Genetic Epidemiology of Asthma · Genetic Epidemiology of Asthma David L Duffy INTRODUCTION Asthma is an increasingly common disease in the developed world (1-3). In the United States,

Epidemiologic ReviewsCopyright O 1997 by The Johns Hopkins University School of Hygiene and Public HealthAll rights reserved

Vol. 19, No. 1Printed In USA.

Genetic Epidemiology of Asthma

David L Duffy

INTRODUCTION

Asthma is an increasingly common disease in thedeveloped world (1-3). In the United States, the sec-ond National Health and Nutrition Examination Sur-vey (NHANES II) (1976-1980) estimates of asthmaprevalence (4) were 6.9 percent for whites and 9.2percent for blacks. Most current definitions of asthma(5) characterize it as a (rapidly or slowly) reversiblenarrowing of the conducting airways of the lungs,usually associated with evidence of airway wall in-flammation and eosinophilia (6, 7). Acute and delayedbronchoconstriction can be precipitated by inhalationby the asthmatic individual of a number of agents,such as ozone, cold dry air, proteins (mainly en-zymes), or low molecular weight chemicals, and pre-sents clinically as development (and resolution) ofdyspnea, wheezing, and cough.

Asthma is clinically heterogenous, but the bulk ofcases tend to be characterized by childhood onset,markers of allergic hypersensitiveness such as wheez-ing and bronchoconstriction following inhalation ofantigen (allergen), and elevated levels of total serumimmunoglobulin E. There may be remission of symp-toms for up to 20 years after childhood, so that earliermild episodes may be forgotten by the individual,though not by the parent (8, 9). Most genetic studieshave concentrated on this type of asthma (i.e., allergicor extrinsic asthma). However, a few studies haveattempted to determine if different genes might beacting in the 20 percent of asthmatics who do notexhibit any signs of allergy (i.e., nonallergic or intrin-sic asthma).

Received for publication August 30, 1996, and accepted forpublication April 15, 1997.

Abbreviations: Cl, confidence interval; LOD score, decimal loglikelihood ratio; OR, odds ratio; 6 , recombination fraction betweentwo genetic loci.

From the Genetic Epidemiology Division, Department of Epide-miology, The Johns Hopkins University School of Hygiene andPublic Health, Baltimore, MD; and the Epidemiology and PopulationHealth Unit, Queensland Institute of Medical Research, Brisbane,Australia.

Reprint requests to Dr. D. L Duffy, Epidemiology Unit, Queens-land Institute of Medical Research, 300 Herston Road, Herston,Brisbane, Queensland 4029, Australia.

Because of the intermittent nature of asthma symp-toms, some difficulties in standardizing the diagnosisof asthma for epidemiologic purposes have been en-countered. Currently, there is increasing emphasis onbiologic markers of disease, such as nonspecific bron-chial responsiveness or sputum eosinophilia, to con-firm the diagnosis.

There has been long-standing interest in the geneticsof other phenotypes that tend to be associated withasthma. These include the other two diseases of the"atopic triad," allergic rhinitis and atopic dermatitis,and the immunologic markers of atopy such as serumimmunoglobulin E level (both total and allergen-specific), skin prick test response to allergen, and,more recently, T-cell receptor repertoires. Indeed, farmore genetic analyses of total serum immunoglobulinE concentration have been performed than of asthma.Several family studies of nonspecific bronchial re-sponsiveness have also been carried out.

General considerations

There are several questions a genetic epidemiologicmodel of asthma and atopy must address:

1. The nature of the interaction between environ-mental and genetic risk factors. Studies in Africa(10) and Oceania (11) suggest that asthma and otheratopic diseases increase markedly with Westerniza-tion, possibly in concert with a fall in mean total serumimmunoglobulin E concentration (12, 13). The criticalexposure has been variously hypothesized as being tohigh levels of potent aeroallergens, such as the housedust mite (14, 15), to increased dietary sodium (thoughan effect on allergic sensitization is less plausible)(16), to exposure (17) or lack of exposure to pathogens(18), or to air pollution (19). Nevertheless, the familystudies reviewed below suggest that an individualmust be genetically predisposed. to respond to theexposure.

2. The selection pressures that kept genes predis-posing to disease common in the preindusthal envi-ronment. One hypothesis, that the atopic phenotypemay protect against parasitic disease, has been previ-ously reviewed (20). The low prevalence of allergicdisease in less industrially developed societies sug-

129

130 Duffy

gests that the costs of carrying such genes in thepopulation may not have been high.

3. 777e clinical heterogeneity of asthma. It is attrac-tive to hypothesize that genetic heterogeneity mayunderlie this condition. As reviewed below, geneticstudies to date have supported both "lumping" and"splitting." A related concept is one of genetic heter-ogeneity of atopic disease, such that specific genes(rather than specific environmental exposures) mightdetermine why individuals might develop asthma asopposed to hay fever or atopic dermatitis.

FAMILY STUDIES

Earlier family studies

Familial aggregation of asthma was recognized his-torically (21, 22). Salter, in his classic text (1860),states "Is Asthma Hereditary?—I think that there canbe no doubt that it is . . .the number of cases in whichthere is a family history of asthma is greater than willbe found. . .on the mere doctrine of chance. . . . Out ofthirty-five cases. . .1 find distinct traces of inheritancein fourteen.. .two cases out of every five" (23, p. 109).

Following the rise of Mendelian genetics, RobertCooke performed two large studies of the inheritanceof atopy, one reported in 1916 (24) and the other in1924 (25). The first study examined asthma, hay fever,urticaria, angioneurotic edema, and acute gastroenter-itis in 504 subjects. In the second study, only asthmaand hay fever were used to define atopy in a further462 individuals. All of these subjects exhibited posi-tive intradermal tests. A control series of 115 non-atopic probands was also recruited. Family history ofatopy was determined by interviewing the proband"and in many cases, other members of the family" (25,p. 522). This is an early example of a "Weinbergproband-control" study, a case-control study compar-ing risk of disease to relatives of a case to that of acontrol. A family history of atopy was present in 48.4percent of cases in the 1916 study and in 58.4 percentof cases in the 1924 study. Only 7 percent of the 115normals reported a family history of atopy (crude oddsratio (OR) = 19; 95 percent confidence interval (CI)9-39).

Cooke used these findings to buttress his classicdefinition of atopy as "[a] sub-group of hypersensi-tiveness.. .restricted to the hay-fever and asthmagroup [of diseases]. . .inherited, subject to a dominantgen[e]. . ." (26, pp. 166, 168).

The numerous family studies of atopy performed inthe 1920s and 1930s generally replicated the finding offamilial aggregation, but often found a simple domi-nant hypothesis untenable. Wiener et al. (21), for ex-ample, proposed a plausible codominant (or additive)

model, with a decreasing age at onset with increasingnumber of disease alleles.

The large Weinberg study of Schwartz (27), per-formed in the 1940s, examined both idiopathic asthma(191 subjects) and a particular occupational asthma(baker's asthma). These subjects were compared with200 controls (polio patients and medical students).Schwartz interviewed 2,352 relatives and obtainedquestionnaire data from a further 1,463. The asthmat-ics underwent examination of peripheral blood, in-cluding eosinophil count, and allergen skin testing.These, along with the history, were used to dichoto-mize the asthma cases into atopic and nonatopic.Atopic relatives and a small number of willing healthyrelatives also underwent skin atopy testing and eosin-ophil count.

Schwartz concluded that asthma and atopic diseaseaggregated within families, and that some pedigreeswere consistent with a dominant hypothesis. A criticalfinding was that there was no difference in risk ofatopic disease between the relatives of allergic andnonallergic asthmatics. Similarly, no differences werefound for the prevalence of skin atopy or eosinophiliain these two groups in either unaffected relatives orthose manifesting atopic disease. For asthmatic rela-tives of allergic and nonallergic asthmatic probands,for example, the prevalences of a positive allergenskin test were 20/30 and 29/39, respectively (p = 0.5).Finally, the probands with baker's asthma and theirfamilies were not found to differ significantly from theother asthmatic series, thus confirming that a geneticpredisposition was necessary to develop this occupa-tional disease (gene by environment interaction).

The studies discussed thus far demonstrate familialcorrelations between asthma, hay fever, and atopicdermatitis. Although such coaggregation was con-firmed in the 361 families examined by Schnyder (28),an important additional finding was that particularatopic diseases tended to "breed true," with childhoodatopic dermatitis being found more frequently in fam-ilies of dermatitis probands than respiratory allergyprobands. Similar patterns have been seen in morerecent family (13, 29-32) and twin (33) studies. Theinheritance of atopy in families "strongly supported]the hypothesis of a single, autosomal dominant genewith reduced penetrance (40-50%)" (28, p. 90).

Sibbald et al. (34) described familial risks for 77asthmatic children (404 relatives) and 87 controls (302relatives) recruited via a single general practice. Allprobands and approximately 25 percent of their rela-tives underwent skin prick testing. Diagnosis in pro-bands was reached via clinical examination and inrelatives by interview. This allowed asthma in all the

Epidemiol Rev Vol. 19, No. 1, 1997

Genetic Epidemiology of Asthma 131

probands and a subset of relatives to be divided intoallergic and nonallergic types.

The prevalence of asthma in the relatives of atopicasthmatics was slightly higher than that among rela-tives of nonatopic asthmatics. The authors also notedthat "in relatives of asthmatics, the prevalence ofatopic asthma exceeded the prevalence of nonatopicasthma, irrespective of the atopie status of the pro-band" (34, p. 672), but this difference is of marginalsignificance (x*2

= 4.88, p = 0.09, compared withdistribution in control families). Positive skin testswere more common in relatives of atopic asthmaticscompared with all controls. Closer examination re-veals that this is due largely to the relatives of thenonatopic controls, where the prevalence is signifi-cantly lower (42 percent) than that for the relatives ofthe atopic asthmatics (60 percent). The difference be-tween risk to relatives of atopic and nonatopic asth-matics (36 percent) is almost significant (p = 0.07),but the small number of nonatopic asthmatics in thestudy prevents more being made of this. The authorsconcluded that asthma and atopy are inherited inde-pendently, but atopy increases the risk of expressingthe asthmatic trait.

Recent studies

A large number of recent epidemiologic studieshave collected information on family history of asthmaas a risk factor for disease. Most, however, are not assophisticated as the studies reviewed above, in termsof the depth of information collected on relatives, andmerely confirm the magnitude of the parent-offspringor, less often, sib-sib recurrence risk as being approx-imately twofold larger than the population baselinerisk. However, a few complex segregation analyseshave been published.

Complex segregation analysis uses maximum like-lihood methods, implemented in computer programssuch as POINTER (Population Genetics Laboratory,University of Hawaii, Honolulu, HI), PAP (PedigreeAnalysis Package) (Department of Human Genetics,University of Utah, Salt Lake City, ITT), and SAGE(Statistical Analysis for Genetic Epidemiology) (De-partment of Epidemiology and Biostatistics, CaseWestern Reserve University, Cleveland, OH), to testthe hypothesis that aggregation of a trait in a familycould be due to the action of a single major locus. Thealternative model that the single major locus is oftencompared with is a polygenic model, where multiplegenes, each of small effect, underlie familial resem-blances. For a dichotomous trait, the usual polygenicmodel is the multifactorial threshold model, whichassumes multiple genes of small effect act multiplica-tively on risk (act additively on the probit of risk). A

"mixed model" posits a single major locus acting inconcert with a polygenic background. The heritabilityis an index often referred to when discussing thesemodels and represents the proportion of populationtrait variance estimated as due to genetic effects (oftenon the probit scale). In conventional family studies,but less in the classical twin or adoption study, it isdifficult to differentiate the effects of polygenes fromthose of the unmeasured shared family environment.

Several studies of nonspecific bronchial responsive-ness, that is to inhaled methacholine or histamine inrelatives of asthmatic probands, were performed in themid 1980s. The interest in these is that increasednonspecific bronchial responsiveness in the absence ofclinical asthma or allergic disease might indicate thatthe relative is carrying an asthma predisposing geno-type.

The largest of these studies, by Townley et al. (35),described 51 families (467 individuals) ascertainedthrough an asthmatic, 32 (291 individuals) familiesascertained through a nonatopic control, and in 26cases for a negative family history of allergic disease.Methacholine responsiveness was found to be bimod-ally distributed in the asthma families, a finding thatmight be consistent with the action of a major gene (oran all-or-nothing trait or disease state). However, seg-regation analysis strongly rejected a single major locusmodel, although familial aggregation was supported.A mixed model was not fitted however. Smaller stud-ies (36-40) also replicated the bimodal distribution ofnonspecific bronchial responsiveness, and demon-strated asymptomatic nonspecific bronchial respon-siveness to be increased in relatives of asthmatics.Most recently, in the British families described below(41), the heritability of nonspecific bronchial respon-siveness (log PD20 (provocative dose of agent thatcaused a 20 percent fall in lung volume) to histamine)was estimated at 27 percent (95 percent CI 13-40),very similar to that seen in the Busselton, WesternAustralia, study (log PD20 to methacholine) at 28percent (95 percent CI 17-39) (42).

There have been many recent studies of total serumimmunoglobulin E concentration (42-58). Most ofthese conclude that immunoglobulin E level is highlyheritable (60-70 percent), more so than nonspecificbronchial responsiveness, and single major locus mod-els with high allele frequencies often fit the pattern offamilial transmission. Borecki et al. (59) described abivariate segregation analysis of atopic disease (asth-ma or eczema) and total serum immunoglobulin E.The 173 Saskatchewan families originally describedby Gerrard et al. (45) were used, in which the preva-lence of so-defined atopy was 21 percent. This samplewas unselected for a history of atopic disease. A single


132 Duffy

major locus model was found to best fit the data, withan estimated allele frequency of 42 percent, the geneacting in either an additive (intermediate) or a reces-sive fashion for liability to clinical disease, and as arecessive for serum immunoglobulin E. No unsharedenvironmental factor unique to atopy (as opposed toone common to total serum immunoglobulin E con-centration) was detected.

Mrazek et al. (60) performed segregation analysisusing POINTER on 145 families ascertained via afemale asthmatic proband. Significant familial aggre-gation was detected, and under the polygenic modelthe heritability was 0.96 (asymptotic standard error(ASE)) = 0.05). The best single major locus modelwas a common recessive model, with a homozygotepenetrance of 62 percent, and 1 percent for the othergenotypes, a gene frequency of 0.22, and a proportionof phenocopies estimated at 17 percent. This modelgives a population prevalence for asthma of 3.5 per-cent.

Lawrence et al. (41) reported a study of 131 families(631 completely phenotyped individuals) fromSouthampton (United Kingdom). Asthma was repre-sented as a combination (first principal component) ofa respiratory questionnaire score (including resultsfrom a video questionnaire) and level of bronchialresponsiveness to histamine inhalation. The pheno-typic correlation between asthma score and total serumimmunoglobulin E concentration was 0.4. Heritabilityof the resulting asthma score was approximately 28percent, while that for total serum immunoglobulin Econcentration was 61 percent.

Segregation analysis using POINTER was support-ive of a dominant gene for asthma, with no residualheritability (increasing allele frequency 0.24). Similarresults were obtained fitting a two-locus model (withone locus acting as a pseudopolygene) via the programCOMDS (61). The models fitted to bronchial hyper-responsiveness alone and wheezing (questionnaire)alone differed in having gene frequencies closer to 0.5,and acting in a recessive fashion.

A segregation analysis of physician-diagnosedasthma in the Tucson (Arizona) Children's Respira-tory Study was presented by Holberg et al. (62). Di-agnosis was available via questionnaire for a total of3,369 individuals in 906 nuclear families. Class Dregressive logistic segregation analyses were per-formed, with age at onset included in the penetrancefunction. Strong familial aggregation was confirmed.A single major locus model did not fit the data well,with the closest fitting Mendelian model being of acommon recessive gene explaining only part of theaggregation. Ethnicity and personal and passive smok-

ing did not contribute detectably to the residual cor-relation between family members.

These segregation analyses suggest that a majorgene could be involved in the etiology of asthma.However, different "best" genetic models were ob-tained in the studies. A similar finding for total serumimmunoglobulin E concentration has been interpretedas evidence for genetic heterogeneity, that is multiple,individually uncommon, genes of large effect. Themultivariate analyses are most consistent with the ac-tion of genes specific to particular forms of atopicdisease. Nonallergic asthma aggregates within fami-lies, but it is still unclear whether it is geneticallydistinct from allergic asthma.

TWIN STUDIES

The twin design has two main advantages over theconventional family studies described above: One isthe estimation of the effects of family environment,which in other designs are confounded with geneticeffects, and the second is the detection of nonadditivegenetic effects (63), either dominance or epistasis,which otherwise may require the collection of reliabledata on third-degree relatives (64). I will concentrateon the several large community-based twin studies ofasthma reported in the literature.

Edfors-Lubs (65) and Lubs (66) reported resultsfrom a questionnaire study of 7,000 pairs of twins bornbetween 1886 and 1925 from the Swedish Twin Reg-istry. The monozygotic recurrence risk for asthma was33.7 percent; that for the dizygotic twins was 8.6percent. The large monozygotic:dizygotic risk ratio isconsistent with strong dominance or epistasis. Underthe multifactorial threshold model, the heritability ofasthma in this study was 63 percent for asthma (95percent CI 54-73 percent).

A similar study was carried out in the Finnish TwinCohort, which contains 13,888 pairs of same-sex twinsborn between 1905 and 1957 and was assembled usingcentralized birth and death information (67). In thebaseline questionnaire survey, monozygotic twin re-currence risks for doctor-diagnosed asthma ranged inthe different sex and age groups from 32 percent to 87percent, and the dizygotic risks ranged from 0 percentto 24 percent. Refinement of the asthma diagnosis wasreported by Nieminen et al. (68) via linkage of thetwin register with databases on hospital admission andasthma medication usage. Using this definition, themonozygotic male and female recurrence risks were8.2 percent and 16.5 percent, respectively; for thedizygotic twins, 8.6 percent and 4.9 percent, respec-tively. The estimated heritability (multifactorialthreshold model) combining the sexes was 36 percent(with broad confidence limits) (69).



In unpublished preliminary data from the VirginiaTwin Registry (Eaves LJ, and coworkers, VirginiaInstitute for Psychiatric and Behavioral Genetics, Vir-ginia Commonwealth University, Richmond, Virginia,1990) for 4,310 twin-pairs (and 14,380 relatives),doctor-diagnosed asthma was reported by 5.5 percentof questionnaire respondents. Under the multifactorialthreshold model the heritability of asthma was 54percent.

Duffy et al. (33) describe a questionnaire survey ofthe Australian Twin Registry carried out in 1981.Responses were received from 3,808 twin-pairs. Forwheezing, the monozygotic recurrence risks were 54.9percent for monozygotic males, 42 percent formonozygotic females, and 21.8-24.0 percent for thethree dizygotic groups, with no significant birth cohorteffects. The monozygotic:dizygotic ratios are consis-tent with dominance in males but not in females; theheritability under the multifactorial threshold modelwas 60 percent for women and 75 percent for men.Multivariate genetic analyses of asthma, hay fever,and eczema replicated the finding of trait-specific ge-netic influences in addition to common genetic deter-minants (such as might underlie atopy).

In all these twin studies, the effects of family orhousehold environment are estimated to be small orabsent compared with the genetic contribution to fa-milial resemblance. The estimates of heritability areconsistent with those from other types of studies.

LINKAGE AND ASSOCIATION ANALYSES

The complexity of the immunologic network in-volved in the allergic response, and the great variety ofpathogenetic mechanisms hypothesized for asthma,mean that there are numerous plausible candidategenes for this disease. Many of these genes overlapwith other immunologic diseases. A number of suchcandidates have been examined, and more recentlyresults of whole genome scans have been reported.Two types of genometric studies have been used.AUelic association studies test for correlation betweena specific allele at a locus with the phenotype, such asasthma or nonspecific bronchial responsiveness. Thiscan be performed using unrelated cases and controls (aprocedure open to the possibility of confounding dueto ethnic difference), or family material. Genetic link-age analysis by contrast requires pedigree data, and isperformed either via maximum likelihood (decimallog likelihood ratio (lod score)) methods, orregression-based relative pair methods.

Early studies

The paper of Zieve et al. "On the linkage relations ofthe genes for allergic disease and the genes determin-

ing the blood groups, MN types and eye colour inman" (70), is the first of its kind. A sample of 66nuclear families (383 individuals) was ascertained.Diagnosis of asthma, hay fever, urticaria, eczema,angioneurotic edema, and gastroenteritis was made bythe investigators based on personal interview. Linkageanalysis carried out using the methods of the time didnot detect linkage to blood group or eye color. Reanal-ysis of the pedigrees using MLINK (71) and SIBPAL(Department of Epidemiology and Biostatistics, CaseWestern Reserve University, Cleveland, OH) actuallyfinds weak hints of linkage to the ABO blood group(mean identity by descent sharing among affectedsib-pairs is 0.61, p = 0.02).

No further linkage studies were reported in the lit-erature until 1985, when an abstract by Eiberg et al.(72) reported a maximum lod score of 2.07 betweentotal serum immunoglobulin E concentration and es-terase D (recombination fraction = 0.00, chromosome13ql4). Linkage to this chromosomal region was notreplicated by Cookson et al. (73) but was by Daniels etal. (74).

Linkage and association to the humanlymphocyte antigen region

The human lymphocyte antigen literature on asthmais quite extensive and inconclusive. Strong evidence ofassociation between human lymphocyte antigen hap-lotypes and allergic sensitization to particular aller-gens exists, most notably ragweed pollen (75, 76) andhouse dust mite feces (77, 78). However, the bulk ofstudies find no evidence for linkage of asthma to theregion with one exception (see table 1).

Caraballo and Hernandez (79) examined humanlymphocyte antigen segregation in 20 Columbian nu-clear families (n = 107) ascertained to contain twoatopic asthmatic probands. All probands were allergicto Dermatophagoides farinae on allergen skin pricktest. No associations between human lymphocyte an-tigen haplotypes and asthma were detected. In sib-pairanalyses, there was evidence of linkage to asthma (j?2

= 21.9, p = 1.8 X 10~5). In several family studiespublished subsequently, no linkage has been found(80-82).

Allelic association between human lymphocyte an-tigen class I and II genes and asthma have often beenreported (81-88). Because the alleles at the varioushuman lymphocyte antigen loci are so numerous, theseresults are often vitiated by the large number of sta-tistical tests required to detect the effect, and havebeen difficult to replicate (82, 88). A very recent paperby Moffatt and Cookson (89) has reported an associ-ation between the tumor necrosis factor-a gene (TNF,


134 Duffy

TABLE 1. Selected studies of linkage of asthma or atopio disease to the human leukocyte antigen (HLA) region

Studyandyear

(reference no.)Caraballo and Hernandez,

1990(79)Amelung et al., 1992(80)Rich eta]., 1992(81)HoDoway et al., 1996(82)

No. ofsubjects(no. ot

ta mites)107 (20)

117(20)67(3)

220(22)

Result Comment

Increased affected sib pair sharing HLA-B No association detected

Lod score* < -2.0 within 6 cM of D6S105 Families linked to chromosome 5Lod score < -2.0 within 23 cM of HLA-A.BNo affected sib pair sharing HLA-D No DRB1 *O4 or *07 association

* Lod score, decimal log likelihood ratio.

6p21.3) and asthma (see discussion of Animal modelsbelow).

Association with other chromosome 6 loci

Platelet activating factor is a potent bronchocon-stricting mediator. Miwa et al. (90) reported the resultsof biochemical analyses and a small family study ofthe enzyme platelet activating factor acetylhydrolase(recently mapped to chromosome 6p 12—21 (91)), oneof the two enzymes responsible for the degradationand resulting brief half-Life of platelet activating factorin serum. Enzyme activity was assayed in 816 adultsand 211 children. The frequency distribution of activ-ity in the children was definitely bimodal, while theadult results were unimodal. There were 40 healthysubjects found to exhibit no acetylhydrolase activity atall. These results were shown to be consistent with asingle major locus model.

Enzyme levels were next examined in 175 asthmaticchildren. Lower amounts of activity were associatedwith increasingly severe asthma. Complete absence ofactivity was found significantly more often in the mostsevere classes of asthma (asthmatics experiencing 10or more episodes of wheeze per year versus nonasth-matic controls (OR = 3.43, 95 percent CI 1.03-11.76)).

Recently, Stafforini et al. (91) have identified aVal279Phe mutation that is associated with plateletactivating factor acetylhydrolase deficiency present atan estimated frequency of 0.17, agreeing with theestimate of Miwa et al. (90). Interestingly, the muta-tion was not found in 108 US subjects. It seemspossible that this genetically determined enzyme defi-ciency may modulate the severity of symptoms expe-rienced in asthma, and the partial deficiency seemsquite common, at least in the Japanese population.

Linkage to chromosome 11q

The report of strong evidence for linkage of asthmaand atopy to the marker D11S97 (formerly A-MS51)on the long arm of chromosome 11 by Cookson et al.

(73) was the first of a number of studies examiningthis chromosomal region. In this initial report, sevenextended pedigrees were ascertained either through anatopic asthmatic attending a clinic or via media ap-peals and advertising seeking multiplex (i.e., multiplyaffected) families. The trait chosen for analysis was abroadly defined "atopy," characterized as either one ormore skin prick wheals with a mean diameter 1 mmgreater than the negative control, or a total serumimmunoglobulin E more than two standard deviationsabove the population mean, or one or more positiveallergen-specific immunoglobulin E assay (radio-allergosorbent test) results. Four families contributed amaximum lod score of 5.24 at a recombination frac-tion of 6 percent. All the families gave a lod score of6.39 at a recombination fraction of 10 percent—theother families, where cigarette use was high, were lessinformative.

The same group (92) described an analysis of 64newly recruited nuclear families. These families wereascertained through children (under the age of 15years) reporting eczema, hay fever, or asthma. Thesame diagnostic criteria for atopy described abovewere used. Under an assumption of equal recombina-tion distance in both sexes, the maximum lod scorewas 3.8, but there was strong evidence for a male-female difference. Relaxing the equality constraint ledto a maximum lod score of 5.2, with 6 M = 0.18 and0 F = 0.001 (where 0 stands for recombination frac-tion between two genetic loci).

In an effort to explain uiis sex difference, Cooksonet al. (93) described sib-pair reanalyses that examinedsharing of maternal and paternal marker alleles (n =723). Affected sib-pairs were significantly more likelyto share a maternal chromosome 11 than a paternalchromosome 11, using any of several different defini-tions of affection. In families where the fattier wasatopic and die mother nonatopic, the sharing of ma-ternal alleles was 19/32 (0.59 percent, exact binomial95 percent CI 0.41-0.76), which trends in the samedirection as the previous results. The authors note thatexplanations for this include genetic imprinting, or



phenotypic maternal effects, so a pair of sibs carryingthis atopy gene are more likely to express it aftertransplacental or breast milk exposure to maternalimmunoglobulin E or other factors. Since the presenceof atopy in offspring of atopic fathers is increased,they also concluded that atopy must be geneticallyheterogenous.

Of six linkage studies performed by other groupspublished up to 1992, only one supported linkage tochromosome 1 lq, but more recently, another two con-firmations have been published (see table 2). Morecritically, Sandford et al. (107) localized a strongcandidate gene to chromosome 1 lq, that for the high-affinity immunoglobulin E receptor /3-subunit. Poly-morphisms of the gene were associated with atopy andasthma (107a).

Association to chromosome 11q

Shirakawa et al. (107a) first sequenced the FceRI/3-subunit gene in six atopic and six nonatopic sub-jects. Three (6th exon) mutations were found in oneatopic individual leading to substitutions of leucine forisoleucine at position 181, and leucine for valine atposition 183. A polymerase chain reaction-based assayfor these two mutations was developed.

In a "random sample" of 163 patients unselected forallergic disease (undergoing venesection for other pur-poses), 25 were found to carry the Ilel81Leu mutation,

but none, the Vall83Leu. A total serum immunoglob-ulin E greater than 100 IU/ml was present in 41 (25percent), of whom 11 carried the Leul81 mutation(OR = 3.1, 95 percent CI 1.2-7.5). A similar associ-ation was found for the presence of grass pollen spe-cific immunoglobulin E (OR = 2.6, 95 percent CI1.1-6.4).

The Leul81 mutation was also found to segregate in10 of 60 of the atopic nuclear families. In each family,the mutation was transmitted from the mother (andwas present in the proband). Among 14 nonprobandoffspring, four were atopic and two carried Leu 181;none of the 10 nonatopic offspring carried the muta-tion. Furthermore, the two sporadics arose in bilinealatopy—that is the father was atopic and did not carryLeul81.

In the Busselton family study (108), 232 unselectednuclear families (1,020 individuals, 556 children) weretyped for the Leul81 mutation in the FceRI /3-subunitgene. This gene was found in 28 subjects. There wereeight children carrying the gene where the parent oforigin was the mother—three with asthma, the remain-ing five with hay fever. Specific immunoglobulin Eand skin prick test wheals to house dust mite (as wellas a sum of radio-allergosorbent test scores) weresignificantly higher than in controls on Wilcoxon test.Further analysis (109) found similar levels of associ-ation to another polymorphism (7th exon) in the same

TABLE 2. Stucfiee of linkage of atopic disease to chromosome region 11q

Studyandyear

(reference no.)

No. ofsubjects(no. of

turtles)

Decimal logIkelhood

ratio(tod score)

Comment

Cookson eta]., 1992(93)Mxingetal., 1992(92)Inacioetal., 1991 (94)Shirakawa et al., 1991,

1994 (95, 96)Amelung et al., 1992 (80)Hizawaetal., 1992(97)Lympany et al., 1992(98)Ffichetal., 1992(81)Coteman et al., 1992 (99)

Colleeetal., 1993(100)Brereton et al., 1994(101)van Herwerden et al.,

1995 (102)Duffy at ai., 1994(103)

Watson et al., 1995(104)Martinet'etal., 1996(105)

(7)281 (64)83(17)

(4)

117(20)60(4)89(9)67(3)

407(95)

52(26)(12)

246 (123)

424 (212)

560(131)213(45)

Neelyetal., 1996(106) 218(12)

6.4toD1iS97(10cM)5.2 to D11S97 (M 18 cM, F 0.1 cM)No linkage to D11S974.88toDi iS97

< -2.0 within 12 cM of FGF3< -2.0 wfthin 4 cM of D11S97< -2.0 within 4 cM of FGF3< -2.0 within 5 cM of D11S97-7.88 multipoint D11S97, PGYM, CD20

Increased affected sib pair sharing D11S97Negative lod scoreIncreased affected sib pair sharing

FceRI-pcaNo increased affected sib pair sharing

FcsRI-pcaNegative lod scores for three markersNo increased affected sib pair sharing

FceRI-pcaIncreased affected sib pair sharing (and

positive TDT) INT2, D11S1369

Selected pedigrees: used 4 of 136

Families linked to chromosome 5

12 families with affected mothers andunaffected fathers gave lod score of 0.8

Combined segregation-linkageUnable to detect Leu18i mutation


136 Duffy

gene, Gly237Glu. In this case, there was no evidenceof a maternal effect.

In a subsequent case-control study (110, 111), 500atopic patients (allergic (early and late onset) andnonallergic asthma, hay fever, or eczema) ascertainedvia Osaka hospital clinics were compared with 100controls attending a health examination company. Re-striction fragment length polymorphisms in FceRI/3(Rsa 1), CD20, and GIF (the latter are two neighbor-ing loci that might be alternative candidates) wereexamined. Significant differences in genotype fre-quencies were detected for the FceRI/3 polymorphism(alone) between the controls and several different sub-groups of the patients, most strongly for childhoodonset allergic asthma and not at all for the nonallergicasthma group. The Leul81 mutation was not detectedat all in the sample (111), but Gly237Glu was found in6 percent of controls, 8 percent of nonallergic asth-matics, and 18 percent of atopic asthmatics (OR =3.43, 95 percent CI 1.50-9.35).

Further evidence for association in the region wasdescribed by Doull et al. (112). As part of theSouthampton study described above, allelic effects oftwo markers were detected on different phenotypes:allele 168 of D11S527 with nonspecific bronchialresponsiveness {p = 0.0003, Bonferroni corrected for13 alleles/? = 0.004), and allele 235 of D11S534 withtotal immunoglobulin E (p = 0.007, Bonferroni cor-rected for 14 alleles p = 0.09).

In contrast, an Australian twin study (103) of 215dizygotic twin-pairs ascertained for wheezing did notdetect linkage of atopy (skin test or serum immuno-globulin E), nonspecific bronchial responsiveness, orserum immunoglobulin E to FceRI/3ca. No evidenceof the Leul81 mutation was found in the twins, theirparents, or a set of controls (n = 939), a findingconfirmed by sequencing in 19 most atopic individuals(113). Similar findings were reported in three otherstudies (105, 114, 115).

We can conclude that the Leul81 and Glu237 aretwo mutations in the FceRI /3-subunit associated withatopy. They were present in 3 and 5 percent, respec-tively, of the general population sample from Bussel-ton. The maternal effect seems to hold up for Leul81.

Linkage to chromosome 5

In 1994, Marsh et al. (55) reported a linkage be-tween total serum immunoglobulin E and markers inthe strong candidate region on chromosome 5q31 con-taining the interleukin gene cluster, /3-adrenoceptor,and granulocyte-macrophage colony stimulating factorgenes. This study was carried out in a sample of 11 oldorder Amish families (subsequently extended to 12)unselected for clinical atopy (though containing at

least one member with detectable allergen specificimmunoglobulin E).

This finding was swiftly confirmed by Meyers et al.(116) in 92 Dutch families (538 individuals), all as-certained through a hospital-diagnosed asthmatic par-ent (in the case of nuclear families) originally studiedin 1962 to 1970. Linkage to serum immunoglobulin Eand nonspecific bronchial responsiveness (117) wasdetected both by sib-pair and lod score analyses.

Blumenthal et al. (118) reported an absence of link-age of chromosome 5q markers to serum immunoglob-ulin E level in their four large atopic families (110typed individuals). Maximum likelihood linkage anal-ysis (under a common dominant gene model) andHaseman-Elston analysis using SIBPAL were per-formed. The maximum lod score among 12 markersspanning chromosome 5 was 0.06 (0 = 0.23), withlod scores of -2 .3 at 0 = 0 for IL-9, and -0 .4 at 0= 0 for IL4-R1. Similarly, the best p value fromsib-pair analysis was 0.29. Linkage to the region wasalso not detected in the genome scan of Daniels et al.(74) (see below).

Association to chromosome 5

Weak evidence of allelic association for IL9 andtotal serum immunoglobulin E concentration has beenpresented by Doull et al. (112) and Borish et al. (119).A polymorphism in the distal IL4 promoter region wasassociated with wheezing (p = 0.03) and presence ofhouse dust mite-specific immunoglobulin E (p =0.01), but not several other related phenotypes, wasreported by Walley and Cookson (120). Interest is nowmore focussed on the proximal promoter region.

Szentivanyi (121) is usually cited as the first workerto propose the unifying hypothesis that asthma repre-sents an underresponsiveness of the lungs to sympa-thetic neurotransmitters. He argued that the defect inasthma must be "nonimmunologic," and suggested the/3-adrenoceptor gene as the most likely site for this. Anumber of recent studies have looked at functionalmutations in /3-adrenoceptor gene (chromosome5q31-32) in asthmatics (122-127). Although the twomost common mutations do not seem to be signifi-cantly more common in asthmatics than nonasthmat-ics, three studies have detected effects in particularasthma subgroups. Reishaus et al. (123) found thatsevere asthmatics, requiring oral steroids or immuno-therapy, were more Likely (75 percent) to be Argl6Glyhomozygotes than controls (59 percent). Turki et al.(124), compared 23 nocturnal asthmatics (who tendedto have a lower mean diurnal forced expiratory volumein one second (FEVj), and to be more likely to besteroid dependent) to 23 "normal" asthmatics. TheArgl6Gly frequency was 80.4 percent in the nocturnal



group and 50.0 percent in the nonnoctumal group(p = 0.004). Hall et al. (125) found Gln27Glu ho-mozygotes to be more responsive to inhaled metha-choline than Ghi27 homozygotes (p - 0.03). How-ever, they did not detect any such effect for Glyl6.Functional studies (126) suggest that mutations doalter receptor kinetics in clinically meaningful ways.

Linkage to chromosome 12

Gamma interferon (chromosome 12ql3) plays animportant role in T-cell regulation that makes it an-other good candidate gene for involvement in allergicdisease. The first published examination of this regionwas that of Watson et al. (104). The maximal lod scorefor IGF1 (tightly linked to D12S318 and PAH andapproximately 30 cM distal to gamma interferon) was0.09 at 25 percent recombination distance.

Barnes et al. (128) have reported evidence of link-age and association between asthma and total serumimmunoglobuUn E concentration over a region stretch-ing from close to IGF1 up to gamma interferon in twodifferent populations, the Amish families originallyused to detect linkage to chromosome 5q and a sampleof 29 Barbadian pedigrees (693 individuals) ascer-tained through a proband with a history of asthma.

Asthmatic sib-pairs exhibited increased ibd sharingat D12S379 (61.8 percent), D12S95 (67.2 percent),PAH (58.5 percent), and D12S360 (58.6 percent).Similarly, Haseman-Elston regression analyses weresignificant for PAH and D12S360. In the Amish sam-ple, where a diagnosis of asthma was not available onany family members, there were significant Haseman-Elston results for D12S360, IGF1, as well as PLA2.Several replications of this finding have recently beenpresented (129).

Chromosome 14

Linkage between the T-cell receptor aS-subunitgenes and atopy was reported by Moffatt et al. (130).Families containing at least two atopic siblings wereascertained in Busselton (413 subjects), as well as inthe United Kingdom (410 subjects). Mean ibd sharingwas increased for T-cell receptor a-subunit alleles insib-pairs concordant for several different phenotypesincluding high immunoglobulin E level (p = 0.002),house dust mite (p = 2 X 10~4), cat {p = 6 X 10~5),and timothy grass (p = 0.02) allergen sensitization.No such linkage was seen for a T-cell receptor /3-subunit polymorphism.

There has been one interesting case report (131) ofa highly atopic individual with multigene deletions inthe heavy chain immunoglobulin gene region, but thisis not a common association (132). Oxelius et al. (133,

134), have examined Gm type (a electrophoreticallydetectable polymorphism in the heavy chain constantregion) in atopic children and controls. Particular phe-notypes were associated with atopy and with elevatedimminuglobulin G4 subclass levels.

a-1-antitrypsin

Complete and partial deficiencies of a-1-antitrypsin(a-l-AT, chromosome 14q31.2 (135)) are associatedwith obstructive lung disease (136, 137). The commonnormal allele is M; those associated with deficiencystates are the S and Z alleles. The Pi7.7, phenotype(phenotype associated with the Z/Z genotype) is themost severely affected (by panacinar emphysema andcirrhosis), while the PiSZ and PiSS phenotypes are atincreased risk of lung disease, especially followingexposures such as cigarette smoking. The disease as-sociation with PiMS and PiMZ is smaller and is notseen in all studies.

Fagerhol and Hauge (138) have reported high pro-portions of PiMS and PiSS subjects among asthmapatients, while Hyde et al. (139) also found that PiMSand MZ asthmatic children required more intense drugtreatment than control PiMM asthmatics. Buist et al.(140) found an excess of asthma in heterozygotes in amatched 2:1 case-control study (3 of 21 cases and 2 of42 controls).

Townley et al. (141) have recently described resultsof a-1-antitrypsin phenotyping of 723 subjects fromthe Natural History of Asthma study. These includedthe families of asthmatic probands and families ascer-tained for a three-generation absence of atopic disease.Asthmatics made up 36 percent of the PiMS group and21 percent of the PiMM group (p = 0.04). The PiMSgroup also had significantly lower levels of nonspe-cific bronchial responsiveness, and removing the cur-rent and former asthmatics did not alter this finding.Serum immunoglobulin E level was the same for allthree Pi phenotypes.

Gaillard et al. (142) reported similar findings forsubgroupings of the MM phenotype. The M2M2 phe-notype was found to be more frequent in 90 asthmaticsthan in 240 controls. Asthmatics were also found tohave higher plasma levels of a-1-antitrypsin, butlower levels of elastase inhibitory capacity. Thesefindings are suggestive, especially when one notes thatelastase inhibitory capacity level and the elastase in-hibitory capacity/a-1-antitrypsin level ratio (a measureof molar efficiency of the different Pi types) are lowestfor PiM2M2 of the MM subtypes, with the exceptionof the M2M3 subtype (143).

In Nigerian asthmatics, Awotedu and Adelaja (144)found 74 of 99 asthmatics to be MM phenotype com-pared with 98 of 100 controls. The MZ phenotype was


138 Duffy

present in 19 asthmatics and one control. A similarstudy of Puerto Rican asthmatics (145) in New Yorkfound 41 of 55 nonsmoking asthmatics to be MM, and49 of 61 controls—not a significant difference.

Lindmark has performed a number of studies ofa-1-antichymotrypsin, deficiencies of which lead to aclinical syndrome similar to that associated with a-1-antitrypsin deficiency (146). In a recent study (147),Lindmark screened 12 women heterozygous for a-1-antichymotrypsin and their relatives for asthma andallergic rhinitis, comparing them with a group of 58controls. The index cases were three times more likelyto report asthma (95 percent CI for OR 1.05-9.80) butnot hay fever. Relatives with decreased levels of a-1-antichymotrypsin (n = 15) were also more likely toreport asthma (OR = 3.1; 95 percent CI 0.96-9.83).Since this deficiency occurs in only 0.5-1 percent ofthe Swedish population, it cannot be a major determi-nant of asthma.

Linkage to other chromosomal regions

The first genome scan to be published (74) was oftwo subsamples from the Busselton (80 families, 364subjects) and Oxford, England (77 families) studyfamilies, with the Oxford families being used to rep-licate any positive findings in the first panel. Thechromosome 11 findings were, of course, replicated.Different regions of the genome gave positive resultsfor differing phenotypes: Nonspecific bronchial re-sponsiveness was linked to chromosomes 4, 7, and 16;total serum immunoglobulin E concentration to chro-mosomes 6, 7, and 16; eosinophil count to chromo-somes 6 and 7; and atopy (combining skin tests andimmunoglobulin E) to chromosomes 6 and 13. Repli-cation in the second panel was of atopy to chromo-some 13 and of asthma to chromosome 16. Interest-ingly, maternal linkage was stronger than paternallinkage for, in addition to FceRI/3, markers on chro-mosomes 4 and 16.

Other genetic associations

The relation between several other electrophoreti-cally or serologically typeable polymorphisms havebeen investigated. Ronchetti et al. (148) described astudy of asthma and polymorphisms in the adenosinedeaminase gene (ADA, chromosome 20ql3). Adeno-sine is a bronchoconstrictor on inhalation, and thedifferent ADA gene isozymes differ in enzymatic ac-tivity. The ADA gene 1-1 phenorype was significantlymore common among 347 asthmatic children thanamong controls (OR = 1.6, 95 percent CI 1.45-1.7).Other unreplicated case-control studies have foundassociations with the angiotensin converting enzyme

I/D polymorphism (149), haptoglobin (150), and com-plement C3 type (151).

Disease associations

A number of studies have observed the co-occurrence of allergic disease with other nonallergicdiseases. Such findings, if reproducible, might answerquestions about the evolution and persistence of genespredisposing to allergic disease in the population. Inaddition, if the mechanism is a genetic correlation, itsuggests further possible candidate genes.

1. The oldest reported association is a lower thanexpected number of individuals suffering from bothasthma and juvenile-onset diabetes melhtus (152).This was not replicated in one recent smaller study(153).

2. Asthma has been reported to be variously in-creased or decreased in a number of cancers, bothrespiratory and nonrespiratory (154-156). The corre-lation between asthma and increased lung cancerseems to be the most replicable (157) but is unlikely tobe genetic—one would suspect an effect of chronicinflammation.

3. The genetic disease familial Mediterranean fever(chromosome 16pl3) has been reported to be protec-tive against asthma (158, 159). One study (160) hassuggested that heterozygotes also have a decreasedincidence (OR = 0.5, 95 percent CI 0.1-2.1).

4. Similar claims have been made recently for theheterozygote carriers of the common AF508 mutationof the cystic fibrosis gene (CFTR, chromosome 7q37)(161).

ANIMAL MODELS

As in other complex diseases, a number of usefulanimal models of asthma have been developed. Theuse of standard crosses and recombinant inbred linesgreatly increases the power of linkage methods todetect genes of small effect, especially in the case ofquantitative traits associated with asthma. The traitscurrently most intensely studied in mice have beenbronchial responsiveness to methacholine (nonspecificbronchial responsiveness) (162), ovalbumin or sheeperythrocytes (allergic), mediators such as platelet ac-tivating factor (163), and chemical irritants such asozone (164). Such analyses have detected the presenceof at least four significant loci, the effects of someoverlapping over the different challenges, the loca-tions of which correspond to good human candidategene regions.

De Sanctis et al. (162) recently described such alinkage analysis in 321 mice from high nonspecificbronchial responsiveness (A/J strain) with a low non-



specific bronchial responsiveness (C57BL/6J) strainbackcrosses. Three loci were detected, explaining 26percent of the genetic interstrain variance (backcrossheritability was 50 percent). The mouse candidategenes in these regions were tumor necrosis factoralpha (TNF, human chromosome 6p), interleukin (IL2,IL3) receptors (chromosome 22q), and platelet derivedgrowth factor B chain (chromosome 22q). The A/Jstrain develops increased nonspecific bronchial re-sponsiveness and airway inflammation followingozone exposure (164), which is attenuated by anti-tumor necrosis factor antibody pretreatment (165).Ewart et al. (165) added a fourth locus, possibly IL5(chromosome 5q), in another study of nonspecificbronchial responsiveness in the A/J strain. This latterfinding may be regarded as further support for thehuman studies of chromosome 5q (55, 116, 117), asmay the report of another mouse locus syntenic tochromosome 5q31.1 modifying the ThlATh2 responses(166).

CONCLUSIONS

The segregation, linkage, and association studiesreviewed above suggest a substantial genetic contri-bution to the familial aggregation of asthma and aller-gic disease, almost certainly involving the action ofmultiple common genes. Several genome scans forasthma close to publication also detect linkage tomultiple regions.

One can divide the genes detected to date into"minor" genes (where the disease allele is uncommonor the effect size is small, so that the proportion ofcases attributable is small) and major genes, explain-ing a larger proportion of population variation. Exam-ples of minor genes might include the a-l-AT and/3-adrenoceptor polymorphisms.

A finding common to other complex diseases is theoften inconsistent results of linkage and associationstudies of asthma. In the case of chromosome 1 lq, forinstance, the positive studies have come from mark-edly different ethnic backgrounds. For every positivestudy however, there have been one or more negativestudies in the same ethnic groups. Several hypotheseshave been suggested to explain these differences (167,168). These include the presence of nonstandard ge-netic mechanisms, such as imprinting, environmentalconfounding in the presence of gene by environmentinteraction, and genetic heterogeneity.

The usual use of the term genetic heterogeneityposits uncommon disease alleles present at the differ-ent loci, each of which is sufficient to cause disease inits own right in the appropriate environment. Thealternative polygenic model assumes that disease al-leles are relatively common (either high allele fre-

quency and low number of involved loci, or viceversa), effect sizes associated with any one locus rel-atively small, and gene effects interact in a simplefashion (e.g., multiplicatively in the multifactorialthreshold model). In this case, an affected individualwill often carry disease alleles at multiple differentloci, and linkage analysis may be difficult. Such poly-genic effects seem to underlie nonspecific bronchialresponsiveness in the mouse models discussed above.

There are now sufficient positive studies to con-clude that a gene on chromosome llq, probably theFceRI /3-subunit, exerts some influence on total serumimmunoglobulin E concentration, nonspecific bron-chial responsiveness, and asthma. The inconsistencybetween studies means that an attributable risk due tothis single locus could be low or high. There is less(though increasing) evidence for the strong candidateregions on chromosomes 5q and 12q. The allelic as-sociations with the /3-adrenoceptor gene suggest it tobe more a modifier of asthma severity, and cannotexplain linkage to total serum immunoglobulin E con-centration level. Similarly, the human lymphocyte an-tigen allelic associations seem to modify response tospecific allergens, but have not consistently been im-plicated in the presence or absence of asthma, exceptpossibly for that due to particular occupational expo-sures (87). There is a lack of linkage results applicableto asthma subtypes, especially nonallergic asthma.

Finally, a large number of environmental determi-nants of asthma are recognized, and must be invokedto explain its increasing incidence in developed coun-tries, and among migrants to these countries. Theseenvironmental factors will interact with multiple genesin a complex fashion, as discussed elsewhere in thisissue of Epidemiologic Reviews.

ACKNOWLEDGMENTS

The author is an Australian National Health and MedicalResearch Council Neil Hamilton Fairley Fellow.

REFERENCES

1. Burr ML, Butland BK, King S, et al. Changes in asthmaprevalence: two surveys 15 years apart Arch Dis Child1989;64:1452-6.

2. Bumey PGJ, Chinn S, Rona RJ. Has the prevalence ofasthma increased in children? Evidence from the nationalstudy of health and growth 1973-86. BMJ 1990;300:1306-10.

3. Whincup PH, Cook DG, Strachan DP, et al. Time trends inrespiratory symptoms in childhood over a 24 year period.Arch Dis Child 1993,68:729-34.

4. Turkeltaub PC, Gergen PJ. Prevalence of upper and lowerrespiratory conditions in the US population by social andenvironmental factors: data from the second National Health


140 Duffy

and Nutrition Examination Survey, 1976 to 1980 (NHANESII). Ann Allergy 1991;67:147-154.

5. Standards for the diagnosis and care of patients with chronicobstructive pulmonary disease (COPD) and asthma. Thisofficial statement of the American Thoracic Society wasadopted by the ATS Board of Directors, November 1986.Am Rev Respir Dis 1987;136:225-44.

6. Bousquet J, Chanez P, Lacoste JY, et al. Eosinophilic in-flammation in asthma. N Engl J Med 1990;323:1033-9.

7. Djukanovic R, Wilson JW, Britten KM, et al. Quantitation ofmast cells and eosinophils in the bronchial mucosa of symp-tomatic atopic asthmatics and healthy control subjects usingimmunohistochemistry. Am Rev Respir Dis 1990; 142:863-71.

8. Jenkins MA, Hopper JL, Bowes G, et al. Factors in childhoodas predictors of asthma in adult life. BMJ 1994;309:90-3.

9. Strachan DP, Anderson HR, Bland JM, et al. Asthma as alink between chest illness in childhood and chronic coughand phlegm in young adults. Br Med J (Clin Res Ed) 1988;296:890-3.

10. Godfrey RC. Asthma and IgE levels in rural and urbancommunities of The Gambia. Clin Allergy 1975;5:201-7.

11. Waite DA, Eyles EF, Tonkin SL, et al. Asthma prevalence inTokelauan children in two environments. Clin Allergy 1980;10:71-5.

12. Emanuel MB. Hay fever, a post industrial revolutionepidemic: a history of its growth during the 19th century.Clin Allergy 1988;18:295-304.

13. Gerrard JW, Geddes CA, Reggin PL, et al. Serum IgE levelsin white and Metis communities in Saskatchewan. Ann Al-lergy 1976;37:91-100.

14. Voorhoost R, Spieksma FTM, Varekamp H. House-dustatopy and the house-dust mite Dermatophagoides pteronys-sinus (Trouessart 1897). Leiden, The Netherlands: StafleuScientific Publication Company, 1969.

15. Sporik R, Holgate ST, Platt-Mills TAE, et al. Exposure tohouse-dust mite allergen (Der p T) and the development ofasthma in childhood: a prospective study. N Engl J Med1990;323:502-7.

16. Burney PG. The causes of asthma—does salt potentiatebronchial activity? Discussion paper. J R Soc Med 1987;80:364-7.

17. Busse WW. Respiratory infections: their role in airway re-sponsiveness and the pathogenesis of asthma. J Allergy ClinImmunol 1990;85:671-83.

18. Shirakawa T, Enomoto T, Shimazu S, et al. The inverseassociation between tuberculin responses and atopic disor-der. Science 1997;275:77-9.

19. Lebowitz MD. Epidemiological studies of the respiratoryeffects of air pollution. Eur Respir J 1996;9:1029-54.

20. Masters S, Barrett-Connor E. Parasites and asthma—predictive or protective? Epidemiol Rev 1985;7:49-58.

21. Wiener AS, Zieve I, Fries JH. The inheritance of allergicdisease. Ann Eugen 1936;7:141-62.

22. Sakula A. A history of asthma. The FitzPatrick lecture 1987.J R Coll Physicians Lond 1988;22:36-44.

23. Salter HH. On asthma: its pathology and treatment. London,England: John Churchill, 1860.

24. Cooke RA, Vander Veer A Jr. Human sensitization. J Im-munol 1916;l:201-305.

25. Spain WC, Cooke RA. Studies in specific hypersensitive-ness. XI. The familial occurrence of hay fever and bronchialasthma. J Immunol 1924;9:521-69.

26. Coca AF, Cooke RA. On the classification of the phenomenaof hypersensitiveness. J Immunol 1923;8:163-82.

27. Schwartz M. Heredity in bronchial asthma. Acta Allergol1952;5(suppl II): 1-288.

28. Schnyder von UW. Neurodermitis, asthma-rhinitis. Eine ge-netisch-allergologische Studie. (In German). Int Arch Al-lergy Appl Immunol 1960;17(suppl):l-106.

29. von Mutius E, Nicolai T. Familial aggregation of asthma ina South Bavarian population. Am J Respir Crit Care Med

1996;153:1266-72.30. Fergusson DM, Horwood LJ, Shannon FT. Parental asthma,

parental eczema and asthma and eczema in early childhood.J Chronic Dis 1983;36:517-24.

31. Dold S, Wjst M, von Mutius E, et al. Genetic risk for asthma,allergic rhinitis, and atopic dermatitis. Arch Dis Child 1992;67:1018-22.

32. Aberg N. Familial occurrence of atopic disease: geneticversus environmental factors. Clin Exp Allergy 1993;23:829-34.

33. Duffy DL, Martin NG, Battistutta D, et al. Genetics ofasthma and hay fever in Australian twins. Am Rev RespirDis 1990;142:1351-8.

34. Sibbald B, Horn MEC, Brain EA, et al. Genetic factors inchildhood asthma. Thorax 1980;35:67t-4.

35. Townley RG, Bewtra A, Wilson AF, et al. Segregationanalysis of bronchial response to methacholine inhalationchallenge in families with and without asthma. J Allergy ClinImmunol 1986;77:101-7.

36. Verity CM, Vanheule B, Carswell F, et al. Bronchial labilityand skin reactivity in siblings of asthmatic children. Arch DisChild 1984;59:871-6.

37. Clifford RD, Pugsley A, Radford M, et al. Symptoms, atopy,and bronchial response to methacholine in parents withasthma and their children. Arch Dis Child 1987;62:66-73.

38. Longo G, Strinati R, Poli F, et al. Genetic factors in nonspe-cific bronchial hyperreactivity: an epidemiologic study. Am JDis Child 1987;141:331-4.

39. Hopp RJ, Bewtra AK, Biven R, et al. Bronchial reactivitypattern in nonasthmatic parents of asthmatics. Ann Allergy1988;61:184-6.

40. Bruderman I, Cohen R, Schachter J. Bronchial response tomethacholine in "healthy" children of asthmatic parents:effect of treatment with cromolyn sodium. Chest 1990;97:285-7.

41. Lawrence S, Beasley R, Doull I, et al. Genetic analysis ofatopy and asthma as quantitative traits and ordered poly-chotomies. Ann Hum Genet 1994;58:359-68.

42. Palmer LJ, Burton PR, Gurrin LC, et al. Variance compo-nents modelling of quantitative traits related to the asthmaphenotype in a population-based sample of nuclear families.(Abstract). Genet Epidemiol 1996;13:304.

43. Marsh DG, Bias WB, Ishizaka K. Genetic control of basalserum immunoglobulin E level and its effect on specificreaginic sensitivity. Proc Natl Acad Sci U S A 1974;71:3588-92.

44. Turner KJ, Rosman DL, O'Mahony J. Prevalence and famil-ial association of atopic disease and its relationship to serumIgE levels in 1,061 school children and their families. IntArch Allergy Appl Immunol 1974;47:650-64.

45. Gerrard JW, Rao DC, Morton NE. A genetic study of im-munoglobulin E. Am J Hum Genet 1978;3046-58.

46. Blumenthal MN, Namboodiri K, Mendell N, et al. Genetictransmission of serum IgE levels. Am J Med Genet 1981;10219-28.

47. Meyers DA, Bias WB, Marsh DG. A genetic study of totalIgE levels in the Amish. Hum Hered 1982;32:15-23.

48. Hasstedt SJ, Meyers DA, Marsh DG. Inheritance of immu-noglobulin E genetic model fitting. Am J Med Genet 1983;14:61-6.

49. Lebowitz MD, Barbee R, Burrows B. Family concordance ofIgE, atopy and disease. J Allergy Clin Immunol 1984;73:259-64.

50. Meyers DA, Freidhoff LR, Marsh DG. Predicting skin testsensitivity and total serum IgE levels in family members. JAllergy Clin Immunol 1986;77:608-15.

51. Meyers DA, Beaty TH, Freidhoff LR, et al. Inheritance oftotal serum IgE (basal levels) in man. Am J Hum Genet1987;41:51-62.

52. Silverman EK, Province MA, Campbell EJ, et al. Biochem-ical intermediates in a,-antitrypsin deficiency: residual fam-ily resemblance for total a^-antitrypsin, oxidized a,-anti-



trypsin, and immunoglobulin E after adjustment for the effectof the Pi locus. Genet Epidemiol 1990;7:137-49.

53. Parks T, Felix K, Rice T, et al. A genetic study of immuno-globulin E and atopic disease based on families ascertainedthrough asthmatic children. Hum Hered 1990;40:69-76.

54. Meyers DA, Beaty TH, Colyer CR, et al. Genetics of totalserum IgE levels: a regressive model approach to segregationanalysis. Genet Epidemiol 1991;8:351-9.

55. Marsh DG, Neely JD, Breazeale DR, et al. Linkage analysisof IL4 and other chromosome 5q31.1 markers and totalserum immunoglobulin E concentrations. Science 1994;264:1152-6.

56. Martinez FD, Holberg CJ, Halonen M, et al. Evidence forMendelian inheritance of serum IgE in Hispanic and non-Hispanic white families. Am J Hum Genet 1994;55:555-65.

57. Xu J, Levitt RC, Panhuysen CTM, et al. Evidence for twounlinked loci regulating total serum IgE levels. Am J HumGenet 1995;57:425-30.

58. Dizier MH, Hill M, James A, et al. Detection of a recessivemajor gene for high IgE levels acting independently of spe-cific response to allergens. Genet Epidemiol 1995;12:93-105.

59. Borecki IB, Rao DC, Lalouel JM, et al. Demonstration of acommon major gene with pleiotropic effects on immuno-globulin E levels and allergy. Genet Epidemiol 1985;2:327-38.

60. Mrazek D, Pauls D, Anderson I, et al. Segregation analysis of145 asthmatic families. (Abstract). Am J Human Genet 1989;45(suppl):A245.

61. Morton NE, Shields DC, Collins A. Genetic epidemiology ofcomplex phenotypes. Ann Hum Genet 1991;55:301—14.

62. Holberg CJ, Elston RC, Halonen M, et al. Segregation anal-ysis of physician-diagnosed asthma in Hispanic and non-Hispanic white families: a recessive component? Am J Re-spir Crit Care Med 1996;154:144-50.

63. Penrose LS. The genetical background of common diseases.Acta Genet Stat Med 1953;4:257-65.

64. Risch N. Linkage strategies for genetically complex traits. I.Multilocus models. Am J Hum Genet 1990;46:222-8.

65. Edfors-Lubs M-L. Allergy in 7000 twin pairs. Acta Allergol1971;26:249-85.

66. Lubs MLE. Empiric risks for genetic counseling in familieswith allergy. J Pediatr 1972,80:26-31.

67. Kaprio J, Sarna S, Koskenvuo M, et al. The Finnish TwinRegistry: baseline characteristics. Section U. History ofsymptoms and illnesses, use of drugs, physical characteris-tics, smoking, alcohol and physical activity. Helsinki,Finland: University of Helsinki, 1978.

68. Nieminen MM, Kaprio J, Koskenvuo M. A population-basedstudy of bronchial asthma in adult twin pairs. Chest 1991;100:70-5.

69. Duffy DL. A population-based study of bronchial asthma inadult twin pairs. (Letter). Chest 1992;102:654-5.

70. Zieve I, Wiener AS, Fries JH. On the linkage relations of thegenes for allergic disease and the genes determining theblood groups, MN types and eye colour in man. Ann Eugen1936;7:163-78.

71. Lathrop GM, Lalouel JM, Julier C. Strategies for multilocuslinkage analysis in humans. Proc Natl Acad Sci U S A1984:81:3443-6.

72. Eiberg H, Lind P, Mohr J, et al. Linkage relationship betweenhuman immunoglobulin-E polymorphism and marker sys-tems. (Abstract). Cytogenet Cell Genet 1985;14:622.

73. Cookson WOCM, Sharp PA, Faux JA, et al. Linkage be-tween immunoglobulin E responses underlying asthma andrhinitis and chromosome llq. Lancet 1989;l:1292-5.

74. Daniels SE, Bhattacharrya S, James A, et al. A genome-widesearch for quantitative trait loci underlying asthma. Nature1996;383:247-50.

75. Marsh DG, Hsu SH, Roebber M, et al. HLA-Dw2: a geneticmarker for human immune response to short ragweed pollenallergen Ra5. I. Response resulting primarily from antigenic

exposure. J Exp Med 1982;155:1439-51.76. Zwollo P, Ansari AA, Marsh DG. Association of class II

DNA restriction fragments with responsiveness to Ambrosiaartemisiifolia (short ragweed)-pollen allergen Amb a V inragweed-allergic patients. J Allergy Clin Immunol 1989;83:45-54.

77. O'Hehir RE, Eckels DD, Frew AJ, et al. MHC class Hrestriction specificity of cloned human T lymphocytes reac-tive with Dermatophagoides farinae (house dust mite). Im-munology 1988;64:627-31.

78. O'Hehir RE, Garman RD, Greenstein JL, et al. The speci-ficity and regulation of T-cell responsiveness to allergens.Annu Rev Immunol 1991;9:67-95.

79. Caraballo LR, Hernandez M. HLA haplotype segregation infamilies with allergic asthma. Tissue Antigens 1990;35:182-6.

80. Amelung PJ, Panhuysen CIM, Postma DS, et al. Atopy andbronchial hyperresponsiveness: exclusion of linkage to mark-ers on chromosomes 1 lq and 6p. Clin Exp Allergy 1992;22:1077-84.

81. Rich SS, Roitman-Johnson B, Greenberg B, et al. Geneticanalysis of atopy in three large kindreds: no evidence oflinkage to D11S97. Clin Exp Allergy 1992;22:1070-6.

82. Holloway JW, Doull I, Begishvili B, et al. Lack of evidenceof a significant association between HLA-DR, DQ and DPgenotypes and atopy in families with HDM allergy. Clin ExpAllergy 1996;26:1142-9.

83. De Week AL, Blumenthal M, Yunis E, et al. HLA andallergy. In: Dausset J, Svejgaard A, eds. HLA and disease.Baltimore, MD: Williams and Wilkins, 1977:196-211.

84. Turton CWG, Morris L, Buckingham JA, et al. Histocom-patibility antigens in asthma: population and family studies.Thorax 1979;34:670-6.

85. Blumenthal M, Marcus-Bagley D, Awdeh Z, et al. HLA-DR2(HLA-B7, SC31, DR2), and (HLA-B8, SC01, DR3) haplo-types distinguish subjects with asthma from those with rhi-nitis only in ragweed pollen allergy. J Immunol 1992;148:411-16.

86. Lympany PA, Welsh KI, Christie PE, et al. An analysis withsequence-specific oligonucleotide probes of the associationbetween aspirin-induced asthma and antigens of the HLAsystem. J Allergy Clin Immunol 1993;92:114-23.

87. Bignon JS, Aron Y, Ju LY, et al. HLA class U alleles inisocyanate-induced asthma. Am J Respir Crit Care Med1994; 149:71-5.

88. Aron Y, Desmazes-Dufeu N, Matran R, et al. Evidence of astrong, positive association between atopy and the HLA classH alleles DR4 and DR7. Clin Exp Allergy 1996;26:821-8.

89. Moffatt MF, Cookson WOCM. Tumour necrosis factor hap-lotypes and asthma. Hum Mol Genet 1997;6:551-4.

90. Miwa M, Miyake T, Yamanaka T, et al. Characterization ofserum platelet-activating factor (PAF) acetylhydrolase: cor-relation between deficiency of serum PAF acetylhydrolaseand respiratory symptoms in asthmatic children. J Clin Invest1988;82:1983-91.

91. Stafforini DM, Satoh K, Atkinson DL, et al. Platelet-activating factor acetylhydrolase deficiency: a missense mu-tation near the active site of an anti-inflammatory phospho-lipase. J Clin Invest 1996;97:2784-91.

92. Young RP, Sharp PA, Lynch JR, et al. Confirmation ofgenetic linkage between atopic IgE response and chromo-some I lq l3 . J Med Genet 1992;29:236-8.

93. Cookson WOCM, Young RP, Sandford AJ, et al. Maternalinheritance of atopic IgE responsiveness on chromosomellq. Lancet 1992,340:381-4.

94. Inacio F, Perichon B, Desvaux FX, et al. Genetic transmis-sion study of grass pollen sensitivity in 17 Portugese fami-lies. (Abstract). J Allergy Clin Immunol 1991;87(suppl):2O4.

95. Shirakawa T, Morimoto K, Hashimoto T, et al. Linkagebetween IgE responses underlying asthma and rhinitis (ato-py) and chromosome 1 lq in Japanese families. (Abstract).Cytogenet Cell Genet 1991;58:1970-1.


142 Duffy

96. Shirakawa T, Hashimoto T, Furuyama J, et al. Linkagebetween severe atopy and chromosome I lq l3 in Japanesefamilies. Clin Genet 1994;46:228-32.

97. Hizawa N, Yamaguchi E, Ohe M, et al. Lack of linkagebetween atopy and locus I lq l3 . Clin Exp Allergy 1992;22:1065-9.

98. Lympany P, Welsh KL MacCochrane GM, et al. Geneticanalysis of the linkage between chromosome 1 lq and atopy.Clin Exp Allergy 1992;22:1085-92.

99. Coleman R, Trembath RC, Harper JI. Chromosome I lq l3and atopy underlying atopic eczema. Lancet 1992;341:1121-2.

100. Collet JM, ten Kate LP, de Vries HG, et al. Allele sharing onchromosome I lq l3 in sibs with asthma and atopy. (Letter).Lancet 1993;342:936.

101. Brereton HM, Ruffin RE, Thompson PJ, et al. Familial atopyin Australian pedigrees: adventitious linkage to chromosome8 is not confirmed nor is there evidence of linkage to the highaffinity IgE receptor. Clin Exp Allergy 1994;24:868-77.

102. van Herwerden L, Harrap SB, Wong ZYH, et al. Linkage ofhigh-affinity IgE receptor gene with bronchial hyperreactiv-ity, even in the absence of atopy. Lancet 1995;346:1262-5.

103. Duffy DL, Healey SC, Martin NG. Path analytic, sib-pairlinkage and cotwin-control studies of asthma and atopy.(Abstract). Am J Hum Genet 1994;55(suppl):A332.

104. Watson M, Lawrence S, Collins A, et al. Exclusion fromproximal l l q of a common gene with megaphenic effect onatopy. Ann Hum Genet 1995;59:403-ll.

105. Martinati LC, Trabetti E, Casartelli A, et al. Affected sib-pairand mutation analyses of the high affinity IgE receptor betachain locus in Italian families with atopic asthmatic children.Am J Respir Crit Care Med 1996;153:1682-5.

106. Neely JD, Duffy DL, Breazeale DR, et al. Linkage analysisof IgE and chromosome l lq 13 markers. Presented at theAmerican Lung Association/American Thoracic Society In-ternational Conference, New Orleans, Louisiana, May10-15, 1996.

107. Sandford AJ, Shirakawa T, Moffatt MF, et al. Localisation ofatopy and 0 unit of high-affinity IgE receptor (FceRI) onchromosome llq. Lancet 1993;341:332-4.

107a. Shirakawa T, Li A, Dubowitz M, et al. Association betweenatopy and variants of the /3 subunit of the high-affinityimmunoglobulin E receptor. Nat Genet 1994;7:125-9.

108. Hill MR, James AL, Faux JA, et al. FceRI|3 polymorphismand risk of atopy in a general population sample. BMJ1995;311:776-9.

109. Hill MR, Cookson WOCM. A new variant of the 0 subunitof the high-affinity receptor for immunoglobulin E (FceRI-8E237G): associations with measures of atopy and bronchialhyperresponsiveness. Hum Mol Genet 1996;5:959-62.

110. Shirakawa T, Mao XQ, Sasaki S, et al. Association betweenFee RIB and atopic disorder in a Japanese population. (Let-ter). Lancet 1996;347:394-5.

111. Shirakawa T, Mao XQ, Sasaki S, et al. Association betweenatopic asthma and a coding variant of FceRI0 in a Japanesepopulation. Hum Mol Genet 1996;5:1129-30.

112. Doull IJM, Lawrence S, Watson M, et al. Allelic associationof gene markers on chromosome 5q and 1 lq with atopy andbronchial hyperresponsiveness. Am J Respir Crit Care Med1996;153:1280-4.

113. Duffy DL, Healey SC, Chenevix-Trench G, et al. Atopy inAustralia. (Letter). Nat Genet 1995; 10:260.

114. Hall IP, Wheatley A, Dewar J, et al. Fc«RI-8 polymorphismsunlikely to contribute substantially to genetic risk of allergicdisease. (Letter). BMJ 1996;312:311.

115. Kofler H, AichbeTger S, Ott G, et al. Lack of associationbetween atopy and the Del81 Leu variant of the beta-subunitof the high-affinity immunoglobulin E receptor. Int ArchAllergy Immunol 1996; 111:44-7.

116. Meyers DA, Postma DS, Panhuysen CM, et al. Evidence fora locus regulating total serum IgE levels mapping to chro-mosome 5. Genomics 1994;23:464-70.

117. Postma DS, Bleecker ER, Amelung PJ, et al. Genetic sus-ceptibility to asthma—bronchial hyperresponsiveness coin-herited with a major gene for atopy. N Engl J Med 1995;333:894-900.

118. Blumenthal MN, Wang Z, Weber JL, et al. Absence oflinkage between 5q markers and serum IgE levels in fourlarge atopic families. Clin Exp Allergy 1996;26:892-6.

119. Borish LC, Mascali JJ, Klinnert M, et al. Polymorphisms inthe chromosome 5 gene cluster. (Abstract). J Allergy ClinImmunol 1994;93:220.

120. Walley AJ, Cookson WOCM. Investigation of an interleu-kin-4 promoter polymorphism for associations with asthmaand atopy. J Med Genet 1996;33:689-92.

121. Szentivanyi A. The beta adrenergic theory of the atopicabnormality in bronchial asthma. J Allergy 1968;42:203-32.

122. Potter PC, Van Wyk L, Martin M, et al. Genetic polymor-phism of the 0-2 adrenergic receptor in atopic and non-atopicsubjects. Clin Exp Allergy 1993;23:874-7.

123. Reishaus E, Innis M, Maclntyre N, et al. Mutations in thegene encoding for the beta 2-adrenergic receptor in normaland asthmatic subjects. Am J Respir Cell Mol Biol 1993;8:334-9.

124. Turki J, Pak J, Green SA, et al. Genetic polymorphisms ofthe Pj-adrenergic receptor in nocturnal and nonnocturnalasthma. J Clin Invest 1995;95:1635-41.

125. Hall IP, Wheatley A, Wilding P, et al. Association of Glu27/Ij-adrenoceptor polymorphism with lower airway reactivityin asthmatic subjects. Lancet 1995;345:1213—14.

126. Green SA, Turki J, Innis M, et al. Amino-terminal polymor-phisms of the human flj-adrenergic receptor impart distinctagonist promoted regulatory properties. Biochemistry 1994;33:9414-19.

127. Kowalski ML, Woszczek G, Borowiec M. Distribution of thePj-adrenoceptor polymorphisms in atopic and non-atopicpatients. (Abstract). J Allergy Clin Immunol 1997;99:S338.

128. Barnes KC, Neely JD, Duffy DL, et al. Linkage of asthmaand total serum IgE concentration to markers on chromo-some 12q: evidence from Afro-Caribbean and Caucasianpopulations. Genomics 1996;37:41-50.

129. Holgate ST. Asthma genetics: waiting to exhale. Nat Genet1997;15:227-9.

130. Moffatt MF, Hill MR, Corndlis F, et al. Genetic linkage ofT-cell receptor a/8 complex to specific IgE responses. Lan-cet 1994;343:1597-600.

131. Walter MA, Chambers CA, Zimmerman B, et al. A multi-gene deletion in the immunoglobulin heavy chain region in ahighly atopic individual. Hum Genet 1990;85:643-7.

132. Fujii H, Kondo H, Agata H, et al. Genetic analysis of IgE andthe IGHE, IGHEP1 and IGHEP2 genes in atopic families. IntArch Allergy Immunol 1995;106:62-8.

133. Oxelius VA. Correlation between atopy and Gm allotypes.Int Arch Allergy Appl Immunol 1990;91:54-7.

134. Oxelius VA, Hultquist C, Husby S. Gm allotypes as indica-tors of non-atopic and atopic bronchial asthma. Int ArchAllergy Appl Immunol 1993;101:66-71.

135. Kalsheker N, Morgan K. The a, antitrypsin gene and chroniclung disease. Thorax 1990;45:759-64.

136. Eriksson S. Pulmonary emphysema and alpha,-antitrypsindeficiency. Acta Med Scand 1964;175:197-205.

137. Lieberman J. Heterozygous and homozygous alpha,-antitrypsin deficiency in patients with pulmonary emphy-sema. N Engl J Med 1969;281:279-84.

138. Fagerhol MK, Hauge HE. Serum Pi types in patients withpulmonary diseases. Acta Allergol 1969;24:107-14.

139. Hyde JS, Werner P, Kumar CM, et al. Protease inhibitorvariants in children and young adults with chronic asthma.Ann Allergy 1979;43:8-13.

140. Buist AS, Sexton GJ, Azzam AMH, et al. Pulmonary func-tion in heterozygotes for alpha,-antitrypsin deficiency: acase-control study. Am Rev Respir Dis 1979;12O:759-66.

141. Townley RG, Southard JG, Radford P, et al. Association ofMS Pi phenotype with airway hyperresponsiveness. Chest



1990;98:594-9.142. GaiUard MC, Kilroe-Smith TA, Nogueira C, et al. Alpha-1-

protease inhibitor in bronchial asthma: phenotypes and bio-chemical characteristics. Am Rev Respir Dis 1992;145:1311-15.

143. Oakeshott JG, Muir A, Clark P, et al. Effects of the proteaseinhibitor (Pi) polymorphism on alpha-1-antitrypsin concen-tration and elastase inhibitory capacity in human serum. AnnHum Biol 1985;12:149-60.

144. Awotedu AA, Adelaja AB. Alpha-1-antitrypsin levels andprevalence of Pi variant phenotypes in adult Nigerian asth-matics. Afr J Med Med Sci 1990;19:285-90.

145. Colp C, Talavera W, Goldman D, et al. Profile of broncho-spastic disease in Puerto Rican patients in New York City: apossible relationship to alpha-1-antitrypsin variants. Ann In-tern Med 1990;150:2349-54.

146. Eriksson S, Lindmark B, Lilja H. Familial a,-antichymotrypsin deficiency. Acta Med Scand 1986;220:447-53.

147. Lindmark B. Asthma and heterozygous a,-antichymotrypsindeficiency: a possible association. J Intern Med 1990;227:115-18.

148. Ronchetti R, Lucarini N, Lucarelli P, et al. A genetic basisfor heterogeneity of asthma syndrome in pediatric ages:adenosine deaminase phenotypes. J Allergy Clin Immunol1984;74:81-4.

149. Benessiano J, Crestani B, Mestari F, et al. High frequency ofa deletion polymorphism of the angiotensin-converting en-zyme gene in asthma. J Allergy Clin Immunol 1997;99:53-7.

150. Frohlander N, Stjernberg N. Association between haptoglo-bin groups and hereditary predisposition for bronchialasthma. Hum Hered 1989;39:7-11.

151. Srivastava N, Gupta SP, Srivastava LM. Association be-tween C3 complement types and bronchial asthma. HumHered 1985;35:263-4.

152. Helander E. Asthma and diabetes. Acta Med Scand 1958;162:165-74.

153. Stromberg LGE, Ludvigsson GJ, Bjorksten B. Atopic allergyand delayed hypersensitivity in children with diabetes. JAllergy Clin Immunol 1995;96:188-92.

154. Vena JE, Bona JR, Byers TE, et al. Allergy-related diseasesand cancer an inverse association. Am J Epidemiol 1985;122:66-74.

155. Reynolds P, Kaplan GA. Asthma and cancer. (Letter). Am JEpidemiol 1987;125:539-40.

156. Robinette CD, Fraumeni JF Jr. Asthma and subsequent mor-tality in World War U veterans. J Chronic Dis 1978;31:619-24.

157. Huovinen E, Kaprio J, Vesterinen E, et al. Mortality of adultswith asthma: a prospective cohort study. Thorax 1997;52:49-54.

158. Danon YL, Laor A, Shlezinger M, et al. Decreased incidenceof asthma in patients with familial Mediterranean fever. IsrJ Med Sci 1990;26:459-60.

159. Ozyilkan E, Simsek H, Telatar H. Absence of asthma inpatients with familial Mediterranean fever. Isr J Med Sci1994;30:237-8.

160. Brenner-Ullman A, Melzer-Ofir H, Daniels M, et al. Possibleprotection against asthma in heterozygotes for familial Med-iterranean fever. Am J Med Genet 1994;53:172-5.

161. Schroeder SA, Gaughan DM, Swift M. Protection againstbronchial asthma by CFTR AF508 mutation: a heterozygoteadvantage in cystic fibrosis. Nat Med 1995; 1:703-5.

162. De Sanctis GT, Merchant M, Beier DR, et al. Quantitativelocus analysis of airway hyperresponsiveness in A/J andC57BL/6J mice. Nat Genet 1995;11:150-4.

163. Zhang LY, Levitt RC, Kleeberger SR. Differential suscepti-bility to ozone-induced airways hyperreactivity in inbredstrains of mice. Exp Lung Res 1995;21:503-18.

164. Young C, Bhalla DK. Effects of ozone on the epithelial andinflammatory responses in the airways: role of tumor necro-sis factor. J Toxicol Environ Health 1995;46:329-42.

165. Ewart SL, Mitzner W, DiSilvestre DA, et al. Airway hyper-responsiveness to acetylcholine: segregation analysis andevidence for linkage to murine chromosome 6. Am J RespirCell Mol Biol 1996;14:487-95.

166. Gorham JD, Gfller ML, Steen RG, et al. Genetic mapping ofa murine locus controlling development of T helper 1/Thelper 2 type responses. Proc Natl Acad Sci U S A 1996;93:12467-72.

167. Moffatt MF, Sharp PA, Faux JA, et al. Factors confoundinggenetic linkage between atopy and chromosome llq. ClinExp Allergy 1992;22:1046-51.

168. Morton NE. Major loci for atopy? (Editorial). Clin ExpAllergy 1992;22:1041-3.


Documents

Genetic Epidemiology of Asthma · Genetic Epidemiology of Asthma David L Duffy INTRODUCTION Asthma is an increasingly common disease in the developed world (1-3). In the United States,