The descriptive epidemiology of accelerometer-measured physical activity in older adults
Berkemeyer et al. International Journal of Behavioral Nutrition and
The descriptive epidemiology of accelerometer-measured physical activity in older adults
K. Berkemeyer 0
K. Wijndaele 0
T. White 0
A. J. M. Cooper 0
R. Luben 2
K. Westgate 0
S. J. Griffin 0
K. T. Khaw 1
N. J. Wareham 0
S. Brage 0
0 MRC Epidemiology Unit, University of Cambridge, School of Clinical Medicine, Box 285 Institute of Metabolic Science , Cambridge Biomedical Campus, Cambridge CB2 0QQ , UK
1 Department of Gerontology, University of Cambridge, School of Clinical Medicine, Addenbrooke's Hospital , Cambridge CB2 2QQ , UK
2 Strangeways Research Laboratory, University of Cambridge , Worts Causeway, Cambridge CB1 8RN , UK
Background: Objectively measured physical activity between older individuals and between populations has been poorly described. We aimed to describe and compare the variation in accelerometry data in older UK (EPIC-Norfolk) and American (NHANES) adults. Methods: Physical activity was measured by uniaxial accelerometry in 4,052 UK (49-91 years) and 3459 US older adults (49-85 years). We summarized physical activity as volume (average counts/minute), its underlying intensity distribution, and as time spent <100counts/minute, ≥809counts/minute and ≥2020counts/minute both for total activity and that undertaken in ≥10-min bouts. Results: In EPIC-Norfolk 65 % of wear-time was spent at <100 counts/minute and 20 % spent in the range 100-500 counts/minute. Only 4.1 % of this cohort accumulated more than 30 min/day of activity above 2020 counts/minute in 10-min bouts. If a cut-point of >809 counts/minute is used 18.7 % of people reached the 30 min/day threshold. By comparison, 2.5 % and 9.5 % of American older adults accumulated activity at these levels, respectively. Conclusion: As assessed by objectively measured physical activity, the majority of older adults in this UK study did not meet current activity guidelines. Older adults in the UK were more active overall, but also spent more time being sedentary than US adults.
Accelerometry; Activity intensity distribution; Older adults; Guidelines
Physical activity changes throughout the life-course,
especially in the transition to older adulthood [
Given the importance of physical activity for the primary
and secondary prevention of major chronic diseases and
maintenance of independence in older age [
], it is
important to understand the levels and patterns of physical
activity in this age group.
The majority of population-based studies have used
self-report methods to assess physical activity, which
have a number of limitations including problems of
recall and reporting bias [
]. These limitations may
be exacerbated in older populations, in whom cognitive
impairment is more likely [
]. In addition, due to the
ubiquitous nature of low intensity activities and sedentary
behaviors, these tend to be particularly difficult to recall
]; however, such behaviors may be of particular
importance in older adults as low intensity activities are more
prevalent in this age group [
]. Objective physical activity
assessment methods, such as accelerometry, enable more
accurate assessment of the entire movement intensity
spectrum and may therefore have particular merits in
There are, however, challenges in interpreting data
from objective physical activity measurement methods.
Several of these arise from classifying activity into broad
intensity categories, often labeled as ‘sedentary, light,
moderate, and vigorous’, rather than using the full
continuous distribution of intensity. For accelerometry data,
the cut-points used to define these classical intensity
categories are typically based on calibration studies using
energy expenditure as a criterion method during specific
lab-based activities, which may or may not apply to all
age groups and populations [
Physical activity guidelines suggest that older adults
should be ‘as active as they can’ while aiming to achieve
the recommended physical activity levels for adults, i.e.,
30 min of moderate-to-vigorous physical activity (MVPA)
per day in bouts of ≥10 min [
]. The evidence base for
the bout requirement, however, is limited and it is also
uncertain how to appropriately define intensity, especially in
an older age group  in whom there is large
interindividual variation in the intensity associated with a given
activity. Therefore, a single cut-point above which an
activity might be regarded as ‘moderate’ for older adults is
difficult to establish, which explains the diversity of
cut-points used to define MVPA [
Activities of at least moderate intensity have been the
main focus of several studies objectively describing
physical activity in older adults [
3, 8, 11
], but the focus on
this specific cut-point may disregard a whole range of
relevant activities of lesser intensity that could be
important contributors to overall physical health and
]. While a number of studies have described
time spent at different activity intensities in older
], to our knowledge the continuous
distribution of activity intensity in this age group, using
objective activity assessment, has only been described in
the British 1946 birth cohort, showing that activity time
is inversely associated with intensity [
Another area of uncertainty is how populations from
different countries compare in terms of key activity
characteristics. International comparisons of physical
activity levels may point to certain determinants of
physical activity and are essential to guide global public
health promotion. So far, however, only a few studies in
adults have used objective accelerometry data and
implemented the same processing procedures to maximize
Therefore, the main aim of this study was to provide a
description of objectively measured physical activity in
older UK adults, both in terms of overall physical activity
volume and time distribution across the entire physical
activity intensity spectrum. We also compared these
estimates to US adults of a similar age from the NHANES
study using the same accelerometry processing procedures
in both cohorts.
Participants and protocol
The Norfolk arm of the European Prospective
Investigation of Cancer (EPIC-Norfolk) has been described in detail
]. In brief, 25,639 men and women aged 40–
79 years, who were recruited from 35 general practices in
the area of Norfolk, consented to participate in the
EPICNorfolk study and attended a first health check between
1993–1997. A 2nd health check was conducted 1997–2000
at which physical activity was measured by self-report
only. Between 2006 and 2011, 8,623 men and women
participated in the 3rd health check and a subsample of 4,148
were asked to wear an accelerometer (ActiGraph model
GT1M, ActiGraph, Pensacola, FL) for the assessment of
free-living physical activity [
At the 3rd health check visit, height and weight were
measured using standard anthropometric techniques.
Employment status, education level and self-rated health
were assessed by self-report questionnaires. At the
end of the clinical visit, participants were asked to
wear the accelerometer on a belt around the right hip
during all waking hours for 7 days, and to only take
it off for water-based activities (e.g., when showering)
and during sleep. Accelerometers were set up to capture
data in 5-s epochs.
To compare objectively measured physical activity levels
between UK and American older adult populations, a
subsample of 3459 participants of the NHANES 2003–2004
cohort was used, excluding participants aged <49 years
(the lowest age in the EPIC-Norfolk 3rd health check).
NHANES is a national population-based study of the
civilian, non-institutionalized US population aged ≥6 years,
as previously described in detail [
]. As in the
EPICNorfolk protocol, NHANES 2003–2004 participants were
asked to wear an accelerometer (ActiGraph AM7164) for
7 consecutive days on a belt around the hip, initialized to
capture data in 60-s epochs. Accelerometer records
deemed to be out-of-calibration or unreliable by the
NHANES team were excluded.
Accelerometry data processing
To maximize comparability between studies, we
processed the accelerometry data of both studies using the
same criteria, after first collapsing the 5-s EPIC-Norfolk
data to 60-s epoch resolution. The two models of
ActiGraph use either a piezo-electric (AM7164) or a
MEMSbased (GT1M) acceleration sensor, both of which have a
frequency-dependent response to acceleration magnitude
]. Both have been shown to be approximately
comparable in the human movement range, except at
the extreme low end ; this necessitates a slight
revision of the wear/non-wear classification compared
to what has been used elsewhere for AM7164 data
]. Continuous zero strings of ≥90 min were
defined as non-wear time, which is similar to other
studies using MEMS-based Actigraph versions [
for this age-group this threshold resulted in a more
realistic pattern of 2–4 wear/ non-wear transitions per day,
compared to using zero strings ≥60 min (see example in
Additional file 1: Figure S1). One would expect at least
two transitions from getting up (non-wear to wear) and
going to bed (wear to non-wear). A valid day was defined
as ≥600 min of wear time . Only participants with ≥4
valid days of data were included in analyses. As more than
200 participants (mainly in the NHANES) were found to
have acceleration data for longer than 19 h/day (indicating
monitor wear during sleep), we truncated wear-time to
19 h/day and time spent in the <100 counts/minute
intensity category for both cohorts to normalize the data.
We derived overall physical activity volume, defined
as total counts divided by wear time, as well as the
movement intensity distribution as a continuous construct
in 15 systematic intervals. In addition, we report time
spent <100 counts/minute, as well as time spent above the
commonly used MVPA cut-points of 2020 counts/minute
] and 809 counts/minute [
]. The latter cut-point was
identified as the lower boundary for MVPA in a validation
study of 20 older adults (aged 60–90 years) performing
treadmill-based and free-living walking activities
around 3 METs [
]. To indicate the degree of activity
accumulation occurring in bouts, we also analyzed time
spent ≥2020 and ≥809 counts/minute in continuous bouts
lasting ≥10 min.
Due to the non-normal distribution of physical activity,
most descriptive characteristics are described using
medians and interquartile ranges (IQR) and some using
means and standard deviation (i.e., age). Univariate
non-parametric tests for trend were used to examine
differences in physical activity by age group, employment
status, education level, BMI, and self-rated health, in
men and women separately. We also examined the
robustness of these differences by mutually adjusting
for all stratifiers. Stata (StataCorp LP) version 13.1 was
used for all analyses.
A total of 4,052 EPIC-Norfolk participants had ≥4 valid
days (>10 h/day) of accelerometer data and were
included in analyses (98.0 % of the whole accelerometry
sample, Table 1). Mean accelerometer wear-time was
14.5 h/day (SD 1.1 h/day) and was slightly higher in men
than women (14.7 vs 14.4 h/day, P < 0.05). Participants
with valid accelerometer data did not differ significantly
from those who did not wear an accelerometer with
respect to age, sex, BMI, education level and self-rated
health (Additional file 2: Table S1).
Total physical activity volume (average movement
intensity, counts/minute) did not differ between men and
women (254.8 vs 254.0, p = 0.8). As shown in Table 2,
physical activity volume was significantly lower across
increasing strata of age and BMI and across decreasing
strata for self-rated health and education, in both men
and women. Physical activity volume was also significantly
higher in those who reported paid employment compared
to those who were not employed or were retired. A
Absolute numbers do not add up to the column total due to missing data
comparable pattern of association with all stratifiers was
found in both sexes for time spent ≥809 and ≥2020 counts/
minute, irrespective of whether this was accumulated in
bouts of ≥10 min or not. Time spent <100 counts/minute
followed the opposite pattern for age, BMI and
selfreported health, showing increasingly higher levels of
sedentary time across increasing strata for age and BMI
and across decreasing strata for self-reported health.
Participants who reported being employed spent less time
sedentary compared to those not employed. The
association between sedentary time and education level was less
strong, and not significant in women. Mutual adjustment
for stratifiers indicated that the strongest factors
associated with physical activity volume and time spent at the
lowest (<100 counts/minute) and highest physical activity
intensities (≥809 or ≥2020 counts/minute) were age, BMI
and self-rated health (P < 0.05).
The movement intensity distribution, expressed in 15
intervals of 100 counts/minute width, is presented in
Fig. 1. As expected, the distribution is positively skewed
with the vast majority of activity occurring in the lower
intensity ranges. Most recorded time (64.5 %) was spent
being sedentary (<100 counts/minute). The majority of
the remaining non-sedentary time (18.2 % in men,
21.2 % in women) was of very light intensity between
100 and 500 counts/minute. Only about 2.7 % of all
activity in men and 2.2 % of all activity in women occurred
above the cut-point of 2000 counts/minute. There were
differences between men and women for activity
intensities. Men spent 44 min/day longer sedentary whereas
women spent more time in all categories up to 2000
counts/minute. However, men spent more time than
women with activity greater than 2000 counts/minute
(P < 0.001).
Comparison between EPIC-Norfolk and NHANES
A total of 3459 NHANES participants met the
accelerometry inclusion criteria and were compared to the EPIC
Norfolk sample. Mean (SD) age of included NHANES
participants was 66.4 (10.1) years and 50.7 % were men,
comparable to the EPIC-Norfolk cohort (Table 1). Weight
status was comparable between the two populations, with
72 % and 64 % of participants being overweight or obese
in NHANES and EPIC-Norfolk, respectively. In terms of
education, 31.1 % of the NHANES participants had less
than a High School degree (24.7 % High School
graduates), whereas 37.3 % of EPIC-Norfolk participants were
educated to less than A-level (45.3 % A-level graduates).
Monitor wear-time did not differ between the two
studies (EPIC-Norfolk: 872 min/day, NHANES: 871 min/
day, P < 0.05). Activity volume was higher in the UK
sample compared to the US (238.2 vs 205.2 counts/
minute, P < 0.05). As shown in Fig. 2, EPIC-Norfolk
participants spent 13 min/day more being sedentary
than NHANES participants (566.3 min/day vs 553.4 min/
day, P < 0.05), NHANES participants spent more time in
the lower intensity categories than EPIC-Norfolk
participants. EPIC-Norfolk participants spent more time than
those in NHANES undertaking physical activity >750
The adherence to physical activity guidelines for older
adults, as determined by ≥30 min/day of movement
intensity ≥2020 counts/minute undertaken in bouts
of ≥10 min, was low in both cohorts, with 4.1 % of all
participants in EPIC-Norfolk and 2.5 % in NHANES
accumulating activity at this level. Using the cut-point of ≥809
counts/minute, the adherence level was 18.7 % in
EPIC-Norfolk and 9.5 % in NHANES. When counting all
activity ≥2020 counts/minute whether or not it was
undertaken in bouts lasting more than 10 min, 26.7 % of
EPIC-Norfolk participants and 13.9 % of NHANES
participants spent ≥30 min/day in this intensity spectrum. The
equivalent numbers for the ≥809 counts/minute definition
were 86.8 % (EPIC-Norfolk) and 73.4 % (NHANES).
Figures 3a-c show the minutes of activity spent <100,
≥809 and ≥2020 counts/minute for EPIC-Norfolk and
NHANES participants, stratified by sex and age group.
Overall, sedentary time (Fig. 3a) was lower in women
than in men in both cohorts. In EPIC-Norfolk, men and
women below the age of 60 years and between 60 and
70 years of age spent less time sedentary compared to
the older age groups. However, overall the differences
between age groups were greater in NHANES than in
EPIC-Norfolk participants. The population variance of
sedentary time was larger in the NHANES study.
EPICNorfolk men and women of all ages generally spent
more time above 809 and 2020 counts/minute than the
NHANES participants (Fig. 3b & c) with the exception
of men in the youngest age group in NHANES.
The aim of this study was to provide a detailed
description of objectively measured physical activity in older
adults by socio-demographic characteristics, BMI and
self-rated health, as well as to directly compare similarly
aged older adults from the UK and the US. We have
shown that the continuous physical activity intensity
distribution is patterned differently, both within and
between populations. Women from the EPIC-Norfolk
study accumulate their overall activity through activities
of lower intensities while men accumulate a similar
activity volume by spending more time sedentary as well
Fig. 3 Median and IQR of Average Daily Physical Activity between 0 and 100 counts/minute, ≥809 counts/minute and ≥2020 counts/minute,
stratified by Gender, Age Category and Cohort. a. Time between 0 and 100 counts/minute (adjusted for non-normal awake wear time). b. Time
above 809 counts/minute. c. Time above 2020 counts/minute
as in activity intensities above 2000 counts/minute.
When examining differences by age, we found a stronger
decline in higher intensity activity when comparing UK
participants aged <70 years with those aged >70 years,
which was accompanied by a more pronounced increase
in sedentary time. BMI and self-rated health appeared as
additional important correlates. In comparison with the
US, the UK older adults generally performed more activity
of higher intensity (≥809 and ≥2020 counts/minute), and
showed a less steep decline in higher intensity activity
across age groups, especially for activity ≥809
counts/minute. The EPIC-Norfolk population also displayed a less
steep increase in sedentary time across increasing age
groups, which was driven in part by the younger
participants having greater sedentary time than in NHANES.
Recently, Jefferis et al. [
] demonstrated that 10 % and
15 % of older British women and men >70 years
accumulated the recommended level of ≥30 min/day of
MVPA in ≥10 min bouts, using an accelerometer
threshold of >1040 counts/minute, which was reduced to 3 %
and 7 % respectively when they used a threshold of >1952
counts/minute. These estimates are comparable to our
prevalence estimates in EPIC-Norfolk 16.1 % in women
and 22.2 % in men for the ≥809 counts/minute cut-point
and 3.2 % in women and 5.4 % in men for the ≥2020
counts/minute cut-point, respectively when using a
10min minimal bout duration criterion. Compared with
Swedish adults aged 60–75 years old who spent on
average 29 (men) or 23 (women) minutes/day in activity
intensity ≥2020 counts/minute, EPIC-Norfolk men and women
accumulated less time in this intensity category; 23.6 and
18.9 min/day, respectively. Furthermore, in Swedish older
adults, around 55 % of the time spent ≥2020
counts/minute was accumulated in bouts longer than 10 min,
whereas in EPIC-Norfolk participants only around 31 % of
time ≥2020 counts/minute was spent in 10-min bouts [
Different processing criteria obviously result in stark
differences in prevalence estimates when considering all
MVPA time, regardless of occurrence in bouts, e.g., 19 %
and 56 % of women accumulated ≥30 min/day of MVPA
at the higher and lower MVPA cut-point in the study by
Jefferis et al., with corresponding numbers in men being
27 % and 62 % [
]. This concept of time spent in MVPA
regardless of durations of the bouts was also examined in
the British 1946 birth cohort of older adults in 2006–
2010, in which 43 % of women and 60 % of men
accumulated over 30 min/day above 3 METs, as estimated by
combined heart rate and movement sensing [
equivalent percentage of the population spending more
than 30 min/day above the cut-points used in
EPICNorfolk (≥809 and ≥2020 counts/minute) were 86.8 % and
26.7 %, respectively.
The differences that we have observed between strata
and studies for achievement of physical activity targets
raise the question about whether a cut-point of 2020
counts/minute is appropriate to characterize “moderate”
activity as defined in physical activity guidelines.
Differences in the degree of adherence to recommendations
between men and women or between age groups can be
explained by variation in how activity time is spent.
However, a number of studies have demonstrated that
there are health benefits of physical activity in the light
intensity range below 2020 counts/minute [
]. It is
currently difficult to define a lower boundary for the
minimum amount of activity that is beneficial for health. In
the future, revision of guidelines will need to be based
on epidemiological data, which includes the full range of
movement intensity, rather than data that have been
condensed into broad categories. This is a particular
issue in older participants for whom activity of lower
intensity may be especially important.
This study has several strengths. Firstly, it provides a
high-resolution description of the full intensity
distribution of physical activity, as well as a detailed description
of higher intensity physical activity, the accumulation of
bouts of physical activity at two different levels of intensity
and sedentary time in EPIC-Norfolk and NHANES.
Secondly, with more than 4000 participants, the
EPICNorfolk cohort is the largest cohort in UK adults in which
physical activity was measured by hip accelerometry.
Finally, implementation of the same processing procedures
allowed a direct comparison between the two cohorts.
There are also limitations of this study that need to be
considered when interpreting results. At baseline,
EPICNorfolk participants were found to be representative of
the Health Survey for England population for age and
sex as well as all measured anthropometric
]. However, participants who returned to the 3rd
health check were, on average, younger and had a
healthier cardiovascular risk profile and higher
socioeconomic status than the baseline population, although
the sample still represents a variety of socio-economic
backgrounds and lifestyle factors [
]. A comparison of
basic socio-demographic characteristics between
participants with and without accelerometry at the 3rd health
check did not show any difference between the main
socio-demographic and health characteristics.
Furthermore, as this was the first time point at which physical
activity was measured objectively in this cohort, the data
are cross-sectional and therefore yield no information
on within-individual change of activity with age. It is
possible that not wearing the monitor during sleep or
water-based activities, and the use of an automated
nonwear algorithm may have resulted in some
misclassification between non-wear and sedentary time. In addition,
we have collapsed the raw EPIC data in 5-s resolution to
60-s resolution for the purpose of comparison to other
population data; intensity estimates using 5-s data are
likely to differ [
]. The GT1M model of the
accelerometer used in EPIC-Norfolk is less sensitive at the very
low end of the movement continuum than the model
used in NHANES, leading to potential differential
properties of the zero-strings used to classify non-wear [
which additionally supports our decision to relax the
zero-string criteria for defining non-wear time from 60 to
90 min as this produced similar wear time estimates
between cohorts. Non-normally long wear-times were found,
mainly in the NHANES dataset, leading to a possible
overestimation of wear-time and time spent in the
lowest intensity category. Therefore, sedentary time was
normalized for participants with a wear-time >19 h/day in
both cohorts. Finally, activities which are less well
detected by hip-accelerometry (e.g., cycling) may have a
different prevalence between UK and US older adults, in
this way potentially biasing our comparison for overall
physical activity volume and higher intensity activity
between the cohorts [
The semi-continuous distribution of movement intensity
showed that UK women spent significantly more time in
lower intensity physical activity categories whereas men
spent more time being sedentary and in high intensity
physical activity categories, while overall physical activity
volume was not significantly different between sexes. As
a large proportion of time is spent in the light intensity
range, the entire movement distribution should be
considered when analyzing physical activity as it is possible
that activities of this intensity may also confer health
benefits in addition to those already established for
higher intensity. Finally, this study also showed that
there are differences between UK and US older adults,
with the latter having lower overall activity volume and
less time in higher intensity physical activity, as well as a
steeper age gradient in time spent in higher intensity
physical activity. However, and maybe contrary to
expectations, American older adults spent less time
being sedentary. This suggests that effective physical
activity promotion in older adults should specifically
aim to delay the decline of physical activity with age.
With this in mind, consideration of the whole
intensity distribution may aid our ability to detect subtle
but important effects of physical activity promotion in
the globally growing population of older people.
Additional file 1: Figure S1. Example accelerometer file showing time
segments classified as non-wear using the 90-min and the 60-min zero
string criterions. (JPG 80 kb)
Additional file 2: Table S1. Differences between participants with and
without accelerometry at 3rd health check, EPIC-Norfolk. (XLSX 17 kb)
BMI: Body Mass Index; EPIC: European Prospective Investigation into Cancer;
MET: Metabolic equivalent of task; MVPA: Moderate-to-vigorous physical
activity; NHANES: National Health and Nutrition Examination Survey.
The authors declare that they have no compering interests.
KB conducted all analyses and wrote the paper. KWi and SB wrote parts of
the paper and gave substantial input at all stages of research. KB, KWi and SB
interpreted the results. KWe and TW contributed to the analysis of
accelerometry data. AJMC, KWe, TW and NJW gave critical input on the
manuscript. RL was responsible for EPIC-Norfolk data management. All
authors read and approved the final version of the manuscript.
We thank all EPIC-Norfolk and NHANES participants and data collection
teams. Special thanks to Amit Bhaniani for assisting with data management
for EPIC-Norfolk data.
This work was supported by programme grants from the Medical Research
Council [G9502233; G0401527] and Cancer Research UK [C864/A8257]. A
grant from Research into Ageing  funded the 3rd health check clinic.
KW is supported by a British Heart Foundation Intermediate Basic Science
Research Fellowship [FS/12/58/29709], and AJMC, SJG, NJW, and SB are
supported by MRC programme grants [MC_UU_12015/3 and
Submit your next manuscript to BioMed Central
and we will help you at every step:
1. Aresu M , Bécares L , Brage S , Chaudhury M , Doyle-Francis M , Esliger D , et al. Health Survey for England 2008 : Volume 1 Physical activity and fitness . NHS, The Information Centre for Health and Social Care . 2011 . p. 894 - 910 . Available from: http://www.hscic.gov.uk/pubs/hse08physicalactivity.
2. Department of Health. Start Active, Stay Active: A report on physical activity for health from the four home countries' Chief Medical Officers . Strategy . 2011 . https://www.gov.uk/government/publications/start -active-stay-active-a-reporton-physical-activity-from-thefour-home-countries-chief-medical-officers.
3. Sun F , Norman IJ , While AE . Physical activity in older people: a systematic review . BMC Public Health . 2013 ; 13 ( January 2000 ): 449 .
4. Rait G , Fletcher A , Smeeth L , Brayne C , Stirling S , Nunes M , et al. Prevalence of cognitive impairment: Results from the MRC trial of assessment and management of older people in the community . Age Ageing . 2005 ; 34 ( 3 ): 242 - 8 .
5. Wareham NJ , Rennie KL . The assessment of physical activity in individuals and populations: why try to be more precise about how physical activity is assessed? Int J Obes Relat Metab Disord . 1998 ; 22 Suppl 2 : S30 - 8 .
6. Janz KF . Physical activity in epidemiology: moving from questionnaire to objective measurement . Br J Sports Med . 2006 ; 40 ( 3 ): 191 - 2 .
7. NHS Choices . Factsheet 5: Physical activity guidelines for Older Adults (65+ years) . NHS Choices ; 2013 . Available from: http://www.hscic.gov.uk/ pubs/hse08physicalactivity.
8. Jefferis BJ , Sartini C , Lee I-M , Choi M , Amuzu A , Gutierrez C , et al. Adherence to physical activity guidelines in older adults, using objectively measured physical activity in a population-based study . BMC Public Health . 2014 ; 14 ( 1 ): 382 .
9. Hagströmer M , Troiano RP , Sjöström M , Berrigan D . Levels and patterns of objectively assessed physical activity-a comparison between Sweden and the United States . Am J Epidemiol . 2010 ; 171 ( 10 ): 1055 - 64 .
10. Troiano RP , Berrigan D , Dodd KW , Mâsse LC , Tilert T , Mcdowell M . Physical activity in the United States measured by accelerometer . Med Sci Sports Exerc . 2008 ; 40 ( 1 ): 181 - 8 .
11. Tucker JM , Welk GJ , Beyler NK . Physical activity in U.S. adults: Compliance with the physical activity guidelines for Americans . Am J Prev Med . 2011 ; 40 ( 4 ): 454 - 61 .
12. Buman MP , Hekler EB , Haskell WL , Pruitt L , Conway TL , Cain KL , et al. Objective light-intensity physical activity associations with rated health in older adults . Am J Epidemiol . 2010 ; 172 ( 10 ): 1155 - 65 .
13. Lohne-Seiler H , Hansen BH , Kolle E , Anderssen S a. Accelerometer-determined physical activity and self-reported health in a population of older adults (65-85 years): a cross-sectional study . BMC Public Health . 2014 ; 14 ( 1 ): 284 .
14. Evenson KR , Buchner DM , Morland KB . Objective measurement of physical activity and sedentary behavior among US adults aged 60 years or older . Prev Chronic Dis . 2012 ; 9 ( 2 ): E26 .
15. Copeland JL , Esliger DW . Accelerometer assessment of physical activity in active, healthy older adults . J Aging Phys Act . 2009 ; 17 ( 1 ): 17 - 30 .
16. Davis MG , Fox KR . Physical activity patterns assessed by accelerometry in older people . Eur J Appl Physiol . 2007 ; 100 ( 5 ): 581 - 9 .
17. Golubic R , Martin KR , Ekelund U , Hardy R , Kuh D , Wareham N , et al. Levels of physical activity among a nationally representative sample of people in early old age: results of objective and self-reported assessments . Int J Behav Nutr Phys Act . 2014 ; 11 ( 1 ): 58 .
18. Corder K , Van Sluijs EMF. Invited commentary: Comparing physical activity across countries-current strengths and weaknesses . Am J Epidemiol . 2010 ; 171 ( 10 ): 1065 - 8 .
19. Hayat S a , Luben R , Keevil VL , Moore S , Dalzell N , Bhaniani A , et al. Cohort profile: A prospective cohort study of objective physical and cognitive capability and visual health in an ageing population of men and women in Norfolk (EPIC-Norfolk 3) . Int J Epidemiol . 2014 ; 43 ( 4 ): 1063 - 72 .
20. Atienza A a , Moser RP , Perna F , Dodd K , Ballard-Barbash R , Troiano RP , et al. Self-reported and objectively measured activity related to biomarkers using NHANES . Med Sci Sports Exerc . 2011 ; 43 ( 5 ): 815 - 21 .
21. Brage S , Brage N , Wedderkopp N , Froberg K. Reliability and Validity of the Computer Science and Applications Accelerometer Model 7164 in a Mechanical Setting . Meas Phys Educ Exerc Sci . 2003 ; 7 ( 2 ): 101 - 19 .
22. Rothney MP , Apker GA , Song Y , Chen KY . Comparing the performance of three generations of ActiGraph accelerometers . J Appl Physiol . 2008 ; 105 ( 4 ): 1091 - 7 .
23. Ried-Larsen M , Brønd JC , Brage S , Hansen BH , Grydeland M , Andersen LB , et al. Mechanical and free living comparisons of four generations of the Actigraph activity monitor . Int J Behav Nutr Phys Act . 2012 ; 9 : 113 . PMCID: PMC3463450 doi: 10.1186/ 1479 -5868-9-113.
24. Davis MG , Fox KR , Hillsdon M , Sharp DJ , Coulson JC , Thompson JL . Objectively measured physical activity in a diverse sample of older urban UK adults . Med Sci Sports Exerc . 2011 ; 43 ( 4 ): 647 - 54 .
25. Orme M , Wijndaele K , Sharp SJ , Westgate K , Ekelund U , Brage S . Combined influence of epoch length, cut-point and bout duration on accelerometryderived physical activity . Int J Behav Nutr Phys Act . 2014 ; 11 ( 1 ): 34 .
26. Hall KS , Howe C a , Rana SR , Martin CL , Morey MC . METs and accelerometry of walking in older adults: Standard versus measured energy cost . Med Sci Sports Exerc . 2013 ; 45 ( 3 ): 574 - 82 .
27. Ensrud KE , Blackwell TL , Cauley J a, Dam T-TL , Cawthon PM , Schousboe JT , et al. Objective Measures of Activity Level and Mortality in Older Men . J Am Geriatr Soc . 2014 ; 62 ( 11 ): 2079 - 87 .
28. Day N , Oakes S , Luben R , Khaw KT , Bingham S , Welch A , et al. EPIC-Norfolk: study design and characteristics of the cohort . European Prospective Investigation of Cancer. Br J Cancer . 1999 ; 80 Suppl 1 : 95 - 103 .
29. Corder K , Brage S , Ramachandran A , Snehalatha C , Wareham N , Ekelund U . Comparison of two Actigraph models for assessing free-living physical activity in Indian adolescents . J Sports Sci . 2007 ; 25 ( 14 ): 1607 - 11 .
30. Bassett DR , Pucher J , Buehler R , Thompson DL , Crouter SE . Walking, cycling, and obesity rates in Europe, North America, and Australia . J Phys Act Health . 2008 ; 5 ( 6 ): 795 - 814 .