RESEARCH ARTICLE

Cognitive stimulation in cognitively impaired individuals

and cognitively healthy individuals with a family history of
dementia: short-term results from the “Allena-Mente”
randomized controlled trial
Letizia Polito1†, Simona Abbondanza1†, Roberta Vaccaro1, Eleonora Valle1, Annalisa Davin1, Alessandro Degrate2,
Simona Villani2 and Antonio Guaita1
1
“Golgi Cenci” Foundation, Abbiategrasso, Italy
2
Unit of Biostatistics and Clinical Epidemiology, Department of Public Health, Experimental and Forensic Medicine, University of Pavia,
Pavia, Italy
Correspondence to: A. Guaita. E-mail: a.guaita@golgicenci.it

†
These authors contributed equally to the study.

Objective: We evaluated the short-term efficacy of a protocol of cognitive stimulation (CS), compared with
a sham intervention, on cognitive performance in cognitively healthy individuals with a family history of
dementia (NDFAM) and in non-demented individuals with cognitive impairment (CI).
Methods: We performed a randomized controlled trial of CS in NDFAM and CI. CS consisted in 10 twice
weekly meetings of CS focused on a specific cognitive area. CS was compared with a sham intervention
(CT) using Mini-mental state examination (MMSE), Montreal Cognitive Assessment (MoCA), and the
Corsi test. All study participants were typed for the presence of apolipoprotein E (APOE)-Ɛ4.
Results: Cognitively healthy NDFAM showed a higher net cognitive gain after CS, as reflected in their
MoCA score, and a borderline significant net increase in visuospatial memory (Corsi test) compared
with those receiving the CT. APOE-Ɛ4 carriers showed a less significant improvement on the Corsi test
with respect to APOE-Ɛ4 non-carriers. In the CI sample, the MoCA and Corsi test results did not differ
between the cognitively stimulated subjects and the controls. No changes in MMSE scores were found
in either sample of subjects.
Conclusions: These findings suggest that CS as structured in this study is an effective treatment in cognitively
healthy individuals, whereas it is less effective in individuals with CI. Moreover, evaluation of APOE-Ɛ4 status
provided evidence of a substantial genetic contribution to the efficacy of CS on visuospatial memory as
measured using the Corsi test. Copyright # 2014 John Wiley & Sons, Ltd.
Key words: cognitive stimulation; mild cognitive impairment; dementia; aging; APOE
History: Received 8 January 2014; Accepted 31 July 2014; Published online in Wiley Online Library
(wileyonlinelibrary.com)
DOI: 10.1002/gps.4194

Introduction dementia. A growing number of studies are evaluating

The rapid aging of the population is expected to bring
an increase in the rate of people developing dementia
(Alzheimer’s Association, 2010). The lack of effective
therapies places a great burden on society. It is important
to develop strategies designed to delay cognitive decline
onset in healthy individuals at risk of developing
the efficacy of cognitive stimulation (CS) in older sub-
jects with cognitive impairment (CI) (Martin et al.,
2011). Martin and co-workers concluded that cognitive
interventions do lead to performance gain in subjects
with mild CI (MCI), even though none of the observed
effects could be attributed specifically to cognitive train-
ing. A recent meta-analysis of 10 cognitive training trials

Copyright # 2014 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2014

in MCI subjects showed moderate-to-large beneficial ef-
fects of cognitive exercises on memory-related outcomes
(Gates et al., 2011). According to the authors, previous
reports of lack of efficacy of cognitive training in MCI
may be due to poorly defined interventions and incor-
rect study designs.
In an attempt to clarify the conflicting evidence in
the literature with regard to the efficacy of CS in
preventing or slowing down cognitive deterioration,
a randomized control trial (called the “Allena-Mente”
study, NCT01793493 ClinicalTrial.gov) was conducted.
The aim of this trial was to evaluate the short-term,
medium-term, and long-term efficacy of CS, versus
sham treatment, on cognitive performance in cogni-
tively healthy subjects with a family history of dementia
(NDFAM) and in non-demented subjects affected by
MCI.
In this paper, we specifically evaluate the short-term
efficacy of a CS protocol, compared with a sham inter-
vention (CT), on cognitive performance, and investigate
the influence of apolipoprotein E (APOE)-Ɛ4 carrier
status.
Methods
Participants
Two samples of subjects, each meeting specific eligibility
criteria, were recruited:
• cognitively healthy individuals with self-reported first-
degree family history of dementia (Diagnostic and
Statistical Manual of Mental Disorders, fourth edition,
criteria) including those who reported Alzheimer’s
disease and vascular dementia, but excluding other
subtypes of dementia (NDFAM sample) and
• individuals with a diagnosis of MCI according to
Petersen’s criteria (Petersen, 2004) or “cognitive im-
pairment, no dementia,” for individuals who met the
MCI criteria except memory complain (Guaita et al.,
2013) (CI sample). Amnesic CI and non-amnesic CI
were equally present in CS (13 CIa and 9 CIna) and
CT group (12 MCIa and 10 MCIna).
Subjects with clinical conditions that may reduce their
active participation in the intervention were excluded.
Both samples were recruited from subjects included
in the “InveCe.Ab” population study (NCT01345110
ClinicalTrial.gov). A more detailed description of
diagnostic criteria was reported in the methodological
article of the “InveCe.Ab” study (Guaita et al., 2013).
The study was approved by the Ethical Review
Board of the University of Pavia. The subjects enrolled
Copyright # 2014 John Wiley & Sons, Ltd.
L. Polito et al.
gave their written informed consent before entering the
study, including consent to the use of genetic information.
Interventions
Participants were divided into small groups (seven to
eight subjects) for CS or CT. The CS intervention
consisted of 10 twice weekly sessions, led by a trained
neuropsychologist. Each session included 10 min devoted
to body awakening and 90 min of cognitive activities
divided into two parts. The main part, lasting about
60 min, was focused on reality orientation and CS to
activate strategies for improving cognitive functions.
The second part consisted of cognitive training exercises
focused on a specific cognitive function that changed
from session to session (Olazarán et al., 2004; Belleville
et al., 2006; Willis et al., 2006; Wolinsky et al., 2006).
During CS, the approach of “mediated learning” was
applied (Feuerstein et al., 1999). With the help of the
trainer, participants were required to compare differ-
ent strategies in order to identify the most efficient
and effective ones. In the training part of the session,
participants carried out exercises designed to stimulate
attention (auditory and visual) on the targeted neuro-
psychological function (executive reasoning, language–
verbal fluency, semantic memory, visual perception,
encoding, information storage, nonverbal learning, and
executive problem solving). At the end of each meeting,
after discussing difficulties and possible new strategies,
the group was encouraged to use the technique of
“cognitive bridging” in relation to their real-life activi-
ties. An observer attended all the sessions to guarantee
homogeneity. The CT group attended two interactive
60-min meetings (lesson and questions by a geriatrician)
that focused on lifestyle education and brain functioning
in older people.
A pilot study was conducted on healthy older indi-
viduals (n = 31) to set up the training protocol to stan-
dardize methodologies. Data were not included.
Cognitive endpoints
The primary outcome was the change in the Mini-mental
state examination (MMSE; Magni et al., 1996) score from
baseline to post-intervention, whereas the secondary out-
comes were the baseline to post-intervention changes in
the Montreal Cognitive Assessment (MoCA; Nasreddine
et al., 2005) and Corsi test scores (Spinnler and Tognoni,
1987).
The neuropsychological test battery took 30–40 min
per subject and was administered before and after the
intervention period.
Int J Geriatr Psychiatry 2014
Cognitive stimulation in non-demented individuals
Baseline assessments were completed 1–2 weeks
prior to the first CS or CT session, whereas the post-
intervention assessments were performed 2 weeks after
program completion.
Other variables
For the purposes of the present work, age, gender, years
of formal education, and APOE-Ɛ4 genotype were
considered.
Apolipoprotein E Ɛ4 genotyping was conducted on
participants’ genomic DNA with real-time polymerase
chain reaction using custom TaqMan probes (Applied
Biosystems, Inc., Foster City, CA) specific for single
nucleotide polymorphisms (SNPs) of rs7412 and
rs429358 at nucleotides 112 and 158 of the APOE gene,
respectively. APOE genotype was determined from the
combination of alleles present at the 112 and 158
polymorphisms. Genotype information was simplified
to a binary classification on the basis of the absence
or presence of one or more Ɛ4 alleles at this locus.
Sample sizes
We report the sample size calculation for the primary
aim of the “Allena-Mente” study (the long-term effi-
cacy of CS measured with MMSE). Expected effect
sizes were estimated on the basis of a pilot study on
31 healthy volunteers (data not reported), published
data when available, or otherwise the consensus opin-
ion of clinicians according to their own experience.
Mild cognitive impairment and NDFAM were con-
sidered separately for sample size calculation.
NDFAM: Taking into account an average MMSE score
of 27 points, 36 subjects were needed in each arm to
detect a 0-point net change at long time (4 years) with
respect to baseline on MMSE score in CS individuals and
a 1-point net decrease in sham individuals (standard
deviation [SD] = 1.5 points in both groups), assuming
a statistical power of 80% and two-sided alpha error
of 0.05. To compensate for a discontinuation rate of
about 10% in the follow-up period, not less than 40
subjects in each study group were needed.
CI: Similarly, 22 subjects were needed in each arm to
ensure a statistical power of 80% (0.05 two-sided alpha
error) of detecting a 1-point net increase at long time
with respect to baseline on MMSE score in CS individ-
uals and a 1-point net decrease in sham individuals
(SD = 2 points in both groups), supposing an average
MMSE at baseline equal to 22 points. To compensate
Copyright # 2014 John Wiley & Sons, Ltd.
for a discontinuation rate of about 20% in the follow-up
period, 27 individuals were enrolled in each arm.
NDFAM and CI were separately randomly allocated
to CS and control groups.
Randomization and blinding
A stratified randomization procedure was performed
using random allocation. Stratification was performed
by birth cohort (≤1937 and ≥1938) and level of education
(≤5 and >5 years of education); the allocation ratio to the
CS and CT interventions was set at 1:1. The randomiza-
tion was performed by an independent statistician.
To ensure blinding, the neuropsychologist who ad-
ministered the cognitive tests was not aware of the
samples or groups to which the subjects belonged.
For the same reason, the enrolled subjects were told
that the two interventions were different cognitive
strategies, not real versus sham treatments.
Statistical analyses
For the purposes of the present study, the following
analyses were conducted 2 weeks after the end of the
CS or CT treatment. The CI and NDFAM samples were
analyzed separately using the same statistical approach.
Quantitative variables were reported as mean values ± SD,
whereas qualitative variables were reported as percentages.
Differences in qualitative baseline variables between
the CS and CT groups were investigated using the chi-
square test. Unpaired t-tests with Satterthwaite’s correc-
tion, if necessary, were applied to compare quantitative
(age and years of education) characteristics and end-
points (MMSE, MoCA, and Corsi test scores) between
the CS and CT subjects both at baseline and 2 weeks
after the CS or CT. Short-term efficacy was measured
as net change from baseline in the MMSE score (pri-
mary endpoint) and in the MoCA and Corsi test scores
(secondary endpoints). Effect sizes of between-group
changes according to Cohen (1992) were also reported.
A paired t-test was also applied to compare within-
group changes over time in cognitive scores (MMSE,
MoCA, and Corsi test). A linear regression model was
used to adjust the effect of intervention by APOE-Ɛ4
and gender. If the interaction of intervention with
APOE-Ɛ4 was significant, the linear regression analysis
was reported separately by the presence of at least one
E4 allele and by the absence of the E4 allele (dominant
model). A p-value less than 0.05 was considered signif-
icant (two-sided). Analyses were conducted using STATA
version 12 StataCorp. 2011 (Stata Statistical Software:
Int J Geriatr Psychiatry 2014
 L. Polito et al.

Release 12. College Station, TX: StataCorp LP ) and SPSS Baseline to post-intervention changes in primary and
(release 11, Inc., Chicago, IL) software. secondary endpoints

In the NDFAM sample, both treatment groups re-

Results
A total of 77 cognitively healthy NDFAM and 44 non-
demented individuals with CI subjects successfully
completed the study. Their baseline characteristics are
summarized in Table 1. In the NDFAM sample,
APOE-Ɛ4 allele was significantly more frequent in the
CT group than in the CS group. No differences were
found between the subjects who were lost to follow-up
and those who completed the study (data not shown).
corded, on average, higher scores after than before
the intervention, with the exception of the Corsi test
scores in the CT group, which declined significantly
(Table 2). This trend toward higher scores was more
pronounced in the CS group, reaching statistical signif-
icance for the MMSE and MoCA scores. In the CT
group, only the MoCA score increased significantly.
In the CI sample, the behavior of the CS group was
similar to that recorded in the same arm in the
NDFAM sample, even though the MMSE and MoCA

Table 1 Baseline characteristics of two samples of individuals (cognitively healthy and cognitively impaired) receiving cognitive stimulation or sham stimulation

NDFAM CI

CS (n = 38) CT (n = 39) Test p CS (n = 22) CT (n = 22) Test p

Age in year, mean (SD) 73.8 (1.2) 73.8 (1.3) 0.16 0.870 74.0 (1.4) 74.3 (1.7) 0.58 0.56
Education in years, mean (SD) 7.7 (3.0) 7.6 (3.0) 0.10 0.920 7.6 (3.3) 7.3 (3.1) 0.33 0.75
Male, n (%) 17.0 (44.7) 12.0 (30.8) 0.240 17.0 (77.3) 14.0 (63.6) 0.51
APOE-Ɛ4, n (%) 7.0 (18.4) 16.0 (42.1) 0.045 6.0 (27.3) 5.0 (22.7) 0.99
Cognitive endpoints
MMSE, mean (SD) 28.2 (1.5) 27.9 (1.6) 0.86 0.390 25.8 (1.6) 25.7 (2.3) 0.19 0.85
MoCA, mean (SD) 23.2 (3.6) 23.2 (3.6) 0.02 0.980 19.6 (3.3) 19.0 (3.0) 0.62 0.53
Corsi test, mean (SD) 3.7 (0.9) 4.0 (0.8) 1.74 0.090 3.4 (1.2) 3.8 (0.8) 1.27 0.21

NDFAM, cognitively healthy individuals having at least one first-degree relative affected by dementia; CI, individuals with cognitive impairment
but no dementia; CS, cognitive stimulation; CT, sham intervention.
Test and significance levels (p) for the comparison of CS and CT are reported. For quantitative variables, unpaired t-test was applied, and for qual-
itative variables Fisher’s exact test.
Bold figures denote significant difference.

Table 2 Pre-treatment and post-treatment mean neuropsychological test scores in cognitively healthy and cognitively impaired subjects receiving
cognitive stimulation or sham stimulation

NDFAM CI

Mean (SD) Mean (SD)

Baseline Post t p Baseline Post t p

CS
MMSE 28.22 (1.52) 28.77 (1.41) 2.25 0.0300 25.83 (1.62) 26.50 (2.12) 2.39 0.027
MoCA 23.21 (3.56) 25.61 (3.10) 6.30 0.0001 19.60 (3.28) 20.73 (3.48) 2.87 0.009
Corsi test 3.67 (0.93) 3.75 (0.97) 0.45 0.6500 3.37 (1.19) 3.69 (0.75) 1.32 0.200
CT
MMSE 27.91 (1.60) 28.33 (1.80) 1.81 0.0790 25.71 (2.30) 26.54 (1.82) 2.39 0.026
MoCA 23.23 (3.57) 24.13 (3.34) 2.18 0.0360 19.00 (3.00) 20.23 (3.58) 3.53 0.002
Corsi test 4.02 (0.82) 3.69 (0.73) 2.24 0.0310 3.76 (0.79) 3.66 (1.14) 0.42 0.678

NDFAM, cognitively healthy individuals having at least one first-degree relative affected by dementia; CI, individuals with cognitive impairment
but no dementia; CS, cognitive stimulation; CT, sham intervention; Baseline, pre-treatment; Post, post-treatment; MMSE, Mini-mental state ex-
amination; MoCA, Montreal Cognitive Assessment.
Results of paired t-tests and significance levels (p) are also reported.
Bold figures denote significant difference.

Copyright # 2014 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2014
Cognitive stimulation in non-demented individuals
mean scores were, at both time points, lower than
those recorded in the NDFAM sample. The CT group
in the CI sample showed significant MMSE and MoCA
score improvements, whereas the improvements re-
corded by the CT subjects in the NDFAM sample were
significant only on the MMSE (Table 2). The CS and
CT groups in the CI sample displayed similar behavior.
Short-term efficacy of cognitive stimulation with respect
to sham stimulation
The mean net changes in MMSE and Corsi test scores
were not found to differ significantly between the CS
and CT arms in the NDFAM sample, whereas the CS
subjects recorded a significantly greater improvement
in the MoCA score with respect to the CT subjects
(Table 3): in detail, the CS group showed a net MoCA
score increase of almost 2.4 points, versus an increase
of less than one point in the CT group.
In the CI sample, the two intervention groups (CS
and CT) recorded similar, positive post-stimulation net
changes for all three neuropsychological tests (Table 3).
Because a different distribution of the well-known
cognitive decline risk factor APOE-Ɛ4 was found at
baseline in the NDFAM sample, the possible influence
of APOE-Ɛ4 carrier status on outcomes was investigated.
The subjects in the NDFAM sample showed a differ-
ent effect of CS according to APOE-Ɛ4 carrier status. In
the APOE-Ɛ4 non-carriers, the CS subjects showed a
significantly higher positive net change in MoCA score
than did the CT group and a more marked Corsi test
Table 3 Mean net changes (post-intervention minus baseline) in neuropsychological test scores following cognitive stimulation or sham stimulation in
cognitively healthy and cognitively impaired subjects
Mean net change (SD)
CSa CTb
NDFAM
MMSE 0.56 (1.46) 0.42 (1.52)
MoCA 2.39 (2.34) 0.90 (2.57)
Corsi test 0.08 (1.07) 0.33 (0.93)
CI
MMSE 0.68 (1.33) 0.83 (1.63)
MoCA 1.14 (1.86) 1.23 (1.63)
Corsi test 0.32 (1.13) 0.10 (1.14)
NDFAM, cognitively healthy individuals having at least one first-degree relative affected by dementia; CI, individuals with cognitive impairment
but no dementia; CS, cognitive stimulation; CT, sham intervention; MMSE, Mini-mental state examination; MoCA, Montreal Cognitive
Assessment; SD, standard deviation.
The results of unpaired t-tests with significance levels and effect sizes are also reported.
a
For NDFAM, n = 38; for CI, n = 22.
b
For NDFAM, n = 39; for CI, n = 22.
Bold figures denote significant difference.
Copyright # 2014 John Wiley & Sons, Ltd.
score improvement (Table 4). Conversely, in the
APOE-Ɛ4 carriers, no differences were found between
the CS and CT groups, even though the mean net
changes in the MMSE and MoCA scores were greater
in the CS group than in the CT group.
In the APOE-Ɛ4 non-carriers belonging to the CI
sample, no significant differences in cognitive test
score net changes were found between the CS and
CT subjects. Similarly, no treatment effect was found
in the APOE-Ɛ4 carriers (Table 4).
In the NDFAM sample, multiple regression analysis
failed to confirm a confounding effect of APOE-Ɛ4 carrier
status on the response to cognitive treatment as measured
using either the MoCA or the MMSE (Table 5); con-
versely, the APOE-Ɛ4 genotype was shown to modify the
response to cognitive treatment as measured using the
Corsi test (p for interaction = 0.019, data not shown;
Table 6). In detail, in the NDFAM sample, a significant
net improvement in the mean MoCA score was found
in the CS versus CT groups also after adjustment for
APOE-Ɛ4 and gender (Table 5), whereas in the APOE-
Ɛ4 non-carriers, the CS group showed a significant net
improvement in the Corsi test score compared with the
CT group irrespective of gender (Table 6). On the contrary,
in the APOE-Ɛ4 carriers, a mean net worsening in the Corsi
test score was found in the CS group compared with the
CT group, even though the difference was not significant.
Discussion
The findings of this study indicate that, at short-term
follow-up, cognitively stimulated subjects do not show
t p Effect size (Cohen’s d)
0.39 0.698 0.01
2.26 0.009 0.61
1.80 0.075 0.41
0.34 0.733 0.10
0.17 0.864 0.05
1.23 0.226 0.37
Int J Geriatr Psychiatry 2014
 L. Polito et al.

Table 4 Mean net changes (post-intervention minus baseline) in neuropsychological test scores following cognitive stimulation or sham stimulation in
cognitively healthy and cognitively impaired subjects stratified for APOE- 4 status

APOE-Ɛ4 non-carriers APOE-Ɛ4 carriers

Mean net change (SD) Mean net change (SD)

CSa CTb t p CSc CTd t p

NDFAM
MMSE 0.5 (1.5) 0.60 (1.4) 0.32 0.750 0.9 (1.9) 0.20 (1.6) 0.88 0.39
MoCA 2.5 (2.4) 1.20 (2.0) 2.06 0.045 2.0 (2.0) 0.30 (3.2) 1.30 0.21
Corsi test 0.2 (1.0) 0.60 (0.8) 3.02 0.004 0.4 (1.2) 0.06 (1.0) 0.75 0.46
CI
MMSE 0.7 (1.4) 1.00 (1.7) 0.61 0.540 0.7 (1.2) 0.20 (1.5) 0.55 0.60
MoCA 1.1 (2.0) 1.70 (1.4) 0.97 0.340 1.2 (1.5) 0.40 (1.5) 1.73 0.12
Corsi test 0.4 (1.3) 0.01 (1.1) 0.95 0.350 0.2 (0.7) 0.40 (1.3) 0.89 0.40

NDFAM, cognitively healthy individuals having at least one first-degree relative affected by dementia; CI, individuals with cognitive impairment
but no dementia; CS, cognitive stimulation; CT, sham intervention; MMSE, Mini-mental state examination; MoCA, Montreal Cognitive
Assessment; SD, standard deviation.
The results of unpaired t-tests and significance levels for the comparison of CS and CT are reported.
a
For NDFAM, N = 31; for CI, N = 16.
b
For NDFAM, N = 22; for CI, N = 17.
c
For NDFAM, N = 7; for CI, N = 5.
d
For NDFAM, N = 16; for CI, N = 6.
Bold figures denote significant difference.

Table 5 Multiple regression model: adjusted mean net changes in neuropsychological test scores following cognitive stimulation or sham stimulation in
cognitively healthy subjects

NDFAM

Adjusted mean (95%CI) t p

Net change MMSE

Stimulation (CS vs. CT) 0.06 [ 0.66, 0.78] 0.16 0.87
APOE-Ɛ4 (carriers vs. non-carriers) 0.13 [ 0.91, 0.65] 0.34 0.74
Sex (female vs. male) 0.23 [ 0.95, 0.49] 0.63 0.53
Net change MoCA
Stimulation (CS vs. CT) 1.39 [0.22, 2.57] 2.36 0.021
APOE-Ɛ4 (carriers vs. non-carriers) 0.72 [ 1.99, 0.55] 1.13 0.26
Sex (females vs. males) 0.13 [ 1.30, 1.04] 0.23 0.82

NDFAM, cognitively healthy individuals having at least one first-degree relative affected by dementia; CS, cognitive stimulation; CT, sham intervention;
MMSE, Mini-mental state examination; MoCA, Montreal Cognitive Assessment.
Bold figure denotes significant difference.

Table 6 Multiple regression model: adjusted mean net changes in Corsi test scores following cognitive stimulation or sham stimulation according to
APOE- 4 status

Net change Corsi test Adjusted mean 95%CI t p

APOE-Ɛ4 non-carriers
Stimulation (CS vs. CT) 0.77 [0.25, 1.30] 2.97 0.005
Sex (females vs. males) 0.18 [ 0.70, 0.35] 0.66 0.510
APOE-Ɛ4 carriers
Stimulation (CS vs. CT) 0.50 [ 1.59, 0.59] 0.96 0.350
Sex (females vs. males) 0.42 [ 1.47, 0.63] 0.83 0.420

CS, cognitive stimulation; CT, sham intervention.

Bold figure denotes significant difference.

Copyright # 2014 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2014
Cognitive stimulation in non-demented individuals
a major improvement in their MMSE score with respect to
controls. Conversely, in the cognitively healthy NDFAM
sample, the CS group showed a significantly higher net
cognitive gain as measured using the MoCA score and a
borderline higher net cognitive gain as measured using
the Corsi test, compared with the CT group. In the non-
demented individuals with CI sample, on the other hand,
no difference in MoCA and Corsi test performance was
found between the subjects receiving CS and the controls.
The failure of the MMSE score to detect differences
between CS and CT subjects, in either sample, is prob-
ably due to the well-known limitations of this test.
Although the MMSE is a widely accepted screening
tool for demented subjects, when used with high-
functioning individuals, it is less sensitive than the
MoCA in revealing mild cognitive modifications, espe-
cially in executive functioning (Nasreddine et al., 2005;
Tardif and Simard, 2011; Larner, 2012; Markwick et al.,
2012). Furthermore, the MMSE does not investigate
visuospatial short-term memory ability assessed by
the Corsi test. Our results confirm the aforementioned
higher sensitivity of MoCA compared with MMSE. We
are aware that our findings must be interpreted with
caution because the study could be not well dimen-
sioned for secondary endpoints.
Our short-term results indicate that CS is an effec-
tive treatment in NDFAM, but not in CI subjects. This
finding is in agreement with the conclusion of the
meta-analysis of the Cochrane collaboration group
(Martin et al., 2011). A possible explanation for this
difference is that our training programs addressed
cognitive functions with a comprehensive approach,
whereas subjects with CI might benefit more from in-
terventions focusing on the learning of specific infor-
mation relevant to the individual, as highlighted by a
recent systematic review (Stott and Spector, 2011).
Another hypothesis is that sample heterogeneity could
have influenced the lack of efficacy of CS compared with
CT in cognitive impaired group. CI group comprised
different types of CI, involving amnesic and non-
amnesic subjects. We can argue that in more homoge-
neous groups, different results could be found, but till
now, very few studies were focused on homogeneous
subtypes of MCI (Jean et al., 2010; Moro et al., 2012).
The “Allena-Mente” study has a planned follow-up
period that will allow us to evaluate what happens to
these subjects over time. It can be hypothesized that in
CI subjects, the benefits of CS will manifest themselves
through a slower cognitive decline with respect to that
shown by controls. Moreover, we will evaluate the sta-
bility of the effects observed in the NDFAM sample.
Another interesting result of this study is that the
subjects receiving the control intervention recorded
Copyright # 2014 John Wiley & Sons, Ltd.
increased performance levels 2 weeks after the intervention
ended. There are several possible explanations for this
finding. We may hypothesize a sort of “placebo” effect,
given that these patients knew they were receiving a cogni-
tive intervention but did not know that it was CT. It is also
possible that the lifestyle education received by these
patients contributed to their improved cognitive per-
formance, as it may have encouraged them to adopt
healthier behaviors during the intervention period.
The present study is, to our knowledge, the first to
report the influence of APOE-Ɛ4 carrier status on the
efficacy of CS, measured as a gain in cognitive perfor-
mance. Stratifying the subjects for APOE-Ɛ4 carrier
status, we found that the non-carriers in the cogni-
tively healthy group benefited from CS, as shown by
improved cognitive performance measured using both
the MoCA and the Corsi test, whereas the APOE-Ɛ4
carriers showed no significant improvement. The most
striking finding concerned the Corsi test. Although
borderline significant in the entire population, the
Corsi test score revealed a marked improvement in
cognitive performance after CS in the APOE-Ɛ4 non-
carriers belonging to the NDFAM sample. Multivariate
analysis suggested that the presence of the APOE-Ɛ4
allele could be an effect modifier for performance on
the Corsi test. A possible interpretation of the different
behaviors of APOE-Ɛ4 non-carriers and carriers with
respect to the Corsi test is that the CS protocol applied
in the present study particularly targeted attentional–
executive abilities; these include visual attention, which
is crucial for performing the Corsi test. Because the
APOE-Ɛ4 genotype is reported to affect visuospatial at-
tention (Greenwood et al., 2005), we suggest that the
presence of this allele could have made the CS inter-
vention less effective in stimulating this cognitive area
and that the Corsi test was able to register this.
For the other tests used to measure the effect of CS
(MMSE and MoCA), APOE-Ɛ4 was found to be nei-
ther an effect modifier nor a confounding factor. How-
ever, there emerged an interesting trend that, although
not statistically significant, deserves further assessment.
The net cognitive gain recorded by the CT group, seen
in the global analysis, was absent when analyzing only
the APOE-Ɛ4 carriers. Indeed, the APOE-Ɛ4 carriers
(both NDFAM and CI) belonging to the CT group re-
corded a net change value close to zero. This applied to
both the MMSE and MoCA results. A possible explana-
tion for this finding may lie in the concept of cognitive
reserve (i.e., the space between basic and maximum
cognitive function). We hypothesize that the two
interventions have different mechanisms of action de-
termining their efficacy. Whereas the CT mildly stimu-
lates basic cognitive abilities usually needed for the
Int J Geriatr Psychiatry 2014

autonomy in the everyday life, the CS intervention
probably exploits cognitive reserve to solve complex
and more intensive tasks. The presence of the APOE-
Ɛ4 allele could amplify this difference—APOE-Ɛ4 status
has been reported to influence cognitive functions
(Pettigrew et al., 2013)—with the result that carriers
and non-carriers benefit differently from experiences
that enhance cognition. This hypothesis could also ex-
plain why the NDFAM subjects were more likely to im-
prove after CS than after CT, whereas the CI subjects
recorded similar improvements after the two interven-
tions. It is possible that the CI group were less able to
engage cognitive reserve mechanisms to benefit from
more intensive and complex stimuli.
In summary, in a sample of cognitively healthy in-
dividuals, CS produced, in the short term, a higher
net gain in MoCA scores than did the control inter-
vention; however, it was less effective in individuals
with MCI. Finally, this work demonstrates, for the first
time, that there is a substantial genetic contribution to
CS efficacy: the APOE-Ɛ4 genotype was found to mod-
ify the efficacy of CS on visuospatial memory perfor-
mance measured using the Corsi test.
Conflict of interest
None declared.
Key points
•
Cognitive stimulation is an effective treatment in
cognitively healthy individuals, whereas it is less
effective in individuals with cognitive impairment.
• APOE-Ɛ4 status has a significant impact on the
efficacy of cognitive stimulation on visuospatial
memory.
Acknowledgements
The authors are grateful to “Federazione Alzheimer
Italia,” Milan, for supporting Golgi Cenci Foundation
research activities.
Copyright # 2014 John Wiley & Sons, Ltd.
L. Polito et al.
References
Alzheimer’s Association. 2010. 2010 Alzheimer’s disease facts and figures. Alzheimers
Dement 6(2): 158–194.
Belleville S, Gilbert B, Fontaine F, et al. 2006. Improvement of episodic memory in
persons with mild cognitive impairment and healthy older adults: evidence from
a cognitive intervention program. Dement Geriatr Cogn Disord 22(5-6): 486–499.
Cohen J. 1992. A power primer. Psychol Bull 112(1): 155–159.
Feuerstein R, Klein PS, TAJ. 1999. Mediated Learning Experience (MLE): Theoretical,
Psychosocial and Learning Implication. Freund Publishing House Ltd.: Tel Aviv.
Gates NJ, Sachdev PS, Fiatarone Singh MA, Valenzuela M. 2011. Cognitive and mem-
ory training in adults at risk of dementia: a systematic review. BMC Geriatr 11: 55.
DOI: 10.1186/1471-2318-11-55
Greenwood PM, Lambert C, Sunderland T, Parasuraman R. 2005. Effects of apolipo-
protein E genotype on spatial attention, working memory, and their interaction in
healthy, middle-aged adults: results From the National Institute of Mental Health’s
BIOCARD study. Neuropsychology 19(2): 199–211.
Guaita A, Colombo M, Vaccaro R, et al. 2013. Brain aging and dementia during the
transition from late adulthood to old age: design and methodology of the
“Invece.Ab” population-based study. BMC Geriatr 13: 98. DOI: 10.1186/1471-
2318-13-98
Jean L, Bergeron M-E, Thivierge S, Simard M. 2010. Cognitive intervention programs
for individuals with mild cognitive impairment: systematic review of the literature.
Am J Geriatr Psychiatry 18(4): 281–296. DOI: 10.1097/JGP.0b013e3181c37ce9
Larner AJ. 2012. Screening utility of the Montreal Cognitive Assessment (MoCA): in
place of—or as well as—the MMSE? Int Psychogeriatr 24(3): 391–396. DOI:
10.1017/S1041610211001839
Magni E, Binetti G, Bianchetti A, Rozzini R, Trabucchi M. 1996. Mini-mental state
examination: a normative study in Italian elderly population. Eur J Neurol 3(3):
198–202. DOI: 10.1111/j.1468-1331.1996.tb00423.x
Markwick A, Zamboni G, de Jager CA. 2012. Profiles of cognitive subtest impairment
in the Montreal Cognitive Assessment (MoCA) in a research cohort with normal
Mini-mental state examination (MMSE) scores. J Clin Exp Neuropsychol 34(7):
750–757. DOI: 10.1080/13803395.2012.672966
Martin M, Clare L, Altgassen AM, Cameron MH, Zehnder F. 2011. Cognition-based
interventions for healthy older people and people with mild cognitive impairment.
Cochrane Database Syst Rev (1): CD006220. DOI: 10.1002/14651858.CD006220.
pub2
Moro V, Condoleo MT, Sala F, et al. 2012. Cognitive stimulation in a-MCI: an exper-
imental study. Am J Alzheimers Dis Other Demen 27(2): 121–130. DOI: 10.1177/
1533317512441386
Nasreddine ZS, Phillips NA, Bédirian V, et al. 2005. The Montreal Cognitive Assess-
ment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr
Soc 53(4): 695–699.
Olazarán J, Muñiz R, Reisberg B, et al. 2004. Benefits of cognitive–motor inter-
vention in MCI and mild to moderate Alzheimer disease. Neurology 63(12):
2348–2353.
Petersen RC. 2004. Mild cognitive impairment as a diagnostic entity. J Intern Med
256(3): 183–194.
Pettigrew C, Soldan A, Li S, et al. 2013. Relationship of cognitive reserve and APOE
status to the emergence of clinical symptoms in preclinical Alzheimer’s disease.
Cogn Neurosci 4(3–4): 136–142. DOI: 10.1080/17588928.2013.831820. Epub 2013
Aug 25
Spinnler H, Tognoni G. 1987. Standardizzazione e taratura Italiana di test
neuropsicologici. Ital J Neurol Sci 8(6): 44–46.
Stott J, Spector A. 2011. A review of the effectiveness of memory interventions in mild
cognitive impairment (MCI). Int Psychogeriatr 23(4): 526–538. DOI: 10.1017/
S1041610210001973. Epub 2010 Oct 15.
Tardif S, Simard M. 2011. Cognitive stimulation programs in healthy elderly: A
review. Int J Alzheimers Dis 2011: 378934. DOI: 10.4061/2011/378934. Epub
2011 Aug 15
Willis SL, Tennstedt SL, Marsiske M, et al. 2006. Long-term effects of cognitive training
on everyday functional outcomes in older adults. JAMA 296(23): 2805–2814. DOI:
10.1001/jama.296.23.2805
Wolinsky FD, Unverzagt FW, Smith DM, et al. 2006. The effects of the ACTIVE
cognitive training trial on clinically relevant declines in health-related quality of
life. J Gerontol B Psychol Sci Soc Sci 61(5): S281–287.
Int J Geriatr Psychiatry 2014