The association between free-living daily activity and aging is unclear because nonexercise movement and its energetic equivalent, nonexercise activity thermogenesis, have not been exhaustively studied in the elderly. We wanted to address the hypothesis that free-living nonexercise movement is lower in older individuals compared with younger controls matched for lean body mass. Ten lean, healthy, sedentary elderly and 10 young subjects matched for lean body mass underwent measurements of nonexercise movement and body posture over 10 days using sensitive, validated technology. In addition, energy expenditure was assessed using doubly labeled water and indirect calorimetry. Total nonexercise movement (acceleration arbitrary units), standing time, and standing acceleration were significantly lower in the elderly subjects; this was specifically because the elderly walked less distance per day despite having a similar number of walking bouts per day compared with the young individuals. The energetic cost of basal metabolic rate, thermic effect of food, total daily energy expenditure, and nonexercise activity thermogenesis were not different between the elderly and young groups. Thus, the energetic cost of walking in the elderly may be greater than in the young. Lean, healthy elderly individuals may have a biological drive to be less active than the young.
- nonexercise activity thermogenesis
- physical activity monitoring system
the aging population is increasing in size. For example, in the US, by census data in 2005 there were ∼37 million people aged ≥65 yr. By 2020, it is predicted, there will be ∼54 million, and by 2050 ∼87 million people in the US will be age 65 yr and over (10). Worldwide in 2000, there were 606 million people aged 60 yr and above, and it is predicted that by 2050 this will increase to two billion (9). Overall, the population and number of elderly people is increasing worldwide.
Free-living nonexercise movement declines with aging in several animal species, including worms, fruit-flies, rats, mice, cats, and dogs (17, 23, 25–27, 37, 39). The decline in nonexercise movement with aging could be important in human science. This is because low levels of overall nonexercise movement may be associated with the loss of muscle associated with aging, termed sarcopenia. The sarcopenia of aging, in turn, is associated with physical disability in the elderly, resulting in increased morbidity (1, 15, 16, 24). Thus, free-living nonexercise movement is important both for understanding the biology of aging and for its health implications (22).
Nonexercise movement has been widely assumed to decrease in humans with aging; however, this has not been objectively assessed. The reason that so little is known about nonexercise movement in the elderly is the difficulty of measuring this variable in free-living people. However, validated tools have recently become available to do this (18–20), and so we are able to detail free-living nonexercise movement and its components in healthy elderly people.
Nonexercise movement is all the activity an individual performs throughout the day that is not planned, intentional exercise (activity performed with the goal of improving health and/or well being). Nonexercise movement can be separated into posture and the movement performed in a given posture. Nonexercise movement is measured in acceleration units for the given posture. The energy equivalent of nonexercise movement is nonexercise activity thermogenesis (NEAT). NEAT is quantified either by using a combination of the measures of nonexercise movement and calorimetry, or by using doubly labeled water (DLW).
Data suggest that nonexercise movement may be regulated through central circuits, and so we conjecture that nonexercise movement declines in aging independently of muscle mass. Conversely, the lower nonexercise movement with aging may reflect decreased muscle mass alone, as occurs in other muscle-wasting states (4, 30, 44). Thus, if nonexercise movement were less in elderly individuals compared with young control individuals that are matched for lean body mass, it would suggest that the difference in nonexercise movement was independent of lean body mass. On the other hand, if nonexercise movement were similar between these groups it would suggest that lean body mass could be pivotal in the decline in nonexercise movement associated with aging.
We addressed the hypothesis that daily free-living nonexercise movement is lower in older than in younger controls matched for lean body mass. This information might provide insight into the mechanism that underlies human sarcopenia but also provide information that is relevant for improving health care provision and quality of life for our aging population.
MATERIALS AND METHODS
Ten lean, healthy, sedentary elderly volunteers and ten lean, healthy, sedentary young volunteers were recruited. Subjects were excluded if they exercised more than twice a week or >100 min/wk, smoked, abused alcohol, were pregnant, or reported unstable body weight (>2 kg fluctuation for the 3 mo before the study). The young group was free of illness and medication at the time of the study and within 6 mo of the study. The elderly group was healthy, although one subject each took risedronate, atorvastatin, or the combination of atorvastatin, lisinopril, and hydrochlorothiazide at the time of the study and within 6 mo of the study.
Subjects were studied as outpatients for 10 days receiving weight maintenance meals. Subjects wore a physical activity monitoring system (PAMS) on days 2–11 inclusive. DLW was administered on day 2. Subjects were admitted to the General Clinical Research Center (GCRC) on the evening of day 11. Inpatient testing occurred on days 12–14, including basal metabolic rate (BMR) and activity energy expenditure on days 12 and 13, and BMR and thermic effect of food on day 14. Dual-energy X-ray absorptiometry (DEXA) was performed on day 13. Informed written consent was obtained after the nature and possible consequences of the study had been explained. The study was approved by the Mayo Clinic Institutional Review Board.
Weight Maintenance Dietary Provision
Subjects were initially studied as outpatients for 10 days. Meals were prepared in the metabolic kitchen at the Mayo Clinic GCRC. All foods were weighed to within 1 g. Volunteers were fed so as to establish the dietary intake necessary to maintain steady-state body weight. The diet composition was 45% carbohydrate, 35% fat, and 20% protein. The volunteer's body weight was measured each morning under standardized conditions (with an empty bladder, without shoes, and wearing consistent, light clothing); these measurements were made by trained GCRC personnel using the same calibrated scale (ScaleTronix 5005; S/N 5-1700, Wheaton, IL). “Lean” was defined as BMI between 18 and 26 kg/m2. Subjects were instructed not to adopt new exercise practices and to continue their usual daily activities and occupation.
For body weight, the average of the last 7 days' weights for the period the subject was in the PAMS was taken to represent the body weight for that subject.
Measurement of total daily body posture and movement.
We used a PAMS that we had devised and validated (18–20). This system is unique, because it captures data on body posture and movement in duplicate continuously every half-second. PAMS included six sets of sensors, four inclinometers [each of which captures two axes of acceleration against the earth's gravitational field (CXTA02; Crossbow Technology, San Jose, CA)], and two triaxial accelerometers [each captures motion in x-, y-, and z-axes (CXL02LF3-R, Crossbow Technology)]. The 10 axes of data were binned and stored every half-second on two data loggers (Ready DAQ AD2000, Crossbow Technology). The inclinometers were attached to the right and left outer upper aspect of the trunk and right and left outer aspect of the thigh. The two accelerometers were placed over the base of the spine. Specially designed underwear was used to attach the sensors. The two data loggers were stored in a fanny pack worn around the waist. The PAMS weighed less than 1 kg.
Every 24 h, study personnel removed the sensors while the subject showered for 15 min. During this time, data from the data loggers were downloaded to a personal computer and analyzed using Matlab programs (The MathWorks, Natick, MA). The time taken showering was taken to represent “standing” for this period of time. PAMS sensors were tested daily using electronic testing equipment we had devised. PAMS was also tested daily on each subject for posture and for graded ambulation at 0, 1.0, 2.0, and 3.0 mph for the young subjects and 0, 0.8, 1.6, and 2.4 mph for the elderly subjects. The difference in speeds was because in our preliminary testing several elderly subjects were not able to tolerate walking at 3 mph.
Posture allocation (daily lie, sit, stand/ambulatory time) was calculated for each subject as the average of the 10 days of PAMS data for the number of minutes spent lying, sitting, and standing/ambulating.
Nonexercise movement was calculated for each subject as the average of the 10 days of PAMS data for the sum of the number of acceleration units spent lying, sitting, and standing/ambulating. Data are expressed as acceleration arbitrary units (AAU) for the stated posture.
Standing acceleration was calculated for each subject as the average of the 10 days of PAMS data for the acceleration for each recorded bout of standing posture of >0.5 s duration.
FREE-LIVING WALKING VELOCITY.
Free-living walking velocity was calculated for each subject by plotting the subject's standing acceleration against the subject's velocity/acceleration calibration curve. The calibration curve was defined each day using the 12-min PAMS daily walk calibration procedure described above.
Walking bout is defined as when a person is standing for more than 0.5 s and takes more than 1 step.
Walking distance was calculated for each subject as the product of the average of the 10 days of PAMS walking velocity and total walk time for that subject.
Measurements of BMR.
BMR was measured on three consecutive mornings (days 12, 13, 14) between 0600 and 0800 in the subjects who had slept uninterrupted the previous nights in the GCRC (subjects were admitted to the GCRC on the evening of day 11
BMR was the average of the final 25 min for the 3 consecutive days of data (days 12–14).
Measurements of energy expenditure.
On days 12 and 13, subjects were admitted to the GCRC overnight (from the night of day 11). After the measurements of BMR had been completed as described above, energy expenditure was measured while subjects wore their PAMS sensors and either sat motionless, stood motionless, moved between different postures, or walked at 1.0, 2.0, and 3.0 mph (young subjects) or 0.8, 1.6, and 2.4 mph (elderly subjects) on a calibrated treadmill.
ENERGY COMPONENTS OF NEAT.
the regression equations between velocity, energy expenditure, and accelerometer output allows free-living ambulatory accelerometer output to be translated into energy expenditure.
NEATPAMS is the sum of the energy components of NEAT for that individual.
Measurements of thermic effect of food.
Postprandial thermogenesis was measured on day 14 after the measurement of BMR. Subjects were given a meal that provided one-third of their daily intake (45% carbohydrate, 35% fat, 20% protein). Energy expenditure was measured using the indirect calorimeter for 15 of every 30 min (to prevent agitation) for 450 min.
The thermic effect of food (TEF) for the subject was taken as the area under the postprandial curve, using the BMR as baseline and then multiplying by 3 to give the daily TEF for the subject.
Measurements of total daily energy expenditure.
Total daily energy expenditure (TDEE) was measured for days 2–11 using DLW (7, 8, 32, 33, 35). Baseline urine samples were collected, and after timed administration of the isotopes urine samples were collected at 0700, 1200, and 1800 each day for 10 days. The urine samples were analyzed by mass spectroscopy, and the resultant enrichments were processed using equations (7, 34, 40) to calculate TDEE.
NEATDLW is calculated using the DLW determined total daily energy expenditure minus the sum of the subject's BMR and TEF.
Measurements of body composition.
Each volunteer was weighed daily as described above. Body fat and mineral mass were measured in duplicate using DEXA (Lunar, Madison, WI) after the 11 days of baseline feeding. Each elderly individual was matched for fat-free mass (FFM) with a young individual as being within 10%. The matching of individual for individual was sex specific. To ensure that our measurements of body composition were reproducible and precise, 1) we used the same DEXA scanner throughout the study, 2) we calibrated the DEXA scanner before each measurement with tissue phantoms, and 3) we calibrated the DEXA scanner weekly against tissue blocks of known composition.
Fat mass was the average of the two DEXA fat mass results for the subject.
FFM was calculated as the difference between the subject's body weight and combined fat and mineral mass as determined by DEXA.
Data on posture and motion were analyzed using Matlab. To address our hypothesis that daily free-living nonexercise movement is lower in older lean, healthy, sedentary subjects compared with younger lean, healthy, sedentary controls matched for lean body mass, a two-tailed unpaired t-test was used assuming normal distribution. Where data were not normally distributed, nonparametric testing was performed. Data are expressed as means ± SD, and statistical significance was defined as P < 0.05.
The elderly subjects were 76 ± 5 yr (5 males, 5 females). The baseline weight of the subjects was 69 ± 9 kg, and BMI was 24 ± 2 kg/m2. The elderly subjects were healthy and had limited illnesses and medications at the time of the study and within 6 mo before the study. Two subjects had dyslipidemia, one hypertension, and one asymptomatic, fracture-free osteoporosis. None of the elderly had symptomatic degenerative arthritis. The elderly subjects were retired and active. They participated in community volunteering on average 6 ± 8 h/wk, attended to their own errands (banking, mailing, and shopping) and, in general, considered themselves “too busy to sit and watch television.” Many had hobbies such as assisting with church errands and events, doing the shopping for other elderly persons, traveling internationally, renovating their houses, participating in craft activities, and preparing events for family celebrations. Several of the elderly individuals commented on “not knowing how I previously had time for work,” given that they now find it very difficult to complete all the activities they currently would like to do.
The young subjects were 38 ± 10 yr (5 males, 5 females) and weighed 65 ± 9 kg with a BMI of 22 ± 2 kg/m2. They were matched for lean body mass to the elderly subjects. Of the 10 young subjects recruited to participate, eight were working full time in occupations that would be considered sedentary, such as phlebotomist, and of the other two, one was a full-time housekeeper and one was a full-time student. One young subject participated in community volunteering, and all were independent in their activities of daily living. Thus, the young subjects might be considered representative of healthy, sedentary, free-living young individuals.
PAMS was used to gather detailed information on free-living physical activity and posture over 10 days. The PAMS instruments provided 1,728,000 data points per day per subject. We analyzed ∼17 million data points for each subject studied and 345,600,000 data points for the entire study. PAMS was tested daily in each subject both for electronic validity and for posture detection and velocity determination. Thus the accuracy and precision of posture allocation was confirmed for all 20 subjects for each of the 200 days of testing. There were no detected errors in 288,000 test measurements. There were log linear relationships between accelerometer output and velocity with r2 >0.95 in all cases. PAMS collected data in duplicate, and for all subjects positional and velocity data showed a coefficient of variation of <1%.
For the indirect calorimetry, repeated alcohol burn experiments yielded CO2 and O2 recoveries of >98%. The SD of the respiratory quotient (RQ) for the last 15 min of these measurements was <1% of the mean. Test-retest differences for duplicate measurements of BMR were <4%. The test-retest differences for duplicate DEXA measurements was <2%.
Table 1 presents the body composition data for the two groups. By design, FFM for the elderly and young were comparable: 45 ± 10 compared with 46 ± 9 kg (P = 0.80). There was no significant difference between the two groups in weight or BMI. As expected (13), the elderly subjects had, on average, 5.5 kg more fat than the young subjects (P = 0.04).
Weight maintenance energy intake for the elderly was 2,165 ± 435 kcal/day, and for the young it was 2,325 ± 335 kcal/day (P = 0.34). TDEE calculated from DLW was 2,324 ± 469 kcal/day in the elderly and 2,348 ± 336 kcal/day in the young subjects (P = 0.90). Measurements of TDEE derived using DLW were in excellent agreement with the measurements of baseline weight-maintenance dietary intake (r2 = 0.83, P < 0.001), with an intercept not different from 0 and a slope not different from 1.
BMR in the elderly was 1,356 ± 201 kcal/day and in the young 1,354 ± 179 kcal/day (P = 0.97). Thermic effect of food in the elderly was 247 ± 67 kcal/day compared with 284 ± 63 kcal/day in the young (P = 0.22).
To examine the energy cost of walking in the elderly, we examined the free-living energy expenditure of daily activity, called NEAT. NEAT calculated by DLW in the elderly was not different (P = 0.93) from that in the young subjects (720 ± 289 kcal/day vs. 711 ± 203 kcal/day, respectively).
There were differences in posture allocation between the groups. The elderly subjects sat 557 ± 100 min/day compared with 426 ± 103 min/day in the young subjects (P < 0.01; Fig. 1). The elderly subjects stood 376 ± 90 min/day compared with the young 486 ± 122 min/day (P = 0.04). There was no difference in daily lie time between the two groups. Thus, the elderly sat more and stood less than the young subjects.
Daily movement was lower in the elderly. Total nonexercise movement in the elderly was 329 ± 55 AAU/day compared with 466 ± 110 AAU/day in the young (P < 0.01). This difference in nonexercise movement was also true when corrected for FFM. Nonexercise movement per kilogram of FFM in the elderly was 7.5 ± 1.6 AAU/kg FFM compared with 10.5 ± 3.2 AAU/kg FFM in the young (P = 0.02). Accelerometery output per minute standing was different between the two groups, with the elderly being 0.52 ± 0.10 AAU/min and the young 0.64 ± 0.09 AAU/min (P < 0.01). There was no relationship between standing acceleration and percent fat mass for the group as a whole. Thus, total nonexercise movement (AAU) and accelerometer output per minute of standing were significantly less in the elderly compared with the young.
Because all the data were time stamped, we were able to define the partitioning of body acceleration throughout the day. Having identified that standing acceleration was less in the elderly, we were able to examine whether this was because of differences in the acceleration per minute of walking (velocity) or the time spent walking (distance traveled) or both.
The elderly walked at a similar free-living velocity (P = 0.66) compared with the young (1.2 ± 0.3 vs. 1.2 ± 0.2 mph, respectively). The elderly subjects had a similar number (P = 0.92) of walking bouts: 50.9 ± 8.1 vs. 51.3 ± 8.9 walks/day in the young subjects (Table 2). For the elderly, the duration of each walk averaged 7.1 ± 2.0 min/walk compared with the young 9.8 ± 3.0 min/walk (P = 0.06). Overall, the elderly walked less distance per day; the elderly walked 6.5 ± 1.2 miles/day and the young 9.7 ± 2.1 miles/day (P < 0.01). In conclusion, the elderly walked, on average, 140 min less per day (P = 0.01) and traveled, on average, 3.1 miles less per day than the young subjects (P < 0.01; Fig. 2, A and B).
NEAT derived from PAMS in the elderly was 715 ± 258 kcal/day and from DLW 720 ± 289 kcal/day (ICC 0.77). NEAT derived from PAMS in the young was 845 ± 278 kcal/day and from DLW 711 ± 203 kcal/day (ICC 0.63). The slope and intercept were not significantly different from the line of identity.
Free-living walking energy expenditure was derived from NEATDLW minus the sum of the free-living standing and sitting energy expenditure. We recognize that this includes the energy cost of fidgeting. Walking EE was 638 ± 268 kcal/day in the elderly subjects and 592 ± 191 kcal/day in the young subjects; these were not significantly different (P = 0.66). The calculated energy expended per mile walked in the free-living state each day in the elderly group was 154 ± 60 kcal compared with 125 ± 13 kcal in the young group (P = 0.16). The energy expended per mile walked each day in the elderly group was 3.4 ± 0.8 kcal/kg FFM compared with 2.8 ± 0.4 kcal/kg FFM in the young group (P = 0.05). Thus, the elderly walk less distance than the young; however, the energy cost per mile walked per kilogram of FFM is likely to be greater in the elderly than the young.
Nonexercise movement declines in many animal species with aging. We objectively examined whether nonexercise movement, independently of lean body mass, declines in the elderly. We compared nonexercise movement, using a valid PAMS, between healthy young and elderly individuals matched for lean body mass. We found that nonexercise movement (AAU) was lower in the elderly by 29%, which was explained by the elderly walking 3 miles/day less than the young controls. Thus, aging is associated with lower nonexercise movement and, specifically, less locomotion. The implication is that elderly individuals may have a physiological drive to be less active than the young that is independent of lean body mass.
There have not been detailed prior studies looking specifically at nonexercise movement in the healthy elderly population. Questionnaire-based studies have identified that age is inversely related to energy expenditure for occupation, that women may overvalue housework/caretaking activities in questionnaires, and that the utility in assessing free-living activity with questionnaires may still need to be established (3, 11, 21, 36). There is one study focusing on nocturnal nonexercise movement in healthy elderly subjects that utilizes a wrist-worn activity monitor, suggesting that there is no effect of age in males on nocturnal nonexercise movment and that the duration of nocturnal immobility periods for females is greater than in males (47). Thus, prior to these studies little was known about nonexercise movement and healthy aging.
The data reported here demonstrate that nonexercise movement and in particular walking is lower in healthy elderly people. The mechanism of this nonexercise movement downregulation may include alterations in central, peripheral, or humoral factors. Changes in, or dysregulation of, central mechanisms that modulate or regulate activity may occur with aging (14, 41, 45). Alternatively, there may be a change in the central perception of fatigue, resulting in a lower threshold for activity cessation (2, 12, 31, 42). There may be dysfunction or dysregulation in peripheral neurons or intrinsic muscle function limiting the activity that a subject may be able to perform. Several authors have also found that there is mitochondrial dysfunction with aging (5, 6, 28, 29, 38, 43, 46). A reduction in mitochondrial function through reducing ATP synthesis may contribute to energy inefficiency leading to lower nonexercise movement with aging. Humoral factors may act through central or peripheral sites modulating nonexercise movement. Specifically, we suggest humoral factors related to the increased fat mass associated with aging. We are aware that short-term weight gain of 10 lb does not alter posture allocation and so is unlikely to explain the differences in nonexercise movement and posture allocation with aging. We suggest that perhaps an adipokine may contribute to the altered posture allocation and possible energy inefficiency of aging. Calculation of nonexercise movement from this study as a proportion of fat mass identifies that the elderly have significantly less (P < 0.001) movement (16.9 ± 5.5 AAU/kg fat mass) compared with the young (32 ± 9.6 AAU/kg fat mass), thus suggesting that fat mass may contribute to less nonexercise movement in the elderly.
We recognize several limitations of this study. First, it was not longitudinal. Conducting longitudinal studies would be a logical next step; however, this may be problematic considering the life expectancy of the subjects and the prevalence of illness occurrence in the study target population, considering that herein we studied healthy elderly people. It may be thought that lower nonexercise movement in the elderly is a reflection of societal pressures to walk less with retirement. However, the fact that the elderly subjects were all retired negates the constraints of a work environment where limited mobility is often required. Our subjects, interestingly, took the same number of walking bouts per day as the young subjects, although they did not travel as far with each walk. Thus, despite the lack of a work environment constraint on nonexercise movement, the elderly subjects walked less for each period of walking, supporting a biological etiology to their lower walking distance per day. Another potential limitation of this study is that the results are applicable only to lean, healthy, sedentary elderly individuals. However, this was the goal of the study, namely to examine the effect of aging (rather than illness) on free-living nonexercise movement. Thus, we believe that the elderly subjects recruited for this study are representative of lean, healthy, sedentary subjects in our community. Further and larger studies are required to confirm our results and assess the implications of these findings for the elderly population as a whole. Future studies are also required to address nonexercise activity thermogenesis and energy expenditure in the increasing elderly obese population and to investigate the etiology of our findings.
In conclusion, the population is aging, and it is universally agreed that this will precipitate a demographic shift of substantial economic and societal importance. Therefore, understanding core components of the aging process has great implications. Here, we describe that healthy, active elderly people show one-third less nonexercise movement (AAU) compared with sedentary younger people. The reason for this is specific: elderly people walk 3 miles a day less than matched young controls. The difference in nonexercise movement that we report suggests that with aging the biological drive to be active may be lower. This age-related decline in physical activity has implications for societal adaptation to an aging population. On an individual level, consideration will need to be given to providing resources that are more geographically compact and that will permit greater sitting periods for elderly people. This may impact societal infrastructure should more elderly enter the workforce. Societal expectations of the elderly population may need to be adjusted too.
This work was supported by a Robert and Arlene Kogod Grant, National Institutes of Health grants AG-26117, AG-09538, DK-56650, DK-63226, DK-662760, and M01-RR-00585, and the Mayo Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by American Physiological Society