Relationship between asthma and cognition: the Cooper Center Longitudinal Study
E. Sherwood Brown, Department of Psychiatry, The University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd, MC 8849, Dallas, TX 75390-8849, USA.
Minimal data are available on the relationship between asthma and cognitive performance. In this report, we examine the relationship between asthma and cognitive performance in older adults, a subpopulation with elevated risk of cognitive impairment.
We conducted a cross-sectional, retrospective analysis of 1380 participants age ≥55 who completed preventive health examinations at the Cooper Clinic in Dallas, TX. Cognition was assessed using the Montreal Cognitive Assessment (MoCA), a brief test for mild cognitive impairment. Data were analyzed in a multiple logistic regression using MoCA scores suggestive of cognitive impairment as the dependent variable.
When controlling for demographic characteristics, self-rated health status, inhaled corticosteroid use, and FEV1/FVC, asthma were associated with 78% increased risk of cognitive impairment (P = 0.02) as defined by MoCA score.
In the largest sample examined to date, we have identified a significant relationship between asthma and cognitive impairment in older people.
Asthma is a common medical condition with increasing prevalence . Clinical characteristics, such as transient hypoxia, could lead to poor cognition in asthma patients as hypoxia is associated with hippocampal neurotoxicity . Patients with asthma also have high rates of depression  and anxiety disorders , which can be associated with poor cognition. The treatment of asthma could compromise cognition, as corticosteroid therapy is associated with cognitive changes . Cognition is important in asthma patients because it is a predictor of medication nonadherence in medically ill patients [6, 7].
Minimal data are available on cognition in asthma. In adults, performance on the Bender–Gestalt test suggestive of cognitive impairment was observed in 65% of hospitalized asthma patients . Medications were not discussed in this study, but clinical practice suggests patients were likely receiving systemic corticosteroids while hospitalized. However, cognitive performance may also be poor in asthma patients not receiving systemic corticosteroids. We reported that adults with asthma had mean memory performance in the low-normal to mildly impaired range prior to receiving corticosteroid bursts .
In the current study, we hypothesized that individuals age ≥55 with asthma seen at a preventive medicine clinic would have greater cognitive impairment than those without an asthma history.
Materials and methods
From 02/09 to 09/12, 1380 generally healthy patients age ≥55 years were seen at the Cooper Clinic in Dallas, Texas for preventive medical examinations that included extensive medical history, Center for Epidemiologic Studies Depression (CES-D) scale, physical examination, laboratories analyses, pulmonary function testing, and the Montreal Cognitive Assessment (MoCA). Participants gave informed consent for use of their data in the Cooper Center Longitudinal Study database maintained by The Cooper Institute, and approved by the Institute's Institutional Review Board.
Montreal Cognitive Assessment
Cognition was assessed using the MoCA, a validated eight-item test that assesses eight cognitive domains including attention, executive functions, and memory. MoCA scores range from 0 to 30 with a score of <26 defining possible mild cognitive impairment (MCI) . As recommended, one additional point was added to scores of participants with ≤12 years of education.
Koko PFT System was used to measure lung function with the best of three trials recorded in the database. Hankinson equations were used to calculate forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC). An FEV1% score, the ratio of FEV1 and FVC, of <70 is suggestive of obstructive lung disease .
Vitamin D blood levels were measured following standardized procedures as described in Hoang et al. . A level <30 ng/ml suggests insufficient vitamin D . The 10-item CES-D quantified depressive symptoms during the previous week. A score of ≥10 suggests clinically significant depressive symptoms . Global health rating of each participant was assessed using a four-point (poor, fair, good, excellent) self-report assessment developed for clinical assessment by the Cooper Clinic. History of depression and anxiety was also assessed by self-report.
Multiple logistic regression predicted the dependent variable of MoCA score (≥26 or <26) based on covariates (e.g., demographic/health information as in other Cooper Clinic data analyses , pulmonary function values, variables potentially related to cognition and/or asthma [e.g., vitamin D levels , diabetes , and asthma medications]) and is presented in Table 2. SPSS® software version 14.0 (SAS Institute, Cary, NC) was used for the analyses, including summary statistics using chi-squared analyses in Table 1.
Table 1. Demographic information of participants (N = 1380), Cooper Center Longitudinal Study
|Sex, n (%)|
|Female||28 (27.5)||325 (25.4)||0.64|
|Male||74 (72.5)||953 (74.6)|
|Age, n (%)|
|&!hairsp; <65 years||70 (68.6)||911 (71.3)||0.57|
|≥65 years||32 (31.4)||367 (28.7)|
|Body mass index, n (%)|
|Underweight or normal||30 (29.4)||433 (33.9)||0.39|
|Overweight or obese||72 (70.6)||845 (66.1)|
|Education, n (%)|
|≤12 years||16 (15.7)||242 (18.9)||0.51|
|&!hairsp; >12 years||86 (84.3)||1036 (81.1)|
|Ethnicity, n (%)|
|Non-Hispanic Caucasian||98 (96.1)||1224 (95.8)||1.00|
|Other||4 (3.92)||54 (4.23)|
|Smoking status, n (%)|
|Current smoker||10 (9.8)||97 (7.6)||0.44|
|Nonsmoker or past smoker||92 (90.2)||1181 (92.4)|
|Alcohol use status, n (%)|
|Current drinker||81 (79.4)||1031 (80.7)||0.80|
|Nondrinker or past drinker||21 (20.6)||247 (19.3)|
|Health rating, n (%)|
|Poor or fair||23 (22.5)||177 (13.8)||0.03|
|Good or excellent||79 (77.5)||1101 (86.2)|
|Vitamin D level, n (%)|
|<30 ng/ml||28 (27.5)||378 (29.6)||0.74|
|≥30 ng/ml||74 (72.5)||900 (70.4)|
|Diabetes, n (%)|
|Yes||9 (8.8)||60 (4.7)||0.09|
|No||93 (91.2)||1218 (95.3)|
|CES-D, n (%)|
|<10||94 (92.2)||1191 (93.2)||0.68|
|≥10||8 (7.84)||87 (6.81)|
|Depression, n (%)|
|Yes||13 (12.7)||151 (11.8)||0.75|
|No||89 (87.3)||1127 (88.2) |
|Anxiety, n (%)|
|Yes||16 (15.7)||154 (12.1)||0.27|
|No||86 (84.3)||1124 (87.9)|
|FEV1/FVC (FEV1%), n (%)|
| <70||25 (24.5)||78 (6.1)||<0.001|
|≥70||77 (75.5)||1200 (93.9)|
|Inhaled corticosteroid (ICS) treatment, n (%)|
|No ICS use||74 (72.5)||1269 (99.3)||<0.001|
|Either ICS or combined ICS||28 (27.5)||9 (0.7)|
Participant characteristics are shown in Table 1 for those with and without asthma. Mean MoCA score was 26.9 ± 2.3 (range 10–30, 25‰ =26, 50th‰ <27, 75‰ <29). Groups were similar demographically except for subjective health rating (P = 0.03), corticosteroid or combined inhaler use (P < 0.001) and FEV1/FVC (P < 0.001). The vast majority of patients had mild asthma, based on FEV1/FCV, current use of corticosteroid/combined inhaler, and subjective health ratings.
Table 2 shows the association between asthma and cognition. When controlling for the noted covariates, asthma was significantly associated with MCI (OR 1.78, [95% CI, 1.09–2.91]) as defined by the MoCA. Variables potentially related to asthma severity (e.g., corticosteroid inhaler use, FEV1/FVC, health rating) were not associated with MoCA performance.
Table 2. Multiple logistic regression models of the Montreal Cognitive Assessment score, Cooper Center Longitudinal Study
|Female vs male||0.87 (0.63–1.20)||0.40|
|<65 vs ≥65 years||2.20 (1.68–2.89)||<0.001|
|Body mass index|
|Underweight or normal vs overweight or obese||1.11 (0.84–1.49)||0.46|
|Other race/ethnicity vs non-Hispanic Caucasian||0.43 (0.24–0.75)||0.003|
|Current smoker vs nonsmoker or past smoker||1.53 (0.98–2.39)||0.06|
|Alcohol use status|
|Current drinker vs nondrinker or past drinker||1.06 (0.77–1.46)||0.74|
|Poor or fair vs good or excellent||1.06 (0.73–1.54)||0.77|
|Vitamin D level|
|<30 vs ≥30 ng/ml||0.68 (0.51–0.89)||0.006|
|Yes or no||1.29 (0.74–2.25)||0.36|
|<10 vs ≥10||1.07 (0.62–1.84)||0.81|
|Yes or no||0.84 (0.53–1.35)||0.47|
|Yes or no||1.08 (0.69–1.70)||0.74|
|<70 vs ≥70||1.14 (0.70–1.85)||0.59|
|Inhaled corticosteroid (ICS) treatment|
|No ICS vs ICS or combined ICS||1.10 (0.49–2.46)||0.83|
|Yes vs no||1.78 (1.09–2.91)||0.02|
Prior cognition research in adults with asthma is limited to reports in inpatients who were likely receiving systemic corticosteroids , and patients with asthma exacerbations scheduled to initiate oral corticosteroid therapy . The current report adds to these findings by examining older, generally healthy patients with relatively mild asthma. The 78% increased risk of MCI in the asthma patients is concerning and suggests that older asthma patients may be a target population for cognitive impairment screening.
The etiology of increased risk of MCI with asthma is not clear. Groups were similar related to demographic and co-morbid factors other than those related to asthma. Given the generally mild asthma in the sample, hypoxia and recent corticosteroid exposure are also not likely causes of the cognitive changes.
The study has limitations. While asthma diagnosis was confirmed by a clinic physician, the patients were not assessed by an allergist or pulmonologist and thus lacked extensive asthma history (e.g., years) and evaluation. While the MoCA is a widely used and sensitive instrument for assessing cognition , formal neurocognitive assessment was not performed. As the sample was almost exclusively non-Hispanic Caucasian and men were overrepresented (74.4%), the findings may not be generalizable to other populations. Medication use, spirometry, and subjective health rating did not raise suspicion for chronic obstructive lung disease (COPD), and smoking was comparable in both groups. However, in a sample of older asthma patients, the diagnosis may actually represent COPD.
In a large sample of older participants, asthma was associated with a significantly increased risk of cognitive impairment. Additional research is needed to confirm this finding and determine the etiology and impact of cognitive changes in asthma.
We thank the Cooper Clinic for data collection and The Cooper Institute for data management.
Conflicts of interest
There are no conflicts of interest for any of the authors to report.
G Caldera-Alvarado assisted with data analysis and wrote portions of the manuscript. D Khan assisted with design of the study and edited the manuscript. L DeFina was involved in data collection, study design, and wrote portions of the manuscript. A Pieper wrote portions of the manuscript and extensively edited the manuscript. ES Brown designed the study, performed the data analysis, and wrote portions of the manuscript.