Cohort Profile: Burden of Obstructive Lung Disease (BOLD) study

The Burden of Obstructive Lung Disease (BOLD) study was established to assess the prevalence of chronic airflow obstruction, a key characteristic of chronic obstructive pulmonary disease, and its risk factors in adults (≥40 years) from general populations across the world. The baseline study was conducted between 2003 and 2016, in 41 sites across Africa, Asia, Europe, North America, the Caribbean and Oceania, and collected high-quality pre- and post-bronchodilator spirometry from 28 828 participants. The follow-up study was conducted between 2019 and 2021, in 18 sites across Africa, Asia, Europe and the Caribbean. At baseline, there were in these sites 12 502 participants with high-quality spirometry. A total of 6452 were followed up, with 5936 completing the study core questionnaire. Of these, 4044 also provided high-quality pre- and post-bronchodilator spirometry. On both occasions, the core questionnaire covered information on respiratory symptoms, doctor diagnoses, health care use, medication use and ealth status, as well as potential risk factors. Information on occupation, environmental exposures and diet was also collected. Collaborative research proposals and access to data requests should be submitted to Dr Andre F S Amaral [a.amaral@imperial.ac.uk]. For more information on the BOLD study, please visit [https://www.imperial.ac.uk/nhli/bold]. At the end of the 20th century, chronic obstructive pulmonary disease (COPD) was already considered a leading cause of morbidity and mortality.1–3 Yet, little was known about its prevalence and aetiology, particularly in low- and middle-income countries (LMICs). This information is important to improve the understanding of the impact of the disease on quality of life and health care cost, as well as to identify ways to reduce its risk. The Burden of Obstructive Lung Disease (BOLD) study was set up and launched across several regions of the world as a network of population-based surveys using a standardized protocol. The initial main aims of the study were to assess the worldwide prevalence of chronic airflow obstruction, which is a defining characteristic of COPD, and to identify its main risk factors. The main aims of the follow-up study were: (i) to quantify the rate of lung function decline during adulthood; (ii) to assess the risk factors associated with lung function decline; and (iii) to understand the relationship between lung function and mortality, particularly across different ethnic groups. The baseline study was funded in part by a grant from the Wellcome Trust (085790/Z/08/Z), which supported the coordinating centre in London, UK, and in part by unrestricted educational grants from University of Kentucky, Aventis, AstraZeneca, Boehringer-Ingelheim, Chiesi, GlaxoSmithKline, Merck, Novartis, Pfizer, Schering-Plough and Sepracor, which supported the initial coordinating centre in Portland, OR, USA. Additional support was provided to several sites in the baseline study (please see Funding for details). The follow-up study in LMICs was funded by the UK Medical Research Council (MR/R011192/1) and in European countries by AstraZeneca AB (ESR-17-13417). The baseline study was conducted, between 2003 and 2016, in 41 sites in 34 countries across Africa, Asia, Europe, North America, the Caribbean and Oceania [https://www.imperial.ac.uk/nhli/bold] (Figure 1).4 These were selected to represent most of the regions covered by the Global Burden of Disease Programme,5 while over-representing larger regions such as South Asia and excluding Latin America, which had a separate study (PLATINO).6 Participants were non-institutionalized adults, aged 40 years and over, recruited from the general population around sites with at least 150 000 inhabitants. Sampling varied across sites, with some using simple random sampling and others using either stratified random sampling or cluster sampling (Figure 1). For each site and participant, weights were derived to account for sampling design. Burden of Obstructive Lung Disease (BOLD) study sites At baseline, 77 640 contact attempts were made to recruit participants to the study, but 28 901 were ineligible (either died before clinic/home visit or left catchment area or were under 40 years old or were institutionalized or were untraceable or could not be contacted). Of the 48 739 eligible people, 14 482 (29.7%) were non-responders (actively refused to participate or provided partial data) and 34 257 (70.3%) were responders (completed the core questionnaire and post-bronchodilator spirometry, regardless of quality control score) (Table 1). Overall, at baseline the proportions of males and smokers were slightly higher among responders, and the proportion of people with a diagnosis of respiratory disease was slightly higher among non-responders. Non-responders were also older (Table 2). Numbers of contact attempts, ineligible people, non-responders, responders, response rate and cooperation rate in each Burden of Obstructive Lung Disease cohort site at baseline and at follow-up At baseline: ineligible people were those who died before clinic/home visit, who left the catchment area, who were under 40 years old, who were institutionalized, who were untraceable or who could not be contacted. Non-responders were those who actively refused to participate and those who provided partial data. Responders were those who completed the core questionnaire and post-bronchodilator spirometry, regardless of quality control score. Response rate was defined as the number of responders divided by the number of invites or attempts to contact potential participants. Cooperation rate was defined as the number of responders divided by the number of responders plus the number of non-responders. At follow-up: invites (attempts to contact) were made to those who responded at baseline and had useable spirometry. Ineligible people were those who died between baseline and follow-up, who left the catchment area, who were untraceable or who could not be contacted. Non-responders were those who actively refused to participate and those who provided partial data. Responders were those who completed the core questionnaire. Response rate was defined as the number of responders divided by the number of invites or attempts to contact potential participants. Cooperation rate was defined as the number of responders divided by the number of responders plus the number of non-responders. Numbers of contact attempts, ineligible people, non-responders, responders, response rate and cooperation rate in each Burden of Obstructive Lung Disease cohort site at baseline and at follow-up At baseline: ineligible people were those who died before clinic/home visit, who left the catchment area, who were under 40 years old, who were institutionalized, who were untraceable or who could not be contacted. Non-responders were those who actively refused to participate and those who provided partial data. Responders were those who completed the core questionnaire and post-bronchodilator spirometry, regardless of quality control score. Response rate was defined as the number of responders divided by the number of invites or attempts to contact potential participants. Cooperation rate was defined as the number of responders divided by the number of responders plus the number of non-responders. At follow-up: invites (attempts to contact) were made to those who responded at baseline and had useable spirometry. Ineligible people were those who died between baseline and follow-up, who left the catchment area, who were untraceable or who could not be contacted. Non-responders were those who actively refused to participate and those who provided partial data. Responders were those who completed the core questionnaire. Response rate was defined as the number of responders divided by the number of invites or attempts to contact potential participants. Cooperation rate was defined as the number of responders divided by the number of responders plus the number of non-responders. Selected characteristics comparing non-responders and responders at baseline in the Burden of Obstructive Lung Disease cohort Sites names (n = 41): Albania (Tirana); Algeria (Annaba); Australia (Sydney); Austria (Salzburg); Benin (Sèmè-Kpodji); Cameroon (Limbe); Canada (Vancouver); China (Guangzhou); England (London); Estonia (Tartu); Germany (Hannover); Iceland (Reykjavik); India (Kashmir); India (Mumbai); India (Mysore); India (Pune); Jamaica; Kyrgyzstan (Chui); Kyrgyzstan (Naryn); Malawi (Blantyre); Malawi (Chikwawa); Malaysia (Penang); Morocco (Fes); Netherlands (Maastricht); Nigeria (Ile-Ife); Norway (Bergen); Pakistan (Karachi); Philippines (Manila); Philippines (Nampicuan-Talugtug); Poland (Krakow); Portugal (Lisbon); Saudi Arabia (Riyadh); S. Africa (Uitsig & Ravensmead); Sri Lanka; Sudan (Gezeira); Sudan (Khartoum); Sweden (Uppsala); Trinidad & Tobago; Tunisia (Sousse); Turkey (Adana); USA (Lexington, KY). Sites in bold type have follow-up data (n = 18). Missing for 11 194 non-responders (41 sites); missing for 1164 non-responders (18 sites). Missing for 4487 non-responders (41 sites); missing for 767 non-responders (18 sites). Selected characteristics comparing non-responders and responders at baseline in the Burden of Obstructive Lung Disease cohort Sites names (n = 41): Albania (Tirana); Algeria (Annaba); Australia (Sydney); Austria (Salzburg); Benin (Sèmè-Kpodji); Cameroon (Limbe); Canada (Vancouver); China (Guangzhou); England (London); Estonia (Tartu); Germany (Hannover); Iceland (Reykjavik); India (Kashmir); India (Mumbai); India (Mysore); India (Pune); Jamaica; Kyrgyzstan (Chui); Kyrgyzstan (Naryn); Malawi (Blantyre); Malawi (Chikwawa); Malaysia (Penang); Morocco (Fes); Netherlands (Maastricht); Nigeria (Ile-Ife); Norway (Bergen); Pakistan (Karachi); Philippines (Manila); Philippines (Nampicuan-Talugtug); Poland (Krakow); Portugal (Lisbon); Saudi Arabia (Riyadh); S. Africa (Uitsig & Ravensmead); Sri Lanka; Sudan (Gezeira); Sudan (Khartoum); Sweden (Uppsala); Trinidad & Tobago; Tunisia (Sousse); Turkey (Adana); USA (Lexington, KY). Sites in bold type have follow-up data (n = 18). Missing for 11 194 non-responders (41 sites); missing for 1164 non-responders (18 sites). Missing for 4487 non-responders (41 sites); missing for 767 non-responders (18 sites). In total, high-quality post-bronchodilator spirometry data are available for 28 828 participants (52.6% females, 47.4% males). The mean age of participants was 55 years, the mean body mass index (BMI) was 26.7 kg/m2, 39.8% had ever smoked and 25.8% had higher education. The mean post-bronchodilator FVC (forced vital capacity) was 3.24 L and the mean post-bronchodilator FEV1 (forced expiratory volume in one second)/FVC was 77.7%. Table 3 shows the distribution of these characteristics by sex. Selected characteristics of the Burden of Obstructive Lung Disease cohort participants who completed the core questionnaire and had provided useable spirometry at baseline (41 sites; n = 28 828) FEV1 (forced expiratory volume in one second); FVC, forced vital capacity; SD, standard deviation. Selected characteristics of the Burden of Obstructive Lung Disease cohort participants who completed the core questionnaire and had provided useable spirometry at baseline (41 sites; n = 28 828) FEV1 (forced expiratory volume in one second); FVC, forced vital capacity; SD, standard deviation. The follow-up study was planned for 23 sites, but it was not feasible in five sites due to the COVID-19 pandemic. Participants from the BOLD baseline study were followed up once, between 2019 and 2021, in 14 sites in LMICs and four sites in Northern Europe, with a median follow-up time of 8.4 years. At follow-up, 12 502 individuals who had completed the core questionnaire, and had provided acceptable and repeatable lung function measurements at baseline, were invited to participate. The number of participants who completed the core questionnaire was 5936, and of these 4044 were able to perform spirometry and provided high-quality post-bronchodilator measurements for at least one lung function parameter (Figure 2). Slightly more than half of these participants were females (55.6%), the mean age was 61 years, the mean BMI was 26.5 kg/m2 and 30.7% had ever smoked. Table 4 shows the distribution of these characteristics by sex for participants who were followed up and had high-quality lung function data. Selection of participants in the Burden of Obstructive Lung Disease (BOLD) cohort Selected characteristics of the Burden of Obstructive Lung Disease cohort participants who completed the core questionnaire and had provided useable spirometry both at baseline and follow-up (n = 4044) FEV1 (forced expiratory volume in one second); FVC, forced vital capacity; SD, standard deviation. Selected characteristics of the Burden of Obstructive Lung Disease cohort participants who completed the core questionnaire and had provided useable spirometry both at baseline and follow-up (n = 4044) FEV1 (forced expiratory volume in one second); FVC, forced vital capacity; SD, standard deviation. Among participants lost to follow-up, 1155 had died, 3658 had migrated or were unreachable and 1237 refused to participate. To explore reasons for not being able to participate in the follow-up, we investigated potential explanatory variables (sex, age, BMI, smoking status, education level, self-reported doctor diagnosis of cardiovascular disease, self-reported doctor diagnosis of diabetes, a history of tuberculosis, dyspnoea, chronic cough, chronic phlegm and wheezing) using logistic regression within each site and then pooling together estimates using random effects meta-analysis. Older participants, current smokers and those with lower BMI were more likely to be lost to follow-up (Table 5). Based on this information, we calculated inverse probability weights to correct for loss-to-follow up in future analyses.7,8 Pooled odds ratio (OR) and 95% confidence interval (CI) of being lost to follow-up Adjusted simultaneously for all variables in the table. P-value from chi square test for heterogeneity across sites. Pooled odds ratio (OR) and 95% confidence interval (CI) of being lost to follow-up Adjusted simultaneously for all variables in the table. P-value from chi square test for heterogeneity across sites. In both surveys, participants were asked to answer a set of standardized questionnaires and undergo a series of measurements. The questionnaires were translated into the local language of each site, back-translated to English and checked for accuracy before administration by trained staff. Questionnaires were developed to obtain information about respiratory symptoms, respiratory and cardiometabolic diagnoses, health care use, medication use, activity limitation and health status, as well as about potential risk factors, including tobacco smoking, occupational and environmental exposures and diet (Table 6). In addition, measurements of several anthropometric parameters, blood pressure and pulse rate (M2 Basic, Omron), and lung function were taken. Lung function testing was performed using a spirometer (EasyOne, ndd Medizintechnik AG), before and after the administration of 200 μg of salbutamol via an inhalation spacer (Able, Clement Clarke International). All spirometry curves were checked centrally at the BOLD Operations Centre, and to be considered useable, tests had to include at least three acceptable curves (no hesitation, complete blow, no artefact affecting lung function readings), with the two best blows being within 200 mL of each other. Prior to the start of the surveys, study site staff underwent a 1-week intensive training which covered consenting, questionnaire data collection, spirometry testing and quality control, anthropometry measurements and data transfer. Information collected in the Burden of Obstructive Lung Disease cohort at baseline and follow-up COPD, chronic obstructive pulmonary disease; GPS, global positioning system. Some sites only. Information collected in the Burden of Obstructive Lung Disease cohort at baseline and follow-up COPD, chronic obstructive pulmonary disease; GPS, global positioning system. Some sites only. BOLD has published extensively on the prevalence and aetiology of chronic airflow obstruction. By April 2023, there were 102 publications in peer-reviewed journals [https://www.imperial.ac.uk/nhli/bold/publications]. Here we highlight the main findings from this study to date. The prevalence of chronic airflow obstruction varies widely across world regions but is, on average, slightly lower in LMICs and more common among males (11.2%) than among females (8.6%). Among males, the prevalence of chronic airflow obstruction ranges from 3.5% in Riyadh (Saudi Arabia) to 23.2% in Uitsig and Ravensmead (South Africa), and among females from 2% in Sousse (Tunisia) to 19.4% in Salzburg (Austria).4; The main risk factors for chronic airflow obstruction are tobacco smoking (both active and passive), which accounts for approximately 46% of the prevalence in males and 26% in females. The next most important risk factors are a poor education level and poverty, followed by a history of tuberculosis (where tuberculosis is common), a low BMI and exposure to dust in the workplace for more than 10 years.4,9–12 Ambient particulate matter and the use of solid fuels for cooking and heating are unlikely to explain a substantial amount of the prevalence of chronic airflow obstruction. These findings are equally true for males and females.13,14 The prevalence of small airways obstruction in the presence of what is usually considered normal lung function (i.e. isolated small airways obstruction) is common in general populations across the world. In addition, the main risk factors for isolated small airways obstruction are the same as for chronic airflow obstruction, suggesting that the former has the potential to predict the latter.15 There is a large proportion of people with low forced vital capacity (FVC), suggestive of low lung volumes, in LMICs, particularly in sub-Saharan Africa. In addition, this low FVC is associated with cardiometabolic diseases (i.e. cardiovascular disease, hypertension and diabetes)16 and cardiometabolic risk factors.17 The BOLD study has brought together a strong and widespread international collaboration to address the epidemiology of COPD and is today a reference in the field of chronic respiratory diseases. In addition, this study has had an important impact on capacity building and promotion of equity and inclusion across the study sites. The main strengths of the BOLD study are the: broad coverage of world regions and ethnic groups; large sample of representative population-based data; use of a standardized protocol, including the same questionnaires and same model of spirometers to test lung function, across study sites; centralized training and certification of interviewers and spirometry technicians. The quality of the data was monitored throughout the study, and re-training of staff was carried out if necessary; high quality of pre- and post-bronchodilator lung function measurements, with centralized quality control and assessment of all spirometry curves. In the follow-up study, we also developed and implemented an algorithm to monitor lung function data collection and provide feedback to spirometry technicians in ‘real-time’. The main weaknesses are the: self-reported doctor diagnoses of COPD, asthma, heart disease, hypertension and diabetes, which are subject to recall bias and local diagnostic guidelines and patterns; limited data on causes of death in sites located in LMICs; except for four sites in Northern Europe, the lack of biological samples, including DNA samples, for investigation of biological mechanisms. These will be sought in the next wave of the study. To compensate for non-response at baseline and at follow-up, we derived weights for respondents to the core questionnaire. However, we cannot dismiss the possibility of bias in our estimates due to unmeasured factors. The study data are not freely accessible. However, proposals for collaboration will be considered. Requests should be directed to the project lead, Dr Andre F S Amaral [a.amaral@imperial.ac.uk]. Albania: Hasan Hafizi (principal investigator [PI]), Anila Aliko, Donika Bardhi, Holta Tafa, Natasha Thanasi, Arian Mezini, Alma Teferici, Dafina Todri, Jolanda Nikolla, and Rezarta Kazasi (Tirana University Hospital Shefqet Ndroqi, Albania); Algeria: Hamid Hacene Cherkaski (PI), Amira Bengrait, Tabarek Haddad, Ibtissem Zgaoula, Maamar Ghit, Abdelhamid Roubhia, Soumaya Boudra, Feryal Atoui, Randa Yakoubi, Rachid Benali of of and of and (PI), (PI), of Medical (PI), and of Medical (PI), of and Research in and University of (PI), and for and University of [PI]), and of Hospital of Medical (PI), and University (PI), and of Medical and of Lung (PI), and of and University Medical A (PI), A and of Medical (PI), and of (PI), and (PI), and Research (PI), of the (PI), and and of and and Malawi Wellcome (PI), and of in and University Malaysia and The F and University Medical the (PI), and of of University of and of University of A (PI), Hasan F S A and of the Philippines and Centre, and the A S S F S and of the (PI), and of of University of (PI), and of Saudi (PI), (PI), A and L (Saudi Saudi South and of Lung South Sri (PI), Research Sri A (PI), and of Medical and Trinidad and (PI), of the Trinidad and and Hospital (PI), and of University of and Andre F S Amaral and Lung London, London, A & Portland, for Portland, of A L of (PI), and of Kentucky, of All sites from local the follow-up study was also by Research and participants provided of the the and conducted data All and to the of the This was supported by the Wellcome Trust (085790/Z/08/Z), University of Portland, AstraZeneca Portland, Portland, Portland, South Africa, Uitsig and Portland, Portland, Schering-Plough Portland, Portland, China Portland, OR, and Salzburg Research for the Research Centre, the South Medical Research the South and the University of Lung (South Africa, Uitsig and Ravensmead); University Hospital and The Research AstraZeneca and of for and University Medical Research for of of and of Research Clement Clarke and Lung and Lung and University The BOLD follow-up study was supported by the UK Medical Research Council (MR/R011192/1) and AstraZeneca AB (ESR-17-13417). all participants and staff for time and into this study. is a to GlaxoSmithKline, AstraZeneca and a of GlaxoSmithKline, and an on of people the tobacco has from AstraZeneca, and for and at the 3 years, and has published research with in the grants from of Research Collaborative Research with Merck, during the of the from GlaxoSmithKline, and from the submitted

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.