Effect of Water Temperature during Aquatic Exercise on Cognitive Impairment in Mice with Chronic Cerebral Hypoperfusion

Abstract

Introduction

Given the lack of effective pharmacotherapy for vascular dementia (VaD) and the reported benefits of exercise, this study examined the impact of water temperature during aquatic exercise (AE) on cognitive and pathological outcomes in a mouse model of VaD.

Methods

Twelve-month-old male mice underwent bilateral common carotid artery stenosis (BCAS) to induce chronic cerebral hypoperfusion. Groups included sham, BCAS control (no exercise), and BCAS mice assigned to AE in water baths at 25°C, 30°C, or 35°C. Cognitive performance was assessed using the elevated plus maze, Y-maze, and novel object recognition test (NORT). Brain tissues were analyzed for microglial marker CD68, astrocytic marker GFAP, and myelin basic protein (MBP) in the corpus callosum.

Results

The 30°C AE group showed the greatest improvement in NORT performance and swim activity. CD68 expression was unchanged across groups, but GFAP expression was significantly reduced at 30°C, suggesting suppressed astrocyte activation. Furthermore, the decline in MBP expression after BCAS was attenuated in this group, indicating preserved white matter integrity.

Conclusion

AE at approximately 30°C alleviated cognitive deficits in a VaD model, likely by reducing neuroinflammation and protecting myelin. These findings highlight the therapeutic potential of optimizing AE conditions, particularly water temperature, for dementia-related rehabilitation.

Introduction

Cerebrovascular diseases are mainly caused by decreased blood flow to the brain, which results in chronic cerebral hypoperfusion (CCH). CCH not only causes hypoxia but also impairs neurovascular coupling and leads to complex metabolic changes in the central nervous system (CNS), resulting in cerebral cell death and increased vascular inflammatory factors [1]. The second most frequent form of dementia after Alzheimer’s disease (AD) is vascular dementia (VaD), a type of vascular cognitive impairment resulting from cerebrovascular dysfunction [2]. The pathophysiology of VaD is diverse and often coexists with AD indicators, making diagnosis and treatment complex [2–4].

VaD accounts for a significant portion of dementia cases globally (15–30%), and the resulting decline in daily living capacity leads to a greater care burden than AD [5, 6]. Unlike AD, no approved drugs are pharmacologically effective for VaD. Therefore, the current primary treatment involves controlling vascular risk factors, and symptom-modifying drug treatment is the primary pharmacological approach for dementia [7].

However, non-pharmacological treatments, particularly exercise therapy, have been widely reported to improve cognitive function in elderly people and patients with dementia [8–10]. Given that reduced activity often accompanies dementia, efficient exercise can play a positive role in both treating the condition and preventing complications [11–17]. Aquatic exercise (AE), in particular, has been shown to be more beneficial than land-based exercise in improving balance, gait, and quality of life [18, 19]. Some clinical cases have shown that therapeutic AE, such as immersion in a 32°C pool, can relieve symptoms in severe AD patients where conventional drug therapy has failed [20–23].

Although AE can be effectively used for CNS diseases, reports on its application to VaD are rare, and research comparing the results based on differences in water temperature is lacking. Therefore, this study used a well-known VaD mouse model to investigate the impact of different water temperatures (25°C, 30°C, and 35°C) during voluntary AE on changes in cognitive function and neuroinflammation-related cell markers in mice with vascular cognitive impairment.

Materials and Methods

Animals

Eight-week-old male C57Bl/6 mice were purchased from SAMTAKO (Gyeonggi-do, Republic of Korea) and acclimatized to the animal breeding room. In order to simulate elderly humans, the mice were reared until they were 12 months old. Throughout the breeding and experimentation phase, the mice were kept in an environment with a light-dark cycle of 12 h, a room temperature of approximately 24 ± 4°C, and a 50% humidity level. They were also given unlimited access to food and water. Mice were weighed once every 2 weeks during the experimental period.

Reagents

Bio Basic Inc. (Markham, ON, Canada) provided the phosphate-buffered saline (PBS), Sigma-Aldrich Co. (St. Louis, MO, USA) supplied the 4% paraformaldehyde, Junsei Chemical Co. Ltd. (Chuo-ku, Tokyo, Japan) produced the sucrose, and Choongwae Pharmaceutical (Seocho-gu, Seoul, Korea) supplied the saline solution. The company Scigen Scientific, Inc. (Gardena, CA, USA) provided the compound with the optimal cutting temperature.

Bilateral Common Carotid Artery Stenosis-Induced CCH

By using micro-coils (0.18-mm inner diameter, Hamamatsu, Japan) to restrict both carotid arteries, bilateral common carotid artery stenosis (BCAS) causes a persistent reduction in blood flow to both hemispheres and causes VaD in mice (Fig. 1a). During BCAS, 2.0% isoflurane was administered through the respiratory system along with a 70% N₂O and 30% O₂ mixture to provide anesthesia. Following anesthesia, the mice were placed on the operating table in a supine posture. Their body temperature was maintained at 36.5 ± 0.5°C using an animal rectal homeothermic monitoring system (Harvard Instruments, Holliston, MA, USA) and a heated pad connected to the homeothermic system.

Fig. 1.

Schematic diagram of CCH using micro-coils and experimental processes. a Micro-coils are wrapped around both CCAs to induce CCH in experimental mice. b A schematic view of experimental process, including adaptation to the breeding room, experimental surgery on the mice, AE in water bath, behavior test, and euthanasia for brain collection. CCH, chronic cerebral hypoperfusion.

A depilator and clipper were used to remove the hair from the mouse’s front neck and surrounding area, which served as the surgical site. Povidone and 70% ethanol were used for disinfection (Deoksan General Science, Seoul, Korea). In order to properly separate the two common carotid arteries (CCAs) from the surrounding connective tissue, a midline incision was performed in the anterior neck under a stereomicroscope. After that, a micro-coil was used to wrap the CCAs on both sides. After applying a suture line to the skin, povidone was used to sterilize the surrounding tissue. The operated mice were transferred to a recovery cage. After confirming that they had sufficiently recovered from anesthesia, they were reared in an animal breeding room.

Division of Experimental Groups and Load of AE

The experimental group was divided into five groups: a sham surgery group (sham group), a control group in which BCAS surgery was performed (BCAS control group), and three experimental groups (BCAS + AE25, AE30, and AE35, respectively) in which exercise was performed in a water tank at 25, 30, and 35°C along with BCAS surgery. The AE was performed freely for 30 min at three individual water temperatures three times a week for 2 weeks. The water level was set such that the 4 feet of the mouse could touch the bottom of the water bath, thereby encouraging voluntary AE and avoiding the forced nature of a traditional swim test.

Measurement of Body Weight, Physiological Parameters, and Forepaw Grip Strength

The mice were weighed every 2 weeks during the experiment, and following behavioral assessments, blood was drawn from inferior vena cava while the mice were deeply sedated. Blood samples were centrifuged at 1,500 g for 15 min at 4°C in order to extract plasma. Potential electrolyte imbalances were tracked by utilizing an electrolyte analyzer (Dri-Chem 3500i, Fuji, Tokyo, Japan) to measure the plasma concentrations of electrolytes such as sodium, potassium, and chloride.

The experimental mice were positioned on a wire grid attached to a specially designed grip test apparatus. The mice grasped the grid using both front paws. Subsequently, the grip is gradually released. The maximum force produced was measured in gram-force. Each test was conducted in triplicate for each mouse, and the mean value was recorded as the measured value.

Behavioral Disorder Assessment Using Elevated Plus Maze Test, Y-Maze Spontaneous Alternation Test (Y-Maze Test), and Novel Object Recognition Test

At week 3, 1 week before the euthanizing for brain collection, the mice were acclimated to the elevated plus maze (EPM) and Y-maze for 3 days. The adaptive training allowed the subjects to explore the maze for 10 min. It is acknowledged that this repeated exposure may compromise the assessment of anxiety-like and spatial working memory behaviors due to the loss of novelty. Thus, EPM and Y-maze results will primarily be interpreted in terms of locomotor activity. Full-scale EPM and Y-maze tests were performed at the end of week 4, 1 day before euthanizing the brain (Fig. 1b).

In rodent models of CNS disorders, anxiety-related behavior has been evaluated using the EPM test [21, 24]. The EPM apparatus comprised a central area, two oppositely positioned open arms, and two oppositely positioned closed arms (online suppl. Fig. S1A; for all online suppl. material, see https://doi.org/10.1159/000550096). To gauge anxiety-related behavior, the preference for open arms as opposed to closed arms for 5 min was calculated and expressed as a percentage of entries or time spent in the open arms [21, 24].

A behavioral test called the Y-maze test was used to gauge how eager rodents were to explore their novel surroundings. Rodents usually explore a new section of the maze rather than return to a previously explored section [25]. The experimental mice in this study spent a total of 10 min in the Y-maze. The mice explored two of the three arms for the first 2 min to get used to the maze, and then they freely explored the entire Y-maze for 8 min (online suppl. Fig. S1A). Each experimental mouse received a score for entering the Y-maze’s three arms, A, B, and C. The score was then computed using the following formula as a percentage of spontaneous alternation: when successive choices of a triplet set (such as C–A–B, B–C–A, and A–B–C) led to admission into the three arms, this was deemed to be spontaneous Y-maze arm alternation.

A rodent’s preference for new things over well-known ones can be evaluated using novel object recognition test (NORT) [26, 27]. Each mouse was given 5 min to explore the open-field arena (40 × 40 × 40 cm [height] gray box) object-free during the adaptation period (online suppl. Fig. S1A). In the first trial, the mouse was given 10 min to investigate the two identical objects (the red circles in online suppl. Fig. S3B) that were positioned in two opposing corners of the test arena. Twenty minutes later, the mice were put back in the arena for the second trial, and a novel object (novel [N]) was given in place of one of the identical objects (familiar [F]) from the first. For 10 min, each mouse’s exploration of each object was captured on video. For every experimental group, item search time and discrimination rate analyses were conducted using the following formula: N h/(N h + F h) × 100 is the total. The experimental animals’ motions in the maze or arena were captured using a digital camcorder, and the analysis was conducted using video tracking software (SMART, Panlab Harvard Apparatus, Holliston, MA, USA) (online suppl. Fig. S1B).

Euthanasia of Experimental Mice, Cardiac Perfusion, Preparation of Brain Tissue for Frozen Section

The mice’s abdomens were excised, and PBS was used to provide cardiac perfusion. A 21-gauge needle was placed into the left ventricle and fastened to the ascending aorta after the pulmonary artery was occluded. The right atrium was cut out with scissors as soon as perfusion began. For fixation and perfusion, respectively, 4% paraformaldehyde and PBS were utilized. The brains were placed in a refrigerator at −80°C and frozen in optimal cutting temperature compound after being successively submerged in 10%, 20%, and 30% sucrose solutions for cryoprotection. A cryostat (Leica, Wetzlar, Germany) was then used to create brain sections that were 25-μm thick. Sections were kept in a refrigerator at −80°C until they were needed after being on glass slides for 12 h.

Brain Tissue Immunofluorescence Staining

After drying and reacting with blocking buffer (5% BSA) for 1 h at 25°C, the brain tissue slices were rinsed, and other materials such as neuronal nuclei (Cat No. 94403, Cell Signaling Technology, Danvers, MA, USA), cluster of differentiation 68 (CD68; Cat No. ab283654, Abcam, Cambridge, UK), glial fibrillary acidic protein (GFAP; Cat No. 12389, Cell Signaling Technology), tumor necrosis factor-α (TNF-α; Cat No. ab183218, Abcam), and myelin-associated glycoprotein (MAG; Cat No. ab277524; Abcam) and myelin basic protein (MBP; Cat No. ab313827; Abcam, Cambridge, UK) antibodies were added. The samples were then allowed to react for 12 h at 4°C. After washing with PBS three times for 5 min to remove excess primary antibody, the tissue sections were incubated with diluted secondary antibodies for 2 h at 25°C (Alexa Fluor 488; Cat No. ab150113 and ab150117; Alexa Fluor 555; Cat. No. ab150114 and ab150118; Abcam) to develop fluorescence. Subsequently, the secondary antibody was rinsed three times in PBS for 5 min each, and the tissue was covered with a cover slide and sealed with nail polish. The glass-fixing solution contained diamidino-2-phenylindole (Cat. No. ab104139, Abcam). Tissue observations were performed using a fluorescence microscope (Ni-U; Nikon, Tokyo, Japan), and ImageJ software (NIH, MD, USA) was used to merge the tissue images. After examination under a fluorescence microscope, the samples were stored for long-term preservation at 4°C in a refrigerator. The positive cell counts of immune intensity per unit area of the brain were used to calculate marker expression. The fluorescence intensity after antibody staining, determined using Digimizer version 4.6.1 (MedCalc Software Ltd., Ostend, Belgium), was used to determine protein expression.

Statistical Analysis

SigmaPlot 15.0 (Systat Software Inc., IL, USA) was used to analyze the experimental results. A one-way ANOVA was used to determine the statistical significance of the variation in the means of each group with equal variance, and the Holm-Sidak method was adopted as a post hoc test. When the equal variance test failed, a one-way ANOVA on rank method was adopted, and Tukey’s post hoc analysis was used. Only cases having a p value of less than 5% were judged statistically significant as a consequence of the analysis. The experimental result’s mean ± standard deviation was used to express all data.

Results

Body Weight Change, Plasma Electrolytes, and Forepaw Grip Strength Change Rate

Physiological changes may occur because of anesthesia and surgery, and these changes may affect research results. Therefore, body weight was measured at 2-week intervals, starting immediately before surgery, and blood was collected when all behavioral measurements were completed. Changes in the plasma electrolytes were also observed. Throughout the study period, the body weights of the experimental mice tended to increase (online suppl. Fig. S2A), and the plasma electrolyte content showed no significant difference among the groups (online suppl. Fig. S2B). Therefore, anesthesia and surgery did not affect the study results. If blood vessel damage occurs during BCAS surgery to induce CCH or if brain damage occurs due to the failure of the micro-coil to properly wrap the CCA, the motor function of the body may be affected. Therefore, changes in the forepaw grip strength were used to assess brain damage. No significant differences in forepaw grip strength were found between the groups (online suppl. Fig. S2C). Therefore, it was confirmed that no factors in the surgical method and process used in this study could have affected the results.

Changes in Behavior and Cognitive Impairment due to BCAS-Induced CCH and the Effect of AE on These Changes

The EPM test is a popular rodent behavioral test that is effective in identifying the brain regions and mechanisms responsible for anxiety-related behavior, as well as in evaluating the antianxiety effects of pharmaceuticals and steroid hormones [21, 24]. In this study, mice were observed in the EPM for 5 min (Fig. 2a) and the distance traveled, and number of entries into the open arm was significantly reduced (Fig. 2b, c) in mice with CCH. In the case of mice that exercised in a water bath, the distance moved significantly increased when the water temperature was above 30°C (Fig. 2b), but movement to the open arm showed no significant change (Fig. 2c). These results show that activities at high water temperatures may have a positive effect on physical ability. However, due to the acclimation protocol, no definitive conclusions regarding anxiety-like behavior or a depressive-like state can be drawn from the open arm entries/time data.

Fig. 2.

Assessment of CCH-induced depression-related behavioral changes using EPM. a Total distance traveled in the EPM. b Total number of open arm entries in 5 min. Due to BCAS-induced CCH, both the total distance (b) traveled in the EPM and the total number of times (c) entering the open arm were significantly reduced. AE in 30–35°C water bath significantly increased the total distance the mice moved per unit of time but did not affect the total number of times they entered the open arm. ^###p < 0.001 vs. sham group; *p < 0.05 vs. BCAS control group; ***p < 0.001 vs. BCAS control group. Data are presented as mean ± SD (n = 7 in each group). CCH, chronic cerebral hypoperfusion; EPM, elevated plus maze; BCAS, bilateral common carotid artery stenosis; AE, aquatic exercise; SD, standard deviation.

Mice were tested for short-term memory using the Y-maze [25]. Allowing mice to explore all three arms of the maze allows for the assessment of spontaneous alternation, a measure of spatial working memory motivated by the natural eagerness of rodents to investigate locations that they have not visited before. An intact working memory mouse, which also has intact prefrontal cortex function, will recall the arms it has previously visited and exhibit a propensity to enter a less-visited arm [25]. In this study, we examined how CCH and AE affected mouse behavior in the Y-maze (online suppl. Fig. S3A). The total distance moved, and the number of times each arm was moved showed no significant difference in all groups (online suppl. Fig. S3B, S3C). Due to the repeated acclimatization protocol, the assessment of spontaneous alternation (a measure of spatial working memory) is compromised. The spontaneous alternation percentage was reduced by CCH, and AE did not improve this measure (online suppl. Fig. S3D).

Nonspatial learning of object identity can be assessed using the NORT [26, 27]. To evaluate the effect of AE on cognitive impairment caused by CCH in mice, an open-field arena was used to analyze the results (Fig. 3a). CCH significantly reduced the distance traveled (Fig. 3b) and ability to identify novel objects (Fig. 3c) in an open-field arena. However, mice that exercised in a water bath containing 30°C water showed significant improvements in moving distance and cognitive impairment, and mice that exercised in a water bath containing 35°C water showed a significant increase only in total moving distance (Fig. 3c). The results that mice undergoing AE in a 35°C water bath had no effect on improving cognitive function but significantly increased total moving distance in the maze or arena (Fig. 2b, 3b) suggest a selective effect on locomotor activity without a corresponding cognitive benefit.

Fig. 3.

Open-field arena and analysis results for NORT evaluation. a The left panel shows the path the mouse moved, and the right panel shows the time the mouse stayed. b Total distance traveled by mice in the open-field arena. c The result of analyzing the recognition ratio of novel objects compared to familiar objects. The total distance traveled in an open field and the ability to recognize novel objects were reduced due to BCAS-induced CCH in BCAS control group mice, and the experimental group that conducted AE in a water bath with a water temperature of 30°C showed an effect of suppressing these pathological changes. In the case of a water temperature of 30°C, the distance traveled in the open field was significantly increased, but there was no improvement in cognitive ability. ^#p < 0.05 vs. sham group; ^###p < 0.001 vs. sham group; *p < 0.05 vs. BCAS control group; ***p < 0.001 vs. BCAS control group. Data are presented as mean ± SD (n = 7 in each group). NORT, novel object recognition test; BCAS, bilateral common carotid artery stenosis; CCH, chronic cerebral hypoperfusion; AE, aquatic exercise; SD, standard deviation.

Activation of Microglia and Astrocytes due to BCAS-Induced CCH and Regulatory Effect of AE on Microglia and Astrocyte Activation

In the CNS, microglia are a specialized population of cells that resemble macrophages. They are immune sentinels with the ability to orchestrate powerful inflammatory responses [28, 29]. CD68 is a lysosomal protein that is expressed at low levels in resting microglia and at high levels in macrophages and activated microglia [28]. Astrocytes are among the many glial cell types found in the CNS and are involved in both health and sickness. However, in neurodegenerative diseases, they appear to acquire toxic capabilities in addition to normal homeostatic tasks [29, 30]. The monomeric intermediate filament protein GFAP is exclusive to the CNS and is present in the astroglial cytoskeleton [31, 32]. In this study, it was confirmed that microglia and astrocytes were activated in the corpus callosum region when CCH was induced by BCAS (Fig. 4b, 5b, c), and when exercising in a 30°C water bath, the activation of astrocytes and the expression of TNF-α protein were reduced (Fig. 5b, c). A water bath at 25°C did not suppress the activation of these inflammatory cells in the brain at all, and at AE in a 35°C water bath, it did not suppress the activation of inflammatory cells, but it appeared to reduce the expression of TNF-α protein (Fig. 5c), not suppressing the activation of inflammatory cells. Thus, it is possible to reduce the inflammatory response via other mechanisms.

Fig. 4.

Changes in CD68-positive microglial expression due to BCAS-induced CCH and the effect of AE. Different cell types can be distinguished through immunofluorescent staining (a), and microglial cells with CD68 as a marker were activated by BCAS-induced CCH (b). No significant effect was seen in the experimental group that conducted AE in the water bath. DAPI, which can stain the nucleus, was expressed in blue; NeuN, an antibody which can stain neurons, was confirmed in green; and CD68, a marker for microglial cells, was expressed in red. CD68 was expressed in a relatively wide area in the BCAS control group. Since the equal variance test failed, ANOVA rank sum analysis was performed. ^###p < 0.001 vs. sham group. Data are presented as mean ± SD (n = 7 in each group). CD68, cluster of differentiation 68; BCAS, bilateral common carotid artery stenosis; CCH, chronic cerebral hypoperfusion; AE, aquatic exercise; DAPI, 4′,6-diamidino-2-phenylindole; NeuN, neuronal nuclei; ANOVA, analysis of variance; SD, standard deviation.

Fig. 5.

Changes in GFAP-positive astrocyte expression induced by CCH and effects of AE. Different cell types can be distinguished through immunofluorescent staining (a), and astrocytes with GFAP as a marker were activated by BCAS-induced CCH (b). The expression of both GFAP and TNF-α was suppressed in mice that exercised in a heated water bath at 30 and 35°C (b, c). DAPI, which can stain the nucleus, is shown in blue; GFAP, a marker of astrocytes, is shown in red; and the expression of TNF-α is shown in green. In the BCAS control group, the expression of GFAP was evident around the corpus callosum, a consisting region of WM. ^###p < 0.001 vs. sham group; *p < 0.05 vs. BCAS control group; ***p < 0.001 vs. BCAS control group. Data are presented as mean ± SD (n = 7 in each group). GFAP, glial fibrillary acidic protein; CCH, chronic cerebral hypoperfusion; AE, aquatic exercise; TNF-α, tumor necrosis factor-α; WM, white matter; DAPI, 4′,6-diamidino-2-phenylindole; NeuN, neuronal nuclei; SD, standard deviation.

Demyelination Caused by CCH and the Inhibitory Effect of AE on It

The fatty layer that surrounds nerve axons and promotes electrical impulse conduction is called myelin and is a crucial component of the white matter that keeps nerves healthy and is able to transmit and receive signals throughout the body [33, 34]. Recently, Bouhrara et al. [35] demonstrated that demyelination is significantly increased in patients with mild cognitive impairment and dementia. It has been found that among people who are cognitively intact, a lower myelin content is linked to a faster rate of cognitive deterioration [36]. As the expression of MAG and MBP in the corpus callosum and MBP in the dorsal striatum of mice was decreased by BCAS (Fig. 6a–e), it was confirmed that the cognitive impairment caused by CCH was caused by demyelination. When experimental mice were subjected to AE, only MBP expression in the corpus callosum was suppressed in mice that exercised in a 30°C water bath (Fig. 6c).

Fig. 6.

Change in myelin-related protein (MAG and MBP) expression. Due to BCAS-induced CCH, the expression of myelin-related proteins was decreased not only in the WM (a–c) but also in the dorsal striatum (d, e), and the decrease in MBP expression was suppressed in the AE conducted group in a 30°C water bath (c). AE did not suppress the decrease in MBP expression in the dorsal striatum (e). The level of expression can be confirmed in blue for DAPI, which can stain the nucleus, in red for MAG, and in green for MBP. ^##p < 0.01 vs. sham group; ^###p < 0.001 vs. sham group; **p < 0.01 vs. BCAS control group. Data are presented as mean ± SD (n = 7 in each group). MAG, myelin-associated glycoprotein; MBP, myelin basic protein; BCAS, bilateral common carotid artery stenosis; CCH, chronic cerebral hypoperfusion; AE, aquatic exercise; WM, white matter; DAPI, 4′,6-diamidino-2-phenylindole; SD, standard deviation.

Discussion

It was confirmed that cognitive function was reduced by wrapping micro-coils around both CCA of mice to induce CCH, and it was found that AE of more than 30°C suppressed this cognitive impairment. This inhibitory effect on cognitive impairment was confirmed to occur through the suppression of the activation of astrocytes, one of the cells involved in inflammatory responses in the brain. In this study, the most pronounced activity was shown at a water temperature of 30°C. However, it was confirmed that even at a water temperature of 35°C, it positively improved motor skills, such as increasing the distance traveled in the maze.

In aging populations, cerebrovascular disease and cognitive impairment are highly prevalent, with risk factor reduction potentially preventing up to 25% of dementia cases linked to cardiovascular factors [37]. VaD is a key pathological outcome associated with this cognitive decline [38, 39].

In this study, the BCAS mouse model was utilized to induce CCH and mimic VaD pathology. BCAS-induced CCH resulted in ischemic white matter lesions, chronic hypoxic supply, hippocampal atrophy, and inflammatory responses such as gliosis and blood-brain barrier disruption [40, 41]. As the BCAS model consistently induces these characteristic pathological changes, it is considered the most suitable model for VaD research in mice [42–45].

Consistent with established findings in VaD patients, our BCAS model exhibited impaired cognitive function (Fig. 2c, 3c), reactive gliosis (Fig. 4b, 5b), and decreased expression of MBP and MAG due to demyelination in the corpus callosum (Fig. 6b, c). Interestingly, MBP expression also decreased in the dorsal striatum region (Fig. 6e), a finding previously observed in similar models [46, 47].

Crucially, we must acknowledge the limitations of our behavioral assessment protocol. Due to the repeated acclimatization of the mice to the EPM and Y-maze, the results from these tests cannot be interpreted as reflecting anxiety-like behavior or spatial working memory. The behavioral findings are thus limited to the significant decline in NORT and the changes in total locomotor activity (distance moved, Fig. 2b, 3b). While some studies link reduced handgrip strength to VaD [48, 49], we observed no differences in forepaw grip strength among any of the groups (online suppl. Fig. S2C). This stability in gross motor function confirms that the observed cognitive and pathological changes were specific to CCH and not due to acute physical incapacitation.

It is well established that physical exercise improves cognitive function [50], and AE offers advantages over land-based exercise by reducing load on damaged tissues, making voluntary exercise easier [18–20, 22, 23]. Regular swimming pool temperatures range from 26 to 29°C [51, 52]. Our findings, however, demonstrate that the water temperature plays a major role in the effectiveness of underwater exercise. The optimal therapeutic effect, characterized by suppressed cognitive impairment and significantly reduced reactive astrocytes (Fig. 5b), was observed when AE was performed at 30°C.

This finding is biologically significant due to the thermal neutral point of mice. For C57BL/6 mice, the thermal neutral point – the ambient temperature range where metabolic rate is minimized – is approximately 29–33°C during the resting phase when our experiments were conducted. The 30°C bath falls squarely within this range, suggesting that minimizing thermal stress is crucial for AE’s therapeutic effect in VaD. In contrast, 25°C requires energy expenditure for thermogenesis, and 35°C may induce hyperthermic stress [53]. This concept holds clinical relevance, as human aquatic therapy guidelines also consider temperature for optimal effect, suggesting that temperatures slightly higher than a regular pool may be more advantageous for neurological rehabilitation.

Clear guidelines on exercise methods or the amount of exercise for dementia patients are lacking. However, studies show that while activity reduces the risk of dementia [54, 55], the amount of exercise and cognitive improvement are not necessarily proportional [56]. Therefore, performing moderate-intensity exercise that is practically feasible may be more appropriate for dementia patients rather than excessive high-intensity exercise. This is consistent with our use of voluntary AE, which is easily sustained by the mice.

While physical exercise has been reported to suppress VaD-induced cognitive impairment in rats [57], research on the specific mechanism of action for AE, particularly the influence of water temperature, is novel. This study is believed to be the first to examine the presence or absence of effects due to differences in water temperature. Our results confirm that performing AE at 30°C effectively controls the activation of immune response cells, thereby minimizing the inflammatory response and suppressing cognitive impairment.

A key limitation of this study is the lack of individual quantification of the AE load, as we allowed free movement in the water bath rather than forced swimming. Furthermore, this initial study’s scope was limited to assessing microglia and astrocyte markers. Therefore, future follow-up studies will investigate the effects of increasing AE load and will confirm the specific underlying mechanism of action by analyzing markers for apoptosis, neurogenesis, and hypoxia-induced neovascularization, along with the types and expression levels of relevant proteins. Lastly, using only male mice in the present study is a limitation [58, 59]; we plan to confirm these results using both males and females in a follow-up study.

In summary, CCH was induced by BCAS in mice. When measured 4 weeks later, they showed impaired novel object recognition and significant pathological changes in the corpus callosum region, including increased reactive cells and reduced myelin proteins. The AE conducted at the optimal temperature of 30°C significantly reduced the expression of reactive astrocytes and suppressed cognitive impairment. This study confirmed that the temperature suitable for performing AE in patients with VaD should be set higher than that of a regular swimming pool, aligning with the concept of the thermal neutral zone.

Acknowledgments

We would like to thank Yunyoung Cho and Jeongha Lee, students at Brookline High School in MA, USA, for their assistance in analyzing the research results under blinded conditions.

Statement of Ethics

All procedures, including surgery, were approved by the Laboratory Animal Ethics Regulations of the Pusan University Animal Experimental Ethics Committee (Approval Nos. PNU-2023-03673 and PNU-2024-0462), conducted in compliance with relevant provisions, and followed in the handling and utilization of laboratory animals.

Conflict of Interest Statement

The authors declare that they have no conflicts of interest in relation to this study.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

Chiyeon Lim: conceptualization, methodology, formal analysis, data curation, and writing – original draft, review, and editing. Sehyun Lim: conceptualization, formal analysis, data curation, validation, visualization, and writing – original draft. Sung Min Moon: validation, visualization, and writing – original draft. Suin Cho: conceptualization, methodology, investigation, formal analysis, and writing – original draft, review, and editing.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.