Case–Control Study of Eczema in Relation to IL4Rα Genetic Polymorphisms in Japanese Women: The Kyushu Okinawa Maternal and Child Health Study

Authors


Correspondence to: Dr Y. Miyake, Department of Preventive Medicine and Public Health, Faculty of Medicine, Fukuoka University, 814-0180 Fukuoka, Japan. E-mail: miyake-y@fukuoka-u.ac.jp

Abstract

Epidemiological research on the relationship between single nucleotide polymorphisms (SNPs) in the IL4Rα gene and eczema is sparse. We investigated the associations between IL4Rα SNPs rs1805011, rs1805015 and rs1801275 and risk of eczema in young adult Japanese women. Included were 188 women who met the criteria of the International Study of Asthma and Allergies in Childhood (ISAAC) for eczema. Controls were 635 women without eczema according to the ISAAC criteria who also had not been diagnosed with asthma, atopic eczema and/or allergic rhinitis by a doctor. Adjustment was made for age, region of residence, number of children, smoking and education. Under the additive model, SNP rs1805011 was significantly related to eczema: the adjusted OR was 0.55 (95% CI: 0.31−0.99). SNP rs1805015 was significantly associated with eczema in the additive and dominant models: the adjusted ORs were 0.55 (95% CI: 0.30−0.98) and 0.55 (95% CI: 0.30−0.997), respectively. There was no significant association between SNP rs1801275 and eczema. None of the haplotypes were significantly related to eczema. Significant associations between SNPs rs1805011 and rs1805015 and eczema were reported in women who had never smoked, but not in those who had ever smoked; the multiplicative interactions, however, were not significant. This is the first study to demonstrate significant associations between IL4Rα SNPs rs1805011 and rs1805015 and eczema. We do not find evidence for interactions affecting eczema between IL4Rα SNPs and smoking.

Introduction

According to a recent nationwide hospital-based survey, atopic eczema is the second most common skin disease in Japan, following miscellaneous eczema [1]. Its age distribution was biphasic, peaking at 0−5 and 21−35 years of age [1]. Both environmental and genetic factors with complex interactions are thought to be responsible for the development of atopic eczema.

Interleukin (IL)-4 and IL-13, secreted predominantly by Th2 cells, are key cytokines in the pathogenesis of allergic disorders and share many biological and biochemical characteristics [2]. Both IL-4 and IL-13 use the IL-4 receptor α chain (IL-4Rα) as a component of their receptors and transmit their signals through IL-4Rα [3]. Thus, the gene encoding IL-4Rα represents an ideal candidate gene for susceptibility to allergic disorders. Multiple single nucleotide polymorphisms (SNPs) in the IL4Rα gene including rs1805010, rs1805011, rs1805015 and rs1801275 have been significantly associated with asthma, atopy or total serum IgE levels in several genetic association studies, although these studies have produced conflicting results [4-13]. Epidemiological evidence for the relationship between SNPs in the IL4Rα gene and eczema has been limited, however [12, 14-19].

Using data from the Kyushu Okinawa Maternal and Child Health Study (KOMCHS), we conducted a case–control study of the association between IL4Rα SNPs and eczema in a Japanese population of young adult women. In addition, haplotype analyses were performed, and the possibility of interaction between the SNPs and smoking was investigated.

Methods

Study population

The KOMCHS is an ongoing prospective prebirth cohort study that investigates risk and preventive factors for maternal and child health problems such as allergic disorders. From April 2007 to March 2008, the KOMCHS requested that 131 obstetric hospitals in Fukuoka Prefecture, the largest prefecture on Kyushu Island in southern Japan, with a total population of approximately 5.04 million, provide as many pregnant women as possible with a set of leaflets explaining the KOMCHS, an application to participate in the study and a self-addressed stamped envelope in which to return the application. From May 2007 to March 2008, the KOMCHS also requested that 40 obstetric hospitals in Okinawa Prefecture, one of the southernmost islands of Japan, with a total population of almost 1.37 million, provide as many pregnant women as possible with a comparable set of documents. In addition, to increase the sample size, from August 2007 to March 2008, pregnant women living in six prefectures on Kyushu Island (other than Fukuoka Prefecture), with a total population of approximately 8.22 million, were provided with a similar set of materials at 252 obstetric hospitals. Pregnant women who intended to participate in the KOMCHS returned the application form to the data management centre. When all responses were tallied, a total of 1757 pregnant women between the 5th and 39th weeks of pregnancy had provided their fully informed written consent to participate and had completed the baseline survey. Around 4 months after delivery, 1492 women gave informed consent to genotyping. The ethics committee of the Faculty of Medicine, Fukuoka University, approved the KOMCHS.

Selection of cases and controls

In the baseline survey, each participant filled out a self-administered questionnaire and mailed the completed questionnaire to the data management centre. Research technicians completed missing or illogical data by telephone interview.

The questionnaire included questions on eczema based on the International Study of Asthma and Allergies in Childhood (ISAAC) Phase One questionnaire [20]. Participants were considered to have eczema if they gave affirmative answers to the following three questions: ‘Have you ever had an itchy rash which was coming and going for at least 6 months?’ ‘Have you had this itchy rash at any time in the last 12 months?’ and ‘Has this itchy rash at any time affected any of the following places: the folds of the elbows, behind the knees, in front of the ankles, under the buttocks, or around the neck, ears, or eyes?’ The questionnaire also yielded information on age, region of residence, number of children, smoking habits and education. A history of smoking was defined as having smoked at least once per day for at least one year.

Based on the definition provided above, 188 cases of eczema were identified among the 1492 women whose DNA samples were available. Among the 1304 remaining women who were eligible to serve as control subjects, 668 women were excluded who were not considered to have eczema as defined by the ISAAC criteria but who had answered ‘yes’ to any of the following three questions: ‘Have you ever been diagnosed by a physician as having asthma?’ ‘Have you ever been diagnosed by a physician as having atopic eczema?’ and/or ‘Have you ever been diagnosed by a physician as having allergic rhinitis?’ After the further exclusion of one woman with incomplete data on smoking, 635 women were identified as control subjects for the final analysis.

DNA extraction and genotyping

Research technicians or subjects themselves collected buccal specimens with BuccalAmp swabs (Epicenter BioTechnologies, Madison, WI, USA). Genomic DNA was extracted using a QIAmp DNA mini kit (Qiagen, Inc., Valencia, CA, USA). Genotyping of the IL4Rα SNPs was performed using TaqMan SNP Genotyping Assays on the StepOnePlus machine (Applied Biosystems, Foster City, CA, USA), according to the manufacturer's instructions.

Statistical analysis

Departures from Hardy–Weinberg equilibrium were tested among the control subjects using the chi-squared test. Linkage disequilibrium was examined using haploview software version 4.2 (Broad Institute, Cambridge, MA, USA) [21]. Estimations of crude odds ratios (ORs) and 95% confidence intervals (CIs) for eczema associated with the SNPs under study were made by means of logistic regression analysis, with the reference category being the homozygote of the major allele. Because there were few subjects with the homozygote of the minor allele, not only the co-dominant model but also the additive and dominant models were tested. For the additive model, the IL4Rα SNPs were entered into the logistic model as a count (0, 1 or 2) corresponding to the number of minor alleles the subject carried. For the dominant model, the comparison was between the homozygote of the major allele and the combination of the heterozygote and the homozygote of the minor allele. Multiple logistic regression analysis was used to control for age, region of residence, number of children, smoking and education. The statistical power calculation was performed using quanto version 1.2 [22]. Haplotypes and their frequencies were inferred according to the expectation maximization algorithm. For differences in haplotype frequency between the cases and control groups, crude ORs and 95% CIs were estimated based on the frequency of each haplotype relative to all other haplotypes combined. The interaction was tested using a term of the product of two variables in a multiple logistic regression model. Excluding the calculation of linkage disequilibrium and statistical power calculation, all statistical analyses were performed using stata/se software version 12.0 (StataCorp, College Station, TX, USA).

Results

There were no differences between cases and control subjects with regard to age, region of residence, having one or more children, smoking or education (Table 1).

Table 1. Characteristics of the study population
VariableCases (N = 188) n (%)Controls (= 635) n (%)P valuea
  1. a

    chi-squared test or Wilcoxon rank sum test.

Age, years, mean ± SD30.7 ± 4.631.4 ± 4.20.05
Region of residence  0.35
Fukuoka Prefecture112 (59.6)357 (56.2) 
Other than Fukuoka Prefecture in Kyushu61 (32.5)204 (32.1) 
Okinawa Prefecture15 (8.0)74 (11.7) 
Having one or more children107 (56.9)402 (63.3)0.11
Having ever smoked63 (33.5)199 (31.3)0.57
Education, years  0.79
<1344 (23.4)135 (21.3) 
13−1462 (33.0)209 (32.9) 
≥1582 (43.6)291 (45.8) 

In this study, we investigated three SNPs from the IL4Rα gene: rs1805011 (E400A), rs1805015 (S503P) and rs1801275 (Q576R). SNP rs1805010 (I75V) was excluded from this study because the genetic distribution among our control subjects deviated significantly from Hardy–Weinberg equilibrium (P = 0.0003). The distributions of the IL4Rα SNPs rs1805011, rs1805015 and rs1801275 among control subjects were in Hardy–Weinberg equilibrium (P = 0.50, 0.50 and 0.87, respectively). The three SNP pairs were in strong linkage disequilibrium (D’ = 0.98–1.00, r2 = 0.37–0.98) (Table 2).

Table 2. Pairwise linkage disequilibrium of IL4Rα polymorphisms (r2 below and D' above the diagonal)
 rs1805011rs1805015rs1801275
rs1805011 1.000.98
rs18050150.98 0.98
rs18012750.370.37 

None of the cases carried the CC genotype of SNPs rs1805011 and rs1805015, and only one control subject with the CC genotype of SNPs rs1805011 and rs1805015 was identified (Table 3). Under the additive model, SNP rs1805011 was significantly related to the risk of eczema after adjustment for confounders: the adjusted OR was 0.55 (95% CI: 0.31−0.99). No such significant relationship was found in the co-dominant or dominant model. SNP rs1805015 was significantly associated with the risk of eczema in the additive and dominant models: the adjusted ORs were 0.55 (95% CI: 0.30−0.98) and 0.55 (95% CI: 0.30−0.997), respectively. There was no significant association between SNP rs1801275 and eczema in any genetic model. With respect to SNP rs1801275, the statistical power calculation revealed that, using our sample size, we could detect the gene–disease association for an OR of 0.598 with an accuracy of more than 80% at a significance level of 0.05 with a two-sided alternative hypothesis under the log-additive model.

Table 3. ORs and 95% CIs for eczema associated with IL4Rα polymorphisms in Japanese women
SNPModelGenotypeCases (= 188) n (%)Controls (= 635) n (%)Crude OR (95% CI)Adjusted OR (95% CI)a
  1. CI, confidence interval; OR, odds ratio.

  2. a

    Adjusted for age, region of residence, number of children, smoking and education.

rs1805011Co-dominantAA174 (92.6)555 (87.4)1.001.00
  AC14 (7.5)79 (12.4)0.57 (0.31−1.02)0.57 (0.31−1.03)
  CC0 (0.0)1 (0.2)Not calculableNot calculable
 Additive   0.56 (0.31−1.002)0.55 (0.31−0.99)
 Dominant   0.56 (0.31−1.01)0.56 (0.31−1.01)
rs1805015Co-dominantTT174 (92.6)554 (87.2)1.001.00
  TC14 (7.5)80 (12.6)0.56 (0.31−1.01)0.56 (0.31−1.01)
  CC0 (0.0)1 (0.2)Not calculableNot calculable
 Additive   0.55 (0.30−0.99)0.55 (0.30−0.98)
 Dominant   0.55 (0.30−0.995)0.55 (0.30−0.997)
rs1801275Co-dominantAA147 (78.2)465 (73.2)1.001.00
  AG40 (24.6)156 (24.6)0.81 (0.55−1.20)0.79 (0.53−1.17)
  GG1 (0.5)14 (2.2)0.23 (0.03−1.73)0.20 (0.03−1.54)
 Additive   0.74 (0.52−1.06)0.71 (0.50−1.03)
 Dominant   0.76 (0.52−1.12)0.74 (0.50−1.09)

When haplotypes with a frequency of <1% in either cases or control subjects were deleted, three constructed haplotypes remained; however, none of these were significantly related to the risk of eczema (Table 4).

Table 4. Haplotype analysis of 3 IL4Rα polymorphisms associated with eczema in Japanese womena
HaplotypebFrequency, n (%)Crude OR (95% CI)c
Cases (2= 376)Controls (2= 1270)
  1. CI, confidence interval; OR, odds ratio.

  2. a

    Rare haplotypes (frequency <1% in either cases or controls) were deleted.

  3. b

    Haplotype order is rs1805011, rs1805015 and rs1801275.

  4. c

    Crude OR for each haplotype is relative to all other haplotypes combined.

ATA334 (88.8)1085 (85.4)1.36 (0.94−1.99)
ATG28 (7.4)103 (8.1)0.91 (0.57−1.42)
CCG14 (3.7)80 (6.3)0.58 (0.30−1.04)

Under the dominant model, significant associations between SNPs rs1805011 and rs1805015 and eczema were reported in women who had never smoked, but not in those who had ever smoked (Table 5). The multiplicative interactions between these SNPs and smoking affecting eczema were not statistically significant, however.

Table 5. Association between IL4Rα polymorphisms and eczema, stratified by smoking history in Japanese women
SNPGenotypeSmoking historyP for interaction
NoYes
No. cases/controlsAdjusted OR (95% CI)aNo. cases/controlsAdjusted OR (95% CI)a
  1. CI, confidence interval; OR, odds ratio.

  2. a

    Adjusted for age, region of residence, number of children and education.

rs1805011AA117/3781.0057/1771.00 
 AC + CC8/580.46 (0.21−0.99)6/220.83 (0.31−2.18)0.96
rs1805015TT117/3771.0057/1771.00 
 TC + CC8/590.45 (0.21−0.98)6/220.83 (0.31−2.18)0.88
rs1801275AA101/3311.0046/1341.00 
 AG + GG24/1050.75 (0.45−1.23)17/650.75 (0.39−1.43)0.46

Discussion

The current study demonstrated that the IL4Rα SNPs rs1805011 and rs1805015 were significantly associated with the risk of eczema, while SNP rs1801275 and all examined haplotypes were not materially related to eczema. The multiplicative interactions between SNP rs1805011 or rs1805015 and smoking did not have a statistically significant effect on the risk of eczema. Previously, no significant association was shown between SNP rs1805011 and eczema in four case–control studies of Japanese adults (27 cases and 29 controls [14] and 302 cases and 122 controls [15]), Japanese children (45 cases and 125 controls) [19], Taiwanese adults (94 cases and 186 controls) [16], a birth cohort study of 762 US infants with at least one atopic parent [18] and a study using data from 13 countries participating in Phase Two of the ISAAC (4242 subjects) [12]. Similarly, there was no significant relationship between SNP rs1805015 and eczema in a case–control study of Taiwanese adults [16], a study using data from 13 countries participating in Phase Two of the ISAAC [12], and a cross-sectional study of 309 Swedish children [17]. These results are at variance with our findings regarding SNPs rs1805011 and rs1805015. In four of the above-mentioned studies [12, 15-17], SNP rs1801275 was not significantly related to eczema: these results are in agreement with our findings. In contrast, a case–control study of Japanese adults showed a significant relationship between SNP rs1801275 and eczema [14]. The inconsistency in our findings with those of some previous studies may be at least partly explained by differences in the genetic backgrounds of the populations examined, definitions of eczema and statistical power.

A case–control study of Dutch adults found a significant relationship between asthma and SNP rs1805011, but not SNP rs1805015 or rs1801275 (118 cases and 102 controls) [13]. In another Dutch study among adults (200 probands with asthma and 201 spouses), SNP rs1805015, but not rs1801275, was significantly associated with total serum IgE levels, asthma, bronchial hyper-responsiveness and positive skin tests for sensitization to common allergens [8]. A study of 181 Danish children showed significant relationships between SNPs rs1805015 and rs1801275 and total IgE levels, while in vivo immunoassays revealed that only in cases of double polymorphisms was the phosphorylation of insulin receptor substrate molecules increased [6].

The German Multicenter Atopy Study found a significant interaction between SNP rs1801275 and prenatal exposure to environmental tobacco smoke with respect to sensitization to cat allergens [23]. In the present study, smoking did not significantly modify the associations between SNPs rs1805011, rs1805015 or rs1801275 and eczema.

The current study offers two main methodological advantages. Study subjects were homogeneous in that they were all pregnant women and adjustment was made for several confounders.

Several weaknesses should be considered. First, the participation rate cannot be calculated because the exact number of eligible pregnant women who were provided with the above-mentioned KOMCHS documents is not available. In addition, we were not able to assess differences between participants and non-participants, because information on personal characteristics such as age, socio-economic status and history of allergic disorders among the non-participants was not available. Our subjects were probably not representative of Japanese women in the general population, however: for example, the distribution of educational status among participants differed considerably from that among the general population. According to the 2000 population census of Japan, the proportions of women aged 30–34 years in Fukuoka Prefecture with years of education of <13, 13−14, ≥15 and unknown were 52.0%, 31.5%, 11.8% and 4.8%, respectively [24]. The corresponding figures for the control group in the current study were 21.3%, 32.9%, 45.8% and 0.0%, respectively. Therefore, the present population might have had greater awareness concerning health-related matters compared with the general population. Nevertheless, the distribution of all three SNPs under study was consistent with Hardy–Weinberg equilibrium, and the selection bias associated with genotype distribution would be negligible.

Second, the definition of eczema was based on the questions in the ISAAC questionnaire, although validation tests of such questions have not been performed for Japanese young adults. Neither data on serum-specific IgE levels nor skin prick test results were available. The possibility of non-differential outcome misclassification would have given rise to an underestimation of our results.

Third, the current study size was rather small for a valid genetic association study, although significant associations were detected between SNPs rs1805011 and rs1805015 and eczema. The lack of a significant relationship between SNP rs1801275 and eczema and the significant multiplicative interactions under investigation might be attributable to an insufficient statistical power.

Fourth, correction for multiple testing, an appropriate element in initial exploratory analyses, was not performed in this study. As this is a hypothesis-testing study and part of the current findings is a replication of previously published results, we think that correction for multiple testing would cause us to underestimate our results.

Fifth, although adjustment was made for some confounders, residual confounding effects could not be ruled out.

To our knowledge, this is the first study to demonstrate significant associations between the IL4Rα SNPs rs1805011 and rs1805015 and the risk of eczema. We could not find evidence for interactions affecting eczema between IL4Rα SNPs and smoking. Because epidemiological research on this topic is still sparse, further evidence from well-designed case–control and prospective cohort studies is required, and functional studies are also needed.

Acknowledgment

The authors would like to thank the Kyushu Branch of the Japan Allergy Foundation, the Fukuoka Association of Obstetricians & Gynecologists, the Okinawa Association of Obstetricians & Gynecologists, the Miyazaki Association of Obstetricians & Gynecologists, the Oita Association of Obstetricians & Gynecologists, the Kumamoto Association of Obstetricians & Gynecologists, the Nagasaki Association of Obstetricians & Gynecologists, the Kagoshima Association of Obstetricians & Gynecologists, the Saga Association of Obstetricians & Gynecologists, the Fukuoka Society of Obstetrics and Gynecology, the Okinawa Society of Obstetrics and Gynecology, the Fukuoka City Government and the Fukuoka City Medical Association for their valuable support, as well as Mrs. Yukari Hayashi for her technical assistance. This study was supported by KAKENHI grants (19590606, 20791654, 21590673, 22592355 and 24390158), by Health and Labour Sciences Research Grants for Research on Allergic Disease and Immunology from the Ministry of Health, Labour, and Welfare of Japan, by the Central Research Institute of Fukuoka University and by the Takeda Science Foundation.

Ancillary