Table 1_Association of memory function with COVID-19 outcomes in adults aged 50 years and older: analysis of three prospective cohorts.docx

BackgroundPatients with Alzheimer’s disease or dementia are at increased risk for COVID-19 hospitalization and mortality. However, no study has examined whether memory function is associated with COVID-19 outcomes in general older adults.
MethodsData were obtained from SHARE (the Survey of Health, Ageing and Retirement in Europe), HRS (the Health and Retirement Study), and ELSA (the English Longitudinal Study of Ageing), three prospective and representative cohorts of non-institutionalized adults aged 50 years and older in 25 European countries plus Israel, the United States, and the United Kingdom, respectively. Memory function was measured with immediate and delayed 10-words recall tests. Associations of 10-words recall with COVID-19 hospitalization and mortality were assessed using logistic models adjusted for age, sex, race, body mass index, smoking, physical activity, household income, education level, and chronic conditions.
ResultsA total of 4,062 participants with COVID-19 infection from SHARE, 1349 from HRS, and 278 from ELSA were included in the analysis. 610 (15.0%) in SHARE, 142 (10.5%) in HRS, and 39 (14.0%) in ELSA were hospitalized, and 102 (2.5%) died of COVID-19 or related complications in SHARE. The adjusted odds ratios (aORs) for COVID-19 hospitalization were 1.15 (95% CI, 1.09–1.22) in SHARE, 1.07 (95% CI, 0.94–1.21) in HRS, and 1.34 (95% CI, 1.02–1.77) in ELSA, per word decrease in immediate 10-words recall. For delayed 10-words recall, the corresponding aORs were 1.11 (95% CI, 1.06–1.17), 1.12 (95% CI, 1.01–1.24), and 1.25 (95% CI, 1.01–1.55), respectively. The aORs for COVID-19 mortality were 1.06 (95% CI, 0.93–1.20) and 1.14 (95% CI, 1.01–1.28) per word decrease in immediate and delayed 10-words recall in SHARE, respectively. Results were relatively robust to missing data of covariates, exclusion of cases based on symptoms alone, or exclusion of cases with Alzheimer’s disease or dementia.
ConclusionThis study shows that low memory performance, as measured by 10-words recall, is independently associated with an increased risk of COVID-19 hospitalization and mortality in adults aged 50 years and older.

背景：阿尔茨海默病或痴呆患者感染新型冠状病毒肺炎（COVID-19）后，住院及死亡风险均有所升高。然而，目前尚无研究针对普通老年人群探讨记忆功能与COVID-19预后的关联。
方法：本研究数据来自欧洲健康、老龄化与退休调查（Survey of Health, Ageing and Retirement in Europe, SHARE）、健康与退休研究（Health and Retirement Study, HRS）以及英国老龄化纵向研究（English Longitudinal Study of Ageing, ELSA）三个前瞻性代表性队列，分别纳入了25个欧洲国家加以色列、美国、英国的50岁及以上非机构化成人。记忆功能通过即刻和延迟10词回忆测验进行评估。采用校正了年龄、性别、种族、体质量指数、吸烟状况、体力活动、家庭收入、教育水平及慢性疾病的logistic回归模型，分析10词回忆得分与COVID-19住院、死亡的关联。
结果：本研究共纳入SHARE队列的4062名、HRS队列的1349名及ELSA队列的278名COVID-19感染者进行分析。其中，SHARE队列中有610例（15.0%）患者住院，HRS队列142例（10.5%），ELSA队列39例（14.0%）；SHARE队列中有102例（2.5%）因COVID-19或相关并发症死亡。即刻10词回忆得分每降低1分，SHARE队列COVID-19住院的校正比值比（adjusted odds ratios, aOR）为1.15（95%置信区间[CI]：1.09~1.22），HRS队列1.07（95%CI：0.94~1.21），ELSA队列1.34（95%CI：1.02~1.77）。延迟10词回忆得分每降低1分，对应的aOR分别为1.11（95%CI：1.06~1.17）、1.12（95%CI：1.01~1.24）及1.25（95%CI：1.01~1.55）。在SHARE队列中，即刻和延迟10词回忆得分每降低1分，COVID-19死亡的aOR分别为1.06（95%CI：0.93~1.20）与1.14（95%CI：1.01~1.28）。敏感性分析显示，当存在协变量缺失、仅基于症状排除病例或排除阿尔茨海默病/痴呆病例时，本研究结果均保持稳健。
结论：本研究表明，以10词回忆测验评估的记忆功能低下，与50岁及以上成人COVID-19住院及死亡风险升高独立相关。