In this analysis, the majority (60.7%) of the COPD patients had severe or very severe airflow limitation. Biomass smoke exposure was significantly more common among patients with very severe airflow limitation. Similarly, all and moderate COPD exacerbations were significantly more frequent in those with very severe airflow limitation. Regardless of airflow limitation severity, the HRQOL of the patients was markedly impaired compared to healthy individuals. Even though patients with more severe airflow limitation had more frequent severe exacerbations, higher dyspnea symptom score, and poorer HRQOL, these were not significantly different across the severity of airflow limitation groups. Patients with mild airflow limitation had fewer respiratory symptoms than those with very severe airflow limitation. PB-FEV1 was only weakly associated with exacerbation frequency, dyspnea symptom, and HRQOL; exacerbation frequency was moderately correlated with dyspnea symptom and HRQOL; while dyspnea symptom was strongly correlated with HRQOL.
In the Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points (ECLIPSE) study, COPD exacerbations were more frequent and more severe as the severity of airflow limitation increased.(17) The COPD in Five Latin American Cities (PLATINO), and COPD and Systemic Consequences-Comorbidities Network (COSYCONET) reported significantly more frequent exacerbations among patients with severe to very severe airflow limitation.(21, 22) In the Study to Understand Mortality and Morbidity in COPD (SUMMIT), COPD patients in the lowest quantiles (FEV1 < 53.5% predicted) of airflow limitation reported a significantly higher risk of moderate-to-severe and severe exacerbations.(23) The Understanding Potential Long-term Impacts on Function with Tiotropium (UPLIFT) study and a systemic review by Westwood et al reported an inverse relationship between exacerbation and FEV1 but the correlation was small (r = − 0.12, p < 0.0001; and r = − 0.27, p = 0.049, respectively).(24, 25) The findings of significantly more frequent exacerbations in our patients with very severe airflow limitation and a small correlation with PB-FEV1 are consistent with the findings of these studies.
With regard to associations between FEV1 and dyspnea symptom, Oga et al reported a medium correlation (r = − 0.37, p < 0.05),(26) Huang et al reported a small correlation (r = − 0.234, p < 0.001),(27) while the post-hoc analysis of the Effect of Glycopyrronium or Indacaterol Maleate and Glycopyrronium Bromide Fixed-Dose Combination on Symptoms and Health Status in Patients with Moderate COPD (CRYSTAL) study did not show any correlation (r = 0.121).(28) This analysis showed a small correlation between PB-FEV1 and dyspnea symptom which is only similar to that reported by Huang et al.(27)
Unlike the present study, Izquierdo et al and Hong et al reported that HRQOL was significantly different between patients with different COPD severity.(29, 30) The Clinical COPD Questionnaire (CCQ) was used in the former study while the European Quality of Life Five Dimensions (EQ-5D) was used in the latter. Meanwhile, Agrawal et al and Balcells et al reported a significantly poorer HRQOL assessed by SGRQ-c in patients with stage III /IV and stage IV COPD, respectively.(31, 32) For the association between FEV1 and HRQOL, Ahmed et al,(33) Burgel et al,(34) and Zamzam et al (35) reported a large correlation (SGRQ-c, r = − 0.86, p < 0.001; SGRQ-c, r = − 0.56, p < 0.001; and r = − 0.65, p < 0.001, respectively); while Deslee et al,(36) Oga et al,(26) and Westwood et al (25) (SGRQ-c, r = − 0.372, p < 0.0001; r = − 0.44, p = 0.007; and r = − 0.46, p < 0.001, respectively ), as well as Kim et al (CCQ, r = − 0.35; and EQ-5D, r = 0.30) and Garrido et al (12-Item Short Form Survey: physical component, r = 0.38, p < 0.001) reported a medium correlation.(37, 38) Sundh et al and Bentsen et al reported a small correlation between FEV1 and HRQOL (CAT, r = − 0.13, p = 0.001; and EQ-5D, r = 0.008, p < 0.0001 and 36-Item Short Form Survey: physical, r = 0.19, p = 0.007; mental, r = − 0.14, p = 0.043, respectively),(39, 40) which were similar to our findings.
Studies have shown patients with frequent COPD exacerbations have significantly worse dyspnea symptom and HRQOL.(6, 41) Our analysis shows a medium correlation between exacerbation frequency and dyspnea symptom as well as HRQOL, consistent with that reported by Kelly et al (mMRC, r = 0.31; and CAT, r = 0.42, both p < 0.0001),(42) Burgel et al (SGRQ-c, r = 0.31, p < 0.001),(34) and Deslee et al (SGRQ-c, r = 0.391, p < 0.0001).(36) To date, a strong correlation between mMRC and HRQOL has been reported by the majority of the studies including that by Deslee et al (SGRQ-c, r = 0.602, p < 0.0001),(36) Ahmed et al (SGRQ-c, r = 0.59, p < 0.001),(33) Burgel et al, (SGRQ-c, r = 0.53, p < 0.01),(34) and Kelly et al (CAT, r = 0.50, p < 0.0001).(42) The present study and another study in Japan by Horita et al (CAT, r = 0.88, p < 0.001) show a very large correlation between mMRC and HRQOL.(43)
FEV1 is a unidimensional measurement that reflects the pathophysiology of COPD while exacerbation, dyspnea symptom, and HRQOL are multidimensional measurements of COPD consequences from the patients’ perspective.(44) This explains the absence or small correlation between FEV1 with exacerbation, dyspnea symptom, and HRQOL. The interaction between smoking, air pollution, respiratory tract infection, bronchiectasis, blood eosinophil count, the severity of airflow limitation, prior exacerbation, and comorbidities leads to the occurrence of COPD exacerbation.(45) For dyspnea symptom, the mechanisms encompass the interaction of physiological, psychological, and emotional factors of COPD patients.(46) Meanwhile, the HRQOL of COPD patients depends on the interaction of their physical, functional, emotional, social, and economic well-being.(47) The moderate correlation between COPD exacerbation and dyspnea symptom or HRQOL is attributed to the partial overlap between their dimensions. On the other hand, most of the dimensions are overlapped between dyspnea symptom and HRQOL which explains the large correlation. Even though symptoms such as cough, sputum, and wheezing are also assessed, the majority of questions in CAT (CAT 3 – CAT 8) and SQRQ-c (part of symptoms component, most of activity and impact component) still assess dyspnea symptom or their complications as the main outcomes.(18, 19)
The findings from this study suggest that FEV1 is not a reliable parameter to measure during the follow-up of COPD patients due to its weak correlation with exacerbation, dyspnea symptom, and HRQOL. Exacerbation should be routinely assessed as it is the prognosis hallmark of COPD and is only moderately reflected by the dyspnea symptom or HRQOL. Dyspnea symptom strongly predicts the value of HRQOL. Therefore, during a busy clinic, a simpler and time-saving tool such as mMRC should precede CAT and SGRQ-c. In short, exacerbation and mMRC are the recommended parameters to evaluate during the follow-up of COPD patients. This is in line with the Global Initiative for Chronic Obstructive Lung Disease guidelines that recommend grouping of COPD patients based on exacerbation frequency in the past one year and mMRC or CAT score at diagnosis and evaluating exacerbation and dyspnea symptom during follow-up visits.(1) Only in selected circumstances, spirometry is used to identify alternative diagnoses, suitability for interventional procedures, and to detect a rapid decline in FEV1.
This study evaluates the core parameters of COPD simultaneously, namely FEV1, exacerbation, dyspnea symptom, and HRQOL. Even though studies looking at these parameters have been conducted in other parts of the world, this is one of the very few studies in the South-East Asia region. Data from different regions of the world is needed for various reasons. First, the etiology of COPD can be different. In the present study, biomass smoke exposure was reported in more than one-third of the patients. Even though FEV1 decline due to biomass smoke exposure is slower,(48) biomass smoke exposure and cigarette smoking have an additive adverse effect on airflow obstruction.(49) Second, genetic heterogeneity can affect the presentation and outcomes of COPD. This study included the population from Peninsular Malaysia and the Island of Borneo. Third, the culture and economic activity of the population could have an impact on the perceived symptom and HRQOL. Other strengths of this study include representative sample was obtained from both the primary and tertiary care centers, as well as a similar methodology was used in both studies, therefore, minimizing any data bias. However, there are several limitations to this study. First, this was a cross-sectional study. HRQOL may vary over time and such variation may not be reflected in a cross-sectional study. Serial changes of FEV1 may give a better understanding of the exacerbation, dyspnea symptom, and HRQOL. Second, the inclusion criteria of the studies were slightly different in terms of age and definition of fixed airway obstruction. The use of PB-FVC6 in the second study potentially excludes a proportion of patients with mild COPD. Third, the study outcome was decided later in this post-hoc analysis. However, this study fulfilled the minimum sample size of 208 patients as calculated. Fourth, recall errors in exacerbation frequency cannot be discounted but minimized by counter checks with the medical records and family members. Fifth, the result of this study is only novel for South-East Asia. A future study that prospectively evaluates the FEV1, exacerbation, dyspnea symptom and HRQOL of COPD patients is expected to mitigate these limitations.