Abstract
Objectives
To describe the prevalence and correlates of chronic obstructive pulmonary disease (COPD) in a multicentre international cohort of persons living with HIV (PLWH).
Methods
We performed a cross-sectional analysis of adult PLWH, naïve to HIV treatment, with baseline CD4 cell count >500 cells/μL enrolled in the Pulmonary Substudy of the Strategic Timing of AntiRetroviral Treatment trial. We collected standardised, quality-controlled spirometry. COPD was defined as forced expiratory volume in one second/forced vital capacity (FEV1/FVC) ratio less than the lower limit of normal.
Results
Among 1026 participants from 80 sites and 20 countries, median (IQR) age was 36 (30, 44) years, 29% were female, and time since HIV diagnosis was 1.2 (0.4, 3.5) years. Baseline CD4 cell count was 648 (583, 767) cells/μL, viral load was 4.2 (3.5, 4.7) log10 copies/mL, and 10% had viral load ≤400 copies/mL despite lack of HIV treatment. Current/former/never smokers comprised 28%/11%/61% of the cohort, respectively. COPD was present in 6.8% of participants, and varied by age, smoking status, and region. 48% of those with COPD reported lifelong nonsmoking. In multivariable regression, age and pack-years of smoking had the strongest associations with FEV1/FVC ratio (p<0.0001). There were significant differences between the effect of region on FEV1/FVC ratio (p=0.010).
Conclusions
Our data suggest that among PLWH, naïve to HIV treatment and with CD4 cell count >500 cells/μL, smoking and age are important factors related to COPD. Smoking cessation should remain a high global priority for clinical care and research in PLWH.
Keywords: HIV, pulmonary disease, spirometry, smoking, START trial
INTRODUCTION
Emerging data in the era of effective antiretroviral treatment (ART) suggest that pulmonary complications of HIV are common, especially chronic obstructive pulmonary disease (COPD) (1). COPD is a major global health problem, being the third leading cause of death and fifth leading cause of disability in 2010 (2, 3). Although the most common cause of COPD is cigarette smoking, and smoking rates are often high in populations of persons living with HIV (PLWH), studies have consistently identified HIV infection as an independent risk factor for COPD, even when adjusted for smoking (4–7). The mechanism explaining how HIV infection increases COPD risk is not clear, but hypotheses include frequent respiratory infections, alterations in the lung microbiota, pulmonary inflammation (including alveolar CD8 T-cell recruitment and local upregulation of matrix metalloproteinase expression from HIV infection of alveolar macrophages) and oxidative stress (1, 8–10).
The role of ART as a factor in HIV-associated COPD is unclear, as existing studies disagree. Two studies implicated ART as associated with an increased risk of COPD (11,12), another study showed a lower incidence of COPD among ART users versus non-users (5), and others showed no association (6,13,14). In these studies, ART use might be confounded by other factors such as socioeconomic status and adherence to therapy (e.g., PLWH of low socioeconomic status may have difficulty accessing ART, and prescribers may be less willing to offer ART to those who are not likely to adhere).
To address this important knowledge gap, we are conducting a nested pulmonary substudy in the Strategic Timing of AntiRetroviral Treatment (START) trial to determine whether ART initiated at CD4 cell counts >500 cells/μL, compared with deferred ART to 350 cells/μL, slows decline in lung function among HIV-positive persons. The substudy also provides novel information about lung function in HIV patients from low-income countries. Current lung function data in HIV have exclusively come from North American or European cohorts, with no published data from other regions of the world where HIV infection is highly prevalent, such as Africa and Asia. Here, we report on baseline spirometry data collected from around the world in the START Pulmonary Substudy.
METHODS
The design and methods of the START trial have been previously published (15). All substudy participants provided additional informed consent specific to their substudy participation and all site institutional review boards/ethics committees approved the substudy. We selected substudy sites on the basis of their interest and ability to participate. We set no specific criteria for regional distribution of sites. Initial recruitment began at 36 sites during the pilot phase of the main START trial, and we added 44 sites during the definitive expansion phase of START.
Pulmonary Substudy Inclusion and Exclusion Criteria
In addition to the entry criteria for the START trial, additional pulmonary substudy criteria included the requirement that participants be ≥25 years old, the age at which lung function, the primary outcome, begins to decline in most normal adults. Additional exclusion criteria were directed towards spirometry accuracy and safety and included: 1) an episode of respiratory illness with two or more symptoms of cough, wheezing, breathlessness, or increase in sputum production within the six weeks before baseline spirometry; 2) use of asthma medications (bronchodilator, inhaled corticosteroid, leukotriene inhibitor, or theophylline) for two or more consecutive weeks within the six months before baseline spirometry; 3) relative contraindications to spirometry, such as chest, abdominal, or eye surgery within the three months before baseline spirometry, or known retinal detachment at the time of baseline spirometry; 4) known allergy to albuterol/salbutamol, relative contraindications to albuterol/salbutamol, such as resting heart rate of >110 beats per minute, and 5) a known serious or recurrent or uncontrolled cardiac condition (such as unstable coronary artery disease, decompensated heart failure, or recurrent tachyarrhythmias).
Spirometry methods
Sites performed spirometry using the EasyOne ultrasonic flow device (ndd Medical, Zurich, Switzerland). Sites calibrated the device on the day of spirometry, using a standardised 3L calibration syringe (Hans Rudolph, Inc., Shawnee, Kansas, United States) to confirm device volume measurement to within ±3% accuracy. Participants inhaled 180 mcg of albuterol/salbutamol via a metered dose inhaler, then performed spirometry at least 15 minutes later.
Participants performed spirometry while seated, using a noseclip, with ndd brand spirettes™. Per American Thoracic Society (ATS)/European Respiratory Society (ERS) standards, participants performed at least three and no more than eight maneuvers. The EasyOne device automatically generates a suggested quality grade, with a grade of “A” and “B” both meeting the most updated ATS/ERS 2005 spirometry standards, which include the requirement that the two best measurements of both forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) be reproducible to within 150 mL of each other (16). A grade of “C” meets the older ATS 1994 standards, where FEV1 and FVC reproducibility had to be within 200 mL of each other (17). A grade of “D” or “F” did not meet accepted quality standards. The substudy goal was to achieve a grade of “A” or “B,” although we considered a grade of “C” acceptable.
Site staff training included formal didactic education and hands-on spirometry training. Due to the large numbers and global distribution of sites, we employed a train-the-trainer model. A core group of expert trainers was centrally trained in the United States by the study chair (author Ken Kunisaki), and these trainers returned to their home regions and trained technicians. When possible, the study chair also participated in regional training sessions. Once training was completed, technicians submitted spirometry tracings for study chair review for certification. Following certification, technicians also participated in a spirometry skill maintenance program and were asked to submit monthly spirometry tracings performed on spirometry-naïve volunteers.
Printed spirometry tracings were submitted to the INSIGHT Statistical and Data Management Center, where the study chair performed quality control review. If a device-assigned value for FEV1, FVC, reproducibility, or quality grade was determined to be incorrect, corrected final values were assigned in the study database. The study chair provided written feedback to the study technician on every tracing and requested repeat spirometry in cases of unacceptable quality. In this analysis, we report baseline spirometry data only for tests with final grades of “A,” “B,” and “C,” as determined after study chair quality control review.
We used Global Lung Function Initiative (GLI) 2012 normative equations to determine predicted FEV1, predicted FVC, and lower limit of normal (LLN) (i.e., fifth percentile) of FEV1/FVC ratio (18). The GLI group pooled spirometry data collected from over 74,000 healthy individuals around the world to derive predictive equations that account for an individual’s age, height, gender, and ethnicity. GLI ethnic categories include Caucasian, African-American, South East Asian, North East Asian, and other. Of the countries and participants in the START Pulmonary Substudy, none were from the North East Asian region. GLI notably lacked data from Sub-Saharan Africa and India. We therefore analysed African blacks and non-African blacks in the African-American category; Indians were analysed in the “other” category. GLI also had no data from Peru and Argentina, but other Latin-American data (Brazil, Chile, Uruguay, Venezuela) led to Latin-Americans being classified as Caucasians, so we chose to include participants from Peru and Argentina as Caucasians for purposes of GLI predictive equations. There were no GLI data from Morocco, but data from Algeria and Tunisia led to North Africans being classified as Caucasians, so we chose to include participants from Morocco as Caucasians for purposes of GLI predictive equations.
Other data collection
Site staff also asked participants specific questions related to tobacco use to assess current and former cigarette use. We defined current smokers as having smoked more than 10 cigarettes in the past 30 days. We defined lifelong nonsmokers as having reported smoking less than 120 cigarettes in their lifetime. We calculated smoking pack-years as: (average number of cigarettes per day/20) x (number of years smoked). We also collected data about noncigarette tobacco use, but only six participants reported exclusive use of cigarillos (n=2), cigars (n=2) or pipes (n=2).
Statistical analysis
Descriptive statistics of the cohort are means ± standard deviation or median (interquartile range) as appropriate. Statistical significance tests included t-tests, nonparametric tests, or chi-square tests as appropriate.
COPD is commonly defined using one of two definitions: 1) FEV1/FVC ratio <0.70 or 2) FEV1/FVC ratio <lower limit of normal (LLN), defined as being <5th percentile of a normal adult population, adjusted for age, height, gender, and ethnicity. There is disagreement about which definition is preferable, but the fixed ratio (i.e., ratio <0.70) definition tends to underdiagnose COPD at young ages and overdiagnose COPD at older ages. Given the young age of the START cohort, we used the LLN definition as our primary COPD definition, but we present data using both definitions. We classified COPD severity according to Global Initiative for Chronic Obstructive Lung Disease criteria of mild (FEV1 ≥80% of predicted normal), moderate (FEV1 between 50% to <80% of predicted), severe (FEV1 between 30% and <50% of predicted), and very severe (FEV1 <30% of predicted) (19).
We additionally performed multivariable linear regression analysis, using FEV1/FVC ratio as the dependent variable and region, age, height, gender, ethnicity, and pack-years of smoking as independent variables.
The sample size of the substudy was determined based on the planned prospective randomised comparison of rate of FEV1 decline between the immediate and deferred ART arms of START. Due to the strong effects of smoking on FEV1 decline, we preplanned separate analyses for smokers and nonsmokers. Assumptions of the sample size determination included a standard deviation for rate of FEV1 decline of 60 mL/year, a minimal clinically important difference of 15.5 mL/year in nonsmokers and 19.0 mL/year in smokers, a two-tailed alpha error rate of 0.05, and a desired power of 0.85. This resulted in a target sample size of 580 non-smokers and 332 smokers (total sample size was increased to 1,000 to allow for 10% study withdrawals and losses to follow-up).
RESULTS
The substudy enrolled 1026 participants from 80 sites in 20 countries between March 2010 and September 2013. Substudy participants were similar to the main START trial participants with respect to age, gender, CD4, HIV RNA, BMI, and comorbidities. Compared to the main START trial, the substudy had a higher percentage of participants from African sites. Substudy participants were also more likely to be female, have acquired HIV via heterosexual contact, and less likely to be current smokers.
The region with highest enrolment was Africa (n=328), followed by Europe/Israel/Australia (n=313), Mexico/South America (n=191), Asia (n=103), and the United States (n=91) (Table 1). Overall, pulmonary substudy participants had a median age of 36 years, and 29.1% were female. Most reported HIV acquisition via sexual contact (49.2% men who have sex with men, 43.9% heterosexual contact), and median time since HIV diagnosis of 1.2 years. Median baseline CD4 cell count was 648 cells/μL, median viral load was 4.2 log10 copies/mL, and 9.5% had suppressed viral load (≤400 copies/mL) despite lack of ART use. Current/former/never smokers comprised 28.3%/10.8%/60.9% of the cohort, respectively, and the most common comorbidity was hypertension (18.7%).
TABLE 1.
Baseline characteristics of START Pulmonary Substudy participants by region
| Total | Africa | Asia | Europe/Israel/Australia | Mexico/South America | United States | |
|---|---|---|---|---|---|---|
| 1026 (100.0%) | 328 (32.0%) | 103 (10.0%) | 313 (30.5%) | 191 (18.6%) | 91 (8.9%) | |
| Demographics | ||||||
| Age (years) | 36 (30, 44) | 37 (32, 44) | 36 (30, 41) | 38 (31, 45) | 34 (29, 40) | 36 (29, 47) |
| Female | 299 (29.1%) | 211 (64.3%) | 27 (26.2%) | 26 (8.3%) | 26 (13.6%) | 9 (9.9%) |
| Race | ||||||
| Black | 387 (37.7%) | 321 (97.9%) | 0 (0.0%) | 22 (7.0%) | 0 (0.0%) | 44 (48.4%) |
| Latino/Hispanic | 179 (17.4%) | 0 (0.0%) | 0 (0.0%) | 14 (4.5%) | 153 (80.1%) | 12 (13.2%) |
| Asian | 105 (10.2%) | 0 (0.0%) | 101 (98.1%) | 1 (0.3%) | 0 (0.0%) | 3 (3.3%) |
| White | 344 (33.5%) | 6 (1.8%) | 1 (1.0%) | 268 (85.6%) | 38 (19.9%) | 31 (34.1%) |
| Other | 11 (1.1%) | 1 (0.3%) | 1 (1.0%) | 8 (2.6%) | 0 (0.0%) | 1 (1.1%) |
| HIV Factors | ||||||
| Likely mode of HIV infection | ||||||
| Injecting drug use | 14 (1.4%) | 0 (0.0%) | 1 (1.0%) | 5 (1.6%) | 2 (1.0%) | 6 (6.6%) |
| Male sexual contact with person of same sex | 505 (49.2%) | 5 (1.5%) | 59 (57.3%) | 238 (76.0%) | 138 (72.3%) | 65 (71.4%) |
| Sexual contact with person of opposite sex | 450 (43.9%) | 306 (93.3%) | 34 (33.0%) | 46 (14.7%) | 46 (24.1%) | 18 (19.8%) |
| Other/unknown | 57 (5.6%) | 17 (5.2%) | 9 (8.7%) | 24 (7.7%) | 5 (2.6%) | 2 (2.2%) |
| Time since first HIV diagnosis, years | 1.2 (0.4, 3.5) | 1.5 (0.5, 4.8) | 0.8 (0.2, 3.4) | 1.2 (0.5, 3.5) | 0.6 (0.3, 2.2) | 1.5 (0.4, 4.6) |
| CD4 cell count (cells/μL) | 648 (583, 767) | 695 (603, 814) | 618 (561, 728) | 634 (581, 738) | 632 (574, 718) | 674 (582, 773) |
| CD8 cell count (cells/μL) | 1019 (758, 1394) | 922 (692, 1275) | 1079 (805, 1452) | 1069 (800, 1547) | 1023 (770, 1383) | 959 (765, 1400) |
| CD4/CD8 ratio | 0.66 (0.48, 0.94) | 0.79 (0.56, 1.06) | 0.58 (0.44, 0.81) | 0.62 (0.44, 0.81) | 0.63 (0.45, 0.91) | 0.66 (0.51, 0.99) |
| Nadir CD4 (cells/μL) | 552 (486, 661) | 605 (519, 740) | 542 (477, 618) | 524 (452, 608) | 545 (492, 630) | 564 (479, 688) |
| Log10 HIV RNA (copies/mL) | 4.2 (3.5, 4.7) | 3.8 (3.0, 4.6) | 4.5 (3.9, 4.9) | 4.3 (3.8, 4.7) | 4.4 (3.9, 4.8) | 3.9 (3.3, 4.5) |
| HIV RNA ≤ 400 copies/mL | 97 (9.5%) | 62 (19.0%) | 6 (5.9%) | 9 (2.9%) | 9 (4.7%) | 11 (12.1%) |
| Medical History | ||||||
| Body mass index, kg/m2 | 24.8 (22.2, 28.4) | 25.7 (22.2, 31.8) | 22.8 (20.7, 25.7) | 23.8 (22.2, 26.6) | 25.5 (23.2, 28.2) | 26.5 (23.1, 30.7) |
| Prior cardiovascular disease | 7 (0.7%) | 0 (0.0%) | 1 (1.0%) | 3 (1.0%) | 1 (0.5%) | 2 (2.2%) |
| Hypertension | 192 (18.7%) | 65 (19.8%) | 9 (8.7%) | 73 (23.3%) | 17 (8.9%) | 28 (30.8%) |
| Diabetes | 37 (3.6%) | 20 (6.1%) | 2 (1.9%) | 5 (1.6%) | 6 (3.1%) | 4 (4.4%) |
| Hepatitis B or C | 72 (7.1%) | 22 (6.7%) | 10 (9.7%) | 23 (7.5%) | 12 (6.3%) | 5 (5.6%) |
| Osteoporosis | 2 (0.2%) | 0 (0.0%) | 0 (0.0%) | 1 (0.3%) | 0 (0.0%) | 1 (1.1%) |
| Smoking status | ||||||
| Current smoker | 290 (28.3%) | 46 (14.0%) | 20 (19.4%) | 140 (44.7%) | 54 (28.3%) | 30 (33.0%) |
| Former smoker | 111 (10.8%) | 19 (5.8%) | 11 (10.7%) | 42 (13.4%) | 27 (14.1%) | 12 (13.2%) |
| Never smoker | 625 (60.9%) | 263 (80.2%) | 72 (69.9%) | 131 (41.9%) | 110 (57.6%) | 49 (53.8%) |
| Pack years smoking (current and former smokers only) | 6.0 (2.0, 15.0) | 2.5 (1.2, 5.0) | 4.3 (1.2, 9.5) | 10.0 (5.0, 20.0) | 4.3 (1.0, 10.5) | 4.6 (1.5, 12.0) |
Data shown as n(%) or median (interquartile range).
Prior cardiovascular disease = history of stroke, myocardial infarction, or coronary revascularisation.
Hypertension= use of blood pressure-lowering medications, systolic blood pressure ≥140 mmHg, or diastolic blood pressure ≥90 mmHg
Diabetes= prior diagnosis of diabetes, use of diabetes medication, or fasting baseline glucose ≥126 mg/dL.
Hepatitis B = positive Hepatitis B surface antigen test in the year prior to randomisation
Hepatitis C = positive Hepatitis C core antibody test any time prior to randomisation
Osteoporosis= prior diagnosis of osteoporosis or use of osteoporosis medication
There were significant regional differences in many baseline demographic characteristics (Table 1). For example, of the 299 women in the study, 211 came from sites in Africa; of the 290 current smokers, only 66 came from sites in Africa and Asia. The proportion of men who reported having sex with men was over 70% in Europe/Israel/Australia, Mexico/South America, and the Unites States, was 57.3% in Asia, but only 1.5% in Africa.
Spirometry was obtained in 1025 participants (99.9%) and overall quality was very good: 775 (75.6%) spirometry tests met ATS/ERS 2005 criteria (grade “A” or “B”) and 214 (20.9%) met ATS 1994 criteria (grade “C”). Thirty-six (3.5%) failed to meet quality standards.
Of the 989 high-quality spirometry tests, evidence of COPD, as defined by the LLN for the FEV1/FVC ratio, was present in 6.8% of this cohort, with regional variation ranging from 2.0% in Asia to 9.1% in Europe/Israel/Australia (p=0.033) (Table 2). Participants with COPD generally had mild or moderate airflow obstruction, with 52.2% having an FEV1 >80% of predicted, 43.3% having an FEV1 between 50%–80% of predicted, 4.5% demonstrating an FEV1<50% of predicted and none with an FEV1 <30% of predicted. The prevalence of COPD varied by smoking status with COPD demonstrated in 11.7% of current smokers, 2.8% of former smokers, and 5.3% of never smokers (p<0.001) (Table 3). The prevalence of COPD also increased with age quartiles, from 3.9% (≤30 years) to 6.7% (31–36 years) to 7.6% (37–44 years) to 9.1% (>44 years), but was not statistically significant (p=0.14) (Table 3). COPD prevalence did not differ between women (6.2%) and men (7.0%) (p=0.63).
TABLE 2.
Spirometry results and comparison of chronic obstructive pulmonary disease (COPD) prevalence/severity by region.
| Total (n=989) | Africa (n=322) | Asia (n=102) | Europe/Israel/Australia (n=298) | Mexico/South America (n=182) | USA (n=85) | p-values | |
|---|---|---|---|---|---|---|---|
| FEV1 %predicted | 96.2 (85.4, 103.8) | 96.1 (84.2, 104.9) | 91.2 (81.4, 100.0) | 96.0 (87.0, 103.9) | 100.1 (91.8, 105.3) | 90.0 (82.1, 101.7) | - |
| FEV1/FVC ratio | 0.83 (0.78, 0.86) | 0.83 (0.78, 0.86) | 0.87 (0.84, 0.89) | 0.80 (0.76, 0.85) | 0.84 (0.81, 0.86) | 0.83 (0.79, 0.86) | - |
| FEV1/FVC <LLN | 67 (6.8%) | 25 (7.8%) | 2 (2.0%) | 27 (9.1%) | 6 (3.3%) | 7 (8.2%) | p=0.033 |
| FEV1 ≥80% predicted | 35 (52.2%) | 10 (40.0%) | 2 (100.0%) | 18 (66.7%) | 4 (66.7%) | 1 (14.3%) | - |
| FEV1 50%–79% predicted | 29 (43.3%) | 15 (60.0%) | 0 (0.0%) | 7 (25.9%) | 2 (33.3%) | 5 (71.4%) | - |
| FEV1 30%–49% predicted | 3 (4.5%) | 0 (0.0%) | 0 (0.0%) | 2 (7.4%) | 0 (0.0%) | 1 (14.3%) | - |
| FEV1 <30% predicted | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | - |
| FEV1/FVC <0.70 | 54 (5.5%) | 16 (5.0%) | 0 (0.0%) | 27 (9.1%) | 5 (2.7%) | 6 (7.1%) | p=0.002 |
| FEV1 ≥80% predicted | 26 (48.1%) | 4 (25.0%) | 0 (0.0%) | 18 (66.7%) | 3 (60.0%) | 1 (16.7%) | - |
| FEV1 50%–79% predicted | 25 (46.3%) | 12 (75.0%) | 0 (0.0%) | 7 (25.9%) | 2 (40.0%) | 4 (66.7%) | - |
| FEV1 30–49% predicted | 3 (5.6%) | 0 (0.0%) | 0 (0.0%) | 2 (7.4%) | 0 (0.0%) | 1 (16.7%) | - |
| FEV1 <30% predicted | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | - |
Data shown as n(%) or median (interquartile range). The primary COPD definition was FEV1/FVC <lower limit of normal (LLN) by Global Lung Function Initiative prediction equations; the secondary COPD definition was FEV1/FVC <0.70. FEV1 % predicted was calculated using Global Lung Function Initiative prediction equations. p-values were calculated by chi-square test.
TABLE 3.
Prevalence of chronic obstructive pulmonary disease (COPD) overall, by smoking status and by age quartiles.
| Total | FEV1/FVC <LLN | FEV1/FVC <0.70 | |
|---|---|---|---|
| Overall cohort | 989 (100%) | 67 (6.8%) | 54 (5.5%) |
| Smoking status | |||
| Current | 273 (27.6%) | 32 (11.7%) | 28 (10.3%) |
| Former | 109 (11.0%) | 3 (2.8%) | 4 (3.7%) |
| Never | 607 (61.4%) | 32 (5.3%) | 22 (3.6%) |
| p-value | - | <0.001 | <0.001 |
| Age (years) quartile | |||
| 25 – 30 | 254 (25.7%) | 10 (3.9%) | 4 (1.6%) |
| 31 – 36 | 255 (25.8%) | 17 (6.7%) | 13 (5.1%) |
| 37 – 44 | 250 (25.3%) | 19 (7.6%) | 11 (4.4%) |
| >44 | 230 (23.3%) | 21 (9.1%) | 26 (11.3%) |
| p-value | - | 0.14 | <0.001 |
Data shown as n(%). The primary COPD definition was FEV1/FVC <lower limit of normal (LLN) by Global Lung Function Initiative prediction equations; the secondary COPD definition was FEV1/FVC <0.70. p-values were calculated by chi-square test.
Comparing those with and without COPD, participants with COPD were a median of three years older (p=0.07) and had a higher proportion of current smokers (47.8% vs. 26.1%; p<0.001). Among those with COPD, while 47.8% were current smokers, the exact same proportion (47.8%) reported lifelong nonsmoking, and the median pack-year burden of smoking in those with COPD was 0.8 pack-years. There was no difference in HIV factors such as current/nadir CD4, CD4/CD8 ratio, CD8, viral load, or years since HIV diagnosis between those with and without COPD (all p-values >0.30).
Data regarding varying COPD prevalence by region, smoking status, and age were similar when using a fixed-ratio definition of COPD, although the fixed ratio definition showed a statistically stronger difference in COPD prevalence by age quartiles (p <0.001).
In multivariable linear regression analysis, increasing age and increasing pack-years of smoking were strong independent correlates of lower FEV1/FVC ratio (Table 4). Height, white race, and region were also independently associated with FEV1/FVC ratio. Black race and gender had no significant association with FEV1/FVC ratio. There were significant differences between the effect of region on FEV1/FVC ratio (p=0.011).
TABLE 4.
Multivariable linear regression analysis, with FEV1/FVC ratio as dependent variable.
| Variable | Coefficient | Standard Error | p-value |
|---|---|---|---|
| Age, per 10 years | −0.021 | 0.002 | <0.001 |
| Pack-years of smoking, per 10 pack-years | −0.007 | 0.002 | <0.001 |
| Height, per 6 inches | −0.009 | 0.004 | 0.03 |
| Female gender | 0.008 | 0.006 | 0.18 |
| Region* | 0.01 | ||
| North America | −0.018 | 0.011 | 0.11 |
| South America | −0.024 | 0.007 | 0.001 |
| Europe/Israel/Australia | −0.023 | 0.010 | 0.02 |
| Africa | −0.030 | 0.012 | 0.01 |
| Race** | |||
| White | −0.024 | 0.008 | 0.002 |
| Black | −0.015 | 0.010 | 0.16 |
Referent region: Asia
Referent race: Other (Asian, Latino/Hispanic, and Other)
DISCUSSION
In our large, multisite, international cohort of mostly recently diagnosed PLWH, naïve to ART with CD4 cell count >500 cells/μL, we found an overall spirometry-confirmed prevalence of COPD of 6.8%. Importantly, smoking status had a substantial association with COPD prevalence, with current smokers demonstrating a COPD prevalence of nearly 12%. Regression analysis also identified increasing cumulative pack-years of smoking as an independent predictor of lower FEV1/FVC ratio. These data continue to support strongly the need for continued focus on smoking cessation in PLWH.
Our finding that 6.8% of our overall cohort has COPD is within the range reported by other studies. Published data regarding spirometry in PLWH in the current ART era come from nine studies from six countries (Table 5). The four studies from the United States demonstrated COPD prevalence ranging from 7% to 21% in studies with samples ranging from n=98 to n=316. COPD prevalence in PLWH in other countries has been reported as 23% in Italy (n=106), 19% in Denmark (n=63), 17% in Spain (n=275), 10% in Japan (n= 49), and 3% in Canada (n=119). This wide heterogeneity in regional COPD prevalence may be partially due to differing characteristics of each cohort. All previously reported studies have been single-centre studies and may therefore not fully reflect clinical characteristics of PLWH globally. Even within a single city or country, PLWH seeking care at one hospital or clinic may be very different from PLWH in the wider city or country. For example, although the COPD prevalence reported in the cohort study from Baltimore, Maryland, United States was 17% (13), this cohort was selected on the basis of a history of injecting drug use, perhaps explaining the very high prevalence (94%) of current or former smoking. Therefore, while that cohort has provided important data regarding COPD in PLWH, extrapolating these single-centre data to the larger population of PLWH globally, in the United States, or even within Baltimore, should be done cautiously.
TABLE 5.
Summary of regional chronic obstructive pulmonary disease (COPD) prevalence estimates among persons living with HIV in the highly active antiretroviral therapy era.
| Study | Country | Sample size (n) |
Sites (n) | Age, yrs | Female, % |
ART use, % |
Years of known HIV |
Current smoker, % |
Former smoker, % |
COPD prevalence, % |
|---|---|---|---|---|---|---|---|---|---|---|
| Drummond 2013(13) | USA | 316 | 1 | 48.0 ± 6.5 | 34.2% | 55.0% | nr | 84.2% | 9.5% | 16.5% (fixed ratio) |
| George 2009(11) | USA | 234 | 1 | 44.1 ± 9.4 | 17.5% | 83.3% | 8–10 (IQR nr) | 37.1% | 22.6% | 6.8% (fixed ratio) 8.6% (LLN) |
| Gingo 2010(12) | USA | 167 | 1 | 46 (IQR nr) | 26.4% | 80.7% | 13.0 (IQR nr) | 52.7% | 23.4% | 21.0% (fixed ratio) 19% (LLN) |
| Hirani 2011(24) | USA | 98 | 1 | 44.8 ± 11.2 | 16.3% | 87.7% | 8.3 ± 6.5 | 21.0% | 34.0% | 16.3% (fixed ratio)* |
| Samperiz 2013(25) | Spain | 275 | 1 | 48.6 ± 6.6 | 21.8% | 95.6% | 11.9 ± 5.4 | 61.5% | 25.1% | 17.2% (fixed ratio) |
| Cui 2010(26) | Canada | 119 | 1 | 43.4 ± 8.4 | 21.0% | 84.0% | 9.0 ± 6.6 | 43.7% | 19.3% | 3.4% (fixed ratio) |
| Madeddu 2013(14) | Italy | 111 | 1 | 42.3 ± 8.1 | 30.6% | 78.4% | nr | 56.8% | nr | 23.4% (fixed ratio) |
| Kristoffersen 2012(23) | Denmark | 63 | 1 | 43.3 ± 9.0 | 11.1% | 88.9% | 9.3 ± 5.1 | 47.6% | nr | 9.5% (fixed ratio) at baseline; 19.0% at 4.4 year follow-up |
| Nakamura 2014(7) | Japan | 49 | 1 | 40 (IQR nr) | 0.0% | 97.9% | nr | 44.9% | 16.3% | 10.2% (fixed ratio) |
| START Pulmonary Substudy | Multiple | 989 | 80 | 36 (30, 44) | 29.4% | None | 1.2 (0.4, 3.5) | 27.5% | 11.0% | 5.5% (fixed ratio) 6.8% (LLN) |
| Africa | 322 | 7 | 37 (32, 44) | 64.0% | “ | 1.5 (0.5, 4.8) | 14.0% | 5.9% | 5.0% (fixed ratio) 7.8% (LLN) |
|
| Asia | 102 | 8 | 36 (30, 41) | 26.5% | “ | 0.8 (0.2, 3.4) | 19.6% | 9.8% | 0.0% (fixed ratio) 2.0% (LLN) |
|
| Europe/Israel/Australia | 298 | 35 | 38 (31, 45) | 8.7% | “ | 1.2 (0.5, 3.5) | 44.0% | 14.1% | 9.1% (fixed ratio) 9.1% (LLN) |
|
| Mexico/S. America | 182 | 10 | 34 (29, 40) | 12.6% | “ | 0.6 (0.3, 2.0) | 26.9% | 14.8% | 2.7% (fixed ratio) 3.3% (LLN) |
|
| USA | 85 | 20 | 36 (29, 45) | 10.6% | “ | 1.5 (0.4, 4.6) | 32.9% | 12.9% | 7.1% (fixed ratio) 8.2% (LLN) |
FEV1/FVC <0.7 and FEV1 <80% of predicted
ART=antiretroviral therapy, COPD=chronic obstructive pulmonary disease, LLN=lower limit of normal, nr = not reported.
Studies are ordered in decreasing sample sizes of country-level data. Estimates from this current study are shown at the bottom. COPD prevalence reported as both FEV1/FVC <0.70 (fixed ratio) and FEV1/FVC <lower limit of normal (LLN). Continuous variables reported as mean ± standard deviation or median (interquartile range).
Compared to the existing studies, our cohort was not only the largest and included the most study sites and countries, but also had the youngest age and were ART naïve and therefore not subject to potential pulmonary effects of ART. Therefore, while our COPD prevalence estimate was on the lower end of those reported by other studies, we enrolled a young cohort of PLWH with relatively recent diagnosis of HIV infection. Our regression analysis showed a significant independent relationship between increasing age and lower FEV1/FVC ratio, reinforcing the notion that as more PLWH age, COPD is likely to begin to emerge as a major comorbidity.
A major strength of our study was the inclusion of many sites across many regions of the world. We found significant regional differences in COPD prevalence. Some of this may be explained by differences in smoking rates (e.g., 19% current smokers in Asia compared to 45% in Europe/Israel/Australia). However, Africa had the lowest prevalence of current smokers (14%), yet had higher COPD prevalence than regions with more smokers, such as Asia and Mexico/South America. Regional differences persisted after adjustment for smoking status and other COPD risk factors in regression analysis. Among those with COPD, a high proportion (48%) also reported being lifelong nonsmokers. These data suggest that other nontobacco exposures may be playing a role in COPD pathogenesis in this population.
Indoor air pollution is a common cause of COPD (20), and use of biomass fuels to heat the home and cook inside the home may contribute to a significant proportion of COPD in our participants, particularly those from Africa. Our study was not designed to address the topic of indoor air pollution or air quality, so we did not collect baseline information on these factors. Among the participants enrolled from Africa, 64% were women. Women traditionally have more biomass smoke exposure, due to women often performing more indoor labor than men, and some data also suggest that women are more susceptible to COPD than men (21). We found no difference in COPD prevalence by gender, but because 71% of our female sample was recruited from African sites, analyses of gender in our cohort are also significantly influenced by region.
Our primary definition of COPD has limitations. We used the common LLN definition of COPD, which assigns a COPD diagnosis when airflow obstruction (FEV1/FVC ratio) is lower than the fifth percentile of predicted FEV1/FVC ratio. Predicted FEV1/FVC ratio is typically related to factors such as age, height, gender, and ethnicity. International studies have faced challenges due to predictive equations coming from samples with limited ethnic representation. However, we used newly available GLI equations to determine the fifth percentile (18). The GLI project pooled international data to derive unifying prediction equations for spirometry interpretation. Therefore, our lung function measures are based on the best available data for international studies. GLI did not have adequate data from Africa and India, so we can’t determine how applicable our definitions are to those regions. The GLI project noted little ethnic variation in FEV1/FVC ratio, suggesting that ethnicity does not affect our COPD prevalence estimates, but without further data from African and Indian samples, the true LLN in persons from these countries remains unclear.
One advantage of the LLN definition relates to our analysis of COPD by age quartile. The fixed-ratio definition would naturally be expected to show more COPD in older groups, due to normal age-related decline in FEV1/FVC ratio. Our data in Table 3 exemplify this, where a fixed-ratio definition shows a stronger age-related increase in COPD compared to a LLN definition that adjusts for the effect of normal aging. The LLN definition showed a statistically insignificant difference in COPD prevalence by age (p=0.14), although there did appear to be a trend for higher COPD prevalence with each increase in age quartile.
Although our point estimate of COPD prevalence varied by which COPD definition we used, COPD was generally quite mild, with most FEV1 values being generally only mildly impaired (FEV1 ≥80% of predicted). This was also not surprising, given the young age of our study sample. Some data do suggest that PLWH lose lung function faster than predicted by aging alone (13, 22) and one small cohort study (n=63) in Denmark showed that an initial prevalence of COPD of 10% had increased to 19% at a median of 4.4 years of follow-up (23). Such data raise concerns that as PLWH continue to age, COPD may begin to emerge as a major comorbidity, with its attendant impact on quality of life and mortality (2,3). Our longitudinal follow-up, with each person serving as his or her own control, will provide a better study design to understand the effect of aging on COPD prevalence in PLWH.
CONCLUSION
In our novel international spirometry study of relatively young PLWH with CD4 cell count >500 cells/μL and naïve to ART, overall COPD prevalence was 6.8%. COPD prevalence varied substantially across global regions, as did demographics such as gender and smoking behaviour. Our data suggest that smoking and age are important factors related to COPD in PLWH. Smoking cessation should remain a high priority for clinical care and research in PLWH.
The START study (NCT00867048) and the START Pulmonary Substudy (NCT01797367) are registered at clinicaltrials.gov.
Acknowledgments
Funding
The START study is primarily funded by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health under Award Number UM1-AI068641, the Department of Bioethics at the NIH Clinical Center and five NIH institutes: the National Cancer Institute, the National Heart, Lung, and Blood Institute, the National Institute of Mental Health, the National Institute of Neurological Disorders and Stroke and the National Institute of Arthritis and Musculoskeletal disorders. Financial support is also provided by the French Agence Nationale de Recherches sur le SIDA et les Hépatites Virales (ANRS), the German Ministry of Education and Research, the European AIDS Treatment Network (NEAT), the Australian National Health and Medical Research Council, and the UK Medical Research Council and National Institute for Heath Research. Six pharmaceutical companies (AbbVie, Inc., Bristol-Myers Squibb, Gilead Sciences, GlaxoSmithKline/ViiV Healthcare, Janssen Scientific Affairs, LLC, and Merck Sharp and Dohme Corp.) donate antiretroviral drugs to START.
The START Pulmonary Substudy reported here was supported by the National Heart Lung and Blood Institute (R01-HL096453). The Veterans Health Administration Office of Research and Development also provided protected research time in support of this study.
The authors thank the START Pulmonary Substudy participants for their dedicated participation and contribution to our scientific understanding of lung disease in HIV infection. See INSIGHT START Study Group, 2015, this supplement for a complete list of START investigators.
Footnotes
Disclosures
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The views expressed in this article are those of the authors and do not reflect the views of the Unites States Government, the Department of Veterans Affairs, or any of the authors’ affiliated academic institutions. The University of Minnesota, the sponsor of START, receives royalties from the use of abacavir, one of the HIV medicines that can be used in START.
Author contributions:
KMK, DENiewoehner, and JEC were responsible for study conceptualisation.
KMK wrote the manuscript.
GC and JEC performed the primary statistical analysis.
KMK, DENiewoehner, JEC, DENixon, and ET designed the study.
KMK, DE Niewoehner, GC, DENixon, ET, CA, CK, HK, AL, and JEC revised the manuscript critically for important intellectual content, and approved the final version.
KMK, GC and JEC take responsibility for the integrity of the data and the accuracy of the data analysis.
Conflicts of interest: None of the authors have conflicts of interest to declare.
References
- 1.Morris A, George MP, Crothers K, et al. HIV and chronic obstructive pulmonary disease: is it worse and why? Proc Am Thorac Soc. 2011;8(3):320–325. doi: 10.1513/pats.201006-045WR. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lozano R, Naghavi M, Foreman K, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–2128. doi: 10.1016/S0140-6736(12)61728-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2163–2196. doi: 10.1016/S0140-6736(12)61729-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Crothers K, Butt AA, Gibert CL, et al. Increased COPD among HIV-positive compared to HIV-negative veterans. Chest. 2006;130(5):1326–1333. doi: 10.1378/chest.130.5.1326. [DOI] [PubMed] [Google Scholar]
- 5.Crothers K, Huang L, Goulet JL, et al. HIV infection and risk for incident pulmonary diseases in the combination antiretroviral therapy era. Am J Respir Crit Care Med. 2011;183(3):388–395. doi: 10.1164/rccm.201006-0836OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Drummond MB, Kirk GD, Astemborski J, et al. Association between obstructive lung disease and markers of HIV infection in a high-risk cohort. Thorax. 2012;67(4):309–314. doi: 10.1136/thoraxjnl-2011-200702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nakamura H, Tateyama M, Tasato D, et al. The prevalence of airway obstruction among Japanese HIV-positive male patients compared with general population; a case-control study of single center analysis. J Infect Chemother. 2014;20(6):361–364. doi: 10.1016/j.jiac.2014.02.001. [DOI] [PubMed] [Google Scholar]
- 8.Yearsley MM, Diaz PT, Knoell D, Nuovo GJ. Correlation of HIV-1 detection and histology in AIDS-associated emphysema. Diagn Mol Pathol. 2005;14(1):48–52. doi: 10.1097/01.pas.0000142168.72253.11. [DOI] [PubMed] [Google Scholar]
- 9.Twigg HL, 3rd, Weiden M, Valentine F, et al. Effect of highly active antiretroviral therapy on viral burden in the lungs of HIV-infected subjects. J Infect Dis. 2008;197(1):109–116. doi: 10.1086/523766. [DOI] [PubMed] [Google Scholar]
- 10.Kanner RE, Anthonisen NR, Connett JE. Lower respiratory illnesses promote FEV1 decline in current smokers but not ex-smokers with mild chronic obstructive pulmonary disease. Results from the Lung Health Study. Am J Respir Crit Care Med. 2001;164(3):358–364. doi: 10.1164/ajrccm.164.3.2010017. [DOI] [PubMed] [Google Scholar]
- 11.George MP, Kannass M, Huang L, Sciurba FC, Morris A. Respiratory symptoms and airway obstruction in HIV-infected subjects in the HAART era. PLoS ONE. 2009;4(7):e6328. doi: 10.1371/journal.pone.0006328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gingo MR, George MP, Kessinger CJ, et al. Pulmonary function abnormalities in HIV-infected patients during the current antiretroviral therapy era. Am J Respir Crit Care Med. 2010;182(6):790–796. doi: 10.1164/rccm.200912-1858OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Drummond MB, Merlo CA, Astemborski J, et al. The effect of HIV infection on longitudinal lung function decline among injection drug users: a prospective cohort. AIDS. 2013;27(8):1303–11. doi: 10.1097/QAD.0b013e32835e395d. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Madeddu G, Fois AG, Calia GM, et al. Chronic obstructive pulmonary disease: an emerging comorbidity in HIV-infected patients in the HAART era? Infection. 2013;41(2):347–353. doi: 10.1007/s15010-012-0330-x. [DOI] [PubMed] [Google Scholar]
- 15.Babiker AG, Emery S, Fatkenheuer G, et al. Considerations in the rationale, design and methods of the Strategic Timing of AntiRetroviral Treatment (START) study. Clin Trials. 2013;10(1 Suppl):S5–S36. doi: 10.1177/1740774512440342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Miller MR, Hankinson J, Brusasco V, et al. Standardisation of spirometry. Eur Respir J. 2005;26(2):319–338. doi: 10.1183/09031936.05.00034805. [DOI] [PubMed] [Google Scholar]
- 17.Standardization of spirometry, 1994 update. American Thoracic Society. Am J Respir Crit Care Med. 1995;152(3):1107–1136. doi: 10.1164/ajrccm.152.3.7663792. [DOI] [PubMed] [Google Scholar]
- 18.Quanjer PH, Stanojevic S, Cole TJ, et al. Multi-ethnic reference values for spirometry for the 3–95-yr age range: the Global Lung Function 2012 equations. Eur Respir J. 2012;40(6):1324–1343. doi: 10.1183/09031936.00080312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Global Initiative for Chronic Obstructive Lung Disease. [accessed 06 October 2014];Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. doi: 10.1164/ajrccm.163.5.2101039. (updated 2014). Available at http://www.goldcopd.org/ [DOI] [PubMed]
- 20.Salvi SS, Barnes PJ. Chronic obstructive pulmonary disease in non-smokers. Lancet. 2009;374(9691):733–743. doi: 10.1016/S0140-6736(09)61303-9. [DOI] [PubMed] [Google Scholar]
- 21.Gan WQ, Man SF, Postma DS, Camp P, Sin DD. Female smokers beyond the perimenopausal period are at increased risk of chronic obstructive pulmonary disease: a systematic review and meta-analysis. Respir Res. 2006;7:52. doi: 10.1186/1465-9921-7-52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Morris AM, Huang L, Bacchetti P, et al. Permanent declines in pulmonary function following pneumonia in human immunodeficiency virus-infected persons. Am J Respir Crit Care Med. 2000;162(2):612–616. doi: 10.1164/ajrccm.162.2.9912058. [DOI] [PubMed] [Google Scholar]
- 23.Kristoffersen US, Lebech AM, Mortensen J, Gerstoft J, Gutte H, Kjaer A. Changes in lung function of HIV-infected patients: a 4. 5-year follow-up study. Clin Physiol Funct Imaging. 2012;32(4):288–295. doi: 10.1111/j.1475-097X.2012.01124.x. [DOI] [PubMed] [Google Scholar]
- 24.Hirani A, Cavallazzi R, Vasu T, et al. Prevalence of obstructive lung disease in HIV population: a cross sectional study. Respir Med. 2011;105(11):1655–1661. doi: 10.1016/j.rmed.2011.05.009. [DOI] [PubMed] [Google Scholar]
- 25.Samperiz G, Guerrero D, Lopez M, et al. Prevalence of and risk factors for pulmonary abnormalities in HIV-infected patients treated with antiretroviral therapy. HIV Med. 2014;15(6):321–329. doi: 10.1111/hiv.12117. [DOI] [PubMed] [Google Scholar]
- 26.Cui Q, Carruthers S, McIvor A, Smaill F, Thabane L, Smieja M. Effect of smoking on lung function, respiratory symptoms and respiratory diseases amongst HIV-positive subjects: a cross-sectional study. AIDS Res Ther. 2010;7:6. doi: 10.1186/1742-6405-7-6. [DOI] [PMC free article] [PubMed] [Google Scholar]