Institute of Gerontology, Wayne State University, Detroit, MI, United States; Department of Psychology, Wayne State University, Detroit, MI, United States.
Institute of Gerontology, Wayne State University, Detroit, MI, United States.
Dev Cogn Neurosci. 2021 Dec;52:101037. doi: 10.1016/j.dcn.2021.101037. Epub 2021 Nov 17.
Functional MRI (fMRI) is a key tool for investigating neural underpinnings of cognitive development. Yet, in recent years, the reliability of fMRI effects has come into question and with it, the feasibility of using task-based fMRI to identify developmental changes related to cognition. Here, we investigated the reliability of task-based fMRI activations with a widely used subsequent memory paradigm using two developmental samples: a cross-sectional sample (n = 85, age 8-25 years) and a test-retest sample (n = 24, one-month follow up, age 8-20 years). In the large cross-sectional sample, we found good to excellent group-level reliability when assessing activation patterns related to the encoding task and subsequent memory effects. In the test-retest sample, while group-level reliability was excellent, the consistency of activation patterns within individuals was low, particularly for subsequent memory effects. We observed consistent activation patterns in frontal, parietal, and occipital cortices, but comparatively lower test-retest reliability in subcortical regions and the hippocampus. Together, these findings highlight the limitations of interpreting task-based fMRI effects and the importance of incorporating reliability analyses in developmental studies. Leveraging larger and densely collected longitudinal data may help contribute to increased reproducibility and the accumulation of knowledge in developmental sciences.
功能磁共振成像（fMRI）是研究认知发展神经基础的重要工具。然而，近年来，fMRI 效应的可靠性受到质疑，使用基于任务的 fMRI 来识别与认知相关的发展变化的可行性也受到质疑。在这里，我们使用一种广泛使用的后续记忆范式，通过两个发展性样本（一个横断面样本，n=85，年龄 8-25 岁；一个测试-重测样本，n=24，一个月后随访，年龄 8-20 岁）来研究基于任务的 fMRI 激活的可靠性。在大的横断面样本中，我们发现评估与编码任务和后续记忆效应相关的激活模式时，组水平的可靠性较好到极好。在测试-重测样本中，虽然组水平的可靠性极好，但个体内部的激活模式一致性较低，特别是对于后续记忆效应。我们观察到在额、顶和枕叶皮层中存在一致的激活模式，但在皮质下区域和海马体中的测试-重测可靠性相对较低。总之，这些发现强调了解释基于任务的 fMRI 效应的局限性，以及在发展研究中纳入可靠性分析的重要性。利用更大和更密集收集的纵向数据可能有助于提高发展科学的可重复性和知识积累。
