Background: Subjective Memory Impairment (SMI) may hold value in the elderly as a predictor of dementia. There is yet to exist any standard definition of SMI for use in research or for clinical practice.
Objective: This study aims to identify previous and current definitions of SMI used in published research and to propose a set of criteria that may help increase SMI's predictive power of future cognitive decline.
Methods: Literature searches were conducted across a number of electronic databases including Medline.
Results: 515 citations were identified, 336 papers were obtained, of which 44 were selected for containing definitions for SMI. These definitions varied widely in terms of the types of questions used to determine SMI and additional features pertaining to memory complaints included in the definition.
Conclusion: There is no consistency in how SMI is defined. We propose a set of criteria aimed to increase specificity of memory complainers for those at increased risk of dementia. Further research is required to refine and validate the different criteria suggested. An international consent on the necessary criteria by experts in the field might be useful.