In modern biomedical research of complex diseases, a large number of demographic and clinical variables, herein called phenomic data, are often collected and missing values (MVs) are inevitable in the data collection process. Since many downstream statistical and bioinformatics methods require complete data matrix, imputation is a common and practical solution. In high-throughput experiments such as microarray experiments, continuous intensities are measured and many mature missing value imputation methods have been developed and widely applied. Numerous methods for missing data imputation of microarray data have been developed. Large phenomic data, however, contain continuous, nominal, binary and ordinal data types, which void application of most methods. Though several methods have been developed in the past few years, not a single complete guideline is proposed with respect to phenomic missing data imputation. In this paper, we investigated existing imputation methods for phenomic data, proposed a self-training selection (STS) scheme to select the best imputation method and provide a practical guideline for general applications. We introduced a novel concept of "imputability measure" (IM) to identify missing values that are fundamentally inadequate to impute. In addition, we also developed four variations of K-nearest-neighbor (KNN) methods and compared with two existing methods, multivariate imputation by chained equations (MICE) and missForest. The four variations are imputation by variables (KNN-V), by subjects (KNN-S), their weighted hybrid (KNN-H) and an adaptively weighted hybrid (KNN-A). We performed simulations and applied different imputation methods and the STS scheme to three lung disease phenomic datasets to evaluate the methods. An R package "phenomeImpute" is made publicly available. Simulations and applications to real datasets showed that MICE often did not perform well; KNN-A, KNN-H and random forest were among the top performers although no method universally performed the best. Imputation of missing values with low imputability measures increased imputation errors greatly and could potentially deteriorate downstream analyses. The STS scheme was accurate in selecting the optimal method by evaluating methods in a second layer of missingness simulation. All source files for the simulation and the real data analyses are available on the author's publication website.
15 November 2023 in American Journal of Respiratory and Critical Care Medicine
Chronic obstructive pulmonary disease is a major health problem with a high prevalence, a rising incidence, and substantial morbidity and mortality. Its course is punctuated by acute episodes of increased respiratory symptoms, termed exacerbations of chronic obstructive pulmonary disease (ECOPD). ECOPD are important events in the natural history of the disease, as they are associated with lung function decline and prolonged negative effects on quality of life. The present-day therapy for ECOPD with short courses of antibiotics and steroids and escalation of bronchodilators has resulted in only modest improvements in outcomes. Recent data indicate that ECOPD are heterogeneous, raising the need to identify distinct etioendophenotypes, incorporating traits of the acute event and of patients who experience recurrent events, to develop novel and targeted therapies. These characterizations can provide a complete clinical picture, the severity of which will dictate acute pharmacological treatment, and may also indicate whether a change in maintenance therapy is needed to reduce the risk of future exacerbations. In this review we discuss the latest knowledge of ECOPD types on the basis of clinical presentation, etiology, natural history, frequency, severity, and biomarkers in an attempt to characterize these events.
15 August 2023 in American Journal of Respiratory and Critical Care Medicine
Rationale: Cigarette smoking contributes to the risk of death through different mechanisms. Objectives: To determine how causes of and clinical features associated with death vary in tobacco cigarette users by lung function impairment. Methods: We stratified current and former tobacco cigarette users enrolled in Genetic Epidemiology of Chronic Obstructive Pulmonary Disease (COPDGene) into normal spirometry, PRISm (Preserved Ratio Impaired Spirometry), Global Initiative for Chronic Obstructive Lung Disease (GOLD) 1–2 COPD, and GOLD 3–4 COPD. Deaths were identified via longitudinal follow-up and Social Security Death Index search. Causes of death were adjudicated after a review of death certificates, medical records, and next-of-kin interviews. We tested associations between baseline clinical variables and all-cause mortality using multivariable Cox proportional hazards models. Measurements and Main Results: Over a 10.1-year median follow-up, 2,200 deaths occurred among 10,132 participants (age 59.5 ± 9.0 yr; 46.6% women). Death from cardiovascular disease was most frequent in PRISm (31% of deaths). Lung cancer deaths were most frequent in GOLD 1–2 (18% of deaths vs. 9–11% in other groups). Respiratory deaths outpaced competing causes of death in GOLD 3–4, particularly when BODE index ≥7. St. George’s Respiratory Questionnaire score ≥25 was associated with higher mortality in all groups: Hazard ratio (HR), 1.48 (1.20–1.84) normal spirometry; HR, 1.40 (1.05–1.87) PRISm; HR, 1.80 (1.49–2.17) GOLD 1–2; HR, 1.65 (1.26–2.17) GOLD 3–4. History of respiratory exacerbations was associated with higher mortality in GOLD 1–2 and GOLD 3–4, quantitative emphysema in GOLD 1–2, and airway wall thickness in PRISm and GOLD 3–4. Conclusions: Leading causes of death vary by lung function impairment in tobacco cigarette users. Worse respiratory-related quality of life is associated with all-cause mortality regardless of lung function.
Chronic obstructive pulmonary disease (COPD) is characterized by both airway remodeling and parenchymal destruction. The identification of unique biomarker patterns associated with airway dominant versus parenchymal dominant patterns would support the existence of unique phenotypes representing independent biologic processes. A cross-sectional study was performed to examine the association of serum biomarkers with radiographic airway and parenchymal phenotypes of COPD. Serum from 234 subjects enrolled in a CT screening cohort was analyzed for 33 cytokines and growth factors using a multiplex protein array. The association of serum markers with forced expiratory volume in one second percent predicted (FEV1%) and quantitative CT measurements of airway thickening and emphysema was assessed with and without stratification for current smoking status. Significant associations were found with several serum inflammatory proteins and measurements of FEV1%, airway thickening, and parenchymal emphysema independent of smoking status. The association of select analytes with airway thickening and emphysema was independent of FEV1%. Furthermore, the relationship between other inflammatory markers and measurements of physiologic obstruction or airway thickening was dependent on current smoking status. Airway and parenchymal phenotypes of COPD are associated with unique systemic serum biomarker profiles. Serum biomarker patterns may provide a more precise classification of the COPD syndrome, provide insights into disease pathogenesis and identify targets for novel patient-specific biological therapies.
Introduction Muscle loss is an important extrapulmonary manifestation of COPD. Dual energy X-ray absorptiometry (DXA) is the method of choice for body composition measurement but is not widely used for muscle mass evaluation. The pectoralis muscle area (PMA) is quantifiable by CT and predicts cross-sectional COPD-related morbidity. There are no studies that compare PMA with DXA measures or that evaluate longitudinal relationships between PMA and lung disease progression. Methods Participants from our longitudinal tobacco-exposed cohort had baseline and 6-year chest CT (n=259) and DXA (n=164) data. Emphysema was quantified by CT density histogram parenchymal scoring using the 15th percentile technique. Fat-free mass index (FFMI) and appendicular skeletal mass index (ASMI) were calculated from DXA measurements. Linear regression model relationships were reported using standardised coefficient (β) with 95% CI. Results PMA was more strongly associated with DXA measures than with body mass index (BMI) in both cross-sectional (FFMI: β=0.76 (95% CI 0.65 to 0.86), p<0.001; ASMI: β=0.76 (95% CI 0.66 to 0.86), p<0.001; BMI: β=0.36 (95% CI 0.25 to 0.47), p<0.001) and longitudinal (ΔFFMI: β=0.43 (95% CI 0.28 to 0.57), p<0.001; ΔASMI: β=0.42 (95% CI 0.27 to 0.57), p<0.001; ΔBMI: β=0.34 (95% CI 0.22 to 0.46), p<0.001) models. Six-year change in PMA was associated with 6-year change in emphysema (β=0.39 (95% CI 0.23 to 0.56), p<0.001) but not with 6-year change in airflow obstruction. Conclusions PMA is an accessible measure of muscle mass and may serve as a useful clinical surrogate for assessing skeletal muscle loss in smokers. Decreased PMA correlated with emphysema progression but not lung function decline, suggesting a difference in the pathophysiology driving emphysema, airflow obstruction and comorbidity risk.
12 August 2025 in European Journal of Internal Medicine
Most patients with a chronic disease are multimorbid. This is particularly important in patients with chronic obstructive pulmonary disease (COPD), who on average have five other identified comorbidities that independently impact their health and increase their mortality risk. Using a modified Delphi method, we selected the 20 most important diseases associated with COPD and clustered them into five domains: mental, respiratory, cardiovascular, metabolic and multiple organs loss of tissue. We then developed a systematic approach to characterise the impact and clinical presentation of individual diseases within each cluster, and to define the priority and timing of measurement of the potential markers of disease presence and severity. Given the absence of integrated guidelines to treat multimorbid patients, we reviewed and selected individual disease guidelines or recommendations that can be accessed for specific information related to the management of each disease. In addition, we built a multimorbidity 'Health Dashboard' that, completed by the patient or health practitioner, can help identify the presence and severity of comorbid diseases. By using a practical comprehensive approach, it is possible to identify and characterise important comorbid diseases in patients with COPD, and to implement management tools that should help improve their outcome. This expert consensus commentary summarises patient-centred recommendations to manage comorbidities in COPD patients, aiming to improve quality-of-life and reduce disease burden through a holistic approach. Prospective pragmatic trials comparing such an approach with usual care for multimorbid patients with COPD including long-term follow-up are urgently needed.