Article Text

Isolated small airways obstruction predicts future chronic airflow obstruction: a multinational longitudinal study
  1. Ben Knox-Brown1,
  2. James Potts1,
  3. Valentina Quintero Santofimio1,
  4. Cosetta Minelli1,2,
  5. Jaymini Patel1,
  6. Najlaa Mohammed Abass3,
  7. Dhiraj Agarwal4,
  8. Rana Ahmed5,
  9. Padukudru Anand Mahesh6,
  10. Jayaraj BS6,
  11. Meriam Denguezli7,
  12. Frits Franssen8,9,
  13. Thorarinn Gislason10,11,
  14. Christer Janson12,
  15. Sanjay K Juvekar4,
  16. Parvaiz Koul13,
  17. Andrei Malinovschi12,
  18. Asaad Ahmed Nafees14,
  19. Rune Nielsen15,
  20. Stefanni Nonna M Paraguas16,17,
  21. Sonia Buist18,
  22. Peter GJ Burney1 and
  23. Andre F S Amaral1
  1. 1National Heart and Lung Institute, Imperial College London, London, UK
  2. 2NIHR Imperial Biomedical Research Centre, London, UK
  3. 3College of Medicine, Alshaab Teaching Hospital, Bahri University, Khartoum, Sudan
  4. 4Vadu Rural Health Program, KEM Hospital Pune Research Centre, Pune, India
  5. 5The Epidemiological Laboratory, Khartoum, Sudan
  6. 6Respiratory Medicine, JSS Medical College, Mysore, Karnataka, India
  7. 7Faculte de Medecine de Sousse, Universite de Sousse, Sousse, Tunisia
  8. 8Respiratory medicine, Maastricht University Medical Centre+, Maastricht, Netherlands
  9. 9Research and Education, CIRO, Horn, Netherlands
  10. 10Department of Sleep, Landspitali University Hospital, Reykjavik, Iceland
  11. 11Faculty of Medicine, University of Iceland, Reykjavik, Iceland
  12. 12Department of Medical Sciences Respiratory, Allergy and Sleep Research, Uppsala University, Uppsala, Sweden
  13. 13Sher-i-Kashmir Institute of Medical Sciences, Srinagar, Jammu and Kashmir, India
  14. 14Department of Community Health Sciences, The Aga Khan University, Karachi, Sindh, Pakistan
  15. 15Department of Thoracic Medicine, University of Bergen, Bergen, Hordaland, Norway
  16. 16Philippine College of Chest Physicians, Quezon City, Philippines
  17. 17Philippine Heart Center, Quezon City, Manila, Philippines
  18. 18Oregon Health & Science University, Portland, Oregon, USA
  1. Correspondence to Mr Ben Knox-Brown; b.knox-brown20{at}imperial.ac.uk

Abstract

Background Chronic airflow obstruction is a key characteristic of chronic obstructive pulmonary disease. We investigated whether isolated small airways obstruction is associated with chronic airflow obstruction later in life.

Methods We used longitudinal data from 3957 participants of the multinational Burden of Obstructive Lung Disease study. We defined isolated small airways obstruction using the prebronchodilator mean forced expiratory flow rate between 25% and 75% of the forced vital capacity (FVC) (FEF25–75) if a result was less than the lower limit of normal (<LLN) in the presence of a normal forced expiratory volume in 1 s to FVC ratio (FEV1/FVC). We also used the forced expiratory volume in 3 s to FVC ratio (FEV3/FVC) to define small airways obstruction. We defined chronic airflow obstruction as post-bronchodilator FEV1/FVC<LLN. We performed mixed effects regression analyses to model the association between baseline isolated small airways obstruction and chronic airflow obstruction at follow-up. We assessed discriminative and predictive ability by calculating the area under the receiver operating curve (AUC) and Brier score. We replicated our analyses in 26 512 participants of the UK Biobank study.

Results Median follow-up time was 8.3 years. Chronic airflow obstruction was more likely to develop in participants with isolated small airways obstruction at baseline (FEF25-75 less than the LLN, OR: 2.95, 95% CI 1.02 to 8.54; FEV3/FVC less than the LLN, OR: 1.94, 95% CI 1.05 to 3.62). FEF25-75 was better than the FEV3/FVC ratio to discriminate future chronic airflow obstruction (AUC: 0.764 vs 0.692). Results were similar among participants of the UK Biobank study.

Conclusion Measurements of small airways obstruction can be used as early markers of future obstructive lung disease.

  • COPD epidemiology
  • Lung Physiology

Data availability statement

Data are available upon reasonable request. Deidentified participant data and questionnaires of the BOLD study may be shared, after publication, on a collaborative basis upon reasonable request made to Dr Amaral (a.amaral@imperial.ac.uk). Requesting researchers will be required to submit an analysis plan.

https://creativecommons.org/licenses/by/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Small hospital-based studies and studies of symptomatic smokers have suggested that isolated small airways obstruction predicts future chronic airflow obstruction.

WHAT THIS STUDY ADDS

  • In general populations, individuals with isolated small airways obstruction are at greater risk of lung function decline and development of chronic airflow obstruction.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • These findings will lead to more research into the role of the small airways in chronic obstructive pulmonary disease and how to target them to prevent disease. They will also raise awareness of clinicians to the potential benefit of keeping record of patients’ forced expiratory flows, in addition to other more common spirometry measures.

Introduction

Chronic obstructive pulmonary disease (COPD) is a heterogeneous condition associated with reduced lifespan,1 increased disability2 and a greater dependence on healthcare services.3 A primary feature of COPD is chronic airflow obstruction, which is defined by an abnormal postbronchodilator forced expiratory volume in 1 s to forced vital capacity ratio (FEV1/FVC).4 The FEV1/FVC ratio is non-specific, reflecting the presence of airflow obstruction in both the large and small airways. However, in early disease, damage is largely confined to the small airways.5

Parameters that are universally measured by spirometry devices but seldom reported are the mean forced expiratory flow rate between 25% and 75% of the forced vital capacity (FEF25-75) and the forced expiratory volume in three seconds as a ratio of the forced vital capacity (FEV3/FVC). There is increasing interest in the use of these parameters to identify airflow obstruction in the small airways,6 and studies in ever smokers have shown that individuals with atypical measurements for these parameters have evidence of functional small airways disease, gas trapping and emphysema on CT, even when lung function is normal according to traditional measurement indices (ie, isolated small airways obstruction).7–10 Despite these findings, it is still widely believed that these parameters are neither sensitive nor specific to changes within the small airways.11

Studies in selected clinical populations and symptomatic ever smokers have found that individuals with isolated small airways obstruction are at greater risk of developing chronic airflow obstruction in later life.8 12 To the best of our knowledge, no study has attempted to determine whether this is also true in general populations. Using longitudinal data from the multinational Burden of Obstructive Lung Disease (BOLD) study, we aimed to investigate if having isolated small airways obstruction at baseline was associated with progression to chronic airflow obstruction at follow-up and to compare results for two different definitions of small airways obstruction. We also aimed to replicate our findings using data from the UK Biobank study.

Methods

Main study

Study population

The BOLD study is a multinational observational cohort study whose protocol has been published previously.13 14 Between January 2003 and December 2016, non-institutionalised adults ≥40 years of age were recruited from 41 municipalities, across 34 countries. Site-specific sampling strategies were implemented to randomly recruit representative samples of the populations studied. Participants from 18 sites were then followed up between January 2019 and October 2021. For the present study, participants were included if they had completed the study core questionnaire and had acceptable spirometry at both baseline and follow-up. Participants were excluded if they had a contraindication for lung function testing at either visit.

Procedures

Demographic data and information on respiratory symptoms, health status and exposure to potential risk factors were collected by trained staff, who administered standardised questionnaires translated into the local language. Lung function, including FEV1, FVC, FEV3 and FEF25-75, was measured using the ndd EasyOne Spirometer (ndd Medizintechnik AG, Zurich, Switzerland), before and 15 min after inhaled salbutamol (200 μg). Spirograms were centrally reviewed and assigned a quality score based on acceptability and reproducibility criteria.15

Definitions of spirometric abnormalities

At baseline, we defined isolated small airways obstruction as a prebronchodilator FEF25-75 less than the lower limit of normal (LLN), with an FEV1/FVC ratio equal to or greater than the LLN. Due to the perceived lack of clinical utility for FEF25-75 and its large between subject variation in normal populations,11 16 we investigated a second parameter, the FEV3/FVC ratio using the same definition. At follow-up, we defined chronic airflow obstruction as postbronchodilator FEV1/FVC ratio below the LLN. To calculate the LLN, we used reference equations for European Americans in the third US National Health and Nutrition Examination Survey.17 18

Statistical analysis

We calculated the incidence rate of chronic airflow obstruction per 1000 person years. To estimate the association between having isolated small airways obstruction at baseline and chronic airflow obstruction at follow-up, we performed multilevel (mixed effects) logistic regression analyses to account for clustering by study site. We also used multilevel linear regression to estimate the association between isolated small airways obstruction and postbronchodilator FEV1/FVC ratio as a continuous measure. We initially considered all potential risk factors for chronic airflow obstruction,19 however, as risk factors for progression from isolated small airways obstruction to chronic airflow obstruction are largely unknown, we then used a backward elimination procedure, keeping only those variables that were significant in the final model: sex (male/female), age (years), body mass index (BMI; kg/m²), smoking status (never/former/current), and pack years of smoking. We modelled the association of isolated small airways obstruction with chronic airflow obstruction using a random slope to allow the magnitude of the association to vary by study site. We did not directly model the effect of follow-up time as this was determined at site level. However, to check for effect modification, we performed stratified analyses in those with less than 5-year follow-up and those with equal to or greater than 5-year follow-up. We also performed stratified analyses by sex to investigate possible effect modification. Finally, we performed a sensitivity analysis on never smokers to investigate any residual confounding due to smoking. All analyses were conducted using inverse probability weights20 to account for missing data at follow-up.

Receiver-operating characteristic curves were constructed, and the area under the curve (AUC) calculated for both FEF25-75 and the FEV3/FVC ratio to determine their sensitivity and specificity in predicting chronic airflow obstruction. The AUC values of the two parameters were compared as previously described.21 In addition, we evaluated the incremental value of both parameters to determine if they conveyed an improvement in classification accuracy over a model containing age, sex, BMI and smoking history.22 To assess the overall predictive performance of the parameters, we calculated the Brier score, which ranges between 0 and 1, with 0 indicating a perfect prediction and 1 a poor predictive ability.23 All results were considered significant if the p value was below 0.05. Analyses were performed using Stata V.17 (Stata Corp.).

Replication study

Study population

The UK Biobank study recruited over 500 000 adults, aged 40–69 years, across 22 different sites covering England, Wales and Scotland, between 2006 and 2010.24 Participants completed a baseline assessment with a detailed health questionnaire and clinical measurements, which included spirometry. Between 2014 and 2020, individuals living within close proximity of an assessment site were invited for repeat assessment

Procedures

Participants were included in this study if they had acceptable spirometry at both baseline and follow-up. Spirometry was performed prebronchodilator (participants were not instructed to withhold their usual inhaled medications) using a calibrated Vitalograph Pneumotrac 6800. We included only those with the highest quality spirometry manoeuvres, defined as a having minimum of two spirograms with no cough, back-extrapolated volume<5% FVC (or>5% but <150 mL), reproducible FEV1 and FVC, and a forced expiratory time of≥6 s on the best curve (curve with highest FEV1 and FVC). FEF25-75 and FEV3 were derived from the raw data as previously described.25 For participants who attended for more than one follow-up visit, airflow obstruction was defined at its first presentation.

Statistical analysis

We conducted the same analysis used for the BOLD data, further adjusting for follow-up time, as this was not determined at site level. We also performed sensitivity analyses excluding those with a self-reported doctor diagnosis of asthma at baseline. This was done to make the results comparable to the postbronchodilator estimates in the BOLD study.

Patient and public involvement

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Results

Main study

At baseline, 26 448 participants, across 41 sites, completed the core study questionnaire and had acceptable measurements for FEF25-75 and FEV3/FVC ratio. Eighteen study sites took part in follow-up, with 12 520 eligible participants. At follow-up, 1155 participants had died, 3658 had migrated or were unreachable, and 1237 refused to participate. Five thousand nine hundred and thirty-six participants completed the core questionnaire at follow-up, from which 1979 participants were excluded due to not performing spirometry (n=855) or poor-quality spirometry (n=1124). A total of 3957 participants with a median (IQR) follow-up time of 8.3 years (6.1–11.0) were included in the present analysis (table 1). There were fewer males than females (1733 vs 2224). Mean age ranged from 46.1 to 61.9 years across study sites. Mean BMI was lowest in Chikwawa, Malawi (21.9 kg/m²) and highest in Jamaica (30.2 kg/m²). The site with the largest proportion of never smokers was Sémé-Kpodji, Benin (99%, 150 of 152 participants), and the smallest proportion was Bergen, Norway (34%, 81 of 237 participants). At baseline, the prevalence of isolated small airways obstruction for FEF25-75 ranged from 0% in Tartu, Estonia and Fez, Morocco to 27.9% in Mysore, India. For FEV3/FVC ratio, prevalence ranged from 0% in Jamaica and Fes, Morocco to 11.6% in Reykjavik, Iceland. The prevalence of chronic airflow obstruction at baseline ranged from 4.3% in Naryn, Kyrgyzstan to 24.6% in Kashmir, India. At follow-up, the prevalence of chronic airflow obstruction was similar to baseline, ranging from 3.8% in Sémé-Kpodji, Benin to 25.0% in Kashmir, India. Mean follow-up time ranged from 4.4 years in Karachi, Pakistan to 14.7 years in Reykjavik, Iceland (table 1).

Table 1

Characteristics and prevalence estimates for chronic airflow obstruction (CAO) and isolated small airways obstruction (SAO) for Burden of Obstructive Lung Disease (BOLD) study participants according to participation at follow-up

Table 2 shows the characteristics of BOLD study participants with isolated small airways obstruction at baseline. Participants with isolated small airways obstruction for FEF25-75, were on average younger, more likely to be female, and to be never smokers than those with isolated small airways obstruction for FEV3/FVC ratio. They also had a smaller smoking pack year history, higher FEV1/FVC ratio and lower FVC.

Table 2

Characteristics of those with isolated small airways obstruction (SAO) at baseline in the BOLD and UK Biobank studies according to participation at follow-up

Of those with isolated small airways obstruction for FEF25-75, 26 of 448 (6%) progressed to chronic airflow obstruction at follow-up. While for FEV3/FVC ratio, 14 of 233 (6%) progressed to chronic airflow obstruction. In participants with no evidence of any airway obstruction at baseline, 72 of 2545 (3%) progressed to chronic airflow obstruction. The incidence rates for progression to chronic airflow obstruction were 7.1/1000 person years (95% CI 4.9 to 10.5) for FEF25-75 less than the LLN and 6.9/1000 (95% CI 4.1 to 11.6) for FEV3/FVC ratio less than the LLN. In individuals with no evidence of airflow obstruction at baseline (ie, FEF25-75, FEV3/FVC, and FEV1/FVC greater than or equal to the LLN), incidence of progression to chronic airflow obstruction was 3.2/1000 (95% CI 2.6 to 4.1). For both parameters, incidence rates were higher in males than females and in ever smokers compared with never smokers (figure 1).

Figure 1

Incidence rate per 1000 person years for progression from isolated small airways obstruction to chronic airflow obstruction for A) FEF25-75 and B) FEV3/FVC ratio.

When stratifying by WHO region, incidence rates for progression to chronic airflow obstruction for FEF25-75, ranged from 2.4/1000 (95% CI 0.6 to 9.6) in the African region to 23.0/1000 (95% CI 12.7 to 41.5) in the European region. While for FEV3/FVC ratio, incidence ranged from 0.0/1000 in the Eastern Mediterranean region to 7.6/1000 (95% CI 3.6 to 15.9) in the European region (online supplemental etable 1, appendix p2).

Tables 3 and 4 display the results of the mixed effects regression analyses. At baseline, isolated small airways obstruction for FEF25-75 was associated with a lower FEV1/FVC ratio (β: −4.16, 95% CI −5.70 to –2.63), and significantly increased odds of chronic airflow obstruction at follow-up (OR: 2.95, 95% CI 1.02 to 8.54). Similarly, isolated small airways obstruction for FEV3/FVC ratio was associated with a lower FEV1/FVC ratio (β: −2.87 95% CI −4.27 to –1.47) and chronic airflow obstruction at follow-up (OR: 1.94, 95% CI 1.05 to 3.62). After excluding ever smokers from the analysis, isolated small airways obstruction for FEF25-75 (β: −3.41, 95% CI −4.88 to –1.95) and FEV3/FVC ratio (β: −2.58, 95% CI −3.97 to –1.21) were associated with a lower FEV1/FVC ratio but not chronic airflow obstruction. When stratifying by follow-up time, there were no significant differences in the association with FEV1/FVC ratio for either parameter. Isolated small airways obstruction for both FEF25-75 and FEV3/FVC ratio were associated with a significantly lower FEV1/FVC ratio at follow-up in both males and females. However, only in males was there an association with subsequent chronic airflow obstruction (tables 3 and 4). Of note, 44% of male participants reported a smoking history, compared with just 14% of females. The results for postbronchodilator FEF25-75 and FEV3/FVC ratio were not materially different from prebronchodilator (online supplemental table 2 and 3, appendix p3).

Table 3

Association between baseline isolated small airways obstruction (SAO) and chronic airflow obstruction (CAO) at follow-up for FEF25-75

Table 4

Association between baseline isolated small airways obstruction (SAO) and chronic airflow obstruction (CAO) at follow-up for FEV3/FVC ratio.

The AUC to discriminate progression to chronic airflow obstruction was 0.764 for FEF25-75 and 0.692 for FEV3/FVC ratio (figure 2). There was a significant difference in the AUC between the two parameters (p=0.0017). When compared with the AUC for a model containing just age, sex, BMI and smoking history (AUC=0.686), FEF25-75 significantly improved discrimination (p=0.0006), while FEV3/FVC ratio did not (p=0.3816). The Brier scores assessing the predictive accuracy of the parameters was 0.0322 for FEF25-75 and 0.0320 for FEV3/FVC ratio, indicating good predictive accuracy.

Figure 2

Receiver operator characteristic curve and area under the curve (AUC) comparing ability of FEF25-75 and FEV3/FVC ratio to a model containing age, sex, BMI and smoking history alone to discriminate future chronic airflow obstruction. *P-value less than 0.05 indicates significant difference between models according to X2 test.

Replication study

Two hundred and fifty-two thousand five hundred and sixty participants had high-quality spirometry at baseline. Of these, 26 512 did not have airflow obstruction at baseline, had high quality spirometry at follow-up and were included in this analysis. Forty-two per cent were male, with a mean age of 55.5 years. Mean BMI was 26.7 kg/m² and 60% were never smokers. At baseline, 549 (2%) participants had isolated small airways obstruction for FEF25-75 and 162 (1%) for FEV3/FVC ratio. Median (IQR) follow-up time was 8.0 years (IQR: 5.0–10.0). Like the BOLD study participants, UK Biobank participants with isolated small airways obstruction for FEF25-75 were generally younger, more likely to be females, and have a lower FVC than those with isolated small airways obstruction for FEV3/FVC ratio. Interestingly, unlike in the BOLD study, those with isolated small airways obstruction for FEF25-75 had the lower FEV1/FVC ratio. When comparing across cohorts, the characteristics of those with isolated small airways obstruction were similar. The exceptions being that there were less never smokers and more former smokers in the UK Biobank study, as well as a larger smoking pack year history (table 2).

Of those with normal lung function at baseline, 1877 of 25 832 (7%) progressed to airflow obstruction (9.82/1000 person years, 95% CI 9.38 to 10.27). For FEF25-75, 116 of 549 (21%) progressed to airflow obstruction (28.77/1000 person years, 95% CI 23.98 to 34.91), and for FEV3/FVC ratio, 17 of 162 (10%) progressed to airflow obstruction (14.22/1000 person years, 95% CI 8.84 to 22.88).

Isolated small airways obstruction for FEF25-75 was associated with a lower FEV1/FVC ratio (β: −4.45, 95% CI −5.05 to –3.85) and greater odds of progression to airflow obstruction at follow-up (OR: 3.79, 95% CI 3.10 to 4.71). This association was seen in both males and females, and those who had never smoked (online supplemental etable 4, appendix). There was no association between isolated small airways obstruction for FEV3/FVC ratio and the FEV1/FVC ratio at follow-up (online supplemental etable 5, appendix). When stratifying by follow-up time, there were no significant differences in the associations for either parameter. After excluding those with a self-reported doctor diagnosis of asthma (online supplemental etable 6, appendix), associations with isolated small airways obstruction for FEF25-75 did not materially change. However, those with isolated small airways obstruction for FEV3/FVC ratio had a lower FEV1/FVC ratio (β: −2.02, 95% CI −3.61 to –0.43) and greater odds of progressing to airflow obstruction (OR: 2.65, 95% CI 1.45 to 4.82).

The AUC for FEF25-75 was 0.692, which improved discrimination compared with FEV3/FVC ratio (AUC: 0.629) (online supplemental figure 1, appendix). For both parameters, isolated small airways obstruction improved discrimination compared with a model containing age, sex, BMI and smoking history. The brier scores were 0.096 and 0.1042 for FEF25-75 and FEV3/FVC ratio, respectively. Restricting the analyses to those without a self-reported history of asthma did not improve discrimination (online supplemental efigure 2, appendix).

Discussion

To the best of our knowledge, this is the first general-population study to investigate whether isolated small airways obstruction is associated with progression to chronic airflow obstruction over time. Our study shows that isolated small airways obstruction is associated with having a lower FEV1/FVC ratio and increased odds of chronic airflow obstruction later in life. In addition, we found that isolated small airways obstruction measured using FEF25-75 was a better predictor of future obstruction than the FEV3/FVC ratio. We successfully replicated these findings using data from the UK Biobank study.

In the present study, when compared with those with no evidence of airflow obstruction at baseline, isolated small airways obstruction for FEF25-75 was associated with a 4% lower FEV1/FVC ratio and three times greater odds of chronic airflow obstruction at follow-up. We found similar in the UK Biobank study. Only one study has previously investigated this association. Kwon et al12 showed, in a hospital-based population of South Korean adults, that an isolated abnormality in FEF25-75 was associated with increased risk of airflow obstruction over a 10-year period. They also showed that risk of progression was higher for those with a smoking history. Unlike our study, Kwon et al12 used the fixed ratio of 0.70 for FEV1/FVC to define chronic airflow obstruction. The limitations of using a fixed ratio in general populations are well known and relate to overestimation of incidence. For this reason, its use is no longer recommended by the American Thoracic Society and European Respiratory Society.26 Our study adds to their findings by showing that the same association is seen when using the LLN to define abnormality in general populations.

At baseline, isolated small airways obstruction for FEV3/FVC ratio was associated with a 3% lower FEV1/FVC ratio and two times greater odds of progression to chronic airflow obstruction at follow-up. We replicated these findings in UK biobank participants but only after exclusion of those with a self-reported history of asthma. No previous studies have investigated this association. However, Dilektasli et al9 reported that in ever smokers from the COPDGene study, when the FEV1/FVC ratio was greater than 0.70, FEV3/FVC ratio less than the LLN was associated with increased emphysema and gas trapping on CT imaging, supporting our finding that an abnormal FEV3/FVC ratio is a precursor to future obstructive lung disease. In addition, studies in ever smokers have shown that having a FEV3/FEV6 ratio less than the LLN is also associated with increased risk of chronic airflow obstruction.8 27 The FEV3/FVC ratio and FEV3/FEV6 ratio are highly correlated and shown to give very similar prevalence estimates for isolated small airways obstruction.28 The rationale behind using the FEV6 in place of the FVC is uncertain providing spirometry has been performed correctly.

Interestingly, in the BOLD study, we found that isolated small airways obstruction was associated with progression to chronic airflow obstruction only in males. However, it is not clear if this is a genuine interaction, as in contrast with the logistic regression, the results of the linear regression showed that regardless of sex, those with isolated small airways obstruction had a lower FEV1/FVC ratio at follow-up. A potential explanation is the far smaller proportion of females that reported a smoking history compared with males; an important difference considering smoking is the strongest risk factor for chronic airflow obstruction.19 29 In our replication study, females with isolated small airways obstruction had a lower FEV1/FVC ratio and greater odds of progression to airflow obstruction at follow-up. This makes it likely that despite the association being stronger in males, universally, isolated small airways obstruction is a good predictor of future airflow obstruction.

Our finding that in never smokers, isolated small airways obstruction was associated with having a significantly lower FEV1/FVC ratio at follow-up is novel and was successfully replicated in participants of the UK Biobank study. It is well known that cigarette smoke damages the small airways, eventually leading to chronic airflow obstruction.19 We have previously shown that risk factors for isolated small airways obstruction also include occupational exposures to dust, previous TB diagnosis, low education level and family history of COPD.29 The causal pathways in never smokers are less clear and deserve further research, especially on the impact of intrauterine exposures, childhood growth, and ambient and indoor air pollution.

No previous studies have reported incidence rates for progression of isolated small airways obstruction to chronic airflow obstruction. In the BOLD study, we found that incidence of progression was similar for both parameters. It was higher in males compared with females, and ever smokers compared with never smokers. We found considerable variation in incidence rates across WHO regions. Despite this, incidence of progression to chronic airflow obstruction was generally highest in the European region. The incidence rates for progression from isolated small airways obstruction to airflow obstruction were significantly higher in our replication study. However, they were similar to BOLD sites in the European region. This finding is likely related to tobacco smoking, as in the European region of the BOLD study and the UK Biobank study, 43% and 40% of participants, respectively, were smokers.

Due to lack of agreement as to which spirometry parameters best reflects changes within the small airways, we used both FEF25-75 and FEV3/FVC. Both parameters have been shown to correlate with functional small airways disease on CT imaging,7 9 10 and we found that when less than the LLN, both are also associated with progression to chronic airflow obstruction. When we calculated the AUC of the parameters, FEF25-75 was significantly better at discriminating future chronic airflow obstruction than the FEV3/FVC ratio. In addition, when compared with a model containing age, sex, BMI and smoking history alone, only FEF25-75 significantly improved discrimination. Predictive accuracy using the Brier score was good for both parameters. We found a similar pattern in the UK Biobank study; however, overall discriminative ability was weaker. This could be explained by the optimisation of the statistical model for participants of the BOLD study, meaning there could be additional covariates that influence discriminative ability in the UK Biobank study that are not significant or are unmeasured in the BOLD study. Despite this, our findings suggest that while both FEF25-75 and FEV3/FVC ratio can be used to identify those at risk of chronic airflow obstruction, FEF25-75 performs the best.

Our study has strengths, including a wide geographical coverage, samples that are representative of the general population, and quality assured spirometry. We also compared the predictive ability of two different spirometry parameters. Our decision to prioritise prebronchodilator spirometry to define isolated small airways obstruction means at-risk individuals can be identified without the need for postbronchodilator spirometry, which is time consuming, costly and not widely available in low resource settings. There are also limitations. First, in a previous publication, we found that there was minimal difference between the FVC and forced expiratory volume in six seconds (FEV6) in UK Biobank partiscipants.25 This suggests that the FVC may be underestimated, which would falsely increase the FEF25-75 and FEV3/FVC ratio. As a result, prevalence if isolated small airways obstruction would be underestimated and the power to find an association reduced, particularly for the FEV3/FVC ratio. Second, due to the COVID-19 pandemic, the BOLD study had considerable loss to follow-up at some sites. However, we used inverse probability weighting in our analyses to account for this.

Conclusion

People with isolated small airways obstruction, particularly when measured using FEF25-75, are more likely to develop chronic airflow obstruction over time. As chronic airflow obstruction is a key component of a COPD diagnosis, our findings have implications for early detection and prevention of disease.

Data availability statement

Data are available upon reasonable request. Deidentified participant data and questionnaires of the BOLD study may be shared, after publication, on a collaborative basis upon reasonable request made to Dr Amaral (a.amaral@imperial.ac.uk). Requesting researchers will be required to submit an analysis plan.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by all sites from their local ethics committee, the follow-up study was also approved by Imperial College London Research Ethics Committee (ref. 17IC4272), and participants provided informed consent.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Twitter @RanaAttaElmula

  • Contributors BK-B and AFSA conceived the study. Under the supervision of AFSA, BK-B performed data analysis and prepared the initial draft with input from PGJB and CM. JPotts and VQS assisted with the preparation of the databases and analyses. All authors contributed to further drafting and final approval of the paper. BK-B is guarantor for this paper, accepts full responsibility for the work and the conduct of the study, had access to the data, and controlled the decision to publish.

  • Funding The Burden of Obstructive Lung Disease (BOLD) study has been supported by grants from the Wellcome Trust (085790/Z/08/Z) and Medical Research Council (MR/R011192/1). This research has been conducted using the UK Biobank Resource under application number 80005. We thank all participants and field workers/research assistants for their time and effort put into this study.

  • Competing interests RN reports grants and personal fees from AstraZeneca and GlaxoSmithKline and grants from Boehringer Ingelheim and Novartis, outside of the submitted work.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.