Fax +41 61 306 12 34E-Mail [email protected] Original Research Article Dement Geriatr Cogn Disord 2010;30:238–244 DOI: 10.1159/000320137 Predictive Utility of Type and Duration of Symptoms at Initial Presentation in Patients with Mild Cognitive Impairment DeidreJ.Devier  a NicoleVillemarette-Pittman  a PatrickBrown  b,c GregoryPelton  b,c YaakovStern  b,c MarySano  d,e D.P.Devanand  b,c  a  Department of Neurology, Louisiana State University Health Sciences Center, New Orleans, La. , b  Division of Geriatric Psychiatry, Columbia University College of Physicians and Surgeons and New York State Psychiatric Institute, c  Department of Neurology, Sergievsky Center, Taub Institute for Research in Alzheimer’s Disease and the Aging Brain, and d  Department of Psychiatry, and Alzheimer’s Disease Research Center, Mount Sinai School of Medicine, New York, N.Y. , and e  Department of Neurology, James J. Peters VA Medical Center, Bronx, N.Y. , USA Introduction Mild cognitive impairment (MCI) is often a transi-tional stage between age-related memory decline andAlzheimer’s disease (AD) [1, 2] , though many individuals with MCI remain cognitively stable. Identifying the earli-est symptoms and their relative temporal onset may help to improve early diagnosis, estimate prognosis and even-tually impact treatment. Efforts have been made to stan-dardize methods of determining the date of onset and the types of symptoms present [3–7] . Symptoms appear well before one seeks treatment, and clarifying that time course by the employment of thorough interview tech-niques yields reliable findings [7] . Although memory de-cline is often the first symptom identified, other symp-toms may also be the earliest to appear. Oppenheim [3] found in a retrospective review of patients diagnosed with AD that about half of them reported early changes in psychiatric, neurological and functional modalities. Kang et al. [8] interviewed caregivers of persons with AD and found that, in addition to memory decline, changes in orientation, judgment, depression and language pre-ceded diagnosis. Key Words Age at onset ⴢ Alzheimer’s disease ⴢ Behavioral symptoms ⴢ Mild cognitive impairment ⴢ Neuropsychological assessment Abstract Background/Aims: To assess (1) the duration and symptoms present in participants with mild cognitive impairment (MCI) and (2) the impact of these variables on predicting conver-sion to Alzheimer’s disease (AD). Methods: Participants with MCI (n = 148) were assessed and followed systematically. Re-sults: Decline in memory was reported as the first symptom in 118 of the cases. Converters had more symptoms (e.g. lan-guage decline, depression), and the combination of decline in memory and in performance of high-order social/cogni-tive activities as well as disorientation more often than non-converters (p = 0.036). In an age-stratified Cox model, predic-tors of conversion to AD were shorter time since onset of memory decline and lower baseline MMSE score. Conclu-sions: Recent onset of memory decline with older age, de-creased MMSE score, change in performance and disorienta-tion indicate a greater likelihood of short-term conversionto AD. Copyright © 2010 S. Karger AG, Basel Accepted: August 6, 2010 Published online: September 15, 2010 Deidre J. Devier, PhD Department of Neurology, Louisiana State University Health Sciences Center 1542 Tulane Avenue, Floor 7 New Orleans, LA 70112 (USA) Tel. +1 504 939 8906, Fax +1 504 285 8078, E-Mail ddevie  @  lsuhsc.edu © 2010 S. Karger AG, Basel Accessible online at:www.karger.com/demOnset of Illness in MCI Dement Geriatr Cogn Disord 2010;30:238–244239 In the development of the ‘Onset of Illness Interview’, our group used a standardized semistructured inter-view to assess time of onset of specific domains of symp-toms, and tested it with informants of 36 individuals with dementia [4] . Open-ended questions about the first iden tified symptom referenced calendar dates and as-sisted informants to recall sa lient personal events to pin-point specific occurrences. The interview format also helped informants identify the date when that symptom was not present [4] . This semistructured interview pro-cess showed good interrater and interinformant reliabil-ity [4] and has been used effectively in other studies [8–12] . In addition to the reliability of the semistructured interview, there is a considerable literature to suggest that eliciting information from informants is more reli-able than self-report in patients with MCI or AD [13 –18] . In this study that examined several predictors of con-version from MCI to AD, the Onset of Illness Interview [4] was administered to the informants of patients who presented with cognitive impairment, but not dementia. The type of informant-reported symptoms present at ill-ness onset and the duration of these symptoms from on-set to study entry were also examined as predictors of conversion to AD. We hypothesized that a greater num-ber of early symptoms and a shorter time course of these symptoms would be associated with a quicker conver-sion to AD. In addition, we planned to explore the rela-tionship between early informant-reported symptoms and objectively measured baseline cognitive perfor-mance. Materials and Methods Participants Patients presented with memory complaints to either the Memory Disorders Clinic or the Center for Memory and Behav-ioral Disorders at the New York State Psychiatric Institute/Co-lumbia University. All participants met study criteria for cogni-tive impairment without dementia and provided informed con-sent in this New York State Psychiatric Institute/Columbia University institutional-review-board-approved protocol. Eligibility Criteria Inclusion criteria identified a wide group of cognitively im-paired individuals, aged 41–85 years, with memory complaints and without dementia with onset of cognitive impairment be-tween 6 months and 10 years prior to enrollment, and Folstein MMSE scores of 6 22/30 [19] . Cognitive screening inclusion guidelines were described previously [20] and included Folstein MMSE recall of ^ 2/3 objects at 5 min, or a Buschke Selective Re-minding Test delayed recall score 1 1 SD below norms, or a Wechsler Adult Intelligence Scale-Revised performance IQ score 6 10 points