diagnosed in 1987–1990 for study inclusion. On the basis of three exposed cases, a multivariate analysis revealed a slightly increased risk of MM with exposure to carbamate insecticides (as a class) for those reporting exposure at any point in time and for those who used the insecticides in their profession. Although an a priori matrix was used in assessing exposure, which probably reduced recall bias, there is still the potential for nondifferential misclassification of exposure, which would tend to reduce the risk estimate. The authors also stated that the MM diagnosis may have been less accurate among residents who lived farther from the main specialized hospitals in Italy. The small number of exposed cases also limits the value of the findings of this study.
Two studies of MM mortality and occupation as farmer obtained from death certificates in Iowa (Burmeister et al., 1983) and Wisconsin (Cantor and Blair, 1984) used regional production of crops or livestock as the basis of estimating relative amounts of insecticide use by county. Although the risk of MM was increased in both studies in areas where insecticide use was estimated to be high, the committee did not consider those two studies critical in forming their conclusions, because exposure and disease outcome data were based on death certificates only and there was no further specification or validation of exposure or disease status. Furthermore, the exposure estimates used are approximations of county-wide agricultural practices and do not apply to all residents of every county.
Possible sex-related differences in MM risk were examined by Zahm and colleagues (1992) in a case-control study of agricultural exposures in eastern Nebraska. Exposure to insecticides yielded no association in men (OR=0.6, 95% CI=0.2–1.4) on the basis of 11 cases, but women experienced an increased MM risk (OR=2.8, 95% CI=1.1–7.3) on the basis of 21 cases. The increased MM risk among women is difficult to explain but could have been due to chance or to confounding by other exposures.
The Brown and colleagues study described above found a slightly increased risk of MM among persons reporting exposure to and use of insecticides (OR=1.2, 95% CI=0.9–1.8) (Brown et al., 1993).
A number of studies have looked at MM risk among workers potentially exposed to pesticides. Exposures were broadly characterized, and separate analyses are not provided on more relevant exposures of concern, such as exposure to insecticides. Cohort mortality studies (Ritter et al., 1990; Sperati et al., 1999; Viel and Richardson, 1993), incidence studies (Kristensen et al., 1996; Wiklund and Dich, 1994, 1995; Wiklund et al., 1989), and case-control studies (Brownson and Reif, 1988; Costantini et al., 2001; Demers et al., 1993; Eriksson and Karlsson, 1992; Fritschi and Siemiatycki, 1996; Pearce et al., 1985) of MM and exposure to pesticides have shown mixed results. Although some studies show an increased MM risk with pesticide use, they do not evaluate the specific agents of interest needed to draw conclusions for the purposes of this review.
Several studies that have examined the association between MM risk and exposure to specific insecticides, classes of insecticides, and insecticides in general show small positive associations. However, this body of evidence is limited by the lack of specific or valid exposure determinations, small numbers, and the likelihood that the small associations are due to bias and chance. It is interesting that a number of studies show slight increases in risk; future research to explore the relationship between exposure to insecticides and MM needs additional statistical power—larger numbers of subjects and better measures of