Journal Logo


Accelerometer-based Physical Activity

Total Volume per Day and Standardized Measures


Author Information
Medicine & Science in Sports & Exercise: April 2015 - Volume 47 - Issue 4 - p 833-838
doi: 10.1249/MSS.0000000000000468
  • Free


The use of accelerometers in physical activity (PA) research has increased exponentially over the past two decades (Fig. 1). Accelerometer-based activity monitors have several advantages over questionnaires. First, accelerometers provide an objective measure of PA that avoids biases due to subjective recall of past events, varied interpretation of questions, and the participants’ desire to please investigators (4,14,37,38). Second, accelerometers can capture PA performed across a variety of domains (i.e., leisure time, transportation, occupation, and domestic) whereas many questionnaires only capture one or two domains. Third, as shown by Atienza et al. (1), who examined the National Health and Nutrition Examination Survey (NHANES) 2003–2006 data, accelerometer-derived moderate-to-vigorous PA (MVPA) accumulated in bouts of 10 min or longer was more highly related to cardiometabolic biomarkers compared with questionnaire measures of this variable. Fourth, accelerometers can capture sedentary behaviors and incidental PA performed at the lower end of the intensity spectrum in contrast to questionnaires, which have low validity for assessing incidental ambulatory activity (3,38,41).

Accelerometer articles published per year, obtained from Scopus, over a 21-yr period (1993–2013). The search terms were “exercise” or “PA” and “accelerometer” or “accelerometry.”

Despite the many advantages of accelerometers in PA assessment, questionnaires are superior in other ways. Questionnaires are easier and less costly to administer, and they can provide information on activity types and reasons for performing PA (45). Questionnaires can also assess activities that are difficult to capture with accelerometers, including muscle-strengthening, bone loading, and flexibility-enhancing activities (34). In addition, questionnaires can capture other subjective or perceived aspects of PA (such as location and context) (15).

PA is not a single entity, but rather, it is a multidimensional construct. The dimensions of PA include frequency, intensity, duration, and types of activities performed (41). However, these dimensions can be condensed down into a single variable that reflects the total volume of PA. Other methods for classifying PA exist, and these are valued by the research community (Fig. 2). In this article, we will give a brief historical overview of the use of accelerometers in PA research, advocate for using accelerometers to estimate the total volume of PA as a primary metric, and suggest ways to standardize accelerometer data.

PA is a complex multidimensional human behavior that can be classified in a number of different ways. The total volume of PA accumulated over time is an important variable because it comprises the frequency, intensity, and duration of activity bouts and reduces them down to a single metric. To obtain a reliable assessment of a person’s habitual PA, data must be collected over a number of days.


The first commercially available accelerometer-based PA monitor (the Caltrac) estimated PA energy expenditure (PAEE) in kilocalories (kcal) (30). User characteristics (i.e., age, height, weight, and gender) were entered, enabling resting metabolic rate (RMR) to be estimated. The Caltrac activity monitor then computed total daily energy expenditure (TDEE) using the formula TDEE = PAEE + RMR + thermic effect of feeding (23).

After publication of the 1995 Centers for Disease Control and Prevention/American College of Sports Medicine summary statement on PA (33), researchers began to pay increased attention to MVPA. This document stated, “Every US adult should accumulate 30 min or more of moderate-intensity PA on most, preferably all, days of the week” (33). It stimulated interest in measuring the number of minutes of MVPA per day, and PA measurement researchers seized the challenge of converting acceleration data into these units (9,18,42). Because the studies that served as the basis of the guidelines captured self-reported PA in bouts of 10 min or longer, the summary statement specified 10 min as the minimum bout duration.

Because the focus on MVPA accumulated in bouts of 10 min or more, other types of PA such as light-intensity PA (LPA) and intermittent MVPA were often ignored. In recent years, however, evidence concerning the health hazards of sedentary behavior has accumulated (2,7,12,16,17,44). These studies also point to the health benefits of LPA because there is a strong inverse correlation (r = −0.96) between sedentary behavior and LPA (16). Some researchers have proposed that sedentary behavior is related to cardiometabolic biomarkers even after adjustment for PA (16,17,24). However, these studies only adjusted for MVPA and they did not take into consideration LPA (16,17,24). Had the total PA volume been adjusted for, it is likely that a significant association between sedentary activities and health variables would not have been found. For instance, a NHANES study by Maher et al. (28) found that after adjusting for the total volume of PA, the significant independent associations between sedentary behavior and health variables virtually disappeared.

Today, most PA monitors contain microelectromechanical system accelerometers, real-time clocks, and 500 MB to 2 GB of memory, enabling the storage of second-by-second summary measures (commonly expressed as monitor-specific activity counts) or even raw acceleration data for 40 d or more. Research-grade accelerometers allow investigators to apply various methods to convert “activity counts” (or raw acceleration data) to PA metrics (e.g., time spent in various intensity categories, kilocalories, kilojoules, and MET-minutes). Currently, the waist-mounted ActiGraph is the most commonly used device to objectively assess PA in research studies (Fig. 3). However, a major problem today is the proliferation of algorithms used to convert “activity counts” to PA metrics, which hinders the ability to draw comparisons between studies (29). For example, using adult data from NHANES 2003–2006, estimates ranged from 17 to 59 min·d−1 of MVPA and the percentage of people meeting PA guidelines ranged from 6.2% to 59.3%, depending on which algorithms were applied to the data (26).

PubMed citations (as of September 9, 2013) for different PA monitors.

As new methods of measuring PA are developed, the two important questions are as follows: 1) If researchers are to use just one PA metric, which one should they use? and 2) Is it possible to standardize measurement of PA to enable comparisons between studies?


The total volume of PA may be a better metric than the number of minutes per day spent in various PA intensity categories because it incorporates the full continuum of intensities. Epidemiological studies and behavioral interventions have focused on MVPA accumulated in 10-min bouts because that was consistent with the public health recommendations. However, accelerometers have made the examination of LPA and nonbout MVPA possible. Recent evidence suggests that LPA, moderate PA (MPA), and vigorous PA (VPA) all have health benefits. Kim et al. (22) reported that individuals who accumulated more LPA were at lower risk for the metabolic syndrome, independent of the amount of MVPA performed. This result is generally consistent with emerging evidence in the field of “inactivity physiology,” which suggests that LPA for several hours per day can lessen the deleterious health effects of prolonged sitting (7,12,13,32). The vital importance of accounting for LPA was recently shown by Maher et al. (28), who observed that there is no association between sedentary behavior and cardiometabolic health after adjusting for total activity (i.e., light + moderate + vigorous). With regard to bout duration, Glazer et al. (10) and Strath et al. (40) showed that short bouts of MVPA were associated with cardiometabolic health, just as bouts of 10 min or longer were. Holman et al. (19) reported the same findings in youths. Loprinzi et al. (25) also found that short bouts of MVPA were inversely related to the metabolic syndrome, suggesting that they have positive effects on cardiometabolic health. Taken together, these studies suggest that LPA and short bouts of MVPA have health benefits and should be accounted for in PA and health research.

There is much anticipation around the newly emerging technologies and methods for accelerometer-based PA assessment (6,8). It has been demonstrated that triaxial accelerometers (5), multiple monitors worn in various locations (51), and sophisticated pattern-recognition techniques (21,35,39,46) can enhance PA characterization and lead to more accurate assessment of PA. However, a large volume of current technology is still in use and valuable data sets have been collected using waist-worn legacy count-based monitors, especially the ActiGraph. The widespread use of the ActiGraph and availability of US population reference data in a diverse sample has several advantages in the short term, which will be discussed in the paragraphs that follow. It is unlikely that the vertical axis acceleration of a waist-mounted ActiGraph will remain the standard forever. Nevertheless, total activity counts per day (TAC/d) obtained from the ActiGraph may provide a good starting point for promotion of standardized measures of PA volume across studies.

In the current landscape of count-based objective monitoring, TAC/d (averaged over 4–7 d) can serve as a proxy for the total volume of PA. Wolff et al. (48,49) have noted that TAC/d has the advantage of integrating the frequency, intensity, and duration of ambulatory movement and combining them into an overall measure of PA. Because activity counts are already weighted according to intensity, researchers can simply sum all of the counts accumulated throughout the day to obtain a proxy for total volume. (On the other hand, if researchers use ActiGraph-derived MET-minutes of LPA, MPA, and VPA, they would need to use multipliers to adjust for the mean MET level of intensity categories (20).) A second advantage of accelerometer-derived “activity counts” is that it is a step closer to what the accelerometer actually measures (body acceleration) rather than being a derived variable based on regression algorithms or cut points. A third advantage is that TAC/d is more closely related to cardiometabolic biomarkers than minutes of MVPA accumulated in 8- to 10-min bouts (49). Finally, a fourth advantage is that TAC/d tends to be normally distributed (i.e., skewness and kurtosis within ±2) unlike many PA metrics (Table 1). This allows researchers to conduct parametric statistics without violating the assumption of normality (11,31).

Skewness and kurtosis of various accelerometer-based measures of PA.

Although TAC/d is a proxy for the overall volume of PA performed, it is likely that the research community will remain interested in other dimensions and characteristics of PA. It may be that the total volume of PA is not the only important variable and that other PA dimensions also affect health, independent of total volume. Thus, we propose that TAC/d be considered as an additional variable, not one that replaces all other PA metrics.

Although TAC/d has several strengths, it also has limitations. One is that TAC/d lacks intuitive meaning, in and of itself. However, TAC/d can be converted to age- and gender-specific percentiles, yielding a score (from 1 to 100) that is based on population reference data (50). This percentile score depicts how active a person is relative to others of the same age and gender. Eventually, criterion-referenced standards for total volume of PA could be established by studying the relation between these percentile scores and health outcomes. Anthropometric data and physical fitness data are frequently translated to percentiles for population surveillance and use as reference data for evaluating individuals (e.g., growth curves). There may be advantages for researchers in using this approach because PA changes over the lifespan (50).


Accelerometer-based monitors quantify PA on the basis of accelerations measured at a fixed body location (e.g., hip or wrist). Conceptually, these devices are optimized for assessment of the total volume of PA “performed above a zero baseline,” as has been recommended (36). The ability to standardize accelerometer-based measures of PA could assist in harmonizing data across studies. However, there are several challenges to standardization.

The first challenge, which has already been mentioned, is the proliferation of cut points used to translate activity counts to minutes of light, moderate, and vigorous PA. One way to achieve standardization is to report TAC/d. TAC/d is a more direct measure of what the accelerometer measures versus transforming activity counts into energy expenditure. This helps provide standardization by circumventing the problem of proliferation of activity count cut points.

A second challenge is that the counts from various devices (e.g., ActiGraph, Actical, GENEA) are not equivalent. There are substantial differences in activity counts recorded by different accelerometer models due to differences in accelerometer design, filtering, and signal processing (48). As a result, a limitation of TAC/d is that activity counts vary according to the brand of activity monitor used. Fortunately, the fact that the many biomedical and public health researchers are currently using a single device lessens the extent of the problem. In addition, it might be possible to align the output from different PA monitors by doing equivalence testing (using a shaker table that allows the frequency and amplitude of accelerations to be varied under precise conditions).

Research data collection is progressing toward storage of raw acceleration signals rather than preprocessed activity counts as a result of the 2010 releases of the GT3X+ (ActiGraph LLC, Pensacola, FL) and GeneActiv (ActivInsights, Kimbolton, Cambridgeshire, England) devices. As new methods of PA assessment are developed, how can assessment of the total volume of PA be preserved? It might be possible to use standardized procedures, including accelerometers with specific device characteristics worn in standard locations. In addition, triaxial acceleration (i.e, vector sum magnitude) could be expressed in Système International units, such as meters per second squared. These raw data could be full-wave rectified, integrated, and summed to express total PA volume in standardized activity counts expressed in Système International units, rather than relying on a single manufacturer’s proprietary activity counts. As future advances in machine learning algorithms occur, it would be possible to analyze the activity count data using sophisticated algorithms to estimate activity types, body postures, PA components (frequency, intensity, and duration), and PAEE.

A third challenge to standardization is that activity monitors can be worn on different body sites. For years, the traditional location was the waist but activity monitors are now being placed on the wrist, ankle, and thigh. Emerging evidence suggests that when devices are moved from one body site to another, the PA outputs change. Obviously, taking algorithms designed to convert acceleration to PA metrics at one body site and applying them to a different site would probably yield noncomparable data. Although there is no simple solution to the problems imposed by different body sites, researchers should, at the very least, be aware of the pros and cons of various locations and they should only use algorithms that have been calibrated and validated for a particular location.

An important challenge for exercise scientists who use accelerometers is to decide whether they want to standardize the measurement of PA. There is a precedent for scientific organizations standardizing measurement of blood pressure (27,43) and blood glucose (47). However, these are physiological variables that have “gold standards” and lend themselves to direct measurement in clinical settings, whereas PA is a complex multidimensional human behavior that must be assessed over a longer time (34,41). Even so, the lack of standardization in activity monitoring research is a major limitation that presents a significant barrier to progress in research. At the very least, if all researchers who use the same monitor were to agree to place it on the same body site and report a standard metric (such as TAC/d) as a proxy for total volume of PA, it would be a step in the right direction.


In summary, complete standardization of PA measurement is extremely challenging because of the wide range of questions that researchers want to answer. These questions require multiple PA variables to be measured. Various devices that use various combinations of physiological and movement sensors will continue to be developed and used. However, for many purposes, triaxial accelerometers have been found to provide an adequate compromise between high accuracy and reliability on the one hand and low participant and researcher burden on the other. For now, TAC/d is a variable that seems to have some advantages over minutes of MVPA. As new methods of PA measurement are developed, researchers should continue to consider metrics that reflect the total volume of PA performed. If researchers who elect to use accelerometers can standardize the device’s characteristics (i.e., accelerometer range, digital filter, and sampling rate), wear site, measurement period, and data reduction methods, this could allow the measurement of PA volume to be harmonized across studies. Alternatively, if procedures can be developed for translating data obtained with various devices, wear sites, etc. so that the data are interchangeable, this may come close to achieving the same goal.

This study did not receive funding from other agencies.

David R. Bassett, Jr., is a member of the ActiGraph Scientific Advisory Board. For the remaining authors, no conflicts of interest were declared.

The results of this study do not constitute endorsement by the American College of Sports Medicine.


1. Atienza AA, Moser RP, Perna F, et al. Self-reported and objectively measured activity related to biomarkers using NHANES. Med Sci Sports Exerc. 2011; 43 (5): 815–21.
2. Bankoski A, Harris TB, McClain JJ, et al. Sedentary activity associated with metabolic syndrome independent of physical activity. Diabetes Care. 2011; 34 (2): 497–503.
3. Bassett DR. Validity and reliability issues in objective monitoring of physical activity. Res Q Exerc Sport. 2000; 71 (2): 30–6.
4. Brownson RC, Eyler AA, King AC, Shyu YL, Brown DR, Homan SM. Reliability of information on physical activity and other chronic disease risk factors among US women aged 40 years or older. Am J Epidemiol. 1999; 149 (4): 379–91.
5. Butte NF, Ekelund U, Westerterp KR. Assessing physical activity using wearable monitors: measures of physical activity. Med Sci Sports Exerc. 2012; 44 (1 Suppl): S5–12.
6. Chen K, Janz K, Zhu W, Brychta R. Redefining the roles of sensors in objective physical activity monitoring. Med Sci Sports Exerc. 2012; 44 (1 Suppl): S13–23.
7. Dunstan DW, Howard B, Healy GN, Owen N. Too much sitting—a health hazard. Diabetes Res Clin Pract. 2012; 97 (3): 368–76.
8. Freedson P, Bowles HR, Troiano R, Haskell W. Assessment of physical activity using wearable monitors: recommendations for monitor calibration and use in the field. Med Sci Sports Exerc. 2012; 44 (1 Suppl): S1–4.
9. Freedson PS, Melanson E, Sirard J. Calibration of the Computer Science and Applications, Inc accelerometer. Med Sci Sports Exerc. 1998; 30 (5): 777–81.
10. Glazer N, Lyass A, Esliger D, et al. Sustained and shorter bouts of physical activity are related to cardiovascular health. Med Sci Sports Exerc. 2012; 45 (1): 109–15.
11. Grimm L, Yarnold P. Reading and Understanding More Multivariate Statistics. Washington (DC): American Psychological Association; 2000.
12. Hamilton MT, Hamilton DG, Zderic TW. Role of low energy expenditure and sitting in obesity, metabolic syndrome, type 2 diabetes, and cardiovascular disease. Diabetes. 2007; 56 (11): 2655–67.
13. Hamilton MT, Healy GN, Dunstan DW, Zderic T, Owen N. Too little exercise and too much sitting: inactivity physiology and the need for new recommendations on sedentary behavior. Curr Cardiovasc Risk Rep. 2008; 2 (4): 292–8.
14. Harris TJ, Owen CG, Victor CR, Adams R, Ekelund U, Cook DG. A comparison of questionnaire, accelerometer, and pedometer: measures in older people. Med Sci Sports Exerc. 2009; 41 (7): 1392–402.
15. Haskell WL. Physical activity by self-report. J Phys Act Health. 2012; 9 (1 Suppl): S5–10.
16. Healy GN, Matthews CE, Dunstan DW, Winkler EAH, Owen N. Sedentary time and cardio-metabolic biomarkers in US adults: NHANES 2003–06. Eur Heart J. 2011; 32 (5): 590–7.
17. Healy GN, Wijndaele K, Dunstan DW, et al. Objectively measured sedentary time, physical activity, and metabolic risk: the Australian Diabetes, Obesity and Lifestyle study (AusDiab). Diabetes Care. 2008; 31 (2): 369–71.
18. Hendelman D, Miller K, Baggett C, Debold E, Freedson P. Validity of accelerometry for the assessment of moderate intensity physical activity in the field. Med Sci Sports Exerc. 2000; 32 (9 Suppl): S442–9.
19. Holman RM, Carson V, Janssen I. Does the fractionalization of daily physical activity (sporadic vs. bouts) impact cardiometabolic risk factors in children and youth? PLoS One. 2011; 6 (10): e25733.
20. Janssen I, Ross R. Vigorous intensity physical activity is related to the metabolic syndrome independent of the physical activity dose. Int J Epidemiol. 2012; 41 (4): 1132–40.
21. Johannsen DL, Calabro MA, Stewart J, Franke W, Rood JC, Welk GJ. Accuracy of armband monitors for measuring daily energy expenditure in healthy adults. Med Sci Sports Exerc. 2010; 42 (11): 2134–40.
22. Kim J, Tanabe K, Yokoyama N, Zempo H, Kuno S. Objectively measured light-intensity lifestyle activity and sedentary time are independently associated with metabolic syndrome: a cross-sectional study of Japanese adults. Int J Behav Nutr Phys Act. 2013; 10 (1): 30.
23. Klesges R, Klesges L, Swenson A, Pheley A. A validation of two motion sensors in the prediction of child and adult physical activity levels. Am J Epidemiol. 1985; 122 (3): 400–10.
24. Koster A, Caserotti P, Patel KV, et al. Association of sedentary time with mortality independent of moderate to vigorous physical activity. PLoS One. 2012; 7 (6): e37696.
25. Loprinzi PD, Cardinal BJ. Association between biologic outcomes and objectively measured physical activity accumulated in ≥10-minute bouts and <10-minute bouts. Am J Health Promot. 2013; 27 (3): 143–51.
26. Loprinzi PD, Lee H, Cardinal BJ, Crespo CJ, Andersen R, Smit E. The relationship of Actigraph accelerometer cut-points for estimating physical activity with selected health outcomes. Res Q Exerc Sport. 2012; 83 (3): 422–30.
27. Luders S, Kruger R, Zemmrich C, Forstner K, Sturm CD, Bramlage P. Validation of the Beurer BM 44 upper arm blood pressure monitor for home measurement, according to the European Society of Hypertension International Protocol 2002. Blood Press Monit. 2012; 17 (6): 248–52.
28. Maher C, Olds T, Mire E, Katzmarzyk PT. Reconsidering the sedentary behavior paradigm. PLoS One. 2013; 9 (1): e6403.
29. Matthews C. Calibration of accelerometer output for adults. Med Sci Sports Exerc. 2005; 37 (11 Suppl): S512–22.
30. Montoye HJ, Washburn R, Servais S, Ertyl A, Webster JG, Nagle FJ. Estimation of energy expenditure by a portable accelerometer. Med Sci Sports Exerc. 1983; 15 (5): 403–7.
31. Newton RR, Rudestam KE. Your Statistical Consultant: Answers to Your Data Analysis Questions. Thousand Oaks (CA): SAGE Publications, Inc.; 1999.
32. Owen N, Healy GN, Matthews CE, Dunstan DW. Too much sitting: the population-health science of sedentary behavior. Exerc Sport Sci Rev. 2010; 38 (3): 105.
33. Pate RR, Pratt M, Blair SN, et al. Physical activity and public health: a recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA. 1995; 273 (5): 402–7.
34. Pettee Gabriel KK, Morrow JR, Woolsey AL. Framework for physical activity as a complex and multi-dimensional behavior. J Phys Act Health. 2012; 9 (1 Suppl): S11–8.
35. Pober DM, Raphael C, Freedson PS. Novel technique for assessing physical activity using accelerometer data (Abstract). Med Sci Sports Exerc. 2004; 36 (5 Suppl): S198.
36. Powell KE, Pallluch AE, Blair SN. Physical activity for health: what kind? how much? how intense? On top of what? Annu Rev Public Health. 2011; 32: 349–65.
37. Schmidt MD, Cleland VJ, Thomson RJ, Dwyer T, Venn AJ. A comparison of subjective and objective measures of physical activity and fitness in identifying associations with cardiometabolic risk factors. Ann Epidemiol. 2008; 18 (5): 378–86.
38. Shephard RJ. Limits to the measurement of habitual physical activity by questionnaires. Br J Sports Med. 2003; 37 (3): 197–206.
39. Staudenmayer J, Pober D, Crouter S, Bassett D, Freedson P. An artificial neural network to estimate physical activity energy expenditure and identify physical activity type from an accelerometer. J Appl Physiol (1985). 2009; 107 (4): 1300–7.
40. Strath SJ, Holleman RG, Ronis DL, Swartz AM, Richardson CR. Objective physical activity accumulation in bouts and nonbouts and relation to markers of obesity in US adults. Prev Chronic Dis. 2008; 5 (4): A131.
41. Strath SJ, Kaminsky LA, Ainsworth BE, et al. Guide to the assessment of physical activity: clinical and research applications: a scientific statement from the American Heart Association. Circulation. 2013; 128 (20): 2259–79.
42. Swartz AM, Strath SJ, Bassett DR, O’Brien WL, King GA, Ainsworth BE. Estimation of energy expenditure using CSA accelerometers at hip and wrist sites. Med Sci Sports Exerc. 2000; 32 (9 Suppl): S450–6.
43. Takahashi H, Yoshika M, Yokoi T. Validation of home blood pressure-monitoring devices, Omron HEM-1020 and Omron i-Q132 (HEM-1010-E), according to the European Society of Hypertension International Protocol. Blood Press Monit. 2011; 16 (4): 203–7.
44. Tremblay MS, Colley RC, Saunders TJ, Healy GN, Owen N. Physiological and health implications of a sedentary lifestyle. Appl Physiol Nutr Metab. 2010; 35 (6): 725–40.
45. Troiano RP, Pettee Gabriel KK, Welk GJ, Owen N, Sternfeld B. Reported physical activity and sedentary behavior: why do you ask? J Phys Act Health. 2012; 9 (1 Suppl): S68–75.
46. Trost S, Wong W, Pfeiffer K, Zheng Y. Artificial neural networks to predict activity type and energy expenditure in youth. Med Sci Sports Exerc. 2012; 44 (9): 1801–9.
47. Vesper H, Myers G. Approaches for improving glucose monitor measurements for self-monitoring of blood glucose: from measurement harmonization to external quality assessment programs. J Diabetes Sci Technol. 2007; 1 (2): 153–7.
48. Wolff DL, Fitzhugh EC, Bassett DR, Churilla JR. Total activity counts and bouted minutes of and moderate-to-vigorous physical activity: relationships with cardiometabolic biomarkers using 2003–2006 NHANES. J Phys Act Health. In press.
49. Wolff DL, Fitzhugh EC, Bassett DR, Churilla JR. Waist-worn actigraphy: population-referenced percentiles for total activity counts in US adults. J Phys Act Health. In press.
50. Zhang K, Pi-Sunyer FX, Boozer CN. Improving energy expenditure estimation for physical activity. Med Sci Sports Exerc. 2004; 36 (5): 883–9.
51. Zhang K, Werner P, Sun M, Pi-Sunyer FX, Boozer CN. Measurement of human daily physical activity. Obes Res. 2003; 11 (1): 33–40.


© 2015 American College of Sports Medicine