中国临床试验注册中心 - 世界卫生组织国际临床试验注册平台一级注册机构

注册号： Registration number：	ChiCTR2600120470
最近更新日期： Date of Last Refreshed on：	2026-03-16 09:28:29
注册时间： Date of Registration：	2026-03-16 00:00:00
注册号状态：	预注册
Registration Status：	Prospective registration
注册题目：	基于垂直领域大模型的AI智能体对运动员焦虑与抑郁的影响
Public title：	A Domain-Specific Large Language Model–Based AI Agent for Reducing Depression and Anxiety in Athletes: A Randomized Controlled Trial
注册题目简写：
English Acronym：
研究课题的正式科学名称：	基于垂直领域大模型的AI智能体对运动员焦虑与抑郁的影响
Scientific title：	A Domain-Specific Large Language Model–Based AI Agent for Reducing Depression and Anxiety in Athletes: A Randomized Controlled Trial
研究课题代号(代码)： Study subject ID：
在二级注册机构或其它机构的注册号： The registration number of the Partner Registry or other register：

申请注册联系人：	汪琦一	研究负责人：	董晓晓
Applicant：	Qiyi Wang	Study leader：	Xiaoxiao Dong
申请注册联系人电话： Applicant telephone：	+86 155 7925 1232	研究负责人电话： Study leader's telephone：	+86 152 0526 0159
申请注册联系人传真： Applicant Fax：		研究负责人传真： Study leader's fax：
申请注册联系人电子邮件： Applicant E-mail：	wangqiyi979@163.com	研究负责人电子邮件： Study leader's E-mail：	xiaox_dong@163.com
申请单位网址(自愿提供)： Applicant website(voluntary supply)：		研究负责人网址(自愿提供)： Study leader's website(voluntary supply)：
申请注册联系人通讯地址：	中国江苏省扬州市邗江区文汇东路77号	研究负责人通讯地址：	中国江苏省南京市玄武区灵谷寺路8号
Applicant address：	No. 77, Wenhui East Road, Hanjiang District, Yangzhou, Jiangsu, China	Study leader's address：	No. 8, Lingguizhi Road, Xuanwu District, Nanjing, Jiangsu, China
申请注册联系人邮政编码： Applicant postcode：		研究负责人邮政编码： Study leader's postcode：	225009
申请人所在单位：	扬州大学
Applicant's institution：	Yangzhou University
研究负责人所在单位：	南京体育学院
Affiliation of the Leader：	Nanjing Sport Institute

是否获伦理委员会批准：	是
Approved by ethic committee：	Yes
伦理委员会批件文号： Approved No. of ethic committee：	RT-2026-04	伦理委员会批件附件： Approved file of Ethical Committee：	查看附件View
批准本研究的伦理委员会名称：	南京体育学院人体实验伦理委员会
Name of the ethic committee：	Human Experiment Ethics Committee of Nanjing Institute of Physical Education
伦理委员会批准日期： Date of approved by ethic committee：	2026-01-30 00:00:00
伦理委员会联系人：	陈磊
Contact Name of the ethic committee：	Lei Chen
伦理委员会联系地址：	中国江苏省南京市玄武区灵谷寺路8号
Contact Address of the ethic committee：	No. 8, Lingguizhi Road, Xuanwu District, Nanjing, Jiangsu, China
伦理委员会联系人电话： Contact phone of the ethic committee：	+86 25 8475 5776	伦理委员会联系人邮箱： Contact email of the ethic committee：

研究实施负责（组长）单位：

南京体育学院

Primary sponsor：

Nanjing Sport Institute

研究实施负责（组长）单位地址：

中国江苏省南京市玄武区灵谷寺路8号

Primary sponsor's address：

No. 8, Lingguizhi Road, Xuanwu District, Nanjing, Jiangsu, China

试验主办单位(项目批准或申办者)：

Secondary sponsor：

国家：	中国	省(直辖市)：	江苏	市(区县)：
Country：	China	Province：	Jiangsu	City：
单位(医院)：	南京体育学院	具体地址：	中国江苏省南京市玄武区灵谷寺路8号
Institution hospital：	Nanjing Sport Institute	Address：	No. 8, Lingguizhi Road, Xuanwu District, Nanjing, Jiangsu, China

经费或物资来源：

无

Source(s) of funding：

N/A

研究疾病：

无

Target disease：

None

研究疾病代码：

Target disease code：

研究类型：

干预性研究

Study type：

Interventional study

研究所处阶段：

其它

Study phase：

N/A

研究设计：

随机平行对照

Study design：

Parallel

研究目的：

在竞技体育高度专业化与系统化发展的背景下，运动员的心理健康正逐渐成为影响其竞技表现与长期发展的重要议题。其中，焦虑和抑郁被认为是最为常见且具有代表性的心理健康问题。焦虑通常体现为对比赛结果的过度担忧、睡眠障碍或注意力分散，而抑郁则可能表现为持久的情绪低落、动力缺失或自我价值感下降。根据相关研究数据显示，精英运动员中焦虑和抑郁的患病率显著高于一般人群，甚至可达20%-30%，这不仅影响竞技成绩，还可能引发长期健康隐患。然而，现有面向运动员群体的心理健康服务仍存在服务可及性有限、支持连续性短缺以及个体化适配不足等局限性，难以满足运动员在日常训练与比赛周期中有效获得心理支持的实际需求。因此，有必要探索更具可持续性与适应性的心理健康支持路径。近年来，随着人工智能技术的发展，基于垂直领域大模型的AI智能体逐渐被引入心理健康支持领域。该类智能体依托于运动心理健康领域的专业知识体系，具备较高的可获得性、持续交互能力和个体化响应潜力，被认为在运动员焦虑与抑郁的预防与改善方面具有应用前景。因此，本研究旨在系统评估一种基于垂直领域大模型的AI智能体对运动员焦虑与抑郁水平的影响，为运动员心理健康促进措施的合理选择与规范实施提供参考依据，并为人工智能技术在体育心理领域的应用积累实证证据。

Objectives of Study：

Against the backdrop of highly specialized and systematic development in competitive sports, athletes’ mental health has increasingly become a critical factor affecting both their performance and long-term development. Among mental health issues, anxiety and depression are recognized as the most common and representative. Anxiety typically manifests as excessive worry about competition outcomes, sleep disturbances, or difficulty concentrating, whereas depression may present as persistent low mood, lack of motivation, or diminished self-worth. Research indicates that the prevalence of anxiety and depression among elite athletes is significantly higher than in the general population, reaching 20%–30%, which not only impacts competitive performance but may also pose long-term health risks. However, existing mental health services for athletes face limitations such as restricted accessibility, lack of continuity, and insufficient individualization, making it difficult to meet athletes’ practical needs for effective psychological support during training and competition cycles. Therefore, it is necessary to explore more sustainable and adaptive pathways for mental health support. In recent years, with the advancement of artificial intelligence technologies, AI agents based on vertical-domain large language models have gradually been introduced into the field of mental health support. These agents, grounded in the professional knowledge system of sports psychology, offer high accessibility, sustained interactive capacity, and individualized responsiveness, and are considered to hold potential for the prevention and alleviation of anxiety and depression in athletes. Accordingly, this study aims to systematically evaluate the effects of a vertical-domain large language model-based AI agent on athletes’ levels of anxiety and depression, providing a reference for the rational selection and standardized implementation of interventions to promote athletes’ mental health, while also generating empirical evidence for the application of artificial intelligence in the field of sports psychology.

药物成份或治疗方案详述：

Description for medicine or protocol of treatment in detail：

纳入标准：

Inclusion criteria

排除标准：

为保障研究的科学性与被试安全，符合以下情况之一者将不纳入研究： 1. 未满18周岁者； 2. 存在严重心理或精神健康问题，且正在接受专业医疗或心理治疗者； 3. 在研究期间无法完成基本研究流程或无法持续参与者； 4. 对研究内容理解存在明显困难，或无法独立作出知情同意决定者。

Exclusion criteria：

To ensure the scientific nature of the research and the safety of the participants, those meeting any of the following conditions will not be included in the study: 1. Individuals under the age of 18; 2. Those with severe psychological or mental health issues and who are undergoing professional medical or psychological treatment; 3. Those who cannot complete the basic research procedures or cannot remain as participants during the study period; 4. Those who have significant difficulties understanding the research content or who cannot independently make an informed consent decision.

研究实施时间：

Study execute time：

从 From 2026-03-01 00:00:00至 To 2027-03-01 00:00:00

征募观察对象时间：

Recruiting time：

从 From 2026-03-25 00:00:00 至 To 2026-04-25 00:00:00

干预措施：

Interventions：

组别：	试验组	样本量：	30
Group：	Experimental group	Sample size：	30
干预措施：	本研究的干预措施为基于垂直领域大模型的AI心理支持智能体，旨在为运动员提供持续、个性化且情境化的心理支持，以改善焦虑和抑郁水平。干预工具为经过运动员心理健康语料微调的Qwen3大语言模型，封装为可通过Web或移动应用访问的智能体，提供非临床、非诊断性质的心理支持和心理教育功能。干预周期为28天，每位被试需完成16次有效互动，每次约10分钟，互动由运动员自主发起，内容包括自然语言对话、情绪表达与觉察指导、压力情境应对建议以及低强度心理练习。智能体通过情绪标签化反馈、共情回应和澄清性提问帮助运动员提升情绪觉察能力，并在训练、比赛及自我评价压力情境下提供结构化、可选择的心理调节建议，如呼吸放松、注意力控制、赛前心理准备及日常恢复策略。为降低干预门槛和依赖风险，智能体采用低频、短时、自主启动的支持模式，同时内置关键词与上下文安全监测机制，当检测到高强度负性情绪或潜在自伤风险时，会立即触发安全提示，引导运动员寻求线下专业心理支持或紧急帮助。所有互动日志均进行匿名化记录，仅用于干预效果评估，不涉及可识别个人信息。	干预措施代码：
Intervention：	The intervention in this study involves an AI psychological support agent driven by a vertical-domain large language model, designed to provide athletes with continuous, personalized, and context-specific psychological support to alleviate anxiety and depression. The intervention tool is a fine-tuned version of the Qwen3 large language model, adapted using athlete-specific psychological health dialogue corpora, and deployed via Web or mobile applications to provide non-clinical, non-diagnostic psychological support and psychoeducation. The intervention lasts 28 days, during which each participant completes 16 effective interactions, each lasting approximately 10 minutes, initiated by the athlete. Interactions include natural language conversations, guided emotional expression and awareness, stress-context coping suggestions, and low-intensity psychological exercises. The agent enhances emotional awareness through emotion labeling, empathic responses, and clarifying questions, while providing structured, optional strategies for managing training, competition, and self-evaluation stress, such as breathing relaxation, attention control, pre-competition mental preparation, and daily recovery strategies. To reduce intervention barriers and dependency risks, the agent operates in a low-frequency, short-duration, user-initiated mode and incorporates keyword- and context-based safety monitoring. When high-intensity negative emotions or potential self-harm risks are detected, safety alerts are triggered, clearly guiding the athlete to seek offline professional psychological support or emergency assistance. All interaction logs are anonymized and used solely for intervention evaluation, with no personally identifiable information involved.	Intervention code：

组别：	对照组	样本量：	30
Group：	Control group	Sample size：	30
干预措施：	等待组的参与者在研究期间不会接受基于垂直领域大模型的AI心理支持智能体干预，其干预措施为常规观察与标准化信息提供。参与者仅需按照研究要求完成与实验组相同的问卷测评，包括前测、后测及相关心理健康评估工具，如PHQ-9、GAD-7、PSQI及运动员心理健康服务需求量表，以确保数据可比性。等待组的设定旨在作为对照，评估AI智能体干预的特异性效应，同时不对参与者施加任何心理干预、训练或生活方式改变。所有测评均通过电子平台完成，参与者在研究过程中可自由退出，且退出权利与奖励机制与实验组一致，以保障伦理性与研究公平性。	干预措施代码：
Intervention：	Participants in the waitlist control group will not receive the AI-based psychological support agent intervention during the study period. Their "intervention" consists of routine observation and standardized information provision. Participants are required only to complete the same set of questionnaires and psychological assessments as the experimental group, including pre- and post-intervention measurements using PHQ-9, GAD-7, PSQI, and the Athlete Psychological Health Service Demand Scale, ensuring comparability of data. The waitlist group serves as a control to evaluate the specific effects of the AI agent intervention, without imposing any psychological support, training, or lifestyle modifications. All assessments are conducted electronically, and participants retain the right to withdraw at any time. Withdrawal procedures and any associated compensation are identical to those in the experimental group, ensuring ethical standards and fairness.	Intervention code：

研究实施地点：

Countries of recruitment and research settings：

国家：	中国	省(直辖市)：	江苏	市(区县)：
Country：	China	Province：	Jiangsu	City：
单位(医院)：	南京体育学院	单位级别：	大学
Institution hospital：	Nanjing Sport Institute	Level of the institution：	University

测量指标：

Outcomes：

指标中文名：	焦虑	指标类型：	主要指标
Outcome：	Anxiety	Type：	Primary indicator
测量时间点：		测量方法：
Measure time point of outcome：		Measure method：

指标中文名：	抑郁	指标类型：	主要指标
Outcome：	Depression	Type：	Primary indicator
测量时间点：		测量方法：
Measure time point of outcome：		Measure method：

指标中文名：	睡眠质量	指标类型：	次要指标
Outcome：	Sleep quality	Type：	Secondary indicator
测量时间点：		测量方法：
Measure time point of outcome：		Measure method：

指标中文名：	心理健康服务需求	指标类型：	次要指标
Outcome：	Mental Health Service Needs	Type：	Secondary indicator
测量时间点：		测量方法：
Measure time point of outcome：		Measure method：

采集人体标本：

Collecting sample(s)
from participants：

标本中文名：	无	组织：
Sample Name：	N/A	Tissue：
人体标本去向	其它	说明
Fate of sample：	0thers	Note：

征募研究对象情况：

Recruiting status：

尚未开始

Not yet recruiting

年龄范围：

Participant age：

最小 Min age	18	岁 years
最大 Max age		岁 years

性别：

男女均可

Gender：

Both

随机方法（请说明由何人用什么方法产生随机序列）：

本研究采用随机分组设计，将符合纳入标准的运动员按1:1比例分配至实验组或等待组。随机序列由不参与干预实施和数据评估的独立研究助理通过计算机生成随机数字表产生，以确保分组的客观性与不可预测性。分组信息在干预开始前由研究协调员统一封装，只有在参与者完成前测问卷后方可揭盲，以防分配偏倚。

Randomization Procedure (please state who generates the random number sequence and by what method)：

This study uses a randomized allocation design, assigning eligible athletes to the experimental group or the waitlist control group at a 1:1 ratio. The random sequence is generated by an independent research assistant, who is not involved in intervention delivery or data assessment, using a computer-generated random number table to ensure objective and unpredictable allocation. Group assignment information is securely sealed by the study coordinator and only revealed after participants complete the baseline assessments, preventing allocation bias.

是否公开试验完成后的统计结果:

Calculated Results after the Study Completed public access:

公开/Public

盲法：	本研究采用单盲设计，评估人员在干预期间对参与者的组别保持盲态。具体而言，负责数据采集、问卷发放与评分的研究人员不参与干预实施，也不知晓参与者是实验组还是等待组，从而减少测量偏倚。参与者由于需要使用AI心理支持智能体或处于等待组，因此无法对自身分组保持盲态，但研究者通过标准化指导和电子平台操作尽量保持数据收集过程的一致性和客观性。此外，数据分析阶段将采用编号化数据，分析人员在初步统计时对分组信息保持盲态，以进一步降低分析偏倚的风险。
Blinding：	This study employs a single-blind design, in which outcome assessors remain blinded to participants’ group assignments during the intervention period. Specifically, research personnel responsible for data collection, questionnaire administration, and scoring do not participate in intervention delivery and are unaware of whether participants are in the experimental or waitlist control group, reducing measurement bias. Participants, due to their engagement with the AI psychological support agent or assignment to the waitlist group, cannot be blinded to their own group; however, standardized instructions and electronic platform procedures are implemented to ensure consistency and objectivity in data collection. Furthermore, during the data analysis phase, all datasets are coded, and analysts remain blinded to group assignments during initial statistical processing to minimize analysis bias.
试验完成后的统计结果（上传文件）：	点击下载
Calculated Results after the Study Completed(upload file)：	download

是否共享原始数据： IPD sharing	否No
共享原始数据的方式（说明：请填入公开原始数据日期和方式，如采用网络平台，需填该网络平台名称和网址）：	无
The way of sharing IPD”(include metadata and protocol, If use web-based public database, please provide the url)：	None
数据采集和管理（说明：数据采集和管理由两部分组成，一为病例记录表(Case Record Form, CRF)，二为电子采集和管理系统(Electronic Data Capture, EDC)，如ResMan即为一种基于互联网的EDC：	电子采集和管理系统
Data collection and Management (A standard data collection and management system include a CRF and an electronic data capture：	Electronic Data Capture
数据与安全监察委员会： Data and Safety Monitoring Committee：	暂未确定/Not yet
注册人： Name of Registration：	2026-03-16 09:28:23