characterizing mild cognitive disorders in the young-old over 8 years: prevalence, estimated...
TRANSCRIPT
Alzheimer’s & Dementia 9 (2013) 640–648
Characterizing mild cognitive disorders in the young-old over 8 years:Prevalence, estimated incidence, stability of diagnosis,
and impact on IADLsKaarin J. Ansteya,*, Nicolas Cherbuina, Ranmalee Eramudugollaa, Kerry Sargent-Coxa,
Simon Eastealb, Rajeev Kumarc, Perminder Sachdevd
aCentre for Research on Ageing, Health and Wellbeing, The Australian National University, Canberra, AustraliabJohn Curtin School of Medical Research, Sydney, Australia
cMedical School, The Australian National University, Canberra, AustraliadSchool of Psychiatry, University of New South Wales, Sydney, Australia
Abstract Objectives: Few studies report incidence of mild cognitive impairment (MCI) and other mild cog-
Prof Anstey, Dr R
performed the statistic
Professor Anstey:
ing of manuscript, sup
design and critical rev
tent. Dr Sargent-Cox a
ical revision of the
Easteal participated i
in data acquisition and
tellectual content. Pro
and critical revision o
*Corresponding au
E-mail address: ka
1552-5260/$ - see fro
http://dx.doi.org/10.10
nitive disorders (MCD) in cohorts in their 60s, at an age when diagnoses are less stable. The authors’goal was to estimate the incidence and prevalence of MCI and MCD, characterize subgroups withstable vs nonstable diagnoses, and evaluate the impact of diagnosis on daily life in a young-old cohort.Methods: A community-based cohort age 60 to 64 years in 1999 (n 5 2551) was monitored for 8years and assessed every 4 years. A two-stage sampling design was used to identify MCI andMCD through a neuropsychological and neurological assessment. A panel of physicians blind toprevious diagnoses reviewed each case using published criteria.Results: The prevalence of MCDs in the cohort aged 68 to 72 years at the last follow-up was approx-imately 10%. An estimated 141 subjects (7.7%) progressed to MCI and 183 subjects (10.0%) pro-gressed to MCD between years 4 and 8. Only eight participants received a dementia diagnosis atany wave, five of whom progressed from MCDs. More than 45% of diagnoses were unstable duringthe 8 years of follow-up. Stable diagnoses were associated with lower Mini-Mental State Examina-tion scores, history of neurological disorder, higher cardiovascular risk, and depression at baseline.MCDs were associated with impairments in instrumental activities of daily living and higher ratesof reporting memory problems prior to diagnosis.Conclusions: MCDs in individuals in their 60s occur in at least 10% of the population and are likelyto be heterogeneous in terms of their etiology and long-term prognosis, but may cause a significantimpact in everyday life.� 2013 The Alzheimer’s Association. All rights reserved.
Keywords: MCI; Incidence study; Cohort study; Dementia
anmalee Eramudugolla, and Dr Kerry Sargent-Cox
al analysis.
study concept, design, analysis, interpretation, draft-
ervision of study. Dr Cherbuin participated in study
ision of the manuscript for important intellectual con-
nd Dr Eramudugolla participated in analysis and crit-
manuscript for important intellectual content. Prof
n study concept and design. Dr Kumar participated
critical revision of the manuscript for important in-
f Sachdev participated in study concept, study design,
f the manuscript for important intellectual content.
thor. Tel.: 1612 61253858; Fax: 1612 61250733.
nt matter � 2013 The Alzheimer’s Association. All rights r
16/j.jalz.2012.11.013
1. Introduction
Most of the research on mild cognitive impairment(MCI) and related definitions of preclinical dementiasyndromes has been conducted in adults age 70 yearsor older [1]. There are few studies available from whichto ascertain the incidence, prevalence, and course of MCIin individuals in their 60s, the stability of diagnoses atthis age, and the impact cognitive impairment has on em-ployment and everyday life. Characterizing MCI clini-cally in its earliest stages is required to providea complete picture of the epidemiology of neurocognitive
eserved.
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648 641
disorders and may assist in identifying their etiology andin guiding treatment decisions.
Conversion rates of MCI to Alzheimer’s disease (AD)and other dementias are highly age dependent and variable[2]. Unless patients who ultimately develop dementia areassessed prior to diagnosis of a mild cognitive disorder(MCD), the true duration of the MCI stage at the popula-tion level is not estimable. Our aim is to characterize cog-nitive impairment at an age when many older adultsremain in the workforce and are actively engaged in theircommunity. Following results from our previous twowaves [3,4], we estimate the incidence of MCI [5] anda broader category of MCD [3] in a young-old cohort intheir 60s and evaluate factors associated with the stabilityof diagnoses in this age range. We also evaluate the asso-ciation between MCDs and perceived memory difficulties,reports of attending a general practitioner for memory dif-ficulties, and problems with instrumental activities of dailyliving (IADLs).
2. Methods
2.1. Study participants
Participants were sampled randomly from the electoralrolls for Canberra, ACT, and Queanbeyan, NSW, Australia,as part of the Personality and Total Health (PATH) ThroughLife Project [6]. They were asked to complete a question-naire under the supervision of a professional interviewer.Cognitive and physical tests were also carried out anda cheek swab was taken from which DNA could be ex-tracted. Written informed consent was obtained from all par-ticipants prior to involvement in the study, and the AustralianNational University Human Research Ethics Committee ap-proved the study. The oldest cohort is the focus of this studyand was 60 to 64 years at wave 1; interviews were conductedin 2001 to 2002 (n 5 2551), 2005 to 2006 (n 5 2222), and2009 to 2010 (n 5 1973). The design of the PATH ThroughLife Project and the methodology for clinical diagnoses havebeen described elsewhere [3,4,6].
2.2. Screening
At each wave, the same predetermined cutoff on a cogni-tive screening battery was used to screen participants intoa substudy on MCDs and dementia. Participants from thefull cohort were selected for clinical assessment if theyhad any of the following: (i) a Mini-Mental State Examina-tion (MMSE) [7] score � 25, (ii) a score below the fifthpercentile score from wave 1 on immediate or delayed recallof the California Verbal Learning Test [8] (immediate ordelayed score of ,4 and ,2, respectively), or (iii) a scorebelow the fifth percentile on two or more of the followingtests: Symbol-Digit Modalities Test [9] (,33) or PurduePegboard with both hands [10] (wave 1, ,8; wave 2, ,7),or simple reaction time [11] (third set of 20 trials; wave 1,.310 ms; wave 2, .378 ms).
2.3. Clinical assessment
The clinical assessment involved a Structured Clinical As-sessment for Dementia (available from us) by one of two phy-sicians, a neuropsychological assessment, and the ClinicalDementia Rating Scale [12]. Information was also gatheredon medical history related to cognitive function, duration ofsymptoms, medical history from medical practitioners andfamily, current treatment, and psychiatric history. Depressionwas assessed using the self-administered Patient HealthQuestionnaire from thePrimaryCareEvaluationofMentalDis-orders (PRIME-MD) [13]. Informant interviews were conduct-ed when possible. Participants receiving any clinical diagnosiswere referred to their family doctor for laboratory investiga-tions. The research protocol included magnetic resonance im-aging for all consenting participants. The neuropsychologicalassessment included frontal executive function (Trails A andB[14],VerbalFluency [15], andClockDrawing [16]), language(Boston Naming Short Form [17]), constructional praxis fromtheConsortium to Establish a Registry for Alzheimer’s Diseasebattery [18], memory (ReyAuditory Verbal Learning Test withverbal recall and recognition [19]) recall of constructionalpraxis for nonverbal memory, and agnosia [20].
Clinicians used clinical checklists, data from the neuro-psychological assessment, neuropsychiatric history, andmedical history to formulate consensus diagnoses. Criteriafor the following diagnoses were applied: MCI [5,21],age-associated memory impairment [22], age-associatedcognitive decline [23], mild neurocognitive disorder [24],impairment on the Clinical Dementia Rating Scale [12],and other cognitive disorder [24]. Diagnostic and StatisticalManual of Mental Disorders, 4th edition, criteria were usedto assess dementia and delirium [24]. The Petersen criteriafor MCI were used at waves 1 and 2 [21], whereas the Win-blad criteria [5] were used at wave 3. For all other categories,the same criteria were used for all three waves and have beenpublished by our group elsewhere [4]. A description of crite-ria is included in the online supplementary material. Impor-tant for this study, clinicians were blind to the presence orlack of diagnosis obtained at waves 1 and 2.
2.3.1. Assessment of self-reported impact of memory onfunction
Atwaves 1 and 2 participants were asked: Do you feel youcan remember things as well as you used to? That is, is yourmemory the same as itwas earlier in life? If they answered no,then theywere asked: Does this memory problem interfere inanyway with your day-to-day life? And have you seen a doc-tor about your memory? IADLs were assessed at wave 3 withitems adapted from the U.S. Health and Retirement Study[25] and included difficulty reading a map, preparing meals,shopping for groceries, and making telephone calls.
2.3.2. Health measuresDuring the interview, data were collected on history of
neurological disorders and thyroid conditions. Information
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648642
on medication for hypercholesterolemia, smoking status,and diabetes mellitus and its treatment were obtained viaself-report, and blood pressure was measured. These vari-ableswere combinedwith age and sex to compute a Framing-ham cardiovascular risk score for each participant [26].
Table 1
Wave 3 prevalence of mild cognitive disorder
Diagnostic category Prevalence wave 3, % 95% CI
MCD 9.97 (4.31–15.63)
MCI 7.68 (3.50–13.33)
MND 4.06 (0.83–7.91)
AACD 5.63 (1.82–10.26)
CDR � 0.5 7.91 (3.09–12.73)
Abbreviations: CI, confidence interval; MCD, mild cognitive disorder;
MCI, mild cognitive impairment; MND, mild neurocognitive disorder;
AACD, age-associated cognitive decline; CDR, Clinical Dementia Rating.
NOTE. A reliable prevalence estimate of age-associated memory impair-
ment (AAMI) and other cognitive disorder (OCD) according to the Diag-
nostic and Statistical Manual of Mental Disorders, 4th edition, could not
be calculated. MCD comprises any person meeting criteria for MCI,
MND, AACD, AAMI, CDR � 0.5, or OCD.
3. Data analysis
Group comparisons were made with analysis of variancefor continuous variables or c2 tests for frequencies. To iden-tify factors associated with diagnostic instability, we evalu-ated whether a range of factors previously associated withcognitive fluctuation and cognitive decline in the literaturediffered among groups. These included both depression atbaseline and variability in depression over time [27], base-line anticholinergic medication and variability in anticholin-ergic burden [28,29], history of thyroid disorder [30,31],history of neurological disorder, and baseline cardio-vascular risk profile [32,33].
To establish predicted prevalence and incidence rates forthe cohort, predictive regression models were built based onthe relationship between the screening measures and theclinical data for the subsample for which diagnostic datawere available using methods described previously [3]. Itwas necessary to compute prevalence among participantswho screened positive but who did not undertake the clinicalassessment, and to estimate the false negatives in the samplethat screened negative. Using the clinical diagnoses as thegold standard, logistic regressions with age, sex, and screen-ing measures as predictor variables were built. For each di-agnostic criterion, a predictive score defined as theprobability of positive diagnosis was derived and a cutoffpoint was chosen so that the number of predicted diagnoseswas the same as the number of observed diagnoses under thecriterion in the subsample. This cutoff point was then ap-plied to the predictive score of those who screened positivebut did not undertake the clinical assessment and thosewho screened negative. The final prevalence estimate wasa sum of those receiving clinical diagnoses, those estimatedto receive a diagnosis among the group that screened positivebut did not receive a diagnosis, and those estimated to havebeen falsely screened as negative. To estimate the influenceof attrition on the overall incidence rates data were imputed(using the expectation maximization algorithm) for 269 par-ticipants not monitored up from wave 2 to wave 3. Imputa-tion was estimated using wave 1 and wave 2 cognitiondata, gender, and age.
The impact of attrition on estimates was evaluated by ap-plying the predictive regression equations to the subgroups’most recent cognitive assessment data (wave 2) to estimateconservatively the proportion lost to follow-up with likelycognitive impairment.
Confidence intervals and standard errors for the preva-lence estimate were estimated using bootstrap techniques.The logistic regression linking the diagnostic data and thescreening data was fitted to 1000 bootstrap samples, and
the population prevalence estimate was then calculated asthe average of prevalence estimates across bootstrap sam-ples. The 95% confidence interval was calculated by takingthe 2.5th and 97.5th percentile of the bootstrap prevalenceestimates. Analyses were conducted in IBM SPSS19(Armok, NY, USA) and STATA version 9.0 (CollegeStation, TX, USA).
4. Results
4.1. Sample size and demographics
At wave 3, the sample had a mean age of 70.60 years(standard deviation [SD], 1.50 years; range, 68–72 years),including 1020 males (51.7%). The sample had a mean of13.98 years (SD, 2.7 years) of education.
4.2. Prevalence and incidence rates derived fromstatistical modeling
Incidence rates were estimated from the 2047 participantswho were free from clinical diagnosis at wave 2. Of these in-dividuals, 1838 were available for assessment at wave 3. Us-ing the statistical estimates based on the screening measuresapplied to the entire sample, an estimated 141 (7.7%) partic-ipants converted to MCI and 183 (10.0%) converted to MCDduring 4 years, indicating an incidence rate of 19.18 cases ofMCI/1000 person-years and 24.89 cases of MCD/1000person-years. Rates for conversion to MCI and MCD werehigher for males (MCI, 9.3%; MCD, 11.7%) than females(MCI, 5.9%; MCD, 8.2%; P, .01). Of participants with sta-tistically derived diagnoses who screened negative andtherefore did not undergo a clinical assessment, 73 receiveda diagnosis ofMCI and 95 received a diagnosis ofMCD. Thestatistically estimated prevalence rates of MCDs in the co-hort at wave 3 (not including imputed data from those lostat wave 2) are shown in Table 1.
In the subsample lost to attrition (n 5 269), it was esti-mated that 33 (12.3%) progressed to MCI and 28 (10.4%)to MCD. Combining these predicted conversions with thefull wave 3 sample, the incidence rate was calculated to
Table 2
Classification of MCI/MCD participants across three waves of the study
Diagnosis
Classification at each wave Total, n (%)*
Wave 1 Wave 2 Wave 3 MCD MCI
Unstable 76 (51) 40 (45.5)
Normal MCD/MCI Normal 35 (23.5) 15 (17.0)
MCD/MCI Normal Normal 17 (11.4) 16 (18.2)
MCD/MCI Normal MCD/MCI 4 (2.7) 3 (3.4)
MCD/MCI MCD/MCI Normal 20 (13.4) 6 (6.8)
Stable
diagnosisy42 (28.2) 15 (17.0)
MCD/MCI MCD/MCI MCD/MCI 17 (11.4) 2 (2.3)
Normal MCD/MCI MCD/MCI 25 (16.8) 13 (14.8)
New diagnosis 31 (20.8) 33 (37.5)
Normal Normal MCD/MCI 30 (20.7) 33 (37.5)
Normal Missing MCD/MCI 1 (0.7)
Abbreviations: MCD, mild cognitive disorder; MCI, mild cognitive im-
pairment.
NOTE. Totals for diagnosis groups marked in bold.
*Percentages based on prevalent cases during the 8-year study period
(MCD: n 5 149; MCI: n 5 93).yStability of classification is defined as consecutive diagnoses across mul-
tiple assessments [34] without reversion to normalcy [35,36].
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648 643
increase to 20.65 MCI cases/1000 person-years (an increaseof 1.47 cases) and 25.04 cases/1000 person-years for MCD(an increase of 0.15 cases).
4.2.1. Classification of stable and unstable MCD and MCIAll participants with an MCI/MCD diagnosis at any point
in the study (MCI, n5 149; MCD, n5 93) were classified asstable, unstable, or new cases (Table 2 [34–36]). Participantswith a first diagnosis of any MCD/MCI at wave 3 wereclassified as new cases (any MCD, n 5 31; MCI, n 5 33).Of the participants with previous diagnoses of any MCD/MCI, stable cases were defined as those with consecutive
Table 3
Demographic and baseline clinical characteristics for the cohort at wave 3 (n 5 1
Diagnostic
group n
Age at wave 3,
Years
Gender,
male, n (%)
No. of years of
education
Undiagnosed 1816 70.61 (1.49) 925 (50.9) 14.11 (2.65)
Unstable MCD* 76 70.79 (1.46) 45 (59.2) 12.45 (2.72)x
Stable MCD* 42 69.95 (1.46) 26 (61.9) 12.29 (2.77)x
New MCD cases* 31 70.77 (1.67) 19 (61.3) 13.22 (3.08)
Dementia 8 71.00 (1.51) 5 (62.5) 12.75 (2.90)
c2 or F statisticy 2.54z 5.41 12.49x
Unstable MCI* 40 70.60 (1.52) 25 (62.5) 12.02 (2.81)x
Stable MCI* 15 69.87 (1.41) 10 (66.7) 11.73 (2.68)x
New MCI cases* 33 70.73 (1.57) 20 (60.6) 13.02 (2.99)
c2 or F statisticy 1.11 4.87 9.92x
Abbreviations: NESB, non-English-speaking background; APOE, apolipoprotei
tistical Manual of Mental Disorders, 4th edition; MCD, mild cognitive disorder;
NOTE. The depression score was on the Patient Health Questionnaire Depressi
*Difference across all groups (omnibus test): undiagnosed, stable/unstable/newyBonferroni adjusted comparisons between MCI/MCD group and undiagnosedzP , .05xP , .01{P , .06
diagnoses across multiple assessments [34] without rever-sion to normalcy [35,36] (any MCD, n 5 42; MCI, n 515). The remaining participants were classified unstableand comprised those with a prior diagnosis who hadreverted to normalcy at wave 3, and those with a currentdiagnosis but nonconsecutive prior diagnoses (any MCD,n 5 76; MCI, n 5 40). Participants who progressed todementia were categorized separately (n 5 8).
4.3. Characteristics of normal and clinical groups
Baseline demographic and cognitive characteristics ofparticipants who remained without diagnosis and those diag-nosed with MCI, any MCD, and dementia are shown inTable 3. During the entire 8 years of follow-up, 187 partici-pants received a diagnosis of any MCD (149 of whom wereavailable at wave 3) and eight participants progressed to de-mentia. Of the participants diagnosed with an MCD duringthe study, approximately half demonstrated diagnostic insta-bility, reverting to normal cognitive status during wave 3(Fig. 1). Table 3 presents the demographic and baseline clin-ical characteristics for all participants available for study atwave 3. The diagnostic groups differed on the basis of levelof education, proportion of participants from a non-English-speaking background, baseline levels of employment, de-pression, and MMSE scores.
Relative to undiagnosed participants, thosewith a diagno-sis of either stable or unstable MCI/MCD had significantlylower education, lower baseline MMSE scores, and weremore likely to be of a non-English-speaking background(6.2% vs 16.0-31.0%). In addition, a stable diagnosis ofMCI/MCD was associated with lower rates of baseline em-ployment (13.3-16.7% vs 43.3%) and more severe depres-sion at baseline (3.55 [SD, 4.39] vs 2.25 [SD, 2.92]). The
973).
NESB
APOE e4
genotype
Baseline mood and cognitive status
Employed at
wave 1 MMSE score
DSM-IV
depression
111 (6.2) 463 (26.9) 787 (43.3) 29.36 (1.04) 2.25 (2.92)
12 (16.0)x 18 (24.0) 30 (39.5) 28.21 (1.95)x 2.63 (3.32)
13 (31.0)x 7 (16.7) 7 (16.7)x 27.12 (2.92)x 3.55 (4.39){
5 (16.1) 13 (43.3)z 8 (25.8) 28.06 (1.41)z 3.03 (3.68)
1 (12.5) 4 (50.0) 4 (50.0) 28.63 (1.06) 2.38 (3.58)
50.00x 8.83{ 16.06x 63.18x 2.63z
13 (33.3) 9 (22.5) 20 (50.0) 27.50 (2.29)x 2.90 (3.39)
7 (46.7) 2 (13.3) 2 (13.3) 26.76 (3.16)x 4.00 (5.25)
4 (12.1) 10 (31.2) 9 (27.3) 27.84 (1.71)x 2.69 (3.91)
77.68x 4.28 9.14z 58.24x 1.78
n E; MMSE, Mini-Mental State Examination; DSM-IV, Diagnostic and Sta-
MCI, mild cognitive impairment.
on Scale.
MCD/MCI, dementia.
group.
Fig. 1. (A, B) Flow diagram depicting progression of participants with mild cognitive disorders through the PATHThrough Life Project. Participant progression
across waves as a function of mild cognitive impairment (MCI) diagnosis (A) and mild cognitive disorder (MCD) diagnosis (B). *Participants reverting to nor-
mal state in wave 3 categorized as unstable diagnosis.
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648644
proportion of those with one or two apolipoprotein E(APOE) e4 alleles did not differ significantly across groups,and was 26.9% in the undiagnosed group, ranged from13.3% to 43.3% in the MCI/MCD categories, and was50% for dementia.
At wave 3, clinical interview and neuropsychologicaltesting of MCI/MCD participants established 26 indi-viduals as amnestic single-domain MCI, 21 as amnesticmultidomain MCI, and four individuals as nonamnesticmultidomain MCI. None were diagnosed with nonamnesticsingle-domain MCI. Of the 51 participants diagnosed withthe wave 3 MCI criteria [5], 47 also met the wave 1 MCIcriteria [21] at wave 3.
4.4. Clinical and demographic factors associated withdiagnostic stability over 8 years
Comparisons of stable and unstable MCD and MCIgroups are shown in Table 4. Diagnostic stability for MCDwas associated with lower baseline MMSE scores andincreased frequency of neurological disorders. Diagnosticstability for MCI was associated with a higher Framinghamcardiovascular 10-year risk profile calculated at baseline andhistory of neurological disorder. None of the other potentialexplanatory influences on cognitive fluctuation differenti-ated diagnostic stability. Participants categorized as stablewere also significantly more likely to report memory
Table 4
Factors differentiating the diagnostic stability of MCIs
Variable
MCD
P value
MCI
P valueUnstable Stable New cases Unstable Stable New cases
Years of education 12.45 (2.72) 12.29 (2.77) 13.22 (3.08) NS 12.02 (2.81) 11.73 (2.68) 13.02 (2.99) NS
MMSE score at baseline 28.21 (1.95) 27.12 (2.92)z 28.06 (1.41) ,.05 28.21 (1.95) 27.12 (2.92) 28.06 (1.41) NS
Depression at baseline 2.63 (3.32) 3.55 (4.39) 3.03 (3.68) NS 2.90 (3.39) 4.00 (5.25) 2.69 (3.91) NS
Variability in depression* 1.84 (1.83) 2.38 (2.13) 2.13 (1.83) NS 1.53 (1.43) 1.39 (1.16) 1.27 (1.24) NS
Variability in anticholinergic burden* 0.30 (0.49) 0.37 (0.51) 0.34 (0.48) NS 0.32 (0.49) 0.34 (0.43) 0.34 (0.54) NS
Framingham relative risk at baseline 0.69 (0.29) 0.71 (0.32) 0.73 (0.22) NS 0.61 (0.23) 0.82 (0.40)z 0.71 (0.27) ,.05
Non-English speaking background, n (%) 12 (16.0) 13 (31.0) 5 (16.1) NS 13 (33.3) 7 (46.7) 4 (12.1)z ,.05
APOE e4 carriage, n (%) 18 (24.0) 7 (16.7) 13 (43.3)z ,.05 9 (22.5) 2 (13.3) 10 (31.2) NS
White, n (%) 71 (93.4) 37 (88.1) 30 (96.8) NS 34 (85.0) 12 (80.0) 32 (97.0) NS
History of neurological disorder, n (%)z 17 (22.4) 14 (33.3) 12 (38.7) NS 7 (17.5) 2 (13.3) 13 (39.4)z ,.05
History of thyroid disorder, n (%) 11 (14.5) 8 (19.0) 6 (19.4) NS 5 (12.5) 1 (6.7) 7 (21.2) NS
Anticholinergic medications at baseline, n (%) 13 (17.1) 5 (11.9) 4 (12.9) NS 6 (15.0) 1 (6.7) 3 (9.1) NS
Subjective memory complaints, yes, n (%) 17 (22.4) 14 (33.3) 7 (22.6) NS 8 (20.0) 3 (20.0) 10 (30.3) NS
Abbreviations: MCD, mild cognitive disorder; MCI, mild cognitive impairment; NS, not significant; MMSE, Mini-Mental State Examination; APOE, apo-
lipoprotein E.
* Mean absolute deviation of individuals’ scores across waves.
y Any history of stroke, epilepsy, head injury, brain tumor, brain infection, or parkinsonism.zP , .05 for pairwise comparison relative to unstable MCD/MCI group.
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648 645
complaints at baseline relative to those with unstable MCDdiagnosis.
4.5. Impact of cognitive impairment on memorycomplaints and IADLs
Impact of cognitive impairment was evaluated by exam-ining medical consultations related to memory, memorycomplaints, and problems with IADLs (Table 5). The stableMCD group had higher rates of reporting they had consulteda doctor about their memory in wave 2, and higher rates ofreporting memory interfering with their daily activities in
Table 5
Self-reported memory and impact on IADLs by diagnostic stability of MCD and
Undiagnosed
MCD
Unstable Stable New diagnoses
Reporting yes, % Reporting yes, %
Memory worse than used to be
Wave 1 24.1 26.3 14.3 19.4
Wave 2 18.4 13.2 16.7 13.3
Memory interferes with daily life
Wave 1 8.8 14.3 27.8* 28*
Wave 2 9.6 16.7 28.6* 34.6*
Consulted doctor about memory
Wave 1 2.3 23.1 16.7 37.5
Wave 2 2.3 21.4 53.8* 40
IADLs, wave 3 Reporting yes, n (%) Reporting yes, n (%)
Any IADL problems 254 (14.0) 11 (14.5) 14 (33.3)*
Problems in map reading 149 (8.2) 7 (9.2) 7 (16.2)
Problems preparing meals 36 (2.0) 3 (3.9) 3 (7.1)
Problems grocery shopping 64 (3.5) 1 (1.3) 3 (1.5)
Problems making phone calls 25 (1.4) 2 (2.6) 2 (4.8)
Problems taking medications 21 (1.2) 1 (1.3) 3 (7.1)*
Abbreviations: MCD, mild cognitive disorder; MCI, mild cognitive impairmen
*P , .05, pairwise comparisons with undiagnosed group.
both waves 1 and 2, relative to the undiagnosed group aswell as to the unstable MCD group.
Participants with a stable MCI or MCD diagnosis at wave 3reportedmore difficulties readingmaps, shopping,making tele-phone calls, and taking medications. All groups reported lowlevels of problemswithmeal preparation and grocery shopping.
5. Discussion
The current study adds to the two previous reports of MCIand any MCD in this sample [3,4] to provide results
MCI
P value
MCI
P value
Unstable Stable New diagnoses
Reporting yes, %
NS 10 13.3 21.2 NS
NS 5 20 21.9 NS
,.01 17.5 33.3* 21.2* ,.001
,.001 20* 26.7* 24.2* ,.01
NS 5 13.3* 9.1* ,.01
,.01 2.5 26.7* 15.2* ,.001
P value Reporting yes, n (%) P value
12 (38.7)* ,.001 7 (17.5) 4 (26.7) 14 (42.4)* ,.001
8 (25.8)* ,.01 4 (10.0) 2 (13.3) 9 (27.3)* ,.01
0 (0) NS 2 (5.0) 1 (6.7) 2 (6.1) NS
3 (9.7) NS 0 (0) 0 (0) 4 (12.1)* ,.05
2 (6.5)* ,.05 1 (2.5) 1 (6.7) 2 (6.1) NS
1 (3.2) ,.01 2 (5.0) 1 (6.7) 2 (6.1) ,.01
t; NS, not significant; IADLs, instrumental activities of daily living.
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648646
characterizing progression to MCD over 8 years ina relatively young cohort. There are several importantcontributions from this study, including the reporting ofincidence rates of MCD in this age group, evaluation ofstability of diagnoses, and documentation of the impact ofMCD on everyday life.
After 8 years of follow-up, approximately 10% of thissample age 68 to 72 years at wave 3 had MCD after correct-ing statistically for false negatives and attrition. As expectedfrom the younger age sample, incidences of MCI and anyMCD were less than one third of the 63.6/1000 recently re-ported in the Mayo Clinic Study of Aging for a sample age70 to 89 years at baseline [1]. The incidence rates were alsoat the lower end of the range reported in a recent review [37].These rates are likely to underestimate the true level of MCDin the population because the sample had relatively few par-ticipants with a low level of education, which is the strongestrisk factor for MCI [38]. Moreover, there is greater attritionamong participants with cognitive disorders in longitudinalstudies [39], particularly at older ages [40]. Previous studieshave also found that the prevalence of MCI is determined bythe criteria used [41]. As noted by Petersen and colleagues[42] and Ritchie [43], epidemiologic study samples includefar greater heterogeneity than clinical samples.
Approximately half of all diagnoses in this age group areunstable. Only five participants with MCD converted to de-mentia during our 8 years of follow-up, which is close to thelow rate of progression reported in another study that com-menced at age 60 [44], and was lower than the progressionrates reported for studies with older age ranges [42]. Theseresults emphasize the importance of considering the age ofthe sample when evaluating rates of progression from MCIto dementia. Although participants received clinical diagno-ses and reported difficulties in IADLs, MMSE scores werenot low.
The clinical indicators of stability of diagnosis includedhigher levels of depression at baseline and lower MMSEscores. Unlike previous studies, we did not find that theAPOE genotype was associated with MCI [45], and thismay be a result of the heterogeneity in our epidemiologicsample. However, 50% of the eight dementia cases had atleast one e4 allele, and further follow-up of the cohort willdetermine whether higher rates of conversion are foundamong APOE-positive MCD-classified participants.
The characteristics of participants at risk of cognitive im-pairment were consistent with the literature. Our finding ofbaseline depression being higher in participants convertingto MCI is consistent with previous research showing thatlate-life depression is associated with increased risk of ADand dementia [46,47]. Our finding of a higher Framinghamrisk score at baseline in the MCD group is consistent withprevious research from the Whitehall study [32] linkingthis measure with poorer cognitive function and cognitivedecline in men. Similar to the Mayo study [1], we foundthat low education and males were at increased risk. Whatis particularly interesting in relation to the protective effect
of education is that the mean difference in years of educationbetween those with MCI/MCD and cognitively normal sub-jects is that it only differed by 2 years. Moreover, this differ-ence occurred among groups with relatively high levels ofeducation (approximately 12 years vs approximately 14years).
This study also documents the impact of MCDs on ev-eryday function in adults in their 60s. Consistent withother studies in older samples, impairment in some butnot all IADLs was found [48], indicating that these adultsliving in the community already experience disability asso-ciated with their cognitive disorder. The greater level ofimpairment in IADLs in the groups with stable diagnosesis consistent with previous research indicating that impair-ment in IADL predicts faster progression to AD andgreater brain atrophy [49]. Compared with the normallyaging participants in our study, we found that the MCDgroup reported higher rates of memory problems interfer-ing with their everyday lives, lower rates of employment,and higher rates of difficulties with IADLs [49]. Criteriafor MCI include minimal impairment on complex instru-mental functions, and this was established during diagnosisusing the Informant Questionnaire for Cognitive Decline inthe Elderly and an informant interview. Our findings indi-cate that those with stable MCD are aware of and morelikely to endorse difficulties with complex IADLs relativeto the healthy population. The items used to assess IADLsin this study were drawn from the U.S. Health and Retire-ment Study [25] and may have been more sensitive to im-pairment in higher level IADLs than items designed forclinical use when impairment is greater. Hence, even inthis relatively highly educated cohort, a small proportionexperience disability in their 60s as a result of cognitiveimpairment.
Limitations included sample attrition and a primarilywhite and highly educated population. Strengths of thisstudy include the population-based sample of relativelyyoung age, a narrow cohort design that allows for aging ef-fects to be studied distinct from cohort effects, the availabil-ity of 8 years of follow-up data, and the thorough clinicalassessment and consensus diagnosis by clinicians using pub-lished criteria. We also used statistical models with boot-strapping to minimize error in our prevalence estimates,adjusted for false negatives, and evaluated statistically theimpact of sample attrition on incidence.
The detection of MCI in this younger age group suggeststhat a proportion of those diagnosed in the older cohort stud-ies may have been impaired for many years, and hence theduration of the clinically detectable MCI phase may be lon-ger than previously considered. This is consistent with themodel of biomarkers of AD showing disease progression de-cades prior to diagnosis [50]. We conclude that MCDs in in-dividuals in their 60s occur in at least 10% of the populationand are likely to be heterogeneous in terms of their etiologyand long-term prognosis, but may cause a significant impacton everyday life.
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648 647
Acknowledgments
We are grateful to the PATH chief investigators, Chantal Me-slin, Patricia Jacomb, Karen Maxwell, and the PATH inter-viewers. The study was supported by National Health andMedical Research Council of Australia unit grant no.973302, program grant no. 179805, NHMRC project grantno. 157125, and grants from the Australian Rotary HealthResearch Fund and the Australian Brewers Foundation. An-stey and Cherbuin are funded by NHMRC fellowships no.1002560 and 471501.
RESEARCH IN CONTEXT
1. Systematic review: We searched reports from largecohort studies and systematic reviews to identifystudies reporting on the epidemiology of mild cogni-tive disorders in early old age. No study with a largecohort aged early 60s was identified. Nearly all re-search on mild cognitive impairment has been onadults in their 70s and 80s.
2. Interpretation: In their late 60s, about 10% of ourcohort had progressed to a mild cognitive disor-der, with 8% meeting diagnostic criteria formild cognitive impairment. Cognitive impairmentwas associated with difficulties in instrumentalactivities of daily living. Stable diagnoses (about45%) were associated with lower Mini-MentalState Examination scores, history of neurologi-cal disorder, higher cardiovascular risk, and de-pression at baseline.
3. Future directions: Correlation of cognitive impair-ment in late middle age and early old age withbiomarkers of Alzheimer’s disease and other demen-tias will assist in characterizing and differentiatingmore completely the course and causes of dementia,optimizing targeting of interventions.
References
[1] Roberts RO, Geda YE, Knopman DS, Cha RH, Pankratz VS,
Boeve BF, et al. The incidence of MCI differs by subtype and is higher
in men: the Mayo Clinic Study of Aging. Neurology 2012;78:342–51.
[2] Bruscoli M, Lovestone S. Is MCI really just early dementia? A sys-
tematic review of conversion studies. Int Psychogeriatr 2004;
16:129–40.
[3] Anstey KJ, Cherbuin N, Christensen H, Burns R, Reglade-Meslin C,
SalimA, et al. Follow-up of mild cognitive impairment and related dis-
orders over four years in adults in their sixties: the PATH Through Life
Study. Dement Geriatr Cogn Disord 2008;26:226–33.
[4] Kumar R, Dear KB, Christensen H, Ilschner S, Jorm AF, Meslin C,
et al. Prevalence of mild cognitive impairment in 60- to 64-year-old
community-dwelling individuals: the Personality and Total Health
through Life 601 Study. Dement Geriatr Cogn Disord 2005;19:67–74.
[5] Winblad B, Palmer K, Kivipelto M, Jelic V, Fratiglioni L,
Wahlund LO, et al. Mild cognitive impairment: beyond controversies,
towards a consensus: report of the International Working Group on
Mild Cognitive Impairment. J Intern Med 2004;256:240–6.
[6] Anstey KJ, Christensen H, Butterworth P, Easteal S, Mackinnon A,
Jacomb T, et al. Cohort profile: The PATH Through Life project. Int
J Epidemiol 2012;41:951–60.
[7] Folstein MF, Robins LN, Helzer JE. The Mini-Mental State Examina-
tion. Arch Gen Psychiatry 1983;40:812.
[8] Delis DC, Kramer JH, Kaplan E, Ober BA. California Verbal Learning
Test. San Antonio, TX: Psychological Corporation Harcourt Brace Jo-
vanovich; 1987.
[9] Smith A. Symbol Digit Modalities Test (SDMT)manual. Los Angeles,
CA: Western Psychological Services; 1982.
[10] Tiffin J. Purdue Pegboard: examiners manual. Rosemont, IL: London
House; 1968.
[11] Anstey KJ, Dear K, Christensen H, Jorm AF. Biomarkers, health, life-
style and demographic variables as correlates of reaction time perfor-
mance in early, middle and late adulthood. Q J Exp Psychology 2005;
58A:5–21.
[12] Morris JC. The Clinical Dementia Rating (CDR): current version and
scoring rules. Neurology 1993;43:2412–4.
[13] Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-
report version of PRIME-MD: the PHQ primary care study: primary
care evaluation of mental disorders: patient health questionnaire.
JAMA 1999;282:1737–44.
[14] Reitan RM. Trail Making test results for normal and brain-damaged
children. Percept Motor Skills 1971;33:575–81.
[15] Lezak MD. Neuropsychological assessment. Oxford: Oxford Univer-
sity Press; 1995.
[16] Sunderland T, Hill JL, Mellow AM, Lawlor BA, Gundersheimer J,
Newhouse PA, et al. Clock drawing in Alzheimer’s disease: a novel
measure of dementia severity. J Am Geriatr Soc 1989;37:725–9.
[17] Mack WJ, Freed DM, Williams BW, Henderson VW. Boston Naming
Test: shortened versions for use in Alzheimer’s disease. J Gerontol
1992;47:P154–8.
[18] Morris JC, Heyman A, Mohs RC, Hughes JP, van Belle G,
Fillenbaum G, et al. The Consortium to Establish a Registry for
Alzheimer’s Disease (CERAD). Part I: clinical and neuropsycholog-
ical assessment of Alzheimer’s disease. Neurology 1989;
39:1159–65.
[19] Rey A. Psychological examination of traumatic encephalopathy. Arch
Psychol 1941;28:286–340.
[20] Kertesz A. Western aphasia battery. San Antonio, TX: Psychological
Corporation; 1983.
[21] Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG,
Kokmen E. Mild cognitive impairment: clinical characterization and
outcome. Arch Neurol 1999;56:303–8.
[22] Crook T, Bahar H, Sudilovsky A. Age-associated memory impairment:
diagnostic criteria and treatment strategies. Int J Neurol 1987;
21–22:73–82.
[23] Kral VA. Senescent forgetfulness: benign andmalignant. CanMed As-
soc J 1962;86:257–60.
[24] American Psychiatric Association. Diagnostic and statistical manual
of mental disorders (DSM-IV). Washington, DC: American Psychiat-
ric Association; 1994.
[25] Freedman VA, Martin LG, Schoeni RF, Cornman JC. Declines in late-
life disability: the role of early- and mid-life factors. Soc Sci Med
2008;66:1588–602.
[26] Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular
disease risk profiles. Am Heart J 1991;121:293–8.
[27] Mackin RS, Insel P, Aisen PS, Geda YE, Weiner MW. Longitudinal
stability of subsyndromal symptoms of depression in individuals
with mild cognitive impairment: relationship to conversion to demen-
tia after 3 years. Int J Geriatr Psychiatry 2012;27:355–63.
[28] Jessen F, Kaduszkiewicz H, Daerr M, Bickel H, Pentzek M, Riedel-
Heller S, et al. Anticholinergic drug use and risk for dementia: target
for dementia prevention. Eur Arch Psychiatry Clin Neurosci 2010;
260:S111–5.
K.J. Anstey et al. / Alzheimer’s & Dementia 9 (2013) 640–648648
[29] Low LF, Anstey KJ, Sachdev P. Use of medications with anticholiner-
gic properties and cognitive function in a young-old community sam-
ple. Int J Geriatr Psychiatry 2009;24:578–84.
[30] Harper PC, Roe CM. Thyroid medication use and subsequent develop-
ment of dementia of the Alzheimer type. J Geriatr Psychiatry Neurol
2010;23:63–9.
[31] Wahlin A, Bunce D, Wahlin TB. Longitudinal evidence of the impact
of normal thyroid stimulating hormone variations on cognitive func-
tioning in very old age. Psychoneuroendocrinology 2005;30:625–37.
[32] Kaffashian S, Dugravot A, Nabi H, Batty GD, Brunner E, KivimakiM,
et al. Predictive utility of the Framingham general cardiovascular dis-
ease risk profile for cognitive function: evidence from the Whitehall II
Study. Eur Heart J 2011;32:2326–32.
[33] Yaffe K, Blackwell T, Kanaya AM, Davidowitz N, Barrett-Connor E,
Krueger K. Diabetes, impaired fasting glucose, and development of
cognitive impairment in older women. Neurology 2004;63:658–63.
[34] Vandermorris S, Hultsch DF, HunterMA,MacDonald SWS, Strauss E.
Including persistency of impairment in mild cognitive impairment
classification enhances prediction of 5-year decline. Arch Clin Neuro-
psychol 2011;26:26–37.
[35] Manly JJ, Tang M-X, Schupf N, Stern Y, Vonsattel J-PG, Mayeux R.
Frequency and course of mild cognitive impairment in a multiethnic
community. Ann Neurol 2008;63:494–506.
[36] Ritchie K, Artero S, Touchon J. Classification criteria for mild cogni-
tive impairment: a population-based validation study. Neurology
2001;56:37–42.
[37] Ward A, Arrighi HM, Michels S, Cedarbaum JM. Mild cognitive im-
pairment: disparity of incidence and prevalence estimates. Alzheimers
Dement 2012;8:14–21.
[38] Barnes DE, Yaffe K. The projected effect of risk factor reduction on
Alzheimer’s disease prevalence. Lancet Neurol 2011;10:819–28.
[39] Anstey KJ, Luszcz MA. Selective non-response to clinical assessment
in the longitudinal study of aging: implications for estimating popula-
tion levels of cognitive function and dementia. Int J Geriatr Psychiatry
2002;17:704–9.
[40] Kukull WA, Higdon R, Bowen JD, McCormick WC, Teri L,
SchellenbergGD,etal.Dementia andAlzheimerdisease incidence:apro-
spective cohort study. Arch Neurol 2002;59:1737–46.
[41] Ganguli M, Chang CC, Snitz BE, Saxton JA, Vanderbilt J, Lee CW.
Prevalence of mild cognitive impairment by multiple classifications:
theMonongahela-Youghiogheny Healthy Aging Team (MYHAT) pro-
ject. Am J Geriatr Psychiatry 18:674–83.
[42] Petersen RC, Roberts RO, Knopman DS, et al. Mild cognitive impair-
ment: ten years later. Arch Neurol 2009;66:1447–55.
[43] Ritchie K. Mild cognitive impairment: an epidemiological perspec-
tive. Dialogues Clin Neurosci 2004;6:401–8.
[44] Farias ST, Mungas D, Reed BR, Harvey D, DeCarli C. Progression of
mild cognitive impairment to dementia in clinic- vs community-based
cohorts. Arch Neurol 2009;66:1151–7.
[45] Caselli RJ, Reiman EM, Locke DE, Hutton ML, Hentz JG,
Hoffman-Snyder C, et al. Cognitive domain decline in healthy apo-
lipoprotein E epsilon4 homozygotes before the diagnosis of mild
cognitive impairment. Arch Neurol 2007;64:1306–11.
[46] Li G, Wang LY, Shofer JB, Thompson ML, Peskind ER,
McCormick W, et al. Temporal relationship between depression and
dementia: findings from a large community-based 15-year follow-up
study. Arch Gen Psychiatry 2011;68:970–7.
[47] Ownby RL, Crocco E, AcevedoA, John V, Loewenstein D. Depression
and risk for Alzheimer disease: systematic review, meta-analysis, and
metaregression analysis. Arch Gen Psychiatry 2006;63:530–8.
[48] Binegar DL, Hynan LS, Lacritz LH,WeinerMF, CullumCM. Can a di-
rect IADL measure detect deficits in persons with MCI? Curr Alz-
heimer Res 2009;6:48–51.
[49] Chang YL, Bondi MW, McEvoy LK, Fennema-Notestine C,
Salmon DP, Galasko D, et al. Global clinical dementia rating of 0.5
in MCI masks variability related to level of function. Neurology
2011;76:652–9.
[50] Jack CR Jr., Knopman DS, Jagust WJ, Shaw LM, Aisen PS,
Weiner MW, et al. Hypothetical model of dynamic biomarkers of
the Alzheimer’s pathological cascade. Lancet Neurol 2010;9:119–28.