Age-related differences in associations between uncontrolled asthma, comorbidities and biomarkers in adult-onset asthma

Abstract Objective Adult-onset asthma is a recognized but heterogeneous phenotype and has been described to associate with poor asthma control. Knowledge about associations between clinical characteristics including comorbidities and control of adult-onset asthma is limited, especially in older populations. We aimed to study how clinical biomarkers and comorbidities are associated with uncontrolled asthma among middle-aged and older individuals with adult-onset asthma. Methods Clinical examinations including structured interview, asthma control test (ACT), spirometry, skin prick test (SPT), blood sampling, and measurement of exhaled fractional nitric oxide (FeNO) was performed in a population-based adult-onset asthma cohort in 2019–2020 (n = 227, 66.5% female). Analyses were performed among all included, and separately in middle-aged (37–64 years, n = 120) and older (≥65 years, n = 107) participants. Results In bivariate analysis, uncontrolled asthma (ACT ≤ 19) was significantly associated with a blood neutrophil count ≥5/µl, BMI ≥30, and several comorbidities. In multivariable regression analysis, uncontrolled asthma was associated with neutrophils ≥5/µl (OR 2.35; 95% CI 1.11–4.99). In age-stratified analysis, BMI ≥30 (OR 3.04; 1.24–7.50), eosinophils ≥0.3/µl (OR 3.17; 1.20–8.37), neutrophils ≥5/µl (OR 4.39; 1.53–12.62) and allergic rhinitis (OR 5.10; 1.59–16.30) were associated with uncontrolled asthma among the middle-aged. Among the older adults, uncontrolled asthma was only associated with comorbidities: chronic rhinitis (OR 4.08; 1.62–10.31), ischemic heart disease (OR 3.59; 1.17–10.98), malignancy (OR 3.10; 1.10–8.73), and depression/anxiety (OR 16.31; 1.82–146.05). Conclusions In adult-onset asthma, comorbidities were strongly associated with uncontrolled asthma among older adults, while clinical biomarkers including eosinophils and neutrophils in blood were associated with uncontrolled asthma among middle-aged.


Introduction
The clinical characteristics of asthma in childhood and young adulthood are to date well described, while only a few studies address this matter among those with adult-onset asthma (1)(2)(3).Asthma among children and adolescents is strongly associated with sensitization to airborne allergens (4), and this association persists often into adulthood (5).The incidence of allergic sensitization is however low after middle-age (6), and other factors like obesity and smoking gain importance as risk factors for asthma with onset in adult age (7,8).Adult-onset asthma is associated with worse asthma control and poorer quality of life (7,9) and may be difficult to distinguish from other airway or cardiac conditions appearing in older age (10).In addition, the increasing appearance of comorbidities with increasing age might affect correct diagnosis, treatment, and control of symptoms in older patients with asthma (11).
The definition of "uncontrolled asthma" differs between international guidelines, and is mostly based on poor symptom control and frequent exacerbations (12,13).Furthermore, the distinction between "uncontrolled asthma, " difficult-to-treat, " and "severe asthma" may be challenging.Management in accordance to the GINA report by healthcare-professionals including adherence to treatment, correct inhalation technique, patient education, and knowledge of factors triggering asthma exacerbations have all been described to affect asthma control (14,15).The asthma control test (ACT) has become an important tool in assessing patients' asthma control (16), and clinical biomarkers including markers for Type-2-inflammation, such as measurement of fractional exhaled nitric oxide (FeNO) and blood eosinophils are used to guide and monitor treatment of asthma (13,17,18).However, these biomarkers as well as the impact of comorbidities on asthma control have rarely been studied specifically among individuals with adult-onset asthma.Especially in older patients, where several comorbidities might present with symptoms similar to asthma, epidemiological studies focusing on these associations are rare.
Thus, our aim was to study how clinical biomarkers and comorbidities associate with asthma control among middle-aged and older individuals with adult-onset asthma.

Study population and sample
The present study was performed within the Obstructive Lung diseases in Northern Sweden (OLIN) studies in the county of Norrbotten, Sweden.
The study sample consists of individuals participating in a follow-up of two randomly selected adult population-based cohorts initially recruited in 1996 and 2006, respectively, and a third cohort recruited in 2016.All cohorts were invited to a postal survey, performed in 2016 (19)(20)(21).Based on the 2016 survey of the three cohorts, individuals with asthma could be identified.These were invited to clinical examinations in 2019-2020 including a structured interview, skin prick testing (SPT), and spirometry.In addition, blood sampling for cell counts in peripheral blood and measurements of FeNO were performed.Written informed consent was obtained from all participants, and the local Ethical Review Board at Umeå University (DNR 2018-396-31) approved the study.
In total, n = 421 individuals underwent the examinations, of which n = 251 were classified as adult-onset asthma, defined as physician-diagnosis of asthma after the age of 16 years.As we aimed to characterize asthma in middle and older age, we chose to include individuals at least 35 years old of which n = 227 had answered the ACT questionnaire.

Structured interview
The interview questionnaire included detailed questions about airway diseases, respiratory symptoms, comorbidities, and medication use.Furthermore, questions about potential risk factors for obstructive airway disease, such as smoking habits, living conditions, and exposures during childhood and adulthood, education level, and occupational status were included.In addition, the participants also answered the ACT and detailed questions about medication use for asthma in the last 12 months.

Lung function
Spirometry was performed using a Jaeger Masterscope pneumotach spirometer according to European Respiratory Society/American Thoracic Society guidelines (22).Pre-bronchodilatory FEV 1 ≤ LLN (lower limit of normal) and FEV 1 /FVC ≤ LLN were presented based on the OLIN reference values for spirometry (23).

Skin prick test (SPT)
SPT was performed according to the guidelines of the European Academy of Allergy and Clinical Immunology as a single test on one forearm.Standardized solutions for a panel of ten in Sweden common specific airborne allergens (Solu-Prick, ALK, Denmark) were used.The panel of allergens included pollen (birch, timothy, mugwort), furred animals (dog, cat, horse), mite (Dermatophagoides pteronyssinus and Dermatophagoides farinae), and mold (Cladosporium and Alternaria alternata).A mean wheal size of ≥3 mm was defined as a positive reaction.

Fractional exhaled nitric oxide (FeNO)
FeNo was measured using NIOX VERO™, and a value of at least 25 ppb was defined as elevated.

Definitions
Adult-onset asthma was defined as a physician diagnosis of asthma after the age of 16 years.An ACT-score of at least 20 was defined as controlled asthma, while a score of 19 or lower was defined as uncontrolled asthma.Allergic sensitization was defined as at least one positive SPT reaction.The use of inhaled corticosteroids (ICS use) was defined as maintenance treatment, i.e.ICS use daily.ICS dose was categorized as low, medium, or high dose according to GINA (18).Ever smoking was defined as current or former smoking.In the age-stratified analysis, subjects between 37 and 64 years of age (n = 120) were defined as middle-aged (none of the participants was 35 or 36 years old), and those at least 65 years of age (n = 107) were defined as older age.

Statistical analysis
The IBM Statistical Package for the Social Sciences (SPSS) Statistics 27 was used for statistical analysis.Differences in proportions were assessed by the Chi-square-test and Fisher's exact test when appropriate, while the student's t-test was used to assess differences in means.A p-value <0.05 was considered statistically significant.Odds ratios (OR) with 95% confidence intervals (95% CI) for clinical biomarkers and comorbidities associated with uncontrolled asthma were estimated by multivariable logistic regression analysis.The multivariable models evaluating associations between uncontrolled asthma and clinical biomarkers as well as comorbidities, respectively, among all were adjusted for age, sex, ever smoking, BMI ≥30, and ICS use/-dose.In the age stratified analyses, the multivariable models were adjusted for sex and ever smoking only.

Results
Of all participants 66.5% were women, and the mean age was 63.9 years, range 37-82 years.The demography of the study population is shown in Table 1.

Clinical biomarkers associated with uncontrolled adult-onset asthma
In total, 27.8% (n = 63) had uncontrolled asthma (ACT ≤ 19).Respiratory symptoms and high dose ICS use were more common in uncontrolled than in controlled asthma (Table 2).Blood neutrophils ≥5/µl, BMI ≥30, and FEV 1 ≤ LLN were also more common among those with uncontrolled than controlled asthma, however, this was not seen for FEV 1 /FVC ≤ LLN (Table 3).
In the unadjusted logistic regression analyses, BMI ≥30 was associated with uncontrolled asthma (OR 1.85, 95% CI 1.02-3.35),while the significance was lost when adjusting for age, sex, and ever smoking as well as for ICS use and dose, respectively.A blood neutrophil count ≥5/µl was significantly associated with uncontrolled asthma in both the unadjusted analysis (OR 2.38; 1.17-4.83),as well as in the adjusted models (Table 4).A similar pattern of associations was seen when including only participants who reported current use of any asthma medication (n = 209, 92% of included) in the analysis, as this should exclude participants possibly being in remission of asthma (Supplemental Table 1).

Comorbidities associated with uncontrolled adult-onset asthma
Allergic and chronic rhinitis, ischemic heart disease, depression and anxiety, having a chronic pain condition as well as malignancy were significantly more common in uncontrolled than in controlled asthma (Figure 1).With the exception of chronic pain conditions, these associations were also significant in the multivariable logistic regression model adjusted for age, sex, ever smoking and ICS use accordingly: chronic rhinitis OR 3.16 (95% CI 1.66-5.70),allergic rhinitis OR 2.42 (1.25-4.69),ischemic heart disease OR 2.81 (1.05-7.52),depression/anxiety OR 3.32 (1.19-9.26)and malignancy OR 3.13 (1.31-7.51).The significance remained for all associations when adjusting the model for ICS dose instead of ICS use (Table 4).
A different pattern was seen in the older age group (65-82 years, n = 107) where chronic rhinitis, ischemic heart disease, malignancy, and having depression/ anxiety were all associated with uncontrolled asthma in the adjusted analyses.No associations with uncontrolled asthma were found for BMI ≥30, blood eosinophils ≥0.3/µl or blood neutrophils ≥5/µl (Table 5).

Discussion
In this population-based study of adult-onset asthma, we found associations between uncontrolled asthma and chronic rhinitis, ischemic heart disease, malignancy as well as depression and anxiety disorders in older individuals.Of the included clinical biomarkers, a higher blood neutrophil count was associated with uncontrolled asthma in the study population.Among middle-aged with adult-onset asthma, uncontrolled asthma was associated with elevated eosinophil and neutrophil blood counts, but less so with comorbidities.
In our study of uncontrolled adult-onset asthma, a blood eosinophil count ≥0.3/µl was indeed associated with uncontrolled asthma in the middle-aged, but not among the older participants.One other study has convincingly shown that higher levels of blood eosinophils associate with uncontrolled asthma (24), but generally, uncontrolled adult-onset asthma has to date been rarely described with regards to objective clinical biomarkers, especially in middle-and older age (1).Other studies based primarily on middle-aged with adult-onset asthma report conflicting findings on associations between higher levels of blood eosinophils and uncontrolled (7) or severe asthma (25,26).Thus, adult-onset asthma might be a phenotype with a less clear pattern regarding the association between blood eosinophils and asthma control.
In our study, no significant association between elevated FeNO and uncontrolled asthma was found.Elevated FeNO has previously been shown to associate with an increased frequency of asthma exacerbations as a surrogate for uncontrolled asthma in adults (27), but not to severe adult-onset asthma (26).Especially among the elderly, our results do not indicate any strong relationship between FeNo as a marker of Type-2-inflammation and asthma control.
We found significant associations between uncontrolled asthma and blood neutrophils of at least 5/µl,  BmI: body mass index; feno: fractional exhaled nitric oxide.results were presented as odds ratios (or) with 95% confidence intervals (95% CI).all models adjusted for sex and ever smoking.Bold font indicates p < 0.05.and this association was particularly strong in middle-aged individuals.Importantly, only a few subjects were current smokers, and about 43% were former smokers in our study.Excluding ever smoking from the logistic regression models did not change the significance of the association between uncontrolled asthma and neutrophils ≥5/µl.An elevated blood neutrophil count has in previous studies been associated with a specific asthma phenotype characterized by female sex and obesity, which tended to relate to poorer asthma control (1,28).Also smoking status might contribute to different types of airway inflammation in asthma, and associations between higher levels of blood neutrophils and FEV 1 decline (29), uncontrolled asthma (7), and frequent asthma exacerbations (30) have been reported.Although the neutrophilic severe asthma phenotype has been repeatedly identified, varying cutoff levels for blood neutrophils are used throughout the literature, and the role of neutrophilic inflammation in asthma control is still unclear.
Furthermore, we found an association between a BMI of at least 30 and uncontrolled asthma.A study from the US showed an increasing risk of adult-onset asthma with increasing BMI, the latter also associated with worse asthma control (31).Results of the Seinäjoki Adult Asthma Study also highlight the risk of uncontrolled asthma in obese patients in the long-term (3,7), and in 2005, the OLIN-studies reported the risk for incident adult asthma related to obesity regardless of sex or atopic status (8).However, a recent review emphasizes the complexity of the relationship between asthma and obesity whereas also other factors, such as a family history of both asthma and obesity have to be considered (32).The same review also highlights the influence of several increased inflammatory markers in obesity as well as the mechanistic effects of increased body fat on lung function regarding the pathogenesis of the obese asthma phenotype.
The awareness to assess and treat comorbidities in adult asthma is increasing (12,33), and our results further highlight the importance of reviewing the patients' total medical history as several comorbidities were associated with uncontrolled asthma, especially among individuals older than 60 years.The association between uncontrolled asthma and chronic rhinitis was observed in both middle-aged and older individuals, in middle-aged also for allergic rhinitis.The co-existence of asthma and chronic rhinitis with nasal polyposis has been described as a specific phenotype in asthma that is difficult-to-control (34), as previously seen in the large GA2LEN-survey in all ages (35).Symptoms related to chronic rhinitis and nasal polyposis might be overlapping with those related to asthma in the clinical setting, and there is a need to identify and treat this condition correctly.
A Dutch study found associations between uncontrolled asthma and in particular cardiovascular disease and depression/anxiety disorders in a cohort of asthma patients with comparable age distribution (11).In our study, only very few subjects with uncontrolled asthma had a co-existing depression/anxiety disorder, thus the analysis was less powered, but also our results suggest strong associations.Similarities in the presentation and perception of symptoms might however partly explain this association, as discussed also by others (36) whereas also asthma might be misclassified as in fact heart disease or vice versa.This highlights the importance of correctly identifying and treating diseases with a similar symptom profile, to reduce misclassification of asthma among elderly as well as improving asthma control in adult-onset asthma.
The association between ischemic heart disease and uncontrolled asthma in particular among older individuals in our study has also been reported by others (37,38).Current as well as former smoking is a common risk factor for both uncontrolled asthma and ischemic heart disease (11).We adjusted our age-stratified analysis of the association between uncontrolled asthma and comorbidities for ever smoking, and excluding the variable from the model did not change the significance of the association.Smoking is thus unlikely the sole explanation for our result.Others reported an increased risk for incident ischemic cardiovascular disease in persistent asthma (39), where systemic inflammation potentially may contribute.The cross-sectional design of our study did not allow the analysis of a similar causal association.
The association between clinical biomarkers, such as blood eosinophils and neutrophils, and uncontrolled asthma was not significant in older age in our study, in contrast to comorbidities whereof several were associated with uncontrolled asthma.Our results are partly supported by the findings of a previous study conducted among participants with mild to moderate asthma in a comparable age range (40).In that study, elevated inflammatory markers in both sputum and peripheral blood were found in all ages, with a change in the inflammatory pattern from dominating eosinophilic inflammation among middle-aged individuals toward a higher neutrophilic inflammation in older age.Effects of aging on the immune system, in particular the function of eosinophils, have been described earlier (41) and might contribute to explain our result.
Strengths of our study are the population-based design, including both phenotypes with Th-2-inflammation and without increased Th-2inflammatory markers, and representation of degrees of severity from mild to severe asthma.Clinically well-established biomarkers were used for the assessment of participants, and the ACT for assessment of asthma control is a validated and widely used tool (16).
We did not include the frequency of asthma exacerbations when defining uncontrolled asthma as suggested by some guidelines, and the lack of a universal definition of uncontrolled asthma might be critical when comparing our results to similar studies.Age at onset of asthma may be affected by recall-bias in particular among the older participants and those with long duration of asthma.Some of the participants might have been in remission of asthma, however, a logistic regression analysis of those participants reporting current use of asthma medication as a surrogate for prevalent asthma showed consistent results (Supplemental Table 1).Furthermore, the respiratory symptoms assessed by ACT might also be related to other conditions, such as heart disease or mental disorders.A possible misclassification of COPD in elderly participants has to be considered, however, only 11% of all participants had a pre-bronchodilator FEV 1 /FVC lower than 0.70, and the smoking prevalence was lower than in the general Swedish population.Thus, a possible misclassification of disease is unlikely to significantly influence our results.Furthermore, we defined adult-onset asthma as an asthma-diagnosis after the age of 16.Varying use of terms like adult-onset asthma or late onset asthma and lack of a common definition in the current literature have to be considered when discussing characteristics of asthma diagnosed in adult age.Furthermore, only 14% of the participants with uncontrolled asthma in our study reported the use of high dose ICS, and about 56% reported maintenance treatment with ICS.Undertreatment with ICS might contribute to our result of a quite large number of participants with uncontrolled asthma.

Conclusions
In conclusion, we found differences between middle-aged and older individuals with adult-onset asthma regarding associations between uncontrolled asthma, comorbidities, and clinical biomarkers.Among older individuals with adult-onset asthma, chronic rhinitis, ischemic heart disease, malignancy as well as depression and anxiety disorders were associated with uncontrolled asthma, but none of the biomarkers expressing Type-2-inflammation.Among middle-aged, clinical biomarkers including blood eosinophils and neutrophils were more strongly related to uncontrolled asthma than in the elderly.In addition to the assessment of established biomarkers, lung function, and sensitization status, our results emphasize the importance of assessing and treating comorbidities with similar symptom presentation as adult-onset asthma.

Figure 1 .
Figure 1.Prevalence of comorbidities (%) and p-values for test of association with uncontrolled asthma among subjects with adult-onset asthma.

Table 1 .
Characteristics of the study population.

Table 2 .
symptoms and medication use in controlled and uncontrolled adult-onset asthma.

Table 4 .
associations between (a) biomarkers and (B) comorbidities and uncontrolled adult-onset asthma, were analyzed by logistic regression.
feno: fractional exhaled nitric oxide.results were presented as odds ratios (or) with 95% confidence intervals (95% CI).model 1: adjusted for age, sex, BmI (Body mass index) ≥30, ICs use (inhaled corticosteroids used most days per week) and ever smoking.model 2: adjusted for age, sex, BmI ≥30, ICs dose (categorized into none or low, medium, high dose according to GIna) and ever smoking.Bold font indicates p < 0.05.

Table 3 .
Characteristics and biomarkers in controlled and uncontrolled adult-onset asthma.

Table 5 .
associations between (a) biomarkers and (B) comorbidities and uncontrolled adult-onset asthma in middle-aged and older adults separately, analyzed by logistic regression models.