Skip to navigation Skip to main content Skip to footer

Approved research

Exploring Diet/Lifestyle factors as causes and modifiers of genetic determinants of obesity and metabolic traits

Principal Investigator: Dr Zoltan Kutalik
Approved Research ID: 16389
Approval date: December 1st 2015

Lay summary

We would like to investigate how environmental factors (e.g. (reported) dietary intake and essential lifestyle factors) correlate with obesity/metabolic trait. We will establish which factors are causes and which are consequences of obesity (i.e. preventive measures) through Mendelian randomisation. Moreover we'd explore which of these environmental factors modify the effect of the genetic risk score on obesity/metabolic traits. The research aims at better understanding of obesity genetics and identify modifiable lifestyle factors causally influencing obesity and other metabolic traits We will apply statistical methods to explore the causal effect of obesity/metabolic traits on environment and visa versa. These methods require genetic data and lifestyle information.This will inform us which modifiable environmental factor are more likely causes/consequences of obesity. Furthermore, we'll identify which of these factors modify association strength between body-mass-index associated SNPs and obesity. We need the entire cohort for this analysis as lifestyle and dietary factors are self reported, thus less reliable. In addition, since individual genetic variants have small impact on these lifestyle factors, we have to maximise the sample size to ensure robust estimates.

Scope extension:

We would like to investigate how environmental factors (e.g. (reported) dietary intake and essential lifestyle factors) correlate with obesity/metabolic trait. We will establish which factors are causes and which are consequences of obesity (i.e. preventive measures) through Mendelian randomisation. Moreover we'd explore which of these environmental factors modify the effect of the genetic risk score on obesity/metabolic traits.

As part of the investigation of the genetic underpinnings of obesity, we would like to extend our search from imputed SNPs to all sequence variants (exome and eventually full genome sequencing) and copy number variations (CNVs) identified from genotype (and eventually sequencing) data.

As we are exploring the consequences of obesity, (common and rare forms of) cardio-metabolic diseases are of central interest, for which we would like to dissect the contribution of environmental, genetic factors (including polygenic risk scores) and their interactions.

We would like to investigate how environmental factors (e.g. (reported) dietary intake and essential lifestyle factors) correlate with obesity/metabolic trait. We will establish which factors are causes and which are consequences of obesity (i.e. preventive measures) through Mendelian randomisation. Moreover we'd explore which of these environmental factors modify the effect of the genetic risk score on obesity/metabolic traits.

 

As part of the investigation of the genetic underpinnings of obesity, we would like to extend our search from imputed SNPs to all sequence variants (exome and eventually full genome sequencing) and copy number variations (CNVs) identified from genotype (and eventually sequencing) data.

 

As we are exploring the consequences of obesity, (common and rare forms of) cardio-metabolic diseases are of central interest, for which we would like to dissect the contribution of environmental, genetic factors (including polygenic risk scores) and their interactions.

 

In order to reliably assess the causes and consequences of obesity [presented in the original scope], we have to be aware of possible biases that may distort the Mendelian randomisation estimates which rely on unbiased SNP-trait associations. One key factor that may bias these association is participation bias and self reporting errors. Hence, we would estimate how these biases can be predicted as a function of cardio-metabolic, obesity and socio-economic traits, which are known to influence participation and self-reporting error.

 

In the original proposal we put strong emphasis on the interplay between genetic and environmental factors influencing complex traits. However, we noticed that these environmental factors are often poorly measured. A key environmental factor is geographic location, which can also proxy other environmental exposures (e.g. SES, air/noise pollution, access to education, etc.) that may be relevant. Therefore, in the future, we plan to extend the scope of our research to leverage the geographical (spatial) aspect and examine its interplay with genetic predisposition (and genetic risk scores) to diseases. Such geographic-genetic interactions impact the proposed Mendelian randomisation analyses, as they can give rise to environment-specific causal effects. Also, this will require more accurate geographical information on the current address of the participants.

 

A further type of environment we would like to explore is parental rearing. Recent studies showed that part of the genetic effects impacting educational attainment and obesity are indirect and may represent parental/dynastic effects. Mendelian randomisation is particularly sensitive to such assumption violations and hence dissecting direct and parental effects has two-fold benefit for our originally proposed aims: (1) working with direct effects provides less biased MR estimates between obesity and cardio-vascular traits [original goal]; (2) elucidating indirect effects allows for the assessment of family environment, which is an important determinant and interaction partner for complex exposures [original goal of the project].

Furthermore, we aim to explore non-linear effects of obesogenic exposures, as we realised that classical MR estimates assuming linear effect only can give misleading estimates. Finally, we would extend the probed environment to molecular traits (gene expression, methylation, metabolomics, proteomics), which can be proxied by their polygenic risk scores and are/will be available in the UK Biobank. These omics layers capture key  environmental exposures, which may not have been directly measured.