GPS tracking methods for spatiotemporal air pollution exposure assessment: comparison and challenges in study implementation.

Abstract

Background: Epidemiological studies investigating long-term health effects of air pollution typically only consider the residential locations of the participants, thereby ignoring the space-time activity patterns that likely influence total exposure. This paper, part of a study in which residential-only and mobility-integrated exposures were compared in two tracking campaigns, reflects on GPS device choice, privacy, and recruitment strategy.

Methods: Tracking campaigns were conducted in Switzerland and the Netherlands. Participants completed a baseline questionnaire, carried a GPS device (SODAQ) for 2 weeks, and used a smartphone app for a time activity diary. The app also tracked GPS, albeit less frequently. Tracks were combined with air pollution surfaces to quantify NO₂ and PM_2.5 exposure by activity.

Results: In Switzerland, participants were recruited from the COVCO-Basel cohort (33% recruitment rate; 489 of 1,475). In the Netherlands, -random recruitment was unsuccessful (1.4% rate; 41 of 3,000). Targeted recruitment with leaflets and a financial incentive (25 Euro voucher) increased participation to 189. Comparisons between smartphone app and SODAQ device data showed moderate to high correlations (R2 > 0.57) for total NO₂ exposure and NO₂ exposure at home in both study areas. Activity-specific correlations ranged from 0.43 to 0.63. PM_2.5 correlations in Switzerland were moderate to high, but lower in the Netherlands (R² = 0.28-0.58), due to smaller spatial contrast in observed PM_2.5 levels (RMSE < 0.68 µg/m³).

Conclusions: Tracking can be effectively conducted using a mobile app or GPS device. The app's low-frequency GPS readings (every 3-4 min) were sufficient for long-term air pollution exposure assessment. For finer-scale readings, a dedicated GPS device is recommended. Tracking campaigns are crucial for studying personal exposure to air pollution but face challenges due to low recruitment rates and strict privacy regulations. Leveraging an existing cohort can improve recruitment, while targeted leaflet distribution with financial incentives can enhance participation in studies without a pre-recruited group.