Europe PMC

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

Total daily physical activity, brain pathologies, and parkinsonism in older adults.

Author information

Affiliations

  • 1. Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, Illinois, United States of America.
      Authors
    • Oveisgharan S 1
    • Dawe RJ 1
    • Leurgans SE 1
    • Yu L 1
    • Schneider JA 1
    • Bennett DA 1
    • Buchman AS 1
    (7 authors)

ORCIDs linked to this article

Plos one, 29 Apr 2020, 15(4):e0232404
https://doi.org/10.1371/journal.pone.0232404 PMID: 32348372 PMCID: PMC7190120

This article is in the Europe PMC Open access subset. Refer to the copyright information in the article for licensing details.
Free full text in Europe PMC

Abstract 

Objective

We examined the association of physical activity, postmortem brain pathologies, and parkinsonism proximate to death in older adults.

Methods

We studied the brains of 447 older decedents participating in a clinical-autopsy cohort study. We deployed a wrist worn activity monitor to record total daily physical activity during everyday living in the community-setting. Parkinsonism was assessed with 26 items of a modified motor portion of Unified Parkinson's Disease Rating Scale (UPDRS). We used linear regression models, controlling for age and sex, to examine the association of physical activity with parkinsonism with and without indices of Alzheimer's disease and related disorders (ADRD) pathologies. In separate models, we added interaction terms to examine if physical activity modified the associations of brain pathologies with parkinsonism.

Results

Mean age at death was 90.9 (SD, 6.2), mean severity of parkinsonism was 14.1 (SD, 9.2, Range 0-59.4), and 350 (77%) had evidence of more than one ADRD pathologies. Higher total daily physical activity was associated with less severe parkinsonism (Estimate, -0.315, S.E., 0.052, p<0.001). The association of more physical activity with less severe parkinsonism persisted after adding terms for ten brain pathologies (Estimate, -0.283, S.E., 0.052, p<0.001). The associations of brain pathologies with more severe parkinsonism did not vary with the level of physical activity.

Conclusion

The association of higher physical activity with less severe parkinsonism may be independent of the presence of ADRD brain pathologies. Further work is needed to identify mechanisms through which physical activity may maintain motor function in older adults.

Free full text 

Logo of plosoneLink to Publisher's site
PLoS One. 2020; 15(4): e0232404.
Published online 2020 Apr 29. https://doi.org/10.1371/journal.pone.0232404
PMCID: PMC7190120
PMID: 32348372

Total daily physical activity, brain pathologies, and parkinsonism in older adults

Shahram Oveisgharan, Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing,1,2,* Robert J. Dawe, Methodology, Writing – review & editing,1,3 Sue E. Leurgans, Formal analysis, Methodology, Writing – review & editing,1,2 Lei Yu, Formal analysis, Methodology, Writing – review & editing,1,2 Julie A. Schneider, Investigation, Methodology, Project administration, Writing – review & editing,1,2,4 David A. Bennett, Funding acquisition, Investigation, Methodology, Project administration, Resources, Writing – review & editing,1,2 and Aron S. Buchman, Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing1,2
Diane K. Ehlers, Editor
Author information Article notes Copyright and License information Disclaimer

Associated Data

Supplementary Materials
Data Availability Statement

Abstract

Objective

We examined the association of physical activity, postmortem brain pathologies, and parkinsonism proximate to death in older adults.

Methods

We studied the brains of 447 older decedents participating in a clinical-autopsy cohort study. We deployed a wrist worn activity monitor to record total daily physical activity during everyday living in the community-setting. Parkinsonism was assessed with 26 items of a modified motor portion of Unified Parkinson’s Disease Rating Scale (UPDRS). We used linear regression models, controlling for age and sex, to examine the association of physical activity with parkinsonism with and without indices of Alzheimer’s disease and related disorders (ADRD) pathologies. In separate models, we added interaction terms to examine if physical activity modified the associations of brain pathologies with parkinsonism.

Results

Mean age at death was 90.9 (SD, 6.2), mean severity of parkinsonism was 14.1 (SD, 9.2, Range 0–59.4), and 350 (77%) had evidence of more than one ADRD pathologies. Higher total daily physical activity was associated with less severe parkinsonism (Estimate, -0.315, S.E., 0.052, p<0.001). The association of more physical activity with less severe parkinsonism persisted after adding terms for ten brain pathologies (Estimate, -0.283, S.E., 0.052, p<0.001). The associations of brain pathologies with more severe parkinsonism did not vary with the level of physical activity.

Conclusion

The association of higher physical activity with less severe parkinsonism may be independent of the presence of ADRD brain pathologies. Further work is needed to identify mechanisms through which physical activity may maintain motor function in older adults.

Introduction

Parkinsonism, a complex aging phenotype, includes impaired gait and balance, bradykinesia, rigidity, and tremor and may affect 50% or more of adults 85 years or older[1]. The presence of clinical parkinsonism is associated with an increased risk of disabilities[2], mild cognitive impairment[3], and dementia[4]. Given the magnitude of the personal and social consequences of parkinsonism, modifiable risk factors like physical activity are being intensely studied for their potential efficacy to maintain or reduce the severity of parkinsonism in older adults[5]. The neurobiology underlying the potential efficacy of a higher level of physical activity to reduce the severity of parkinsonism is unknown.

In a previous study, we found that a higher level of total daily physical activity in older adults is associated with less severe parkinsonism [6]. In several other studies, we have also shown that postmortem indices of brain pathology such as macroinfarcts are related to more severe parkinsonism[7,8], and to less total daily physical activity[9]. Together, these prior studies show that both physical activity and brain pathologies are related to the severity of parkinsonism. As illustrated in Fig 1, these reports provide the scientific framework and support for testing the hypothesis that brain pathologies link (mediates) physical activity with parkinsonism in older adults. If we do not find evidence for mediation, this would suggest that physical activity and brain pathologies are independently associated with parkinsonism in older adults. Thus, testing this hypothesis would provide novel data about a potential mechanism underlying the motor benefits of physical activity in older adults. In turn this would provide novel targets for interventions to maintain motor function in old age.

An external file that holds a picture, illustration, etc. Object name is pone.0232404.g001.jpg
Scientific framework for Hypothesis Testing.

Our prior work has shown that a higher level of physical activity is related to less severe parkinsonism (A [6]), physical activity and brain pathologies are related (B [9]), and brain pathologies are related to the severity of parkinsonism (C [7,8,10]). Based on these findings, this study tested the hypothesis that brain pathology links (mediates) the association of physical activity with parkinsonism. If the hypothesized sequence of events shown in D is correct, then adding the term for brain pathology to the model would attenuate the association of physical activity with parkinsonism, which would no longer be significant. If brain pathology does not mediate this association, both physical activity and brain pathology would be independently associated with parkinsonism.

Prior investigators did not test this mediation hypothesis since it is uncommon to obtain postmortem indices from large numbers of well-characterized older adults prior to death. To test the hypothesis that brain pathologies link physical activity with the severity of parkinsonism, we used novel data from 447 deceased participants from the Rush Memory and Aging Project[11].

Methods

Participants

The Rush Memory and Aging Project (MAP) is an ongoing cohort study of community-dwelling healthy older adults 65 years older or older who agree to annual clinical assessments and organ donation at the time of death. The study was approved by an Institutional Review Board of Rush University Medical Center. Written informed consent was obtained from all study participants as was an Anatomical Gift Act for organ donation. Participants are recruited from retirement facilities, subsidized housings, and individual homes across Chicago Metropolitan area. While the study’s recruitment started in 1997, multi-day physical activity recordings started in 2005.

This study focuses on the associations between total daily physical activity metrics extracted from multiday activity recordings obtained in the community setting, brain indices of ADRD obtained at autopsy, and the severity of clinical parkinsonism based on the United Parkinson’s Disease Rating Scale (UPDRS). As such, inclusion was limited to decedents with completed brain autopsy with valid measures of quantitative physical activity metrics and clinical parkinsonism assessment proximate to death. Since 2005, 613 participants with physical activity metrics died and 537 of them underwent autopsy (88% autopsy rate). Of these, 519 (97%) had completed autopsy results at the time of these analyses, 72 had missing clinical data (e.g. parkinsonism) which left 447 for these analyses. Comparison of participants included in the current analyses with participants excluded (S1 Table) showed that the two groups were not different in age, education, vascular risk factors and diseases, and levels of physical activity, but were different in the percentage of women and neuroleptic medication consumption (S1 Table).

Assessment of parkinsonism

Trained nurse clinicians assessed parkinsonism annually using 26 items of a modified version of UPDRS[12,13]. This assessment has high inter-rater reliability and short-term stability among nurses and compared with a movement disorders specialist[12,13].

This testing assessed four common parkinsonian signs: parkinsonian gait, bradykinesia, rigidity, and tremor. Each of the 26 items was scored on a similar 0–4 scale, with 0 indicating no impairment and 4 indicating very severe impairment. We summarized scores for each of the four parkinsonian signs and the scores for these four signs were then averaged together to form a summary global parkinsonism score for the overall severity of parkinsonism as previously described. This measure was the primary outcome measure in our analyses[12,13].

Assessment of total daily physical activity

Each participant wore an activity monitor (Actical; Philips Healthcare, Andover, MA) over the non-dominant hand 24 hours/day for up to 10 days. The Actical is an omni-directional accelerometer which generates a signal proportional to the magnitude of all movements it registers. The device digitizes the signal and expresses the average activity for each recording epoch as activity counts (15s in this study; 5760 epochs/day). More physical activity is reflected in higher activity counts.

In prior work we constructed and validated two measures which summarize the mean total daily physical activity based on multiday recording[6,14]. In this study, we analyzed total daily activity for all days with complete 24 hours recordings. Total daily physical activity is the mean sum of all activity counts for up to 10 days of complete recordings. Since the benefits of physical activity may be related to the intensity of activity, we constructed a second measure. Intensity of daily physical activity was calculated by dividing the total daily activity counts by the total duration of non-zero 15 second epochs to yield a metric of the intensity of activity counts per hour of activity. These two metrics were validated in prior studies predicting adverse health outcomes and are independently associated with these outcomes when controlling for self-reported physical activity[6,14].

Movement during sleep and non-rest periods may make different contributions to parkinsonism. Sleep diaries are not available in this cohort and many participants nap extensively during the daytime making it difficult to filter out movements related to periods of sleep/rest. In prior studies, we developed and validated a sleep fragmentation metric kRA, from actigraphic recordings [15]. KRA captures the temporal organization of human rest-activity patterns in terms of transition probabilities between periods of rest (when recorded activity is zero) to recording epochs in which movement (non-zero activity) is recorded. Conceptually, kRA is a measure of the tendency to fragment sustained rest periods by physical activity. A higher value of kRA represents more fragmented rest, and a lower value represents a more consolidated rest. Since most rest periods are recorded during sleep, a higher kRA is associated with more frequent movement during sleep. KRA is weakly related to total daily physical activity (Spearman correlation coefficient = 0.15, p = 0.003), suggesting that they measure different constructs.

These continuous multiday recordings were obtained during routine living in the community-setting and captured all movements including both exercise and habitual physical activity. While these inform on total levels of physical activity, the particular device employed cannot be used to quantify the contributions of specific activities. Thus, these metrics quantify total daily physical activity which shows how active an individual’s lifestyle was during these multiday recordings.

Assessment of postmortem brain pathologies

Structured brain autopsies were performed by staff blinded to clinical data. Following brain removal, tissue preparation and sectioning, gross inspection, and tissue blocks were obtained from pre-specified brain regions as previously described[16].

Alzheimer’s disease pathology

A modified Bielschowsky silver stain was used to visualize diffuse plaques, neuritic plaques, and neurofibrillary tangles, the AD pathological indices, in the frontal, temporal, parietal, entorhinal and hippocampal cortical areas[16]. Each of the three AD pathological indices was counted in each of the brain regions. For each AD pathological index, we created a summary measure by counting the index in each brain region and standardizing the count followed by making an average of the standardized scores across the 5 brain regions. Then, we constructed the global AD pathology score by averaging the summary measures of the three AD pathologic indices. We previously showed that each regional standardized score, e.g. neuritic plaques in the frontal cortex, was correlated with the other standardized AD indices’ regional scores (r range: 0.47–0.73, median r = 0.67) and the Cronbach α coefficient was 0.90, indicating high internal consistency and supporting use of the global AD pathology score[17].

Lewy body pathology

Immunohistochemistry with α-synuclein immunostain (Zymed; 1:50) was used for assessment of Lewy bodies on sections from the frontal, temporal, parietal, anterior cingulate, entorhinal, hippocampus, basal ganglia, and midbrain[16]. In this study, we used a dichotomous summary variable indicative of the presence or absence of Lewy bodies pathology.

Nigral neuronal loss

Dissection of the diagnostic blocks included a hemisection of the midbrain containing subtantia nigra. Nigral neuronal loss was assessed in the substantia nigra at the level of the 3rd nerve exit using H&E staining. We used a semi-quantitative scale (none, mild to severe) for the assessment of nigral neuronal loss, as described previously[8].

Transactive response DNA-binding protein 43 (TDP-43)

TDP-43 was assessed using immunostaining with monoclonal antibody to phosphorylated TDP-43 (pS409/410; 1:100) in 6 brain regions: amygdala, hippocampus, dentate gyrus, entorhinal, frontal, and temporal cortices. Each of the 6 regions was assessed for the presence of the TDP-43 cytoplasmic inclusions in the glia or neurons, as described previously[16]. In this study, we used a dichotomous variable of the presence or absence of TDP-43 in the limbic or neocortical regions.

Hippocampal Sclerosis (HS)

HS was evaluated unilaterally in a coronal section of the midhippocampus at the level of lateral geniculate body, as described elsewhere[18]. It was graded as absent or present based on severe neuronal loss and gliosis in CA1 and/or subiculum.

Macroinfarcts

Uniformed inspection for cerebral infarcts was conducted by naked eyes on the fixed slabs from one hemisphere and pictures of fresh slabs from the other hemisphere, and found lesions were confirmed microscopically[19]. For this analysis, we only included chronic infarcts, as acute or subacute infarcts were unlikely to affect parkinsonism measured on average two years prior to death. We used a dichotomous variable summarizing the presence or absence of the macroinfarcts.

Microinfarcts

Microinfarcts are not visible to the naked eyes and can only be identified under microscopy, and have been shown to be associated with the parkinsonian signs[7]. A minimum of 9 regions in 1 hemisphere were examined for the microinfarcts: 6 cortical (frontal, temporal, entorhinal, hippocampal, parietal, and anterior cingulate), 2 subcortical (anterior basal ganglia and thalamus), and mid brain. Like macroinfarcts, only chronic microinfarcts were included for these analyses through a dichotomous variable (presence vs. absence of the microinfarcts).

Atherosclerosis

Atherosclerosis severity was assessed by evaluation of circle of Willis vessels at the base of the brain (vertebral, basilar, posterior, middle, and anterior cerebral arteries). Severity of atherosclerosis was scaled semi-quantitatively (none, mild to severe) on the basis of atherosclerosis severity in each artery and number of affected arteries, as described previously[16]. For this study, we constructed a dichotomous variable for the presence or absence of moderate to severe atherosclerosis.

Arteriolosclerosis

Arteriosclerosis severity was assessed by evaluation of the vessel walls in small arterioles of the anterior basal ganglia. It was based on the concentric hyaline thickening and narrowing of the examined vessels, and was scored semi-quantitatively (none, mild to severe, complete occlusion)[16]. For this study, we constructed a dichotomous variable for the presence or absence of moderate to severe arteriolosclerosis.

Cerebral Amyloid Angiopathy (CAA)

CAA was examined using amyloid-β immunostaining of meningeal and parenchymal vessels in four regions (frontal, temporal, angular, and calcarine cortices). CAA was scaled semi-quantitatively in each brain region (from 0 to 4) and CAA score was the average of the regional scores. From this continuous CAA score, we have also constructed a semi-quantitative (none, mild to severe) CAA scale[16]. For this analysis, we constructed a dichotomous variable for the presence or absence of moderate to severe CAA.

Other clinical covariates

Sex and years of education were recorded at study entry. Age was calculated from date of birth to date of physical activity assessment. Chronic health conditions included the sum of three self-reported vascular risk factors (hypertension, diabetes mellitus, and smoking) and sum of four self-reported vascular diseases (myocardial infarction, congestive heart failure, stroke, and claudication). Self-reported marital status was assessed and classified into five categories: never married, married, widowed, divorced, and separated. All clinical data used in these analyses were obtained from the same visit proximate to death.

As done in prior studies, a diagnosis of Parkinson disease (PD) was based on self-reported medical history of a clinical diagnosis of PD for which the participants had received levodopa or dopamine agonists[20]. At each visit, study personnel reviewed and recorded all medications including neuroleptics taken by the study participants.

Statistical analysis

The global parkinsonism scores were positively skewed, and these scores were square root transformed for these analyses [6,20]. As illustrated in Fig 1, we employed a series of multivariable linear regressions, adjusted for age at death and sex, to assess whether indices of brain pathologies link (mediate) the association of physical activity with the severity of parkinsonism proximate to death (Fig 1 and 1D). First, in a model without postmortem indices, we replicated the finding that physical activity was related to the severity of parkinsonism (Fig 1 and 1A). Second, we examined the association of total daily physical activity with ADRD pathology indices (Fig 1 and 1B). Third, we examined the association of postmortem indices with the severity of parkinsonism in a model without total daily physical activity (Fig 1 and 1C). Fourth, we examined a model that included both physical activity and brain pathologies together (Fig 1 and 1D). If the association of physical activity with parkinsonism is attenuated in model D, it would suggest that brain pathologies may link (mediate) the association of total daily physical activity with the severity of parkinsonism. In contrast, if the association of physical activity with the severity of parkinsonism does not change in the presence of brain pathology indices it will suggest that physical activity and brain pathologies are independently associated with the severity of parkinsonism. To examine potential collinearity among pathology indices, we calculated their variance inflation factor (VIF). VIF greater than 10 strongly suggests collinearity[21]. Since other health conditions and covariates may affect parkinsonism and physical activity, in further analyses we controlled for possible confounding effects of education, marital status, vascular risk factors and diseases, and use of neuroleptic medications. Finally, we repeated the models after exclusion of decedents with a clinical diagnosis of PD or decedents receiving neuroleptic medications that can affect the severity of parkinsonism.

Next, we repeated these analyses, but replaced total daily physical activity with the intensity of total daily physical activity, a second metric that captures aspects of daily physical activity that may be important for the benefit of physical activity. As the association of parkinsonism with sleep related movement may be different compared to movement during wake periods, we repeated the analyses by further adjustment for kRA [15], which is the probability of sleep disruption by movement.

Multiple mechanisms may account for the motor benefits of higher levels of physical activity. Prior studies have reported that life style factors are associated with less cognitive and motor impairment through modifying association of ADRD pathologies with cognitive and motor outcomes [22–25]. Therefore, in secondary analyses, we added interaction terms between physical activity and brain pathologies in model D to determine if the motor benefit of physical activity was due to modification of the association between brain pathologies and parkinsonism. The analyses were done using SAS version 9.4.

Results

Characteristics of study participants

Clinical Characteristics: There were 447 participants and their clinical and postmortem indices are summarized in Table 1. On average total daily physical activity was measured 2.1 (SD = 2.0) years prior to death, and the average length of recording was for 9.4 (SD = 1.8) days.

Table 1

Clinical and postmortem measures of the participants in these analyses.
CovariatesSummary measure
DemographicMean (SD) or N (%)
Age at death (years)90.9 (6.2)
Age at the actigraphic recording88.8 (6.2)
Female316 (71)
Years of education14.7 (2.9)
Marital status
Never married34(8)
Married97(23)
Widowed261(62)
Divorced26(6)
Separated0
Clinical
Sum of history of vascular risk factors1.3 (0.8)
Hypertension315 (70)
Diabetes Mellitus97 (22)
Smoking (ever in life)176 (40)
Sum of vascular diseases0.8 (0.9)
Stroke95 (22)
Heart attack87 (19)
Heart failure57 (15)
Lower extremities claudication121 (27)
Neuroleptic medication use86 (19)
Severity of parkinsonismMean (range)
Global parkinsonism score14.1 (0–59.4)
Parkinsonian gait score33.5 (0–100)
Bradykinesia score17.2 (0–80)
Rigidity score3.9 (0–60)
Tremor score3.1 (0–45.5)
Quantitative physical activity metricsMean (SD)
Total Daily physical activity (activity counts/day)1.44 × 105 (1.13× 105)
Intensity of physical activity (activity counts/active hours)0.18× 105 (0.09× 105)
KRA(an indirect measure of sleep time movement)0.028 (0.008)*
Postmortem IndicesN (%)
Neurodegenerative
NIA-Reagan AD pathological diagnosis309 (69)
Nigral neuronal loss (moderate or severe)49 (11)
Lewy bodies (present in one or more sites)125 (28)
Limbic or neocortical Transactive response DNA-binding protein-43 (TDP-4 3)159 (36)
Presence of hippocampal sclerosis47 (11)
Cerebrovascular pathologies
Macroinfarcts (1 or more present)175 (39)
Microinfarcts (1 or more present)144 (32)
Atherosclerosis (moderate or severe)116 (26)
Arteriolosclerosis (moderate or severe)134 (30)
Cerebral Amyloid Angiopathy (moderate or severe)149 (33)
Number of pathologies present3.1 (1.7)

*KRA was available for 378 of the participants.

Postmortem indices

The average postmortem interval was 9.0 (SD = 7.7) hours. For descriptive purposes we dichotomized the presence or absence of each of the 10 indices of brain pathologies as summarized in Table 1. The average participant showed evidence of 3 different pathologies (SD = 1.7, median = 3).

Total daily physical activity, brain pathologies, and parkinsonism

First, we replicated prior findings that a higher level of total daily physical activity proximate to death was associated with less severe parkinsonism (Fig 1A; Table 2, Model 1-TDPA). Second, we replicated prior findings that a higher level of physical activity was associated with lower odds of nigral neuronal loss and macroinfarcts (Fig 1B; S2 Table).

Table 2

Total daily physical activity, indices of brain pathologies and parkinsonism proximate to death*.
Model TermsModel 1-TDPAModel 2-PathModel 3-TDPA+PathModel 4-Intensity+Path
Est. (SE)Est. (SE)Variance inflation factorEst. (SE)Est. (SE)
p-valuep-valuep valuep-value
Total daily physical activity (TDPA)-0.315 (0.052)-0.283 (0.052)
<0.001<0.001
KRA
Intensity of total daily physical activity (intensity)-4.022 (0.669)
<0.001
AD pathology-0.157 (0.105)1.2-0.151 (0.102)-0.144 (0.101)
0.135
0.1380.153
Lewy body pathology0.057 (0.143)1.20.029 (0.138)0.006 (0.137)
0.689
0.8340.964
Nigral neuronal loss0.558 (0.203)1.20.472 (0.197)0.500 (0.195)
0.006
0.0170.011
TDP-43-0.124 (0.135)1.3-0.058 (0.131)-0.050 (0.130)
0.357
0.6590.701
Hippocampal sclerosis0.256 (0.206)1.20.210 (0.200)0.232 (0.198)
0.215
0.2940.243
Macroinfarcts0.233 (0.125)1.10.177 (0.122)0.196 (0.121)
0.063
0.1470.106
Microinfarcts0.025 (0.128)1.10.012 (0.124)-0.009 (0.123)
0.8430.9250.939
Arteriolosclerosis0.196 (0.133)1.10.195 (0.128)0.154 (0.128)
0.140
0.1300.229
Atherosclerosis0.376 (0.140)1.10.380 (0.135)0.388 (0.134)
0.007
0.0050.004
Cerebral Amyloid Angiopathy0.005 (0.129)1.10.034 (0.125)0.027 (0.124)
0.970
0.7840.829
Additional Explained Variance**7.2%5.8%11.5%12.7%

*All the models are linear regressions and included terms for age and sex which alone accounted for 2.5% of the variance of the outcome which is square root transformation of the global parkinsonism score. Cells’ parameters are estimates (SE, p value) derived from the linear regressions.

**In each model, an additional 2.5% of the variance is due age at death and sex.

Third, we found that a higher burden of nigral neuronal loss and atherosclerosis were associated with more severe parkinsonism proximate to death (Fig 1C; Table 2, Model 2-Path). The VIFs in this model were less than 1.5 indicating that collinearity did not exist among the examined pathology indices [21].

Fourth, we examined the association of total physical activity with the severity of parkinsonism in a model which also included terms for brain pathologies. In this model, the association of total daily physical activity with parkinsonism was not attenuated by the presence of brain pathologies (Fig 1D; Table 2, Model 3-TDPA+Path). In the Model 3-TDPA-Path model, total daily physical activity was independently associated with parkinsonism accounting for an additional 5.7% of the variance of parkinsonism as compared to 8.3% contributed by age, sex and brain pathologies.

In a series of sensitivity analyses, we examined the association of total daily physical activity with parkinsonism severity in models adjusting for education, marital status, vascular risk factors and diseases, and consumption of neuroleptic medications as possible confounding variables. Adjustment for these covariates did not affect the finding that higher levels of total daily physical activity was associated with less severe parkinsonism independent of brain pathology indices (S3 Table). Participants with a clinical diagnosis of PD (N = 17) or those taking neuroleptics (N = 86) might have more severe parkinsonism. After excluding these participants, our results were unchanged (S4 Table).

Other physical activity metrics, brain pathologies, and parkinsonism

In a series of similar models, we replaced total daily physical activity with the intensity of daily activity. Controlled for age and sex, more intense physical activity was associated with less severe parkinsonism (estimate = -4.406, SE 0.671, p < 0.001), and the association was unaffected by the presence of the terms of brain pathologies (Table 2, Model 4-Intensity+Path).

The association of parkinsonism with sleep related movement may be different compared to movement during wake periods. To quantify sleep-related movement, we used kRA [15], which is the probability of sleep disruption by movement. We examined if adding a term for kRA to model A confounded the association between total daily physical activity and parkinsonism. In this model, a higher level of total daily physical activity remained associated with less severe parkinsonism, while a higher level of kRA was associated with more severe parkinsonism (S5 Table). These finding suggest that a higher level of physical activity is associated with less severe parkinsonism and increased sleep fragmentation, due to more frequent movement during sleep, is associated with more severe parkinsonism. These associations persisted in the presence of brain pathologies (S5 Table).

Total daily physical activity and modification of the association between brain pathologies and parkinsonism

Physical activity might also be associated with less severe parkinsonism through a second mechanism by modifying the untoward contribution of brain pathologies to the severity of parkinsonism. Therefore, we examined if there was evidence to support this mechanism by addition of interaction terms for total daily physical activity and brain pathologies to Model D (Fig 1). The association of the pathology indices with the parkinsonism did not vary with the level of total daily physical activity (nigral neuronal loss × total daily physical activity [estimate = -0.323, SE 0.233, p = 0.166]; and atherosclerosis × total daily physical activity [estimate = -0.216, SE 0.114, p = 0.060]). These additional analyses suggest that the association of a higher level of physical activity with less severe parkinsonism was not due to physical activity modifying the association of ADRD pathologies with the severity of parkinsonism.

Discussion

Using novel data from almost 450 community-dwelling older participants, a higher level of total daily physical activity proximate to death was associated with less severe parkinsonism. Postmortem indices of ADRD pathologies did not mediate the association of physical activity with parkinsonism. Rather, we found that both physical activity and ADRD pathologies were independently associated with the level of parkinsonism proximate to death. These findings were robust and did not change when controlling for a wide range of clinical and health covariates. Further analyses did not show evidence for an interaction between physical activity and brain pathology in their association with the severity of parkinsonism, and hence, the interaction does not underlie the motor benefits of physical activity. Together, these findings highlight the need for further studies to elucidate the biology underlying the motor benefits of a more active lifestyle in older adults.

While the health benefits of higher levels of physical activity are well-known, the mechanisms underlying its potential benefits to reduce the severity of parkinsonism is unknown. Prior mouse studies suggest that higher levels of physical activity were associated with less amyloid and phosphorylated tau deposition, higher levels of markers of neurogenesis, and better spatial memory [26,27]. However, due to limited brain tissues available in well-characterized older adults, few studies have examined the association of physical activity and ADRD pathologies in older adults, and extant amyloid brain imaging studies have shown conflicting results [28].

Our prior studies have shown that a higher level of physical activity at study baseline in this cohort is associated with a slower progression of parkinsonism as well as a decreased incident of parkinsonism during an average of four years of follow-up. Additional studies have shown indices of brain pathology are associated with lower levels of total daily physical activity and more severe parkinsonism [9,10]. Together, these studies provided the scientific framework of the current study suggesting the possibility that brain pathologies may link (mediate) the association of higher levels of physical activity with less severe parkinsonism.

The current study replicated previous findings showing that a higher level of physical activity is associated with less severe parkinsonism [6]. While this association was significant, total daily physical activity explained only 7.2% of the variance of parkinsonism. This finding is similar with a previous study which reported that total daily physical activity was mildly correlated with a different motor phenotype which summarized several motor abilities, including gait speed, grip strength, and finger tapping and explained 9% of its variance [29].

Total daily physical activity and motor phenotypes may be weakly correlated as they assess different facets of motor function, a complex multidimensional phenotype. It has been suggested that total daily physical activity measures the quantity of volitional activity that occurs during the day. In contrast, other motor phenotypes may assess innate motor abilities. A person with poor motor abilities might nonetheless be very active and, conversely, an individual with good motor abilities might voluntarily elect not to move.

Similar to prior studies, in the current study brain pathologies were related to parkinsonism, but only accounted for a minority of its variance [30]. The accumulation of ADRD pathologies in motor regions outside the brain that were not measured in this study [31] may account for the small amount of the variance of parkinsonism explained by brain ADRD pathologies. In addition, while the amount of parkinsonism variance explained by the physical activity or ADRD pathologies is small at the individual level their effect at the public health level is not inconsequential. Given the extent of motor impairment in old age, even the modest effect sizes observed in the current study are likely to be very important.

The current analyses did not find evidence that ADRD pathologies link physical activity with the severity of parkinsonism. Rather, we found that both ADRD pathologies and physical activity were independently associated with parkinsonism. Recent studies have suggested that some clinical covariates may modify indices of brain pathology and mitigate its untoward clinical effects [22–25]. The current study did not find evidence that the motor benefits of physical activity are due the interaction and modification of indices of ADRD brain pathologies with the severity of parkinsonism in older adults.

The current results with parkinsonism are similar to a prior study, in this cohort, in which we found that a higher level of total daily physical activity in older adults was associated with better cognition and this association was also independent of the presence of brain ADRD pathologies [29]. Thus, both the potential cognitive and motor benefits of a higher level of physical activity in older adults were unrelated to the presence of indices of brain pathologies, and there was also no evidence that physical activity modify the association of brain pathologies with parkinsonism or cognition.

As discussed below, these cross-sectional findings derive from observational data and conclusive causal inferences require further interventional studies. Yet, the current study’s findings together with prior longitudinal studies which show that higher levels of baseline physical activity are associated with slower progressive parkinsonism[6] and cognitive decline[14] may lend support for interim public health strategies until more conclusive data is available. Currently, there are no treatments for AD and other degenerative brain pathologies. Findings that the potential clinical benefits of physical activity are unrelated to indices of brain pathology lend support public health efforts to increase the level of physical activity as a means to maintain both physical and cognitive function in older adults, even in the absence of efficacious treatments for ADRD pathologies.

If the results of the current study are replicated what potential mechanisms might underlie the motor benefits of physical activity? Most prior studies about the beneficial effects of life style factors have examined their association with cognition, not motor function. Like physical activity, varied factors such as education[32,33], late-life cognitive[34] and physical activities[29], and psychological factors like purpose in life[23] and social networks[35] were associated with slower cognitive decline independent of AD pathology. These data like the current data highlight that physical activity and other lifestyle factors may be linked with cognition via unidentified mechanisms which lack a known “pathologic footprint”. Results of some human brain imaging studies suggest that increased physical activity may lead to better memory through vascular and structural brain changes [36]. Recently, system biology and network-based approaches have been exploited to leverage transcriptome data to identify genes and proteins i.e. molecular brain mechanisms that may drive cognition [37]. For example, higher levels of inositol 1, 3, 4-triphosphate 5/6-kinase (ITPK1) was associated with slower cognitive decline, but its association with cognition was independent of the presence of ADRD pathologies [38]. A similar approach may be useful to elucidate novel molecular mechanisms that underlie the potential motor and cognitive benefits of a more active lifestyle in older adults.

The biology underlying the association of life style factors with motor function is unclear. Two prior studies have suggested that education and physical activity modify the association of white matter hyperintensities with motor function [24,39]. These studies examined brain imaging indices in living adults, but did not examine indices of ADRD pathologies. While the current study found that physical activity did not modify the association of ADRD with motor function, this study did not examine brain imaging indices. Therefore, further studies will need to examine brain imaging together with ADRD pathologies indices. In examining the underlying mechanisms of the association between life style factors and motor function, it is also important to consider the widespread distribution of motor-related CNS regions. In contrast to cognition, the motor system extends beyond the brain to influence spinal cord structures which via the peripheral nervous system regulate muscle structure, the final effector of all movement. Damage to any portion of this distributed motor system can impair motor function. Recent work has shown that indices of ADRD extend beyond the brain and brainstem to reach the spinal cord and are related to the severity of parkinsonism in older adults [31]. Further work is needed to extend the collection of indices of ADRD pathologies in postmortem studies to outside of the brain and also include brainstem and spinal cord motor structures. Finally, no prior work has examined the role of motor unit structures (spinal motor neuron, peripheral nerve and muscle) in linking ADRD phenotypes with parkinsonism in older adults [40]. These motor tissues will also be need to be examined to identify genes and proteins that may underlie the motor benefits of physical activity without manifesting a pathologic footprint.

Another question raised by the current study is the extent to which the association of higher levels of physical activity with better cognition and motor function derive from similar or distinct central nervous system (CNS) loci or mechanisms. For example recent work suggests that physical activity may cause muscle to release a hormone-like factor, irisin, that may reach cognitive brain regions via systemic circulation to protect cognition[40,41]. Finally, recent work suggests that molecular brain mechanisms without a known “pathologic” footprint may drive both cognitive and motor function[38,42].

Our study has several strengths. We analyzed quantitative metrics of physical activity derived from multiday continuous recordings obtained during everyday living in the community-setting. Structured and validated methods were employed to assess parkinsonism and brain autopsies. The combination of antemortem measures of physical activity and parkinsonism with neuropathological indices only available postmortem is especially challenging from a logistics perspective and adds to the strength of this work.

Our study has important limitations. The study was composed of highly educated Caucasians limiting generalizability of the study findings. However, the frequency of ADRD pathologies is this cohort is similar to other clinical-pathological longitudinal studies of aging (S6 Table) [43]. Another limitation is the cross sectional design of this study limiting causal inferences, as we cannot rule out reverse causality i.e., that reduced parkinsonism might lead to reduced physical activity. However, a prior longitudinal study in this same cohort showed that higher baseline total daily physical activity was associated with slower rate of progressive parkinsonism over time lending support for the potential beneficial effect of physical activity on the severity of parkinsonism [6]. Future longitudinal physical activity interventional studies employing repeated imaging of brain pathologies together with clinical data might be needed and might identify links between physical activity and parkinsonism not observed in the current study.

Moreover, prior data about how active the individuals in the current study were over the course of their lifespan was not available. Therefore, it is unclear if the benefit of a more active lifestyle is due to lifestyle choices over many years or due to choices during this study. Brain imaging indices were not examined in these analyses. Since the motor system extends beyond the brain, further work is needed to determine if ADRD pathologies in motor tissues outside the brain might mediate the association of total daily physical activity with the severity of parkinsonism [41,44]. Finally, the activity monitors employed in this study do not differentiate between different physical activities, and further studies are needed to determine whether the specific types of movements associated with less severe parkinsonism. In addition, removal of the device employed in the current study cannot always be distinguished from periods of no activity.

Despite its limitations, accumulating more than 450 autopsies from well-characterized older adults with both quantitative multi-day activity metrics as well as assessments of parkinsonism proximate to death provides novel data which has potential to inform on the design of both clinical and interventional studies of physical activity in older adults. Further studies are needed to elucidate the knowledge gaps about the mechanisms underlying the benefits of physical activity. This is crucial to facilitate interventions that modify lifestyles as a means to maintain physical and cognitive function in our aging population.

Supporting information

S1 Table

Comparison of participants included and excluded from these analyses due to missing clinical data.

(DOCX)

S2 Table

Association of total daily physical activity proximate to death with postmortem brain pathology indices*.

(DOCX)

S3 Table

Association of total daily physical activity and indices of brain pathologies with parkinsonism proximate to death controlling for potential confounders*.

(DOCX)

S4 Table

Association of total daily physical activity and indices of brain pathologies with parkinsonism proximate to death after exclusion of participants with a clinical diagnosis of Parkinson Disease (PD) or those receiving neuroleptic medications.

(DOCX)

S5 Table

Association of total daily physical activity and indices of brain pathologies with parkinsonism proximate to death controlling for kRA, probability metric of sleep disruption due to movement)*.

(DOCX)

S6 Table

Distribution of indices of Alzheimer’s Disease and Related Disorders (ADRD) pathologies in the Memory and Aging Project compared to 2 other community-based longitudinal clinical-pathological studies of aging(1).

(DOCX)

Acknowledgments

We thank participants of the Rush Memory and Aging project. Also, we appreciate staff of the Rush Alzheimer’s Disease Center.

Funding Statement

This work was supported by National Institute of Health grants: A.S.B.: R01AG47679; R01AG056352; R01AG059732 D.A.B.: P30AG10161; R01AG017917; R01AG15819; R01AG043379 The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability

All data included in these analyses are available via the Rush Alzheimer’s Disease Center Research Resource Sharing Hub, which can be found at www.radc.rush.edu. It has descriptions of the studies and available data. Any qualified investigator can create an account and submit requests for deidentified data.

References

1. Louis ED, Luchsinger JA, Tang MX, Mayeux R. Parkinsonian signs in older people: prevalence and associations with smoking and coffee. Neurology. 2003. July 8;61(1):24–8. 10.1212/01.wnl.0000072330.07328.d6 [Abstract] [CrossRef] [Google Scholar]
2. Louis ED, Tang MX, Schupf N, Mayeux R. Functional correlates and prevalence of mild parkinsonian signs in a community population of older people. Arch Neurol. 2005. February;62(2):297–302. 10.1001/archneur.62.2.297 [Abstract] [CrossRef] [Google Scholar]
3. Louis ED, Schupf N, Manly J, Marder K, Tang MX, Mayeux R. Association between mild parkinsonian signs and mild cognitive impairment in a community. Neurology. 2005. April 12;64(7):1157–61. 10.1212/01.WNL.0000156157.97411.5E [Abstract] [CrossRef] [Google Scholar]
4. Louis ED, Tang MX, Schupf N. Mild parkinsonian signs are associated with increased risk of dementia in a prospective, population-based study of elders. Mov Disord. 2010. January 30;25(2):172–8. 10.1002/mds.22943 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
5. Intlekofer KA, Cotman CW. Exercise counteracts declining hippocampal function in aging and Alzheimer’s disease. Neurobiol Dis. 2013. September;57:47–55. 10.1016/j.nbd.2012.06.011 [Abstract] [CrossRef] [Google Scholar]
6. Oveisgharan S, Yu L, Dawe RJ, Bennett DA, Buchman AS. Total daily physical activity and the risk of parkinsonism in community-dwelling older adults. J Gerontol Biol Sci Med Sci. 2019; [Europe PMC free article] [Abstract] [Google Scholar]
7. Buchman AS, Leurgans SE, Nag S, Bennett DA, Schneider JA. Cerebrovascular disease pathology and parkinsonian signs in old age. Stroke. 2011. November;42(11):3183–9. 10.1161/STROKEAHA.111.623462 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
8. Buchman AS, Shulman JM, Nag S, Leurgans SE, Arnold SE, Morris MC, et al. Nigral pathology and parkinsonian signs in elders without Parkinson disease. Ann Neurol. 2012. February;71(2):258–66. 10.1002/ana.22588 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
9. Buchman AS, Dawe RJ, Yu L, Lim A, Wilson RS, Schneider JA, et al. Brain pathology is related to total daily physical activity in older adults. Neurology. 2018. May 22;90(21):e1911–9. 10.1212/WNL.0000000000005552 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
10. Buchman AS, Yu L, Wilson RS, Leurgans SE, Nag S, Shulman JM, et al. Progressive parkinsonism in older adults is related to the burden of mixed brain pathologies. Neurology. 2019. April 16;92(16):e1821–30. 10.1212/WNL.0000000000007315 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
11. Bennett DA, Buchman AS, Boyle PA, Barnes LL, Wilson RS, Schneider JA. Religious Orders Study and Rush Memory and Aging Project. J Alzheimers Dis. 2018;64(s1):S161–89. 10.3233/JAD-179939 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
12. Bennett DA, Shannon KM, Beckett LA, Goetz CG, Wilson RS. Metric properties of nurses’ ratings of parkinsonian signs with a modified Unified Parkinson’s Disease Rating Scale. Neurology. 1997. December;49(6):1580–7. 10.1212/wnl.49.6.1580 [Abstract] [CrossRef] [Google Scholar]
13. Bennett DA, Shannon KM, Beckett LA, Wilson RS. Dimensionality of parkinsonian signs in aging and Alzheimer’s disease. J Gerontol Biol Sci Med Sci. 1999. April;54(4):M191–6. [Abstract] [Google Scholar]
14. Buchman AS, Boyle PA, Yu L, Shah RC, Wilson RS, Bennett DA. Total daily physical activity and the risk of AD and cognitive decline in older adults. Neurology. 2012. April 24;78(17):1323–9. 10.1212/WNL.0b013e3182535d35 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
15. Lim ASP, Yu L, Costa MD, Buchman AS, Bennett DA, Leurgans SE, et al. Quantification of the Fragmentation of Rest-Activity Patterns in Elderly Individuals Using a State Transition Analysis. Sleep. 2011. November;34(11):1569–81. 10.5665/sleep.1400 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
16. Oveisgharan S, Arvanitakis Z, Yu L, Farfel J, Schneider JA, Bennett DA. Sex differences in Alzheimer’s disease and common neuropathologies of aging. Acta Neuropathol. 2018. December;136(6):887–900. 10.1007/s00401-018-1920-1 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
17. Bennett DA, Wilson RS, Schneider JA, Evans DA, Aggarwal NT, Arnold SE, et al. Apolipoprotein E epsilon4 allele, AD pathology, and the clinical expression of Alzheimer’s disease. Neurology. 2003. January 28;60(2):246–52. 10.1212/01.wnl.0000042478.08543.f7 [Abstract] [CrossRef] [Google Scholar]
18. Nag S, Yu L, Capuano AW, Wilson RS, Leurgans SE, Bennett DA, et al. Hippocampal sclerosis and TDP-43 pathology in aging and Alzheimer disease. Ann Neurol. 2015. June;77(6):942–52. 10.1002/ana.24388 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
19. Schneider JA, Wilson RS, Cochran EJ, Bienias JL, Arnold SE, Evans DA, et al. Relation of cerebral infarctions to dementia and cognitive function in older persons. Neurology. 2003. April 8;60(7):1082–8. 10.1212/01.wnl.0000055863.87435.b2 [Abstract] [CrossRef] [Google Scholar]
20. Buchman AS, Leurgans SE, Yu L, Wilson RS, Lim AS, James BD, et al. Incident parkinsonism in older adults without Parkinson disease. Neurology. 2016. September 6;87(10):1036–44. 10.1212/WNL.0000000000003059 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
21. Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography. 2013. January;36(1):27–46. [Google Scholar]
22. Bennett DA, Wilson RS, Schneider JA, Evans DA, Mendes de Leon CF, Arnold SE, et al. Education modifies the relation of AD pathology to level of cognitive function in older persons. Neurology. 2003. June 24;60(12):1909–15. 10.1212/01.wnl.0000069923.64550.9f [Abstract] [CrossRef] [Google Scholar]
23. Boyle PA, Buchman AS, Boyle PA, Yu L, Schneider JA, Bennett DA. Effect of Purpose in Life on the Relation Between Alzheimer Disease Pathologic Changes on Cognitive Function in Advanced Age. Arch Gen Psychiatry. 2012. May 1;69(5):499 10.1001/archgenpsychiatry.2011.1487 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
24. Fleischman DA, Yang J, Arfanakis K, Arvanitakis Z, Leurgans SE, Turner AD, et al. Physical activity, motor function, and white matter hyperintensity burden in healthy older adults. Neurology. 2015. March 31;84(13):1294–300. 10.1212/WNL.0000000000001417 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
25. Buchman AS, Yu L, Boyle PA, Schneider JA, De Jager PL, Bennett DA. Higher brain BDNF gene expression is associated with slower cognitive decline in older adults. Neurology. 2016. February 23;86(8):735–41. 10.1212/WNL.0000000000002387 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
26. Maliszewska-Cyna E, Xhima K, Aubert I. A Comparative Study Evaluating the Impact of Physical Exercise on Disease Progression in a Mouse Model of Alzheimer’s Disease. J Alzheimers Dis. 2016. May 6;53(1):243–57. 10.3233/JAD-150660 [Abstract] [CrossRef] [Google Scholar]
27. Tapia-Rojas C, Aranguiz F, Varela-Nallar L, Inestrosa NC. Voluntary Running Attenuates Memory Loss, Decreases Neuropathological Changes and Induces Neurogenesis in a Mouse Model of Alzheimer’s Disease. Brain Pathol. 2016. January;26(1):62–74. 10.1111/bpa.12255 [Abstract] [CrossRef] [Google Scholar]
28. de Souto Barreto P, Andrieu S, Rolland Y. Physical Activity and beta-Amyloid Brain Levels in Humans: A Systematic Review. J Prev Alzheimers Dis. 2015;2(1):56–63. 10.14283/jpad.2015.34 [Abstract] [CrossRef] [Google Scholar]
29. Buchman AS, Yu L, Wilson RS, Lim A, Dawe RJ, Gaiteri C, et al. Physical activity, common brain pathologies, and cognition in community-dwelling older adults. Neurology. 2019. February 19;92(8):e811–22. 10.1212/WNL.0000000000006954 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
30. Buchman AS, Yu L, Wilson RS, Boyle PA, Schneider JA, Bennett DA. Brain pathology contributes to simultaneous change in physical frailty and cognition in old age. J Gerontol Biol Sci Med Sci. 2014. December;69(12):1536–44. [Europe PMC free article] [Abstract] [Google Scholar]
31. Buchman AS, Leurgans SE, Nag S, VanderHorst V, Kapasi A, Schneider JA, et al. Spinal Arteriolosclerosis Is Common in Older Adults and Associated With Parkinsonism. Stroke. 2017. October;48(10):2792–8. 10.1161/STROKEAHA.117.017643 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
32. Del Ser T, Hachinski V, Merskey H, Munoz DG. An autopsy-verified study of the effect of education on degenerative dementia. Brain. 1999. December;122 (Pt 12):2309–19. [Abstract] [Google Scholar]
33. Farfel JM, Nitrini R, Suemoto CK, Grinberg LT, Ferretti RE, Leite RE, et al. Very low levels of education and cognitive reserve: a clinicopathologic study. Neurology. 2013. August 13;81(7):650–7. 10.1212/WNL.0b013e3182a08f1b [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
34. Wilson RS, Boyle PA, Yu L, Barnes LL, Schneider JA, Bennett DA. Life-span cognitive activity, neuropathologic burden, and cognitive aging. Neurology. 2013. July 23;81(4):314–21. 10.1212/WNL.0b013e31829c5e8a [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
35. Bennett DA, Schneider JA, Tang Y, Arnold SE, Wilson RS. The effect of social networks on the relation between Alzheimer’s disease pathology and level of cognitive function in old people: a longitudinal cohort study. Lancet Neurol. 2006. May;5(5):406–12. 10.1016/S1474-4422(06)70417-3 [Abstract] [CrossRef] [Google Scholar]
36. Maass A, Duzel S, Brigadski T, Goerke M, Becke A, Sobieray U, et al. Relationships of peripheral IGF-1, VEGF and BDNF levels to exercise-related changes in memory, hippocampal perfusion and volumes in older adults. Neuroimage. 2016. May 1;131:142–54. 10.1016/j.neuroimage.2015.10.084 [Abstract] [CrossRef] [Google Scholar]
37. Mostafavi S, Gaiteri C, Sullivan SE, White CC, Tasaki S, Xu J, et al. A molecular network of the aging human brain provides insights into the pathology and cognitive decline of Alzheimer’s disease. Nat Neurosci. 2018. June;21(6):811–9. 10.1038/s41593-018-0154-9 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
38. Yu L, Petyuk VA, Gaiteri C, Mostafavi S, Young-Pearse T, Shah RC, et al. Targeted brain proteomics uncover multiple pathways to Alzheimer’s dementia. Ann Neurol. 2018. July;84(1):78–88. 10.1002/ana.25266 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
39. Elbaz A, Vicente-Vytopilova P, Tavernier B, Sabia S, Dumurgier J, Mazoyer B, et al. Motor function in the elderly: Evidence for the reserve hypothesis. Neurology. 2013. July 30;81(5):417–26. 10.1212/WNL.0b013e31829d8761 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
40. Delezie J, Handschin C. Endocrine Crosstalk Between Skeletal Muscle and the Brain. Front Neurol. 2018;9:698 10.3389/fneur.2018.00698 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
41. Zhang J, Zhang W. Can irisin be a linker between physical activity and brain function? Biomol Concepts. 2016. August 1;7(4):253–8. 10.1515/bmc-2016-0012 [Abstract] [CrossRef] [Google Scholar]
42. Buchman AS, Yu L, Petyuk VA, Gaiteri C, Tasaki S, Blizinsky KD, et al. Cognition may link cortical IGFBP5 levels with motor function in older adults. Ginsberg SD, editor. PLOS ONE. 2019. August 12;14(8):e0220968. [Europe PMC free article] [Abstract] [Google Scholar]
43. Rahimi J, Kovacs GG. Prevalence of mixed pathologies in the aging brain. Alzheimers Res Ther. 2014. December;6(9):82 10.1186/s13195-014-0082-1 [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
44. Whitham M, Parker BL, Friedrichsen M, Hingst JR, Hjorth M, Hughes WE, et al. Extracellular Vesicles Provide a Means for Tissue Crosstalk during Exercise. Cell Metab. 2018. January 9;27(1):237–251 e4. 10.1016/j.cmet.2017.12.001 [Abstract] [CrossRef] [Google Scholar]
2020; 15(4): e0232404.
Published online 2020 Apr 29. 10.1371/journal.pone.0232404.r001

Decision Letter 0

Diane K. Ehlers, Academic Editor
Copyright and License information Disclaimer

19 Feb 2020

PONE-D-19-27901

Total daily physical activity, brain pathologies, and parkinsonism in older adults

PLOS ONE

Dear Dr. Ovelsgharan,

Thank you for submitting your manuscript to PLOS ONE. We apologize that review of this manuscript took longer than usual. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The reviewer comments are provided below, and the following are additional editor comments:

  • There are a number of concerns related to the statistical analyses that require careful revision in the re-submission:

    • The introduction is well-written, but does not support the methods or results sections as presented in the abstract or the manuscript. The introduction implies that a mediation analysis was conducted to test the mediating effect of brain pathologies on the relationship between PA and parkinsonism. However, the statistical analyses appear to test the association between PA and parkinsonism independent of brain pathologies and to test the interaction between PA and brain pathologies. These analyses do not reflect the research question as stated in the introduction section.

    • The number of analyses is unfocused and raises concerns whether hypotheses were developed a priori.

    • It appears that the four parkinsonism signs were tested in separate regression models. This is problematic, as the four subscales are likely correlated.

    • As Reviewer 1 noted, the collinearity between brain pathology variables is not addressed. 

    • Were the physical activity data also positively skewed?

  • While Figure 1 is pretty, it does not offer additional information not already available in Table 2.

  • Please provide further information on the choice of covariates - it seems that only age at death and sex were included in the models, but other covariates (e.g., education, marital status, medications) may also be relevant.

  • While the authors cite two previously published studies utilizing the physical activity data processing methods, it is not clear how sleep was accounted for in the data processing. The protocol utilizes 24-hour monitoring, and does not mention how sleep was filtered out. Movement during sleep time may represent disrupted sleep and should not be included in total activity counts. Please include a statement on how sleep time was handled in the data processing.

  • Please pay specific attention to Review 1's comments related to the discussion section of the paper. 

  • Review 1 and I thought the presentation of the scientific evidence utilizing a logical argument of ABC was sufficient and helpful; however, Review 2 found it unclear. Please consider including a figure to illustrate the proposed scientific framework/analyses for the research question.

  • Minor comments:

    • Please include ranges in parentheses for parkinsonism variables in Table 1.

We would appreciate receiving your revised manuscript by March 23, 2020. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Diane K. Ehlers, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Journal requirements met. Although Reviewer 2 felt that the data were not publicly available, the authors have adequately provided evidence of the availability of the data.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript leverages a rich autopsy dataset with associated clinical measures to test a mediation hypothesis that, brain pathology will mediate the relationship between physical activity (PA) and parkinsonism.

The introduction is clear, concise and well written. I especially appreciated the clear articulation of the mediation model and hypothesis. Thank you.

Some light editing is still required throughout for spelling and grammar.

I'm not sure I would say r of .4-.7 is "highly correlated". Might be better not to use an adjective here (p 7, line 134)

I think it would be important to note if there were any differences in PA (or other demographics) or other characteristics between those who had autopsy data, or UPDRS, and those who did not. I don't suspect there are but one never knows how selection bias might surface.

Help me understand the basis for your claim that the data support the idea that pathologies explain parkinsonism (Table 2, Model 2). 5.8% variance explained beyond age and sex seems a bit underwhelming (and there is no discussion of variance inflation amongst correlated neuropathologies, or presentation of the fully model, just the components). 7.8% variance explained by PA is a bit underwhelming too. However, I get that people are noisy.

I think my biggest concern in this manuscript is what seems to be the authors conclusion that the PAy measured can somehow be considered protective or "affording reserve". I know a veteran and established group such as this knows the inherent difficulty in identifying causality in regressions. Here one could easily interpret the data to mean that PAprotects against parkinsonism, or people with parkinsonism are less active. Thus the statement on p13,line 238 about physical activity affording reserve is ambitious at best. And the conclusion that PA in not interacting with brain pathology, when PA was measured so close to death is perhaps overly pessimistic for the field.

Again, on page 14, there is the hint of a conclusion about physical activity not affection neuropathology when the data provide no clear way to test this over time, or even with sufficient separation of activity data and neuropathology work up.

Given these concerns, the bulk of the discussion about reserve seems ambitious.

Reviewer #2: This manuscript sets out to describe the association of total daily physical activity and degree of parkinsonism in older adults near the end of life and to determine if that association is maintained when postmortem brain pathology indices are applied. The authors have previously established that higher levels of daily physical activity in older adults is associated with less severe parkinsonism. They have also previously shown that certain brain pathology indices related to more severe parkinsonism and in a separate publication that these brain pathology indices related to less total daily physical activities. The uniqueness of this current manuscript combines all three factors (totally daily physical activity, brain pathology, parkinsonism) in a single analysis in order to determine if higher level of physical activity relates to lower degree of parkinsonism and that this relationship is not diminished by the addition of brain pathology indices into the analysis. This analysis provides unique data to the field of exercise in the aging population and is in proper context of previous publications.

The analysis includes data from 447 subjects who participated in a community-based brain bank. The degree of parkinsonism is derived from a modified United Parkinson’s Disease Rating Scale (UPDRS) that has previously been verified. Total daily activity measures come from an accelerometer worn by participants which has also been previously verified. Brain pathology information obtained from blinded pathologist who evaluated for degree of Alzheimer’s disease and related disorders (ADRD) pathologies. A series of multivariable age and sex-adjusted linear regressions were used to assess the whether the association of daily physical activity with parkinsonism persisted when indices of brain pathology were included. The results of these analyses show higher levels of daily activity were associated with less severe parkinsonism; higher burden of brain pathology was associated with more severe parkinsonism; and that the association between increased physical activity resulting in less parkinsonism was not reduced by presence of brain pathology. The data they present do support the claims of the authors and the summary of current data in the field nicely demonstrates the importance of this current work. I do have the following minor recommendations to improve upon the clarity and quality of the publication, presented in the order they occurred in the manuscript:

#1: In the introduction (page 3; starting line 57) the terms “physical activity” “parkinsonism” and “brain pathology” are given A, C and B labeling but I found this confusing to follow while reading the text and did not aid in the understanding of the framework. Initially I thought this was going to relate into an equation but this labeling was not mentioned again and thus would recommend removing it.

#2: Page 4; Methods sections: I think it is pertinent to explicitly include if the cohort from the MAP is a “aging healthy control” database, or if these participants were recruited in the brain bank because they had cognitive decline or parkinsonism. This will help to provide further insight in how to interpret the data.

#3; Table 1 displays the clinical and postmortem characteristics of the cohort. It may be useful to add in this section that these characteristics (if indeed true) are similar to other reported clinical and brain bank findings in similarly aged subjects. This would serve to boost the generalizability of the data provided. For example – does the literature report similar findings of predominately (69%) AD pathologic findings in an aging cohort? Likewise is the degree of global parkinsonism similar to those also reported? Since this journal will attract more of a general audience; I think this will be helpful to put the results into perspective to those readers that are not as intricately aware of this data.

#4; Table 2: Would consider re-naming the Models with a more descriptive title to increase the clarity of what the linear regression models are representing.

#5: Table 2: in the final row: “Additional Explained Variance” are these values shown without the 2.5% addition of the variance due to age/sex as referenced in the caption? I am nearly certain that is what is trying to be related, but the language in the caption and Table 2 could be more clear. Perhaps, “an additional 2.5% of the variance is due to age/sex” or something similar could replace the “compared to age and sex” language.

#6; Page 13, Line 223: Regarding the statement (higher burden of brain pathology related to more severe parkinsonism) made about Model 2 – I am unclear about which data in Table 2 represent this statement. There does not appear to be a “global pathology burden” index to represent the overall burden of pathology; instead each individual pathology is listed and its individual relationship to parkinsonism. Additionally the linear regression models are not significant as P values are greater than 0.05 except for Nigral neuronal loss and Atherosclerosis. The association between these 2 pathologies and parkinsonism is mentioned in the text but I feel that the statement “higher burden of brain pathologies was associated with more severe parkinsonism” is unfounded because the bulk of the brain pathology did not show this same association.

#7: Page 13, line 229: Then sentence starting “In the latter model…” It is not clear which model you are referring to and again would be more clear if the models were renamed or at least directly referenced in this sentence. Is this referring to Model 2? If so, the variance mentioned in the text of 5.7% does not match the corresponding variance in Table 2 of 5.8%.

#8: Figure 1 section: The text starting on 233 and ending on 237: “The blue regression line….pathologies to the first model” seem more appropriate for the Figure 1 caption.

#9 Figure 1 section: Two points – 1) line 235 “A second red line..” are there supposed to be two red lines or should this be worded “The red line…” ? 2) The red regression line is displaying Model 3 from the description provided but the text references Models 1 and 2 at the start of this section. It is not clear to me which model this red line is representing- I think it is Model 3 but this should be more clearly stated.

#9: Figure 1: Overall this figure doesn’t really add to any new information except for a different way to represent the data from Table 2. I would recommend cutting this figure all together. If the goal is present a more “graphic” form of the information (which in my opinion is more digestible) – then the opposite approach could be considered and would suggest creating figures to represent the other models as well. If this approach was done, then Table 2 could be made available as a supplement table so as not to give redundant data in the body of the text.

#10: Page 15, line 267-268: The claim that higher level of daily activity is associated with lower odds of tremor does not appear to be true as the p value is 0.066

#11: Table 3: It is not clear what is meant by predictors “with TDPA” and “without TDPA.” It seems like these rows are representing data with and without the inclusion of the brain indices and perhaps better labeling would add clarity. Further explanation is needed to explain this figure and the significance it has in terms of the rest of the data presented.

#12: The PLOS One journal states that all data underlying the findings should available. There are 3 instances where the authors state that “data is not shown.” These occur in Line 248 as well as twice in Table 3. I would recommend sharing this data as a supplement.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at gro.solp@serugif. Please note that Supporting Information files do not need this step.

    2020; 15(4): e0232404.
    Published online 2020 Apr 29. 10.1371/journal.pone.0232404.r002

    Author response to Decision Letter 0

    Copyright and License information Disclaimer

    24 Mar 2020

    Dear Dr. Ehlers

    Thank you for the opportunity to revise and resubmit our manuscript. We thank you and the reviewer’s for the careful reading and helpful comments. A point by point response to each comment is included below. Material changes to the manuscript are noted below and changes are highlighted in the revised manuscript.

    Kind regards,

    Shahram Oveisgharan, MD

    EDITORS’ COMMENTS:

    E.1 The introduction is well-written, but does not support the methods or results sections as presented in the abstract or the manuscript. The introduction implies that a mediation analysis was conducted to test the mediating effect of brain pathologies on the relationship between PA and parkinsonism. However, the statistical analyses appear to test the association between PA and parkinsonism independent of brain pathologies and to test the interaction between PA and brain pathologies.

    We hope that the changes to the revised the manuscript align the introduction, methods and results to support the proposed mediation hypothesis tested in our primary analysis (Introduction, (P3-4, lines 53-73), Methods (P10-11, lines 217-231), and Results (P14-17, lines 263-280)). As suggested by the reviewer as well as the editor (E.10), we have added a figure that summarizes and illustrates the scientific framework and hypothesis tested in this manuscript (Figure 1).

    E.2 The number of analyses is unfocused and raises concerns whether hypotheses were developed a priori.

    We hope that the scientific framework outlined in the Introduction (Page 3-4, lines 53-73) and illustrated in Figure 1 underscore that the hypothesis tested was developed a priori and that changes in the main text focus the manuscript on the proposed mediation hypothesis. To further focus the paper, we have reduced the number of secondary analyses by removing the analysis of parkinsonian signs. The goal of our primary analysis was to elucidate whether mediation is a potential mechanism underlying the motor benefits of higher levels of physical activity. Multiple pathways likely underlie the health benefits of physical activity. To increase the scientific impact of this manuscript and maximally leverage the unique data used in these analyses, we retained the secondary analysis, during which a second potential mechanism distinct from mediation analysis was examined. In this secondary analysis, we examined whether the motor benefits of physical activity was via effect-modification of pathologies underlying parkinsonism (P18-19, lines 310-319)

    E.3 It appears that the four parkinsonism signs were tested in separate regression models. This is problematic, as the four subscales are likely correlated.

    We agree with the editor that these signs are related and this is why in our earlier publications, supported by principal component analysis (PCA), these signs have been combined into the global parkinsonian score (1). Nonetheless, as additional analyses were requested (E5, E7) and to maintain the focus of this manuscript (E2), we have removed the analysis of the four individual parkinsonian signs from the manuscript.

    E.4 As Reviewer 1 noted, the collinearity between brain pathology variables is not addressed.

    We examined collinearity among the brain pathology indices by calculating their variance inflation factor (VIF) (Table 2, model 2). The VIfs were less than 1.5 indicating that collinearity did not exist among the examined pathology indices(2). We added relevant sentences to the text indicating lack of collinearity among the examined brain pathologies (P11, lines 231-233; P17, lines 275-276).

    E.5 Were the physical activity data also positively skewed?

    The skewness of the total daily physical activity was 1.3 and of the intensity of daily activity was 0.9. As physical activity metrics were predictors, not outcome variables, in our study that had a sample size of 447 providing reliable mean and standard deviations for our variables (3) we did not use the square-root transformation of the physical activity data. Nonetheless, we repeated our analyses after using square-root transformation of both metrics of physical activity (total daily physical activity and intensity of the physical activity). Using their square root transformation did not change the conclusion that a higher level of physical activity was associated with less severe parkinsonism. In linear regressions controlled for age at death and sex, a higher level of square root of total daily physical activity (estimate = -0.868, SE = 0.123, p<0.001) or square root of intensity of physical activity (estimate = -4.183, SE = 0.567, p<0.001) were associated with less severe parkinsonism proximate to death.

    E.6 While Figure 1 is pretty, it does not offer additional information not already available in Table 2.

    We removed the previous Figure 1 and replaced it with the Figure suggested by the editor in comment E.1.

    E.7 Please provide further information on the choice of covariates - it seems that only age at death and sex were included in the models, but other covariates (e.g., education, marital status, medications) may also be relevant.

    Additional analyses have been added adjusting our models for education, marital status, vascular risk factors and diseases, and use of neuroleptic medications and our results were unchanged. These analyses are summarized in the main text [methods (P11, lines 233-235), results (P17, lines 284-289), Table 1] and included in a supplementary table (Supplementary Table e-3).

    E.8 While the authors cite two previously published studies utilizing the physical activity data processing methods, it is not clear how sleep was accounted for in the data processing. The protocol utilizes 24-hour monitoring, and does not mention how sleep was filtered out. Movement during sleep time may represent disrupted sleep and should not be included in total activity counts. Please include a statement on how sleep time was handled in the data processing.

    We agree that some movements during sleep may represent disrupted sleep and we have developed a metric to capture and quantify movements related to periods of sleep/rest. Conventional sleep diaries are not employed in MAP because of participant’s burden, difficulty in minimizing missing data, and progressive cognitive impairment that affects the validity and reliability of these data. Moreover, in this cohort sleep or naps are not restricted to the night. On average 60% of the 24 hour recording shows no activity (either naps or sleep). Thus, periods of rest or sleep are not restricted to the “night”, but are present throughout the 24 recording making it impractical to filter out movement during “rest” or “sleep”.

    To address the editor’s concern, we used a previously validated metric that quantifies the probability of rest/sleep disruption by movement, and is derived from the continuous 24 hour actigraphic recordings. KRA is the transition probability once sustained rest is achieved. This metric examines the temporal organization of human rest-activity patterns in terms of transition probabilities between periods of rest and activity. Conceptually, kRA is a measure of the tendency for disruption of rest (no activity) by physical activity. A higher value of kRA represents more fragmented sleep/rest - more movement during periods of sleep/rest. A lower value represents a more consolidated sleep/rest. Therefore, kRA indirectly captures movement/physical activity during sleep. KRA was weakly correlated with total daily physical activity (Spearman correlation coefficient = 0.15, p = 0.003).

    To adjust for sleep-related movement, we added a term for kRA to our primary model and the association of total daily physical activity with parkinsonism in presence of brain pathology indices was unchanged. These data are summarized in the Results (P18, lines 297-307), and the full models are included in the Supplementary Table e-5. Kra is described in the methods (P6-7, lines 124-135), and has been added to Table 1.

    E.9 Please pay specific attention to Review 1's comments related to the discussion section of the paper.

    We hope that the changes to the discussion described below in response to Reviewer 1’s comments have adequately addressed these concerns.

    E.10 Review 1 and I thought the presentation of the scientific evidence utilizing a logical argument of ABC was sufficient and helpful; however, Review 2 found it unclear. Please consider including a figure to illustrate the proposed scientific framework/analyses for the research question.

    We added a figure, Figure 1, illustrating the proposed scientific framework and proposed hypothesis testing in this manuscript.

    E.11 Please include ranges in parentheses for parkinsonism variables in Table 1.

    We included ranges in parentheses for the parkinsonism variables in Table 1.

    REVIEWER 1

    R1.1 Some light editing is still required throughout for spelling and grammar.

    Thank you for pointing this out and we hope we have corrected all spelling and grammar errors.

    R1.2 I'm not sure I would say r of .4-.7 is "highly correlated". Might be better not to use an adjective here (p 7, line 134).

    We deleted “highly” in the mentioned sentences.

    R1.3 I think it would be important to note if there were any differences in PA (or other demographics) or other characteristics between those who had autopsy data, or UPDRS, and those who did not. I don't suspect there are but one never knows how selection bias might surface.

    These data have been added to the Supplementary Table e-1 comparing included and excluded participants in the demographics and clinical variables and summarized in the main text (P5, lines 92-96).

    R1.4 Help me understand the basis for your claim that the data support the idea that pathologies explain parkinsonism (Table 2, Model 2). 5.8% variance explained beyond age and sex seems a bit underwhelming (and there is no discussion of variance inflation amongst correlated neuropathologies, or presentation of the fully model, just the components).

    We agree with the reviewer. While this manuscript is focused on physical activity, our prior publications have focused on this important question raised by the reviewer (4–9). In prior reports which examining different motor phenotypes, we have found that indices of brain pathologies account for a only a small minority of the motor phenotype variance (5-15%)(9) much less than the 30-50% observed with cognitive phenoytpes (10). This question is the current focus of our research efforts that seek to determine the pathologic basis underlying progressive parkinsonism and motor decline in old adults. While networks that subserve cognition are found in the brain, the networks underlying motor function extend from the brain through the CNS to reach muscle in the periphery. Our recent work shows that ADRD pathologies extend beyond the brain to brainstem and spinal cord regions(11). Thus, to determine the full extent of the pathologic basis for parkinsonism as well as the mechanisms underlying the motor benefits of physical activity one needs to investigate the pathologies not only in the brain but also interrogate vital motor-related tissues outside the brain. This point is addressed in the Discussion, (P20-21, lines 359-363).

    Moreover, while the amount of parkinsonism variance explained by the physical activity or ADRD pathologies is small at the individual level the effect at the public health level is not inconsequential. Given the extent of motor impairment in old age, that half of the adults 85 years or older have parkinsonian signs, even the modest effect sizes observed in the current study are important (12). These details are noted in the Discussion (P21, lines 363-366).

    Variance Inflation: We also examined collinearity among the brain pathology indices by calculating their variance inflation factor (VIF) (Table 2, model 2). The VIfs were less than 1.5 indicating that variance inflation did not occur among the examined pathology indices(2). These data have been added to the text (P10, lines 201-203; P14, lines 230-231).

    R1.5 7.2% variance explained by PA is a bit underwhelming. However, I get that people are noisy.

    Thank you for raising this important point. The variance accounted for by physical activity of parkinsonism is similar to that observed in a prior study(13) in which physical activity accounted for about 8% of the variance of a summary 10 motor performances. Some of the motor phenotypes assessed examine motor abilities. In contrast, total daily physical activity is a volitional behavior. Damage to any portion of the distributed motor system has potential to degrade motor abilities underlying movement while leaving motor decision-making regions necessary for the initiation of movement intact. Thus, an individual with poor motor abilities may nonetheless have higher levels of total daily physical activity compared to an individual with good motor abilities, who might elect to sit in a chair for the entire day. As illustrated in the scatterplot on the right(13), this may explain why the quantity of daily physical activity is only modestly related to motor abilities and not everyone with a high level of physical activity necessarily has good motor abilities.

    Finally, in prior publications most single risk factors significantly associated with motor phenotypes account for 2-3% of the variance of a motor phenotype (14). Therefore, at the population level, the fact that total daily physical activity accounts for 7% of the variance of parkinsonism suggests that it is an important risk factors for parkinsonism in older adults. We have added these points to the discussion (P20, lines 353-358; P21, lines 363-366).

    R1.6 I think my biggest concern in this manuscript is what seems to be the authors conclusion that the PAy measured can somehow be considered protective or "affording reserve". I know a veteran and established group such as this knows the inherent difficulty in identifying causality in regressions. Here one could easily interpret the data to mean that PAprotects against parkinsonism, or people with parkinsonism are less active. Thus the statement on p13, line 238 about physical activity affording reserve is ambitious at best.

    We agree that a well-known limitation of all cross-sectional observational studies is that causal inferences are limited due to the potential for reverse causality. This and other limitations affecting the current study are included in the discussion (P24, lines 443-451). However, as noted in the current discussion, we have published a longitudinal study in this same cohort which showed that a higher level of physical activity at study baseline was associated with a slower rate of progressive parkinsonism and decreased incidence of parkinsonism during four years of follow-up on average. In the absence of results from an interventional study supporting a causal relationship, our results are not conclusive. Yet, the results of the previous and current study lend support for considering the potential public health consequences of our current findings. Our findings inform on the need for further interventional studies of physical activity, and suggest potential utility of research efforts to identify molecular mechanisms without a pathologic footprint that drive the motor benefit of physical activity in older adults. The changes can be found at the following places in the manuscript (P21-22, lines 381-390; P24, lines 448-451).

    R1.7 And the conclusion that PA is not interacting with brain pathology, when PA was measured so close to death is perhaps overly pessimistic for the field. Again, on page 14, there is the hint of a conclusion about physical activity not affection neuropathology when the data provide no clear way to test this over time, or even with sufficient separation of activity data and neuropathology work up.

    We agree with the reviewer that there are multiple limitations to our current study design. No single study is going to elucidate the complex biology underlying the benefits of physical activity in older adults. Despite the limitations noted by the reviewer, by using the same approach as the current study used we have found the following clinical covariates as modifiers of the relationship between postmortem pathology and cognition proximate to death: education, purpose in life, physical activity and postmortem BDNF gene expression levels derived from prefrontal cortex (15–18). Thus, in the absence of prior human data, despite limitations these observational cross sectional data provide novel data about an important potential mechanism that could account for the motor benefits of physical activity. We hope that the changes in the revised text highlights the limitations of this study in examining the interaction of total daily physical activity with ADRD pathologies in their association with parkinsonism (P12, lines 244-248; P19, lines 317-319; P21, lines 369-373), and address in part the concerns of the reviewer.

    REVIEWER 2

    R2.1 In the introduction (page 3; starting line 57) the terms “physical activity” “parkinsonism” and “brain pathology” are given A, C and B labeling but I found this confusing to follow while reading the text and did not aid in the understanding of the framework. Initially I thought this was going to relate into an equation but this labeling was not mentioned again and thus would recommend removing it.

    Following the editor’s comment, E.1, we added a figure, Figure 1, illustrating scientific framework underlying current analyses. In addition, we revised the introduction to illustrate more clearly the underlying scientific framework of the study (P3, lines 53-59).

    R2.2 Page 4; Methods sections: I think it is pertinent to explicitly include if the cohort from the MAP is a “aging healthy control” database, or if these participants were recruited in the brain bank because they had cognitive decline or parkinsonism. This will help to provide further insight in how to interpret the data.

    The study is a cohort of aging healthy community-dwelling participants, and we added this to the manuscript (P4, lines 76-77).

    R2.3 Table 1 displays the clinical and postmortem characteristics of the cohort. It may be useful to add in this section that these characteristics (if indeed true) are similar to other reported clinical and brain bank findings in similarly aged subjects. This would serve to boost the generalizability of the data provided. For example – does the literature report similar findings of predominately (69%) AD pathologic findings in an aging cohort? Likewise is the degree of global parkinsonism similar to those also reported? Since this journal will attract more of a general audience; I think this will be helpful to put the results into perspective to those readers that are not as intricately aware of this data.

    We added the Supplementary Table e-6 including frequencies of ADRD pathology indices in our cohort in comparison to two other clinical-pathological studies. In addition, we added the limitations the current study has in the generalizability of its findings (P24, lines 440-443).

    R2.4 Table 2: Would consider re-naming the Models with a more descriptive title to increase the clarity of what the linear regression models are representing.

    At Table 2, we added abbreviations to the models’ names indicating included predictors. The changes are: model 1 is changed to model 1-TDPA, model 2 to model 2-Path, model 3 to model 3-TDPA+Path, and model 4 to model 4-Intensity+path. In addition, TDPA and intensity are defined in the “Model Term” column of Table 2.

    R2.5 Table 2: in the final row: “Additional Explained Variance” are these values shown without the 2.5% addition of the variance due to age/sex as referenced in the caption? I am nearly certain that is what is trying to be related, but the language in the caption and Table 2 could be more clear. Perhaps, “an additional 2.5% of the variance is due to age/sex” or something similar could replace the “compared to age and sex” language.

    We appreciate the suggestion. We added the suggested sentence to the footnote of Table 2, and deleted the “compared to age and sex” from the table.

    R2.6 Page 13, Line 223: Regarding the statement (higher burden of brain pathology related to more severe parkinsonism) made about Model 2 – I am unclear about which data in Table 2 represent this statement. There does not appear to be a “global pathology burden” index to represent the overall burden of pathology; instead each individual pathology is listed and its individual relationship to parkinsonism. Additionally the linear regression models are not significant as P values are greater than 0.05 except for Nigral neuronal loss and Atherosclerosis. The association between these 2 pathologies and parkinsonism is mentioned in the text but I feel that the statement “higher burden of brain pathologies was associated with more severe parkinsonism” is unfounded because the bulk of the brain pathology did not show this same association.

    We deleted the “higher burden of brain pathologies was associated with more severe parkinsonism” from the text.

    R2.7 Page 13, line 229: Then sentence starting “In the latter model…” It is not clear which model you are referring to and again would be more clear if the models were renamed or at least directly referenced in this sentence. Is this referring to Model 2? If so, the variance mentioned in the text of 5.7% does not match the corresponding variance in Table 2 of 5.8%.

    We apologize. We named the model, which is model 3-TDPA-Path. In this model, 8.3% of the parkinsonism severity variance was explained by age, sex, and pathologies and an additional 5.7% by total daily physical activity.

    R2.8 Figure 1 section: The text starting on 233 and ending on 237: “The blue regression line….pathologies to the first model” seem more appropriate for the Figure 1 caption.

    We deleted Figure 1 and its section in the text, following E.6 and R2.9.

    R2.9 Figure 1 section: Two points – 1) line 235 “A second red line..” are there supposed to be two red lines or should this be worded “The red line…” ? 2) The red regression line is displaying Model 3 from the description provided but the text references Models 1 and 2 at the start of this section. It is not clear to me which model this red line is representing- I think it is Model 3 but this should be more clearly stated. Figure 1: Overall this figure doesn’t really add to any new information except for a different way to represent the data from Table 2. I would recommend cutting this figure all together. If the goal is present a more “graphic” form of the information (which in my opinion is more digestible) – then the opposite approach could be considered and would suggest creating figures to represent the other models as well. If this approach was done, then Table 2 could be made available as a supplement table so as not to give redundant data in the body of the text.

    We deleted Figure 1 and its section in the text, following E.6 and R2.9.

    R2.10 Page 15, line 267-268: The claim that higher level of daily activity is associated with lower odds of tremor does not appear to be true as the p value is 0.066

    Following E.2 comment, we deleted Table 3 and related segment from the paper to make the analyses more focused, and the paper clearer.

    R2.11 Table 3: It is not clear what is meant by predictors “with TDPA” and “without TDPA.” It seems like these rows are representing data with and without the inclusion of the brain indices and perhaps better labeling would add clarity. Further explanation is needed to explain this figure and the significance it has in terms of the rest of the data presented.

    Following E.2 comment, we deleted Table 3 and related segment from the paper to make the analyses more focused, and the manuscript clearer.

    R2.12 The PLOS One journal states that all data underlying the findings should available. There are 3 instances where the authors state that “data is not shown.” These occur in Line 248 as well as twice in Table 3. I would recommend sharing this data as a supplement.

    This data has been added as Supplementary Table e-4. Table 3 are deleted following E.2 comments.

    REFERENCES

    1. Bennett DA, Shannon KM, Beckett LA, Wilson RS. Dimensionality of parkinsonian signs in aging and Alzheimer’s disease. J Gerontol Biol Sci Med Sci. 1999 Apr;54(4):M191-6.

    2. Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography. 2013 Jan;36(1):27–46.

    3. Piovesana A, Senior G. How Small Is Big: Sample Size and Skewness. Assessment. 2018 Sep;25(6):793–800.

    4. Buchman AS, Leurgans SE, Nag S, Bennett DA, Schneider JA. Cerebrovascular disease pathology and parkinsonian signs in old age. Stroke. 2011 Nov;42(11):3183–9.

    5. Buchman AS, Shulman JM, Nag S, Leurgans SE, Arnold SE, Morris MC, et al. Nigral pathology and parkinsonian signs in elders without Parkinson disease. Ann Neurol. 2012 Feb;71(2):258–66.

    6. Buchman AS, Yu L, Boyle PA, Levine SR, Nag S, Schneider JA, et al. Microvascular brain pathology and late-life motor impairment. Neurology. 2013 Feb 19;80(8):712–8.

    7. Buchman AS, Yu L, Wilson RS, Boyle PA, Schneider JA, Bennett DA. Brain pathology contributes to simultaneous change in physical frailty and cognition in old age. J Gerontol Biol Sci Med Sci. 2014 Dec;69(12):1536–44.

    8. Buchman AS, Nag S, Shulman JM, Lim ASP, VanderHorst VGJM, Leurgans SE, et al. Locus coeruleus neuron density and parkinsonism in older adults without Parkinson’s disease. Mov Disord. 2012 Nov;27(13):1625–31.

    9. Buchman AS, Yu L, Wilson RS, Leurgans SE, Nag S, Shulman JM, et al. Progressive parkinsonism in older adults is related to the burden of mixed brain pathologies. Neurology. 2019 Apr 16;92(16):e1821–30.

    10. Boyle PA, Wilson RS, Yu L, Barr AM, Honer WG, Schneider JA, et al. Much of late life cognitive decline is not due to common neurodegenerative pathologies: Cognitive Decline. Ann Neurol. 2013 Sep;74(3):478–89.

    11. Buchman AS, Leurgans SE, Nag S, VanderHorst V, Kapasi A, Schneider JA, et al. Spinal Arteriolosclerosis Is Common in Older Adults and Associated With Parkinsonism. Stroke. 2017 Oct;48(10):2792–8.

    12. Louis ED, Luchsinger JA, Tang MX, Mayeux R. Parkinsonian signs in older people: prevalence and associations with smoking and coffee. Neurology. 2003 Jul 8;61(1):24–8.

    13. Buchman AS, Yu L, Wilson RS, Lim A, Dawe RJ, Gaiteri C, et al. Physical activity, common brain pathologies, and cognition in community-dwelling older adults. Neurology. 2019 Feb 19;92(8):e811–22.

    14. Oveisgharan S, Yu L, Dawe RJ, Bennett DA, Buchman AS. Total daily physical activity and the risk of parkinsonism in community-dwelling older adults. J Gerontol Biol Sci Med Sci. 2019;

    15. Bennett DA, Wilson RS, Schneider JA, Evans DA, Mendes de Leon CF, Arnold SE, et al. Education modifies the relation of AD pathology to level of cognitive function in older persons. Neurology. 2003 Jun 24;60(12):1909–15.

    16. Buchman AS, Yu L, Boyle PA, Schneider JA, De Jager PL, Bennett DA. Higher brain BDNF gene expression is associated with slower cognitive decline in older adults. Neurology. 2016 Feb 23;86(8):735–41.

    17. Boyle PA, Buchman AS, Boyle PA, Yu L, Schneider JA, Bennett DA. Effect of Purpose in Life on the Relation Between Alzheimer Disease Pathologic Changes on Cognitive Function in Advanced Age. Arch Gen Psychiatry. 2012 May 1;69(5):499.

    18. Fleischman DA, Yang J, Arfanakis K, Arvanitakis Z, Leurgans SE, Turner AD, et al. Physical activity, motor function, and white matter hyperintensity burden in healthy older adults. Neurology. 2015 Mar 31;84(13):1294–300.

    Attachment

    Submitted filename: Response to Reviewers-TDPA-path-park-v3.docx

      2020; 15(4): e0232404.
      Published online 2020 Apr 29. 10.1371/journal.pone.0232404.r003

      Decision Letter 1

      Diane K. Ehlers, Academic Editor
      Copyright and License information Disclaimer

      15 Apr 2020

      Total daily physical activity, brain pathologies, and parkinsonism in older adults

      PONE-D-19-27901R1

      Dear Dr. Oveisgharan,

      We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

      Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

      Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at gro.solp@gnillibrohtua.

      If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact gro.solp@sserpeno.

      With kind regards,

      Diane K. Ehlers, PhD

      Academic Editor

      PLOS ONE

      Additional Editor Comments (optional):

      Please review for minor grammatical errors prior to production.

      Reviewers' comments:

      Reviewer's Responses to Questions

      Comments to the Author

      1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

      Reviewer #1: All comments have been addressed

      Reviewer #2: All comments have been addressed

      **********

      2. Is the manuscript technically sound, and do the data support the conclusions?

      The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

      Reviewer #1: Yes

      Reviewer #2: Yes

      **********

      3. Has the statistical analysis been performed appropriately and rigorously?

      Reviewer #1: Yes

      Reviewer #2: Yes

      **********

      4. Have the authors made all data underlying the findings in their manuscript fully available?

      The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

      Reviewer #1: Yes

      Reviewer #2: Yes

      **********

      5. Is the manuscript presented in an intelligible fashion and written in standard English?

      PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

      Reviewer #1: Yes

      Reviewer #2: Yes

      **********

      6. Review Comments to the Author

      Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

      Reviewer #1: The authors have addressed my initial concerns. I have no further comments that I think would improve the readability or utility of the manuscript.

      Reviewer #2: (No Response)

      **********

      7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

      If you choose “no”, your identity will remain anonymous but your review may still be made public.

      Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

      Reviewer #1: Yes: Eric Vidoni

      Reviewer #2: Yes: Mara Seier

        2020; 15(4): e0232404.
        Published online 2020 Apr 29. 10.1371/journal.pone.0232404.r004

        Acceptance letter

        Diane K. Ehlers, Academic Editor
        Copyright and License information Disclaimer

        20 Apr 2020

        PONE-D-19-27901R1

        Total daily physical activity, brain pathologies, and parkinsonism in older adults

        Dear Dr. Oveisgharan:

        I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

        If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact gro.solp@sserpeno.

        For any other questions or concerns, please email gro.solp@enosolp.

        Thank you for submitting your work to PLOS ONE.

        With kind regards,

        PLOS ONE Editorial Office Staff

        on behalf of

        Dr. Diane K. Ehlers

        Academic Editor

        PLOS ONE

          Articles from PLOS ONE are provided here courtesy of PLOS

          Full text links 

          Read article at publisher's site: https://doi.org/10.1371/journal.pone.0232404

          Read article for free, from open access legal sources, via Unpaywall: https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0232404&type=printableUnpaywall

          Citations & impact 

          Impact metrics

          8
          Citations
          Jump to Citations

          Citations of article over time

          Alternative metrics

          Altmetric item for https://www.altmetric.com/details/80890660
          Altmetric
          Discover the attention surrounding your research
          https://www.altmetric.com/details/80890660

          Article citations

          Go to all (8) article citations

          Data 

          Data behind the article

          This data has been text mined from the article, or deposited into data resources.

          BioStudies: supplemental material and supporting data

          Similar Articles 

          To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.

          Funding 

          Funders who supported this work.

          NIA NIH HHS (8)

          National Institutes of Health (7)