Comparison of three methods for measuring the time spent in physical activity

Area of Science:

• Exercise Physiology
• Biomedical Engineering
• Public Health

Background:

• Accurate measurement of physical activity (PA) is crucial for health research and interventions.
• Various methods exist to assess PA, including objective measures (e.g., accelerometers) and subjective measures (e.g., logs, surveys).
• Understanding the agreement between different PA measurement tools is essential for data interpretation and study design.

Purpose of the Study:

• To compare three methods for measuring daily physical activity (PA) over a 21-day period in adults.
• To evaluate the agreement between a motion sensor (CSA monitor), a PA log, and a telephone survey in quantifying PA.
• To assess the consistency of PA estimates across different intensity levels (resting/light, moderate, hard/very hard).

Main Methods:

• 83 adults participated, wearing a CSA monitor and completing a daily PA log.
• Participants also completed weekly telephone surveys on walking and moderate-to-vigorous PA.
• Data were analyzed using descriptive statistics and Spearman rank-order correlations to assess agreement between methods.

Main Results:

• Modest to good agreement was found between different CSA cut-point methods (r = 0.43-0.94).
• Correlations between CSA monitors and PA logs ranged from r = 0.22 to r = 0.36.
• Correlations between survey data and PA logs varied by activity intensity, with lower agreement for hard/very hard activities (r < 0.09).

Conclusions:

• Motion sensors, PA logs, and surveys are all capable of reflecting physical activity.
• These methods do not consistently provide similar estimates for time spent in different PA intensity categories.
• Validation studies highlight the need for careful consideration of measurement tools in PA research.