Association between birthweight and later body mass index: an individual-based pooled analysis of 27 twin cohorts participating in the CODATwins project

Aline Jelenkovic, Yoshie Yokoyama, Reijo Sund, Kirsi H. Pietiläinen, Yoon‐Mi Hur, Gonneke Willemsen, Meike Bartels, Maties Torrent, Syuichi Ooki, Kimberly J. Saudino, Maria A. Stazi, Corrado Fagnani, Cristina D’Ippolito, Tracy L. Nelson, Keith E. Whitfield, Ariel Knafo‐Noam, David Mankuta, Lior Abramson, Kauko Heikkilä, Tessa L. Cutler, John L. Hopper, Jane Wardle, Clare Llewellyn, Abigail Fisher, Robin P. Corley, Brooke M. Huibregtse, Cathérine Derom, Robert Vlietinck, Ruth J. F. Loos, Morten Bjerregaard-Andersen, Henning Beck‐Nielsen, Morten Sodemann, Ádám Domonkos Tárnoki, Dávid László Tárnoki, S. Alexandra Burt, Kelly L. Klump, Juan R. Ordoñana, Juan F. Sánchez-Romera, Lucía Colodro‐Conde, Lise Dubois, Michel Boivin, Mara Brendgen, Ginette Dionne, Frank Vitaro, Jennifer R. Harris, Ingunn Brandt, Thomas Sevenius Nilsen, Jeffrey M. Craig, Richard Saffery, Finn Rasmussen, Per Tynelius, Gombojav Bayasgalan, Narandalai Danshiitsoodol, Claire M. A. Haworth, Robert Plomin, Fuling Ji, Feng Ning, Zengchang Pang, Esther Rebato, Robert F. Krueger, Matt McGue, Shandell Pahlen, Dorret I. Boomsma, Thorkild I. A. Sørensen, Jaakko Kaprio, Karri SilventoinenInternational Journal of Epidemiology 2017 · DOI

TL;DR

This study leverages a massive twin dataset to investigate the long-term impact of birthweight on Body Mass Index (BMI). It reveals that while birthweight is consistently linked to later BMI, unique environmental factors, rather than shared genetics, play a significant role in this association, particularly during childhood and adolescence.

What Problem Does This Paper Address?

The increasing global prevalence of obesity has spurred significant research into its origins. One area of interest is the "fetal origins hypothesis," which suggests that conditions during pregnancy, like birthweight, can have lasting effects on adult health, including obesity risk. However, it's been challenging to disentangle the influence of the intrauterine environment from shared genetic factors that also contribute to body size. This paper aims to address this by using twin studies, a powerful tool for separating genetic and environmental influences, to understand the association between birthweight and BMI from infancy through adulthood.

Methods

The researchers employed a pooled analysis of data from 27 twin cohorts across 17 countries, creating a large dataset of over 78,000 twin individuals. They used linear regression analyses at two levels:

  • Individual Level: Examining the association between an individual's birthweight and their later BMI, controlling for factors like birth year and twin cohort.
  • Within-Pair Level: Analyzing the differences in birthweight and BMI between co-twins. This is crucial because it controls for shared genetic and family environmental factors, isolating the impact of individual-specific environmental influences.

They also conducted secondary analyses using birthweight standardized for gestational age and ponderal index (PI) at birth to explore these relationships further.

Datasets

The study utilized data from the CODATwins project, which pooled information from 27 twin cohorts worldwide. The dataset included:

  • 78,642 twin individuals (20,635 monozygotic and 18,686 same-sex dizygotic twin pairs).
  • Birthweight information for these individuals.
  • A total of 214,930 BMI measurements collected at ages ranging from 1 to 49 years.

The paper does not introduce new datasets or benchmarks.

Key Insights and Ideas

  • Birthweight is a persistent predictor of BMI: The study confirms a robust, positive, and linear association between birthweight and BMI that extends from infancy into adulthood.
  • Environmental factors are key: The most striking finding is that the association between birthweight and later BMI is largely explained by environmental factors unique to each individual, particularly evident in the within-pair analyses of monozygotic (MZ) twins (who are genetically identical). This suggests that differences in the intrauterine environment, rather than shared genes, are driving this link.
  • Age matters: The strength of the birthweight-BMI association is strongest in childhood and early adolescence and tends to attenuate in adulthood. This implies that while early life conditions set a trajectory, other factors become more influential later in life.
  • Sex and zygosity differences are minimal for the core association: The association between birthweight and BMI was similar in males and females, and the within-pair analyses showed comparable results for monozygotic (MZ) and dizygotic (DZ) twins, further supporting the role of individual-specific environments.

Key Results

  • At the individual level, a 1-kg increase in birthweight was associated with up to a 0.9 kg/m² higher BMI.
  • Within twin pairs, this association was even stronger, with regression coefficients generally greater (up to 1.2 kg/m² per kg birthweight).
  • The within-pair associations were similar in both MZ and DZ twins, indicating that shared genetic factors do not strongly explain the link.
  • The association was generally stronger in childhood and early adolescence and weaker in adulthood.

Limitations and Future Work

  • Generalizability: While the study boasts a large sample size, the database is heavily weighted towards Caucasian populations with Westernized lifestyles, limiting generalizability to other populations.
  • Data Source: Many measures, particularly birth data, were parentally reported, which can introduce recall bias, although the authors note high kappa values for birthweight recall in singletons.
  • Measurement of Body Composition: Birthweight is a proxy and doesn't fully capture body composition. While ponderal index was used, BMI itself is also a proxy for body fat.
  • Future Directions: The authors suggest that identifying specific intrauterine environmental factors that influence later BMI could be crucial for understanding and potentially intervening in the development of obesity. Further research into epigenetic mechanisms and the role of placental differences in twin pregnancies could also be valuable.

Summary updated

Abstract

These findings indicate that environmental factors unique to each individual have an important role in the positive association between birthweight and later BMI, at least until young adulthood.
BibTeX
@article{jelenkovic2017association,
  author = {Aline Jelenkovic and Yoshie Yokoyama and Reijo Sund and Kirsi H. Pietiläinen and Yoon‐Mi Hur and Gonneke Willemsen and Meike Bartels and Maties Torrent and Syuichi Ooki and Kimberly J. Saudino and Maria A. Stazi and Corrado Fagnani and Cristina D’Ippolito and Tracy L. Nelson and Keith E. Whitfield and Ariel Knafo‐Noam and David Mankuta and Lior Abramson and Kauko Heikkilä and Tessa L. Cutler and John L. Hopper and Jane Wardle and Clare Llewellyn and Abigail Fisher and Robin P. Corley and Brooke M. Huibregtse and Cathérine Derom and Robert Vlietinck and Ruth J. F. Loos and Morten Bjerregaard-Andersen and Henning Beck‐Nielsen and Morten Sodemann and Ádám Domonkos Tárnoki and Dávid László Tárnoki and S. Alexandra Burt and Kelly L. Klump and Juan R. Ordoñana and Juan F. Sánchez-Romera and Lucía Colodro‐Conde and Lise Dubois and Michel Boivin and Mara Brendgen and Ginette Dionne and Frank Vitaro and Jennifer R. Harris and Ingunn Brandt and Thomas Sevenius Nilsen and Jeffrey M. Craig and Richard Saffery and Finn Rasmussen and Per Tynelius and Gombojav Bayasgalan and Narandalai Danshiitsoodol and Claire M. A. Haworth and Robert Plomin and Fuling Ji and Feng Ning and Zengchang Pang and Esther Rebato and Robert F. Krueger and Matt McGue and Shandell Pahlen and Dorret I. Boomsma and Thorkild I. A. Sørensen and Jaakko Kaprio and Karri Silventoinen},
  title = {{Association between birthweight and later body mass index: an individual-based pooled analysis of 27 twin cohorts participating in the CODATwins project}},
  journal = {International Journal of Epidemiology},
  year = {2017},
  doi = {10.1093/ije/dyx031},
}
Figure