INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthmatic patients often suffer from nocturnal symptoms, varying from cough to severe asthma attacks. Many of them have asthma arousals at least one night a week and often more frequently (1, 2). Fitzpatrick and coworkers (3) have reported that asthmatics with inhaled corticosteroids as maintenance treatment still may have nocturnal airway obstruction. With the introduction of new drugs such as long-acting ₂-agonist and a new inhaled glucocorticosteroid fluticasone propionate (FP), good effects have been reported on diurnal variation in peak expiratory flow (PEF) (3, 4). Salmeterol xinafoate (SLM) reduces both PEF variations and nocturnal symptoms (5). Greening and coworkers (12) compared twice the maintenance dose of inhaled corticosteroids with adding SLM to the maintenance treatment of inhaled corticosteroids. PEF variation was reduced to a similar extent, but doubling the inhaled corticosteroid dose seemed to be less effective with regard to nocturnal symptoms.

It has been shown that the occurrence of nocturnal symptoms, especially awakening, is associated with impaired daytime cognitive performance as tested by a battery of psychometric test (3). These nocturnal symptoms seem to be due to nocturnal airway obstruction and/or nocturnal increase in bronchial hyperresponsiveness (BHR). Therefore, we hypothesized that prevention of nocturnal airway obstruction and/or nocturnal increase in BHR might improve daytime cognitive performance. As it is yet unclear which treatment is the best one for reducing nocturnal airway obstruction, we designed a study that compares three different treatment approaches, i.e., using SLM or FP or their combination. In the present study, we investigated the therapeutic effects of SLM, FP, and their combination on the reduction in circadian PEF variation and the increase in FEV₁ and in BHR to methacholine (MCh) in relation to these effects on daytime cognitive performance. A battery of psychometric tests was used to assess this performance; the results were compared with the results in healthy volunteers.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Design

The study was performed in a randomized, double-blind, double-placebo, parallel manner. One week prior to the study, for 3 consecutive days, PEF values were recorded at home with mini-Wright peak flow meter at 4:00 A.M., 8:00 A.M., and 4:00 P.M. No inhaled short-acting ₂-agonists were allowed 4 h before the PEF measurements during these 3 d and 12 h before every test in our laboratory.

At the first study visit a battery of psychometric tests in a fixed order were performed starting at 8:00 A.M. followed by lung function and a MCh challenge. In the afternoon the lung function and MCh challenges were repeated at 4:00 P.M. and during the consecutive night at 4:00 A.M. Subsequently, subjects were randomly allocated to one of the three double-blind therapy protocols for a period of 6 wk. One group was treated with SLM 50 µg twice a day, one with FP 250 µg twice a day, and one with both drugs in the same dose twice daily. The three drugs were administered by a multidose dry-powder delivery system (Diskhaler; Glaxo, Middlesex, UK). All subjects were instructed to take their study medication at 8:00 A.M. and 8:00 P.M. Short-acting ₂-agonists were allowed during these 6 wk for relief of asthma symptoms. In the last week of the 6-wk treatment period, PEF values were recorded again at home at 4:00 A.M., 8:00 A.M., and 4:00 P.M. for 3 consecutive days, and no short-acting ₂-agonists were allowed 4 h before the measurements. The visit after the treatment period had the same design as the first study visit. The use of study medication was continued during study Visit 2. The healthy volunteers came to the hospital at 8:00 A.M. twice within a 6-wk interval.

Subject Characteristics

Nonsmoking, atopic asthmatic subjects 18 to 45 yr of age were selected according to the following criteria: a history of episodic dyspnea or wheezing consistent with the clinical diagnosis of asthma and no concomitant diseases, bronchial hyperresponsiveness to methacholine (PC₂₀ < 9.6 mg/ml), elevated specific immunoglobulin (IgE) against house dust mite (RAST  2), or positive intracutaneous tests against house dust mite or two other common aeroallergens (ALK, Benelux; Groningen, The Netherlands). Subjects were not allowed to enter the study when using oral corticosteroids or when they had a respiratory tract infection or acute asthma attack during the 2 mo prior to the study. Inhaled corticosteroids, if used, were stopped 4 wk prior to the study, whereas nedocromil sodium and long-acting ₂-agonists were discontinued for 2 wk. Subjects were recruited from our outpatient clinic and from advertisements in local newspapers, and they participated in the study after giving written informed consent. The study was approved by the hospital medical ethics committee. Sixteen healthy volunteers served as the control group. All were nonsmokers, were receiving no medication, and had no history of any respiratory illness. The control group was matched for age, sex, and education level.

Lung Function Tests

Bronchial hyperresponsiveness was assessed with a methacholine bromide challenge (MCh) as described previously (13). Briefly, after measuring a baseline FEV₁ with a dry rolling-seal spirometer (Vica-test 5; Jaeger BV, Breda, The Netherlands), a MCh provocation test was measured using the 2-min tidal breathing method adapted from Cockcroft and coworkers (14). The challenge was started with a saline inhalation followed by inhalation of doubling concentrations (DC) of MCh (0.038 to 19.6 mg/ml), which were inhaled at 5-min intervals, Single FEV₁ values were measured 30 and 90 s after each inhalation of MCh. The inhalations were stopped as soon as the fall in FEV₁ was 20% or more from baseline FEV₁ or when the upper dose for MCh (19.6 mg/ml) was reached.

Psychometric Tests

The test lasted about 45 min and consisted of the following: Stroop Color Word test, Trail making A and B, and the PASAT test. The tests were performed in this fixed order.

The Stroop Color Word test is a test of focused attention that makes use of response interference. The task consist of three subtests. The first one presents the subject with 100 simple words, i.e., the color names yellow, red, green, and blue printed in black. Subjects have to read aloud these 100 words as quickly as possible. The next one presents 100 colored blocks (again yellow, red, green, and blue) and the subjects have to name these colors as quickly as possible. Finally, in the third subtest subjects look at 100 color names that have been printed in conflicting colors, e.g., the word "green" in red letters. Again, they name the colors of the 100 stimulus words. However, because reading is completely automatized in adults, the word meaning creates a strong interference effect. Thus, considerable mental effort and concentration are necessary to overcome this interference. As a result, subjects perform much slower on the interference task than on the simple color naming task in the second subtest, eventhough their verbal production is identical in both tasks (15, 16).

The Trail making A and B from the Halstead-Reitan battery is a visual search task requiring hand-eye coordination and mental flexibility (17). Trail A is a test in which a series of numbers has to be connected with a pencil line in consecutive order (e.g., 1-2-3, etc.) as fast as possible, measured in seconds. Trail B is a test with a series of alternate numbers and letters to be connected in consecutive order (e.g., A-1-B-2, etc.) as fast as possible. This test requires flexibility as alphabetic letters and numbers have to be used alternately.

The PASAT test (paced auditory serial addition test) is a test of concentration and attention (18). The subjects hear from a tape recorder a series of 61 digits. Each time the last two digits have to be added. The optimal score is 60 correct responses. The time interval between the digits is decreased over the series, the interstimulus intervals being 3.2, 2.4, and 1.6 s. Thus, by presenting a long series of stimuli under time pressure the PASAT test give information about processing capacity.

Data Analysis

Results are presented as mean values with standard errors of the mean (SEM), unless stated otherwise. FEV₁ values were expressed as percent, predicted. The circadian variation in PEF was calculated as: 100*(highest - lowest 24 h value)/mean 24-h value per day. The mean circadian variation was then calculated over 3 d.

Skewedness of distribution was assessed using Kolmogorov-Smir-nov tests. Normalization of distribution was reached for PC₂₀ MCh and circadian variation of PEF by logarithmic transformation. Differences between healthy control subjects and asthmatics were analyzed by using Student's t test. Multiple regression analyses (stepwise) were used to assess the importance of possible contributing factors to explain lower scores of the psychometric tests in the asthmatic group. The regression equation was built up with each of the psychometric tests as the dependent variable and the lung function parameters as the independent variables (FEV₁ day and night, PC₂₀ MCh day and night, and circadian variation in PEF). Two-sided p values < 0.05 were considered statistically significant. The PC₂₀ was calculated by linear interpolation of the last two points of the log dose-response curve. In all three treatment groups some asthmatics reacted on saline inhalation with a 20% fall or more in FEV₁. As a consequence MCh was not administered to these asthmatic, and for analytic purposes the PC₂₀ was considered to be half of the lowest dose of MCh (0.015). All analyses were carried out with the SPSS/PC⁺ V4.0 statistical package (SPSS, Inc., Chicago, IL).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A total of 50 atopic asthmatic participated in the study of whom four dropped out: three because of an asthma exacerbation requiring oral corticosteroids and one because of a surgical procedure unrelated to the study. The three dropouts with an exacerbation each came from a different treatment group, the fourth was allocated to the combination treatment group. These four subjects were excluded for statistical analysis. A total of 46 subjects completed the study: 16 in the SLM group, 16 in the FP group, and 14 in the combination group. Subject characteristics are summarized in Table 1. The three treatment groups were comparable to each other for all baseline values measured at Day 1 of the first study visit at 8:00 A.M. (p < 0.05).

The circadian variation in PEF was 22.9% (geometric mean; range, 15 to 59%) as measured at home. The day/night variation in FEV₁ was 15.9 ± 2.3% as measured at the hospital. A significant decline of 1.6 ± 0.2 doubling concentrations was seen in PC₂₀ MCh overnight. The more extended clinical results are published elsewhere (19).

Treatment Effects on Lung Function and BHR

The three treatment groups were comparable to each other for all baseline values measured at study visit 1 at 8:00 A.M. No significant differences in therapeutic effect were observed between the three treatment protocols with respect to FEV₁ expressed as percent predicted. FEV₁ was above 90% predicted both day and night in all three groups after 6 wk of treatment. The three treatment arms were equally effective to reduce circadian PEF variation below 10% (Figure 1). The circadian PEF variation decreased from 22.4% (geometric mean; range, 15 to 49) to 9.1% (3 to 21; p < 0.001) for SLM, from 24.0% (16 to 53) to 7.9% (3 to 23; p < 0.0001) for FP, and from 24.0% (15 to 58) to 10% (4 to 31; p < 0.0001) for their combination.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Effect of the three treatment modalities on the circadian variation in PEF. No significance in therapy was observed.

We found no significant differences in therapeutic effect on PC₂₀ MCh both day and night between the three treatment arms. PC₂₀ MCh values improved significantly within each treatment group, 1.5 ± 0.5 DC (p = 0.02) at 4:00 P.M. for SLM, 2.1 ± 0.5 DC (p = 0.001) for FP, and 2.5 ± 0.6 DC (p = 0.003) for their combination. At night (4:00 A.M.), improvement was 2.4 ± 0.5 DC (p < 0.001) for SLM, 3.0 ± 0.5 DC (p = < 0.001) for FP, 2.9 ± 0.8 DC (p = 0.003) for their combination.

Psychometric Tests

Before treatment. The PASAT score was significantly lower in asthmatics than in control subjects at all three time intervals (Table 2). For the color-word-chart subtest, asthmatics needed significantly more time to finish reading the chart than did control subjects, 83.2 ± 2.4 versus 75.3 ± 2.8 s (p = 0.04). No difference was observed between asthmatics and healthy control subjects for the word-chart subtest and color-chart subtest. The asthmatics did not differ significantly from control subjects in completing Trail making A and B. 

The circadian variation in PEF was the only independent factor significantly associated with performance of the word-charge, color-chart, and all three PASAT tests. After dividing the asthmatic group in a group with a circadian variation < 20% and a circadian variation  20%, the association was much stronger for the high variation group and abolished for the low one (Figures 2 and 3). Those with a large PEF variation tended to perform slower on color naming and word-color naming (r = 0.61, r = 0.62, respectively). For all three PASAT time intervals, a lower score was associated with a larger PEF variation (3.2 s, r = 0.52; 2.4 s, r = 0.66; 1.6 s, r = 0.55). The other lung function parameters (FEV₁ at day and night, PC₂₀ MCh at day and night) were not associated with a lower score on the psychometric tests.

View larger version (13K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 2.   Association between circadian PEF variation and the Stroop Color-word test (top panel: word-color; bottom panel: color-color). For the whole group of asthmatics a significant association was observed, implying that asthmatics with lower circadian variations need less time to finish the Stroop Color Word test. This association disappeared when the circadian PEF variation was less than 20% and became stronger with a variation > 20%. The latter may have clinical importance for asthmatics.

View larger version (13K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 3.   Association between circadian PEF variation and the PASAT test (top panel: PASAT, 3.2 s; bottom panel: PASAT, 2.4 s; above: PASAT, 1.6 s). For the whole group of asthmatics a significant association was observed, implying that asthmatics with lower circadian variations give more correct answers in the PASAT test for all three time intervals (3.2, 2.4, 1.6 s). This association disappeared when the circadian PEF variation was less than 20% and became stronger with a variation > 20%. The latter may have clinical importance for asthmatics.

After treatment. As mentioned above the three treatment protocols were equally effective in improving all lung function parameters. A significant improvement was seen in the score on the psychometric tests for the three treatment groups, with no difference between the treatments. For the healthy control subjects, a significant test-retest effect was observed after 6 wk, except for Trail making A and B (Table 2). The test-retest effect did not differ significantly between the asthmatics and the healthy controls subjects for the psychometric tests. However, the end-score of the color-word chart, and of all three PASAT time intervals improved to the levels of the healthy control subjects.

The circadian variation in PEF was no longer significantly associated with performance of the word-chart, color-chart, and the three PASAT time intervals after 6 wk of treatment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study has shown that asthmatic subjects with mild to moderate airway obstruction, but with severe BHR to MCh, score lower on psychometric tests than do healthy control subjects, suggesting a lower daytime cognitive performance in the asthmatic group. The only independent factor associated with a lower score on some of the psychometric test was a circadian PEF variation  20%. After 6 wk of therapy, FP, SLM, and their combination were equally effective in reducing circadian PEF variation and improving FEV₁ both day and night and bronchial hyperresponsiveness both day and night. As asthmatics with moderate to severe asthma were included in this study there may be discriminatory differences between the three treatment groups. When the FEV₁ for the group as a whole is nearly 80% of predicted normal, then it would be difficult to discriminate between these treatment groups. Within either of the three groups only a small improvement was necessary to place the group into normal range for airflow. However, a severe BHR to MCh was assessed in all three groups, and even an increase in this end parameter did not discriminate between the three groups.

The psychometric tests we used in this study were informative about focused attention (the Stroop Color-Word test) (16), hand-eye coordination and mental flexibility (the Trail making test) (17), and concentration and attention (the PASAT test) (18). Detecting small differences in cognitive ability is difficult because most people can call on reserve capacity, even when fatigued. The Trail making test did not separate asthmatic subjects from healthy control subjects. In contrast, the PASAT test and one of the Stroop Color-Word tests, the word-color chart, showed that asthmatics had a lower score on these tests than did healthy control subjects, demanding a high level of concentration. These discriminating tests might implicate that daytime cognitive performance, specified by concentration and attention, was impaired in the asthmatic group before treatment was started. It cannot be explained by age, sex, and premorbid intelligence, known to influence performance of psychometric tests (17, 18). We matched the control group for age, sex, and premorbid intelligence by education level.

Our study has shown that this was associated with a circadian PEF variation > 20%. It is still under discussion whether nocturnal airway obstruction has to be defined as a circadian variation in PEF  15% or  20%. For this reason, we calculated for both if an association could be considered. The circadian PEF variation was assessed at the time points 4:00 A.M. or 8:00 A.M. to measure the greatest airflow limitation and at 4:00 P.M. to measure the smallest airflow limitation. Thus, the circadian PEF variation was calculated most carefully.

The study of Fitzpatrick and coworkers (3) concluded that asthmatics have a poorer daytime cognitive performance than do control subjects. This group of asthmatics still suffered from nocturnal symptoms and awakenings, and all of them had a documented overnight fall in PEF. This finding is in accordance with our study results. The explanation for their observed deficits in daytime cognitive performance in asthmatic subjects include circadian variation in PEF only. Circadian PEF variation is the strongest factor influencing time spent awake at night in asthmatics (3). It is known that sleep deprivation, and in particular REM sleep, leads to demonstrable impairment of daytime cognitive performance (20, 21). Because no assessments of either quality of sleep, the length of sleep, or the number of interruptions of sleep were incorporated, further studies are necessary to confirm this. Both healthy control subjects and asthmatics improved their test score significantly after 6 wk by doing each test twice, a learning process so-called the test-retest effect. Although the test-retest effect did not differ significantly between the asthma and control groups, the end-score of the asthmatic subjects became comparable to the end-scores of the healthy control subjects. The lack of significance of the test-retest effect between the control subjects and the asthmatics may be explained by the small group of asthmatics we entered into the study.

Drug treatment may also affect sleep quality. Steward and coworkers (22) observed that administration of an oral ₂-agonist to asthmatics did not improve sleep quality despite reduction in overnight bronchoconstriction. They explained their findings by the known side effects of oral ₂-agoists such as agitation. In our study, however, there was no suggestion of these side effects by salmeterol as monotherapy for 6 wk, most likely because inhalation of ₂-agonists will lead to less systemic side effects than will oral ones.

In conclusion, circadian PEF variation  20% in asthmatics was associated with a poorer daytime cognitive performance when compared with healthy control subjects. After 6-wk of three modalities of treatment, the asthmatics had a level of daytime cognitive performance comparable to that of the healthy control subjects, which was parallelled by a reduction of circadian PEF variation below 10%.

Our study suggests, therefore, that it may be of great importance to reduce circadian PEF variation below 10% by antiasthmatic drugs because a high PEF variability has a great impact on daytime cognitive performance in asthmatic subjects.