Abstract
Introduction
Various neurological and neuropsychiatric conditions, such as multiple sclerosis, Parkinson’s disease, and dementias are frequently characterized by important deficits that affect patients’ skills and on their quality of life (Borghi et al., 2016; Cavallo, Enrici, & Adenzato, 2011; Enrici et al., 2015; Ostacoli et al., 2013). Alzheimer’s disease (AD) typically includes significant neuropsychological deficits such as memory problems, frequently associated to other cognitive deficits such as aphasia, apraxia, and/or agnosia, which significantly interfere with everyday life (McKhann, Knopman, et al., 2011).
In recent years, cognitive training in AD has started to show its possible benefits, mainly in the realm of memory (Cavallo, Cavanna, et al., 2013; Cavallo, Hunter, van der Hiele, & Angilletta, 2016; Clare, Wilson, Carter, Hodges, & Adams, 2001). Converging evidence clearly indicates that some cognitive subsystems (e.g., procedural memory) remain relatively intact (Cavallo & Angilletta, 2018), whereas others (e.g., episodic memory) are dramatically impaired (Pause et al., 2013; Salmon & Bondi, 2009). These dissociations are indeed supported also by a developing understanding of the role played by different brain areas in the cognitive processes of memory encoding, storing, and retrieval, in both normal and pathological conditions (Pagani & Cavallo, 2014).
Computerized cognitive training has started to show interesting evidence in this clinical domain. In one of the first studies on this topic, Gaitán et al. (2013) investigated the effect of a computer-based cognitive training (CBCT) program, adjunctive to traditional cognitive training (TCT) based on pen-and-paper exercises. Patients in the combined treatment group (CBCT + TCT) showed less anxiety symptoms and less disadvantageous choices in decision making than the TCT group at 12 months. In another study, Lee, Yip, Yu, and Man (2013) investigated the effects of a computerized errorless learning-based memory training program (CELP) for persons with early AD, and compared its outcomes with those of a therapist-led errorless learning program (TELP) group and a waiting-list control group. However, the small group size was a significant limit of this study. More recently, a systematic review of the literature (Coyle, Traynor, & Solowij, 2015) focused on CCT and virtual reality cognitive training (VRCT) for individuals at high risk of cognitive decline. General limitations of the studies included the need to improve study design by including larger samples, to apply longitudinal designs, and to assess the wider effect of cognitive training on cognitive decline.
To avoid these limitations, in our recent Randomized Controlled Trial (RCT) in the field (Cavallo, Hunter, van der Hiele & Angilletta, 2016), we recruited a large group of early stage AD patients (N = 80), and randomly assigned them to two groups: an experimental group (n = 40), undergoing a computerized structured cognitive training using the rehabilitation software Brainer©, which had been fruitfully used in an our previous study (Cavallo, Trivelli, et al., 2013) and included memory, attention, executive function, and language tasks of increasing difficulty and tailored on patient’s performance, and a control group (n = 40), undergoing a computerized general cognitive intervention encompassing different exercises (such as reading online newspaper articles and discussing them with the neuropsychologists, navigating websites of interest, etc.). Interestingly, we were able to demonstrate a significant and stable positive effect of the computerized training on several neuropsychological tests, such as digit span forward and backward, two-syllable words test, Rivermead Behavioural Memory Test (RBMT), and Brixton test. However, to date, little is known about the effects once the training had been interrupted. Thus, in the present study, we monitored over time patients’ neuropsychological profiles 6 and 12 months after the end of the intervention.
Method
Participants
The present study involved patients with early stage AD included in our previous RCT (Cavallo, Hunter, van der Hiele & Angilletta, 2016). All patients received a diagnosis of early stage AD within 12 months prior the beginning of the study. Diagnosis was made by independent neurologists by considering patients’ medical history, screening of cognitive and functional measures, degree of cerebral atrophy, and symptoms reported by patients’ caregivers. Exclusion criteria were the following: the additional presence of other neurological and/or psychiatric disorders such as traumatic brain injuries, strokes or psychosis, a positive history of alcohol or drug abuse, the presence of any significant general health comorbidities (e.g., diabetes or hypertension), and the presence of significant sensorial impairments and/or extremely severe communication problems that could seriously compromise both the administration of cognitive tests and the interpretation of the relative results, and the implementation of the computerized training. Before the beginning of the present study, a comprehensive clinical assessment, including neurological examination, neuropsychological assessment, and consecutive brain magnetic resonance imaging (MRI) scans, was arranged by consultant neurologists, who confirmed a diagnosis of early stage probable AD, according to standard National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) diagnostic criteria (McKhann, Drachman, et al., 1984; McKhann, Knopman, et al., 2011) and after the exclusion of possible neuropsychiatric confounders. They were consecutively recruited over 3 years (from January 2011 until October 2013) in the Assisted Health Residence “Ville Roddolo” (Moncalieri, Italy). The study was granted approval by the local Research Ethics Committee. Informed written consent was obtained from all patients and from their caregivers.
Neuropsychological assessment
All participants underwent detailed neuropsychological assessments by experienced neuropsychologists blinded to patients’ allocation before training, after training, after 6 months, and after 12 months, to monitor their profile over time.
Neuropsychiatric assessment
Emotional disturbances were investigated by administering the Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983).
Computerized training (experimental group)
Each patient received an individual computerized training (three 30-min sessions per week, for 12 consecutive weeks). The software Brainer© (https://www.brainer.it/) was used (see Cavallo, Hunter, van der Hiele & Angilletta, 2016, for details).
Control cognitive intervention (control group)
Patients in the control group followed the same frequency of sessions (three 30-min session per week, for 12 weeks) together with a neuropsychologist. During the sessions, a computer connected to the Internet was used and the patient was free to choose to read electronic newspaper articles and discuss them with the neuropsychologist, or to play games and solve puzzles, and/or reach sites and contents of interest for him or her.
Statistical Analyses
Descriptive statistical analyses were performed using IBM SPSS Statistics (Statistical Package for the Social Sciences) version 22.0. As the graphical and statistical exploration of the data by means of box plots, histograms, Q-Q plots, and normality tests indicated normal distributions, parametric tests were used. To investigate the effect of training on neuropsychological performance, a Repeated Measures General Linear Model (RM-GLM) was run by considering participants’ performance on each test at the four assessment points (i.e., T0, T1, T2, and T3) as the within-subjects variable, and “group” (i.e., patients and controls) as the between-subjects factor. Last, single-case analysis via modified t test (Crawford & Garthwaite, 2002) was conducted too, to investigate computerized training longitudinal effects also at individual patient’s level. A p value < .05 was considered statistically significant throughout the analyses.
Results
Two AD patients per group were missing at the 6-month assessment and four patients per group were not present at the 12-month follow-up, as their families preferred to move them from our Health Assisted Residence to another one closer to their places. As a result, 36 out of 40 patients for both group were available at the 12-month follow-up. To investigate the possible longitudinal effect of treatment on neuropsychological measures, an RM-GLM was run by considering participants’ performance on each test at the four assessment points (before training, after training, and at the 6-month and 12-month follow-ups) as the within-subjects variable, and “group” (i.e., experimental vs. control groups) as the between-subjects factor. The p values were adjusted for multiple comparisons using the false discovery rate approach. There was a statistically significant interaction between time and group on patients’ performance on the following tests: digit span forward, F(3, 70) = 2.841, p = .03, d = .43, and backward, F(3, 70) = 3.258, p = .02, d = .48; two-syllable words test, F(3, 70) = 3.874, p = .004, d = .54; RBMT story immediate, F(3, 70) = 2.981, p = .03, d = .45; RBMT story delayed, F(3, 70) = 3.541, p = .003, d = .47; Token test, F(3, 70) = 4.879, p = .001, d = .57; and Brixton test, F(3, 70) = 7.245, p < .001, d = .64. Simple main effects analysis showed that patients’ cognitive performance in the experimental group was more influenced positively by the training than patients in the control group both at the posttreatment assessment and at follow-ups. Table 1 reports patients’ neuropsychological assessments at baseline and at the 12-month follow-up. Figure 1 reports groups’ performances on two of these tasks (e.g., digit span forward and Brixton tests).
Patients’ Performance on Neuropsychological Measures Before the Cognitive Training, and at the 12-Month Follow-Up.
Note. MMSE = Mini-Mental State Examination; RBMT = Rivermead Behavioural Memory Test; GNT = Graded Naming Test; VOSP = Visual Object and Space Perception Battery.
p < .05.
Groups’ performance at the four assessment points (before, after, follow-ups) on the (a) digit span forward and (b) Brixton tests.
To investigate neuropsychological individual changes in the treatment group versus controls’ scores, single-case analyses were performed using the procedure formalized by Crawford and Garthwaite (2002) to deal with single cases in cognitive neuropsychology appropriately. More precisely, modified t tests were used to determine whether each individual’s performance at follow-up was significantly better than the corresponding control group’s scores for the seven cognitive tasks that were significantly influenced by the computerized training. This is considered a very conservative approach, and aims at making it more difficult (and then more reliable) to reject the null hypothesis of absence of differences between a single patient and a control group. Interestingly, the vast majority of patients belonging to the experimental group got better test scores at follow-up, as compared with control group. More precisely, for each neuropsychological tests of interest, the number of patients showing a better performance at follow-up was the following: for digit span forward, 30/36; for digit span backward, 32/36; for two-syllable words test, 32/36; for RBMT story immediate, 30/36; for RBMT story delayed, 32/36; for Token test, 33/36; and for Brixton test, 35/36.
Discussion
Computerized cognitive training has started to show interesting evidence, even if at this point in time evidence in favor of it still remains weak and in need of more robust studies. The two important issues at hand pertain to the possibility for patients to acquire new procedural skills, and to maintain them as long as possible once the training comes to an end. In the present study, we investigated the stability of the effects of a previously administered computerized training over a long period of time (12 months after the end of the intervention). To take into account the possibility that cognitive performance was influenced by neuropsychiatric factors, we investigated the presence of possible differences between groups in terms of levels of anxiety and depressive symptoms at each assessment point, as measured by the HADS. We were able to see a significant effect of the structured computerized cognitive intervention on different neuropsychological measures. More precisely, as throughout the computerized intervention main cognitive functions typically compromised in AD were repeatedly trained to become more effective, we were able to show a significant improvement at the posttreatment and follow-up neuropsychological assessments. Interestingly, this improvement was maintained at the 6-month follow-up, but tended to decrease at the 12-month follow-up by reaching the baseline performance. In other words, it is like the training allowed patients in the experimental group to contrast the decay of cognitive functioning associated to AD, by delaying for 1 year the worsening of patients’ performance. Due to the use of the same medium (computer), and as the length and duration of training sessions were comparable in both groups, we were confident that our results depended mainly on the different interventions implemented, and not on other factors. However, it is not possible to rule out completely the possibility that other factors played a role in this sense. Further research should specifically address this important point.
To the best of our knowledge, our study is one of the first monitoring the effects over time of the computerized training implemented.
In conclusion, this study suggests an acceptable stability of computerized training in AD, even if the main lesson learned highlights the clinical need of a continuation of the training to maintain its positive effects on patients’ neuropsychological profile. At the same time, it is still important to take into account also patients’ quality of life, to design and realize clinical interventions tailored on patients’ needs and aimed at improving as much as possible their activities of daily living and ultimately their personal satisfaction.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.