Biomonitoring of manganese in blood, urine and axillary hair following low-dose exposure during the manufacture of dry cell batteries

Objectives: A cross-sectional study was carried out on 100 workers from three different workplace areas in a dry cell battery manufacturing plant and on 17 currently nonexposed referents, to examine the relationship between the external exposure to manganese dioxide (MnO₂) and the body burden of manganese in blood, urine and hair. Methods: Inhalable dust was measured gravimetrically after stationary active sampling. Manganese was analyzed in dust samples, blood, urine and axillary hair by atomic absorption spectro- metry. Results: The average air concentrations of manganese in the three workplace areas were 4 μg/m³ (range: 1–12 μg/m³), 40 μg/m³ (12–64 μg/m³) and 400 μg/m³ (137–794 μg/m³). Manganese in blood and axillary hair correlated with airborne manganese in group-based calculations but not on an individual level. The manganese concentrations varied between 3.2 μg/l and 25.8 μg/l in the blood (mean: 12.2 ± 4.8 μg/l) and between 0.4 μg/g and 49.6 μg/g in hair (mean: 6.2 ± 6.2 μg/g in the proximal sequence), respectively. The results for the nonexposed referents were 7.5 ± 2.7 μg/l (mean) in the blood (range: 2.6–15.1 μg/l) and 2.2 ± 1.8 μg/g (mean) in axillary hair (range: 0.4–6.2 μg/g). In these matrices, manganese differed significantly between the highly exposed workers and both the reference and the low-exposure group. Manganese in blood revealed the lowest background variance. No differences for manganese in urine were observed between workers (mean: 0.36 ± 0.42 μg/l, range: 0.1–2.2 μg/l) and referents (mean: 0.46 ± 0.47 μg/l, range: 0.1–1.7 μg/l). Conclusions: Manganese in blood is a specific and suitable parameter for the biomonitoring of MnO₂ exposure, although its validity is limited to group-based calculations. Urinary manganese failed to allow a differentiation between exposed workers and referents. The suitability of manganese analysis in hair for biomonitoring purposes suffers from a relatively great background variation as well as from analytical problems. Received: 11 December 1998 / Accepted: 17 July 1999