Effects of computerized cognitive training on cognitive function,activity,and participation in individuals with stroke: A randomized controlled trial

Abstract

BACKGROUND:

Computerized cognitive training (CCT) is an emerging alternative intervention for stroke survivors.

OBJECTIVE:

This study investigated the effects of CCT on the cognition, activity, and participation of stroke survivors and compared the findings with those of match-dosed conventional cognitive training.

METHODS:

This randomized controlled trial included 39 patients with stroke who were divided into the intervention group (n = 19; receiving CCT with Lumosity software) and the control group (n = 20; receiving conventional cognitive training). Both the groups were trained for 20 min, twice a week, for 12 weeks. Participants were evaluated at pretest, posttest, and 4-week follow-up. Outcome measures included various cognitive function tests and the Stroke Impact Scale scores.

RESULTS:

The CCT group exhibited significant improvement in global cognitive function (evaluated using the Mini-Mental State Examination and Montreal Cognitive Assessment) and specific cognitive domains: verbal working memory (backward digit span test), processing speed (Symbol Digit Modalities Test), and three MoCA subtests (attention, naming, and delayed recall). CCT exerted no significant effect on activities and participation. No significant between-group differences in changes in cognitive function were noted. However, CCT significantly improved cognitive function domains immediately after training, and these effects were sustained at the 4-week follow-up.

CONCLUSIONS:

Cognitive function of individuals with chronic stroke could improve after administration of CCT. However, future studies with a more rigorous design and higher training dose are warranted to validate our findings.

Keywords

cerebral vascular accident computerized cognitive training cognitive function quality of life

1 Introduction

Cognitive impairment is a severe clinical deficit after stroke. Approximately 30% of stroke survivors may experience vascular cognitive impairment 3 months after the onset of stroke (Kalaria & Ballard, 2001). Vascular cognitive impairment is the second leading cause of dementia and manifests as attention deficiency, slow information processing, memory impairment, and executive dysfunction (Barba et al., 2000; Wong et al., 2013). Cognitive impairment can considerably impair stroke survivors’ balance, daily life functioning, and community participation (Hyndman & Ashburn, 2003; Lipskaya-Velikovsky et al., 2018; Spitzer et al., 2011).

Increasing attention is being focused on restoring and reinforcing cognitive capacities through cognitive training. Many cognitive training programs have been progressively integrated into computerized training and provided through technological devices, such as smartphones and tablets, because computerized training is considered a lower-cost alternative than conventional cognitive training using paper-and-pencil or table-top tasks (König et al., 2017). Computerized cognitive training (CCT) programs use brain training exercises focused on improving specific cognitive skills and generally allow the adjustment of task difficulty according to individual performance (Hill et al., 2017). Although companies providing CCT claim its effectiveness, CCT programs exhibited limited generalization effects (Simons et al., 2016). Relevant studies on brain training examining its near- and far-transfer effects have indicated that CCT programs mostly produce near-transfer effects (improvement in tasks that are similar to training tasks), especially on closely related skills such as visuomotor processing and memory and processing speed in the age-related population. However, little evidence is available for far-transfer effects (improvement in tasks that are dissimilar to training tasks in nature or in the real world) (Nguyen et al., 2021; Simons et al., 2016). The aforementioned findings indicate that most of the CCT programs cannot result in broad changes in cognition, raising suspicion regarding the efficacy of the programs.

Near-transfer improvements in cognition may be beneficial for stroke survivors. Stroke commonly causes cognitive impairment associated with more specific deficits in areas such as attention, working memory, and executive function (Cumming et al., 2013). These conditions might be more responsive to targeted CCT interventions; hence, stroke survivors can benefit from specific cognitive training. Consistent with the aforementioned findings, recent systematic reviews have demonstrated that CCT can lead to near-transfer improvements in domain-specific cognitive processes, such as information processing speed, executive function, memory, and attention, in stroke survivors (Bogdanova et al., 2016; Sigmundsdottir et al., 2016; van de Ven et al., 2016). Although studies have investigated the cognitive function outcomes of CCT in patients with stroke, outcomes related to far-transfer improvements in activity and participation remain underexplored (Rogers et al., 2018; Sigmundsdottir et al., 2016). Cognitive impairment negatively affects independence in the basic activities of daily living (ADLs) and instrumental ADLs (IADLs) in patients with chronic stroke (Lipskaya-Velikovsky et al., 2018). Therefore, the effect of CCT on the activities and participation of patients with stroke must be investigated.

This study examined the effects of a multidomain CCT program, namely Lumosity cognitive training, on patients with stroke. Lumosity cognitive training is a commercially available CCT program designed to improve cognitive function in the general population (Hardy et al., 2015). Although a study reported specific controversial findings for Lumosity cognitive training (Simons et al., 2016), CCT provides more advantages compared with conventional cognitive training. CCT requires a lower proportion of trained workforce, can provide immediate feedback through the use of software, and involves user performance– specific cognitive challenges and interesting training modules for better motivation and participation (Kueider et al., 2012). A previous study reported that Lumosity cognitive training exerted no effect on the cognitive function, quality of life, and self-efficacy of patients with chronic stroke and subjective cognitive impairment after training and at follow-ups, except for near-transfer effects on working memory and speed (Wentink et al., 2016). However, this study included patients who received usual rehabilitation into the passive control group, which only corrected for retest effects and spontaneous recovery but not for placebo effects. Drawing conclusions regarding the nature of any effects without including appropriate controls would be difficult. Thus, studies examining the effects of CCT on patients with stroke must include an active control group.

This study investigated whether the multidomain Lumosity cognitive training program improves the cognitive function and activity and participation of stroke survivors. The findings were compared between stroke survivors receiving CCT and those receiving conventional cognitive training (the active control group). The study results can be used as an empirical basis for cognitive rehabilitation after stroke.

2 Methods

2.1 Design

In this randomized controlled study, following baseline assessment, eligible participants were randomly assigned to the experimental group (who received CCT) or the control group (who received conventional cognitive training). The participants were allocated to the experimental or control group by using a simple randomization method that involved the use of sealed opaque envelopes. The participants were evaluated at the pretest, posttest, and 4 weeks after the end of the interventions (i.e., the follow-up assessment). This study was approved by the Institutional Review Board (KMUH-IRB-20130213), and written informed consent was obtained from all the participants before study commencement.

2.2 Participants

We recruited volunteers through verbal announcements from the department of rehabilitation at a municipal hospital located in southern Taiwan. Inclusion criteria were as follows: (1) being aged between 45 and 85 years, (2) being a stroke survivor, (3) having mild to moderate cognitive impairment, as determined on the basis of the Mini-Mental State Examination (MMSE) scores of 20– 26 (Crum et al., 1993; Petersen, 2004), and (4) being able to follow the therapist’s instructions. Exclusion criteria were as follows: (1) having receptive or expressive aphasia, which may affect the ability to perform cognitive training tasks or outcome measures, and (2) being unable to perform cognitive training tasks due to a visual perception problem (line bisection test score < 7) (Wilson et al., 1987).

According to a previous study investigating the effects of CCT on older adults (Kueider et al., 2012), we calculated the sample size for a statistical power of 80%, an alpha value of 0.05, and a moderate to large effect size (Cohen’s d = 0.8). A sample size of 20 participants per group was determined.

2.3 Intervention

The intervention group received CCT, which was administered through a touch-screen computer equipped with Lumosity software. Lumosity (www.lumosity.com) is a neuropsychological software product particularly designed for the various domains of cognitive functions. This software enables users to perform various cognitive activities and provides immediate feedback on the users’ performance, allowing the user to participate in challenging activities that match their capabilities at their own pace of progress. Each participant attended 20 min CCT sessions twice a week for 12 weeks (i.e., 24 sessions in total). A total of 10 games suitable for the participants’ difficulty level, culture, and language were selected from Lumosity. The name of games corresponding to domains were as follows: information processing speed (speed match, spatial speed match, and speed pack), attention (lost in migration, space junk, start search, and train of thought), and memory (memory matrix, rotation matrix, and money comb). To balance the training exposure to these domains, each training session consisted of three to four games covering all the three domains. Lumosity changes the difficulty level of games according to the participant’s performance. The training was conducted using a one-on-one approach, with an occupational therapist providing encouragement and addressing technical problems when required.

The use of a touch-screen computer markedly reduced the difficulty of using a mouse and keyboard and increased the participants’ acceptance. Moreover, the touch-screen computer enabled the participants to smoothly play games by using their nondominant hand, resulting in the enhanced adoption and feasibility of CCT for the stroke survivors.

The control group received conventional cognitive training involving paper-and-pencil tasks and table-top tasks, such as board games, puzzles, card games, and memory games, aimed at improving attention, memory, information processing speed, and executive function. The therapist selected the appropriate game according to the participants’ difficulty level and adjusted the difficulty according to their performance; the therapist provided cues and feedback. The duration and number of training sessions were the same as those in the intervention group and involved one-on-one supervision by an occupational therapist.

All cognitive training sessions in both the groups were conducted in an appropriate space in the rehabilitation department. Missed sessions were rescheduled; thus, all the participants completed 24 training sessions. During the trial, both the groups continued their routine rehabilitation programs developed by occupational therapists, including upper limb motor function training and ADL training.

2.4 Measurements

Outcome measures included the scores of cognitive function tests and scales related to activity and participation. Cognitive function tests were conducted to examine attention, memory, processing speed, and executive function.

2.4.1 Cognitive function tests

Mini-Mental State Examination. The MMSE is used to evaluate cognitive impairment (Folstein et al., 1975). The MMSE assesses the cognitive domains of orientation, memory, attention, calculation, registration, and language (including reading, writing, naming, comprehension, and repetition). The total score is 30 points, with a score of≤23 suggesting cognitive impairment (Schultz-Larsen et al., 2007).

Symbol Digit Modalities Test. The Symbol Digit Modalities Test (SDMT) is used to evaluate the speed of information processing. Using a reference key, the participant has 90 s to pair specific numbers with given geometric figures as quickly as possible. The score is based on the number of completed correct matches, and responses can be provided in writing or orally (Smith, 1973). Considering that the affected side of the patients may affect their writing speed, we conducted the test orally, with the test administrator writing down the responses.

Digit Span Test. The Digit Span Test is among the main tests of the Wechsler Adult Intelligence Scale-III and is used to assess verbal working memory (Tulsky et al., 1997). The test administrator reads a list of numbers as a digit sequence, and participants repeat them until they provide an incorrect answer. The digit span is measured for the forward and backward recall of the digit sequence. The total score is based on the number of lists reported correctly in the forward and backward spans.

Spatial span test. The spatial span test is among the main tests of the Wechsler Memory Scale-III (Tulsky et al., 1997) and is used to examine changes in spatial working memory. The participants were asked to watch the test administrator tap a certain number of blocks in an increasingly longer sequence and then tap the blocks in the same (forward) or inverse (backward) order. The score is based on the highest number of digits that a participant can identify.

Taiwanese Version of the Montreal Cognitive Assessment. The Montreal Cognitive Assessment (MoCA) is a valid tool used to evaluate cognitive impairment (Julayanont et al., 2013). This study used the Taiwanese version of the MoCA that focuses on visuospatial/executive function, naming, attention, language ability, abstraction, delayed recall, and orientation. The maximum score is 30 points, with a score of≤26 indicating cognitive impairment (Tsai et al., 2012).

2.4.2 Activity and participation measure

Stroke impact scale. The Stroke Impact Scale (SIS) is a stroke-specific self-reported health-related quality of life as well as activity and participation measure that includes 64 items across eight domains (i.e., strength, ADLs/IADLs, mobility, hand function, memory, emotion, communication, and participation) as well as the global perception of the percentage of recovery (Duncan et al., 1999). The test– retest reliability of the SIS has been established, and satisfactory concurrent validity was reported between the SIS and independence functioning measured using the Lawton IADL scale, Barthel Index, and Functional Independence Measure (Carod-Artal et al., 2008; Duncan et al., 1999). Considering that the patients may not be able to complete the test by themselves, a therapist explained them the questions or helped them in filling the scores.

Table 1
Demographic characteristics

Variable Intervention (n = 19) Control (n = 20) t/X² p

Age (year), mean (SD) 63.63 (11.27) 65.50 (8.28) t = – 0.59 0.56

Gender (male/female), n 12/7 14/6 X² = 0.21 0.65

Education (y), mean (SD) 9.42 (3.44) 10.65 (3.33) t = – 1.13 0.26

Lesion side, n (%) X² = 0.65 0.42

Left 11 (57.9) 9 (45.0)

Right 8 (42.1) 11 (55.0)

Stroke type, n (%) X² = 3.31 0.07

Hemorrhage 5 (26.3) 11 (55.0)

Infarction 14 (73.7) 9 (45.0)

Months after stroke, mean (SD) 19.89 (18.95) 31.78 (23.96) t = – 1.68 0.10

Br. stage– arm mean (SD) 4.37 (1.08) 4.03 (1.14) t = 0.97 0.34

Br. stage-hand, mean (SD) 4.32 (1.25) 4.08 (1.10) t = 0.64 0.53

Baseline MMSE, mean (SD) 24.88 (2.53) 24.94 (3.07) t = 0.09 0.93

Variable	Intervention (n = 19)	Control (n = 20)	t/X²	p
Age (year), mean (SD)	63.63 (11.27)	65.50 (8.28)	t = – 0.59	0.56
Gender (male/female), n	12/7	14/6	X² = 0.21	0.65
Education (y), mean (SD)	9.42 (3.44)	10.65 (3.33)	t = – 1.13	0.26
Lesion side, n (%)			X² = 0.65	0.42
Left	11 (57.9)	9 (45.0)
Right	8 (42.1)	11 (55.0)
Stroke type, n (%)			X² = 3.31	0.07
Hemorrhage	5 (26.3)	11 (55.0)
Infarction	14 (73.7)	9 (45.0)
Months after stroke, mean (SD)	19.89 (18.95)	31.78 (23.96)	t = – 1.68	0.10
Br. stage– arm mean (SD)	4.37 (1.08)	4.03 (1.14)	t = 0.97	0.34
Br. stage-hand, mean (SD)	4.32 (1.25)	4.08 (1.10)	t = 0.64	0.53
Baseline MMSE, mean (SD)	24.88 (2.53)	24.94 (3.07)	t = 0.09	0.93

Br. stage, Brunnstrom stage.

2.5 Statistical analysis

Demographic differences at baseline between the groups were examined using the chi-square test for categorical data or the t test for continuous data. The independent t test was conducted to examine the comparability of the baseline performance of the measures of interest. No significant differences in baseline measures were noted between the groups, except for the mobility subdomain in the SIS (p < 0.05). A repeated-measures analysis of variance (ANOVA) was used to compare changes among different time points in the outcome variables between the groups. Group, time, and group-by-time interaction effects were determined. Considering the significant difference in the mobility subdomain in the SIS between the groups, a repeated-measures ANOVA with baseline measures as covariates was performed for this variable. Subsequently, planned comparisons with paired t tests were used to examine changes between pre- and post-tests (i.e., T0 vs. T1) and between pretest and follow-up tests (i.e., T0 vs. T2) for different outcome measures in both the groups. All statistical analyses were performed using SPSS 22.0 (IBM, Armonk, NY, USA). A two-tailed significance level of p < 0.05 was set.

3 Results

We recruited 39 participants with stroke who were randomly allocated to either the intervention group (n = 19) or the control group (n = 20). Table 1 presents a summary of the participants’ demographic characteristics, which were not significantly different between the groups. All the participants had a mean age of 64.55±10.19 years, and the average duration after stroke onset was 29.97±22.34 months. The average baseline MMSE score was 24.91±2.78. Thirty-three participants (16 in the intervention group and 17 in the control group) completed the study intervention and posttest assessment (Fig. 1). One participant from each group did not complete the follow-up assessment. A few participants in the intervention group reported that the pace of some games was too fast to complete and that they had sore eyes at the early phase of the intervention. The conditions were improved after a few sessions, and all the participants completed the intervention program.

Fig. 1

CONSORT study flowchart.

For cognitive function outcomes (Table 2), significant time effects were noted on the score of the MMSE (p < 0.001) and the total score of the Digit Span Test (p = 0.02). For changes between T0 and T1, the intervention group exhibited significant improvement in the MMSE score and the backward Digit Span Test score. For changes between T0 and T2, significant changes were noted in the MMSE score for both the groups and in the SDMT score only for the intervention group.

Table 2

Performance in MMSE, processing speed, and working memory in the intervention and control groups at the pretest (T0), posttest (T1), and follow-up test (T2)

Outcomes	Intervention			Control			Group		Time		Group×Time
	T0	T1	T2	T0	T1	T2	F	p	F	p	F	p
MMSE	24.88 (2.53)	26.06^* (2.05)	26.47^* (2.85)	24.94 (3.07)	25.71 (3.18)	26.63^* (2.90)	0.00	0.99	11.00	< 0.001^*	0.34	0.66
SDMT	24.50 (10.11)	25.31 (9.74)	26.40^* (7.64)	21.35 (7.54)	21.53 (7.12)	21.94 (7.72)	1.15	0.29	1.76	0.18	1.14	0.33
Digit span-forward	11.81 (3.15)	12.25 (2.96)	12.80 (2.34)	11.35 (2.69)	11.71 (2.73)	11.44 (2.16)	1.11	0.30	2.22	0.12	0.69	0.51
Digit span-backward	4.50 (1.41)	5.44^* (1.41)	5.07 (1.83)	5.00 (2.42)	5.12 (1.93)	5.38 (3.07)	0.01	0.92	2.72	0.07	0.86	0.43
Digit-span-total	16.31 (3.53)	17.69 (3.09)	17.87 (2.90)	16.35 (4.55)	16.82 (4.17)	16.81 (4.65)	0.48	0.49	4.22	0.02^*	0.28	0.76
Spatial span-forward	7.94 (2.05)	7.94 (1.81)	7.53 (1.73)	6.41 (1.87)	6.65 (2.03)	6.56 (2.39)	3.36	0.08	0.21	0.81	0.50	0.61
Spatial span-backward	5.63 (2.22)	6.31 (1.96)	5.27 (1.67)	5.18 (1.98)	4.88 (1.62)	4.75 (1.95)	1.91	0.18	2.18	0.12	2.52	0.09
Spatial span-total	13.56 (2.97)	14.25 (3.17)	12.80 (3.08)	11.59 (3.26)	11.53 (3.08)	11.31 (4.08)	3.29	0.08	1.99	0.15	1.03	0.36

Values are the mean (SD). T0, pre-test; T1, post-test; T2, follow-up test. ^*in the T1 and T2 indicates significant difference compared with T0. p < 0.05. MMSE, Mini-Mental Status Examination; SDMT, Symbol Digit Modalities Test.

Regarding the MoCA (Table 3), significant time effects were noted on the subscores of naming (p = 0.01) and delayed recall (p = 0.004) as well as the total score (p < 0.001). Furthermore, both the groups exhibited significant improvement in the naming subscore between T0 and T1, and only the intervention group demonstrated significant changes in the attention and total scores. For changes between T0 and T2, both the groups presented significant improvement in the delayed recall and total scores, and only the intervention group exhibited significant changes in the attention subscore.

Table 3

Performance in the MoCA in the intervention and control groups at the pretest (T0), posttest (T1), and follow-up test (T2)

Outcomes	Intervention			Control			Group		Time		Group×Time
	T0	T1	T2	T0	T1	T2	F	p	F	p	F	p
Visuospatial/executive	3.75 (1.06)	3.69 (1.08)	3.93 (0.96)	3.41 (0.94)	3.76 (1.15)	3.69 (1.25)	0.19	0.67	1.44	0.25	1.44	0.25
Naming	2.50 (0.63)	2.88^* (0.34)	2.73 (0.46)	2.65 (0.49)	2.94^* (0.24)	2.81 (0.40)	0.66	0.43	5.76	0.01^*	0.14	0.87
Attention	3.81 (1.11)	4.63^* (1.26)	4.73^* (1.62)	4.71 (1.10)	4.71 (1.45)	4.75 (1.44)	0.66	0.43	2.97	0.06	1.83	0.17
Language	1.31 (1.14)	1.31 (0.87)	1.07 (0.88)	1.41 (1.00)	1.35 (1.17)	1.31 (1.14)	0.04	0.84	0.50	0.61	0.65	0.52
Abstraction	0.88 (0.89)	0.88 (0.72)	1.00 (0.65)	1.06 (0.75)	0.71 (0.69)	1.00 (0.63)	0.01	0.94	1.13	0.32	0.53	0.56
Delayed recall	1.88 (1.36)	2.25 (1.65)	3.20^* (1.70)	1.94 (1.43)	2.59 (1.70)	2.81^* (1.76)	0.07	0.79	6.08	0.004^*	1.12	0.33
Orientation	5.38 (1.41)	5.31 (1.08)	5.27 (1.16)	5.29 (1.16)	5.35 (0.86)	5.56 (1.03)	0.22	0.64	0.20	0.83	0.25	0.78
Total	20.31 (4.19)	21.81^* (4.26)	22.73^* (4.28)	21.24 (3.54)	22.18 (4.56)	22.69^* (4.96)	0.14	0.71	11.58	< 0.001^*	0.82	0.45

Values are the mean (SD). T0, pretest; T1, posttest; T2, follow-up test. ^*in T1 and T2 indicates a significant difference compared with T0. p < 0.05.

Regarding SIS outcomes (Table 4), a significant time effect was observed on the mobility (p = 0.04) and participation subscores (p = 0.02). The control group exhibited significant improvement in the strength subscore between T0 and T1; however, a significant decrease in participation was noted between T0 and T2. No significant group and group-by-time interaction effects were observed in any outcome measures.

Table 4

Performance in the SIS in the intervention and control groups at the pretest (T0), posttest (T1), and follow-up test (T2)

Outcomes	Intervention			Control			Group		Time		Group×Time
	T0	T1	T2	T0	T1	T2	F	p	F	p	F	p
Strength	40.42 (13.95)	37.50 (13.98)	43.75 (17.99)	29.30 (17.78)	36.72^* (21.15)	36.33 (21.19)	1.84	0.19	2.22	0.12	1.25	0.30
Memory	67.26 (21.26)	69.28 (21.59)	72.58 (22.47)	63.23 (20.52)	65.76 (20.13)	64.73 (20.49)	1.14	0.29	0.85	0.43	0.05	0.95
Emotion	54.44 (20.25)	52.85 (15.17)	60.35 (12.86)	56.47 16.83)	59.15 (12.83)	59.20 (17.08)	0.22	0.64	1.35	0.27	0.77	0.47
Communication	74.84 (25.61)	81.90 (19.68)	79.05 (19.60)	73.11 (8.95)	80.46 (16.85)	72.99 (17.59)	0.45	0.51	2.77	0.07	0.16	0.80
ADL	67.59 (18.91)	67.50 (18.27)	69.17 (20.93)	57.83 (20.51)	63.53 (21.23)	62.22 (20.42)	1.44	0.24	1.36	0.26	0.80	0.46
Mobility#	78.70 (16.29)	82.78 (15.83)	79.63 (19.55)	67.57 (17.70)	70.59 (18.76)	71.18 (18.42)	0.13	0.72	3.38	0.04^*	0.33	0.72
Hand function	44.33 (32.12)	32.00 (31.89)	36.33 (32.76)	32.16 (31.70)	35.00 (31.12)	31.88 (34.44)	0.36	0.55	0.70	0.50	2.21	0.12
Participation	50.63 (31.52)	35.42 (28.54)	45.83 (30.47)	40.26 (23.72)	32.35 (27.13)	26.17^* (25.72)	1.77	0.19	4.38	0.02^*	1.35	0.27
Recovery	47.00 (13.86)	51.00 (19.84)	48.67 (19.59)	44.71 (24.72)	46.18 (25.83)	42.19 (23.31)	0.53	0.47	1.05	0.34	0.10	0.86

Values are the mean (SD). T0, pretest; T1, posttest; T2, follow-up test. ^*in T1 and T2 indicates a significant difference compared with T0. p < 0.05 #baseline was used as a covariate in the repeated-measures ANOVA.

4 Discussion

In this study, 24 CCT sessions with Lumosity cognitive training improved cognition, including in global and specific cognitive domains, in the stroke survivors. Although changes in cognitive function did not significantly differ between the intervention and control groups, CCT significantly improved cognitive domains, with near-transfer effects at posttest and follow-up. However, CCT failed to improve activities and participation measured using the SIS in the stroke survivors.

Our findings indicated that CCT effectively improved global cognitive function, as measured using the MMSE and MoCA at the posttest and follow-up, although we administered fewer training sessions (only 20 min per session for 24 sessions) than did the previous studies (Chen et al., 2015; Prokopenko et al., 2013; Tang et al., 2019). The significant changes observed in this study are in agreement with those of the aforementioned studies. The MoCA is a sensitive measurement of the cognitive function of individuals with mild cognitive impairment; therefore, it can effectively reflect changes in overall cognitive function before and after cognitive training (Nasreddine et al., 2005). In this study, the MoCA score tended to increase stably, suggesting that CCT effectively improved the global cognitive function of the stroke survivors after training and at follow-up.

In the present study, only the intervention group exhibited improvement in specific cognitive tasks, including the SDMT, digit span backward, or attention subtest in the MoCA. Notably, the transfer effect on the attention subtest in the MoCA in the CCT group was observed after cognitive training and remained significant after 4 weeks of follow-up. The visuospatial attention ability was among the main contents of the CCT group. Significant improvement in attention performance observed in the digit span, vigilance, and serial 7s tests of the MoCA indicated very near-transfer effect after training and at follow-up. Moreover, only the CCT group exhibited a significant transfer effect on the backward digit span test after training. This finding is crucial because the backward digit span test is a major measure of working memory capacity associated with visuospatial processing (Weng et al., 2019). Apart from visuospatial attention ability, CCT also focused mainly on the memory ability in terms of cognitive function. Some researchers have reported that computer-based memory and attention tasks can result in transfer to working memory test for stroke survivors (van de Ven et al., 2016) Therefore, the significantly improved performance of our CCT group in the digit span backward task after training indicated improvement in the related trained tasks domains. The training effect on working memory was not observed at follow-up, which might be due to insufficient training. The improvement in performance in the SDMT, digit span backward test, or MoCA’s attention subtest after CCT compared with conventional cognitive training suggests that the CCT group exhibited improvement in information processing speed, attention, and working memory ability due to the ability of the Lumosity system to provide immediate feedback and high repeatability within a shorter training time (20 min per day). However, more research is required to confirm the effects of CCT on the cognitive function of stroke survivors.

This study revealed time effects on global cognitive function in some specific cognitive domains (i.e., digit span, naming, and delayed recall), suggesting that both the types of cognitive training can lead to positive outcomes in individuals with stroke. Performance in the naming and delayed recall subtasks of the MoCA in both the active control group and intervention group significantly differed between the posttest and at follow-up, respectively. This finding might be due to the true benefits of active control activity that are equal in magnitude to those of the cognitive training program or practice effect in both the groups. However, because the time interval between two tests was short (12– 16 weeks), the practice effects of the measurements could not be ruled out as a cause. In addition, the main significant improvement in the control group was observed at the follow-up assessment (in terms of MMSE and MoCA total scores). Compared with the CCT group, the active control group tended to require longer time to experience improvement, possibly because of the insufficient intensity or number of cognitive training tasks in the usual program. However, the reason for this difference in the onset of beneficial effects remains unclear (Merriman et al., 2019).

No significant changes in activities and participation measured using the SIS were observed after CCT for the stroke survivors, which is consistent with the findings of previous studies (Hazelton, 2020; Prokopenko et al., 2013; Wentink et al., 2016). This may be partly because the intervention training time was too short to induce obvious changes (van de Ven et al., 2016). Furthermore, many factors affect activities and participation, including health, cognitive and physical function, and contextual factors (i.e., personal and environmental) (Cawood et al., 2016; Hoyle et al., 2012). Several studies have indicated that the greatest overall improvement in real-world abilities requires sustained engagement in a paradigm of complex, sophisticated, and integrated or holistic interventions (Cicerone et al., 2008; Martelli et al., 2012; Mateer et al., 2005). Such wide-ranging holistic intervention is unlikely to be provided by even a relatively large set of computer-based tasks involving an increasing diversity of tasks. Therefore, improved cognitive function may not have sufficient transfer effects on activities and participation under the lack of comprehensive cognitive intervention. Moreover, this finding suggested that any of outcome measures assess the real world function might be inconsistent with any cognitive domain evidence. By contrast, the control group exhibited significant improvement in strength, which may be related to the involvement of the upper extremity and hand activities in the table-top cognitive training regimen. Moreover, a longer duration of disease might lead to a largely negative response phenomenon in the participation domain compared with other domains in the SIS for stroke survivors (Ytterberg et al., 2017). In our control group, a higher proportion of participants with chronic disease experienced a significant negative effect on participation in the follow-up period.

This study adopted an active control design and provided the same paired CCT training doses to the control group; however, the results revealed no significant between-group differences. Our results are consistent with those of previous studies including a passive control group (Wentink et al., 2016). However, our patients demonstrated changes in the MoCA total score after training (from 20.31 to 21.81, difference = 1.5), and these changes were sustained at follow-up (from 20.31 to 22.73, difference = 2.42), which exceeded the minimal clinically important difference based on both anchor-based (1.22) and distribution-based (2.15) methods (Wu et al., 2019). Our control group exhibited a larger than minimal clinically important difference only in MoCA total scores at follow-up (from 21.24 to 22.69, difference = 1.45). These findings suggested that the patients in the CCT group who received the training, even at a lower dose, experienced more beneficial and meaningful improvement in cognitive function than did those in the control group. However, a higher training frequency and a longer CCT duration (i.e., 30 min per day) might improve cognitive outcomes in stroke survivors. And we could not exclude the possibility that CCT could have a stimulation effect on cognition that perhaps has potential to facilitate spontaneous recovery in mildly impaired individuals, like our study participants.

Our study has several advantages. First, an active control group was included to compare the effectiveness of CCT with that of conventional cognitive training. Second, in addition to cognitive function, the effects of CCT on activities and participation were investigated in adults with stroke. Third, follow-up assessments were performed to examine the maintenance of the effectiveness of CCT.

Some limitations should be considered when interpreting the findings of this study. First, our sample size was small. Second, the primary participants were in the chronic stroke phase with mild to moderate upper limb and cognitive impairment. Therefore, the generalization of these results to participants with other characteristics (i.e., severe impaired populations) should be performed with caution. Third, we did not implement any blinding procedure during the evaluation; however, we used standardized procedures to reduce errors. Forth, the outcome measures in our study did not always represent the real-world function, especially executive ability. A more valid surrogate, for example, instrumental activities of daily function will be suggested in the further study. Finally, in the absence of the waiting-list control condition, improvements in both the groups could be due to either both programs being effective or nonspecific elements. Thus, both an active control condition and a waiting-list control condition should be included to control for all training-unspecific effects (van de Ven et al., 2016).

5 Conclusion

In this study, cognition of stroke survivors improved after administration of CCT with Lumosity. The cognitive function improved included both global cognitive function and specific functions related to trained task domains, including working memory, attention, processing speed, naming, and delayed recall abilities. However, a rigorous study design with a larger small sample size is limited to validate the findings. Furthermore, studies examining the effects of a combination of holistic cognitive therapy and a higher training dose on stroke survivors with severe impairment are warranted to confirm our findings.

Footnotes

Acknowledgments

This study was supported by a grant from Kaohsiung Municipal Ta-Tung Hospital, Taiwan, R.O.C (grant number kmtth-103-036). The authors would like to thank all participants and their relatives for participating in the study.

Author contributions

All authors provided critical intellectual interpretations, revised the manuscript, and approved the final manuscript.

Conflict of interest

The authors declare no conflicts of interest.

References

Barba,

, Martinez-Espinosa,

, Rodriguez-Garcia,

, Pondal,

, Vivancos,

, & Del Ser,

(2000) Poststroke dementia: clinical features and risk factors, Stroke, 31(7), 1494–1501.

Bogdanova,

, Yee,

M. K.

, Ho,

V. T.

, & Cicerone,

K. D.

(2016) Computerized cognitive rehabilitation of attention and executive function in acquired brain injury: a systematic review, Journal of Head Trauma Rehabilitation, 31(6), 419–433.

Carod-Artal,

F. J.

, Coral,

L. F.

, Trizotto,

D. S.

, & Moreira,

C. M.

(2008) The stroke impact scale 3.0: evaluation of acceptability, reliability, and validity of the Brazilian version, Stroke, 39(9), 2477–2484.

Cawood,

, Visagie,

, & Mji,

(2016) Impact of post-stroke impairments on activities and participation as experienced by stroke survivors in a Western Cape setting, South African Journal of Occupational Therapy, 46(2), 10–15.

Chen,

C.-X.

, Mao,

R.-H.

, Li,

S.-X.

, Zhao,

Y.-N.

, & Zhang,

(2015) Effect of visual training on cognitive function in stroke patients, International Journal of Nursing Sciences, 2(4), 329–333.

Cicerone,

K. D.

, Mott,

, Azulay,

, Sharlow-Galella,

M. A.

, Ellmo,

W. J.

, Paradise,

, & Friel,

J. C.

(2008) A randomized controlled trial of holistic neuropsychologic rehabilitation after traumatic brain injury, Archives of Physical Medicine and Rehabilitation, 89(12), 2239–2249.

Crum,

R. M.

, Anthony,

J. C.

, Bassett,

S. S.

, & Folstein,

M. F.

(1993) Population-based norms for the Mini-Mental State Examination by age and educational level, Jama, 269(18), 2386–2391.

Cumming,

T. B.

, Marshall,

R. S.

, & Lazar,

R. M.

(2013) Stroke, cognitive deficits, and rehabilitation: still an incomplete picture, International Journal of Stroke, 8(1), 38–45.

Duncan,

P. W.

, Wallace,

, Lai,

S. M.

, Johnson,

, Embretson,

, & Laster,

L. J.

(1999) The stroke impact scale version 2.0: evaluation of reliability, validity, and sensitivity to change, Stroke, 30(10), 2131–2140.

10.

Folstein,

M. F.

, Folstein,

S. E.

, & McHugh,

P. R.

(1975) Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician, Journal of Psychiatric Research, 12(3), 189–198.

11.

Hardy,

J. L.

, Nelson,

R. A.

, Thomason,

M. E.

, Sternberg,

D. A.

, Katovich,

, Farzin,

, & Scanlon,

(2015) Enhancing cognitive abilities with comprehensive training: A large, online, randomized, active-controlled trial, PloS One, 10(9), e0134467.

12.

Hazelton,

(2020) Can cognitive rehabilitation improve attention deficits following stroke?-A Cochrane Review summary with commentary. NeuroRehabilitation 47(3), 355–357.

13.

Hill,

N. T.

, Mowszowski,

, Naismith,

S. L.

, Chadwick,

V. L.

, Valenzuela,

, & Lampit,

(2017) Computerized cognitive training in older adults with mild cognitive impairment or dementia: a systematic review and meta-analysis, American Journal of Psychiatry, 174(4), 329–340.

14.

Hoyle,

, Gustafsson,

, Meredith,

, & Ownsworth,

(2012) Participation after stroke: Do we understand all the components and relationships as categorised in the ICF? Brain Impairment, 13(1), 4–15. .

15.

Hyndman,

, & Ashburn,

(2003) People with stroke living in the community: Attention deficits, balance, ADL ability and falls, Disability and Rehabilitation, 25(15), 817–822. https://doi.org/10.1080/0963828031000122221

16.

Julayanont,

, Phillips,

, Chertkow,

, & Nasreddine,

Z. S.

(2013) Montreal Cognitive Assessment (MoCA): Concept and Clinical Review. In A. J. Larner (Ed.), Cognitive Screening Instruments: A Practical Approach (pp. 111-151). Springer London. https://doi.org/10.1007/978-1-4471-2452-8_6

17.

König,

, Francis,

L. E.

, Joshi,

, Robillard,

J. M.

, & Hoey,

(2017) Qualitative study of affective identities in dementia patients for the design of cognitive assistive technologies, Journal of Rehabilitation and Assistive Technologies Engineering, 4, 1–15.

18.

Kalaria,

R. N.

, & Ballard,

(2001) Stroke and cognition, Current Atherosclerosis Reports, 3(4), 334–339.

19.

Kueider,

A. M.

, Parisi,

J. M.

, Gross,

A. L.

, & Rebok,

G. W.

(2012) Computerized Cognitive Training with Older Adults: A Systematic Review, PLoS One, 7(7), e40588. https://doi.org/10.1371/journal.pone.0040588

20.

Lipskaya-Velikovsky,

, Zeilig,

, Weingarden,

, Rozental-Iluz,

, & Rand,

(2018) Executive functioning and daily living of individuals with chronic stroke: measurement and implications, International Journal of Rehabilitation Research, 41(2), 122–127.

21.

Martelli,

M. F.

, Zasler,

N. D.

, & Tiernan,

(2012) Community based rehabilitation: Special issues, NeuroRehabilitation, 31(1), 3–18.

22.

Mateer,

C. A.

, Sira,

C. S.

, & O’Connell,

M. E.

(2005) Putting Humpty Dumpty together again: the importance of integrating cognitive and emotional interventions, The Journal of Head Trauma Rehabilitation, 20(1), 62–75.

23.

Merriman,

N. A.

, Sexton,

, McCabe,

, Walsh,

M. E.

, Rohde,

, Gorman,

, ... Williams,

D. J.

(2019) Addressing cognitive impairment following stroke: systematic review and meta-analysis of non-randomised controlled studies of psychological interventions, BMJ Open, 9(2), e024429.

24.

Nasreddine,

Z. S.

, Phillips,

N. A.

, Bédirian,

, Charbonneau,

, Whitehead,

, Collin,

, ... Chertkow,

(2005) The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment, Journal of the American Geriatrics Society, 53(4), 695–699.

25.

Nguyen,

, Murphy,

, & Andrews,

(2021) A game a day keeps cognitive decline away? a systematic review and meta-analysis of commercially-available brain training programs in healthy and cognitively impaired older adults. Neuropsychology Review, 1-30.

26.

Petersen,

R. C.

(2004) Mild cognitive impairment as a diagnostic entity, Journal of Internal Medicine, 256(3), 183–194.

27.

Prokopenko,

, Mozheyko,

, Petrova,

, Koryagina,

, Kaskaeva,

, Chernykh,

, ... Bezdenezhnih,

(2013) Correction of post-stroke cognitive impairments using computer programs, Journal of the Neurological Sciences, 325(1-2), 148–153.

28.

Rogers,

J. M.

, Foord,

, Stolwyk,

R. J.

, Wong,

, & Wilson,

P. H.

(2018) General and Domain-Specific Effectiveness of Cognitive Remediation after Stroke: Systematic Literature Review and Meta-Analysis [journal article], Neuropsychology Review, 28(3), 285–309. https://doi.org/10.1007/s11065-018-9378-4

29.

Schultz-Larsen,

, Lomholt,

R. K.

, & Kreiner,

(2007) Mini-Mental Status Examination: a short form of MMSE was as accurate as the original MMSE in predicting dementia, Journal of Clinical Epidemiology, 60(3), 260–267.

30.

Sigmundsdottir,

, Longley,

W. A.

, & Tate,

R. L.

(2016) Computerised cognitive training in acquired brain injury: A systematic review of outcomes using the International Classification of Functioning (ICF, Neuropsychological Rehabilitation, 26(5-6), 673–741.

31.

Simons,

D. J.

, Boot,

W. R.

, Charness,

, Gathercole,

S. E.

, Chabris,

C. F.

, Hambrick,

D. Z.

, & Stine-Morrow,

E. A.

(2016) Do “brain-training” programs work? Psychological Science in the Public Interest, 17(3), 103–186.

32.

Smith,

(1973) Symbol digit modalities test. Western Psychological Services Los Angeles.

33.

Spitzer,

, Tse,

, Baum,

C. M.

, & Carey,

L. M.

(2011) Mild impairment of cognition impacts on activity participation after stroke in a community-dwelling Australian cohort, OTJR: Occupation, Participation and Health, 31(1_suppl), S8–S15.

34.

Tang,

, Xing,

, Zhu,

, He,

, Li,

, Yang,

, ... Zhao,

(2019) The effects of -week cognitive training in patients with vascular cognitive impairment, no dementia (the Cog-VACCINE study): A randomized controlled trial, Alzheimer’s & Dementia, 15(5), 605–614.

35.

Tsai,

C.-F.

, Lee,

W.-J.

, Wang,

S.-J.

, Shia,

B.-C.

, Nasreddine,

, & Fuh,

J.-L.

(2012) Psychometrics of the Montreal Cognitive Assessment (MoCA) and its subscales: validation of the Taiwanese version of the MoCA and an item response theory analysis, International Psychogeriatrics, 24(4), 651–658.

36.

Tulsky,

, Zhu,

, & Ledbetter,

(1997) Wechsler Adult Intelligence Scale-andWechsler Memory Scale-Technical Manual. In: The Psychological Corporation, San Antonio, TX.

37.

van de Ven,

R. M.

, Murre,

J. M.

, Veltman,

D. J.

, & Schmand,

B. A.

(2016) Computer-based cognitive training for executive functions after stroke: A systematic review, Frontiers in Human Neuroscience, 10, 150.

38.

Weng,

, Liang,

, Xue,

, Zhu,

, Jiang,

, Wang,

, & Chen,

(2019) The transfer effects of cognitive training on working memory among Chinese older adults with mild cognitive impairment: a randomized controlled trial, Frontiers in Aging Neuroscience, 11, 212.

39.

Wentink,

, Berger,

, de Kloet,

, Meesters,

, Band,

, Wolterbeek,

, ... Vliet Vlieland,

(2016) The effects of an -week computer-based brain training programme on cognitive functioning, QoL and self-efficacy after stroke, Neuropsychological Rehabilitation, 26(5-6), 847–865.

40.

Wilson,

, Cockburn,

, & Halligan,

(1987) Development of a behavioral test of visuospatial neglect, Archives of Physical Medicine and Rehabilitation, 68(2), 98–102.

41.

Wong,

, Xiong,

Y. Y.

, Wang,

, Lin,

, Chu,

W. W.

, Kwan,

P. W.

, ... Mok,

(2013) The NINDS-Canadian stroke network vascular cognitive impairment neuropsychology protocols in Chinese, J Neurol Neurosurg Psychiatry, 84(5), 499–504. https://doi.org/10.1136/jnnp-2012-304041

42.

Ytterberg,

, Dyback,

, Bergstrom,

, Guidetti,

, & Eriksson,

(2017) Perceived impact of stroke six years after onset, and changes in impact between one and six years, Journal of Rehabilitation Medicine, 49(8), 637–643.