Associations of socioeconomic status with transport-related physical activity: combining a household travel survey and accelerometer data using random forests

Abstract

International audienceBackgroundSocioeconomic disparities in active transport have been documented in household travel surveys. However, active transport in these studies was operationalized with self-reported measures, which poorly approximate physical activity. Unfortunately, objective accelerometer data are very expensive to obtain in large-scale travel studies.PurposeTo benefit from a large sample and objective physical activity data, this study linked a cross-sectional household travel survey with accelerometer data from a small sample to investigate the association between socioeconomic disadvantage and the daily level of transport-related moderate-to-vigorous physical activity (T-MVPA) in an adult population (35–83 years).MethodsAccelerometer data for participants’ trips over 7 days from the RECORD GPS Study (7138 trips, 229 participants) were combined with information on participants’ trips over 1 day from the Global Transport Survey (Enquête Globale Transport, EGT) (82084 trips, 21332 participants). Trip-level T-MVPA data from the RECORD sample were used to train a random forests prediction model, enabling the prediction of T-MVPA for each participant׳s trip from EGT. The associations between socioeconomic indicators and daily T-MVPA were analyzed with negative binomial regression models.ResultsAn average time of 18.9 min (95% confidence interval: 18.6–19.2) of T-MVPA was found for these 35–83 year old adults. The education level had a positive association with T-MVPA. Household income had a negative association with T-MVPA, especially for those people without a motorized vehicle.ConclusionsThis study developed a methodology exporting precise sensor-based knowledge to a large survey sample to shed light on population-level socioeconomic disparities in transport-related physical activity

An evaluation of transport mode shift policies on transport-related physical activity through simulations based on random forests

“…Transport interventions to promote physical activity traditionally focus on walking and biking [ 11 ]. Recently, public transport has been found to also contribute significantly to population physical activity levels [ 4 , 11 , 28 , 32 ]. The results in this study confirm this finding, with a mean of 6.8 min of daily MVPA related to public transport, corresponding to 36% of the total daily T-MVPA.…”

“…‘minutes of walking during one week’). We therefore used data integration to add an accelerometer measure of transport-related physical activity to the large survey dataset [ 16 , 28 ]. Even though the prediction model had a high accuracy, real measured accelerometer data would definitely be preferable.…”

“…The data integration consisted of predicting T-MVPA in EGT based on the data of the RECORD GPS Study [ 28 ], using a random forest prediction model [ 29 ]. Random forest is machine learning prediction model based on decision trees that can be used for predicting both continuous and categorical variables [ 30 ].…”

“…With the EGT data originating from the same population and the same region, we hypothesized that the model provided accurate predictions of T-MVPA for the trips in the EGT dataset. The data integration process was previously described in detail [ 28 ].…”

Methods, theory, and behaviour change

“…Motivational messages were included as information about what the user has accomplished. Brondeel et al (2016) linked household travel survey data with accelerometry in a sub-sample to investigate socio-economic differences in transport-related physical activity in adults. They used random forest prediction plots.…”