Height and cancer risk: energy restriction and genetic variation

Matty Weijenberg's research aims to shed light on the mechanisms behind adult-attained height and cancer risk, specifically the role of early life exposure to energy restriction and genetic variation

  • Topic: Colorectal cancer
  • Institution: Maastricht University
  • Country: Netherlands
  • Status: Completed

Scientific abstract

(View plain language abstract)

Background

Height increases the risk of colorectal (CRC), breast, and ovarian cancer. Tallness in itself is unlikely to cause cancer, but is rather a proxy of related underlying processes or exposures. There is a broad understanding of how genetic and environmental processes contribute to adult-attained height, and it is interesting to speculate that similar processes are related to cancer risks. As the variation in adult height is determined by genetic and environmental interactions during the first 20 years of life, similar early life gene-environment interactions might be related to cancer risk, thus explaining the observed non-causal links between height and cancer. In 2010, the Genetic Investigation of Anthropometric Traits (GIANT) Consortium published a large meta-analysis of genome-wide height association results, and identified gene loci that cluster together in biologically relevant pathways, especially the Insulin Growth Factor (IGF), Hedgehog, Transforming Growth Factor (TGF)-beta, and Growth Hormone signalling pathways. The Netherlands Cohort Study (NLCS) has collected data on childhood and adolescent exposure to energy (caloric) restriction, providing a unique opportunity to consider gene-environment interactions that may shed further insight on the mechanistic pathways that link height and cancer.

Hypothesis and objectives

Energy restriction early in life influences cancer through mechanisms that also determine adult-attained height, which is an established risk factor for colorectal (CRC), post-menopausal breast, and ovarian cancer, although not causal. The association between energy restriction early in life and cancer is modified and/or mediated by height and is modified by genetic variants relevant for growth identified through genome-wide association studies (GWAS) of height. With respect to cancer risk, we aim to assess:

  1. The interaction and mediation effects of early life exposure to energy restriction and height.
  2. The interaction between early life exposure to energy restriction and genetic variants and between height and these genetic variants.

Setting and methods

The NLCS was initiated in 1986 among 120,852 men and women aged 55–69 years who completed a self-administered questionnaire on dietary habits and other risk factors for cancer, including self-reported height and various indicators of exposure to energy restriction during childhood and adolescence. At baseline, a subcohort of 5,000 individuals was randomly selected and has been followed up to estimate person time at risk in the entire cohort. Toenail clippings were collected at baseline and are available for DNA isolation and genotyping. It was previously shown that toenails contain DNA of sufficient quality and quantity for analysis of single nucleotide polymorphisms (SNPs). After 20.3 years of follow-up, we expect 3,106 incident CRC, 2,594 breast, and 405 ovarian cancer cases with DNA material. Criteria for gene variant selection are their relevance for growth pathways also implicated in cancer development. We will use graphical modelling and mediation analysis to investigate mediation. Interactions will be explored in a complementary approach using random forest and multifactor dimensionality reduction (MDR) techniques followed by Cox proportional hazards survival analysis adapted to the case-cohort design.

Potential impact

This study will contribute to the limited knowledge base surrounding the association between height and cancer by providing valuable insights about potential mechanisms involved in cancer aetiology. Although we cannot control our adult-attained height, understanding how height is related to cancer is important for public health strategies and for expanding our knowledge about the pathways to cancer development. Average heights are continuing to increase in most countries, particularly in those undergoing rapid socio-economic improvement. Therefore, understanding this association may allow some anticipation of future health needs. Furthermore, as events in childhood are thought to influence adult height, with over-eating and energy restriction at the far ends of the continuum, it will be important for future studies to explore the predictive capabilities of these exposures. As the NLCS is one of the few studies in which data on early life nutritional exposure has been collected, our study paves the way for such research.

In addition, findings from this study could be of use for identifying susceptible individuals both in terms of genetic variants in relevant pathways, as in terms of human height or extreme exposures in youth at both ends of the continuum (over-eating and energy restriction). Understanding the underlying pathways will provide opportunities for intervening in these pathways in the future, and targeting high-risk individuals with stratified prevention strategies.

Plain language abstract

Background

Taller people have a greater chance of developing cancer than shorter people; however, the reason for this is unclear. As it has been observed that height increases the risk of a number of different types of cancers, a common mechanism might be at play. Likely, being tall is a marker for the "true" mechanism, rather than a causal factor. For example, taller people have more cells in their body and therefore may have a greater chance of developing cancerous cell changes. It has also been suggested that early life nutrition and/or hormone levels related to growth could be responsible. Genome-wide association studies (GWAS) are used to identify common genetic factors that influence health and disease. A recent meta-analysis of GWAS on height identified gene loci that cluster together in biologically relevant pathways. Intriguingly, several of these pathways have also been implicated in cancer development. As the variation in adult height is determined by genetic and environmental interactions during the first 20 years of life, similar early life gene-environment interactions might be related to cancer risk, thus explaining the observed non-causal link between height and cancer.

Aims and goals

We aim to investigate how energy restriction early in life influences colorectal, post-menopausal breast, and ovarian cancer through the pathways that also determine adult-attained height. This may lend support to the hypothesis that early life exposures play an important role in the association between height and cancer risk. This will shed light on potential mechanisms linking height and cancer, for which knowledge is currently lacking.

How it will be done

The unique Netherlands Cohort Study (NLCS) began in 1986 among 120,852 men and women aged 55–69 years who completed a self-administered questionnaire on dietary habits and other risk factors for cancer, including self-reported height and various indicators of exposure to energy restriction during childhood and adolescence. Toenail clippings were collected at baseline and are available for DNA isolation and genotyping. After 20.3 years of follow-up, we expect 3,106 incident CRC, 2,594 breast, and 405 ovarian cancer cases with DNA material. We will use this data to assess whether energy restriction acts in conjunction with or through height when influencing cancer risk. Similarly, we will investigate whether genetic variation relevant to growth affects the associations between energy restriction and height on the risk of cancer.

Potential impact

Although we cannot control our adult-attained height, understanding how height is related to cancer is important for public health strategies and for expanding our knowledge about the pathways to cancer development. Average heights are continuing to increase in most countries, particularly in those undergoing rapid socio-economic improvement. Therefore, understanding this association may allow some anticipation of future health needs. Furthermore, as events in childhood are thought to influence adult height, it will be important for future studies to explore the predictive.

Grant publications