Feasibility of remote biospecimen collection among rural study participants to enhance monitoring of dietary chemical exposures: A cohort study

Background: Given limited access to fresh foods, rural residents have high rates of canned food intake, potentially resulting in exposures to bisphenols, which have been linked to diabetes. The goal of this study was to evaluate the feasibility of remotely collecting urine samples to assess rural residents for dietary chemical exposures. Methods: We collected in-person and remote urine samples among rural residents to measure bisphenol levels. Day 0 samples were collected in-person, immediately frozen, and transported on dry ice to prevent any sample loss. Rural residents then remotely collected Days 1 to 3 urine samples, froze samples at home, and mailed them back using insulated cold packs. We examined daily bisphenol levels among 30 participants to assess the stability of biospecimens collected remotely among rural residents. Results: Of 304 participants, return rate of remotely collected urine collection kits was 97.4%. For the subsample, daily average bisphenol levels were not higher on Day 0 compared to those collected on Days 1 to 3. The intraclass correlation coefficient (ICC) for bisphenol A was 0.063 and bisphenol F was 0.095, demonstrating no correlation between daily values among study participants. While the ICC was statistically significant for bisphenol S, correlation was low at 0.238. Conclusions: We found high sample return rates and no significant sample loss or degradation of bisphenols, demonstrating the feasibility of collecting biospecimen remotely among rural residents. The low correlation between daily levels suggests high daily variability consistent with the short half-lives of these chemicals, highlighting the need to collect multiple samples over time.