文献速递糖尿病

性别校准多基因风险评分用于儿童早期1型糖尿病预筛

Polygenic Risk Scores for Islet Autoantibody Positive Early‐Stage Type 1 Diabetes Pre‐Screening
RSS 订阅 · Diabetes, Obesity and Metabolism · 2026-10-09

结果显示,高PRS儿童中男孩早期T1D的估计风险高于女孩,性别校准PRS可减少但未消除该性别差异,AUC约90%。

要点

  • 儿童早期T1D与3期T1D的PRS高度可比(GPPAD GRS中位数均为12.8),提示遗传风险在抗体阳性阶段已稳定。
  • 在99.2百分位阈值下,男孩早期T1D估计风险为12.6%,女孩为9.3%,存在性别偏倚。
  • 性别校准PRS(包括150-SNP全版与38-SNP简版)区分病例与对照的AUC约90%。

核心结论

  • PRS在早期T1D与3期T1D儿童中高度可比,但提示高PRS儿童中早期T1D风险存在性别偏倚(男孩风险更高)。
  • 性别校准PRS预计可减少但无法消除预测风险中的性别差异。
  • 该PRS主要在欧洲血统的儿童人群中开发,需要独立外部验证(包括其他祖源人群)。

研究设计与样本

  • 病例包括240名早期T1D儿童、1094名胰岛自身抗体阳性3期T1D儿童及695名临床T1D患者的验证队列。
  • 基于14个HLA DR-DQ单体型及143个T1D易感区域共150个SNP,开发男孩和女孩各自的性别校准PRS。
  • 评估既往开发的PRS,并在相同个体中比较男性和女孩的预测能力。

主要发现

  • 早期T1D与3期T1D儿童的PRS相似:GPPAD GRS中位数均为12.8(p=0.81),GRS2中位数分别为13.8和13.9(p=0.1)。
  • 在背景人群99.2百分位对应阈值下,GPPAD GRS估计的男孩早期T1D风险为10.7%,女孩为6.8%。
  • 在相同阈值下,GRS2的男孩估计风险为16.4%,女孩为8.6%。
  • 性别校准的早期T1D PRS(全版150-SNP和简版38-SNP)判别病例与对照的AUC在男孩和女孩中均约90%。
  • 使用性别校准PRS后,99.2百分位阈值下男孩和女孩的估计风险趋于收敛,但差异未消除(男孩12.6%,女孩9.3%)。

关键数值与推荐

  • 性别校准PRS从150个SNP中定义14个HLA DR-DQ单体型和143个T1D遗传易感区域。
  • 简版性别校准PRS包含38个SNP。

临床怎么用

  • 在无一级亲属家族史的普通儿童人群中,可使用性别校准PRS进行早期T1D预筛,识别高风险个体。
  • 男女性别应分开解读风险,尤其关注高PRS男孩群体的更高早期T1D风险。

局限性

  • PRS主要在儿科人群且以欧洲血统为主的人群中开发,需要独立外部验证(包括其他祖源群体)。
  • 性别校准PRS未能完全消除高PRS人群中的性别差异。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

ABSTRACT Aims To develop a sex-aligned polygenic risk score (PRS) for early-stage type 1 diabetes (T1D) in children without a first-degree family history of T1D. Materials and Methods Global Single Nucleotide Polymorphism (SNP) Array was performed in 16 335 individuals without a first-degree family history of T1D, including 13 956 newborns from the background population of Bavaria, Germany, 240 children with early-stage and 1094 with islet autoantibody-positive Stage 3 T1D from Bavaria as cases and a validation cohort of 695 patients with clinical T1D. Previously developed PRS were assessed and sex-aligned PRS for boys and for girls were developed from 150 SNPs defining 14 HLA DR-DQ haplotypes and 143 T1D genetic susceptibility regions. Results PRS were similar between children with early-stage T1D and stage 3 T1D for the GPPAD GRS (median 12.8 vs. 12.8, p = 0.81) and the GRS2 (median 13.8 vs. 13.9, p = 0.1). At thresholds corresponding to the 99.2 centile of the background population, the estimated risk for early-stage T1D was higher in boys than girls using both the GPPAD GRS (10.7% in boys vs. 6.8% in girls) and GRS2 (16.4% vs. 8.6%). Novel full (150 SNP) and short (38 SNP) sex-aligned early-stage T1D PRS discriminated cases from controls with around 90% area under the receiver operator characteristic curve in boys and girls. Estimated risks for boys and girls at the 99.2 centile of the background population using the sex-aligned PRS converged, but differences were not eliminated (12.6% for boys, 9.3% for girls). Conclusions PRS in children with early-stage and Stage 3 T1D are highly comparable, but suggest a sex bias in early-stage Type 1 diabetes risk in children with high PRS. A sex-aligned PRS was predicted to reduce, but not eliminate, sex-related differences in predicted risk. The PRS was developed primarily in a paediatric population of predominantly European ancestry and requires independent external validation, including in other ancestry groups.

来源:Diabetes, Obesity and Metabolism,查看原文。版权归原作者所有。

引用

  1. Diabetes, Obesity and Metabolism:Polygenic Risk Scores for Islet Autoantibody Positive Early‐Stage Type 1 Diabetes Pre‐Screening

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。