What this is
- This research investigates the impact of ambient air pollution on respiratory symptoms in people with HIV (PWH) and chronic lung disease.
- It compares outcomes between PWH and people without HIV (PWoH) with similar lung conditions.
- The study finds that higher exposure to fine particulate matter (PM) correlates with worsened respiratory symptoms in PWH, particularly those with () or impaired lung function.
Essence
- Greater exposure to fine particulate matter (PM) is linked to worse respiratory symptoms in people with HIV and chronic lung disease, particularly in those with or impaired diffusion capacity.
Key takeaways
- In PWH with , a 1 µg/m³ increase in PM exposure correlates with a 4.04-point increase in the St. George's Respiratory Questionnaire (SGRQ) score, indicating worse respiratory health.
- For PWH with impaired diffusion capacity, a 1 µg/m³ increase in PM is associated with a 2.14-point increase in SGRQ score, also reflecting deteriorating respiratory symptoms.
- No significant associations were found between PM exposure and respiratory symptoms in PWoH with or impaired diffusion capacity.
Caveats
- Pollution exposure estimates based on residential address may not accurately reflect individual exposure levels.
- The sample size for subgroup analyses was small, potentially affecting the robustness of the findings.
Definitions
- Chronic obstructive pulmonary disease (COPD): A progressive lung disease characterized by airflow obstruction and breathing difficulties.
- Diffusing capacity for carbon monoxide (DLCO): A measure of how well oxygen and carbon dioxide are exchanged in the lungs, with lower values indicating impaired lung function.
Simplified
Background
Chronic lung disease is a common comorbidity in people with HIV (PWH). PWH experience a greater incidence of chronic obstructive lung disease (COPD) and a faster decline in lung function than people without HIV (PWoH), even after accounting for cigarette smoking [1–3]. Emphysema is more common in PWH who smoke tobacco than in people without HIV who smoke [4]. Impairment in diffusing capacity for carbon monoxide (DLCO), with or without airflow obstruction, is the most prevalent abnormal pulmonary function finding in PWH; low DLCO may represent early emphysema, pulmonary fibrosis, or pulmonary vascular abnormalities, and is associated with a history of respiratory infections [5, 6]. Airflow obstruction and impaired diffusing capacity are both independent predictors of mortality in PWH [7].
Proposed mechanisms for the higher incidence of lung disease in PWH include direct effects of the virus, a history of pneumonia and opportunistic infections, and injury from lung toxins due to altered immune responses [8]. HIV persists in the lungs even after viral suppression and can shift macrophages to produce pro-inflammatory cytokines and proteases, disrupt cell-cell adhesion, and induce oxidative stress [9–12]. HIV-mediated lung disease is also related to dysfunctional T-cell responses as there is an association between lower CD4 + T-cell counts and COPD [13]. Immune dysregulation from HIV may enhance the toxic effects of smoking on the lungs [14]. Alveolar macrophage production of MMP9 and MMP12, immune-modulating cytokines and proteases associated with emphysema, are higher in PWH who smoke as compared to PWoH who smoke [15, 16]. It is possible that air pollution-mediated lung injury, similar to tobacco smoking injury, is amplified in the presence of HIV. In general populations, long-term exposure to fine particulate matter (PM2.5) is associated with accelerated decline in lung function and progression of emphysema [17, 18]. In people with COPD, exposure to higher levels of ambient air pollution is associated with increased exacerbations [19, 20]. However, there are no data regarding the effects of PM2.5 and ozone exposures on respiratory disease in PWH. We sought to explore associations between ambient air pollution exposures and respiratory disease in PWH and comparable populations of PWoH at increased vulnerability for HIV.
In this cross-sectional study, we focused on respiratory symptom scores as a measure of respiratory disease activity and also evaluated associations between pollution exposures and lung function. We included people with and without lung disease but hypothesized that greater 1-year average ambient air pollution exposures would be associated with worse respiratory symptoms in people with chronic lung disease and HIV.
Methods
Study population
We included participants of the Multicenter AIDS Cohort Study (MACS) and the Women’s Interagency HIV Study (WIHS), prospective observational cohorts of U.S. men and women, respectively, with or at increased vulnerability for HIV [21, 22], merged into the MACS/WIHS Combined Cohort Study (MWCCS) in 2020 [23]. Clinical trial number: not applicable. Individuals were eligible for this analysis if they completed pulmonary function tests (PFTs) and respiratory questionnaires at one study visit in 2017–2020 and lived in a single census tract during the year prior to testing. Individuals’ residential addresses were geocoded and assigned to census tracts using ArcGIS (Esri, Redlands, CA), which were then linked to census tract-level daily average ambient fine particulate matter (PM2.5) and 8-hour maximum ozone (O3) concentrations from the U.S. Environmental Protection Agency’s Community Multiscale Air Quality Modeling System (CMAQ) [24]. To determine how long an individual had lived at a given location, due to MACS and WIHS protocol differences, we used self-reported duration of residence for men and longitudinal (annual) geocoding data for women.
Exposures and outcomes
Concentrations of PM2.5 (µg/m³) and O3 (parts per billion, ppb) were the two exposures of interest, each averaged across the year preceding the date of the PFTs. Since we had one year of confirmed location data, we selected one-year average pollution estimates to incorporate potential lag or cumulative effects over the given exposure time interval. A total of six respiratory-related outcomes were examined. Four were measures of lung function: postbronchodilator % predicted forced expiratory volume at 1 s (FEV1), % predicted forced vital capacity (FVC), FEV1/FVC ratio, and % predicted DLCO. Two were measures of symptom burden: St. George’s Respiratory Questionnaire (SGRQ) score [25] and modified Medical Research Council (mMRC) dyspnea scale score [26]. The SGRQ total score was used (0–100 range), with higher scores indicating worse respiratory health status [27]. Ordinal mMRC scores were whole numbers from 0 (“I only get breathless with strenuous exercise”) to 4 (“I am too breathless to leave the house, or I am breathless when dressing”) [28]. We also created a binary mMRC variable reflecting a low burden of symptoms (score 0 or 1) or a high burden of symptoms (score 2, 3, or 4), as higher scores predict mortality independent of lung function [29]. Spirometry measurements were made with an EasyOne Pro or Easy-on-PC spirometer (ndd Medizintechnik AG, Zurich, Switzerland); bronchodilation was by albuterol (360 µg via metered-dose inhaler). Quality assessment was according to American Thoracic Society/European Respiratory Society (ATS/ERS) standards [30], and equations derived from the Third National Health and Nutrition Examination Survey (NHANES) were used to calculate spirometry reference values [31]. DLCO measurements were made with an EasyOne Pro device and adjusted for hemoglobin and carboxyhemoglobin following the ATS/ERS standards; [32] equations from NHANES I were used to calculate DLCO reference values [32, 33]. Spirometry and DLCO reference equations were chosen to align with previously published lung function analyses in these cohorts [34, 35].
Statistical analysis
Linear regression models were used to quantify unadjusted and adjusted associations between selected increases in the two exposures (3 ppb for O3, and 1 µg/m³ for PM2.5) and the six outcomes, separately for MACS PWH, MACS PWoH, WIHS PWH, and WIHS PWoH. We also performed sensitivity analyses that pooled PWH and PWoH in each cohort for each outcome and analyzed an interaction between HIV serostatus and PM2.5 or O3 exposure. Covariates posited a priori as potentially associated with the outcomes were included in adjusted models: age (5-knot restricted cubic spline), race (Black, White, or other), Hispanic ethnicity (binary), current cannabis use (binary), current tobacco smoking (binary), smoking pack-years history (continuous), and body mass index (continuous). In models restricted to PWH, CD4 count (continuous, cells/mm3), CD4/CD8 ratio (continuous), and log-transformed HIV viral load (copies/mL), obtained within six months of the PFT study visit, were additionally included. For the binary mMRC outcome, we used logistic regression, with all other modeling details identical. Because we hypothesized that people with underlying chronic lung disease would be most susceptible to the effects of ambient pollution exposure, we next conducted two subgroup analyses, the first restricted to individuals with COPD (defined as post-bronchodilator FEV1/FVC ratio < 0.7) [36], and the other restricted to individuals with impaired diffusing capacity (DLCO < 80% predicted, as previously defined in this population) [34]. The subgroup analyses were not stratified by cohort (MACS, WIHS) due to smaller sample sizes; as such, sex (cohort) was included in regression models which were otherwise as specified above, including separate models for MACS/WIHS PWH and MACS/WIHS PWoH. Missing covariate data were minimal; a small number of values were carried forward from the prior study visit. Analyses were completed in SAS version 9.4 (SAS Institute Inc., Cary, NC).
Results
Baseline characteristics
In the MACS cohort, 338 men met inclusion criteria [Table 1]. One hundred forty-six (43%) were PWH, with well-controlled HIV (median viral load undetectable and median CD4 + T-cell count 674 cells/mm3). In the MACS, PWH were younger (59 vs. 64 years), were more likely to be people who currently smoke (15% vs. 10%), had greater smoking history (median of 16 pack-years vs. 9 pack-years), and were more likely to have COPD (12% vs. 6%), as compared to PWoH. In MACS there was a higher percentage of Blacks and Hispanics in the PWH group compared to PWoH (30% vs. 12% and 16% vs. 4%, respectively). In the WIHS cohort, 1073 women met inclusion criteria. PWH in the WIHS also had well-controlled HIV, with median viral load undetectable and CD4 + T-cell count 726 cells/mm3. PWH and PWoH in WIHS were of similar age (52 vs. 50 years, respectively) and had similar smoking history, current smoking status, and prevalence of COPD.
As compared to MACS participants, WIHS participants were younger (52 years vs. 62 years), were more often Black (66% vs. 20%), were more likely to have no more than a high school education (62% vs. 11%), had higher BMI, and were more likely to be people who currently smoke (36% vs. 12%). The annual average pollution exposures between MACS and WIHS participants were similar, with median O3 exposures 39.0 and 36.4 ppb, and median PM2.5 exposures 9.4 and 9.2 µg/m3, respectively. Median pollution exposures were also similar between PWH and PWoH in both cohorts.
Pollution levels by site are listed in Supplementary Table 1. There were generally higher levels of pollution exposures at the California sites whereas the other sites showed similar distributions.
| Cohort characteristics | MACS | WIHS | ||||
|---|---|---|---|---|---|---|
| Total | PWH | PWoH | Total | PWH | PWoH | |
| No. of participants | 338 | 146 | 192 | 1073 | 769 | 304 |
| Age in years, median (IQR) | 62 (56–68) | 59 (53–65) | 64 (58.5–70) | 52 (44–58) | 52 (45–58) | 50 (42–57) |
| Sex, No. (%) | ||||||
| Female | 1073 (100.0) | 769 (100.0) | 304 (100.0) | |||
| Male | 338 (100.0) | 146 (100.0) | 192 (100.0) | |||
| Race, No. (%) | ||||||
| Black | 66 (19.5) | 43 (29.5) | 23 (12.0) | 707 (65.9) | 502 (65.3) | 205 (67.4) |
| White | 242 (71.6) | 81 (55.5) | 161 (83.9) | 107 (10.0) | 87 (11.3) | 20 (6.6) |
| Other | 30 (8.9) | 22 (15.1) | 8 (4.2) | 259 (24.1) | 180 (23.4) | 79 (26.0) |
| Hispanic ethnicity, No. (%) | 31 (9.2) | 24 (16.4) | 7 (3.6) | 138 (12.9) | 96 (12.5) | 42 (13.8) |
| <=High school education, No. (%) | 36 (10.7) | 22 (15.1) | 14 (7.3) | 669 (62.4) | 494 (64.2) | 175 (57.8) |
| Smoking status, No. (%) | ||||||
| Never | 128 (38.4) | 56 (39.2) | 72 (37.9) | 372 (34.7) | 272 (35.4) | 100 (32.9) |
| Current | 39 (11.7) | 21 (14.7) | 18 (9.5) | 382 (35.6) | 266 (34.6) | 116 (38.2) |
| Former | 166 (49.8) | 66 (46.2) | 100 (52.6) | 319 (29.7) | 231 (30.0) | 88 (28.9) |
| Pack-years of smoking (median, IQR)* | 11.9 (0.4–31.1) | 15.6 (4.4–28.3) | 9.1 (0.1–31.5) | 9.9 (3.9–17.1) | 9.7 (4.0–16.8) | 10.3 (3.9–17.8) |
| BMI (median, IQR) | 26.9 (24.6–29.8) | 26.5 (23.9–29.7) | 27.2 (25.2–29.8) | 31.9 (26.7–38.5) | 31.6 (24.6–38.0) | 32.4 (27.1–39.0) |
| HIV markers, median (IQR) | ||||||
| HIV viral load (copies/mL) | ND (ND–20) | ND (ND–20) | ||||
| CD4 count (cells/mm)3 | 674 (494–878) | 726 (514–955) | ||||
| CD4/CD8 ratio | 0.9 (0.6–1.3) | 1.0 (0.6–1.4) | ||||
| FVC % predicted, mean (SD) | 94 (15) | 96 (14) | 93 (15) | 89 (18) | 89 (17) | 91 (20) |
| FEV% predicted, mean (SD)1 | 97 (16) | 98 (16) | 97 (17) | 89 (19) | 89 (18) | 91 (20) |
| FEV/FVC, mean (SD)1 | 0.78 (0.07) | 0.78 (0.07) | 0.78 (0.06) | 0.80 (0.09) | 0.80 (0.08) | 0.80 (0.09) |
| DLCO % predicted, mean (SD) | 85 (14) | 86 (15) | 85 (13) | 86 (16) | 84 (16) | 92 (17) |
| COPD, No. (%) | 29 (8.6) | 17 (11.6) | 12 (6.3) | 111 (10.3) | 79 (10.3) | 32 (10.5) |
| Air pollution | ||||||
| O, ppb, median (IQR)3 | 39.0 (36.9–39.9) | 38.7 (36.7–39.9) | 39.3 (37.0–40.0) | 36.4 (34.1–37.8) | 36.4 (34.1–37.8) | 36.2 (34.0–37.8) |
| O, ppb, mean (SD)3 | 39.2 (3.2) | 39.1 (3.3) | 39.3 (3.2) | 35.8 (3.1) | 35.9 (3.0) | 35.5 (3.3) |
| PMµg/m, median (IQR)2.53 | 9.4 (8.5–10.3) | 9.4 (8.6–10.4) | 9.3 (8.4–10.3) | 9.2 (8.7–10.1) | 9.2 (8.7–10.1) | 9.3 (8.7–10.0) |
| PMµg/m, mean (SD)2.53 | 9.5 (1.3) | 9.5 (1.2) | 9.5 (1.3) | 9.4 (1.2) | 9.4 (1.2) | 9.4 (1.1) |
Pollution exposure and lung function
We assessed associations between lung function and 1-year average PM2.5 and O3 exposures [Table 2]. While some findings reached statistical significance, there were no clear trends. In the MACS, PWoH had slightly higher % predicted DLCO with higher O3 exposure (2.39%, 95% confidence interval [CI] 0.65 to 4.14), while in the WIHS, PWH had higher % predicted DLCO with higher O3 exposure (1.36%, 96% CI 0.2 to 2.51). In the WIHS, higher PM2.5 exposure was associated with lower FEV1/FVC ratio in PWoH (−0.012, 95% CI −0.021 to −0.003) but not in PWH. In summary, we did not find any clinically significant associations between 1-year average PM2.5 or O3 exposures and cross-sectional lung function.
| MACSN=338 (PWoH 192, PWH 146)Effect estimate(95% CI)a | WIHSN=1073 (PWoH 304, PWH 769)Effect estimate(95% CI)a | ||||
|---|---|---|---|---|---|
| Unadjusted | Adjustedb | Unadjusted | Adjustedb | ||
| PM2.51 µg/m3increase | FEV/FVC1 | ||||
| PWoH | 0.001 (−0.006, 0.008) | −0.001 (−0.008, 0.006) | −0.012 (−0.021, −0.002) | −0.012 (−0.021, −0.003) | |
| PWH | 0.009 (0, 0.019) | 0.003 (−0.006, 0.012) | −0.002 (−0.007, 0.004) | 0.001 (−0.004, 0.006) | |
| % predicted FEV1 | |||||
| PWoH | 0.40 (−1.41, 2.22) | 0.71 (−1.13, 2.55) | −0.04 (−2.07, 1.98) | −0.21 (−2.18, 1.76) | |
| PWH | 1.38 (−0.72, 3.49) | 1.83 (−0.43, 4.08) | 0.45 (−0.66, 1.56) | 0.65 (−0.46, 1.76) | |
| % predicted FVC | |||||
| PWoH | 0.98 (−0.62, 2.58) | 0.93 (−0.70, 2.55) | 1.47 (−0.57, 3.50) | 1.31 (−0.70, 3.33) | |
| PWH | 0.77 (−1.13, 2.68) | 1.39 (−0.55, 3.32) | 0.46 (−0.55, 1.48) | 0.46 (−0.56, 1.47) | |
| % predicted DLCO | |||||
| PWoH | 1.38 (−0.07, 2.83) | 0.73 (−0.74, 2.19) | 0.36 (−1.96, 2.68) | −0.93 (−3.27, 1.41) | |
| PWH | 1.27 (−0.85, 3.39) | 1.60 (−0.57, 3.77) | 0.38 (−0.83, 1.59) | 1.36 (0.20, 2.51) | |
| O33 ppbincrease | FEV/FVC1 | ||||
| PWoH | 0.004 (−0.005 0.012) | 0.001 (−0.007, 0.01) | −0.002 (−0.011, 0.008) | −0.003 (−0.012, 0.006) | |
| PWH | 0.013 (0.002, 0.023) | 0.009 (−0.0001, 0.018) | −0.0002 (−0.006, 0.006) | −0.004 (−0.01, 0.002) | |
| % predicted FEV1 | |||||
| PWoH | 0.03 (−2.19, 2.24) | 0.82 (−1.44, 3.08) | −1.17 (−3.20, 0.87) | −0.88 (−2.85, 1.09) | |
| PWH | 1.01 (−1.34, 3.37) | 1.51 (−0.83, 3.86) | −0.15 (−1.47, 1.17) | −0.13 (−1.45, 1.20) | |
| % predicted FVC | |||||
| PWoH | −0.15 (−2.11, 1.81) | 0.67 (−1.34, 2.67) | −0.69 (−2.74, 1.37) | −0.51 (−2.53, 1.51) | |
| PWH | −0.29 (−2.42, 1.84) | 0.35 (−1.66, 2.37) | 0.25 (−0.96, 1.45) | 0.27 (−0.93, 1.48) | |
| % predicted DLCO | |||||
| PWoH | 2.28 (0.54, 4.02) | 2.39 (0.63, 4.14) | −0.03 (−2.24, 2.18) | −0.15 (−2.31, 2.00) | |
| PWH | 0.89 (−1.52, 3.30) | 1.76 (−0.55, 4.07) | 0.33 (−1.03, 1.69) | −0.80 (−2.10, 0.50) | |
Pollution exposure and symptom scores
In both MACS and WIHS populations, there were no notable trends in respiratory symptom scores with higher PM2.5 or O3 exposure when stratified by HIV serostatus [Table 3]. Both PWH and PWoH in the WIHS had worse SGRQ scores with higher PM2.5 exposure; however, these increases did not remain significant after adjustment for covariates. PWoH in the WIHS had worse mMRC scores with greater PM2.5 exposure when mMRC was specified as a binary outcome. SGRQ scores were slightly lower with greater O3 exposure only in PWoH in the WIHS. Among MACS participants, there were no significant associations between pollution exposure and respiratory symptom scores.
| MACSN=338 (PWoH 192, PWH 146)Effect estimate(95% CI)a | WIHSN=1073 (PWoH 304, PWH 769)Effect estimate(95% CI)a | ||||
|---|---|---|---|---|---|
| Unadjusted | Adjustedb | Unadjusted | Adjustedb | ||
| PM2.51 µg/m3increase | SGRQ | ||||
| PWoH | −0.18 (−1.30, 0.95) | 0.22 (−0.88, 1.32) | 1.63 (0.13, 3.14) | 1.37 (−0.05, 2.79) | |
| PWH | 0.69 (−1.04, 2.43) | −0.04 (−1.92, 1.85) | 1.24 (0.24, 2.24) | 0.61 (−0.37, 1.59) | |
| mMRC | |||||
| PWoH | 0.01 (−0.07, 0.09) | −0.02 (−0.09, 0.05) | 0.07 (−0.06, 0.19) | 0.05 (−0.07, 0.17) | |
| PWH | 0.08 (−0.04, 0.20) | −0.03 (−0.15, 0.09) | 0.07 (−0.01, 0.15) | 0.05 (−0.03, 0.12) | |
| mMRC binaryc | |||||
| PWoH | 0.98 (0.61, 1.58) | 0.64 (0.25, 1.66) | 1.24 (0.99, 1.55) | 1.36 (1.04, 1.78) | |
| PWH | 1.37 (0.89, 2.11) | 0.95 (0.51, 1.80) | 1.09 (0.97, 1.24) | 1.06 (0.92, 1.21) | |
| O33 ppbincrease | SGRQ | ||||
| PWoH | −0.51 (−1.91, 0.90) | −0.002 (−1.40, 1.39) | −1.78 (−3.30, −0.27) | −1.72 (−3.13, −0.31) | |
| PWH | 0.15 (−1.81, 2.10) | −0.54 (−2.50, 1.41) | −0.31 (−1.49, 0.89) | 0.11 (−1.06, 1.28) | |
| mMRC | |||||
| PWoH | 0.02 (−0.08, 0.11) | −0.01 (−0.09, 0.08) | −0.02 (−0.14, 0.11) | −0.03 (−0.15, 0.09) | |
| PWH | 0.08 (−0.05, 0.21) | 0.03 (−0.10, 0.17) | 0.04 (−0.06, 0.13) | 0.05 (−0.04, 0.15) | |
| mMRC binaryc | |||||
| PWoH | 1.07 (0.62, 1.86) | 0.45 (0.12, 1.77) | 0.92 (0.73, 1.15) | 0.85 (0.65, 1.10) | |
| PWH | 1.42 (0.94, 2.14) | 1.48 (0.75, 2.91) | 1.05 (0.91, 1.22) | 1.09 (0.93, 1.28) | |
Symptoms in participants with COPD
Within the MACS and WIHS cohorts, there were a total of 140 participants diagnosed with COPD, 96 PWH and 44 PWoH [Table 4]. In the COPD group 79% were women, reflective of the larger number of WIHS than MACS participants included in the overall population. More PWH had no smoking history compared to PWoH (26% vs. 18%) and smoking history was less in PWH than in PWoH (median of 14 vs. 22 pack-years). Within the COPD subgroup, a 1 µg/m3 increase in 1-year average exposure to PM2.5 was associated with an increase in 4.11 points in the SGRQ score in PWH (95% CI 0.57 to 7.65) [Table 5]. Worse SGRQ score with greater PM2.5 exposure persisted in the PWH group after adjusting for relevant covariates (4.04 points, 95% CI 0.36 to 7.72). The association between poor respiratory symptom scores and PM2.5 exposure in PWH persisted after additionally adjusting for degree of airflow obstruction (FEV1% predicted) (3.62 points, 95% CI 0.07 to 7.17). In contrast, there were no significant associations between PM2.5 exposure and SGRQ score in the PWoH group.
Similar results were found with mMRC scores: for a 1 µg/m3 increase in 1-year average PM2.5 exposure, mMRC scores worsened by 0.26 points in PWH with COPD (95% CI 0.01–0.51); the association persisted after adjustment for covariates including baseline airflow obstruction (0.27 points, 95% CI 0.01 to 0.52). Higher PM2.5 exposure was also associated with higher mMRC (binary specification), although this association did not hold after adjustment for covariates. There was no significant association between PM2.5 exposure and mMRC score in PWoH with COPD.
Unlike PM2.5, there were no significant associations between higher O3 exposure and respiratory symptoms in either PWH with COPD or PWoH with COPD, aside from lower SGRQ in PWoH with increased O3 exposure on unadjusted analyses only.
| Cohort characteristics | COPD | Impaired Diffusion Capacity (DLCO <80% predicted) | ||||
|---|---|---|---|---|---|---|
| Total | PWH | PWoH | Total | PWH | PWoH | |
| No. of participants | 140 | 96 | 44 | 240 | 149 | 91 |
| Age, median (IQR), y | 56 (49–63) | 55 (47–61) | 61 (53–65) | 59 (53–65) | 57 (50–61) | 64 (56–71) |
| Sex, No. (%) | ||||||
| Female | 111 (79.3) | 79 (82.3) | 32 (72.7) | 135 (56.3) | 110 (73.8) | 25 (27.5) |
| Male | 29 (20.7) | 17 (17.7) | 12 (27.3) | 105 (43.7) | 39 (26.2) | 66 (72.5) |
| Race, No. (%) | ||||||
| Black | 77 (55.0) | 53 (55.2) | 24 (54.5) | 108 (45.0) | 83 (55.7) | 25 (27.5) |
| White | 34 (24.3) | 22 (22.9) | 12 (27.3) | 97 (40.4) | 40 (26.8) | 57 (62.6) |
| Other | 29 (20.7) | 21 (21.9) | 8 (18.2) | 35 (14.6) | 26 (17.4) | 9 (9.9) |
| Hispanic ethnicity, No. (%) | 10 (7.1) | 7 (7.3) | 3 (6.8) | 26 (10.8) | 22 (14.8) | 4 (4.4) |
| <=High school education, No. (%) | 70 (50.0) | 52 (54.2) | 18 (40.9) | 107 (44.6) | 82 (55.0) | 25 (27.5) |
| Smoking status, No. (%) | ||||||
| Never | 32 (23.2) | 24 (25.5) | 8 (18.2) | 71 (29.7) | 46 (31.1) | 25 (27.5) |
| Current | 59 (42.3) | 42 (44.7) | 17 (38.6) | 81 (33.9) | 62 (41.9) | 19 (20.9) |
| Former | 47 (34.1) | 28 (29.8) | 19 (43.2) | 87 (36.4) | 40 (27.0) | 47 (51.6) |
| Pack-years of smoking (median, IQR)* | 15.5 (5.7–30.2) | 14.3 (6.1–26.9) | 21.9 (4.8–40.3) | 13.3 (4.4–26.5) | 12.2 (5.0–20.7) | 22.0 (1.1–38.3) |
| BMI (median, IQR) | 29.4 (24.2–34.0) | 28.4 (23.9–32.6) | 31.2 (25.4–36.7) | 28.3 (25.2–33.7) | 28.1 (24.4–33.8) | 28.6 (25.7–33.1) |
| HIV markers, median (IQR) | ||||||
| HIV viral load (copies/mL) | ND (ND–20) | ND (ND–20) | ||||
| CD4 count (cells/mm)3 | 704 (480–876) | 652 (459–874) | ||||
| CD4/CD8 ratio | 1.0 (0.7–1.3) | 0.7 (0.5–1.2) | ||||
| FVC % predicted, mean (SD) | 91 (25) | 91 (21) | 91 (32) | 86 (15) | 85 (15) | 87 (16) |
| FEV% predicted, mean (SD)1 | 71 (17) | 71 (16) | 70 (20) | 87 (17) | 86 (16) | 88 (17) |
| FEV/FVC, mean (SD)1 | 0.62 (0.08) | 0.62 (0.08) | 0.60 (0.09) | 0.79 (0.07) | 0.80 (0.07) | 0.77 (0.08) |
| DLCO % predicted, mean (SD) | 80 (18) | 79 (20) | 82 (15) | 69 (8) | 69 (8) | 71 (7) |
| COPD, No. (%) | 25 (10.4) | 14 (9.4) | 11 (12.1) | |||
| Air pollution | ||||||
| O, ppb, median (IQR)3 | 36.9 (34.9–38.1) | 36.7 (34.8–37.9) | 37.1 (35.1–38.4) | 37.9 (35.9–39.3) | 37.7 (35.6–39.1) | 38.4 (36.5–39.5) |
| O, ppb, mean (SD)3 | 36.4 (3.1) | 36.3 (3.0) | 36.7 (3.5) | 37.2 (3.5) | 36.8 (3.5) | 38.0 (3.5) |
| PMµg/m, median (IQR)2.53 | 9.5 (8.8–10.2) | 9.4 (8.7–10.1) | 9.9 (8.9–10.4) | 9.3 (8.6–10.3) | 9.5 (8.6–10.3) | 9.3 (8.4–10.3) |
| PMµg/m, mean (SD)2.53 | 9.6 (1.2) | 9.4 (1.1) | 9.8 (1.5) | 9.6 (1.4) | 9.6 (1.4) | 9.5 (1.3) |
| MACS/WIHS COPD SubgroupN=140 (PWoH 96, PWH 44)Effect estimate(95% CI)a | ||||
|---|---|---|---|---|
| Unadjusted | Adjustedb | Adjusted for FEV% predicted1c | ||
| PM2.51 µg/m3increase | SGRQ | |||
| PWoH | 2.58 (−0.92, 6.07) | 1.80 (−1.55, 5.16) | 1.99 (−1.36, 5.34) | |
| PWH | 4.11 (0.57, 7.65) | 4.04 (0.36, 7.72) | 3.62 (0.07, 7.17) | |
| mMRC | ||||
| PWoH | 0.08 (−0.1, 0.37) | −0.11 (−0.34, 0.13) | −0.12 (−0.35, 0.11) | |
| PWH | 0.26 (0.01, 0.51) | 0.28 (0.02, 0.53) | 0.27 (0.01, 0.52) | |
| mMRC binaryd | ||||
| PWoH | 1.30 (0.84, 2.02) | 3.96 (0.52, 30.05) | ||
| PWH | 1.55 (1.02, 2.36) | 1.57 (0.96, 2.57) | ||
| O33 ppbincrease | SGRQ | |||
| PWoH | −4.70 (−8.98, −0.42) | −3.00 (−7.70, 1.70) | −3.18 (−7.86, 1.50) | |
| PWH | −1.06 (−4.95, 2.83) | −1.33 (−5.46, 2.80) | −1.76 (−5.72, 2.21) | |
| mMRC | ||||
| PWoH | 0.02 (−0.35, 0.38) | −0.01 (−0.34, 0.32) | 0.005 (−0.32, 0.33) | |
| PWH | −0.13 (−0.41, 0.14) | −0.09 (−0.38, 0.20) | −0.10 (−0.39, 0.19) | |
| mMRC binaryd | ||||
| PWoH | 1.01 (0.60, 1.69) | 0.95 (0.17, 5.42) | ||
| PWH | 0.71 (0.45, 1.11) | 0.72 (0.40, 1.30) | ||
Symptoms in participants with impaired DLCO
There were a total of 240 participants in MACS and WIHS with DLCO < 80% predicted, 149 PWH and 91 PWoH [Table 4]. Within this subgroup, 10% also had COPD by spirometry. In the impaired DLCO group, smoking history was less in PWH than in PWoH (median 12 pack-years vs. 22 pack-years).
In PWH with impaired DLCO, there was an association between 1 µg/m3 greater 1-year average PM2.5 exposure and worse respiratory symptoms, both by SGRQ (2.77 points, 95% CI 0.81 to 4.73) and mMRC (0.21 points, 95% CI 0.05 to 0.36) [Table 6]. This finding persisted after adjustment for covariates (SGRQ 2.14 points, 95% CI 0.2 to 4.08; and mMRC 0.16 points, 95% CI 0.02 to 0.29). Similarly, increased PM2.5 exposure was associated with higher mMRC (binary specification), although the estimate for PWH was attenuated after adjustment for HIV variables. In PWoH with impaired DLCO, there was no significant association between increased PM2.5 exposure and respiratory symptom scores. With higher O3 exposure, there were no significant differences in respiratory symptoms with increased O3 exposure in either PWH or PWoH in the adjusted models.
| MACS/WIHS Impaired Diffusion Capacity Subgroup=240 (PWoH 91, PWH 149)Effect estimate(95% CI)Na | |||
|---|---|---|---|
| Unadjusted | Adjustedb | ||
| PM2.51 µg/m3increase | SGRQ | ||
| PWoH | 0.17 (−2.33, 2.66) | −0.04 (−2.40, 2.33) | |
| PWH | 2.77 (0.81, 4.73) | 2.14 (0.20, 4.08) | |
| mMRC | |||
| PWoH | 0.06 (−0.12, 0.25) | 0.01 (−0.14, 0.17) | |
| PWH | 0.21 (0.05, 0.36) | 0.16 (0.02, 0.29) | |
| mMRC binaryc | |||
| PWoH | 1.18 (0.82, 1.70) | 1.11 (0.54, 2.28) | |
| PWH | 1.37 (1.06, 1.78) | 1.37 (0.99, 1.87) | |
| O33 ppbincrease | SGRQ | ||
| PWoH | −4.55 (−7.27, −1.83) | −2.58 (−5.49, 0.34) | |
| PWH | −1.33 (−3.75, 1.10) | −0.32 (−3.04, 2.41) | |
| mMRC | |||
| PWoH | −0.27 (−0.47, −0.07) | −0.09 (−0.28, 0.10) | |
| PWH | −0.09 (−0.28, 0.10) | −0.02 (−0.21, 0.17) | |
| mMRC binaryc | |||
| PWoH | 0.54 (0.33, 0.88) | 0.57 (0.22, 1.48) | |
| PWH | 0.89 (0.67, 1.18) | 0.96 (0.64, 1.44) | |
Sensitivity analyses pooling PWH and PWoH
In sensitivity analyses pooling PWH and PWoH groups, the p-value for the interaction term between HIV serostatus and the pollution exposure reached statistical significance (< 0.05) only for PM2.5 exposure and lower FEV1/FVC ratio. The p-value was not statistically significant for the interaction term between HIV serostatus and PM2.5 or O3 exposure in any other lung function measure or symptom score outcome in the MACS or WIHS, or in any symptom score outcome in the COPD and impaired diffusion subgroups. These data are not adjusted for HIV-related covariates (Supplementary Tables 2–4).
Discussion
We found that among participants in the MACS and WIHS cohorts, greater exposure to PM2.5 was associated with worse respiratory symptoms in PWH with COPD or impaired diffusion capacity, while this was not the case in PWoH with similar lung disease characteristics. To our knowledge, our study is the first to examine the potential negative effects of the combination of HIV infection and PM2.5 air pollution, using standardized questionnaires to measure respiratory symptoms, impact, and functional disability, and finding an enhanced negative effect in those with underlying COPD or impaired DLCO. These results demonstrate the relevance of air pollution exposure among susceptible populations, as the average PM2.5 concentrations were not excessively high. It is notable that an increase of 1 µg/m3 in average outdoor pollution exposure was associated with a 4-point increase in SGRQ, the threshold considered clinically relevant for patient outcomes and management [27]. Worse SGRQ scores are associated with increased healthcare utilization, hospital readmissions, and exacerbations [37, 38]. Our data supports the American Thoracic Society findings from the Health of the Air Report [39] that calls for a reduction in outdoor air standards to 8 µg/m3, and provides evidence that lower standards may result in a reduction in respiratory morbidities.
Our results align with prior small studies showing that PWH may experience worse respiratory symptoms when exposed to other forms of air pollution. In a study in Malawi [40], respiratory symptoms such as breathlessness and cough were common in PWH who were exposed to household air pollution (HAP). In PWH exposed to HAP, there was a trend toward increased proinflammatory cytokines and vascular injury markers, as well as lower levels of IL-2 and IL-16, cytokines associated with HIV clearance. Inflammation, vascular injury, and viral persistence are potential mechanisms of HAP-related COPD in PWH. Similarly, a study assessing carbon monoxide (CO) exposure from biomass fuel use in Uganda demonstrated that higher CO exposure was associated with worse respiratory symptoms in PWH but not in PWoH [41].
Interestingly, we found that while PM2.5 exposure was associated with worse symptoms in PWH with lung disease, O3 exposure was not. Although this is in keeping with a study showing that short-term exposure to PM2.5, but not O3 or other pollutants, was associated with COPD exacerbation [42], our results are in contrast to a SubPopulations and InteRmediate Outcome Measures in COPD (SPIROMICS) study wherein greater 10-year O3 exposure was associated with COPD morbidity, including reduced lung function, worse respiratory symptoms, increased exacerbations, and greater emphysema [14]. However, in the SPIROMICS study, when PM2.5 exposure was included in the model, there was no longer a significant association between O3 exposure and SGRQ score or exacerbations. It is possible that particulate exposure is more closely associated with airways disease and respiratory symptoms, while O3 contributes to progression of emphysema [18, 43, 44].
We did not find that greater pollution exposure was associated with worse respiratory symptoms in the overall population or in PWoH with chronic lung disease. While it is generally accepted that air pollution worsens outcomes in COPD, the data are variable based on type and degree of exposure: for example, a meta-analysis of studies of air pollution in COPD showed significant heterogeneity in the effects of PM2.5 on respiratory symptoms, and no overall association between gaseous pollutants including O3 and either lung function or symptoms [45]. Our negative results in the overall population and the PWoH with chronic lung disease subgroup may be related to narrow IQR ranges of pollution exposures, 1-year duration of exposure estimation, and limited sample size. Although our study was underpowered to detect small differences, we found the strongest associations in people with multiple susceptibilities to air pollution (HIV plus chronic lung disease). We speculate that the effects of air pollution are amplified in the setting of underlying airway inflammation, as mediated by both COPD and HIV.
We did not find any clinically significant associations between 1-year average PM2.5 or O3 exposures and cross-sectional lung function. Our findings are in concordance with some prior cohort studies of air pollution and lung function that showed inconclusive results [46]. We surmise that our results are attributable to the multitude of factors beyond air pollution exposure that contribute to acquired adult lung function and that could not be accounted for in this study, such as childhood respiratory disease, nutritional factors, and occupational exposures. Future analyses of longitudinal lung function in the MACS and WIHS cohorts will be informative to determine if ambient pollution exposure is associated with loss of lung function in these populations, and if HIV serostatus or the presence of baseline chronic lung disease modify any associations.
This cross-sectional study has several limitations which impact potential conclusions about air pollution and lung function. We acknowledge pollution exposure estimates using residential address may be imprecise; however, previous studies comparing pollution estimates using residential address versus time-activity patterns have shown high correlations both in degree of exposure and associated health effects [47]. In addition, there is precedent for using residential address as well as 1-year pollution exposure estimates in the COPD and air pollution literature [48].
Another limitation of our study is the use of two cohorts with inherent differences between the populations, most notably sex but also COPD prevalence. As with other studies utilizing these cohorts [49, 50], we chose to stratify by sex (cohort) and, in the subgroup analyses with small sample sizes, combine the MACS and WIHS participants while adjusting for sex (cohort). We acknowledge that the structure of the cohorts and sample size constraints in subgroup analyses may affect interpretability of our results. Sensitivity analyses pooling PWH and PWoH generally did not yield statistically significant interactions between HIV serostatus and pollution exposures. However, these unstratified data must be interpreted with caution given the differences between the PWH and PWoH groups that challenge direct comparison [51], and the lack of adjustment for HIV-related covariates in the pooled analyses. Similar to prior publications in the literature we draw conclusions from the analyses stratified by HIV status [41, 52].
Additionally, our analysis is limited by definition of diffusion impairment as < 80% predicted based on previously published work in the area [35], which may include some people who would be characterized as normal using lower limit of normal criteria, although this misclassification if present would only reduce our effect estimate. We did not further restrict DLCO to moderate impairment (< 60% predicted) due to small sample sizes.
Nevertheless, the strengths of our results include concordant findings in SGRQ and mMRC outcomes, as well as between COPD and impaired DLCO subgroups, supporting a potential pathophysiologic process for worse respiratory symptoms from PM2.5 air pollution in the setting of HIV and underlying chronic lung disease. In addition, our findings notably demonstrate an association between PM2.5 exposure and respiratory symptoms that is independent of tobacco smoking and degree of airflow obstruction in PWH with COPD.
Conclusions
In summary, we demonstrated that PM2.5 pollution exposure was associated with respiratory morbidity in a sample of PWH with chronic lung disease. The combination of chronic HIV infection and air pollution may accelerate or activate lung disease. Further studies are needed to determine how chronic HIV infection predisposes individuals to respiratory symptom exacerbation and whether ambient air pollution plays a role in loss of lung function in PWH.
Supplementary Information
Acknowledgements
The authors gratefully acknowledge the contributions of the study participants and dedication of the staff at the MWCCS sites.
Abbreviations
Authors’ contributions
IB had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. MLW, AE, and IB designed the study, performed and interpreted the analyses, and wrote the initial draft of the manuscript. KPK, JE, AGA, SPB, RGB, MHC, RD, MBD, MAF, RF, HDH, LH, KMK, MM, AM, SR, CR, JR, VS, DT, AV, SDW, and GW contributed substantially to the study design, data analysis and interpretation, and the writing of the manuscript. All authors substantially revised the draft manuscript and approved the final manuscript.
Funding
Data in this manuscript were collected by the Multicenter AIDS Cohort Study (MACS) and the Women’s Interagency HIV Study (WIHS), now the MACS/WIHS Combined Cohort Study (MWCCS). The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health (NIH). MWCCS (Principal Investigators): Atlanta CRS (Ighovwerha Ofotokun, Anandi Sheth, and Gina Wingood), U01-HL146241; Baltimore CRS (Todd Brown and Joseph Margolick), U01-HL146201; Bronx CRS (Kathryn Anastos, David Hanna, and Anjali Sharma), U01-HL146204; Brooklyn CRS (Deborah Gustafson and Tracey Wilson), U01-HL146202; Data Analysis and Coordination Center (Gypsyamber D’Souza, Stephen Gange and Elizabeth Topper), U01-HL146193; Chicago-Cook County CRS (Mardge Cohen, Audrey French, and Ryan Ross), U01-HL146245; Chicago-Northwestern CRS (Steven Wolinsky, Frank Palella, and Valentina Stosor), U01-HL146240; Northern California CRS (Bradley Aouizerat, Jennifer Price, and Phyllis Tien), U01-HL146242; Los Angeles CRS (Roger Detels and Matthew Mimiaga), U01-HL146333; Metropolitan Washington CRS (Seble Kassaye and Daniel Merenstein), U01-HL146205; Miami CRS (Maria Alcaide, Margaret Fischl, and Deborah Jones), U01-HL146203; Pittsburgh CRS (Jeremy Martinson and Charles Rinaldo), U01-HL146208; UAB-MS CRS (Mirjam-Colette Kempf, James B. Brock, Emily Levitan, and Deborah Konkle-Parker), U01-HL146192; UNC CRS (M. Bradley Drummond and Michelle Floris-Moore), U01-HL146194. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute (NHLBI), with additional co-funding from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD), National Institute on Aging (NIA), National Institute of Dental & Craniofacial Research (NIDCR), National Institute of Allergy and Infectious Diseases (NIAID), National Institute of Neurological Disorders and Stroke (NINDS), National Institute of Mental Health (NIMH), National Institute on Drug Abuse (NIDA), National Institute of Nursing Research (NINR), National Cancer Institute (NCI), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute on Minority Health and Health Disparities (NIMHD), and in coordination and alignment with the research priorities of the National Institutes of Health, Office of AIDS Research (OAR). MWCCS data collection is also supported by UL1-TR000004 (UCSF CTSA), UL1-TR003098 (JHU ICTR), UL1-TR001881 (UCLA CTSI), P30-AI-050409 (Atlanta CFAR), P30-AI-073961 (Miami CFAR), P30-AI-050410 (UNC CFAR), P30-AI-027767 (UAB CFAR), P30-AI-124414 (ERC-CFAR), P30-MH-116867 (Miami CHARM), UL1-TR001409 (DC CTSA), KL2-TR001432 (DC CTSA), and TL1-TR001431 (DC CTSA). This material is also the result of work supported with resources and the use of facilities at the Boise VA Medical Center.
Data availability
Access to individual-level data from the MACS/WIHS Combined Cohort Study Data (MWCCS) may be obtained upon review and approval of a MWCCS concept sheet. Links and instructions for online concept sheet submission are on [the study website](https:/statepi.jhsph.edu/mwccs/work-with-us↗).
Declarations
Ethics approval and consent to participate
Research was conducted in accordance with the Declaration of Helsinki. All MACS and WIHS participants consented to cohort participation and use of their data for research, with approval obtained from relevant institutional review boards.
Consent for publication
Not applicable.
Competing interests
SB: Funding from NIH, Nuvaira, Sanofi/Regeneron, Genentech. Fees from Sanofi, Regeneron, Boehringer Ingelheim, Apreo, Genentech, Chiesi, AstraZeneca, GSK, Merck, Verona Pharma, Polarean, Integrity CE, Medscape, Horizon CME, Illuminate Health, Integratias Communications.RGB: Funding from NIH/NHLBI, University of California Office of the President, NIH/NCATS, VA/HSR&D, COPD Foundation, and Foundation of the NIH. Fees from DynaMed/American College of Physicians, 2ndMD, Chiesi, Optum, Pri-Med.KMK: Fees from Nuvaira.AM: Funding from Pfizer.SR: Funding from NIH and Areteia Therapeutics.DT: Funding from NIH/NHLBI, Foundation of the NIH, and COPD Foundation.IB: Funding from Theravance and Viatris, Aerogen, Alpha-1 Foundation, Johnny Carson’s Foundation, Takeda, Amgen. Fees from AstraZeneca, Sanofi/Regeneron, Grifols, Verona Pharma, Inhibrx, Takeda, Genentech, Aerogen, Therevance and Viatris.
Footnotes
References
Associated Data
Supplementary Materials
Data Availability Statement
Access to individual-level data from the MACS/WIHS Combined Cohort Study Data (MWCCS) may be obtained upon review and approval of a MWCCS concept sheet. Links and instructions for online concept sheet submission are on [the study website](https:/statepi.jhsph.edu/mwccs/work-with-us↗).