Daily mood, shortness of breath, and lung function in asthma: Concurrent and prospective associations

Abstract

Objective

Previous asthma diary studies have yielded equivocal findings on concurrent associations between lung function and mood, and prospective associations have rarely been explored. We therefore examined concurrent and prospective associations between daily mood, shortness of breath, and lung function, and studied between-individual variability and stability of concurrent associations across different times of the day.

Method

Twenty asthma patients and 20 healthy controls recorded their positive and negative mood, shortness of breath, physical activity, peak expiratory flow (PEF) and forced expiratory volume in the first second (FEV₁) using an electronic pocket spirometer with diary functions three times per day for about 21 days.

Results

For both groups, PEF showed positive concurrent associations with ratings of various mood states, whereas FEV₁ was only associated with positive mood. Both indices correlated negatively with shortness of breath. Within-individual concurrent associations varied significantly in both groups and their stability varied across time of the day, with overall higher stability for associations with shortness of breath in asthma and PEF for both groups. Prospectively, higher shortness of breath consistently predicted lower lung function later during the day and on the subsequent day.

Conclusion

The relationship between normal mood variations and lung function is highly variable across individuals and times of the day, limiting the predictive value of average group associations. Shortness of breath is predictive of future lung function decline in asthma. Future longitudinal research should focus on extreme emotional states, effort-independent measures of lung function, and additional indicators of asthma control.

Introduction

Self-monitoring of lung function on a daily basis has been recommended for asthma patients to improve the management of their chronic condition [1], [2]. Peak expiratory flow (PEF) diaries have been used frequently in longitudinal studies of lung function changes and perceived symptoms (e.g., Refs. [3], [4], [5], [6], [7]). Some studies have also explored the relationship between daily mood and lung function changes [6], [7], [8], [9], [10], [11], [12], [13], [14]. Evidence from most of these studies suggests that associations between mood, symptoms and lung function are only substantial in some patients and can be highly idiosyncratic. However, little is known about the stability of such idiosyncratic associations in daily life. One study with a small sample of asthma patients found different correlations between mood and PEF for different times of the day in the same individuals [9]. Because lung function is dependent on the circadian rhythm [5], [15], it could be differentially sensitive to psychosocial influences across the day. More specific information about associations of symptoms and mood states with airway obstructions could eventually assist patients' self-care. Individually salient symptomatic or emotional states could function as early warning signs to adjust medication intake or seek professional help.

Most lung function and mood diary studies have explored concurrent associations (analysis of the associations without time lag). Thus, little is known about the prospective association between these variables: are mood and symptom changes associated with subsequent lung function changes or do lung function changes precede mood and symptom changes? For example, it could be speculated that patients report more symptoms and negative mood after becoming aware of a lung function decline [8], [16]. Three earlier studies have used adequate methods to explore the time-sequence between ratings of stress, negative affect, and PEF [7], [8], [11], but the evidence from these studies is inconclusive. In one preliminary study [8], PEF did not significantly predict mood at the next of three daily self-assessments points, but pleasant mood was found to be related to higher PEF. A further analysis with a larger asthma patient sample did not yield any significant prospective association in either direction for daily mood levels and PEF [7]. A third study found that prospective associations for negative mood states or stress on day one with PEF on day two were highly idiosyncratic in direction and strength [11]. The reverse association, lung function predicting psychological variables, was not tested. Thus, the limited evidence has not supported a somatopsychological hypothesis of lung function affecting mood, but some there is some indication for the validity of a psychosomatic hypothesis of mood affecting lung function [16].

In planning the current study, we sought to address a number of methodological issues. First, the majority of PEF diary studies had employed paper and pencil methods for data collection (for exceptions, Refs. [7], [8], [14]), which may have affected the validity of the findings [17]. Without control over the time of sampling, diaries could include a certain proportion of mistimed or even invented entries [18], [19]. A particular problem with invented values is that diaries of mood and lung function could simply reflect the patients' lay theories about correlations between symptoms, mood and airway obstruction. In the present study, we employed an electronic pocket spirometer combined with diary functions to insure that the collection of self report data corresponded in time with the assessments of lung function. Second, whereas most previous studies have only measured PEF, the electronic spirometer used in the present study was able to determine forced expiratory volume in the first second (FEV₁), which is generally seen as less effort-dependent than PEF and more sensitive to smaller changes in airway obstruction [20]. By comparing both indices we sought to examine the degree to which the potential mood-lung function relationship could be replicated with a more sensitive and less effort-dependent measure. Third, an adequate assessment of mood states should reflect the possibility that emotional states of both positive and negative valence are capable of compromising lung function, as suggested by patients' self-report and experimental studies (for reviews see Refs. [21], [22]). An exclusive focus on negative emotionality or distress as shown by earlier diary studies [6], [16] would not reflect the variety of potential emotional influences on the airways. In addition, designing mood ratings both along a categorical and dimensional model of emotions is more informative, because prior studies have focused mostly on either mood categories (such as global positive and negative mood, happiness, sadness, anxiety, or anger [8], [11], [12]) or bipolar dimensional models (such as positive vs. negative affect, pleasantness, or arousal [7], [9], [13]). A dimensional perspective may be particularly suitable in studying associations of the respiratory system activation with emotion or mood [23], and ratings of arousal have been shown to covary with PEF recordings in asthma patients [7], [8].

In the current study, we combined these methodological improvements with a comparison of asthma patients and healthy controls, using Hierarchical Linear Modeling (HLM) to accommodate the correlated responses inherent in a longitudinal repeated measures design. In an earlier analysis of the current data set, we had focused on the association of extreme positive and negative mood states with lung function and its relationship with respiratory resistance increases in a laboratory emotion-induction task [24]. Because the laboratory assessment was scheduled in the afternoon, we had only used afternoon data from the lung function diaries. The current analyses use the full data set of self-assessments three times a day, morning, afternoon, and evening/night, and studied the dependency of the mood, symptoms, and lung function associations on time of the day. Because we expected substantial between-individual variation of the relationship between mood and lung function in both asthmatic and nonasthmatic participants, one major aim was to study the stability of the individual lung function–mood relations throughout the day and across various patient characteristics, such as age, gender, PEF variability (as an indicator of asthma severity [1], [25]), and susceptibility to psychological asthma triggers.

In summary, we planned to study (i) concurrent mood–lung function associations in their dependency on time of the day, choice of spirometric index, positive vs. negative states, dimensional vs. categorical ratings, and asthmatic vs. healthy status; (ii) the stability of within-individual relations in such associations across times of the day; (iii) moderators of individual differences in mood-lung function associations; and (iv) prospective associations between symptoms, mood and lung function.