Periodic Reporting for period 5 - MobiliSense (Air pollution and noise exposure related to personal transport behaviour: short-term and longer-term effects on health)
Reporting period: 2021-07-01 to 2021-12-31
- Building on methods from Epidemiology, Geography, and Transport sciences, the objectives of MobiliSense are to quantify the contribution of personal transport to the air pollution and noise exposure of individuals; to compare the air pollution and noise exposure in the different transport modes; and to investigate whether total and transport-related personal exposure to air pollutants and noise are associated with short-term and two-year changes in respiratory and cardiovascular health.
- The MobiliSense project is strongly related to ongoing policy efforts both at the national and European levels (i) in the field of Transport, (ii) for the regulation of air pollution, and (iii) for the regulation of noise. Overall, in the context of the insufficient evidence available attributable to the limitations of scientific studies, it is important to develop innovative research strategies to derive more reliable data on transport-related environmental exposures and on their health effects, to support international, European, and national policy efforts. The main and critical limitation of previous literature on this topic is that studies have evaluated the influence of air pollutants and noise separately; previous studies may therefore be heavily biased.
- Overall, the MobiliSense project explores the short-term and longer-term effects that air pollution and noise exposures related to personal transport behaviour may have on respiratory health and cardiovascular health. Its final aim is to better understand how traveling with the different transport modes implies exposures generating health effects.
- The MobiliSense project is strongly related to ongoing policy efforts both at the national and European levels (i) in the field of Transport, (ii) for the regulation of air pollution, and (iii) for the regulation of noise. Overall, in the context of the insufficient evidence available attributable to the limitations of scientific studies, it is important to develop innovative research strategies to derive more reliable data on transport-related environmental exposures and on their health effects, to support international, European, and national policy efforts. The main and critical limitation of previous literature on this topic is that studies have evaluated the influence of air pollutants and noise separately; previous studies may therefore be heavily biased.
- Overall, the MobiliSense project explores the short-term and longer-term effects that air pollution and noise exposures related to personal transport behaviour may have on respiratory health and cardiovascular health. Its final aim is to better understand how traveling with the different transport modes implies exposures generating health effects.
The following work was performed over the course of the project:
- We conceived the sampling frame of the MobiliSense cohort (sample size, Grand Paris as the study territory, definition of the geographic strata for sampling, inclusion and exclusion criteria, protocol for taking non-response into account).
- We developed and performed an extensive testing of the different questionnaires (including spatial and non-spatial questionnaires, and smartphone-based questionnaires). We developed electronic versions of these questionnaires.
- We conceived an additional application for managing the recruitment of participants (Participant manager, Device manager).
- We validated the final list of sensors to use in the data collection, and prepared the measurement protocols for these sensors.
- We obtained all the legal allowances for conducting this data collection and we were able to obtain the official label from the National Statistical Information Committee.
- We obtained the right to sample our participants from the French Population Census, and performed the sampling with the French National Institute of Statistical Studies.
- We prepared all the operational and management procedures for the recruitment of participants.
- We hired survey technicians and trained them to the many complex technologies to be used during the data collection.
- We collected data for a sample of MobiliSense participants using all the planned web and smartphone applications and sensors, for the first wave of the study.
- A large fraction of these participants was seen for the second wave of the study between one and two years after the first wave, with the same applications and sensors.
- For the analysis of data, we hired one post-doctorate student, one PhD student, and two study engineers. We also hired one Master student per year to assist us in the analysis of the MobiliSense data.
- With this team of investigators, we progressively processed all the sources of data (questionnaires, sensors, etc.); we implemented procedures of processing and cleaning, especially for each of the sensor’s data; and we defined the numerous variables needed for the analyses.
- We published a protocol article of the MobiliSense project in BMJ Open.
- We already explored how the exposure to the various air pollutants, and also noise, differ by transport modes (see publication in Environment International). We explored the relationship between the various air pollutants and both resting blood pressure and ambulatory blood pressure. We explored the relationship between noise exposure and stress. Numerous analyses still have to be conducted.
- We conceived the sampling frame of the MobiliSense cohort (sample size, Grand Paris as the study territory, definition of the geographic strata for sampling, inclusion and exclusion criteria, protocol for taking non-response into account).
- We developed and performed an extensive testing of the different questionnaires (including spatial and non-spatial questionnaires, and smartphone-based questionnaires). We developed electronic versions of these questionnaires.
- We conceived an additional application for managing the recruitment of participants (Participant manager, Device manager).
- We validated the final list of sensors to use in the data collection, and prepared the measurement protocols for these sensors.
- We obtained all the legal allowances for conducting this data collection and we were able to obtain the official label from the National Statistical Information Committee.
- We obtained the right to sample our participants from the French Population Census, and performed the sampling with the French National Institute of Statistical Studies.
- We prepared all the operational and management procedures for the recruitment of participants.
- We hired survey technicians and trained them to the many complex technologies to be used during the data collection.
- We collected data for a sample of MobiliSense participants using all the planned web and smartphone applications and sensors, for the first wave of the study.
- A large fraction of these participants was seen for the second wave of the study between one and two years after the first wave, with the same applications and sensors.
- For the analysis of data, we hired one post-doctorate student, one PhD student, and two study engineers. We also hired one Master student per year to assist us in the analysis of the MobiliSense data.
- With this team of investigators, we progressively processed all the sources of data (questionnaires, sensors, etc.); we implemented procedures of processing and cleaning, especially for each of the sensor’s data; and we defined the numerous variables needed for the analyses.
- We published a protocol article of the MobiliSense project in BMJ Open.
- We already explored how the exposure to the various air pollutants, and also noise, differ by transport modes (see publication in Environment International). We explored the relationship between the various air pollutants and both resting blood pressure and ambulatory blood pressure. We explored the relationship between noise exposure and stress. Numerous analyses still have to be conducted.
- Major progresses were made beyond the state of the art in the highly innovative data collection that was implemented. A strength of the project is that it systematically relied on objective measurement approaches for the assessment of exposures, confounders, and health outcomes. The protocol used passive sensors of location, behaviour, environmental conditions, and health (GPS receivers, accelerometers, air pollution and noise sensors, heart rate monitors, and ambulatory blood pressure monitors) and active monitors requiring an action of the subject (blood pressure at rest and spirometry). The use of some of these monitors is extremely innovative: the recently developed ambulatory blood pressure monitor that was used (TensioMed Arteriograph) measures central blood pressure (which is more predictive of target organ damage than brachial blood pressure) and aortic pulse wave velocity and the augmentation index as markers of arterial stiffness; the BioPatch has allowed us to assess heart rate variability but also accelerometry and the respiratory rate; and no study had relied on a repeated smartphone survey to assess respiratory symptoms as close as possible from their onset. A particularly innovative aspect of the project was also the simultaneous monitoring of air pollution and noise with personal monitors, as almost no study had done.
- A second measurement strength of the project, integrating methodologies from Public health and Transport sciences, was related to the precise measurement of personal transport behaviour over 6 days using GPS receivers and a GPS data-based electronic survey of visited places, trips, and transport modes. This approach allowed us to decompose in a precise way the 6-day follow-up period into time spent at the different visited places and trips and trip stages (segments of trips with a unique mode), permitting to ascribe the data collected with the behavioural, environmental, and health sensors to each trip or visited place time segment of the mobility survey.
- Progresses of the project beyond the state of the art is that with these data we have been able to document the personal exposure to air pollutants and noise according to the different transport modes at an unprecedented level of accuracy and using data from a very large number of trips made by participants from all the Grand Paris area. Other progresses accomplished so far are that we were able to accurately explore the relationship between the exposure to air pollutants (either single pollutants or cocktails or mixtures of air pollutants, taking into account inhaled quantities and not only concentrations) and the acute blood pressure response. With its sensor-based accurate data, MobiliSense provides high quality and novel information on this topic. Given the massive amount of data that we still have to analyze, other innovative findings on the health effects of transport exposures will be produced in the coming months and years.
- Finally, we have continued to develop algorithms for the smart recognition of transport, which is critical for the future assessment of transport effects on health, and critical for various disciplines that have interest in mobility (geography, urban planning, physical activity science, behavioral social sciences, etc.). A progress beyond the state of the art will be to integrate the new algorithms that we have developed during the project for the automatic recognition of mobility, trips, and transport modes into a smartphone application dedicated to collecting data in studies. We intend to apply to Proof of concept funding to expand and finalize this innovative work.
- A second measurement strength of the project, integrating methodologies from Public health and Transport sciences, was related to the precise measurement of personal transport behaviour over 6 days using GPS receivers and a GPS data-based electronic survey of visited places, trips, and transport modes. This approach allowed us to decompose in a precise way the 6-day follow-up period into time spent at the different visited places and trips and trip stages (segments of trips with a unique mode), permitting to ascribe the data collected with the behavioural, environmental, and health sensors to each trip or visited place time segment of the mobility survey.
- Progresses of the project beyond the state of the art is that with these data we have been able to document the personal exposure to air pollutants and noise according to the different transport modes at an unprecedented level of accuracy and using data from a very large number of trips made by participants from all the Grand Paris area. Other progresses accomplished so far are that we were able to accurately explore the relationship between the exposure to air pollutants (either single pollutants or cocktails or mixtures of air pollutants, taking into account inhaled quantities and not only concentrations) and the acute blood pressure response. With its sensor-based accurate data, MobiliSense provides high quality and novel information on this topic. Given the massive amount of data that we still have to analyze, other innovative findings on the health effects of transport exposures will be produced in the coming months and years.
- Finally, we have continued to develop algorithms for the smart recognition of transport, which is critical for the future assessment of transport effects on health, and critical for various disciplines that have interest in mobility (geography, urban planning, physical activity science, behavioral social sciences, etc.). A progress beyond the state of the art will be to integrate the new algorithms that we have developed during the project for the automatic recognition of mobility, trips, and transport modes into a smartphone application dedicated to collecting data in studies. We intend to apply to Proof of concept funding to expand and finalize this innovative work.