Associations of Physical Activity with Alzheimer’s Disease Pathologies and Cognition: The CABLE Study

Abstract

Background:

The associations of physical activity with Alzheimer’s disease (AD) pathologies remain controversial.

Objective:

To quantitatively assess the association between the frequency of physical activity with cerebrospinal fluid (CSF) biomarkers in AD and further explore the mechanism by which AD pathologies regulate the correlation between physical activity and cognition.

Methods:

A total of 918 participants without dementia from Chinese Alzheimer’s Biomarker and Lifestyle (CABLE) were examined in this population-based cross-sectional study. Multiple linear models were used to evaluate the associations of physical activity with CSF biomarkers and cognition. Moreover, mediation analyses were conducted to investigate the potential relationships between physical activity, AD pathologies, and cognitive function.

Results:

Regular physical activity was positively associated with CSF Aβ₄₂ (p < 0.001) and Aβ_42/40 (p < 0.001), while it was negatively associated with p-tau/Aβ₄₂ (p < 0.001) and t-tau/Aβ₄₂ (p < 0.001). Of all participants, regular physical activity was associated with increased cognitive function (p < 0.001). The interaction effect indicated that age moderated the association between physical activity frequency and CSF Aβ₄₂ (p = 0.014) and p-tau/Aβ₄₂ (p = 0.041). The impact of physical activity on cognition was mediated in part by amyloid pathologies, accounting for 4.87% to 21.56% of the total effect (p < 0.05).

Conclusion:

This study showed the beneficial impact of physical activity on AD pathologies and cognition in participants without dementia.

Keywords

Alzheimer’s disease biomarkers cerebrospinal fluid cognition physical activity

INTRODUCTION

Alzheimer’s disease (AD) is the most common clinical form of dementia, and it poses great challenges to public health and society. Today more than 47.5 million people are suffering from dementia and the number may reach 132 million by 2050, according to the World Alzheimer Report 2015 [1, 2]. AD pathogenesis is related to environmental exposure and genetic susceptibility. Substantial epidemiological evidence suggests that self-management of controllable risk factors can effectively improve cognition and reduce the risk of AD [3]. Barnes et al. noted that if seven controllable risk factors including physical inactivity, depression, education, smoking, obesity, midlife hypertension, and diabetes mellitus were reduced by 10–25%, as many as 1.1 million to 3.0 million cases of AD could be prevented worldwide [4].

Researchers found that physical activity reduces the incidence of AD and is often accompanied by improvements in cognitive function. The following summarizes current research on the underlying biological mechanisms. Firstly, many cardiovascular risk factors are associated with the risk of cognitive decline [5]. Physical activity can decrease or retard neurodegeneration by reducing the risk of cardiovascular disease, stroke, type 2 diabetes, and other conditions such as cancer [6, 7]. Secondly, physical activity promotes the secretion of brain-derived neurotrophic factor (BDNF), which further increases the length and number of dendritic connections, as well as the survival of hippocampal cells and new brain capillaries [8, 9]. Moreover, long-term physical activity increases size of hippocampal volume while improving cognitive function [10 –12]. Thirdly, oxidative stress is a detrimental process shared by brain aging and AD. Physical activity can exert anti-inflammatory effects and improve redox status of the brain, thereby improving the pathophysiological features of AD and improving cognitive function [13, 14]. Finally, based on the gene-environment interaction theory, physical activity may ameliorate the effect of risk genes on AD dementia, like apolipoprotein ɛ4 (APOE ɛ4) [15] and BDNF [16] gene.

The pathological features of AD are nerve fiber tangles, neuroinflammation, and amyloid-β (Aβ) plaques [17]. In AD patients, decreased concentrations of the 42-aminoacid form of Aβ (Aβ₄₂) imply cortical amyloid deposition, elevated total tau (t-tau) concentrations due to cortical neuronal loss, and phosphorylated tau (p-tau) reflects the formation of high concentrations of cortical tangles [18, 19]. Applying CSF biomarkers and their ratios improves the accuracy of AD diagnosis and treatment [20]. However, findings regarding the effect of physical activity on AD pathologies are mainly from animal studies. In transgenic mouse models of AD, cognitive performance was improved in the long-term exercise group, whereas amyloid deposition was decreased [21, 22] or constant [23] or increased [24]. The differences in conclusions may be ascribed to different genetic mice models, as well as different types and durations of physical activity.

To date, the associations of physical activity with AD pathologies and cognition has remained largely unknown. Therefore, this study aimed to quantitatively assess the effect of physical activity on early pathological changes of CSF biomarkers in non-demented participants and its underlying mechanisms, providing insights for the prevention and early intervention of AD.

METHODS

CABLE database

The Chinese Alzheimer’s Biomarker and Lifestyle (CABLE) study is a large-scale cohort study in the Chinese Han population, focusing on CSF biomarkers and AD-risk factors. CABLE aims to explore the genetic and environmental factors that influence AD pathogenesis and provide specific guidelines for AD prevention and therapy. The protocol for the CABLE study was granted by the Institutional Ethics Committee of Qingdao Municipal Hospital, and the study followed the Helsinki declaration. The written consent form was obtained from all subjects or authorized persons.

Participants

All CABLE participants were Han Chinese between the ages of 40 and 90. The exclusion criteria are as follows: 1) central nervous system infection, multiple sclerosis, head trauma, neurodegenerative diseases other than AD (such as epilepsy, Parkinson’s disease), or other major neurological disorders; 2) severe systemic disease (such as malignancy) that may influence AD biomarker level in cerebrospinal fluid; 3) major psychological diseases; or 4) genetic or family history [25].

Finally, 918 northern Han participants without dementia from the CABLE study were included. Each participant received clinical and cognitive assessments, biochemical tests (e.g., blood lipids and homocysteine etc.) and collection of blood and CSF samples. Demographic information (e.g., age, sex, educational background, social relationship, etc.), the profile of AD risk factors (e.g., cigarette consumption, alcohol habit, overweight, etc.) and medical history (e.g., stroke, hypertension, coronary heart disease, etc.) were obtained from a self-reported questionnaire and an electronic patient record system. Professionally trained neurologists assessed participants’ overall cognitive function through the Mini-Mental State Examination (MMSE). The diagnoses of each participant were strictly aligned with the National Institute on Aging-Alzheimer’s Association (NIA-AA) workgroup diagnostic criteria [26].

The assessment of physical activity

For each participant, the data of physical activity was measured by a professional clinician with a clinical questionnaire. Professional clinician divided the participants into five groups with a blinded method according to their frequency of physical activity over the past year. The details are as follows: 0 = I never participate in physical activity (n = 205, 22.3%); 1 = I occasionally participate in physical activity (n = 112, 12.2%); 2 = I participate in physical activity once a week (n = 45, 4.9%); 3 = I participate in physical activity multiple times a week (n = 229, 24.9 %); 4 = I participate in physical activity daily (n = 327, 35.6%). To ensure statistical robustness of the sample analysis and to increase the sample size of the different frequency groups, we combined categories 0, 1 and 2 into infrequent group (N = 362) and categories 3 and 4 into frequent group (N = 556). These categories were determined for similar-sized groups where data distribution allowed [27].

AD biomarkers

Standard lumbar puncture was performed to collect CSF samples in a fasting state, and then the samples were processed within 2 h. We removed cells and other indissoluble contaminants by centrifugation at 2000×g for 10 min, then the samples were quick-frozen at –80°C for use. No sample was thawed/frozen more than twice. CSF Aβ₄₂, Aβ₄₀, t-tau, and p-tau were measured by the enzyme-linked immunosorbent assay (ELISA) kit (Innotestβ-AMYLOID (1–42), β-AMYLOID (1–40), PHOS-PHO-TAU (181p), and hTAU-Ag; Fujirebio, Ghent, Belgium) on the microplate reader (Thermo ScientificTM Multiskan™ MK3). All the professionals responsible for the operation were blinded to all information about the study participants. The within-batch coefficient of variation (CV) was controlled below 5%. The inter-batch CV was controlled below 20%.

APOE genotyping

DNA was extracted using the QIAamp®DNA blood mini-kit from nocturnal fasting blood samples and stored in an enzyme-free EP tube at –80°C until the APOE ɛ4 genotyping. Rs7412 and rs429358 which define APOE ɛ2/ɛ3/ɛ4 isoforms were genotyped with the restriction fragment length polymorphism (RFLP) technology.

Statistical analysis

Among demographic characteristics, categorical variables were depicted as numbers (column percentages), and continuous variables were depicted as mean±SD (standard deviation). The inter-group differences were calculated by using Chi-square test for categorical variables, and t-test (normal distribution) or the Mann-Whitney U test (skewed distribution) for continuous variables. We normalized the dependent variable values of the linear regression model using “car” package of R software and standardized by z-scale. Participants with missing basic information in the CABLE database were removed to ensure accuracy.

First of all, we used multiple linear regression models to explore the relationship between physical activity and CSF biomarkers. In this model physical activity was used as the independent variable and CSF biomarkers were used as the dependent variable. Models were adjusted for study covariates (Model 1: adjusted for gender, age, education, and APOE ɛ4 status; Model 2: adjusted for variables in model 1 and body mass index, cigarette consumption, alcohol habit; Model 3: adjusted for variables in model 2 and take other medical factors into account such as diabetes mellitus, stroke, hypertension, coronary heart disease. After that, we applied interaction analysis to determine the influence of major AD risk factors (gender, age, education, and APOE ɛ4 status) on these associations, then performed subgroup analysis. We did find a moderating effect based on the interaction results. It will explore whether specific demographic characteristics influence potential associations to determine who will benefit the most.

Next, we performed mediation analyses to estimate whether the relationship between physical activity and cognition was mediated by AD pathologies. The establishment criteria of the mediation effect are as follows: 1) physical activity was significantly correlated with CSF biomarkers; 2) physical activity was significantly correlated with cognition; 3) CSF biomarkers were significantly correlated with cognition; and 4) the associations between physical activity and cognition were altered when CSF biomarkers (the mediator) were added to the regression model. All mediational tests performed 10,000 bootstrap replications. Each pathway in the model was corrected for gender, age, education, and APOE ɛ4 status. p < 0.05 (two-tailed) was considered statistically significant. R (version 3.6.2) and SPSS (version 25.0, IBM, Armonk, NY, USA) software programs were used in all statistical analyses.

RESULTS

Basic characteristics in different subgroups

Table 1 and Supplementary Table 1 showed the demographic characteristics of the study population. A total of 918 (mean MMSE score = 27.70) non-demented participants were included. The mean age of the participants was 61.27 years (SD = 10.89). The mean year of education was 10.42 years (SD = 4.09). And 329 participants were female, and 148 participants were APOE ɛ4 allele carriers.

Table 1

Characteristics of included participants in the CABLE database

Characteristics	All participants	Physical activity		p
		Frequent	Infrequent
Number	918	556	362	–
Age, mean (SD), y	61.27 (10.89)	61.48 (10.90)	60.96 (10.87)	0.479
Sex (Female/Male)	329/589	174/382	155/207	<0.001
Education, mean (SD), y	10.42 (4.09)	10.72 (3.99)	9.96 (4.19)	0.007
APOE ɛ4, yes (%)	148 (16.12)	93 (16.73)	55 (15.19)	0.498
Cigarette consumption, yes (%)	285 (31.04)	185 (20.15)	100 (10.89)	0.083
Alcohol habit, yes (%)	293 (31.90)	193 (21.02)	100 (10.89)	0.029
Hypertension, yes (%)	340 (37.03)	202 (22.00)	138 (15.03)	0.631
Diabetes, yes (%)	138 (15.03)	84 (9.15)	54 (5.88)	0.999
Stroke, yes (%)	33 (3.59)	23 (2.50)	10 (1.08)	0.362
CHD, yes (%)	120 (13.07)	76 (8.27)	44 (4.29)	0.572
BMI, mean (SD), kg/m²	25.47 (3.53)	25.31 (3.57)	25.72 (3.45)	0.087
MMSE score, mean (SD)	27.70 (2.63)	27.99 (2.42)	27.27 (2.88)	<0.001
CSF biomarkers, mean (SD), pg/ml
Aβ₄₂	262.21 (155.27)	287.11 (158.47)	224.13 (142.27)	<0.001
Aβ₄₀	6492.22 (3529.45)	6371.63 (3665.83)	6680.67 (3301.49)	0.195
t-tau	196.57 (96.09)	198.59 (102.34)	193.45 (85.58)	0.424
p-tau	43.33 (15.06)	43.15 (15.39)	43.59 (14.56)	0.674
Ratios of biomarkers, mean (SD)
Aβ_42/40	0.05 (0.04)	0.05 (0.03)	0.05 (0.07)	0.058
t-tau/Aβ₄₂	1.06 (0.92)	0.93 (0.84)	1.25 (1.00)	<0.001
p-tau/Aβ₄₂	0.24 (0.19)	0.21 (0.16)	0.29 (0.21)	<0.001

Bold indicated that the results were statistically significant. p values for comparisons between frequent and infrequent groups. APOE ɛ4, apolipoprotein E ɛ4; CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau; SD, standardized deviation; BMI, body mass index; CHD, coronary heart disease; MMSE, Mini-Mental State Examination.

As shown in Table 1, gender (p < 0.001), education (p = 0.007) and MMSE score (p < 0.001) were statistically significant differences between frequent and infrequent groups. In terms of cognitive level, the frequent group had greater MMSE scores. Compared with the infrequent group, levels of CSF Aβ₄₂ (p < 0.001) increased significantly in the frequent group, while levels of CSF t-tau/Aβ₄₂ (p < 0.001), and CSF p-tau/Aβ₄₂ (p < 0.001) decreased significantly. There were no differences in levels of CSF Aβ_42/40, CSF t-tau, CSF p-tau, and CSF Aβ₄₀ between frequent and infrequent groups (p > 0.05). We then divided the participants into age groups to compare differences in CSF AD biomarker levels. As expected, CSF biomarker (Aβ₄₀, t-tau, p-tau, t-tau/Aβ₄₂, p-tau/Aβ₄₂) levels were higher in older (>65 years) subgroup than in younger (≤65 years) subgroup (all p < 0.001). Although CSF Aβ₄₂ and Aβ_42/40 levels were higher in older (>65 years) subgroup than in younger (≤65 years) subgroup, we did not find differences in CSF Aβ₄₂ (p = 0.457) and Aβ_42/40 (p = 0.196) between the age subgroups (Supplementary Table 1).

Associations of physical activity with cognition and CSF AD biomarkers

Figure 1 showed the results of multiple linear regression. In Model 1, we find that regular physical activity was correlated with improved cognitive function (β= 0.079, p < 0.001). The frequency of physical activity was positively correlated with CSF Aβ₄₂ (β= 0.161, p < 0.001) and Aβ_42/40 (β= 0.156, p < 0.001), but negatively correlated with CSF t-tau/Aβ₄₂ (β= –0.142, p < 0.001) and p-tau/Aβ₄₂ (β= –0.156, p < 0.001). And frequency of physical activity was not correlated with CSF Aβ₄₀ (β= –0.040, p = 0.061), t-tau (β= 4.884e-05, p = 0.998), or p-tau (β= –0.014, p = 0.488) in all participants. All these results for Model 2 and Model 3 were barely unchanged.

Fig. 1

Associations between physical activity with global cognition and CSF biomarkers. A-C) The multiple linear regression analysis results of models 1–Model 3, respectively. Model 1: adjusted for gender, age, education, and APOE ɛ4 status; Model 2: adjusted for variables in model 1 and body mass index, cigarette consumption, alcohol habit; Model 3: adjusted for variables in model 2 and medical factors. MMSE, Mini-Mental State Examination; CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau.

Influence of age on associations of physical activity with cognition and CSF biomarkers

As shown in Table 2, the interaction analyses suggested that age moderated the association between physical activity frequency and CSF Aβ₄₂ (p = 0.014) and p-tau/Aβ₄₂ (p = 0.041). However, there were no interaction effects with gender, education, or APOE ɛ4 status. For subgroup analyses stratified by age, associations of physical activity with CSF Aβ₄₂-related biomarkers (Aβ₄₂, Aβ_42/40, t-tau/Aβ₄₂, p-tau/Aβ₄₂) and cognition still remained in two subgroups (Supplementary Table 2). Among all participants, CSF Aβ₄₂ (Fig. 2A: p < 0.001) and Aβ_42/40 (Fig. 2D: p < 0.05) levels were higher in the frequent group than in the infrequent group, while lower levels of t-tau/Aβ₄₂ (Fig. 2E: p < 0.05) and p-tau/Aβ₄₂ (Fig. 2F: p < 0.05). Levels of CSF Aβ₄₂ and p-tau/Aβ₄₂ changed more rapidly with increasing frequency of physical activity in older compared to younger (Supplementary Figure 1).

Table 2

Interaction analyses of associations between physical activity frequency and CSF biomarkers

Biomarkers	Interactions (p)
	Age	Gender	Education	APOE ɛ4
CSF Aβ₄₂	0.014	0.786	0.676	0.248
CSF Aβ₄₀	0.653	0.571	0.755	0.421
CSF t-tau	0.236	0.914	0.863	0.441
CSF p-tau	0.809	0.843	0.647	0.115
CSF Aβ_42/40	0.290	0.400	0.551	0.630
CSF t-tau/Aβ₄₂	0.167	0.996	0.432	0.928
CSF p-tau/Aβ₄₂	0.041	0.983	0.566	0.565

CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau.

Fig.2

Group differences in levels of CSF biomarkers between age subgroups. p values were obtained from Mann-Whitney U test or t test. CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau.

Mediation analysis

Higher frequency of physical activity was correlated with improved cognitive function as indicated by changes in MMSE scores (β= 0.079, p < 0.001). Finally, we performed the mediation analysis to further explore whether the associations between physical activity and MMSE score was mediated through CSF biomarkers. In total participants, mediation analysis showed that CSF Aβ₄₂-related biomarkers (Aβ₄₂, Aβ_42/40, t-tau/Aβ₄₂, p-tau/Aβ₄₂) significantly and partially mediated the associations. Mediation proportions were 4.87% (Aβ_42/40: p = 0.030), 17.18% (p-tau/Aβ₄₂: p = 0.012), 17.83% (t-tau/Aβ₄₂: p = 0.004) to 21.56% (Aβ₄₂: p < 0.001), respectively (Fig. 3). More precisely, this partial mediation effect was only significant in individuals≤65 years old (proportion: 19.60% to 22.77%) (Supplementary Figure 2).

Fig. 3

Mediation analyses with MMSE as cognitive outcome. A–D) Mediation analyses with 10,000 bootstrapped iterations were used to examine mediation effects of CSF Aβ-related biomarkers (Aβ₄₂, Aβ_42/40, t-tau/Aβ₄₂, p-tau/Aβ₄₂) on cognition. Each path of the model corrected for gender, age, education, and APOE ɛ4 status. MMSE: Mini-Mental State Examination; CSF, cerebrospinal fluid; Aβ, amyloid-β; t-tau, total tau; p-tau, phosphorylated tau. IE, indirect effect.

DISCUSSION

This is a population-based cross-sectional study revealed the beneficial effects of physical activity on AD pathologies and cognition. To be specific, frequent physical activity was correlated with low levels of CSF p-tau/Aβ₄₂ and CSF t-tau/Aβ₄₂, while high levels of CSF Aβ₄₂ and CSF Aβ_42/40 (all p < 0.001). Notably, these associations were significantly age-type dependent. Additionally, the protective effect of physical activity on cognitive decline was partially mediated by CSF Aβ₄₂-related biomarkers.

Biomarkers (Aβ₄₂, t-tau, p-tau) have high diagnostic accuracy for AD, with sensitivity and specificity as high as 85–90%. These biomarkers have now been incorporated into modern diagnostic research criteria and are applicable to patients with mild cognitive impairment in AD [28]. The ratios of CSF biomarkers also reflect a range of AD-related conditions during the preclinical period, suggesting an increased risk of impending disease [29]. Lifestyle and vascular risk factors (smoking, overweight, hypertension, diabetes, etc.) were discovered to contribute to the development of dementia and AD [4 , 31]. In our analysis, the associations between physical activity and AD remained stable and significant even after adjusting for various lifestyle and vascular risk factors, suggesting an independent role for regular physical activity. Our study demonstrated that physical activity could effectively regulate the level of CSF Aβ. One explanation is that physical activity promotes anatomical, neurochemical, and electrophysiological changes in neurons by modulating multiple gene products at the mRNA and protein levels, ultimately activating multiple pathways to directly or indirectly regulate amyloid levels. For example, physical activity affects the degradation process of amyloid-β protein precursor (AβPP) proteolytic fragments through induced upregulation of proteasome activity. Because Aβ is produced by the proteolytic process of AβPP, researchers argue that physical activity may be acting downstream of AβPP [21 , 33]. Nichol et al. found that regular physical activity triggered an immune response associated with decreased neurotoxic cytokines and increased Aβ clearance in the brains of AD transgenic mouse models [34]. Taken together, these findings illustrate that there are many potential mechanisms which may increase our understanding of how physical activity affects AD pathology.

In our study, the benefits of physical activity were greater in frequent group (≥2 times/week) than in the infrequent group (<2 times/week). Barnard et al. showed that three 40-min brisk walks per week as regular activity could improve cognitive function [35]. Lucia et al. recommended that older adults (≥65 years) do moderate-intensity aerobic exercise≥30 min 5 days per week or vigorous-intensity aerobic exercise≥20 min 3 days per week to prevent risks such as cognitive deterioration [36]. Although accumulating evidence supports the protective effects of physical activity on AD, recommendations for physical activity are different as to duration, frequency, type, and intensity. Due to the limitations in the type of physical activity and the race of the study population, we cannot figure out which combination (type, frequency, duration, and intensity) is the best. Future research could further quantify physical activity to maximize the protection of physical activity against neurodegeneration.

Demographic characteristics (age, genotype, etc.) will also be an important consideration when examining the role of physical activity in regulating CSF biomarkers of AD levels. Our study shows that the effect of physical activity in improving AD pathological deposition is better in older subgroup and in improving cognitive ability in the younger subgroup. The likely cause is the decreased ability of the nervous system in older adults to change its tissues to adapt to changing needs and circumstances, which is associated with poorer physical function in older adults. Of the many risk factors for neurodegeneration, the aging process itself has the greatest impact [37]. Aging is associated with worsening physical condition, leading to an increased chance of developing other dangerous diseases such as stroke and diabetes, which in turn can increase the risk of AD. Our study suggests that advanced age may have different degrees of “protective” properties in AD pathologies. The explanation for the “protective” properties of advanced age may be that elderly people are susceptible to AD and are therefore more dependent on lifestyle-related factors to protect them from dementia and AD. Interestingly, Luck et al. found that APOE ɛ4 carriers may particularly benefit from a physically active lifestyle [10]. Rovio et al. demonstrated that midlife physical activity was protective against the risk of dementia and AD only in APOE ɛ4 carriers [32]. These findings suggest that populations with genetic susceptibility have a reduced risk of developing AD if they adopt an active lifestyle, which provided an optimistic outlook for AD prevention and treatment.

Studies have shown that physical activity can promote the release of hormones beneficial to neural development and improve neurocognition through increased metabolism, oxygenation, and blood flow [38, 39]. Moreover, physical activity was related to higher pathological levels of Aβ 42. Mediation effects demonstrated that exercise may improve cognition by improving AD pathological deposition, which is an important area for clinical research. Physical activity-induced cognitive improvement was not mediated by AD pathology in people over 65 years, supporting the hypothesis that physical activity improves brain health during aging through multiple pathways (increased neurotrophic levels, improved blood vessel formation, promote synapse formation, modulate inflammation, reduce disordered protein deposition) [14]. Thus, it can be speculated that physical activity affects cognitive function at various stages of life, where the mechanisms change during each period.

Several limitations of this study are as follows. First, the assessment of physical activity relied on subjective assessments through questionnaires or self-reports. Subjective measures may be influenced by social expectation bias and recall bias, and regular records of daily physical activity can therefore be subjective and sketchy. In addition, the number of cases, quality of physical activity, and duration of physical activity can also cause heterogeneity. Our study lacked a quantitative analysis of duration and type of physical activity. Second, conclusions were derived from cross-sectional studies that were unable to demonstrate causality. Third, we didn’t look at diversity among different ethnic groups because the study’s population was limited to the Han nationality. Fourth, physical activity plays a role in modulating AD pathological features, Aβ and tau levels (in the brain, CSF, and blood). Unfortunately, we did not measure the concentration of biomarkers in blood samples, so no information is available in the database. Finally, several longitudinal studies have demonstrated a positive relationship between exercise and cognitive health [40, 41]. Future studies need to combine CSF and neuroimaging indicators to detect the longitudinal effects of physical activity on AD neuropathology.

In conclusion, the CABLE study suggests that physical activity have positively effect on AD by modulating the effect of preclinical amyloid deposition on AD among non-demented participants. Individuals over 65 years old can benefit more from this lifestyle. The frequency group (≥2 times/week) improved the deposition of CSF Aβ-related biomarkers better than the infrequency group (<2 times/week). Finally, we further confirmed that the association between physical activity and MMSE cognition is partially mediated by CSF Aβ-related biomarkers. So far, the use of pharmacological interventions to treat AD has had limited success [42]. This study provides new clues for non-drug prophylaxis and clinical trials to slow or prevent the disease progression of AD.

ACKNOWLEDGMENTS

The authors thank all participants of the present study as well as all members of staff of the CABLE study for their role in data collection.

This study was supported by grants from the National Natural Science Foundation of China (82071201, 81971032, 82001133), Shanghai Municipal Science and Technology Major Project (No.2018SHZDZX01), Research Start-up Fund of Huashan Hospital (2022QD002), Excellence 2025 Talent Cultivation Program at Fudan University (3030277001), and ZHANGJIANG LAB, Tianqiao and Chrissy Chen Institute, and the State Key Laboratory of Neurobiology and Frontiers Center for Brain Science of Ministry of Education, Fudan University.

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/22-0389r1).

Footnotes

The supplementary material is available in the electronic version of this article: .

References

Wimo

, Guerchet

, Ali

, Wu

, Prina

, Winblad

, Jönsson

, Liu

, Prince

(2017) The worldwide costs of dementia 2015 and comparisons with 2010. Alzheimers Dement 13, 1–7.

Wimo

, Jönsson

, Bond

, Prince

, Winblad

(2013) The worldwide economic impact of dementia 2010. Alzheimers Dement 9, 1–11.E13.

, Wang

, Wan

, Tan

, Yu

, Tan

(2017) Leisure time physical activity and dementia risk: A dose-response meta-analysis of prospective studies. BMJ Open 7, E014706.

Barnes

, Yaffe

(2011) The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol 10, 819–828.

Baumgart

, Snyder

, Carrillo

, Fazio

, Kim

, Johns

(2015) Summary of the evidence on modifiable risk factors for cognitive decline and dementia: A population-based perspective. Alzheimers Dement 11, 718–726.

Shi

, Li

, Wang

, Zhou

, Qin

, Yin

, Wei

, Liu

, Nie

(2015) Household physical activity and cancer risk: A systematic review and dose-response meta-analysis of epidemiological studies. Sci Rep 5, 14901.

Stephen

, Hongisto

, Solomon

, Lönnroos

(2017) Physical activity and Alzheimer’s disease: A systematic review. J Gerontol A Biol Sci Med Sci 72, 733–739.

Colcombe

, Erickson

, Scalf

, Kim

, Prakash

, Mcauley

, Elavsky

, Marquez

, Hu

, Kramer

(2006) Aerobic exercise training increases brain volume in aging humans. J Gerontol A Biol Sci Med Sci 61, 1166–1170.

Cotman

, Engesser-Cesar

(2002) Exercise enhances and protects brain function. Exerc Sport Sci Rev 30, 75–79.

10.

Erickson

, Voss

, Prakash

, Basak

, Szabo

, Chaddock

, Kim

, Heo

, Alves

, White

, Wojcicki

, Mailey

, Vieira

, Martin

, Pence

, Woods

, Mcauley

, Kramer

(2011) Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A 108, 3017–3022.

11.

Rovio

, Spulber

, Nieminen

, Niskanen

, Winblad

, Tuomilehto

, Nissinen

, Soininen

, Kivipelto

(2010) The effect of midlife physical activity on structural brain changes in the elderly. Neurobiol Aging 31, 1927–1936.

12.

Tan

, Spartano

, Beiser

, Decarli

, Auerbach

, Vasan

, Seshadri

(2017) Physical activity, brain volume, and dementia risk: The Framingham Study. J Gerontol A Biol Sci Med Sci 72, 789–795.

13.

García-Mesa

, Colie

, Corpas

, Cristòfol

, Comellas

, Nebreda

, Giménez-Llort

, Sanfeliu

(2016) Oxidative stressis a central target for physical exercise neuroprotection againstpathological brain aging. J Gerontol A Biol Sci Med Sci 71, 40–49.

14.

Valenzuela

, Castillo-García

, Morales

, De La Villa

, Hampel

, Emanuele

, Lista

, Lucia

(2020) Exercise benefits on Alzheimer’s disease: State-of-the-science. Ageing Res Rev 62, 101108.

15.

Luck

, Riedel-Heller

, Luppa

, Wiese

, Köhler

, Jessen

, Bickel

, Weyerer

, Pentzek

, König

, Prokein

, Ernst

, Wagner

, Mösch

, Werle

, Fuchs

, Brettschneider

, Scherer

, Maier

(2014) Apolipoprotein E epsilon 4 genotype and aphysically active lifestyle in late life: Analysis ofgene-environment interaction for the risk of dementia andAlzheimer’s disease dementia. Psychol Med 44, 1319–1329.

16.

Kim

, Stewart

, Bae

, Kim

, Yang

, Park

, Shin

, Yoon

(2011) Role of BDNF Val66met polymorphism on the associationbetween physical activity and incident dementia. NeurobiolAging 32, 551.e5–12.

17.

Salvadó

, Milà-Alomà

, Shekari

, Minguillon

, Fauria

, Niñerola-Baizán

, Perissinotti

, Kollmorgen

, Buckley

, Farrar

, Zetterberg

, Blennow

, Suárez-Calvet

, Molinuevo

, Gispert

(2021) Cerebral amyloid-β load is associatedwith neurodegeneration and gliosis: Mediation by p-tau andinteractions with risk factors early in the Alzheimer’s continuum. Alzheimers Dement 17, 788–800.

18.

Buerger

, Ewers

, Pirttilä

, Zinkowski

, Alafuzoff

, Teipel

, Debernardis

, Kerkman

, Mcculloch

, Soininen

, Hampel

(2006) CSF phosphorylated tau protein correlates with neocortical neurofibrillary pathology in Alzheimer’s disease. Brain 129, 3035–3041.

19.

Seppälä

, Nerg

, Koivisto

, Rummukainen

, Puli

, Zetterberg

, Pyykkö

, Helisalmi

, Alafuzoff

, Hiltunen

, Jääskeläinen

, Rinne

, Soininen

, Leinonen

, Herukka

(2012) CSF biomarkers for Alzheimer disease correlate with cortical brain biopsy findings. Neurology 78, 1568–1575.

20.

Hansson

, Lehmann

, Otto

, Zetterberg

, Lewczuk

(2019) Advantages and disadvantages of the use of the CSF amyloid β (Aβ) 42/40 ratio in the diagnosis of Alzheimer’s disease. Alzheimers Res Ther 11, 34.

21.

Adlard

, Perreau

, Pop

, Cotman

(2005) Voluntary exercise decreases amyloid load in a transgenic model of Alzheimer’s disease. J Neurosci 25, 4217–4221.

22.

Kim

, Shin

, Kim

, Baek

, Ko

, Kim

(2014) Treadmill exercise improves short-term memory by enhancing neurogenesis in amyloid beta-induced Alzheimer disease rats. J Exerc Rehabil 10, 2–8.

23.

Arendash

, Garcia

, Costa

, Cracchiolo

, Wefes

, Potter

(2004) Environmental enrichment improves cognition in aged Alzheimer’s transgenic mice despite stable beta-amyloid deposition. Neuroreport 15, 1751–1754.

24.

Liang

, Mintun

, Fagan

, Goate

, Bugg

, Holtzman

, Morris

, Head

(2010) Exercise and Alzheimer’s disease biomarkers in cognitively normal older adults. Ann Neurol 68, 311–318.

25.

, Tan

, Bi

, Shen

, Xu

, Ma

, Li

, Dong

, Yu

(2020) Dynamic changes of CSF Strem2 in preclinical Alzheimer’s disease: The CABLE Study. Mol Neurodegener 15, 25.

26.

Mckhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr., , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association Workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

27.

Haskell

, Lee

, Pate

, Powell

, Blair

, Franklin

, Macera

, Heath

, Thompson

, Bauman

(2007) Physical activity and public health: Updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation 116, 1081–1093.

28.

Olsson

, Lautner

, Andreasson

, Öhrfelt

, Portelius

, Bjerke

, Hölttä

, Rosén

, Olsson

, Strobel

, Wu

, Dakin

, Petzold

, Blennow

, Zetterberg

(2016) CSF and bloodbiomarkers for the diagnosis of Alzheimer’s disease: A systematicreview and meta-analysis. Lancet Neurol 15, 673–684.

29.

Racine

, Koscik

, Nicholas

, Clark

, Okonkwo

, Oh

, Hillmer

, Murali

, Barnhart

, Betthauser

, Gallagher

, Rowley

, Dowling

, Asthana

, Bendlin

, Blennow

, Zetterberg

, Carlsson

, Christian

, Johnson

(2016) Cerebrospinal fluid ratios with Aβ42 predict preclinical brain β-amyloid accumulation. Alzheimers Dement (Amst) 2, 27–38.

30.

Kivipelto

, Helkala

, Laakso

, Hänninen

, Hallikainen

, Alhainen

, Soininen

, Tuomilehto

, Nissinen

(2001) Midlife vascular risk factors and Alzheimer’s disease in later life: Longitudinal, population based study. BMJ 322, 1447–1451.

31.

Launer

(2002) Demonstrating the case that AD is a vascular disease: Epidemiologic evidence. Ageing Res Rev 1, 61–77.

32.

Cotman

, Berchtold

(2002) Exercise: A behavioral intervention to enhance brain health and plasticity. Trends Neurosci 25, 295–301.

33.

Schmitz

, Schneider

, Kummer

, Herzog

(2004) Endoplasmic reticulum-localized amyloid beta-peptide is degraded in the cytosol by two distinct degradation pathways. Traffic 5, 89–101.

34.

Nichol

, Poon

, Parachikova

, Cribbs

, Glabe

, Cotman

(2008) Exercise alters the immune profile in Tg2576 Alzheimer mice toward a response coincident with improved cognitive performance and decreased amyloid. J Neuroinflammation 5, 13.

35.

Barnard

, Bush

, Ceccarelli

, Cooper

, De Jager

, Erickson

, Fraser

, Kesler

, Levin

, Lucey

, Morris

, Squitti

(2014) Dietary and lifestyle guidelines for the prevention of Alzheimer’s disease. Neurobiol Aging 35(Suppl 2), S74–78.

36.

Lucia

, Ruiz

(2011) Exercise is beneficial for patients with Alzheimer’s disease: A call for action. Br J Sports Med 45, 468–469.

37.

Hou

, Dan

, Babbar

, Wei

, Hasselbalch

, Croteau

, Bohr

(2019) Ageing as a risk factor for neurodegenerative disease. Nat Rev Neurol 15, 565–581.

38.

Khan

, Hillman

(2014) The relation of childhood physical activity and aerobic fitness to brain function and cognition: A review. Pediatr Exerc Sci 26, 138–146.

39.

Whiteman

, Young

, He

, Chen

, Wagenaar

, Stern

, Schon

(2014) Interaction between serum BDNF andaerobic fitness predicts recognition memory in healthy young adults. Behav Brain Res 259, 302–312.

40.

Kirk-Sanchez

, Mcgough

(2014) Physical exercise and cognitive performance in the elderly: Current perspectives. Clin Interv Aging 9, 51–62.

41.

Van Gelder

, Tijhuis

, Kalmijn

, Giampaoli

, Nissinen

, Kromhout

(2004) Physical activity in relation to cognitive decline in elderly men: The FINE Study. Neurology 63, 2316–2321.

42.

Salloway

(2009) Current and future treatments for Alzheimer’s disease.4-7; Discussion. CNS Spectr 14, 16–18.