Liebers David T, Pirooznia Mehdi, Ganna Andrea, Goes Fernando S
Harvard Medical School, Boston, MA02115, USA.
Department of Psychiatry and Behavioral Sciences, Johns Hopkins Institute of Medicine, Baltimore, MD21205, USA.
Psychol Med. 2021 Jul;51(9):1451-1458. doi: 10.1017/S003329172000015X. Epub 2020 Feb 17.
Although accurate differentiation between bipolar disorder (BD) and unipolar major depressive disorder (MDD) has important prognostic and therapeutic implications, the distinction is often challenging based on clinical grounds alone. In this study, we tested whether psychiatric polygenic risk scores (PRSs) improve clinically based classification models of BD v. MDD diagnosis.
Our sample included 843 BD and 930 MDD subjects similarly genotyped and phenotyped using the same standardized interview. We performed multivariate modeling and receiver operating characteristic analysis, testing the incremental effect of PRSs on a baseline model with clinical symptoms and features known to associate with BD compared with MDD status.
We found a strong association between a BD diagnosis and PRSs drawn from BD (R2 = 3.5%, p = 4.94 × 10-12) and schizophrenia (R2 = 3.2%, p = 5.71 × 10-11) genome-wide association meta-analyses. Individuals with top decile BD PRS had a significantly increased risk for BD v. MDD compared with those in the lowest decile (odds ratio 3.39, confidence interval 2.19-5.25). PRSs discriminated BD v. MDD to a degree comparable with many individual symptoms and clinical features previously shown to associate with BD. When compared with the full composite model with all symptoms and clinical features PRSs provided modestly improved discriminatory ability (ΔC = 0.011, p = 6.48 × 10-4).
Our study demonstrates that psychiatric PRSs provide modest independent discrimination between BD and MDD cases, suggesting that PRSs could ultimately have utility in subjects at the extremes of the distribution and/or subjects for whom clinical symptoms are poorly measured or yet to manifest.
尽管双相情感障碍（BD）和单相重度抑郁症（MDD）之间的准确区分具有重要的预后和治疗意义，但仅基于临床依据进行区分往往具有挑战性。在本研究中，我们测试了精神科多基因风险评分（PRSs）是否能改善基于临床的BD与MDD诊断分类模型。
我们的样本包括843名BD患者和930名MDD患者，他们使用相同的标准化访谈进行了相似的基因分型和表型分析。我们进行了多变量建模和受试者工作特征分析，测试了PRSs对一个基线模型的增量效应，该基线模型包含已知与BD相关的临床症状和特征，并与MDD状态进行比较。
我们发现BD诊断与从BD全基因组关联荟萃分析（R2 = 3.5%，p = 4.94×10-12）和精神分裂症（R2 = 3.2%，p = 5.71×10-11）中得出的PRSs之间存在强烈关联。与最低十分位数的个体相比，处于最高十分位数BD PRS的个体患BD而非MDD的风险显著增加（优势比3.39，置信区间2.19 - 5.25）。PRSs区分BD与MDD的程度与许多先前显示与BD相关的个体症状和临床特征相当。与包含所有症状和临床特征的完整综合模型相比，PRSs提供了适度改善的区分能力（ΔC = 0.011，p = 6.48×10-4）。
我们的研究表明，精神科PRSs在BD和MDD病例之间提供了适度的独立区分，这表明PRSs最终可能对分布极端的个体和/或临床症状测量不佳或尚未显现的个体有用。
