Back pain is the primary cause of disability worldwide yet surprisingly little is known of the underlying pathobiology. We conducted a genome-wide association study (GWAS) meta-analysis of chronic back pain (CBP). Adults of European ancestry from 15 cohorts in the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium, and UK Biobank were studied. CBP cases were defined as reporting back pain present for ≥3–6 months; non-cases were included as comparisons (“controls”). Each cohort conducted genotyping followed by imputation. GWAS used logistic regression with additive genetic effects adjusting for age, sex, study-specific covariates, and population substructure. Suggestive (p<5×10–7) & genome-wide significant (p<5×10–8) variants were carried forward for replication in an independent sample of UK Biobank participants. Discovery sample n = 158,025 individuals, including 29,531 CBP cases.Purpose
Methods
Osteoarthritis (OA) is a prevalent, age-related joint disease, characterized by diverse progressive changes in articular cartilage and subchondral bone. Disease management is severely hampered by the absence of tools to classify patients based on underlying disease mechanisms. For that matter, increased BMI is a known risk factor for OA in the weight bearing knee joint, but also for hand OA.1The increased risk for OA is therefore thought to be influenced by systemic factors accompanying BMI. It was hypothesized that differences in metabolic state could be underlying OA phenotypes. In the current study we set out to explore the potential role of a large range of metabolites in blood as sensitive biomarker of OA. Plasma samples were taken from the Rotterdam Study, CHECK-, GARP/NORREF- and the LUMC-arthroplasty cohorts. OA was defined as having had arthroplasty for primary OA, stratified per location (any, hip or knee). In total 647 persons with Total Joint Arthroplasty (TJA) were included and 2125 persons were considered as controls (i.e. they had a Kellgrenn-Lawrence Score of <2 indicating no radiographic OA was present) in any of the studied joints. A total of 231 different metabolites were assessed by using the BrainShake NMR platform. Since parts of the metabolites were highly correlated, we used Principal Component Analyses (PCA) to reduce the data. 23 factors were identified, accounting for 91,4% of the variance in the data. Logistic regression models were applied to investigate the identified factors for their association to arthroplasty for primary OA, independent of age, sex, BMI and cholesterol-lowering medication (statins). The models showed two different factors robustly associated to arthroplasty as result of primary OA. A table represents the associations of these factors to arthroplasty adjusted for age, sex and BMI, as the information on statin-use was not known for all subjects. Analyses showed that additional correction for statins did not change the results. When stratifying the arthroplasty phenotypes for joint location, factor 11, characterized by e.g. linoleic acid, was found to be associated to arthroplasty in the hip (THA). Similarly, Factor 22, representing saturated fatty acids and degree of unsaturation, was consistently associated with arthroplasty, independent of the site. When analyzing the metabolites involved in the factors individually these associations were confirmed for most contributors of the factors, except the ratio of saturated fatty acids to total fatty acids. Our preliminary analyses showed that persons with arthroplasty for primary OA compared to controls have different values for factors composed for fatty acids. The identification of groups of fatty acid metabolites as being connected to OA phenotypes indicates an inflammation driven pathway which might give a better understanding of the mechanisms behind OA.
