Health care education, delivery, and quality
Seasonal allergic rhinitis is associated with a
detrimental effect on examination performance
in United Kingdom teenagers: Case-control
study
Samantha Walker, PhD,a,c Saba Khan-Wasti, BSc,a Monica Fletcher, MSc,a
Paul Cullinan, MD,b Jessica Harris, MSc,b and Aziz Sheikh, MDa,c
Warwick, London, and Edinburgh, United Kingdom
Background: Seasonal allergic rhinitis is common globally, and
symptoms have been shown to impair learning ability in
children in laboratory conditions. Critical examinations in
children are often held in the summer during the peak grass
pollen season.
Objective: To investigate whether seasonal allergic rhinitis
adversely impacts examination performance in United
Kingdom teenagers.
Methods: Case-control analysis of 1834 students (age 15-17
years; 50% girls) sitting for national examinations. Cases were
those who dropped 1 or more grades in any of 3 core subjects
(mathematics, English, and science) between practice (winter)
and final (summer) examinations; controls were those whose
grades were either unchanged or improved. Associations
between allergic rhinitis symptoms, clinician-diagnosed allergic
rhinitis, and allergic rhinitis–related medication use, recorded
on examination days immediately before the examination, were
assessed using multilevel regression models.
Results: Between 38% and 43% of students reported symptoms
of seasonal allergic rhinitis on any 1 of the examination days.
There were 662 cases (36% of students) and 1172 controls.
After adjustment, cases were significantly more likely than
controls to have had allergic rhinitis symptoms during the
examination period (odds ratio [OR], 1.4; 95% CI, 1.1-1.8;
P 5 .002), to have taken any allergic rhinitis medication
(OR, 1.4; 95% CI, 1.1-1.7; P 5 .01), or to have taken
sedating antihistamines (OR, 1.7; 95% CI, 1.1-2.8; P 5 .03).
From aEducation for Health, Warwick; bthe Department of Occupational and
Environmental Medicine, Imperial College of Science, Technology and
Medicine, London; and cthe Allergy and Respiratory Research Group,
Division of Community Health Sciences, General Practitioner Section,
University of Edinburgh.
Supported by a restricted grant from Schering-Plough UK.
Disclosure of potential conflict of interest: S. Walker has consulting arrangements
with and has received grant support from Schering-Plough. M. Fletcher has
received grant support from GlaxoSmithKline, AstraZeneca, and Novartis.
The rest of the authors have declared that they have no conflict of interest.
Received for publication November 2, 2006; revised March 23, 2007; accepted
for publication March 26, 2007.
Available online June 13, 2007.
Reprint requests: Samantha Walker, PhD, Education for Health, 10 Church
Street, Warwick, United Kingdom CV34 4AB. E-mail: s.walker@
educationforhealth.org.uk.
0091-6749/$32.00
Ó 2007 American Academy of Allergy, Asthma & Immunology
doi:10.1016/j.jaci.2007.03.034
Conclusion: Current symptomatic allergic rhinitis and rhinitis
medication use are associated with a significantly increased risk
of unexpectedly dropping a grade in summer examinations.
Clinical implications: This is the first time the relationship
between symptomatic allergic rhinitis and poor examination
performance has been demonstrated, which has significant
implications for clinical practice. (J Allergy Clin Immunol
2007;120:381-7.)
Key words: Seasonal allergic rhinitis, examination performance,
learning outcomes
The International Study of Asthma and Allergies in
Childhood found that seasonal allergic rhinoconjunctivitis
(more commonly referred to as allergic rhinitis) each year
affects between 1.4% and 39.7% of the pediatric popula-
Health care education,
tion,1 with a peak age of onset in adolescence.2,3 Common
delivery, and quality
symptoms include sneezing, itching, watery rhinorrhea,
and nasal blockage, which may lead to sleep disturbance,
limitations in activity, and both practical and emotional
problems.4 More specifically, a few studies have investi-
gated the effects of allergic rhinitis on learning ability by
using computer simulation in a classroom situation.
They showed significant learning impairment in children
with symptomatic allergic rhinitis compared with asymp-
tomatic control subjects,5 effects that were compounded
by sedating antihistamines.6
However, as far as we are aware, there are no studies
looking at the effect of seasonal allergic rhinitis on actual
examination performance. This is a crucial question in
populations such as the United Kingdom (UK), where a
quarter of adolescents experience seasonal allergic rhinitis
and critical student examinations (General Certificate of
Secondary Education [GCSE] examinations) in children
15 to 17 years old, which have a major bearing on future
educational and employment trajectories, take place dur-
ing a 6-week period (mid May to the end of June) when
grass pollen counts are at their highest.7
Depending on their academic abilities, GCSE students
sit for between 5 and 15 examinations in different subjects
during a 5-week period between early May and the end
of June. Examinations in 3 subjects—mathematics,
English, and science—are taken by all students. Practice
381
 382 Walker et al
Abbreviations used
GCSE: General Certificate of Secondary Education
OR: Odds ratio
UK: United Kingdom
examinations in the winter (November through January)
are structured similarly to the final examinations and are
used to give students a trial run before their finals. Students
are encouraged to treat practice examinations seriously
and to study as they would for their finals. Both sets of
examinations are marked on an 8-point scale. The results
of practice examinations, together with other information
about the child such as any learning difficulties, home
pressures, or school problems, allow teachers to predict
what results students might expect in their finals if they
study at a similar level. The expectation is that most
children will either achieve their predicted grades or, with
increased effort, improve on them when sitting for finals.
Any drop in grade is therefore unexpected.
A recent editorial highlighted the potential effect of
allergic rhinitis on examination results and called for
further investigation.8 This study addresses the important
and topical question of whether allergic rhinitis adversely
affects summer examination performance in adolescents
by examining the association between symptomatic aller-
gic rhinitis and unexpectedly poor performance in GCSE
examinations.
METHODS
Health care education,
delivery, and quality
We performed a case-control study of students age 15 to 17 years
who were sitting for GCSE examinations in 3 core subjects: math-
ematics, English, and science.
Inclusion and exclusion criteria
State schools in the West Midlands area of the UK were identified
from Local Education Authority records. Schools were invited to
participate if they had a large (>100 pupils) relevant-aged English-
speaking student population and held practice GCSE examinations in
the winter (November through January) and finals in the summer
(May through June). All students age 15 to 17 years in the last year of
study for their GCSE examinations were invited to participate via a
letter sent by the school to the child and their parents; students were
excluded or withdrawn from the study if at any point they expressed
reluctance to participate or in the event of parental or school
objections.
Definition of cases and controls
Case status was defined by comparison of the performance of each
student in winter practice and final summer examinations. If students
dropped at least 1 grade in any of the 3 core subjects, they were
considered cases. Controls were students whose grades in their final
examinations were at least as good as those in their practice
examinations in all 3 subjects.
Reporting of allergic rhinitis, medication use,
and other confounders
Two questionnaires were administered, 1 before the grass pollen
season and 1 on the day of each relevant examination. The first, in
J ALLERGY CLIN IMMUNOL
AUGUST 2007
April 2004, ascertained whether students had ever received a diag-
nosis of seasonal allergic rhinitis from a physician; we also used it to
collect information on potential confounders or effect modifiers:
d Postcode (zip code) of the student, from which we derived an
index of social deprivation9
d Treatments used for allergic rhinitis (particularly sedating
antihistamines)
d Ethnicity using 2001 census categories10
d Smoking status
d History of current asthma symptoms and/or treatment and a
history of clinician diagnosed asthma ever
d Other long-term health problems
d Recent life events that may contribute to poor examination
performance
d Season of birth (spring, March through May; summer, June
through August; autumn, September through November;
winter, December through February)
Our main exposure of interest was the presence of symptomatic
allergic rhinitis on the day of the examinations. In an attempt to avoid
recall bias, this information was collected using a short questionnaire
administered immediately before the examination on each of the final
examination days in May and June 2004. The questionnaire also
asked about the severity of these allergic rhinitis symptoms—using
a Likert scale from 0 to 10—and medication use. The questionnaire
was designed so that only students who reported symptoms were
asked to complete the question on medication. We assumed that
students who did not report symptoms—and those who failed to
complete the question on medication—were not using treatment on
that day. Students who reported oral medication use on an examina-
tion day and who had reported taking chlorpheniramine for their
allergic rhinitis in the April questionnaire were considered to have
taken sedative medication. All other antihistamines were considered
to be nonsedating on the basis of current evidence.11
Practice and final GCSE examination results were obtained from
schools. These data were entered into a database by a researcher who
was unaware of whether students had allergic rhinitis.
Formal National Health Service ethical/institutional review was
deemed unnecessary because of the setting of the study in schools
rather than within the healthcare system (Multi-Center Ethics
Committee, Personal communication, October 2003). In keeping
with Good Clinical Practice guidelines,12 the protocol was subjected
to rigorous peer review, approvals were obtained from the relevant
Local Educational Authorities and schools, informed consent was
obtained, and we complied with the Data Protection Act13 to ensure
confidentiality throughout.
Outcomes
Our primary comparison was of the proportions of cases and con-
trols with symptomatic allergic rhinitis on the day of any examination.
We further looked for evidence of a dose-response relationship be-
tween average symptom scores and the odds of dropping a grade.
Not all students with current symptomatic allergic rhinitis will
have received a physician-diagnosis of allergic rhinitis, and those
who have may represent the more severe end of the disease spectrum.
Thus, we repeated the analyses after restricting children to those who
had reported physician-diagnosed allergic rhinitis.
To test the specificity of the association between symptoms and
(sedative) medication use on dropping a grade, the reverse scenario
was also considered—that is, the effect of symptoms and medication
use on the risk of increasing at least 1 grade in any subject.
Pollen counts
Pollen counts were measured by the Pollen Research Unit,
Worcester, UK. Briefly, tree and grass pollen counts were recorded
J ALLERGY CLIN IMMUNOL Walker et al 383
VOLUME 120, NUMBER 2

daily at 3 sites in the West Midlands area during the entire TABLE I. Characteristics of cases and controls
examination period by using a Burkard spore trap at an elevation of
18 m.14 Pollen counts are reported as daily averages. Cases Controls
(n 5 662) (n 5 1172) P value
Statistical methods Sex, n (%) male 359 (54.2%) 552 (47.1%) .003
Sample size. On the basis of the conservative assumption that 15% School
of adolescents have allergic rhinitis, and on pilot data that 20% of 1 56 (8.5%) 91 (7.8%) <.001
students fail to achieve predicted GCSE grade in at least 1 subject, to 2 32 (4.8%) 126 (10.8%)
have 80% power at the 5% level (2-tailed test), sample size estimates 3 73 (11.0%) 58 (5.0%)
suggested that 1560 students (ratio of cases: controls of 1:5) were 4 55 (8.3%) 160 (13.7%)
required to detect an odds ratio of at least 1.6 for dropping a grade in at 5 19 (2.9%) 92 (7.9%)
least 1 examination. 6 68 (10.3%) 40 (3.4%)
Statistical analysis. Univariate comparisons between case status, 7 20 (3.0%) 65 (5.6%)
sex, smoking, and other variables were made by using x2 tests, Fisher 8 46 (7.0%) 92 (7.9%)
exact tests, or Mann-Whitney tests. Mean symptom scores were cal- 9 108 (16.3%) 79 (6.7%)
culated from all symptom scores reported on examination days. 10 26 (3.9%) 129 (11.0%)
Associations with case status were quantified using multilevel regres- 11 74 (11.2%) 94 (8.0%)
sion models, with the school defined as a nested cluster. Evidence of 12 37 (5.6%) 101 (8.6%)
effect modification was investigated by performing stratified analy- 13 48 (7.3%) 45 (3.8%)
ses, and by including interaction terms in the multilevel models. Deprivation score, 50 (7.6%) 83 (7.1%) .71
All statistical analyses were undertaken by using SAS (SAS Institute, n (%) high
Inc, Cary, NC) and Stata (StataCorp LP, College Station, Tex). Smokers, n (%) 163 (24.8%) 205 (17.6%) <.001
Ethnic group
White 572 (87.1%) 999 (85.8%) .53
RESULTS Black 21 (3.2%) 33 (2.8%)
Study population Mixed 18 (2.7%) 27 (2.3%)
South Asian 43 (6.5%) 91 (7.8%)
Of the 18 schools approached, 14 (78%) agreed to Chinese 1 (0.2%) 7 (0.6%)
participate, with an eligible population of 3295 students. Other 2 (0.3%) 7 (0.6%)
Ninety-seven students (3%) declined to take part. One Recent personal 189 (29.7%) 325 (28.6%) .64
school (167 students) was subsequently excluded because events, n (%)
they forgot to administer questionnaires on some exam- Season of birth
ination days. Students were included in the analysis if they Spring 160 (24.2%) 303 (25.9%) .48
completed the April questionnaire. Summer 179 (27.0%) 288 (24.6%)
Autumn 153 (23.1%) 295 (25.2%)
A total of 1834 students (57% of the available popula-

Health care education,

Winter 170 (25.7%) 286 (24.4%)

delivery, and quality

tion and 80% of participating individuals) from 13 schools
contributed complete information; 923 (50%) were girls
(Table I).
Case status
One hundred thirty (7%) students dropped at least
1 grade in mathematics, 281 (15%) in English, and 363
(20%) in science. This resulted in 662 (36%) cases and
1172 controls. Cases were more likely to be boys, to be
smokers, and to be at certain schools (Table I). Univariate
analyses failed to demonstrate any association between
deprivation score, ethnic group, recent personal events,
or season of birth and case status.
Allergic rhinitis symptoms and
medication use
Allergic rhinitis symptoms on different examination
days were reported by between 38% and 43% of students;
between 19% and 23% reported medication use (Table II).
Almost 2/3 (n 5 1001; 63%) of students reported allergic
rhinitis symptoms, and 542 (37%) reported medication use
at least once during the examination period. A total of 84
of 1425 (6%) students took sedative medication at some
point during the examination period. Daily pollen counts
varied according to the examination day (Table II).
Students who were smokers and those who reported
recent personal events that might have affected their
examination performance were more likely to report
allergic rhinitis symptoms, use of allergic rhinitis medi-
cation, and use of sedative medication during the exam-
ination period (Table III). Season of birth was related to
reporting of (sedative) medication use but not symptoms.
There were no associations between deprivation and re-
ported allergic rhinitis symptoms or medication use.
Reported allergic rhinitis during examinations was
higher for those who reported taking allergic rhinitis
medication in the April questionnaire (Table III); average
symptom scores were also higher (data not shown).
Allergic rhinitis symptoms during the examination were
also more likely to be reported by students with diagnosed
allergic rhinitis (94% vs 53%; P < .001) and with diag-
nosed asthma (77% vs 59%; P < .001). Average symptom
scores were significantly higher for those who had a diag-
nosis of allergic rhinitis (median, 4.83; range, 0-10; vs
median, 0; range, 0-9.83; P < .001) or asthma (median, 1.83;
range, 0-10; vs median, 0.17; range, 0-9.83; P < .001).
Relationship among allergic rhinitis,
medication use, and case/control status
Results from the adjusted multilevel models revealed
that cases were more likely than controls to have had
allergic rhinitis symptoms (odds ratio [OR], 1.43; 95% CI,
 384 Walker et al J ALLERGY CLIN IMMUNOL
AUGUST 2007

TABLE II. Allergic rhinitis symptoms, medication use, and pollen counts over the examination period

Date Examination Allergic rhinitis symptoms Medication use Pollen count (grains/m3)

07/06/2004 Science 1 (n 5 1769) 724 (40.9%) 335 (19.0%) 205

08/06/2004 Maths 1 (n 5 1737) 742 (42.7%) 357 (20.6%) 306
10/06/2004 English 1 (n 5 1569) 594 (37.9%) 307 (19.6%) 83
14/06/2004 Science 2 (n 5 1667) 712 (42.7%) 357 (21.5%) 357
14/06/2004 English 2 (n 5 1632) 705 (43.2%) 370 (22.7%) 357
15/06/2004 Maths 2 (n 5 1649) 700 (42.4%) 374 (22.8%) 66
22/06/2004 Science 3 (n 5 1118) 464 (41.5%) 221 (19.9%) 100
Any (n 5 1583) 1001 (63.2%) 542 (36.9%)

TABLE III. Associations between demographic and allergic factors with reported symptoms or medication use during
the examination period

Allergic rhinitis Medication use during Sedating antihistamine

during examinations examinations use during examinations
Demographic factors N (%) P value N (%) P value N (%) P value

Sex .13 .47 .89

Female 528 (65.0%) 285 (37.8%) 44 (6.0%)
Male 473 (61.4%) 257 (35.9%) 40 (5.8%)
Deprivation score
Low 926 (63.2%) .94 501 (36.9%) .99 80 (6.1%) .30
High 75 (63.6%) 41 (36.9%) 4 (3.7%)
Smoking
No 768 (61.2%) .001 417 (35.7%) .07 66 (5.5%) .23
Yes 229 (71.6%) 121 (41.4%) 21 (7.4%)
Ethnic group
White 848 (62.7%) .12 455 (36.2%) .64 71 (5.8%) .87
Black 38 (77.6%) 21 (47.7%) 3 (7.3%)
Mixed 29 (72.5%) 13 (36.1%) 3 (8.3%)
South Asian 71 (61.2%) 42 (39.3%) 7 (6.8%)
Chinese 5 (71.4%) 3 (42.9%) 0
Health care education,
delivery, and quality

Other 3 (37.5%) 2 (25.0%) 0

Recent personal events <.001 .004 .02
Absent 637 (59.4%) 343 (34.4%) 48 (5.0%)
Present 328 (71.6%) 180 (42.5%) 34 (8.2%)
Season of birth
Spring 261 (65.7%) .33 149 (40.0%) .01 23 (6.4%) .05
Summer 261 (64.4%) 157 (41.6%) 26 (7.1%)
Autumn 245 (63.0%) 127 (35.4%) 25 (7.1%)
Winter 234 (60.0%) 109 (30.2%) 10 (2.9%)
Allergic history
Diagnosed allergic rhinitis
Absent 628 (53.0%) <.001 229 (21.0%) <.001 22 (2.1%) <.001
Present 366 (94.1%) 308 (82.8%) 62 (17.7%)
Diagnosed asthma
Absent 685 (58.5%) <.001 354 (32.8%) <.001 45 (4.3%) <.001
Present 308 (77.0%) 182 (48.2%) 37 (10.2%)
Previous allergic rhinitis
Absent 241 (34.6%) <.001 71 (10.9%) <.001 1 (0.2%) <.001
Present 750 (86.0%) 466 (58.0%) 81 (10.5%)
Previous allergic rhinitis
medication use
Absent 603 (52.3%) <.001 206 (19.5%) <.001 5 (0.5%) <.001
Present 380 (96.5%) 329 (85.9%) 79 (22.2%)

1.13-1.81; P 5 .002), to have taken any allergic rhinitis Cases were also significantly more likely to have a diagno-
medication (OR, 1.36; 95% CI, 1.08-1.73; P 5 .01), or sis of asthma (Table IV).
to have taken sedating antihistamines on any examination Repetition of the adjusted analyses among those
day (OR, 1.71; 95% CI, 1.06-2.75; P 5 .03; Table IV). students who had reported physician-diagnosed allergic
J ALLERGY CLIN IMMUNOL Walker et al 385
VOLUME 120, NUMBER 2

TABLE IV. Relationship between self-reported allergic rhinitis, allergic rhinitis medication, and asthma
and case/control status

Case Controls Crude odds ratio Adjusted odds

(n 5 662) (n 5 1172) (95% CI) P value ratio* (95% CI) P value

Allergic rhinitis symptoms during 385 (67.3%) 616 (60.9%) 1.32 (1.06-1.64) .01 1.43 (1.13-1.81) .002
examinations, n (%)
Allergic rhinitis medication use 221 (41.1%) 321 (34.4%) 1.33 (1.07-1.65) .01 1.36 (1.08-1.73) .01
during examinations, n (%)
Sedating antihistamine medication 40 (7.7%) 44 (4.9%) 1.64 (1.05-2.55) .03 1.71 (1.06-2.75) .03
use during examinations, n (%)
Mean symptom score, n (%)
0 271 (41.4%) 549 (47.2%) 1.00 .04 1.00 .02
>0-2 121 (18.5%) 221 (19.0%) 1.11 (0.85-1.45) 1.27 (0.96-1.68)
>2-5 148 (22.6%) 228 (19.6%) 1.32 (1.02-1.69) 1.39 (1.06-1.82)
>5 115 (17.6%) 165 (14.2%) 1.41 (1.07-1.87) 1.48 (1.09-1.99)
Diagnosed allergic rhinitis, n (%) 147 (22.3%) 252 (21.6%) 1.04 (0.83-1.31) .73 1.09 (0.86-1.40) .47
Diagnosed asthma, n (%) 185 (28.1%) 265 (22.8%) 1.32 (1.06-1.65) .01 1.31 (1.04-1.66) .02
Allergic rhinitis symptoms (ever), n (%) 326 (49.8%) 606 (52.2%) 0.91 (0.75-1.10) .31 0.96 (0.78-1.17) .67
Medication use (ever), n (%) 150 (23.3%) 251 (21.9%) 1.08 (0.86-1.36) .50 1.14 (0.89-1.45) .30

*Odds ratios adjusted by sex and smoking history, with school as a nested cluster.

rhinitis in the spring questionnaire (n 5 1010) generated
similar results; for any allergic rhinitis symptoms, OR, 1.86
(95% CI, 1.18-2.95; P 5 .01); any allergic rhinitis medica-
tion, OR, 1.42 (95% CI, 1.03-1.96; P 5 .03); and any sedat-
ing medication, OR, 1.91 (95% CI, 1.15-3.18; P 5 .01).
Dose-response relationship
Within the whole group, there was a dose-response
relationship with cases reporting higher average symptom
reported a history of allergic rhinitis in previous years
and who had symptoms on any examination day of > 2 (on
a 10-point Likert scale) were more than twice as likely to
drop a grade. These associations are clinically and statis-
tically significant, biologically plausible, and remained
largely unaffected by adjustment for a range of important
confounding exposures.
The number of children taking sedating antihistamines
in this study was high (28% of those who reported taking

Health care education,

scores (median, 0.67; range, 0-9.83; vs median, 0.33; any allergic rhinitis medication). The only antihistamine

delivery, and quality

range, 0-10; P 5 .01; Table IV). This exposure-response
association was stronger when restricted to the 1010 stu-
dents who had reported allergic rhinitis in the spring ques-
tionnaire (OR, 1.64; 95% CI, 1.01-2.68; OR, 2.07; 95%
CI, 1.33-3.24; OR, 2.07; 95% CI, 1.31-3.26; P 5 .004).
Specificity of the association
There was, using multilevel modeling, no relationship
between improvement in grades and any exposure index:
any allergic rhinitis symptoms, adjusted OR, 0.96; 95%
CI, 0.72-1.28; P 5 .78; any allergic rhinitis medication
use, OR, 0.89; 95% CI, 0.66-1.19; P 5 .43; any sedative
medication use, OR, 0.86; 95% CI, 0.46-1.59; P 5 .62.
DISCUSSION
In this study, young people with reported allergic
rhinitis symptoms on an examination day were, in com-
parison with their fellow students without symptoms, 40%
more likely to drop a grade between their practice and final
GCSE examinations, and 70% more likely to drop a grade
if they reported taking sedating antihistamines at the time
of their examinations. The relationship between symp-
toms and poor examination performance was specific and
exhibited a dose-response relationship. Students who
we defined as being sedating was chlorpheniramine. This
was based on a number of studies investigating the
sedative properties of antihistamines that showed signif-
icant sedative effects of first-generation but limited sedat-
ing effects of second-generation antihistamines.11,15-17
The number of children taking sedating antihistamines
in this study was high in spite of current guidelines18
that recommend treatment with nonsedating prepara-
tions.19 The significant effect of sedating medications on
examination performance observed in this study should
encourage prescribers to recommend the use of nonsedat-
ing alternatives in routine practice.
Although sedation is an important issue when consid-
ering the choice of antihistamine treatment, first-generation
antihistamines have also been associated with significant
effects on learning capacity. Treatment with diphenhy-
dramine (a first-generation antihistamine with similar
sedative properties to chlorpheniramine17) resulted in
substantial adverse effects on attention span, working
memory, vigilance, and speed, as well as higher levels of
fatigue, lower levels of motivation, and lower levels of ac-
tivity compared with placebo.20 The effects of antihista-
mines are compounded by the fact that allergic rhinitis
itself has been shown to interrupt sleep21 and decrease
learning capacity in children,22,23 causing weakness,
malaise, irritability, fatigue, headache, and anorexia.24
 386 Walker et al
The systemic effects of allergic rhinitis may result in di-
minished functional capacity, which may be responsible
for the 2 million school days that are lost each year to al-
lergic rhinitis24 and for the increasing direct and indirect
costs ($2.3 billion in 1996) of managing symptoms.25
The study was adequately powered, and we achieved a
high response rate. We made attempts to minimize the risk
of information (recall) bias by measuring symptomatic
allergic rhinitis immediately before each examination.
Case status was established without knowledge of whether
a student had allergic rhinitis.
Potential limitations include the lack of formal vali-
dation—when used in this context—of the method we
used to measure the presence and severity of allergic
rhinitis symptoms, reporting bias, and the generalizability
of findings. Summer allergic rhinitis symptoms may not
always be the result of pollen exposure, and although the
question, ‘‘Do you have any of the following allergic
rhinitis symptoms of sneezing, itchy eyes, a runny or
blocked nose today?’’ has been validated elsewhere,1 we
used it in isolation of other supporting questions because
of time and space limitations of the questionnaires admin-
istered on examination days. Thus, students’ responses
may not have been specific. For this reason, we repeated
the analysis in those who reported a previous diagnosis
of allergic rhinitis. True allergic rhinitis is likely to affect
learning ability during the whole examination period (with
persistent symptoms occurring over an extended period)
rather than just a single day; this may explain the greater
impact of allergic rhinitis on examination performance
in those who report previous allergic rhinitis. Because of
this seasonal effect of allergic rhinitis, efforts to control
Health care education,
delivery, and quality
symptoms prophylactically and throughout the entire
grass pollen season are likely to be more effective than
attempting to manage symptoms on a daily basis.
For the purposes of the analysis, we assumed that
students who did not complete the medication question
on the questionnaire were not using treatment on that
day, even if they reported having symptoms. There was,
however, a possibility that students had in fact taken
medication, but not answered that particular question. To
investigate this possibility, we undertook a sensitivity
analysis and found that repeating the analysis after
removing those symptomatic students who did not com-
plete the medication questions (n 5 101) had negligible
effects on the estimates (OR for any seasonal allergic
rhinitis symptoms for the whole sample, 1.43; 95% CI,
1.13-1.81; OR for any seasonal allergic rhinitis symptoms
confining analysis to those completing the medication
question, 1.44; 95% CI, 1.14-1.82). We also assumed,
again because of time and space limitations, that those
who did not report symptoms were not using medication
and that those previously reporting using sedating med-
ication were still doing so.
Seasonal allergic rhinitis is a common condition across
the economically developed and developing world,1 and it
is important to consider the generalizability of our find-
ings. We studied a predominantly middle class population
of children from large urban and rural schools in the UK,
J ALLERGY CLIN IMMUNOL
AUGUST 2007
which may not be representative of other populations
globally. However, we see no obvious reason why our
findings could not be expected in any country in which
students of this age take examinations during a period of
high allergen exposure. Certainly they are consistent
with studies of the effects of allergic rhinitis and allergic
rhinitis treatments on simulated examinations,5,6 although
this is the first time that the effect has been measured on
actual examination performance.
We anticipated there would be other factors influencing
poor performance in the final examinations: sex, concom-
itant asthma, smoking status, ethnic group, other long-
term illness, and disruptive life events. Of these, only male
sex, smoking, and school were associated with case status.
Male sex is known to affect educational performance at
certain ages26; we suspect that the effect of smoking is in-
direct and explained by its association with other behav-
iors or demographic variables.27
The effect of school on examination performance is
understandable given the inherent variability of teachers
and teaching methods between schools. We did not
replicate previous research that has shown an effect of
ethnic group on educational performance,26 nor did we
find an effect of social deprivation, although there was
little variability in this measure in our population.
This work has substantially progressed understanding
of the social effect of seasonal allergic rhinitis in this
population. Important questions that remain unanswered
include whether intensive case management of sympto-
matic children leading up to the examination season or
shifting of examinations to a time that does not coincide
with the peak pollen season can remove the bias operating
against those with allergic rhinitis.
We thank the staff and students of the schools for their time and
effort in participating in the study, and Mrs Beryl Lockwood at
Warwick Local Education Authority for her help in implementing it.
We also thank the staff at Education for Health, Warwick, for their
encouragement and support, and particularly Kim Esslemont for her
help with the study administration and data collection. We thank the
anonymous reviewers for their helpful comments and suggestions.
REFERENCES
1. Strachan DP, Sibbald B, Weiland SK, Ait-Khaled N, Anabwani G, Ander-
son HR, et al. Worldwide variations in prevalence of symptoms of allergic
rhinoconjunctivitis in children: the International Study of Asthma and Al-
lergies in Childhood (ISAAC). Pediatr Allergy Immunol 1997;8:161-76.
2. Burr ML, Butland BK, King S, Vaughan-Williams E. Changes in asthma
prevalence: two surveys 15 years apart. Arch Dis Child 1989;64:1452-6.
3. Frossard N, Melac M, Benabdesselam O, Pauli G. Consistency of the re-
activity of cetirizine and ebastine on skin reactivity. Ann Allergy Asthma
Immunol 1998;80:61-5.
4. Juniper EF, Guyatt GH. Development and testing of a new measure of
health status for clinical trials in rhinoconjunctivitis. Clin Exp Allergy
1990;21:77-83.
5. Vuurman EPF, van Veggel LMA, Uiterwijk MMC, Leutner D, O’Han-
lon JF. Seasonal allergic rhinitis and anti-histamine effects on childrens’
learning. Ann Allergy 1993;71:121-6.
6. Vuurman EPF, van Veggel LMA, Sanders RL, Muntjewerff ND, O’Han-
lon JF. Effects of semprex-D and diphenhydramine on learning in young
adults with seasonal allergic rhinitis. Ann Allergy Asthma Immunol
1996;76:247-52.
J ALLERGY CLIN IMMUNOL
VOLUME 120, NUMBER 2
7. Emberlin J, Mullins J, Corden J, Jones S, Millington W, Brooke M, et al.
Regional variations in grass pollen seasons in the UK, long-term trends
and forecast models. Clin Exp Allergy 1999;29:347-56.
8. Warner JO. Seasonal rhinoconjunctivitis and examinations: is this unfair
discrimination? Paed Allergy Immunol 2000;11:129-30.
9. Indices of deprivation 2000. Available at: http://www.statistics.gov.uk/
StatBase/Product.asp?vlnk59421&Pos51&ColRank52&Rank5272. Ac-
cessed May 30, 2007.
10. UK Census 2001. Available at: http://www.statistics.gov.uk/census2001/
pdfs/key_statistics_final.pdf. Accessed May 30, 2007.
11. Harvey RP, Comer C, Sanders B, Westley R, Marsh W, Shapiro H, et al.
Model for outcomes assessment of antihistamine use for seasonal allergic
rhinitis. J Allergy Clin Immunol 1996;97:1233-41.
12. European Parliament. Official journal L 121, 01/05/2001 P.0034-0044.
2001. Available at: http://europa.eu.int/eur-lex/lex/LexUriServ/LexUriServ.
do?uri5CELEX:32001L0020:EN:HTML. Accessed March 22, 2007.
13. Office of Public Sector Information. Data protection act. London: Sta-
tionary Office Ltd; 1998.
14. A guide to trapping pollen and spores. British Aerobiology Federation; 1995.
15. Ramaekers JG, Uiterwijk MM, O’Hanlon JF. Effects of loratadine and
cetirizine on actual driving and psychometric test performance, and
EEG during driving. Eur J Clin Pharmacol 1992;42:363-9.
16. Mann RD, Pearce GL, Dunn N, Shakir S. Sedation with ‘‘non-sedating’’
antihistamines: four prescription-event monitoring studies in general
practice. BMJ 2000;320:1184-6.
17. Yanai K, Ryu JH, Watanabe T, Iwata R, Ido T, Sawai Y, et al. Histamine
H1 receptor occupancy in human brains after single oral doses of hista-
mine H1 antagonists measured by positron emission tomography. Br J
Pharmacol 1995;116:1649-55.
Walker et al 387
18. Bousquet J, Van Cauwenberge P, Khaltaev N. Allergic rhinitis and its
impact on asthma. J Allergy Clin Immunol 2001;108:s147-333.
19. Simons FER. Drug therapy: advances in H1-antihistamines. N Engl J
Med 2004;351:2203-17.
20. Kay GG, Berman B, Mockoviak SH, Morris CE, Reeves D, Starbuck V,
et al. Initial and steady-state effects of diphenhydramine and loratadine
on sedation, cognition, mood, and psychomotor performance. Arch In-
tern Med 1997;157:2350-6.
21. Lavie P, Gertner R, Zomer J, Podoshin L. Breathing disorders in sleep
associated with ‘‘microarousals’’ in patients with allergic rhinitis. Acta
Otolaryngol 1981;92:529-33.
22. Storms WW. Treatment of allergic rhinitis: effects of allergic rhinitis
and antihistamines on performance. Allergy Asthma Proc 1997;18:
59-61.
23. Stokes J, Fenstad E, Casale TB. Managing impairment in patients with
allergic rhinitis. Allergy Asthma Proc 2006;27:12-6.
24. American Academy of Allergy Asthma and Immunology. Task force on
allergic disorders: promoting best practice: raising the standard of care
for patients with allergic disorders. Milwaukee (WI): American Academy
of Allergy, Asthma & Immunology; 1998.
25. Blaiss M. Current concepts and therapeutic strategies for allergic rhinitis
in school-age children. Clin Ther 2004;26:1876-89.
26. Department for Education and Skills. Youth cohort study: the
activities and experiences of 16 year olds: England and Wales 2004.
Available at: http://www.dfes.gov.uk/rsgateway/DB/SFR/s000560/index.
shtml. Accessed March 22, 2007.
27. Gibbons S. Geography, resources and primary school performance.
London: London School of Economics and Political Science;
2002.
Health care education,
delivery, and quality