Edinburgh Dementia Prevention, Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, UK.
Faculty of Health and Life Sciences, Department of Psychology, Northumbria University, Newcastle, UK.
Age Ageing. 2021 May 5;50(3):868-874. doi: 10.1093/ageing/afaa232.
Cognitive dispersion, or inconsistencies in performance across cognitive domains, has been posited as a cost-effective tool to predict conversion to dementia in older adults. However, there is a dearth of studies exploring cognitive dispersion in the oldest-old (>80 years) and its relationship to dementia incidence.
The main aim of this study was to examine whether higher cognitive dispersion at baseline was associated with dementia incidence within an 8-year follow-up of very old adults, while controlling for established risk factors and suggested protective factors for dementia.
Participants (n = 468) were from the Origins of Variance in the Old-Old: Octogenarian Twins study, based on the Swedish Twin Registry. Cox regression analyses were performed to assess the association between baseline cognitive dispersion scores and dementia incidence, while controlling for sociodemographic variables, ApoEe4 carrier status, co-morbidities, zygosity and lifestyle engagement scores. An additional model included a composite of average cognitive performance.
Cognitive dispersion and ApoEe4 were significantly associated with dementia diagnosis. These variables remained statistically significant when global cognitive performance was entered into the model. Likelihood ratio tests revealed that cognitive dispersion and cognitive composite scores entered together in the same model was superior to either predictor alone in the full model.
The study underscores the usefulness of cognitive dispersion metrics for dementia prediction in the oldest-old and highlights the influence of ApoEe4 on cognition in very late age. Our findings concur with others suggesting that health and lifestyle factors pose little impact upon cognition in very advanced age.
认知分散，或在认知领域表现不一致，被认为是一种预测老年人向痴呆转化的具有成本效益的工具。然而，关于最年长的老年人（>80 岁）认知分散的研究很少，也没有研究其与痴呆发病率的关系。
本研究的主要目的是检验在对非常年长的老年人进行 8 年随访中，基线时较高的认知分散是否与痴呆发病率有关，同时控制痴呆的既定风险因素和潜在保护因素。
参与者（n=468）来自基于瑞典双胞胎登记处的老年双胞胎研究的起源变异：80 岁以上双胞胎研究。使用 Cox 回归分析来评估基线认知分散评分与痴呆发病率之间的关系，同时控制社会人口统计学变量、ApoEe4 携带状态、合并症、同卵性和生活方式参与度评分。一个额外的模型包括平均认知表现的综合得分。
认知分散和 ApoEe4 与痴呆诊断显著相关。当将整体认知表现纳入模型时，这些变量仍然具有统计学意义。似然比检验显示，在完整模型中，认知分散和认知综合评分一起进入同一个模型比任何一个预测因子单独进入模型的效果都要好。
该研究强调了认知分散指标在预测最年长老年人痴呆中的有用性，并强调了 ApoEe4 对非常高龄人群认知的影响。我们的研究结果与其他研究一致，表明健康和生活方式因素对非常高龄人群的认知影响不大。
