Cumulative chemical risk and associated toxic effect classes across US industries from regulatory inspections

Abstract Objectives Workers are often exposed to multiple chemical agents, which should be accounted for in risk assessment. However, few data are yet available to prioritize specific agents, toxic effects, or industries for cumulative risk assessment (CRA). This study aimed to characterize the cumulative chemical risk in US workplaces using over 600,000 exposure measurements carried out in numerous industries by the Occupational Safety and Health Administration (OSHA), to quantify the frequency of cumulative risk exceedances, and to identify recurring combinations of agents associated with these elevated risks. Methods Data on airborne exposure to 195 chemical agents from the OSHA database for the period 1971 to 2021 were grouped by workplace situation (WS, a job title within a company within a calendar year) to observe the co-presence of chemicals. WSs were then linked via their chemicals to one or several toxicological classes from MiXie, a tool which systematically associates several hundreds of chemicals to a standard list of 24 toxicological classes. CRA was then conducted for all WSs across toxicological classes by calculating the agent-specific hazard quotients (HQ, ratio of a chemical measurement to its occupational exposure limit) and summing them into the WS- and class-specific hazard index (HI, sum of HQ of chemicals relevant to a toxicological class). WSs with HI 1 were flagged as “hazardous,” and hazardous WSs with all HQ 1 were flagged as “missed risk” as assessment of a single agent would not have detected overexposure. Frequent itemset mining (FIM) was used to identify frequent combinations of chemicals in hazardous multiexposed WSs, by toxicological class and across 51 industry groups. Results From 607,676 measurements, 38,512 unexposed WSs and 123,118 exposed WSs were identified. Among the latter, 33% were considered multiexposed (median 3 agents/WS). The following toxicological classes were associated with more than half of all exposed WSs: Carcinogenicity and/or mutagenicity (60%), central nervous system damage (58%), lower airway damage (54%), and upper airway damage (53%). Toxicological classes co-occurred frequently across WSs, with 94% of WSs linked to ≥2 classes. Clustering analyses revealed strong co-association patterns, especially between irritation effects (eyes/airways) and between metal-related effects (blood/kidney/nervous system). Across the 24 toxicological classes, the proportion of hazardous WSs among multiexposed WSs was highest for carcinogenicity and/or mutagenicity (65%), while the proportion of missed risk among hazardous multiexposed WSs was highest for ototoxicity (19%). Among all industries, the highest proportion of hazardous multiexposed WSs was in boat building and repairing (81%), while the highest proportion of missed risk was in general industrial machinery and equipment (30%). FIM revealed recurring hazardous combinations of agents—primarily involving solvents and metals—which varied across toxicological classes and industries (eg toluene/xylene and copper compounds/iron oxides frequent for upper airway damage; toluene/styrene/acetone frequent for boat building and repairing). Conclusions Overall, our study suggests that workers in US workplaces are often exposed to hazardous combinations of chemical agents targeting a wide variety of organ systems. Our results should help prioritize industries and agents likely necessitating CRA.