您的位置：山东大学 -> 科技期刊社 -> 《山东大学学报（医学版）》

山东大学学报 (医学版) ›› 2026, Vol. 64 ›› Issue (9): 88-99.doi: 10.6040/j.issn.1671-7554.0.2025.1017

• 公共卫生与预防医学 • 上一篇    

中国儿童青少年抑郁症疾病负担趋势及危险因素分析:基于SHAP的可解释性机器学习

蒋秀玉1,郑凤家2,刘璇3,李素云4,吕翠霞4,刘昭璐4,郑守娟4,周晗5,于连龙3,6   

  1. 1.山东第一医科大学附属中心医院健康管理中心, 山东 济南 250013;2.山东省疾病预防控制中心卫生检验检测所, 山东 济南 250014;3.山东第二医科大学公共卫生学院, 山东 潍坊 261053;4.山东省疾病预防控制中心健康管理所, 山东 济南 250014;5.山东省疾病预防控制中心食品与营养所, 山东 济南 250014;6. 山东省疾病预防控制中心公共卫生监测评价所, 山东 济南 250014
  • 发布日期:2026-09-09
  • 通讯作者: 于连龙. E-mail:lianlong00a@163.com
  • 基金资助:
    山东省医药卫生科技发展计划项目(202512031110,202412021229,202412011216);齐鲁卫生人才杰青项目(全生命周期健康影响因素研究);齐鲁青耕人才项目

Disease burden trends and risk factors of depressive disorders among children and adolescents in China: an explainable machine learning analysis based on SHAP

JIANG Xiuyu1, ZHENG Fengjia2, LIU Xuan3, LI Suyun4, LYU Cuixia4, LIU Zhaolu4, ZHENG Shoujuan4, ZHOU Han5, YU Lianlong3,6   

  1. 1. Health Management Center, Central Hospital Affiliated to Shandong First Medical University, Jinan 250013, Shandong, China;
    2. Institute of Health Inspection and Testing, Shandong Provincial Center for Disease Control and Prevention, Jinan 250014, Shandong, China;
    3. School of Public Health, Shandong Second Medical University, Weifang 261053, Shandong, China;
    4. Institute of Health Management, Shandong Provincial Center for Disease Control and Prevention, Jinan 250014, Shandong, China;
    5. Institute of Food and Nutrition, Shandong Provincial Center for Disease Control and Prevention, Jinan 250014, Shandong, China;
    6. Institute of Public Health Monitoring and Evaluation, Shandong Provincial Center for Disease Control and Prevention, Jinan 250014, Shandong, China
  • Published:2026-09-09

PDF (PC)

15

摘要： 目的 分析1990—2021年中国20岁以下儿童青少年抑郁症疾病负担的长期趋势及年龄、性别差异,并探索相关暴露特征的模型贡献。 方法 基于全球疾病负担研究2021数据,采用估计年度变化百分比(estimated annual percentage change, EAPC)和Joinpoint回归分析发病率、患病率及伤残调整寿命年(disability-adjusted life years, DALYs)的长期趋势,采用贝叶斯年龄-时期-队列(Bayesian age-period-cohort, BAPC)模型预测2021—2049年发病率的变化。采用轻量级梯度提升机(light gradient boosting machine, LightGBM)结合沙普利加性解释(Shapley additive explanations, SHAP)及负二项回归,探索相关总暴露值(summary exposure value, SEV)特征及其关联强度。 结果 1990—2021年,中国儿童青少年抑郁症发病率、患病率和DALYs总体下降,而全球总体上升。我国疾病负担随年龄增加而升高,女性负担高于男性;9岁及以下儿童抑郁症负担呈上升趋势,而15~19岁组呈下降趋势。SHAP分析显示,儿童体质量不足对模型预测结果的贡献最高,铁缺乏、针对儿童的性暴力及职业性颗粒物与气体烟雾等特征亦具有较高贡献。负二项回归结果显示,针对儿童的性暴力关联强度最高(RR=2.212,95%CI:2.099~2.331)。 结论 中国儿童青少年抑郁症负担总体下降,但低龄儿童和女性群体仍需重点关注。这提示应加强早期筛查,并从暴力预防、营养改善和生活环境等方面开展综合干预。

关键词: 儿童青少年, 抑郁症, 疾病负担, 可解释性机器学习, 总暴露值

Abstract: Objective To analyze the long-term trends and age- and sex-specific differences in the burden of depressive disorders among children and adolescents aged <20 years in China from 1990 to 2021, and to explore the contributions of relevant exposure features to model predictions. Methods Data were obtained from the Global Burden of Disease Study 2021. The estimated annual percentage change(EAPC)and Joinpoint regression were used to analyze long-term trends in incidence, prevalence, and disability-adjusted life years(DALYs). A Bayesian age-period-cohort(BAPC)model was used to project changes in incidence from 2021 to 2049. Light gradient boosting machine(LightGBM), combined with Shapley additive explanations(SHAP)and negative binomial regression, was used to explore relevant summary exposure value(SEV)features and the strengths of their associations. Results From 1990 to 2021, the incidence, prevalence, and DALYs of depressive disorders among children and adolescents in China decreased overall, whereas the corresponding global indicators increased. In China, the disease burden increased with age and was higher among females than among males. The burden of depressive disorders increased among children aged ≤9 years but decreased among those aged 15-19 years. SHAP analysis showed that childhood underweight made the greatest contribution to model predictions, while iron deficiency, sexual violence against children, and occupational exposure to particulate matter, gases, and fumes also showed relatively high contributions. Negative binomial regression showed that sexual violence against children had the strongest association(RR=2.212, 95%CI: 2.099-2.331). Conclusion The burden of depressive disorders among children and adolescents in China has decreased overall, but younger children and females remain priority populations. These findings support strengthening early screening and implementing comprehensive interventions targeting violence prevention, nutritional improvement, and living environments.

Key words: Children and adolescents, Depressive disorders, Disease burden, Explainable machine learning, Summary exposure value

中图分类号: 

  • R179
[1] Thapar A, Eyre O, Patel V, et al. Depression in young people[J]. Lancet, 2022, 400(10352): 617-631
[2] GBD 2021 Diseases and Injuries Collaborators. Global incidence, prevalence, years lived with disability(YLDs), disability-adjusted life-years(DALYs), and healthy life expectancy(HALE)for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021[J]. Lancet, 2024, 403(10440): 2133-2161. DOI:10.1016/S0140-6736(24)00757-8[PubMed]
[3] Kieling C, Buchweitz C, Caye A, et al. Worldwide prevalence and disability from mental disorders across childhood and adolescence: evidence from the global burden of disease study[J]. JAMA Psychiatry, 2024, 81(4): 347-356.
[4] Piao J M, Huang Y Q, Han C, et al. Alarming changes in the global burden of mental disorders in children and adolescents from 1990 to 2019: a systematic analysis for the Global Burden of Disease study[J]. Eur Child Adolesc Psychiatry, 2022, 31(11): 1827-1845.
[5] Deng Y, Sun S W, Wu S X, et al. Burden and trends of mental disorders in China from 1990 to 2019: findings from the Global Burden of Disease Study 2019[J]. Soc Psychiatry Psychiatric Epidemiol, 2024, 59(9): 1563-1576.
[6] Zhu F Y, Yang Y, Yin T L, et al. The Burdenof adolescent depression and the impact of COVID-19 across 204 countries and regions from 1990 to 2021: results from the 2021 global burden of disease study[J]. Sci Rep, 2025, 15: 5658. DOI:10.1038/s41598-024-84843-w
[7] Liu X Z, Cang R, Zhang Z H, et al. Identification of depressive symptoms in adolescents using machine learning combining childhood and adolescence features[J]. BMC Public Health, 2025, 25(1): 264. DOI:10.1186/s12889-025-21506-z
[8] Yoo A, Li F Z, Youn J, et al. Prediction of adolescent depression from prenatal and childhood data from ALSPAC using machine learning[J]. Sci Rep, 2024, 14(1): 23282. DOI:10.1038/s41598-024-72158-9
[9] Patel J, Hung C, Katapally T R. Evaluating predictive artificial intelligence approaches used in mobile health platforms to forecast mental health symptoms among youth: a systematic review[J]. Psychiatry Res, 2025, 343: 116277. DOI:10.1016/j.psychres.2024.116277
[10] Brauer M, Roth G A, Aravkin A Y, et al. Global burden and strength of evidence for 88 risk factors in 204 countries and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021[J]. Lancet, 2024, 403(10440): 2162-2203.
[11] 曾四清, 李艳, 刘珺, 等. 两组序列数据趋势变化特征对比分析的Joinpoint回归模型方法及应用[J]. 中国卫生统计, 2021, 38(2): 307-311.
[12] Riebler A, Held L. Projecting the future burden of cancer: Bayesian age-period-cohort analysis with integrated nested Laplace approximations[J]. Biom J, 2017, 59(3): 531-549.
[13] Steare T, Gutiérrez Muñoz C, Sullivan A, et al. The association between academic pressure and adolescent mental health problems: a systematic review[J]. J Affect Disord, 2023, 339: 302-317. DOI:10.1016/j.jad.2023.07.028
[14] Fassi L, Thomas K, Parry D A, et al. Social media use and internalizing symptoms in clinical and community adolescent samples: a systematic review and meta-analysis[J]. JAMA Pediatr, 2024, 178(8): 814-822.
[15] 中共中央国务院印发《“健康中国2030”规划纲要》[J]. 中华人民共和国国务院公报, 2016(32):5-20.
[16] 国务院办公厅. 国务院办公厅关于转发卫生计生委等部门全国精神卫生工作规划(2015—2020年)的通知[EB/OL].(2015-06-18)[2026-03-21]. https://www.gov.cn/zhengce/content/2015-06/18/content_9860.htm
[17] 吴霞民, 马宁, 陈润滋, 等. 中国内地精神卫生防治技术管理机构现状调查[J]. 中国公共卫生, 2023, 39(4): 485-488. Wu Xiamin, Ma Ning, Chen Runzi, et al. Mental health prevention and control management facilities and personnel in the mainland of China, 2020: a cross-sectional survey[J]. Chinese Journal of Public Health, 2023, 39(4): 485-488.
[18] Liu Z, Kuai M N. The global burden of depression in adolescents and young adults, 1990-2021: systematic analysis of the global burden of disease study[J]. BMC Psychiatry, 2025, 25(1): 767. DOI:10.1186/s12888-025-07201-8
[19] Cagney J, Spencer C, Flor L, et al. Prevalence of sexual violence against children and age at first exposure: a global analysis by location, age, and sex(1990-2023)[J]. Lancet, 2025, 405(10492): 1817-1836.
[20] Spencer C N, Khalil M, Herbert M, et al. Health effects associated with exposure to intimate partner violence against women and childhood sexual abuse: a Burden of Proof study[J]. Nat Med, 2023, 29(12): 3243-3258.
[21] Yu Z Y, Cao Y H, Shang T, et al. Depression in youths with early life adversity: a systematic review and meta-analysis[J]. Front Psychiatry, 2024, 15: 1378807. DOI:10.3389/fpsyt.2024.1378807
[22] Song QQ, Yuan T S, Hu Y Y, et al. The effect of peer victimization during adolescence on depression and gender differences: a systematic review and meta-analysis[J]. Trauma Violence Abuse, 2024, 25(4): 2862-2876.
[23] Newbury J B, Heron J, Kirkbride J B, et al. Air and noise pollution exposure in early life and mental health from adolescence to young adulthood[J]. JAMA Netw Open, 2024, 7(5): e2412169. DOI:10.1001/jamanetworkopen.2024.12169
[24] Shoari N, Blangiardo M, Pirani M. Neighborhood characteristics and mental health from childhood to adolescence[J]. JAMA Netw Open, 2025, 8(4): e254470. DOI:10.1001/jamanetworkopen.2025.4470
[25] Zeng J, Xu Y B. The impact of family socioeconomic status on depression in Chinese adolescents: a comparison of life course models[J]. J AffectDisord, 2024, 356: 97-104. DOI:10.1016/j.jad.2024.04.010
[26] Talib M,Rachdi M, Papazova A, et al. The role of dietary patterns and nutritional supplements in the management of mental disorders in children and adolescents: an umbrella review of meta-analyses[J]. Can J Psychiatry, 2024, 69(8): 567-589.
[27] Fiani D, Chahine S, Zaboube M, et al. Psychiatric and cognitive outcomes of iron supplementation in non-anemic children, adolescents, and menstruating adults: a meta-analysis and systematic review[J]. Neurosci Biobehav Rev, 2025, 178: 106372. DOI:10.1016/j.neubiorev.2025.106372
[28] 况利, 徐何雁. 从精准评估到整合干预: 儿童青少年心理健康问题诊疗范式的革新与展望[J]. 山东大学学报(医学版), 2026, 64(3): 10-16. Kuang Li, Xu Heyan. From precision assessment to integrated intervention: innovations and future perspectives in the diagnostic and therapeutic paradigm for child and adolescent mental health disorders[J]. Journal of Shandong University(Health Science), 2026, 64(3): 10-16.
[29] Zhang Q Y, Wang J, Neitzel A. School-based mental health interventions targeting depression or anxiety: a meta-analysis of rigorous randomized controlled trials for school-aged children and adolescents[J]. J Youth Adolesc, 2023, 52(1): 195-217.
[30] Piantadosi S,Byar D P, Green S B. The ecological fallacy[J]. Am J Epidemiol, 1988, 127(5): 893-904.
[31] Stevens G A, Alkema L, Black R E, et al. Guidelines for accurate and transparent health estimatesreporting: the GATHER statement[J]. Lancet, 2016, 388(10062): 19-23.
[32] Collins G S, Moons K G M, Dhiman P, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods[J]. BMJ, 2024: e078378. DOI:10.1136/bmj-2023-078378
[1] 冯建, 杨丽, 王永俊, 刘长青, 田美娜, 周晗, 于连龙. 血清尿酸与儿童青少年代谢综合征的关联:基于山东、河北两省的横断面研究[J]. 山东大学学报 (医学版), 2026, 64(8): 96-106.
[2] 况利,徐何雁. 从精准评估到整合干预:儿童青少年心理健康问题诊疗范式的革新与展望[J]. 山东大学学报 (医学版), 2026, 64(3): 10-16.
[3] 李芳华,刘娜,李晓宇,许殿花,王倩. 儿童青少年抑郁障碍患者服药依从性研究进展[J]. 山东大学学报 (医学版), 2026, 64(3): 24-34.
[4] 王倩,胡晓婧,王爱青,孙强,孟春燕,李芳华. 儿童青少年抑郁障碍患者非自杀性自伤文献的可视化分析[J]. 山东大学学报 (医学版), 2026, 64(3): 93-107.
[5] 陈婵,李巨章,何稳,吴巧珍. 基于两样本孟德尔随机化研究抑郁症和抗抑郁药物靶基因与睡眠呼吸暂停的关联[J]. 山东大学学报 (医学版), 2026, 64(1): 28-36.
[6] 曹丹凤,李慧. 儿童期不良经历对妊娠期抑郁症状的影响:母胎依恋的中介作用[J]. 山东大学学报 (医学版), 2024, 62(3): 87-91.
[7] 王新起,徐晶晶,许旺旺,张洋,高进. 25(OH)D、IGF-I及焦虑、抑郁症状与门诊青少年精神障碍患者非自杀性自伤的关系[J]. 山东大学学报 (医学版), 2024, 62(12): 59-64.
[8] 张蒙,马伟. 1990—2019年中国人类免疫缺陷病毒/获得性免疫缺陷综合征流行趋势及疾病负担[J]. 山东大学学报 (医学版), 2023, 61(5): 84-89.
[9] 韩梅,孟维静,陶子琨,杨希,徐雅琪,穆华夏,卜伟晓,王素珍,石福艳. 基于G-计算的高血压、抑郁在2型糖尿病与认知功能之间的因果多中介分析[J]. 山东大学学报 (医学版), 2023, 61(10): 101-108.
[10] 曹丹凤,刘宗花,王君芝,张国翔. 女性儿童期家庭功能障碍对成年后妊娠期抑郁症状的独立和累积作用[J]. 山东大学学报 (医学版), 2022, 60(6): 97-101.
[11] 刘盈,杨淑霞,佘凯丽,程传龙,房启迪,韩闯,崔峰,李秀君. 2019年山东省淄博市居民死因及疾病负担探析[J]. 山东大学学报 (医学版), 2022, 60(6): 114-121.
[12] 孙富云,王维鹏,张会会,耿艳,安小霞,李双双,张彬彬. 结直肠癌术后患者人格特质与抑郁、焦虑症状的关联性[J]. 山东大学学报 (医学版), 2021, 59(7): 91-96.
[13] 许怀悦,王情,马润美,班婕,李湉湉. 温度相关的疾病负担研究进展[J]. 山东大学学报 (医学版), 2021, 59(12): 13-19.
[14] 吕岩,于潇,蔺新英,赵琦,王保珍. 聊城市老年女性膳食模式与抑郁症状的关系[J]. 山东大学学报 (医学版), 2020, 58(11): 103-108.
[15] 王颖,杨蓓,王娟,李艳艳,曹枫林. 累积儿童期逆境对护士抑郁症状的影响[J]. 山东大学学报 (医学版), 2019, 57(11): 83-89.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2018 《山东大学学报(医学版)》编辑部　
地址:山东大学科技期刊社（济南市历城区山大南路27号山东大学中心校区明德楼B座721室） 电话: 0531-88366918 E-mail: xbyxb@sdu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发