Pulmonary Function Tests and Their Associated Factors in HIV-infected Patients at Jimma Medical Center; Ethiopia: A Comparative Cross-Sectional Study

Background HIV-infected persons have a greater risk of developing respiratory disorders. Poor socio-economic status, High viral load, low CD4 counts, and anti-retroviral therapy are linked with the problems. Despite its high prevalence the association between retroviral infection and pulmonary function status as well as their associated factors has not yet well established in resource-scarce countries in general and study setting in particular Methods A comparative cross-sectional study was conducted from September 24 to October 15 of 2020 at Jimma Medical Center among HIV-positive patients and matched control group. Data were collected using a pretested structured questionnaire administered via face-to-face interview. The collected data included Socio-demographic, respiratory, retroviral infection, and substance use related. Pulmonary function tests were also conducted using SP10 spirometer. Collected data were entered and analyzed using SPSS version 26. Independent t-test and multiple linear regressions were carried out to identify factors independently associated with the pulmonary function status of the study participants while controlling for potential confounders.


Introduction
Acquired immune de ciency syndrome is a syndrome caused by a retrovirus that predominantly targets and destroys CD4 T cells [1,2].It is a pandemic that affects 36 million people worldwide [3].
The lung is a chief target organ for retroviral infection, rendering it vulnerable to an inclusive array of infectious and noninfectious complications [4].Innovation and provision of highly active antiretroviral treatment (HAART) for Retroviral infected patients has been a new horizon in the reduction of Acquired immune de ciency Syndrome related mortality and prolong life expectancy [5].As a result, the spectrum of infectious and noninfectious diseases signi cantly altered [6].
The incidence of non-communicable diseases particularly, chronic non-infectious pulmonary diseases is dramatically increased in RVI patients in the aftermath of HAART [7].Obstructive lung diseases (OLD) especially, chronic obstructive pulmonary diseases (COPD) and asthma are an emerging pulmonary burden [8].Chronic lung diseases are categorized based on the pattern of anomaly detected with pulmonary function test indices.
The cause of Obstructive lung disease (OLD) in RVI patients is to date an obscure concept [8].The prevalence of tobacco use in RVI patients is higher than the prevalence in the general USA population.Despite this fact, HIV infection is independently associated with OLD incidence and prevalence without smoking exposure.Such query highlight OLD in RVI patients is mechanistically beyond the risk factor that has been considered in the general population [5].
Recent studies try to link the physiological mechanism between pulmonary diseases and RVI deep down to the defense mechanism of the pulmonary system.HIV persistently infected alveolar macrophage-more often smaller alveolar macrophages.As a result, alveolar macrophage function altered and secretes dozen of cytokine and chemokine which calls deregulated Cytotoxic T-cells.The alveolar airspace is in ltrated with malfunctioned cytotoxic T-cells that intensify the activation and reactivation of proteases.
Proteases disrupt the extracellular matrix component elastin and collagen which end up with alveolar parenchyma destruction [6].HIV also associated with a rapid decline of muscle mass and strength of the respiratory muscle due to the chronic in ammatory process and chronic malnutrition [9] Socio-demographic factors (age, sex), history of pulmonary tuberculosis, smoking and substance use, time of ART initiation, CD4 count, viral load, and HAART have been crudely associated with pulmonary disease development and progression in RVI patients [10].Despite these facts, the screening, diagnosing, and managing of noninfectious chronic lung diseases by interpreting pulmonary function test indices has been underemphasized in people living with HIV AIDS [5].
Despite the emergent of the RVI outbreak in Ethiopia, to date, scarce data is found to associate RVI and pulmonary function tests pattern.
In this regard, this study will ll this gap by determining pulmonary function tests and associated factors among RVI patients, and data from this study will be an input to establish a routine pulmonary function test platform for PLWH.

Study area
This study was conducted in Jimma University Medical Center (JUMC), from September 24 to October 15 of 2020.JUMC is located in Jimma city 352 km southwest of Addis Ababa.It provides services for approximately 15,000 inpatient, 160,000 outpatient attendants, 11,000 emergency cases, and 4500 deliveries in a year for patients coming to the hospital from the catchment population of about 15 million people.Currently, 3029 RVI patients are registered and receive service at in ART clinic

Study design
An institution-based comparative cross-sectional study design was used for this study

Source population
The source population was, all RVI patients enrolled in ART clinic for chronic follow up and age-sex matched RVI negative attendants who came to JUMC as a caregiver for their cognate and relatives aged > 18 years.

Study population
All selected RVI patients who visited ART clinic and age-sex matched RVI negative attendants, age > 18 years during the data collection time.

Participant's eligibility criteria
All RVI patients aged > 18 years, and who had regular chronic follow up in ART clinic, HAART treatment for > 6 months, complete personal information, and clinical pro le were recruited into the study.RVI patients who had diagnosed with chronic co-morbidities (hypertension, DM, cancer, PCP, ILD, oral candidiasis, chest deformity, asthma, COPD), pregnancy, < 6 months of HAART TX, severe illness, history of retinal detachment, abdominal surgery < 3 months, chest /abdominal surgery < 3 months, presume COVID -19 as well as.

Sample size determination and sampling technique
The sample size was calculated using the double proportion population formula by taking a proportion of obstructive lung diseases among RVI patients (P1= 10%) and non RVI respondents (P2= 3%) taken from a study done in a rural part of South Africa [10].The con dence interval, level of signi cance, and power of a study was 95%, 5%, and 90% respectively.After considering the 10% non-response rate, the total sample sizes were 192 (96 for a study group and 96 for a comparison group).During the time of data collection, both RVI patients and RVI negative respondents were chosen using a consecutive sampling technique.
Data collection procedure Data were collected using an interviewer-administered structured questionnaire.It had ve distinct parts, socio-demographic characteristics, Respiratory related questionnaire, substance use, anthropometry, and spirometer measurements.RVI related questionnaires were reviewed from study participant medical records.Respiratory-related questionnaires were adapted from ST. George's Respiratory Questionnaire and medical research council's committee (MRC) on environmental and Occupational Health.
Informed and consented respondents from the comparison group were had HIV tests using a rapid HIV diagnostic kit.The procedure and result interpretation was based on the Ethiopian health and nutrition research institute (EHNRE) [11].Weight was measured after the study respondents were barefoot and wear light cloth using a digital weight scale.Height was measured using a stadiometer after the respondents were in a standing position, barefooted and the shoulder was in normal alignment.

Pulmonary function testing procedure
The pulmonary function tests were measured using (CONTEC MEDICAL SYSTEMS CO., LTD, China) SP10 digital spirometer.The procedure, precaution, acceptability, reproducibility, and interpretation of the spirometer output were based on the CDC-National Health and Nutrition Examination Survey of respiratory health, and research article [12].

Infection control
The body of the spirometer and turbine was cleaned using 70% isopropyl alcohol at the end of each procedure.The data collectors were worn N-95 facemask after demonstrating the maneuver for study respondents and powder-free disposable gloves.Each study participant was having their mouthpiece with a bacterial viral lter which was discarded by themselves at the end of the procedure.In line with COVID-19 prevention, all preventive measure and recommendation emanate from World Health Organization was kept in practice.

Instrument calibration
The calibration of the instrument was based on SP-10 pocket spirometer user Manuel which was prepared by the manufacturer.Under the calibration interface veri cation was checked for pre-calibration with a 3L syringe.
Spirometer examination was done in a quiet room in a sitting position by quali ed personnel.The procedure was done in the morning between 8:00 AM to 12:00 AM daily at room temperature.Participants' ID, age, sex, weight, height, race, and smoking status was feed into SP10 spirometer for each participant before the maneuver.A very explicit instruction was delivered to the study participants until a common understanding was grounded for spirometer examination.Well thought of the purpose of the study, emphasize the need for extra effort from them to get a maximal result, avoidance of tight clothing and denture was encouraged.

Maneuver
The study participant was sitting upright with chin elevated and neck extended slightly, Feet at on the ground with uncrossed leg and Hold their nostrils tightly by left/right thumb and index ngers.Then take a big deep breath and ll their lung with air and Hold the spirometer, place the mouthpiece into the mouth above the tongue and between the teeth.The respondent Blast out the air as hard and as fast as possible.Keep blowing out for the rst 6 seconds to empty the lung.The procedure was repeated until three acceptable and reproducible measurement was obtained for a maximum of eight attempts.The procedure was stopped and rescheduled if a respondent was unable to produce an acceptable and repeatable spirometer output after eight attempts.
Data were collected by four BSc nurse professionals who had previous experience in data collection.

Operational de nition
Respiratory symptom -Saint George respiratory questionnaire score >24 is considered a prevalent respiratory symptom Alcohol taking -drinking ≥ 2 bottles of beer per drinking session for males and ≥ 1 bottle of beer per drinking session for females ≥ 3 sessions/week within 1 month before the study.
Khat chewer -chewing ≥ 1 bundle of khat leaves per chewing session ≥ 3 days/week with 1 month before the study.
Using SPSS, continuous variables were summarized as mean and standard deviation using descriptive statistics.Categorical variables were summarized as frequency and percentage using cross-tabulation.
An Independent T-test was used for the comparison of the mean of pulmonary function test indices between RVI and non-RVI study respondents.The assumption of independent sample t-test was checked using Shapiro-Wilk and Levine's tests.
A simple linear regression was employed to determine the factors which predict pulmonary function tests indices of RVI respondents.Variables that were signi cant predictors of PFTs in simple linear regression were entered into multiple linear regression.
Multiple linear regression analysis were utilized to obtain the best t linear combination of sex, weight, BMI, respiratory symptom, history of pulmonary TB, duration of RVI, duration of treatment, current CD4 cell count, alcohol taking, and khat chewing which were signi cant in simple linear regression.
The assumption of multiple linear regression (linearity, normality, homoscedasticity, outliers, and multicollinearity) was checked and met.Normality was checked using histogram, P-P, and Q-Q plot.
Multicollinearity was checked using variance in ation factors (VIF) and VIF >5 shows an association between predictor variables.Recruited variables in multiple linear regressions which had a P-value of ≤ 0.05 were said to be a signi cant predictor of pulmonary function.

Data quality management
To assure the clarity of the questionnaire pretest was done on 5% of the study respondents in shanen gibe hospital.Training for data collectors and supervisor were given for two consecutive days about the purpose of the study, interview, measurement technique, and ethical issues by the principal investigator.Questionnaires were translated to Afan Oromo and Amharic language and retranslated back to English for consistency purposes by language experts.

Results
Socio-demographic characteristics of the respondents One hundred and ninety two respondents (96 = RVI and 96 = non RVI) were recruited in the study.Of them 65% were females.The mean age of RVI and non-RVI respondents were 37.56 (SD =±7.017) and 36.06 (SD =±7.20) respectively.Regards to the level of education, 3.1% and 4.7% of RVI respondents and 7.3 and 15.1% of non-RVI respondents were unable to read and write and above diploma respectively (table 1) Twenty-ve percent of RVI respondents were former smokers and one percent of non-RVI respondents were current smokers.The proportions of khat chewing and alcohol taking in RVI and non-RVI respondents were (9.4% vs 11.5%) and (11.5% vs 8.3%) respectively (Table 2).

Simple linear regression presenting predictors of PFTs among RVI
A simple linear regression was employed to determine the factors which predict FVC (l), FEV1 s (l) , FEV 1 /FVC, FEF 25-75 (l/s), and PEFR (l/s) of RVI respondents.As result, a simple linear regression model revealed a signi cant association between the PFTs and sex (male being used as reference), weight, BMI, respiratory symptom (SGRQ>0 being used as reference), history of pulmonary TB (yes being used as reference), duration of RVI, duration of treatment, current CD4 cell count, alcohol taking (yes being used as reference) and khat chewing (yes being used as reference).
Conversely, age, height, current viral load, history of pneumonia, drug adherence, type of treatment regimen, and history of smoking were not statistically signi cant predictors of pulmonary function tests among RVI patients.

Multiple linear regression model presenting predictors of PFTs among RVI respondents
Multiple linear regression analysis was utilized to obtain the best t linear combination of sex, weight, BMI, respiratory symptom, history of pulmonary TB, duration of RVI, duration of treatment, current CD4 cell count, alcohol taking, and khat chewing which were signi cant in simple linear regression (p<0.05).
Being a male has 0.761 liters greater FVC (l) than being a female.Similarly being a male has 0.69 liters greater of FEV1 s (l) than being a female.
Having frequent respiratory symptoms decreases FVC (l) by 0.38 liters compared to devoid of respiratory symptoms.Respondents who reported respiratory symptom decreases FEV1 s (l) by 0.378 liters compared to devoid of respiratory symptom.Similarly, having frequent respiratory symptoms decreases PEFR by 0.94(l/s) compared to devoid of respiratory symptoms.
Participants who reported a history of pulmonary TB decreases FVC (l) by 0.551 liters compared to those who never report it.Equivalently, participants who report a history of TB was had FEF 25-75 of 0.9 l/s less compared to respondents who never reports.

Discussion
In the current study, the mean of FEV1 was signi cantly reduced in RVI study participants than non-RVI participants.This report is in concordant with a study done in a rural part of South Africa [10] and a study done in Copenhagen [13].This observation might be due to HIV persistently infected alveolar macrophage-more often smaller alveolar macrophages.As a result, alveolar macrophage function altered and secretes dozen of cytokine and chemokine which calls deregulated Cytotoxic T-cells.The alveolar airspace is in ltrated with malfunctioned cytotoxic T-cells that intensify the activation and reactivation of proteases.Proteases disrupt the extracellular matrix component elastin and collagen which end up with alveolar parenchyma destruction.Physiologically, the recoil force of the alveoli is signi cantly reduced and the compliance of alveoli enhanced and expiratory work of breathing is increased.As a result of this, pulmonary function test indices might be declined [6] Conversely, this nding is contradicted with the research done in Amsterdam Netherlands, which reported the mean of FEV1 was not signi cantly different between RVI and non-RVI respondents [9].The reason might be due to relatively high smoking status and alcohol abuse in the controls [9].Moreover, other socio-demographic variations and methodological choices between the two groups might be perpetuating the difference.
Our ndings revealed that the mean of FVC was signi cantly reduced in RVI study participants than non-RVI respondents.This observation is in line with a study done in Cape Town of South Africa and Amsterdam [9,14].This might be linked to the fact that HIV had been associated with the rapid development of sarcopenia [9].Sarcopenia is usually a disease of geriatrics concerned with progressive loss of skeletal muscle mass and muscle strength, despite this fact, it occurs at a young age in RVI patients [15].The reason behind this might be due to persistent chronic immune activation and in ammation.As results of this, the muscle of expiration loss their mass and strength to perform forced expiratory maneuvers which end up with reduced FVC and a high ratio of FEV1/FVC [16].
In the current study, being male was a positive predictor of FVC (l), FVC (%), FEV1 (l), and FEV1%.This nding was consistent with a study carried out in the USA [17].The reason might be since the male has by far a bigger chest cavity, bigger lung, large diaphragm, large number of alveoli and alveoli surface area, and high recoil pressure over the same stature as female.Those physiological advantages are helpful during forced expiratory maneuvers [18].Our observation went against a study done in South Africa and Pittsburg [10,19].Such discrepancy might be due to variability in sample size, substance abuse, and occupational exposure between males and females.
In the present study, weight was a negative predictor of PEFR.This might be due to increment in weight decrease chest and lung compliance which negatively affects pulmonary function parameters [20] Similarly, our study reveals that BMI was a positive predictor of peak expiratory ow rate.This observation was in line with a study carried out in Harare of Zimbabwe [21].The scienti c argument behind this observation is HIV had been associated with chronic under-nutrition.The primary cause of under-nutrition is anorexia secondary to profoundly elevated pro-in ammatory cytokine (IL-1, 6, TNF-α) [22].This results in inadequate energy intake coupled with energy catabolism.Entirely, respiratory muscle mass and strength are weakening and maximal expiratory pressure is weakening to achieve peak expiratory ow rate [22].
This observation went against the study done in South Africa [23].This might be due to the variability socio-demographic and nutritional status of the participants.
The current study also reported that the duration of HAART treatment was a negative predictor of FEV1 (l) and predicted FEV1 %.This result goes inconsistent with a study done in Minneapolis of USA and Pittsburg [17,24].This might be due to HAART induces oxidative stress by triggering massive production of reactive oxygen species and inhibit antioxidant production in many cells of a pulmonary system [25].
The Oxidative stress-causing effect of HAART has been positively associated with increasing the duration of treatment.As remarked in the above statement oxidative stress damage airways and alveolar parenchyma.In addition to this, anorexia is an unavoidable adverse effect of HAART which is contributing factor for chronic under-nutrition.So, the strength and mass of respiratory muscle are lost to generate adequate expiratory pressure [22].
On the contrary, this nding went against the study done in Zimbabwe [21] which says early initiation of ART prevents the decline of FEV1.The reason is probably due to the difference in duration of treatment and treatment regimen.
The nding of the present study also reveals that the Duration of RVI is a negative predictor of FEV1 and FEV1%.This observation is consistent with a study carried out in Zimbabwe and Denmark [13,26].The reason for this observation is quite multifactorial, in the aftermath of HAART provision, RVI patients are living with a low level of viremia.This consistent low level of viremia induces persistent long-term in ammation which causes oxidative stress, disruption of mucociliary clearance mechanism, sarcopenia, airway remodeling, and undernutrition [9,14,21,27].On top of this, the duration of HIV infection is associated with the reduction of serum alpha-1-antitrypsin concentrations which are serine protease inhibitors (protect extracellular matrix of alveolar parenchyma) [28,29].Taken together, the duration of HIV infection on a pulmonary system is manifested as impaired pulmonary function tests and pulmonary diseases.This nding went against research carried out in South Africa which dictates that the duration of HIV has no association with the decline of pulmonary function in RVI respondents [10].This is possibly due to variation in sample size, duration of HIV infection, and other RVI characteristics.
Current CD4 cell counts were positive predictors of pulmonary function in our ndings.The nding is in line with research carried out in France [30].The primary role of ART is viral suppression and T cell function improvement [31].
CD4 cell count increment is immunity to pulmonary function by preventing bronchial colonization, secondary in ammation, and orchestrates in ammatory response to lessen persistent viral in ammation and immune activation to airways and alveolar parenchyma [30,32].The nding is contradicted with a study done in the USA, Pittsburg, South Africa [10,19,24,33].This is might be due to dissimilarity in methodology, current CD4 count, current viral load, ART uses, and high smoking status.History of TB is the negative predictor of FVC (l), FVC%, and FEF 25-75 .This study is in line with a study done in South Africa [10,14,33].The reason for such observation is in TB endemic area TB follows the footsteps of HIV and vice versa [33].Worldwide nearly, 30% of RVI patients estimated to have TB and 30-40% in Africa [10].In Ethiopia, up to 52% of RVI patients estimated to have TB and vice versa [34].Both RVI and TB are chronic in ammatory disorders which trigger persistent and consistent in ammation and immune reactivation of pulmonary defense mechanism [35].
Similarly, a nding of the present study shows the presence of respiratory symptoms was a negative predictor of FVC (l), FVC%, FEV1 (l), FEV1%, PEFR, and PEFR%.This nding is in agreement with a study carried out in Nigeria and Baltimore [36,37].It is probably because HIV as a virus is associated with persistent and consistent immune-activation; HIV-activated deregulated adaptive immune response and lifetime low-grade CD8-T cells lymphocytic alveolitis might contribute to this association [37,38].
Our nding also shows that khat chewing was a negative predictor f PEFR and predicted PEFR.The association between khat chewing and pulmonary function impairment was not elucidated very well so far.As a researcher, it can be suggested that the negative association between khat chewing and reduced pulmonary function test is probably because khat chewing could interfere with drug adherence and feeding appetite (anorexia) of RVI patients.This speculation got a fulcrum from a study done in Jimma which says -that chewers have a 4.2 times more likelihood of missing at least one dose of HAART medication than ever chewers of RVI patients [39].
Finally, the present study revealed that Alcohol taking is a negative predictor of FEF 25-75 , PEFR, and predicted PEFR.This study is in agreement with a study done in Barcelona [40].The general argument for this observation is Alcohol consumption is associated with increase HIV-1 replication and decrease HAART response [41].Apart from this, alcohol consumption impairs alveolar epithelial cell surfactant production, mucociliary transport system, Lowers Glutathione Levels, fosters oxidative injury, and acts as respiratory depressant [40].
As part of our strength -this paper will be the rst paper that studied the effect of RVI on pulmonary function and compare it with their age-sex matched non-RVI respondents.We extensively explore the possible risk factors which affect pulmonary function including khat and alcohol uses, RVI characteristics, drug adherence, and respiratory symptom.By doing so, the current study contributes its part to the present understanding of RVI and its impact on pulmonary function.Similarly, the current study highlights possible predictors of pulmonary function in RVI patients.

Conclusion
The current study reveals pulmonary functions were signi cantly reduced in RVI respondents compared to age-sex matched non-RVI respondents.Sex, weight, BMI, duration of RVI, duration of HAART treatment, current CD4 cell count, khat chewing, alcohol taking, respiratory symptom, and history of TB were signi cant predictors of pulmonary function in RVI patients.Current viral load, type of treatment regimen, history of pneumonia, and degree of adherence had no signi cant association with pulmonary function of RVI participants.Based on the current results of the study, all RVI patients should be screened for pulmonary function at their initial visits and every 2 years to detect a change in pulmonary function and all RVI patients with long duration of RVI, longer duration of RX, low CD4 cell count, history of TB and khat chewing should be screened for a pulmonary function to verify their lung function status.

Table 1 :
Socio-demographic characteristics of RVI and non-RVI respondents in Jimma town, Southwest Ethiopia, 2020.

Table 2
Anthropometric, medical history, and Substance use-related characteristics of RVI and non-RVI respondents in Jimma town, southwest Ethiopia, 2020.

Table 4 :
Comparison of mean of pulmonary function indices in RVI and non-RVI participants in Jimma town, Ethiopia, 2020.

Table
Predictors of pulmonary function tests among RVI respondents in Jimma town, southwest