Boutlength: Towards Identification and Presentation of Bout Lengths

Using Physical Activity Counts Data

Muhammad Asad Ullah Khan

, Francesca Gallè

and Giuliana Valerio

Biomedical Engineering Department, University of Engineering & Technology, Lahore, Pakistan

Department of Medical, Movement and Wellbeing Sciences, University of Naples “Parthenope”, Naples, Italy

Keywords: Physical Activity Bouts, Bout Identification Tool, Bout Length Breakdown, Activity Cut Points.

Abstract: Boutlength is an open program developed in a spreadsheet environment to identify and display activity bout

lengths derived from accelerometer count data. The program classifies counts into an active or inactive level

based on user-defined cut-off thresholds. Using a bout qualification criterion specified by a minimum length,

the program identifies, prints, sorts, and computes basic statistics on resulting valid bout string. The program

was tested on a sample of actual database using cut-off values (Sedentary: < 100 counts; Activity: > 2019

counts) and bout qualification criteria (Sedentary: ≥ 5 minutes; Activity: ≥ 2 minutes). Output of the program

is a display of identified Sedentary or Activity bouts depending on the application. Program’s parameters are

modifiable, and the script was designed in a spreadsheet software. Boutlength is still in development and yet

it may be an interesting resource for researchers analysing bout-based physical activity pattern measures.

1 INTRODUCTION

Sedentary behaviour is associated with adverse health

conditions (Bontrup et al., 2019; Millard et al., 2021;

Stoner et al., 2019). Sedentary behaviour is typically

assessed using metrics based on time spent in

inactivity, sitting or reclining (Tremblay et al., 2017).

The patterns of inactive time accumulation

throughout a day are generally described in terms of

bouts (Boerema et al., 2020). A bout is period of

activity level above or below a certain threshold

criterion (Migueles et al., 2017). While the impact of

sedentary bout length on human well-being remains

debatable (De Vries et al., 2022), bout length based

activity evaluation remains important in physical

activity (PA) research (Kuster et al., 2020).

Bout assessment has traditionally relied on self-

reported estimates (Stamatakis et al., 2019) of PA

patterns. Recently, it has shifted on accelerometer-

measured data and devices with software from several

brands including Actical, ActiGraph, and

Promove3D, etc., have gained popularity in activity

pattern representation (Boerema et al., 2020; Kuster

et al., 2020). An increasing number of brands are

offering devices and software allowing analysis of

activity based on different bout data processing

methods. As far as bout characteristics are concerned,

ActiGraph’s accelerometers and software have

become a standard in evaluation of sedentary and

activity bout patterns and statistics.

ActiGraph accelrometers rely on pre-processed

“counts” as indicators of movement intensity to

calculate bout length statistics (LaMunion et al.,

2017). The software Actilife processes raw

movement intensity data to sum it over a period of

measurement (Altenburg et al., 2021) into ‘counts’ of

PA. To estimate a bout, an appropriate cut point is

applied to call the uninterrupted window of time

accumulating prescribed level of activity counts. A

typical sedentary bout is a minimum 10-minute length

of uninterrupted counts of <100 counts per minute

(cpm) (Altenburg et al., 2015) and a few statistical

derivations of sedentary bout lengths have been

linked to health outcomes (Boerema et al., 2015;

Leeger et al., 2019; Peterson et al., 2015).

The growing relevance of bout analysis to health

indicators has led to bout length evaluation in

accelerometer software. Our laboratory had a chance

to test the bout calculation feature in Actigraph’s

Actilife program. It featured specification of cut

point, minimum valid bout length, and display of

basic bout length statistics such as the number, total

time spent in bouts. However, it did not display a

breakdown of individual bout lengths scanned from

the input count data. While Actilife's bout summary

150

Khan, M. A. U., Gallè, F. and Valerio, G.

Boutlength: Towards Identiﬁcation and Presentation of Bout Lengths Using Physical Activity Counts Data.

DOI: 10.5220/0013674000003988

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 13th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2025), pages 150-154

ISBN: 978-989-758-771-9; ISSN: 2184-3201

statistics are useful enough, it did not print individual

bout lengths in its output. We navigated the software

to best of our ability but could not find the feature to

show the length of each bout in the Actilife package

(ver. 6.13.4) at our disposal.

To address this gap, we developed and tested a

program, BoutLength, aimed to compute and display

the length of each bout from counts data. The

program transforms each count into two activity

levels based on a cutoff or threshold specification. It

then scans for consecutive sequences of a specified

level to compute and display series of valid bouts

using a bout qualification criterion. We tested the

program’s capability using actual counts database

developed as part of the work in progression by Khan

et al. (2023). By classifying each indexed count as per

Troiano et al. (2008) into 0 (<100: Sedentary) or 1

(>2019: moderate to vigorous), Boutlength identified

respective bouts by summing the length of classified

sequences. The program can compute and display

individual bout lengths and basic bout statistics on

counts distributed over any unit of time, cutoff

threshold, or desired bout qualification criteria. With

the bout length breakdown printed on screen, users

can conveniently derive additional statistics by

making minor additions to the program built within

the popular spreadsheet environment.

2 METHODOLOGY

The program was scripted in a standard Microsoft

Excel Macro framework with runtime libraries

enabled to allow single-click execution. To ensure

that the program functions as expected, it is important

to refer the data columns in the macro script correctly.

For instance, the first column in the sheet may contain

time index with the counts vector in the second

column. The third column has classified activity

levels (e.g., '1' indicating counts < 100 and '0'

indicating counts > 100). The fourth column shows

an index of the bout number, and the fifth column

displays bout length computed by the program. The

sixth column has sorted bout lengths printed in

ascending order. Successive columns may display

bout statistics calculated from the bout series. It may

be appropriate to name each column in the first row

of the input data sheet as shown in Figure 2.

Input: Counts, Threshold (100 or 2019 or any),

Minimum Bout Length (2 or 5 or any)

Process:

For each count: If count < Threshold, record

"activity"; else, record "inactivity".

Find Bouts:

For recorded activity levels:

If "inactivity", add position to "current

bout".

If "activity": If "current bout" length >=

Bout Length, save as valid bout; clear

"current bout".

Show Results:

If bouts found:

Display bout lengths, total bouts, total

length, longest bout.

Else: Restart.

Algorithm 1: Program flow sequence. Flow chart can be

found in Figure 1. Navigate to availability section for the

link to the script.

3 DISCUSSION

Boutlength was designed for calculating and

displaying individual bout lengths in a simple way.

The program is short, easy to modify, and quick in

operation. Unlike paid bout analysis software, it

provides open access to individual bout lengths all

within the widely used spreadsheet software.

The program enables use of popular activity cut

points by Troiano et al. (2008) as well as common

bout length specification, e.g., Healy et al. (2011) or

Dunstan et al. (2012). It allows direct utilization of

fixed epoch sized activity counts data of any length.

PA counts data can be directly copied into the input

column with minimum prior processing. The

algorithm does not restrict time units over which

counts are accumulated thus making it possible to

calculate bout lengths in minutes or hours and so on.

Another key aspect of the program is that it can

scan for any ‘type’ of bouts. Depending on the type

of activity label to look for, Boutlength can provide

the length of the contrasting activity sequences. In

other words, it can identify ‘breaks’ of moderate to

vigorous PA (MVPA) in sedentary behaviour data or

inactivity bouts enclosed within physical activity

episodes.

The program is our pilot effort towards open-

source bout length analysis. Despite it is currently

limited in features, code accessibility enables

parameter adjustments. This program allows printing

individual bout lengths; a feature absent in Actilife

(ver. 6.13.4) as tested.

Boutlength takes advantage of the spreadsheet

software’s built-in statistical capabilities. Since the

computed bout lengths are displayed in a column,

user can conveniently calculate any number of

Boutlength: Towards Identiﬁcation and Presentation of Bout Lengths Using Physical Activity Counts Data

151

Figure 1: The flow chart of Boutlength algorithm. After insertion of input as counts vector, the program classifies activity into

levels as per cut-off thresholds adapted from Troiano et al. (2008). Next, it scans for sequences of a level as per minimum

bout qualification criteria to print valid bout lengths. For sedentary, a 5-minute and for MVPA, a 2-minute length was specified

as reference from Healy et al. (2011) and Dunstan et al. (2012). Navigate to availability section for the link to the script.

statistics and sketch plots from the bout length series

output.

There is still room for improvements and

additions in the program. For example, the use of a

single cut point restricts activity categorization in two

levels only. Some users may desire multilevel activity

classification. Having enough labels for each activity

type using multiple cut points makes bout length

computation for each activity level a powerful feature

for practical applications. Moreover, some users may

also be interested in ‘interruption’ or ‘tolerance’

feature in the program’s bout length specification. On

the usability aspect, macro programming may seem

complicated to new users. Modifications to the macro

needs correct referencing variables to the data column

in the worksheet.

Given the importance of bout analysis in physical

activity research, not much remains freely available

for bout length presentation to date. A study by Salim

et al. (2024) proposed a tool for bout length

estimation from count data and its code was openly

available for use and modification. Apart from that,

there is no other tool for the provision of individual

bout lengths to the user, to our best knowledge.

Limited number of accessible tools for bout

length presentation may hinder interested users from

calculating bout statistics, such as the median bout

length, or the half-life bout duration, as described in

Boerema et al. (2020) and Chastin et al. (2015).

Boutlength intends to address the gap by enabling

the computation and display of bout length

breakdown from physical activity counts data. The

program is structured for simplicity and

modification in classification threshold, bout

qualification criteria, and activity level

characterization. For future, we are considering

transforming the code into a graphical user interface

for easier navigation.

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

152

Figure 2: Snippet of program input and output display space in the spreadsheet. Top (Green): The first row contains header

information. The first two columns are input comprising time unit index (minutes) and a sample observation as counts data

from Khan et al. (2023). The third column is a result of activity classification based on MVPA cut-off (counts > 2019: 1,

otherwise: 0). The subsequent columns are identified MVPA bouts. From the 50-minute recording for a participant, it found

4 bouts of MVPA activity that are at least 2 minutes (or longer). The longest of them was 8 uninterrupted minutes. The next

columns are basic statistical operations applied on the identified bout length series. Bottom (Yellow): Another participant’s

count data was analysed for sedentary bout identification in the second column. The operations of third column onwards use

Troiano’s sedentary threshold for level assignment (counts < 100: 1, otherwise: 0) and bout length specification (5 minutes

or longer). From the 50-minute recording, it identified 2 sedentary bouts of 5 and 8 minutes. Likewise, successive columns

apply basic statistics on the identified sedentary bout string.

4 CONCLUSION

Boutlength was designed to print individual bout

lengths derived from activity count data. It allows the

use of a cut-off threshold to identify sequences of

activity levels into usable bout strings as per defined

bout qualification criteria. The program operates in a

commonly used spread sheet software environment

that allows easy data handling, and changes to bout

classification parameters. The printed bout length

series can be used for statistical analysis, deriving

new measures of bout patterns, or creating plots. The

open availability of code promotes transparency,

scrutiny, and adaptability. Future work will prioritize

the development of a graphical user interface, making

the script more accessible and convenient for a wide

range of users.

5 APPLICATION

The program was demonstrated using two

observations of Actigraph wGT3X-BT vector

magnitude time series (30Hz, 1 minute epoch)

obtained from Khan et al. (2023). Each of the two

participants had one ankle mounted unit and sat in a

classroom lesson for 50 minutes. Both sedentary and

MVPA cut-off values were adapted from Troiano et

al. (2008) as Sedentary: 0-99 cpm, MVPA: > 2019

cpm).

Minimum valid bout lengths were adapted from

Dunstan et al. 2012 (MVPA: length ≥ 2 min) and

Healy et al. 2011 (Sedentary: length ≥ 5 min). The

program was executed once for each observation by

changing the specification for both types of activities.

As a result, cells of third column and onwards were

filled with output as shown in Figure 2. Indexed and

sorted bout length breakdown are printed along with

basic statistics.

Depending on the computer performance, the

spreadsheet macro processes the input counts column,

performs classification, computes bouts, and prints

output within seconds.

AVAILABILITY

Boutlength is available in a public repository at

www.github.com/asadkh21/fidgeting.git

Counts database used for the demonstration can

be provided upon reasonable request to Khan et al.

(2023).

ACKNOWLEDGEMENT

The authors acknowledge Osama Malik, a Calgary-

based Software Developer, for his contribution to the

Boutlength: Towards Identiﬁcation and Presentation of Bout Lengths Using Physical Activity Counts Data

153

development of the pilot version of the program.

REFERENCES

Altenburg, Teatske M., Wang, X., van Ekris, E., Andersen,

L. B., Møller, N. C., Wedderkopp, N., & Chinapaw, M.

J. M. (2021). The consequences of using different

epoch lengths on the classification of accelerometer

based sedentary behaviour and physical activity. PLOS

ONE, 16, e0254721.

Altenburg, T.M., de Niet, M., Verloigne, M., De

Bourdeaudhuij, I., Androutsos, O., Manios, Y., …

M.J.M., C. (2015). Occurrence and duration of various

operational definitions of sedentary bouts and cross-

sectional associations with cardiometabolic health

indicators: The ENERGY-project. Preventive Medicine,

71, 101–106.

Boerema, S., Essink, G., Tönis, T., Van Velsen, L., &

Hermens, H. (2015). Sedentary Behaviour Profiling of

Office Workers: A Sensitivity Analysis of Sedentary

Cut-Points. Sensors, 16, 22.

Boerema, S. T., van Velsen, L., Vollenbroek, M. M., &

Hermens, H. J. (2020). Pattern measures of sedentary

behaviour in adults: A literature review. DIGITAL

HEALTH, 6, 205520762090541.

Bontrup, C., Taylor, W. R., Fliesser, M., Visscher, R.,

Green, T., Wippert, P.-M., & Zemp, R. (2019). Low

back pain and its relationship with sitting behaviour

among sedentary office workers. Applied Ergonomics,

81, 102894.

Chastin, S. F. M., Winkler, E. A. H., Eakin, E. G., Gardiner,

P. A., Dunstan, D. W., Owen, N., & Healy, G. N. (2015).

Sensitivity to Change of Objectively-Derived Measures

of Sedentary Behavior. Measurement in Physical

Education and Exercise Science, 19, 138–147.

De Vries, L. P., Pelt, D. H. M., Van Der Ploeg, H. P.,

Chinapaw, M. J. M., De Geus, E. J. C., & Bartels, M.

(2022). The association between well-being and a large

variation of accelerometer-assessed physical activity

and sedentary behavior measures. Mental Health and

Physical Activity, 22, 100446.

Khan, M., Gallè, F., Ballarin, G., Calella, P., Cerullo, G.,

Liguori, G., & Valerio, G. (2023). Accelerometer Based

Body Movement Quantification in Classroom Lectures:

Seated Activity Comparison Between Body Regions:

Proceedings of the 11th International Conference on

Sport Sciences Research and Technology Support, 144–

150. Rome, Italy: SCITEPRESS - Science and

Technology Publications.

Kuster, R. P., Grooten, W. J. A., Baumgartner, D., Blom, V.,

Hagströmer, M., & Ekblom, Ö. (2020). Detecting

prolonged sitting bouts with the ActiGraph GT3X.

Scandinavian Journal of Medicine & Science in Sports,

30, 572–582.

LaMunion, S. R., Bassett, D. R., Toth, L. P., & Crouter, S.

E. (2017). The effect of body placement site on

ActiGraph wGT3X-BT activity counts. Biomedical

Physics & Engineering Express, 3, 035026.

Leeger-Aschmann, C. S., Schmutz, E. A., Zysset, A. E.,

Kakebeeke, T. H., Messerli-Bürgy, N., Stülb, K., …

Kriemler, S. (2019). Accelerometer-derived physical

activity estimation in preschoolers – comparison of cut-

point sets incorporating the vector magnitude vs the

vertical axis. BMC Public Health, 19, 513.

Migueles, J. H., Cadenas-Sanchez, C., Ekelund, U., Delisle

Nyström, C., Mora-Gonzalez, J., Löf, M., … Ortega, F.

B. (2017). Accelerometer Data Collection and

Processing Criteria to Assess Physical Activity and

Other Outcomes: A Systematic Review and Practical

Considerations. Sports Medicine, 47, 1821–1845.

Millard, L. A. C., Tilling, K., Gaunt, T. R., Carslake, D., &

Lawlor, D. A. (2021). Association of physical activity

intensity and bout length with mortality: An

observational study of 79,503 UK Biobank participants.

PLOS Medicine, 18, e1003757.

Peterson, N. E., Sirard, J. R., Kulbok, P. A., DeBoer, M. D.,

& Erickson, J. M. (2015). Validation of Accelerometer

Thresholds and Inclinometry for Measurement of

Sedentary Behavior in Young Adult University Students:

MEASUREMENT OF SEDENTARY BEHAVIOR.

Research in Nursing & Health, 38, 492–499.

Salim, A., Brakenridge, C. J., Lekamlage, D. H., Howden,

E., Grigg, R., Dillon, H. T., … Winkler, E. A. H. (2024).

Detection of sedentary time and bouts using consumer-

grade wrist-worn devices: A hidden semi-Markov

model. BMC Medical Research Methodology, 24, 222.

Stamatakis, E., Ekelund, U., Ding, D., Hamer, M., Bauman,

A. E., & Lee, I.-M. (2019). Is the time right for

quantitative public health guidelines on sitting? A

narrative review of sedentary behaviour research

paradigms and findings. British Journal of Sports

Medicine, 53, 377–382.

Stoner, L., Willey, Q., Evans, W. S., Burnet, K., Credeur, D.

P., Fryer, S., & Hanson, E. D. (2019). Effects of acute

prolonged sitting on cerebral perfusion and executive

function in young adults: A randomized cross-over trial.

Psychophysiology, 56, e13457.

Tremblay, M. S., Aubert, S., Barnes, J. D., Saunders, T. J.,

Carson, V., Latimer-Cheung, A. E., … Chinapaw, M. J.

M. (2017). Sedentary Behavior Research Network

(SBRN) – Terminology Consensus Project process and

outcome. The International Journal of Behavioral

Nutrition and Physical Activity, 14, 75.

Troiano, R. P., Berrigan, D., Dodd, K. W., Masse, L. C.,

Tilert, T., & McDowell, M. (2008). Physical activity in

the United States measured by accelerometer. Medicine

and Science in Sports and Exercise, 40(1), 181-188.

Dunstan, D. W., Kingwell, B. A., Larsen, R., Healy, G. N.,

Cerin, E., Hamilton, M. T., Shaw, J. E., Bertovic, D. A.,

Zimmet, P. Z., Salmon, J., & Owen, N. (2012). Breaking

up prolonged sitting reduces postprandial glucose and

insulin responses. Diabetes care, 35(5), 976–983.

Healy, G. N., Matthews, C. E., Dunstan, D. W., Winkler, E.

A., & Owen, N. (2011). Sedentary time and cardio-

metabolic biomarkers in US adults: NHANES 2003-

06. European heart journal, 32(5), 590–597.

icSPORTS 2025 - 13th International Conference on Sport Sciences Research and Technology Support

154