Potential serum biomarkers from a metabolomics study of autism.
Potential serum biomarkers from a metabolomics study of autism.
复制标题
DOI：
10.1503/jpn.140009
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
Han Wang;S. Liang;Maoqing Wang;Jingquan Gao;Caihong Sun;Jia Wang;W. Xia;Shiying Wu;S. Sumner;Fengyu Zhang;Changhao Sun;Lijie Wu
中科院分区：
文献类型：
--
作者：
Han Wang;S. Liang;Maoqing Wang;Jingquan Gao;Caihong Sun;Jia Wang;W. Xia;Shiying Wu;S. Sumner;Fengyu Zhang;Changhao Sun;Lijie Wu
背景早期发现和诊断对于自闭症非常重要。目前对自闭症的诊断主要依赖于一些观察性问卷和访谈工具，这些工具可能涉及很大的变异性。我们对血清进行了代谢组学分析，以确定自闭症早期诊断和临床评估的潜在生物标志物。方法采用超高效液相色谱-四极杆飞行时间串联质谱（UPLC/Q-TOF MS/MS）检测中国汉族人群中孤独症患者和非孤独症患者的血清代谢变化。在额外的独立病例和对照队列中单独验证了生物标志物的潜在代谢物候选物。我们建立了一个多元逻辑回归模型来评估验证的生物标志物。结果我们在发现队列中纳入了73例患者和63例对照，在验证队列中纳入了100例病例和100例对照。发现阶段的血清代谢组学分析鉴定了17种代谢物，其中11种在独立队列中得到验证。基于11种经验证的代谢物建立的多元逻辑回归模型在两个队列中拟合良好。该模型一致表明，自闭症与2种特定的代谢物有关：1-磷酸鞘氨醇和二十二碳六烯酸。局限性虽然自闭症主要诊断为男孩，但由于难以招募足够的女性患者，我们无法按性别进行分析。其他限制包括需要在同一个体内进行重测评估和样本量相对较小。结论两种代谢产物具有作为孤独症临床诊断和评估的生物标志物的潜力。
BACKGROUND Early detection and diagnosis are very important for autism. Current diagnosis of autism relies mainly on some observational questionnaires and interview tools that may involve a great variability. We performed a metabolomics analysis of serum to identify potential biomarkers for the early diagnosis and clinical evaluation of autism. METHODS We analyzed a discovery cohort of patients with autism and participants without autism in the Chinese Han population using ultra-performance liquid chromatography quadrupole time-of-flight tandem mass spectrometry (UPLC/Q-TOF MS/MS) to detect metabolic changes in serum associated with autism. The potential metabolite candidates for biomarkers were individually validated in an additional independent cohort of cases and controls. We built a multiple logistic regression model to evaluate the validated biomarkers. RESULTS We included 73 patients and 63 controls in the discovery cohort and 100 cases and 100 controls in the validation cohort. Metabolomic analysis of serum in the discovery stage identified 17 metabolites, 11 of which were validated in an independent cohort. A multiple logistic regression model built on the 11 validated metabolites fit well in both cohorts. The model consistently showed that autism was associated with 2 particular metabolites: sphingosine 1-phosphate and docosahexaenoic acid. LIMITATIONS While autism is diagnosed predominantly in boys, we were unable to perform the analysis by sex owing to difficulty recruiting enough female patients. Other limitations include the need to perform test-retest assessment within the same individual and the relatively small sample size. CONCLUSION Two metabolites have potential as biomarkers for the clinical diagnosis and evaluation of autism.