Using the Mini-Mental State Examination for Tracking Cognition in the Older Population Based on Longitudinal Data Mark Chatfield, MSc, à Fiona E. Matthews, PhD, wà Carol Brayne, MD, à and the Medical Research Council Cognitive Function and Ageing Study z OBJECTIVES: To estimate population norms for use in assessment of individuals in relation to their age-matched peers using true longitudinal patterns of decline. DESIGN: Longitudinal study of 10 years of follow-up data from the Medical Research Council Cognitive Function and Ageing Study (MRC CFAS) on the most commonly used cognitive test across clinical and research settings. SETTING: England and Wales. PARTICIPANTS: Thirteen thousand four people were seen in five sites at baseline, with follow up at 2, 5, and 10 years. MEASUREMENTS: Mini-Mental State Examination (MMSE) score at three interviews over 10 years. A total of 42,777 MMSE scores were used in the analysis. RESULTS: MMSE norms are presented according to age and split according to sex using longitudinal data. Potential cohort effects and dropout of individuals with low MMSE scores have been accounted for. CONCLUSION: It is likely that the cognitive MMSE scale will continue to be used in many settings and across the age range. The figures presented here can be used to plot indi- vidual performance and chart where there is change in the relative position of one individual compared with others. J Am Geriatr Soc 55:1066–1071, 2007. Key words: population norms; cognition; MRC CFAS; attrition T he maintenance of good cognition is fundamental to health in older age. Loss of cognitive ability is one of the core measures relevant to identification of dementing conditions. No test of cognition has been created that serves all of the purposes of cognitive researchers and clinicians. The ideal test needs to investigate general cognitive ability and identify specific deficits and to be reliable, sensitive to change, easily standardized, brief, and acceptable. The Mini-Mental State Examination (MMSE) 1 is the only test to gain international acceptance for researching cognition and dementia despite frequently voiced discussions of limita- tions and drawbacks. Hence, it is of interest to provide detailed information on the test for specific populations and according to age group, both cross-sectionally and longi- tudinally. Until better measures become as widely accepted, the MMSE will continue to be used, and in this context, norms are particularly useful. They can indicate the posi- tion of any given individual relative to others in the same population and that individual’s change over time com- pared with what is expected for that population. MMSE NORMS IN THE LITERATURE Many studies have published population norms from cross- sectional data. 2–14 These norms have come from a variety of nations, and in several instances the MMSE has been trans- lated into other languages. Even in the English version, items on the MMSE can vary, such as the three words to repeat and recall, and whether serial sevens, spelling ‘‘WORLD’’ backward, or the highest of these items is used. The majority of papers present the norms separately for the factors most thought to explain the variation in MMSE scores, which gives readers the ability to find the norms most appropriate for their particular patient. Norms are most often presented split according to education and age group 3–6,13,14 and sometimes also according to sex. 9,11,15 Norms on People with and without Dementia Together? Some researchers have produced norms for the population without dementia, 2,3,6–9,11,13,14 because the reference pop- ulation is then ‘‘normal’’ or healthy. It can also be difficult to obtain a score on persons with dementia. It is argued that producing norms after excluding individuals with dementia is more useful for the purpose of screening dementia, 3,6 although it is necessary also to know the scores of people with dementia and how these overlap with the scores of those without for good screening. 6 For this reason, one set Address correspondence to Dr. Fiona Matthews, MRC Biostatistics Unit, Institute of Public Health, Cambridge CB2 0SR, UK. E-mail: fiona.matthews@mrc-bsu.cam.ac.uk DOI: 10.1111/j.1532-5415.2007.01216.x z www.cfas.ac.uk From the à Department of Public Health and Primary Care, University of Cambridge, Cambridge, United Kingdom; w MRC Biostatistics Unit, Institute of Public Health, Cambridge, United Kingdom. JAGS 55:1066–1071, 2007 r 2007, Copyright the Authors Journal compilation r 2007, The American Geriatrics Society 0002-8614/07/$15.00