Depressive symptoms and over-perception of airflow obstruction in older adults with asthma

Abstract

Objective:

Older adults are at increased risk for depression and poor asthma outcomes. We examined whether depressive symptoms are associated with over-perception of airflow obstruction and a pattern of worse asthma control, but not pulmonary function.

Methods:

We recruited a cohort of adults with asthma aged 60 years and older in East Harlem and the Bronx, NY. Baseline measures included the Geriatric Depression Scale, Asthma Control Questionnaire, and mini Asthma Quality of Life Questionnaire. Spirometry was conducted at baseline to assess pulmonary function. Perception of airflow obstruction was assessed for 6 weeks following baseline by participants entering estimates of peak expiratory flow (PEF) into a programmable peak flow meter followed by PEF blows. Participants were blinded to actual PEF values. The percentage of time that participants were in the over-perception zone was calculated as an average.

Results:

Among the 334 participants (51% Hispanic, 25% Black), depressive symptoms were associated with over-perception of airflow obstruction (β = .14, p = .029), worse self-reported asthma control (β = .17, p = .003), and lower asthma-related quality of life (β = −.33, p < .001), but not with lung function (β = −.01, p = .82). Over-perception was also associated with worse self–reported asthma control (β = .14, p = .021), but not lung function (β = −.05, p = .41).

Conclusions:

Depressive symptoms were associated with greater perceived impairment from asthma, but not pulmonary function. Over-perception of asthma symptoms may play a key role in the relationship between depression and asthma outcomes in older adults.

INTRODUCTION

Asthma in older adults is associated with poor asthma control (1), high rates of emergency department visits and hospitalizations for asthma (2–4), and poor asthma-related quality of life (5). Older adults have the highest rates of asthma morbidity and mortality across all age groups (6, 7). Despite improvements in the care of asthma for younger age groups, older adults with asthma remain at risk for negative outcomes (8). There has been limited research dedicated to understanding and improving this major, intersectional public health concern.

There are high rates of comorbidity of depression and asthma across age groups, and this comorbidity is associated with worse asthma outcomes. Results from the World Mental Health Organization survey indicated that the occurrence of asthma alongside depression is ubiquitous across countries despite considerable cultural and environmental variability (9). In a large, community-based sample, having an asthma diagnosis was associated with a 64% increased likelihood of depressive symptoms (10). Even higher rates of depression exist in ethnic minority, inner-city samples of patients with asthma (11, 12). Older adults with asthma are at particularly high risk for depression (13), which has a synergistic effect on poor asthma control (14–18). Depression in older adults is a risk factor for future asthma exacerbations, emergency department visits, sleep disturbances from asthma, and worse health-related quality of life (15, 16, 19–21).

Depression may affect asthma morbidity via perceptions of asthma control, which are often the basis for treatment and daily self-management decisions. While there is some evidence that depression is associated with pulmonary function (22, 23), several studies have demonstrated that depression is more strongly linked to reports of asthma symptoms and control, but not pulmonary function (15, 24–26). This finding has significant clinical implications given the reliance on self-report measures of asthma control to guide treatment decision making in clinical practice. Depression can influence physician ratings of asthma control based on worse patient self-report in the absence of differences on pulmonary function (13). Therefore, it is critical to incorporate both self-report of asthma control and airflow obstruction at the same time points in patients with asthma and depression.

Perception of airflow obstruction is central to effective asthma control. Negative emotions and catastrophic thinking are associated with over-perception of asthma symptoms and beliefs about excessive avoidance of activities (27–30). Rumination and attentional biases might shift attention towards greater perceived general health-related threats in patients with asthma and depressed mood (31). Affective cues (positive or negative) can have a significant impact on asthma symptom perception accuracy (32). Therefore, depression can lead to over-perception via hypervigilance to somatic sensations, increased sensitivity to internal cues, and catastrophizing. Over-perception of airflow obstruction in older adults with depression might also result from medical comorbidity and confusion between symptoms of asthma and other chronic diseases (33).

The present study examined a primarily minority sample of older adults with asthma living in the Bronx and East Harlem areas of New York City. Low income, older adults with asthma from racial and ethnic minorities experience an even higher burden of asthma and high rates of health care utilization and hospitalizations for asthma (34, 35). The primary aim of the present study was to examine the relationship between depressive symptoms, perception of airflow obstruction, and asthma outcomes in older adults. We hypothesized that depressive symptoms would be associated with over-perception of airflow obstruction and worse reports of asthma control and asthma-related quality of life, but not pulmonary function. Similarly, we hypothesized that over-perception of airflow obstruction would be associated with worse asthma control and asthma-related quality of life, but not pulmonary function.

METHODS

Participants

The sample included 334 patients who were at least 60 years old with a diagnosis of asthma by a healthcare provider identified by medical chart review recruited from the outpatient practices of hospitals in the Bronx (N = 179) and East Harlem (N = 155), New York serving large numbers of low income and ethnic minority populations. Participants who spoke English or Spanish were eligible for the study. Exclusion criteria included: physician’s diagnosis of dementia; chronic obstructive pulmonary disease (COPD) or other chronic respiratory illness; smoking ≥ 15 pack-years owing to possible undiagnosed COPD; moderate or severe cardiac disease (ejection fraction < 35%); dependence on assistance for medication administration; and uncorrectable visual impairment. Participants were recruited by mailing letters from providers inviting study participation followed by phone calls from research staff.

The current manuscript is focused on data collected from a baseline visit followed by data collection at home for 6 weeks on perception of airflow obstruction. The larger study is examining perception and cognition in older adults across a one-year follow-up time period. Data were collected between January 2017 to January 2020. All participants provided written informed consent. The study was approved by the Institutional Review Boards at the Albert Einstein College of Medicine and Icahn School of Medicine at Mount Sinai.

Measures

Perception of airflow obstruction was calculated over 6 weeks at home following the baseline session. Each participant’s prediction of peak expiratory flow (PEF) was compared with actual PEF at the same time points. Participants entered their PEF guesses directly into a programmable peak flow monitor (AM2, eResearch Technology, Germany). Then, they performed three maximal effort PEF blows (the actual peak flow values did not appear on the device to keep participants blinded to prevent learning effects). During the baseline visits, research coordinators trained participants to perform maximal effort PEF blows. A colored sticker was attached to the peak flow monitor showing the participant’s predicted PEF values that correspond to the go (green), caution (yellow), and danger (red) zones of asthma control, as per national guidelines (36). The asthma risk grid was developed to quantify perceptual accuracy of airflow obstruction in asthma (37–41). Each objective PEF is plotted against the corresponding estimate of PEF to categorize each time point in the accurate, over-perception, or under-perception zones. Raw values of PEF prediction that are within 10% of actual raw PEF values are considered in the accurate zone. All other calculations are based on matches between the color zones using the percentages. For example, a PEF guess by a patient corresponding to 60% predicted value (yellow) with an actual PEF value of 90% personal best (green) would be categorized as over-perception. Over-perception of airflow obstruction is defined by under-estimation of PEF when conducting guesses. The percentage of the guesses in each zone are treated as a continuous variable, and thus, participants are not categorized as ‘under-perceivers’ or ‘over-perceivers.’ Data reduction steps eliminated extremely high or low PEF values. Twenty data points are needed to provide reliable estimates of symptom perception (39, 40). Validity for this methodology in children has been provided across several studies (40, 42, 43) showing greater perceptual accuracy of pulmonary function is associated with less asthma morbidity across a 1-year follow-up (37).

The Geriatric Depression Scale (GDS) is one of the most widely used self-report measures of depressive symptoms in the elderly (44, 45). The GDS-15 assesses depressive symptoms (e.g., feel helpless, satisfaction with life, energy, afraid something bad is going to happen) in the prior week. The GDS-15 is reliable (KR20 = .71) and has good sensitivity (90–92%) and specificity (81–88%) for detecting major depression (46–49). The GDS has been validated in English and Spanish (50), and among African Americans (51). Greater depressive symptoms are indicated by higher GDS scores. Scores on the GDS between 5 to 9 predict minor depression, and scores ≥ 10 predict major depression (49).

The asthma control questionnaire (ACQ-6) was used to assess self–reported asthma control during the past week (52). The ACQ-6 consists of 5 items about asthma symptoms (e.g., nighttime awakenings, activity limitations, shortness of breath) and 1 item on rescue medication use. It is available in English and Spanish and has good reliability (intraclass correlation coefficient = .90), internal consistency (α = .88-.92), and construct validity (53, 54). Higher scores on the ACQ indicate worse asthma control. Scores on the ACQ ≤ 0.75 predict well controlled asthma, and scores ≥ 1.5 predict poorly controlled asthma (55). The Mini Asthma Quality of Life Questionnaire (AQLQ) contains 15 questions that assess quality of life including activity limitations (e.g., physical, social, work–related activities), symptoms (e.g., coughing, chest tightness), emotional function (e.g., feel frustrated, concerned about asthma) and environmental stimuli (e.g., dust, cigarette smoke). The English and Spanish version of the Mini AQLQ have good reliability (intraclass correlation coefficient = .83), construct and criterion validity (56, 57). Higher scores represent better asthma-related quality of life and scores ≥ 4.7 have been shown to predict less acute asthma exacerbations and emergency healthcare use (58).

Pulmonary function testing was performed at the baseline session using a spirometer (CareFusion, USA). The testing was done as per the guidelines set by the American Thoracic Society and European Respiratory Society (59). The measure of pulmonary function analyzed for this study was the percent predicted forced expiratory volume in one second (FEV₁ % predicted), which is the volume of air that is exhaled during the first second of a forced vital capacity maneuver. The predicted value was based on norms for age, sex, race, and height.

Demographic variables were collected by self-report. Participants were asked if they were a current or former cigarette smoker, or never smoked. Medical chart reviews were conducted to calculate a medical comorbidity score based on the classification of Charlson et al. (60). This classification system assigns weights for each comorbid condition ranging from 1 (e.g., cerebrovascular disease) to 6 (e.g., metastatic solid tumor) and has been shown to predict mortality rates in longitudinal studies (60).

Procedures

The baseline session included administration of questionnaires and spirometry, which was followed by 6 weeks of PEF assessments at home preceded on each occasion by recorded perception of airflow obstruction. A follow-up phone call was conducted within the first few days to remind the participants about maximal effort PEF blows and reinforce coaching techniques.

Statistical Analyses

Version 9.4 of the SAS System for Windows (copyright 2017, SAS Institute Inc.) was used for all analyses. Descriptive statistics assessed the distributional characteristics of the data; means and standard deviations for continuous variables and proportions for categorical variables. Bivariate correlations were examined by Spearman’s rho. Linear regression models were conducted for continuous outcome measures (over–perception, ACQ, mini AQLQ, FEV₁ % predicted). Depressive symptoms were also treated as a continuous measure. Covariates entered into all models included age, sex, education, race/ethnicity, income, currently taking controller medication for asthma, and the Charlson comorbidity index.

RESULTS

A total of 552 participants were screened for eligibility and 218 participants were excluded from the study. The most common reasons for ineligibility were due to COPD (N = 65), smoking ≥ 15 pack-years (N = 38), other chronic respiratory illness (N = 36), and dependence on assistance for medication administration (N = 22). Descriptive characteristics of the final sample (N = 334) are presented in Table 1. The majority of participants were female and either Hispanic or Black, non-Hispanic. The mean age of participants was 67.9 years old (range: 60 – 95 years). Asthma control measured by pulmonary function during the baseline session (FEV₁ % predicted = 73.0 ± 19.2) indicated that, on average, asthma was not well controlled. Participants under-perceived 25% of the time and over-perceived airflow obstruction 13% of the time. The overall mean for actual PEF collected at home was 76.4% ± 9.7 and the overall mean for PEF estimates at home was 81.2% ± 18.7. No relationship existed between depressive symptoms and actual PEF (r_s = −.09) or estimates of PEF (r_s = −.10). Overall, 25% of participants reported symptoms suggestive or indicative of depression on the GDS. The ACQ revealed a pattern of poorly controlled asthma (M = 1.56 ± 1.08; skewness = .71; kurtosis = .30). Asthma–related quality of life was rated on the AQLQ (skewness = −.22; kurtosis = −.84) with a mean of 4.79 ± 1.26 on a 7-point Likert with higher scores indicating better quality of life.

Table 1.

Sample characteristics at baseline

Variables	N	% or Mean (SD), Range
Age, years	334	67.9 (6.7), 60–95
Sex, % Female	282	84.4
Education, %  Some high school  High school graduate  Some college  ≥ College graduate	82 62 76 109	24.9 18.8 23.1 33.1
Race/Ethnicity, %  White, non-Hispanic  Black, non-Hispanic  Other, non-Hispanic  Hispanic	52 83 28 171	15.6 24.9 8.9 51.2
Income, $1500/month or less, %	171	54.8
Language, % Spanish	94	28.1
Cigarette Smoking, %  Current smoker  Former smoker  Never smoked	18 110 197	5.5 33.9 60.6
Charlson comorbidity index	332	2.95 (1.43), 0 – 8.0
On controller medication, %	277	82.9
% FEV1	309	73.0 (19.2), 22.0 – 144.0
Asthma Control Questionnaire	334	1.56 (1.08), 0 – 5.33
Asthma Quality of Life Questionnaire	334	4.79 (1.26), 1.54 – 7.00
% Over-Perception	261	13.4 (21.1)
% Under-Perception	261	24.6 (26.3)
% Accurate Perception	261	62.4 (24.3)
No Depression (< 5 on GDS), % Suggestive of Depression (5–9) Indicative of Depression (≥10)	252 65 17	75.4 19.5 5.1

GDS = Geriatric Depression Scale

Perception data were available for 261 participants out of 334 (Number of data points: M = 65.9 ± 25.3). Reasons for missing data included having less than 20 data points (N = 35), device issues (N = 13; e.g., device failure), never used the device (N = 17), and lost to follow-up (N = 8). Participants with missing data on perception (N = 73) were more likely to have lower levels of education (p < .001) and income (p < .001), and more likely to be Hispanic (p < .001) and speak Spanish (p < .001) in comparison to participants who completed perception data (N = 261). Participants who failed to provide perception data also had worse asthma control (M = 1.92 ± 1.09 vs M = 1.46 ± 1.06, p < .001) and lower asthma-related quality of life (M = 4.33 ± 1.26 vs M = 4.92 ± 1.24, p < .001) in comparison to participants with perception data. No differences were found between completers and non-completers of perception data on age, sex, being on a controller medication, %FEV₁, or depressive symptoms. Table 2 presents the correlations among the continuous measures showing that depressive symptoms and over- perception were both associated with worse asthma control and asthma–related quality of life. Over-perception was not associated with age, sex, education, race/ethnicity, income, language, cigarette smoking, taking a controller medication, or medical comorbidity.

Table 2.

Correlations between continuous study variables

	1.	2.	3.	4.	5.	6.
1. Geriatric Depression Scale
2. Over-perception	0.11
3. Asthma Control	0.24***	0.25***
4. Asthma-Related Quality of Life	−0.39***	−0.21***	−0.74***
5. FEV1 % Predicted	0.01	−0.03	−0.21***	0.17**
6. Age	−0.06	−0.09	−0.10	0.15**	0.03
7. Charlson comorbidity index	0.05	0.03	0.06	−0.06	0.14*	0.25***

Results reflect bivariate correlations.

^**p < .01

^***p < .001

Table 3 shows the associations of depressive symptoms with asthma outcomes. Higher levels of depressive symptoms were associated with greater over-perception of airflow obstruction in both the unadjusted (β = .13, p = .034) and adjusted (β = .14, p = .029) model. Greater depression was associated with worse asthma control (ACQ) in both the unadjusted (β = .24, p <.001) and adjusted model (β = .17, p = .003). Participants with more depressive symptoms reported worse asthma–related quality of life (mini AQLQ) in the unadjusted (β = −.41, p < .001) and adjusted (β = −.33, p < .001) models. Despite worse self-reports of asthma control, depressive symptoms were not associated with FEV₁ % predicted at the baseline session in the unadjusted (β = −.04, p =.44) or adjusted (β = −.01, p =.82) models.

Table 3.

Associations of depressive symptoms with over-perception of airflow obstruction, asthma control, asthma-related quality of life, and FEV₁

Variable	Over-perception						Asthma Control						Asthma-Related Quality of Life						FEV1 % Predicted
	Unadjusted Model			Adjusted Model			Unadjusted Model			Adjusted Model			Unadjusted Model			Adjusted Model			Unadjusted Model			Adjusted Model
	β	t	p	β	t	p	β	t	p	β	t	p	β	t	p	β	t	p	β	t	p	β	t	p
Geriatric Depression Scale	.13	2.13	.034	.14	2.19	.029	.24	4.51	<.001	.17	3.03	.003	−.41	−8.12	<.001	−.33	−6.35	<.001	−.04	−0.77	.44	−.01	−0.22	.82
Age				−.02	−0.38	.70				−.09	−1.68	.094				.10	2.04	.042				.02	0.32	.75
Sex				.01	0.17	.87				−.02	−0.39	.70				−.05	−1.08	.28				.06	1.08	.28
Education  Some high school  High school graduate  Some college  College or higher				Ref  −.07  −.01  −.08	−0.92 −0.07 −0.81	.36 .94 .42				Ref  −.04  −.09  −.10	−0.67 −1.26 −1.23	.50 .21 .22				Ref  .07  .05  .14	1.14 0.79 1.86	.25 .43 .063				Ref  .05  .08  .16	0.69 1.04 1.86	.49 .30 .065
Race/ethnicity  White/other, non-Hispanic  Black, non-Hispanic  Hispanic				Ref  −.02  −.12	−0.24 −1.48	.81 .14				Ref  .07  .10	1.03 1.34	.30 .18				Ref −.09 −.21	−1.45 −3.04	.15 .003				Ref  −.05  −.07	−0.61 −0.80	.54 .42
Income				−.01	−0.14	.89				.06	0.96	.34				−.06	−1.09	.28				.01	0.01	.99
On Controller Medication				.01	0.08	.94				.03	0.52	.60				−.04	−0.72	.47				−.01	−0.18	.85
Charlson Comorbidity Index				.07	1.18	.24				.05	0.97	.33				−.04	−0.86	.39				.12	2.16	.032

Results are from linear regression analyses. Standardized beta values are reported.

We also examined the relationship between over–perception of airflow obstruction and asthma morbidity given that this is the first study to use the perception methodology in older adults. Table 4 shows that over–perception was associated with worse asthma control (ACQ) in unadjusted (β = .14, p =.026) and adjusted (β = .14, p =.021) models. A trend was found for over–perception and worse asthma–related quality of life in unadjusted (β = −.12, p = .063) and adjusted (β = −.11, p = .058) models. As expected, over-perception was not related to pulmonary function at baseline in unadjusted (β = −.03, p = .61) or adjusted (β = −.05, p = .41) models.

Table 4.

Associations of over-perception of airflow obstruction with asthma control, asthma-related quality of life, and FEV₁

Variable	Asthma Control						Asthma-Related Quality of Life						FEV1 % Predicted
	Unadjusted Model			Adjusted Model			Unadjusted Model			Adjusted Model			Unadjusted Model			Adjusted Model
	β	t	p	β	t	p	β	t	p	β	t	p	β	t	p	β	t	p
Over-perception	.14	2.24	.026	.14	2.33	.021	−.12	−1.87	.063	−.11	−1.91	.058	−.03	−0.51	.61	−.05	−0.82	.41
Age				−.11	−1.78	.076				.13	2.14	.034				.02	0.23	.82
Sex				−.02	−0.32	.75				−.04	−0.70	.48				.02	0.26	.79
Education  Some high school  High school graduate  Some college  College or higher				Ref  −.04  −.10  −.09	−0.57 −1.12 −0.93	.57 .26 .35				Ref  .05  .06  .19	0.72 0.75 1.99	.47 .45 .047				Ref  .01  .05  .13	0.16 0.59 1.25	.87 .55 .21
Race/ethnicity  White/other, non-Hispanic  Black, non-Hispanic  Hispanic				Ref  .02  .10	0.31 1.23	.76 .22				Ref  −.04  −.19	−0.50 −2.43	.62 .016				Ref  −.04  −.06	−0.51 −0.64	.61 .52
Income				.09	1.33	.18				−.13	−2.03	.044				−.05	−0.67	.51
On Controller Medication				−.04	−0.65	.52				.01	0.22	.83				.01	0.10	.92
Charlson Comorbidity Index				.01	0.20	.84				−.04	−0.60	.55				.17	2.49	.014

Results are from linear regression analyses. Standardized beta values are reported.

DISCUSSION

Depressive symptoms in older adults with asthma were associated with over-perception of airflow obstruction (PEF) at home and self–reported worse asthma control and lower asthma–related quality of life. However, depressive symptoms were not associated with pulmonary function collected by spirometry at the baseline session (% FEV₁ predicted). Similarly, over-perception, more pronounced among patients with depressive symptoms, was only associated with self-reports of asthma symptoms. These findings support the hypothesis that depressive symptoms are linked with greater perceived impairment from asthma when compared with measures of airflow obstruction. This is consistent with prior findings examining respiratory function at a single time point in a clinical setting (11, 26). This study presents a unique contribution to the older adult literature because it utilizes methodology which compares estimates of PEF to actual PEF in a naturalistic setting. This study suggests that the influence of depressive symptoms and emotional state on asthma symptoms also exists in ethnic minority older adults, a group that is particularly vulnerable to depression. Our findings demonstrate the role of asthma symptom perception in the relationship between depression and asthma outcomes, and the importance of taking into account multiple determinants of asthma outcomes including emotional factors.

There are several clinical implications for over–perception of airflow obstruction that may affect patient care and self-management decisions. Over-perception has been linked to overuse of quick-relief medications and iatrogenic side effects (e.g., tachycardia, tremors), excessive avoidance of activities, overuse of health care resources, and worse physician ratings of asthma control (11, 61). Providers should consider the possibility of depression in patients who consistently report high levels of asthma symptoms despite normal pulmonary function. Patients’ asthma symptom perception can influence their frequency of rescue medication use and when to utilize healthcare resources. Psychological distress and general somatic symptoms in older adults with depression may be misinterpreted as symptoms of asthma (e.g., fatigue, self-perception as sick) and influence healthcare seeking decisions. Moreover, self-reports of asthma symptoms are used by healthcare providers to determine the need for and to titrate the dose of controller medications. Such clinical implications may be addressed using mindfulness-based cognitive therapy (62) or cognitive behavior therapy (63), which have both shown some promise as potential interventions for over-perception of airflow obstruction. These interventions target the affective and sensory components of dyspnea and may promote self-management behaviors. Older adults may also benefit by receiving feedback on their estimates of PEF, which has been shown to improve perception of airflow obstruction in children (42). Asthma self-management programs can improve quality of life (64) and thus, a comprehensive program focused on trigger avoidance, accuracy of asthma symptom perception, medication management, and coping with stress may all be important to address for patients with asthma and depression. It is important that clinicians not minimize or disregard self-reported asthma control or quality-of-life ratings given that these ratings may reflect genuine distress or actual asthma exacerbations that are occurring at times separate from measurement of normal pulmonary function.

The effect of negative emotions on the perception of breathlessness (i.e., dyspnea) is more pronounced on the affective component (unpleasantness) of dyspnea than the sensory (intensity) component (65). Depressed patients are more likely to manage their emotional distress with catastrophizing and other emotion-focused coping strategies (66) at the expense of problem–focused coping and adaptive self-management behaviors (67). These copings strategies commonly used by depressed patients reinforce a negative feedback loop wherein patients maintain inflexible affective processing, which inhibits their adaptive asthma management behaviors (32). Such inflexible affective processing can be understood as a maladaptive lens through which individuals anticipate a negative asthma experience and interpret potentially innocuous bodily sensations as an asthma flare-up. Moreover, catastrophic beliefs and activity avoidance are linked to poorly perceived asthma control (28). Optimal medical management of asthma is associated over time with improvements in both depression and asthma control (23).

The present study also provides preliminary evidence for construct validity of measurement of over-perception using PEF prediction in older adults. Over-perception of airflow obstruction was associated with poorer self–reported asthma control and may also be related to lower asthma-related quality of life. The pessimistic discrepancy between self-reported measures of asthma control versus actual airflow obstruction in the present study existed for both standard measures of asthma control (ACQ) and pulmonary function (spirometry) assessed in a clinic setting, and also for estimation and measurement of PEF at home. These data support the importance of integrating measures of pulmonary function into assessment of asthma control in older adults, particularly patients with depressive symptoms. Given time constraints and lack of spirometry equipment in primary care settings, the ease of access to peak flow meters at home might be an ideal way to incorporate assessments of airflow obstruction into asthma self–management for individuals with inaccurate symptom perception. Although our participants were blinded to PEF for the purposes of the study, guessing PEF and seeing how close estimates are to actual PEF can be more interactive and engaging for patients than traditional home peak flow monitoring (42).

Limitations of the study include the cross-sectional nature of the data; future analyses should examine longitudinal relationships between these variables. Another significant limitation is the missing data on the electronic devices that measured perception. Participants with worse asthma control and lower asthma–related quality of life were less likely to provide perception data. Participants who were Hispanic and Spanish-speaking with less education and income were also more likely to have missing data. Similarly, a pediatric study of asthma perception using the same methodology showed a lower completion rate in Hispanic children versus non-Hispanic white children (43). Nevertheless, the rate of completion of perception data in this study (78.1%) compares favorably to pediatric studies that have used this methodology (range: 65.1–78.6%) (42, 43). Participants who are less compliant with guessing PEF and using their peak flow meter at home may be at even greater risk for inaccurate perception. Finally, PEF is highly effort dependent and it is possible that patients with more depressive symptoms might exert less effort and lead to inaccurate, lower readings versus a skillfully coached spirometry test. We attempted to mitigate this concern with a reminder phone call. We found no relationship between depressive symptoms and home collected actual PEF values.

CONCLUSIONS

In summary, this study shows that depressive symptoms in older, minority adults are linked with a pattern of greater impairment due to asthma and over-perception of airflow obstruction. This is the first study to demonstrate this finding by measuring perception of pulmonary function in a naturalistic setting in older adults. Decision-making by providers for asthma treatment often relies on self-report of asthma symptoms, which may be influenced by psychological factors including depression. These negative emotions in older adults may create a cycle of excessive restriction of activities and exercise, which could further exacerbate depression. Interventions are needed to target asthma and depression comorbidity with a focus on improving perception of pulmonary function and health-related quality of life.

Acronyms:

ACQ

Asthma Control Questionnaire

AQLQ

Asthma Quality of Life Questionnaire

COPD

chronic obstructive pulmonary disease

GDS

Geriatric Depression Scale

FEV₁

forced expiratory volume in 1 second

PEF

peak expiratory flow