TY - JOUR T1 - Validity of retrospective occupational exposure estimates of lead and manganese in a case–control study JF - Occupational and Environmental Medicine JO - Occup Environ Med SP - 680 LP - 687 DO - 10.1136/oemed-2019-105744 VL - 76 IS - 9 AU - Jean-François Sauvé AU - Joemy M Ramsay AU - Sarah J Locke AU - Pamela J Dopart AU - Pabitra R Josse AU - Dennis D Zaebst AU - Paul S Albert AU - Kenneth P Cantor AU - Dalsu Baris AU - Brian P Jackson AU - Margaret R Karagas AU - GM Monawar Hosain AU - Molly Schwenn AU - Alison Johnson AU - Mark P Purdue AU - Stella Koutros AU - Debra T Silverman AU - Melissa C Friesen Y1 - 2019/09/01 UR - http://oem.bmj.com/content/76/9/680.abstract N2 - Objectives The validity of surrogate measures of retrospective occupational exposure in population-based epidemiological studies has rarely been evaluated. Using toenail samples as bioindicators of exposure, we assessed whether work tasks and expert assessments of occupational metal exposure obtained from personal interviews were associated with lead and manganese concentrations.Methods We selected 609 controls from a case–control study of bladder cancer in New England who had held a job for ≥1 year 8–24 months prior to toenail collection. We evaluated associations between toenail metal concentrations and five tasks extracted from occupational questionnaires (grinding, painting, soldering, welding, working near engines) using linear regression models. For 139 subjects, we also evaluated associations between the toenail concentrations and exposure estimates from three experts.Results We observed a 1.9-fold increase (95% CI 1.4 to 2.5) in toenail lead concentrations with painting and 1.4-fold increase (95% CI 1.1 to 1.7) in manganese concentrations with working around engines and handling fuel. We observed significant trends with increasing frequency of both activities. For lead, significant trends were observed with the ratings from all three experts. Their average ratings showed the strongest association, with subjects rated as possibly or probably exposed to lead having concentrations that were 2.0 and 2.5 times higher, respectively, than in unexposed subjects (ptrend <0.001). Expert estimates were only weakly associated with manganese toenail concentrations.Conclusions Our findings support the ability of experts to identify broad contrasts in previous occupational exposure to lead. The stronger associations with task frequency and expert assessments support using refined exposure characterisation whenever possible. ER -