Spontaneous abortion in women employed in plastics manufacture.

An analysis was made of 193 current and previous pregnancies of women employed at time of conception in the plastics industry, derived from a survey of work and pregnancy in 56,012 women in Montreal, 1982-84. The ratio of observed to expected spontaneous abortions, corrected by logistic regression for seven nonoccupational confounding variables, was elevated (1.27), but not significantly (90% confidence interval [CI] 0.91-1.72), in women engaged in process work. The ratio was significantly raised (1.58, 90% CI 1.02-2.35) in women whose work included the processing of polystyrene. Nonsignificant excesses in women working with polyolefine (1.30) and polyvinyl (1.39) were present only when exposure also included other types of plastic; no excess was observed in women whose work did not include polystyrene. The number of women (17) exposed to polyurethane was too small for any conclusion to be drawn about this group.