MATERIALS AND METHODS
Patient Population
Twenty children (11 boys, nine girls) with mild,
asymptomatic asthma between the ages of 6 and 12
Behavior
Abnormalities
and Poor School
Performance
Due to Oral Theophylline
Use
Gary S. Rachelefsky, MD, Julie Wo, RN, Judith Adelson, MA,
M. Ray Mickey, PhD, Sheldon L. Spector, MD, Roger M. Katz, MD,
Sheldon C. Siegel, MD, and Albert S. Rohr, MD
From the Divisions of Allergy/Immunology, Departments of Pediatrics and Biomathematics, University of California at Los Angeles
ABSTRACT. Studies evaluating adverse effects of oral
theophylline on learning and behavior have been per-formed on children with asthma receiving long-term the-ophylline therapy. To further differentiate the effects of asthma itself from the drugs used, we evaluated 20 asth-matic children (6 to 12 years of age) who had not received oral bronchodilators for at least 6 months. A double blind, placebo-controlled, parallel format was used with a 4-week theophylline or placebo period preceded by a 2-week baseline. Theophylline serum levels were maintained be-tween 10 to 20 tg/mL. During baseline and treatment periods, the child’s home and school behavior/perform-ance were monitored independently by their parents and teachers using standardized report forms. A battery of psychologic tests was administered at the end of baseline and treatment periods. Seven children receiving theoph-ylline were noted to have a change in school behavior and/or performance during their 4 weeks on drug com-pared to baseline, whereas none of the children receiving placebo were noted to be different
(P
= .004). Thus, the short-term administration of theophylline to asympto-matic asthmatic children not receiving oral bronchodi-lators can adversely affect school performance and be-havior. Because this population represents the majority of asthmatic children, one needs to use theophylline cautiously in this age group, monitor school performance closely, or seek other treatment modalities. Pediatrics1986;78:1133-1138; theophylline, behavior problems, learning problems, asthma.
Oral theophylline has been used frequently dur-ing the past decade as a chronic maintenance med-ication for the treatment and/or the prevention of
childhood asthma. Theophylline has been noted to
have various side effects including CNS stimulation
which may lead to nervousness, restlessness,
tremors, insomnia, muscle incoordination, and
speech disturbances. Recent studies have been de-signed to determine the effect of maintenance doses (serum theophylline levels between 10 and 20 tg/ mL) of oral theophylline on children’s learning and behavior. Furukawa et al’ reported on the improve-ment in six children with learning and behavior
problems when their theophylline treatment was
changed to cromolyn sodium. Springer et a!2
re-ported the results of a double-blind, randomized, crossover study involving a series of psychologic tests performed on 13 asthmatic children taking
theophylline and cromolyn sodium for 4 weeks
each. Visual-spatial planning tests were better for the four children with low IQs (87 to 105) when
they were receiving cromolyn sodium compared
with theophylline. Suess and Chai3 reported visual retention and spatial configuration difficulties in asthmatic children treated with multiple asthma
medications.
All of the prior studies13 involved children with moderate asthma already receiving oral broncho-dilators on a long-term basis. The present investi-gation was designed to explore the relationship of theophylline and its effects on standard tests of memory, learning and attention, school behavior
and learning, and home behavior in a group of
children with mild asthma who did not require long-term oral medication in an attempt to eliminate the complicating effects of both varying asthma
symp-tomatology and complex medication regimens on
the test results.
Received for publication Dec 13, 1985; accepted March 19, 1986. Reprint requests to (G.S.R.) 11645 Wilshire Blvd, No. 600, Los
Angeles, CA 90025.
History Physical
Behavior and
Learntng Tests
Behavior and Learning Tests
T
I
ATA
bib
TAT
Mi
BAS(LINE
WEEK 1 2 3
VISIT 1 2 3
Five days
ACTIVE OR PLACEBO THEOPHYLL1HE
4 5 6
4 5
years (9.8 ± 2.1 years, mean ± SD), were selected for study. Selection was on the basis of having mild asthma without the need for long-term oral asthma medications. None of the children had received an oral bronchodilator for at least 6 months, the ma-jority (14 of 20) for 9 months or longer. In addition,
none of the children were taking oral antihista-mines and/or decongestants. None had learning disabilities or behavior disorders.
Protocol
The trial was designed to take place during a 6-week period during the patient’s school year
(Fig-ure). Weeks 1 and 2 provided a baseline period
during which the parents evaluated the child’s be-havior and physical symptoms. During this baseline period, the child’s school performance was evalu-ated by the child’s teacher. At the beginning of
week 3, subjects received a battery of tests (Table 1) administered by an educational therapist (J.A.). The children then entered the double-blind, ran-domized, parallel phase of the study. Ten subjects received sustained-release theophylline, and ten subjects received a matching placebo every 8 to 12 hours. All children had theophylline levels (Seroan-alyzer, Miles Laboratory Ames Division, Elkhart, IN) obtained prior to entering study and at the end of baseline period to be sure they were not receiving theophylline. Once starting the treatment phase, the serum level was checked in all children at day 3 and every three days as required for those patients receiving active drug to maintain a level between 10 to 20 sg/mL. Those children receiving placebo had no adjustment in placebo dose. The theophyl-line dose was adjusted independently by a physician not involved in the study. Theophylline serum
1ev-Figure. Diagram depicting protocol and day of psychologic testing, completion of parent and teacher evaluations, and determination of serum theophylline levels. L, dispense diary cards and parent and teacher assessment forms; A, retrieve diary cards and parent and teacher assessment forms; T, theophylline level determination; D, dispense study drug; U, retrieve study drug.
TABLE 1. Psychologic Tests Administered Tests
Attention tests
Digital Span (Wechsler Intelli-gence Scale for Children-Revised [WISC-R]
Coding (WISC-R)
Memory and learning tests Benton Visual Retention
Wechsler Memory Scale
Vocabulary (WISC-R)
Mazes (WISC-R)
Description
Repeating digits sequence both forward and backward. Index of immediate re-call which measures capacity to retain several unrelated elements, attention and concentration.
Requires learning and reproducing a code substituting novel symbols for numer-als. Test is an index of attention and concentration, which measures psy-chomotor speed, short-term visual memory, and visual motor coordination.
Ten geometric designs are studied by the child for 10 seconds and then they are drawn from memory. This test meas-ures visual memory and visual percep-tion.
Seven subjects are administered in ap-proximately 15 to 20 mm.
Child defines a list of words, which meas-ures learning ability.
els were rechecked at the end of the 4-week treat-ment period in all subjects.
The study was approved by the Human Subject Protection Committee of the University of Califor-nia, Los Angeles. Appropriate informed consent forms were signed by parents and permission was obtained from the child.
Psychologic Evaluation (Table 1)
At the end of the baseline period and after the 4-week treatment period, a battery of psychologic tests was administered. These tests were selected to measure memory and learning, attention span, and spatial visualization. Initially, a full Wechsler Intelligence Scale for Children-Revised (WISC-R)4 was administered to define each child’s IQ.
Patient Monitoring
At the end of the baseline period and at 2 weeks and 4 weeks of treatment, the parents completed a standardized Child Behavior Check List (Thomas Achenbach, PhD, Burlington, VT, unpublished) and a check list evaluating activity, attention prob-lems and associated behavior5 (modified version of ANSER system). The parents evaluated the child’s behavior and characteristics on 26 associated
be-TABLE 2. Patient Evaluation Performed by Parents
Associated Behaviors
Is moody; Has a bad temper; Cries easily; Is a worrier; Has bad dreams; Is often sad; Is often very quiet; Doesn’t seem to like himself/herself (no self-respect); Has many fears; Is often tired; Stutters or stammers; Has stomach aches or headaches often; Wets bed or has accidents with bowel movements; Has poor appe-tite; Overeats often; Worries about his/her health; Has
nervous twitches; Complains of feeling ill often; Has constipation; Is often too concerned about cleanliness and/or neatness; Has friends who get him/her in trou-ble; Constantly fights with brothers and/or sisters; Gets his/her friends into trouble; Destroys property; Gets into trouble with neighbors; Gets into fights often Activity-Attention Problems
His/her body is in constant motion; His/her body is underactive; His/her mind seems overactive; He/she has trouble sitting through a meal; He/she does things without thinking; He/she starts things but doesn’t finish them; At times, he/she doesn’t seem to hear what you say; He/she does things in the wrong order; He/she does not realize when he/she’s made a mistake; He/she has trouble falling asleep at night; He/she has trouble staying asleep at night; He/she yawns often during the day; He/she breaks things around the home; He/she seems to do things the hard way; He/she stares at things for long periods of time; He/she listens to outside noises for long periods; He/she gets distracted easily; He/she likes to keep changing games; He/she is hard to control on a long trip; He/she can’t keep his! her hands to himself/herself; He/she seems to want things all the time (is seldom satisfied)
havior and 21 activity-attention problems (Table 2). The items were scored as follows: 1, does not apply; 2, applies somewhat; 3, definitely applies. In addition, daily diaries were kept by the child/par-ents detailing all medications consumed and signs and symptoms, including headaches, shaking, night awakening, trouble going to sleep, nervousness, de-crease in appetite, stomachache, change in stool pattern, vomiting, bed-wetting, and day and night-time coughing and wheezing. These thirteen symp-toms were scored on a daily diary by child/parents as follows: 0, none; 1, mild, one time or episode; 2, moderate, two or three times or episodes, causing some interference with sleep and activities; 3, se-vere, more than three times or episodes with con-siderable interference with sleep and activities.
The child’s teacher(s) (Table 3) (contacted prior to the child’s entering the study), completed a standardized learning and behavior form5 (modified ANSER system) at the end of the baseline period and at the end of the 4-week treatment. Twenty activity-attention behavioral observations and 33 associated behavioral observations were scored as follows: 1, as much as other students; 2, somewhat more than other students; 3, much more than other
TABLE 3. Patient Evaluation Performed by Teachers
Behavioral Observations (Activity-Attention)
Keeps getting out of seat; Seems to do things fast, without thinking*; Learns best on a one-to-one basis; Is unaware of own mistakes; Seems to have too much energy; Has trouble finishing a task*; Seems to “tune out” intermittently; Is not sensitive to punishment; Is very impatient for rewards or approval; Is easily dis-tracted from work*; Does tasks at the wrong time; Hand and/or feet in motion; Tires easily during a task; Does not plan work; Has marked variation in moods; Stares for long periods; Seems underactive or lethargic; Is slow to take up a new task; Makes careless mistakes; Has trouble during unstructured time
Behavioral Observations (Associated)
Frequently goes to health office; Appears unkempt or unclean; Complains of aches or pains; Seems accident prone (gets many injuries); Misses school frequently; Seems self-conscious about body; Is not popular with other students; Seems isolated-does things alone; Is quiet; Most friends are younger children; Is passive in relationships; Gets called names by other students; Is
aggressive towards other students; Tells lies; Gets into fights*; Has trouble accepting discipline; Shows de-structive behavior; Gets angry easily*; Has a bad tem-per; Has to be sent to principal for discipline; “Cuts” classes during school; Appears worried; Gets upset or cries easily; Appears anxious or fearful; Needs reassur-ance from adults; Is hard on himself/herself; Appears sad or depressed; Lacks self-respect; Is a class clown; Does not appear to care about making mistakes or failing test; Tries to copy work of others (cheats); Is defiant toward teacher(s); Acts tough or “cool”
students. In addition, the control of the patient’s asthma was assessed by the physician at entry into the study, at the end of the baseline period, and at 2 and 4 weeks of treatment.
Statistics
Comparability of the two groups of patients was evaluated by two-group Student’s t tests computed using baseline assessments for each background variable and each assessment variable. Drug and placebo treatment effects were compared by two-group Student’s t tests of change scores for each of the several response assessments. Additional, more sensitive but more complex analyses were also com-puted for all assessments including analysis of the individual scale scores for the Child Behavior Check List; covariance analysis of change scores with pretreatment score as covariate; covariance analysis of change scores with pretreatment score, age, gender, physician’s assessment of severity, medication, and WISC-R full IQ as covariates; two-way analysis of variance, treatment by gender; and two-way analysis of variance, treatment by IQ, with
IQ categorized as above and below median.
Calcu-lations were carried out using BMDP computer
programs,6 P3D for t tests, P7D for analysis of variance, and P1V for analysis of covariance.
We report results using Student’s t tests, al-though the more complex tests were also used for data analysis but results did not differ from the t
tests.
RESULTS
Symptoms’ scores were averaged for the 2-week baseline period and also for the 4-week treatment period. Change scores were computed as treatment minus baseline average. There was no significant
difference in the change score between the two
groups.
Overall scores for the parent evaluations at base-line and at 2 and 4 weeks of treatment were ob-tamed by summing the item scores for both sets of items separately and combined for pretreatment evaluations, midtreatment evaluation, and
post-treatment evaluation. Change scores were
com-puted as midtreatment overall score minus
pre-treatment overall score and average of mid- and posttreatment overall scores minus pretreatment overall score. Nine summary scores and six change scores summarized the parent evaluation of behav-ior. Parents also completed the Child Behavior Check List at baseline and 4 weeks of treatment. Total scores for pretreatment and posttreatment
were obtained by summing the item scores of 121
items questioned. Change scores were computed as posttreatment total scores minus pretreatment to-tal scores. The results of the parent evaluations
and the Child Behavior Check List completed by
the parent were not significantly different between the theophylline and the placebo groups.
Each teacher completed at baseline and 4 weeks of treatment an evaluation (Table 3) designed to characterize the student’s ability to control his or her activity or attention. Total scores were obtained by summing the item scores. Change scores of the
53 items were computed as posttreatment score
minus pretreatment score. The intergroup compar-isons of change scores (Table 4) of the teacher’s evaluations revealed a significant difference
(P
= .004) in favor of the placebo group. Seven childrenreceiving theophylline were identified by the
teacher correctly and all had significant
(P
< .001) changes in behavior-attention. None of the placebo-treated children were identified as being treatedwith theophylline, and none had significant
changes in behavior-attention.
For the psychologic tests, raw scores were used for each of the scored items. Change scores were computed for each item separately. Change scores were computed as posttreatment score minus pre-treatment score for each category scaled separately. The number of category scales for which the post-treatment scores were greater than the pretreat-ment scores was computed. There were no signifi-cant differences between treatment groups for any of the tests administered.
All of the children treated with theophylline had
serum levels greater than 10 g/mL; none of the
placebo-treated children had theophylline present. Initially, 22 patients were enrolled; two dropped out, one because of theophylline intolerance and
TABLE 4. Teacher Behavio ral Assessment Score Pre- and Posttreatment* Group Pretreatment Posttreatment Post- Minus
Pretreatment
Time
P Valuet
Drug Placebo
Tre atment
P
valuet14.0 ± 14.3 9.0 ± 12.0
.44
19.3 ± 18.2 5.5 ± 9.9
.08
5.3 ± 5.9 -3.5 ± 4.7
.004
.04 .04
* Results are average scores ± SD. Score is sum for 53 items, each of which was scored as: 1, as much as other students; 2, somewhat more than other students; 3, much more than other students.
the other because of other time commitments.
None of the background variables evaluated in-fluenced the results. Specifically, there was no re-lationship between IQ level and any of the psycho-logic tests or parent or teacher evaluations.
DISCUSSION
We have demonstrated that when high-dose
sus-tamed-release theophylline is administered to
school-aged children, their teachers can easily dis-criminate between those on theophylline compared with those receiving placebo. Parent evaluations and detailed psychologic testing were not discrimi-nating. The children selected for study had a history of asthma but had not received oral bronchodilators for at least 6 months prior to their evaluation. They were all clinically well and were not taking any other medications. Thus, our population’s results could not be affected by concomitant disease and/ or medications.
The present investigation confirms the results of previously reported studies that evaluated behavior and learning in asthmatic children while using oral theophylline. Two recent studies”2 compared the effect of oral theophylline and cromolyn sodium on learning and behavior in children. Both medica-tions were equally effective in controlling symp-toms of chronic asthma in children. However, the two studies revealed learning and behavior prob-lems in the children during the theophylline treat-ment period.
Springer et al2 evaluated 13 children with chronic bronchial asthma who were receiving long-term theophylline therapy during 4-week treatment pe-nods (double-blind, crossover trial) of theophylline and cromolyn. They performed a multitude of psy-chologic tests (including the WISC-R and the Ben-ton Visual Retention Test) and observed that, while receiving a regimen of theophylline, a small number (four of 13) of children with lower IQ scores (87 to
105) had abnormal visual-spatial planning
com-pared with the nine children with higher IQ scores (111 to 134). They did not evaluate school perform-ance, and the parents did not perform home behav-ior/attention evaluations.
Furukawa et al1 evaluated six asthmatic children on and off theophylline treatment (for 2 weeks) with a battery of psychologic tests (including the WISC-R) and parents’ and teachers’ observations. The psychologic evaluation 2 weeks off theophyl-line revealed significant improvement in attention,
concentration, and/or memory compared with the
period during which they were receiving theophyl-line. The parents reported a number of significant changes in behavior while off theophylline with positive changes reported for five items: less
rest-less, distractable, irritable, more manageable, and slept better. Teachers’ evaluations were incomplete; therefore, they were not reported.
In addition, allergic children have been reported to rate lower in reading, auditory perception, and visual perception.7 In another study, 6- to 11-year old allergic children were judged less proficient in intellectual, academic, emotional, and social areas; however, achievement, intelligence, and visual-mo-tor test performances failed to distinguish them.8 McLaughlin et al9 evaluated the relationship of allergies and its treatment in 400 children using a standardized school history form completed by par-ents. There were no differences between allergic and nonallergic children in terms of academic and language performance, retention, or behavior prob-lems. Parents who considered their children’s be-havior as inappropriate attributed this to medica-tion side effects. Specifically, receiving theophylline was correlated significantly with inattentiveness, hyperactivity, irritability, and difficult to handle.
Brumback et al1#{176}have suggested that theophyl-line might cause asthmatic children to suffer from dysphoria or depression which would account for theophylline-induced restlessness, hyperactivity, and irritability. Our study was not designed to evaluate this possible adverse effect of theophylline.
The results of this study demonstrate that, when high-dose sustained-release theophylline is admin-istered to school-aged children, their teachers can easily discriminate between those receiving theoph-ylline compared with those receiving placebo. The teachers observed changes in multiple facets of attention-activity (Table 3) with the most discrim-mating being “has trouble finishing a task,” “seems to do things fast without thinking,” “is easily dis-tracted from work,” “gets into fights,” and “gets
angry
easily.” Because of the design of the study,this population’s results cannot be attributed to concomitant disease and/or medications. The chil-then selected for the study had a history of asthma but had not received oral bronchodilators for at least 6 months prior to their evaluation and were all clinically well and not taking any other medi-cations. Although teacher evaluations were able to distinguish which child was receiving placebo and which was receiving theophylline, parent evalua-tions were not able to distinguish the two groups.
adverse effect of theophylline was on attention
span. Subtle problems with attention are least
likely to be observed by parents, because children are not required to concentrate on such details at home. The fact that none of the psychologic tests discriminated between placebo v active medication suggests either that they were not sensitive enough or that longer treatment (more than 4 weeks) would be necessary for these tests to pick up significant changes.
At the present time, there is no exact knowledge on the mechanism(s) of theophylline-induced CNS side effects. In the range of therapeutic serum con-centrations, decreases in cerebral blood flow have
been demonstrated.” The resultant reduction in
cerebral oxygen tension could conceivably lead to hypoxia of brain tissues.
Thus, the published data suggest that allergic
asthmatic children do not have problems with
school performance/behavior unless they are being treated with oral bronchodilators, especially the-ophylline. The present data lend support for this concept. This study is different from the previously published studies in that an asthmatic population was evaluated while eliminating the potential
ad-verse effects of asthma and its treatment. The
results have demonstrated, indeed, that the initia-tion of oral theophylline treatment to achieve serum levels between 10 and 20 tg/mL will cause adverse effects in school behavior and performance, easily identified by the child’s teacher.
This study should encourage physicians, parents, and teachers to institute more systematic observa-tion of child behavior in their evaluation procedures of treatment. Because a large majority of asthmatic children do require long-term use of medication, school performance and behavior should be moni-tored closely, especially when the child is being treated with oral theophylline, to alter treatment modalities if necessary. Improved communication
between all concerned is imperative to clarify the impact of current medical treatment on school per-formance.
ACKNOWLEDGMENTS
This work was supported by a grant from Allergy Research Foundation, Inc, Los Angeles.
We thank Drs Melvin Levine (University of North Carolina at Chapel Hill) and Michele Hindi-Alexander (State University of New York at Buffalo) for advice in designing of our investigation. The Psychological Cor-poration (Harcourt Brace Javanovich, San Diego) sup-plied materials for the psychologic tests, and Jane Ya-mashita assisted us in finalizing the manuscript.
REFERENCES
1. Furukawa CT, Shapiro GG, Duhamel T, et a!: Learning and behavior problems associated with theophylline therapy.
Lancet 1984;1:621
2. Springer C, Goldenberg B, Dov IB, et al: Clinical, physiologic
and psychologic comparison of treatment by cromolyn or theophylline in childhood asthma. J Allergy Clin Immunol
1985;76:64-69
3. Suess WM, Chai H: Neuropsychological correlation of asthma brain damage on drug effects. J Consult Gun Psychol
1981;49:135
4. Wechsler D: Wechsler Intelligence Scale for Children, Re-vised. New York, The Psychological Corporation, 1957 5. Levine MD: ANSER System. Cambridge, MA, Educators
Publishing Service, 1980
6. Dixon WJ (ed): BMDP Statistical Software. Los Angeles, University of California Press, 1983
7. Havard JG: Relationship between allergic conditions
lan-guage/or learning disabilities. Dissert Abst
mt
1975;35:6940 8. Rawls DJ, Rawls JR, Harrison DW: An investigation of sixto 15 year old children with allergic disorders. J Consult Clin Psychol 1971;35:160-264
9. McLaughlin J, Nall M, Isaacs B, et al: The relationship of allergies and allergy treatment to school performance and student behavior. Ann Allergy 1983;51:506-510
10. Brumback RA, Wilson H, Station RD: Behavior problems in children taking theophylline. Lancet 1984;1:958
