intervention in PAH among older adults. Scaling metabolic‐syndrome screening and lifestyle/weight‐management programs could mitigate pulmonary vascular risk in high‐risk elders.

4.6. Implications and Outlook

The persistent rise of PAH in adults aged ≥ 60 years has become a major challenge for global health equity. Low‐SDI countries face the dual constraints of limited diagnostic capacity and scarce resources, whereas high‐SDI countries must balance population aging with resource equity. The causal signal for BMI strengthens the rationale for obesity control in older adults. In anticipation of a potential “140‐thousand‐patient era” by 2050, global public‐health strategies should advance three fronts: (i) early screening and risk stratification, including cardiopulmonary testing and right‐heart catheterization in high‐risk elders; (ii) access to targeted therapies and comprehensive care, with international support (e.g., WHO and PAH networks) to improve drug availability in low‐income countries; and (iii) global actions on obesity and metabolism, integrating weight management into geriatric chronic‐disease programs via cross‐country collaboration. A tri‐pillar strategy centered on early detection, therapeutic access, and obesity control may effectively curb this high‐fatality vascular disease over the coming decades. Our findings are limited by the quality and completeness of the underlying GBD estimates and GWAS summary statistics.

5. Conclusion

This study shows that the absolute global burden of PAH among adults aged ≥ 60 years increased from 1990 to 2021 and is projected to continue rising through 2050. This increase is mainly attributable to demographic expansion and population aging, while age‐standardized mortality and DALY rates have declined. The burden varies substantially across regions, SDI strata, and age groups, and the prevalence‐ and YLD‐based metrics for PAH‐HF. MR results provide suggestive genetic evidence that higher BMI may increase PAH risk. Public‐health strategies should prioritize improved PAH surveillance, diagnostic capacity, equitable access to targeted therapies, geriatric cardiopulmonary care, and metabolic‐risk prevention.

Author Contributions

Y.F. and B.W. wrote the main manuscript text. F.L. and X.T. prepared figures. Y.G. and J.G. guided writing and modified figures. Y.G., Y.F., and J.G. integrated design and provided financial support. Y.F. is the guarantor of this work. All authors reviewed the manuscript.

Ethics Statement

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File

Fu Y., Wang B., Li F., Tian X., Gu J., and Gao Y., “Global Burden of Pulmonary Arterial Hypertension in Older Adults and Its Causal Association With Body Mass Index,” Pulmonary Circulation 61 (2026): e70416. 10.1002/pul2.70416.

Data Availability Statement

The GBD 2021 data used in this study are publicly available from the Global Health Data Exchange (http://ghdx.healthdata.org/gbd-2021).