Assessment of physical activity during commuting to school with devices in Spanish youthState of the art, methodological approaches, and practical applications

Résumé

Introduction: The regular practice of physical activity (PA) is associated with numerous health benefits in children and adolescents. Nevertheless, in recent years there has been a worrying increase in the prevalence of physical inactivity and increased sedentary time in these populations. Thus, it is critical to implement interventions to address this situation and it has been observed that schools can play a key role in this. In particular, active commuting to and from school (ACS), such as walking or cycling, is considered an effective strategy to increase daily PA levels in the young population. However, to date, there is no standardized protocol or agreed methodology for device-measured PA related to ACS (ACS-PA). This may result in an overestimation or underestimation of ACS-PA and difficulties in comparing and interpreting studies in this topic. Aims: Therefore, this International Doctoral Thesis has a threefold aim: (i) to review and compile methodologies used to assess ACS-PA, as well as to analyze ACSPA levels by conducting a meta-analysis; (ii) to examine trip characteristics and movement behaviors (i.e., sedentary time, light physical PA [LPA], and moderate- to-vigorous PA [MVPA]) related to ACS using accelerometry and Global Positioning System (GPS) data, as well as to analyze the level of agreement between tools to identify the start and end times of the ACS; and (iii) to study the effects of a school-based cycling intervention on movement behaviors device-measured in different space-time domains (i.e., commuting to/from school, overall transportation, and total day) and other health-related outcomes in adolescents. Methods: To address these objectives seven studies were conducted: three review studies, three cross-sectional studies, and one Cluster Randomized Clinical Trial (CRCT). For two of the cross-sectional studies and the intervention study, data from the PACO project were used, which were complemented with data from the CiudActiva project. The three cross-sectional studies and the CRCT study used Actigraph GT3X+ accelerometers and a Qstarz BT-Q1000XT GPS placed on the hip of the participants for seven days for the former device, and for seven or two days for the latter device. After their combination through HABITUS software and R package PALMSplusR, detailed information was obtained on PA levels and trips characteristics to and from school, as well as sedentary time and PA levels in other space-time domains. In addition, other health-related outcomes were analyzed such as Body Mass Index, sleep duration, and breakfast composition. Statistical analyses included meta- regression for meta-analysis, mixed models, and Bland & Altman plots for cross-sectional and intervention studies. Main findings: Regarding the objective i, an overview of the methodologies used for the device-measured ACS-PA was provided, as well as practical considerations to be considered for its correct assessment. In addition, a meta-analysis was carried out which indicated that ACS could provide up to 48% of the PA recommendations during school days. Regarding the objective ii, through the combination of accelerometry and GPS data, this Thesis observed that depending on the trip direction, PA levels and trip characteristics varied, and in addition, those adolescents who walked to and/or from school showed a lower association with sedentary time and higher levels of PA in other space-time domains compared to those adolescents who used other modes of commuting. Moreover, it showed that self-reported measures had a high level of agreement for establishing the start and end times of walking trips to/from school compared to GPS. Finally, in the objective iii, the present Thesis observed no significant differences post school-based cycling intervention between intervention and control groups in sedentary time and PA levels in different space-time domains measured all with accelerometry and GPS. Nor were differences found in other health-related outcomes. Conclusion: The results of this International Doctoral Thesis provided an overview of the methodologies used for device-measured ACS-PA and offer practical considerations for assessing when, where, and how much PA is related to ACS. In addition, the meta-analysis study concludes that ACS can help complete up to 48% of school-day PA recommendations. Moreover, after the application of the practical recommendations, it is shown that the benefits of PA differ according to the trip direction, where adolescents who walked to and/or from school showed less sedentary time and higher PA levels associated with other space-time domains compared to those who used multimodal or motorized-vehicle modes. Moreover, similarities and differences between the methodologies to report ACS-PA related to identifying the start and end times of walking trips to/ from school were highlighted. Finally, no significant differences were obtained in sedentary time and PA levels in different space-time domains or in other health- related outcomes after a school-based cycling intervention between the intervention and control groups, thus future interventions should improve methodological aspect of the intervention to obtain the expected improvements.