To examine associations between serum oxylipins, which regulate tissue repair and pain signalling, and knee pain/radiographic osteoarthritis (OA) at baseline and knee pain at 3 year follow-up. Baseline, and 3 year follow-up, knee pain phenotypes were assessed from 154 participants in the Knee Pain in the Community (KPIC) cohort study. Serum and radiographic Kellgren and Lawrence (KL) and Nottingham line drawing atlas OA scores were collected at baseline. Oxylipin levels were quantified using liquid chromatography coupled with mass spectrometry. Associations were measured by linear regression and receiver operating characteristics (ROC). Serum levels of 8,9-epoxyeicosatrienoic acid (EET) (β(95% confidence intervals (CI)) = 1.809 (−0.71 to 2.91)), 14,15-dihydroxyeicosatrienoic acid (DHET) (β(95%CI) = 0.827 (0.34–1.31)), and 12-hydroxyeicosatetraenoic acid (HETE) (β(95%CI) = 4.090 (1.92–6.26)) and anandamide (β(95%CI) = 3.060 (1.35–4.77)) were cross-sectionally associated with current self-reported knee pain scores (numerical rating scale (NRS) item 3, average pain). Serum levels of 9- (β(95%CI) = 0.467 (0.18–0.75)) and 15-HETE (β(95%CI) = 0.759 (0.29–1.22)), 14-hydroxydocosahexaenoic acid (β(95%CI) = 0.483(0.24–0.73)), and the ratio of 8,9-EET:DHET (β(95%CI) = 0.510(0.19–0.82)) were cross-sectionally associated with KL scores. Baseline serum concentrations of 8,9-EET (β(95%CI) = 2.166 (0.89–3.44)), 5,6-DHET (β(95%CI) = 152.179 (69.39–234.97)), and 5-HETE (β(95%CI) = 1.724 (0.677–2.77) showed positive longitudinal associations with follow-up knee pain scores (NRS item 3, average pain). Combined serum 8,9-EET and 5-HETE concentration showed the strongest longitudinal association (β(95%CI) = 1.156 (0.54–1.77) with pain scores at 3 years, and ROC curves distinguished between participants with no pain and high pain scores at follow-up (area under curve (95%CI) = 0.71 (0.61–0.82)). Serum levels of a combination of hydroxylated metabolites of arachidonic acid may have prognostic utility for knee pain, providing a potential novel approach to identify people who are more likely to have debilitating pain in the future.

中文翻译：

---

花生四烯酸羟基化代谢物的血清水平横向和纵向预测膝关节疼痛进展：一项观察性队列研究

旨在检查血清氧脂素（调节组织修复和疼痛信号传导）与基线时的膝关节疼痛/放射学骨关节炎 (OA) 以及 3 年随访时的膝关节疼痛之间的关联。对社区膝痛 (KPIC) 队列研究的 154 名参与者进行了基线和 3 年随访评估。在基线时收集血清和射线照相 Kellgren 和 Lawrence (KL) 以及诺丁汉线图谱 OA 评分。使用液相色谱结合质谱法对氧脂质水平进行定量。通过线性回归和受试者工作特征（ROC）来测量关联性。 8,9-环氧二十碳三烯酸 (EET) 的血清水平 (β(95% 置信区间 (CI)) = 1.809 (−0.71 至 2.91))、14,15-二羟基二十碳三烯酸 (DHET) (β(95%CI) = 0.827 (0.34–1.31)) 和 12-羟基二十碳四烯酸 (HETE) (β(95%CI) = 4.090 (1.92–6.26)) 和 anandamide (β(95%CI) = 3.060 (1.35–4.77)) 交叉-与当前自我报告的膝盖疼痛评分部分相关（数字评定量表（NRS）第3项，平均疼痛）。 9- (β(95%CI) = 0.467 (0.18–0.75)) 和 15-HETE (β(95%CI) = 0.759 (0.29–1.22))、14-羟基二十二碳六烯酸 (β(95%CI) 的血清水平) = 0.483(0.24–0.73))，8,9-EET:DHET 的比率 (β(95%CI) = 0.510(0.19–0.82)) 与 KL 分数具有横断面相关性。 8,9-EET (β(95%CI) = 2.166 (0.89–3.44))、5,6-DHET (β(95%CI) = 152.179 (69.39–234.97)) 和 5-HETE 的基线血清浓度（β(95%CI) = 1.724 (0.677–2.77) 显示与随访膝关节疼痛评分（NRS 第 3 项，平均疼痛）呈正纵向相关。血清 8,9-EET 和 5-HETE 综合浓度显示最强的纵向相关性3 年时疼痛评分与相关性 (β(95%CI) = 1.156 (0.54–1.77)，ROC 曲线区分随访时无疼痛和高疼痛评分的参与者（曲线下面积 (95%CI) = 0.71 (0.61–0.82)) 花生四烯酸羟基化代谢物组合的血清水平可能对膝盖疼痛具有预测作用，为识别未来更有可能患有衰弱性疼痛的人提供了一种潜在的新方法。