Key factors for selecting PM2.5 and ozone exposure assessment methods in epidemiological studies

Environmental epidemiological studies often use both station-monitored and personal air pollutant exposures, which frequently yield different results. We aimed to identify key considerations when choosing between these measures. In a panel study of 37 college students assessed six times across three seasons for cardiorespiratory outcomes, personal PM2.5 and O3 exposures were monitored for 5 days with wearable sensors before each health assessment, alongside concurrent measurements from nearby monitoring stations. The association between station-monitored and personal concentrations was stronger for PM2.5 (regression coefficient: 0.51 ± 0.16) than for O3 (regression coefficient: 0.19 ± 0.15). Both station-monitored and personal PM2.5 were associated with decreased forced expiratory volume in the first second (FEV1), forced vital capacity (FVC), and increased fractional exhaled nitric oxide (FeNO). In contrast, only station-monitored O3 was associated with decreased FEV1, FVC, increased FeNO, and worsening augmentation index (AI) and blood pressure. Personal O3 showed mostly null associations or even “seemingly beneficial” associations with AI, FEV1, and FVC. These findings suggest station-monitored PM2.5 can serve as a reasonable proxy for personal exposure in studies with minimal indoor PM2.5 sources. However, this may be unsuitable for O3, given its high spatial variability and potential differences in exposure to ozone-derived reaction products.