目的: 对65~105岁老年人以年龄进行分层，探究照料需求的影响因素特点。方法: 共纳入12 244名2018年中国老年健康影响因素跟踪调查(Chinese longitudinal healthy longevity survey，CLHLS)横断面数据中的老年人，分为低龄(65~79岁)、中龄(80~89岁)、高龄(90~105岁)三组。利用综合失能指数(disability index，DI)从四个维度评估老年人的失能水平，反映其照料需求，采用无序多分类Logistic回归进行影响因素分析。结果: 12 244名老年人中，中、高照料需求的老年人占比达43.4%。老年人照料需求高的影响因素包括年龄更大、患慢性病数量更多、无锻炼习惯、睡眠时间过长(≥9 h/d)、有抑郁倾向、与子女同住、空巢合居、未受过教育(P < 0.05)。此外，低龄组老年人还有吸烟史(OR=2.009，95% CI：1.019~3.959)、饮酒史(OR=2.213，95% CI：1.141~4.291)和自评贫穷(OR=2.051，95% CI：1.189~3.540)；中龄组老年人还有女性(OR=1.373，95% CI：1.038~1.817)、从不饮酒(OR=1.551，95% CI：1.059~2.269)和无医疗保险(OR=1.598，95% CI：1.053~2.426)；高龄组老年人还有女性(中照料需求vs.低照料需求：OR=1.412，95% CI：1.062~1.878；高照料需求vs.低照料需求：OR=1.506，95% CI：1.137~1.993)、自评贫穷(OR=2.064，95% CI：1.282~ 3.323)和无医疗保险(OR=1.621，95% CI：1.148~2.291)。结论: 各年龄组老年人照料需求的共性影响因素包括年龄、慢性病、锻炼、睡眠、抑郁倾向、居住方式和受教育情况，低龄组的特有因素包括吸烟、饮酒和经济水平，中、高龄组的特有因素包括性别和医疗保险。建议针对上述影响因素开展前瞻性队列研究和重点人群的干预研究，为政策制定提供可靠依据。

Objective: To identify the factors associated with the care needs of the older adults aged 65-105 by age groups, and to compare these factors across different age groups. Methods: A total of 12 244 older adults from the Chinese longitudinal healthy longevity survey (CLHLS) conducted in 2018 were included in the analyses. The participants were categorized into three age groups: young-old (aged 65-79), middle-old (aged 80-89), and oldest-old (aged 90-105). The level of disability was measured by the disability index (DI) in four dimensions, reflecting their care needs. Potential factors associated with care needs were selected based on the health ecological model (HEM), including perspectives of personal characteristics, behavioral characteristics, interpersonal network, living and working conditions, and policy environment. Multifactor analysis was performed using multinomial Logistic regression. Results: Among China ' s 12 244 older adults, 43.4% had medium or high care needs. Factors for higher care needs of older adults included higher age, higher number of chronic diseases, no exercise habit, excessive sleep duration (≥9 h/d), depressive tendency, living with children or spouse, and uneducated (all P < 0.05). In addition, the young-old group who were past smokers (OR=2.009, 95% CI: 1.019-3.959), were past drinkers (OR=2.213, 95% CI: 1.141-4.291), and reported self-perceived poverty (OR=2.051, 95% CI: 1.189-3.540), had higher level of care needs. The middle-old group who were female (OR=1.373, 95% CI: 1.038-1.817), never drank alcohol (OR=1.551, 95% CI: 1.059-2.269), and were lack of medical insurance (OR=1.598, 95% CI: 1.053-2.426), and had higher level of care needs. The oldest-old group who were female (medium care needs vs. low care needs: OR=1.412, 95% CI: 1.062-1.878; high care needs vs. low care needs: OR=1.506, 95% CI: 1.137-1.993), reported self-perceived poverty (OR=2.064, 95% CI: 1.282-3.323), and were lack of medical insurance (OR=1.621, 95% CI: 1.148-2.291), and had higher level of care needs. Conclusion: The identical factors associated with care needs across different age groups include age, chronic disease, exercise, sleep, depression, living arrangement, and education. Smoking, alcohol consumption, and economic status are specific factors among the young-old group of the older adults, while gender and medical insurance are specific factors among the middle-old and the oldest-old group of the older adults. We recommend conducting prospective cohort studies and intervention studies among specific age groups on the above factors to provide reliable evidence for policy formulation.