The Validity and Reliability of Pedometers for Assessing Physical Activity in Children
Updated: Feb 26, 2022
November 2020
Introduction
With over a quarter (27%) of children aged between two to fifteen being classified as overweight or obese (Corrigan, and Scarlett, 2020), it is crucial that we understand the levels of physical activity in this population. As physical activity is a multi-dimensional construct, it is difficult to find a highly valid measure which applies to every mode of physical activity (Sylvia et al., 2013). And with each measure, there is also a difficulty in analysing the data to determine an individual’s level of physical activity. Blend this with inaccuracies in data collection, obtaining valid and reliable results can be challenging. This literature review aims to assess the validity and reliability of pedometers as a tool to measure levels of physical activity in children and make recommendations about the usefulness of these devices to researchers.
Purpose and Engineering of Pedometers
A pedometer is an instrument that counts the number of steps taken and may also attempt to estimate distance too. However, even with stride length adjustment, greater error is expected versus counting steps. Modern pedometers are clipped to a belt on the hip, or less commonly worn on the ankle (Montoye, et al., 1996).
Electronic pedometers involve a spring suspended, horizontal lever arm that moves up and down due to the vertical accelerations of the hip during walking and running. This consequentially opens and closes the electrical circuit, allowing the computer to count the number of steps taken, which is then presented on the digital display (Welk, 2002).
The Reliability of Pedometers for Measuring Steps
There are a variety of models of pedometer, each with their own proprietary mechanism and software, giving them varying levels of reliability. Schneider et al. (2003) tested ten pedometers over four-hundred metre walk, and found that the Kenz Lifecorder (KZ), New Lifestyles NL-2000 (NL) and the Yamax Digiwalker SW-701 (DW) were the most reliable, with values within ±3% of the actual steps taken, 95% of the time. Intra-model reliability was also remarkably high in these models (>0.99) and also the Omron HJ-105 (OM), However, the OM was the worst performing unit, along with the Sportline 330 (SL330)), with values over an order of magnitude lower than the best performing pedometers- within ±37% of the actual step value 95% of the time. Therefore, out of the ten pedometers tested, KZ, NL and DW are the most reliable- with OM being highly reliable also, but relatively inaccurate. Nevertheless, OM is a variable sensitivity pedometer, which was simply set at the middle setting. Calibrating this device to the correct sensitivity would likely yield more accurate results.
In children specifically, Mitre et al. (2009) found that in twenty-seven children aged eleven (±1) the OM and the Yamax Digi-Walker SW-200 (SW-200) pedometers were deemed inaccurate; underreporting step count- particularly at lower speeds and in children with a higher BMI. The study found that accelerometer-based step count (Stepwatch) had a significantly lower percentage error and was more reliable for both normal and overweight children. However, unlike the study by Schneider et al. (2003), there was no clarification of where the variable sensitive switch was placed on the OM. Nevertheless, the study found that the percentage error of the OM was 36 ±8%- very similar to the ±37% from Schneider et al. Therefore, it can be argued that both studies did not appropriately use the OM, but this is inexplicit in the study by Mitre et al.
The Relationship between the Sensitivity of Pedometers and Reliability
Specify and sensitivity are concepts commonly used in medicine; meaning the “ability of a test to correctly classify an individual as disease- free” and the “ability of a test to correctly classify an individual as ′diseased′”, respectively (Parikh et al., 2008). Translate this to pedometers, the specificity of a pedometer is its ability to correctly classify each step, and discount false positive steps and the sensitivity is its ability to capture all steps and minimise false negatives. In effect, a highly specific pedometer will output a lower number of steps than a highly sensitive pedometer.
Melanson et al. (2004) found that both spring levered pedometers and piezoelectric pedometers had high accuracy (96%) at speeds above three miles per hour, however, at speeds below this, piezoelectric pedometers are more sensitive and therefore are able to count more of the false negatives than the spring levered counterparts. Therefore, piezoelectric pedometers may be more appropriate for children under five, who walk at speeds around 1 mile per hour at average, but offers little benefit to children over the age of five who walk at 2.9 miles per hour on average (Pinheiro et al., 2014).
The correct balance between sensitivity and specificity for maximum accuracy and reliability is a conundrum that is difficult to solve. The sensitivity required for children depends on their age (Melanson et al., 2004), BMI and walking speed (Mitre et al., 2009). The variable sensitivity switch on the OM may help to tailor the sensitivity to each user, however in practice, accurately determining the correct sensitivity may be very time consuming. In addition, walking speed is dynamic in free living children, as they frequently transition from high speed to low speed locomotion, hence the sensitivity can never be correct in all situations.
The Validity of Pedometers for Assessing Physical Activity in Children
The validity of pedometers can be determined firstly by determining how strong the relationship is between step count and levels of physical activity, and secondly determining whether step count is an appropriate measurement of physical activity in children.
Determining the Relationship between Step Count and Levels of Physical Activity
There are no government guidelines for step count in children, rather five to eighteen-year olds should engage in sixty minutes of moderate to vigorous physical activity each day (Davis, 2019). Several papers have attempted to convert this guideline into steps.
Adams et al. (2013) suggested a minimum step per day recommendation of 9000 pedometer steps per day for children and adolescents. However, this study was based on the ActiGraph 7164 accelerometer data from the NHANES Physical Activity Monitor (PAM) dataset, which is more sensitive than a research grade pedometer. The stated pedometer steps were a result of deducting steps that were taken below 500 activity counts/min and therefore, the results of this study likely have reduced validity in comparison to one that directly used a research grade pedometer.
Colley et al. (2012) also suggested lowering the step target in Canada from 13,500 steps per day to 12,000, as the 13,500-step cut point was underreporting the number of children achieving the sixty minutes per day physical activity guidelines. However, this guideline is still 33% higher than what Adams et al. (2013) suggested. Tudor-Locke et al. (2011) disagrees with one defined cut point and suggests specifically applied different targets each age group and gender, ranging from 11,000 to 15,000 steps per day.
Evidently, there is an approximate ratio that correlates step count to minutes of physical activity. Though unclear, these results show that reliable pedometers can be a valid indicator of a child’s level of physical activity.
The Appropriateness of Step Count to the Modes of Activity in Children
Many children partake in activities that do not involve walking or running- such as cycling and swimming. These activities will not contribute to the step count, and therefore levels of physical activity may be underestimated in children. However, in 2019 41% of all children aged five to sixteen usually walked to school, whereas only 3% of children usually cycled (Slocombe, 2020). This is an indicator that many children’s activities- whether that be walking, running, or playing, include steps, and hence will be used to indicate their level of physical activity. It is likely that step count will frequently be underestimated due to these activities, and therefore the use of other means of recording physical activity would be useful in creating more valid results, as any activities that were done without the use of the pedometer can be manually added to the total number of minutes being physically active.
Conclusion
Pedometers were found to be highly accurate and reliable measures of steps in children, with lower accuracy in those with higher BMI, and at lower walking speeds. To get correct results, a pedometer with suitable sensitivity must be selected, and although adjustable sensitivity pedometers, such as the OM, seem useful, they may be more challenging in practice, and can result in large percentage error.
Step count appears to be a highly valid measure of determining a child’s level of physical activity, as it accounts for a large proportion of their activity. However, it still seems unclear where the cut points should be placed to achieve the recommended sixty minutes of moderate to vigorous physical activity each day, presenting challenges in analysing step data.
With this strong validity and reliability, along with the relatively low cost of around £20 per unit (Yamax, 2020), ease of use, and likely higher adherence rate because of this, pedometers may be a strong tool to use to assess levels of physical activity in children in real world applications. However, pedometers must be used with caution, and stronger results may arise when used in conjunction with other means of measurement.
References
Adams, M.A., Johnson, W.D. and Tudor-Locke, C. (2013) 'Steps/day translation of the moderate-to-vigorous physical activity guideline for children and adolescents', International Journal of Behavioral Nutrition and Physical Activity, 10(49) [Online]. Available at: https://doi.org/10.1186/1479-5868-10-49 (Accessed: 6th November 2020).
Colley, R.C., Janssen, I. and Tremblay, M.S. (2012) 'Daily step target to measure adherence to physical activity guidelines in children', Medicine and Science in Sports and Exercise, 44(5), pp. 977-982 [Online]. Available at: https://doi.org/10.1249/MSS.0b013e31823f23b1 (Accessed: 10th November 2020).
Corrigan, D. and Scarlett, M. (2020) Health Survey Northern Ireland: First results 2018/19, Belfast: Department of Health [Online]. Available at: https://www.health-ni.gov.uk/news/health-survey-northern-ireland-first-results-201819 (Accessed: 5th November 2020).
Davis, S.C., Atherton, F., McBride, M. and Calderwood, C. (2019) UK Chief Medical Officers' physical activity guidelines, United Kingdom: Department of Health and Social Care, Llwodraeth Cymru Welsh Government, Department of Health Northern Ireland and the Scottish Government [Online]. Available at: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/832868/uk-chief-medical-officers-physical-activity-guidelines.pdf (Accessed: 12th November 2020).
Mitre, N., Lanningham-Foster, L., Foster, R. and Levine, J.A. (2009) 'Pedometer accuracy in children: Can we recommend them for our obese population?', Pediatrics, 123(1), pp. 127-131 [Online]. Available at: https://doi.org/10.1542/peds.2008-1908 (Accessed: 12th November 2020).
Montoye, H.J., Kemper, H.C.G., Satis, W.H.M., Washburn, R.A. (1996) Measuring physical activity and energy expenditure, Champaign: Human Kinetics.
Parikh, R., Mathai, A., Parikh, S., Chandra Sekhar, G. and Thomas, R. (2008) 'Understanding and using sensitivity, specificity and predictive values', Indian journal of ophthalmology, 56(1), pp. 45-50 [Online]. Available at: https://doi.org/10.4103/0301-4738.37595 (Accessed: 12th November 2020).
Pinheiro, A.T.K, Hokugo, A. and Nishino. T. (2014) 'Walking speed of children by age under the lead of nursery school teacher on a sidewalk and overpass bridge stairs during an urban evacuation drill', Journal of Architecture and Planning, 79(697), pp. 583-588 [Online]. Available at: https://doi.org/10.3130/aija.79.583 (Accessed: 12th November 2020).
Schneider, P.L., Crouter, S.E., Lukajic, O. and Bassett, D.R. (2003) 'Accuracy and reliability of 10 pedometers for measuring steps over a 400-m walk', Medicine & Science in Sports & Exercise, 35(10), pp. 1779-1784 [Online]. Available at: https://doi.org/10.1249/01.MSS.0000089342.96098.C4 (Accessed: 11th November 2020).
Slocombe, M. (2020) Walking and cycling statistics, England: 2019, England: Department for Transport [Online]. Available at: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/906698/walking-and-cycling-statistics-england-2019.pdf (Accessed: 12th November 2020).
Sylvia, L.G., Bernstein, E.E., Hubbard, J.L., Keating, L., and Anderson, E.J. (2013) 'Practical guide to measuring physical activity', Journal of the Academy of Nutrition and Dietetics, 114(2), pp. 119-208 [Online]. Available at: https://jandonline.org/article/S2212-2672(13)01426-3/fulltext (Accessed: 6th November 2020).
Tudor-Locke, C., Craig, C.L., Beets, M.W., Belton, S., Cardon, G.M., Duncan, S., Hatano, Y., Lubans, D.R., Olds, T.S., Raustorp, A., Rowe, D.A., Spence, J.C., Tanaka, S. and Blair, S.N. (2011) 'How many steps/day are enough? for children and adolescents', International Journal of Behavioral Nutrition and Physical Activity, 8(78) [Online]. Available at: https://doi.org/10.1186/1479-5868-8-78 (Accessed: 10th November 2020).
Welk, G.J. (2002) Physical activity assessment for health-related research, Champaign: Human Kinetics.
Yamax (2020) Yamax SW200 Digi-walker pedometer, Available at: https://www.yamax.co.uk/yamax-pedometers/sw200-digi-walker/ (Accessed: 12th November 2020).
Notes:
This paper was originally written in November 2020 as part of my undergraduate degree. It is not published in an academic journal, nor peer-reviewed. It was posted here on the 24th February 2022.