This is a retrospective, observational, analytical case–control study conducted with patients diagnosed with either established or early SSc who attended the outpatient rheumatology clinic at the Hospital Universitario Nacional (HUN), Bogotá, Colombia.

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and complied with all relevant national and international guidelines for research involving human participants. The research ensured that participants’ privacy and confidentiality were maintained throughout the study. Prior to the initiation of the study, ethical approval was obtained from the appropriate ethics committee. From January 2019 to December 2023, the medical records of patients attending the rheumatology clinic at HUN were reviewed for eligibility. Patients were eligible if they met the 2013 ACR/EULAR criteria for SSc or the 2011 VEDOSS criteria for early disease. Additionally, they must have had at least one echocardiogram report estimating the likelihood of PH and information regarding their residence during the preceding five years. Patients with polyautoimmunity (excluding Sjögren's syndrome), those with other non-SSc causes for PH, and those classified under PH Group 5 were excluded. Patients without a complete echocardiogram report or unclear residency data in the last five years were also excluded. A non-probabilistic convenience sampling method was used. A minimum sample size of 114 cases and 228 controls was estimated, considering an HA exposure probability of 0.6 and an estimated minimum risk with an odds ratio (OR) of 2.

Patients were considered to reside at HA if their places of residence in the last five years were all at an altitude>2500m.a.s.l. Those residing at altitudes<2500m were considered non-HA residents.

The included patients needed to have a complete echocardiogram report by a cardiology specialist. The echocardiograms must have been performed in geographic locations that matched the patient's altitude of residence.

Patients were considered to have a high probability of PH if the pulmonary artery systolic pressure (PASP) was>39mmHg, if the peak tricuspid regurgitation velocity (PTRV) was>3.4 m/s, or if the PASP was between 33–39mmHg with two additional echocardiographic signs of PH.3

Patients were classified as having high probability of group 2 PH if, in addition to findings indicating a high probability of PH, clinical or echocardiographic findings of heart disease were present, such as systolic or diastolic heart failure, defined as a left ventricular ejection fraction (LVEF)<50%, segmental wall motion abnormalities, or type 2 or type 3 diastolic dysfunction. Additionally, they were considered to have a high probability of group 2 PH if they had a history of coronary artery disease or atrial fibrillation. Patients were classified as high probability of group 3 PH if, in addition to findings indicating a high probability of PH, there were signs of lung disease, such as interstitial lung disease (ILD) on chest CT, pulmonary function impairment with FVC<80% or DLCO<70%, chronic obstructive pulmonary disease (COPD) classified as moderate to severe, moderate to severe obstructive sleep apnea (OSA) or an indication for the use of positive pressure devices. As part of the analysis, ILD was considered severe if the extent was>20% or if the FVC was<70%. Finally, patients were classified as having high probability of group 4 PH if they had a history of chronic thromboembolic pulmonary disease or if compatible abnormalities were detected via ventilation/perfusion scintigraphy or angiography. Patients could be classified as both groups into high probability of PH groups 2 and 3. High probability of Group 1 PH was assigned only to patients with a high probability of PH who did not meet the criteria for any other group.

Patients who met the criteria for a high probability of PH were classified as cases.

Patients who did not meet the criteria for a high probability of PH were classified as controls.

Information was collected from the medical records registered in the reference hospital's system. Data on sociodemographic variables such as age and sex were extracted, along with clinical variables such as weight, height, past medical history, and prior pharmacological treatment, among others. Additionally, information on laboratory tests, diagnostic imaging, and pulmonary function tests were gathered.

Quantitative variables are expressed as the means and standard deviations (SDs) or medians and interquartile ranges (IQRs). Categorical variables are expressed as frequencies and/or percentages. A bivariate analysis was conducted to assess differences between groups with and without high probability of PH, as well as between HA residents and non-HA residents. For categorical variables, Chi-square or Fisher's exact tests were used, as appropriate. For quantitative variables, the Mann–Whitney U test was performed.

A subgroup analysis was performed to determine the relationship between heart or lung disease and HA residence. Subsequently, HA residence and other risk factors were analyzed via a multivariate logistic regression model adjusted for both conventional risk factors for PH found in the literature and those identified in this study. ORs and 95% confidence intervals (95% CIs) were estimated. A p value of <0.05 was considered to indicate statistical significance. No data imputation was performed. Statistical analysis was performed with Stata 17® software.

Table 1 presents the sociodemographic, clinical, and laboratory characteristics of the patients included in the analysis. Additionally, a comparison between the groups with and without high probability of PH is shown. Overall, 62% (n=231) of the patients were HA residents, with a higher prevalence among those with high probability of PH (75% vs 57%, p=0.001). Patients with high probability of PH were younger (56 vs 62 years, p=0.005). The sex distribution was similar in both groups, with >90% of the patients in each group being female. There were no statistically significant differences in sex distribution.

The most common form of SSc was lcSSc (48.9%), followed by the early/sine scleroderma subtype, with diffuse cutaneous SSc (dcSSc) being the least common subtype. There was a greater number of patients with dcSSc among those with high probability of PH (p=0.005).

No differences were found in the number of years with Raynaud's phenomenon or joint involvement. In the group of patients with high probability of PH, dyspepsia and dyspnea were more common (74% vs 58% and 70%vs28%, respectively, p<0.05). No differences were found between the groups with and without high probability of PH in terms of the frequency of puffy fingers, digital ulcers, calcinosis, or telangiectasias. With respect to missing data, 25% of patients had no recorded information about the presentation of puffy fingers.

The hemoglobin concentration was greater in patients without high probability of PH (13.7 vs 13.3mg/dL, p=0.03). No differences were found between the groups regarding positivity for antinuclear antibodies (ANAs), the ANAs centromere pattern, or the presence of anti-Scl-70 antibodies. The presence of ILD and extensive ILD on chest CT was more common in patients with high probability of PH (51% vs 33%, p=0.001, and 20% vs 10%, p=0.009, respectively). Eighty percent of patients had no information regarding chest CT. There were no statistically significant differences in the frequency of patients with a FVC<80%. Thirty percent of patients had no information on spirometry. The PASP was greater in patients with high probability of PH (51 vs 25mmHg, p=0.001).

The most common pattern on nailfold capillaroscopy (NFC) was the active scleroderma pattern (36%), followed by the early scleroderma pattern (30%), and finally the late scleroderma pattern (13%). No significant differences were found between the groups with and without high probability of PH. Fifty-three percent of patients had no information regarding NFC.

Among the patients with high probability of PH, 50% (n=57) were classified into high probability of group 2 PH, and 54% (n=62) were classified into high probability of group 3 PH. A total of 46% (n=52) were classified as high probability of group 1 PH. None of the patients had high probability of group 4 PH.

Patients with high probability of PH had a higher frequency of pulmonary and cardiac comorbidities (54% vs 34% and 50% vs 39%, respectively).

Overall, 62% (n=229) of the patients were receiving some form of pharmacological treatment for SSc. There were no differences between the groups with and without high probability of PH. However, the use of cyclophosphamide was more common in the high probability of PH group (9.6% vs 8%, p=0.001).

When patients were compared based on whether they were HA residents or not (Supplementary Table 1), a greater percentage of patients with high probability of PH were found in the HA resident group (37% vs 20%, p=0.001). Additionally, HA residents had a greater percentage of dyspepsia (68% vs 55%, p=0.019), puffy fingers (68% vs 45%, p=0.001), telangiectasias (66% vs 52%, p=0.009), extensive ILD (16% vs 9%, p=0.001), and treatment with cyclophosphamide (5% vs 0%). They also had higher hemoglobin levels (13.9 vs 13.1mg/dL, p=0.001) and greater positivity for ANAs (42% vs 28%). The non-HA resident group had a greater percentage of patients with abnormal NFC patterns (89% vs 74%, p=0.01) and early scleroderma patterns (47% vs 21%), a greater percentage of patients receiving pharmacological treatment for SSc (72% vs 56%), and a greater frequency of methotrexate use (8% vs 5%). No differences were found in the frequency of cardiac or pulmonary diseases when these groups were compared.

Considering the differences between groups, a bivariate and multivariate analysis was performed to study the factors associated with high probability of PH, adjusted for variables identified in this study and previously recognized: sex, age, dcSSc, presence of anti-centromere antibodies, anti-Scl-70 antibodies, joint involvement, gastrointestinal symptoms, telangiectasias, puffy fingers, and cyclophosphamide use (Supplementary Table 2). The variables of anti-centromere antibodies and severe pulmonary disease were excluded from the multivariate logistic regression model because of the absence of significant differences in the bivariate analysis. We found that being an HA resident was associated with the presence of high probability of PH of any type, even after adjusting for other factors (OR 2.0, 95% CI: 1.16–3.5, p=0.012). Furthermore, it remained a risk factor for high probability of PH even after adjusting for pulmonary disease (OR: 2.4, 95% CI: 1.2–5.1, p=0.001), cardiac disease (OR: 2.6, 95% CI: 1.3–5.2, p=0.004), and even the combination of these two conditions (OR: 2.8, 95% CI: 1.4–5.4, p=0.003). The combination of being an HA resident and having pulmonary disease further increased the risk of high probability of PH (OR: 3.38).