Accelerometer-assessed light physical activity is protective of future cognitive ability: A longitudinal study among community dwelling older adults

Article  in  Experimental gerontology · March 2017 DOI: 10.1016/j.exger.2017.03.003 CITATIONS 10 READS 88 5 authors, including:

Some of the authors of this publication are also working on these related projects:

CRIS output - dementia and mental health of older adults View project

Integrating physical activity in low resourced mental health settings View project Brendon Stubbs

King's College London

456PUBLICATIONS   5,078CITATIONS   

SEE PROFILE

Li- Jung Chen

National Taiwan University of Sport, Taiwan 68PUBLICATIONS   590CITATIONS   

SEE PROFILE

Wen- Jung Sun

41PUBLICATIONS   324CITATIONS   

SEE PROFILE

Po-Wen Ku

National Changhua University of Education 90PUBLICATIONS   687CITATIONS   

future cognitive ability: A longitudinal study among community dwelling older adults

Brendon Stubbs, Li-Jung Chen, Chun-Yi Chang, Wen-Jung Sun, Po-Wen Ku

PII: S0531-5565(16)30296-0

DOI: doi:10.1016/j.exger.2017.03.003

Reference: EXG 10015

To appear in: Experimental Gerontology

Received date: 30 August 2016

Revised date: 24 February 2017

Accepted date: 2 March 2017

Please cite this article as: Brendon Stubbs, Li-Jung Chen, Chun-Yi Chang, Wen-Jung Sun, Po-Wen Ku , Accelerometer-assessed light physical activity is protective of future cognitive ability: A longitudinal study among community dwelling older adults. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Exg(2017), doi:10.1016/j.exger.2017.03.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Accelerometer-assessed light physical activity is protective of future cognitive ability:

A longitudinal study among community dwelling older adults

Brendon Stubbs, Li-Jung Chen, Chun-Yi Chang, Wen-Jung Sun*, Po-Wen Ku*

Running title: Accelerometer-assessed light physical Activity and cognitive impairment

Submission to Experimental Gerontology

1st author: Brendon Stubbs, PhD

Physiotherapy Department, South London and Maudsley NHS Foundation Trust, Denmark Hill, London SE5 8AZ, UK

Health Service and Population Research Department, Institute of Psychiatry, Psychology and Neuroscience, King's College London, De Crespigny Park, London, Box SE5 8AF, UK Faculty of Health, Social Care and Education, Anglia Ruskin University, Chelmsford, UK E-mail: brendon.stubbs@kcl.ac.uk

2nd Author: Li-Jung Chen, PhD

Department of Exercise Health Science, National Taiwan University of Sport, Taiwan E-mail: ljchen@ntupes.edu.tw

3rd Author: Chun-Yi Chang, PhD

Department of Physical Education, National Hsinchu University of Education, Hsinchu, Taiwan

E-mail: cjy@mail.nhcue.edu.tw

4th and corresponding author: Wen-Jung Sun, MD, PhD

Family Medicine Department, Taipei City Hospital Zhongxing Branch, Taiwan E-mail: das48@tpech.gov.tw

5th and corresponding Author: Po-Wen Ku, PhD (Epidemiology), PhD (Exercise & Health) Graduate Institute of Sports and Health, National Changhua University of Education, Taiwan Email: powen.ku@gmail.com

TEL: +886 4 7232105 ext. 1991 Fax: +886 4 7211155

ACCEPTED MANUSCRIPT

Abstract

Objective (246/250)

Physical activity (PA), especially moderate-to-vigorous intensity, could protect older adults

from cognitive impairment. However, most literature is based on self-reported PA which is

limited by recall bias. Light PA is popular among older adults, but a paucity of objective

longitudinal data has considered the relationship between light PA and cognitive ability. We

examined if a higher level of objectively measured light PA, independent of

moderate-to-vigorous physical activity (MVPA), was prospectively associated with better

cognitive ability in older adults.

Methods

A longitudinal study over 22.12 (±1.46) months including 274 community-dwelling older

adults across 14 regions in Taiwan was undertaken. Cognitive ability was obtained using a

Chinese version of the Ascertain Dementia 8-item Questionnaire (AD8) and light PA and

MVPA captured by 7 days accelerometer positioned on waist. Multivariable negative

ACCEPTED MANUSCRIPT

274 participants (74.52 years, 45.6% male) attended the follow-up (96.1%). Higher light PA,

independent from MVPA, was associated with a reduced rate of decline in cognitive ability

(rate ratio 0.75 [0.60-0.92]). MVPA, was also associated with a reduced decline in cognitive

ability (rate ratio 0.85 [0.75-0.95]). Light PA was protective of cognitive ability in

sensitivity analyses removing participants with activities of daily living difficulties,

depressive symptoms and cognitive impairment at baseline.

Conclusion

Our data suggest that light PA may offer a protective influence of future cognitive ability in

community dwelling older adults. The promotion of light PA may be a valuable means to

maintain cognitive ability in older age.

Key words: Cognitive decline, dementia, light physical activity, moderate to vigorous

ACCEPTED MANUSCRIPT

Dementia is a common neurodegenerative condition among older adults that typically leads to

a loss of independence, reduced quality of life, premature mortality, caregiver burden and

high levels of healthcare utilization and cost (Fiest et al. 2016; Prince et al. 2013). Given the

aging population, the total number of dementia cases will inevitably increase, and there are

pressing calls to prevent cognitive decline and ultimately dementia (Deckers et al. 2015).

Within the last twenty years there has been a rapid increase in interest for the potential of

physical activity to prevent cognitive decline and maintain good cognitive ability (Barnes and

Yaffe 2011; Deckers et al. 2015; Hamer and Chida 2009). Two recent systematic reviews

stipulated that physical activity is one of the top seven modifiable risk factors for cognitive

impairment in older age (Barnes and Yaffe 2011; Deckers et al. 2015). A recent

meta-analysis of longitudinal studies found that higher baseline levels of physical activity

were associated with a 14% reduced risk of future dementia (Relative risk (RR) 0.86, 95% CI

0.76-0.97) (Blondell et al. 2014). The authors (Blondell et al. 2014) only identified one

study capturing physical activity utilizing an objective measurement of physical activity

(Buchman et al. 2012), which found that higher total physical activity was independently associated with a reduced risk of Alzheimer’s disease over 4 years.

ACCEPTED MANUSCRIPT

of altered cognitive ability or cognitive impairment, is challenging. In fact, a recent

systematic review of psychometric and measurement properties of self-report physical activity

measures in older adults noted that most self-report measures have little evidence of any

validity and reliability (Falck et al. 2016). Whilst a small number of cross sectional studies

have investigated objective physical activity and cognition in older age (Doi et al. 2015;

Makizako et al. 2015), there remains a paucity of longitudinal studies investigating

objectively measured physical activity and cognitive impairment/ability. Another key

limitation in the literature is that international physical activity guidelines recommend that

older adults should do at least 150 minutes of moderate-intensity aerobic physical activity or

do at least 75 minutes of vigorous intensity aerobic physical activity throughout the week.

This appears to imply that little or no health benefits can be derived from light physical

activities (World Health Organization 2010). Light physical activity is a popular, relatively

safe and effective form of activity for older adults (Tse et al. 2015). Physical activity

conducted at moderate to vigorous intensity, whilst conferring potentially greater benefits

may also carry a greater risk of injury and potential dropout from exercise (Tse et al. 2015).

Thus, understanding if light intensity physical activity (e.g. casual walking, stretching, and

light yard/house work etc.), independent of moderate-to-vigorous physical activity (MVPA),

might offer protective effects on cognitive ability could provide clinically useful information.

ACCEPTED MANUSCRIPT

physical activity and cognitive abilities using objective devices, the current study examined

whether objectively measured light physical activity, independent of MVPA, is associated

with a reduced risk of decline in cognitive ability in older adults. To test for confounding and

reverse causation, sensitivity analyses were also conducted.

Methods

Study design and sample

The current longitudinal study utilized data from two-waves of a community-based

project conducted in Hunei District, Kaohsiung, which is the second largest city in Taiwan. In

total, 285 community-dwelling older adults who were aged 65 years or older were recruited

and assessed from August to October 2012. Participants were recruited from 14 village

regions, with approximately 20 people being recruited from each community center utilizing

quota sampling in which, participants were drawn based on a national distribution according

to sex and age in 2011 (Taiwan Ministry of Interior 2012). Follow up interviews were

conducted between May to July 2014. From the baseline sample, 274 participants (96.1%)

attended the follow-up after a mean of 22.12 ± 1.46 months. The reasons why participants (n=

11) did not attend the assessment of the second wave included: Inpatient (n= 2), not traceable

(n= 2), deceased cases (n = 4) and refuse to attend (n= 3).

ACCEPTED MANUSCRIPT

ethical approval from National Taiwan University of Sport Institutional Review Board,

Taiwan. All data was collected using a standardized interview format, conducted with

face-to-face interviews at each person’s house.

Measures

Cognitive ability

Cognitive ability was obtained using a Chinese version of the Ascertain Dementia 8-item

Questionnaire (AD8) by participants with a potential range between 0 and 8 (Galvin et al.

2005; Galvin et al. 2007). The AD8 comprises 8 items that ask the respondents to rate

change in memory, problem-solving abilities, orientation, and daily activities (yes=1, no=0).

Higher scores represent cognitive ability decline (Galvin et al. 2005; Ganguli et al. 2014).

The Chinese version of the AD8 has demonstrated adequate reliability and validity among

community-dwelling older Taiwanese adults (Yang et al. 2011). Within the current study,

the Cronbach’s alpha reliability coefficients for the AD8 ranged between 0.79 (first-wave)

and 0.81 (second-wave) across the two waves of data collection.

Objective physical activity

Physical activity were captured using waist worn triaxial accelerometer monitors

(GT3X+, ActiGraph, Pensacola, FL, USA) for 7 days. In order to be included in the study,

participants had to wear the accelerometer for a minimum of 10 hours of monitoring on at

ACCEPTED MANUSCRIPT

software (ActiGraph, Pensacola, FL, USA). Periods of 60 min of consecutive zero counts

were considered as non-wearing time and were excluded from the analyses. Physical activity

parameters were then computed using established cut offs, comprising time spent in light

physical activity (100–1951 counts/min), and moderate-to-vigorous activity (>1951

counts/min) (Gorman et al. 2014), together with total physical activity energy expenditure

(kcal/week). Research evidence has demonstrated that there is a strong correlation between

the indirect calorimetry estimates of energy expenditure and energy expenditure (kcal)

assessed by GT3X+ (r = 0.82) (McMinn et al. 2013). For descriptive purposes, the various

physical activity categories were converted into tertiles, but the primary analyses were

conducted on continuously distributed variables.

Covariates

A number of covariates were collected at baseline in line with previous literature (Coley

et al. 2008; Hamer and Chida 2009): including (i) socio-demographic factors: sex, age (65-74

years, 75+ years), educational attainment (no formal schooling, primary school, secondary

school+), marital status (married/cohabitating, others), main source of income (from offspring

vs. self [e.g. pension/savings]); (ii) lifestyle behaviors: smoking status (current, never, or

former smokers), alcohol consumption (yes vs. no), and; (iii) health status: body mass index

(BMI) (<18.50, 18.50-23.99, 24-26.99, 27+),(Taiwan Department of Health 2003) number of

ACCEPTED MANUSCRIPT

cancer, chronic obstructive pulmonary disease (COPD), liver disease, renal disease, and

arthritis; difficulties with activities of daily living (ADLs, no difficulties at all vs. some or

great difficulties); depressive symptoms assessed by the 15-item Geriatric Depression Scale

(GDS) using cutoff of 5 (no vs. yes), (Brink et al. 1982; Yesavage and Sheikh 1986); mean

daily accelerometer wear time (Hamer et al. 2014), and baseline AD8 scores.

Data analysis

Descriptive statistics were used to describe the features of the study sample. Given the

violation of normality, Mann Whitney U tests and Kruskal-Wallis tests were adopted to check

for differences in cognitive ability in 2014 across levels of accelerometer-derived parameters

(all in tertiles), and covariates. Previous research indicated that the use of the conventional

level (p-value = 0.05) may fail to identify variables known to be important. Variables with a p

value less than 0.25 were included in the subsequent regression models for adjustment

(Hosmer et al. 2013).

To examine the bi-correlations between objectively assessed physical activity parameters

and subsequent cognitive ability after controlling for accelerometer wear time, partial

correlation coefficients between physical activity energy expenditure (kcal/week as a

continuous variable), time (hours/day as a continuous variable) spent in physical activity at

ACCEPTED MANUSCRIPT

Multivariable negative binomial regression was conducted because the outcome variable

was an over-dispersed count with a highly skewed distribution. All accelerometer-derived

physical activity parameters were log-transformed before conducting regression analyses due

to non-normality (Tudor-Locke et al. 2011). Two separate unadjusted regression models

(single-factor models) for light and moderate-to-vigorous activity were conducted to assess

the associations between each intensity categories and cognitive ability. Then, two separate

multivariable regression models (single-factor models) for light and moderate-to-vigorous

activity (without mutual adjustment) were conducted to assess the associations between each

intensity categories and cognitive ability after adjusting for baseline cognitive ability, wear

time of accelerometer, socio-demographic variables, lifestyle behaviors, and chronic

conditions. Finally, one multivariable regression model (a two-factor model) for light

activity was fitted to examine the relationships in more detail after adjusting for MVPA and

other covariates.

Sensitivity analyses were carried out to evaluate confounding and reverse causation. In

first stage, we considered the possibility that ADL difficulties might influence physical

activity behaviors at baseline and subsequent cognitive ability, so the negative binomial

regressions were repeated after excluding the 13 participants with impaired ADLs. At the

second stage, we mitigated the potential impact of depressive symptoms on physical activity

ACCEPTED MANUSCRIPT

(Guo et al. 1988). At the third stage, participants with suspected mild cognitive ability or

dementia (i.e. AD8 scores equal or greater than 2) at baseline (n=65) which may have been

not only associated with baseline physical inactivity and subsequent health conditions but also

associated with the accuracy of survey responses. Sensitivity analysis was conducted

excluding those with suspected mild cognitive impairment or dementia at baseline.

All analyses were conducted using IBM SPSS 20.0 software and a p value < 0.05 was

considered statistically significant.

Results

Full details of the sample with follow up data are presented alongside AD8 cognition scores

in Table 1. At baseline, the mean score of AD8 was 0.80 (SD= 0.93). The mean age of the

sample was 74.52 (SD= 6.12) years, just under half of the sample were male (45.6%), while

the majority of the sample did not smoke (86.8% nonsmokers) or drink alcohol (95.2%

non-drinkers). In the univariate analyses, higher AD8 scores at follow-up (i.e. worse

cognition) were associated with female gender, increasing age, lower education status, low

light PA, MVPA, total physical activity, more chronic conditions, and having depressive

ACCEPTED MANUSCRIPT

Table 1 Distribution of cognitive ability in 2014 between levels of descriptors in 2012

Variables in 2012 N AD8 scores in 2014

Mean (SD) p-valuea Socio-demographic Sex 0.003 Male 125 0.85 (1.31) Female 149 1.29 (1.61) Age 0.001 65-74 156 0.81 (1.22) 75+ 118 1.45 (1.74) Education level 0.002 Secondary school+ 48 0.46 (0.46) Primary school 113 1.17 (1.17) No formal schooling 113 1.27 (1.27) Marital status 0.002 Married 192 0.92 (0.93) Others 82 1.48 (1.48)

Main source of income 0.001

Offspring 139 0.83(1.34)

Self (pension/savings) 135 1.35(1.61)

Lifestyle behaviors

Total energy expenditures (kcal/wk) < 0.001

High 92 0.48(0.75) Medium 91 0.90 (1.27) Low 91 1.89 (1.89) Moderate-to-vigorous PA (hour/day) < 0.001 High 92 0.55 (0.75) Medium 88 1.08 (1.51) Low 94 1.62 (1.83) Light PA (hour/day) <0.001 High 93 0.54 (0.67) Medium 91 1.04 (1.61) Low 90 1.70 (1.76) Smoking 0.122 Never smoker 238 1.03 (1.46) Current smoker 22 1.36 (1.65) Former smoker 14 1.71 (1.77)

ACCEPTED MANUSCRIPT

No 261 1.08 (1.49)

Yes 13 1.31 (1.75)

Health Status

Body mass index 0.888

Underweight <18.5 11 1.18 (1.78)

Normal 18.5-23.99 101 1.06 (1.26)

Overweight 24-26.99 101 1.12 (1.70)

Obese 27+ 61 1.07 (1.47)

Number of chronic diseases 0.041

0 118 0.86 (1.30) 1 104 1.13 (1.37) 2+ 52 1.52 (2.00) Depressive symptoms 0.018 No 225 0.99 (1.41) Yes 49 1.53 (1.78)

Activities of daily living 0.012

No difficulty at all 261 1.03 (1.44)

Some or great difficulties 13 2.23 (2.13)

Baseline cognitive ability (AD8) 274 ρ= 0.455b < 0.001b

AD8= The Ascertain Dementia 8-item Informant Questionnaire (AD8) a: Mann Whitney U test or Kruskal-Wallis test

b: Spearman correlation

Correlation analyses between physical activity and subsequent cognitive ability

Table 2 summarizes the partial correlations between each physical activity category at

baseline, and subsequent cognitive ability. Briefly, worse cognition was negatively

ACCEPTED MANUSCRIPT

Table 2 Partial correlation coefficients between baseline physical activity parameters and

cognitive ability at follow-up adjusting for mean daily accelerometer wear time (n=274) Variables

(continuous)

Baseline Follow-up

Energy expenditures

MVPA LPA Cognitive

ability

Energy expenditures 1.00

MVPA 0.75*** 1.00

LPA 0.79*** 0.39*** 1.00

Cognitive ability -0.35*** -0.25*** -0.39*** 1.00

MVPA: Moderate-to-vigorous physical activity; LPA: Light physical activity; *** p < 0.001

Multivariate relationship between physical activity and subsequent cognitive score

The crude and adjusted estimates of single and multifactor models are presented in Table 3.

In the crude and adjusted analyses of single-factor models, light physical activity and MVPA

were both significantly associated with a reduced rate of cognitive ability decline. In the

adjusted estimates of two-factor model, light physical activity was independent from MVPA

and associated with a reduced rate of cognitive ability decline with a rate ratio of 0.75

(0.60-0.92). Similarly, MVPA was also independent from light PA and associated with a

ACCEPTED MANUSCRIPT

Physical activity (continuous)

Single-factor models Two-factor model

Crude RR (95% CI)a P Adjusted RR (95% CI)b p Adjusted RR (95% CI)c p

MVPA 0.74 (0.66-0.82) < 0.001 0.81 (0.72-0.91) < 0.001 0.85 (0.75-0.95) 0.006

LPA 0.53 (0.45-0.64) < 0.001 0.67 (0.53-0.84) 0.001 0.75 (0.60-0.92) 0.007

RR= rate ratio; MVPA= moderate-to-vigorous physical activity; LPA= light physical activity

a: The two (single-factor) regression models for estimating the crude RR represent the unadjusted association of MVPA or LPA with cognitive impairment without mutual adjustment.

b: The two (single-factor) regression models for estimating the adjusted RR represent the multivariable association of MVPA or LPA with

cognitive ability without mutual adjustment. Covariates in the two models: baseline cognitive scores, sex, age, educational attainment, marital status, income source, smoking, number of chronic diseases, depressive symptoms, activities of daily living, and wear time of accelerometer c: The (two-factor) regression model for estimating the adjusted RR represent the multivariable association of LPA with cognitive ability.

Covariates in the model: baseline cognitive scores, sex, age, educational attainment, marital status, income source, smoking, number of ability chronic diseases, depressive symptoms, activities of daily living, wear time of accelerometer, and MVPA,

ACCEPTED MANUSCRIPT

Sensitivity analyses

The crude and adjusted analyses of single and multifactor models in sensitivity analyses are

shown in Table 4. To examine confounding and reverse causation, sensitivity analyses were

conducted to exclude people with ADL difficulty, people with both ADL difficulty and

depressive symptoms, and those with suspected cognitive impairment at baseline respectively.

However, the effect of light physical activity on subsequent cognitive ability remained. The

patterns of physical activity with cognitive ability at follow-up were similar to those in Table

ACCEPTED MANUSCRIPT

Stage 1: Excluding participants with ADL difficulties at baseline (n= 261)

Physical activity (continuous)

Single-factor models Two-factor model

Crude RR (95% CI)a p Adjusted RR (95% CI)b p Adjusted RR (95% CI)c p

MVPA 0.73 (0.65-0.82) < 0.001 0.81 (0.71-0.92) 0.002 0.85 (0.74-0.97) 0.018

LPA _{0.53 (0.44-0.64) < 0.001 0.68 (0.54-0.86) 0.001} 0.74 (0.58-0.95) 0.017

Stage 2: Excluding participants with ADL difficulties and depressive symptoms at baseline (n= 216)

Physical activity _{Crude RR (95% CI)}a

p Adjusted RR (95% CI)b p Adjusted RR (95% CI)c p

MVPA 0.75 (0.67-0.85) < 0.001 _{0.81 (0.70-0.94) 0.005 0.86 (0.74-0.99) 0.043}

LPA 0.50 (0.39-0.64) < 0.001 _{0.62 (0.45-0.84) 0.002 0.69 (0.51-0.94) 0.018}

Stage 3: Excluding participants with suspected cognitive impairment at baseline (n=209)

Physical activity _{Crude RR (95% CI)}a

p Adjusted RR (95% CI)b p Adjusted RR (95% CI)c p

ACCEPTED MANUSCRIPT

LPA 0.57 (0.45-0.74) < 0.001 0.60 (0.44-0.81) 0.001 0.69 (0.51-0.95) 0.021

RR= rate ratio; MVPA= moderate-to-vigorous physical activity; LPA= light physical activity

a: The two (single-factor) regression models for estimating the crude RR represent the unadjusted association of MVPA or LPA with cognitive ability without mutual adjustment.

b: The two (single-factor) regression models for estimating the adjusted RR represent the multivariable association of MVPA or LPA with cognitive ability without mutual adjustment. Covariates in the two models: baseline cognitive scores, sex, age, educational attainment, marital status, income source, smoking, number of chronic diseases, depressive symptoms, activities of daily living, and wear time of accelerometer

c: The (two-factor) regression model for estimating the adjusted RR represent the multivariable association of LPA with cognitive ability. Covariates in the model: baseline cognitive scores, sex, age, educational attainment, marital status, income source, smoking, number of chronic diseases, depressive symptoms, activities of daily living, wear time of accelerometer, and MVPA.

ACCEPTED MANUSCRIPT

The current study established that objective light physical activity offers a protective

effect from future decline in cognitive abilities over approximately two years among

community dwelling older adults. The protective effects of light physical activity were

independent of MVPA and the results remained robust in sensitivity analyses removing those

people with difficulties with ADL, depressive symptoms and suspected mild cognitive

impairment at baseline.

With the rapidly growing number of older people across the world, our data that objective

light physical activity confers a protective effect for future cognitive abilities are welcome and

has potential public health implications. Previous reviews (Blondell et al. 2014; Hamer and

Chida 2009; Paterson and Warburton 2010) have repeatedly highlighted the absence of

studies considering objective physical activity data and future cognitive impairment. Thus,

the overwhelming reliance of self-report physical activity from previous studies has infiltrated

a bias within the literature and such questionnaires cannot accurately disentangle the potential

individual benefits of different intensities of physical activity. The unique protective

influences of light physical activity on cognition are welcome for a number of reasons. First,

light physical activity such as casual walking, gardening and household chores are a preferred

method of accumulating physical activity for older people (Farren et al. 2015). Moreover,

ACCEPTED MANUSCRIPT

and friends and reduce the risk of social isolation. In addition, light physical activity is

associated with better wellbeing, physical health (Buman et al. 2010; Ku et al. 2016) and

improvements in other health outcomes such as reducing blood pressure, body fat, cholesterol

and improving cardiorespiratory fitness (Hanson and Jones 2015). Moreover, the risk of

injury and adverse outcomes are typically less with light physical activity versus more

vigorous intensities (Paterson and Warburton 2010). However, there are clearly benefits for

engaging in higher intensity physical activity and our data also establish that MVPA has a

protective effect on future cognitive impairment. In particular, higher intensity physical

activity that improves cardiorespiratory fitness confers particular benefits on health and brain

function (Erickson et al. 2014; Erickson et al. 2012).

The potential mechanisms by which physical activity confers a cognitive benefit in older age

are yet to be fully disentangled. One potential mechanism is through improving brain

structure and grey matter volume and in particular stimulating hippocampal neurogenesis,

with more recent evidence also suggesting the caudate nucleus and thalamus may be

positively impacted (Erickson et al. 2014; Erickson et al. 2012; Erickson et al. 2011; Kramer

and Erickson 2007). There is also accumulating evidence that participating in exercise may

improve cognitive outcomes through numerous biomarkers. In particular, recent data

ACCEPTED MANUSCRIPT

dehydroepiandrosterone may be associated with improved cognition and reduced dementia

risk following exercises in older age (Jensen et al. 2015; Maass et al. 2016). Clearly, future

research utilizing objective measures of physical activity are required to disentangle potential

mechanistic changes associated with cognitive status in older age.

Whilst our data are novel, one should note the observation nature of the data, which cannot

make claims regarding causality. The study follow up time was relatively short. Future

longitudinal research of longer duration is required to verify our findings. In addition, future

interventional work should seek to establish if changing physical activity levels can improve

cognitive (and other) outcomes. Another potential limitation is that objectively assessed

physical activity was only measured at baseline, which clearly limits the ability to explore the

reciprocal relationships between light physical activity and cognitive ability. Furthermore,

there is heterogeneity of comorbid conditions or health status in the population aged 65 or

above (e.g., different age groups). Future studies are encouraged to assess the relationships of

objectively measured physical activity with cognition, stratified by age groups. In addition,

we included some participants at baseline who had some degree of cognitive impairment,

which may influence the physical activity status at baseline. However, the independent

impact of light physical activity was evident in sensitivity analyses when such participants

ACCEPTED MANUSCRIPT

with and without such participants may actually increase representativeness, given the high

numbers of older people who are affected by cognitive impairment. This is the first

longitudinal paper to investigate light intensity physical activity and cognitive ability in older

adults. Strengths of our paper include the objective measurement of physical activity,

adjustment for multiple underlying confounders (e.g. baseline cognitive status, depressive

symptoms, and ADL difficulties) known to influence both physical activity and cognitive

ability, and test for reverse causation.

In conclusion, our data suggest that engagement in light intensity physical activity,

independent of MVPA, is associated with a reduced rate of cognitive ability decline in

community dwelling older adults. In addition, objectively assessed MVPA is associated with

better cognitive status. This extends the existing evidence for the benefits of physical

ACCEPTED MANUSCRIPT

The authors declare no conflicts of interest and acknowledge funding support from Taiwan

Ministry of Science and Technology (104-2410-H-018-028).

Conflict of interest statement

ACCEPTED MANUSCRIPT

Barnes, D.E.; Yaffe, K. The projected effect of risk factor reduction on Alzheimer's disease

prevalence. Lancet Neurol. 10:819-828; 2011

Blondell, S.J.; Hammersley-Mather, R.; Lennert Veerman, J. Does physical activity prevent

cognitive decline and dementia?: A systematic review and meta-analysis of

longitudinal studies. BMC Public Health. 14:1036-1061; 2014

Brink, T.L.; Yesavage, J.A.; Lum, O.; Heersema, P.H.; Adey, M.; Rose, T.L. Screening tests

for geriatric depression. Clin Gerontol. 1:37-43; 1982

Buchman, A.S.; Boyle, P.A.; Yu, L.; Shah, R.C.; Wilson, R.S.; Bennett, D.A. Total daily

physical activity and the risk of ad and cognitive decline in older adults. Neurology.

78:1323-1329; 2012

Buman, M.P.; Hekler, E.B.; Haskell, W.L.; Pruitt, L.; Conway, T.L.; Cain, K.L. Objective

light-intensity physical activity associations with rated health in older adults. Am J

Epidemiol. 172:1155-1165; 2010

Coley, N.; Andrieu, S.; Gardette, V.; Gillette-Guyonnet, S.; Sanz, C.; Vellas, B.; Grand, A.

Dementia prevention: Methodological explanations for inconsistent results.

Epidemiologics Review. 30:35-66; 2008

Deckers, K.; van Boxtel, M.P.; Schiepers, O.J.; de Vugt, M.; Munoz Sanchez, J.L.; Anstey,

ACCEPTED MANUSCRIPT

Yaffe, K.; Irving, K.; Verhey, F.R.; Kohler, S. Target risk factors for dementia

prevention: A systematic review and delphi consensus study on the evidence from

observational studies. Int J Geriatr Psychiatry. 30:234-246; 2015

Doi, T.; Makizako, H.; Shimada, H.; Tsutsumimoto, K.; Hotta, R.; Nakakubo, S.; Park, H.;

Suzuki, T. Objectively measured physical activity, brain atrophy, and white matter

lesions in older adults with mild cognitive impairment. Exp Gerontol. 62:1-6; 2015

Erickson, K.I.; Leckie, R.L.; Weinstein, A.M. Physical activity, fitness, and gray matter

volume. Neurobiol Aging. 35:S20-S28; 2014

Erickson, K.I.; Miller, D.L.; Weinstein, A.M.; Akl, S.L.; Banducci, S.E. Physical activity and

brain plasticity in late adulthood: A conceptual review. Ageing Research. 3:34-47;

2012

Erickson, K.I.; Voss, M.W.; Prakash, R.S.; Basak, C.; Szabo, A.; Chaddock, L.; Kim, J.S.;

Heo, S.; Alves, H.; White, S.M.; Wojcicki, T.R.; Mailey, E.; Vieira, V.J.; Martin, S.A.;

Pence, B.D.; Woods, J.A.; McAuley, E.; Kramer, A.F. Exercise training increases size

of hippocampus and improves memory. PNAS Proceedings of the National Academy

of Sciences of the United States of America. 108:3017-3022; 2011

Falck, R.S.; McDonald, S.M.; Beets, M.W.; Brazendale, K.; Liu-Ambrose, T. Measurement of

physical activity in older adult interventions: A systematic review. Br J Sports Med.

ACCEPTED MANUSCRIPT

Farren, L.; Belza, B.; Allen, P.; Brolliar, S.; Brown, D.R.; Cormier, M.L.; Janicek, S.; Jones,

D.L.; King, D.K.; Marquez, D.X.; Rosenberg, D.E. Mall walking program

environments, features, and participants: A scoping review. Prev Chronic Dis. 12:E129;

2015

Fiest, K.M.; Jette, N.; Roberts, J.I.; Maxwell, C.J.; Smith, E.E.; Black, S.E.; Blaikie, L.;

Cohen, A.; Day, L.; Holroyd-Leduc, J.; Kirk, A.; Pearson, D.; Pringsheim, T.;

Venegas-Torres, A.; Hogan, D.B. The prevalence and incidence of dementia: A

systematic review and meta-analysis. Can J Neurol Sci. 43 Suppl 1:S3-s50; 2016

Galvin, J.; Roe, C.; Powlishta, K.; Coats, M.; Muich, S.; Grant, E.; Miller, J.; Storandt, M.;

Morris, J. The AD8: A brief informant interview to detect dementia. Neurology.

65:559; 2005

Galvin, J.E.; Roe, C.M.; Coats, M.A.; Morris, J.C. Patient's rating of cognitive ability: Using

the AD8, a brief informant interview, as a self-rating tool to detect dementia. Arch

Neurol. 64:725-730; 2007

Ganguli, M.; Lee, C.-W.; Snitz, B.; Hughes, T.; McDade, E.M.; Chang, C.-C.H. How well do

MCI criteria predict progression to severe cognitive impairment and dementia?

Alzheimer Dis Assoc Disord. 28:113-121; 2014

Gorman, E.; Hanson, H.; Yang, P.; Khan, K.; Liu-Ambrose, T.; Ashe, M. Accelerometry

ACCEPTED MANUSCRIPT

Guo, N.W.; Liu, H.C.; Wong, P.F.; Liao, K.K.; Yan, S.H.; Lin, K.P.; Chang, C.Y.; Hsu, T.C.

Chinese version and norms of the Mini-Mental State Examination. Journal of

Rehabilitation Medicine Association:52-59; 1988

Hamer, M.; Chida, Y. Physical activity and risk of neurodegenerative disease: A systematic

review of prospective evidence. Psychol Med. 39:3-11; 2009

Hamer, M.; Coombs, N.; Stamatakis, E. Associations between objectively assessed and

self-reported sedentary time with mental health in adults: An analysis of data from the

Health Survey for England. BMJ Open. 4; 2014

Hanson, S.; Jones, A. Is there evidence that walking groups have health benefits? A systematic

review and meta-analysis. Br J Sports Med. 49:710-715; 2015

Hosmer, D.W.; Lemeshow, S.; Sturdivant, R.X. Applied logistic regression. NJ: John Wiley &

Sons; 2013

Jensen, C.S.; Hasselbalch, S.G.; Waldemar, G.; Simonsen, A.H. Biochemical markers of

physical exercise on mild cognitive impairment and dementia: Systematic review and

perspectives. Front Neurol. 6:187; 2015

Kramer, A.F.; Erickson, K.I. Capitalizing on cortical plasticity: Influence of physical activity

on cognition and brain function. Trends in Cognitive Sciences. 11:342-348; 2007

ACCEPTED MANUSCRIPT

assessed physical activity at different intensities with subjective well-being in older

adults. Quality of Life Research. 25:2909–2919; 2016

Maass, A.; Duzel, S.; Brigadski, T.; Goerke, M.; Becke, A.; Sobieray, U.; Neumann, K.;

Lovden, M.; Lindenberger, U.; Backman, L.; Braun-Dullaeus, R.; Ahrens, D.; Heinze,

H.J.; Muller, N.G.; Lessmann, V.; Sendtner, M.; Duzel, E. Relationships of peripheral

igf-1, vegf and bdnf levels to exercise-related changes in memory, hippocampal

perfusion and volumes in older adults. Neuroimage. 131:142-154; 2016

Makizako, H.; Liu-Ambrose, T.; Shimada, H.; Doi, T.; Park, H.; Tsutsumimoto, K.; Uemura,

K.; Suzuki, T. Moderate-intensity physical activity, hippocampal volume, and memory

in older adults with mild cognitive impairment. J Gerontol A Biol Sci Med Sci.

70:480-486; 2015

McMinn, D.; Acharya, R.; Rowe, D.A.; Gray, S.R.; Allan, J.L. Measuring activity energy

expenditure: Accuracy of the GT3X+ and Actiheart monitors. Int J Exerc Sci.

6:217-229; 2013

Paterson, D.H.; Warburton, D.E. Physical activity and functional limitations in older adults: A

systematic review related to Canada's physical activity guidelines. Int J Behav Nutr

Phys Act. 7:38; 2010

Prince, M.; Bryce, R.; Albanese, E.; Wimo, A.; Ribeiro, W.; Ferri, C.P. The global prevalence

ACCEPTED MANUSCRIPT

Taiwan Department of Health. Identification, evaluation, and treatment of overweight and

obesity in adults in Taiwan. Taipei: Taiwan Department of Health; 2003

Taiwan Ministry of Interior. Monthly bulletin of interior statistics: Resident population by age

group. Taipei: Taiwan Ministry of Interior; 2012

Tse, A.; Wong, T.W.; Lee, P.H. Effect of low-intensity exercise on physical and cognitive

health in older adults: A systematic review. Sports Medicine-Open. 1:1-13; 2015

Tudor-Locke, C.; Leonardi, C.; Johnson, W.D.; Katzmarzyk, P.T.; Church, T.S. Accelerometer

steps/day translation of moderate-to-vigorous activity. Prev Med. 53:31-33; 2011

World Health Organization. Global recommendations on physical activity for health. Geneva:

World Health Organization; 2010

Yang, Y.-H.; Galvin, J.E.; Morris, J.C.; Lai, C.-L.; Chou, M.-C.; Liu, C.-K. Application of

AD8 questionnaire to screen very mild dementia in Taiwanese. Am J Alzheimer's Dis

Other Dementias. 26:134-138; 2011

Yesavage, J.A.; Sheikh, J.I. Geriatric depression scale (gds): Recent evidence and

development of a shorter version. in: Brink T.L., ed. Clinical gerontologist: A guide to

assessment and intervention. New York: The Haworth Press; 1986

ACCEPTED MANUSCRIPT

29

Highlight Points

 A paucity of studies have considered objective longitudinal physical activity and future cognitive ability decline in older adults.

 Our data suggests that higher light physical activity, independent of

moderate-to-vigorous physical activity, is associated with a reduced risk of future

cognitive decline.

 This is the first longitudinal paper to investigate light intensity physical activity and cognitive decline in older adults.

 This extends the existing evidence for the benefits of physical activity for preventing cognitive deterioration in later life.