Testing for cognitive function in animals in a regulatory context

- Tests of cognitive function are often assumed to be more sensitive to chemical exposure than screening tests.
- Published reports using both kinds of tests are reviewed for acute studies in adults and for developmental exposures.
- Of 35 chemicals compared in adults, cognitive function tests detected effects of 3 chemicals that screening tests missed, and yielded lower no-effect levels for 5 more.
- In developmental studies, cognitive function tests detected effects of 2 chemicals that screening tests did not; the tests were equivalent for 3 others; and conflicting evidence was found for 1.
- Despite not being consistently more sensitive, cognitive function tests do provide data to characterize effects that screening tests cannot.

Superior cognitive functions have allowed the human species to proliferate in a world of incredible biological diversity. Threats to these essential capacities cannot be ignored, and a strategy is needed to evaluate the hazard posed by exposure to chemical and other agents. Because people exposed to chemicals often complain about confusion and forgetfulness, it is commonly thought that cognitive functions should be sensitive indicators of adverse consequences of chemical exposure. For these reasons, complex tests of cognitive function have been developed and deployed in experimental animal laboratories for decades. However, the results of these tests are rarely used as points of departure for chemical risk assessments. Due to their high cost in time, animals, and equipment, the efficacy and utility of these tests need to be evaluated in relation to cheaper and faster whole-animal screening methods. This review examines evidence for the assertions that cognitive functions represent uniquely sensitive indicators of chemical exposure, and that animal models of these functions are necessary to detect and quantify the neurotoxicity of chemicals. Studies conducted since the early 1980s to compare these approaches to assess the neurotoxicity of chemicals are reviewed for both adult and perinatal exposures in experimental rodents. Forty-one studies of 35 chemicals were found that directly compared acute effects using complex tests (i.e., tests that require training animals) with acute effects using screening tests (i.e., tests that do not require training animals) in adult rodents. Complex tests detected effects of three substances (bitertanol, iso-amyl nitrite, and Pfiesteria toxin) that had no effect on screening tests; for an additional five chemicals (carbaryl, deltamethrin, methyl mercury, tetraethyl tin, and Isopar-C), complex tests identified effects at lower doses than did screening tests. Fewer comparable cases were found for developmental exposures: screening and complex tests were found to be equivalent for trimethyltin, n-propylthiouracil (PTU), and elemental mercury. Analysis of two studies yielded an inconclusive case for lead. Evidence for the insufficiency of screening tests was found for PCBs and inhaled ethanol, though it is not clear that the measured effects of these chemicals reflected cognitive deficits per se. Whether these benefits are worth the additional time and expense of conducting complex tests is a matter for discussion in the research and risk management communities.