Effectiveness of Simultaneous Combined Intervention for Enhancing Cognitive Function in Patients with Moderate Alzheimer’s Disease

Design of the study

A randomized controlled trial of 8 weeks of training intervention was carried out at the Laboratory of Exercise Physiology and Physiopathology (LR19ES09) in the Faculty of Medicine of Sousse, Tunisia. Cognitive assessments were conducted at baseline (W1), after 4 weeks (W4), at the end of the training (W8), and after a 1-month follow-up (4 weeks of detraining, W12). The cognitive evaluation was counterbalanced and randomized for all patients with AD. These evaluations were performed under the supervision of a trained PhD student. For the cognitive assessment of patients with AD in this study, a comprehensive neuropsychological evaluation was administered by a qualified neuropsychologist. Neuropsychological assessments were administered to all patients with moderate AD at baseline and at the end of the training (see Table 1).

Participants

The preliminary sample size determination was guided by Beck’s [30] methodology, utilizing G *Power software (version 3.1.9.2, Faul et al. [31], from Kiel University, Germany). The analysis framework chosen was an F test (ANOVA: repeated measures, within-between interaction) involving three different groups over four-time points. We set the Type I error probability (α) at 0.05 and aimed for a power (1-β) of 0.95, with an assumed correlation of 0.5 among repeated measures. Drawing from Ben Ayed et al. [32] and collaborative discussions, we anticipated a medium effect size, designated as partial eta-squared of 0.06 (equivalent to an f value of 0.25), necessitating a minimum of 45 participants. A total of 79 patients with moderate AD voluntarily participated in the study. Four patients in the control group dropped the intervention after 8-week period and two other patients in CG dropped after 4 weeks of detraining. All patients were clinically diagnosed with moderate AD at the Department of Neurology, Sahloul Hospital, Sousse, Tunisia according to the International Working Group-2 (IWG-2) diagnostic criteria [33].

The study’s eligibility criteria included individuals aged 65 to 75 years, living independently, outside of institutional settings, scoring between 10 and 19 on the MMSE [34], possessing normal or corrected-to-normal vision and color perception, and capable of daily movement without technical support.

Exclusion criteria were as follows: inability to provide written informed consent; cognitive or functional impairments hindering study participation; cardiovascular diseases; acute lower back or extremity pain; peripheral neuropathy; orthopedic and rheumatic conditions; and severe psychiatric disorders (depression score < 5). Additionally, potential participants were required to undergo a risk factor assessment, including cardiovascular, and other comorbidities, prior to enrollment.

The study’s protocol received approval from the Research Ethics Committee of the Faculty of Medicine, Sousse, Tunisia (IRB 00008931).

Seventy-nine patients with moderate AD provided informed consent. Patients with AD were randomly assigned to 3 groups: a simultaneous combined training group (ACT group, n = 25), an aerobic training group (AT group, n = 27), or simply reading for control group (CG group, n = 26).

Training protocol

All patients were randomly assigned to 8 weeks, twice-weekly of either AT group, ACT group, or a reading control group (total of intervention group was 16 sessions).

The Aerobic Training consisted of an aerobic physical activity program. Before the physical exercise, the seat on the ergocycle (Labyrinth, Paris, France) was regulated to the patient height. After a 5-min warm-up, patients with AD performed 20 min of exercise at a moderate intensity, i.e., 60% of the maximal heart rate (MHR) measured at the end of the 6-Minute Walk Test (6MWT) [35, 36], followed by a 5–10-min cooldown. One week preceding the experimentation period, MHRs were measured during a familiarization session at the end of the 6MWT. During the physical exercise sessions, patients were monitored with heart rate monitors (Polar V800, USA), and they were requested to report their rates of perceived exertion (RPE) at the conclusion of each aerobic training session.

In line with the findings of a meta-analysis by Zhu et al. [24], the aerobic-cognitive training conducted was a form of dual task training. Patients with AD in the combined group performed the same training regimen as those in the AT group, with the additional incorporation of three simple cognitive tasks displayed on a computer screen. These tasks, including visual scanning, working memory, information processing speed, selective attention, arithmetic, and perception exercises, were designed as straightforward and engaging cognitive activities engaging various cognitive processes within the same session. A team of medical experts, particularly neuropsychologists, developed these games to target cognitive functions in patients with AD, with each task lasting approximately 5 min. To avoid repeated boot effects, three different tasks were introduced at each session from those used previously. The scores for these cognitive games were not recorded, as they were intended to be a supplementary component of the AT.

Patients with moderate AD in the control group were asked to read a religious book for 20 min instead of engaging in physical training. All participants were advised to refrain from engaging in strenuous physical activities or any particularly fatiguing efforts prior to each session. Furthermore, to guarantee uniformity in testing conditions and minimize the influence of circadian rhythms on outcomes [37, 38], all sessions were conducted at the same time of day during the morning hours.

Measures

Neuropsychological assessments were used for subjective evaluations of daily activity performance, including the MMSE test [34, 39], the 16-item free and cued recall test (RL/RI-16) [40], the clock test [41], and the fluency test [42]. Additionally, the 15-item Geriatric Depression Scale (GDS) [43] was used for depression screening, and the World Health Organization Quality of Life-BREF (WHOQOL-BREF-100) [44] was utilized for quality of life; the full procedure is defined in detail in a previous publication by our team [32, 45].

Cognitive function (i.e., working memory, selective attention, and problem-solving) was assessed 5–10 min after finished interventions to permit heart rate to return to a resting value (≤10% above the individual baseline rate). To avoid an order effect, cognitive assessments were randomly counterbalanced for all AD patients. Following the instructions for the cognitive tasks, the experimenter sat alongside the patients, observing their activity without any visible reaction. Furthermore, these tests were maintained as investigational scenarios.

The French version of Victoria of Stroop color-word task was used to evaluate selective attention [46] and resistance to interference [47]. The Victoria Stroop Test (VST) presents a short version of the original Stroop test. It consists of three subtests, each presenting a stimulus card with six rows of four items. The first subtest (VST1) displays 24 dots in green, blue, red and yellow colors in a pseudorandom sequence. The second subtest (VST2) presents 24 common words, written in the four different colors. The third subtest (VST3) shows 24 color words (green, yellow, blue, and red) written in a color that does not match the word’s meaning (e.g., “green” is written in blue ink). Additionally, we incorporated an experimental condition requiring patients to read color names written in black ink. This condition was designed to investigate automatic reading behavior with a limited number of items and to create interference scenarios.

Working memory was explored using the forward and backward digits span test [48–50]. On the computer screen, digits were displayed at a pace of one digit per second. Subjects participated in two experimental conditions: in the forward condition, they were required to recall and list the digits in the same sequence as presented, while in the backward condition, they had to list the digits in reverse order. Performance in both conditions was assessed based on the number of correctly recalled items [49, 50].

The problem-solving function was assessed using the Tower of Hanoi test [51]. In this test, patients were required to rebuild a pyramid with three disks, using the middle disk for necessary moves, to shift it from the right-hand peg to the left-hand peg. The rules stipulated patients could only move one disk at a time and were not allowed to place a larger disk on top of a smaller one. Time taken and the number of moves made were recorded. Seven disk displacements were required, at minimum, to resolve this test.

The complete description of cognitive tests is provided in thorough detail in our prior publication [32, 45].

Statistical analysis

All Analyses were conducted using SPSS software (IBM® SPSS® Statistics version 26.0). Descriptive statistics were summarized as mean (M) and standard deviation (SD) for the MMSE, Stroop, tower of Hanoi, and digit span tests. Furthermore, we calculated baseline and post-intervention values (at W4, W8, and W12) to assess the magnitude of significant differences. The percentage of performance gain was calculated as follows: [(post value –pre value)/ (pre value)]×100. The Shapiro–Wilks W-test was used to check the normality of the distribution.

The main effect of groups was explored using the Kruskal–Wallis non-parametric test. When appropriate, post hoc pairwise comparisons were performed, and the results were interpreted using a Bonferroni correction. The Friedman non-parametric test was employed to explore the time assessment (baseline, W4 and W8) of each group. The paired Wilcoxon test was used to interpret within-group results of neuropsychological measures (Baseline and W8) and the persistent effect (W8 and W12). Statistical significance for all analyses was accepted at the level of p < 0.05.

Participant characteristics

Patients with moderate AD were randomly assigned to one of three groups. Across all participants, the mean age was 70.14±4.61 years, body mass averaged 75.38±8.57 kg, body mass index (BMI) was 28.60±3.84, and the distribution between female and male participants was 65.4% and 34.6%, respectively. No significant differences were found between the groups regarding age, body mass, and BMI (p > 0.05), indicating the sample was homogeneous with respect to these parameters.

As shown in Table 2 and taking into account the standards for different neuropsychological scales as well as in accordance with the literature, the three groups of participants showed similar cognitive disorders. Regarding their emotional profile, AT and ACT participants had mild depressive symptoms, which were more marked compared to the CG. The results obtained for quality of life indicated that the majority of participants were not satisfied in the physical, psychological, and environmental domains but, on the opposite, rather satisfied in the field of social relations. However, a greater overall satisfaction was observed in the AT (p < 0.05).

Neuropsychological results

After the 8 weeks of aerobic and combined training, a significant improvement of all neuropsychological assessments was observed (all p < 0.05; detailed results in Table 3). The patients in the control group showed no significant difference from pre- to post-reading session (detailed results in Table 3).

In addition to the results presented in Table 3, changes in depression and quality of life score were evaluated immediately following the intervention and after a four-week detraining period. We found depression levels decreased significantly, and quality of life improved at week 8 (p < 0.001) in both experimental groups. The positive effects of aerobic training (AT) and aerobic-cognitive training (ACT) were maintained after four weeks of detraining, as evidenced by the absence of significant differences between W12 and W8 (p > 0.05).

Training-detraining effects on cognitive functions

As shown in Fig. 1, there were no significant differences between groups at baseline, regardless of the cognitive task (detailed results presented in Supplementary Table 1). After four weeks of training (W4), both the AT and ACT groups significantly improved the majority of cognitive performances (except Hanoi moves) compared to the CG group (all p < 0.001), with ACT achieving greater cognitive performance than AT (p = 0.000). The performance of the AT and ACT groups continued to improve at W8 compared to W4 (all p < 0.01; see detailed statistics in Supplementary Table 2), with ACT outperforming AT in all tested skills (all p < 0.001) except attention (p > 0.05). It is worth noting that the CG group’s performance remained unchanged throughout the 8-week intervention program.

OPEN IN VIEWER

Fig. 1

Cognitive performances in the different tests for control, aerobic and combined groups at baseline, after 4 weeks (W4), after 8 weeks (W8) of training, and after 4 weeks of detraining (W12). S: significant differences compared with CG ^*: significant difference between ACT and AT; a: significant compared to baseline (W4 versus baseline and W8 versus baseline); and b: significant differences between W4 and W8.

Following 4 weeks of detraining, a lasting effect was observed in the AT and ACT groups in all tested tasks (Fig. 2), as evidenced by the absence of a significant difference between W8 and W12 (detailed results presented in Supplementary Table 3). The improvements induced by training at W8 persisted at W12.

OPEN IN VIEWER

Fig. 2

Percent gain from baseline at W8 and W12 for the different tests among the aerobic and combined groups. ^*significant difference between experimental groups.

The improvements in cognitive function at W8 and W12, compared with the baseline values are presented in Fig. 2. For all tested functions after training and detaining periods, higher percentages of gain were recorded for the two experimental groups. For example, at both test sessions, our results showed significant improvements in Hanoi task performance with percent gains around 29% and 24% for ACT and AT, respectively (all p < 0.05). Additionally, significant enhancements were observed for the digit span forward condition at W12 (p = 0.04) with percentage of gain ≈19% versus ≈15% for ACT and AT, respectively.

Limitations and perspectives

While the results presented above are positive, there are some limitations to the present study. First, we experienced some attrition, particularly among participants in the CG. Nevertheless, all participants, except for four patients in the CG, successfully completed the eight training sessions. Second, we did not have access to quantitative electroencephalogram, therefore future studies are encouraged to utilize functional and structural brain imaging to quantitively measure adaptations in the brain structure and function. Third, while beneficial effects persisted after four weeks, we did not monitor the duration of these changes. This highlights the need for follow-up assessments at 3, 6, or even 12 months to better understand the longevity of the intervention’s impact. Finally, future studies, incorporating larger sample sizes and accounting for potential confounding factors including sleep patterns, physical activity behaviors, and sex differences [68], are crucial to comprehensively understanding the impact of combined interventions. This research is necessary to identify the most effective training strategies and evaluate the sustained impact of such regimens on the cognitive functions of patients with AD.

Conclusion

The observed beneficial effects of combined training on cognitive functions, coupled with improvements in quality of life and even a reduction in depressive symptoms, underscores the potential utility, safety, and enduring efficacy of combined training as a strategy to mitigate the impairments associated with AD. To solidify these findings and explore potential implications for early cognitive decline prevention, further large-scale studies with extended follow-up periods are recommended.