Background/Purpose:  Gout has been reported to be a risk factor for development of diffuse idiopathic skeletal hyperostosis (DISH), a condition characterized by abnormal bone formation of the soft tissues affecting the spine and also peripheral skeleton.  However, it is unknown whether gout has a direct effect on development of DISH or whether this association occurs due to shared risk factors, such as raised body mass index or other features of the metabolic syndrome.  The aim of this study was to determine whether gout has a direct effect on the development of DISH.

Methods:   Participants in the Genetics of Gout in Aotearoa Study with lateral chest or thoracic spine radiographs were included in this analysis.  All participants in this analysis were 30 years or older at the time of the imaging test.  Each radiograph was scored for the presence of DISH by a trained rheumatologist or musculoskeletal radiologist.  For the purposes of this analysis, DISH was defined as definite and probable DISH, according to the criteria of Ustinger et al, 1985.  Participants with imaging features of spondyloarthropathy; spondylosis with osteophytes bridging 2 consecutive vertebrae; or features of DISH and another pathology (primarily disc space narrowing) were not included in the analysis.  Genotypes were available for the two major gout-associated single nucleotide polymorphisms (SNPs); ABCG2 rs2231142 and SLC2A9 rs734553. Genome-wide genotype data generated on the Illumina CoreExome platform were used to estimate the heritability of DISH using the Genome-wide Complex Trait Analysis (GCTA) tool with prevalence estimates of 41% in Māori and Pacific people and 22% in European (Bateman et al 2018).  Clinical and genetic associations with DISH were analysed by logistic regression adjusted by age at recruitment, age at the time of the imaging test, sex and, as appropriate, by metabolic co-morbidity.  

Results: The analysis included 816 participants, including 535 (65.6%) with gout. DISH was present in 205/816 (25.1%) of all participants, including 138/453 (30.5%) of those of Polynesian ancestry (Māori or Pacific people) and 67/363 (18.5%) participants of European ancestry. DISH was present in 155/535 (29.0%) participants with gout, and 50/281 (17.8%) participants without gout; unadjusted odds ratio (95% CI) 1.85 (1.29-2.65), P=7.6 x 10^-4.  However, after adjusting for body mass index, DISH was not associated with gout; adjusted odds ratio (95% CI) 1.28 (0.86-1.91), P=0.22.  Consistent with this, no association was observed between DISH and the gout-associated ABCG2 and SLC2A9 SNPs (P>0.52 for all comparisons).  The heritability (percent variance in phenotype explained by the genome-wide genotype data) was estimated to be 0.37 (se=0.56) in people of Polynesian ancestry and 0.24 (se=2.1) in people of European ancestry.  

Conclusion:   DISH is more common in people with gout.  Although DISH does demonstrate a heritable component, genetic analysis does not support the hypothesis that gout is causal for DISH.  The relationship between gout and DISH is likely due to shared risk factors, particularly raised body mass index and other features of the metabolic syndrome.