Non-communicable airway disease and air pollution in three African Countries: Benin, Cameroon and The Gambia

SUMMARY BACKGROUND Air pollution exposure can increase the risk of development and exacerbation of chronic airway disease (CAD). We set out to assess CAD patients in Benin, Cameroon and The Gambia and to compare their measured exposures to air pollution. METHODOLOGY We recruited patients with a diagnosis of CAD from four clinics in the three countries. We collected epidemiological, spirometric and home air pollution data. RESULTS Of the 98 adults recruited, 56 were men; the mean age was 51.6 years (standard deviation ±17.5). Most (69%) patients resided in cities and ever smoking was highest in Cameroon (23.0%). Cough, wheeze and shortness of breath were reported across the countries. A diagnosis of asthma was present in 74.0%; 16.3% had chronic obstructive pulmonary disease and 4.1% had chronic bronchitis. Prevalence of airflow obstruction was respectively 77.1%, 54.0% and 64.0% in Benin, Cameroon, and Gambia. Across the sites, 18.0% reported >5 exacerbations. The median home particulate matter less than 2.5 μm in diameter (PM2.5) was respectively 13.0 μg/m3, 5.0 μg/m3 and 4.4 μg/m3. The median home carbon monoxide (CO) exposures were respectively 1.6 parts per million (ppm), 0.3 ppm and 0.4 ppm. Home PM2.5 differed significantly between the three countries (P < 0.001) while home CO did not. CONCLUSION Based on these results, preventive programmes should focus on ensuring proper spirometric diagnosis, good disease control and reduction in air pollution exposure.

Asthma is a highly prevalent chronic respiratory disease affecting 4.4% of adults worldwide, which translates to about 343 million people in 2020. 1 Using the Global Initiative for Chronic Obstructive Lung Disease (GOLD) definition, 212.3 million people had chronic obstructive pulmonary disease (COPD) worldwide in 2019. 2 From 1990 to 2019, the global prevalence of COPD decreased by 8.7%.During the same period, the number of deaths due to COPD had decreased by 41.7%.However, most of the deaths occurred in low-and-middle income countries (LMICs). 2 Less is known about the prevalence of asthma and COPD in adults living in sub-Saharan African countries but, in a biomass exposed population in rural Cameroon, the prevalence of COPD was 18.4%, 3 while the prevalence of adult asthma in Cameroon was 3.9%. 4In Benin, a recent study reported an asthma prevalence of 4.1% in the general population. 5Air pollution is an important driver of non-communicable diseases such as stroke, lung cancer, COPD and asthma.It is a major public health concern worldwide.Besides smoking, exposure to occupational pollution, biomass cooking and heating are other important risk factors for COPD. 6Some studies also suggest air pollution contributes to asthma causation. 7urthermore, the WHO estimates that around 7 million people die every year from exposure to fine particles in polluted air. 8,9Global mortality due to ambient particulate matter less than 2.5 lm in diameter (PM 2.5 ) increased from 1990 to 2015.Attributable deaths rose from 3$5 million in 1990 to 4$2 million in 2015, representing 7.6% of total global deaths, with ambient PM 2.5 being the fifth-ranking mortality risk factor.Air pollution has been reported to exacerbate existing CAD, leading to increase of admissions in the emergency department. 10,11ome studies have assessed the outdoor air quality index (AQI) in the living area of patients with CAD in order to help them to regulate their daily activity according to the local AQI report and help prevent exacerbation. 11In sub-Saharan Africa, studies assessing air pollution in patients with noncommunicable chronic airway diseases such as asthma and COPD are scarce.We present here one of the first studies to have related air pollution exposures in people with CAD in three sub-Saharan countries in West Africa.

Study design and setting
The study was a multicentre clinic-based crosssectional study that recruited from four clinics in three sub-Saharan African countries: Benin, Cameroon and Gambia.

Study setting
Patients were recruited from 1) National Teaching Hospital of Pulmonology, Cotonou, Benin, a tertiary care and teaching centre; 2) Douala General Hospital, Cameroon, a tertiary care and teaching centre; 3) Kanifing General Hospital, The Gambia, a secondary care hospital; and 4) Fajara site of the Medical Research Council Gambia at London School of Hygiene & Tropical Medicine, a tertiary facility serving the country.
Benin is a West African country of 115,000 km 2 with a population of 13 million and a gross domestic product (GDP) of $11.4 billion. 12The population prevalence of COPD from the Benin BOLD study was 7.0%. 13Cameroon is a west-central African country bordering the Gulf of Guinea with a GDP of US$44.8 billion and a population of 29.3 million. 14eported asthma prevalence in Cameroon is 3.9% 15 and reported COPD prevalence ranges from 2.4% to 18.4%. 3,11,16The Gambia is the smallest country in mainland Africa (10,689 km 2 ), with a population of 1,900,000. 16There is paucity of CAD data.Prior to this study, none of these clinics had routine spirometry services.
During a 13-month period (August 2019 to March 2020 in Benin and Cameroon, November 2020 to April 2021 in Gambia), outpatients aged �18 years with a physician diagnosed chronic airway disease (asthma, COPD, chronic bronchitis) were recruited consecutively.

Data collection and quality control
An interviewer administered a respiratory questionnaire based on the Burden of Obstructive Lung Disease (BOLD) questionnaire, with additional questions developed as part of the National Institute for Health Research International Multidisciplinary Programme to Address Lung Health and TB in Africa ''IMPALA'' Lung health across the Life Course (LuLi) initiative. 17[19]

Spirometry
This was performed according to the Pan African Thoracic Society (PATS), American Thoracic Society (ATS)/European Respiratory Society (ERS) standards by trained and PATS certified technicians. 20The EasyOne ® spirometer (ndd, Z ürich, Switzerland) with a daily calibration check was used.The procedure included at least three acceptable and repeatable forced vital capacity manoeuvres out of up to eight performed while sitting.Inhaled salbutamol 400 lg was administered via a spacer device and postbronchodilator (BD) blows executed 15 min later.Manual overreading of all the traces was performed.Spirometric outcomes were compared against African American predictive Global Lung Function Initiative (GLI) 2012 equations. 21The main spirometric parameters recorded were the forced expiratory volume in one second (FEV 1 ), the forced vital capacity (FVC) and the ratio (FEV 1 /FVC).As per GOLD guidelines, post-BD airflow obstruction (FEV 1 /FVC , 0.7) was categorised into mild (FEV 1 �80% predicted), moderate (50% � FEV 1 , 80% predicted), severe (30% � FEV 1 , 50% predicted) and very severe (FEV 1 ,30% predicted). 20,21ormal spirometry was defined as FEV 1 and FVC .80%predicted with FEV 1 /FVC.0.7.A restrictive pattern was defined as FVC ,80% predicted with FEV 1 /FVC .0.7. 21me air quality monitoring Twenty-four-hour home measurements of particulate matter 2.5 lg/m 3 (PM 2.5 ) were performed with RTI MicroPEM� (RTI MicroPEM; Diamond Bar, CA, USA) (Cotonou and Douala) and PurpleAir-II-SD (Fajara and Kanifing).Gravimetric calibration for quality control was conducted monthly for the RTI MicroPEM, and two-weekly routine maintenance for the PurpleAir-II-SD (PurpleAir; Draper, UT, USA) sensor. 22Carbon monoxide (CO) measurement was performed with Lascar ® (Lascar Electronics, Whiteparish, England, UK) loggers.Measurement of the air pollution biomarker: exhaled carbon monoxide (eCO) was with a pre-calibrated CO check plus device.The latter is a short-term (,2 h) biomarker of exposure both outdoor and indoor air pollution exposure. 23

Non-communicable airway disease
Second-hand smoke exposure was assessed with the following questions: "Does anyone smoke cigarettes or tobacco inside the building where you sleep (do not include yourself)?"and "Does anyone smoke cigarettes or tobacco inside any building where you work or spend other time?" and "Are you ever exposed to smoke from burning refuse (waste, rubbish)?".Traffic exposure was assessed with: "How many hours of your average day do you spend on the roadside of a major road?(in hours)" and "How many hours of your average day do you spend travelling on a main road?(include walking)".Also, work exposure was assessed with: "In your job, do you breathe in vapours, dusts, gases or fumes for more than 15 h per week?"Finally, Crop cultivation and pesticide use were assessed with: "Do you cultivate crops?" and "Do you use pesticides in cultivation of your plants?"Statistics A sample of 50 patients was planned from each centre based on feasibility, and timeline to get sufficient preliminary data as a prelude to larger, more definitive studies.
The primary exposure variables are home air pollution exposure levels: PM 2.5 in lg/m 3 , CO in parts per million (ppm) and exhaled CO (in ppm and percentage of carboxyhaemoglobin).Values were presented as median with interquartile ranges (IQRs).
Secondary exposure variables included: exposures at home/daily life, traffic and occupational exposure.Primary outcome is the presence of CAD.Secondary outcomes are disease type based on spirometry test (e.g., chronic airflow obstruction, restriction etc), measure of disease control/ impact of control (ACT, CAT). 18,19etween-group comparison were performed using v 2 , Fisher's Exact test, t-tests or analysis of variance, as appropriate; P , 0.05 was considered statistically significant.Analyses was performed with R statistical package (R Computing, Vienna, Austria).

Ethical considerations
The

Sociodemographic characteristics of study participants
Benin patients were the oldest and patients in Cameroon were the youngest.Most patients lived in cities.Most Gambian patients were in employment (70.0%), while Benin patients were the least likely employed (29.0%).The rates of ever smoking were similar between the three countries (Table 1).Workplace particulate matter exposure was highest in Cameroon (71.0%) and lowest in Gambia (26.0%).Reported passive smoking exposure was 38.0% in Cameroon, 26.0% in Gambia and 14.0% in Benin; 54% of the Cameroonian cohort spent more than one hour in traffic daily, while only 4.0% of Gambian participants reported traffic exposure greater than 1 h (Table 1).Aerosol use in the home was reported by respectively 16 (46.0%),10 (77.0%) and 18 (36.0%) of the respondents in Benin, Cameroon and Gambia.

Clinical characteristics of study participants
The most frequently reported symptom across all sites was wheeze (55.0%); shortness of breath was least reported (30.0%); 42% of patients reported cough.Asthma was the most common physician diagnosis across all sites (80%), followed by COPD (16.0%), and then chronic bronchitis (4.0%).Comparatively, Benin had the highest proportion of people with asthma (83.0%), followed by Gambia (80.0%).Cameroon had the highest proportion of people with COPD (23.1%), followed by Gambia (18.0%).Over half of the cohort had at least one exacerbation in the preceding year.In the previous year, �1 exacerbation was reported by 79.5% of the people with asthma and 68.7% of COPD patients, with no statistical difference in the rates of asthma exacerbation (P ¼ 0.255) and COPD exacerbation (P ¼ 0.519) across all sites.Furthermore, about twothird (66.0%) of the people with asthma in this cohort have poor disease control (ACT score �19).In 91.0% of those with COPD, the disease had at least a moderate impact on their lives.

Disease exacerbation, disease control and spirometry patterns by respondents' diagnoses among all cohorts across the sites
About half (52.1%) of the patients with asthma and three-quarters (75.0%) of the obstructive airway disease patients had FEV 1 ,80% predicted.Nearly 60% of respondents with physician diagnosis of asthma fulfilled the GOLD definition of COPD, whereas 44.0% of those with diagnosed COPD did not meet the GOLD spirometric definition of COPD of FEV 1 /FVC,0.7 (Table 3).

DISCUSSION
This is, to the best of our knowledge, the first study comparing non-communicable airway disease and air pollution exposure in multiple West African countries.
In this cross-sectional study, we found that most patients living with CAD attending these city-based clinics are usually males, older than 50 years, residing in urban settlements, with up to three-quarters being symptomatic: cough, shortness of breath and wheeze.Asthma was the most diagnosed CAD across the countries, followed by COPD.
A notable finding of this study is that 58.0% of those diagnosed with asthma fulfilled the spirometric criteria for COPD, while 44.0% of those with physician diagnosis of COPD do not meet the spirometric criteria for COPD.This finding is similar to that reported by Binegdie et al. in their study of 519 chronic respiratory disease patients in outpatient clinics in Kenya, Ethiopia and Sudan.They found that 39.0% of clinician-diagnosed asthma patients fulfilled the criteria for COPD.Furthermore, despite the physician diagnosis of COPD in 7.9% of the patients, spirometry consistent with COPD was present in 35.0%. 24Possible explanation for these findings include: chronic airway remodelling in asthma and asthma overdiagnosis due to misdiagnosis of COPD as asthma because of reliance on clinical features in the absence of spirometry.Similarly, COPD underdiagnosis is a consequence of lack of spirometry and reliance on clinical suspicion.Clinical diagnosis without spirometry is not reliable in establishing chronic airway disease.
The overall prevalence of physician-diagnosed airway obstruction in this study was 100%.The overall spirometry confirmed prevalence of airway obstruction was 67.3%.This highlights the tendency of clinical diagnosis to over-estimate obstructive lung disease.The place of spirometry in establishing the diagnosis of chronic obstructive airway disease cannot be over-emphasised. 25he Gambia had majorly mild and moderate CAD severity, while Benin and Cameroon reported severe and very severe CAD in their respondents.Across all the sites, 11.8% of asthma patients and 38.0% of COPD patients have at least severe airway obstruction.About two-third (66.0%) of the asthma patients had poor disease control (ACT score �19).In 91.0% of those with COPD, the disease had at least a moderate effect on their lives.The finding of poor disease control also featured in a North African study of over 7,000 eligible people with asthma, where only 16.0% were found to have good asthma control (mean ACT score of 21.8). 26,27In addition, a study of 449 adults with asthma in Uganda by Kirenga et al. highlighted the fact that 28.1% (n ¼ 126) were uncontrolled with an ACT score of ,15. 28e found the between country differences in PM 2.5 median exposure among asthmatic respondents was statistically significant (P , 0.001), while PM 2.5 of COPD patients and CO exposure of asthma patients were statistically insignificant (P ¼ 0.197 and P ¼ 0.173, respectively).Also, there was no significant association between air pollution and disease exacerbation, asthma control and the impact of COPD on patients' lives.In the absence of very similar studies in sub-Saharan Africa, Dionisio et al. reported fixed site household mean PM 2.5 of 275 lg/m 2 and 219 lg/m 2 for the mothers and the children in their study in Gambia. 29The CO averages were respectively 2.4 ppm and 1.5 ppm in mothers and children.This study and ours were both urban in location but the latter did a 48-hour measurement while we measured for 24 hours. 30In addition, there was a 13-year difference which would likely affect the energy source used at home then as compared with now.
The proportion of ever smoking tobacco, passive smoking and home aerosol use (atomised spray chemicals such as mosquito insecticides and air fresheners) were highest in Cameroon (23.0%, 38.0% and 77.0%, respectively).Ever smoking was 23.0% or less at each site among the respondents.Most of the respondents spend significant time in and around their homes as they are mainly traders and technicians, thus exposing them to locally used aerosols and fumes from the mostly biomass-powered kitchen which is usually near the house. 29,31The study by Binegdie et al. in East Africa reported that 83.0% of the respondents never smoked but 19.0% and 32.0% were exposed to passive smoking at home and at work. 24ur study was not without limitations.The total number of respondents were less than 100, thus affecting the power of the study and limiting the generalisability to the cohort studied.In addition, the instruments used in The Gambia differed from those used in Benin and Cameroon: PurpleAir-II-SD and Lascar CO devices in Gambia and, RTI Micro-PEM and Lascar CO devices at the other two locations.This device difference between the sites can affect data comparability.However, RTI MicroPEM uses both gravimetric and optical laser methods to measure fine particulate matter while PurpleAir uses optical laser method, thus providing some methodical overlap for the two equipments.We recognise the latter as a significant limitation of our study.
The study had some strengths: it was a multicountry, multi-locational study in an environment with paucity of information on CAD, thus allowing for inter-country comparisons.Furthermore, PurpleAir-II-SD outside/ambient device was used for home/ indoor air pollution measurement in our Gambian cohort, a use to which it is not usually put to.We had the confidence to apply this innovation because the equipment has been shown to perform consistently and accurately at very high and very low particulate matter exposure. 32This has added another affordable tool for air quality researchers to measure indoor or home PM 2.5 exposure measurement, especially in resource-constrained settings like sub-Saharan Africa.

Table 1 .
Socio-demographic characteristics of study participants.

Table 2 .
Clinical characteristics of study participants.

Table 3 .
Disease exacerbation, disease control and spirometry pattern by respondents' diagnoses among all cohorts across the sites. 22