Memory training alters hippocampal neurochemistry in healthy elderly Michael J. Valenzuela, 1,2,CA Megan Jones, 1 Wei Wen, 1 Caroline Rae, 3 Scott Graham, 4 Ronald Shnier 4 and Perminder Sachdev 1,2 1 Neuropsychiatric Institute, The Prince of Wales Hospital, Euroa Centre, Randwick, NSW 2031; 2 School of Psychiatry, The University of New South Wales, Sydney; 3 Department of Biochemistry, The University of Sydney; 4 St George MRI, St George Hospital, Sydney, Australia CA,1 Corresponding Author and Address: michaelv@unsw.edu.au Received 2 January 2003; accepted12 February 2003 DOI: 10.1097/01.wnr.0000077548.91466.05 Accumulating epidemiological evidence supports the notion of brain reserve, but there has been no investigation of neurobiologi- cal change associated with brief mental activation training in hu- mans. Healthy older individuals were therefore investigated with magnetic resonance spectroscopy (MRS) in di¡erent brain regions before and after 5 weeks of focused memory training. Recall of a test-word list of 4 23 items was achieved accompanied by eleva- tion of creatine and choline signals in the hippocampus. Those at risk for neural dysfunction, as indicated by lower neurometabolites at baseline, demonstrated the largest MRS increases after training. Biochemical changes related to cellular energy and cell-membrane turnover were found to increase after structured memory exer- cises and were limited to the medial temporal lobe. NeuroReport 14 :1333^1337  c 2003 Lippincott Williams & Wilkins. Key words: Ageing; Brain reserve; Creatine; Dementia; Magnetic resonance spectroscopy; Memory; Training INTRODUCTION How can lifetime patterns of mental activity modulate the pathogenesis or clinical manifestation of neurodegenerative disease? This concept, commonly referred to as brain reserve, has arisen from epidemiological studies showing that activities such as advanced education, occupational complexity, greater pre-morbid IQ and increased partici- pation in post-retirement leisure activities independently relate to lower risk for cognitive decline and incident dementia [1]. Neuroimaging studies have shown that cognitive performance can be preserved in individuals with Alzheimer’s disease (AD) perfusion deficits who also hold complex occupational histories or high educational levels [2]. Discovering the mechanisms by which protracted habits of mental activity should offer neuroprotection in late life is extremely challenging, with no satisfactory answers at present. Studying the effects of mental stimulation over short- term periods, say a number of weeks, is one way of making this problem simpler. In rats, brief periods of enriched mental activity lead to a number of beneficial neurotrophic changes, including neurogenesis [3], enhanced synaptic budding and dendritic arborization complexity [4] and even protection from brain disease [5,6]. Induction of the ARC gene and increased brain-derived neuro- trophic factor activity have been implicated in these brain changes [7]. Post-mortem human studies also con- firm the close link between brain reserve indices like education and pre-morbid IQ and synaptic density [8]. The neurobiological effect of structured mental activity programs in adult humans, however, remains untested. One cognitive memory program that is both brief and highly successful is the ancient method of loci (MOL; see Fig. 1a,b). First described in Cicero’s De Oratore B40 BC, it asks the subject to link features of a familiar environment with items from a list requiring memorization. Sequential retrieval is aided by walking though one’s mental land- scape, each landmark acting as a cue for a list item via a self-generated mental association. MOL performance there- fore depends on several cognitive functions including the generation of imagery, linguistic association, working memory, and in particular, mental map retrieval and navigation. Use of the MOL strategy can increase standard free recall from 7–10 word items to 30–40 items in sequence [9]. We used localized proton magnetic resonance spec- troscopy (MRS) to measure biochemistry in three different brain regions, before and after five weeks of MOL exercises. The right hippocampus [10], midline parietal–occipital region [11] and left frontal lobes [12] were chosen for spectroscopic evaluation because of their implication with cognitive processes identified as critical for MOL. MRS allows measurement of several metabolic products central to neural energy pathways, cell membrane integrity and neural function [13] (Fig. 2a–c). 0959- 4965  c Lippincott Williams & Wilkins Vol 14 No 10 18 July 2003 1333 AGEING NEUROREPORT Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.