INTRODUCTION
The coronavirus disease 2019 (COVID-19) pandemic required a return to the literature
in search of ways to stimulate the cognition of healthy elderly people who adhered
to distancing measures. Consequently, many situations involving interpersonal contact
were conducted using information and communication technologies (ICTs).[1]
In this scenario, in a systematic review[2] on the use of tablets by the elderly, cognitive benefits were reported in most of
the reviewed articles. In one of the studies,[3] which involved 16 elderly people, the authors concluded that electronic devices
provided results similar to those of standard cognitive training. In another study[4] testing the effect of the tablet on cognitive skills, which included 22 adults,
the results showed that engagement in the intervention was associated with an increase
in processing speed and acquisition of new skills. Thus, ICTs have been proven to
be useful to improve performance in several different cognitive domains.[2]
[3]
[4]
[5]
[6]
[7]
However, internet use is also linked to improved well-being, empowerment, autonomy,
and independence.[6] Thus, digital inclusion yields different benefits, but cognitive gains have been
the focus of a large number of studies.[6]
[7] During the use of digital tools, many processes and steps can benefit cognition:
when the correct procedure to perform a task is remembered, procedural memory is triggered,
for example; when tracking the information and actions performed, the working memory
is being activated; structuring action in the correct order requires activating executive
functions; locating relevant information on the screen requires visual perception;
information management takes place to assess what information is relevant; and the
attention process is necessary to focus on relevant items, ignoring irrelevant items
and stimuli.[6]
[7]
In this context, few studies[8]
[9]
[10] have assessed the subjective impacts among the elderly who use such tools and digital
platforms. Da Silva et al.[7] highlighted that the combination of cognitive training and other types of interventions
to promote health yields multiple benefits to the quality of life of the elderly,
such as improvement in metamemory and metacognition, lower risk of frailty, and possible
reduction in the risk of dementia. Therefore, the aim of the present study was to
determine the impact of computerized cognitive training on mood, frequency of forgetfulness,
memory complaints, and quality of life of elderly participants of USP 60 + , a program
for the elderly offered by Universidade de São Paulo in partnership with Instituto
Supera de Educação.
METHODS
Participants
A total of 66 elderly individuals enrolled in the computerized cognitive stimulation
workshop offered free of charge by the USP60+ program. They were randomized with an
allocation ratio of 1:1 into a training group (n = 33) and a control group (n = 33).
In the first semester, the training group received the intervention and, after six
months, with the end of the intervention, the control group was able to participate
in the online workshop. Comparative statistical tests confirmed that the sociodemographic
data of the two groups were similar; therefore, the groups were homogeneous before
the start of the intervention. All of them signed the Informed Consent Form and were
informed that during the study they could take part in any other study that could
potentially have an effect on their cognitive functions.
The sample size was calculated using the G*Power software, version 3.1, from which
an alpha significance level of 5% and an effect size of 0.5 were established, with
33 participants in each group, resulting in a power of the sample of 84%.
Inclusion criteria
The participants were required to be older than 60 years of age and enrolled in USP60 + .
Exclusion criteria
Individuals under 60 years of age, who were not enrolled in the USP60+ program and
did not own a cell phone, tablet or computer with internet access were excluded.
Supera Online digital platform
A French company (HAPPYneuron, Inc., Lyon, France) , through its partner Supera Online
(São José dos Campos, SP, Brazil), provided the computerized cognitive stimulation
test for 12 weeks for free analysis and evaluation. A web-based cognitive training
game platform, it provided participants with a virtual environment with didactic and
attractive explanations. Nine exercises developed by Happyneuron for cognitive training
were applied for practice during the present study. The brain training cognitive game
platform aims to stimulate various cognitive aspects, including memory, attention,
language, executive functions (reasoning, logical thinking), and visual and spatial
skills, with significant results in several international studies.[11]
[12]
[13]
[14] The platform in Portuguese is available on the following website ([Figure 1]): www.superaonline.com.br.[15]
Figure 1 Image from the Supera Online platform.
The program offers training for four categories of users: all age brackets; monitors;
coordinators; and managers. Each registered user is assigned a login and password
that provide unlimited access to the platform via computers connected to the internet
or using mobile devices, such as notebooks, tablets and cell phones.[11]
[12]
[13]
[14]
[15]
In 2016, the Supera Online Digital Platform underwent the process of cultural adaptation
and validation for use by mature and older Brazilian adults. The participants in this
validation process were 124 healthy mature and older adults who were users of social
and health services at the Integrated Center for Health and Education (Centro Integrado
de Saúde e Educação, CISE, in Portuguese) in the city of São Caetano do Sul, state
of São Paulo. The participants used the platform weekly during 90-minute meetings
over a 12-week period. The results of the study, which was conducted and published
by Ordonez et al.,[6] showed improvements in cognitive performance and mood among the participants.
The current version of the Supera Online training program consists of a digital platform
for cognitive stimulation, brain training, and promotion of quality of life. Based
on exercises and challenges in the form of digital games, the program develops and
stimulates the main cognitive functions: memory, attention, language, logical reasoning,
and visuospatial abilities, which contribute to longevity and improved quality of
life.[8]
[9]
[10]
There are twenty modalities of exercises/games grouped into blocks which train specific
cognitive abilities: memory, attention, logical reasoning, visuospatial ability and
language, which can optimize the cognitive development of participants through play-based
interactive activities. The teaching-learning process occurs in a systematized manner,
and it is monitored individually by a virtual instructor. Depending on individual
progress, users can successively move on to other, more difficult, stages of the tasks
programmed on the platform.[8]
[9]
[10]
Assessment protocol
Sociodemographic questionnaire
Through the sociodemographic questionnaire, we collected data on age, gender, marital
status, level of schooling, and healthcare provision.
Memory Complaints Questionnaire (MAC-Q)
The Memory Complaints Questionnaire (MAC-Q) is composed of six items related to memory
functioning in everyday activities. The higher the score, the greater the severity
of the memory-related complaints, and scores ≥ 25 indicate age age-associated impairment,
classifying the respondent as having a “negative” memory complaint.[16]
Frequency of Forgetfulness Scale (McNair and Kahn)
The Frequency of Forgetfulness Scale includes 15 questions on different situations
that characterize memory failures, with the following answer options: never (0 points);
sometimes (1 point); often (2 points); and always (3 points). The score ranges from
0 to 45 points, with higher scores indicating greater frequency of forgetting.[17]
Control, Autonomy, Self-realization, and Pleasure (CASP-19) questionnaire
The Control, Autonomy, Self-realization, and Pleasure (CASP-19) questionnaire comprises
19 items measuring perceived quality of life in individuals aged 55 years or older.
Each item has 4 response options: never (0 points); not often (1 point); sometimes
(2 points); and often (3 points). The score ranges from 0 to 57, and higher scores
indicate better perceived quality of life. A total of 6 items (1, 2, 4, 6, 8, and
9) are recorded as negative values and subsequently inverted in the data analysis.[18]
Geriatric Depression Scale (GDS)
The Geriatric Depression Scale (GDS) quantifies depressive symptoms among the elderly.
It consists of a 15-item scale with dichotomous responses: yes or no. Scores < 6 points
are defined as normal; from 6 to 10, as mild-to-moderate depression; and > 10, as
severe depresssion.[19]
[20]
Geriatric Anxiety Inventory (GAI)
The Geriatric Anxiety Inventory (GAI) is used to measure anxiety symptoms; it comprises
twenty descriptive statements of anxiety symptoms to be answered subjectively. The
total score, which ranges from 0 to 20 points, is the sum of the questions with answers
marked as “yes”. 20. Scores higher than 10/11 suggest the presence of generalized
anxiety disorder.[21]
Investigation venue
The intervention was carried out at the School of Arts, Sciences and Humanities of
Universidade de São Paulo, within a program called Open University for the Elderly,
which is currently referred to as USP60 + .
Data analysis
The Kolmogorov-Smirnov test documented the absence of a normal distribution between
continuous and ordinal variables, so we chose to use non-parametric statistics. Therefore,
for the descriptive statistics, tables were compiled containing absolute and relative
frequencies, as well as measures of position and dispersion (median, interquartile
range, minimum and maximum values). The groups were compared using the Chi-squared
test for the categorical variables. The Mann-Whitney U test was used for the analysis
of the continuous variables, and the Wilcoxon test were used for the analysis of paired
samples,[22]
[23] both followed by the effect size.
Finally, multivariate logistic regression was used to analyze the association regarding
the groups and the deltas of the total scores (postest–pretest) of the scales used.
All analyses were performed using the Jeffreys's Amazing Statistics Program (JASP,
open source), version 0.16.3. The significance level adopted for the statistical tests
was of 5%, that is, p < 0.05. In addition to the p-value, we used the Vovk-Sellke maximum p-ratio: based on a two-sided p-value, the maximum possible odds in favor of H1 over H0.[22]
[23]
Ethical aspects
The present study was submitted to the Ethics Committee for Research in Humans of
the Teaching Hospital of the School of Medicine at Universidade de São Paulo The study
was approved under no. 4.357.429.
RESULTS
Regarding the profile of the 66 older adults interviewed, the sample comprised predominantly
female participants, aged between 60 and 92 years, who were married or separated/divorced,
and whose level of schooling was higher (39.39%) or postgraduate (42.42%) education.
Most of the interviewees had private healthcare insurance and were retired. The participants
were divided into two groups (control and training) and, as shown in [Tables 1] and [2], were closely matched statistically in terms of sociodemographic and psychosocial
data. The similar profiles enabled the measurement of changes between baseline and
postintervention values in the training group.
Table 1
Comparison of the sociodemographic data of the study groups
|
Variable
|
Total
|
Training group
|
Control group
|
P-value
|
|
n = 66
|
%
|
n = 33
|
%
|
n = 33
|
%
|
|
Sex
|
Female
|
60
|
90.91
|
30
|
90.91
|
30
|
90.91
|
0.937a
|
|
Male
|
6
|
9.09
|
3
|
9.09
|
3
|
9.09
|
|
Age
|
Median (interquartile range)
|
66.00 (6.75)
|
66.00 (5.00)
|
66.00 (8.00)
|
0.893b
|
|
Minimum-maximum
|
60.00-92.00
|
61.00-81.00
|
60.00-92.00
|
|
Level of schooling
|
Median (interquartile range)
|
17.00 (4.00)
|
17.00 (3.00)
|
16.00 (5.00)
|
0.343b
|
|
Minimum-maximum
|
8.00-23.00
|
11.00-22.00
|
8.00-23.00
|
|
Retired
|
Yes
|
62
|
93.94
|
32
|
96.97
|
30
|
90.91
|
0.332a
|
|
No
|
4
|
6.06
|
1
|
3.03
|
3
|
9.09
|
Notes: aChi-squared test; bMann-Whitney U-Test.
Table 2
Comparison of group performance pre- and postintervention
|
Variables and statistics
|
General (n = 66)
|
Training group (n = 33)
|
Control group (n = 33)
|
Mann-Whitney test
|
P-value
|
Effect size
|
VS-MPR
|
|
Median
|
IQR
|
Median
|
IQR
|
Median
|
IQR
|
|
MAC-Q (pre)
|
24.50
|
4.00
|
24.00
|
4.00
|
25.00
|
3.00
|
564.500
|
0.801
|
0.037
|
1.000
|
|
MAC-Q (post)
|
24.00
|
3.75
|
22.00
|
6.00
|
25.00
|
2.00
|
266.000
|
< 0.001
|
-0.511
|
137.170
|
|
Wilcoxon test
|
974.000
|
384.500
|
114.500
|
|
|
|
|
|
P-value
|
0.151
|
0.002
|
0.123
|
|
|
|
|
|
Effect size
|
0.221
|
0.654
|
-0.348
|
|
|
|
|
|
VS-MPR
|
1.288
|
32.787
|
1.427
|
|
|
|
|
|
McNair (pre)
|
8.50
|
4.00
|
9.00
|
3.00
|
8.00
|
4.00
|
680.500
|
0.080
|
0.250
|
1.820
|
|
McNair (post)
|
8.00
|
4.75
|
8.00
|
4.00
|
9.00
|
5.00
|
487.500
|
0.467
|
-0.105
|
1.000
|
|
Wilcoxon test
|
1063.000
|
377.000
|
181.500
|
|
|
|
|
|
P-value
|
0.107
|
< 0.001
|
0.441
|
|
|
|
|
|
Effect size
|
0.243
|
0.733
|
-0.166
|
|
|
|
|
|
VS-MPR
|
1.540
|
96.086
|
1.000
|
|
|
|
|
|
CASP-19 (pre)
|
30.50
|
7.75
|
32.00
|
8.00
|
30.00
|
5.00
|
658.500
|
0.145
|
0.209
|
1.315
|
|
CASP-19 (post)
|
31.00
|
6.00
|
32.00
|
6.00
|
30.00
|
6.00
|
653.000
|
0.165
|
0.199
|
1.238
|
|
Wilcoxon test
|
794.500
|
212.500
|
191.500
|
|
|
|
|
|
P-value
|
0.496
|
0.688
|
0.580
|
|
|
|
|
|
Effect size
|
-0.102
|
-0.086
|
-0.120
|
|
|
|
|
|
VS-MPR
|
1.000
|
1.000
|
1.000
|
|
|
|
|
|
GDS-15 (pre)
|
3.00
|
2.00
|
3.00
|
2.00
|
2.00
|
2.00
|
594.500
|
0.518
|
0.092
|
1.000
|
|
GDS-15 (post)
|
2.00
|
2.75
|
2.00
|
2.00
|
2.00
|
3.00
|
531.500
|
0.870
|
-0.024
|
1.000
|
|
Wilcoxon test
|
919.000
|
1.524
|
232.00
|
|
|
|
|
|
P-value
|
0.067
|
0.124
|
0.299
|
|
|
|
|
|
Effect size
|
0.284
|
0.342
|
0.228
|
|
|
|
|
|
VS-MPR
|
2.026
|
1.422
|
1.019
|
|
|
|
|
|
GAI (pre)
|
3.00
|
4.00
|
3.00
|
6.00
|
4.00
|
3.00
|
545.500
|
0.995
|
0.002
|
1.000
|
|
GAI (post)
|
2.00
|
4.00
|
2.00
|
4.00
|
2.00
|
4.00
|
536.000
|
0.917
|
-0.016
|
1.000
|
|
Wilcoxon test
|
1072.000
|
275.500
|
272.500
|
|
|
|
|
|
P-value
|
0.011
|
0.037
|
0.113
|
|
|
|
|
|
Effect size
|
0.393
|
0.458
|
0.342
|
|
|
|
|
|
VS-MPR
|
7.590
|
3.013
|
1.495
|
|
|
|
|
Abbreviations: CASP-19, Control, Autonomy, Self-Realization, and Pleasure questionnaire;
GAI, Geriatric Anxiety Inventory; GDS-15, Geriatric Depression Scale; IQR, interquartile
range; MAC-Q, Memory Complaints Questionnaire; McNair, McNair and Kahn's Frequency
of Forgetfulness Scale; VS-MPR, Vovk-Sellke maximum p-ratio.
The analysis of the psychosocial data comparing the total scores on the scales showed
that the training group scored lower on the MAC-Q, the Frequency of Forgetfulness
Scale, and the GAI. Moreover, there was a significant difference between the groups
in the total score on the MACQ postintervention ([Table 2]).
Finally, a multivariate analysis was performed with the aid of logistic regression,
in which the groups were categorized (training group = 1; and control group = 0),
enabling the association based on the deltas of the total scores (posttest–pretest)
of the scales used ([Table 3]). The logistic regression model generated was statistically significant (χ2[64] = 14,310; p < 0.001), and it correctly classified 73.0% of the cases of participants who had
few memory complaints as belonging to the training group.
Table 3
Model summary of the logistic regression
|
Model
|
Deviance
|
AIC
|
BIC
|
df
|
ΔΧ2
|
P-value
|
McFadden R2
|
Nagelkerke R2
|
Tjur R2
|
Cox and Snell R2
|
|
1
|
91.495
|
93.495
|
95.685
|
65
|
|
|
0.000
|
|
0.000
|
|
|
2
|
77.186
|
81.186
|
85.565
|
64
|
14.310
|
< .001
|
0.156
|
0.260
|
0.199
|
0.195
|
|
3
|
73.694
|
79.694
|
86.263
|
63
|
3.492
|
0.062
|
0.200
|
0.315
|
0.244
|
0.236
|
Abbreviations: AIC, Akaike Information Criteria; BIC, Bayesian Information Criteria.
Notes: AIC and BIC are measures of fit for the model; the best model will have the
lowest AIC and BIC values. Four pseudo-R2 values are calculated in the JASP: McFadden, Nagelkerke, Tjur, and Cox and Snell.
These are analogous to R2 in linear regression, and all yield different values. What constitutes a good R2 value varies; however, they are useful when comparing different models for the same
data. The model with the largest R2 statistic is considered the best.
Table 4
Coefficients of the logistic regression
|
Model
|
Parameter
|
Estimate
|
Robust standard error
|
Z
|
Wald test
|
VS-MPR
|
|
Wald statistic
|
df
|
P-value
|
|
1
|
(Intercept)
|
0.000
|
0.246
|
3,33E-13
|
1,11E-28
|
1
|
1.000
|
1.000
|
|
2
|
(Intercept)
|
-0.207
|
0.290
|
-0.712
|
0.536
|
1
|
0.476
|
1.000
|
|
MAC-Q
|
-0.260
|
0.085
|
-3.069
|
10.179
|
1
|
0.002
|
27.859
|
|
3
|
(Intercept)
|
-0.268
|
0.281
|
-0.952
|
0.840
|
1
|
0.341
|
1.003
|
|
MAC-Q
|
-0.225
|
0.088
|
-2.559
|
7.223
|
1
|
0.010
|
7.697
|
|
McNair
|
-0.155
|
0.087
|
-1.774
|
3.085
|
1
|
0.076
|
1.878
|
Abbreviations: MAC-Q, Memory Complaints Questionnaire; McNair, McNair and Kahn's Frequency
of Forgetfulness Scale; VS-MPR, Vovk-Sellke maximum p-ratio.
Note: Training group coded as class 1.
DISCUSSION
The objective of the present study was to determine the subjective impact of twelve
weeks of computerized cognitive training on mood, frequency of forgetfulness, memory
complaints, and quality of life of active elderly people enrolled in a program for
the elderly in the city from São Paulo. The results of the postintervention assessments
and statistical analyses showed improvements in the quality of life of participants
in the training group. In addition, there was an improvement in depressive and anxious
symptoms, as well as a reduction in memory complaints and related forgetfulness.
These findings corroborate a systematic review[2] on computerized cognitive training, which found a significant improvement in the
cognitive and psychological characteristics of healthy older adults. Likewise, the
study by Ordonez et al.,[6] involving cognitive training with computer games in the elderly, also observed positive
effects in the assessment of memory and quality of life.
Thus, the results of the present study pointed to improvements in quality of life,
suggesting that training cognitive functions using an online platform can alleviate
forgetfulness and memory complaints, leading to improved mental and physical well-being.
Menascu et al.[10] suggest that cognitive game training has a beneficial effect on cognitive performance
in multiple sclerosis patients with mild cognitive impairment. However, further evaluation
is needed to assess the longevity of this effect.
In another review[24] on computerized cognitive training combined with psychoeducation and pen-and-paper-based
cognitive training, effects were found for general cognition, working memory, attention,
learning, and depressive symptoms; but, in older adults with mild cognitive impairment,
there was a lack of effectiveness for non-verbal memory, executive function, processing
speed, and visuospatial skills.
Previous studies[6]
[7] on older patients showed that, in terms of frequency and time of computer use or
exposure to different forms of technology, age-related differences were not as great
as expected. Among the elderly, individuals showed improvement in some mental skills,
including memory, attention, executive functions, and information processing speed.[6]
[7]
In conclusion, the Supera Online Digital Platform offers a range of activities that
can help improve cognitive skills, representing an alternative for people with reduced
mobility. Other advantages of the program are the lower cost of transport and the
convenience of training at home, which contributes to increase the engagement and
digital inclusion of the participants. Finally, there is a lack of computerized cognitive
training studies with older adults evaluating the subjective effects of these interventions
on mood, depression, anxiety, and memory complaints. Future investigations should
be conducted to build on the results of the present study.
