Abstract

Capsule summary

INTRODUCTION

Asthma is reported in 4–13% of persons >60 years of age^1–5. Asthma in the aged is not only more prevalent than previously considered, but morbidity and mortality in this group are high with 50–66% of all asthma deaths occurring in adults >65 years of age^{2, 6–9}. Although co-existing lung disease¹⁰, co-morbidities^11–12 and under-treatment^{2, 8} contribute to worse outcomes, the effect of aging on asthma-related airway inflammation and its regulation is also likely a major factor in heightened morbidity. Because of increases in aging of the worldwide population, the number of people >60 years is projected to grow to 72 million by the year 2030¹³; and with this increase so will the unmet needs of aged patients with asthma.

The limited data available on airway inflammation in aged adults with asthma suggest that there are increased airway neutrophils^14–15. However, these studies have not considered the potential impact of “inflamm-aging,” that is, the low-grade basal systemic inflammation (e.g., increased interleukin [IL]1-β, IL-6 and tumor necrosis factor [TNF]-α) often observed with aging¹⁶, nor have they been conducted in inner-city subjects, a population with a higher asthma morbidity^17–18. Additionally, analyses have not correlated airway inflammation with clinical features of disease or appropriately controlled for inhaled corticosteroid (ICS) use, a factor that may influence airway cellularity¹⁹. Establishing the characteristics of airway inflammation in aged asthma patients is an important step towards a greater understanding of the high morbidity and mortality rates in this population, and to begin to formulate more effective treatments.

To address these unanswered questions, we recruited aged and younger asthma patients from an inner-city population; defined their clinical characteristics; and obtained sputum samples to measure patterns of airway inflammation. Furthermore, to more fully assess the age independent effects of asthma, control populations of aged-matched subjects without asthma were included. Thus, our study design allowed us to evaluate three aspects of aging and airway inflammation: a) the effect of aging on asthma (i.e., aged vs. younger patients with asthma), b) the effect of asthma on aging (i.e., aged patients with asthma vs. aged controls), and c) the effect of aging itself (i.e., aged vs. younger controls).

METHODS

RESULTS

Past and present asthma history

Aged vs. younger asthma patients had a significantly lower FEV₁ (69.1% vs. 81.0%, respectively, P <0.01; Table III). Although all patients with asthma had prior evidence of reversibility of obstruction as documented in their medical records, the mean change in mL of FEV₁ after bronchodilator therapy was significantly greater in the younger asthma patients (P =0.03). Aged patients, on average, received their diagnosis at an older age than younger patients (P <0.01), and had asthma for a greater duration (P <0.01). In the past 12 months, no significant differences in the percentage of patients with a history of past intubation, or the number of hospitalizations, urgent care visits and exacerbations requiring oral or intravenous corticosteroids were observed among aged vs. younger patients (P >0.05 for all comparisons).

TABLE III

Asthma: past and present characteristics

	Asthma
	Aged	Younger	P value
Pulmonary Function
Pre-BD FEV1, % predicted, mean (SD)	69.1 (17.7)	81.0 (13.1)	<0.01
FEV1/FVC, mean (SD)	84.4 (10.5)	86.7 (10.9)	0.37
Max β agonist FEV1 response (ml), mean (SD)*	131 (135)	271 (326)	0.03
Max β agonist FEV1 response (% change), mean (SD)**	5.8 (7.3)	7 (7.5)	0.53
Asthma History
Age of asthma onset, median [IQR]	34 [14.5–50.5]	5 [2–9]	<0.01
Years with asthma, median [IQR]	35 [20–51]	23 [20–27]	<0.01
Positive Family History of Asthma***	21 (60.0%)	26 (70.3%)	0.5
History of intubation	7 (20%)	4 (10.8%)	0.45
History of hospital admission	22 (62.9%)	19 (51.4%)	0.45
Asthma Outcomes in past 12 months
Freq. ER visits/patient, mean (SD)	1.03 (1.8)	1.3 (2.4)	0.63
Freq. Hospitalizations/patient, mean (SD)	0.3 (0.8)	0.1 (0.4)	0.25
Freq. Urgent Care visits/patient, mean (SD)	1.9 (2.9)	1.2 (2.1)	0.29
Freq. Patients with >1 oral/IV CS burst	28 (80%)	27 (75%)	0.83
ACT score, mean (SD)	15 (4.9)	18 (4.3)	0.01
Mini-AQLQ, mean (SD)	3.8 (1.4)	4.4 (1.3)	0.07
Symptom Score, mean (SD)	3.7 (1.5)	4.6 (1.5)	0.01
Activity Score, mean (SD)	4.2 (1.8)	5 (1.5)	0.05
Environment Score, mean (SD)	3.3 (1.7)	3.6 (1.5)	0.50
Emotional Score, mean (SD)	4.2 (1.7)	4.3 (1.7)	0.86
Current Treatment
ICS Dose: Low	4 (15.4%)	4 (16%)	0.93
Medium	11 (42.3%)	12 (48%)
High	11 (42.3%)	9 (36%)
None	9 (25.7%)	12 (32.4%)	0.53
Long-acting β-agonist	22 (62.9%)	12 (32.4%)	0.02
Leukotriene Receptor Antagonist	13 (37.1%)	6 (16.2%)	0.08

Open in a separate window

Data are presented in absolute numbers (percentages), means (Standard Deviation=SD) or median [Interquartile range=IQR]; ACT=Asthma Control Test; AQLQ=Asthma quality of life questionnaire; BD-bronchodilator; CS=Corticosteroids; ER=emergency room; FEV₁=Forced expiratory volume in 1 second; Freq.=Frequency; FVC= Forced vital capacity; ICS=inhaled corticosteroid; IV=intravenous.

^*Absolute ml improvement of FEV₁ after bronchodilator treatment;

^**Percent change in FEV₁ after bronchodilator treatment;

^***Family history defined as at >1 first-degree relative.

Aged asthma patients had significantly worse asthma control as measured by the ACT (15.0 in the aged vs. 18.0 in the young, P =0.01). There was not a significant difference in the overall impact of asthma on the quality of life in the aged compared to the younger group (mini-AQLQ scores of 3.8 vs. 4.4, respectively, P =0.07). However, there was a statistically significant difference in aged vs younger groups in the AQLQ domains of Symptom and Activity Scores (3.7 vs. 4.6, P=0.01, and 4.2 vs. 5.0, P=0.05, respectively).

Most aged and younger patients with asthma were taking an ICS (74.3% vs. 67.6%, P =0.71). ICS dose was not significantly different between the two patient groups; most received medium to high doses.

Comparison of sputum inflammatory cell patterns between aged and younger patients with asthma

The total numbers of sputum cells were not significantly different between all 4 study groups (P=0.75; Fig 1A). The total numbers of sputum neutrophils was significantly elevated in the aged (30.5 ×10⁴/mL) compared to the younger asthma patients (13.0 ×10⁴/mL, P <0.01; Fig 1A). Additionally, the percentage of sputum neutrophils was significantly elevated in the aged asthma patients (23.1%) compared to younger patients (6.9%, P <0.01) and to aged-matched controls (1.9%, P <0.01; Fig 1B). However, no significance differences in the total numbers (3.0 ×10⁴/mL vs. 7.0×10⁴/mL, respectively, P =0.12) and percentages of sputum neutrophils (1.9% vs. 3.5%, respectively, P =0.20), were observed between aged and younger controls suggesting that these changes are specific to asthma. After adjusting for ICS adherence, ICS dose, years with asthma, atopy, race, sex, BMI and pack-years of smoking, the associations remained significant.

Open in a separate window

Figure 1

Aged patients with asthma have increased sputum neutrophils and eosinophils. (A) Total cell numbers per mL of sputum for neutrophils, eosinophils macrophages and total cells. (B) Percentages of neutrophils, eosinophils, macrophages and lymphocytes. Box-plot diagram representing median (mid-line), interquartile range (shaded boxes) and range (whiskers). Groups not sharing the same letter are significantly different (P <0.05).

Similarly, the absolute number and percentage of sputum eosinophils was significantly elevated in the aged (7.0×10⁴/mL and 3.8%, respectively) compared to younger (2.0×10⁴/mL, P <0.01 and 1.2%, P <0.01, respectively) asthma patients and to aged controls (1.0×10⁴/mL, P <0.01 and 0.7%, P <0.01, respectively). This increase was not secondary to aging as suggested by the non-significant difference between eosinophil numbers (P =0.69) and percentages (P =0.85) in aged and younger control subjects. Analyses adjusting for ICS adherence, ICS dose, years with asthma, atopy, race, sex, BMI and pack-years of smoking showed similar results.

Association between clinical characteristics of asthma and inflammatory markers

The mean difference in ACT scores among asthma patients with high vs. low IL-6 levels in aged subjects was −4.5 units compared to 2.4 units among younger patients (P=0.01); suggesting a differential relationship between IL-6 and asthma control according to age (Fig 2A). Similarly, aged vs. younger patients with high vs. low MIP-3α/CCL20 levels had an ACT score differences of −3.6 vs. 2.3 units (P=0.03), also suggesting an age-specific effect of this cytokine. No significant age-associated interactions between neutrophil percentage, eosinophil percentage, IL-1β, IL-5, IL-8, IL-17A, IL-17F, IL-23, IL-27, GM-CSF and ACT were observed (P>0.05 for all other comparisons, Fig 2A and Table EI).

Open in a separate window

Figure 2

Association between sputum inflammation and asthma control and hospitalizations according to age group. Asthma patients were classified by sputum cellular percentages and cytokine expression into “high” or “low” to evaluate the association between (A) Asthma Control Test (ACT) and (B) the Odds Ratio (OR) for asthma hospitalization in the previous 12 months. Interaction is significant at P<0.05.

The odds ratio (OR) for asthma hospitalizations (over the past 12 months) of aged patients with high neutrophil percentages was 6.7 vs. 0.1 in younger patients (P<0.01 for interaction; Fig 2B). Additionally, aged asthma subjects with high sputum IL-1β, IL-6, MIP-3α/CCL20 had increased OR for hospitalization at 7.0, 60.0, 28.0 vs. 0.4 (P=0.03), 2.5 (P=0.05), 0.7 (P=0.01) in the younger. Conversely, no significant interaction between age group and hospitalization was observed for eosinophil percent, IL-5, IL-8, IL-17A, IL-17F, IL-23, IL-27, GM-CSF (P>0.05 for all comparisons, Fig 2B and Table EI).

DISCUSSION

Our study investigated the relationship of aging on airway inflammation in asthma, as measured in induced sputum, and its association with clinical outcomes, including disease control. We found that aged asthma patients have a different pattern of sputum inflammation, which is characterized by increases in both neutrophils and eosinophils and elevated protein expression of IL-1β, IL-6, IL-8, eotaxin-1, GM-CSF and cytokines associated with Th17 cells. These identified markers have potential clinical relevance. In patients >60 years, elevated sputum IL-6 and MIP3α/CCL20 were significantly associated with decreased asthma control; and elevated sputum neutrophils, IL-1β, IL-6 and MIP3α/CCL20 correlated to an increased risk of hospitalization in the past 12 months. Our findings suggest that selective age-related inflammatory markers have relationships with clinical features of asthma. Further research of these pathways may provide new management strategies for this vulnerable group of patients.

The limited data to define airway inflammation in aged adults suggest that this group of patients tends to have increased airway neutrophils^29–31. The sputum neutrophils that we found in our groups of asthma patients is slightly lower than previous reports^29–32, but these differences may relate to methods of sputum collection³³, ICS doses or underlying severity of asthma. Nonetheless, our groups of younger and aged patients received comparable doses of ICS, and no relationship was found between ICS and sputum neutrophils. Although previous studies provided an initial insight to features of airway inflammation in the aged, these reports did not match study populations for disease severity and intensity of treatment^{14–15, 34–35}, did not specifically address the effect of aging on immune markers of inflammation³⁴, or evaluate inner-city patients, a highly at-risk asthma population in the United States.

Aging is associated with several immune system changes which may affect inflammation³⁶. Increased concentrations of pro-inflammatory cytokines (e.g., TNF-α, IL-6), in the absence of an underlying infection or systemic inflammatory disorder, is a process referred to as “inflamm-aging”¹⁶, and is frequently observed in aged adults. In the lung, the impact of “inflamm-aging” is reflected by an elevation of bronchoalveolar lavage fluid (BALF) neutrophils and IL-6, even without obvious airway disease^37–38; and changes proposed to be secondary to increased systemic inflammation which “spills” over into the airways^39–40. These aged-related changes highlight the need to evaluate appropriate controls to differentiate age vs. asthma-related variability. Although we did not find evidence of “inflamm-aging” in our aged control population, our data offers insight into how aging may affect inflammation in asthma. A key future step will be to identify and establish mechanisms leading to these changes, how they may affect the increased morbidity in this population, and seek age-specific treatments to potentially improve outcomes.

One interpretation of the findings in our population of aged asthma patients is that the “inflamm-aging” patterns were reflected in sputum samples by increased neutrophils, IL-6, IL-8, and cytokines associated with Th17 cells. However, the lack of significant difference in neutrophils and levels of these cytokines among aged and younger controls suggests that these findings were not necessarily due to aging itself, but rather through asthma and aging. There are intricate interactions between IL-6, IL-8 and Th17 cells, which may contribute to, and explain, age-related differences in sputum neutrophils and are reflected in clinical outcomes, including diminished asthma control. For example, IL-6 differentiates naïve T-cells towards a Th17 subset (rather than Tregs), and increases the stability of committed Th17 cells^41–49. Conversely, neutralization of IL-6 in T-cell co-cultures decreases Th17 cells⁵⁰, and administration of anti-IL-6 to patients with rheumatoid arthritis significantly decreases peripheral Th17 cells⁴⁴. IL-27 is a complex cytokine with both pro-and anti-inflammatory properties; inhibiting the development of Th17 from naïve T-cells, but with limited effect on the stability of committed Th17 cells^51–52. Furthermore, IL-27 has been reported to inhibit Treg cells^51–52. Th17 cells promote neutrophilia through multiple mechanisms^53–54, including secretion of the neutrophil chemo-attractant, CXCL8 (IL-8)⁵⁵, and are associated with asthma in younger adults that is less responsive to ICS^56–57. Although we did not measure clinical responsiveness to ICS in our asthma subjects, aged patients, despite similar ICS treatments, had significantly decreased asthma control compared to younger patients, which was associated with increased IL-6 and the Th17 cytokine, MIP3α/CCL20. Furthermore aged asthma patients with elevated sputum neutrophils, IL-1β, IL-6 and MIP3α/CCL20 had a greater risk of hospitalization.

The significantly increased IL-6 in the aged asthma patients is of particular interest and potential clinical relevance. For example, elevated systemic levels of IL-6 are associated with increased frailty and higher morbidity of chronic diseases in the aged^58–60. Of note, our population of aged controls did not have elevated sputum IL-6, which may be due to the fact that we measured IL-6 in the sputum and not plasma. In younger patients, variants of the IL-6 receptor⁶¹, polymorphisms of the IL-6 gene⁶² and elevated sputum IL-6 are associated with more severe asthma⁶³. Moreover, whether an over-expression of IL-6 contributes to diminished corticosteroid responsiveness with aging has yet to be established.

We also found an increase in sputum eosinophils in the aged asthma patients compared to the younger asthma population, which may reflect poorer asthma control in the latter group. Although increases in eosinophils reflect disease severity, risks for exacerbations, and a T2 immune profile, our initial prejudice, and that in the literature, suggested that the dominant inflammatory cell would be the neutrophil in the aged group. However, allergen sensitized and airway challenged aged mice developed both elevated airway eosinophil and neutrophil populations^64–65. A potential explanation for increased eosinophils in the aged asthma subjects, is a high frequency of allergic sensitization, elevated eotaxin and IL-5, all implying a T2 profile. In younger patients with asthma, combined elevated airway neutrophils and eosinophils often identify patients whose asthma is more difficult to control⁶⁶. The persistence of sputum eosinophils in the aged asthma likely implies more severe disease, diminished responsive to ICS and a possible need for another approach, i.e. biologics, to achieve greater disease control.

The function of Treg cells in asthma is not completely understood, but has been proposed to suppress and regulate inflammation including decreasing Th17 cells^{43, 67}. Studies conducted in younger patients with asthma have shown that compared to age-matched controls, the numbers of peripheral and airway Treg cells are decreased^68–71, a concept found in our younger patients. The role of Treg cells in aged patients with asthma, however, has yet to be defined and has not been evaluated from airways samples. Most prior work suggests that Treg cells increase with aging, however the effect of aging on their function is less clear^72–75. The underlying mechanisms for increased Treg cells with aging may include increased clonal expansion, in particular due to chronic viral infection (e.g., cytomegalovirus), increased IL-15 protein and receptor expression (promoting Treg development and maintenance), and decreased expression of pro-apoptotic molecules^76–78. One study has specifically analyzed markers of peripheral Treg cells in aged asthma patients and reported decreased numbers compared to normal controls³⁴. However, younger patients were not included for comparison. In our study, we noted significantly decreased sputum IL-15 in the younger patients with asthma compared to age-matched controls (Table IV), but no significant difference between aged asthma patients and controls. Although we did not find statistical differences between the Foxp3 MFI (as a marker of Treg function) of aged vs. younger patients with asthma and between patients with asthma vs. aged-matched controls, it may be that Treg suppressive activity is more accurately determined by co-culture proliferation experiments.

Our study has several strengths and limitations that are worth acknowledging. The study population was from a single institution, thus limiting the generalization of our findings. However, our center is one of the largest providers of medical care to the East Harlem, New York population, which has one of the highest rates of asthma in the United States^79–80. We recruited patients from an inner-city minority population, a group with increased morbidity and mortality^81–82. Previous studies of aged patients with asthma have not specifically addressed inner-city populations. Our study is also strengthened by the comparison of age-matched normal controls and younger patients with asthma, which allowed us to examine the effect of asthma on aging, aging on asthma, and the effect of aging itself. Additionally, as higher doses of ICS may increase sputum neutrophils¹⁹, we included a run-in period to accurately measure and then control our analyses for ICS dose and adherence. We excluded subjects with a smoking history of greater than 10 pack-years to remove subjects with COPD or possible asthma-COPD overlap³⁵. Limitations to our study include a one-time sputum collection, thus, we could not evaluate the longitudinal stability of our findings. Additionally, by nature of their age, aged asthma patients had a longer duration of disease.

Our study of aged inner-city adults with asthma, suggest that both aging and asthma affect airway inflammation. Establishing the characteristics of airway inflammation in aged adults and underlying mechanisms of these processes is an important first step in the development of age-specific or personalized therapy. Current treatment of asthma applies the pathophysiology and therapeutic principles of asthma largely based on work in younger patients to aged individuals. We feel that our study helps set the stage for further investigation on the role of aging, asthma and airway inflammation in a high risk group with significant unmet treatment needs.

Supplementary Material

Acknowledgments

Abbreviations used

Footnotes

REFERENCES