Mendelian randomisation analysis provides no evidence for a relationship between adult height and testicular cancer risk.
Date
2017-09-01ICR Author
Author
Levy, M
Hall, D
Sud, A
Law, P
Litchfield, K
Dudakia, D
Haugen, TB
Karlsson, R
Reid, A
Huddart, RA
Grotmol, T
Wiklund, F
Houlston, RS
Turnbull, C
Type
Journal Article
Metadata
Show full item recordAbstract
Observational studies have suggested anthropometric traits, particularly increased height are associated with an elevated risk of testicular cancer (testicular germ cell tumour). However, there is an inconsistency between study findings, suggesting the possibility of the influence of confounding factors. To examine the association between anthropometric traits and testicular germ cell tumour using an unbiased approach, we performed a Mendelian randomisation study. We used genotype data from genome wide association studies of testicular germ cell tumour totalling 5518 cases and 19,055 controls. Externally weighted polygenic risk scores were created and used to evaluate associations with testicular germ cell tumour risk per one standard deviation (s.d) increase in genetically-defined adult height, adult BMI, adult waist hip ratio adjusted for BMI (WHRadjBMI), adult hip circumference adjusted for BMI (HIPadjBMI), adult waist circumference adjusted for BMI (WCadjBMI), birth weight (BW) and childhood obesity. Mendelian randomisation analysis did not demonstrate an association between any anthropometric trait and testicular germ cell tumour risk. In particular, despite good power, there was no global evidence for association between height and testicular germ cell tumour. However, three SNPs for adult height individually showed association with testicular germ cell tumour (rs4624820: OR = 1.47, 95% CI: 1.41-1.55, p = 2.7 × 10-57 ; rs12228415: OR = 1.17, 95% CI: 1.11-1.22, p = 3.1 × 10-10 ; rs7568069: OR = 1.13, 95% CI: 1.07-1.18, p = 1.1 × 10-6 ). This Mendelian randomisation analysis, based on the largest testicular germ cell tumour genome wide association dataset to date, does not support a causal etiological association between anthropometric traits and testicular germ cell tumour aetiology. Our findings are more compatible with confounding by shared environmental factors, possibly related to prenatal growth with exposure to these risk factors occurring in utero.
Subject
Humans
Neoplasms, Germ Cell and Embryonal
Testicular Neoplasms
Body Mass Index
Body Height
Waist-Hip Ratio
Models, Statistical
Risk Factors
Random Allocation
Genotype
Polymorphism, Single Nucleotide
Adult
Male
Research team
Cancer Genomics
Molecular & Population Genetics
Clinical Academic Radiotherapy (Huddart)
Language
eng
Date accepted
2017-05-04
License start date
2017-09
Citation
Andrology, 2017, 5 (5), pp. 914 - 922
Publisher
WILEY