Combining sensor tracking with a GPS-based mobility survey to better measure physical activity in trips: public transport generates walking

Table 5 Accelerometer-derived number of steps taken in trips according to the main mode^a used in the trip

Classifications of trips according to the mode	Number of steps taken per trip: median (10th and 90th percentiles)	Cumulated number of steps in trips per individual per 8 h of accelerometer wear time:^b median (10th and 90th percentiles)	% of trip-related steps taken attributable to these trips
Crude classification
Entirely walked trips	425 (4, 1587)	554 (61, 2176)	43.2
Other active modes	283 (41, 1007)	0 (0, 123)	3.3
Public transport	1352 (499, 2453)	397 (0, 1913)	37.3
Private motorized	86 (10, 453)	120 (0, 549)	11.7
Other^c	14 (2, 26)	0 (0, 0)	0.0
Multi-mode	1213 (445, 2402)	0 (0, 234)	4.4
Detailed classification
Entirely walked trips	425 (4, 1587)	554 (61, 2176)	43.2
Other active modes	283 (41, 1007)	0 (0, 123)	3.3
Bus/coach	771 (169, 1910)	0 (0, 243)	4.3
Metro	1267 (576, 2292)	0 (0, 892)	12.6
Suburban train	1933 (859, 2876)	0 (0, 265)	5.4
Tramway	1286 (271, 1957)	0 (0, 0)	1.5
Private motorized (driver)	83 (10, 420)	72 (0, 476)	9.4
Private motorized (passenger)	118 (12, 705)	0 (0, 105)	2.2
Other^c	14 (2, 26)	0 (0, 0)	0.0
Multi-mode	1466 (638, 2480)	0 (0, 1165)	18.1

RECORD MultiSensor Study, 284 participants, 8728 trips, 20,564 trip stages
^aWe consider the steps accumulated over the entire trip (including with non-walking modes and during transfer episodes)
^bThe overall number of days of mobility survey could not be used as the denominator as there is nonwear of the accelerometer within it. As a consequence, we chose as the denominator the number of units of 8 h of accelerometer wear time
^cLong distance train and plane

ISSN: 1479-5868