Effects of Physical Exercise on Alzheimer’s Disease Biomarkers: A Systematic Review of Intervention Studies

Abstract

Physical exercise may be an important adjunct to pharmacological treatment of Alzheimer's disease (AD). Animal studies indicate that exercise may be disease modifying through several mechanisms including reduction of AD pathology. We carried out a systematic review of intervention studies of physical exercise with hippocampal volume (on MRI), amyloid-β, total tau, phosphorylated tau in cerebrospinal fluid (CSF), ¹⁸F-FDG-PET or amyloid PET as outcome measures in healthy subjects, patients with subjective memory complaints, mild cognitive impairment, or AD. We identified a total of 8 studies of which 6 investigated the effects of exercise on hippocampal volume in healthy subjects and 1 on CSF biomarkers and 1 on hippocampal volume in AD, and none investigating the remaining outcome measures or patient groups. Methodological quality of identified studies was generally low. One study found a detrimental effect on hippocampal volume and one found a positive effect, whereas the remaining studies did not find an effect of exercise on outcome measures. The present systematic study identified a relatively small number of studies, which did not support an effect of exercise on hippocampal volume. Methodological issues such small to moderate sample sizes and inadequate ramdomization procedures further limits conclusions. Our findings highlight the difficulties in conducting high quality studies of exercise and further studies are needed before definite conclusions may be reached.

Keywords

Alzheimer’s disease amyloid-β dementia hippocampus magnetic resonance imaging phosphorylated tau physical activity physical exercise total tau

INTRODUCTION

Alzheimer’s disease (AD) is the leading cause of dementia worldwide [1]. At present, state-of-the-art treatment combines pharmacological treatment, chiefly with cholinesterase inhibitors and memantine, and supportive measures such as counseling, social support, and care. No disease-modifying treatment exists as of yet, but efforts within the last decades has aimed at developing pharmacological therapies targeted at reducing cortical amyloid-β (Aβ) and tau pathology, which are believed to be key in the pathogenesis of AD [2].

Physical exercise has received attention as an addition to treatment and a possible disease-modifying therapeutic approach. This presumption is based on promising findings in animal studies [3 –7] (reviewed in [8]) and in patients with AD [9, 10]. With regards to underlying mechanisms of action, several have been proposed. These include induction of angiogenesis [11], changes in brain connectivity [12], stimulation of neurotrophic factors, such as BDNF [13, 14], and anti-inflammatory effects [15, 16]. Moreover, interest has focused on a direct effect on pathological events in the amyloid cascade hypothesis, i.e., cortical deposition of Aβ, cortical atrophy, predominantly in the hippocampus, and tau. Adlard et al. demonstrated a decrease in cortical Aβ in a transgenic mouse model following 5 months of voluntary exercise [4]. Similar finding have been demonstrated in other AD animal models [5 –8]. Hyperphosphorylation of tau, which most likely also plays an important role in the pathophysiology of AD, was reduced after exercise in another study using a transgenic mouse model [17]. Furthermore, equally promising results have been reported regarding effects on hippocampal volume, possibly by induction of neurogenesis[18, 19].

In humans, a few observational studies have also lent support to an effect of physical activity on AD pathology. Liang et al. retrospectively demonstrated an association between physical exercise engagement and low concentration of total tau (t-tau) and phosphorylated tau (p-tau) in the cerebrospinal fluid (CSF) and low cortical Aβ in healthy elderly subjects [20]. These findings have been replicated [21, 22], including in autosomal dominant AD [23] and extended to include a similar beneficial effect on brain glucose metabolism and hippocampal volume [21]. However, the relationship may not be straightforward regarding physical activity and amyloid metabolism. Brown et al. [24] found that cortical amyloid levels were only influenced by physical activity level in ApoE ɛ4 positive individuals, whereas plasma Aβ was affected in ApoE ɛ4 negative. Moreover, in an MCI population, physical activity level did not attenuate cortical Aβ levels [25], although it was found to modify amyloid-driven cognitive impairment in a similarpopulation [26].

A relatively large number of systematic reviews and meta-analyses of intervention and observational studies have been published on the effects of physical exercise and activity on cognition in aging and dementia [10 , 27–31], the overall conclusion of which suggests a positive effect on cognition and risk of dementia. However, to our knowledge no systematic review or meta-analysis has been conducted examining the effects of physical exercise on AD biomarkers. Therefore, the objective of the present study was to assess the possible effects of physical activity on AD pathophysiology. Hence, we undertook a systematic review of randomized, controlled trials, which examined the effects of physical activity, on validated AD biomarkers, which are thought to reflect pivotal processes in the pathophysiological cascade in AD.

METHODS

The present study reports the results of a systematic review. The study was carried out according to the principles laid out in the PRISMA statement and checklist [32], and also follows the aforementioned in regards to reporting.

A protocol was written prior to the conduction of the systematic review.

Eligibility criteria for studies, participants, and outcome measures

Studies eligible for inclusion were intervention studies of physical exercise with one or more of the following AD biomarkers as outcome measures: 1) Aβ_1-42, t-tau, and/or p-tau in CSF; 2) cortical metabolism measured by ¹⁸F-FDG-PET imaging, 3) cortical amyloid deposition measured by amyloid-PET imaging; or 4) hippocampal volume measured on MRI (including studies using voxel-based morphometry analysis). The study had to be carried out in healthy subjects, subjects with subjective cognitive complaints, patients with MCI (according to any of the following criteria: NIA-AA-diagnostic criteria for MCI [33], preclinical AD [34] or International working group (IWG) criteria for preclinical AD [35], or Petersen criteria for MCI [36]) or patients with AD dementia (NINCDS-ADRDA [37], NIA-AA-diagnostic criteria for AD [38] or IWG criteria [39]) of above 17 years of age. Physical activity is defined according to the World Health Organization’s definition which states that ”Physical activity is defined as any bodily movement produced by skeletal muscles that requires energy expenditure” [40], whereas physical exercise is “Physical activity that is planned, structured, and repetitive for the purpose of conditioning any part of the body” [35]. For the present study, we adopted these definitions. An intervention period of more than 2 weeks was required. Intervention studies were included if a suitable control situation such as usual care or stretching was included. Furthermore, interventions could include activities with mostly aerobic components such as walking or running or non-aerobic components such as strength exercises. Studies had to be published from January 1984 to August 2017 in English. Only full-length articles which reported on original data were eligible for inclusion (i.e., reviews, editorials, case reports, conference papers, etc., were not considered for eligibility).

Search strategy, study selection, data collection, and evaluation of risk of bias

The following databases were searched: MEDLINE, EMBASE, Cochrane Register of Controlled Clinical Trials, PsycInfo and Web of Science. Limits were: English language, publication date 1984-August 2017. Search words were identified and refined by identifying key concepts and words in relevant publications together with the use of controlled vocabularies and thesauruses such as Medical Subject Headings. Detailed electronic searches for MEDLINE and EMBASE databases are published as supplementary data (Supplementary Table 1).

Bibliographies of eligible studies included in the final systematic review were hand searched for further eligible studies (backwards citation search). For forward citation search, studies included in the review were searched in Web of Science. Lastly, any references, which were known to the authors but not identified by the searches were considered for inclusion. Evaluation of eligibility was assessed independently by two reviewers (KF, LG). Studies returned by searches were screened based on title and/or abstract and full-text articles of selected studies were retrieved for further evaluation. Eligibility was assessed using a standardized check-list developed specifically for the present study (available upon request from the corresponding author), and any disagreement was resolved by consensus, and if this was not possible 1 further reviewer was consulted with the purpose of arbitration. Data was extracted from all identified reports independently by two reviewers (KF, LG) in a piloted data extraction sheet. Piloting was carried out on four studies. Protocol, search strings and data extraction sheets are available, upon request, from the corresponding author.

For intervention studies, the following data items were retrieved from individual reports: 1) title, authors, year of publication; 2) participant characteristics (age, gender, global cognitive function (e.g., MMSE score), dementia severity (e.g., scale assessing activities of daily living), diagnosis at baseline); 3) randomization procedure and blinding; 4) recruitment process and setting; 5) inclusion and exclusion criteria; 6) description of intervention and control situation (type, duration, intensity, compliance/attendance); 7) method of assessment of physical fitness; 8) biomarkers used; 9) specific method used for assessing biomarkers; 10) baseline and follow-up values for the biomarker and physical fitness in each group; and 11) changes in biomarkers and physical fitness from baseline tofollow-up.

Assessment of risk of bias was carried out independently by two reviewers in a rigorous manner to detect possible sources of bias (KF, LG). For the assessment of the risk of bias within studies, the Cochrane Collaboration’s tool for assessing risk of bias Checklist was used [41]. The checklist assesses several different domains, which may be sources of bias, and acted as a supplement to the assessment of bias within studies. No compound score is calculated since this has been shown to be insufficient to prevent bias [42]. Furthermore, rigorous reviews of evidence, manual searches and focus on possible missing data was carried out to reduce the risk of publication bias and reviewer selection bias.

RESULTS

Results of searches

A total of 55,114 citations were identified through the performed searches. Of these, 26 full-text papers were retrieved for further evaluation of eligibility, and 8 papers were included in the present study [43 –50]. The PRISMA flow-chart shows the flow and selection process of citations (Fig. 1). No additional studies were identified through forward searching or backwards citation searches.

Fig.1

Flow chart for included studies.

Characteristics of included studies

The 8 studies included in the systematic review contained a total of 584 participants (323 in intervention group, 261 in control group). Six studies included healthy subjects and two studies patients with AD (Table 1). No studies with patients with subjective memory complaints or MCI, which were eligible for inclusion, were identified. Study samples were small to moderate in size (34 to 120). Most studies reported a mean age around 70 years (Range of means in intervention group: Healthy subjects: 65.5 to 72.0) but one study reported a mean age of 25.0. One study included only females and one only male participants. Four studies reported mean MMSE scores (Range of means in intervention group: healthy subjects: 28.6 (±1.5) –29.0 (±1.2); AD: 25.5 (±2.3) –25.8 (±3.3)). Participants were generally reported to be sedentary prior to the studies. Further inclusion and exclusion criteria varied across studies but generally included no medical conditions, which prohibited exercise (all studies), no other neurological or psychiatric conditions (all studies) and MMSE score above a certain score (three studies). Recruitment setting and process were in general poorly described, with most studies reporting community setting, and memory clinics for two studies (Table 2).

Table 1

Characteristics of studies included in the systematic review

Healthy subjects
Reference	Number of subjects	Age, mean (SD)	Percent female	Baseline MMSE, mean (SD)
	(intervention/control)	(intervention/control)	(intervention/control)	(intervention/control)
Best et al., [43]	26/32/25*	69.4 (±3.0)/ 69.5 (±2.7)/	100/100/100*	28.6 (±1.5)/ 28.5 (±1.3)/
		70 (±3.3)*		28.8 (±1.2)^*
Colcombe et al., [45]	30/29	65.5/66.9^§	53/57	29 (±1.2)/ 29.4 (±1.4)
Erickson et al., [49]	60/60	67.6 (±5.8)/ 65.5 (±5.4)	73/60	Not reported
Niemann et al., [46]	19/17/13**	68.2 (2.61)/ 69.6 (5.1)	71/68/54	Not reported
		/68.8 (2.6)
Tamura et al., [47]	75/35	72.0 (±4.1)/72.0 (±4.7)	61/40	28.9 (±1.3)/ 28.0 (±2.0)
Wagner et al., [48]	17/17	25.0 (±3.3) /23.7 (±1.7)	0/0	Not reported
AD patients
Jensen et al., [44]	26/27	68.1 (±6.8)/ 69.2 (±3.9)	67/33	25.5 (±2.3)/ 25.1 (±3.9)
Morris et al., [50]	39/37	74.4 (±6.7)/ 71.4 (±8.4)	43/57	25.8 (±3.3/ 25 (±3.2)

*The study included three groups. The intervention groups consisted of 2 or 1 weekly session of resistance training. The latter number indicates data for the balance and toning group. **The study included three groups. The intervention group consisted of Nordic walking whereas the 2nd number indicates number of subjects in the group which engaged in resistance training and the latter had no intervention. ^§SD not reported.

Table 2

Study design and methodology of studies included in the systematic review

Studies with healthy subjects
Reference	Study design	Inclusion and exclusion criteria¹	Recruitment and population	Randomization	Study length,number of sessions, duration of sessions	Description of intervention	Description ofcontrol situation
Best et al., [43]	Single-blinded RCT	Female, intact cognition, no medical condition contraindicating physical exercise, no neurological disease present, participation in resistance training in the past 6 months, not receiving estrogen or testosterone therapy.	Community dwelling, elderly healthy	Not reported	12 months	Supervised, group-based. Resistance, free weights, Keiser pressurized air system and non-machine exercises.	Resistance training once weekly, balance and toning twice weekly.
Colcombe et al., [45]	Single-blinded RCT	Cognitively intact, no history of neurological diseases, obtain consent from their personal physician, sedentary	Community dwelling elderly healthy	Not reported	6 months, 3 sessions/week, 60 min/session	Individual aerobic exercise program. Intensity: Incremental from 40–50% to 60–70% of PHR.	Stretching and toning
Erickson et al., [49]	Single-blinded RCT	Age 55–80 y, cognitively intact, no history of neu- rological diseases or major vasculature problems, obtain consent from their personal physician, sedentary	Community dwelling elderly healthy	Not reported	12 months. 3 sessions/week, 50 min/session	Supervised, group-based. Walking. Intensity: Incremental from 50–60% to 60–75% of PHR.	Stretching, balance and yoga.
Niemann et al., [46]	RCT²	No history of cardiovascular disease, neurological disorder, or motor or cognitive restriction, MMSE≥27.	Community dwelling elderly healthy	According to place of residence	12 months, 3 sessions/week, 45–60 min/session	Supervised, group-based. Nordic walking. Intensity: Individual in a zone at and above AGE thresholds	Coordination, stretching and relaxation
Tamura et al., [47]	RCT²	No history of psychiatric disease, stroke, or grade 3 on the Fazekas scale. MMSE≥24, age≥65	Community dwelling elderly healthy	Participant could decide which group to participate in	2 y. Calisthenics: 3 times daily for 10 min, Aerobic 1 h/month	Japanese calisthenics, individual; Aerobic exercise, group-based	Not reported
Wagner et al., [48]	RCT²	Male, no present or past history of psychiatric, neurologic, or other clinically significant disorders.	Community dwelling university students	Not reported	6 weeks, 3 sessions/week, 60 min/session	Supervised aerobic exercise program. Intensity: Between individuals aerobic and anaerobic thresholds	Training as usual
Studies with AD patients
Jensen et al., [44]	Multicenter, single-blinded RCT	Mild AD (NINCDS-ADRDA criteria), MMSE > 19, age 50–90, no other neurologic or psychiatric disease, no medical condition which prevents participating in exercise, sedentary, on stable or no anti-dementia medication	Community-dwelling patients with AD recruited from memory clinics	Central randomization procedure using computer generated random sequences	16 weeks, 3 sessions/week, 60 min/session	Supervised aerobic exercise in groups on exercise machines. 3 sessions per week, 60 mins. per session. Intensity: 70–80% of PHR.	Usual care
Morris et al., [50]	Single-blinded RCT	MCI/Mild AD (NIA-AA criteria) CDR≤1, afe > 54, no other neurologic or psychiatric disease, no medical condition which prevents participating in exercise, sedentary	Community-dwelling patients with MCI/AD recruited from memory clinics and community	Block randomized (stratified by age and gender)	26 weeks, 3–5 sessions, 60–150 mins per week, incremental.	Supervised aerobic exercise. Target intensity: 60–75%	Strength exercises, modified Tai-Chi and yoga

RCT, randomized controlled trial; PHR, peak heart rate; AGE, aerobic gas exchange. Single-blinded studies indicates studies in which assessors were blinded to group allocation, but not participants. ¹Major inclusion and exclusion criteria reported. ²Blinding of assessors to group allocation not reported.

Intervention and control situation varied across studies. In one study, the intervention was a stretching/weights program, whereas the rest were aerobic exercise programs. Both group and individual training programs were used and most were supervised. Study period ranged from 6 weeks to 2 years (3 studies were 12 months), whereas frequency of training sessions was 3 times per week for most studies. Exercise intensity varied from low to high. Control situations included training as usual, balance training, coordination training, yoga and relaxations. Five studies measured aerobic fitness (VO2 max), two studies did not report on any measures of physical fitness, and one study reported on muscular strength. Adherence to the exercise program was reported in 3 studies, and 6 studies reported drop-out rates (0% in 5 studies, 10.5% in 1 study).

Methodology of included studies

All included studies were randomized studies. Five studies were single-blinded studies, where outcome assessors were blinded to group allocation (Table 2).

Outcome measures in 7 of the studies (6 in healthy subjects, 1 in AD) were change in brain volumes, including hippocampal volume, on MRI. Analysis of MRI included manual tracing of the hippocampus (1 study), whole-brain voxel-based morphometry (VBM) (two study), semi-automated quantification of hippocampal volume (3 studies), and 1 study performed both VBM and semi-automated quantification of brain region volumes. One study, which used CSF biomarkers as outcome measure, was identified and could be included. This study included subjects with AD (Table 3). We did not identify any studies, which were eligible for inclusion with the remaining biomarkers.

Table 3

Results and outcome measures of studies included in the systematic review

Studies with healthy subjects
		Control group				Intervention group
Reference	Outcome measure and method or assessment of outcome measure	Outcome, Baseline Mean (SD) cm^3a	Outcome, Follow-up Mean (SD) cm^3a	Outcome, Change from baseline % change	Outcome, Baseline Mean (SD) cm^3a	Outcome, Follow-up Mean (SD) mm³ &	Outcome, Change from baseline % change	Statistical analysis (between groups)
Best et al., [43]	Hippocampal volume. Siemens 3 Tesla MR system. Semi-automatic using freesurfer software.	BAT: R hippo 3.67 (0.29), L hippo: 3.6 (0.24)	BAT: R hippo 3.72 (0.31), L hippo 3.58 (0.24)	Not reported	RTx 2: R hippo 4.07 (0.43), L hippo 3.87 (0.47); RTx1 R hippo 3.84 (0.32), L hippo 3.80 (0.44)	RTx 2: R hippo 4.08 (0.35), L hippo 3.86 (0.44); RTx1 R Hippo 3.9 (0.33), L Hippo 3.82 (0.38)	Not reported	No significant between-group effects
Colcombe et al., [45]	VBM analysis of hippocampus. Siemens 3 Tesla and 1.5 GE MR system. Semi-automatic VBM analysis.	NA ^b	NA ^b	NA ^b	NA ^b	NA ^b	NA ^b	No significant between-group effects
Erickson et al., [49]	Hippocampal volume. Siemens 3 Tesla. Semi-automatic analysis using FMRIB Integrated Registration and Segmentation Tool	L hippo: 4.90 (0.80), R Hippo 4.92 (0.80)	L hippo 4.90 (0.80) R Hippo 4.92 (0.80)	L hippo 1.40 % R Hippo 1.43 % decline	L hippo 4.89 (0.74) R Hippo 5.00 (0.67)	L hippo 4.98 (0.69) R hippo 5.09 (0.63)	L hippo 2.12% increase, R hippo 1.97% increase	Time×Group interaction for L and R hippocampus (p < 0.001)
Niemann et al., [46]	Hippocampal volume. Allegra 3 T MR system. Manual delineation of hippocampal volume using in-house software.	Coordination: Total hippo 4.23 (0.40) L hippo 2.12 (0.22) R hippo 2.11 (0.22) Stretching: Total hippo 4.17 (0.48) L hippo 2.11 (0.27) R hippo 2.06 (0.27)	Coordination: Total hippo 4.34 (0.44) Left 2.16 (0.24) R Hippo 2.19 (0.24) Stretching: T hippo 4.24 (0.53) L hippo 2.15 (0.30) R hippo 2.10 (0.27)	Coordination: Total hippo 2.84 % Stretching Total hippo 1.67%	Total hippo 4.22 (0.41) L hippo 2.13 (0.22) R hippo 2.09 (0.21)	Total hippo 4.37 (0.45) L hippo 2.21 (0.24) R hippo 2.16 (0.25)	Total hippocampal volume increased by 3.6%	No significant between-group effects
Tamura et al., [47]	VBM analysis of hippocampus. Siemens 1.5 T MR system. Semi-automatic VBM analysis using SPM8	NA ^b	NA ^b	NA ^b	NA ^b	NA ^b	NA ^b	No significant between-group effects
Wagner et al., [48]	Hippocampal volume. Siemens 3 Tesla MR system. Semi-automatic VBM analysis (SPM8) and volumetric analysis using freesurfer software.	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported for total hippocampus. Right CA2/3 decreased by 2.2%, subiculum by 1.3%, and DG/CA4 by 1.9%. ^c	Significant between-group effects reported. Right CA2/3 (p = 0.01), subiculum, (P = 0.01), DG/CA4 (p = 0.005).
AD subjects
Jensen et al., [44]	CSF Aβ₄₂: V-plex Peptide Panel (Meso Scale Discovery system); t-tau AND tau phosphorylated at amino acid 181 (p-tau): ELISA assay (INNOTES)	Aβ₄₂: 145 (±69) T-tau: 693 (±373)P-tau: 72 (±30)	Aβ₄₂: 194 (±84) T-tau: 747 (±639)P-tau: 83 (±35)	Not reported	Aβ₄₂: 175 (±107) T-tau: 722 (±386)P-tau: 83 (±31)	Aβ₄₂: 200 (±121) T-tau: 868 (±664)P-tau: 75 (±32)	Not reported	No significant between-group effects
Morris et al., [50]	Hippocampal volume. Siemens 3 Tesla MR system. Semi-automatic volumetric segmentation using FreeSurfer	Bilateral hippocampus: 6.2 (±1.3)	Not reported	Bilateral hippocampus: 1.6%	Bilateral hippocampus 6.1 (±1.3)	Not reported	Bilateral hippocampus 0.8%	No significant between-group effects

VBM, voxel-based morphometry; ELISA, enzyme-linked immunosorbent assay; BAT, balance and training; RTx1, resistance training once weekly; RTx2, resistance training twice weekly. ^aResults are pg/mL for Jensen et al., [44]. ^bNot applicable, since MRIs were analyzed using VBM analysis. ^cVBM analysis confirmed findings in hippocampal subregions.

Methodological quality of included studies

Methodological quality and risk of bias was assessed using the Cochrane Collaboration’s tool for assessing risk of bias [41]. For most studies, insufficient reporting hampered assessment of bias (Supplementary Table 2). However, two studies were identified as having a high risk of bias, specifically regarding sequence generation and allocation concealment. One study was rated as having low risk of bias for all items.

Results of studies in healthy subjects

Baseline hippocampal volume ranged from 2.06 to 4.92 cm³ in the control groups and 2.09 to 5.0 cm³ in the intervention group (reported for three studies) (Table 3). Change from baseline (hippocampal volume) was reported for 3 studies and ranged from a decline of 1.43% to an increase of 2.84% in the control group. In the intervention group, changes ranged from a decline of 1.3% (reported for subiculum) to an increase of 3.6%.

Whereas 4 studies found no effect, two studies reported significant between-group effects in change from baseline in hippocampal volume. Erickson et al. reported an increase in volume for both hippocampi in the intervention group (left hippocampus: 2.12% increase; right hippocampus: 1.97%) and a decline in the control group (left hippocampus: 1.40%; right hippocampus 1.43%), indicating an effect of the intervention [49]. Wagner et al. reported a decline in volume in the intervention group (right CA2/3 decreased by 2.2%, subiculum by 1.3%, and Dentate gyrus/CA4 by 1.9%) relative to the control group, indicating a detrimental effect of the intervention on hippocampal volume [48].

Results of studies in AD patients

In the study which investigated CSF biomarkers in AD patients, baseline and follow-up values for CSF biomarkers were reported for the study sample (Table 3) but change from baseline to follow-up values were not [44]. There were no significant between-group differences in change from baseline to follow-up, indicating no effect of the intervention on Aβ, t-tau, and p-tau levels in the CSF.

Morris et al. [50] reported both left and right hippocampi as one volume and reported a change in the intervention group of –0.8% and –1.6% in the control group. This change was not significant.

DISCUSSION

We carried out a systematic review of physical exercise intervention studies examining the effects of such interventions on AD biomarkers. We identified a total of 8 studies, comprising 6 studies in healthy subjects with hippocampal volume as outcome and two studies in AD patients with CSF AD biomarkers and hippocampal volume, respectively, as outcome. To our knowledge this is the first systematic review focusing on AD biomarkers and physical exercise. Two of the identified studies found a significant effect of physical exercise on hippocampal volume, with one study finding an increase in hippocampal volume, and one a detrimentaleffect.

Physical exercise may provide an important supplement to pharmacological treatment of AD, and a large number of studies in animal models of AD and aging support a possible effect on cognition and brain pathology [3 , 19]. In recent years, several intervention studies in elderly, MCI and AD populations have been carried out. Results have been inconsistent with regards to effect on cognition and behavioral symptoms [9 , 51–55]. Furthermore, quality of studies has been varying with some studies having relatively small sample sizes, lack of definite diagnosis and low methodological quality [56]. Moreover, variance in study design has hampered comparisons across studies [29]. Similarly, the studies identified in the present study were also heterogeneous in methodology and design. Studies varied in length from 6 weeks to 2 years, and type of exercise also varied both regarding type (resistance versus aerobic exercise, use of exercise machines, Nordic walking) as well as intensity (low to high). The identified studies also varied with regards to the effects found on hippocampal volume as evidenced by one study finding a positive effect [49], one a detrimental effect [48] and 5 studies found no effect [43, 45–47 , 50]. Regarding the study, which identified a positive effect, the intervention period was 12 months with 3 weekly sessions of walking. Interestingly, Niemann et al. used an almost identical exercise program, but failed to find an effect [46]. Both studies did find an increase in hippocampal volume (1.97% versus 3.6%, highest in Niemann et al.) in the intervention group, but where Erickson et al. found a decline in hippocampal volume in the control group, Niemann et al. conversely found an increase in hippocampal volume in both control groups. It may be speculated that the different findings may be due to the control situation selected, although both studies used stretching/coordination-type controls. Supporting the importance of selecting an appropriate control situation is the finding in the study of Niemann et al. of a numerical difference in the increase in hippocampal volume between the two different control groups, which were used. One other study also applied a study period of 12 months [43] and another study a period of 2 years [47], without finding a significant effect on hippocampal volume. This may indicate that increasing the length of the intervention does not lead to a benefit. However, it must be taken into account that the latter study used very low intensity exercise and a quasi-randomized design. It is surprising that a single study found a detrimental effect of exercise on hippocampal volume [48]. The study used a very short intervention period in a relatively young population, and found that the hippocampus had decreased in volume by up to 2%. Although results were not reported for the control group, a significant between-group difference was reported. The authors point to the relatively young age of participants as a factor, as well as the fact that all participants were male, and speculate whether there may be age-related differences in response of the hippocampus to exercise. Moreover, due to findings of relatively large between-subjects variations in volume changes, the authors further speculate that individuals harbor intrinsic factors, which mediate a difference in the effects of exercise on the hippocampus, which may underlie the results. Another possible mechanism underlying individual differences may be whether participants benefitted from the exercise in the sense of improved fitness. Wagner et al. [48] reported a correlation between change in workload and hippocampal volume in their study in the intervention group, which could be interpreted as lending support to such a mechanism, and should be explored further. From a neurobiological aspect, it is difficult to imagine a pathway by which exercise may lead to brain atrophy, but further studies are needed to confirm or refute a possible detrimental effect on the brain as well as the possibility of age-related differences in hippocampal response to exercise. In the meantime, the findings in this single study does not warrant advising against exercising, which is also underscored by the fact that the majority of studies showed no effects of the intervention on hippocampal volumes, and thus at least does not support a detrimental effect. Wagner et al. also looked at specific subregions of the hippocampus and found that the effects were in anterior regions [48]. In parallel, Erickson et al. found that effects on hippocampal volume were driven by a beneficial effect in the anterior part of the hippocampus [49]. Whether this is due to a topographically based susceptibility of the hippocampus to exercise should be explored in further studies. Lastly, non-significant results may also be a result of underpoweredstudies.

We identified two studies in patients with AD, one of which was also the only study in which effects of exercise on CSF biomarkers, specifically Aβ, t-tau, and p-tau, was investigated [44]. The study reported no effects of the intervention on the examined biomarkers. The length of intervention was 16 weeks, and despite the relatively high intensity exercise used in the study, this may be too short an intervention to affect AD pathology. It may also be that the dementia stage of AD is too late in a pathophysiological process which has most likely been ongoing for year prior [57], to intervene with the aim of a disease-modifying effect. In parallel, studies with pharmacological agents (e.g., vaccines aimed at Aβ [2]) have shifted to predementia stages, and a similar shift may be relevant when it comes to exercise. The second study in AD patients found no effect of 26 weeks of aerobic exercise on hippocampal volume, in line with most of the studies in healthy elderly subjects [50]. Interestingly, in line with the findings of Wagner et al. [48] a correlation between improvement in physical fitness and bilateral hippocampal volume was reported.

We chose to include studies in any age group (from 18 years and up) for two reasons. Firstly, studies in recent year, such as those carried out in asymptomatic carriers of autosomal dominant AD [57], have found that pathological changes in AD most likely begin many years before symptom onset and the dementia stage. Therefore, it is relevant to examine whether exercise may have an effect on AD pathology many years prior to the usual onset of symptoms and dementia. Secondly, the concept of cognitive reserve, which may explain individual differences in resilience to AD pathology, may, to some degree, be mediated by brain reserve such as a larger hippocampus [58]. This would have the theoretical implication that increasing the size of the hippocampus may translate to a higher resilience against AD pathology, and ultimately to progression to dementia. It should be highlighted here, that hippocampal atrophy, although often observed in AD and highly correlated to symptomatology of AD, in that is subserves memory, is not considered a specific biomarker of AD, as reflected for example, in the omission from the revised International working group-2 diagnostic criteria [35]. Nevertheless, for the reasons stated above, we chose to include it in the analysis, and is warranted as a marker of progressionin AD.

Attendance and adherence to the exercise program, including whether the intended intensity was achieved, are important parameters in assessment of interventions such as physical exercise. Although most of the identified studies used supervised exercise, little information was given regarding efforts to assess these parameters. Furthermore, we employed the Cochrane Collaboration’s tool for assessing risk of bias, and found that the methodological quality was difficult to judge for most studies due to insufficient reporting. Risk of bias was high for two studies [46, 47] regarding sequence generation and allocation concealment, whereas 4 studies [43–45 , 50] received low risk of bias scores for 1 to 6 of 6 items. This should be taken into account when interpreting the present findings.

We identified a large number of records in our searches. Despite carrying out preliminary searches in order to refine search strings, we may have failed to achieve a balance between sensitivity and specificity in our searches. However, this also reduces the risks of missing relevant studies. We applied a rigorous methodology to our searches and the systematic review, according to published standards, thereby reducing risk of bias. We did not carry out a quantitative synthesis of data as in a meta-analysis. Although such an approach is less susceptible to bias through interpretation of data from individual studies, exercise studies, especially so in dementia populations, tend to vary greatly in design and methodology. This may lead to synthesis of very heterogeneous data. For example, in the present study, we included data from studies with duration from 6 weeks to 2 years, exercise programs varied in intensity etc. and is it likely that effects are greatly dependent on such factors. This is further underlined by meta-analyses on exercise and cognition failing to reach conclusions due to the heterogeneity of studies [29, 55].

In conclusion, this systematic review did not uncover evidence to support an effect of physical exercise on AD biomarkers. However, the fact that we were able to identify a relatively small number of studies, of varying quality, also does not permit to conclude that such an effect does not exist, especially regarding biomarkers other that hippocampal volume. Future exercise studies are underway, such as the rrAD study (NCT02913664; at-risk healthy elderly subjects hippocampal volume), Fit-AD (NCT 01954550; AD patients, hippocampal volume), APEx study (NCT02000583; healthy elderly subjects, brain amyloid measured by amyloid PET), and BIM2 study (NCT03035851, at-risk healthy elderly subjects hippocampal volume) which may help to address these questions. Mechanisms of actions most likely involve several factors, but animal studies support a direct effect of exercise on AD pathology, and further studies exploring this avenue are needed in humans. The conclusions of this systematic review should be interpreted in light of other evidence indicating beneficial effects of exercise for patients with AD with regards to reducing symptoms of dementia and in elderly persons with regards to reducing the risk of AD. Hence, exercise should still be recommended as an add-on to pharmacological treatment of AD. Future studies should be conducted with a rigorous methodology in large populations and include both well-established biomarkers which reflects AD pathology as well as novel biomarkers.

DISCLOSURE STATEMENT

Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/17-0567r1).

Footnotes

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

References

Prince

, Comas-Herrera

, Knapp

, Guerchet

, Karagiannidou

(2016) World Alzheimer Report 2016 Improving healthcare for people living with dementia. Coverage, quality and costs now and in the future, Alzheimer’s Disease International, London, pp. 1–140.

Mangialasche

, Solomon

, Winblad

, Mecocci

, Kivipelto

(2010) Alzheimer’s disease: Clinical trials and drug development. Lancet Neurol 9, 702–716.

, Kang

, Leem

, Cho

, Yang

, Chae

, Hwang

, Cho

(2008) Exercise training acts as a therapeutic strategy for reduction of the pathogenic phonetypes for Alzheimer’s disease in an NSE/APPSw-transgenic model. Int J Mol Med 22, 529–539.

Adlard

, Perreau

, Pop

, Cotman

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

Yuede

, Zimmerman

, Dong

, Kling

, Bero

, Holtzman

, Timson

, Csernansky

(2009) Effects of voluntary and forcedexercise on plaque deposition, hippocampal volume, and behavior inthe Tg2576 mouse model of Alzheimer's disease. Neurobiol Dis 35, 426–432.

Herring

, Münster

, Metzdorf

, Bolczek

, Krüssel

, Krieter

, Yavuz

, Karim

, Roggendorf

, Stang

, Wang

, Hermann

, Teuber-Hanselmann

, Keyvani

(2016) Late running is not too late against Alzheimer’s pathology. Neurobiol Dis 94, 44–54.

Moore

, Girens

, Larson

, Jones

, Restivo

, Holtzman

, Cirrito

, Yuede

, Zimmerman

, Timson

(2016) A spectrum of exercise training reduces soluble Aβ in a dose-dependent manner in a mouse model of Alzheimer’s disease. Neurobiol Dis 85, 218–224.

Van Praag

(2008) Neurogenesis and exercise: Past and future directions. Neuromolecular Med 10, 128–140.

Hoffmann

, Sobol

, Frederiksen

, Beyer

, Vogel

, Vestergaard

, Brændgaard

, Gottrup

, Lolk

, Wermuth

, Jacobsen

, Laugesen

, Gergelyffy

, Hogh

, Bjerregaard

, Andersen

, Siersma

, Johannsen

, Cotman

, Waldemar

, Hasselbalch

(2015) Moderate-to-high intensity physical exercise in patientswith Alzheimer’s disease: A randomized controlled trial. JAlzheimers Dis 50, 443–453.

10.

Groot

, Hooghiemstra

, Raijmakers

, van Berckel

, Scheltens

, Scherder

, van der Flier

, Ossenkoppele

(2016) The effect of physical activity in cognitive function in patients with dementia: A meta-analysis of randomized controlled trials. Ageing Res Rev 25, 13–23.

11.

Paillard

, Rolland

, de Souto Barreto

(2015) Protective effects of physical exercise in Alzheimer’s disease and Parkinson’s disease: A narrative review. J Clin Neurol 11, 212–219.

12.

Smith

, Lancaster

, Nielson

, Woodard

, Seidenberg

, Durgerian

, Sakaie

, Rao

(2016) Interactive effects of physical activity and ApoE ɛ4 on white matter tract diffusivity in healthy elders. Neuroimage 131, 102–112.

13.

Coelho

FGDM

, 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.

14.

Xiong

, Li

, Sun

, Zhang

, Dong

, Zhong

, Sun

(2015) Long-term treadmill exercise improves spatial memory of male APPswe/PS1dE9 mice by regulation of BDNF expression and microglia activation. Biol Sport 32, 295–300.

15.

Nascimento

CMC

, Teixeira

CVL

, Gobbi

LTB

, Stella

(2012) A controlled clinical trial on the effects of exercise on neuropsychiatric disorders and instrumental activities in women with Alzheimer’s disease. Rev Bras Fisioter 16, 197–204.

16.

Ryan

, Kelly

(2016) Exercise as a pro-cognitive, pro-neurogenic and anti-inflammatory intervention in transgenic mouse models of Alzheimer’s disease. Ageing Res Rev 27, 77–92.

17.

Kang

, Cho

(2015) Effect of treadmill exercise on PI3K/AKT/mTOR, autophagy, and Tau hyperphosphorylation in the cerebral cortex of NSE/htau23 transgenic mice. J Exerc Nutr Biochem 19, 199–209.

18.

Intlekofer

, Cotman

(2013) Exercise counteracts declining hippocampal function in aging and Alzheimer’s disease. Neurobiol Dis 57, 47–55.

19.

Van Praag

, Christie

, Sejnowski

, Gage

, Van Praag

(1999) Running enhances neurogenesis, learning, and long-term potentiation in mice. Proc Natl Acad Sci U S A 96, 13427–13431.

20.

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.

21.

Okonkwo

, Schultz

, Oh

, Larson

, Edwards

, Cook

, Koscik

, Gallagher

, Dowling

, Carlsson

, Bendlin

, LaRue

, Rowley

, Christian

, Asthana

, Hermann

, Johnson

, Sager

(2014) Physical activity attenuates age-related biomarker alterations in preclinical AD. Neurology 83, 1753–1760.

22.

Merrill

, Siddarth

, Raji

, Emerson

, Rueda

, Ercoli

, Miller

, Lavretsky

, Harris

, Burggren

, Bookheimer

, Barrio

, Small

(2016) Modifiable risk factors and brain PET Measures of amyloid and tau in non-demented adults with memory complaints. Am J Geriatr Psychiatry 24, 1–9.

23.

Brown

, Sohrabi

, Taddei

, Gardener

, Rainey-Smith

, Peiffer

, Xiong

, Fagan

, Benzinger

, Buckles

, Erickson

, Clarnette

, Shah

, Masters

, Weiner

, Cairns

, Rossor

, Graff-Radford

, Salloway

, Vöglein

, Laske

, Noble

, Schofield

, Bateman

, Morris

, Martins

; Dominantly Inherited Alzheimer Network (2017) Habitual exercise levels are associated with cerebral amyloid load in presymptomatic autosomal dominant Alzheimer’s disease. Alzheimers Dement, ;doi: 10.1016/j.jalz.2017.03.008

24.

Brown

, Peiffer

, Taddei

, Lui

, Laws

, Gupta

, Taddei

, Ward

, Rodrigues

, Burnham

, Rainey-Smith

, Villemagne

, Bush

, Ellis

, Masters

, Ames

, Macaulay

, Szoeke

, Rowe

, Martins

(2012) Physical activity and amyloid-β plasma and brain levels: Results from the Australian Imaging, Biomarkers and Lifestyle Study of Ageing. Mol Psychiatry 18, 875–881.

25.

De Souto Barreto

, Andrieu

, Payoux

, Demougeot

, Rolland

, Vellas

(2015) Physical activity and amyloid-β brain levels in elderly adults with intact cognition and mild cognitive impairment. J Am Geriatr Soc 63, 1634–1639.

26.

Schultz

, Boots

, Almeida

, Oh

, Einerson

, Korcarz

, Edwards

, Koscik

, Dowling

, Gallagher

, Bendlin

, Christian

, Zetterberg

, Blennow

, Carlsson

, Asthana

, Hermann

, Sager

, Johnson

, Stein

, Okonkwo

(2015) Cardiorespiratory fitness attenuates the influence of amyloid on cognition. J Int Neuropsychol Soc 21, 841–850.

27.

Heyn

, Abreu

, Ottenbacher

(2004) The effects of exercise training on elderly persons with cognitive impairment and dementia: A meta-analysis. Arch Phys Med Rehabil 85, 1694–1704.

28.

Forbes

, Forbes

, Morgan

, Markle-Reid

, Wood

, Culum

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

29.

Forbes

, Forbes

, Blake

, Thiessen

, Forbes

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

30.

Aarsland

, Sardahaee

, Anderssen

, Ballard

; Alzheimer’s Society Systematic Review group (2010) Is physical activity a potential preventive factor for vascular dementia? A systematic review. Aging Ment Health 14, 386–395.

31.

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.

32.

Moher

, Liberati

, Tetzlaff

, Altman

(2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. J Clin Epidemiol 62, 1006–1012.

33.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment 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, 270–279.

34.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

, Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292.

35.

Dubois

, Feldman

, Jacova

, Hampel

, Molinuevo

, Blennow

, DeKosky

, Gauthier

, Selkoe

, Bateman

, Cappa

, Crutch

, Engelborghs

, Frisoni

, Fox

, Galasko

, Habert

M-O

, Jicha

, Nordberg

, Pasquier

, Rabinovici

, Robert

, Rowe

, Salloway

, Sarazin

, Epelbaum

, de Souza

, Vellas

, Visser

, Schneider

, Stern

, Scheltens

, Cummings

(2014) Advancingresearch diagnostic criteria for Alzheimer’s disease: The IWG-2criteria. Lancet Neurol 13, 614–629.

36.

Petersen

, Smith

, Waring

, Ivnik

, Tangalos

, Kokmen

(1999) Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 56, 303–308.

37.

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.

38.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

, 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 and the Alzheimer’s Association workgroup. Alzheimers Dement 7, 263–269.

39.

Cummings

, Dubois

, Molinuevo

, Scheltens

(2013) International Work Group criteria for the diagnosis of Alzheimer disease. Med Clin North Am 97, 363–368.

40.

World Health Organization. Health topics: Physical activity. http://www.who.int/topics/physical_activity/en/

41.

Higgins

JPT

, Altman

, Gøtzsche

, Jüni

, Moher

, Oxman

(2011) The Cochrane Collaboration ’ s tool for assessing risk of bias in randomised trials. BMJ 343, 1–9.

42.

Jüni

, Altman

, Egger

(2001) Systematic reviews in health care: Assessing the quality of controlled clinical trials. BMJ 323, 42–46.

43.

Best

, Chiu

, Liang Hsu

, Nagamatsu

, Liu-Ambrose

(2015) Long-term effects of resistance exercise training on cognition and brain volume in older women: Results from a randomized controlled trial. J Int Neuropsychol Soc 21, 745–756.

44.

Jensen

, Portelius

, Siersma

, Høgh

, Wermuth

, Blennow

, Zetterberg

, Waldemar

, Hasselbalch

, Simonsen

(2016) Cerebrospinal fluid amyloid beta and tau concentrations are not modulated by 16 weeks of moderate- to high-intensity physical exercise in patients with Alzheimer disease. Dement Geriatr Cogn Disord 42, 146–158.

45.

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 61A, 1166–1170.

46.

Niemann

, Godde

, Voelcker-Rehage

(2014) Not only cardiovascular, but also coordinative exercise increases hippocampal volume in older adults. Front Aging Neurosci 6, 1–24.

47.

Tamura

, Nemoto

, Kawaguchi

, Kato

, Arai

, Kakuma

, Mizukami

, Matsuda

, Soya

, Asada

(2015) Long-term mild-intensity exercise regimen preserves prefrontal cortical volume against aging. Int J Geriatr Psychiatry 30, 686–694.

48.

Wagner

, Herbsleb

, Cruz

, Schumann

, Brünner

, Schachtzabel

, Gussew

, Puta

, Smesny

, Gabriel

, Reichenbach

, Bär

K-J

(2015) Hippocampal structure, metabolism, and inflammatory response after a 6-week intense aerobic exercise in healthy young adults: A controlled trial. J Cereb Blood Flow Metab 35, 1570–1578.

49.

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.

50.

Morris

, Vidoni

, Johnson

, Van Sciver

, Mahnken

, Honea

, Wilkins

, Brooks

, Billinger

, Swerdlow

, Burns

(2017) Aerobic exercise for Alzheimer’s disease: A randomized controlled pilot trial. PLoS One 12, 1–14.

51.

Baker

, Frank

, Foster-Schubert

, Green

, Wilkinson

, McTiernan

, Plymate

, Fishel

, Watson

, Cholerton

, Duncan

, Mehta

, Craft

(2010) Effects of aerobic exercise on mild cognitive impairment: A controlled trial. Arch Neurol 67, 71–79.

52.

Baker

, Frank

, Foster-Schubert

, Green

, Wilkinson

, McTiernan

, Cholerton

, Plymate

, Fishel

, Watson

, Duncan

, Mehta

, Craft

(2010) Aerobic exercise improvescognition for older adults with glucose intolerance, a risk factorfor Alzheimer’s disease. J Alzheimers Dis 22, 569–579.

53.

Rolland

, Pillard

, Klapouszczak

, Reynish

, Thomas

, Andrieu

, Rivière

, Vellas

(2007) Exercise program for nursing home residents with Alzheimer’s disease: A 1-year randomized, controlled trial. J Am Geriatr Soc 55, 158–165.

54.

Eggermont

LHP

, Swaab

, Hol

, Scherder

(2009) Walking the line: A randomised trial on the effects of a short term walking programme on cognition in dementia. JNNP 80, 802–804.

55.

Öhman

, Savikko

, Strandberg

, Pitkälä

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

56.

Littbrand

, Stenvall

, Rosendahl

(2011) Applicability and effects of physical exercise on physical and cognitive functions and activities of daily living among people with dementia: A systematic review. Am J Phys Med Rehabil 90, 495–518.

57.

Bateman

, Xiong

, Benzinger

TLS

, Fagan

, Goate

, Fox

, Marcus

, Cairns

, Xie

, Blazey

, Holtzman

, Santacruz

, Buckles

, Oliver

, Moulder

, Aisen

, Ghetti

, Klunk

, McDade

, Martins

, Masters

, Mayeux

, Ringman

, Rossor

, Schofield

, Sperling

, Salloway

, Morris

(2012) Clinical and biomarker changes in dominantly inherited Alzheimer’s disease. N Engl J Med 367, 795–804.

58.

Stern

(2012) Cognitive reserve in ageing and Alzheimer’s disease. Lancet Neurol 11, 1006–1012.