On the association between canine malignant lymphoma and opportunity for exposure to 2,4-dichlorophenoxyacetic acid.

In response to criticisms raised regarding a case-control study of canine malignant lymphoma, the results of several ancillary analyses are reported. The case-control study demonstrated a significant association between risk for canine malignant lymphoma and the opportunity for exposure to 2,4-dichlorophenoxyacetic acid herbicides. It is demonstrated that risk estimates do not vary by type of control group (i.e., tumor control or nontumor control group), by method of response (i.e., self-administered or telephone interview), or by geographic area. Questions related to the potential for referral bias, supposed inconsistencies in subject responses regarding frequency of herbicide use, and ambiguities regarding exposure classification are also examined.