Computational modeling and empirical studies of hippocampal neurogenesis-dependent memory: Effects of interference, stress and depression.

Prolonged stress causes dysregulation in the hypothalamic-pituitary-adrenal axis and may contribute to the pathogenesis of major depressive disorder (MDD). MDD is associated with pathological changes in several brain regions, particularly the prefrontal cortex and hippocampus. Evidence from animal research suggests that one of the earliest signs of pathological change after exposure to stress is a reduction in hippocampal neurogenesis. We therefore sought to test the prediction that people in the earliest stages of a first episode of depression would show selective memory deficits on neurogenesis-dependent tasks. Our computational model predicts that new neurons are important for representing distinct contexts; thus, when overlapping memories are learned over an interval of several days, during which time some neuronal turnover has taken place, the neurogenesis should reduce the potential for interference between the overlapping memories. At much shorter time scales, within the span of a single memory episode, rather than contributing to pattern separation, neurogenesis might play more of an integrative role in mediating contextual associative learning. Consistent with this, empirical evidence from animal studies suggests a role for the new neurons in forming complex event memories that bridge across time delays. This leads us to predict selective memory deficits on putative neurogenesis-dependent tasks in the earliest pre-clinical stages of a first episode of depression, before a clinical diagnosis has been made and prior to the development of more serious pathological brain changes. We present the results of new simulations with the model, lending further support to the prediction that neurogenesis reduces interference when memory events are separated by several days. We also report findings from an empirical study in which we tested a large number of undergraduates on a set of cognitive and memory tests from the CANTAB battery, and also administered neuropsychological inventories for stress, depression and anxiety. One of the subtests in the CANTAB battery, the delayed match to sample (DMS) task, was of particular interest as delayed non-match to sample has been found in animal studies to be dependent upon neurogenesis. Our empirical results indicate that as predicted, participants scoring high on the Beck Depression Inventory show a selective deficit on the DMS at long delays while performing on par with non-depressed participants on all other tasks. The potential to detect very early signs of major depression using simple neurogenesis-dependent cognitive tests could have important implications for the diagnosis and treatment of this debilitating and highly prevalent disorder.