Oregon Health & Science University, Portland, OR
Physical Activity Actigraphy Among Truck Drivers: Accounting for Seat Vibrations
- Presented on May 30, 2014
Background: Measuring and increasing physical activity among commercial truck drivers is a health imperative due to elevated rates of obesity and associated health problems in the population. However, actigraphy measures of physical activity are likely inﬂated by large truck seat vibrations during driving periods.
Purpose: In preliminary studies recordings from two identical accelerometers, one placed on the seat and the other on a driver’s hip, were poorly correlated. The current study evaluated a method to remove truck seat vibration error variance from actigraphy data in order to produce accurate measures of caloric expenditure.
Methods: A subsample of truck drivers enrolled in a randomized trial of a weight loss and health promotion intervention participated in the study (n = 47). Drivers wore an Actigraph GT3x+ on their hip and self-recorded daily driving times for 7-10 days. A second accelerometer (HOBO Pendant G Logger) was placed on the truck seat. An algorithm to detect yes/no driving time from seat accelerometer data was developed, and results were compared to yes/no driving time records from driver logs. Both sources of driving information were used to estimate caloric expenditure during non-driving periods, and these estimates were compared to raw unfiltered actigraphy data.
Results: Twenty-eight drivers provided sufﬁcient self-reported driving data for analyses. Raw agreement between seat accelerometer data and driver logs averaged 86.9% (SD = 9.2). Sensitivity of the seat accelerometer for identifying driving time as reported by drivers was .85. Speciﬁcity of the seat accelerometer for identifying non-driving time was .93. Daily physical activity caloric expenditure estimates were 531.9 (SD = 289.1) unﬁltered, 449.5 (SD = 283.7) when ﬁltered with driver data, and 385.2 (SD = 260.1) when ﬁltered with seat accelerometer data.
Conclusions: Unﬁltered actigraphy measures inﬂated caloric expenditure estimates by 146.7 kcal/day on average. A HOBO pendant G logger on the truck seat reliably identiﬁed driving periods and subsequently allowed more accurate estimates of caloric expenditure during non-driving periods among commercial truck drivers.
- Brad Wipﬂ i 1
- Ryan Olson 1
- Uche Uba 2
- Louis Moore, III 1
- Noal Clemons 1
- Sharon Thompson 1
- Kevin Bransford 1
Stanford University, Stanford, CA
Sponsor: Diane Elliot, MD, FACSM
ACSM 2014 Annual Meeting