The comprehensive impact of exercise interventions on cognitive function and quality of life in alzheimer’s disease patients: a systematic review and meta-analysis

Abstract

Background

Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder marked by cognitive decline and reduced quality of life. Exercise interventions are considered a promising non-pharmacological strategy to enhance cognition and well-being in AD patients. However, effects across different cognitive domains and quality of life remain unclear due to variations in study designs, intervention types, and outcome measures.

Objective

This systematic review and meta-analysis evaluated the effects of exercise interventions on cognitive function, executive function, memory, and quality of life in AD patients by synthesizing evidence from randomized controlled trials (RCTs).

Methods

We conducted a comprehensive search in PubMed, Embase, and Web of Science. Forty-five studies met inclusion criteria and were included in the quantitative synthesis. Primary outcomes included global cognition, executive function, memory, and quality of life. Meta-analyses were performed using standardized mean differences (SMD) and 95% confidence intervals (CI). Risk of bias was assessed via the Cochrane tool, and heterogeneity was quantified using the I² statistic.

Results

Exercise interventions Significantly improved global cognition, with SMDs of 0.30 (95% CI: 0.17–0.43, p < 0.00001) for MMSE and 0.39 (95% CI: 0.08–0.70, p = 0.01) for MoCA, with low heterogeneity (I² = 36% and 0%). ADAS-Cog showed a small, non-significant effect (SMD = − 0.09, 95% CI: − 0.35 to 0.17, p = 0.49). Executive function (SMD = 0.13, p = 0.09) and memory (SMD = 0.09, p = 0.34) did not improve significantly. Quality of life improved significantly (SMD = 0.17, 95% CI: 0.03–0.31, p = 0.02; I² = 5%), measured using tools such as QOL-AD and Qualidem.

Conclusion

Evidence suggests that structured, multimodal exercise programs—combining aerobic (≥ 150 min/week), resistance (2–3 times/week), and balance training (2–3 times/week) for at least 12 weeks with individualized intensity—may improve global cognition and quality of life in Alzheimer’s disease patients.

Introduction

Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder that currently affects over 50 million people worldwide, and this number is expected to triple by 2050 as the global population continues to age [1–7]. It is characterized by a progressive decline in cognitive functions, particularly in memory, language, and executive functions, which ultimately leads to impairments in daily living and a reduced quality of life [8–12]. Current pharmacological treatments, including cholinesterase inhibitors and NMDA receptor antagonists, provide only symptomatic relief and have limited effectiveness in slowing disease progression [13–19]. As a result, there is increasing interest in non-pharmacological interventions such as exercise, which has demonstrated potential to improve cognitive function and overall health in both healthy older adults and individuals with mild cognitive impairment [20–25]. Exercise may exert its cognitive benefits through neuroplasticity, enhanced cerebral perfusion, and anti-inflammatory pathways, which are discussed in detail in later Sects. [26–31].

While exercise has been widely studied as a potential intervention to improve cognitive function in healthy aging populations, its effects on Alzheimer’s Disease patients remain unclear and inconsistent [23, 32, 33]. Some randomized controlled trials (RCTs) have demonstrated significant cognitive improvements in AD patients, particularly in executive function and memory [20, 34, 35]. However, other studies have reported either minimal or non-significant effects. Additionally, while exercise is believed to enhance quality of life, the variability in reported outcomes raises questions about the generalizability of these findings across different AD populations [32].

While exercise has been widely studied as a potential intervention to improve cognitive function in healthy aging populations, its effects on patients with Alzheimer’s Disease remain unclear and inconsistent. These inconsistencies are especially evident across different domains—for example, global cognition as assessed by MMSE or MoCA tends to show significant improvement, while executive function and memory domains often yield small or statistically non-significant effects [23, 33], These ‘non-significant’ findings in prior research typically refer to small positive effect sizes that fail to achieve statistical significance in executive function and memory, likely due to short intervention durations, limited sample sizes, and heterogeneity in cognitive assessment tools [23, 33]. Moreover, variability in intervention intensity, frequency, and participant adherence may further contribute to inconsistent domain-specific outcomes. Furthermore, variability exists across exercise modalities, with aerobic and multimodal interventions generally showing more promising outcomes than resistance training alone [24, 32]. Previous reviews, such as those by Chen, WQ et al. [36] Groot, C et al. [33] and Jia, RX et al. [23], provided useful insights but were limited by smaller sample sizes, lack of domain-specific analyses, and inconsistent inclusion of quality-of-life outcomes. However, previous meta-analyses were limited by small sample sizes, heterogeneous intervention protocols, and lack of domain-specific cognitive assessments, which restrict the generalizability of their conclusions. Among the different exercise modalities evaluated in previous literature, aerobic exercise and multimodal interventions that combine aerobic, resistance, and balance training have been most consistently associated with improvements in cognitive function, whereas resistance training alone has yielded less consistent results. Our study aims to address these limitations by incorporating a broader range of RCTs and analyzing outcomes across specific cognitive domains. Therefore, a comprehensive meta-analysis is needed to clarify the overall impact of exercise on cognitive function and quality of life in Alzheimer’s Disease patients.

This study evaluates the impact of exercise interventions on global cognitive function, assessed through standardized measures such as the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog), the Mini-Mental State Examination (MMSE), and the Montreal Cognitive Assessment (MoCA).

Methods

Systematic review approach

This study employed a systematic review and meta-analysis approach to evaluate the effects of exercise interventions on cognitive function and quality of life in patients with Alzheimer’s Disease. This study is registered under the identifier CRD42024599246. The review (Fig. 1) follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to ensure methodological rigor and transparency. Finally, 45 studies met the inclusion criteria and were included in the quantitative synthesis (meta-analysis).

Fig. 1.

Illustrates the entire process of study selection for the systematic review and meta-analysis, following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines

Literature search strategy

A comprehensive literature search was conducted in multiple electronic databases, including PubMed, Embase, and Web of Science. The search encompassed studies published from 1985 to November 2024. The following keywords and MeSH terms were used (as detailed in Table 1).

Table 1.

Search terms

Search term classifcation	Term
Intervention method	“Physical activity“[tiab] OR “Exercise“[tiab] OR “sport“[tiab] OR “Exercise“[mesh]OR “exergam*“[tiab] OR “train*“[tiab] OR “activity“[tiab]
Study subjects	“Alzheimer disease“[mesh] OR “Alzheimer disease“[tiab] OR “Alzheimer“[tiab] OR“dementia“[tiab]
Outcome indicator	“Randomized controlled trial“[pt] OR “Controlled clinical trial“[pt] OR “Randomizedcontrolled trial*“[tiab] OR “Trial“[tiab] OR “Random*“[tiab] OR “controlgroup“[tiab] OR “Single blind*“[tiab] OR “Double blind*“[tiab] OR “Tripleblind*“[tiab] OR “Randomized Controlled Trials as topic“[mesh] OR “RandomizedControlled Trial“[pt] OR “Controlled Clinical Trials as topic“[mesh] OR “ControlledClinical Trial“[pt] OR “Random Allocation“[mesh] OR “Single-Blind Method“[mesh]OR “Double-Blind Method“[mesh]
Cognition/Qualityof life	“Cognition“[mesh] OR “Cognition“[tiab] OR “cognitive“[tiab] OR “executivefunction“[tiab] OR “quality of life“[mesh] OR “quality of life“[tiab] OR “livingquality“[tiab] OR “quality living“[tiab]

Inclusion and exclusion criteria

Studies was included in this review if they meet the following criteria:

1. Participants: Patients diagnosed with Alzheimer’s Disease (AD) according to established diagnostic criteria (e.g., DSM-5, or NIA-AA guidelines). The DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, 5th Edition) was used for clinical cognitive assessments, ensuring consistency across included trials. If available, NIA-AA 2011/2023 criteria (National Institute on Aging – Alzheimer’s Association) were additionally used to confirm AD diagnosis based on biomarker evidence, such as amyloid PET imaging or cerebrospinal fluid analysis.

2. Interventions: Exercise interventions that include aerobic, strength, or stretching exercises;

3. Outcomes: Cognitive function was assessed using MMSE, MoCA, and ADAS-Cog.

Executive function was assessed using the Clock Drawing Test (CDT) and the Frontal Assessment Battery (FAB).

Memory was assessed using the Rey Auditory Verbal Learning Test (RAVLT) and the WHO–UCLA Auditory Verbal Learning Test.

• 4.Study Design: Randomized controlled trials (RCTs) with a control group receiving non-exercise interventions;
• 5.Language: Published in English.

  Exclusion Criteria:

  Study Type: Case reports, reviews, editorials, and non-empirical studies were excluded.

  Methodological Limitations: Studies without a control group were excluded.

  Confounding Factors: Studies involving confounding variables (e.g., additional pharmacological treatments or nutritional supplements in the intervention group compared to the control group) were excluded.

  Unavailable Data: Studies that did not provide complete data (e.g., reporting results only in graphical form without numerical values) were excluded.

• 6.Data Extraction and Quality Assessment.

  Two independent reviewers extracted data from the selected studies, including sample size, intervention details, outcome measures, and follow-up periods. Any discrepancies were resolved through discussion or consultation with a third reviewer. The quality of included studies was assessed using the Cochrane Risk of Bias Tool for RCTs.

Statistical analysis

Meta-analysis was performed using R software. Effect sizes were calculated using standardized mean differences (SMD) with 95% confidence intervals (CI). Heterogeneity was assessed using the I² statistic. A random-effects model was applied if significant heterogeneity is detected; otherwise, a fixed-effects model will be used.

Results

As shown in Table 2., the included randomized controlled trials employed diverse exercise interventions, primarily combining aerobic and resistance training, with several also incorporating balance, flexibility, and cognitive tasks. Most interventions lasted between 8 and 26 weeks, were conducted 2 to 5 times per week, and had session durations ranging from 30 to 60 min. The exercise intensity was typically moderate, though some studies applied either high-intensity or low-intensity protocols. Control groups mainly received usual care, no intervention, or non-exercise social activities.

Table 2.

Overview of exercise modalities and protocols in the included randomized controlled trials

Study	Intervention Method
	Experimental group	Duration (week)	Frequency (times/week)	Time (min/session)	Experimental group exercise intensity
Aguiar, 2014 [47]	Rivastigmine patch + combined exercise (aerobic, strength, balance, flexibility)	24	2	60	Moderate
Arcoverde,2014 [58]	Treadmill aerobic training	16	2	30	Moderate (60%VO2max)
Barreto, 2017 [53]	Exercise training (aerobic + strength)	24	2	60	-
Bossers, 2015 [46]	Combined training (aerobic + strength)	9	4	30	Moderate
Bossers, 2016 [61]	Aerobic training	9	4	30	Moderate
Coelho, 2012 [59]	Coelho, 2012	16	3	60	Moderate
Conradsson, 2010 [48]	High-intensity functional exercise (balance, strength, coordination)	12	5	60	High-intensity
De Oliveira Silva,2019 [60]	Multimodal training (aerobic, strength, balance, flexibility)	12	2	60	-
Deandrade,2013 [57]	Multimodal exercise (balance + cognitivetasks)	16	3	60	-
Dechamps, 2010 [54]	Individualized Cognition-Action (social interaction + exercise)	12	5	30
Eggermont, 2009a [52]	Hand movement program (fine motortasks)	6	5	30	Low (self-paced)
Eggermont,2009b [92]	Walking program (self-selected pace)	6	5	30	Low (self-paced)
El-Kader, 2016 [62]	Treadmill aerobic exercise (moderate intensity)	8	-	30	Moderate (60–70% max HR)
Gaitán, 2019 [63]	Progressive treadmill walking (70–80% heart rate reserve)	26	3	50	Moderate (70–80% heart ratereserve)
Henskens,2018a [64]	ADL training, multicomponent aerobic + strength training, or combined	24	-	-	-
Henskens,2018b [65]	ADL training (focus onself-care tasks)	24	-	-	-
Hoffmann, 2016 [66]	Supervised moderate-to-high intensity aerobic exercise (treadmill/cycle)	16	3	60	Moderate-to-vigorous (60–80% max HR)
Holthoff, 2015 [67]	Home-based leg training (passive/assisted/active resistive)	12	-	-	Low-to-moderate (adjusted to tolerance)
Huang, 2019 [56]	Modified Tai Chi (combined with cognitive tasks)	40	3	-	Low (gentle movements)
Kemoun, 2010 [68]	Walking, balance, and endurance training	15	3	60	Moderate (adjusted to ability)
Kim, 2024 [93]	Physical exercise (aerobic, strength) + multicomponent cognitive program	24	3	60	Unspecified
Kwak, 2008 [70]	Aerobic + strength training (chair exercises, stepping, balance)	52	2-3	30-60	Moderate (VO2max30%)
Lee, 2015 [69]	Fine motor skills + cognitive therapy (coloring, card games)	12	3	60	-
Morris, 2017 [71]	Aerobic exercise (treadmill/cycle)	26	5	30	Moderate (60–75% HRmax)
Nagy, 2021 [72]	Nagy, 2021	12	3	60	-
Öhman, 2016 [73]	Home-based (HE) or group-based (GE) exercises (strength, balance, flexibility)	52	2	60	-
Parvin, 2020 [74]	Dual-task training (muscle endurance, balance, flexibility, aerobic exercises with visual stimulation)	12	2	-	Progressive
Sampaio, 2019 [75]	Multicomponent training (aerobic, resistance, flexibility, postural exercises)	24	2	45-55	Moderate
Sanders, 2020 [76]	Aerobic + strength training	24	3	-	(12 weeks low→ 12 weeks high intensity)
Suttanon, 2013 [78]	Home-based balance, strength, walking exercises	24	-	-	-
Telenius, 2015 [77]	Functional exercise (strength & balance)	12	2	60	High (individualized intensity)
Todri, 2019 [79]	Global Postural Reeducation (GPR, spine/breathingexercises)	24	3	-	Low (postural control)
Toots, 2016 [80]	Lower limb strength & balance training	16	2	45	High (near maximal capacity)
Toots, 2017 [55]	Lower limb strength & balance training	16	2	45	High (near maximal capacity)
Vandewinckel,2004 [81]	Music-based exercises (rhythmic movements)	12	5	30	Moderate (music-paced)
Venturelli, 2011 [82]	Supervised walking program	24	5	30	Moderate (progressive)
Venturelli, 2016 [83]	Aerobic exercise ± cognitive training	12	5	60	Moderate (heart rate monitored)
Vidoni, 2019 [85]	Aerobic exercise (walking/cycling)	26	3	50	Moderate (60–80% max HR)
Vidoni, 2021 [84]	Aerobic exercise (150 min/week)	52	3-5	30-50	Moderate (target HRzone)
Vreugdenhil,2012 [86]	Community-based exercise (walking + ADL training)	16	Daily	30	Moderate (tailored)
Yágüez, 2011 [87]	Non-aerobic movement training (coordination tasks)	6	2	60	Low-moderate (movement complexity)
Yang, 2015 [88]	Cycling (70% max capacity)	12	3	40	Moderate (70% VO2max)
Yang, 2023 [89]	Sport StackingTraining: instructor-led sessions + home-based self-practice sessions	12	Clinic sessions: 1/week (first 7 weeks); Home practice: 5/week	Clinic: Unspecified; Home: 30	Low intensity (standing or seated
Yu, Vock, 2021 [91]	Supervised cycling	24	3	20-50	Moderate
Yu, Salisbury,2021 [90]	aerobic exercise (e.g.,walking, cycling)	24	3	30-50	Moderate (70% VO2max)

Figure 2 illustrates the effects of exercise interventions compared to control on cognitive function, assessed using the Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog). After excluding Toots 2017 [55] and Vreugdenhil 2012, a total of four randomized controlled trials remained for analysis. The pooled standardized mean difference (SMD) was − 0.09, with a 95% confidence interval (CI) of − 0.35 to 0.17, indicating a statistically significant improvement in cognitive function in the exercise group compared to the control group. The overall effect was not statistically significant (Z = 0.69, p = 0.49), and heterogeneity was low (I² = 0%, p = 0.50), suggesting consistent but non-significant results across the included studies. Exercise interventions did not significantly improve cognitive function in older adults as measured by ADAS-Cog, although the effect direction favored exercise and showed minimal heterogeneity.

Fig. 2.

Presents a forest plot analyzing the effect of exercise interventions on cognitive function in Alzheimer’s Disease patients, using the Alzheimer’s Disease Assessment Scale-Cognitive subscale (ADAS-Cog)

Figure 3 illustrates the effects of interventions compared to control on cognitive function, assessed using the Mini-Mental State Examination (MMSE). After excluding Toots, 2017 [55] and Vreugdenhil, 2012, a total of 21 randomized controlled trials remained for analysis. The pooled standardized mean difference (SMD) was 0.30, with a 95% confidence interval (CI) of 0.17 to 0.43, indicating a statistically significant improvement in cognitive function in the intervention group compared to the control group. The overall effect was statistically significant (Z = 4.67, p < 0.00001), and heterogeneity was moderate (I² = 36%, p = 0.05), suggesting relatively consistent results across the included studies with some variability.

Fig. 3.

Forest plot of effect of exercise intervention on cognitive function (MMSE)

Interventions significantly improve cognitive function in the analyzed population as measured by MMSE.

Figure 4: Assessment (MoCA). After excluding Parvin, 2020, three randomized controlled trials (56, 57, 72) were included in the analysis, with sample Sizes of 80 and 82 for the experimental and control groups, respectively. The pooled standardized mean difference (SMD) was 0.39, with a 95% confidence interval (CI) of 0.08 to 0.70, indicating a statistically significant improvement in cognitive function in the intervention group compared to the control group (Z = 2.46, P = 0.01).

Fig. 4.

Forest plot of effect of exercise intervention on cognitive function (MoCA)

Heterogeneity testing showed χ² = 0.51 (df = 2, P = 0.77) and I² = 0%, suggesting extremely low heterogeneity and excellent consistency across studies. Based on MoCA assessment, interventions have a small-to-moderate effect in improving cognitive function, with highly consistent results across different studies, providing robust evidence for the effectiveness of the interventions.

Figure 5: A total of 12 randomized controlled trials were included in the analysis of executive function outcomes, involving 393 participants in the exercise groups and 347 in the control groups. The pooled standardized mean difference (SMD) was 0.13 (95% CI: − 0.02 to 0.27), and the overall effect was not statistically significant (Z = 1.67, p = 0.09). Heterogeneity among the studies was very low (Chi² = 8.82, df = 11, p = 0.64; I² = 0%), indicating consistent findings across trials.

Fig. 5.

Forest plot of effect of exercise intervention on executive function

Figure 6: This forest plot illustrates the effects of interventions compared to the control group on memory function, assessed using an unspecified scale (details not explicitly mentioned in the document). After excluding Holthoff, 2015, nine randomized controlled trials were included in the analysis, with sample Sizes of 256 and 245 for the experimental and control groups, respectively. The pooled standardized mean difference (SMD) was 0.09, with a 95% confidence interval (CI) of − 0.09 to 0.26, indicating no statistically significant difference in memory function between the intervention and control groups (Z = 0.95, P = 0.34). Heterogeneity testing showed a chi-squared value of 5.67 (degrees of freedom = 8, P = 0.68) and an I² of 0%, suggesting extremely low heterogeneity and high consistency across studies. While most studies exhibited positive effect sizes (e.g., rcoverde, 2014; Bossers, 2015) [46], the overall effect did not reach statistical significance, possibly due to sample size limitations or differences in intervention types.

Fig. 6.

Forest plot of effect of exercise intervention on memory function

Figure 7: This forest plot illustrates the effects of interventions compared to the control group on quality of life. After excluding Conradsson, 2010 [48]; Holthoff, 2015; and Toots, 2016, a total of 11 randomized controlled trials were included in the analysis, with sample Sizes of 407 and 392 for the experimental and control groups, respectively. The pooled standardized mean difference (SMD) was 0.17, with a 95% confidence interval (CI) of 0.03 to 0.31, indicating a statistically significant improvement in quality of life in the intervention group compared to the control group (Z = 2.39, P = 0.02).

Fig. 7.

Forest plot of effect of exercise intervention on quality of life

Heterogeneity testing showed a chi-squared value of 10.51 (degrees of freedom = 10, P = 0.40) and an I² of 5%, suggesting extremely low heterogeneity and high consistency across studies. The results indicate that interventions have a small but significant positive effect on quality of life, with minimal variability in outcomes among the included trials.

Figure 8 provides visual evidence of publication bias for the meta-analysis of exercise interventions in Alzheimer’s disease. Regardless of its symmetry, its core value lies in informing the completeness of evidence and the robustness of conclusions—symmetric funnels strengthen the conclusion that exercise improves cognition and quality of life, while asymmetric funnels call for more high-quality studies to fill evidence gaps. Ultimately, integrating the full analysis, the positive effects of exercise interventions are highly credible within a reasonable methodological framework, supporting their role as an important adjunctive clinical intervention.

Fig. 8.

Funnel plot for publication bias assessment of exercise interventions on cognitive function and quality of life in Alzheimer’s disease patients

Figure 9: Egger’s test was used to quantify publication bias in the meta-analysis of exercise interventions on cognitive function and quality of life in Alzheimer’s disease (AD) patients. This test assesses the asymmetry of the funnel plot by regressing effect sizes against their precision (inverse of standard error), providing statistical evidence for potential publication bias.

Fig. 9.

Egger’s test assessing publication bias in studies of exercise interventions for Alzheimer’s disease patients

Figure 10 illustrates the detailed assessment of the methodological quality of the included studies, evaluating several bias-related domains: randomization process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selection of reported results. Each study is classified as either “low risk,” “some concerns,” or “high risk” in each domain, represented by green, yellow, and red symbols, respectively.

Fig. 10.

Methodological quality of the included studies

Figure 11 presents a summary of the risk of bias assessment for the studies included in the meta-analysis. The assessment covers key domains such as randomization, deviations from intended interventions, missing outcome data, measurement of outcomes, selection of reported results, and overall bias. The bars indicate the proportion of studies categorized as “low risk,” “some concerns,” or “high risk” in each domain.

Fig. 11.

Risk of bias summary (Intention-to-treat analysis)

Discussion

Principal findings

This systematic review and meta-analysis aimed to evaluate the effects of exercise interventions on cognitive function, executive function, memory, and quality of life in Alzheimer’s Disease (AD) patients. The results across various outcome measures suggest that exercise interventions, particularly aerobic and resistance exercises, may offer cognitive benefits for AD patients, although the magnitude and consistency of these effects vary across different domains. However, the magnitude and consistency of these effects vary across different cognitive domains, as evidenced by the meta-analyses conducted on ADAS-Cog, MMSE, MoCA, and memory function. The MMSE and MoCA demonstrated stronger effects than ADAS-Cog, which may be attributed to their broader sensitivity in detecting global cognitive changes, especially in early to moderate stages of Alzheimer’s Disease. In contrast, ADAS-Cog, although widely used, is primarily designed for detecting changes in more advanced stages and may exhibit a ceiling effect in earlier or less impaired populations.

Effects on cognitive function

This systematic review and meta-analysis synthesized evidence from 45 randomized controlled trials (RCTs) and indicated that exercise interventions can improve global cognitive function in patients with Alzheimer’s disease (AD), although effect sizes varied across cognitive domains. The pooled effect size for the Montreal Cognitive Assessment (MoCA) was moderate (SMD = 0.39, 95% CI: 0.08–0.70, p = 0.01, I² = 0%), indicating high sensitivity and consistency in capturing cognitive changes induced by exercise. The Mini-Mental State Examination (MMSE) also showed a statistically significant effect (SMD = 0.30, 95% CI: 0.17–0.43, p < 0.00001, I² = 36%), suggesting a general benefit of exercise on cognitive decline in AD. Although the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog) yielded a smaller effect size (SMD = − 0.09, 95% CI: − 0.35 to 0.17, p = 0.49), its low heterogeneity (I² = 0%) indicated consistent results across studies. These findings represent a core contribution of the present review. Compared with previous meta-analysis [33, 37], our study includes a larger number of RCTs and provides a more comprehensive examination of domain-specific cognitive outcomes, particularly executive function and memory. This enhances the precision of effect estimates and highlights the necessity for standardized exercise protocols in future trials. Physiologically, the beneficial effects of exercise on cognition in AD are attributed to multisystem neurobiological mechanisms, including enhanced neuroplasticity, improved cerebral perfusion, metabolic regulation, inflammation control, and modulation of the gut-brain axis. Several studies have suggested that exercise may upregulate brain-derived neurotrophic factor (BDNF), which is involved in synaptic plasticity and neuronal survival [38–40], which promotes neurogenesis, synaptic plasticity, and neuronal survival through activation of TrkB receptors and downstream PI3K/Akt and MAPK/ERK pathways—processes known to counteract hippocampal atrophy in AD. Several studies have demonstrated that exercise increases vascular endothelial growth factor (VEGF) and endothelial nitric oxide synthase (eNOS), thereby enhancing cerebrovascular function and mitigating cerebral hypoperfusion commonly observed in Alzheimer’s disease [41, 42]. Exercise has been shown to activate the AMPK–PGC1α signaling pathway, thereby enhancing mitochondrial biogenesis, energy metabolism, and oxidative phosphorylation, which collectively reduce metabolic dysfunction and insulin resistance in the Alzheimer’s disease brain [25]. Several studies have demonstrated that exercise suppresses neuroinflammation by shifting microglial activation from the pro-inflammatory M1 phenotype toward the anti-inflammatory M2 phenotype and downregulating cytokines such as IL-6 and TNF-α, thereby protecting synaptic integrity [40, 43, 44]. Exercise has also been shown to modulate the gut microbiota, enhance the production of short-chain fatty acids (SCFAs), and regulate neuroinflammation and neurotransmitter synthesis through the gut-brain axis. Collectively, these neurobiological mechanisms likely underlie the cognitive benefits of exercise in AD [29]. Collectively, these neurobiological mechanisms likely underlie the cognitive benefits of exercise in AD. From a clinical perspective, multimodal exercise programs—combining aerobic, resistance, and balance training, performed for at least 12 weeks at ≥ 150 min per week—are considered safe, cost-effective, and feasible in community or home settings, particularly for patients in the early to moderate stages of AD [45, 46]. This review also found that exercise significantly improved quality of life. Several studies have confirmed the positive effects of exercise on mood, social engagement, and daily functioning [47, 48]. Although heterogeneity was generally low in most analyses, variability in exercise types, frequencies, and intensities across studies remains a limitation. Moreover, the reliance on subjective scales in some trials calls for future studies to incorporate objective biomarkers such as BDNF, fMRI, or PET imaging. Stratified analyses based on sex, disease stage, and genetic background are also warranted to develop personalized and precise exercise prescriptions for AD. These efforts will further support the integration of exercise into routine clinical care for cognitive rehabilitation in AD.

Effects on executive function and memory function

This study reveals differential effects of exercise interventions on cognitive subdomains in patients with Alzheimer’s disease (AD), particularly executive and memory functions. This discrepancy suggests that exercise may have a more consistent effect on higher-order cognitive processes—such as planning, inhibitory control, and cognitive flexibility—whereas long-term memory, especially episodic memory, appears less responsive to physical activity.

Compared with prior reviews [33], which predominantly focused on global cognitive outcomes, our study is among the first to differentiate executive and memory functions in a domain-specific meta-analysis. This provides greater precision and clinical interpretability, advancing the field toward more personalized exercise prescriptions for AD. Mechanistically, improvements in executive function are likely supported by exercise-induced functional remodeling of the prefrontal cortex and enhanced cerebral perfusion. One study reported that aerobic exercise increases cerebral blood flow in the prefrontal regions, thereby enhancing neurovascular coupling and metabolic support for executive control networks [20]. Furthermore, exercise upregulates brain-derived neurotrophic factor (BDNF), which activates TrkB receptors and triggers PI3K/Akt and MAPK/ERK pathways, facilitating dendritic spine growth, Several studies have reported that exercise enhances synaptic plasticity, cortical integration, and task-related activation in the prefrontal cortex [49, 50], improving the efficiency of neural modulation during executive tasks, as shown through fNIRS measurements.

In contrast, the lack of significant improvement in memory function may reflect the early and irreversible structural damage to the hippocampus and entorhinal cortex in AD. These regions are among the first to be affected by amyloid-beta deposition and tau-related neurodegeneration, leading to neuronal apoptosis and reduced hippocampal neurogenesis. While animal studies have shown that exercise promotes hippocampal plasticity, including increased long-term potentiation (LTP) and neurogenesis in the dentate gyrus, the translational efficacy in human subjects—especially those with moderate to severe hippocampal atrophy—remains limited [25]. Additionally, most trials included in this review had relatively short intervention durations (8–12 weeks), which may not be sufficient to induce structural remodeling in memory-related circuits. Moreover, emerging evidence suggests that white matter integrity, particularly within frontoparietal tracts such as the superior longitudinal fasciculus, may improve with exercise, supporting executive function, whereas recovery of memory-related tracts such as the fornix may require neurogenesis and structural restoration.

Clinically, these findings support the use of structured exercise interventions to selectively improve executive function, which may translate into enhanced independence, emotional regulation, and adaptive behavior in patients with AD. Previous research [51] has reported that even modest improvements in executive function are positively associated with enhanced performance in instrumental activities of daily living (IADLs). However, the limited effect on memory underscores the need for combinational strategies—such as exercise plus cognitive training or nutritional supplementation—to more effectively engage and preserve memory circuits.

Despite encouraging results, this study has several limitations. The exercise modalities varied considerably, and the frequency, intensity, and duration were inconsistent across trials. Most studies relied on neuropsychological assessments without objective validation through neuroimaging or biomarkers. Furthermore, individual characteristics such as APOE ε4 status, age, or educational background were not consistently controlled. Future research should employ neuroimaging tools (e.g., fMRI, DTI), molecular assays (e.g., BDNF, IL-6), and long-term, high-frequency training protocols tailored to cognitive phenotypes. Such integrative approaches may help identify precise neural mechanisms underlying domain-specific improvements and advance the development of targeted exercise interventions in AD.

Effects on quality of life

The pooled analysis on quality of life demonstrated a statistically significant but modest improvement in Alzheimer’s disease (AD) patients receiving exercise interventions. This finding highlights a key outcome of the present analysis: exercise interventions in Alzheimer’s disease are associated not only with cognitive improvements but also with broader benefits to psychosocial well-being and functional health, consistent with previous findings reported in [46, 47]. Compared with earlier meta-analyses that emphasized cognitive gains, our study provides a domain-specific quantification of QoL effects using consistently low-heterogeneity evidence, thus establishing QoL as a reliable and clinically relevant endpoint.

The mechanistic basis for this improvement is multifactorial. Firstly, physical activity modulates the hypothalamic-pituitary-adrenal (HPA) axis by lowering cortisol reactivity, thereby reducing stress and anxiety symptoms commonly present in AD [26]. Secondly, exercise reduces peripheral and central inflammatory markers such as IL-6 and TNF-α, which are closely linked to behavioral disturbances and affective dysregulation [27]. At the neurochemical level, exercise increases brain-derived neurotrophic factor (BDNF), enhancing serotonergic and dopaminergic neurotransmission pathways that are essential for emotional regulation and mood stability [49, 50]. Furthermore, aerobic training has been shown to normalize circadian rhythms and improve sleep quality, which are often disrupted in dementia and critically influence perceived life satisfaction [52].

A novel and emerging explanation involves the role of the gut-brain axis. One study reported that regular exercise promoted beneficial shifts in gut microbiota composition, which in turn modulated neuroinflammation and neuroimmune balance through microbial metabolites such as short-chain fatty acids [37]. These systemic effects may contribute to a more stable neuropsychiatric profile, improving patients’ subjective experience and caregiver-reported quality of life. Structured group-based physical activities were shown to promote social engagement and improve self-efficacy among individuals in institutional settings, according to two randomized trials [53, 54].

From a clinical application perspective, the results reinforce that QoL should be treated as a core outcome rather than an ancillary benefit in exercise-based dementia interventions. Improvements in QoL not only reflect cognitive gains but also translate into reduced caregiver burden, delayed institutionalization, and increased adherence to non-pharmacological treatments. This has profound implications for long-term disease management strategies. One study reported that improvements in quality of life (QoL) following a 9-week aerobic and strength training program were associated with enhanced daily functioning and mobility [46]. Importantly, our study found this effect to be consistent across diverse geographic and intervention settings, with minimal heterogeneity, increasing its generalizability.

However, certain limitations must be acknowledged. While the heterogeneity was low, the measurement instruments for QoL varied across studies, including QOL-AD, EQ-5D, SF-36, and WHOQOL, potentially introducing conceptual discrepancies. Moreover, the absence of long-term follow-up data limits our ability to assess the sustainability of QoL improvements. Future studies should adopt unified QoL assessment tools and incorporate mixed-method designs (e.g., qualitative interviews) to better capture the nuanced impacts of exercise.

In summary, this analysis highlights quality of life as a multidimensional outcome that may be modestly improved through exercise interventions in AD patients. These improvements are underpinned by physiological, neuroimmune, and psychosocial mechanisms, providing a compelling rationale for incorporating structured, multimodal physical activity into routine dementia care.

Strengths and limitations

One of the strengths of this review is the inclusion of a large number of studies, allowing for a comprehensive evaluation of the impact of exercise interventions across multiple cognitive domains. The studies from [46, 55, 56]contributed valuable insights into the effects of exercise on cognition, executive function, and quality of life. Moreover, the use of standardized cognitive assessments such as ADAS-Cog, MMSE, and MoCA provides reliable insights into how exercise affects specific cognitive functions in AD patients.

Limitations. Several factors warrant cautious interpretation. First, exercise prescriptions varied widely in type, frequency, intensity, and duration, which likely contributed to between-study variability despite generally low-to-moderate heterogeneity in the pooled estimates. Second, many trials were of relatively short duration (often 8–12 weeks), potentially underpowering detection of changes in domains such as episodic memory. Third, most outcomes relied on neuropsychological scales without concurrent objective biomarkers or neuroimaging, limiting mechanistic inference. Fourth, risk-of-bias assessments (Fig. ) indicated concerns in randomization and missing data handling in some studies. Finally, a majority of studies were classified as having ‘some concerns’ in various domains (Fig. ). While funnel plots and Egger’s tests were performed where ≥ 10 studies were available, these methods are underpowered and cannot definitively rule out small-study effects or publication bias. Finally, although the included studies mainly involved participants aged ≥ 65 years, one study reported a minimum age of 51 years, which may have introduced some heterogeneity. However, as the vast majority of participants were older adults, the overall impact was likely limited. These factors collectively suggest that while exercise interventions are promising, their clinical benefits should be interpreted with caution.

Implications for practice and future research

The findings from this meta-analysis suggest that exercise interventions can be a valuable non-pharmacological approach to improving cognitive function and quality of life in Alzheimer’s Disease patients. However, given the variability in results, future studies should aim to standardize exercise protocols, ensuring consistent reporting of intervention details such as duration, intensity, and type of physical activity. Additionally, studies should strive for better methodological rigor, particularly in terms of randomization processes and handling of missing data.

Conclusion

In summary, this meta-analysis suggests that a multimodal exercise protocol combining aerobic (150 min/week moderate intensity), strength (2–3 days/week), and balance training (2–3 days/week), delivered over 12 + weeks with individualized intensity adjustments, may provide modest benefits in cognition and quality of life. Given modest effects and methodological variability, exercise should be framed as a complementary, low-risk option. Where feasible, multimodal programs performed ≥ 12 weeks may be considered, with expectations aligned to small average benefits.

Authors’ contributions

All authors contributed to the study design, data analysis, and manuscript writing. The final version was reviewed and approved by all co-authors.

Funding

This study received no external funding.

Data availability

The datasets generated during this study are not publicly available due to confidentiality restrictions but may be obtained from the corresponding author upon reasonable request.

Data availability

The datasets generated during this study are not publicly available due to confidentiality restrictions but may be obtained from the corresponding author upon reasonable request.

Declarations

Ethical approval and consent to participate

This study did not involve human participants, animal experiments, or clinical trials. Not applicable.

Consent for publication

All authors have read and approved the final manuscript and consent to its publication.

Consent for publication refers to consent for the publication of identifying images or other personal or clinical details of participants that compromise anonymity.

Competing interests

The authors declare no competing interests.