Effect of Physical Activity on the Progression of Alzheimer’s Disease: The Clinical Research Center for Dementia of South Korea Study

Abstract

Background:

There is a lack of research on the effects of physical activity (PA) on the progression of Alzheimer’s disease (AD).

Objectives:

We investigated whether PA is associated with progression of dementia and mortality in AD.

Methods:

In the present study, 934 patients with mild-to-moderate AD were included. PA was evaluated using a questionnaire written by the caregiver. The outcome measures were the Clinical Dementia Rating-Sum of Boxes (CDR-SB), Seoul-Instrumental Activities of Daily Living (S-IADL), Caregiver-Administered Neuropsychiatric Inventory (CGA-NPI), a global composite score of neuropsychological subtests, and mortality. They were evaluated annually and received a maximum of three follow-up examinations.

Results:

Between-group differences compared with the no PA group in the change of CDR-SB scores were –0.431 (95% CI = –0.824∼–0.039; p = 0.031) for the moderate PA group (150–750 minutes per week of moderate intensity PA), and –1.148 (–1.656∼–0.639; p < 0.001) for the high PA group (>750 minutes per week). As PA increased, there was a significant trend to slow the rate of increase in the CDR-SB, S-IADL, and CGA-NPI scores. The patients with ≥150 minutes per week for each of non-recreational and recreational PAs had a lower risk of mortality compared to those with <150 minutes per week for each of the PAs (hazard ratio 0.22, 95% CI = 0.05∼0.88; p = 0.033).

Conclusion:

More PA is associated with slower progression of dementia severity, functional decline, and abnormal behavior, and with a lower risk of mortality in AD.

Keywords

Alzheimer’s disease dementia physical activity progression mortality

INTRODUCTION

Alzheimer’s disease (AD) is the most common cause of dementia. It is a neurodegenerative disorder characterized by an insidious onset and a progressive deterioration in cognition, functional ability, and behavior [1]. In patients with AD, the final year of life is characterized by persistently severe disability [2]. Societal costs associated with AD increase with increasing dementia severity [3]. Current available pharmacological interventions for AD include cholinesterase inhibitors and the N-methyl-D-aspartate receptor antagonist memantine, with the primary goal of symptomatic improvement [4]. Despite extensive research in the field of dementia, no treatment has yet been developed to modify the progression of AD. Alternative therapies, including non-pharmacological approaches, are urgently needed.

A systematic meta-analysis of prospective studies showed that nondemented subjects who performed physical activity (PA) were significantly protected against cognitive decline during follow-up [5]. Poor physical fitness at midlife was associated with increased dementia risk in old age [6]. Moreover, a dose-response has been identified in healthy subjects showing that higher levels of PA are associated with better cognitive performance [5, 7]. A systemic review of randomized controlled trials (RCTs) has shown that aerobic PA is beneficial for cognitive function in healthy older adults [8].

The RCTs with older subjects with mild cognitive impairment (MCI) reported some positive effects of physical exercise on cognition, mainly on global cognition, executive function, attention, and delayed recall [9, 10]. However, most RCTs performed with older subjects with dementia showed no effect of exercise on cognition or depression [9, 11]. A systemic review showed only that exercise programs may improve the ability to perform activities of daily living (ADLs) in those with dementia [11]. A recent RCT did not show positive effects of supervised moderate-to-high intensity exercise on cognitive tests and quality of life in patients with mild AD but showed a positive effect of moderate-to-high exercise on the Neuropsychiatric Inventory (NPI) [12].

Many patients with AD might not have adhered to the exercise program in the RCTs because of indifference and a lack of attention. Indifference or apathy is the most common behavior in AD [13, 14]. Furthermore, in most RCTs, the PA levels outside the exercise sessions were not monitored. In addition, studies on PA have mainly focused on recreational activities, and there is little evidence regarding the effect of other forms of PA, such as housework and daily activity. However, higher non-recreational as well as recreational PA are associated with a lower risk of cardiovascular events [15]. In patients with dementia, recreational PA may be reduced more than daily activities due to abulia.

Long-term follow-up studies will be needed to investigate the effects of exercise on AD, if exercise reduces the progression of the disease rather than improving the symptoms. It is challenging to perform a long-term RCT to observe exercise effects in patients with dementia. The effects of PA on dementia progression or survival have also not been investigated in previous long-term observational studies seeking to detect prognostic factors associated with AD progression or survival [16 –18].

Furthermore, the level of exercise that would benefit patients with AD is not clear. We aimed to investigate whether recreational and non-recreational PAs are associated with progression of dementia over three years and mortality from AD patients enrolled in the Clinical Research Center for Dementia of South Korea (CREDOS) study.

MATERIALS AND METHODS

Participants

In the present study, 934 patients with mild-to-moderate AD with a Clinical Dementia Rating (CDR) scale score of 2 or less [19] were selected from the participants of the CREDOS study, which is a multi-center, hospital-based registry study (November 2005 to January 2015) [20]. The patients included in the present study met the criteria for probable AD proposed by the National Institute of Neurological and Communicative Disorders and Stroke and the AD and Related Disorders Association [21] as well as the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition [22]. The PA questionnaires started being administered in the CREDOS study in 2008. Therefore, in this study we included the AD patients who registered from 2008 to 2013, who had a reliable caregiver who could provide investigators with the requested information, who completed the PA questionnaire at the baseline, and who had an annual follow-up evaluation at least once after the baseline assessment. The exclusion criteria included an abnormal result on the thyroid function test (TFT), a deficiency in vitamin B12 or folate, syphilis, major depressive disorder, psychosis, mental retardation, a history of encephalitis or metabolic encephalopathy, head trauma with loss of consciousness longer than 1 h, a brain tumor, a traumatic intracranial hemorrhage or a subarachnoid hemorrhage, sick-sinus syndrome, a second or third degree atrioventricular block, severe pulmonary, hepatic or renal disease, an active gastric ulcer, uncontrolled diabetes mellitus (DM), or uncured malignancy. Patients who had severe periventricular and deep white matter hyperintensities, as indicated by a score of three on the Fazekas ischemic scale were excluded [23]. We also excluded patients with clinical presentations of other types of dementia such as dementia with Lewy bodies, progressive supranuclear palsy, or frontotemporal lobar degeneration.

The study was performed in accordance with the International Harmonization Conference guidelines on Good Clinical Practice and was approved by the institutional review board of each center prior to beginning the study. Prior to participation in the study, all participants with AD and their legal representatives provided written informed consent to participate in the study.

Clinical assessment

All participants underwent physical and neurological examinations and thorough diagnostic procedures including an interview regarding the participants’ cognition, abnormal behaviors, ADL, demographic characteristics, vascular risk factors, current medication, and other comorbidities; the Mini-Mental State Examination (MMSE) [24]; the Geriatric Depression Scale of 15 items (GDS-15) [25]; the CDR [19]; the Caregiver-Administered NPI (CGA-NPI) [26]; and the Seoul-Instrumental ADL (S-IADL) [27].

The participants underwent an extensive standardized neuropsychological battery, the Seoul Neuropsychological Screening Battery (SNSB) [28]. They also underwent brain MRIs and laboratory tests that included TFT, HbA1c levels, fasting glucose levels, BUN, creatinine, liver and lipid panel, CBC, serum levels of vitamin B12 and folate, and the venereal disease research laboratory test.

Hypertension was defined as systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or use of antihypertensive medication [29]. DM was defined based on current treatment with insulin or oral hypoglycemic medication, 8-h fasting plasma glucose ≥126 mg/dl or HbA1c ≥6.5% [30]. Dyslipidemia was defined as total cholesterol ≥200 mg/dl, LDL-C ≥130 mg/dl, HDL-C<40 mg/dl, triglyceride level ≥150 mg/dl or the use of lipid-lowering drugs [31]. Participants’ weight and height were measured while they were wearing light clothing. Body mass index (BMI) was calculated as their weight (kg) divided by the square of their height (m²).

The participants were assessed annually using the same clinical and neuropsychological protocols as the baseline. All of 934 patients had follow-up evaluation at least once after the baseline assessment. Seven hundred ninety-one patients were reassessed one year after the baseline, 388 were reassessed two years after the baseline, and 136 were also reassessed three years after the baseline. However, the extensive neuropsychological tests were not followed in some patients. Of the total subjects, 186 (19.9%) refused to perform the extensive neuropsychological tests in the follow-up assessment, and 71 (7.6%) were unable to perform the neuropsychological tests due to severe dementia in the follow-up assessment.

Physical activity assessment

The PA was evaluated using the questionnaire written by the caregiver. The International PA Questionnaire (IPAQ) was designed to be used by adults aged 18–65 years [32]. We used a modified version of the IPAQ long form for the elderly. At first, the caregiver replied ‘Yes’ or ‘No’ to the question ‘Is the patient currently doing PA or exercise?’. If the caregiver replied ‘Yes’ to the question, he/she chose all the activities that the patient was performing among light walking, fast walking, bare-hand gymnastics, farming, housework (cleaning, laundry, etc.), running, climbing, aerobic dancing, and biking, and wrote the number of days a week and the number of hours per day. There were patients with moderate intensity PA as well as vigorous intensity PA.

All PA was also transformed into minutes per week of moderate intensity PA using the equation where minutes reported in each PA are weighted relative to moderate intensity PA according to the guidelines of the IPAQ [15, 32]:

Minute (min) per week = 0.825×light walking min + 1.25×fast walking min + 1×bare-hand gymnastics min + 1.375×farming min + 0.75×housework min + 2×running min + 2×climbing min + 2×aerobic dancing min + 1.5×biking min.

Total PA was categorized as low, moderate, and high PA, corresponding to less than 150 minutes per week, 150–750 minutes per week, and more than 750 minutes per week of moderate intensity PA [15]. The PA was also dichotomized as meeting or not meeting current PA guidelines (PA ≥150 minutes per week of moderate intensity PA as per World Health Organization (WHO) [33]) with periods of less than 10 minutes of PA not included as per the IPAQ guidelines [32].

Outcomes

We used 4 summary scores to index dementia progression, namely the CDR-Sum of Boxes (CDR-SB), the S-IADL, the CGA-NPI, and a global composite score. The CDR assesses dementia severity along 5 levels of impairment (rated as 0, 0.5, 1, 2, or 3) in each of 6 domains, namely memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care. The CDR-SB score (range 0–18) is the sum of the ratings in each of the 6 domains [19]. The S-IADL score (range 0–45) was determined using a 15-item questionnaire about activities that included using the telephone, shopping, meal preparation, housekeeping, mode of transportation, travelling a short distance, taking medication, handling money, grooming, using electrical equipment, finding belongings, locking a door, keeping an appointment, talking about a recent event, and hobbies. For each item, a four-point scale (0–3) was used that ranged from 3, ‘unable to do’, to 0, ‘independent activity’ [27]. The CGA-NPI (range 0–144) is a questionnaire form of the original NPI, in which caregivers completed the written form of the NPI worksheet [26]. The global composite score was estimated by averaging the z scores of the Digit Span Forward and Backward test, the Korean version of the Boston Naming Test [34], copying, immediate and 20-min delayed recalls, and the recognition tests of the Rey-Osterrieth Complex Figure, three learning-free recall trials of 12 words, 20-min delayed recall, and the recognition tests of the Seoul Verbal Learning Test, the category fluency test, the Controlled Oral Word Association Test, and the Stroop Color-Word test in the SNSB [28]. These scores were based on the means and standard deviations of each measure in the age- and education-matched control group. Increases in scores represent worsening for the CDR-SB, CGA-NPI, and S-IADL, and improvement in the global composite score.

We also investigated deaths as an outcome variable. Data on mortality in this study was obtained from the nationwide mortality database of Statistics Korea. In South Korea, all deaths are compiled into a national database and managed by Statistics Korea. Each subject in this study was tracked for mortality from the baseline to December 31, 2016.

Statistical analyses

Baseline characteristics were described for all participants and were stratified by no, low, moderate, and high total PA. One-way analysis of variance (ANOVA) and Tukey’s post hoc tests were used to test the heterogeneity of age, education, BMI, GDS-15, MMSE, and the time that the caregiver not living with the patient spent with him/her per week across the four PA levels. For the categorical variables, namely gender, hypertension, DM, dyslipidemia, heart disease, current smoking, current alcohol consumption, and the number of participants living with their caregivers, we calculated the frequencies and compared their differences across PA levels using the χ² test. The association of the variables with the four clinical outcomes was assessed using Pearson or Spearman correlations.

We used linear mixed models with a random subject effect to analyze the relationship between the four clinical outcomes and the four baseline PA levels, controlling for age, sex, education, time, BMI, GDS-15, and baseline MMSE. We also analyzed the relationship between the four clinical outcomes and whether meeting PA guideline, and the relationship between the four clinical outcomes and the types of PA (recreational or non-recreational) by using linear mixed model. The resultant slope estimates reflect the influence of baseline PA on clinical outcomes over time. Linear mixed models included all 4 time points (baseline, 1, 2, and 3 years), thus using all available data to maximize statistical efficiency while also controlling for any existing baseline differences in performance. We used the 1st order autoregressive (AR1) covariance structure in the mixed model, which was selected by the Akaike information criterion (AIC) of the model. The AR1 covariance structure considers correlations to be the greatest for nearby times, with an exponentially decreasing correlation with increasing distance between time points. To examine the effects of different amounts and types of PA on survival, a Cox proportional hazards model was used after controlling for age, gender, and MMSE. Data are presented as hazard ratios (HRs) and 95% confidence intervals (CIs). The statistical analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC, USA) and SPSS 19.0 (SPSS, Chicago, IL, USA). The values of p < 0.05 were considered statistically significant.

RESULTS

A total of 934 patients with AD were included in the present study. Table 1 presents the baseline demographic and clinical characteristics of the patients stratified by no, low, moderate, and high PA. There were 380 (40.7%) patients who did not engage in PA, 155 (16.6%) in the low PA group, 267 (28.6%) in the moderate PA group, and 132 (14.1%) in the high PA group. The mean age was higher in the no and low PA groups than in the moderate and high PA groups, and higher in the moderate PA group than in the high PA group. The proportion of males was higher in the moderate and high PA groups than in the no and low PA groups. The mean educational level and mean MMSE score were higher in the moderate PA group than in the no PA group. The BMI was lower in the high PA group than in the no and low PA groups. There were no significant differences in the prevalence of hypertension, DM, dyslipidemia, heart disease, current smoking, current alcohol consumption, or mean GDS-15 score across the four groups. The caregivers consisted of 339 (36.3%) spouses, 557 (59.6%) children, and 38 (4.1%) others. The number of participants who lived with their caregivers was 284 (74.7%), 104 (67.5%), 203 (77.5%), and 107 (77.5%) in the no, low, moderate, and high PA groups, respectively, and was not different across the four groups (p = 0.118). The mean time that the caregiver not cohabiting with the patient spent with him/her per week was 9.8±17.1 hours. There was no significant difference in the time among the four PA groups (p = 0.263).

Table 1.

Baseline characteristics of the participants stratified by total physical activity

		Physical activity
	Overall	No	Low	Moderate	High	p
	(n = 934)	(n = 380)	(n = 155)	(n = 267)	(n = 132)
Age, y	74.6 (7.7)	75.8 (7.2)^{‡ §}	76.2 (7.5)^{‡ §}	73.6 (8.1)^{* † §}	71.5 (7.6)^{* † ‡}	<0.001
Male	272 (29.1%)	102 (26.8%)	34 (21.9%)	91 (34.1%)	45 (34.1%)	0.022
Education, y	6.8 (5.5)	6.3 (5.3)^‡	6.7 (5.7)	7.6 (5.7)^*	7.0 (5.2)	0.025
Hypertension	500 (53.5%)	211 (55.5%)	86 (55.5%)	133 (49.8%)	70 (53.0%)	0.438
Diabetes mellitus	214 (22.9%)	101 (26.6%)	33 (21.3%)	55 (20.6%)	25 (18.9%)	0.144
Dyslipidemia	488 (52.2%)	218 (57.4%)	79 (51.0%)	129 (48.3%)	62 (47.0%)	0.061
Heart disease	132 (14.1%)	62 (16.3%)	23 (14.8%)	31 (11.6%)	16 (12.1%)	0.305
Current smoking	75 (8.0%)	38 (10.0%)	9 (5.8%)	19 (7.1%)	9 (6.8%)	0.308
Current alcohol consumption	315 (33.7%)	118 (31.1%)	47 (30.3%)	100 (37.5%)	50 (37.9%)	0.196
BMI, kg/m²	23.3 (3.2)	23.5 (3.4)§	23.5 (3.2)§	23.3 (3.2)	22.4 (2.7) * †	0.009
GDS-15	6.4 (4.6)	6.9 (4.8)	6.1 (4.2)	6.0 (4.6)	6.4 (4.6)	0.097
MMSE	18.5 (4.8)	18.0 (4.9)^‡	18.3 (4.7)	19.2 (4.6)^*	18.7 (4.8)	0.013

BMI, body mass index; GDS-15, Geriatric Depression Scale of 15 items; MMSE, Mini-Mental State Examination.

^*p < 0.05 versus no physical activity group,

^†p < 0.05 versus low physical activity group,

^‡p < 0.05 versus moderate physical activity group,

^§p < 0.05 versus high physical activity group.

Age and education were associated with the CDR-SB (r = 0.194, p < 0.001; r = –0.124, p < 0.001), global composite score (r = 0.169, p < 0.001; r = –0.307, p < 0.001), and S-IADL (r = 0.237, p < 0.001; r = –0.121, p < 0.001). Female sex (rho = 0.147, p < 0.001) and BMI (r = 0.085, p = 0.011) were associated with the global composite score. The GDS-15 was associated with the CDR-SB (r = 0.087, p = 0.009) and the CGA-NPI (r = 0.116, p = 0.001). MMSE was associated with CDR-SB (r = –0.544, p < 0.001), global composite score (r = 0.278, p < 0.001), S-IADL (r = –0.456, p < 0.001), and CGA-NPI (r = –0.160, p < 0.001).

Of those without PA, 281(73.9%) still had no PA in the follow-up evaluation. Ninety-seven subjects (62.6%) in the low PA group, 224 (83.9%) in the moderate PA group, and 126 (95.5%) in the high PA group became less active in the follow-up evaluation. There were 391 in the CDR 0.5 group, 448 in the CDR1 group, and 95 in the CDR 2 group. Of the participants in CDR 0.5, 177 (45.3%) progressed to CDR 1, 23 (5.9%) progressed to CDR 2, and 6 (1.5%) progressed to CDR 3 over three years. Of the participants in CDR 1, 139 (31.0%) progressed to CDR 2, 20 (4.5%) progressed to CDR 3, and 13 (2.9%) improved to CDR 0.5 over three years. Of the participants in CDR 2, 13 (13.7%) progressed to CDR 3, and 11 (11.6%) improved to CDR 1 over three years. Approximately 22.4% of patients with mild dementia of CDR 0.5 or 1 progressed to moderate dementia of CDR 2 or severe dementia of CDR 3 over three years.

Table 2 shows the estimated mean difference in the rate of change of each clinical outcome for 1 standardized unit increase in the predictor. Between-group differences compared with the no PA group in the change of CDR-SB scores were –0.431 (95% confidence interval [CI] = –0.824∼–0.039; p = 0.031) for the moderate PA group, and –1.148 (95% CI = –1.656∼–0.639; p < 0.001) for the high PA group. The estimated mean differences in the change of S-IADL scores were –3.192 (95% CI = –4.563∼–1.821, p < 0.001) in the moderate PA group and –5.477 (95% CI = –7.249∼–3.706, p < 0.001) in the high PA group, compared with the no PA group. Between–group differences compared with the no PA group in the change of the CGA-NPI scores were –2.562 (95% CI = –4.839∼–0.284, p = 0.028) for the moderate PA group and –4.373 (95% CI = –7.337∼–1.408, p = 0.004) for the high PA group. There was a trend to slow the rate of increase in the CDR-SB (p < 0.001), S-IADL (p < 0.001), and CGA-NPI (p = 0.002) scores as PA increased. There was no significant difference in the rate of change of the global composite score between patients without PA and those in each PA group. Figure 1 graphically represents the relationship between baseline PA levels and changes over three years in clinical outcomes.

Table 2.

The influence of the predictors on clinical outcomes over time according to the linear mixed models

		CDR-SB		Global composite score		S-IADL		CGA-NPI
		Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p
PA	No	0
	Low	–0.289 (–0.753∼0.175)	0.221	–0.083 (–0.212∼0.046)	0.206	–0.843 (–2.454∼0.768)	0.305	–0.771 (–3.460∼1.919)	0.574
	Moderate	–0.431 (–0.824∼–0.039)	0.031	–0.037 (–0.148∼0.073)	0.508	–3.192 (–4.563∼–1.821)	<0.001	–2.562 (–4.839∼–0.284)	0.028
	High	–1.148 (–1.656∼–0.639)	<0.001	0.004 (–0.138∼0.146)	0.953	–5.477 (–7.249∼–3.706)	<0.001	–4.373 (–7.337∼–1.408)	0.004
	Trend	–1.792 (–2.601∼–0.984)	<0.001	0.030 (–0.196∼0.255)	0.798	–9.391 (–12.207∼–6.575)	<0.001	–7.455 (–12.166∼–2.743)	0.002
Age		0.038 (0.016∼0.060)	0.001	0.014 (0.008∼0.020)	<0.001	0.251 (0.175∼0.326)	<0.001	–0.058 (–0.184∼0.068)	0.366
Male		0.192 (–0.202∼0.586)	0.340	–0.124 (–0.234∼–0.014)	0.027	1.938 (0.564∼3.313)	0.006	1.260 (–1.029∼3.550)	0.280
Education		0.164 (0.127∼0.201)	<0.001	–0.090 (–0.100∼–0.080)	<0.001	0.435 (0.306∼0.563)	<0.001	0.319 (0.104∼0.534)	0.004
BMI		–0.025 (–0.075∼0.025)	0.331	0.010 (–0.004∼0.024)	0.150	0.060 (–0.114∼0.234)	0.499	–0.086 (–0.376∼0.203)	0.558
MMSE		–0.425 (–0.465∼–0.386)	<0.001	0.099 (0.088∼0.110)	<0.001	–1.211 (–1.349∼–1.073)	<0.001	–0.697 (–0.928∼–0.465)	<0.001
GDS-15		0.009 (–0.026∼0.045)	0.610	0.003 (–0.007∼0.013)	0.526	0.021 (–0.103∼0.144)	0.744	0.414 (0.208∼0.621)	<0.001
time		1.225 (1.117∼1.332)	<0.001	–0.154 (–0.185∼–0.123)	<0.001	3.833 (3.459∼4.207)	<0.001	2.404 (1.640∼3.168)	<0.001

PA, physical activity; BMI, body mass index; MMSE, Mini-Mental State Examination; GDS-15, Geriatric Depression Scale of 15 items; CDR-SB, Clinical Dementia Rating-Sum of Boxes; S-IADL, Seoul-Instrumental Activities of Daily Living; CGA-NPI, Caregiver-Administered Neuropsychiatric Inventory.

Fig. 1.

Predicted means of clinical outcomes according to total physical activity (PA) at baseline over three years for the participants with Alzheimer’s disease. The predicted means of clinical outcomes are presented for PA levels adjusted for age, sex, education, time, body mass index, Geriatric Depression Scale of 15 items, and Mini-Mental State Examination at baseline. Lower scores represent worse performance for the global composite score (B), whereas higher scores represent decline in the Clinical Dementia Rating-Sum of Boxes (CDR-SB) (A), the Seoul-Instrumental Activities of Daily Living (S-IADL) (C), and the Caregiver-Administered Neuropsychiatric Inventory (CGA-NPI) (D). The patients with moderate or high PA at the baseline had significantly slower increases in CDR-SB, S-IADL, and CGA-NPI scores compared to those without PA.

Higher education and lower MMSE scores at baseline were associated with faster rates of increase for the CDR-SB, S-IADL, and CGA-NPI scores and with a faster decline in the global composite score. Higher age was associated with faster increasing rates for the CDR-SB and S-IADL scores. The male gender was associated with a faster increase in the S-IADL score and a faster decline in the global composite score compared with the female gender. Higher GDS-15 score was associated with a faster increasing rate of the CGA-NPI score.

The participants meeting the current PA guideline were associated with slower decline in the CDR-SB, S-IADL, and CGA-NPI compared with those not meeting the guideline (Table 3). In comparison to the participants with less than 150 minutes per week for each of the moderate intensity non-recreational and recreational PAs, the participants with 150 minutes or more per week of the moderate intensity non-recreational PA were associated with slower decline in the CDR-SB, S-IADL, and CGA-NPI, and those only with 150 minutes or more per week of the moderate intensity recreational PA were associated with slower decline in the S-IADL (Table 3). There was a trend to slow the rate of increase in the CDR-SB (p = 0.007) and S-IADL (p = 0.001) scores in participant with 150 minutes or more per week for each of non-recreational and recreational PAs rather than those with at least 150 minutes per week for only one from recreational and non-recreational PAs.

Table 3

The influence of meeting the current physical activity (PA) guideline and the types of PA (recreational or non-recreational) on clinical outcomes over time according to the linear mixed models

Time per week of moderate intensity PA	CDR-SB		Global composite score		S-IADL		CGA-NPI
	Estimate (95% CI)	p *	Estimate (95% CI)	p *	Estimate (95% CI)	p *	Estimate (95% CI)	p *
Meeting the current PA guideline^†	–0.569 (–0.901, –0.236)	<0.001	0.0001 (–0.093, 0.093)	0.998	–3.657 (–4.816, –2.498)	<0.001	–2.892 (–4.816, –0.968)	0.003
Recreational PA <150 min &non-recreational PA <150 min	0		0		0		0
Recreational PA ≥150 min &non-recreational PA <150 min	–0.160 (–0.559∼0.239)	0.432	–0.083 (–0.195∼0.029)	0.146	–2.164 (–3.554∼–0.774)	0.002	–1.327 (–3.640∼0.986)	0.261
Recreational PA <150 min &non-recreational PA ≥150 min	–0.956 (–1.480∼–0.432)	<0.001	0.092 (–0.055∼0.239)	0.219	–5.436 (–7.258∼–3.614)	<0.001	–5.212 (–8.249, –2.174)	0.001
Recreational PA ≥150 min &non-recreational PA ≥150 min	–1.213 (–1.881∼–0.545)	<0.001	0.053 (–0.135∼0.241)	0.579	–5.424 (–7.774∼–3.074)	<0.001	–4.526 (–8.415, –0.637)	0.023
Trend	–1.422 (–2.462∼–0.382)	0.007	–0.008 (–0.300∼0.284)	0.958	–6.500 (–10.155∼–2.844)	0.001	–4.847 (–10.899, 1.205)	0.116

CDR-SB, Clinical Dementia Rating-Sum of Boxes; S-IADL, Seoul-Instrumental Activities of Daily Living; CGA-NPI, Caregiver-Administered Neuropsychiatric Inventory; min, minutes.

^* Adjusted for age, sex, education, body mass index, Geriatric Depression Scale of 15 items, Mini-Mental State Examination, and time. † The reference group consisted of the participants not meeting the current PA guideline (<150 min per week of moderate intensity PA).

The mortality risk was not associated with education, BMI and GDS-15, but was significantly associated with age, gender, MMSE, and four PA groups in an unadjusted Cox proportional hazards regression analysis, respectively. However, HRs adjusted for age, gender, and MMSE were not significantly different among the four PA groups. The adjusted HR was not also significantly different between the participants who met the PA guideline and those who did not meet the PA guideline. The participants with 150 minutes and more per week for each of non-recreational and recreational PAs had a lower risk of mortality compared to those with less than 150 minutes per week for each of non-recreational and recreational PAs (HR 0.22, 95% CI = 0.05∼0.88; p = 0.033) (Table 4 and Fig. 2).

Table 4

The hazard ratios of mortality according to the types of physical activity estimated by the Cox proportional hazards model

Time per week of moderate intensity PA	Total	Event	HR (95% CI)	p *
Recreational PA <150 min &	544	134	1 (reference)
non-recreational PA <150 min
Recreational PA ≥150 min &	225	52	1.04 (0.74∼1.45)	0.126
non-recreational PA <150 min
Recreational PA <150 min &	105	13	0.64 (0.36∼1.13)	0.126
non-recreational PA ≥150 min
Recreational PA ≥150 min &non-recreational PA ≥ 150 min	60	2	0.22 (0.05∼0.88)	0.033
p for trend			p = 0.011

PA, physical activity; HR, Hazard ratio; CI, confidence interval; min, minutes.

^* Adjusted for age, gender, and Mini-Mental State Examination.

Fig. 2.

Survival curves comparing the cumulative proportions of participants according to the types of physical activity (PA). Curves are based on Cox proportional hazards regression analysis adjusting for age, gender, and Mini-Mental State Examination as covariates. The participants with 150 minutes or more per week for each of non-recreational and recreational PAs (a purple line) had a lower risk of mortality compared to those with less than 150 minutes per week for each of non-recreational and recreational PAs (a blue line) (Hazard ratio 0.22, 95% confidence interval = 0.05∼0.88; p = 0.033).

DISCUSSION

In this study, moderate or high PA at the baseline was associated with a slower progression of dementia as assessed by the CDR-SB compared to AD patients without PA. Moderate or high PA at the baseline was also associated with a slower functional deterioration as assessed by the S-IADL and a slower increase in abnormal behavior as assessed by the CGA-NPI when compared to those without PA. These associations were independent of age, sex, education, time, BMI, GDS-15, and the baseline MMSE. Moreover, a dose-response was identified in the CDR-SB, S-IADL, and CGA-NPI, and more PA was associated with slower progression of dementia severity, functional decline, and abnormal behaviors. These results are compatible with the previous RCTs that demonstrated that an exercise program improved the ability to perform ADLs or reduced neuropsychiatric symptoms in AD patients [12, 35].

The recommendation for PA for older adults from the WHO and the American College of Sports Medicine is 30 minutes or more of moderate intensity activities at least five days per week [33, 36]. However, the level of exercise that would benefit patients with AD is not clear. In our study, AD patients meeting the WHO guideline of 150 minutes or more per week of moderate intensity PA experienced slower progression of dementia severity, functional decline, and abnormal behavior compared to those who did not engage in PA or compared to those who did not meet the guideline. To slow the progression of dementia severity as well as the functional decline and increase in abnormal behavior, we suggest that a reasonable PA guideline for AD patients is 150 minutes or more of moderate intensity activities per week.

The benefit of PA in slowing functional decline was independent of the type of PA (recreational or non-recreational). The slower progression of dementia severity and abnormal behaviors was not associated with recreational PA but was associated with non-recreational PA (≥150 minutes per week of moderate intensity non-recreational PA). There may not have been many dementia patients performing recreational PA enough to meet the current PA guidelines because of the decrease in motivation. However, incorporating PA into a daily lifestyle whether through housework or farming has the potential to achieve higher PA. There was a trend to slow the rate of increase in the CDR-SB and S-IADL scores in the participants with 150 minutes or more per week for each of non-recreational and recreational PAs rather than those with at least 150 minutes per week for only one from recreational and non-recreational PAs. The participants with 150 minutes or more per week for each of non-recreational and recreational PAs also had a lower risk of mortality compared to those with less than 150 minutes per week for each of non-recreational and recreational PAs. These findings suggest the synergistic effect of recreational and non-recreational PAs.

In our study, PA was not associated with the rate of change of the global composite score estimated by averaging the z scores of the neuropsychological subtests. The systemic reviews also revealed no evidence of a beneficial effect of exercise on cognition [9 , 35]. However, there were some AD patients who refused neuropsychological tests or were unable to take the tests because of their dementia progression during the follow-up in this study. Because patients in whom dementia had worsened might be dropped or might not be administered the neuropsychological tests, the relationship between PA and the rate of change of the global composite score may have been rendered insignificant. Individuals who declined quickly tended to be dropped much earlier in a previous study about AD progression [16].

The animal studies showed that exercise induced the expression of brain-derived neurotrophic factor (BDNF), and BDNF has been shown to improve cognitive function in rats [37, 38]. A previous human study also revealed that aerobic exercise increased BDNF plasma levels in patients with AD and found a significant positive correlation between BDNF levels and the level of PA [39]. Physical exercise enhanced hippocampal neurogenesis in a dose-dependent manner in an animal study [40]. The human studies also demonstrated a dose-response showing that higher levels of PA are associated with better cognitive performance in non-demented healthy elderly individuals [5 , 41]. Our study showed that the more PA, the slower the increase in CDR-SB, S-IADL and CGA-NPI scores in AD patients. To our knowledge, out study is the first study to reveal a dose-response, showing that more PA is associated with slower progression in the severity of dementia, functional decline, and abnormal behaviors in AD patients.

Our study showed that the higher the education level and the lower the MMSE score at the baseline, the faster the functional and cognitive decline, the faster the progression of dementia, and the faster the increase in neuropsychiatric symptoms. Those are compatible with the results of previous studies, which showed that after the clinical onset of AD, patients with high education levels show more rapid cognitive and functional decline than patients with lower education levels [17, 42] and that cognitive test results at baseline were predictors of progression rate in AD [18]. Older patients with AD were associated with faster ADL decline and dementia progression, as assessed by CDR-SB in this study. Most of patients in this study had late-onset AD. Age may be poor prognostic factor in late-onset AD [18]. Male sex was associated with a faster decline in cognition and ADL in our study. In a previous study, the same clinical severity of dementia was associated with greater reductions in cerebral metabolism in men than in women, suggesting a greater amount of brain reserves in men [43]. Male sex may offer an increased ability to buffer pathological changes, and thus males may have a larger pathological burden than females at the disease onset, causing them to experience a faster decline afterward [44]. Depression was associated with physical inactivity in older adults [45]. A rapid progression of dementia was associated with increasing depressive symptoms in a previous study [46]. This study showed that higher depressive symptoms at baseline were associated with a faster increase of neuropsychiatric symptoms in AD.

Our study had several limitations. We did not measure PA with a quantitative method such as actigraphy but rather evaluated it with the caregiver-reported questionnaire. We also did not use self-reporting measures for PA. Thus, PA may be overestimated or underestimated. Second, even though our sample represents a natural history of a real-world population who visited memory clinics with a starting size of 934 AD subjects, the results should be interpreted with caution due to the small final sample. People who experience a faster decline might be dropped from the study at an earlier time point. Third, the study was observational. In the future, the effect of PA on AD progression should be investigated in a long-term follow-up, large-scale RCT. Especially, it may be interesting to have the no PA group exercise and document the clinical improvement. Fourth, the AD diagnoses of the patients in this study were not confirmed with AD biomarkers or biopsy. Fifth, the criterion of the caregiver was not defined strictly in the CREDOS study. Therefore, some of the caregivers who did not meet patients often may have provided incorrect information about the patient’s PA, behavior, and daily activities. Sixth, GDS-15 may not be reliable tool in moderately affected AD [47]. Thus, depression may not have been accurately assessed in patients with moderate stage of AD. Lastly, the participants included in this study usually attended tertiary hospital outpatient clinics focused on cognitive disorders and might not be representative of the general population.

In conclusion, despite the abovementioned limitations, our findings demonstrate that moderate intensity PA for 150 minutes or more per week is associated with slowing the progression of dementia severity, functional decline, and abnormal behaviors in AD. In addition, there was a dose-dependent response showing that the more PA, the slower the progression of dementia severity, functional decline, and abnormal behavior. The AD patients with 150 minutes or more per week for each of non-recreational and recreational PAs also had a lower risk of mortality. Our study may contribute to the formation of guidelines for the recommendation of a beneficial PA level for AD patients.

Footnotes

ACKNOWLEDGMENTS

This study was supported by grants from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (HI18C0479 and HI10C2020) and the Original Technology Research Program for Brain Science through the National Research Foundation of Korea (NRF) funded by the Korean government (MSIP) (NRF-2014M3C7A1064752 and NRF-2018M3A9F1023690).

Authors’ disclosures available online ().

References

Reisberg

, Burns

, Brodaty

, Eastwood

, Rossor

, Sartorius

, Winblad

(1997) Diagnosis of Alzheimer’s disease. Report of an International Psychogeriatric Association Special Meeting Work Group under the cosponsorship of Alzheimer’s Disease International, the European Federation of Neurological Societies, the World Health Organization, and the World Psychiatric Association. Int Psychogeriatr 9(Suppl 1), 11–38.

Gill

, Gahbauer

, Han

, Allore

(2010) Trajectories of disability in the last year of life. N Engl J Med 62, 1173–1180.

Olazarán

, Agüera-Ortiz

, Argimón

, Reed

, Ciudad

, Andrade

, Dilla

(2017) Costs and quality of life in community-dwelling patients with Alzheimer’s disease in Spain: Results from the GERAS II observational study. Int Psychogeriatr 29, 2081–2093.

Tan

, Yu

, Wang

, Tan

, Meng

, Wang

, Jiang

, Zhu

, Tan

(2014) Efficacy and safety of donepezil, galantamine, rivastigmine, and memantine for the treatment of Alzheimer’s disease: A systematic review and meta-analysis. J Alzheimers Dis 41, 615–631.

Sofi

, Valecchi

, Bacci

, Abbate

, Gensini

, Casini

, Macchi

(2011) Physical activity and risk of cognitive decline: A meta-analysis of prospective studies. J Intern Med 269, 107–117.

Kulmala

, Solomon

, Kåreholt

, Ngandu

, Rantanen

, Laatikainen

, Soininen

, Tuomilehto

, Kivipelto

(2014) Association between mid- to late life physical fitness and dementia: Evidence from the CAIDE study. J Intern Med 276, 296–307.

Weuve

, Kang

, Manson

, Breteler

, Ware

, Grodstein

(2004) Physical activity, including walking, and cognitive function in older women. JAMA 292, 1454–1461.

Angevaren

, Aufdemkampe

, Verhaar

, Aleman

, Vanhees

(2008) Physical activity and enhanced fitness to improve cognitive function in older people without known cognitive impairment. Cochrane Database Syst Rev 3, CD005381.

Ohman

, Savikko

, Strandberg

, Pitkälä

(2014) Effect of physical exercise on cognitive performance in older adults with mild cognitive impairment or dementia: A systematic review. Dement Geriatr Cogn Disord 38, 347–365.

10.

Lautenschlager

, Cox

, Flicker

, Foster

, von Bockxmeer

, Xiao

, Greenop

, Almeida

(2008) Effect of physical activity on cognitive function in older adults at risk for Alzheimer disease: A randomized trial. JAMA 300, 1027–1037.

11.

Forbes

, Forbes

, Morgan

, Markle-Reid

, Wood

, Culum

(2008) Physical activity programs for persons with dementia. Cochrane Database Syst Rev 16, CD006489.

12.

Hoffmann

, Sobol

, Frederiksen

, Beyer

, Vogel

, Vestergaard

, Brændgaard

, Gottrup

, Lolk

, Wermuth

, Jacobsen

, Laugesen

, Gergelyffy

, Høgh

, Bjer-regaard

, Andersen

, Siersma

, Johannsen

, Cotman

, Waldemar

, Hasselbalch

(2016) Moderate-to-high intensity physical exercise in patients with Alzheimer’s disease: A randomized controlled trial. J Alzheimers Dis 50, 443–453.

13.

Mega

, Cummings

, Fiorello

, Gornbein (1996) The spectrum of behavioral changes in Alzheimer’s disease. Neurology 46, 130–135.

14.

Choi

, Na

, Kwon

, Yoon

, Jeong

, Ha

(2000) The Korean version of the neuropsychiatric inventory: A scoring tool for neuropsychiatric disturbance in dementia patients. J Korean Med Sci 15, 609–615.

15.

Lear

, Hu

, Rangarajan

, Gasevic

, Leong

, Iqbal

, Casanova

, Swaminathan

, Anjana

, Kumar

, Rosengren

, Wei

, Yang

, Chuangshi

, Huaxing

, Nair

, Diaz

, Swidon

, Gupta

, Mohammadi-fard

, Lopez-Jaramillo

, Oguz

, Zatonska

, Seron

, Avezum

, Poirier

, Teo

, Yusuf

(2017) The effect of physical activity on mortality and cardiovascular disease in 130 000 people from 17 high-income, middle-income, and low-income countries: The PURE study. Lancet 390, 2643–2654.

16.

Barocco

, Spallazzi

, Concari

, Gardini

, Pelosi

, Caffarra

(2017) The progression of Alzheimer’s disease: Are fast decliners really fast? A four-year follow-up. J Alzheimers Dis 57, 775–786.

17.

Wattmo

, Wallin

ÅK

(2017) Early-versus late-onset Alzheimer disease: Long-term functional outcomes, nursing home placement, and risk factors for rate of progression. Dement Geriatr Cogn Dis Extra 7, 172–187.

18.

Carcaillon

, Pérès

, Péré

, Helmer

, Orgogozo

, Dartigues

(2007) Fast cognitive decline at the time of dementia diagnosis: A major prognostic factor for survival in the community. Dement Geriatr Cogn Disord 23, 439–445.

19.

Choi

, Lee

, Kim

, Hahm

, Jeong

, Yoon

, Jeong

, Ha

, Na

(2003) Interchanging scores between clinical dementia rating scale and global deterioration scale. Alzheimer Dis Assoc Disord 17, 98–105.

20.

Park

, Na

, Han

, Kim

, Cheong

, Kim

, Hong

, Kim

, Ku

, Moon

, Lee

, Shim

, Youn

, Kim

, Park

, Cha

, Seo

, Lee

(2011) Clinical characteristics of a nationwide hospital-based registry of mild-to-moderate Alzheimer’s disease patients in Korea: A CREDOS (Clinical Research Center for Dementia of South Korea) study. J Korean Med Sci 26, 1219–1226.

21.

McKhann

, Drachman

, Folstein

, Katzman

, Price

, Stadlan

(1984) Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s disease. Neurology 34, 939–944.

22.

American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders, 4th ed (DSM-IV). American Psychiatric Association, Washington DC.

23.

Fazekas

, Chawluk

, Alavi

, Hurtig

, Zimmerman

(1987) MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol 149, 351–356.

24.

Han

, Jo

, Kim

, Park

, Kang

(2008) An adaptation of the Korean Mini-Mental State Examination (K-MMSE) in elderly Koreans: Demographic influence and population-based norms (the AGE study). Arch Gerontol Geriatr 47, 302–310.

25.

Bae

, Cho

(2004) Development of the Korean version of the Geriatric Depression Scale and its short form among elderly psychiatric patients. J Psychosom Res 57, 297–305.

26.

Kang

, Choi

, Lee

, Jeong

, Hahm

, Han

, Cummings

, Na

(2004) Caregiver-Administered Neu-ropsychiatric Inventory (CGA-NPI). J Geriatr Psychiatry Neurol 17, 32–35.

27.

, Kim

, Kwon

, Kim

, Lee

, Ko

(2004) A study on the reliability and validity of Seoul-Instrumental Activities of Daily Living (S-IADL). J Korean Neuropsy-chiatr Assoc 43, 189–199.

28.

Ahn

, Chin

, Park

, Lee

, Suh

, Seo

, Na

(2010) Seoul Neuropsychological Screening Battery-dementia version (SNSB-D): A useful tool for assessing and monitoring cognitive impairments in dementia patients. J Korean Med Sci 25, 1071–1076.

29.

The Guidelines Subcommittee of the World Health Organization/International Society of Hypertension (WHO/ISH) (1999) 1999 World Health Organization-International Society of Hypertension guidelines for the management of hypertension. J Hypertens 17, 151–183.

30.

The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus (1997) Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 20, 1183–1197.

31.

National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) (2002) Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation 106, 3143–3421.

32.

Craig

, Marshall

, Sjöström

, Bauman

, Booth

, Ainsworth

, Pratt

, Ekelund

, Yngve

, Sallis

, Oja

(2003) International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 35, 1381–1395.

33.

World Health Organization (2010) Global recommendations on physical activity for health. World Health Organization, Geneva.

34.

Kim

, Na

(1999) Normative data on the Korean version of the Boston Naming Test. J Clin Exp Neuropsychol 21, 127–133.

35.

Forbes

, Forbes

, Blake

, Thiessen

, Forbes

(2015) Exercise programs for people with dementia. Cochrane Database Syst Rev 4, CD006489.

36.

Nelson

, Rejeski

, Blair

, Duncan

, Judge

, King

, Macera

, Castaneda-Sceppa

; American College of Sports Medicine; American Heart Association (2007) Physical activity and public health in older adults: Recommendation from the American College of Sports Medicine and the American Heart Association. Circulation 116, 1094–1105.

37.

Gomes da Silva

, Unsain

, Mascó

, Toscano-Silva

, de Amorim

, Silva Araújo

, Simões

, Naffah-Mazzacoratti Mda

, Mortara

, Scorza

, Cavalheiro

, Arida

(2012) Early exercise promotes positive hip-pocampal plasticity and improves spatial memory in the adult life of rats. Hippocampus 22, 347–358.

38.

Hopkins

, Nitecki

, Bucci

(2011) Physical exercise during adolescence versus adulthood: Differential effects on object recognition memory and brain-derived neurotrophic factor levels. Neuroscience 194, 84–94.

39.

Coelho

, Vital

, Stein

, Arantes

, Rueda

, Camarini

, Teodorov

, Santos-Galduróz

(2014) Acute aerobic exercise increases brain-derived neurotrophic factor levels in elderly with Alzheimer’s disease. J Alzheimers Dis 39, 401–408.

40.

Diederich

, Bastl

, Wersching

, Teuber

, Strecker

J-K

, Schmidt

, Minnerup

, Schabitz

(2017) Effects of different exercise strategies and intensities on memory performance and neurogenesis. Front Behav Neurosci 11, 47.

41.

Colcombe

, Kramer

(2003) Fitness effects on the cognitive function of older adults: A meta-analytic study. Psychol Sci 14, 125–130.

42.

Cho

, Jeon

, Kim

, Ye

, Kim

, Noh

, Kim

, Yoon

, Kim

, Park

, Kim

, Lee

, Kang

, Suh

, Chin

, Na

, Kang

, Seo

(2015) Higher education affects accelerated cortical thinning in Alzheimer’s disease: A 5-year preliminary longitudinal study. Int Psychogeriatr 27, 111–120.

43.

Perneczky

, Drzezga

, Diehl-Schmid

, Li

, Kurz

(2007) Gender differences in brain reserve: An (18)F-FDG PET study in Alzheimer’s disease. J Neurol 254,1395–1400.

44.

Roselli

, Tartaglione

, Federico

, Lepore

, Defazio

, Livrea

(2009) Rate of MMSE score change in Alzheimer’s disease: Influence of education and vascular risk factors. Clin Neurol Neurosurg 111, 327–330.

45.

Cho

, Jin

, Kang

(2018) Weight status, physical activity, and depression in Korean older adults. J Epidemiol 28, 292–299.

46.

Barca

, Persson

, Eldholm

, Benth

JŠ

, Kersten

, Knapskog

, Saltvedt

, Selbaek

, Engedal

(2017) Trajectories of depressive symptoms and their relationship to the progression of dementia. J Affect Disord 222,146–152.

47.

Debruyne

, Van Buggenhout

, Le Bastard

, Aries

, Audenaert

, De Deyn

, Engelborghs

(2009) Is the geriatric depression scale a reliable screening tool for depressive symptoms in elderly patients with cognitive impairment? Int J Geriatr Psychiatry 24, 556–562.