Abstract
Employment duration is often used as a surrogate for cumulative occupational exposure, although a single blood specimen reflects a shorter and chemical-specific window. We compared three blood metals and three aromatic volatile organic compounds (VOCs) by recorded construction job duration in pooled National Health and Nutrition Examination Survey (NHANES) data from 2009 to 2012. Using both occupation-questionnaire routes gave 404 adults (234 with ≤15 years and 170 with >15 years). The longer-duration group was older by a mean of 20.7 years. In the observed data, cadmium and lead medians were higher in the longer-duration group, whereas total mercury and VOC patterns varied by analyte. Survey-weighted cadmium detection differed by 17.7 percentage points (95% confidence interval 8.6 to 26.8), and its geometric-mean ratio was 1.40 (1.11 to 1.77); the corresponding ratios were 1.34 (1.11 to 1.62) for lead and 0.98 (0.82 to 1.18) for total mercury. VOC detection differences had intervals spanning zero. Age-overlap and cotinine analyses weakened a simple duration interpretation. Thirty-seven participants reported lifetime cancer history. Exploratory Bayesian kernel machine regression fits are reported in the supplement, and no clear evidence of a tenure-related cancer-mixture interaction was found. The cross-sectional biomarker contrasts cannot attribute exposure to construction work or rank biological residence times.