Clinical Phenotypes of Difficult-to-treat and Mild Asthma Defined by Cluster Analysis

Clinical Phenotypes of Difficult-to-treat and Mild Asthma Defined by Cluster Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Clinical Phenotypes of Difficult-to-treat and Mild Asthma Defined by Cluster Analysis Mohammad Leily, Heena Mistry, Hongmei Zhang, Maria Larsson, Judit Varkonyi-Sepp, and 17 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9238898/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Background Cluster modelling has demonstrated the heterogeneity of asthma but has previously focused mainly on severe disease with limited assessment of mild disease or treatable traits like comorbidities. Objective To identify and characterise difficult-to-treat and mild asthma clusters in two UK cohorts: Wessex AsThma CoHort of Difficult Asthma (WATCH-DA) and a mild-asthma cohort from the Epigenetics of Severe Asthma study (EOSA-MA). Methods Separate K-means clustering was applied to WATCH-DA (n = 498; 11 variables) and EOSA-MA (n = 67; 12 variables). Post-hoc comparisons evaluated demographic, inflammatory, physiological, comorbidity and patient-reported outcome profiles. Results Six difficult-to-treat and two mild asthma clusters were identified respectively, all Type-2 (T2)-predominant. Difficult-to-treat asthma clusters differed by sex, age of asthma-onset, body mass index (BMI) and comorbidities. Two clinically-controlled clusters, cluster-1 (early-onset–clinically-controlled–atopic disease) and cluster-4 (adult-onset–clinically-controlled–least-atopic disease), showed distinct comorbidity patterns despite lower overall morbidity. Three severe, exacerbation-prone, adult-onset, female predominant difficult-to-treat clusters (cluster-2, cluster-5, cluster-6) varied by blood eosinophil counts (BEC), spirometry, BMI, treatment needs, comorbidities, and quality of life. An adolescent-onset–obese–atopic–airflow-obstructive disease (cluster-3) showed fewer exacerbations but high BEC with worst spirometry and poor asthma control. In mild asthma, cluster-1 (early-onset-atopic-mild-asthma) showed worse pathophysiological indices and asthma control than cluster-2 (adolescent-onset-mild-asthma) but similarly high comorbidity prevalence. Conclusion Characterisation of difficult-to-treat and mild asthma clusters reveals diverse associated clinical traits and outcomes across the asthma severity spectrum. Recognition of these clusters and their associated comorbidities should prompt early personalised asthma management to address both airway-centric and comorbid disease aspects. Asthma Cluster Analysis Comorbidities Difficult-to-treat Asthma Mild Asthma Clinical Phenotypes Precision Medicine T2 Inflammation. Figures Figure 1 Figure 2 Background Asthma is a chronic inflammatory airway disease characterised by variable airflow obstruction, bronchial hyperresponsiveness, and a broad range of clinical presentations. Despite structured guidelines, asthma still remains a clinical burden due to disease heterogeneity with substantial variation in clinical expression and natural history. 1 – 3 This heterogeneity spans age of asthma-onset, atopic status, physiological and inflammatory profiles, symptom burden, comorbidities and treatment response. 4 , 5 To better capture this complexity, research has increasingly focused on asthma phenotypes, defined by particular observable clinical characteristics, and endotypes underpinned by distinct pathobiological mechanisms. 6 , 7 Among various approaches, unsupervised cluster analysis has emerged as a useful method for identifying clinically relevant phenotypes without prior assumptions. Landmark cluster studies, have revealed reproducible severe asthma phenotypes, including early-onset allergic asthma, late-onset eosinophilic asthma, and obese non-eosinophilic asthma. 8 – 13 Prior cluster studies have largely focused on patients with more severe or difficult-to-treat asthma (DA) managed in specialist settings. Moreover, limited work has included patients with mild asthma (MA), and few cluster analyses have systematically assessed comorbid treatable traits such as obesity 14 , gastro-oesophageal reflux disease (GORD) 15 , sinonasal disease 16 , breathing pattern disorder 17 , and psychological comorbidities 18 within clusters. Such comorbidities are increasingly recognised as important modifiers of asthma disease burden, adverse patient outcomes and treatment response, often within a complex framework of multimorbidity. 19 – 21 In this study, we sought to address these knowledge gaps by defining asthma clusters in DA and MA patient cohorts through extensive clinical characterisation using comparable methodology. The Wessex AsThma CoHort of Difficult Asthma (WATCH-DA) is a prospective, UK-based cohort of adults with DA (GINA treatment steps 4 or 5). 22 with extensive characterisation data on asthma-focused clinical and pathophysiological parameters, and comorbidities. Conversely, a MA cohort from the Epigenetics of Severe Asthma study (EOSA-MA) (GINA treatment steps 1 or 2) provides a rare opportunity to evaluate data-driven clinical phenotyping in milder disease. The EOSA-MA cohort includes a subset of patients from the long-standing Isle of Wight Birth Cohort (IOWBC), 23 together enabling detailed clinical and biomarker characterisation. The overall aim of this study is to enhance precision medicine in asthma care by identifying distinct DA and MA clinical phenotypes through assessment of their clinical, physiological, inflammatory and comorbidity profiles to inform future approaches to treatable trait-based stratification and management. The four main objectives of this study were: 1) To identify DA and MA phenotypes in the respective WATCH-DA and EOSA-MA cohorts using K-means clustering; 2) To characterise the clinical, physiological, inflammatory, and comorbidity profiles across identified DA and MA clusters; 3) To compare the disease severity spectrum of asthma phenotypes across mild and difficult-to-treat asthma populations; 4) To inform future approaches to treatable trait-based stratification and precision medicine in asthma care. Methods Study Design and Cohorts This was a cross-sectional, observational study involving two independent UK-based asthma cohorts, the WATCH-DA and EOSA-MA cohorts, using comparable methodology for deeper clinical characterisation. WATCH Difficult-to-Treat Asthma Cohort At the time of study, the WATCH-DA study recruited 501 adult DA patients from July 2015 to March 2019 from University Hospital Southampton NHS Foundation Trust (UHSFT) and the David Hide Asthma and Allergy Centre (DHAAC), Isle of Wight, UK. All patients were managed according to the BTS/SIGN step “high-dose therapies” and/or frequent or continuous oral corticosteroids (OCS) use (GINA treatment steps 4 and 5 equivalent). 3 , 24 Ethics approval and written informed consent from patients was obtained (REC reference: 14/WM/1226). Briefly, comprehensive baseline characterisation was performed at enrolment including: demographics; asthma history; medication use; lung function tests; fractional exhaled nitric oxide (FeNO); blood biomarkers (BECs, blood neutrophils (BNC), total immunoglobulin E (IgE)); GINA Type-2 inflammation (T2)-high status - defined by either FeNO ≥ 20 ppb, BEC ≥ 0.2×10⁹/L (threshold chosen due to laboratory reporting to one decimal place), maintenance OCS (mOCS) use, or clinically allergen-driven disease (≥ 1 positive aeroallergen skin-prick test plus a relevant trigger), with sputum eosinophilia excluded due to limited samples); allergy skin prick testing (SPT) to 13 common aeroallergens; comorbidities defined using conventional clinical criteria ( Supplementary Table 1 ); and health and disease-related questionnaires including Asthma Control Questionnaire-6 (ACQ-6), Nijmegen Score, Hospital Anxiety and Depression Scale (HADS), Sino-Nasal Outcome Test-22 (SNOT-22), Hull Cough Hypersensitivity Questionnaire, EuroQoL 5-Dimension 5-Level health today Visual Analogue Scale (EQ5D-5L VAS), and St George’s Respiratory Questionnaire (SGRQ). Data were stored in a central WATCH-DA database housed at UHSFT and exported to SPSS v26 (IBM, NY, USA) The full study protocol is described elsewhere. 22 . Three patients were excluded due to extreme outlier values for total IgE, leaving 498 patients for clustering analysis ( Supplementary Fig. 1 ). EOSA Mild Asthma Cohort The EOSA-MA cohort included 69 participants with physician-diagnosed MA defined by BTS/SIGN treatment steps 1–2 (GINA steps 1–2), 3,24 recruited from August 2018 to July 2019 as part of a larger NIH-funded EOSA study 25 (total n = 193). A subset (n = 23) was drawn from the IOWBC, a whole-population birth cohort initiated in 1989 to study the natural history of asthma and allergy, with detailed longitudinal data from birth to 26-years of age. 23 The remaining participants (n = 46) were identified from local IOW Allergy Clinics, primary care practices via review of electronic health records and community outreach (including use of the DHAAC website) ( Supplementary Fig. 1 ). Ethics approval for the EOSA study was obtained (REC reference: 18/SC/0105). All participants provided written informed consent. All participants underwent identical clinical, physiological, biomarker, comorbidity and questionnaire assessments to the WATCH-DA cohort. Data were stored securely in a DHAAC-based SPSS dataset. Two participants were excluded as outliers (one for extreme heavy smoking, one for an extreme outlier value for total IgE), leaving 67 participants for clustering analysis. Clustering Variables and Analysis K-means cluster analysis was performed separately for each cohort using PROC FASTCLUS in SAS v9.4 (SAS Institute, Cary, NC, USA). The selection of clinical variables used for cluster analysis is detailed in Supplementary Methods . Categorical variables were transformed into numeric form prior to clustering. The number of clusters was determined using the Cubic Clustering Criterion (CCC), pseudo-F-statistic, and the R² statistic, in combination with clinical interpretability. Eleven clustering variables were used for WATCH-DA: age of asthma-onset, BMI, hospitalisations in past 12 months, Intensive Care Unit (ICU) admissions (ever), number of OCS courses in past 12 months, FeNO, post-bronchodilator (BD) Forced Expiratory Volume in 1 second (FEV₁) % predicted, BEC, serum Total IgE, ACQ-6 score, and sum of positive SPTs. Twelve clustering variables were used for EOSA-MA: age of asthma-onset, BMI, ICU admissions (ever), asthma-related GP visits in past 12 months, number of OCS courses in past 12 months, current inhaled corticosteroids (ICS) use, FeNO, pre-BD FEV₁ % predicted, BEC, serum total IgE, ACQ-6 score, and sum of positive SPTs. The differing and additional clinical variables used for the EOSA-MA cohort included GP visits in past 12 months (instead of hospitalisations in past 12 months) and current ICS use respectively to enable capturing the asthma severity spectrum in this population. Post-Hoc Trait Analysis Following cluster assignment, additional post-hoc analyses were conducted to evaluate differences between clusters in demographics, treatment, inflammatory (only a subset of WATCH-DA patients (n = 139) had induced sputum samples available for differential cell count analysis), physiological, and comorbidity profiles plus health outcomes. Statistical Analysis Cluster membership was used as the grouping variable for all post-hoc trait analyses. Categorical variables were represented as frequency; n (%), and compared across clusters using Chi-squared test, or Fisher’s exact test when expected cell counts < 5. Continuous variables were represented as mean (SD) or median (IQR), according to distribution. Between-cluster differences for continuous data were assessed using one-way ANOVA for normally-distributed variables and Kruskal–Wallis test for skewed variables. Where the omnibus test was significant, exploratory pairwise comparisons between clusters were made using independent-sample t-test or Mann–Whitney U test, as appropriate, to identify which clusters differed. A p-value < 0.05 was considered statistically significant. All post-hoc analyses were conducted using SPSS v26 (IBM, NY, USA). Results Difficult-to-treat asthma clusters Main clustering outcomes Cluster analysis using 11 selected variables identified six discrete DA clusters. A six-cluster solution was selected because, across candidate models with differing numbers of clusters, this showed the most favourable combination of cubic clustering criterion and pseudo-F statistics together with a high between-cluster R², whilst also yielding clinically coherent, non-redundant phenotypes (as detailed in Methods). All clustering variables showed significant differences across these clusters ( Table 1 ). Resulting clustering variable-based characterisation is summarised below and in Figure 1a . Cluster-1 (n=171; 34.3%); Early-onset-clinically-controlled-atopic disease: Earliest (childhood)-onset asthma, raised BMI, low T2 inflammatory biomarkers (BEC/FeNO), best spirometry, better than average asthma control plus low asthma exacerbations and asthma-related admissions. Cluster-2 (n=40; 8.0%); Adult-onset-severe-exacerbation-prone disease : Adult-onset disease, lowest BMI, highest BEC and FeNO, average spirometry and asthma control plus highest exacerbation frequency and admissions. Cluster-3 (n=35; 7.0%); Adolescent-onset-obese-atopic-airflow-obstructive disease : Adolescent-onset disease, obese-level BMI, low FeNO but high BEC and highest atopy, worst spirometry, poorer asthma control plus moderate exacerbation frequency and lower acute admissions. Cluster-4 (n=127; 25.5%); Adult-onset-clinically-controlled-least-atopic disease ; Oldest age of onset, raised BMI, moderate FeNO, low BEC, least atopy, average spirometry, best asthma control, lowest exacerbation frequency, and lower admissions. Cluster-5 (n=33; 6.6%); Young-adult-onset- s evere-exacerbation-prone disease : Young adult-onset disease, raised BMI, low BEC and FeNO, moderate atopy, poorer spirometry, worse asthma control, elevated asthma exacerbation frequency and admissions plus highest frequency of asthma-related ICU admissions. Cluster-6 (n=92; 18.5%); Adult-onset-obese-severe-exacerbation-prone disease: Adult-onset disease, obesity-level BMI, lowest BEC and FeNO, moderate atopy, average spirometry, worst asthma control, and high frequency of asthma exacerbations and admissions. Post-hoc phenotypic cluster characteristics Female predominance was seen in four clusters ( early-onset-clinically-controlled-atopic (cluster-1), a dult-onset-severe-exacerbation-prone (cluster-2) , young-adult-onset-severe-exacerbation-prone (cluster-5) and adult-onset-obese-severe-exacerbation-prone (cluster-6) with no sex predominance seen in the remaining 2 clusters ( Table 2 ). Current age was oldest in a dult-onset-clinically-controlled-least-atopic (cluster-4) and youngest in early-onset-clinically-controlled-atopic (cluster-1) matching age of onset associations. Smoking status broadly separated into never-smoking ( early-onset-clinically-controlled-atopic ( cluster-1), adult-onset-severe-exacerbation-prone ( cluster-2) and adolescent-onset-obese-atopic-airflow-obstructive ( cluster-3)), and ex-smoking ( adult-onset-clinically-controlled-least-atopic ( cluster-4)). Current smoking status was low but mostly seen in adult-onset-obese-severe-exacerbation-prone (cluster-6) and young-adult-onset-severe-exacerbation-prone (cluster-5) . Morbidity and treatment need differed. Days lost from work were highest in young-adult-onset-severe-exacerbation-prone (cluster-5) but lowest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3). Intubation for acute severe asthma exacerbations was highest in young-adult-onset-severe-exacerbation-prone (cluster-5), and lowest ina dult-onset-clinically-controlled-least-atopic (cluster-4). T2 biologics use at study enrolment were highest in young-adult-onset-severe-exacerbation-prone (cluster-5) and lowest in adult-onset-obese-severe-exacerbation-prone (cluster-6). Omalizumab use was highest in young-adult-onset-severe-exacerbation-prone (cluster-5), and lowest in a dult-onset-clinically-controlled-least-atopic-disease (cluster-4). Post-study enrolment T2 biologics initiation was highest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3) plus young-adult-onset-severe-exacerbation-prone (cluster-5), whilst lowest in early-onset-clinically-controlled-atopic (cluster-1). Antifungal (itraconazole) use was highest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3), and lowest in adult-onset-obese-severe-exacerbation-prone (cluster-6). Mean ICS dose, mOCS, prophylactic antibiotics and mepolizumab use at study enrolment did not differ significantly by cluster ( Table 2 ). Baseline inflammatory profiles Although GINA-defined T2-high status displayed some cluster variation ( p =0.049), all clusters showed overwhelming T2 predominance ranging from 88.2% in adult-onset-obese-severe-exacerbation-prone (cluster-6) to 100% in a dult-onset-severe-exacerbation-prone (cluster-2) and adolescent-onset-obese-atopic-airflow-obstructive (cluster-3). Aspergillus fumigatus-specific IgE was highest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3) and lowest in adult-onset-obese-severe-exacerbation-prone (cluster-6). In a subset of WATCH-DA patients (n = 139) who had induced sputum samples and differential counts available, sputum eosinophils and eosinophil proportion ≥2% were both highest in adult-onset-severe-exacerbation-prone (cluster-2) and lowest in early-onset-clinically-controlled-atopic (cluster-1). Mean sputum neutrophils, neutrophil proportion >61%, and sputum inflammatory phenotypes did not differ significantly across clusters ( Table 2 ). Baseline physiological characteristics Clinic (post-BD) spirometry differed across clusters ( Supplementary Table 2 ). Early-onset-obese-atopic-airflow-obstructive (cluster-3) showed the worst spirometry (lowest FEV₁ %pred, Forced Vital Capacity (FVC) %pred, FEV₁/FVC ratio and Forced Expiratory Flow between 25% and 75% of vital capacity (FEF₂₅ % – ₇₅ % )), whereas adult-onset-obese-severe-exacerbation-prone ( cluster-6) had the least airflow obstruction (highest FEV₁/FVC and FEF₂₅ % – ₇₅ % ). RV/TLC ratio also differed significantly; worst in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3), and best in early-onset-clinically-controlled-atopic (cluster-1) ( Supplementary Table 2 ). Gas transfer indices did not differ across clusters. Comorbidity characteristics Comorbidities significantly differed by cluster ( Table 3 and Figure 1b ). Obesity (BMI≥ 30) was highest in adult-onset-obese-severe-exacerbation-prone (cluster-6) and lowest in adult-onset-severe-exacerbation-prone (cluster-2). Breathing pattern disorder was highest in young-adult-onset-severe-exacerbation-prone (cluster-5) and lowest in adult-onset-clinically-controlled-least-atopic (cluster-4). Depression/anxiety were highest in adult-onset-obese-severe-exacerbation-prone (cluster-6) but lowest in young-adult-onset-severe-exacerbation-prone (cluster-5) and adult-onset-clinically-controlled-least-atopic (cluster-4). Eczema was highest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3) and lowest in adult-onset-clinically-controlled-least-atopic (cluster-4). Nasal polyps were predominant in adult-onset-clinically-controlled-least-atopic (cluster-4) and lowest in early-onset-clinically-controlled-atopic (cluster-1). Salicylate sensitivity was highest in young-adult-onset-severe-exacerbation-prone ( cluster-5) and lowest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3). Allergic bronchopulmonary aspergillosis (ABPA)/Severe Asthma with Fungal Sensitisation (SAFS) were highest in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3) and lowest in adult-onset-obese-severe-exacerbation-prone (cluster-6). Chronic Obstructive Pulmonary Disease (COPD) was predominant in adult-onset-clinically-controlled-least-atopic (cluster-4) and lowest in early-onset-clinically-controlled-atopic ( cluster-1). Rhinitis and GORD were ubiquitous with no difference between clusters ( Table 3 ). Comorbidity-related questionnaires differed significantly by cluster ( Table 3 ). SNOT-22 was highest in adult-onset-obese-severe-exacerbation-prone (cluster-6) and lowest in adult-onset-clinically-controlled-least-atopic ( cluster-4). Hull cough hypersensitivity score was also highest in adult-onset-obese-severe-exacerbation-prone (cluster-6) but lowest in early-onset-clinically-controlled-atopic (cluster-1). Nijmegen questionnaire scores (and proportion with score >23) were highest in adult-onset-obese-severe-exacerbation-prone (cluster-6), and lowest in adult-onset-severe-exacerbation-prone (cluster-2) and adult-onset-clinically-controlled-least-atopic (cluster-4). HADS-D (depression component) was also highest in adult-onset-obese-severe-exacerbation-prone (cluster-6) with lowest proportions of HADS-D ≥11 in adolescent-onset-obese-atopic-airflow-obstructive (cluster-3) and adult-onset-clinically-controlled-least-atopic (cluster-4). Quality of life (QoL) mirrored symptom burden, with the worst EQ-5D-5L health today VAS and SGRQ total score in adult-onset-obese-severe-exacerbation-prone (cluster-6) ( Table 3 ). Mild asthma clusters Main clustering outcomes Cluster analysis with 12 selected variables identified two discrete MA clusters. A two-cluster solution was selected favouring clinically coherent, non-redundant MA phenotypes. Clustering variables with significant differences between clusters included BEC, Total IgE, FeNO, FEV₁% pred., atopic status, and ACQ-6 ( Table 4 ). Resulting cluster-based characterisation is summarised below and in Figure 1a . Cluster-1 (n=20; 29%); Early-onset-atopic-mild-asthma: Early (childhood)-onset asthma with significantly raised FeNO, BEC and Total IgE, greater atopy, lower FEV 1 plus higher ACQ-6. Cluster-2 (n=47; 70.1%); Adolescent-onset-mild-asthma . Early (adolescent)-onset asthma with lower expression of inflammatory biomarkers and atopy plus better FEV 1 and ACQ-6. Post-hoc phenotypic characteristics Male predominance was observed for early-onset-atopic-mild-asthma (cluster-1) whilst adolescent-onset-mild-asthma (cluster-2) showed female predominance . Other clinical parameters including current age, smoking status, morbidity, and medication use did not differ significantly between clusters ( Table 4 ). Baseline inflammatory profiles Early-onset-atopic-mild-asthma (cluster-1) showed higher Aspergillus fumigatus-specific IgE, BNC and sputum inflammatory markers (% eosinophils, % neutrophils; inflammatory phenotype), but did not significantly differ due to limited induced sputum availability. GINA T2-high status was prevalent across both clusters but highest in early-onset-atopic-mild-asthma (cluster-1) ( Table 4 ). Baseline physiological characteristics Lung function tests (spirometry, lung volumes and gas transfer factors) were within normal range for both clusters but modestly lower in early-onset-atopic-mild-asthma (cluster-1); i.e. pre-BD FEV₁ %pred., FEV₁/FVC ratio, and FEF₂₅ % - ₇₅ % ( Supplementary Table 3 ). Comorbidity characteristics Comorbidities were highly prevalent among both MA clusters ( Table 4 ). Obesity (BMI≥ 30) was more prevalent in a dolescent-onset-mild-asthma (cluster-2). Other comorbidities and comorbidity-related questionnaires outcomes were similar between clusters ( Table 4 ). Summary characterisation of difficult-to-treat and mild asthma cluster phenotypes A summary of the six DA and two MA cluster phenotypes respectively, incorporating clustering and morbidity characteristics is shown in Figure 1a . Discussion Using a common assessment platform across two cohorts spanning the asthma severity spectrum, we identified six difficult-to-treat clusters and two mild-asthma clusters with distinct clinical, inflammatory and comorbidity profiles. A key novel focus of this study was the exploration of comorbidities and highlighting their role in asthma heterogeneity. One important finding is that despite using an identical phenotyping platform for comprehensive clinical characterisation, heterogeneity is already apparent in milder asthma and does not map directly onto patterns seen in more difficult-to-treat disease. Separate K-means clustering analyses for the WATCH-DA and EOSA-MA cohorts using identical clinical and biomarker assessments revealed distinct phenotypic profiles across both groups. Additionally, we move beyond description by linking readily available bedside clinical cues - sex, age of onset, BMI, simple T2 biomarkers and spirometry phenotypes - through a Clinician Recognition Guide ( Table 5 ). Although T2-high biology was common, it did not alone explain variation in symptoms, physiology, or multimorbidity. Among difficult-to-treat clusters, disease severity, sex, age of onset, and BMI varied alongside airflow limitation and healthcare use. The mild asthma clusters, despite preserved spirometry, were clinically and biologically distinct with significant comorbidities suggesting multimorbid burden. Comparisons across asthma severities indicate that burden is shaped by combinations of pulmonary and extra-pulmonary traits rather than a single linear progression from mild to difficult-to-treat disease. This perspective clarifies how the phenotypes relate across the disease spectrum and provides a practical basis for trait-targeted, guideline-aligned care. Our results reflect and extend existing literature showing reproducible, though not identical, asthma clinical phenotypes, including early-onset atopic T2-high, adult-onset eosinophilic, and obese difficult asthma,. 8,9,11 While the cluster characteristics inevitably depend on the clinical and pathobiological parameters collected, these studies consistently demonstrate that clinically recognisable clusters can share T2 features yet diverge in physiology and symptom burden, emphasising the need for multi-domain appraisal rather than reliance on T2-biomarkers alone. Large population and longitudinal cohorts further underscore the breadth of multimorbidity across asthma severities, with adult-onset disease and obesity, particularly among women, associated with greater morbidity over time. 26,27 While our cross-sectional design precludes inference about directionality, the prominence of comorbidities in both the difficult-to-treat and mild asthma cohorts is concordant with these observations and merits routine clinical attention. Obesity, which has long reached epidemic proportions, warrants particular discussion. In our difficult-to-treat clusters, higher BMI co-occurred with greater symptom burden and lower T2 signals in subsets, while in mild asthma a lower-T2, obesity-linked group exhibited good spirometry but notable patient-reported burden. Observational and intervention studies suggest that weight loss can be associated with improvements in asthma control and quality of life in individuals with obesity; however, responses vary and should be considered alongside optimisation of inhaled therapy and comorbidity management. 26-29 We therefore interpret obesity as a clinically relevant trait that frequently co-occurs with asthma across the asthma severity spectrum. Sex and age of onset patterns were also informative. Prior work has described female predominance in obesity-linked phenotypes and higher risk associated with adult-onset disease; our clusters showed similar associations. 8,9,11,27 These readily available cues: sex, age of onset, and BMI, remain practical anchors for recognising phenotypes, particularly when considered alongside simple T2-biomarkers, basic spirometry and a focused review of extra-pulmonary traits. Comparison across both cohorts showed that one mild asthma cluster exhibited greater morbidity but did not map one-to-one onto any single difficult-to-treat asthma cluster. Instead, overlapping mixtures of traits were seen across severities, supporting the view that heterogeneity is also present in upstream milder disease and amplifies with accrued comorbidity regardless of asthma severity. This observation argues for early identification and management of co-existing traits between mild and difficult-to-treat clusters even when spirometry is near-normal. Our findings could provide a framework to support a treatable-traits approach in which pulmonary, extra-pulmonary and behavioural traits are identified systematically and addressed in parallel with guideline-directed care. Randomised and translational studies indicate that trait-targeted pathways can improve outcomes relative to usual clinical care, and that distinct trait profiles differ in burden and responsiveness to systematic assessment. 30-32 Against this backdrop, we frame the clinical management of our clusters according to their predominant component treatable traits. Active T2-high traits: where BEC and/or FeNO remain raised and morbidity persists despite optimised inhaled therapy, consideration of add-on T2-biologics within steroid-sparing pathways and using guideline eligibility criteria, is consistent with current strategy documents. 33,34 Obesity-associated, symptom-dominant presentations with lower T2 signals: first-line emphasis on structured weight reduction management and physical activity is supported by trials showing clinically meaningful improvements in QoL and, in some studies, airway inflammation with ~5–10% weight loss alongside continued inhaled therapy optimisation in parallel. 28,29 Upper-airway comorbidity e.g. chronic rhinosinusitis with nasal polyps [CRSwNP] : where sinonasal disease co-exists, ENT-led care and, for eligible patients, biologics for CRSwNP may reduce overall burden if locally available. Selection should follow the European Position Paper on Rhinosinusitis and Nasal Polyps/ European Forum for Research and Education in Allergy and Airway Diseases (EPOS/EUFOREA) criteria. 35 Breathing pattern disorder: physiotherapy-based breathing retraining can improve asthma-related QoL as an adjunct to standard care; screening and referral are vital where this trait is suspected. 36 GORD and cough hypersensitivity: management should follow guideline-based indications. 37,38 Where T2-biomarkers are low and extra-pulmonary drivers predominate, escalation of corticosteroids is unlikely to help and should be avoided. 1,3 The Clinician Recognition Guide ( Table 5 ) operationalises these principles at the bedside for individual patient clusters. Sex, age of onset, and BMI, together with simple T2-biomarkers, spirometry and a brief scan for sentinel comorbidities; including GORD, obesity, CRSwNP, breathing pattern disorder, inducible laryngeal obstruction, and smoking/asthma-COPD overlap, typically suffice to place a patient within a phenotype and to select the first-line action. The table presents associative cues and management considerations, not treatment mandates; add-on therapies should be chosen using clinical and biomarker criteria, shared decision-making and local availability, alongside non-pharmacological interventions for comorbid traits. Our study has both strengths and limitations. Strengths include harmonised clustering across two well-phenotyped cohorts which have been identically characterised. Use of clinically relevant clustering variables adds further robustness. Also, deep post-hoc characterisation spanning inflammatory, physiology, comorbidities and patient-reported outcomes/HRQoL assessments of both mild and difficult-to-treat asthma patients. One potential limitation is the cross-sectional design, and future studies would benefit from assessing longitudinal cluster stability and changing outcomes based on our recommendations of care provision in Table 5 . Limited induced sputum and airway sampling, particularly in the MA cohort, may be viewed as another shortcoming. Nevertheless, an adequate proportion of airway sampling was achieved in the WATCH-DA cohort permitting valid inferences. While the study might be viewed as single centre, the WATCH-DA cohort is drawn from a tertiary referral site that effectively serves as a multi-centre checkpoint. These considerations do not alter the central practice message that comorbidity-linked treatable traits drive multimorbid burden across phenotypes. Future steps include using these clusters to validate a minimal clinical recognition algorithm (age of onset, BMI, atopy, FeNO/BEC plus 2–3 comorbidity traits) against patient outcomes and treatment response; to assess phenotype-tailored therapeutic interventions; to prospectively evaluate OCS-sparing strategies in high-burden groups; and integrate omics/imaging where comorbidity signals are strongest (obesity, CRSwNP, AERD, ABPA/SAFS), incorporating sex-specific analyses. In summary, across difficult-to-treat and mild asthma cohorts assessed with a comparable clinical characterisation platform, clustering reveals clinically coherent associations between biomarker patterns, physiological measures and multimorbidity on a largely T2-predominant background. Used cautiously in alignment with guidelines and trait-targeted evidence, these patterns can help clinicians recognise asthma phenotypes rapidly and instigate first-line interventions earlier for active treatable traits. Abbreviations ABPA, Allergic bronchopulmonary aspergillosis; ACQ6, Asthma Control Questionnaire-6; BD, bronchodilator; BEC, blood eosinophil count; BMI, Body Mass Index; BN, blood neutrophils count; BTS, British Thoracic Society; CCC, Cubic Clustering Criterion; COPD, Chronic Obstructive Pulmonary Disease; CRSwNP, chronic rhinosinusitis with nasal polyps; DA, Difficult Asthma; DHAAC, David Hide Asthma and Allergy Centre; EOSA, Epigenetics of Severe Asthma; EQ-5D-5L VAS, EuroQol 5-Dimension 5-Level health today Visual Analogue Scale; FEF 25% - 75% , forced expiratory flow between 25% and 75% of vital capacity; FeNO, fractional exhaled nitric oxide; FEV₁, Forced Expiratory Volume in 1 second; FVC, Forced Vital Capacity; GORD, gastro-oesophageal reflux disease; GINA, Global Initiative for Asthma; GP, general practice; HADS, Hospital Anxiety and Depression Scale; ICU, Intensive Care Unit; IgE, immunoglobulin E; IOWBC, Isle of Wight Birth Cohort; MA, mild asthma; NICE, National Institute for Health and Care Excellence; OCS, oral corticosteroids; QoL, Quality of life; SAFS, severe asthma with fungal sensitisation; SGRQ, St George’s Respiratory Questionnaire; SNOT-22, Sino-Nasal Outcome Test; SPT, skin prick testing; T2, Type-2 inflammation; WATCH, Wessex AsThma CoHort of Difficult Asthma. Declarations Ethics approval and consent to participate Ethical approval for the Wessex AsThma CoHort of difficult asthma (WATCH-DA) was granted by the West Midlands – Solihull Research Ethics Committee (REC reference: 14/WM/1226). The Epigenetics of Severe Asthma (EOSA) study was approved by the South-Central Hampshire B – Southampton Research Ethics Committee (REC reference: 18/SC/0105). All participants provided written informed consent prior to inclusion in the studies. Availability of data and materials The datasets generated and/or analysed during the current study are not publicly available due to participant confidentiality and governance restrictions associated with the WATCH-DA and EOSA cohorts but are available from the corresponding author on reasonable request. Conflicts of Interest The authors, ML, HM, HZ, ML, JV, BA, LW, CE, JJH, AL, AF, HMH, PD, SH, GS, PV, MAK, EV, ED, SHA, RJK, declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Professor Ratko Djukanovic reports personal fees, has shares in the company and is a consultant to Synairgen, and personal fees from GlaxoSmithKline and Kymab, outside the submitted work. Funding The Wessex AsThma CoHort of difficult asthma (WATCH) study has been supported by the NIHR Southampton BRC and Clinical Research Facility UHSFT, UK. The WATCH study itself is not externally funded. Funding assistance for database support for the WATCH study was initially obtained from a non-promotional grant from Novartis (£35,000). Funding assistance for patient costs (e.g. parking) was initially provided by a charitable grant (£3500) from the Asthma, Allergy & Inflammation Research (AAIR) Charity. Lastly, funding assistance for the Epigenetics of Severe Asthma (EOSA) study was obtained from a National Institutes of Health (NIH) grant (£400,000) in collaboration with La Jolla Institute of Immunology, La Jolla, California, USA. Author Contributions HM, HZ, SHA, RJK were responsible for the conceptualisation of the study and designed the methodology. Data collection was performed by HM, ML, AF, HMH, PD, MAK, RJK. Data curation and formal analysis were performed by HM, HZ, SHA, RJK. ML, HM and RJK drafted the manuscript and all authors reviewed, edited, and approved the final version of the manuscript. Acknowledgements The authors wish to thank the patients who participated in this study. They also wish to acknowledge the contributions of the wider WATCH study team at University Hospital Southampton NHS Foundation Trust (UHSFT) and University of Southampton, and the research teams at the David Hide Asthma and Allergy Centre, Isle of Wight and La Jolla Institute of Immunology, USA. The authors wish to acknowledge the support of the Southampton NIHR Biomedical Research Centre (BRC) and Clinical Research Facility. The Clinical Research Facility and BRC are funded by Southampton NIHR and are a partnership between the University of Southampton and UHSFT. The authors wish to thank all those who made this study possible. The authors also acknowledge funding support from Novratis, NIH and the AAIR charity. References Global Initiative for A. Global Strategy for Asthma Management and Prevention. GINA; 2022. Boulet L-P. Difficult-to-Treat and Severe Asthma in adolescents and adults: diagnosis and management (GINA Pocket Guide). Global Initiative for Asthma; 2019. British Thoracic S, Scottish Intercollegiate Guidelines N. British guideline on the management of asthma (SIGN 153): SIGN, 2016. Wenzel SE. Asthma phenotypes: the evolution from clinical to molecular approaches. Nat Med. 2012;18(5):716–25. Chung KF. Precision medicine in asthma: linking phenotypes to targeted treatments. Curr Opin Pulm Med. 2018;24(1):4–10. Lötvall J, Akdis CA, Bacharier LB. Asthma endotypes: a new approach to classification of disease entities within the asthma syndrome. J Allergy Clin Immunol. 2011;127(2):355–60. Kuruvilla ME, Lee FE, Lee GB. Understanding Asthma Phenotypes, Endotypes, and Mechanisms of Disease. Clin Rev Allergy Immunol. 2019;56(2):219–33. Haldar P, Pavord ID, Shaw DE. Cluster analysis and clinical asthma phenotypes. Am J Respir Crit Care Med. 2008;178(3):218–24. Lefaudeux D, De Meulder B, Loza MJ. U-BIOPRED clinical adult asthma clusters linked to a subset of sputum omics. J Allergy Clin Immunol. 2017;139(6):1797–807. Newby C, Heaney LG, Menzies-Gow A. Statistical cluster analysis of the British Thoracic Society Severe Refractory Asthma Registry: clinical outcomes and phenotype stability. PLoS ONE. 2014;9(7):e102987. Moore WC, Meyers DA, Wenzel SE. Identification of asthma phenotypes using cluster analysis in the Severe Asthma Research Program. Am J Respir Crit Care Med. 2010;181(4):315–23. Fitzpatrick AM, Teague WG, Meyers DA, et al. Heterogeneity of severe asthma in childhood: confirmation by cluster analysis of children in the National Institutes of Health/National Heart, Lung, and Blood Institute Severe Asthma Research Program. J Allergy Clin Immunol. 2011;127(2):382–e91. Fitzpatrick AM, Moore WC. Severe Asthma Phenotypes - How Should They Guide Evaluation and Treatment? J Allergy Clin Immunol Pract. 2017;5(4):901–8. Bal C, Pohl W, Milger K, et al. Characterization of Obesity in Severe Asthma in the German Asthma Net. J Allergy Clin Immunol Pract. 2023;11(11):3417–e243. Tariq K, Schofield JPR, Nicholas BL, et al. Sputum proteomic signature of gastro-oesophageal reflux in patients with severe asthma. Respir Med. 2019;150:66–73. Langdon C, Mullol J. Nasal polyps in patients with asthma: prevalence, impact, and management challenges. J Asthma Allergy. 2016;9:45–53. Freeman A, Abraham S, Kadalayil L, et al. Associations of Breathing Pattern Disorder and Nijmegen Score With Clinical Outcomes in Difficult-to-Treat Asthma. J Allergy Clin Immunol Pract. 2024;12(4):938–e476. Fong WCG, Rafiq I, Harvey M et al. The Detrimental Clinical Associations of Anxiety and Depression with Difficult Asthma Outcomes. J Pers Med 2022; 12(5). Scelo G, Torres-Duque CA, Maspero J, et al. Analysis of comorbidities and multimorbidity in adult patients in the International Severe Asthma Registry. Ann Allergy Asthma Immunol. 2024;132(1):42–53. Shackleford A, Heaney LG, Redmond C, McDowell PJ, Busby J. Clinical remission attainment, definitions, and correlates among patients with severe asthma treated with biologics: a systematic review and meta-analysis. Lancet Respir Med. 2025;13(1):23–34. Kurukulaaratchy RJ, Freeman A, Bansal AT, et al. Evaluation of the effect of multimorbidity on difficult-to-treat asthma using a novel score (MiDAS): a multinational study of asthma cohorts. Lancet Respir Med. 2025;13(9):821–32. Azim A, Mistry H, Freeman A. Protocol for the Wessex AsThma CoHort of difficult asthma (WATCH): a pragmatic real-life longitudinal study of difficult asthma in the clinic. BMC Pulm Med. 2019;19(1):99. Arshad SH, Patil V, Mitchell F. Cohort Profile Update: The Isle of Wight Whole Population Birth Cohort (IOWBC). Int J Epidemiol. 2020;49(4):1083–4. Global Initiative for A. Global Strategy for Asthma Management and Prevention. Global Initiative for Asthma (GINA); 2016. Herrera-De La Mata S, Ramírez-Suástegui C, Mistry H, et al. Cytotoxic CD4(+) tissue-resident memory T cells are associated with asthma severity. Med. 2023;4(12):875–e978. Kankaanranta H, Viinanen A, Ilmarinen P, et al. Comorbidity Burden in Severe and Nonsevere Asthma: A Nationwide Observational Study (FINASTHMA). J Allergy Clin Immunol Pract. 2024;12(1):135–e459. Backman H, Stridsman C, Hedman L, et al. Determinants of Severe Asthma - A Long-Term Cohort Study in Northern Sweden. J Asthma Allergy. 2022;15:1429–39. Scott HA, Gibson PG, Garg ML, et al. Dietary restriction and exercise improve airway inflammation and clinical outcomes in overweight and obese asthma: a randomized trial. Clin Exp Allergy. 2013;43(1):36–49. Freitas PD, Ferreira PG, Silva AG, et al. The Role of Exercise in a Weight-Loss Program on Clinical Control in Obese Adults with Asthma. A Randomized Controlled Trial. Am J Respir Crit Care Med. 2017;195(1):32–42. Agusti A, Bel E, Thomas M, et al. Treatable traits: toward precision medicine of chronic airway diseases. Eur Respir J. 2016;47(2):410–9. McDonald VM, Clark VL, Cordova-Rivera L, Wark PAB, Baines KJ, Gibson PG. Targeting treatable traits in severe asthma: a randomised controlled trial. Eur Respir J 2020; 55(3). Lin T, Pham J, Denton E, et al. Trait profiles in difficult-to-treat asthma: Clinical impact and response to systematic assessment. Allergy. 2023;78(9):2418–27. British Thoracic S, National Institute for H, Care E, Scottish Intercollegiate Guidelines N. Asthma: diagnosis, monitoring and chronic asthma management (BTS, NICE, SIGN). London: National Institute for Health and Care Excellence (NICE); 2024. Global Initiative for A. Global Strategy for Asthma Management and Prevention. (2025 update): Global Initiative for Asthma (GINA), 2025. Fokkens WJ, Viskens AS, Backer V, et al. EPOS/EUFOREA update on indication and evaluation of Biologics in Chronic Rhinosinusitis with Nasal Polyps 2023. Rhinology. 2023;61(3):194–202. Bruton A, Lee A, Yardley L, et al. Physiotherapy breathing retraining for asthma: a randomised controlled trial. Lancet Respir Med. 2018;6(1):19–28. Parker SM. British Thoracic Society clinical statement on chronic cough in adults. Thorax. 2023;78(Suppl 6):s3–19. National Institute for H, Care E. Gastro-oesophageal reflux disease and dyspepsia in adults: investigation and management. London: National Institute for Health and Care Excellence (NICE); 2014. Tables Tables 1 to 5 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Additionalfile1.docx Tables.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 25 Apr, 2026 Reviews received at journal 02 Apr, 2026 Reviewers agreed at journal 01 Apr, 2026 Reviewers invited by journal 01 Apr, 2026 Editor assigned by journal 30 Mar, 2026 Submission checks completed at journal 30 Mar, 2026 First submitted to journal 26 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9238898","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":616875460,"identity":"7187fafa-7469-4ccb-a679-20017054b50c","order_by":0,"name":"Mohammad Leily","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Mohammad","middleName":"","lastName":"Leily","suffix":""},{"id":616875461,"identity":"59ba4935-b79c-483d-b4e3-96e2c5723252","order_by":1,"name":"Heena Mistry","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Heena","middleName":"","lastName":"Mistry","suffix":""},{"id":616875462,"identity":"81ecf0b1-1f4c-4b3b-823d-294083b6d08e","order_by":2,"name":"Hongmei Zhang","email":"","orcid":"","institution":"University of Memphis","correspondingAuthor":false,"prefix":"","firstName":"Hongmei","middleName":"","lastName":"Zhang","suffix":""},{"id":616875463,"identity":"b000c112-c468-4920-9d2d-5949935fc86c","order_by":3,"name":"Maria Larsson","email":"","orcid":"","institution":"St Mary’s Hospital, Isle of Wight","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"","lastName":"Larsson","suffix":""},{"id":616875464,"identity":"8efb3641-6d7a-416b-9f8e-c47a27cc79dd","order_by":4,"name":"Judit Varkonyi-Sepp","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Judit","middleName":"","lastName":"Varkonyi-Sepp","suffix":""},{"id":616875465,"identity":"c4f56e5d-99f7-4190-9c41-7641cd9c9c0a","order_by":5,"name":"Ben Ainsworth","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Ben","middleName":"","lastName":"Ainsworth","suffix":""},{"id":616875467,"identity":"13f1619b-2b78-4a6f-9503-4db51b0000ef","order_by":6,"name":"Liuyu Wei","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Liuyu","middleName":"","lastName":"Wei","suffix":""},{"id":616875471,"identity":"2d218c7a-af70-45b3-8c10-7645ae8ed462","order_by":7,"name":"Chellan Eames","email":"","orcid":"","institution":"University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Chellan","middleName":"","lastName":"Eames","suffix":""},{"id":616875472,"identity":"e8425b0b-7938-48b1-bd30-b8e2d2feaf05","order_by":8,"name":"JJ Hudson-Colby","email":"","orcid":"","institution":"University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"JJ","middleName":"","lastName":"Hudson-Colby","suffix":""},{"id":616875473,"identity":"fd871f66-d8c0-4c34-aa76-145784e6274c","order_by":9,"name":"Adam Lewis","email":"","orcid":"","institution":"University of Southampton","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"","lastName":"Lewis","suffix":""},{"id":616875474,"identity":"e502a042-d5d9-468c-a998-084213fa4c75","order_by":10,"name":"Anna Freeman","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"","lastName":"Freeman","suffix":""},{"id":616875475,"identity":"fbf89d46-c6e0-40cc-b356-c8b5d969c61d","order_by":11,"name":"Hans Michael Haitchi","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Hans","middleName":"Michael","lastName":"Haitchi","suffix":""},{"id":616875476,"identity":"7d4b0cfe-d737-4b4a-9390-246b314fac35","order_by":12,"name":"Paddy Dennison","email":"","orcid":"","institution":"University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Paddy","middleName":"","lastName":"Dennison","suffix":""},{"id":616875477,"identity":"f59e9a7e-42e3-457e-af7b-f1d02bba0236","order_by":13,"name":"Ratko Djukanovic","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Ratko","middleName":"","lastName":"Djukanovic","suffix":""},{"id":616875478,"identity":"42cdf783-3a33-4221-8f73-37d85025870b","order_by":14,"name":"Sara Herrera-De la Mata","email":"","orcid":"","institution":"University of California, San Diego","correspondingAuthor":false,"prefix":"","firstName":"Sara","middleName":"Herrera-De la","lastName":"Mata","suffix":""},{"id":616875479,"identity":"582424df-c312-4fbf-a9b4-371f6a5f0d42","order_by":15,"name":"Grégory Seumois","email":"","orcid":"","institution":"San Diego Biomedical Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Grégory","middleName":"","lastName":"Seumois","suffix":""},{"id":616875480,"identity":"4d422ae5-8817-4448-adc1-aa96de8b0d27","order_by":16,"name":"Pandurangan Vijayanand","email":"","orcid":"","institution":"La Jolla Institute For Allergy \u0026 Immunology","correspondingAuthor":false,"prefix":"","firstName":"Pandurangan","middleName":"","lastName":"Vijayanand","suffix":""},{"id":616875481,"identity":"e20c2a29-ace3-47e6-bcec-0232a2e5a98f","order_by":17,"name":"Mohammed A Kyyaly","email":"","orcid":"","institution":"Southampton Solent University","correspondingAuthor":false,"prefix":"","firstName":"Mohammed","middleName":"A","lastName":"Kyyaly","suffix":""},{"id":616875487,"identity":"533c8f18-0022-4c9f-9630-88abbb34fcff","order_by":18,"name":"Elena Vorobeva","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"Vorobeva","suffix":""},{"id":616875491,"identity":"9b9b125b-e8ee-438e-a3f8-be6654ad9dbf","order_by":19,"name":"Elena Dabrio-Reina","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"Dabrio-Reina","suffix":""},{"id":616875492,"identity":"450d6d06-3651-4369-885f-5d16b935596d","order_by":20,"name":"Syed Hasan Arshad","email":"","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Syed","middleName":"Hasan","lastName":"Arshad","suffix":""},{"id":616875493,"identity":"af6d0778-9d10-4754-a43a-0ef5ff161d21","order_by":21,"name":"Ramesh J Kurukulaaratchy","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIie3RMQrCMBSA4VcCcYm4BgRzhZTs9SoRoZ0KjgUH41JH125ewSO8EohLD+AqgrubThqcdEkdBfOPIR+8lwDEYj/Y1DhAkI/J2xkNE4kuQaiIAiBfE0oQOjIz35PW6fZe02K32bbnBWQCeK6DxO+Cdlizct9ZohqYp4bnGCQZHIxNal7u+ZyOGRANvDBBojzxg8lCNC+y6iepHwxZpzUcX8R60jOYf2RtWYWp30UpJg9pzS46TI5OXW8ShdisT2dWLcVokMsgAf4xhez9Fd/I9F6JxWKxf+8JwYdImgHzgGcAAAAASUVORK5CYII=","orcid":"","institution":"National Institute for Health Research (NIHR), University Hospital Southampton NHS Foundation Trust","correspondingAuthor":true,"prefix":"","firstName":"Ramesh","middleName":"J","lastName":"Kurukulaaratchy","suffix":""}],"badges":[],"createdAt":"2026-03-27 01:39:01","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9238898/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9238898/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106307662,"identity":"cbface8b-61cb-4aa4-a983-78e1105fab77","added_by":"auto","created_at":"2026-04-07 10:06:15","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":91404,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea. Summary characterisation of the 6 difficult-to-treat asthma clusters and 2 mild asthma clusters.\u003c/strong\u003eTop: A doughnut chart displaying the 6 difficult-to-treat asthma clusters according to cluster frequency. The boxes summarise each difficult-to-treat asthma cluster phenotype according to age of onset, BMI, morbidity, inflammatory and comorbidity profiles. The colours for each cluster depict the asthma severity spectrum; peach (cluster-1) and green (cluster-4) representing the lower disease severity clusters, whilst orange (cluster-2), red (cluster-3) and burgundy (cluster-5) highlight the more severe asthma spectrum clusters with increasing colour intensity. Pink (cluster-6) represents the more difficult-to-treat asthma cluster phenotype. Bottom: A doughnut chart of the 2 mild asthma clusters according to cluster frequency. The boxes summarise each mild asthma cluster phenotype according to age of onset, BMI, morbidity, inflammatory and comorbidity profiles. Abbreviations: BMI, body mass index; T2, type-2; BEC, blood eosinophil count; FeNO, fractional exhaled nitric oxide; ICU, Intensive Care Unit; IgE, immunoglobulin E; FEV1, forced expiratory volume in 1 second; ACQ-6, Asthma Control Questionnaire-6.\u003c/p\u003e","description":"","filename":"1a.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9238898/v1/a5fc4093d02e067684fbb1a0.jpg"},{"id":106403975,"identity":"414c1ba9-c5ff-4da0-abe5-9de105322a74","added_by":"auto","created_at":"2026-04-08 09:15:18","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":33651,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure 1b. Comorbidity characteristics of the 6 difficult-to-treat asthma cluster phenotypes.\u003c/strong\u003e Heatmap displaying the mean frequency of comorbidities for each difficult-to-treat asthma cluster. All comorbidities except for rhinitis and GORD were statistically significant (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05; see Table 3). Abbreviations: GORD, gastro-oesophageal reflux disease; BPD, breathing pattern disorder; ABPA/SAFS, allergic bronchopulmonary aspergillosis/severe asthma with fungal sensitisation; COPD, chronic obstructive pulmonary disease.\u003c/p\u003e","description":"","filename":"1b.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9238898/v1/59e25778b6e8a9b1c5132d9a.jpg"},{"id":106405975,"identity":"48999510-012b-4cf9-8645-496f1b6a7371","added_by":"auto","created_at":"2026-04-08 09:29:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1115708,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9238898/v1/4de17e0d-becd-4904-841f-7ced9c04c6d0.pdf"},{"id":106307661,"identity":"3d713562-711b-4b57-8869-7c3ba0f72832","added_by":"auto","created_at":"2026-04-07 10:06:15","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":245746,"visible":true,"origin":"","legend":"","description":"","filename":"Additionalfile1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9238898/v1/f388563800b7267be6819cfe.docx"},{"id":106307664,"identity":"a1ee67ec-38bd-4568-b8c1-0ea8c6c2bf46","added_by":"auto","created_at":"2026-04-07 10:06:15","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":84420,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-9238898/v1/333fc3edbf8dbc24dedfad8f.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Clinical Phenotypes of Difficult-to-treat and Mild Asthma Defined by Cluster Analysis","fulltext":[{"header":"Background","content":"\u003cp\u003eAsthma is a chronic inflammatory airway disease characterised by variable airflow obstruction, bronchial hyperresponsiveness, and a broad range of clinical presentations. Despite structured guidelines, asthma still remains a clinical burden due to disease heterogeneity with substantial variation in clinical expression and natural history.\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e This heterogeneity spans age of asthma-onset, atopic status, physiological and inflammatory profiles, symptom burden, comorbidities and treatment response.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eTo better capture this complexity, research has increasingly focused on asthma phenotypes, defined by particular observable clinical characteristics, and endotypes underpinned by distinct pathobiological mechanisms.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e Among various approaches, unsupervised cluster analysis has emerged as a useful method for identifying clinically relevant phenotypes without prior assumptions. Landmark cluster studies, have revealed reproducible severe asthma phenotypes, including early-onset allergic asthma, late-onset eosinophilic asthma, and obese non-eosinophilic asthma.\u003csup\u003e\u003cspan additionalcitationids=\"CR9 CR10 CR11 CR12\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003ePrior cluster studies have largely focused on patients with more severe or difficult-to-treat asthma (DA) managed in specialist settings. Moreover, limited work has included patients with mild asthma (MA), and few cluster analyses have systematically assessed comorbid treatable traits such as obesity\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, gastro-oesophageal reflux disease (GORD)\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, sinonasal disease\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, breathing pattern disorder\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, and psychological comorbidities\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e within clusters. Such comorbidities are increasingly recognised as important modifiers of asthma disease burden, adverse patient outcomes and treatment response, often within a complex framework of multimorbidity.\u003csup\u003e\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e In this study, we sought to address these knowledge gaps by defining asthma clusters in DA and MA patient cohorts through extensive clinical characterisation using comparable methodology.\u003c/p\u003e \u003cp\u003eThe Wessex AsThma CoHort of Difficult Asthma (WATCH-DA) is a prospective, UK-based cohort of adults with DA (GINA treatment steps 4 or 5).\u003csup\u003e22\u003c/sup\u003e with extensive characterisation data on asthma-focused clinical and pathophysiological parameters, and comorbidities. Conversely, a MA cohort from the Epigenetics of Severe Asthma study (EOSA-MA) (GINA treatment steps 1 or 2) provides a rare opportunity to evaluate data-driven clinical phenotyping in milder disease. The EOSA-MA cohort includes a subset of patients from the long-standing Isle of Wight Birth Cohort (IOWBC),\u003csup\u003e23\u003c/sup\u003e together enabling detailed clinical and biomarker characterisation.\u003c/p\u003e \u003cp\u003eThe overall aim of this study is to enhance precision medicine in asthma care by identifying distinct DA and MA clinical phenotypes through assessment of their clinical, physiological, inflammatory and comorbidity profiles to inform future approaches to treatable trait-based stratification and management.\u003c/p\u003e \u003cp\u003eThe four main objectives of this study were: 1) To identify DA and MA phenotypes in the respective WATCH-DA and EOSA-MA cohorts using K-means clustering; 2) To characterise the clinical, physiological, inflammatory, and comorbidity profiles across identified DA and MA clusters; 3) To compare the disease severity spectrum of asthma phenotypes across mild and difficult-to-treat asthma populations; 4) To inform future approaches to treatable trait-based stratification and precision medicine in asthma care.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Design and Cohorts\u003c/h2\u003e \u003cp\u003eThis was a cross-sectional, observational study involving two independent UK-based asthma cohorts, the WATCH-DA and EOSA-MA cohorts, using comparable methodology for deeper clinical characterisation.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eWATCH Difficult-to-Treat Asthma Cohort\u003c/h3\u003e\n\u003cp\u003eAt the time of study, the WATCH-DA study recruited 501 adult DA patients from July 2015 to March 2019 from University Hospital Southampton NHS Foundation Trust (UHSFT) and the David Hide Asthma and Allergy Centre (DHAAC), Isle of Wight, UK. All patients were managed according to the BTS/SIGN step \u0026ldquo;high-dose therapies\u0026rdquo; and/or frequent or continuous oral corticosteroids (OCS) use (GINA treatment steps 4 and 5 equivalent).\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e Ethics approval and written informed consent from patients was obtained (REC reference: 14/WM/1226).\u003c/p\u003e \u003cp\u003eBriefly, comprehensive baseline characterisation was performed at enrolment including: demographics; asthma history; medication use; lung function tests; fractional exhaled nitric oxide (FeNO); blood biomarkers (BECs, blood neutrophils (BNC), total immunoglobulin E (IgE)); GINA Type-2 inflammation (T2)-high status - defined by either FeNO\u0026thinsp;\u0026ge;\u0026thinsp;20 ppb, BEC\u0026thinsp;\u0026ge;\u0026thinsp;0.2\u0026times;10⁹/L (threshold chosen due to laboratory reporting to one decimal place), maintenance OCS (mOCS) use, or clinically allergen-driven disease (\u0026ge;\u0026thinsp;1 positive aeroallergen skin-prick test plus a relevant trigger), with sputum eosinophilia excluded due to limited samples); allergy skin prick testing (SPT) to 13 common aeroallergens; comorbidities defined using conventional clinical criteria (\u003cb\u003eSupplementary Table\u0026nbsp;1\u003c/b\u003e); and health and disease-related questionnaires including Asthma Control Questionnaire-6 (ACQ-6), Nijmegen Score, Hospital Anxiety and Depression Scale (HADS), Sino-Nasal Outcome Test-22 (SNOT-22), Hull Cough Hypersensitivity Questionnaire, EuroQoL 5-Dimension 5-Level health today Visual Analogue Scale (EQ5D-5L VAS), and St George\u0026rsquo;s Respiratory Questionnaire (SGRQ). Data were stored in a central WATCH-DA database housed at UHSFT and exported to SPSS v26 (IBM, NY, USA) The full study protocol is described elsewhere.\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Three patients were excluded due to extreme outlier values for total IgE, leaving 498 patients for clustering analysis (\u003cb\u003eSupplementary Fig.\u0026nbsp;1\u003c/b\u003e).\u003c/p\u003e\n\u003ch3\u003eEOSA Mild Asthma Cohort\u003c/h3\u003e\n\u003cp\u003eThe EOSA-MA cohort included 69 participants with physician-diagnosed MA defined by BTS/SIGN treatment steps 1\u0026ndash;2 (GINA steps 1\u0026ndash;2),\u003csup\u003e3,24\u003c/sup\u003e recruited from August 2018 to July 2019 as part of a larger NIH-funded EOSA study\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e (total n\u0026thinsp;=\u0026thinsp;193). A subset (n\u0026thinsp;=\u0026thinsp;23) was drawn from the IOWBC, a whole-population birth cohort initiated in 1989 to study the natural history of asthma and allergy, with detailed longitudinal data from birth to 26-years of age.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e The remaining participants (n\u0026thinsp;=\u0026thinsp;46) were identified from local IOW Allergy Clinics, primary care practices via review of electronic health records and community outreach (including use of the DHAAC website) (\u003cb\u003eSupplementary Fig.\u0026nbsp;1\u003c/b\u003e). Ethics approval for the EOSA study was obtained (REC reference: 18/SC/0105). All participants provided written informed consent.\u003c/p\u003e \u003cp\u003eAll participants underwent identical clinical, physiological, biomarker, comorbidity and questionnaire assessments to the WATCH-DA cohort. Data were stored securely in a DHAAC-based SPSS dataset. Two participants were excluded as outliers (one for extreme heavy smoking, one for an extreme outlier value for total IgE), leaving 67 participants for clustering analysis.\u003c/p\u003e\n\u003ch3\u003eClustering Variables and Analysis\u003c/h3\u003e\n\u003cp\u003eK-means cluster analysis was performed separately for each cohort using PROC FASTCLUS in SAS v9.4 (SAS Institute, Cary, NC, USA). The selection of clinical variables used for cluster analysis is detailed in \u003cb\u003eSupplementary Methods\u003c/b\u003e. Categorical variables were transformed into numeric form prior to clustering. The number of clusters was determined using the Cubic Clustering Criterion (CCC), pseudo-F-statistic, and the R\u0026sup2; statistic, in combination with clinical interpretability. Eleven clustering variables were used for WATCH-DA: age of asthma-onset, BMI, hospitalisations in past 12 months, Intensive Care Unit (ICU) admissions (ever), number of OCS courses in past 12 months, FeNO, post-bronchodilator (BD) Forced Expiratory Volume in 1 second (FEV₁) % predicted, BEC, serum Total IgE, ACQ-6 score, and sum of positive SPTs. Twelve clustering variables were used for EOSA-MA: age of asthma-onset, BMI, ICU admissions (ever), asthma-related GP visits in past 12 months, number of OCS courses in past 12 months, current inhaled corticosteroids (ICS) use, FeNO, pre-BD FEV₁ % predicted, BEC, serum total IgE, ACQ-6 score, and sum of positive SPTs. The differing and additional clinical variables used for the EOSA-MA cohort included GP visits in past 12 months (instead of hospitalisations in past 12 months) and current ICS use respectively to enable capturing the asthma severity spectrum in this population.\u003c/p\u003e\n\u003ch3\u003ePost-Hoc Trait Analysis\u003c/h3\u003e\n\u003cp\u003eFollowing cluster assignment, additional post-hoc analyses were conducted to evaluate differences between clusters in demographics, treatment, inflammatory (only a subset of WATCH-DA patients (n\u0026thinsp;=\u0026thinsp;139) had induced sputum samples available for differential cell count analysis), physiological, and comorbidity profiles plus health outcomes.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eCluster membership was used as the grouping variable for all post-hoc trait analyses. Categorical variables were represented as frequency; n (%), and compared across clusters using Chi-squared test, or Fisher\u0026rsquo;s exact test when expected cell counts\u0026thinsp;\u0026lt;\u0026thinsp;5. Continuous variables were represented as mean (SD) or median (IQR), according to distribution. Between-cluster differences for continuous data were assessed using one-way ANOVA for normally-distributed variables and Kruskal\u0026ndash;Wallis test for skewed variables. Where the omnibus test was significant, exploratory pairwise comparisons between clusters were made using independent-sample t-test or Mann\u0026ndash;Whitney U test, as appropriate, to identify which clusters differed. A p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant. All post-hoc analyses were conducted using SPSS v26 (IBM, NY, USA).\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eDifficult-to-treat asthma clusters\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMain clustering outcomes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCluster analysis using 11 selected variables identified six discrete DA clusters. A six-cluster solution was selected because, across candidate models with differing numbers of clusters, this showed the most favourable combination of cubic clustering criterion and pseudo-F statistics together with a high between-cluster R², whilst also yielding clinically coherent, non-redundant phenotypes (as detailed in Methods). All clustering variables showed significant differences across these clusters (\u003cstrong\u003eTable 1\u003c/strong\u003e). Resulting clustering variable-based characterisation is summarised below and in \u003cstrong\u003eFigure 1a\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-1\u003c/em\u003e (n=171; 34.3%);\u003cem\u003e\u0026nbsp;Early-onset-clinically-controlled-atopic disease:\u0026nbsp;\u003c/em\u003eEarliest (childhood)-onset asthma, raised BMI, low T2 inflammatory biomarkers (BEC/FeNO), best spirometry, better than average asthma control plus low asthma exacerbations and asthma-related admissions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-2\u003c/em\u003e (n=40; 8.0%); \u003cem\u003eAdult-onset-severe-exacerbation-prone disease\u003c/em\u003e: Adult-onset disease, lowest BMI, highest BEC and FeNO, average spirometry and asthma control plus highest exacerbation frequency and admissions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-3\u003c/em\u003e (n=35; 7.0%); \u003cem\u003eAdolescent-onset-obese-atopic-airflow-obstructive disease\u003c/em\u003e: Adolescent-onset disease, obese-level BMI, low FeNO but high BEC and highest atopy, worst spirometry, poorer asthma control plus moderate exacerbation frequency and lower acute admissions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-4\u003c/em\u003e (n=127; 25.5%); \u003cem\u003eAdult-onset-clinically-controlled-least-atopic disease\u003c/em\u003e\u003cem\u003e;\u003c/em\u003e Oldest age of onset, raised BMI, moderate FeNO, low BEC, least atopy, average spirometry, best asthma control, lowest exacerbation frequency, and lower admissions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-5\u003c/em\u003e (n=33; 6.6%); \u003cem\u003eYoung-adult-onset-\u003c/em\u003es\u003cem\u003eevere-exacerbation-prone disease\u003c/em\u003e\u003cem\u003e:\u003c/em\u003e Young adult-onset disease, raised BMI, low BEC and FeNO, moderate atopy, poorer spirometry, worse asthma control, elevated asthma exacerbation frequency and admissions plus highest frequency of asthma-related ICU admissions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-6\u003c/em\u003e (n=92; 18.5%);\u003cem\u003e\u0026nbsp;Adult-onset-obese-severe-exacerbation-prone disease:\u003c/em\u003e Adult-onset disease, obesity-level BMI, lowest BEC and FeNO, moderate atopy, average spirometry, worst asthma control, and high frequency of asthma exacerbations and admissions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePost-hoc phenotypic cluster characteristics\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFemale predominance was seen in four clusters (\u003cem\u003eearly-onset-clinically-controlled-atopic\u0026nbsp;\u003c/em\u003e(cluster-1), a\u003cem\u003edult-onset-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-2)\u003cem\u003e, young-adult-onset-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-5) and\u003cem\u003e\u0026nbsp;adult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) with no sex predominance seen in the remaining 2 clusters (\u003cstrong\u003eTable 2\u003c/strong\u003e). Current age was oldest in\u0026nbsp;a\u003cem\u003edult-onset-clinically-controlled-least-atopic\u003c/em\u003e (cluster-4)\u0026nbsp;and youngest in \u003cem\u003eearly-onset-clinically-controlled-atopic\u0026nbsp;\u003c/em\u003e(cluster-1) matching age of onset associations. Smoking status broadly separated into never-smoking\u0026nbsp;(\u003cem\u003eearly-onset-clinically-controlled-atopic (\u003c/em\u003ecluster-1), \u003cem\u003eadult-onset-severe-exacerbation-prone (\u003c/em\u003ecluster-2) and\u003cem\u003e\u0026nbsp;adolescent-onset-obese-atopic-airflow-obstructive (\u003c/em\u003ecluster-3)), and\u0026nbsp;ex-smoking\u0026nbsp;(\u003cem\u003eadult-onset-clinically-controlled-least-atopic (\u003c/em\u003ecluster-4)). Current smoking status was low but mostly seen in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-6) and\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-5)\u003cem\u003e.\u003c/em\u003e Morbidity and treatment need differed. Days lost from work were highest in \u003cem\u003eyoung-adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-5) but lowest in\u003cem\u003e\u0026nbsp;adolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3). Intubation for acute severe asthma exacerbations was highest in\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-5), and lowest\u0026nbsp;ina\u003cem\u003edult-onset-clinically-controlled-least-atopic\u003c/em\u003e (cluster-4). \u0026nbsp;T2 biologics use at study enrolment were highest in\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-5) and lowest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6). Omalizumab\u0026nbsp;use was highest in\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-5), and lowest\u0026nbsp;in a\u003cem\u003edult-onset-clinically-controlled-least-atopic-disease\u003c/em\u003e (cluster-4). Post-study enrolment T2 biologics initiation was highest in \u003cem\u003eadolescent-onset-obese-atopic-airflow-obstructive\u0026nbsp;\u003c/em\u003e(cluster-3) plus\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-5), whilst lowest in \u003cem\u003eearly-onset-clinically-controlled-atopic\u0026nbsp;\u003c/em\u003e(cluster-1). Antifungal\u0026nbsp;(itraconazole)\u0026nbsp;use was highest in\u003cem\u003e\u0026nbsp;adolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3), and lowest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-6).\u0026nbsp;Mean ICS dose, mOCS, prophylactic antibiotics and mepolizumab use at study enrolment did not differ significantly by cluster (\u003cstrong\u003eTable 2\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBaseline inflammatory profiles\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAlthough GINA-defined T2-high status displayed some cluster variation (\u003cem\u003ep\u003c/em\u003e =0.049), all clusters showed overwhelming T2 predominance ranging from 88.2% in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) to 100% in a\u003cem\u003edult-onset-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-2) and \u003cem\u003eadolescent-onset-obese-atopic-airflow-obstructive\u0026nbsp;\u003c/em\u003e(cluster-3). Aspergillus fumigatus-specific IgE was highest in\u003cem\u003e\u0026nbsp;adolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3) and lowest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u0026nbsp;\u003c/em\u003e(cluster-6). In a subset of WATCH-DA patients (n = 139) who had induced sputum samples and differential counts available,\u0026nbsp;sputum eosinophils and eosinophil proportion ≥2%\u0026nbsp;were both highest in\u003cem\u003e\u0026nbsp;adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-2) and lowest in \u003cem\u003eearly-onset-clinically-controlled-atopic\u003c/em\u003e (cluster-1). Mean sputum neutrophils, neutrophil proportion \u0026gt;61%, and sputum inflammatory phenotypes did not differ significantly across clusters (\u003cstrong\u003eTable 2\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBaseline physiological characteristics\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eClinic (post-BD) spirometry differed across clusters (\u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). \u003cem\u003eEarly-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3) showed the worst spirometry (lowest FEV₁\u0026nbsp;%pred, Forced Vital Capacity (FVC) %pred, FEV₁/FVC ratio and Forced Expiratory Flow between 25% and 75% of vital capacity (FEF₂₅\u003csub\u003e%\u003c/sub\u003e\u003csub\u003e–\u003c/sub\u003e₇₅\u003csub\u003e%\u003c/sub\u003e)), whereas \u003cem\u003eadult-onset-obese-severe-exacerbation-prone (\u003c/em\u003ecluster-6)\u0026nbsp;had the least airflow obstruction (highest FEV₁/FVC and FEF₂₅\u003csub\u003e%\u003c/sub\u003e\u003csub\u003e–\u003c/sub\u003e₇₅\u003csub\u003e%\u003c/sub\u003e). RV/TLC ratio also differed significantly; worst in \u003cem\u003eadolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3), and best in \u003cem\u003eearly-onset-clinically-controlled-atopic\u0026nbsp;\u003c/em\u003e(cluster-1) (\u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e). Gas transfer indices did\u0026nbsp;not\u0026nbsp;differ across clusters.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eComorbidity characteristics\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eComorbidities significantly differed by cluster (\u003cstrong\u003eTable 3\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Figure 1b\u003c/strong\u003e). Obesity (BMI≥ 30) was highest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) and lowest in \u003cem\u003eadult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-2). Breathing pattern disorder was highest in\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-5) and lowest in \u003cem\u003eadult-onset-clinically-controlled-least-atopic\u003c/em\u003e (cluster-4). Depression/anxiety were highest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) but lowest in\u003cem\u003e\u0026nbsp;young-adult-onset-severe-exacerbation-prone (cluster-5)\u0026nbsp;\u003c/em\u003eand\u003cem\u003e\u0026nbsp;adult-onset-clinically-controlled-least-atopic\u003c/em\u003e (cluster-4). Eczema was highest in \u003cem\u003eadolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3) and lowest in \u003cem\u003eadult-onset-clinically-controlled-least-atopic\u003c/em\u003e (cluster-4). Nasal polyps were predominant in \u003cem\u003eadult-onset-clinically-controlled-least-atopic\u0026nbsp;\u003c/em\u003e(cluster-4) and lowest in \u003cem\u003eearly-onset-clinically-controlled-atopic\u003c/em\u003e (cluster-1). Salicylate sensitivity was highest in \u003cem\u003eyoung-adult-onset-severe-exacerbation-prone (\u003c/em\u003ecluster-5)\u0026nbsp;and lowest in\u003cem\u003e\u0026nbsp;adolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3). Allergic bronchopulmonary aspergillosis (ABPA)/Severe Asthma with Fungal Sensitisation (SAFS) were highest in \u003cem\u003eadolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3) and lowest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6). Chronic Obstructive Pulmonary Disease (COPD) was predominant in \u003cem\u003eadult-onset-clinically-controlled-least-atopic\u0026nbsp;\u003c/em\u003e(cluster-4) and lowest in \u003cem\u003eearly-onset-clinically-controlled-atopic (\u003c/em\u003ecluster-1).\u0026nbsp;Rhinitis\u0026nbsp;and\u0026nbsp;GORD\u0026nbsp;were ubiquitous with no difference between clusters (\u003cstrong\u003eTable 3\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eComorbidity-related questionnaires differed significantly by cluster (\u003cstrong\u003eTable 3\u003c/strong\u003e). SNOT-22 was highest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) and lowest in \u003cem\u003eadult-onset-clinically-controlled-least-atopic (\u003c/em\u003ecluster-4). Hull cough hypersensitivity score was also highest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) but lowest in \u003cem\u003eearly-onset-clinically-controlled-atopic\u003c/em\u003e (cluster-1). Nijmegen questionnaire scores (and proportion with score \u0026gt;23) were highest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6), and lowest in \u003cem\u003eadult-onset-severe-exacerbation-prone\u003c/em\u003e (cluster-2) and \u003cem\u003eadult-onset-clinically-controlled-least-atopic\u0026nbsp;\u003c/em\u003e(cluster-4). HADS-D (depression component) was also highest in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) with lowest proportions of HADS-D ≥11 in \u003cem\u003eadolescent-onset-obese-atopic-airflow-obstructive\u003c/em\u003e (cluster-3) and \u003cem\u003eadult-onset-clinically-controlled-least-atopic\u0026nbsp;\u003c/em\u003e(cluster-4). Quality of life (QoL) mirrored symptom burden, with the worst EQ-5D-5L health today VAS and SGRQ total score in \u003cem\u003eadult-onset-obese-severe-exacerbation-prone\u003c/em\u003e (cluster-6) (\u003cstrong\u003eTable 3\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMild asthma clusters\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMain clustering outcomes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCluster analysis with 12 selected variables identified two discrete MA clusters. A two-cluster solution was selected favouring clinically coherent, non-redundant MA phenotypes. Clustering variables with significant differences between clusters included BEC, Total IgE, FeNO, FEV₁% pred., atopic status, and ACQ-6 (\u003cstrong\u003eTable 4\u003c/strong\u003e). Resulting cluster-based characterisation is summarised below and in \u003cstrong\u003eFigure 1a\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-1\u0026nbsp;\u003c/em\u003e(n=20; 29%);\u003cem\u003e\u0026nbsp;Early-onset-atopic-mild-asthma:\u0026nbsp;\u003c/em\u003eEarly (childhood)-onset asthma with significantly raised FeNO, BEC and Total IgE, greater atopy, lower FEV\u003csub\u003e1\u003c/sub\u003e plus higher ACQ-6.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-2\u003c/em\u003e (n=47; 70.1%); \u003cem\u003eAdolescent-onset-mild-asthma\u003c/em\u003e. Early (adolescent)-onset asthma with lower expression of inflammatory biomarkers and atopy plus better FEV\u003csub\u003e1\u003c/sub\u003e and ACQ-6.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePost-hoc phenotypic characteristics\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMale predominance was observed for \u003cem\u003eearly-onset-atopic-mild-asthma\u0026nbsp;\u003c/em\u003e(cluster-1) whilst\u003cem\u003e\u0026nbsp;adolescent-onset-mild-asthma\u0026nbsp;\u003c/em\u003e(cluster-2) showed female predominance\u003cem\u003e.\u0026nbsp;\u003c/em\u003eOther clinical parameters including current age, smoking status, morbidity, and medication use did not differ significantly between clusters (\u003cstrong\u003eTable 4\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBaseline inflammatory profiles\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEarly-onset-atopic-mild-asthma\u0026nbsp;\u003c/em\u003e(cluster-1) showed higher Aspergillus fumigatus-specific IgE, BNC and sputum inflammatory markers (% eosinophils, % neutrophils; inflammatory phenotype), but did not significantly differ\u0026nbsp;due to\u0026nbsp;limited induced sputum availability. GINA T2-high status was prevalent across both clusters but highest in \u003cem\u003eearly-onset-atopic-mild-asthma\u0026nbsp;\u003c/em\u003e(cluster-1) (\u003cstrong\u003eTable 4\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBaseline physiological characteristics\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLung function tests (spirometry, lung volumes and gas transfer factors) were within normal range for both clusters but modestly lower in \u003cem\u003eearly-onset-atopic-mild-asthma\u003c/em\u003e (cluster-1); i.e. pre-BD FEV₁\u0026nbsp;%pred., FEV₁/FVC ratio,\u0026nbsp;and FEF₂₅\u003csub\u003e%\u003c/sub\u003e\u003csub\u003e-\u003c/sub\u003e₇₅\u003csub\u003e%\u003c/sub\u003e (\u003cstrong\u003eSupplementary Table 3\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eComorbidity characteristics\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eComorbidities were highly prevalent among both MA clusters (\u003cstrong\u003eTable 4\u003c/strong\u003e). Obesity\u0026nbsp;(BMI≥ 30) was more prevalent in a\u003cem\u003edolescent-onset-mild-asthma\u003c/em\u003e (cluster-2). Other comorbidities and comorbidity-related questionnaires outcomes were similar between clusters (\u003cstrong\u003eTable 4\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSummary characterisation of difficult-to-treat and mild asthma cluster phenotypes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA summary of the six DA and two MA cluster phenotypes respectively, incorporating clustering and morbidity characteristics is shown in\u0026nbsp;\u003cstrong\u003eFigure 1a\u003c/strong\u003e.\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eUsing a common assessment platform across two cohorts spanning the asthma severity spectrum, we identified six difficult-to-treat clusters and two mild-asthma clusters with distinct clinical, inflammatory and comorbidity profiles. A key novel focus of this study was the exploration of comorbidities and highlighting their role in asthma heterogeneity. One important finding is that despite using an identical phenotyping platform for comprehensive clinical characterisation, heterogeneity is already apparent in milder asthma and does not map directly onto patterns seen in more difficult-to-treat disease. Separate K-means clustering analyses for the WATCH-DA and EOSA-MA cohorts using identical clinical and biomarker assessments revealed distinct phenotypic profiles across both groups. Additionally, we move beyond description by linking readily available bedside clinical cues - sex, age of onset, BMI, simple T2 biomarkers and spirometry phenotypes - through a Clinician Recognition Guide (\u003cstrong\u003eTable 5\u003c/strong\u003e). Although T2-high biology was common, it did not alone explain variation in symptoms, physiology, or multimorbidity. Among difficult-to-treat clusters, disease severity, sex, age of onset, and BMI varied alongside airflow limitation and healthcare use. The mild asthma clusters, despite preserved spirometry, were clinically and biologically distinct with significant comorbidities suggesting multimorbid burden.\u003c/p\u003e\n\u003cp\u003eComparisons across asthma severities indicate that burden is shaped by combinations of pulmonary and extra-pulmonary traits rather than a single linear progression from mild to difficult-to-treat disease. This perspective clarifies how the phenotypes relate across the disease spectrum and provides a practical basis for trait-targeted, guideline-aligned care.\u003c/p\u003e\n\u003cp\u003eOur results reflect and extend existing literature showing reproducible, though not identical, asthma clinical phenotypes, including early-onset atopic T2-high, adult-onset eosinophilic, and obese difficult asthma,.\u003csup\u003e8,9,11\u003c/sup\u003e While the cluster characteristics inevitably depend on the clinical and pathobiological parameters collected, these studies consistently demonstrate that clinically recognisable clusters can share T2 features yet diverge in physiology and symptom burden, emphasising the need for multi-domain appraisal rather than reliance on T2-biomarkers alone. Large population and longitudinal cohorts further underscore the breadth of multimorbidity across asthma severities, with adult-onset disease and obesity, particularly among women, associated with greater morbidity over time.\u003csup\u003e26,27\u003c/sup\u003e While our cross-sectional design precludes inference about directionality, the prominence of comorbidities in both the difficult-to-treat and mild asthma cohorts is concordant with these observations and merits routine clinical attention.\u003c/p\u003e\n\u003cp\u003eObesity, which has long reached epidemic proportions, warrants particular discussion. In our difficult-to-treat clusters, higher BMI co-occurred with greater symptom burden and lower T2 signals in subsets, while in mild asthma a lower-T2, obesity-linked group exhibited good spirometry but notable patient-reported burden. Observational and intervention studies suggest that weight loss can be associated with improvements in asthma control and quality of life in individuals with obesity; however, responses vary and should be considered alongside optimisation of inhaled therapy and comorbidity management.\u003csup\u003e26-29\u003c/sup\u003e We therefore interpret obesity as a clinically relevant trait that frequently co-occurs with asthma across the asthma severity spectrum.\u003c/p\u003e\n\u003cp\u003eSex and age of onset patterns were also informative. Prior work has described female predominance in obesity-linked phenotypes and higher risk associated with adult-onset disease; our clusters showed similar associations.\u003csup\u003e8,9,11,27\u003c/sup\u003e These readily available cues: sex, age of onset, and BMI, remain practical anchors for recognising phenotypes, particularly when considered alongside simple T2-biomarkers, basic spirometry and a focused review of extra-pulmonary traits.\u003c/p\u003e\n\u003cp\u003eComparison across both cohorts showed that one mild asthma cluster exhibited greater morbidity but did not map one-to-one onto any single difficult-to-treat asthma cluster. Instead, overlapping mixtures of traits were seen across severities, supporting the view that heterogeneity is also present in upstream milder disease and amplifies with accrued comorbidity regardless of asthma severity. This observation argues for early identification and management of co-existing traits between mild and difficult-to-treat clusters even when spirometry is near-normal.\u003c/p\u003e\n\u003cp\u003eOur findings could provide a framework to support a\u0026nbsp;treatable-traits\u0026nbsp;approach in which pulmonary, extra-pulmonary and behavioural traits are identified systematically and addressed in parallel with guideline-directed care. Randomised and translational studies indicate that trait-targeted pathways can improve outcomes relative to usual clinical care, and that distinct trait profiles differ in burden and responsiveness to systematic assessment.\u003csup\u003e30-32\u003c/sup\u003e Against this backdrop, we frame the clinical management of our clusters according to their predominant component treatable traits.\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cem\u003eActive T2-high traits:\u003c/em\u003e where BEC and/or FeNO remain raised and morbidity persists despite optimised inhaled therapy, consideration of add-on T2-biologics within steroid-sparing pathways and using guideline eligibility criteria, is consistent with current strategy documents.\u003csup\u003e33,34\u003c/sup\u003e\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eObesity-associated, symptom-dominant presentations with lower T2 signals:\u003c/em\u003e first-line emphasis on structured weight reduction management and physical activity is supported by trials showing clinically meaningful improvements in QoL and, in some studies, airway inflammation with ~5–10% weight loss alongside continued inhaled therapy optimisation in parallel.\u003csup\u003e28,29\u003c/sup\u003e\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eUpper-airway comorbidity e.g. chronic rhinosinusitis with nasal polyps [CRSwNP]\u003c/em\u003e: where sinonasal disease co-exists, ENT-led care and, for eligible patients, biologics for CRSwNP may reduce overall burden if locally available. Selection should follow the\u0026nbsp;European Position Paper on Rhinosinusitis and Nasal Polyps/ European Forum for Research and Education in Allergy and Airway Diseases (EPOS/EUFOREA) criteria.\u003csup\u003e35\u003c/sup\u003e\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eBreathing pattern disorder:\u003c/em\u003e physiotherapy-based breathing retraining can improve asthma-related QoL as an adjunct to standard care; screening and referral are vital where this trait is suspected.\u003csup\u003e36\u003c/sup\u003e\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eGORD and cough hypersensitivity:\u003c/em\u003e management should follow guideline-based indications.\u003csup\u003e37,38\u003c/sup\u003e Where T2-biomarkers are low and extra-pulmonary drivers predominate, escalation of corticosteroids is unlikely to help and should be avoided.\u003csup\u003e1,3\u003c/sup\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThe\u0026nbsp;Clinician Recognition Guide\u0026nbsp;(\u003cstrong\u003eTable 5\u003c/strong\u003e) operationalises these principles at the bedside for individual patient clusters. Sex, age of onset, and BMI, together with simple T2-biomarkers, spirometry and a brief scan for sentinel comorbidities; including GORD, obesity, CRSwNP, breathing pattern disorder, inducible laryngeal obstruction, and smoking/asthma-COPD overlap, typically suffice to place a patient within a\u0026nbsp;phenotype and to select the first-line action. The table presents associative cues and\u0026nbsp;management considerations, not treatment mandates; add-on therapies should be chosen using clinical and biomarker criteria, shared decision-making and local availability, alongside non-pharmacological interventions for comorbid traits.\u003c/p\u003e\n\u003cp\u003eOur study has both strengths and limitations. Strengths include harmonised clustering across two well-phenotyped cohorts which have been identically characterised. Use of clinically relevant clustering variables adds further robustness. Also, deep post-hoc characterisation spanning inflammatory, physiology, comorbidities and patient-reported outcomes/HRQoL assessments of both mild and difficult-to-treat asthma patients. One potential limitation is the cross-sectional design, and future studies would benefit from assessing longitudinal cluster stability and changing outcomes based on our recommendations of care provision in \u003cstrong\u003eTable 5\u003c/strong\u003e. Limited induced sputum and airway sampling, particularly in the MA cohort, may be viewed as another shortcoming. Nevertheless, an adequate proportion of airway sampling was achieved in the WATCH-DA cohort permitting valid inferences. While the study might be viewed as single centre, the WATCH-DA cohort is drawn from a tertiary referral site that effectively serves as a multi-centre checkpoint. These considerations do not alter the central practice message that\u0026nbsp;comorbidity-linked treatable traits\u0026nbsp;drive multimorbid burden across phenotypes.\u003c/p\u003e\n\u003cp\u003eFuture steps include using these clusters to\u0026nbsp;validate a minimal clinical recognition algorithm\u0026nbsp;(age of onset, BMI, atopy, FeNO/BEC plus 2–3 comorbidity traits) against patient outcomes and treatment response; to assess\u0026nbsp;phenotype-tailored therapeutic interventions; to prospectively evaluate\u0026nbsp;OCS-sparing\u0026nbsp;strategies in high-burden groups; and integrate\u0026nbsp;omics/imaging\u0026nbsp;where comorbidity signals are strongest (obesity, CRSwNP, AERD, ABPA/SAFS), incorporating\u0026nbsp;sex-specific\u0026nbsp;analyses.\u003c/p\u003e\n\u003cp\u003eIn summary, across difficult-to-treat and mild asthma cohorts assessed with a comparable clinical characterisation platform, clustering reveals clinically coherent associations between biomarker patterns, physiological measures and multimorbidity on a largely T2-predominant background. Used cautiously in alignment with guidelines and trait-targeted evidence, these patterns can help clinicians recognise asthma phenotypes rapidly and instigate first-line interventions earlier for active treatable traits.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eABPA, Allergic bronchopulmonary aspergillosis; ACQ6, Asthma Control Questionnaire-6; BD, bronchodilator; BEC, blood eosinophil count; BMI, Body Mass Index; BN, blood neutrophils count; BTS, British Thoracic Society; CCC, Cubic Clustering Criterion; COPD, Chronic Obstructive Pulmonary Disease; CRSwNP, chronic rhinosinusitis with nasal polyps; DA, Difficult Asthma; DHAAC, David Hide Asthma and Allergy Centre; EOSA, Epigenetics of Severe Asthma; EQ-5D-5L VAS, EuroQol 5-Dimension 5-Level health today Visual Analogue Scale; FEF\u0026nbsp;\u003csub\u003e25%\u003c/sub\u003e\u003csub\u003e-\u003c/sub\u003e\u003csub\u003e75%\u003c/sub\u003e, forced expiratory flow between 25% and 75% of vital capacity; FeNO, fractional exhaled nitric oxide; FEV₁, Forced Expiratory Volume in 1 second; FVC, Forced Vital Capacity; GORD, gastro-oesophageal reflux disease; GINA, Global Initiative for Asthma; GP, general practice; HADS, Hospital Anxiety and Depression Scale; ICU, Intensive Care Unit; IgE, immunoglobulin E; IOWBC, Isle of Wight Birth Cohort; MA, mild asthma; NICE, National Institute for Health and Care Excellence; OCS, oral corticosteroids; QoL, Quality of life; SAFS, severe asthma with fungal sensitisation; SGRQ, St George\u0026rsquo;s Respiratory Questionnaire; SNOT-22, Sino-Nasal Outcome Test; SPT, skin prick testing; T2, Type-2 inflammation; WATCH, Wessex AsThma CoHort of Difficult Asthma.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eEthical approval for the Wessex AsThma CoHort of difficult asthma (WATCH-DA) was granted by the West Midlands \u0026ndash; Solihull Research Ethics Committee (REC reference: 14/WM/1226). The Epigenetics of Severe Asthma (EOSA) study was approved by the South-Central Hampshire B \u0026ndash; Southampton Research Ethics Committee (REC reference: 18/SC/0105).\u003c/p\u003e\n\u003cp\u003eAll participants provided written informed consent prior to inclusion in the studies.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are not publicly available due to participant confidentiality and governance restrictions associated with the WATCH-DA and EOSA cohorts but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eConflicts of Interest\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe authors, ML, HM, HZ, ML, JV, BA, LW, CE, JJH, AL, AF, HMH, PD, SH, GS, PV, MAK, EV, ED, SHA, RJK, declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Professor Ratko Djukanovic reports personal fees, has shares in the company and is a consultant to Synairgen, and personal fees from GlaxoSmithKline and Kymab, outside the submitted work.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe Wessex AsThma CoHort of difficult asthma (WATCH) study has been supported by the NIHR Southampton BRC and Clinical Research Facility UHSFT, UK. The WATCH study itself is not externally funded. Funding assistance for database support for the WATCH study was initially obtained from a non-promotional grant from Novartis (\u0026pound;35,000). Funding assistance for patient costs (e.g. parking) was initially provided by a charitable grant (\u0026pound;3500) from the Asthma, Allergy \u0026amp; Inflammation Research (AAIR) Charity. Lastly, funding assistance for the Epigenetics of Severe Asthma (EOSA) study was obtained from a National Institutes of Health (NIH) grant (\u0026pound;400,000) in collaboration with La Jolla Institute of Immunology, La Jolla, California, USA.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eHM, HZ, SHA, RJK were responsible for the conceptualisation of the study and designed the methodology. Data collection was performed by HM, ML, AF, HMH, PD, MAK, RJK. Data curation and formal analysis were performed by HM, HZ, SHA, RJK. ML, HM and RJK drafted the manuscript and all authors reviewed, edited, and approved the final version of the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe authors wish to thank the patients who participated in this study. They also wish to acknowledge the contributions of the wider WATCH study team at University Hospital Southampton NHS Foundation Trust (UHSFT) and University of Southampton, and the research teams at the David Hide Asthma and Allergy Centre, Isle of Wight and La Jolla Institute of Immunology, USA. The authors wish to acknowledge the support of the Southampton NIHR Biomedical Research Centre (BRC) and Clinical Research Facility. The Clinical Research Facility and BRC are funded by Southampton NIHR and are a partnership between the University of Southampton and UHSFT. The authors wish to thank all those who made this study possible. The authors also acknowledge funding support from Novratis, NIH and the AAIR charity.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGlobal Initiative for A. Global Strategy for Asthma Management and Prevention. GINA; 2022.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoulet L-P. Difficult-to-Treat and Severe Asthma in adolescents and adults: diagnosis and management (GINA Pocket Guide). Global Initiative for Asthma; 2019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBritish Thoracic S, Scottish Intercollegiate Guidelines N. British guideline on the management of asthma (SIGN 153): SIGN, 2016.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWenzel SE. Asthma phenotypes: the evolution from clinical to molecular approaches. Nat Med. 2012;18(5):716\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChung KF. Precision medicine in asthma: linking phenotypes to targeted treatments. Curr Opin Pulm Med. 2018;24(1):4\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026ouml;tvall J, Akdis CA, Bacharier LB. Asthma endotypes: a new approach to classification of disease entities within the asthma syndrome. J Allergy Clin Immunol. 2011;127(2):355\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuruvilla ME, Lee FE, Lee GB. Understanding Asthma Phenotypes, Endotypes, and Mechanisms of Disease. Clin Rev Allergy Immunol. 2019;56(2):219\u0026ndash;33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaldar P, Pavord ID, Shaw DE. Cluster analysis and clinical asthma phenotypes. Am J Respir Crit Care Med. 2008;178(3):218\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLefaudeux D, De Meulder B, Loza MJ. U-BIOPRED clinical adult asthma clusters linked to a subset of sputum omics. J Allergy Clin Immunol. 2017;139(6):1797\u0026ndash;807.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewby C, Heaney LG, Menzies-Gow A. Statistical cluster analysis of the British Thoracic Society Severe Refractory Asthma Registry: clinical outcomes and phenotype stability. PLoS ONE. 2014;9(7):e102987.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore WC, Meyers DA, Wenzel SE. Identification of asthma phenotypes using cluster analysis in the Severe Asthma Research Program. Am J Respir Crit Care Med. 2010;181(4):315\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFitzpatrick AM, Teague WG, Meyers DA, et al. Heterogeneity of severe asthma in childhood: confirmation by cluster analysis of children in the National Institutes of Health/National Heart, Lung, and Blood Institute Severe Asthma Research Program. J Allergy Clin Immunol. 2011;127(2):382\u0026ndash;e91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFitzpatrick AM, Moore WC. Severe Asthma Phenotypes - How Should They Guide Evaluation and Treatment? J Allergy Clin Immunol Pract. 2017;5(4):901\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBal C, Pohl W, Milger K, et al. Characterization of Obesity in Severe Asthma in the German Asthma Net. J Allergy Clin Immunol Pract. 2023;11(11):3417\u0026ndash;e243.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTariq K, Schofield JPR, Nicholas BL, et al. Sputum proteomic signature of gastro-oesophageal reflux in patients with severe asthma. Respir Med. 2019;150:66\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangdon C, Mullol J. Nasal polyps in patients with asthma: prevalence, impact, and management challenges. J Asthma Allergy. 2016;9:45\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFreeman A, Abraham S, Kadalayil L, et al. Associations of Breathing Pattern Disorder and Nijmegen Score With Clinical Outcomes in Difficult-to-Treat Asthma. J Allergy Clin Immunol Pract. 2024;12(4):938\u0026ndash;e476.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFong WCG, Rafiq I, Harvey M et al. The Detrimental Clinical Associations of Anxiety and Depression with Difficult Asthma Outcomes. J Pers Med 2022; 12(5).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScelo G, Torres-Duque CA, Maspero J, et al. Analysis of comorbidities and multimorbidity in adult patients in the International Severe Asthma Registry. Ann Allergy Asthma Immunol. 2024;132(1):42\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShackleford A, Heaney LG, Redmond C, McDowell PJ, Busby J. Clinical remission attainment, definitions, and correlates among patients with severe asthma treated with biologics: a systematic review and meta-analysis. Lancet Respir Med. 2025;13(1):23\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKurukulaaratchy RJ, Freeman A, Bansal AT, et al. Evaluation of the effect of multimorbidity on difficult-to-treat asthma using a novel score (MiDAS): a multinational study of asthma cohorts. Lancet Respir Med. 2025;13(9):821\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAzim A, Mistry H, Freeman A. Protocol for the Wessex AsThma CoHort of difficult asthma (WATCH): a pragmatic real-life longitudinal study of difficult asthma in the clinic. BMC Pulm Med. 2019;19(1):99.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArshad SH, Patil V, Mitchell F. Cohort Profile Update: The Isle of Wight Whole Population Birth Cohort (IOWBC). Int J Epidemiol. 2020;49(4):1083\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlobal Initiative for A. Global Strategy for Asthma Management and Prevention. Global Initiative for Asthma (GINA); 2016.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerrera-De La Mata S, Ram\u0026iacute;rez-Su\u0026aacute;stegui C, Mistry H, et al. Cytotoxic CD4(+) tissue-resident memory T cells are associated with asthma severity. Med. 2023;4(12):875\u0026ndash;e978.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKankaanranta H, Viinanen A, Ilmarinen P, et al. Comorbidity Burden in Severe and Nonsevere Asthma: A Nationwide Observational Study (FINASTHMA). J Allergy Clin Immunol Pract. 2024;12(1):135\u0026ndash;e459.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBackman H, Stridsman C, Hedman L, et al. Determinants of Severe Asthma - A Long-Term Cohort Study in Northern Sweden. J Asthma Allergy. 2022;15:1429\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScott HA, Gibson PG, Garg ML, et al. Dietary restriction and exercise improve airway inflammation and clinical outcomes in overweight and obese asthma: a randomized trial. Clin Exp Allergy. 2013;43(1):36\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFreitas PD, Ferreira PG, Silva AG, et al. The Role of Exercise in a Weight-Loss Program on Clinical Control in Obese Adults with Asthma. A Randomized Controlled Trial. Am J Respir Crit Care Med. 2017;195(1):32\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgusti A, Bel E, Thomas M, et al. Treatable traits: toward precision medicine of chronic airway diseases. Eur Respir J. 2016;47(2):410\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcDonald VM, Clark VL, Cordova-Rivera L, Wark PAB, Baines KJ, Gibson PG. Targeting treatable traits in severe asthma: a randomised controlled trial. Eur Respir J 2020; 55(3).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin T, Pham J, Denton E, et al. Trait profiles in difficult-to-treat asthma: Clinical impact and response to systematic assessment. Allergy. 2023;78(9):2418\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBritish Thoracic S, National Institute for H, Care E, Scottish Intercollegiate Guidelines N. Asthma: diagnosis, monitoring and chronic asthma management (BTS, NICE, SIGN). London: National Institute for Health and Care Excellence (NICE); 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlobal Initiative for A. Global Strategy for Asthma Management and Prevention. (2025 update): Global Initiative for Asthma (GINA), 2025.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFokkens WJ, Viskens AS, Backer V, et al. EPOS/EUFOREA update on indication and evaluation of Biologics in Chronic Rhinosinusitis with Nasal Polyps 2023. Rhinology. 2023;61(3):194\u0026ndash;202.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBruton A, Lee A, Yardley L, et al. Physiotherapy breathing retraining for asthma: a randomised controlled trial. Lancet Respir Med. 2018;6(1):19\u0026ndash;28.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParker SM. British Thoracic Society clinical statement on chronic cough in adults. Thorax. 2023;78(Suppl 6):s3\u0026ndash;19.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNational Institute for H, Care E. Gastro-oesophageal reflux disease and dyspepsia in adults: investigation and management. London: National Institute for Health and Care Excellence (NICE); 2014.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 5 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"respiratory-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"rere","sideBox":"Learn more about [Respiratory Research](http://respiratory-research.biomedcentral.com/)","snPcode":"12931","submissionUrl":"https://submission.nature.com/new-submission/12931/3","title":"Respiratory Research","twitterHandle":"@RespiratoryBMC","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Asthma, Cluster Analysis, Comorbidities, Difficult-to-treat Asthma, Mild Asthma, Clinical Phenotypes, Precision Medicine, T2 Inflammation.","lastPublishedDoi":"10.21203/rs.3.rs-9238898/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9238898/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eCluster modelling has demonstrated the heterogeneity of asthma but has previously focused mainly on severe disease with limited assessment of mild disease or treatable traits like comorbidities.\u003c/p\u003e\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eTo identify and characterise difficult-to-treat and mild asthma clusters in two UK cohorts: Wessex AsThma CoHort of Difficult Asthma (WATCH-DA) and a mild-asthma cohort from the Epigenetics of Severe Asthma study (EOSA-MA).\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eSeparate K-means clustering was applied to WATCH-DA (n\u0026thinsp;=\u0026thinsp;498; 11 variables) and EOSA-MA (n\u0026thinsp;=\u0026thinsp;67; 12 variables). Post-hoc comparisons evaluated demographic, inflammatory, physiological, comorbidity and patient-reported outcome profiles.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eSix difficult-to-treat and two mild asthma clusters were identified respectively, all Type-2 (T2)-predominant. Difficult-to-treat asthma clusters differed by sex, age of asthma-onset, body mass index (BMI) and comorbidities. Two clinically-controlled clusters, cluster-1 (early-onset\u0026ndash;clinically-controlled\u0026ndash;atopic disease) and cluster-4 (adult-onset\u0026ndash;clinically-controlled\u0026ndash;least-atopic disease), showed distinct comorbidity patterns despite lower overall morbidity. Three severe, exacerbation-prone, adult-onset, female predominant difficult-to-treat clusters (cluster-2, cluster-5, cluster-6) varied by blood eosinophil counts (BEC), spirometry, BMI, treatment needs, comorbidities, and quality of life. An adolescent-onset\u0026ndash;obese\u0026ndash;atopic\u0026ndash;airflow-obstructive disease (cluster-3) showed fewer exacerbations but high BEC with worst spirometry and poor asthma control. In mild asthma, cluster-1 (early-onset-atopic-mild-asthma) showed worse pathophysiological indices and asthma control than cluster-2 (adolescent-onset-mild-asthma) but similarly high comorbidity prevalence.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eCharacterisation of difficult-to-treat and mild asthma clusters reveals diverse associated clinical traits and outcomes across the asthma severity spectrum. Recognition of these clusters and their associated comorbidities should prompt early personalised asthma management to address both airway-centric and comorbid disease aspects.\u003c/p\u003e","manuscriptTitle":"Clinical Phenotypes of Difficult-to-treat and Mild Asthma Defined by Cluster Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-07 10:06:11","doi":"10.21203/rs.3.rs-9238898/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-25T15:30:13+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-03T01:18:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"211273843737191700165940759569778473403","date":"2026-04-01T14:54:03+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-01T06:45:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-30T17:11:20+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-30T09:06:54+00:00","index":"","fulltext":""},{"type":"submitted","content":"Respiratory Research","date":"2026-03-27T01:31:26+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"respiratory-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"rere","sideBox":"Learn more about [Respiratory Research](http://respiratory-research.biomedcentral.com/)","snPcode":"12931","submissionUrl":"https://submission.nature.com/new-submission/12931/3","title":"Respiratory Research","twitterHandle":"@RespiratoryBMC","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9fd38478-51a1-4d27-9cf8-519764485312","owner":[],"postedDate":"April 7th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T05:39:32+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-07 10:06:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9238898","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9238898","identity":"rs-9238898","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}