Cohort Profile: The Québec Birth Cohort on Immunity and Health (QBCIH)

The Québec Birth Cohort on Immunity and Health (QBCIH) was set up in 2010 to investigate the effect of immune stimulation in early life on the risk of developing selected autoimmune and inflammatory diseases, and the initial focus was on occurrence of asthma and diabetes through early adulthood (20 years of age). The Bacillus Calmette-Guérin (BCG) vaccine is an exogenous modulator of the immune response. Unlike most other vaccines which stimulate Th2 lymphocytes and antibody production, the BCG vaccine induces Th1 lymphocytes and cell-mediated immunity.1–4 Long-lasting effects of BCG vaccination on Th1 and Th17 immune responses were recently shown.5 Overtly predominant Th1 and/or Th17 responses characterize several autoimmune diseases (e.g., multiple sclerosis and type 1 diabetes), whereas an immune response skewed towards Th2 is prototypic of atopic diseases.6 The modulation of the immune maturation and function—as resulting from BCG vaccination—and its role in the development of autoimmune or inflammatory diseases has been the object of considerable research interest. Inconclusive results were however found in a 2008 qualitative review of the epidemiological evidence on a possible link between BCG vaccination and several diseases characterized by autoimmune or inflammatory processes.7 Human population studies8–10 failed to confirm an expected protective effect of BCG vaccination on diabetes which was observed in animal models.11,12 As well, several systematic reviews and meta-analyses of epidemiological studies on the relation between BCG vaccination and childhood asthma showed conflicting results; protective13,14 or no associations15,16 were observed. In this context, the overarching aim of the QBCIH was to assess the occurrence of autoimmune and inflammatory diseases in relation to a non-specific stimulation of the immune function in early age, as resulting from BCG vaccination. More specifically, it was originally set up to address three aims: (i) to estimate the association between BCG vaccination and childhood diabetes; similarly, (ii) to assess the relationship between BCG vaccination and childhood asthma, adjusting for individual and area-level potential confounders obtained from provincial administrative demographic and health databases; and (iii) given the multifactorial aetiology of asthma and hence a greater likelihood of confounding, to evaluate the impact of further adjusting the measure of association between BCG vaccination and childhood asthma by collecting data on additional potential confounders unavailable in administrative databases. Designed and led by one of the authors (M.C.R.) at Institut National de la Recherche Scientifique—Institut Armand-Frappier (INRS-IAF), the QBCIH was supported by an infrastructure grant from the Canada Foundation for Innovation and the Québec Ministry of Education, Leisure and Sports, and by governmental research funds from the Canadian Institutes of Health Research (CIHR), the Fonds de Recherche du Québec-Santé (FRQS), and an in-kind contribution from Statistics Québec (Institut de la Statistique du Québec). The cohort establishment and subsequent analyses were approved by the Commission d’Accès à l’Information (CAI), the governmental body overseeing access to and protection of personal information, and the research ethics committees of INRS, FRQS, the public health care provider (Régie de l’Assurance Maladie du Québec, RAMQ) and Statistics Québec. A confidentiality agreement protecting the privacy of research subjects was signed by all members of the research team. All persons born in 1974 in the province of Québec, Canada, after at least 32 weeks of gestation, were eligible. The QBCIH retrospective birth cohort was assembled through linkage of the provincial Birth Registry and the 2010 Healthcare Registration File (universal public health system) (Figure 1). In the absence of a unique identification number, probabilistic linkage allows combining data from various sources, using several identifiers. Linkage is based on the likelihood, expressed as ‘match weights’, that two records are from the same individual when considering the level of agreement for all the identifiers used. The process of probabilistic record linkage and the underlying statistical framework have been described previously.17 For the QBCIH, probabilistic record linkage was carried out by Statistics Québec analysts using five basic nominal identifiers (surname and given name, date of birth, sex, father’s given name) with the ‘Generalized Record Linkage System’ software developed by Statistics Canada.18 Probabilistic linkage was also carried out to incorporate data from the Québec BCG Vaccination Registry. In the absence of linkage, subjects were considered unvaccinated. When agreement was partial and the estimated match weights did not reach the established threshold for successful linkage, subjects were excluded as their vaccination status could not be determined. Before assembling the current cohort, linkage between the Québec BCG Vaccination Registry and the above-mentioned administrative databases was shown to be highly feasible and successful.19 Establishment of the Québec Birth Cohort on Immunity and Health: overview of record linkage process. BCG, Bacillus Calmette-Guérin; INRS-IAF, Institut National de la Recherche Scientifique – Institut Armand-Frappier; ISQ, Institut de la Statistique du Québec, Statistics Québec; Med-Écho, Maintenance et Exploitation des Données Pour l’Étude de la Clientèle Hospitalière [meta-data on hospitalizations from the public health care provider]; QBCIH, Québec Birth Cohort on Immunity and Health; RAMQ, Régie de l’Assurance Maladie du Québec [public health care provider]; Survey, ‘Survey on childhood environment and the development of allergic diseases’. Of the 90 060 eligible individuals found in the Birth Registry, 82 628 (92%) were linked with the Healthcare Registration File. When comparing them with individuals for whom linkage was unsuccessful (8% of eligible subjects), small differences were observed in terms of sex, birthweight, gestational age and parent’s age. These characteristics varied by less than 2% between the two groups, except for mothers’ age (the mothers of subjects with successful linkage were 3 years older on average, a difference of 11.6%). After linkage with the Québec BCG Vaccination Registry, the QBCIH included 81 496 subjects, representing 90.5% of those who were initially eligible. The QBCIH also benefits from the collection of information on a wide range of additional factors for a subsample of 1643 subjects. To document potential confounders of the BCG vaccination-asthma association which were unavailable in administrative databases, we applied a two-stage sampling strategy with a balanced design.20,–22 We randomly selected subjects among four groups defined by BCG vaccination and asthma status, over-sampling from the smallest cells of the 2 x 2 table (BCG vaccinated and non-vaccinated individuals with asthma), thus maximizing precision for the adjusted estimates of association.23 All randomly selected subjects received an invitation letter explaining the purpose of our ‘Survey on childhood environment and the development of allergic diseases’, and inviting them to participate. Trained interviewers from Statistics Québec then contacted potential participants by phone. After providing a verbal consent, participants completed a computer-assisted telephone interview in either French or English (between 26 April 2012 and 7 July 2012). Answers were entered in real-time using the Interviewer software (Voxco, Montréal, QC). The questionnaire incorporated questions based on and adapted from the standardized validated questionnaire of the International Study of Asthma and Allergy in Childhood,24 other validated questionnaires25,26 and questions developed by Statistics Québec,27 members of our research team28 and other researchers.29–31 A detailed description of the methodology, data collection and evaluation of the validity of the Stage 2 sample have been recently published.32 Briefly, 1643 individuals participated in the Stage 2 survey. Among those with valid phone numbers, participation rates were 58% for vaccinated subjects with asthma (n = 446), 55% for non-vaccinated subjects with asthma (n = 385), 58% for vaccinated subjects without asthma (n = 447) and 54% for non-vaccinated subjects without asthma (n = 365). The overall participation rate, calculated by dividing the number of completed interviews by the number of potentially eligible subjects, was respectively 37%, 32%, 37% and 30%. There were slight differences between Stage 2 participants and the source population in demographic, geographical, socioeconomic and health aspects, but these differences were present across all strata. Consequently, the effect of non-participation on associations between these factors and asthma was small, and the Stage 2 sample was found to be valid.32 There were two different sources of data for subjects in the QBCIH: (i) retrospective follow-up through extraction of data from administrative sociodemographic and health databases; and (ii) self-reported information. In addition to constituting the base population for the cohort, linkage with the Birth Registry made it possible to extract some perinatal and sociodemographic information. Information about BCG vaccination was retrieved from the provincial registry. Other data acquired from administrative databases included medical and pharmaceutical services, hospitalizations and death if applicable. Pertaining to the entire cohort, Table 1 provides an overview of the type of information extracted from the following databases: birth and death registries, the Healthcare Registration File, prescription drug claims, medical services claims, hospitalization data and the Québec BCG vaccination registry. Data available in the QBCIH, acquired from administrative demographic and health databases AHFS, American Hospital Formulary Service; ICD-9, International Classification of Diseases–9th edition; Med-Écho, Maintenance et Exploitation des Données pour l’Étude de la Clientèle Hospitalière; RAMQ, Régie de l’Assurance Maladie du Québec. Data available in the QBCIH, acquired from administrative demographic and health databases AHFS, American Hospital Formulary Service; ICD-9, International Classification of Diseases–9th edition; Med-Écho, Maintenance et Exploitation des Données pour l’Étude de la Clientèle Hospitalière; RAMQ, Régie de l’Assurance Maladie du Québec. Characteristics of the QBCIH study population and the Stage 2 sample, as documented from administrative databases, are shown in Table 2. Half of the QBCIH subjects were men (51%). The vast majority of subjects were born from parents who were themselves born in the province of Québec (88% of the subjects’ mothers and 86% of their fathers). In the Stage 2 sample, 42% of the participants were men and again, most participants’ parents were born in the province of Québec (93% of the subjects’ mothers and 91% of their fathers). Characteristics of the Québec Birth Cohort on Immunity and Health (QBCIH) study population, and the Stage 2 sample Compilation based on data from the ©Government of Québec, Statistics Québec, 2012. Statistics Québec is not responsible for compilations or interpretation of results. Numbers in table do not necessarily sum to the total number of subjects due to missing/invalid values. Birthweight: extremely low (< 1000 g), very low (1000-< 1500 g), low (1500-< 2500 g), normal (2500-≤ 4200 g), and high (> 4200 g). Gestational age: premature (< 37 weeks), at term (37-41 weeks), and post term (> 41 weeks). Determined using the second character of the subjects’ postal code (RAMQ) (0: rural, ≠0: urban). Estimated by ‘median household income’ from the 1991 Canadian census, using the first three characters of subjects’ postal code (RAMQ). Characteristics of the Québec Birth Cohort on Immunity and Health (QBCIH) study population, and the Stage 2 sample Compilation based on data from the ©Government of Québec, Statistics Québec, 2012. Statistics Québec is not responsible for compilations or interpretation of results. Numbers in table do not necessarily sum to the total number of subjects due to missing/invalid values. Birthweight: extremely low (< 1000 g), very low (1000-< 1500 g), low (1500-< 2500 g), normal (2500-≤ 4200 g), and high (> 4200 g). Gestational age: premature (< 37 weeks), at term (37-41 weeks), and post term (> 41 weeks). Determined using the second character of the subjects’ postal code (RAMQ) (0: rural, ≠0: urban). Estimated by ‘median household income’ from the 1991 Canadian census, using the first three characters of subjects’ postal code (RAMQ). The two main outcomes under study were diabetes and asthma. For diabetes, two algorithms validated in Canada were applied. The first one, used by the Canadian Chronic Diseases Surveillance System, consisted in having two or more diabetes-related medical services within 2 years or one or more diabetes-related hospitalization(s).33 The second algorithm, validated among children and found to be the most specific in this population among 16 evaluated algorithms, consisted in four or more diabetes-related medical services within 2 years.34 For asthma, subjects who had two or more asthma-related medical services or at least one asthma-related hospitalization were considered as having asthma. The validity of asthma definitions based on Canadian administrative health data has been previously demonstrated.35,36 Other secondary outcomes included allergic diseases such as allergic rhinitis, atopic dermatitis and urticaria. Markers of medical events related to main and secondary outcomes were retrieved from medical service claims and prescription drugs data (RAMQ), as well as from hospitalization data (Maintenance et Exploitation des Données pour l’Étude de la Clientèle Hospitalière, MED-ÉCHO). Table 3 shows the prevalence of these primary and secondary outcomes in the QBCIH and the Stage 2 sample, based on use of health services until 1994 (subjects at 20 years of age). Prevalence of primary and secondary outcomes in the QBCIH and the Stage 2 sample based on health services use (1983–94) Compilation based on data from the ©Government of Québec, Statistics Québec, 2012. Statistics Québec is not responsible for compilations or interpretation of results. International Classification of Diseases Ninth Revision. A code with an x as the 4th digit indicates that all possible digits were included (e.g. 493.0, 493.1, 493.2, 493.8, 493.9). Corrected for sampling weights from QBCIH to Stage 2 sample. Prevalence of primary and secondary outcomes in the QBCIH and the Stage 2 sample based on health services use (1983–94) Compilation based on data from the ©Government of Québec, Statistics Québec, 2012. Statistics Québec is not responsible for compilations or interpretation of results. International Classification of Diseases Ninth Revision. A code with an x as the 4th digit indicates that all possible digits were included (e.g. 493.0, 493.1, 493.2, 493.8, 493.9). Corrected for sampling weights from QBCIH to Stage 2 sample. Table 4 summarizes the main domains and specific self-reported variables by Stage 2 participants during the telephone interviews. Although the primary motivation was to collect additional confounder information for the association between BCG vaccination and asthma, these are also highly pertinent to other allergic diseases. Data collected in Stage 2 sample, ‘Survey on childhood environment and the development of allergic diseases’ Data collected in Stage 2 sample, ‘Survey on childhood environment and the development of allergic diseases’ Baseline perinatal and sociodemographic data were available at birth (1974), whereas health encounters such as BCG-related (1974-92), medical and pharmaceutical (1983-94) and hospitalization (1987-94) events were documented as they occurred and available over several years, depending on the time frames of data coverage in these databases. Further, the residential geographical area, based on three-digit postal codes, was documented yearly from 1987. This allowed determining if the residence was in a rural or urban area, as well as conducting further linkages with data from the Canadian Census. The data linkage performed to establish this cohort resulted in obtaining information on a wide range of factors and health outcomes for each subject from birth until 1994, when they were aged 20 years. In the study period, the public health coverage of pharmaceutical services and thus the Prescription Drug Claims database were restricted to individuals over 65 years of age, welfare recipients and their children. There were few losses to follow-up, since the vast majority of the population is covered by the public health care programme (96% in 2011-12).37,38 Only 4% of the cohort either had temporary interruptions in health coverage or were lost to follow-up. The median duration of interruptions was 3.4 years. The questionnaire-based interviews were conducted once in 2012 in a subset of 1643 participants (2% of the QBCIH). Research have related to at the of BCG vaccination and the association of BCG vaccination with diabetes and asthma, in addition to a description and evaluation of the validity of the two-stage sample. We the demographic and perinatal of BCG vaccination among children born in the province of Québec in We also these between subjects who received the BCG vaccine in 1974 as of the vaccination and those who were vaccinated after the programme had In the QBCIH, of subjects were BCG during the programme and after it BCG vaccination during the programme was more among children with parents and those in rural Vaccination after the programme was related to vaccination was more than as among children were all of French with validated algorithms based on health services we BCG vaccination in the first of life was with the occurrence of childhood We found no association using either of the two definitions based on the previously algorithms = = and no difference by study was the second one to address an early time of BCG vaccination when a potential association with diabetes research is considering time of in In this cohort with a two-stage sampling results based on the Stage 1 sample a asthma risk in BCG vaccinated with subjects, adjusting for variables available in administrative databases further for potential confounders collected in the Stage 2 sample showed no association = The initially observed risk of asthma among BCG vaccinated subjects was no present adjusting for several additional confounders unavailable in administrative databases, the of a two-stage We a of due to non-participation in the two-stage sample in the This included a detailed description of non-participation and of characteristics between participants and as well as a of in the associations between sociodemographic characteristics and asthma. In addition to the validity of the two-stage sample in our this as a for in There are to using administrative health databases for epidemiological and some pertinent information not be For the analyses with asthma, we have this by a subset of the study population, obtaining information on potential confounders of the association of interest. to the of prescription drug claims data for a of the study population, which if available have the algorithms for the of coverage of health databases for the health encounters for diseases of that occurred from age to For of asthma or which have by age have been asthma from the Stage 2 interviews this results that was agreement between asthma defined from administrative health data and the early in health databases The QBCIH has a number of its sample population follow-up, as well as the of health outcomes using a the to for potential confounders with information by telephone interviews for a subsample of the initial study population for the analyses on BCG vaccination and BCG vaccination status is from an since of BCG vaccination has low and thus in The QBCIH is unique in the of a on the effects of non-specific stimulation of the immune resulting from BCG on the development of diabetes and asthma in a population a of were vaccinated with BCG and for which the individual records of vaccination are The of the BCG vaccination programme present some for these associations in the Québec the vaccine was of but was not resulting in a of the population unvaccinated. were differences in BCG vaccination but they to on factors than on individual for or BCG In it is that for BCG vaccination have the observed associations of interest. from the QBCIH have the potential of providing and to a of the role of early in life in the development of these diseases and other diseases. Further, this infrastructure to address other research to BCG diabetes, asthma and allergic diseases. For we have analyses on early life in relation to asthma agreement between asthma defined from administrative health databases and from and the identification of asthma based on use of health services for asthma and other allergic other research questions be with the available We are the of the QBCIH to study the association between BCG vaccination and diseases. We are thus the cohort and duration of follow-up. The QBCIH of individuals born between and for whom we the BCG vaccination be up until using administrative health databases for the occurrence of diabetes, asthma, other allergic diseases, multiple sclerosis and inflammatory The resulting cohort of This to address in about the aetiology of a of autoimmune and inflammatory diseases. We research and we the and of the study research team. in or further information are to the and of the study and results from the and at be at or our The QBCIH was set up to investigate the effects of a non-specific immune stimulation in early as resulting from BCG on the risk of developing selected autoimmune and inflammatory diseases. in the QBCIH 81 496 individuals born in after at least 32 weeks of gestation, in the province of Québec, Data acquired from administrative databases perinatal and sociodemographic information, BCG medical services hospitalizations and prescription drugs for diabetes, asthma and other allergic diseases, as well as death if applicable. health encounters was documented from the public health databases from to Only 4% of the cohort had either temporary interruptions in health coverage or were lost to follow-up. a two-stage sampling strategy with a balanced among four groups defined by BCG vaccination and asthma status, telephone interviews were conducted with a subsample of 1643 subjects. confounders of the BCG vaccination-asthma association unavailable in administrative databases were is by data privacy governmental but we for For the This was supported by an infrastructure grant from the Canada Foundation for Innovation and the Québec Ministry of Education, Leisure and and research from the Canadian Institutes of Health Research and Fonds de Recherche du Québec-Santé and through a with Statistics Québec. When conducting the described in this and were recipients of from of We and du from Statistics Québec, as well as from RAMQ, for their contribution to various of the establishment of the We also to our at the Health for the diabetes and of the for the asthma

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.