|Home | About | Journals | Submit | Contact Us | Français|
Parent of origin effect is important in understanding the genetic basis of childhood allergic diseases and to improve our ability to identify high risk children.
To investigate parent of origin effect in childhood allergic diseases.
The Isle of Wight Birth Cohort (n=1,456) has been examined at 1, 2, 4, 10 and 18-years. Information on prevalence of asthma, eczema, rhinitis and environmental factors was obtained using validated questionnaires. Skin prick tests were carried out at ages 4, 10 and 18 year, and total IgE at 10 and 18 years. Parental history of allergic disease was assessed soon after the birth of the child when maternal IgE was also measured. Prevalence ratios (PR) and their 95% confidence intervals (CI) were estimated, applying log-linear models, adjusted for confounding variables.
When stratified for sex of the child, maternal asthma was associated with asthma in girls [PR:1.91 (CI:1.34–2.72), p=0.0003], but not in boys [PR:1.29 (CI:0.85–1.96), p=0.23), while paternal asthma was associated with asthma in boys [PR:1.99 (CI:1.42–2.79), p<0.0001], but not in girls [PR: 1.03 (0.59–1.80) p=0.92). Maternal eczema increased the risk of eczema in girls [PR: 1.92 (CI: 1.37–2.68); p=0.0001] only, while paternal eczema did the same for boys (PR: 2.07 (CI:1.32–3.25); P=0.002). Similar trends were observed when the effect of maternal and paternal allergic disease was assessed for childhood atopy and when maternal total IgE was related to total IgE in children at age 10 and 18 years.
The current study indicates a sex dependent association of parental allergic conditions with childhood allergies; maternal allergy increasing the risk in girls and paternal allergy in boys. This has implications for childhood allergy prediction and prevention.
The heritability of allergic diseases has been recognized since the early 20th century. A history of asthma in the immediate family is one of the major risk factors for childhood asthma1, and the same is true for atopic eczema and allergic rhinitis.2 An accurate assessment of the heritable risk is important in providing more accurate diagnosis to parents, identifying at-risk children for preventive measures and also to investigate how environmental factors may interact with the individual’s genetic predisposition. An important unresolved issue in this context is the parent of origin effect; the respective contribution of maternal and paternal allergic disease. A number of studies have investigated this question with conflicting results.
A common perception is that maternal asthma confers a greater risk3–6, although some studies indicated a stronger paternal effect7 or no difference.8 A stronger maternal effect may be explained by a stronger maternal parent of origin effect9, the effects of maternal environmental exposure during pregnancy, or immune interactions between mothers and their offspring in utero.10 For childhood eczema, several studies reported a greater effect of maternal than paternal eczema.7,11–15 However, a number of large studies failed to confirm a greater influence of maternal eczema.2,16–19 For allergic rhinitis, only a few studies have investigated the parent of origin effect but found no significant difference in maternal or paternal rhinitis.2,7
Most studies did not stratify their samples according to the sex of the child, having made the assumption that maternal and paternal effects are identical in boys and girls. This also applies to a recent meta-analysis of the effect of parental history which did not take the sex of the offspring into consideration.6 One cross-sectional study of 9 to 11 year olds found no differential sex-dependent effect of parental allergic disease.7 As allergic diseases tend to relapse and remit throughout childhood, cross-sectional studies run the risk of misclassification as those regarded as not having the disease at the time of assessment may have had it earlier or may develop it later. Only one longitudinal study investigated the effect of parental atopy (but not asthma) on childhood wheeze (but not eczema or rhinitis) up to age 26.20 The association of parental atopy was dependent on age and sex of the child but not in one direction. Maternal atopy was a risk factor for childhood onset wheeze in both sexes, while paternal atopy increased the risk in male only. For adolescent onset wheeze, maternal history of atopy increased the risk in girls only. However, when both ages were combined, paternal history had a stronger risk in girls than boys. This conflicting outcome may be because “parental asthma and rhinitis” was combined as “parental atopy” with childhood asthma as the outcome of interest. It is well known that heritability of childhood allergic conditions is disease-specific so that parental asthma has a greater influence on childhood asthma and the same is true for atopic eczema and allergic rhinitis.7 In addition, early life risk factors, such as respiratory infections, mode of early feeding, and exposure to environmental tobacco smoke might also influence the outcome and thus confound the results.21 Therefore, a comprehensive investigation requires a birth cohort that is studied with prospective phenotyping, including objective tests for atopy, throughout childhood with available information on early life risk factors.
We addressed the following questions utilizing longitudinal data collected from the Isle of Wight birth cohort:
An unselected whole population birth cohort (n=1,536) was recruited in 1989 to prospectively study the natural history of asthma and allergic conditions. After exclusion of adoptions, perinatal deaths and refusal for follow-ups, 1,456 children were enrolled with follow-up assessments conducted at 1, 2, 4, 10 and 18 years. At each stage, validated questionnaires, including International Study of Asthma and Allergy in Childhood,22 were completed on asthma and allergic diseases plus exposure to relevant environmental factors such as tobacco smoke and pets. Information on breastfeeding duration was collected at 1 and 2 years. The majority of participants underwent skin prick testing (SPT) at 4, 10 and 18 years to 14 common food and aeroallergens (ALK-Abello, Horsholm, Denmark).23–25
Serum for IgE assessment was collected at parturition from mothers (n=1,057, 73%) of study subjects. Samples for determining IgE levels at age 10 and 18 years were available for 954 (66%) and 610 (42%) subjects, respectively. Maternal IgE and IgE at age 10 and 18 years were determined using PRIST® (Phadia AB, Uppsala, Sweden) designed to measure IgE between 2.0 to 1000 kU/L. Detailed methodology of recruitment and follow-ups has been published previously.23–26 Ethics approval had been obtained at each follow-up by local research ethics committees and informed consent was obtained by parents and/or participants.
Information on parental history of allergic conditions was collected from mothers soon after birth. Maternal or paternal asthma were defined as those parents responding “yes” to the question; “Have you ever suffered from asthma?” Information on parental reports of eczema and rhinitis was collected in the same way. If one or more of these allergic conditions were present in parents, they were regarded as having “a history of allergy”. The definition used for asthma in a cohort child was a history of physician-diagnosed asthma plus at least one episode of wheezing or asthma treatment in the previous 12 months. Eczema was defined as chronic or chronically relapsing, itchy dermatitis lasting more than 6 weeks with characteristic morphology and distribution, following Hanifin and Rajka criteria.27 Rhinitis was defined by a positive response to; “In the past 12 months have you had a problem with sneezing, or a runny or a blocked nose when you did not have a cold or the flu?” Since asthma and rhinitis cannot be confidently diagnosed in early childhood (1–2 years), to reduce misclassification, we analyzed asthma and rhinitis from 4 years onwards. Eczema was considered from 1–18 years. Atopy was defined by positive SPT (mean wheal diameter 3 mm larger than negative control) to at least one allergen. Maternal and offspring IgE at 10 and 18 years higher than 200 kU/L were considered as raised. A newborn was classified as having “low birth weight” if birth weight was <2.5 kg. Other environmental factors assessed were maternal smoking during pregnancy (yes vs. no), birth order of child in the family (1st versus 2nd or higher), presence of cat or dog in the home at birth, and breast feeding duration. Breast feeding was analyzed as breast fed for at least 3 months versus those where breast feeding ceased before this age.
Data were double entered and analyzed using SPSS version 17 (IBM, NY, USA). The prevalence of asthma, eczema, allergic rhinitis and atopy was calculated. Univariate analyses using Chi square tests (two sided) were used to test for differences in proportions stratified for the sex of the child. Generalized linear mixed models were applied to examine the interaction effect of parental history of disease with sex on a multiplicative scale, after adjusting for the covariates (the conceptual model: logit(P(Asthma in child)) = Maternal asthma (0/1) + Paternal asthma (0/1)+Gender (M/F) + Maternal asthma*gender + Paternal asthma*gender + covariates). This was followed by stratified analyses upon statistically significant findings. For each childhood allergic manifestation (asthma, eczema, rhinitis and atopy), the impact of a parental disease of the same type was analysed separately for mother and the father. Using repeated measurement analyses, changes were investigated in the prevalence of childhood asthma related to parental asthma from age 4 to 18, stratifying for sex of the child. To obtain an overall independent effects of the parental history, we adjusted for early childhood risk factors (in repeated measurement analysis), including maternal smoking during gestation (yes vs. no), low birth weight (<2.5 kg), exposure to a dog or a cat at home, and breastfeeding (3 months or more vs. less).
Since the prevalence of allergic diseases does not present a rare event, odds ratios are likely to overestimate relative risks.28 To directly estimate prevalence ratios (PR), we applied a log-linear model for prevalence. For each observation period (ages 1, 2, 4, 10, and 18 years) we estimated the association with maternal and paternal disease and present PR and their 95% confidence intervals (CI). To assess long-term development in individual children, we had to consider that repeated measurements for each individual child represent correlated observations. Applying the method of generalized estimating equation (GEE) 29 that takes the within-child effect into account, we estimated marginal probabilities for maternal and paternal history of asthma and allergy using GENMOD in the SAS system (SAS, Gary, NC, USA). Details of statistical methods are provided in an on-line data supplement.
Of the 1,456 children enrolled in the study, 1,374 (94%) were seen at the age of 1 year, 1,231 (85%) at 2 years, 1,214 (83%) at 4 years, 1,368 (94%) at 10 years and 1,305 (90%) at 18 years. In a simple cross-tabulation, both maternal and paternal asthma were closely and consistently associated with childhood asthma (Table I). This was further confirmed using repeated measurement analyses with 1,961 observations in boys and 1,926 observations in girls. The interaction of maternal asthma with sex was insignificant (p-value=0.11); however, a significant interaction of paternal asthma with sex was observed (p-value=0.03). Accordingly, the stratified analyses showed that paternal asthma increased the risk of childhood asthma in boys (Table II, repeated measurement analyses: RR=1.99 (1.42–2.79), p <0.0001). Although not statistically significant on a multiplicative interaction scale, maternal asthma tended to increase the risk of childhood asthma in girls. A biparental history did not interact with sex and had a greater impact on the risk and was equally important in boys and girls [Boys: PR: 3.28 (CI: 1.80–6.00)] and [Girls: PR: 3.19 (1.98–5.12)] (data not shown).
In simple cross-tabulation, maternal eczema increased the risk of childhood eczema at most assessments (Table III). In repeated measurement analysis (3,279 observations in boys from age 1 to 18 and 3,206 observations in girls), a maternal effect was observed for the childhood and adolescent period from 1 to 18 years, which was statistically significant on a multiplicative interaction scale (p=0.01). Paternal eczema significantly increased the risk of childhood eczema up to age 10, and an overall effect was also shown with repeated measurement analysis (p=0.02); however, the interaction of paternal history and male offspring sex was not statistically significant on a multiplicative scale. When further stratified for child’s sex, we found that maternal eczema had no statistically significant effect in boys at any age, but in girls, maternal eczema increased the risk of eczema at all ages except at age 1 and a significant overall effect (p=0.009) was noted (Table IV). For paternal effects, although the interaction were statistically insignificant, paternal eczema tended to increase the risk of childhood eczema up to age 10 in boys and an overall effect was observed in repeated measurement analysis (p=0.002).
When analyzing the whole sample, maternal rhinitis was not associated with childhood rhinitis at any age (Table V). In contrast, paternal rhinitis was significantly associated with childhood rhinitis at all ages (except age 18 when it failed to reach statistical significance; p=0.06). A significant association was observed in the repeated measurement analysis adjusted for confounders (P<0.0001). From age 4 to 18, there were 1,962 observations in boys and 1,923 observations in girls. No statistically significant parental rhinitis and sex interaction was detected. This was further supported by stratification on paternal rhinitis and offspring sex (Table VI): paternal rhinitis increased the risk of childhood rhinitis with no differences between boys and girls at 10 years and in the repeated measurement analyses (boys: p<0.001;girls: p=0.004). Maternal history of rhinitis had no effect.
We then examined the parent of origin effect on atopy (SPT positivity) and total IgE. Both maternal and paternal history of allergy increased the risk of childhood atopy from 4 to 18 years. The repeated measurement analysis of 1,416 observations in boys and 1,455 observations in girls corroborated an overall effect (Table VII). A significant interaction of maternal history of atopy with sex was identified (p=0.046); whereas paternal history did not interact with sex (p=0.88). A stratified analysis revealed that in boys, maternal history of allergy had no effect on atopy while in girls it significantly increased the risk at all ages (from 4 to 18 years).
Regarding offspring IgE at 10 years higher than 200 kU/L, we assumed that maternal IgE at birth (>200 kU/L) will predict increased levels in girls but not in boys. In girls, 42% of those with raised maternal IgE also had raised IgE (26% if not), whereas in boys 35% with raised maternal IgE also had raised IgE at age 10 (30% if not). We tested this association using log-linear models The prevalence ratio (PR) of maternal IgE at birth predicted raised IgE in girls at age 10 (PR=1.71, CI: 1.16–2.51), but not in boys (PR=1.1, CI: 0.74–1.63; data not shown). Similar results were seen for IgE measured at age 18 years: 41% of the girls with raised maternal IgE also had raised IgE (26% if not), whereas in boys 33% with raised maternal IgE also had raised IgE at age 10 (31% if not). The PR for raised maternal IgE in boys at age 18 was 1.08 (CI 0.67–1.74) and 1.66 (CI 1.01–2.73) in girls.
We have shown that the effect of maternal and paternal history of allergy varies with sex of the child, where by maternal allergic history increased the risk of asthma and eczema in girls while paternal history of allergy increases this risk in boys. This sex-specific effect was consistently present at most or all ages. When we performed repeated measures analyses and adjusted this for various confounders to get an overall independent effect, we found statistically significant sex specific effects for both asthma and eczema. The effect on the opposite sex (maternal to boys and paternal to girls) was weak and non-significant. On a multiplicative scale of interaction, for asthma the paternal effect was stronger in boys and for eczema the maternal effect was statistically stronger in girls. Rhinitis, however, showed a dominant paternal effect for both boys and girls. The sex-specific effect was further confirmed by extending the observation to objective tests of atopy (allergic sensitization on skin test and total IgE), which had a similar effect. For allergic sensitization assessing differences on a multiplicative scale of interaction, the maternal effect was statistically stronger in girls.
The strengths of our study are longitudinal phenotyping from birth to 18 years of age using validated tools, objective assessment for atopy and availability of information on early life environmental factors, which can confound the effect of genetic factors.21,30 Further, the information on parental history of allergy was ascertained at birth and therefore, avoids the risk of biased reporting following subsequent development of allergic disease in children. In general, all potential misclassification could be affected by the parental history. Hence, parents with symptoms recognize and report symptoms in their offspring more often than those without symptoms (overestimation of heritability). However, this does not account for the differences seen between maternal and paternal history of the disease and not for the association of maternal IgE with IgE in boys and girls at ages 10 and 18 year.
Loss to follow-up in this longitudinal cohort study was remarkably low with information obtained from 83% to 94% children at various follow-ups up to age 18. However, it is possible that among the 10–15% children who did not attend, there was selective loss to follow-up influenced by parental disease or childhood disease. To explore this further, we used maternal asthma as the index disease but did not find any statistically significant difference in attrition rates among children with and without maternal asthma at each follow-up (Table E4). Similarly, we did not find any difference in attrition rates among children with and without an asthma diagnosis at the previous follow-up (Table E5). Thus, there was no selective follow-up due to parental or childhood disease. A limitation of our study is that parental allergic disease was reported (by the mother), rather then physician diagnosed with the risk of over reporting. However, the prevalence in both mother and father was around 10%, which is less than what has been reported previously in UK studies.4,17 Further limitations include unavailability of skin prick tests on all children at ages 1 and 2 and paternal total IgE at birth.
More studies report a closer association of childhood asthma with maternal asthma.5–6 There could be several reasons for this. Inclusion of early childhood wheeze may skew the effect. In our cohort, wheeze at 1 and 2 years was associated only with maternal asthma (Table E3), while maternal and paternal asthma had similar influence for childhood asthma (Table 1). Our results are consistent with the findings of Litonjua et al, who suggested that maternal history of asthma is associated with childhood asthma (wheeze) at all ages while paternal history assumes significance for asthma in later childhood.3 As early childhood wheeze is largely transient and has different etiologies to childhood asthma, we focused on asthma from 4 to 18 years. Similarly, it is not possible to diagnose allergic rhinitis with confidence in early childhood and therefore, we restricted data analysis to children aged 4 and above for rhinitis. Eczema, on the other hand, is common in early childhood and easy to diagnose. Thus, eczema was included in the analysis from 1 to 18 years. Another explanation for a higher effect of maternal asthma might be a biased parental report of their asthma as questionnaires are often completed by mothers, who may not know if their partners suffered from childhood asthma. We can be confident that this was not the case in our study, as in case of a maternal reporting bias, we would expect a higher prevalence of maternal than paternal asthma. In our data, a life time prevalence (a yes answer to “Have you ever suffered from asthma?”) of maternal asthma was 10.7% and paternal asthma 9.7%, with information available from 98.8% and 97.9% of parents, respectively.
Further explanation of a greater presumed maternal effect is that the pre- and postnatal environment of the child is more closely associated with mothers than fathers. Thus, early life environmental effects, such as exposure to cigarette smoke, may cause asthma symptoms in both the mother and the child and could falsely be regarded as a genetic influence. However, adjustment for common early life environmental factors made no significant difference to the association observed for childhood asthma in our study, which further supports the validity of the findings (Tables I to VIII).
Most previous studies investigating the parent of origin effects assessed childhood asthma at one time point.3,7,8,31 Ignoring the sex of the offspring, cross-sectional analysis of our data indicates nearly equal maternal and paternal effects (Table 1), which may conceal the differential effect of parental asthma unless stratified by sex of the child. Similarly, for childhood eczema, maternal and paternal eczema had comparable influence in our study. This is consistent with previous large studies.16,17 To our surprise, childhood rhinitis was associated only with paternal rhinitis. A reporting bias, by the mothers, would amplify maternal, rather than paternal effects. The effect seems to be consistent throughout childhood and we believe that there is a greater paternal rhinitis influence. Previous cross-sectional studies that have investigated parent of origin effect in allergic rhinitis did not find a significant difference.2,7 Given that allergic sensitization, and to a lesser extent total IgE, also show a differential parental effect based on child’s sex, our findings are supportive of an underlying immune mechanism driving the effect on asthma and eczema. Why rhinitis behaves differently to asthma and eczema is not clear.
There are three studies that have investigated the parent of origin effect in asthma with respect to sex of the child. Melen et al, found a significant interaction between male sex and parental asthma but their cohort was confined to age 1 to 4 years and thus dominated by early childhood wheezers.32 Bjerg et al, performed a sex stratified analysis in a cross-sectional study of 7–8 years old but did not find any major difference in the effect of parental asthma.8 Mandhane et al, found a significant effect of maternal atopy (asthma or rhinitis) for childhood wheeze in both sexes but the association of paternal atopy was restricted to boys.20 They did not analyze the effect of parental asthma alone on childhood asthma (which may be more specific); neither did they assess other allergic disease or sensitization. Only one study addressed parent of origin effect stratified by child’s sex for all three allergic conditions and did not find any significant difference.7 However, this was a cross-sectional questionnaire survey with no longitudinal information, no supporting objective tests of allergic sensitization or IgE and analyses were not adjusted for confounding variables. This differential effect of parent of origin may be due to genetic differences, intrauterine programming in mothers, or interaction with environmental factors.9,21 It is plausible that epigenetic factors such as DNA methylation, which provide heritable information beyond the DNA sequence, contributes to the effect of parental disease status on risk of disease in offspring.33,34 Imprinting is an epigenetic modification that is parental origin specific, leading to preferential expression of a specific parental allele in somatic cells of the offspring.35 The vital role of imprinted genes in mammalian prenatal growth and development is shown most clearly by the abnormal development and early demise of embryos that inherit two copies of either a maternal or paternal genome, rather than the usual one of each.36 Furthermore, sex-specific imprinting effects have previously been observed in mammals.37 Other possible explanations of parent of origin effects include trans-generational genetic and epigenetic effects.38 Such effects have been postulated to account for the “missing heritabilty” observed when considering the proportion of heritability accounted for by loci identified in genome-wide association studies of complex disease.39
It is possible that the paternal effect is primarily (epi)-genetic, whereas the maternal effect is also influenced by shared pre- and postnatal environment. These environmental effects could include the effectiveness of the placenta in delivering nutrients to support fetal growth40, maternal smoking impeding antenatal growth41, or exposure of the developing airways to oxidant stress such as acetaminophen.42 These prenatal exposures may disproportionately affect children of asthmatic mothers and may cause wheeze in early childhood in both boys and girls. This is supported by our observations where early childhood wheeze was associated with maternal, but not paternal, asthma (Table E3). In contrast, the effect of paternal asthma on disease status, possibly mediated by paternal imprinting, would manifest later in childhood with development of persistent clinical symptoms of asthma in boys, while the same is true for girls associated with maternal asthma.
An accurate assessment of the influence of parental atopy on various allergic manifestations is critical in facilitating early diagnosis and accurate prognosis. At present, child’s sex is not considered when assessing the influence of parental allergic disease on child risk of the disease. We have shown that this is critical as maternal asthma does not increase the risk of asthma in boys and paternal asthma has no effect on the risk of asthma in girls. Similar effects are observed for eczema, atopy and, to a lesser extent, total IgE. This information is also important to consider when identifying children at high risk for preventive measures. Lastly, it should help us to understand the pathogenesis of the different patterns of childhood asthma and allergies. While early childhood wheeze (1–2 years) was associated with maternal asthma, childhood asthma showed the differential effect. For eczema, atopy and total IgE, this differential effect was observed for all time points that were analyzed. These allergic diseases are driven by atopic immune responses and possibly share genetic and epigenetic mechanisms (atopic diathesis). More studies, focusing on transgenerational epigenetic programming, are needed to critically appraise the differential effect of parental atopy, its origin in (epi)genetics and influence of maternal environment.
We need to consider a child’s sex when assessing the influence of parental atopic history on childhood allergy. This is relevant for clinical assessment, identifying children for prevention and understanding the genetic basis of atopy.
Funding sources: The Isle of Wight birth cohort assessments have been supported by the National Institutes of Health USA (Grant no. R01 HL082925 and R01AI091905-01) and Asthma UK (Grant no. 364).
We would like to acknowledge the generosity of the participants and their families of the 1989 Isle of Wight Birth cohort in being part of these assessments. We appreciate help of all the staff at The David Hide Asthma and Allergy Research Centre in undertaking this assessment. We are grateful to ALK-Abello who kindly supplied the skin prick testing reagents and Phadia who donated kits for measurements of total IgE. These organizations had no input into the study design or analysis of the data presented in the manuscript.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.