Objective: Normative data for older adults may be tainted by inadvertent inclusion of undiagnosed individuals at the very early stage of a neurodegenerative process. To avoid this pitfall, we developed norms for a cohort of older adults without MCI/dementia at 3-year follow-up. Methods: A randomly selected sample of 1041 community-dwelling individuals (age ≥ 65) received a full neurological and neuropsychological examination on two occasions [mean interval = 3.1 (SD = 0.9) years]. Results: Of these, 492 participants (Group 1; 65–87 years old) were without dementia on both evaluations (CDR=0 and MMSE ≥ 26); their baseline data were used for norms development. Group 2 (n = 202) met the aforementioned criteria only at baseline, but not at follow-up. Multiple linear regressions included demographic predictors for regression-based normative formulae and raw test scores as dependent variables for each test variable separately. Standardized scaled scores and stratified discrete norms were also calculated. Group 2 performed worse than Group 1 on most tests (p-values < .001–.021). Education was associated with all test scores, age with most, and sex effects were consistent with the literature. Conclusions: We provide a model for developing sound normative data for widely used neuropsychological tests among older adults, untainted by potential early, undiagnosed cognitive impairment, reporting regression-based, scaled, and discrete norms for use in clinical settings to identify cognitive decline in older adults. Additionally, our co-norming of a variety of tests may enable intra-individual comparisons for diagnostic purposes. The present work addresses the challenge of developing robust normative data for neuropsychological tests in older adults.

中文翻译：

---

为希腊老年人的神经心理学评估制定可靠的规范数据

目的：老年人的规范数据可能会因在神经退行性过程的早期阶段无意中纳入未确诊的个体而受到污染。为了避免这个陷阱，我们为一组在 3 年随访中没有 MCI/痴呆的老年人制定了规范。方法：随机抽取 1041 名社区居民（年龄 ≥ 65 岁）作为样本，接受两次全面的神经学和神经心理学检查[平均间隔 = 3.1 (SD = 0.9) 年]。结果：其中，492 名参与者（第 1 组；65-87 岁）在两项评估中均未患痴呆症（CDR=0 且 MMSE ≥ 26）；他们的基线数据用于制定规范。第 2 组（n= 202）仅在基线时满足上述标准，但在随访时不满足。多元线性回归包括基于回归的规范公式的人口预测因子和原始测试分数，分别作为每个测试变量的因变量。还计算了标准化评分和分层离散规范。在大多数测试中，第 2 组的表现比第 1 组差（p-值 < .001–.021)。教育程度与所有考试成绩相关，年龄影响最大，性别影响与文献一致。结论：我们提供了一个模型，用于为老年人中广泛使用的神经心理学测试开发合理的规范数据，不受潜在的早期、未诊断的认知障碍的污染，报告基于回归的、按比例缩放的和离散的规范，用于临床环境中识别老年人的认知衰退。成年人。此外，我们对各种测试的共同规范可以实现用于诊断目的的个体内比较。目前的工作解决了为老年人神经心理学测试开发可靠的规范数据的挑战。