HUMAN GENOME EPIDEMIOLOGY (HuGE) REVIEW
Systematic Review and Meta-Analysis of the Association between
b
2
-Adrenoceptor Polymorphisms and Asthma: A HuGE Review
Ammarin Thakkinstian
1
, Mark McEvoy
2
, Cosetta Minelli
3
, Peter Gibson
4
, Bob Hancox
5
,
David Duffy
6
, John Thompson
3
, Ian Hall
7
, Joel Kaufman
8
, Ting-fan Leung
9
, Peter Joseph Helms
10
,
Hakon Hakonarson
11
, Eva Halpi
11
, Ruth Navon
12
, and John Attia
2
1
Clinical EpidemiologyUnit, Faculty ofMedicine,Ramathibodi
Hospital, Mahidol University, Bangkok, Thailand.
2
Centre for Clinical Epidemiology and Biostatistics, University
of Newcastle, Newcastle, New South Wales, Australia.
3
Centre for Biostatistics and Genetic Epidemiology,
Department of Health Sciences, University of Leicester,
Leicester, United Kingdom.
4
Department of Respiratory and Sleep Medicine, Hunter
Medical Research Institute, John Hunter Hospital, Newcastle,
New South Wales, Australia.
5
Department of Respiratory Medicine, Waikato Hospital,
Hamilton, New Zealand.
6
Queensland Institute of Medical Research, Brisbane,
Queensland, Australia.
7
Division of Therapeutics and Molecular Medicine, University
Hospital, Nottingham, United Kingdom.
8
Department of Environmental and Occupational
Health Sciences, School of Public Health and
Community Medicine, University of Washington,
Seattle, WA.
9
Department of Paediatrics, The Chinese University of
Hong Kong, Hong Kong, Special Administrative Region,
China.
10
Department of Child Health, Medical School, University of
Aberdeen, Aberdeen, United Kingdom.
11
deCODE Genetics, Inc., Reykjavik, Iceland.
12
Department of Genetics, Sackler Faculty of Medicine, Tel
Aviv University, Tel Aviv, Israel.
Received for publication January 17, 2005; accepted for publication March 14, 2005.
A number of studies have investigated two common polymorphisms in the b
2
-adrenoceptor gene, Arg/Gly16 and
Gln/Glu27, in relation to asthma susceptibility. The authors performed a meta-analysis of each polymorphism, as
well as haplotype analysis, for adult and pediatric populations separately, using published data, supplemented by
additional data requested from the original authors. Individual analysis detected no effect of Arg/Gly16 in adults but
did suggest a recessive protective effect of Gly16 for children, with an odds ratio of 0.71 (95% confidence interval
(CI): 0.53, 0.96) compared with the other genotypes. Results for Gln/Glu27 in adults seem to indicate that hetero-
zygotes are at decreased risk of asthma than either homozygote (odds ratio ¼ 0.73, 95% CI: 0.62, 0.87), although
the studies are heterogeneous; in children, the Glu/Glu genotype has a decreased risk of asthma (odds ratio ¼
0.60, 95%CI: 0.35, 0.99) compared with the other genotypes. Despite the proximity of these two polymorphic sites,
the linkage disequilibrium coefficient of 0.41 was not high (p < 0.001). Haplotype analysis suggests that there may
be an interaction between the two sites, with a lower risk of asthma associated with theGlu27 allele (compared with
Gln27), and that this risk is modified by the allele at position 16.
asthma; epidemiology; genetics; haplotypes; linkage disequilibrium; meta-analysis; polymorphism, genetic;
receptors, adrenergic
Abbreviations: CI, confidence interval; LR, likelihood ratio; OR, odds ratio; SNP, single-nucleotide polymorphism.
Editor’s note: This paper is also available on the website
of the Human Genome Epidemiology Network (http://
www.cdc.gov/genomics/hugenet/).
One of the main thrusts of genetic epidemiology is to
understand the genetic contribution to complex diseases
such as cardiac disease, diabetes, and asthma. One of the
Correspondence to Dr. John Attia, Centre for Clinical Epidemiology and Biostatistics, University of Newcastle, Newcastle, NSW 2300,
Australia (e-mail: John.Attia@newcastle.edu.au).
201 Am J Epidemiol 2005;162:201–211
American Journal of Epidemiology
Copyright ª 2005 by the Johns Hopkins Bloomberg School of Public Health
All rights reserved
Vol. 162, No. 3
Printed in U.S.A.
DOI: 10.1093/aje/kwi184

most popular study designs in this area is a molecular asso-
ciation study in which a polymorphism is linked to the
disease outcome, either in cases and controls or in a cohort.
These studies are often limited by small sample sizes (1), so
there is a role for meta-analysis in pooling these studies,
particularly to detect the small effect sizes that may be
associated with these polymorphisms.
The b
2
-adrenoceptor gene is a key gene to study in
asthma. b
2
-Adrenoceptors are present on many airway cells,
including smooth muscle cells which are hyperreactive in
asthma, and b
2
-adrenoceptor agonists form a major treat-
ment class in asthma. Functional polymorphisms of this
gene may influence both disease susceptibility and treat-
ment response in asthma.
A number of studies have investigated polymorphisms in
the b
2
-adrenoceptor gene in relation to asthma. Two com-
mon polymorphisms are Arg/Gly16 and Gln/Glu27; in the
former polymorphism, glycine is substituted for arginine at
codon 16 (Arg16/Gly) and, in the latter, glutamic acid is
substituted for glutamine at codon 27 (Gln27/Glu) (2, 3).
Invitro studies indicate that theGly16 allele enhances agonist-
induced down regulation of the receptor, whereas the
Glu27 allele enhances resistance to down regulation (4, 5).
It is plausible that these differences in receptor regulation
influence the reactivity of airway smooth muscle in response
to airway inflammation and thereby alter the risk of asthma.
However, epidemiologic studies have yielded conflicting
results, with the direction of the effects not always congruent
with the in vitro results. Several narrative reviews of these
two polymorphisms and asthma (4–6) have been conducted;
however, neither a magnitude nor a mode of gene effect was
provided in these reviews. Furthermore, new studies that
examine this association have been reported since those
reviews, and there have been new developments in the
methodology of meta-analysis of genetic studies (1, 7, 8).
We therefore performed a systematic review of the associa-
tion between Arg/Gly16 and Gln/Glu27 and asthma with the
following objectives: first, to estimate allele frequencies;
second, to ascertain if there is an effect of these polymor-
phisms on asthma susceptibility, and if so to estimate the
magnitude of that effect and the possible mode of inheri-
tance (1, 7, 8); third, to determine linkage disequilibrium
between these two polymorphisms; and fourth, to infer hap-
lotypes of these polymorphisms and link them with asthma
susceptibility.
MATERIALS AND METHODS
Search strategy
Embase and Medline databases (from January 1966 to
March 2004) were searched using the Embase, PubMed,
and Ovid search engines. The search strategy for allele fre-
quency was as follows: beta2* AND prevalence AND gene.
The search strategy for association between gene polymor-
phisms and asthma was the following: asthma AND (beta
receptor or beta-2 or adrenoceptor) AND (polymorph* or
mutation* or variant* or genotype*). Searching was per-
formed in duplicate by two independent reviewers (A. T.
and M. M.).
Inclusion criteria
For allele frequency, any human studies that estimated the
prevalence of b
2
-adrenoceptor polymorphisms at codon
16 (Arg/Gly16) and/or codon 27 (Gln/Glu27) and reported
on ethnically homogeneous populations were included, re-
gardless of size. For assessing association, human studies,
regardless of sample size, were included if they met the
following criteria:

b
2
-Adrenoceptor polymorphisms at codon 16 (Arg/Gly16)
and/or codon 27 (Gln/Glu27) were determined. The wild-
type alleles for these two polymorphisms were Arg and
Gln,respectively.

The outcome was asthma (incident or prevalent), and
there were at least two comparison groups, for example,
asthma versus control (nonasthma) groups.

Participants could be either children or adults, but results
should be reported separately.

There were sufficient results for extraction of data, that
is, number of subjects for each genotype in asthma and
control groups. Where eligible papers had insufficient
information, we contacted authors by e-mail for addi-
tional information.
The reference lists of the articles retrieved were also re-
viewed to identify publications on the same topic. The most
complete and recent results were used where there were
multiple publications from the same study group.
Data extraction
Data were extracted independently and in duplicate by
two reviewers (A. T. and M. M.) who used a standardized
data extraction form. Any disagreement was adjudicated by
a third author (J. A.). Covariables, such as mean age, gender,
and ethnicity, were also extracted for each study.
Quality score assessment
The quality of studies was also independently assessed by
the same two reviewers who used quality assessment scores
that were modified from our previous meta-analysis of mo-
lecular association studies (7) (appendix table 1). These
scores were based on both traditional epidemiologic consid-
erations and genetic issues (1). Total scores ranged from
0 (worst) to 13 (best).
Statistical analysis
Data analyses were performed as follows. First, the fre-
quency of Arg16 and Gln27 alleles in various ethnic groups
was estimated by the inverse variance method, as described
in the Appendix.
Second, estimation of the gene effect on asthma was
performed by a logistic regression approach described pre-
viously (8). In brief, the steps were as follows. Hardy-
Weinberg equilibrium was assessed for each study by use of
the v
2
test or Fisher’s exact test, where appropriate, and only
in control groups. A Q test for heterogeneity was performed
separately for three odds ratios (ORs), that is,Gly/Gly versus
202 Thakkinstian et al.
Am J Epidemiol 2005;162:201–211