Sources of variance in daily physical activity levels as measured by an accelerometer

Charles E. Matthews, Barbara Ainsworth, Raymond W. Thompson, David R. Bassett

Research output: Contribution to journalArticle

353 Citations (Scopus)

Abstract

Purpose: To examine sources of variance in objective measures of physical activity in a group of healthy adults (N= 92) participating in a physical activity measurement study. Methods: Physical activity was assessed for up to 21 consecutive days using the Computer Science Applications (CSA) accelerometer. Day-of-the-week effects were evaluated for activity counts (ct·min-1·d-1, ct·d-1) and time (min·d-1) spent in inactivity (0-499 ct), moderate-1 (500-1951 ct), and moderate-2-vigorous activity (≥1952 ct). Random effects models were employed to estimate variance components for subject, day of the week, and residual error from which the number of days of assessment required to achieve 80% reliability were estimated. Results: Physical inactivity was lower on weekend days, and Saturday was the least inactive day of the week for both men and women. Inter-individual variation, or differences between subjects, was proportionally the largest source of variance (55-60% of total) in accelerometer counts and time spent in moderate to vigorous activity. Differences within subjects (intra-individual variation) accounted for 30-45% of the overall variance, and day-of-the-week effects accounted for 1-8%. For activity counts, and time spent in moderate to vigorous activity, at least 3-4 d of monitoring were required to achieve 80% reliability. Reliable measures of physical inactivity required at least 7 d of monitoring. Conclusion: These findings provide insight for understanding the behavioral variability in the activity patterns of adults and suggest that reliable measures of activity behaviors require at least 7 d of monitoring.

Original languageEnglish (US)
Pages (from-to)1376-1381
Number of pages6
JournalMedicine and Science in Sports and Exercise
Volume34
Issue number8
StatePublished - 2002
Externally publishedYes

Fingerprint

Exercise

Keywords

  • Ambulatory monitoring
  • Assessment
  • Exercise
  • Reliability

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health
  • Physical Therapy, Sports Therapy and Rehabilitation
  • Orthopedics and Sports Medicine

Cite this

Matthews, C. E., Ainsworth, B., Thompson, R. W., & Bassett, D. R. (2002). Sources of variance in daily physical activity levels as measured by an accelerometer. Medicine and Science in Sports and Exercise, 34(8), 1376-1381.

Sources of variance in daily physical activity levels as measured by an accelerometer. / Matthews, Charles E.; Ainsworth, Barbara; Thompson, Raymond W.; Bassett, David R.

In: Medicine and Science in Sports and Exercise, Vol. 34, No. 8, 2002, p. 1376-1381.

Research output: Contribution to journalArticle

Matthews, CE, Ainsworth, B, Thompson, RW & Bassett, DR 2002, 'Sources of variance in daily physical activity levels as measured by an accelerometer', Medicine and Science in Sports and Exercise, vol. 34, no. 8, pp. 1376-1381.
Matthews, Charles E. ; Ainsworth, Barbara ; Thompson, Raymond W. ; Bassett, David R. / Sources of variance in daily physical activity levels as measured by an accelerometer. In: Medicine and Science in Sports and Exercise. 2002 ; Vol. 34, No. 8. pp. 1376-1381.
@article{f6e5da276bc84defa986b8c97c08186a,
title = "Sources of variance in daily physical activity levels as measured by an accelerometer",
abstract = "Purpose: To examine sources of variance in objective measures of physical activity in a group of healthy adults (N= 92) participating in a physical activity measurement study. Methods: Physical activity was assessed for up to 21 consecutive days using the Computer Science Applications (CSA) accelerometer. Day-of-the-week effects were evaluated for activity counts (ct·min-1·d-1, ct·d-1) and time (min·d-1) spent in inactivity (0-499 ct), moderate-1 (500-1951 ct), and moderate-2-vigorous activity (≥1952 ct). Random effects models were employed to estimate variance components for subject, day of the week, and residual error from which the number of days of assessment required to achieve 80{\%} reliability were estimated. Results: Physical inactivity was lower on weekend days, and Saturday was the least inactive day of the week for both men and women. Inter-individual variation, or differences between subjects, was proportionally the largest source of variance (55-60{\%} of total) in accelerometer counts and time spent in moderate to vigorous activity. Differences within subjects (intra-individual variation) accounted for 30-45{\%} of the overall variance, and day-of-the-week effects accounted for 1-8{\%}. For activity counts, and time spent in moderate to vigorous activity, at least 3-4 d of monitoring were required to achieve 80{\%} reliability. Reliable measures of physical inactivity required at least 7 d of monitoring. Conclusion: These findings provide insight for understanding the behavioral variability in the activity patterns of adults and suggest that reliable measures of activity behaviors require at least 7 d of monitoring.",
keywords = "Ambulatory monitoring, Assessment, Exercise, Reliability",
author = "Matthews, {Charles E.} and Barbara Ainsworth and Thompson, {Raymond W.} and Bassett, {David R.}",
year = "2002",
language = "English (US)",
volume = "34",
pages = "1376--1381",
journal = "Medicine and Science in Sports and Exercise",
issn = "0195-9131",
publisher = "Lippincott Williams and Wilkins",
number = "8",

}

TY - JOUR

T1 - Sources of variance in daily physical activity levels as measured by an accelerometer

AU - Matthews, Charles E.

AU - Ainsworth, Barbara

AU - Thompson, Raymond W.

AU - Bassett, David R.

PY - 2002

Y1 - 2002

N2 - Purpose: To examine sources of variance in objective measures of physical activity in a group of healthy adults (N= 92) participating in a physical activity measurement study. Methods: Physical activity was assessed for up to 21 consecutive days using the Computer Science Applications (CSA) accelerometer. Day-of-the-week effects were evaluated for activity counts (ct·min-1·d-1, ct·d-1) and time (min·d-1) spent in inactivity (0-499 ct), moderate-1 (500-1951 ct), and moderate-2-vigorous activity (≥1952 ct). Random effects models were employed to estimate variance components for subject, day of the week, and residual error from which the number of days of assessment required to achieve 80% reliability were estimated. Results: Physical inactivity was lower on weekend days, and Saturday was the least inactive day of the week for both men and women. Inter-individual variation, or differences between subjects, was proportionally the largest source of variance (55-60% of total) in accelerometer counts and time spent in moderate to vigorous activity. Differences within subjects (intra-individual variation) accounted for 30-45% of the overall variance, and day-of-the-week effects accounted for 1-8%. For activity counts, and time spent in moderate to vigorous activity, at least 3-4 d of monitoring were required to achieve 80% reliability. Reliable measures of physical inactivity required at least 7 d of monitoring. Conclusion: These findings provide insight for understanding the behavioral variability in the activity patterns of adults and suggest that reliable measures of activity behaviors require at least 7 d of monitoring.

AB - Purpose: To examine sources of variance in objective measures of physical activity in a group of healthy adults (N= 92) participating in a physical activity measurement study. Methods: Physical activity was assessed for up to 21 consecutive days using the Computer Science Applications (CSA) accelerometer. Day-of-the-week effects were evaluated for activity counts (ct·min-1·d-1, ct·d-1) and time (min·d-1) spent in inactivity (0-499 ct), moderate-1 (500-1951 ct), and moderate-2-vigorous activity (≥1952 ct). Random effects models were employed to estimate variance components for subject, day of the week, and residual error from which the number of days of assessment required to achieve 80% reliability were estimated. Results: Physical inactivity was lower on weekend days, and Saturday was the least inactive day of the week for both men and women. Inter-individual variation, or differences between subjects, was proportionally the largest source of variance (55-60% of total) in accelerometer counts and time spent in moderate to vigorous activity. Differences within subjects (intra-individual variation) accounted for 30-45% of the overall variance, and day-of-the-week effects accounted for 1-8%. For activity counts, and time spent in moderate to vigorous activity, at least 3-4 d of monitoring were required to achieve 80% reliability. Reliable measures of physical inactivity required at least 7 d of monitoring. Conclusion: These findings provide insight for understanding the behavioral variability in the activity patterns of adults and suggest that reliable measures of activity behaviors require at least 7 d of monitoring.

KW - Ambulatory monitoring

KW - Assessment

KW - Exercise

KW - Reliability

UR - http://www.scopus.com/inward/record.url?scp=0036323437&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036323437&partnerID=8YFLogxK

M3 - Article

C2 - 12165695

AN - SCOPUS:0036323437

VL - 34

SP - 1376

EP - 1381

JO - Medicine and Science in Sports and Exercise

JF - Medicine and Science in Sports and Exercise

SN - 0195-9131

IS - 8

ER -