【摘要】 家庭医生签约服务是实现社区全生命周期健康管理的重要工作内容，提升非重点人群家庭医生签约率是目前面临的一大挑战。本文聚焦中青年人群，分别对亚健康状态和亚健康评估方法展开梳理，提出以疲劳为主要症状的疾病前亚健康状态评估或成为促进中青年人群签约的突破口。亚健康状态与慢性疾病的发生发展密不可分，将亚健康状态限定为以疲劳为主要症状的疾病前亚健康状态有助于对中青年人群的亚健康状态开展评估。分析现有亚健康状态评估方法优缺点及中青年亚健康评估研究的不足。当前亚健康状态评估以量表和问卷为主，主观性较强，缺乏客观评价指标，此外相比于传统的统计学建模方法，机器学习算法可以处理大样本量、多维度数据，具有更好的评估效能。针对中青年亚健康评估缺乏客观评估指标及方法，建议纳入多维度客观评估指标，运用集成机器学习方法，构建科学可行的疾病前亚健康状态的评估模型，并以慢性疾病临界指标设定模型阈值。以期通过科学有效的疾病前亚健康状态评估，提高中青年人群主动健康意识，并为其提供个性化健康管理服务，满足其对家庭医生服务的需求，从而提高中青年人群家庭医生签约意愿。更多还原

【Abstract】 Family doctor contract service is an important part of whole life-cycle health management in community. Improving the contract rate of family doctors for normal groups is a major challenge at present. This paper focuses on young and middle-aged people, and sorts out the sub-health status and assessment methods respectively, and proposes that the pre-disease sub-health status assessment with fatigue as the main symptom may become a breakthrough for promoting contracts signing in the young and middle-aged population. Sub-health status is closely related to the occurrence and development of chronic diseases. It is helpful to evaluate the sub-health status of young and middle-aged people by limiting it to pre-disease sub-health states with fatigue as the main symptom. Secondly, we comparatively analyzed the advantages and disadvantages of existing sub-health assessment methods and the shortcomings of sub-health assessment for young and middle-aged people. We found that the current sub-health status assessment is mainly based on scales and questionnaires, with strong subjectivity and lack of objective evaluation indicators. In addition, compared with traditional statistical modeling methods, machine learning algorithms can process large sample sizes and multi-dimensional data, which has better evaluation efficiency. In view of the lack of objective evaluation indicators and methods in sub-health assessment in young and middle-aged people, it is recommended that multi-dimensional objective evaluation indicators should be included to integrate machine learning methods to construct a scientifically feasible assessment model of pre-disease sub-health status, as well as to set the model threshold with chronic disease critical indicators. Expecting that the scientific and effective pre-disease sub-health status assessment can improve the active health awareness of young and middle-aged people and provide personalized health management services to meet their needs for family doctor services, thereby increasing their willingness to sign family doctor contracts.更多还原