Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Number of days required to estimate physical activity constructs objectively measured in different age groups: Findings from three Brazilian (Pelotas) population-based birth cohorts

  • Luiza Isnardi Cardoso Ricardo ,

    Roles Conceptualization, Formal analysis, Writing – original draft, Writing – review & editing

    Affiliation Postgraduate Program in Epidemiology, Federal University of Pelotas, Pelotas, Brazil

  • Andrea Wendt,

    Roles Formal analysis, Writing – original draft, Writing – review & editing

    Affiliation Postgraduate Program in Epidemiology, Federal University of Pelotas, Pelotas, Brazil

  • Leony Morgana Galliano,

    Roles Methodology, Writing – original draft

    Affiliation Postgraduate Program in Physical Education, Federal University of Pelotas, Pelotas, Brazil

  • Werner de Andrade Muller,

    Roles Methodology, Writing – original draft

    Affiliation Postgraduate Program in Physical Education, Federal University of Pelotas, Pelotas, Brazil

  • Gloria Izabel Niño Cruz,

    Roles Writing – original draft

    Affiliation Postgraduate Program in Epidemiology, Federal University of Pelotas, Pelotas, Brazil

  • Fernando Wehrmeister,

    Roles Supervision, Visualization

    Affiliation Postgraduate Program in Epidemiology, Federal University of Pelotas, Pelotas, Brazil

  • Soren Brage,

    Roles Project administration, Supervision, Visualization

    Affiliation MRC Epidemiology Unit, University of Cambridge, Cambridge, England, United Kingdom

  • Ulf Ekelund,

    Roles Conceptualization, Funding acquisition, Project administration, Visualization

    Affiliation Department of Sport Medicine, Norwegian School of Sport Sciences, Oslo, Norway

  • Inácio Crochemore M. Silva

    Roles Formal analysis, Validation, Visualization, Writing – review & editing

    Affiliations Postgraduate Program in Epidemiology, Federal University of Pelotas, Pelotas, Brazil, Postgraduate Program in Physical Education, Federal University of Pelotas, Pelotas, Brazil



The present study aims to estimate the minimum number of accelerometer measurement days needed to estimate habitual physical activity (PA) among 6- (2010), 18- (2011) and 30- (2012) year-old participants, belonging to three population-based Brazilian birth cohorts.


PA was assessed by triaxial wrist-worn GENEActiv accelerometers and the present analysis is restricted to participants with at least 6 consecutive days of measurement. Accelerometer raw data were analyzed with R-package GGIR. Description of PA measures (overall PA, moderate-to-vigorous PA (MVPA), light PA (LPA)) on weekdays and weekend days were conducted, and statistical differences were tested with chi-squared and Kruskal-Wallis tests. Spearman Brown Formulae was applied to test reliability of different number of days of accelerometer use.


Differences between week and weekend days regarding LPA, MVPA and overall PA, were only observed among 30-year-olds. Higher levels of MVPA (p = 0.006) and overall PA (p<0.001) were identified on weekdays. For overall PA, to achieve a reliability coefficient >0.70, two and three days of measurement were needed in adults and children, respectively. For LPA, a reliability coefficient >0.70 was achieved with five days in 6-year-old children, three days in 18-year-old young adults, and four days in 30-year-old adults. Considering MVPA, four days would be necessary to represent a week of measurement among all cohort groups.


Our results show that four and five measurement days are needed to estimate all habitual PA constructs, for children and adults, respectively. Also, among 30-year-old adults, it is important to make efforts towards weekend days measurement.


Physical activity (PA) has positive effect on health and quality of life of individuals and communities [1]. Evidence shows that PA is a protective behavior on major non-communicable diseases, such as coronary heart disease, type II diabetes and cancer [2]. However, global PA prevalence is still low, placing PA promotion as a priority in the public health agenda and, therefore, demonstrating the need of monitoring PA at population levels [3].

Accelerometers (portable motion sensors) have been increasingly adopted in large scale studies, since they provide more accurate physiological and mechanical parameters to estimate PA [4]. In the last decade, the use of accelerometers for this purpose increased significantly [5], due to the capability to quantify duration, frequency and intensity of PA through acceleration signals, movement patterns and magnitude [6]. Also, currently available accelerometers enable large quantity of data storage, a variety of cut-off points for different PA intensities, movement pattern recognition and the possibility of more detailed analyses using the raw data [5].

For a better understanding of the prevalence, levels and impact of PA on health, accurate measures are essential. However, with the increase of accelerometer-based research, the variability among protocols is also rising, since there is no standardized recommendation for data collection. Also, there are different sources of variability, such as accelerometers brands, placements, unit of analysis, minimum number of hours per day and minimum number of measurement days [7,8]. All these decisions will influence final PA estimates at some degree, and therefore must be discussed according to their advantages and disadvantages.

Therefore, the literature regarding the PA monitoring frame presents some evidence that a high percentage of participants are usually excluded from analysis for not being considered as presenting sufficiently valid accelerometer data across different age groups [9]. Furthermore, how representative is the available data when compared to the habitual PA period is not often discussed. Considering that, in reduced sample sizes, differential losses are frequent between subgroups and PA estimates might be biased[10].

Historically, considering that during the first two decades of accelerometer use these measurements have been conducted by waist/hip-worn devices, a decision regarding the minimum number of hours per day was crucial to define valid days of measurement and, consequently, the minimum number of days to estimate habitual physical activity. However, studies using wrist-worn accelerometers usually adopt a 24 hours-protocol, which increases compliance [5,9]. Therefore, the remaining issue is the minimum days of measurement, which is an important decision defining the logistics of the study and representativeness of habitual PA behavior.

The wrist-worn accelerometry literature regarding this issue has not come to a consensus yet, however some studies have analyzed the ideal number of days for accelerometry measurement [1113]. Among adults, around two or three days of measurement are required depending on the evaluated PA construct [11,13,14]. In addition, Scheers et al.[15] suggest that PA should be measured for two weekend days and three weekdays in order to achieve a common week reliability in adults. Also, some studies recommend the inclusion of weekend days on PA estimates [1618]. Among children, the widely accepted recommendation seems to be between three and four measurement days [7,12].

In this sense, it is imperative to discuss not only how many days of measurement will be needed to adequately measure PA, but also how many days will be considered valid. Moreover, scientific literature regarding this subject is mostly from high income settings and low- and middle-income countries (LMICs) must be studied due to potential differences in terms of compliance of accelerometer use as well as regarding PA behavior, such as the distribution of PA according to different domains. Therefore, the present study aims to estimate the minimum number of accelerometer measurement days needed to estimate habitual PA among 6-, 18- and 30-year-old participants, belonging to three population-based Brazilian birth cohorts.

Materials and methods

Study design and participants

This study is based on data from three birth cohorts studies conducted in Pelotas (a southern city in Brazil) where PA was objectively assessed between 2010 and 2012. All babies delivered in hospital in 1982, 1993 and 2004 were identified during daily visits to the maternity hospitals in the city (more than 99% of deliveries occur in hospitals). In each cohort, less than 1% of participants recruited refused to participate. All three cohorts have been followed up at different time points thereafter. More details regarding the methodology of each cohort have been previously described elsewhere [1921].

Recent follow-up visits were performed when cohort participants born in 1982, 1993 and 2004 were approximately 30 years, 18 years and 6 years of age, respectively. At these follow-ups, participants were interviewed and clinically examined by the research team. Data on socioeconomic and demographic characteristics, anthropometry, clinical and biochemical measurements were collected. After the clinical examinations, participants were invited to wear an accelerometer.

Approval for the study was obtained from the School of Medicine Ethics Committee of the Federal University of Pelotas and all participants or their legal representatives voluntarily signed written informed consent.

Physical activity measurement.

PA was assessed by triaxial wrist-worn accelerometers (GENEActiv; ActivInsights, Kimbolton, UK) on the non-dominant wrist. This device has been previously validated and applied in calibration studies among children and adults[2224]. The GENEActiv Accelerometer measures body movements on three axes: vertical (Y), horizontal right-left (X), and horizontal front-back axis (Z), within an acceleration dynamic range of ± 8g, where g represents gravitational unit. Sampling frequency was set at 85.7 Hz.

For practical reasons and to increase compliance, participants were invited to wear the accelerometer using a 24 hours protocol (during awakening and sleeping hours) for 4–7 consecutive days, including at least one weekend day. The total amount of monitored days varied according to the day of the clinical visit. Participants who visited the clinic either on Mondays, Tuesdays or Wednesdays were monitored until the following Monday, whereas those who visited the clinic either on Thursdays, Fridays or Saturdays, were monitored until the following Wednesday. Further information regarding the protocol is available elsewhere [25]. After applying the non-wear time criteria, only part of the cohort groups achieved six 24 hours cycles with valid data. All analyses in the present study were restricted to individuals who provided at least six valid days, respecting the valid day criteria of minimum of 16 hours of measurement.

Data reduction

The device set up and data download was performed in the GENEActiv software. Accelerometer data in binary format were analyzed with R-package GGIR [] [24]. The detailed signal processing scheme included the following steps: verification of sensor calibration error using local gravity as a reference; detection of sustained abnormally high values (pattern non-compatible with human movement) and non-wear detection. Furthermore, calculation of the vector magnitude of activity-related acceleration Euclidian Norm Minus One (ENMO) was used to summarize three-dimensional raw data (from axes x, y, and z) into a single-dimensional signal vector magnitude. The data were further summarized when calculating the average values per 5-second-epochs (ENMO = ∑ | – 1g|)[24].

The summary measures used in the present study were (1) overall PA (the average ENMO per day—expressed in mg)[24], (2) moderate-to-vigorous PA (MVPA) per day (expressed in minutes) with bouts of 10 minutes, and (3) light PA (LPA) (expressed in minutes). MVPA was defined as ENMO records above 100mg, while bouts criterion was defined as consecutive periods in which participants spent at least 80% of time in MVPA. LPA was considered as activities with acceleration between 50mg and 100mg[22].

Statistical analyses

Descriptive analyzes comparing characteristics of individuals with six days of data and the remaining participants of the cohorts were performed using ANOVA or its non-parametric equivalent when necessary. Absolute and relative frequencies distribution and respective p values were presented, according to sex (male/female), weight status (underweight, normal-weight, overweight and obese) and asset index. The BMI estimates were based on the World Health Organization (WHO) classification of BMI for adults (1982 cohort) [26], according to the age and sex Z-scores for BMI for children and young adults (1993 and 2004 cohorts)[27]. Asset index was obtained based on a list of assets. These assets were inserted in a principal component analysis to calculate a score and divided into quintiles where the first one indicates the poorest group[28].

Also, description of PA measures (medians of overall PA, MVPA, LPA) between weekdays and weekend days were conducted, and statistical differences were tested with chi-squared and Kruskal-Wallis tests. The reliability for accelerometer minimum days was obtained using the Spearman Brown Formulae[29], which is based on Intraclass Correlation coefficient (ICC) combined with the Spearman Brown Prophecy, applying the ICC of a single day to a different number of days. The reliability provides the proportion of “real” information about a construct of interest, e.g. if a reliability of 0.70 is found, it can be concluded that 70% of the variability in the measurement represented the construct, and 30% represented random variation[30]. All analyses were stratified by age groups (birth cohorts) and performed in the software Stata 12.0. Statistical significance was set at 5%.


Sample description, as well as comparison between individuals with six measurement days (analytical sample) and the remaining members followed-up on the most recent data collection of each cohort is presented in Table 1. There were no significant differences between the analytical sample and the rest of the cohort in terms of socioeconomic position and BMI groups. The only significant difference observed was between gender among 18-year-old young adults, presenting a higher proportion of females in the analytical sample compared to the remaining cohort participants, which included more males (p = 0.005).

Table 1. Sample characteristics of 6-, 18- and 30-year-old individuals that wore the accelerometer for six and less than six measurement days.

Fig 1 presents the median of minutes spent per day in LPA and MVPA, respectively, and the median of overall PA. Among adults (18- and 30- year-old), overall PA, MVPA and LPA median were lower on Sundays compared to the rest of the week (p<0.05 for all estimates in both age groups). MVPA medians on Sunday for the 30- year-old group was 0. Among children, overall PA and LPA were lowest on Mondays (overall PA: p = 0.003; LPA: p = 0.012). Differences between week and weekend days regarding LPA, MVPA and overall PA, were only observed among 30-year-olds. In this cohort group, more MVPA (p = 0.006) and overall PA (p<0.001) were performed on weekdays (Fig 2).

Fig 1. Median of minutes spent in light and moderate-to-vigorous physical activity and overall acceleration (expressed in mg) at 6-, 18- and 30-year-old individuals (number of observations in brackets).

Fig 2. Description of physical activity median spent on light and moderate-to-vigorous physical activity (minutes) and overall acceleration (expressed in mg) during weekdays and weekend days among 6-, 18- and 30- year-old individuals.

Fig 3A, 3B and 3C present the reliability coefficient for one to six measurement days. Highest reliability coefficient was observed for overall PA in all three groups. The coefficients regarding overall PA ranged from 0.44 to 0.83 in children and from 0.54 to 0.88 in adults. Also, for overall PA, to achieve a reliability coefficient >0.70, two and three days of measurement were needed in adults and children, respectively. Three and four days of measurements of MVPA was needed to achieve a reliability coefficient >0.70 in 30-year-olds and in 18- and 6-year-olds, respectively. Five days of LPA was necessary to reach a reliability coefficient >0.70 in 6-year-old children, three days for 18-year-old young adults, and four days for 30-years-old adults.

Fig 3.

Reliability coefficient of 1 to 6 measurement days on overall, light and moderate-to-vigorous physical activity among 6-year-old children (A), 18-year-old young adults (B) and 30-year-old adults (C).


Our results show that, five and four days of measurement are needed to adequately estimate all constructs of PA among children and adults, respectively. Considering MVPA, to represent a week of measurement, three days would be needed for children and four days for adults. Overall PA follows a stable pattern through the week, resulting in a smaller number of days needed to an admissible reliability coefficient, being two days for adults and three for children.

A better interpretation of our results requires the understanding of the main characteristics of the analyzed PA constructs. For instance, overall PA demonstrates the behavior with minimum arbitrariness, including all daily movement, from ordinary daily life activities to physical exercise. Similarly, light PA as treated in the present study, non-bouted, represents low intensity activities without any duration restriction, but excluding sedentary time and MVPA. Lastly, when using 10 minutes bouts, MVPA estimates attempt to represent structured activities performed for longer periods, such as sports, long periods of walking and running (as exercise), etc. Therefore, it is expected to find different reliability results among the outcomes, once they represent distinct constructs of PA.

Opposite to our results, other studies have reported differences regarding weekdays and weekend days among children and adults [11,31,32]. This may be due to our specific sample, which includes only 6, 18 and 30-year-old individuals, not an age range (e.g. 18 to 60-year-old individuals). However, it is important to highlight that reliability is not just an age-related phenomenon, it is also a cultural aspect and observed reliability estimates in the present study does not necessarily translate to other settings.

Most previous studies examining reliability, or other measures to establish the minimum number of measurement days to achieve a desirable reliability coefficient have usually examined different PA intensities such as light, moderate, vigorous or MVPA[11,15]. However, overall PA assessment is also relevant, because it encompasses all PA intensities with lower arbitrariness, therefore illustrating a continuum of PA [14]. Also, considering the issue of lack of comparability among accelerometry literature, assessing overall PA expressed in raw acceleration (mg) can facilitate a direct comparison between studies[33].

Regarding children, more days of measurement were necessary to achieve reliability coefficient >0.70, for all three PA measures. This could be due to lack of precision related to small sample size in our youngest group compared to adults and to other studies with children. Addy [12] found that 3.6 days of measurement would be needed to achieve an ICC of 0.70 among preschool children. Another study, using wrist-worn accelerometers showed a reliability coefficient of 0.8 with two days evaluating adults’ MVPA[11]. This evidence shows that many factors could impact on the decision of how many days should be included in accelerometry data analyses, such as: a) reliability coefficient ideal threshold, b) study population, and c) the PA construct of interest.

Therefore, it is important to consider an accelerometry protocol including at least a week of measurement, in order to guarantee reliable estimates of PA. Evidence from two weeks of adults’ 24 h-accelerometry-assessed data, shows that seven days are necessary to estimate overall PA, and between six and nine days for PA intensities[34]. For children, the literature is not consistent on how many days would be necessary for a reliable estimate of habitual PA, ranging from four to nine days[35]. Also, different protocols are needed depending on the research question. Surveillance studies may be satisfied with only one day of data while studies about seasonality need many days in different periods of the year. Furthermore, the present analyzes are recommended for each data collection a posteriori, in order to define the minimum of valid days of measurement in each study, balancing between sample size and reliability estimates[10].

We observed no differences in any of the PA constructs between weekdays and weekend days, except for 30-year-old adults, in which MVPA and overall PA levels were lower on weekend days. Likewise, adults presented lower PA on Sundays in the daily comparison analyzes. This may occur due to the decrease in workload and routine activities on Sunday. The Brazilian legislation determines employers must guarantee the Sunday as holiday for most jobs [36] therefore, many people work from Monday to Saturday. For 18-year-old young adults, the analyzes between weekend and weekdays’ PA did not present significant difference, possibly because Saturday presents similar results than the remaining week, diluting the difference between weekdays and weekend days. Also, children may present different patterns of PA then adults when comparing weekdays and weekend days.

Addressing the difference between weekdays and weekends for adults (30-year-old adults), can be more complex than simply including Saturday or Sunday in the final measure. As shown in our study, MVPA on Sunday was much lower compared to Saturday and weekdays. This finding indicates a potential dissimilar PA pattern for adults on Sundays, which may lead researchers to rethink data analysis and interpretation in this case. The recommendation to include at least one weekend day in the analysis is frequent in the literature [35,37]. However, it seems to be important to discuss if it would be better to include a completely different day, in terms of PA patterns, in the weekly averages, which would have an impact on the weekly estimates; or to treat this day separately in the analysis. These topics might lead to different methodological decisions depending on the research question and the evaluated outcomes.

The present study is based on three large population-based birth cohorts born 11 years apart, with objective measures of PA, providing estimates of the number of days needed to obtain a desirable reliability coefficient on different age groups. Also, to our knowledge, this is the first Brazilian study to evaluate reliability of wrist-worn accelerometry-based PA data.

However, some limitations need to be addressed, such as the decrease in the number of observations when restricting the analyzes to those with six consecutive days, mainly among children. This decrease is due to the use of the non-commercial version of the GENEA accelerometer in the beginning of the data collection. Analyses (not showed here) concluded that such information would not be comparable to the GENEActiv and were not analyzed in the present study. Further explanation on this issue is available in a previous publication[25]. This decrease in sample size is an example of the issue discussed by Migueles, et al. (2017), that elevating valid week criteria provides greater reliability but may also compromise sample size [10].

Also, in the present study we did not exclude sleep periods from the analyses, therefore overall PA represents all body movement in the period. This may impact on our results, specially the comparison between weekdays and weekend days, since sleep duration is increased during weekend days[38]. Lastly, we applied general cut-point thresholds for all age groups since validation/calibration studies of wrist-worn accelerometers are still incipient, which might be considered an important limitation for this area of measurement. In this matter, there is need for more calibration studies and independent validation studies for current available cut-points in order to move towards a consensus[39].

Another issue in accelerometry-based literature is the lack of studies in LMICs, where PA practices differ from high income settings [25,40]. Also, the improvement of measurement tools for physical activity surveillance would enhance the quality and coverage of physical activity data in LMICs [41]. In this sense, regarding the generalizability of our results, it is important to keep in mind that we studied a specific sample, including 6-, 18- and 30-year-old individuals. Although our findings are relevant for LMICs settings, they are not necessarily representative of children or adults in general.

In conclusion, our results show that four and five measurement days are needed to estimate all habitual PA constructs, for children and adults, respectively. Also, among 30-year-old adults, it is important to make efforts towards weekend days measurement. However, final decisions on the research protocol of accelerometer data collection should balance according to each research question, available logistic and potential losses of valid days during the process.

What does this article add?

Since we found differences between weekdays and weekend days physical activity among adults, it is important to make efforts towards weekend days measurement in order to well represent the physical activity variability across the week. The key finding is that the number of measurement days required to estimate physical activity depends on the age range and the construct of physical activity evaluated, varying between three and five measurement days. When evaluating children’s physical activity, there is a need for more days of measurement when compared to adults, regardless of the construct of physical activity.

Supporting information

S1 Table. Median of minutes spent in light and moderate-to-vigorous physical activity and overall acceleration (expressed in mg) at six, 18- and 30-years old individuals in different days of the week.


S1 Dataset. Individual accelerometry and sociodemographic data.



This article is based on data from the studies "Pelotas Birth Cohort, 1982", "Pelotas Birth Cohort, 1993" and “Pelotas Birth Cohort, 2004" conducted by Postgraduate Program in Epidemiology at Universidade Federal de Pelotas with the collaboration of the Brazilian Public Health Association (ABRASCO). From 2004 to 2013, the Wellcome Trust supported all three birth cohort studies. National Support Program for Centers of Excellence (PRONEX), the Brazilian National Research Council (CNPq), and the Brazilian Ministry of Health supported previous phases of the studies. The 1982 Birth Cohort were funded by the International Development Research Center, Overseas Development Administration. The World Health Organization funded the 1982 and 2004 Birth cohorts. And the 1982 and 1993 Birth Cohort were also supported by the European Union. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 001. LMG was funded by the Fundação Médica do Rio Grande do Sul; FCW has a research productivity grant, level 2, granted by the Conselho Nacional de Desenvolvimento Científico e Tecnológico—Brasil (CNPq).


  1. 1. Arena R, McNeil A, Sagner M, Hills AP. The Current Global State of Key Lifestyle Characteristics: Health and Economic Implications. Prog Cardiovasc Dis. 2017;59: 422–429. pmid:28216110
  2. 2. Lee I-M, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380: 219–229. pmid:22818936
  3. 3. Sallis JF, Bull F, Guthold R, Heath GW, Inoue S, Kelly P, et al. Progress in physical activity over the Olympic quadrennium. Lancet. 2016;388: 1325–1336. pmid:27475270
  4. 4. Ainsworth B, Cahalin L, Buman M, Ross R. The Current State of Physical Activity Assessment Tools. Prog Cardiovasc Dis. 2015;57: 387–395. pmid:25446555
  5. 5. Troiano RP, McClain JJ, Brychta RJ, Chen KY. Evolution of accelerometer methods for physical activity research. Br J Sports Med. 2014;48: 1019–1023. pmid:24782483
  6. 6. Chen KY, Bassett DR. The technology of accelerometry-based activity monitors: current and future. Med Sci Sports Exerc. 2005;37: S490–500. Available: pmid:16294112
  7. 7. Cain KL, Sallis JF, Conway TL, Van Dyck D, Calhoon L. Using Accelerometers in Youth Physical Activity Studies: A Review of Methods. J Phys Act Heal. 2013;10: 437–450. pmid:23620392
  8. 8. Ward DS, Evenson KR, Vaughn A, Rodgers AB, Troiano RP. Accelerometer use in physical activity: Best practices and research recommendations. Medicine and Science in Sports and Exercise. 2005. pp. S582–8. pmid:16294121
  9. 9. Pedišić Ž, Bauman A. Accelerometer-based measures in physical activity surveillance: current practices and issues. Br J Sports Med. 2015;49: 219–223. pmid:25370153
  10. 10. Migueles JH, Cadenas-Sanchez C, Ekelund U, Delisle Nyström C, Mora-Gonzalez J, Löf M, et al. Accelerometer Data Collection and Processing Criteria to Assess Physical Activity and Other Outcomes: A Systematic Review and Practical Considerations. Sport Med. 2017;47: 1821–1845. pmid:28303543
  11. 11. Dillon CB, Fitzgerald AP, Kearney PM, Perry IJ, Rennie KL, Kozarski R, et al. Number of Days Required to Estimate Habitual Activity Using Wrist-Worn GENEActiv Accelerometer: A Cross-Sectional Study. PLoS One. 2016;11: e0109913. pmid:27149674
  12. 12. Addy CL, Trilk JL, Dowda M, Byun W, Pate RR. Assessing Preschool Children’s Physical Activity: How Many Days of Accelerometry Measurement. Pediatr Exerc Sci. 2014;26: 103–109. pmid:24092773
  13. 13. Hart TL, Swartz AM, Cashin SE, Strath SJ. How many days of monitoring predict physical activity and sedentary behaviour in older adults? Int J Behav Nutr Phys Act. 2011;8: 62. pmid:21679426
  14. 14. Matthews CE, Keadle SK, Troiano RP, Kahle L, Koster A, Brychta R, et al. Accelerometer-measured dose-response for physical activity, sedentary time, and mortality in US adults. Am J Clin Nutr. 2016;104: 1424–1432. pmid:27707702
  15. 15. Scheers T, Philippaerts R, Lefevre J. Variability in physical activity patterns as measured by the SenseWear Armband: how many days are needed? Eur J Appl Physiol. 2012;112: 1653–1662. pmid:21874552
  16. 16. Trost SG, Pate RR, Freedson PS, Sallis JF, Taylor WC. Using objective physical activity measures with youth: How many days of monitoring are needed? Med Sci Sports Exerc. 2000;32: 426–431. pmid:10694127
  17. 17. Roscoe CMP, James RS, Duncan MJ. Accelerometer-Based Physical Activity Levels Differ between Week and Weekend Days in British Preschool Children. J Funct Morphol Kinesiol. 2019;4: 65.
  18. 18. Treuth MS, Catellier DJ, Schmitz KH, Pate RR, Elder JP, McMurray RG, et al. Weekend and weekday patterns of physical activity in overweight and normal-weight adolescent girls. Obesity. 2007;15: 1782–1788. pmid:17636097
  19. 19. Gonçalves H, Assunção MC, Wehrmeister FC, Oliveira IO, Barros FC, Victora CG, et al. Cohort Profile update: The 1993 Pelotas (Brazil) Birth Cohort follow-up visits in adolescence. Int J Epidemiol. 2014;43: 1082–1088. pmid:24729426
  20. 20. Horta BL, Gigante DP, Goncalves H, dos Santos Motta J, Loret de Mola C, Oliveira IO, et al. Cohort Profile Update: The 1982 Pelotas (Brazil) Birth Cohort Study. Int J Epidemiol. 2015;44: 441–441e. pmid:25733577
  21. 21. Santos IS, Barros AJ, Matijasevich A, Zanini R, Chrestani Cesar MA, Camargo-Figuera FA, et al. Cohort Profile Update: 2004 Pelotas (Brazil) Birth Cohort Study. Body composition, mental health and genetic assessment at the 6 years follow-up. Int J Epidemiol. 2014;43: 1437–1437f. pmid:25063002
  22. 22. Hildebrand M, Van Hees VT, Hansen BH, Ekelund U. Age group comparability of raw accelerometer output from wrist-and hip-worn monitors. Med Sci Sports Exerc. 2014;46: 1816–1824. pmid:24887173
  23. 23. Esliger DW, Rowlands A V., Hurst TL, Catt M, Murray P, Eston RG. Validation of the GENEA accelerometer. Med Sci Sports Exerc. 2011;43: 1085–1093. pmid:21088628
  24. 24. van Hees VT, Gorzelniak L, Dean León EC, Eder M, Pias M, Taherian S, et al. Separating Movement and Gravity Components in an Acceleration Signal and Implications for the Assessment of Human Daily Physical Activity. PLoS One. 2013;8: e61691. pmid:23626718
  25. 25. da Silva IC, van Hees VT, Ramires V V, Knuth AG, Bielemann RM, Ekelund U, et al. Physical activity levels in three Brazilian birth cohorts as assessed with raw triaxial wrist accelerometry. Int J Epidemiol. 2014;43: 1959–1968. pmid:25361583
  26. 26. World Health Organization. Obesity: preventing and managing the global epidemic. Geneva; 2000.
  27. 27. De Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bull World Health Organ. 2007;85: 660–667. pmid:18026621
  28. 28. Rutstein SO. Steps to constructing the new DHS Wealth Index. Usaid. 2015. Available:
  29. 29. Eisinga R, Grotenhuis M te, Pelzer B. The reliability of a two-item scale: Pearson, Cronbach, or Spearman-Brown? Int J Public Health. 2013;58: 637–642. pmid:23089674
  30. 30. Landers R. Computing (ICC) as Intraclass Estimates Correlations of Interrater Reliability in SPSS. The Winnower. 2015; 1–4.
  31. 31. Doherty A, Jackson D, Hammerla N, Plötz T, Olivier P, Granat MH, et al. Large Scale Population Assessment of Physical Activity Using Wrist Worn Accelerometers: The UK Biobank Study. PLoS One. 2017;12: e0169649. pmid:28146576
  32. 32. Nilsson A, Anderssen SA, Andersen LB, Froberg K, Riddoch C, Sardinha LB, et al. Between- and within-day variability in physical activity and inactivity in 9- and 15-year-old European children. Scand J Med Sci Sports. 2008;19: 10–18. pmid:18248534
  33. 33. Bassett DR, Troiano RP, Mcclain JJ, Wolff DL. Accelerometer-based physical activity: Total volume per day and standardized measures. Med Sci Sports Exerc. 2015;47: 833–838. pmid:25102292
  34. 34. Jaeschke L, Steinbrecher A, Jeran S, Konigorski S, Pischon T. Variability and reliability study of overall physical activity and activity intensity levels using 24 h-accelerometry-assessed data. BMC Public Health. 2018;18: 530. pmid:29678152
  35. 35. Trost SG, McIver KL, Pate RR. Conducting accelerometer-based activity assessments in field-based research. Med Sci Sports Exerc. 2005;37: S531–43. Available: pmid:16294116
  36. 36. Brasil. Decreto de lei 5.452—Consolidação das Leis do Trabalho (CLT). 1943.
  37. 37. Hislop J, Law J, Rush R, Grainger A, Bulley C, Reilly JJ, et al. An investigation into the minimum accelerometry wear time for reliable estimates of habitual physical activity and definition of a standard measurement day in pre-school children. Physiol Meas. 2014;35: 2213–2228. pmid:25340328
  38. 38. van Hees VT, Sabia S, Jones SE, Wood AR, Anderson KN, Kivimäki M, et al. Estimating sleep parameters using an accelerometer without sleep diary. Sci Rep. 2018;8: 12975. pmid:30154500
  39. 39. Kim Y, Beets MW, Welk GJ. Everything you wanted to know about selecting the “right” Actigraph accelerometer cut-points for youth, but…: A systematic review. J Sci Med Sport. 2012;15: 311–321. pmid:22306372
  40. 40. Bull F, Armstrong T, Dixon T, Ham S, Neiman A, Pratt M. Comparative Quantification of Health Risks. Global and Regional Burden of Disease Attributable to Selected Major Risk Factors: Physical inactivity. Geneva: World Health Organization; 2004.
  41. 41. Barr AL, Young EH, Sandhu MS. Objective measurement of physical activity: Improving the evidence base to address non-communicable diseases in Africa. BMJ Glob Heal. 2018;3. pmid:30364334