Article Text

Download PDFPDF

  1. A. Joshi1,
  2. N. McCormick2,
  3. C. Yokose2,
  4. N. Lu3,
  5. H. Choi2
  1. 1Massachusetts General Hospital, CTEU, Mongan Institute, Department of Medicine, Boston, United States of America
  2. 2Massachusetts General Hospital, Division of Rheumatology, Allergy and Immunology, Boston, United States of America
  3. 3Arthritis Research Canada, Arthritis Research, Vancouver, Canada


Background Serum urate (SU) level is the strongest known causal predictor of clinical gout, but only ~20% with prolonged hyperuricemia develop gout, motivating the need for additional biomarkers for risk prediction and stratification. The metabolome represents a compelling intermediate trait between genome and phenome to elucidate disease mechanisms. Multiple cross-sectional studies of prevalent gout from men in Asia have been conducted, but no prospective data for incident gout (prediagnostic metabolome) are available.

Objectives Our objectives were to (1) conduct a discovery-based metabolome-wide study to identify novel biomarkers of incident gout; and (2) replicate novel metabolomic biomarkers of gout in independent samples.

Methods We conducted a prospective cohort analysis of 105,703 UK Biobank (UKB) participants (46% males, mean age 57.2 years) with targeted NMR metabolomic profiling (N=168 metabolites, including routine lipids and amino acids) available from baseline samples (2006-10), and no prior diagnosis of gout or urate lowering therapy use. Incident cases of gout were documented from linked medical records until gout diagnosis, death, or end of study period (Dec 31/19). We used Cox proportional hazard models to obtain hazard ratios (HR) and 95% confidence intervals (CIs) per standard deviation (SD) increase in each of the 168 metabolites to determine associations with incident gout.

To replicate our findings, we assessed association of metabolome-wide significant metabolites in a replication set, restricted to 4,804 non-overlapping participants who provided blood in the repeat assessment visit (2012-13).

Results During a median 10.4 years follow-up, we documented 1,367 cases of incident gout in the discovery set. After correction for multiple comparisons, glycoprotein acetyls (GlycA) were positively associated with risk of incident gout (multivariable HR per 1SD increase = 1.34 (1.27 to 1.41), P = 9.04x 10-28) after adjusting for age, sex, and lifestyle and clinical covariates (Table 1). This association persisted even after SU adjustment (HR 1.07, P = 0.0091). In the replication set, among 4,804 participants followed for a median of 6.8 years, we documented 22 cases. In this dataset, we replicated GlycA association with incident gout (multivariable HR per 1SD increase =1.56 (1.08 to 2.25), P = 0.017).

Table 1.

Association of glycoprotein acetyls (GlycA) with risk of incident gout in the UK Biobank

Conclusion In this large-scale, prospective metabolomics study, we identified and independently replicated our findings that plasma levels of GlycA are associated with incident gout in UKB participants. GlycA is novel for gout, though this pro-inflammatory biomarker has predicted risk of other cardiometabolic-inflammatory phenotypes, independent of CRP.1 These findings may provide insight into the metabolic-inflammatory pathogenesis of gout, with implications for risk prediction, even beyond SU, but call for further investigation with more extensive metabolome profiling and external replication.

References [1]Kettunen; PMID 30571186

Disclosure of Interests Amit Joshi: None declared, Natalie McCormick: None declared, Chio Yokose: None declared, Na Lu: None declared, Hyon Choi Consultant of: Ironwood, Selecta, Horizon, Takeda, Kowa, Vaxart, Grant/research support from: Ironwood, Horizon

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.