Polyunsaturated fatty acids and risk of melanoma: A Mendelian randomisation analysis

Melanoma Meta-Analysis Consortium

Research output: Contribution to journalArticle

5 Citations (Scopus)
251 Downloads (Pure)

Abstract

Melanoma is the deadliest form of skin cancer, mainly affecting populations of European ancestry. Some observational studies suggest that particular diets reduce melanoma risk, putatively through an increase in polyunsaturated fatty acid (PUFA) consumption. However, interpretation of these observational findings is difficult due to residual confounding or reverse causality. To date, a randomized controlled trial has not been carried out to examine the relationship between PUFAs and melanoma. Hence, we performed a Mendelian randomisation (MR) study to evaluate the link between PUFAs and melanoma. To perform MR, we used summary results from the largest risk genome-wide association study (GWAS) meta-analysis of melanoma, consisting of 12,874 cases and 23,203 controls. As instrumental variables we selected SNPs associated with PUFA levels from a GWAS meta-analysis of PUFA levels, from the CHARGE consortium. We used the inverse variance weighted method to estimate a causal odds ratio. To aid interpretation, we established a benchmark “large” predicted change in PUFAs in which, for example, an increase in docosahexaenoic acid (DPA) of 0.17 units (equal to 1 standard deviation) moves a person from the 17th percentile to the median. Raising PUFA levels by a large amount (increasing DPA by 0.17 units) only negligibly changed melanoma risk: odds ratio [OR] = 1.03 (95% confidence interval [CI] = 0.96–1.10). Other PUFAs yielded similar results as DPA. Our MR analysis suggests that the effect of PUFA levels on melanoma risk is either zero or very small.

Original languageEnglish
Pages (from-to)508-514
Number of pages7
JournalInternational Journal of Cancer
Volume143
Issue number3
DOIs
Publication statusPublished - 1 Aug 2018

Fingerprint

Mendelian Randomization Analysis
Unsaturated Fatty Acids
Melanoma
Genome-Wide Association Study
Odds Ratio
Random Allocation
Meta-Analysis
Benchmarking
Docosahexaenoic Acids
Skin Neoplasms
Causality
Observational Studies
Single Nucleotide Polymorphism
Randomized Controlled Trials
Confidence Intervals
Diet

Cite this

Melanoma Meta-Analysis Consortium. / Polyunsaturated fatty acids and risk of melanoma : A Mendelian randomisation analysis. In: International Journal of Cancer. 2018 ; Vol. 143, No. 3. pp. 508-514.
@article{cb0e500612004df9a50cf9401379ef49,
title = "Polyunsaturated fatty acids and risk of melanoma: A Mendelian randomisation analysis",
abstract = "Melanoma is the deadliest form of skin cancer, mainly affecting populations of European ancestry. Some observational studies suggest that particular diets reduce melanoma risk, putatively through an increase in polyunsaturated fatty acid (PUFA) consumption. However, interpretation of these observational findings is difficult due to residual confounding or reverse causality. To date, a randomized controlled trial has not been carried out to examine the relationship between PUFAs and melanoma. Hence, we performed a Mendelian randomisation (MR) study to evaluate the link between PUFAs and melanoma. To perform MR, we used summary results from the largest risk genome-wide association study (GWAS) meta-analysis of melanoma, consisting of 12,874 cases and 23,203 controls. As instrumental variables we selected SNPs associated with PUFA levels from a GWAS meta-analysis of PUFA levels, from the CHARGE consortium. We used the inverse variance weighted method to estimate a causal odds ratio. To aid interpretation, we established a benchmark “large” predicted change in PUFAs in which, for example, an increase in docosahexaenoic acid (DPA) of 0.17 units (equal to 1 standard deviation) moves a person from the 17th percentile to the median. Raising PUFA levels by a large amount (increasing DPA by 0.17 units) only negligibly changed melanoma risk: odds ratio [OR] = 1.03 (95{\%} confidence interval [CI] = 0.96–1.10). Other PUFAs yielded similar results as DPA. Our MR analysis suggests that the effect of PUFA levels on melanoma risk is either zero or very small.",
keywords = "melanoma, Mendelian randomisation, n-3 fatty acids, n-6 fatty acids, polyunsaturated fatty acids",
author = "{Melanoma Meta-Analysis Consortium} and Liyanage, {Upekha E.} and Law, {Matthew H.} and Ong, {Jue Sheng} and Cust, {Anne E.} and Mann, {Graham J.} and Ward, {Sarah V.} and Puya Gharahkhani and Iles, {Mark M.} and Stuart MacGregor",
year = "2018",
month = "8",
day = "1",
doi = "10.1002/ijc.31334",
language = "English",
volume = "143",
pages = "508--514",
journal = "International Journal of Cancer (Predictive Oncology)",
issn = "0020-7136",
publisher = "Blackwell",
number = "3",

}

Polyunsaturated fatty acids and risk of melanoma : A Mendelian randomisation analysis. / Melanoma Meta-Analysis Consortium.

In: International Journal of Cancer, Vol. 143, No. 3, 01.08.2018, p. 508-514.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Polyunsaturated fatty acids and risk of melanoma

T2 - A Mendelian randomisation analysis

AU - Melanoma Meta-Analysis Consortium

AU - Liyanage, Upekha E.

AU - Law, Matthew H.

AU - Ong, Jue Sheng

AU - Cust, Anne E.

AU - Mann, Graham J.

AU - Ward, Sarah V.

AU - Gharahkhani, Puya

AU - Iles, Mark M.

AU - MacGregor, Stuart

PY - 2018/8/1

Y1 - 2018/8/1

N2 - Melanoma is the deadliest form of skin cancer, mainly affecting populations of European ancestry. Some observational studies suggest that particular diets reduce melanoma risk, putatively through an increase in polyunsaturated fatty acid (PUFA) consumption. However, interpretation of these observational findings is difficult due to residual confounding or reverse causality. To date, a randomized controlled trial has not been carried out to examine the relationship between PUFAs and melanoma. Hence, we performed a Mendelian randomisation (MR) study to evaluate the link between PUFAs and melanoma. To perform MR, we used summary results from the largest risk genome-wide association study (GWAS) meta-analysis of melanoma, consisting of 12,874 cases and 23,203 controls. As instrumental variables we selected SNPs associated with PUFA levels from a GWAS meta-analysis of PUFA levels, from the CHARGE consortium. We used the inverse variance weighted method to estimate a causal odds ratio. To aid interpretation, we established a benchmark “large” predicted change in PUFAs in which, for example, an increase in docosahexaenoic acid (DPA) of 0.17 units (equal to 1 standard deviation) moves a person from the 17th percentile to the median. Raising PUFA levels by a large amount (increasing DPA by 0.17 units) only negligibly changed melanoma risk: odds ratio [OR] = 1.03 (95% confidence interval [CI] = 0.96–1.10). Other PUFAs yielded similar results as DPA. Our MR analysis suggests that the effect of PUFA levels on melanoma risk is either zero or very small.

AB - Melanoma is the deadliest form of skin cancer, mainly affecting populations of European ancestry. Some observational studies suggest that particular diets reduce melanoma risk, putatively through an increase in polyunsaturated fatty acid (PUFA) consumption. However, interpretation of these observational findings is difficult due to residual confounding or reverse causality. To date, a randomized controlled trial has not been carried out to examine the relationship between PUFAs and melanoma. Hence, we performed a Mendelian randomisation (MR) study to evaluate the link between PUFAs and melanoma. To perform MR, we used summary results from the largest risk genome-wide association study (GWAS) meta-analysis of melanoma, consisting of 12,874 cases and 23,203 controls. As instrumental variables we selected SNPs associated with PUFA levels from a GWAS meta-analysis of PUFA levels, from the CHARGE consortium. We used the inverse variance weighted method to estimate a causal odds ratio. To aid interpretation, we established a benchmark “large” predicted change in PUFAs in which, for example, an increase in docosahexaenoic acid (DPA) of 0.17 units (equal to 1 standard deviation) moves a person from the 17th percentile to the median. Raising PUFA levels by a large amount (increasing DPA by 0.17 units) only negligibly changed melanoma risk: odds ratio [OR] = 1.03 (95% confidence interval [CI] = 0.96–1.10). Other PUFAs yielded similar results as DPA. Our MR analysis suggests that the effect of PUFA levels on melanoma risk is either zero or very small.

KW - melanoma

KW - Mendelian randomisation

KW - n-3 fatty acids

KW - n-6 fatty acids

KW - polyunsaturated fatty acids

UR - http://www.scopus.com/inward/record.url?scp=85043326053&partnerID=8YFLogxK

U2 - 10.1002/ijc.31334

DO - 10.1002/ijc.31334

M3 - Article

VL - 143

SP - 508

EP - 514

JO - International Journal of Cancer (Predictive Oncology)

JF - International Journal of Cancer (Predictive Oncology)

SN - 0020-7136

IS - 3

ER -