Volatile nitrosamine contamination of laboratory animal diets.

A number of precautions were taken to ensure the accuracy of the analytical procedures and to avoid artifactual formation of N-nitroso compounds. NDPA (5 ng/g), an NDMA analog which is not normally present in the diets, was added to each powdered sample prior to the vacuum distillation step as an internal (recovery) standard. All reported levels of NDMA and NPYR have been adjusted for the recovery of the NDPA (av erage, 80%; range, 68 to 108%). in addition, sodium ascorbate and a-tocopherol, 10 mg/g each, known inhibitors of nitrosa tion (2, 10), were added to each sample just prior to the NDPA. Other procedures which were used to avoid artifacts and to check on the accuracy of the results are described in a recent review of this subject (8). We first analyzed 9 samples of commercially available pel leted laboratory rodent diets and one canine diet from 3 major suppliers in the United States. These feeds were labeled as containing fish meal (amount unspecified), and all contained between 0.9 and 3.7 ppb of NDMA. One of 2 tested samples of pelieted rabbit food contained NDMA (1.3 ppb) and NPYR (3.2 ppb). The other sample contained no detectable volatile nitrosamines (<0.1 ppb). Neither of the latter 2 samples in ciuded fish meal as a labeled ingredient. We also analyzed several samples of dry pet food. Two samples of peileted dog food and 6 samples of cat food, all marked as containing fish or ‘‘fish flavor,†• were purchased from local markets and analyzed for volatile nitrosamines. One sample (dog food) contained 0.1 ppb NDMA. Five other sam pies contained trace amounts of NDMA (<0.1 ppb), and 2 contained no detectable NDMA. Table 1 shows the results of analyses on several samples of the NIH open formula rat and mouse ration obtained from 2 suppliers during the summer and fall of 1978. All samples contained NDMA at levels ranging up to 52 ppb. Five of 7 samples also contained small amounts of NPYR. The possibility that these findings were the result of nitrosation during the analytical procedure was ruled out by confirming the Sample A data with experiments in which the nitrosation inhibitors (ascorbate and a-tocopheroi) were omitted and/or excess dimethylamine was added. If such artifacts were a problem, one would expect higher values for NDMA in the absence of inhibitors and in the presence of excess dimethylamine. in all cases, the …