Abstract
An investigation was conducted among all female laboratory workers in the county of Funen to test the hypothesis that work in laboratories during pregnancy increases the risk of spontaneous abortion. Laboratory workers from hospital, university, and other public laboratories and from laboratories in industry were investigated. A socially comparable reference group less exposed to chemicals at work included office workers, technical assistants and designers, physiotherapists, and occupational therapists. Information was obtained through a postal questionnaire study in May 1980 and from hospital records. Neither the occupation in any single category of laboratory nor the alleged exposure to any single chemical during pregnancy showed a significant increase in the odds ratio for spontaneous abortion. A selection bias was found among the non-respondents with respect to both occupation and outcome of pregnancy, but the influence on the results was slight, owing to the small number of non-respondents. The higher rates of hospital admission for spontaneous abortion among the cases compared with the controls indicates that register data are not more valid than self reported data. This leads to several methodological problems that ought to be studied in further detail.