Overnight Theophylline Concentrations and Effects on Sleep and Lung Function in Chronic Obstructive Pulmonary Disease 1- 3

RICHARD J. MARTIN and JUNO PAK With the technical assistance of Eliza G. Moore

Introduction SUMMARY Circadian alterations In lung function occur In respiratory disorders, with the nadir

It has been well documented that paduring the sleep-related hours. Higher therapeutic serum theophylline concentrations (STC)during tients with asthma can have marked worthe night have been shown to Improve lung function In reversible airway disease. Todetermine what sening of lung function during sleep and effect higher nocturnal STC would have In patients with chronic obstructive pulmonary disease that increased levels of theophylline at (COPO)on overnight lung function, oxygen saturation, and sleep quality, two different theophylline night improvethis process (1,2). Patients products were used to give higher or lower STCduring the night. We found that with a higher STC (15.0 ± 1.0 versus 11.0 ± 1.0 Itg/ml, p = 0.005) at 7:00 A.M., the overnight changes In FEV, (+7.4 with chronic obstructive pulmonary dis± 5.7% versus -18.9 ± 7.9%, respectively) and FVC (+1.8 ± 7.5% versus -17.2 ± 3.9%, respec· ease (COPO) have been noted to have tlvely) were significantly better. However, there was no apperent effect on oxygen seturatlon (mean marked nocturnal alterations in minute sleep values for higher STC were 85.3 ± 1.2%, and for lower STC they were 86.5 ± 0.8%). The ventilation, tidal volume, and oxygenahigher STC did not adversely affect sleep latency, sleep efficiency, or sleep steglng. We conclude tion during sleep (3-6). Using intravethat a higher therapeutic STCduring sleep will Improve lung function without altering oxygen setunously administered aminophylline in paration in patients with COPO. In this group of patients, the higher STC did not Interfere with sleep tients with COPO to produce averagelevcharacteristics. AM REV RESPIR DIS 1992: 145:540-544 els of 9.2 ug/ml, Ebden and Vathenen (7) did not show an improvement in nocturnal oxygensaturation, FEV 1, or FVC, and they concluded that there was no was "nonreversibility" to an inhaled beta-2- and at 7:00 P.M. to ensure a steady state, i.e., overall benefit from the administration adrenergic agonist, i.e., ~ 15% improvement less than 25% fluctuation comparing similar in the FEV 1 during daytime testing. Exclu- hours on each day (1). This was determined of theophylline in this situation. How- sion criteria were the use of corticosteroids, independently by a person not familiar with ever, if asthma is taken as a model for cromolyn sodium, any acute respiratory pro- the subjects or their study results. If a steady chronotherapeutics, then progressively cess in the preceding 6 wk and any other or- state was not present, then the subjects were higher serum theophylline concentra- gan system disease that was acutely being continued on that phase of the study using tions (STC) during the night do produce treated. Only 14patients met the above criteria the same dose until a steady state was reached. continued improvement in lung function from 24 respondents to an advertisement ask- On Days 3 and 4, the patients wereinterviewed (2). This nocturnal STC response rela- ing for patients with COPD. All participants regarding severity (on a zero to 3 scale) of tionship, interestingly, is not present dur- signed an approved Institutional Review cough, dyspnea, chest tightness, nocturnal symptoms, and side effects. Board consent form. ing the daytime (2). On the fifth day of each phase, the subThe purpose of this study, using the Protocol jects entered the sleep laboratory at 7:00 P.M., recent noted chronotherapeutic informa- To achieve different overnight STC in each and spirometry and STC were determined. tion on the use of theophylline overa 24-h subject, two types of orally administered the- Preparation for a full polysomnographic time period (1, 2), was to determine in ophylline preparations were used (1, 2). The evaluation began approximately 30 min pripatients with COPO whether higher STC TD product was Theo-Dur, which gives a rel- or to the subjects' normal bedtime (10:00 to during the night improve (1) overnight atively constant 24-h STC, and a once-daily 11:30 P.M.). The polysemnogram was evaluchanges in spirometry, (2) oxygen de- (OD) product, Uniphyl, which gives peak STC ated for sleep latency, staging, arousals, oxysaturation, (3)respiratory events, and/or between 4:00and 6:00 A.M. when given at 7:00 gen saturation, respiratory pattern, and cardi(4) sleep quality or, conversely, hinder P.M. Prior to entering into the study, each patient was receiving the TD product with therthese patients' ability to sleep. Methods Patient Selection Fourteen adult patients who met the criteria for COPD (8)and werereceiving a "12-h"theophylline preparation twice daily (TD) were selected for the study if they had greater than a 40/0 decrement from baseline in oxygensaturation for more than 15% of the night on two consecutive nights of home oximetry. An additional criteria for inclusion into the study 540

apeutic STC (between 10 and 20 ug/ml). In each subject, their individual dose of the TD preparation was used in the TD phase of the study given at 7:00 A.M. and at 7:00 P.M. For the OD phase, the total TD dose was used at 7:00 P.M. with a placebo given at 7:00 A.M. This was a double-blind crossover protocol. The patients were randomly divided into the above-noted starting phases of the study. On Days 1 and 2 of the study, medication and placebo were given as described above. On Days 3 and 4 STC were obtained at 7:00 A.M.

(Received in original form April 11, 1991 and in revised form August 5, 1991)

1 From the Department of Medicine, National Jewish Center for Immunology and Respiratory Medicine and the Pulmonary Disease Section, University of Colorado, Denver, Colorado. 2 Supported by a grant from The Purdue Frederick Company. 3 Correspondence and requests for reprints should be addressed to Richard J. Martin, M.D., The National Jewish Center, 1400Jackson Street, Denver, CO 80206.

541

THEOPHYLLINE AND SLEEP IN COPO

ac rate/rhythm. The mean saturation was identified for each 30-sepoch. The mean oxygen saturation for the entire night was then based on the epoch saturations. Additionally, the percent of time spent with an oxygen saturation below 4070 of baseline during nonrapid eye movement (non-REM) and REM were calculated. Apnea was defined as cessation of flow for ~ 10 s, and hypopnea was defined as a ~ 50070 reduction in airflow and chest motion. The heart rate was measured for I min at IS-min intervals from lights out to morning awakening. The subjects had bedtime spirometry and were awakened at 6:55 A.M. for spirometry and SlC. If spontaneous awakening occurred during the night because of complaints of dyspnea, spirometry wasalso performed. After completion of these studies, the subjects were assigned to the alternate preparation, i.e., from TD to OD and vice versa, and the study measurements were carried out again over the next 5 days.

Data Analysis The study design was a two-period, twotreatment crossover. For each variable, a univariate repeated measures ANOVA designed especially for data from a crossover study was fit (9). This model has treatment type, treatment period, and carryover effects as the independent variables. The carryover effect is always tested first, and when the carryover effect is significant, the data from the second period are not used as a bias may exist. The analysis then consists of a two-sample t test using only the first-period data. To test for a relationship between SlC and the sleep quality variables, an analysis of covariance model was used. This model fits parallel lines to each subject's data, and then tests the common slope to see if it is significant (10). Results

Characteristics of the patients are shown in table 1. There were 11 men and three women 64.7 ± 1.6 yr of age. The mean percent predicted FEV 1 was 48.9 ± 4.1070 and FVC was 67.8 ± 4.2%. The FEV 1 bronchodilator response, by definition, was ~ 15%, with a group mean of 4.1 ± 2.6%.

Symptoms and Side Effects There were no differences in morning or daytime symptoms of cough, dyspnea, chest tightness, or nocturnal complaints between both study phases. The combination of all symptoms was also similar: 00 was 0.93 ± 0.14, and TO was 0.92 ± 0.17; p > 0.05. Side effects were similar in both groups. There were no complaints of gastrointestinal symptoms, tremors, or change in sleep patterns with either preparation. Intermittent headaches occurred in five subjects receiving

TABLE 1 PATIENT CHARACTERISTICS From Daytime Screening Subject No.

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Sex M M M M M M F M M M M F F M

Mean SEM

Age (yr)

Height (em)

Weight (kg)

FEV, (L)

FEV, (% predj

FVC (L)

FVC (% predj

65 62 62 51 70 69 58 72 65 59 67 67 67 72 64.7 1.6

180 175 170 173 180 175 163 180 186 180 180 155 152 173 173.0 2.7

109.0 104.0 79.5 73.0 98.0 61.5 96.0 82.0 95.5 77.0 61.5 48.0 53.0 68.0 79.0 5.2

1.33 1.51 0.86 1.12 2.31 1.17 1.68 1.08 2.29 1.00 1.16 0.79 0.82 1.78 1.35 0.13

38 47 29 33 76 41 72 36 68 43 37 51 47 66 48. 9 4.1

1.75 3.42 2.47 3.26 3.24 3.14 2.42 2.94 3.06 1.61 2.65 1.91 1.40 4.00 2.66 0.21

42 75 60 72 72 74 78 66 62 46 57

the 00 preparation and in one subject receivingthe TO preparation, but this did not preclude drug use in any of them. Interference with daily activities was a complaint in one subject from each group.

Serum Theophylline Concentrations The individual STC at 7:00 P.M. and at 7:00 A.M. for both phases of the study are shown in figure 1. The STC at 7:00 P.M. for the TO preparation was 10.9 ± 1.0 ug/ml and for the 00 preparation it was 8.4 ± 0.6 ug/ml; p = 0.07. At 7:00 A.M. the 00 preparation was significantly higher (15.0 ± 1.0 ug/ml) than the TO preparation (11.0 ± 1.0 ug/ml); p = 0.005. The STC at 7:00 A.M. and at 7:00 P.M. were similar for the TO preparation (p > 0.05), but different for the 00 preparation (p = 0.0001). 1900h 20

15

E

Dl

" S~

10

0700h

,.-P= 0.07.....,

•••

• + I

r-P.0.005 .,

• •

••• .1. t

N:'.

• •

•

•

.:-

I

•

.L

•••

•• ' - - P> 0.05-----'

L - P =0.0001 ----.J TO

00

TO

00

THEOPHYLLINE

Fig. 1. The individual serum theophylline concentrations are shown for the twice-daily (TO) and once-daily (00) products,

88 57 100 67.8 4.2

%BD Change FEV,

FVC

13 9 -9 -15 -7 0 3 12 9 15 -3 10 15 5 4.1 2.6

-10 15 -13 -14 2 -10 -2 2 4 0 12 12 16 4 1.3 2.7

Spirometry As shown in table 2, at 7:00 P.M. there were no significant differences between the 00 and the TO preparation for FEV 1 (1.31 ± 0.15versus 1.24 ± 0.15L, respectively; p > 0.05) or for the FVC (2.52 ± 0.27 versus 2.45 ± 0.21 L, respectively; p > 0.05). Within groups, the bedtime spirometry was similar to the 7:00 P.M. values (p > 0.05), and between groups it was also not significantly different (bedtime FEV 1 : 00,1.21 ± 0.13 L; TO, 1.25 ± 0.13 L. FVC: 00, 2.31 ± 0.19 L; TO, 2.46 ± 0.16 L). The morning FVC 00 did not reach significance compared with the TO, but there was a trend toward a higher FVC (2.25 ± 0.17 versus 2.07 ± 0.20, respectively; p = 0.06). The overnight change in FVC improved (figure 2) with the 00 preparation, by 1.8 ± 7.5%, and with the TO preparation there was a decrement of 17.2 ± 3.9% (p = 0.05). A carryover effect was found for the overnight change in FEV 1 from Period 1 to Period 2. Therefore, to statistically analyze the data, only Period 1information was used for the FEV 1 measurements (table 3 givesall the FEV 1 data for both periods). Although the morning FEV 1 values were higher with 00 dosing, this did not reach statistical significance. However, the percentage overnight change in FEV 1 improved by 7.4 ± 5.7% with the 00 preparation compared with a decrement of 18.9 ± 7.9% with the TO dosing (p = 0.02) (figure 2). Polysomnography The sleep latency (onset), sleep efficiency, and sleep staging were similar while receiving both preparations (p > 0.05)

542

MARTIN AND PAK

TABLE 2 SPIROMETRIC VALUES 7:00 P.M. Subject No. Once-daily theophylline dosing 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Twice-daily theophylline dosing 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Morning

Bedtime

FEV,

FVC

FEV,

FVC

FEV,

FVC

1.59 1.42 0.73 0.86 2.27

1.60 3.52 2.29 2.44 3.45

1.60

2.29

1.93 0.82 1.00 0.95 0.82 1.74

3.16 0.88 2.99 2.29 1.42 3.85

1.16 1.21 0.59 0.95 2.09 1.26 1.50 0.95 2.06 0.82 1.13 0.69 0.87 1.68

1.26 2.61 1.62 2.75 2.69 3.22 2.07 2.94 2.77 1.24 2.19 1.94 1.65 3.33

1.51 1.47 0.69 1.00 2.27 1.09 1.60 0.79 2.06 0.78 0.69 0.60 0.74 1.34

2.17 3.00 1.80 2.18 3.10 2.80 2.22 1.70 2.84 1.27 2.61 1.51 1.47 2.83

1.37 1.51 0.73 0.95 2.13

1.62 3.48 2.53 2.22 3.27

1.30 0.78 2.20 0.82 0.97 0.78 0.82 1.74

2.16 2.24 3.10 1.84 2.07 1.82 1.63 3.86

1.37 1.38 0.78 0.91 2.13 0.91 1.40 0.79 2.16 1.08 1.42 0.86 0.73 1.55

1.87 3.12 2.34 2.44 3.19 2.34 2.10 2.24 2.97 1.88 3.05 2.20 1.41 3.30

1.21 1.38 0.60 0.73 2.26 0.86 1.40 0.65 2.24 0.82 0.75 0.60 0.74 1.29

1.59 3.01 1.37 1.56 3.29 2.17 2.03 1.96 2.91 1.26 2.25 1.25 1.38 2.94

• For technical reasons, 7:00 P.M. spirometry was not obtained.

(figure 3). These values (TD versus OD) were:sleep latency, 24.0 ± 5.7 versus 19.2 ± 6.5 min; efficiency, 74.0 ± 3.0% versus 74.5 ± 3.3%; awake, 26.0 ± 3.0% versus 24.9 ± 3.3%; Stage 1, 19.4 ± 2.3% versus 18.7 ± 1.6%; Stage 2,38.3 ± 3.0% versus 37.0 ± 3.1%; Stage 3-4, 3.3 ± 1.0% versus 5.2 ± 1.9%; REM, 13.0 ±

10

1.70/0 versus 14.2 ± 1.4%. To better demonstrate whether the STC had an effect on sleep, an analysis of covariance model was fit for all STC regardless of preparation. There was no statistical significance found with regard to sleep efficiency, percent of awake time, or sleep Stages 1 and 2 and REM sleep in rela-

10

*

5

5

*

-'--

~ IL

TO 0

00

..J

~

~

TO I I +-~__ ---..,.----L_+-_....I..00

:::l C

~ 52

·5

-5

_L...-

z II:

Z

II: W

.,.

0

IL

52 ~

~

W

·10 -r--

~ .,.

·10

·15

-15

·20

-20

T *

*

Fig. 2. The overnight changes in FEV 1 and FVC are shown for the twice-daily (TO) and once-daily (00) products. Asterisks indicate p " 0.05 (see text for exact p value).

tion to the STC. However, the higher the STC the more Stage 3-4 sleep (p = 0.05). Apneas and hypopneas were uncommon during both regimens; the TD index was 5.5 ± 1.8/h, and OD index was 4.6 ± 1.5/h (p > 0.05). The mean heart rate for the night with the TD preparation was 77.0 ± 3.3 beats/min, and with the OD preparation it was 77.7 ± 3.1 beats/min (p > 0.05). The presleep baseline oxygen saturation with the TD preparation compared with the OD preparation was 90.4 ± 0.8% versus 89.0 ± 0.9%, respectively (p > 0.05). The mean overnight oxygen saturation TD versus OD was 86.5 ± 0.8% versus85.3 ± 1.2%, respectively (p >0.05),and the nadir saturation was 73.9 ± 1.9% versus 74.1 ± 2.6%, respectively (p > 0.05). The percent of time during non-REM for TD and OD preparations with an oxygen saturation 4% below baseline was 21.4 ± 6.1% versus 18.0 ± 5.5%, respectively (p > 0.05). The similar variable during REM sleep was 43.7 ± 6.8% versus 40.2 ± 10.0%, respectively (p > 0.05). Discussion

Our results demonstrate that the overnight STC are important in the preservation of lung function. With an OD theophylline preparation given at 7:00 P.M. and producing higher STC at 7:00 A.M. than a TD preparation, there was a significant improvement in the overnight falls in both the FEV 1 and the FVC. However, the higher STC did not change the amount of sleep-related respiratory events or oxygen desaturation. Of importance in this older population (mean age, 64.7 yr), the higher STC did not adversely affect sleep onset, efficiency, or the different sleep stages or increase the heart rate. These findings with regard to theophylline not affecting sleep characteristics or heart rate werealso found by Berry and coworkers (11). They studied a similar group of patients with COPD, comparing placebo with theophylline. With slightly lower morning STC than our study, there were no differences compared with placebo with regard to the amount of awake time, Stages 1, 2, and 3-4, or REM sleep. Additionally, the mean sleep heart rate was not increased during the theophylline phase of the study nor was the apnea plus hypopnea index altered. Both the 9:00 P.M. and 7:00 A.M. FEV 1 and FVC were significantly higher with theophylline compared with placebo, and the overnight fall in FEV 1 was improved. As described in the RESULTS, a car-

543

THEOPHYLLINE AND SLEEP IN COPD

TABLE 3 FEV, MEASUREMENTS' Period II

Period I Subject No.

Bedtime (L)

Morning (L)

%~

Bedtime (L)

Mean SEM

1.16 1.21 0.59 0.95 2.09 1.26 1.50 0.95 1.21 0.16

1.51 1.47 0.69 1.00 2.27 1.09 1.60 0.79 1.30 0.19

1.37 1.38 0.78 0.91 2.13 0.91 1.40 0.79 1.21 0.16

30.2 21.5 16.9 5.3 8.6 -13.5 6.7 -16.8 7.4t 5.7

Mean SEM

2.16 1.08 1.42 0.86 0.73 1.55 1.30 0.22

2.24 0.82 0.75 0.60 0.74 1.29 1.07 0.25

1.21 1.38 0.60 0.73 2.26 0.86 1.40 0.65 1.14 0.20

-11.7 0 -23.1 -19.8 6.1 -5.5 0 -17.7 -9.0 3.8

OD Dosing

TD Dosing

9 10 11 12 13 14

%~

TD Dosing

OD Dosing

1 2 3 4 5 6 7 8

Morning (L)

3.7 -24.1 -47.2 -30.2 1.4 -16.8 -18.9t 7.9

2.06 0.82 1.13 0.69 0.87 1.68 1.21 0.22

2.06 0.78 0.69 0.60 0.74 1.34 1.04 0.23

0 -4.9 -38.9 -13.0 -14.9 -20.2 -15.3 5.6

. Definition of abbtevletlons: 00 = once-daily theophylline dosing;TO = twice-daily theophylline dosIng; %~ = percentage change in FEV, from bedtimeto morning. • See text for explanation of carryovereffect from PeriodI to Period II. t p = 0.02.

ryover effect was found for the overnight change in FEV 1 from Period 1 to Period 2. A significant carryover effect indicates that there may be residual treatment effects in the second period from the first period. This effect may also result from a treatment by period interaction or from inherent differences in the grouping of subjects. It was not possible in our study to ascertain the exact reason for this carryover effect. However, when this type of effect is statistically found, the second period should not be used in data analysis for that specific variable, as it may be biased. The improved overnight change in FEV 1 (taking into account the carryover effect) with the OD dosing regimen follows similar significant improvement in the PVC using all subjects' data. Thus, not only by statistical methodology

is the overnight improvement in FEV 1 shown, but this variable is also associated with another spirometrically improved variable that did not show any carryover effect. The importance of a higher overnight STC compared with daytime values has been previously documented in asthmatic patients (1, 2). This also appears to be true in patients with COPD. At 7:00 P.M. the STC with the OD preparation tended to be lower than those of the TD product, but the FEV 1 and FVC measurements were similar. The spirometric values at 7:00 A.M. tended to be higher, and the overnight decrements in both the FEV 1 and PVC were significantly improved with the higher STC at 7:00 A.M. with the OD preparation. Thus, it appears in various types of obstructive lung

Fig. 3. Sleep characteristics are shown for the twice-daily (closed bars) and once-daily (open bars) theophylline products. No significant differences exist.

SLEEP EFFICIENCY

AWAKE

STAGE t

STAGE2

STAGES3-4

REM

diseases that the nadir in lung function occurs during the night, and this is when the peak STC is needed. It is possible that the STC in our study are merely reflections of day-to-day variability. However, our findings are in general agreement with previous reports in ast~matic patients (1, 2, 12), and we estabhshed a steady-state STC prior to testing. It is of interest that the daytime screening STC with the TD product was greater than 10 ug/ml in all subjects, but during the study, five patients at 7:00 P.M. and six at 7:00 A.M. were below this value. To our knowledge the circadian variability of STC in patients with COPD receivingTD products has not been studied. Severalinvestigators have shown that TD preparations in asthmatics have a decrease in the nocturnal versus the diurnal STC (1, 12), but this has not been established in COPD. D'Alonzo and coworkers (2) demonstrated in asthma that a STC lung function response relationship waspresent between 2:00 and 6:00 A.M. that did not occur between 2:00 and 6:00 P.M. This suggests that more benefit is derived from higher doses of theophylline during the night. This concept is also suggested by our data. At 7:00 A.M. with higher STC with the OD preparation, there was a significant improvement in the overnight decrements in both FEV 1 and FVC. Because the mechanism(s) of action of theophylline is not known, the exact reason why higher STC improve overnight lung function is not understood. The improvement may be related to any of a host of circadian rhythms that potentiate bronchoconstriction during sleep. The possibility existsthat improvement in respiratory muscle function overnight may have also been playing a role in the improvement in lung function with theophylline. Demonstration of increased daytime respiratory muscle strength in COPD has been demonstrated (13, 14). We did not have a specific measurement of respiratory muscle function in our study. Indirectly, the FVC may reflect improved respiratory muscle function, with the overnight fall in FVC improved with the higher STC at 7:00 A.M.. Because respiratory muscle activity diminishes at night (3), as does lung volume (15), theophyllines may improve the respiratory muscle tone during sleep, resulting in the overnight improvement in lung function. Alternately, Chrystyn and coworkers (16) during daytime studies in COPD showed that a specific lung function marker, trapped gas volume (body plethysmograph minus helium dilution measure-

544

ments), was correlated with the STC in patients with COPO. The trapped gas volume progressively decreased from a placebo value of 1.84 ± 0.16 L through incremental increases in STC (5 to 10, 10 to 15, 15to 20 ug/ml) to a value of 0.67 ± 0.10 L. The spirometric values increased, but not to the extent that the trapped gas volume measurement decreased. The trapped gas volume may be a more sensitive indicator of lung function in patients with COPO, and it could place the diaphragm at an improved mechanical advantage with resultant improvement in spirometric values. Although there were distinct pharmacokinetic and pharmacodynamic differences between the two types of theophyllinepreparations, does it reallymatter which therapy is used? Symptoms and side effects were not different nor were polysomnographic variables between therapies. With the exception of oxygen therapy (17, 18), there has been difficulty in documenting the usefulness of any specific form of therapy with regard to objective changes in course or survival of patients with COPO. Obviously, only long-term studies evaluating quality of life and objective variables as well as medication side effects would answer the question if a higher morning FEV, in COPO is important. It has been shown that survivalis closelyrelated to the FEV, level (19-22). Additionally, disease prognosis is related to the rapidity of decrement of functional impairment overa few years of observation (20-22). Perhaps with improvement in overnight ventilatory function the long-term clinical course and mortality rates could be altered. The indicator that is most commonly used to classify COPO as "nonreversible" to bronchodilatation is the daytime FEV, response to an inhaled beta-2adrenergicagonist. As seen by our screening information, the mean increase in FEV, after an inhaled agonist was only 4.1 ± 2.60/0. Thus, these patients were deemed to have "fixed" airway disease. However, these entrance screeningstudies weredone during the daytime. Overnight, the group had a 18.9% fall in the FEV, with the TO theophylline preparation. Unfortunately, A.M. postinhaled bronchodilator treatment was not part of the protocol, but the daytime screening, FEV, at 7:00P.M., and bedtime FEV, were all at least 18% higher than the 7:00 A.M. FEV,. This suggests there is a circadian response in airway caliber sizein patients with COPO and "fixed" daytime bronchodilator response. Anthonisen and

MARTIN AND PAK

Wright (23) demonstrated in a large COPO population followed over 3 yr that there was a marked intraindividual response to inhaled bronchodilators over the course of their study. Thus, the term "fixed" or nonresponsiveness to bronchodilators in actuality depends on the time of day or giventime during the year the test is done. Ebden and Vathenen (7) did not show an effect of aminophylline (STC approximately 9.2 ug/ml) compared with no aminophylline with regard to nocturnal oxygenation in COPO. Berry and coworkers (11) demonstrated a slight but significant improvement in oxygen saturation only during REM sleep in patients with COPO receiving theophylline versus placebo. However, our study did not show a difference between different STC with regard to the mean, nadir, or REM versus non-REM oxygen saturation levels. Perhaps this is due to the mechanism of nocturnal oxygen desaturation in COPO. Fletcher and coworkers (24) showed that in addition to alveolar hypoventilation, deterioration of gas exchange mechanisms wereoccurring. The latter effect may not be altered by theophyllines. Whatever the exact mechanism(s) of action theophylline has at night, higher STC improve the overnight falls in FEV, and FVC while not altering the respiratory events and oxygen saturation in patients with COPO. In this older population the higher STC did not adversely affect sleep as defined by sleep latency, efficiency,and sleepstages. This and other studies reinforcethe concept that treatment of many types of respiratory disorders need to be evaluated on the basis of chronotherapeutic interventions. Acknowledgment The writers wish to thank Drs. David IkIe and Lynn Ackerson for their help in the study design and statistical methods. They also appreciate the word processing skills of Judy Tisdale and the efforts of Patricia Carter in the data analysis.

References I. Martin RJ, Cicutto tc, Ballard RD, Goldenheim PD, Cherniack RM. Circadian variations in theophylline concentrations and the treatment of nocturnal asthma. Am Rev Respir Dis 1989; 139: 475-8. 2. D'Alonzo GE, Smolensky MH, Feldman S, et al. 1\venty-four hour lung function in adult patients with asthma. Am Rev RespirDis 1990; 142:84-90. 3. Hudgel DW, Martin RJ, Capehart M, Johnson B, Hill P. Contribution of hypoventilation to sleep oxygen desaturation in chronic obstructive pulmonary disease. J Appl Physio11983; 55:669-77. 4. Fletcher EC, Gray BA, Levin DC. Nonapneic

mechanisms of arterial oxygen desaturation during rapid-eye-movement sleep. J Appl Physio11983; 54:632-9. 5. Catterall JR, Douglas NJ, Calverley PMA, et al. Transient hypoxemia during sleep in chronic obstructive pulmonary disease is not a sleep apnea syndrome. Am Rev Respir Dis 1983; 138:24-9. 6. DeMarco FJ, Wynne JW, Block AJ, Boysen PG, Taason VC. Oxygen desaturation during sleep as a determinant of the "blue and bloated" syndrome. Chest 1981; 79:621-5. 7. Ebden P, Vathenen AS. Does aminophylline improve nocturnal hypoxia in patients with chronic airflow obstruction. Eur J Respir Dis 1987; 71: 384-7. 8. ACCP-ATS Joint Committee on Pulmonary Nomenclature. Pulmonary terms and symbols. Chest 1975; 67:583-93. 9. Jones B, Kenward M. Design and analysis of cross-over trials. London: Chapman and Hall, 1989; 30-50. 10. Snedecor GW, Cochran WG. Statistical methods. 7th ed. Ames: Iowa State University Press, 1980; 365-92. ll. Berry RB, Dega MM, Branum JP, Light RW. Effect of theophylline on sleep and sleep-disordered breathing in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1991; 143: 245-50. 12. Rogers RJ, Wiener MB, Hill MH, Szefler SJ. Theophylline absorption from two sustained-release products. Am Rev Respir Dis 1987; 136:1168-73. 13. Murciano D, Auclair MH, Pariente R, Aubier M. A randomized, controlled trial of theophylline in patients with severe chronic obstructive disease. N Engl J Med 1989; 320:1521-5. 14. Mahler DA, Matthay RA, Snyder PE, Wells CK, Loke J. Sustained-release theophylline reduces dyspnea in nonreversible obstructive airway disease. Am Rev Respir Dis 1985; 131:22-5. 15. Ballard RD, Irvin CG, Martin RJ, Pak J, Pandey R, White DP. Influence of sleep on lung volume in asthmatic patients and normal subjects. J Appl Physiol 1990; 68:2034-41. 16. Chrystyn H, Mulley BA, Peake MD. Dose response relation to oral theophylline in severe chronic obstructive airways disease. Br Med J 1988; 297: 1506-10. 17. Medical Research Council Working Party. Long-term domiciliary oxygen therapy in hypoxic cor pulmonale complicating chronic bronchitis and emphysema. Lancet 1981; 1:681-6. 18. Nocturnal Oxygen Therapy Trial Group. Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive lung disease. Ann Intern Med 1980; 93:391-8. 19. Traver GA, Cline MG, Burrows B. Predictors of mortality in chronic obstructive pulmonary disease. Am Rev Respir Dis 1979; 119:895-902. 20. Diener CF, Burrows B. Further observations on the course and prognosis of chronic obstructive lung disease. Am Rev Respir Dis 1975; 1ll:719-24. 21. Burrows B, Earle RH. Prediction of survival in patients with chronic airway obstruction. Am Rev Respir Dis 1969; 99:865-71. 22. Anthonisen NR, Wright EC, Hodgkin JE, IPPB Trial Group. Prognosis in chronic obstructive pulmonary disease. Am Rev Respir Dis 1986; 133:14-20. 23. Anthonisen NR, Wright EC, IPPB lrial Group. Bronchodilator response in chronic obstructive pulmonary disease. Am Rev Respir Dis 1986; 133:814-9. 24. Fletcher EC, Gray BA, Levin DC. Nonapneic mechanisms of arterial oxygen desaturation during rapid-eye-movement sleep. J Appl Physio11983; 54:632-9.