Trends in Parent-Child Correlations of Childhood Body Mass Index during the Development of the Obesity Epidemic

Srensen TIA (2014) Trends in Parent-Child Correlations of Childhood Body Mass Index during the Development of the Obesity Epidemic. PLoS ONE 9(10): e109932. doi:10.1371/journal.pone.0109932 Trends in Parent-Child Correlations of Childhood Body Mass Index during the Development of the Obesity Epidemic Teresa A. Ajslev 0 Lars A ngquist 0 Karri Silventoinen 0 Jennifer L. Baker 0 Thorkild I. A. Srensen 0 David Meyre, McMaster University, Canada 0 1 Institute of Preventive Medicine, Frederiksberg and Bispebjerg Hospital , Frederiksberg , Denmark , 2 Novo Nordisk Foundation Center for Basic Metabolic Research, Faculty of Health and Medical Sciences University of Copenhagen , Copenhagen , Denmark , 3 Population Research Unit, Department of Social Research, University of Helsinki , Helsinki , Finland Background: The intergenerational resemblance in body mass index may have increased during the development of the obesity epidemic due to changes in environment and/or expression of genetic predisposition. Objectives: This study investigates trends in intergenerational correlations of childhood body mass index (BMI; kg/m2) during the emergence of the obesity epidemic. Methods: The study population was derived from the Copenhagen School Health Records Register, which includes height and weight measurements since birth year 1930. Mothers and fathers with BMIs available at ages 7 (n = 25,923 and n = 20,972) or 13 years (n = 26,750 and n = 21,397), respectively, were linked through the civil registration system introduced in 1968 to their children with BMIs available at age 7 years. Age- and sex-specific BMI z-scores were calculated. Correlations were estimated across eight intervals of child birth years (1952-1989) separately by sex. Trends in these correlations were examined. Whereas the mother-child correlations reflected the biological relationship, a likely decline in the assignment of non-biological fathers through the registration system across time must be considered when interpreting the father-child correlations. Results: The BMI correlations between mothers and sons ranged from 0.29-0.36 and they decreased marginally, albeit significantly across time at ages 7-7 years (20.002/year, p = 0.006), whereas those at 13-7 years remained stable (,0.0004/ year, p = 0.96). Mother-daughter correlations ranged from 0.30-0.34, and they were stable at ages 7-7 years (0.0001/year, p = 0.84) and at 13-7 years (0.0004/year, p = 0.56). In contrast, father-son correlations increased significantly during this period, both at ages 7-7 (0.002/year, p = 0.007) and at ages 13-7 years (0.003/year, p,0.001), whereas the increase in fatherdaughter correlations were insignificant both at ages 7-7 (0.001/year, p = 0.37) and at ages 13-7 years (0.001/year, p = 0.18). Conclusion: During the obesity epidemics development, the intergenerational resemblance with mothers remained stable, whereas the father-child BMI resemblance increased, possibly reflecting changes in family relationships, and unlikely to have influenced the epidemic. - Data Availability: The authors confirm that, for approved reasons, some access restrictions apply to the data underlying the findings. The data used in this study are based on a combination of data with personal identification numbers from three third parties, the Copenhagen School Health Record Register, developed and hosted by the Institute of Preventive Medicine, the Central Person Register for Denmark, and the so-called Fertility Database at Statistics Denmark. The reason why the data cannot be freely available at a public repository is because it contains family connections, which include personal identification numbers, delivered by the Central Person Register. According to the Danish law it is not permitted to have such data publicly available. Data are placed on a secure data platform at Statistics Denmark, from which the data cannot be downloaded for analyses outside the services of Statistics Denmark. Access to the particular subset of data, generated by the record linkage at Statistics Denmark, will require a project specific permission from the Institute of Preventive Medicine, from Statistics Denmark, and from Danish Data Protection Agency. Director of the Institute of Preventive Medicine, Professor Thorkild I.A. Srensen and Statistics Denmark through either Charlotte Nielsen or Ivan Thaulow with reference to project number 703266. Funding: This study was funded by The UNIK Food, Fitness and Pharma program, grant 28 (www.foodfitnesspharma.ku.dk). The UNIK project is supported by the Danish Ministry of Science, Technology and Innovation. Funding was also received from Helsefonden (grant number 2008B104). We thank the Danish Governmental Archive for their ongoing work in preserving the Copenhagen School Health Records. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Children resemble their adult parents in body mass index (BMI = weight/height-squared; kg/m2) both during childhood and as adults [19]. Studies of individuals adopted-away early in life [10,11] and studies of twins [12,13] show that this resemblance is due mainly to genetic similarities between relatives. The environment shared by family members apparently has little or no influence on the BMI, perhaps except for some influence when the children still live with their parents in the same household [13]. The obesity epidemic is obviously due to major changes in the environmental determinants either directly or indirectly involved in obesity, but it remains unclear whether this has implied changes in the familial resemblance. The prevailing understanding of the causes of the obesity epidemic is that it is due to societal changes in the so-called obesogenic environment that primarily affect those genetically predisposed to obesity. If so, it would be expected that the familial resemblance in the underlying trait, BMI, would increase during the development of the obesity epidemic. On the other hand, influences from the obesogenic environment independent of the genetic predisposition would lead to less familial resemblance. Previous studies of correlations in BMI between brothers and among twins do suggest that the expression of the genetic predisposition and hence the familial resemblance is greater the greater the obesity prevalence [1416]. In Denmark, the obesity epidemic among children and young men developed in phases, alternating between periods of stable and periods of increasing prevalence of obesity [1719]. The phase-wise development was closely related to birth years, with an increasing prevalence of obesity observed in children born after the early 1940s and after the early 1970s. This opens for the possibility that corresponding changes in BMI during childhood in the parental generation may have influenced BMI in the child gen (...truncated)