Identification of metabolites associated with prostate cancer risk: a nested case-control study with long follow-up in the Northern Sweden Health and Disease Study

• Published in: BMC Medicine
• Main Authors: Hanna E. Röhnisch, Cecilie Kyrø, Anja Olsen, Elin Thysell, Göran Hallmans, Ali A. Moazzami
• Format: Article in Journal/Newspaper
• Language: English
• Published: BMC 2020
• Subjects: Prostate cancer; Metabolomics; Nested case-control study; Nuclear magnetic resonance spectroscopy; Mass spectrometry; Risk biomarkers; Medicine; R; Northern Sweden
• Online Access: https://doi.org/10.1186/s12916-020-01655-1; https://doaj.org/article/574a273ca2bd4da8a0a4c9ccdf7b2a2b

Abstract Background Prostate cancer is the second most frequently diagnosed cancer in men. Metabolomics can potentially provide new insights into the aetiology of prostate cancer by identifying new metabolic risk factors. This study investigated the prospective association between plasma metabolite concentrations and prostate cancer risk, both overall and by stratifying for disease aggressiveness and baseline age. Methods In a case-control study nested in the Northern Sweden Health and Disease Study, pre-diagnostic concentrations of 148 plasma metabolites were determined using targeted mass spectrometry- and nuclear magnetic resonance-based metabolomics in 777 prostate cancer cases (follow-up ≥ 5 years) and 777 matched controls. Associations between prostate cancer risk and metabolite concentrations were investigated using conditional logistic regression conditioned on matching factors (body mass index, age and sample storage time). Corrections for multiple testing were performed using false discovery rate (20%) and Bonferroni. Metabolomics analyses generated new hypotheses, which were investigated by leveraging food frequency questionnaires (FFQs) and oral glucose tolerance tests performed at baseline. Results After correcting for multiple testing, two lysophosphatidylcholines (LPCs) were positively associated with risk of overall prostate cancer (all ages and in older subjects). The strongest association was for LPC C17:0 in older subjects (OR = 2.08; 95% CI 1.45–2.98; p < 0.0001, significant also after the Bonferroni correction). Observed associations with risk of overall prostate cancer in younger subjects were positive for glycine and inverse for pyruvate. For aggressive prostate cancer, there were positive associations with six glycerophospholipids (LPC C17:0, LPC C20:3, LPC C20:4, PC ae C38:3, PC ae C38:4 and PC ae C40:2), while there was an inverse association with acylcarnitine C18:2. Moreover, plasma LPC C17:0 concentrations positively correlated with estimated dietary intake of fatty acid C17:0 ...