Memory complaints and risk of cognitive

impairment after nearly 2 decades among

older women

Allison R. Kaup, PhD ABSTRACT

Jasmine Nettiksimmons, Objectives: To investigate the association between subjective memory complaints (SMCs) and
PhD long-term risk of cognitive impairment in aging because most previous studies have followed in-
Erin S. LeBlanc, MD, dividuals for only a few years.
MPH
Methods: Participants were 1,107 cognitively normal, community-dwelling older women (aged 65
Kristine Yaffe, MD
years and older at baseline) in a prospective study of aging. SMCs were assessed shortly after
baseline and repeatedly over time with the yes/no question, “Do you feel you have more problems
Correspondence to
with memory than most?” Cognitive status 18 years later (normal or impaired with mild cognitive
Dr. Kaup: impairment or dementia) was determined by an expert panel. Using logistic regression, we inves-
[email protected]
tigated the association between SMCs over time and risk of cognitive impairment, adjusting for
demographics, baseline cognition, and characteristics that differed between those with and with-
out SMCs.
Results: At baseline, 8.0% of participants (n 5 89) endorsed SMCs. Baseline SMCs were asso-
ciated with increased risk of cognitive impairment 18 years later (adjusted odds ratio [OR] 5 1.7,
95% confidence interval 1.1–2.8). Results were unchanged after excluding participants with
depression. The association between SMCs and cognitive impairment was greatest at the last
SMC assessment time point (18 years before diagnosis: adjusted OR 5 1.7 [1.1–2.9]; 14 years
before diagnosis: adjusted OR 5 1.6 [0.9–2.7]; 10 years before diagnosis: adjusted OR 5 1.9
[1.1–3.1]; 4 years before diagnosis: adjusted OR 5 3.0 [1.8–5.0]).
Conclusions: SMCs are associated with cognitive impairment nearly 2 decades later among older
women. SMCs may be a very early symptom of an insidious neurodegenerative disease process,
such as Alzheimer disease. Neurology® 2015;85:1852–1858

GLOSSARY
AD 5 Alzheimer disease; CI 5 confidence interval; DSM-IV 5 Diagnostic and Statistical Manual of Mental Disorders (Fourth
Edition); GDS 5 Geriatric Depression Scale; MCI 5 mild cognitive impairment; mMMSE 5 modified Mini-Mental State Exam-
ination; OR 5 odds ratio; SMC 5 subjective memory complaint; SOF 5 Study of Osteoporotic Fractures.

Alzheimer disease (AD) and other neurodegenerative disorders manifest with an insidious
course, making it challenging to detect their subtle beginnings in an individual patient.
Indeed, the neuropathologic changes of AD may occur many years before an individual be-
gins to exhibit cognitive impairment.1 Subjective cognitive complaints among cognitively
normal older adults may be an early symptom of AD or other neurodegenerative processes.2
Some studies of cognitively normal older adults have found that individuals with subjective
memory complaints (SMCs) have greater amyloid burden,3,4 while others have found no such
associations5,6 or have argued that SMCs merely reflect symptoms of depression or anxiety.5 A
recent meta-analysis7 of longitudinal studies supports an overall association between SMCs
and the development of mild cognitive impairment (MCI) and dementia, but also revealed
that previous studies have followed individuals for an average of only 4 to 5 years.7 This is a
key limitation in the field given that neurodegenerative processes such as AD can take several
years to decades to unfold.8

From Research Service (A.R.K.), San Francisco VA Medical Center (K.Y.); Departments of Psychiatry (A.R.K., J.N., K.Y.), Neurology (K.Y.), and
Epidemiology and Biostatistics (K.Y.), University of California San Francisco; and Center for Health Research (E.S.L.), Kaiser Permanente
Northwest, Portland, OR.
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

1852 © 2015 American Academy of Neurology

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 Additional research is needed to clarify the Learning Test–II Short Form,12 a measure of learning and
memory; (3) Digit Span–Forward and Backward,13 a measure of
long-term predictive utility of an older adult’s
auditory attention/working memory; (4) Trail Making Test, Part
subjective report of memory problems and the B,14 a measure of executive functioning; and (5) Verbal Asso-
extent to which SMCs may reflect an early ciative Fluency (letter and category),15 a measure of language.
symptom of a neurodegenerative disorder. Year-20 clinical cognitive diagnosis of normal cognition, MCI, or
dementia was determined using a previously described process.16
Our objective was to examine whether SMCs First, the following screening criteria were applied as indicators of
are associated with long-term risk of cognitive possible cognitive impairment: (1) mMMSE score ,8817; (2)
impairment (MCI or dementia) 18 years later California Verbal Learning Test delayed recall score ,412; (3)
Informant Questionnaire on Cognitive Decline in the Elderly
among older women. We also explored the
score $3.618; (4) self-reported dementia diagnosis; or (5) living in
association between SMCs assessed across a nursing home. Women who did not meet any of these screening
time and risk of cognitive impairment, in criteria were considered cognitively normal. Data from women
order to evaluate whether the sensitivity of who met any of the screening criteria were then further examined
by a team of clinicians, who reviewed individuals’ cognitive test
SMCs as a potential early indicator of future
results from year 20 and all prior cognitive assessment data,
cognitive impairment may change over time. functional status, medical history, medications, and depressive
symptoms and then diagnosed individuals as having MCI (based
METHODS Population. Participants were community- on modified Peterson criteria),19 dementia (based on DSM-IV
dwelling older women (aged 65 years and older) in the criteria), or as being cognitively normal.
prospective cohort Study of Osteoporotic Fractures (SOF), Other variables. Participants completed self-report ques-
recruited by study centers in Baltimore, MD, Minneapolis, tionnaires assessing basic demographics and educational history at
MN, Portland, OR, and the Monongahela Valley near baseline. Information about comorbidities was collected repeat-
Pittsburgh, PA. A total of 9,704 Caucasian women were edly over time, based on participants’ self-report of physician
enrolled between 1986 and 1988, and 662 African American diagnoses of hypertension, diabetes, stroke, and myocardial
women were enrolled later, between 1997 and 1998. Women infarction. Body mass index was calculated based on participants’
with a history of hip fracture or bilateral hip replacement or height and weight (kg/m2). Depressive symptoms were assessed
who were unable to walk without assistance were excluded. with the 15-item GDS.10 Because our primary variable of interest
Further details regarding SOF recruitment methods and study (SMCs) was taken from this inventory, we utilized a GDS score
design have been published previously.9 out of 14 that excluded the memory item, and we applied a
As part of the SOF–Women, Cognitive Impairment Study of prorated cutoff (6 of 14) to define depression in keeping with
Exceptional Aging ancillary study, year-20 (2006–2008) clinical traditional cutpoint.10
cognitive diagnosis of normal cognition, MCI, or dementia was
evaluated by an expert panel among 1,338 women from the Cau- Statistical analysis. We first compared basic demographic and
casian cohort. For our primary analyses, we included 1,107 other baseline characteristics between individuals with and without
women who were cognitively normal at baseline (as defined by baseline SMCs using t tests, x2, or Fisher exact tests as appropriate.
a baseline modified Mini-Mental State Examination [mMMSE] We then conducted logistic regression models to investigate
score no more than 1.5 SD below the mean compared with age- whether baseline SMCs predicted clinically significant cognitive
and education-matched peers in the entire Caucasian cohort), impairment (MCI/dementia vs normal cognition) 18 years later.
who completed SMC assessment shortly after baseline (year 2), We conducted this analysis using an unadjusted model as well as a
and whose year-20 cognitive diagnosis was determined. Com- model adjusting for demographics (age, education), baseline
pared with the remaining women in the SOF Caucasian cohort, mMMSE score, and participant characteristics that significantly
these 1,107 women were younger and more highly educated, had differed between groups (p , 0.05). We also explored whether
fewer comorbidities and endorsed fewer depressive symptoms (all the association between SMCs and risk of cognitive impairment
p , 0.001), and had higher body mass index (p 5 0.03). changed over time. Similar to our primary analyses, we used logistic
Included women were less likely to endorse SMCs at baseline regression models to investigate the association between SMCs at
and at all subsequent SMC assessments (all p , 0.001). each time point (years 2, 6, 10, 16) and later diagnosis of cognitive
impairment, unadjusted and adjusted for the same factors as above
Standard protocol approvals, registrations, and patient measured at each respective time point. These models were con-
consents. All participants gave written informed consent to par- ducted among participants who had complete SMC data across all
ticipate in SOF. The study was approved by institutional review time points and who were cognitively normal at the examined time
boards at each study site and at the University of California, San point as determined using the same age- and education-based
Francisco (coordinating center). mMMSE cutoffs as defined above (year 2: n 5 1,025; year 6:
Measures. Subjective memory complaints. SMCs were assessed n 5 990; year 10: n 5 977; year 16: n 5 943).
shortly after baseline at year 2 (henceforth referred to as baseline
SMCs) and again in years 6, 10, and 16 using the following yes/ RESULTS At baseline, 8.0% of participants (n 5 89)
no question from the 15-item Geriatric Depression Scale endorsed SMCs. Demographics and other participant
(GDS)10: “Do you feel you have more problems with memory characteristics are compared between those with and
than most?” without baseline SMCs in table 1. Compared to those
Cognition and cognitive diagnosis. Global cognitive func-
without baseline SMCs, women with SMCs had
tioning was assessed using a 26-item mMMSE11 at baseline and
repeatedly over time. At year 20, a larger cognitive battery was lower education (p 5 0.01), greater myocardial
given that included the following: (1) the mMMSE,9 which is an infarction history (p 5 0.01), and higher depressive
expanded measure of global cognition; (2) the California Verbal symptoms (p , 0.001).

Neurology 85 November 24, 2015 1853

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 Although SMCs measured 14 years before the diag-
Table 1 Baseline characteristics by presence of SMCs among 1,107 older
women
nostic evaluation were not significantly associated
with cognitive impairment (adjusted OR 5 1.6,
Baseline SMCs No SMCs at baseline 95% CI 0.9–2.7, p 5 0.12), the adjusted OR was
Characteristic (n 5 89) (n 5 1,018) p Value
in the same direction and of similar magnitude as the
Age, y 70.5 (2.8) 70.3 (2.9) 0.43
significant association between SMCs measured
Education 18 years before the diagnostic evaluation (adjusted
<High school 25.8 13.8 OR 5 1.7, 95% CI 1.1–2.9, p 5 0.03) and SMCs
High school graduate 38.2 46.5 0.01 measured 10 years before the diagnostic evaluation
Some college/college graduate 36.0 39.8
(adjusted OR 5 1.9, 95% CI 1.1–3.1, p 5 0.01).
The association between SMCs and cognitive
Hypertension 31.5 29.2 0.65
impairment was greatest at the last time point, i.e.,
Diabetes 3.4 2.3 0.46
SMCs measured 4 years before the diagnostic
Stroke 0 0.8 0.99 evaluation (adjusted OR 5 3.0, 95% CI 1.8–5.0,
Myocardial infarction 7.4 2.5 0.01 p , 0.001).
Body mass index, kg/m 2
26.1 (4.3) 26.5 (4.2) 0.41 In sensitivity analyses among individuals with
Depressive symptoms 2.0 (2.4) 0.9 (1.4) ,0.001 complete SMC data who remained cognitively nor-
(GDS total of 14)a mal at year 16 (n 5 943), we explored the indepen-
mMMSE score, total of 26 25.3 (1.0) 25.3 (1.0) 0.89 dent effects of SMCs at particular time points and the
effects of individuals’ pattern of SMC endorsement
Abbreviations: GDS 5 Geriatric Depression Scale; mMMSE 5 modified Mini-Mental State
Examination; SMC 5 subjective memory complaint. over time. For the former, we included all 4 SMC
Data are mean (SD) or %. Characteristics were measured at year 2 (at the same time as assessment time points as predictors in a logistic
baseline SMC assessment), except for education, hypertension, diabetes, and mMMSE, regression model, adjusted for the same covariates as
which were measured at year 1.
a
GDS total score was taken out of 14, excluding the memory item, which was used to
above. There was a significant overall effect of endors-
assess SMCs. ing SMCs at some point on likelihood of cognitive
impairment (p , 0.001) and significant heterogene-
ity of effects among the SMC time points (p 5 0.02).
Women with baseline SMCs were more likely to
Pairwise comparisons revealed that the last SMC time
be diagnosed with cognitive impairment at year 20
point (4 years before the cognitive diagnostic evalua-
compared to women without baseline SMCs
tion) was most strongly associated with cognitive
(52.8% vs 38.0%; x2 5 7.5, p 5 0.01). As shown
impairment compared with other SMC time points
in table 2, baseline SMCs were significantly associated
(the last SMC time point was significantly different
with increased risk of cognitive impairment 18 years
from every other time point [all p , 0.05]; there were
later even after adjustment for demographics, baseline
no other significant pairwise differences [all p .
mMMSE score, history of myocardial infarction, and
0.05]). To explore individuals’ pattern of SMC
depressive symptoms (adjusted odds ratio [OR] 5
endorsement over time, we first compared risk of
1.7, 95% confidence interval [CI] 1.1–2.8, p 5
cognitive impairment between individuals who
0.03). In a sensitivity analysis, we excluded 26 partic-
endorsed SMCs at 0, 1, vs $2 out of the 4 SMC
ipants with baseline depression, and the effect of
assessment visits. Compared with those who never
SMCs was unchanged (fully adjusted OR 5 1.7,
endorsed SMCs, individuals who endorsed SMCs at
95% CI 1.01–2.8, p 5 0.046).
one or more visits were more likely to be diagnosed
The figure displays the association between SMCs
with cognitive impairment (SMCs at only 1 visit
assessed at varying time points before the diagnos-
[10.7% of participants; n 5 101]: adjusted OR 5
tic evaluation and risk of cognitive impairment.
2.1, 95% CI 1.3–3.2, p 5 0.001; SMCs at $2 visits
[7.3% of participants; n 5 69]: adjusted OR 5 2.0,
Table 2 Association between baseline SMCs and cognitive impairment 18 years 95% CI 1.2–3.4, p 5 0.01). We then classified
later among 1,107 older women
individuals as endorsing patterns of (1) persistent
% with Model 1 (unadjusted) Model 2 (adjusted)a SMCs—those who endorsed SMCs at baseline and
cognitive
impairment
$1 additional visit (n 5 47), (2) incident SMCs—
Baseline SMCs 18 y later OR (95% CI), p value those who denied SMCs at baseline but endorsed
No SMCs (n 5 1,018) 38.0 Reference Reference SMCs at $1 later visit (n 5 91), (3) transient
SMCs (n 5 89) 52.8 1.8 (1.2–2.8), p 5 0.007 1.7 (1.1–2.8), p 5 0.03 SMCs—those who endorsed SMCs at baseline but
denied SMCs at all other visits (n 5 32), vs (4) those
Abbreviations: CI 5 confidence interval; OR 5 odds ratio; SMC 5 subjective memory
who never endorsed SMCs (n 5 773). Compared
complaint.
a
Adjusted for demographics (age, education), baseline modified Mini-Mental State Exam- with individuals who never endorsed SMCs, individ-
ination score, history of myocardial infarction, and depressive symptoms. uals with incident SMCs (adjusted OR 5 2.4, 95%

1854 Neurology 85 November 24, 2015

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 Figure SMCs at varying time points before the diagnostic evaluation and risk of cognitive impairment

Results of logistic regression analyses examining the association between SMCs assessed at varying time points before the diagnostic evaluation and risk of
cognitive impairment. Adjusted models include adjustment for demographics (age, education) as well as mMMSE score, history of myocardial infarction, and
depressive symptoms as measured at each SMC assessment time point. At each time point, individuals with an impaired mMMSE score at that time point
were excluded (using 21.5 SD age- and education-based cutoffs). CI 5 confidence interval; mMMSE 5 modified Mini-Mental State Examination; SMC 5
subjective memory complaint.

CI 1.5–3.9, p , 0.001) and persistent SMCs significance of SMCs, we found that even women
(adjusted OR 5 1.9, 95% CI 1.01–3.5, p 5 0.046) who endorsed SMCs at only one point in time were
were more likely to develop cognitive impairment. The at increased risk of cognitive impairment and that
effect of transient SMCs was not statistically significant patterns of persistent SMCs and incident SMCs
(adjusted OR 5 1.4, 95% CI 0.7–3.0, p 5 0.36). over time were both significantly associated with
risk of cognitive impairment. Our findings provide
DISCUSSION We investigated the long-term asso- further evidence that SMCs in aging warrant close
ciation between SMCs and cognitive outcomes attention as a possible early warning sign of future
among older women and found that SMCs were cognitive problems, even several years in advance.
significantly associated with cognitive impairment Indeed, a recent study quantified that older adults
nearly 2 decades later. We also examined the with SMCs who went on to develop MCI
association between SMCs assessed repeatedly over progressed to this diagnosis an average of 9.2 years
time and subsequent diagnosis of cognitive after they first endorsed SMCs,20 documenting a
impairment and found that the strongest association prolonged prodromal period between SMCs and
was present when SMCs were assessed just a few the development of cognitive impairment.
years in advance of participants’ clinical cognitive Our results add further support to the possibility
diagnostic evaluation. While women with SMCs 18 that SMCs may be an early symptom of an under-
years before the diagnostic evaluation had 1.7 times lying neurodegenerative process, such as AD. That
greater odds of receiving an impaired diagnosis, the association between SMCs and risk of cognitive
women with SMCs 4 years before the diagnostic impairment was present very early (i.e., for SMCs
evaluation had 3 times greater odds of receiving an endorsed nearly 20 years before our cognitive
impaired diagnosis. As further evidence of the diagnostic evaluation) and was strongest for SMCs

Neurology 85 November 24, 2015 1855

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 closest in time to the diagnostic evaluation, this Particularly as included women were less likely to
timing pattern suggests that SMCs may signal an have SMCs than those not included, some women
insidious disease process as it first emerges and with SMCs may have actually been lost to follow-up
continues to unfold. This would be consistent with before year 20 related to the development of cogni-
the long, prodromal period of AD, during which tive impairment. Further limitations include that
amyloid deposition, thought to be the beginnings our findings cannot be generalized to men or to
of the AD neuropathology cascade, is present many other racial/ethnic groups.
years before the clinical manifestation of objective It is important to consider that we investigated
cognitive impairment.21 Moreover, neuroimaging SMCs among community-dwelling older women
studies corroborate this possibility in finding that participating in a population-based study. SMCs
older adults with SMCs have greater amyloid bur- endorsed by women in the general population may
den3,4 and patterns of atrophy similar to older reflect a different phenomenon than what occurs
adults with amnestic MCI and AD.22,23 Neverthe- when older adults are concerned enough about their
less, SMCs may not be specific to AD but could cognition to present to a clinic for SMCs. Moreover,
also precede objective cognitive impairment in the because we assessed SMCs using one yes/no ques-
context of the gradual accumulation of other neu- tion, we likely did not capture many complexities
ropathology in aging, such as cerebrovascular dis- inherent in asking an older adult to evaluate his or
ease.24,25 Despite the above possibilities and our her own cognitive functioning. As highlighted by
findings, it should be noted that the associations the 2014 Subjective Cognitive Decline Initiative
we observed (particularly for early years of SMC Workgroup,2 valuable data may be gleaned from a
assessment) are somewhat modest. There may be more comprehensive assessment that asks individu-
other reasons for an older adult to endorse SMCs als to appraise their functioning in multiple cogni-
that do not necessarily lead to the development of tive domains both in comparison to their peers and
MCI or dementia. relative to their own prior ability levels. However, it
Strengths of the present study include our ability appears notable that, despite the simplicity of our
to investigate the predictive utility of SMCs over an SMC measure and despite studying women from the
extended follow-up period of 18 years and to exam- general population rather than clinic patients, we
ine SMCs assessed repeatedly over time. In addition, still found an association between SMCs and risk
our study benefits from the completion of a compre- of cognitive impairment. Therefore, our results sup-
hensive clinical cognitive diagnostic evaluation to port the idea that clinical providers, even in primary
determine participants’ cognitive status (normal, care settings, should consider incorporating assess-
MCI, dementia). Nevertheless, there are some lim- ment of SMCs into their routine checkups of older
itations in the assessment of cognitive status in our patients, as endorsement of SMCs may be informa-
study. Because the clinical cognitive diagnostic eval- tive even when the patient’s primary reason for pre-
uation was only conducted at the final time point, senting to clinic is not their cognition. Our results
we cannot be certain exactly when individuals first also raise the possibility that even a relatively brief
met diagnostic criteria for MCI/dementia. Our uti- SMC assessment could be a valuable screening tool,
lization of global cognitive screen (mMMSE) scores perhaps in the form of a questionnaire administered
to exclude individuals with cognitive impairment at in the waiting room to help minimize provider bur-
baseline and at subsequent SMC assessment time den. Moreover, as we found associations between
points reduces the likelihood that significant cogni- SMCs and risk of cognitive impairment among in-
tive impairment was present at those time points. dividuals with unimpaired scores on a global cogni-
However, it is possible that more subtle cognitive tive screen, we recommend that providers should
deficits were not detected by these screens. An addi- not be entirely reassured by a “normal” cognitive
tional limitation is that we cannot be certain of the screen performance but instead take care to monitor
extent to which survival bias may have influenced individuals with SMCs over time for possible cogni-
our results. Women included in the present study tive decline.
were generally healthier than those who were not, SMCs among cognitively normal older women
were less likely to endorse SMCs, and were required appear to be an early indicator of risk of cognitive
to have completed a 20th year of study participation. impairment and may be a subtle signal of an underly-
While some survival bias in our study is therefore ing neurodegenerative disease process such as AD that
likely, it is perhaps even more compelling that we is still in its earliest stages. Early detection of AD and
found long-term associations between SMCs and other dementias is likely needed to enable potential
cognitive impairment among this relatively healthier interventions to be applied as early in the disease
cohort, as survival bias would most likely bias course as possible. Our results suggest that dementia
our findings toward an underestimate of risk. prevention research trials should target older women

1856 Neurology 85 November 24, 2015

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 with SMCs as a high-risk group, in order to attempt serves on the Beeson Scientific Advisory Board. Go to Neurology.org for
full disclosures.
to intervene among those who may be showing the
earliest symptoms of neurodegeneration. Received April 29, 2015. Accepted in final form July 28, 2015.

AUTHOR CONTRIBUTIONS REFERENCES

Dr. Kaup designed and conceptualized the study, conducted data anal- 1. Nelson PT, Braak H, Markesbery WR. Neuropathology
ysis, interpreted the data, and drafted and revised the manuscript. and cognitive impairment in Alzheimer disease: a complex
Dr. Nettiksimmons assisted with data analysis and data interpretation but coherent relationship. J Neuropathol Exp Neurol
and reviewed and revised the manuscript. Dr. LeBlanc assisted with
2009;68:1–14.
data interpretation and reviewed and revised the manuscript. Dr. Yaffe
2. Jessen F, Amariglio RE, van Boxtel M, et al. A conceptual
designed and conceptualized the study, interpreted the data, reviewed
and revised the manuscript, and supervised the study. All authors have
framework for research on subjective cognitive decline in
approved the manuscript. preclinical Alzheimer’s disease. Alzheimers Dement 2014;
10:844–852.
3. Amariglio RE, Becker JA, Carmasin J, et al. Subjective
ACKNOWLEDGMENT
cognitive complaints and amyloid burden in cognitively
The authors thank the following personnel associated with the Study
normal older individuals. Neuropsychologia 2012;50:
of Osteoporotic Fractures (SOF) Research Group. Investigators in the
2880–2886.
SOF Research Group: San Francisco Coordinating Center (California
Pacific Medical Center Research Institute and University of California 4. Perrotin A, Mormino EC, Madison CM, Hayenga AO,
San Francisco): S.R. Cummings (principal investigator), D.C. Bauer Jagust WJ. Subjective cognition and amyloid deposition
(coinvestigator), D.M. Black (coinvestigator), W. Browner (coinvesti- imaging: a Pittsburgh compound B positron emission
gator), P.M. Cawthon (coinvestigator), N. Lane (coinvestigator), tomography study in normal elderly individuals. Arch
M.C. Nevitt (coinvestigator), C. McCulloch (coinvestigator), Neurol 2012;69:223–229.
A. Schwartz (coinvestigator), K.L. Stone (coinvestigator), G. Tranah 5. Buckley R, Saling M, Ames D, et al. Factors affecting
(coinvestigator), K. Yaffe (coinvestigator), R. Benard, T. Blackwell, subjective memory complaints in the AIBL Aging Study:
L. Concepcion, D. Evans, S. Ewing, C. Fox, R. Fullman, S.L.
biomarkers, memory, affect, and age. Int Psychogeriatr
Harrison, M. Jaime-Chavez, D. Kriesel, W. Liu, L. Lui, L. Palermo,
2013;25:1307–1315.
N. Parimi, K. Peters, M. Rahorst, C. Schambach, and J. Ziarno.
6. Hollands S, Lim YY, Buckley R, et al. Amyloid-b related
University of Maryland: M.C. Hochberg (principal investigator), R.
Nichols (clinic coordinator), and S. Link. University of Minnesota: memory decline is not associated with subjective or infor-
K.E. Ensrud (principal investigator), S. Diem (coinvestigator), mant rated cognitive impairment in healthy adults.
M. Homan (coinvestigator), P. Van Coevering (program coordinator), J Alzheimers Dis 2015;43:677–686.
S. Fillhouer (clinic director), N. Nelson (clinic coordinator), K. Moen 7. Mitchell A, Beaumont H, Ferguson D, Yadegarfar M,
(assistant program coordinator), K. Jacobson, M. Forseth, R. An- Stubbs B. Risk of dementia and mild cognitive impairment
drews, S. Luthi, K. Atchison, and L. Penland-Miller. University of Pitts- in older people with subjective memory complaints: meta‐
burgh: J.A. Cauley (principal investigator), L.H. Kuller (co–principal analysis. Acta Psychiatr Scand 2014;130:439–451.
investigator), J.M. Zmuda (coinvestigator), L. Harper (project director),
8. Sperling RA, Aisen PS, Beckett LA, et al. Toward defining
L. Buck (clinic coordinator), M. Danielson (project administrator),
the preclinical stages of Alzheimer’s disease: recommenda-
D. Cusick, A. Flaugh, M. Gorecki, and C. Newman. The Kaiser Perma-
tions from the National Institute on Aging–Alzheimer’s
nente Center for Health Research, Portland, OR: T. Hillier (principal
investigator), K. Vesco (coinvestigator), K. Pedula (coinvestigator), Association Workgroups on Diagnostic Guidelines for Alz-
J. Van Marter (project director), M. Summer (clinic coordinator), heimer’s disease. Alzheimers Dement 2011;7:280–292.
A. MacFarlane, J. Rizzo, K. Snider, and J. Wallace. 9. Teng EL, Chui HC. The modified Mini-Mental State
Examination (3MS). Can J Psychiatry 1987;41:114–121.
10. Sheikh JI, Yesavage JA. Geriatric Depression Scale (GDS):
STUDY FUNDING
recent evidence and development of a shorter version. In:
The Study of Osteoporotic Fractures (SOF) is supported by NIH fund-
ing. The National Institute on Aging (NIA) provides support under the Clinical Gerontology: A Guide to Assessment and Inter-
following grants: R01 AG005407, R01 AR35582, R01 AR35583, R01 vention. New York: The Haworth Press; 1986:165–173.
AR35584, R01 AG005394, R01 AG027574, and R01 AG027576. 11. Folstein MF, Robins LN, Helzer JE. The Mini-Mental
The research described here was also supported in part by NIA grant State Examination. Arch Gen Psychiatry 1983;40:812.
K24AG031155 (awarded to Kristine Yaffe, MD); by Career Develop- 12. Delis DC, Kramer JH, Kaplan E, Ober BA. California
ment Award 1IK2RX001629 (awarded to Allison Kaup, PhD) from Verbal Learning Test–Second Edition (CVLT-II). San An-
the US Department of Veterans Affairs, Rehabilitation Research and tonio: Psychological Corporation; 2000.
Development Service; and the Department of Veterans Affairs Office
13. Strauss E, Sherman E, Spreen O. General cognitive function-
of Academic Affiliations Advanced Fellowship Program in Mental Illness
ing, neuropsychological batteries, and assessment of premor-
Research and Treatment, the Medical Research Service of the San Fran-
bid intelligence. In: A Compendium of Neuropsychological
cisco Veterans Affairs Medical Center, and the Department of Veterans
Affairs Sierra-Pacific Mental Illness Research, Education, and Clinical Tests: Administration, Norms, and Commentary, 3rd ed.
Center (MIRECC). The contents of this manuscript do not represent New York: Oxford University Press; 2006:283–285.
the views of the US Department of Veterans Affairs or the United States 14. Reitan RM. Trail Making Test: Manual for Administra-
Government. tion and Scoring. Tucson: Reitan Neuropsychology Lab-
oratory; 1992.
DISCLOSURE 15. Strauss E, Sherman E, Spreen O. Executive functions. In:
A. Kaup and J. Nettiksimmons report no disclosures relevant to the man-
A Compendium of Neuropsychological Tests: Administra-
uscript. E. LeBlanc’s institution has received funding from Amgen, tion, Norms, and Commentary, 3rd ed. New York:
Bristol-Myers Squibb, and AstraZeneca for research grants unrelated to Oxford University Press; 2006:499–503.
the current study. K. Yaffe is a consultant for Novartis and Pfizer. She 16. Tranah GJ, Blackwell T, Stone KL, et al. Circadian activ-
serves on DSMBs for Takeda, Inc., and an NIA-sponsored study and also ity rhythms and risk of incident dementia and mild

Neurology 85 November 24, 2015 1857

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 cognitive impairment in older women. Ann Neurol 2011; 21. Jack CR, Knopman DS, Jagust WJ, et al. Update on hypo-
70:722–732. thetical model of Alzheimer’s disease biomarkers. Lancet
17. Espeland MA, Rapp SR, Robertson J, et al. Benchmarks for Neurol 2013;12:207–216.
designing two-stage studies using modified Mini-Mental 22. Peter J, Scheef L, Abdulkadir A, et al. Gray matter atrophy
State Examinations: experience from the Women’s Health pattern in elderly with subjective memory impairment.
Initiative Memory Study. Clin Trials 2006;3:99–106. Alzheimers Dement 2014;10:99–108.
18. Jorm A, Jacomb P. The Informant Questionnaire on Cog- 23. Saykin A, Wishart H, Rabin L, et al. Older adults with
nitive Decline in the Elderly (IQCODE): socio- cognitive complaints show brain atrophy similar to that of
demographic correlates, reliability, validity and some amnestic MCI. Neurology 2006;67:834–842.
norms. Psychol Med 1989;19:1015–1022. 24. Sajjad A, Mirza SS, Portegies ML, et al. Subjective mem-
19. Petersen RC, Doody R, Kurz A, et al. Current concepts ory complaints and the risk of stroke. Stroke 2015;46:
in mild cognitive impairment. Arch Neurol 2001;58: 170–175.
1985–1992. 25. van Norden AG, van Uden IW, de Laat KF, et al. Cere-
20. Kryscio RJ, Abner EL, Cooper GE, et al. Self-reported bral microbleeds are related to subjective cognitive fail-
memory complaints implications from a longitudinal ures: the RUN DMC Study. Neurobiol Aging 2013;34:
cohort with autopsies. Neurology 2014;83:1359–1365. 2225–2230.

Increasing the Value of YOUR AAN Membership

FREE MOC Benefits Starting January 1, 2015

You asked and we listened. As of January 1, 2015, your robust AAN membership package includes
FREE* access to online learning programs designed specifically to help you take the necessary steps
toward fulfilling your maintenance of certification (MOC) requirements as mandated by the
ABPN: NeuroPISM, NeuroSAE®, NeuroLearnSM.

Learn more at AAN.com/view/MOC

*$0 purchase price excludes Student members and Nurse Practitioner/Physician Assistant members at the
lower dues rate. Free access is limited to one course per program at a time.

Neurology® Genetics Call For Papers

Neurology: Genetics is an open access, online only journal that

provides neurologists with outstanding original contributions
that elucidate the role of genetic and epigenetic variation in
diseases and biological traits of the central and peripheral ner-
vous system. We welcome all submissions. For more informa-
tion on how to submit, visit http://www.neurology.org/site/
gen/gen2.xhtml.

1858 Neurology 85 November 24, 2015

ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.