Effect of Physical Exercise on Fluid Intelligence Among the Elderly with or Without Dementia: A Systematic Review of Randomized Control Trials

Abstract

This review examines the impact of physical exercises on fluid intelligence (FI) among elderly individuals with or without dementia. Fluid intelligence declines with age and worsens faster with dementia, and studies suggest that physical exercise, like aerobic and strength training, may improve fluid intelligence by enhancing neurogenesis, promoting cerebral blood flow, and increasing brain plasticity. A comprehensive search was carried out for studies from inception to 31 May 2024 across databases, including PubMed, Scopus, Web of Science, and PEDro, using keywords related to “physical activity”, “physical exercise”, “fluid intelligence”, and “dementia”. The inclusion criteria focused on randomized control trials (RCTs) involving elderly participants with or without dementia, where fluid intelligence was measured using validated tools. The PEDro scale was used for the quality assessment of included studies. Risk of bias assessment was done using the Cochrane risk of bias tool version 2. Out of 1982 screened studies (PubMed: 104; Web of Science: 1676; Scopus: 195; PEDro: six), five RCTs involving 676 participants were included. Interventions lasted 4 to 24 weeks and included aerobic training, flexibility training, combined strength, and cognitive training, amongst others. It is seen that low- and high-intensity aerobic exercises improved FI, whereas another study reported that combined physical and mental activity significantly enhanced executive function and processing speed. Overall, all included studies demonstrated improvements in FI following exercise interventions. Physical exercises may support fluid intelligence in older adults, but current evidence is limited to a small number of trials. More robust studies are required.

1. Introduction

Dementia is defined by chronic, acquired loss of two or more cognitive abilities caused by brain disease or injury [1]. Alzheimer’s disease (AD) is the most frequent cause of dementia, which leads to an overall decline in memory and other thinking skills and contributes to 60–70% of the total cases of dementia. According to studies, one in every nine people 65 or older has AD [2,3]. With aging, the global prevalence of dementia has increased exponentially, and currently, more than 55 million people live with dementia worldwide, and this number is expected to reach 78 million and 139 million by 2030 and 2050, respectively [4].

According to the WHO (2021) global status report, dementia is currently the seventh leading cause of death and one of the major causes of disability and dependency among older people globally. Individuals suffering from the condition experience changes in cognition, function, and behavior, and the symptoms are gradual, persistent, and progressive in nature. The clinical presentation of dementia varies greatly among individuals, and the resulting cognitive deficits cause memory loss, communication and language impairments, agnosia (inability to recognize objects), apraxia (inability to perform previously learned tasks), and impaired executive function (reasoning, judgment, and planning) [5].

The main type of cognitive skill affected in dementia is fluid intelligence (FI), which refers to the ability to think logically and solve problems in new situations independently of acquired knowledge. With dementia or Alzheimer’s disease, there is a marked decline in FI with advancing age [6,7,8]. The concept of intelligence includes two separate, yet highly correlated components [9,10]: (i) a knowledge-based component, referring to abilities acquired through various forms of education and experience and commonly termed “crystallized intelligence” (CI), and (ii) a core of knowledge and language unrelated functions tightly related to the ability for problem solving and abstract reasoning, termed “fluid intelligence”(FI) [11,12,13]. Undermining the functioning of executive abilities results in a gradual decline of FI [14]. Any disturbances, whether CI or the FI, will affect a person’s ability to perform everyday activities independently and thus increase the emotional and physical burden to the caregivers as well as financial expenditure to the healthcare system.

Physical exercise or physical activity has well-documented benefits for general health and well-being, as well as cognitive functions. Epidemiological evidence consistently links physical activity with cognitive benefits, lower risk for dementia, and reduced pathological changes. A meta-analysis of 58 studies with over 250,000 participants reported that higher physical activity significantly reduced the risk of all-cause dementia, Alzheimer’s disease, and vascular dementia, regardless of age, follow-up duration, or study quality [15].

Although scientists are exploring conservative, disease-modifying pharmacological interventions, there is currently no known cure for dementia or AD [16]. Therefore, researchers shift their focus to interventions with the potential to delay the onset of the disease. As physical activity is known to benefit overall cognition and reduce dementia risk, its specific impact on fluid intelligence—especially in elderly individuals with and without dementia—remains underexplored. Very few studies investigated the effects of physical activity or compared the responses of different exercise interventions on cognitive functions, particularly FI. Thus, there is a need to highlight the need for a focused review of how physical activity and exercise interventions cause changes in fluid intelligence among older adults.

Objectives

The objectives of this study were to conduct a systematic review investigating the effects of exercise training on improving fluid intelligence in elderly individuals with or without dementia. Additionally, the review also aimed to evaluate how various training or physical activity programs influence outcomes and to analyze the relationship between the duration and intensity of exercise sessions and changes in FI-related outcomes. This review seeks to determine whether continuous physical activity contributes to improvements in fluid intelligence, focusing on studies that utilize appropriate tools or questionnaires to assess these outcomes.

2. Materials and Methods

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [17]. The review was registered with the International Prospective Register of Systematic Reviews (PROSPERO) on 22 June 2024, under registration number CRD42024554122.

2.1. Procedure

Two of the authors (S.J. and T.K.) exported all the search results into the EndNote reference manager, and duplicates were removed. The authors reviewed the titles and abstracts independently and screened them for suitability according to the criteria. Full texts for studies were screened to check whether the study met the inclusion criteria. PEDro scoring was done for all the RCTs to assess the quality of the studies; it was cross-checked by both authors separately, and any disagreements were resolved by discussion with the third author (M.A.). The PRISMA flowchart depicting the steps involved is shown in Figure 1. The PRISMA checklist of the study is given in Figure 2.

2.2. Eligibility Criteria

The inclusion criteria of the studies were: studies done on elderly people diagnosed with dementia. Studies in which physical activity or exercise protocols were given as an intervention and studies in which fluid intelligence was measured, or along with other outcomes via reliable tools or questionnaires, were included. The type of studies included were randomized controlled trials (RCTs) only in the English language. Studies that are in non-English languages, quasi-experimental studies, non-randomized trials, and those that do not include fluid intelligence as an outcome were excluded.

2.3. Search Strategy

A comprehensive search strategy was developed based on the Population, Intervention, Comparison, Outcome, and Study Type (PICOS) framework. The databases searched were PubMed, Scopus, Web of Science (Core Collection), and PEDro (Physiotherapy Evidence Database) for studies from inception to 31 May 2024, to maintain the objective in mind that we are evaluating the effect of physical activity or exercise on fluid intelligence in elderly individuals, with and without dementia. The search utilized keywords such as “Dementia [MeSH],” “Fluid Intelligence,” “Physical Exercise [MeSH],” “Physical Activity,” and “Elderly,” combined with Boolean operators “AND” and “OR” to ensure precision and comprehensiveness. The process was conducted collaboratively by S.J., T.K., and M.A., and the details of the search strategy are summarized in

Table 1. Search strategy of the databases.

Database	Search Strategy	Filter(s)
Scopus	“Elderly” AND “Physical activity” OR “Exercise” AND “fluid intelligence” AND “Dementia” AND “Physical activity” OR “Exercise” AND “Fluid intelligence”	Document type: Articles including case reports and trials Language: English
Web Of Science	https://www.webofscience.com/wos/woscc/summary/0b483cd4-5924-4dc8-9f5e-3f0e39cdbc4eea056ddb/relevance/5 (accessed on 24 October 2024)	Document type: Articles including case reports and trials Language: English
PubMed	((((Elderly) AND (Physical activity)) OR (Exercise)) AND (fluid intelligence)) AND ((((Dementia) AND (Physical activity)) OR (Exercise)) AND (Fluid intelligence))	Document type: Articles including case reports and trials Language: English
Physiotherapy Evidence Database (PEDro)	Exercise AND Fluid Intelligence	Document type: Articles including case reports and trials Language: English

.

2.4. Study Selection

The two authors (S.J. and T.K.) independently screened the studies for the titles and abstracts during the search. Studies that did not meet the inclusion criteria (e.g., irrelevant outcomes, non-elderly population, lack of physical activity intervention, or absence of fluid intelligence assessment) were excluded at this stage. The full texts of the remaining studies were assessed for eligibility. Whenever there was a disagreement in either the selection process, data extraction, or assessing the risk of bias, a third author (M.A.) reviewed the assessment, and a consensus was reached by discussion.

2.5. Data Extraction

Data extraction was done by two authors (S.J. and T.K.). The data extracted included the title of the study, author/s, year of publication, journals, country, and language of publication. The trial included study design, randomization, allocation concealment, and level of blinding. All the participants must be diagnosed with dementia or must be in the elderly age group. The sample size and participants’ distribution in each arm of the trial are mentioned. Intervention includes components within the intervention, mode of delivery, health professional involved, time and frequency of intervention, intensity of the intervention, and characteristics of the control group. The outcomes are associated with fluid intelligence and dementia, and the study conclusion and any reported adverse events are mentioned within the study.

A meta-analysis was not performed due to substantial heterogeneity across studies. The included trials varied in exercise modalities (aerobic, strength, flexibility, multimodal), intervention frequency, session duration, intervention length (4–24 weeks), and cognitive assessment tools used to measure fluid intelligence. These differences precluded statistical pooling of results and prevented meaningful subgroup or sensitivity analysis; therefore, a narrative synthesis was conducted.

2.6. Quality Assessment

The quality of assessment of the included clinical trials was performed by two independent authors (S.J. and T.K.) using the PEDro rating scale. This scale assesses 11 items related to the study’s internal validity and statistical reporting, except for the first one (eligibility criteria), which is not computed in the total score. Each item is scored as either present (1) or absent (0), leading to a maximum score of 10. The studies with scores ranging between 0 and 3 are considered “poor”, 4–5 are considered “fair”, 6–8 are considered “good,” and 9–10 are considered “excellent” [18]. If there was any disagreement on any criterion, it was re-assessed by each reviewer independently. Unresolved disagreements were identified and discussed in a consensus meeting. Any conflict that remained unresolved was then taken to a third author (M.A.), who was independent of the initial deliberations, and a final consensus was reached.

2.7. Risk of Bias Assessment

Cochrane’s risk of bias tool version 2 (RoB 2) was used to assess the risk of bias in the included studies. RoB 2 is structured into a fixed set of domains of bias, focusing on different aspects of trial design, conduct, and reporting [19]. Within each domain, a series of questions (“signalling questions”) aimed to elicit information about features of the trial that can provide useful insight to the reader about the methodological rigor followed by the investigators. Risk of bias assessment for all the included trials was done by three independent authors (S.P, C.A.S. and M.M.N.). A discrepancy check was later done to determine the overall bias of the included trials.

3. Results

3.1. Search Result

The electronic literature search was conducted across four major databases, PubMed (104), Web of Science (WOS) (1676), Scopus (195), and PEDro (6), resulting in a total of 1982 records found. After removing duplicate records (i.e., 60) through EndNote reference manager, the remaining studies were screened based on their titles and abstracts. A total of 1772 records were removed through the title and abstract screening. Out of the remaining 150 studies, the full texts of three studies were not retrieved. After analyzing the remaining 147 articles, 142 did not meet the inclusion criteria (not an RCT, not measuring fluid intelligence with validated tools, involving participants outside the target population, i.e., non-elderly or without a dementia-related focus, inadequate reporting of intervention details or outcome measures). Finally, five studies were included in the qualitative synthesis and data extraction phase. These studies collectively involved 676 participants and evaluated the effect of various physical activity or exercise interventions on fluid intelligence in elderly individuals with or without dementia.

3.2. Qualitative Assessment

The methodological quality of the included studies was evaluated using the Physiotherapy Evidence Database (PEDro) scale, a validated tool specifically designed to assess the internal validity and statistical reporting of randomized controlled trials. In this review, four studies conducted by Galle et al. (2023) [20], Linde & Alfermann (2014) [21], Sanders et al. (2020) [22], and Brown et al. (2009) [23] achieved a good quality score of 6/10, while Klusmann et al. (2010) [24] received an excellent score of 9/10, as the study implemented all three levels of blinding, contributing to its higher score. All studies met the criteria for random allocation, baseline comparability, outcome reporting, and statistical analysis. However, common methodological limitations included a lack of blinding of subjects, therapists, and assessors, which can introduce performance and detection bias. These findings reflect the overall reasonable quality of the included trials. The details of the qualitative assessment are shown in

Table 2. PEDro scoring of the studies.

Criteria	Galle et al., 2023 [20]	Linde and Alfermann, 2014 [21]	Sanders et al., 2020 [22]	Brown et al., 2009 [23]	Klusmann et al., 2010 [24]
1. Eligibility criteria were specified	yes	yes	yes	yes	yes
2. Subjects were randomly allocated to groups (in a crossover study, subjects were randomly allocated an order in which treatments were received)	yes	yes	yes	yes	yes
3. Allocation was concealed	no	yes	no	yes	yes
4. The groups were similar at baseline regarding the most important prognostic indicators	yes	yes	yes	yes	yes
5. There was blinding of all subjects	no	no	no	no	yes
6. There was blinding of all therapists who administered the therapy	no	no	no	no	yes
7. There was blinding of all assessors who measured at least one key outcome	no	no	no	no	yes
8. Measures of at least one key outcome were obtained from more than 85% of the subjects initially allocated to groups	yes	yes	yes	yes	yes
9. All subjects for whom outcome measures were available received the treatment or control condition as allocated or, where this was not the case, data for at least one key outcome was analyzed by “intention to treat”	yes	yes	yes	no	no
10. The results of between-group statistical comparisons are reported for at least one key outcome	yes	yes	yes	yes	yes
11. The study provides both point measures and measures of variability for at least one key outcome	yes	yes	yes	yes	yes
Total	6/10	6/10	6/10	6/10	9/10

.

3.3. Assessment of Risk of Bias

The risk of bias assessment is illustrated in Figure 3. Among the five studies evaluated, three studies [23,24,25] exhibited some concerns regarding the randomization process. The remaining two studies [22,26] were classified as having a low risk of bias in this domain. All studies [20,21,22,23,24] demonstrated a low risk of bias concerning deviations from the intended intervention (domain 2). Additionally, all studies [20,21,22,23,24] were assessed as having a low risk of bias related to missing outcome data. In the domain of outcome measurement, three studies [21,22,24] were found to have a low risk of bias, while one study [23] also had a low risk of bias, and another study [20] was classified as having a high risk of bias. For the domain concerning the selection of the reported result, four studies [20,21,22,23] exhibited either some concern or low risk, with only one study [24] showing a low risk of bias. Overall, the risk of bias assessment indicated that four studies [20,21,22,23] presented some concerns, while one study [24] was assessed as having a low risk of bias.

3.4. Characteristics of the Participants

The five included RCTs involved a total of 676 elderly participants, aged 55 to 95 years, encompassing a broad spectrum of aging from the “young-old” to the “oldest-old.” Sample sizes ranged from 70 to 259. All studies [20,21,22,23,24] included both male and female participants, though gender distribution varied; for example, Klusmann et al. (2010) [24] focused exclusively on older women, while others, such as Sanders et al. (2020) [22] and Galle et al. (2023) [20], included mixed-gender cohorts. Participants also differed in baseline cognitive status, physical activity levels, and health backgrounds, ranging from sedentary individuals with no prior exercise experience to relatively healthy, low-active elderly adults. This demographic and clinical diversity underscores the heterogeneity of the study populations and highlights the importance of interpreting outcomes within the context of age, gender, and baseline functional status.

3.5. Intervention

The five randomized controlled trials reviewed employed a variety of physical and cognitive interventions, differing in type, intensity, frequency, and duration. Brown et al. (2009) [23] administered a group-based program comprising flexibility exercises and relaxation techniques, conducted for 1 h twice weekly over 6 months. Galle et al. (2023) [20] utilized the COACH protocol, involving moderate-intensity physical activity with integrated coaching elements, conducted twice weekly for 24 weeks, with each session lasting 45–60 min. Sanders et al. (2020) [22] delivered aerobic and strength training interventions twice per week for 12 weeks, with each session lasting 60 min. Linde and Alfermann (2014) [21] divided participants into three groups, one receiving aerobic and strength training (twice weekly, 60 min), one undergoing cognitive activity (once weekly, 30 min), and one combining both, over a 16-week period. Klusmann et al. (2010) [24] implemented a six-month multimodal intervention involving aerobic, strength, flexibility, balance, and coordination exercises, delivered in group settings; the comparison group received computer-based cognitive activity. Session durations across studies ranged from 45 to 60 min, typically twice weekly over intervention periods spanning 4 to 24 weeks. This variability reflects diverse approaches in structuring physical and cognitive activity interventions for elderly populations. The intervention characteristics of the included studies are presented in

Table 3. Characteristics of the RCTs.

Trials	Participant Characteristics	Experimental Group Intervention	Control Group Intervention	Outcome Measures	Major Findings
Galle et al., 2023 [20]	n = 102 Age: 55+ yrs M/F: 25/77	COACH-based moderate physical activity intervention (walking, cycling, housekeeping, gardening)	Stretching exercises	Pedometer, PASE, 15-WT, TMT, Stroop test, SPPB, and 6MWT	Participants in the COACH-based physical activity group maintained cognitive function better than controls, with statistically significant improvements in gait speed and step counts (p < 0.05) and trends toward cognitive benefit
Linde and Alfermann, 2014 [21]	n = 70 Age: 67.06 ± 3.55 yrs M/F: 41/29	Gr1: Aerobic endurance + strength training. Gr2: Cognitive activity Gr3: Combined physical activity + cognitive activity	Usual routine + fitness class	PTS, LPS 50+, d2 test of attention, TMTA, DSST, and CVF	Cognitive speed significantly improved in the combined intervention group compared to control (p = 0.01), with a large effect post-intervention (d = 0.86) and moderate at follow-up (d = 0.56). Concentration improved in all intervention groups post-intervention (highest in the cognitive group, d = 0.88), but at follow-up, only the physical activity group maintained the effect (d = 0.46). No significant changes were observed in short-term memory, reasoning, spatial relations, or processing speed (p > 0.05)
Sanders et al., 2020 [22]	n = 91 Age: 82 ± 7.0 yrs M/F: 32/59	Gr1: Combined aerobic (walking) + lower limb strengthening exercises	Flexibility exercises + recreational activities	6MWT, SPPB, FICSIT-4, MMSE, TMT-A, DSFW, DSBW, Stroop test, VMSFW, and VMSBW	Significant improvements in delayed story recall in the combined aerobic and strength group compared to control (p < 0.001) and improvements in executive function (Trail Making Test B/A) (p = 0.04)
Brown et al., 2009 [23]	n = 154 Age: 62–95 yrs M/F: 19/135	Gr1: General group-based exercise (GE)—resistance training and balance training. Gr2: Flexibility and relaxation (FR)—gentle joint movements, trunk/neck rotation, and rhythmic breathing	No-exercise control group	6MWT, WAIS-R, TMT-B, Stroop–CW, COWAT, WMS-R, GDS, and PANAS-P	Group-based exercise significantly enhanced memory and cognitive flexibility compared to both a stretching control and a passive group, with main effects showing p < 0.05 for most outcomes
Klusmann et al., 2010 [24]	n = 259 Age: 70–93 yrs M/F: 0/259	Gr1: Aerobic endurance, strength, flexibility, balance, and coordination exercise	Computer training; continued routine daily activities	VFA, RBMT, FCSRT, TMT A&B, and Stroop test	Both physical exercise and computer training significantly improved episodic memory (RBMT immediate recall: p = 0.007; delayed recall: p = 0.01) and preserved working memory, with no significant difference between the two active groups

n—sample size, Gr1—group 1, Gr2—group 2, yrs—years, M—male, F—female, 6 MWT—6 Minute Walk Test, GE—group-based Ex, WAIS-R—Wechsler Adult Intelligence Scale—Revised, TMT-B—Trail Making Test (part B), Stroop–CW—Stroop neuropsychological screening color–word test, COWAT—Controlled Oral Word Association Test, WMS-R—Wechsler Memory Scale—Revised, GDS—Geriatric Depression Scale, PANAS-P—Positive and Negative Affect Schedule, PASE—Physical Activity Scale for the Elderly, 15-WT—15-word test delayed recall, SPPB—Short Physical Performance Battery, FICSIT-4—Frail and Injuries: Cooperative Studies of Intervention Techniques, MMSE—Mini-Mental State Examination, VMSFW and VMSBW—Visual Memory Span Forward and Backward, PTS LPS 50+—Performance Test System; Leistungs–Prüf-System 50+, TMT-A—Trial Making Test (part B), DSST—digit symbol substitution test, CVF—Cardiovascular Fitness, VFA—Verbal Fluency Assessment, RBMT—Rivermead Behavioral Memory Test, FCSRT—Free and Cued Selective Reminding Test, RCT—randomized control trial.

, while the detailed exercise prescription parameters—type, frequency, intensity, and duration tailored for elderly populations—are summarized in

Table 4. Exercise prescription parameters in included randomized controlled trials.

Study	Exercise Mode	Intensity	Frequency	Session Duration	Intervention Duration	Relevant to the Elderly Population
Galle et al., 2023 [20]	COACH-based moderate physical activity (walking, cycling, household activities)	Moderate intensity guided by pedometer-based step goals	2x/week + daily activity encouragement	45–60 min	24 weeks + 9-month follow-up	Progressive increase in daily steps made activity feasible and safe for sedentary elderly people
Linde & Alfermann, 2014 [21]	Aerobic endurance + strength training ± cognitive training	Moderate intensity (not quantified)	2x/week (physical); 1x/week (cognitive)	60 min (physical)/30 min (cognitive)	16 weeks	Combined physical + cognitive training showed the largest gains; intensity was kept moderate to avoid fatigue
Sanders et al., 2020 [22]	Walking (aerobic) + lower limb resistance training	Weeks 1–12: 57–63% HRmax (RPE 9–11, low) Weeks 13–24: 83–89% HRmax (RPE 15–16, high)	3x/week	30 min/session	24 weeks	High-intensity phase is tolerated only after 12-week gradual conditioning; low-intensity is a suitable baseline for frail adults with dementia
Brown et al., 2009 [23]	Group-based strength + balance exercises	Not reported explicitly; supervised low-moderate effort	2x/week	60 min	24 weeks	Group format improved adherence; resistance was kept low initially for institutionalized seniors
Klusmann et al., 2010 [24]	Multimodal: aerobic, strength, balance, flexibility, coordination	Moderate, progressively increased	2x/week	50–60 min	24 weeks	Exercises are scalable for very old adults (70–93 yrs); multimodal design avoided exertion overload

HRmax—Maximum Heart Rate; RPE—Rating of Perceived Exertion; COACH—Coaching Older Adults and Caregivers at Home; min—minutes.

.

3.6. Control

Control groups in the included RCTs received either minimal or low-intensity interventions. Brown et al. (2009) [23] used a no-treatment control, while Galle et al. (2023) [20] and Sanders et al. (2020) [22] provided stretching or recreational flexibility exercises, respectively. Linde and Alfermann (2014) [21] offered general fitness activities without cognitive components, and Klusmann et al. (2010) [24] used a computer-based cognitive training program as a non-physical control. These varied controls helped isolate the effects of targeted physical or cognitive interventions.

3.7. Outcome Measures

The outcome measures reported in the included trials were fluid intelligence, executive function, and cognition. The tool and scales used were the Wechsler Adult Intelligence Scale—Revised (WAIS-R), Trial Making Test (TMT-A and TMT-B), Stroop color word test, Wechsler Memory Scale—Revised (WMS-R), Visual Memory Span Forward and Backward (VMSFW and VMSBW), Leistungs–Pruf System 50+ (LPS 50+), and digit symbol substitution test (DSST). Studies selected for the review are summarized in Table 3.

3.8. Effects on Fluid Intelligence

Evidence from five randomized controlled trials suggests that physical activity, cognitive training, or their combination can significantly enhance fluid intelligence in older adults, including those with or without cognitive impairment, as assessed through validated neuropsychological tools. Galle et al. (2023) [20] evaluated the effects of a 9-month COACH-based physical activity intervention in sedentary older adults. Although no group-level effects were observed, participants who increased their daily physical activity by ≥35% showed significant improvements in verbal memory, executive function, and global cognition (p < 0.05), as measured by the Trail Making Test (TMT), Stroop test, and 15-word test, highlighting the importance of adherence and behavioral engagement. Linde and Alfermann (2014) [21] conducted a 16-week RCT comparing physical, cognitive, and combined interventions and found that the combined group showed significant improvements in cognitive speed (p = 0.01, d = 0.86), sustained at 3-month follow-up (d = 0.56). Concentration improved across all intervention groups, most notably in the cognitive group (d = 0.88), while only the physical activity group maintained this gain at follow-up (d = 0.46). These outcomes were assessed using the LPS 50+, d2 Test of Attention, digit symbol substitution test (DSST), and Trail Making Test A. In a dementia population, Sanders et al. (2020) [22] demonstrated that a 24-week combined aerobic and strength training program led to significant gains in delayed story recall (p < 0.001) and executive function (TMT B/A, p = 0.04), as evaluated using the Trail Making Test, Stroop test, and verbal memory assessments, thus supporting cognitive benefits even in impaired individuals. Brown et al. (2009) [23] assessed the impact of a 6-month group-based resistance and balance training program in older adults residing in retirement facilities and observed significant enhancements in fluid intelligence (p < 0.05), as measured by subtests from the Wechsler Adult Intelligence Scale—Revised (WAIS-R), Stroop color–word test, and Controlled Oral Word Association Test (COWAT), compared to both flexibility-based and no-exercise controls. Lastly, Klusmann et al. (2010) [24] compared physical exercise with a computer-based cognitive training program among women aged 70–93 years, finding that both interventions produced significant improvements in episodic memory (Rivermead Behavioral Memory Test: p = 0.007 for immediate recall, p = 0.01 for delayed recall) and preserved working memory, as assessed by the Free and Cued Selective Reminding Test (FCSRT) and Trail Making Tests A and B. No significant difference was observed between intervention types, underscoring the comparable value of both physical and cognitive enrichment in promoting fluid cognitive function. Collectively, these findings support the use of structured physical and cognitive interventions, particularly in combination, to produce measurable and often sustained improvements in fluid intelligence among older adults.

4. Discussion

This systematic review underscores the potential role that physical exercise plays in supporting cognitive health, particularly fluid intelligence (FI) and episodic memory (EM), in elderly individuals with or without dementia; however, given that only five randomized controlled trials met the inclusion criteria, the current evidence remains preliminary. Across numerous studies, regular exercise consistently emerged as a key intervention to slow or mitigate age-related cognitive decline. Notably, this review adds depth to our understanding of how physical activity differentially affects specific cognitive domains such as FI and EM.

The frequency and consistency of exercise appear to be particularly influential. For instance, Kachouri et al. (2022) [25] observed that exercising at least twice per week significantly reduced the negative effects of aging on both FI and EM. These findings are in line with Hess et al. (2014) [26], who reported substantial cognitive improvements on the Mini-Mental State Examination (MMSE) and Alzheimer’s Disease Assessment Scale (ADAS) following regular physical activity. Executive functions, which are central to reasoning, planning, and problem solving, seem especially responsive to the effects of exercise and may serve as a bridge for improvements in both FI and EM [27]. On a biological level, physical activity is linked to enhanced brain function through mechanisms like increased neurogenesis, improved synaptic plasticity, and elevated BDNF levels [28]. These benefits are in line with the Scaffolding Theory of Aging and Cognition (STAC), which suggests that exercise supports the brain’s ability to reorganize and compensate for structural decline [29].

Interestingly, some studies suggest that EM might be more sensitive to the benefits of exercise than FI, likely due to its strong dependence on the prefrontal cortex and executive functioning [8]. These effects are especially pronounced in older adults, whereas younger individuals who typically have higher baseline cognitive performance tend to show limited cognitive gains from exercise [30,31]. Evidence from very old populations further strengthens these findings. In their study of individuals aged 85 and above, Ho et al. (2024) found that targeted exercise, like aerobic and resistance training, significantly improved FI and processing speed [32]. Similarly, Ho et al. (2023) emphasized that the type of exercise plays a critical role in maintaining cognitive flexibility in later life [33].

The mode of physical activity also matters. For example, Lv et al. (2022) demonstrated that Tai Chi, a gentle, coordinated mind–body practice, can lead to notable gains in both FI and executive functioning, making it a suitable intervention for older adults who may not tolerate high-intensity workouts [34]. Similarly, Linde & Alfermann (2014) showed that combining physical and cognitive training resulted in greater cognitive improvements than either intervention alone, supporting the potential of integrated approaches [21].

It is important to recognize that physical activity is not the only path to cognitive enhancement. Tranter & Koutstaal (2008) found that mentally stimulating activities alone can also lead to improvements in FI, suggesting that both cognitive and physical components should be targeted in dementia-prevention strategies [35]. Meanwhile, Brown et al. (2009) showed that group-based exercise benefits cognition independently of mood improvements, confirming the direct effects of physical training [23]. Additionally, Ferrucci et al. (1993) linked lower FI levels in elderly individuals to greater disability and dependence, underscoring the real-world implications of maintaining cognitive function [36].

By reducing the progression of age-related decline in FI and EM, regular exercise provides an accessible and cost-effective strategy for promoting mental well-being. The growing body of evidence highlights the intricate connection between physical activity, executive function, and brain plasticity, making a compelling case for its integration into both public health policies and dementia care frameworks. In summary, this review suggests that physical activity has the potential to support cognitive functioning, including aspects of FI, in older adults; however, due to the small number of available trials and considerable methodological heterogeneity, the evidence is not yet sufficient to establish firm conclusions.

It may be noted that when interpreting these findings, it is important to consider the methodological limitations of the included trials. Several studies lacked blinding of participants, therapists, or outcome assessors, which may have introduced performance or detection bias. Inadequate allocation concealment was also noted in some studies, increasing the risk of systematic differences between groups. Given that cognitive outcomes are susceptible to expectancy effects and assessors’ influence, these sources of bias may have inflated the reported benefits of exercise. Furthermore, many studies rely on self-reported activity levels, which are prone to bias and inaccuracy [25,26]. Consequently, the observed improvements in FI should be interpreted with caution.

4.1. Strengths

This systematic review offers several notable strengths that enhance the credibility and impact of its findings. Firstly, the study followed a rigorous methodological approach, adhering strictly to PRISMA guidelines, ensuring transparency and methodological integrity. A unique strength of this review lies in its focus on fluid intelligence (FI)—a cognitive domain often neglected in prior research despite its critical relevance to aging and independence. The inclusion of diverse physical activity interventions, ranging from aerobic and resistance training to combined cognitive–physical protocols, highlights the flexibility of exercise in promoting cognitive health among elderly individuals with and without dementia. In addition to examining outcomes, the review provides insight into underlying mechanisms, such as increased neurogenesis, synaptic plasticity, and brain-derived neurotrophic factor (BDNF) release, supporting neurobiological theories like the Scaffolding Theory of Aging and Cognition (STAC). The inclusion of age-specific subgroups, particularly the “oldest-old” (aged 85+), expands the relevance and applicability of the findings to more vulnerable aging populations. This systematic review reaffirms the role of physical activity as a cornerstone for preserving and enhancing cognitive health in aging populations.

4.2. Limitations

The overall number of studies that met the inclusion criteria was relatively small, with only five randomized controlled trials specifically addressing fluid intelligence, which limits the statistical power and generalizability of the conclusions. Moreover, there was significant heterogeneity across the included studies in terms of intervention type, frequency, intensity, and duration, making it challenging to determine the most effective exercise parameters for cognitive benefit. Most interventions were short-term, typically lasting 4 to 24 weeks, which may not be sufficient to assess the long-term effects of physical activity on cognitive health. Another drawback of the review is the variability in the tools used to assess fluid intelligence, such as WAIS-R, LPS 50+, and DSST. The heterogeneity of interventions and outcome measures limited the feasibility of conducting a meta-analysis.

The real-world applicability of these interventions remains uncertain. Older adults may face clinical, social, and logistical barriers such as co-morbidities, limited mobility, low motivation, and restricted access to supervised exercise settings. Additionally, adherence and dropout rates were inconsistently reported across the included studies, limiting conclusions about the feasibility of implementing these protocols outside controlled environments.

4.3. Future Scope

Future research should address these limitations by conducting larger, more diverse RCTs with standardized exercise protocols and objective measures of physical activity. Long-term studies are needed to explore the sustained effects of physical activity over the lifespan, and advanced tools such as neuroimaging could offer deeper insights into the structural brain changes, such as the prefrontal cortex and hippocampus, associated with exercise [37,38]. Further exploration into the dose–response relationship between exercise frequency, intensity, and cognitive outcomes is essential to establish evidence-based guidelines.