Polygenic risk for obesity and its interaction with lifestyle and sociodemographic factors in European children and adolescents

Anke Hüls, Marvin N. Wright, Leonie H. Bogl, Jaakko Kaprio, Lauren Lissner, Dénes Molnár, Luís A. Moreno, Stefaan De Henauw, Alfonso Siani, Toomas Veidebaum, Wolfgang Ahrens, Iris Pigeot, Ronja ForaitaInternational Journal of Obesity 2021 · DOI

TL;DR

This study investigates how genetic predisposition to obesity interacts with lifestyle and sociodemographic factors in European children. It reveals that a healthy environment, characterized by good diet and limited screen time, can significantly mitigate the genetic risk for obesity, particularly in children with a high genetic predisposition.

What Problem Does This Paper Address?

Childhood obesity is a complex issue influenced by a combination of genetic and environmental factors. While twin studies have long suggested a significant genetic component, the precise mechanisms by which genes interact with lifestyle and socioeconomic factors, especially during childhood and adolescence, remain incompletely understood. Previous research has often relied on genome-wide significant variants, which capture only a fraction of the genetic influence. This study aims to address this gap by utilizing a genome-wide polygenic risk score (PRS) to capture a broader genetic predisposition and examine its interactions with various environmental factors in a large European cohort of children and adolescents. Understanding these interactions is crucial for developing targeted prevention strategies early in life.

Methods

The researchers employed a genome-wide polygenic risk score (PRS) approach, specifically the PRS-Khera, which aggregates the effects of millions of single nucleotide polymorphisms (SNPs) associated with Body Mass Index (BMI). This PRS was calculated for each participant using summary statistics from large-scale genome-wide association studies (GWAS) of BMI. To analyze the associations and interactions, generalized linear mixed models were used. These models are suitable for handling repeated measurements from individuals and account for genetic relatedness and population stratification, which are important considerations in genetic studies. The study also performed sensitivity analyses using different PRS calculations to ensure the robustness of their findings.

Datasets

The study utilized data from the pan-European IDEFICS/I.Family cohort, which includes 8609 repeated observations from 3098 children aged 2–16 years. This cohort was recruited from various European countries (Belgium, Cyprus, Estonia, Germany, Hungary, Italy, Spain, and Sweden). Genotyping was performed on a subset of participants using the UK Biobank Axiom array. The study collected data on various sociodemographic factors (parental education, region of residence), lifestyle factors (dietary intake including fruit and vegetable score and fiber intake, physical activity measured by accelerometers, and screen time), and anthropometric measurements (BMI and waist circumference).

Key Insights and Ideas

  • PRS as a powerful tool: The study demonstrates that a polygenic risk score (PRS) is a more comprehensive measure of genetic predisposition to obesity than relying solely on genome-wide significant variants. This allows for the detection of interactions that might otherwise be missed.
  • Environmental buffering effect: A healthy childhood environment can significantly "buffer" or offset the genetic predisposition to obesity. This is a crucial finding for public health interventions, suggesting that environmental modifications can have a substantial impact.
  • Differential susceptibility: Children with a higher genetic risk for obesity are not uniformly affected by environmental factors. Their susceptibility is modulated by specific sociodemographic and lifestyle factors, highlighting the importance of personalized approaches.
  • Early life interventions are key: The study emphasizes that these interactions are particularly important during childhood and adolescence, the critical developmental phases for establishing long-term health trajectories.

Key Results

  • The PRS-Khera explained a significant portion of the variance in BMI (r² = 0.11) and waist circumference (r² = 0.09).
  • Children in Southern Europe showed a higher genetic liability to obesity compared to those in Central Europe (interaction p = 0.0066).
  • Children of parents with low educational attainment had a higher genetic liability to obesity compared to those with high educational attainment (interaction p = 0.0012).
  • Increased dietary fiber intake attenuated the genetic liability to obesity (interaction term for BMI: -0.02, p < 0.05).
  • Shorter screen times also attenuated the genetic liability to obesity (interaction term for BMI: 0.02, p < 0.05).
  • Interactions with sex, medium parental education, and Northern European region of residence were not significant.
  • Interactions with fruit and vegetable score and moderate-to-vigorous physical activity (MVPA) were not significant, possibly due to reduced statistical power for MVPA.

Limitations and Future Work

The authors acknowledge several limitations. Measurement errors in self-reported lifestyle behaviors (e.g., screen time) are inevitable and can bias interaction effects towards the null, reducing statistical power. The study also used a complete-case analysis strategy, which could potentially bias estimates. Furthermore, the assessment of physical activity (MVPA) was only available for a subset of the participants, limiting the power to detect interactions with this factor.

Future work could focus on refining dietary and physical activity assessments to improve accuracy and reduce measurement error. Exploring the underlying biological mechanisms of these gene-environment interactions would also be valuable. The authors suggest that system-level interventions, rather than solely individual-focused strategies, might be more effective given the influence of sociodemographic factors like parental education and region of residence.

Summary updated

Abstract

Our results highlight that a healthy childhood environment might partly offset a genetic predisposition to obesity during childhood and adolescence.
BibTeX
@article{huls2021polygenic,
  author = {Anke Hüls and Marvin N. Wright and Leonie H. Bogl and Jaakko Kaprio and Lauren Lissner and Dénes Molnár and Luís A. Moreno and Stefaan De Henauw and Alfonso Siani and Toomas Veidebaum and Wolfgang Ahrens and Iris Pigeot and Ronja Foraita},
  title = {{Polygenic risk for obesity and its interaction with lifestyle and sociodemographic factors in European children and adolescents}},
  journal = {International Journal of Obesity},
  year = {2021},
  doi = {10.1038/s41366-021-00795-5},
}
Figure