Abstract
Individual differences in cognitive aging are, in some cases, correlated with changes in molecular markers at the neuronal level. However, the use of simple correlations to analyze data across multiple age groups has a number of potential pitfalls. When young animals differ from aged animals on both of the dependent variables being assessed with a correlation analysis, the age difference often accounts for the detection of a relationship between the 2 measures. When the age groups are analyzed discretely, the data may exhibit a completely different trend, as suggested in a recent Commentary. In addition to reconsidering the interpretation of recently published data on the relationship between age-related deficits in cognition and hippocampal neurogenesis, the demands of the task should be taken into account when evaluating the contributions of newly-generated neurons.