Establishing a birth cohort to investigate the course and aetiology of asthma and allergies across three generations – rationale, design, and methods of the ACROSSOLAR study

Atopic diseases such as asthma, allergic rhinitis, and atopic dermatitis are the most frequent illnesses among children and adolescents with about one in four children being affected [1, 2]. Their prevalence varies considerably across different regions and populations [3‐5]. While asthma and allergies have traditionally been regarded as morbidities of high-income countries, an increase in their prevalence has recently been observed in low- and middle-income countries – a fact that might lead to severe public health problems for the affected regions [2, 6‐8]. Moreover, these diseases are also causing substantial economic costs, both in terms of direct costs such as treatments costs and indirect costs such as reduction of work productivity [9, 10].

Concerning the aetiology of these diseases, the current state of knowledge assumes that they are caused by a complex interaction between genetic and environmental factors such as pre- and postnatal maternal smoking or exposure to damp and mould [11‐16]. These interactions are understood to be especially powerful during the prenatal period and the first years of life [17‐23]. Therefore, these early years are a crucial period when investigating potential risk factors for asthma and allergies. Such investigations should be especially done under consideration of epigenetic mechanisms as these constitute a novel and promising approach to further disentangle the origin and progression of asthma and allergies [24‐27]. Out of several potential epigenetic markers, DNA methylation is the most prominent and most frequently investigated one [24, 26‐28]. Regarding asthma, there is notable evidence for an association of DNA methylation with various asthma-related phenotypes [29‐31]. However, the role of DNA methylation during asthma development as well as underlying biological mechanisms requires further elucidation [26, 32]. Birth cohort studies following the role of DNA methylation prospectively and across generations are the ideal approach to answer these questions [26].

The cross-generational aspect is especially intriguing since there is some first indication of an effect not only of maternal exposures but also of grandmaternal exposures, especially smoking, on subsequent diseases in the offspring [33‐36]. Such effects across three generations are also assumed to be mediated by epigenetic modifications [37‐39]. Further investigating these aspects can substantially increase the understanding of the aetiology, treatment, and prevention of asthma and allergies [37].

The ACROSSOLAR (Atopic Childhood Disease Risk Factors in the Second Generation of the SOLAR Participants) project described here thus aims at establishing a prospective population-based birth cohort that collects information across three generations. To achieve this aim, the study builds upon an existing cohort which has been recruited for the German branch of Phase II of the International Study of Asthma and Allergies in Childhood (ISAAC) and in the meantime has been followed up twice by the Study on Occupational Allergy Risks (SOLAR) [40‐45]. This approach provides the unique opportunity to assess the effects of genetic predisposition, epigenetic factors, and environmental factors such as exposure to environmental tobacco smoke, living conditions, and parental occupation in a prospective and cross-generational study. Accordingly, the main objectives of ACROSSOLAR are:

The present paper describes the design of the ACROSSOLAR project giving concise and detailed information on the study population, recruitment methods, and study instruments. Furthermore, the challenges in the study design are discussed.

Atopic diseases not only continue to have a high prevalence in high-income countries, but they are also gaining importance in low- and middle-income countries. They are thus a substantial burden of disease on a global scale. At the same time, it is still not fully understood so far which factors contribute to their development. What has been shown rather reliably is that genetic predisposition, exposures during the first years of life, epigenetic mechanisms, and gene-environment interactions may be crucial. Research aiming at precisely disentangling their role is thus urgently needed and requires innovative methods. Birth cohorts are an approach that is especially suited to tackle this issue [67]. They allow investigating the effect of genetic and environmental factors as well as their interactions on the course of atopic diseases from the early life years onwards.

The ACROSSOLAR study presented here provides the unique opportunity to establish such a birth cohort in the offspring of an existing cohort. Including the parent generation, ACROSSOLAR can also benefit of a wide range of detailed socio-demographic, occupational, and health-related data on the parents and even grandparents of the children that have already been collected during several follow-ups. It is thus possible to follow the course of atopic diseases and potential risk factors across three generations. Another strength of our approach is that not only do we build upon the experience and infrastructure of ISAAC phase II and SOLAR phases I and II but also that the ACROSSOLAR questionnaire is based on validated instruments that have been successfully used in similar projects [56, 57]. Providing a web-based version of the questionnaire is another asset of our study since it has been shown that it becomes increasingly popular among participants to complete surveys online [68]. Also the SurveyMonkey software itself was successfully utilised in previous medical research projects [69‐71]. Moreover, integrating epigenetic analysis, namely DNA methylation of isolated CD-4+ T cells, across generations will offer the unparalleled opportunity to investigate underlying mechanisms of asthma development in an unique way.

Regarding potential limitations of our approach, there is some indication from SOLAR II that subjects with higher educational level, higher levels of health consciousness, and higher likelihood of atopic diseases were more likely to participate [42]. This means that the original cohort on which the ACROSSOLAR birth cohort is based might be affected by some form of selection towards higher educated people with higher risk of suffering from asthma or allergies. It seems plausible that the same effect might apply to participation in ACROSSOLAR. On the other side, this could be counterbalanced by the effect that people with lower educational levels conceive children earlier than higher educated people so that more families with lower educational status will enrol into the birth cohort. In any case, the potential of selection bias will be carefully taken into account when analysing and interpreting the collected data. Having said that, it is much more important for the validity of the results in prospective cohort studies to maintain a high level of internal validity [72]. In this context, a possibly more severe form of bias could be introduced by loss to follow-up, especially if it differs between exposed and unexposed subjects [73]. Therefore, maximal effort will be spent to keep loss to follow-up as low as possible, e.g. by sending out two reminders to non-respondents at each phase of follow-up. The same applies to the issue of generalisability: One might argue that especially from a geographic or socio-economic point of view our sample is a rather homogenous group that can hardly be compared with populations in low- and middle-income settings. We are albeit confident that our findings will help to advance the understanding of the underlying biologic principles so that our sample would nevertheless be “biologically representative” [62]. In the past, many important and valid conclusions have been drawn from cohorts that were not geographically representative whatsoever [72, 74].

Another challenge of our study will be to accurately coordinate the enrolment and follow-up of participants as the two will run in parallel. This complex approach requires significant administrative and organisational efforts throughout the study. However, as already mentioned we are building upon the experience and equipment from ISAAC Phase II and SOLAR as well as other large-scale research projects in the field of asthma and allergies.

To conclude, the ACROSSOLAR project offers the unique opportunity to build a birth cohort among the children of an existing large-scale cohort. It allows following the course of asthma, allergic diseases, and a wide range of potential risk factors and underlying mechanisms including epigenetics in a prospective and cross-generational way. The study can thus substantially contribute to further enhance the understanding about the determinants of these diseases.