Background National surveys of the French working population estimate that approximately 15% of all workers are exposed to at least three different chemical agents at work. However, the most prevalent coexposure situations and their associated health risks remain relatively understudied.

Objective To characterize occupational coexposure situations in France using available data from two occupational exposure databases.

Methods We extracted personal measurement data from the Colchic and Scola databases for the period 2010–2019. We selected 118 chemical agents that had ≥100 measurements with detected concentrations, including 31 carcinogens (IARC groups 1, 2A, and 2B). We grouped measurements by work situation (WS, combination of sector, occupation, task, and year). We characterized the mixtures across WS using frequent itemset mining methods.

Results We retained 275,213 samples from 32,670 WS, encompassing 4,692 unique mixtures. Workers in thirty-two percent of all WS were exposed to ≥2 agents (median 3 agents/WS, interquartile interval 2–5) and 13% of all WS contained ≥2 carcinogens (median 2 carcinogens/WS, maximum 14). The most frequent coexposures across all agents were ethylbenzene-xylene (1,550 WS), quartz-cristobalite (1,417 WS), and toluene-xylene (1,305 WS). Prevalent combinations of carcinogens also included hexavalent chromium-lead (368 WS) and benzene-ethylbenzene (314 WS). Agents with the lowest proportions of coexposure were wood dust (6% of WS exposed to at least one other agent) and asbestos (8%). Tasks with the highest proportions of coexposure to carcinogens include electric arc welding (37% of WS with coexposure), polymerization and distillation (34%), construction drilling and excavating (34%), and water collection and treatment (32%).

Conclusion Coexposure to multiple chemical agents, including carcinogens, was highly prevalent in the databases, and should be taken into account when assessing exposure risks in the workplace. However, these databases do not necessarily represent a random sample of the working population, thereby limiting the generalizability of our findings.