Prognostic Significance of Mild Cognitive Impairment Subtypes for Dementia and Mortality: Data from the NEDICES Cohort

Abstract

Background:

The predictive value of diverse subtypes of mild cognitive impairment (MCI) for dementia and death is highly variable.

Objective:

To compare the predictive value of several MCI subtypes in progression to dementia and/or mortality in the NEDICES (Neurological Disorders in Central Spain) elderly cohort.

Methods:

Retrospect algorithmic MCI subgroups were established in a non-dementia baseline NEDICES cohort using Spanish adaptations of the original Mini-Mental State Examination (MMSE-37) and Pfeffer’s Functional Activities Questionnaire (Pfeffer-11). The presence of MCI was defined according two cognitive criteria: using two cut-offs points on the total MMSE-37 score. Five cognitive domains were used to establish the MCI subtypes. Functional capacity (Pfeffer-11) was preserved or minimally impaired in all MCI participants. The incident dementia diagnoses were established by specialists and the mortality data obtained from Spanish official registries.

Results:

3,411 participants without dementia were assessed in 1994-5. The baseline prevalence of MCI varied according to the MCI definition (4.3%–31.8%). The follow-up was a mean of 3.2 years (1997-8). The dementia incidence varied between 14.9 and 71.8 per 1,000/person-years. The dementia conversion rate was increased in almost all MCI subgroups (p > 0.01), and mortality rate was raised only in four MCI subtypes. The amnestic-multi-domain MCI (aMd-MCI) had the best dementia predictive accuracy (highest positive likelihood ratio and highest clinical utility when negative).

Conclusions:

Those with aMd-MCI were at greatest risk of progression to dementia, as in other surveys and might be explored with increased attention in MCI research and in dementia preventive trials.

Keywords

Amnesic multidomain MCI dementia dementia conversion MCI mortality MCI subtypes MMSE-37 NEDICES Pfeffer’s FAQ

INTRODUCTION

Mild cognitive impairment (MCI) is often considered to be a pre-dementia stage of cognitive decline [1]. Although not all individuals with MCI progress to dementia, this condition is associated with increased risk of dementia and mortality and has generated in the last two decades clinical and scientific interest [2]. Recognition of MCI introduces the possibility of intervention to delay progression to dementia conversion with future effective treatments [3]. Currently, MCI is considered a cognitive entity between normal cognition and mild dementia [2, 4], but it has numerous operative definitions [5] and diverse causes and outcomes [2, 6]. As a result, figures of prevalence, incidence [7, 8], and outcomes [9 –11] show discrepancies.

The NEDICES (Neurological Disorders in Central Spain) cohort is a prospective census-based survey that screened 5,278 elderly participants from three areas of Central Spain in 1994-5 [12 –14]. The cohort had a follow-up until 2008, with mortality data collected by means of the Spanish Death Registries in 2009-10 [15]. This study analyzes the progression of several algorithmic definitions of MCI that may better predict progression to dementia and mortality between 1994-5 to 1997-8. These definitions have been frequently employed in previous MCI cohorts [2 , 16–19] and were utilized in others studies of the NEDICES cohort [20, 21]. In one of them, we demonstrated that our retrospective MCI definition of MCI subtypes conferred an increased rate of mortality (and dementia as cause-specific of mortality) over 13 years, suggesting good predictive validity of this approach [22]. Obviously, this algorithmic strategy has limitations; the clinical-based diagnosis of MCI has potentially better face validity [9, 23], and several authors recommend MCI diagnosis only following a comprehensive psychometric evaluation accompanied by objective data of cognitive decline [2 , 24]. In the baseline NEDICES population sample, we were limited to only a simple psychometric and functional evaluation, namely the Spanish version of the Mini-Mental State Examination [25], adapted to populations with low level of education and with modifications of the original items (MMSE-37) [26], and the Pfeffer’s Functional Activities Questionnaire (FAQ) [27] that also had a Spanish adaptation tailored to the Spanish sociocultural background [28]. In the second and third wave (conducted only in rural areas), NEDICES participants received a brief neuropsychological battery [29], with validation for the diagnosis of dementia [30 –33]. The limitation of baseline psychometric and functional evaluation was due to the multiple objectives of the NEDICES study in a population cohort [12 –14]. This type of evaluation was mainly based on the classic MMSE-30 [34 –40] (item analysis, total score or both) or modified MMSE (3MS) [41], which have been used in elderly longitudinal population or community-based surveys [34 , 40–42], or in clinical settings [35 , 39]. In those studies, the MMSE has been used as MCI diagnostic tool [37, 40], or as prognostic tool for short-medium term evolution (3–6 years) of cognitive decline [39, 40], or for dementia prediction [34–36 , 42]. Previous studies suggest that the MMSE-30 for MCI diagnosis has adequate dementia prediction [34, 38], with good specificity but moderate sensitivity [36]. Obviously, more comprehensive psychometric evaluation is required in order to increase its sensitivity [36, 43], However, in meta-analyses of MMSE utility in detecting MCI, there was no data about the efficacy of MMSE-30 items analysis in this diagnosis or on its population-based dementia prediction [6 , 45].

In this study, the MCI algorithmic definition utilized the MMSE-37: total score (1 and 1.5 SD below the mean of non-demented subjects) [7 , 42] along with deficits in specific cognitive domains (one or more deficits) (1.5 SD below the mean) [2 , 40] also obtained from the MMSE-37. Our aim in this investigation is to evaluate the predictive validity for dementia and mortality of these algorithmic MCI classifications in a short period of time (three years), in which the dementia conversion is high [36, 46].

METHODS

Study population

The NEDICES cohort population and methods has been described elsewhere [12 –14]. In summary, the NEDICES population was sampled from the census of three Central Spain communities to obtain a cohort of elderly people (>64 years) with different socioeconomic backgrounds: Las Margaritas, a working-class neighborhood in Getafe (Greater Madrid); Lista, a professional-class neighborhood in Salamanca district (downtown Madrid); and 38 villages of the Arévalo county (125 km northwest of Madrid). The three areas were chosen because they have approximately 2,000 elderly inhabitants, a computer-based registry of elders’ medical data in the primary physician setting, and relationships between the NEDICES team and the local physicians and health authorities. In this survey, the household and nursing home populations of the three communities were covered. Written (signed) informed consent was obtained from all participants and the ethical committees of two university hospitals from Madrid approved its methods [12, 13].

Baseline (1994-1995) and incidence (1997-8) dementia evaluation

An in-person evaluation was performed by lay trained interviewers with the aid of participant’s general practitioners at baseline; this evaluation comprised a 500-items questionnaire face-to-face with the participant, assessing demographic information, health status, lifestyle habits, medical and neurological disorders (including depressive symptoms and complaints of cognitive decline), cardiovascular risk factors, all current medications, and the name of their own family physician. A short form of the questionnaire was mailed to participants who were unavailable to attend a face-to-face or telephone evaluation [12 –14]. The screening instruments for dementia (MMSE-37 and Pfeffer-11) were the same in the baseline evaluation and the incidence wave: the MMSE-37 [26] that was adapted to Spanish from the standard MMSE, adding three additional items: (1) an attention task (i.e., “say 1, 3, 5, 7, 9 backwards”), (2) a visual order (i.e., a man raising his arms), and (3) a simple construction task (i.e., copying two overlapping circles); and the Pfeffer-11 [28] (Two items were removed from the original version [27] and three were added: handles own medications, can go outside alone, and greets appropriately). Both the MMSE-37 and Pfeffer-11 have been tested for validity in dementia screening, demonstrating high sensitivity and specificity [30 –32]. The method of dementia detection was described elsewhere [12 –14]. We established a two-phase diagnosis: phase I in which every participant had the screening for cognitive impairment and, when tested positive, underwent a neurological evaluation at a National Health Service clinic or at home. The screening phase was considered positive if the individual scored <24 points on the MMSE-37 and >5 points on the Pfeffer-11 or the individual could not completed any of both tests; and the individual, proxy, or medical data gave information regarding suspicion of cognitive decline. Every participant with positive screen underwent a phase II with cognitive and medical diagnosis performed by a specialist. The final dementia diagnosis was based on the consensus of three experts according to the DSM-IV criteria [32]. Questionable dementia (QD) cases were also clinically diagnosed according to a recommendation of WHO-AAD project [30 , 47] (all cases had positive screening) but without clear social (or domestic) impairment to be classified as dementia cases according to the expert consensus [32]. This group of cases (with abnormal functional evaluation) and their evolution were closer to mild dementia than to MCI [32] and were also eliminated in other similar survey [36].

MCI diagnosis

In this study, the algorithmic retrospective MCI definition was as described in the International Working Group (IWG) recommendations [4]. The presence of cognitive impairment was based on performances obtained from the MMSE-37 at baseline. Accordingly, the MCI cases showed evidence of cognitive impairment on MMSE-37 with preserved or ‘minimally impaired’ activities of daily living (score on Pfeffer-11 ≤5), but did not meet conventional diagnostic criteria for dementia (or QD) [2, 4]. Subjective memory complaints were not included, since a substantial proportion of participants did not give this information at the baseline survey. MCI cases were subclassified using the MMSE-37as follows:

Total score (TS) of the MMSE-37 For this approach, two cut-off points (1.5 SD: moderate cognitive impairment and 1.0 SD: mild cognitive impairment below the mean score of non-dementia participants) were obtained from the total MMSE-37 TS to determine MCI-1.5 and MCI-1.0 groups respectively [7 , 42]. These cut-off points were adjusted for those who were illiterate [48].

Specific cognitive domain (CD) in the MMSE-37 Algorithmic definitions of the four Petersen’s MCI subtypes [2, 4] were adapted from MMSE-37 cognitive domains as described in previous surveys [2, 7 , 40]. Thus, MCI classification was achievedaccording to the presence of memory impairment (amnesic and non-amnesic single) and the number of domains altered (single or multiple domains). Five composite scores were calculated: 1) spatial-temporal orientation; 2) attention-concentration (serial subtraction 7 from 100, and digits backwards); 3) memory (word recall); 4) language (naming, repeating, comprehension, and writing); and 5) visuo-constructive abilities (visual reproduction of the two figures) summing item performances of the individuals. This classification required the presence of at least one affected cognitive domain [cut-offs 1.5 SD below the mean of the baseline non-dementia cases (demented and QD cases excluded)]. As literacy has a profound effect on the performance of the MMSE, specific cut-off points by domain were calculated for illiterate subjects [48].

Non-dementia cohort

This baseline sub-cohort of NEDICES survey includes all participants that had completed with both measures (MMSE-37 and Pfeffer-11) with the exclusion of cases with a diagnosis of dementia or QD, and participants with abnormal Pfeffer-11 (>5 points) scale.

Statistical analyses

Data analyses were performed using SPSS Version 19.0 (SPSS, Inc., Chicago, IL). The relationship between socio-demographic and biomedical characteristics with the presence of MCI diagnosis was performed using bivariate analyses (first step), and logistic regression analysis (second step) after adjusting for several covariates such as age, gender, and years of education.

Cox proportional-hazards models to estimate hazard ratios (HRs) were used to calculate the risk of dementia in every MCI subgroup (with 95% confidence intervals, CI95%). The time variable for theses analyses was calculated from date at the baseline evaluation (1994-5) to second evaluation (incidence survey) and the participants death was computed if the date of death was previous to the end date of the incidence survey (31 December 1998). The predictive values for dementia (sensitivity, specificity, positive and negative predictive value, likelihood ratios, accuracy and positive and negative clinical utility) were also calculated using several algorithm definitions of MCI.

RESULTS

Non-dementia cohort

Figure 1 describes the flow chart of this study. From the baseline NEDICES cohort (n = 5,278) we excluded 306 participants with dementia and 83 with QD, and 1,478 were eliminated because they had incomplete cognitive screening (MMSE-37 or Pfeffer-11; n = 1,120); also those (n = 356) with abnormal scores on Pfeffer-11 and with missing data on mortality status (n = 2) were excluded. Thus, 3,411 participants were selected at baseline. At follow-up, 323 were dead (included in the mortality analysis), 237 were non-respondents (lost or refused), 83 suffered from incident dementia (and 19 from incident QD), and 433 were not included because they had missing data on screening (MMSE-37/Pfeffer-11) or abnormal Pfeffer-11 score during incidence phase; only 2,316 participants were evaluable for MCI diagnosis at follow-up.

The main socio-demographic and health characteristics of the non-dementia cohort are summarized (Table 1); the main data of the 3,411 selected and the excluded 1,478 subject were depicted. The main differences between the selected participants (n = 3,411) and those subjects who were excluded of the study (n = 1,478) were on age, gender, and education; the excluded group was older (age = 75.4±7.5 years) than the non-dementia cohort (age = 72.8±5.9, p < 0.001), had more females (59.9% versus 54.4% ; p < 0.003) and less years of education (excluded group = 6.0±5.1 versus 6.9±5.2 years of education;p < 0.001).

MCI prevalence according to the algorithmic definitions

Definitions based on MMSE-37 TS

The mean MMSE-37 score (n = 3,411) was 30.1 points (SD = 4.8); subjects with scores over 26 points were considered cognitively “normal” (CN) (n = 2,949 participants). Those subjects who scored <26 points [1.0 SD below the mean; n = 462 (13.5%)] were integrated into the MCI-1.0 subgroup, whereas the subjects who scored <23 points [1.5 SD below de mean were n = 145 (4.3%)] into the MCI-1.5 subgroup.

Specific cognitive domain (CD) definitions

The four MCI subtypes based on MMSE-37 five CD (4MCI-CD) comprised 1,085 (31.8%) participants that showed impairment in at least one domain of the MMSE-37 and 2,326 were considered as CN (CN-CD). Further to this general classification, the 4MCI-CD included the following groups: amnesic MCI-single domain (a-MCI; n = 259; 7.6%), amnesic multiple-domain (aMd-MCI; n = 193; 5.7%), non-amnestic single domain (na-MCI, n = 517; 15.2%), and non-amnestic multiple-domain (naMd-MCI; n = 116; 3.4%). Table 2 describes the subgroups based in CD definitions.

Table 3 depicts the frequency and relationships of the MCI subgroups (four CD and two TS subgroups) and its respective CN subjects. As it is shown, there is a complex overlapping between the groups. The majority of CN individuals classified by domains (N = 2,326) are also normal in TS categorization (only 15 participants scored <1 SD and one <1.5 SD were TS abnormal, mainly in those who were illiterate or had sensory deficit), but 639 subjects with normal MMSE-37 TS showed a cognitive deficit in any CD. The two MCI subgroups with greater overlapping were the MCI-1.5 and the aMd-MCI with 83 common cases.

Conversion to dementia and mortality incidence at follow-up

Table 4 depicts the progression to dementia and incidence of death (per 1,000-person-years) of seven MCI subgroups (MCI 1.5 and 1.0; a-MCI; aMd-MCI, na-MCI, naMd-MCI, and 4MCI-CD summing the four CD subgroups) and their respective CN subgroups. The mean follow-up from the baseline to the incidence wave was 3.2 years (range 0.4–4.5) years (follow-up from baseline to clinical incidence visit or until death during the incidence period).

Looking at MCI subgroups, all subgroups showed a higher dementia incidence rate than the corresponding CN controls, but in the case of naMd-MCI this did not reach a statistical significance. The highest rate of progression to dementia belonged to the aMd-MCI subtype: 71.8 cases (CI 95% : 46.9–105.2) per 1,000 persons-year compared to the CN subgroup (CN-CD): 4.3 (CI 95% : 2.8–6.4) per 1,000 persons-year (p < 0.01). The MCI subgroups with higher mortality rates were 1.5 MCI and aMd-MCI subgroups.

The conversion to dementia subtypes (AD, vascular dementia, and others) was not related to the MCI subtypes; the conversion to AD oscillated between 62.5% –69.5% in the MCI subtypes, but in the a-MCI subgroup there was greater risk (83.3%), although this was not statistically significant. The conversion to vascular dementia and other dementia subtypes were similar in all MCI subgroups (Table 5).

Evolution of the MCI subgroups

The main demographic data and evolution data at 3.2 years of the most significant MCI subgroups are shown in Table 6. Also, this table illustrates the MCI subgroup compared with CN controls, and its dementia conversion and other evolution data. The aMd-MCI subgroup had the lowest reversion to CN followed (10.9%) followed by the MCI-1.5 (24.8%) (p < 0.05, Chi squared test = 11.487).

MCI and associated variables

Baseline age, years of education, and self-rated health were associated with all MCI groups; no other variables were consistently associated with MCI after adjusting by socio-demographics covariates (age, gender, and years of education) (see Supplementary Tables).

Incident dementia and mortality

Table 7 showed the main risk factors for dementia in several MCI subtypes by means Cox’s hazard that demonstrated that the risk of dementia was significantly higher in all MCI subgroups versus their corresponding CN groups (p < 0.001) with the exception of naMd-MCI (not shown in the table for space reasons). The values of HR varied from 3.20 (CI95% : 1.96–5.22) in MCI-1.0, to HR 5.09 (CI95% : 3.00–8.63) for aMd-MCI (p < 0.001). Age (HR = 1.12, CI 95% : 1.08–1.15); years of education (HR 0.92, CI 95% : 0.87–0.98), and bad (poor-very poor) self-rated health (HR = 1.83, CI95% : 1.15–2.90) were significant risk factors for dementia conversion. Gender is not a risk factor for dementia conversion. Scores on Pfeffer-11 were not a significant risk factor (Cox regression analyses) for dementia conversion (not shown in Table 6).

Predictive accuracy of MCI subtypes

Table 8 analyzes the predictive accuracy of the most relevant MCI subgroups for the incident of dementia. The 4MCI-CD showed the highest sensitivity = 0.71, whereas all MCI subgroups had low-intermediate sensitivity values (range: 0.19–0.52). The positive predictive value (PPV) was low in all groups. However, negative predictive value (NPV) was very high suggesting that the absence of MCI diagnosis would effectively rule out the risk of dementia (over 3.2 years). Negative clinical utility used as a rule-out confirmation was strongest for MCI-1.5 and aMd-MCI. The aMd-MCI subgroup showed a high positive likelihood ratio (LR+): 9.54 (CI 95% : 6.93–13.12). The negative likelihood ratio (LR-) was minimal or small for all groups. According to this analysis the aMd-MCI subgroup had the better predictive values for dementia at 3.2 years follow-up.

DISCUSSION

MCI is a well-established cognitive syndrome with evolving definitions [2, 4], that has raised awareness of the spectrum of cognitive impairment short of dementia [2 , 50]. Early discrimination of MCI mainly from mild dementia cases remains a challenge, especially in the general population (51,52).

MCI is a common syndrome with greater prevalence than dementia (10% or higher in people over 65 years) [8 , 54]. MCI confers an increased risk of progression to dementia, but many MCI subjects revert to normal and others maintain this cognitive state many years [6 , 54–57]. The risk of dementia in those with MCI is greater in clinical setting and primary care than in population-based surveys [6 , 57]. The risk factors for this conversion, apart from aging, male gender, education, and basal cognition are debated [10 , 59]. Genetic (ApoE and others) and some comorbidities (diabetes mellitus) had a low predictive capacity [2 , 58–60]. MCI subtypes have different predictive capacity for dementia and mortality [6 , 54–61]. In this context we used several algorithmic MCI definitions including absence of dementia (and QD in this cohort) and diminished cognitive performance but without abnormal function [2, 4]. However, subjective memory loss was not included because many baseline participants did not answer this question and its utility is under discussion [62 –66]. Currently, the IWG increases the memory loss deficit, considering cognitive complaints most précised condition [4].

Our findings can be summarized as follows:

The different MCI definitions determine variable MCI prevalence (4.5% –31.8%), finding observed in studies with different MCI definitions [2 , 54], as well as in dementia prevalence [67].

The MCI subgroups were cognitively heterogeneous and generated frequent overlap between them [5 , 69] (e.g., the MCI-1.5 and aMd-MCI showed an overlap higher than 40%); this finding and the high frequency of a-MCI are possibly due to our simple cognitive exam, although analogous results appear in studies with wider psychometric evaluation [7 , 68–72]; the high prevalence of non-amnesic MCI is due to the inclusion of subjects with minimal cognitive deficits (68.3% had normal MMSE-37 TS).

At follow-up, the MCI subgroups determined higher dementia conversion in all MCI subgroups than in their corresponding CN subgroup (p > 0.01) [2 , 54], but the non-amnesic subgroups had low predictive dementia conversion and death risk (the naMd-MCI did not reach statistical significance). The highest dementia conversion and stability (less conversion to CN) was obtained with the most cognitive affected subgroups: MCI-1.5 and mainly with aMd-MCI [6 , 58], findings observed in population-based and clinical surveys [9–11 , 73]. The suggestion that a change in MCI definition affects MCI prevalence but not outcome [74] is not corroborated in this survey (MCI definitions affect both prevalence and outcome) [17, 55].

The different MCI subgroups did not determine specific dementia subtypes at follow-up, with the exception of a-MCI (to AD), findings previously described [75 –78].

In this study, the MCI risk factors for dementia conversion were age and years of education [2 , 59], but not gender as in NEDICES total cohort [29]; neither disease or comorbidity were risk factors for dementia conversion, with the exception of self-rated health [79], data observed [80, 81].

Our MCI definitions demonstrated low sensitivity (0.19–0.71) [with the exception of 4MCI-CD (0.71) that includes more than the 30% of the non-dementia cohort]. Two definitions with high cognitive deficits (aMd-MCI and MCI-1.5) had an elevated specificity (higher than any other MCI definition), but low sensitivity [76, 77]. This finding is relevant for preventive dementia trials (rule-out diagnosis). The aMd-MCI has the best dementia predictive accuracy (higher PL+; high clinical utility when negative) (Table 8). This observation is consistent with twoprevious analyses of this cohort [20, 21]. In practice, for increasing the rule-in predictive value is necessary to include cases with subjective memory or cognitive loss, non-amnesic subtypes and use with 1.0 SD cutoffs for impaired cognitive function [2 , 82]. Therefore, for preventive AD trials the need is to increase diagnostic specificity (stronger cognitive deficits –1.5 SD cutoffs- or multidomain deficits: aMd-MCI), or to add clinical data, biomarkers, or neuroimaging [2 , 76], although the cognitive markers are stronger dementia predictors than the biomarkers [83].

The most interesting finding of this study is the high rate of dementia conversion in the subgroup with aMd-MCI. This replicated an early study from Bokozi et al. [84] and later confirmed in different settings[6 , 85–88]. Several surveys with similar MCI definitions to our study, but with a more extensive psychometric evaluation [61 , 89] also showed similar results. The aMd-MCI is the only stable MCI subgroup in a longitudinal clinical investigation [88]. This observation has biological plausibility in research [90] showing that the temporal lobe gray matter atrophy has a continuum: from normal cognition, amnesic MCI, aMd-MCI until mild AD.

The main limitation of this study is the retrospective MCI diagnosis, the elemental cognitive evaluation (only MMSE-37 and Pfeffer-11), the evaluation of MCI with the same dementia screening instruments, and also to explore a selected non-dementia sub-cohort (younger and more educated) of NEDICES [29] in which we eliminated all subjects with abnormal Pfeffer-11 (nevertheless this was less restrictive that other surveys [17 , 92]). We had several strengths: the high number of participants, the adequate definition of dementia cases [29, 32], the mortality data (link with the official Spanish Death Registries), and the low-moderate rate of non-respondent participants in NEDICES [93].

Our main objective to compare the predictive accuracy of several MCI subgroups in progression of MCI to dementia and/or mortality may prove valuable given the increasing needs of dementia preventive trials. Our results suggest that MCI should be routinely subclassified [94] and further that the aMd-MCI [2] subgroup was the most stable and had the highest predictive accuracy for dementia conversion. These findings should be considered in future MCI dementia preventive trials.

Footnotes

ACKNOWLEDGMENTS

To all NEDICES cohort collaborators and funding institutions. The Spanish Health Research Agency, the Spanish Office of Science and Technology and CIBERNED have supported NEDICES. More details about collaborators and funding are listed on . This study is jointly financed by ISCII, FEDER and PI12/01602.

Authors’ disclosures available online ().

Appendix

References

Reisberg

, Ferris

, Kluger

, Franssen

, Wegiel

, de Leon

(2008) Mild cognitive impairment (MCI): A historical perspective. Int Psychogeriatr 20, 18–31.

Petersen

, Caracciolo

, Brayne

, Gauthier

, Jelic

, Fratiglioni

(2014) Mild cognitive impairment: A concept in evolution. J Intern Med 275, 214–228.

Cooper

, Li

, Lyketsos

, Livingston

(2013) Treatment for mild cognitive impairment: Systematic review. Br J Psychiatry 203, 255–264.

Winblad

, Palmer

, Kivipelto

, Jelic

, Fratiglioni

, Wahlund

L-O

, Nordberg

, Bäckman

, Albert

, Almkvist

, Arai

, Basun

, Blennow

, de Leon

, DeCarli

, Erkinjuntti

, Giacobini

, Graff

, Hardy

, Jack

, Jorm

, Ritchie

, van Duijn

, Visser

, Petersen

(2004) Mild cognitive impairment–beyond controversies, towards a consensus: Report of the International Working Group on Mild Cognitive Impairment. J Intern Med 256, 240–246.

Matthews

, Stephan

BCM

, Bond

, McKeith

, Brayne

, Medical Research Council Cognitive Function and Ageing Study. (2007) Operationalization of mild cognitive impairment: A graphical approach. PLoS Med 4, 1615–1619.

Mitchell

, Shiri-Feshki

(2009) Rate of progression of mild cognitive impairment to dementia–meta-analysis of 41 robust inception cohort studies. Acta Psychiatr Scand 119, 252–265.

Trittschuh

, Crane

, Larson

, Cholerton

, McCormick

, McCurry

, Bowen

, Baker

, Craft

(2011) Effects of varying diagnostic criteria on prevalence of mild cognitive impairment in a community based sample. J Alzheimers Dis 25, 163–173.

Ward

, Arrighi

, Michels

, Cedarbaum

(2012) Mild cognitive impairment: Disparity of incidence and prevalence estimates. Alzheimers Dement 8, 14–21.

Bruscoli

, Lovestone

(2004) Is MCI really just early dementia? A systematic review of conversion studies. Int Psychogeriatr 16, 129–140.

10.

Farias

, Mungas

, Reed

, Harvey

, DeCarli

(2009) Progression of mild cognitive impairment to dementia in clinic- vs community-based cohorts. Arch Neurol 66, 1151–1157.

11.

Ward

, Tardiff

, Dye

, Arrighi

(2013) Rate of conversion from prodromal Alzheimer’s disease to Alzheimer’s dementia: A systematic review of the literature. Dement Geriatr Cogn Dis Extra 3, 320–332.

12.

Bermejo

, Gabriel

, Vega

, Morales

, Rocca

, Anderson

, Neurological

Disorders in Central Spain (NEDICES) Study Group

(2001) Problems and issues with door-to-door, two-phase surveys: An illustration from central Spain. Neuroepidemiology 20, 225–231.

13.

Morales

, Bermejo

, Benito-León

, Rivera-Navarro

, Trincado

, Gabriel

, Vega

, NEDICES Study Group (2004) Methods and demographic findings of the baseline ine survey of the NEDICES cohort: A door-to-door survey of neurological disorders in three communities from Central Spain. Public Health 118, 426–433.

14.

Bermejo-Pareja

, Benito-León

, Vega-QS , Díaz-Guzmáan

, Rivera-Navarro

, Molina

, Olazarán-Rodríguez

, Morales-Gonzílez

(2008) The NEDICES cohort of the elderly. Methodology and main neurological findings. Rev Neurol 46, 416–423.

15.

Villarejo

, Benito-León

, Trincado

, Posada

, Puertas-Martín

, Boix

, Medrano

MRAJ

, Bermejo-Pareja

(2011) Dementia-associated mortality at thirteen years in the NEDICES Cohort Study. J Alzheimers Dis 26, 543–551.

16.

Ganguli

, Snitz

, Saxton

, Chang

C-CH

, Lee

C-W

, Vander Bilt

, Hughes

, Loewenstein

, Unverzagt

, Petersen

(2011) Outcomes of mild cognitive impairment by definition: A population study. Arch Neurol 68, 761–767.

17.

Manly

, Tang

M-X

, Schupf

, Stern

, Vonsattel

J-PG

, Mayeux

(2008) Frequency and course of mild cognitive impairment in a multiethnic community. Ann Neurol 63, 494–506.

18.

Matthews

, Stephan

BCM

, McKeith

, Bond

, Brayne

, Medical

Research Council Cognitive Function and Ageing Study

(2008) Two-year progression from mild cognitive impairment to dementia: To what extent do different definitions agree?. J Am Geriatr Soc 56, 1424–1433.

19.

Ritchie

, Tuokko

(2010) Patterns of cognitive decline, conversion rates, and predictive validity for 3 models of MCI. Am J Alzheimers Dis Other Demen 25, 592–603.

20.

Olazarám

, Trincado

, Bermejo

, Benito-León

, Díaz

, Vega

(2004) Selective memory impairment on an adapted Mini-Mental State Examination increases risk of future dementia. Int J Geriatr Psychiatry 19, 1173–1180.

21.

Villarejo

, Bermejo-Pareja

, Trincado

, Olazarán

, Benito-León

, Rodríguez

, Medrano

, Boix

, Vega

(2011) Memory impairment in a simple recall task increases mortality at 10 years in non-demented elderly. Int J Geriatr Psychiatry 26, 182–187.

22.

Contador

, Bermejo-Pareja

, Mitchell

, Trincado

, Villarejo

, Sánchez-Ferro

, Benito-León

(2014) Cause of death in mild cognitive impairment: A prospective study (NEDICES).253-e. J Eur Neurol 21, 9–.

23.

McCarten

(2013) Clinical evaluation of early cognitive symptoms. Clin Geriatr Med 29, 791–807.

24.

Petersen

(2011) Clinical practice. Mild cognitive impairment. N Engl J Med 364, 2227–2234.

25.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

26.

Prieto

, Contador

, Tapias-Merino

, Mitchell

, Bermejo-Pareja

(2012) The Mini-Mental-37 test for dementia screening in the Spanish population: An analysis using the Rasch Model. Clin Neuropsychol 26, 1003–1018.

27.

Pfeffer

, Kurosaki

, Harrah

, Chance

, Filos

(1982) Measurement of functional activities in older adults in the community. J Gerontol 37, 323–329.

28.

Olazarán

, Mouronte

, Bermejo

(2005) Clinical validity of two scales of instrumental activities in Alzheimer’s disease. Neurologia 20, 395–401.

29.

Bermejo-Pareja

, Benito-León

, Vega

, Medrano

, Román

, Neurological Disorders in Central Spain (NEDICES) Study Group (2008) Incidence and subtypes of dementia in three elderly populations of central Spain. J Neurol Sci 264, 63–72.

30.

Baldereschi

, Meneghini

, Quiroga

, Albala

, Mamo

, Muscat

, Gabriel

, Bermejo

, Amaducci

, Katzman

(1994) Cognitive versus functional screening for dementia across different countries: Cross-cultural validation of the Mini-Mental State Examination (MMSE) and the Pfeffer activities questionnaire (PFAQ) against the standardized clinical diagnosis of dementia. Neurology 44(Suppl 2), A365.

31.

Villanueva-Iza

, Bermejo-Pareja

, Berbel-Garcia

, Trincado Soriano

, Rivera Navarro

(2003) [Validation of a clinical protocol for the detection of dementia in the population]. Rev Neurol 36, 1121–1126.

32.

Bermejo-Pareja

, Benito-León

, Vega

, Olazarán

, de Toledo

, Díaz-Guzmán

, Sánchez-Sánchez

, Morales-González

, Trincado

, Portera-Sánchez

, Román

(2009) Consistency of clinical diagnosis of dementia in NEDICES: A population-based longitudinal study in Spain. J Geriatr Psychiatry Neurol 22, 246–255.

33.

Serna

, Contador

, Bermejo-Pareja

, Mitchell

, Fernández-Calvo

, Ramos

, Villarejo

, Benito-León

(2015) Accuracy of a brief neuropsychological battery for the diagnosis of dementia and / or mild cognitive impairment. A diagnostic validity analysis from the NEDICES cohort. J Alzheimers Dis 48, 163–173.

34.

Small

, Viitanen

, Bäckman

(1997) Mini-Mental State Examination item scores as predictors of Alzheimer’s disease: Incidence data from the Kungsholmen Project, Stockholm. J Gerontol A Biol Sci Med Sci 52, M299–304.

35.

, Meyer

, Thornby

(2001) Longitudinal follow-up of depressive symptoms among normal versus cognitively impaired elderly. Int J Geriat Psychiat 16, 718–727.

36.

Palmer

, Wang

H-X

, Bäckman

, Winblad

, Fratiglioni

(2002) Differential evolution of cognitive impairment in nondemented older persons: Results from the Kungsholmen Project. Am J Psychiatry 159, 436–442.

37.

Diniz

BSO

, Yassuda

, Nunes

, Radanovic

, Forlenza

(2007) Mini-mental State Examination performance in mild cognitive impairment subtypes. Int Psychogeriatr 19, 647–656.

38.

Sonobe

, Hata

, Ishikawa

, Sonobe

, Matsumoto

, Toyota

, Mori

, Fukuhara

, Komori

, Ueno

S-I

, Tanimukai

, Ikeda

(2011) Risk of progression from mild memory impairment to clinically diagnosable Alzheimer’s disease in a Japanese community (from the Nakayama Study). Int Psychogeriatr 23, 772–779.

39.

Xie

, Mayo

, Koski

(2011) Predictors of future cognitive decline in persons with mild cognitive impairment. Dement Geriatr Cogn Disord 32, 308–317.

40.

Moretti

, De Ronchi

, Palmer

, Forlani

, Morini

, Ferrari

, Dalmonte

, Atti

(2013) Prevalence and characteristics of mild cognitive impairment in the general population. Data from an Italian population-based study: The Faenza Project. Aging Ment Health 17, 267–275.

41.

Hogan

, Ebly

(2000) Predicting who will develop dementia in a cohort of Canadian seniors. Can J Neurol Sci 27, 18–24.

42.

Palmer

, Bäckman

, Winblad

, Fratiglioni

(2008) Mild cognitive impairment in the general population: Occurrence and progression to Alzheimer disease. Am J Geriatr Psychiatry 16, 603–611.

43.

Lin

, O’Connor

, Rossom

, Perdue

, Eckstrom

(2013) Screening for cognitive impairment in older adults: A systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 159, 601–612.

44.

Lonie

, Tierney

, Ebmeier

(2009) Screening for mild cognitive impairment: A systematic review. Int J Geriatr Psychiatry 24, 902–915.

45.

Mitchell

(2009) A meta-analysis of the accuracy of the mini-mental state examination in the detection of dementia and mild cognitive impairment. J Psychiatr Res 43, 411–431.

46.

Johansson

, Zarit

(1997) Early cognitive markers of the incidence of dementia and mortality: A longitudinal study of the oldest old. Int J Geriatr Psychiatry 12, 53–59.

47.

Baldereschi

, Amato

, Nencini

, Pracucci

, Lippi

, Amaducci

, Gauthier

, Beatty

, Klassen

, Galea

, Muscat

, Osuntokun

, Ogunniyi

, Portera-Sánchez

, Bermejo

, Hendrie

, Burdine

, Brashera

, Farlow

, Maggi

, Katzman

(1994) Cross-national interrater agreement on the clinical diagnostic criteria for dementia. Neurology 44, 239–242.

48.

Ardila

, Bertolucci

, Braga

, Castro-Caldas

, Judd

, Kosmidis

, Matute

, Nitrini

, Ostrosky-Solis

, Rosselli

(2010) Illiteracy: The neuropsychology of cognition without reading. Arch Clin Neuropsychol 25, 689–712.

49.

Bensadon

, Odenheimer

(2013) Current management decisions in mild cognitive impairment. Clin Geriatr Med 29, 847–871.

50.

Langa

, Levine

(2014) The diagnosis and management of mild cognitive impairment: A clinical review. JAMA 312, 2551–2561.

51.

Duara

, Loewenstein

, Greig

, Acevedo

, Potter

, Appel

, Raj

, Schinka

, Schofield

, Barker

, Wu

, Potter

(2010) Reliability and validity of an algorithm for the diagnosis of normal cognition, mild cognitive impairment, and dementia: Implications for multicenter research studies. Am J Geriatr Psychiatry 18, 363–370.

52.

Bartley

, Bokde

, O’Neill

(2011) Mild cognitive impairment.1357-1358; author reply. N Engl J Med 365, 1358–1359.

53.

Plassman

, Langa

, Fisher

, Heeringa

, Weir

, Ofstedal

, Burke

, Hurd

, Potter

, Rodgers

, Steffens

, McArdle

, Willis

, Wallace

(2008) Prevalence of cognitive impairment without dementia in the United States. Ann Intern Med 148, 427–434.

54.

Mariani

, Monastero

, Mecocci

(2007) Mild cognitive impairment: A systematic review. J Alzheimers Dis 12, 23–35.

55.

Ganguli

, Dodge

, Shen

, DeKosky

(2004) Mild cognitive impairment, amnestic type: An epidemiologic study. Neurology 63, 115–121.

56.

Koepsell

, Monsell

(2012) Reversion from mild cognitive impairment to normal or near-normal cognition: Risk factors and prognosis. Neurology 79, 1591–1598.

57.

Kaduszkiewicz

, Eisele

, Wiese

, Prokein

, Luppa

, Luck

, Jessen

, Bickel

, Mösch

, Pentzek

, Fuchs

, Eifflaender-Gorfer

, Weyerer

, König

H-H

, Brettschneider

, van den Bussche

, Maier

, Scherer

, Riedel-Heller

, Study on Aging, Cognition, Dementia in Primary Care Patients (AgeCoDe) Study Group (2014) Prognosis of mild cognitive impairment in general practice: Results of the German AgeCoDe study. Ann Fam Med 12, 158–165.

58.

Campbell

, Unverzagt

, LaMantia

, Khan

, Boustani

(2013) Risk factors for the progression of mild cognitive impairment to dementia. Clin Geriatr Med 29, 873–893.

59.

Cooper

, Sommerlad

, Lyketsos

, Livingston

(2015) Modifiable predictors of dementia in mild cognitive impairment: A systematic review and meta-analysis. Am J Psychiatry 172, 323–334.

60.

Adams

, de Bruijn

, Hofman

, Uitterlinden

, van Duijn

, Vernooij

, Koudstaal

, Ikram

(2015) Genetic risk of neurodegenerative diseases is associated with mild cognitive impairment and conversion to dementia. Alzheimers Dement 11, 1277–1285.

61.

Han

, Kim

, Lee

, Park

, Lee

, Huh

, Park

, Jhoo

, Lee

, Kim

(2012) Predictive validity and diagnostic stability of mild cognitive impairment subtypes. Alzheimers Dement 8, 553–559.

62.

Purser

, Fillenbaum

, Wallace

(2006) Memory complaint is not necessary for diagnosis of mild cognitive impairment and does not predict 10-year trajectories of functional disability, word recall, or short portable mental status questionnaire limitations. J Am Geriatr Soc 54, 335–338.

63.

Lenehan

, Klekociuk

, Summers

(2012) Absence of a relationship between subjective memory complaint and objective memory impairment in mild cognitive impairment (MCI): Is it time to abandon subjective memory complaint as an MCI diagnostic criterion?. Int Psychogeriatr 24, 1505–1514.

64.

Benito-León

, Mitchell

, Vega

, Bermejo-Pareja

(2010) A population-based study of cognitive function in older people with subjective memory complaints. J Alzheimers Dis 22, 159–170.

65.

Edmonds

, Delano-Wood

, Galasko

, Salmon

, Bondi

, Alzheimer’s Disease Neuroimaging Initiative (2014) Subjective cognitive complaints contribute to misdiagnosis of mild cognitive impairment. J Int Neuropsychol Soc 20, 836–847.

66.

Mitchell

(2008) The clinical significance of subjective memory complaints in the diagnosis of mild cognitive impairment and dementia: A meta-analysis. Int J Geriatr Psychiatry 23, 1191–1202.

67.

Erkinjuntti

, Ostbye

, Steenhuis

, Hachinski

(1997) The effect of different diagnostic criteria on the prevalence of dementia. N Engl J Med 337, 1667–1674.

68.

Nordlund

, Rolstad

, Hellström

, Sjögren

, Hansen

, Wallin

(2005) The Goteborg MCI study: Mild cognitive impairment is a heterogeneous condition. J Neurol Neurosurg Psychiatr 76, 1485–1490.

69.

Libon

, Xie

, Eppig

, Wicas

, Lamar

, Lippa

, Bettcher

, Price

, Giovannetti

, Swenson

, Wambach

(2010) The heterogeneity of mild cognitive impairment: A neuropsychological analysis. J Int Neuropsychol Soc 16, 84–93.

70.

Alladi

, Arnold

, Mitchell

, Nestor

, Hodges

(2006) Mild cognitive impairment: Applicability of research criteria in a memory clinic and characterization of cognitive profile. Psychol Med 36, 507–515.

71.

Manly

, Bell-McGinty

, Tang

M-X

, Schupf

, Stern

, Mayeux

(2005) Implementing diagnostic criteria and estimating frequency of mild cognitive impairment in an urban community. Arch Neurol 62, 1739–1746.

72.

Kramer

, Nelson

, Johnson

, Yaffe

, Glenn

, Rosen

, Miller

(2006) Multiple cognitive deficits in amnestic mild cognitive impairment. Dement Geriatr Cogn Disord 22, 306–311.

73.

Brodaty

, Heffernan

, Kochan

, Draper

, Trollor

, Reppermund

, Slavin

, Sachdev

(2013) Mild cognitive impairment in a community sample: The Sydney Memory and Ageing Study. Alzheimers Dement 9, 310.e1–317.e1.

74.

Fisk

, Merry

, Rockwood

(2003) Variations in case definition affect prevalence but not outcomes of mild cognitive impairment. Neurology 61, 1179–1184.

75.

Rasquin

SMC

, Lodder

, Visser

, Lousberg

, Verhey

FRJ

(2005) Predictive accuracy of MCI subtypes for Alzheimer’s disease and vascular dementia in subjects with mild cognitive impairment: A 2-year follow-up study. Dement Geriatr Cogn Disord 19, 113–119.

76.

Busse

, Hensel

, Gühne

, Angermeyer

, Riedel-Heller

(2006) Mild cognitive impairment: Long-term course of four clinical subtypes. Neurology 67, 2176–2185.

77.

Yaffe

, Petersen

, Lindquist

, Kramer

, Miller

(2006) Subtype of mild cognitive impairment and progression to dementia and death. Dement Geriatr Cogn Disord 22, 312–319.

78.

Jungwirth

, Zehetmayer

, Hinterberger

, Tragl

, Fischer

(2012) The validity of amnestic MCI and non-amnestic MCI at age 75 in the prediction of Alzheimer’s dementia and vascular dementia. Int Psychogeriatr 24, 959–966.

79.

Fernández-Ruiz

, Guerra-Vales

, Trincado

, Fernández

, Medrano

, Villarejo

, Benito-León

, Bermejo-Pareja

(2013) The ability of self-rated health to predict mortality among community-dwelling elderly individuals differs according to the specific cause of death: Data from the NEDICES cohort. Gerontology 59, 368–377.

80.

Tuokko

, Frerichs

, Graham

, Rockwood

, Kristjansson

, Fisk

, Bergman

, Kozma

, McDowell

(2003) Five-year follow-up of cognitive impairment with no dementia. Arch Neurol 60, 577–582.

81.

Teng

, Tassniyom

, Lu

(2012) Reduced quality-of-life ratings in mild cognitive impairment: Analyses of subject and informant responses. Am J Geriatr Psychiatry 20, 1016–1025.

82.

Artero

, Petersen

, Touchon

, Ritchie

(2006) Revised criteria for mild cognitive impairment: Validation within a longitudinal population study. Dement Geriatr Cogn Disord 22, 465–470.

83.

Gomar

, Conejero-Goldberg

, Davies

, Goldberg

, Alzheimer’s Disease Neuroimaging Initiative (2014) Extension and refinement of the predictive value of different classes of markers in ADNI: Four-year follow-up data. Alzheimers Dement 10, 704–712.

84.

Bozoki

, Giordani

, Heidebrink

, Berent

, Foster

(2001) Mild cognitive impairments predict dementia in nondemented elderly patients with memory loss. Arch Neurol 58, 411–416.

85.

Tabert

, Manly

, Liu

, Pelton

, Rosenblum

, Jacobs

, Zamora

, Goodkind

, Bell

, Stern

, Devanand

(2006) Neuropsychological prediction of conversion to Alzheimer disease in patients with mild cognitive impairment. Arch Gen Psychiatry 63, 916–924.

86.

Baars

MAE

, van Boxtel

MPJ

, Dijkstra

, Visser

, van den Akker

, Verhey

FRJ

, Jolles

(2009) Predictive value of mild cognitive impairment for dementia. The influence of case definition and age. Dement Geriatr Cogn Disord 27, 173–181.

87.

Forlenza

, Diniz

, Nunes

, Memória

, Yassuda

, Gattaz

(2009) Diagnostic transitions in mild cognitive impairment subtypes. Int Psychogeriatr 21, 1088–1095.

88.

Klekociuk

, Summers

(2014) Exploring the validity of mild cognitive impairment (MCI) subtypes: Multiple-domain amnestic MCI is the only identifiable subtype at longitudinal follow-up. J Clin Exp Neuropsychol 36, 290–301.

89.

Facal

, Guárdia-Olmos

, Juncos-Rabadán

(2015) Diagnostic transitions in mild cognitive impairment by use of simple Markov models. Int J Geriatr Psychiatry 30, 669–676.

90.

Brambati

, Belleville

, Kergoat

M-J

, Chayer

, Gauthier

, Joubert

(2009) Single- and multiple-domain amnestic mild cognitive impairment: Two sides of the same coin?. Dement Geriatr Cogn Disord 28, 541–549.

91.

Peres

, Chrysostome

, Fabrigoule

, Orgogozo

, Dartigues

, Barberger-Gateau

(2006) Restriction in complex activities of daily living in MCI: Impact on outcome. Neurology 67, 461–466.

92.

Ganguli

, Lee

, Snitz

, Hughes

, McDade

, Chang

(2014) How well do MCI criteria predict progression to severe cognitive impairment and dementia?. Alzheimer Dis Assoc Disord 28, 113–121.

93.

Vega

, Benito-León

, Bermejo-Pareja

, Medrano

, Vega-Valderrama

, Rodríguez

, Louis

(2010) Several factors influenced attrition in a population-based elderly cohort: Neurological disorders in Central Spain Study. J Clin Epidemiol 63, 215–222.

94.

Hughes

, Snitz

, Ganguli

(2011) Should mild cognitive impairment be subtyped?. Curr Opin Psychiatry 24, 237–242.