The effects of prenatal cocaine exposure on subsequent learning in the rat.
This chapter has reviewed the animal studies related to prenatal cocaine administration and subsequent learning. A summary of these studies and their findings is given in table 1. It is important to note that there is a relative scarcity of reports in this area. Given the potential number of infants that might be exposed to cocaine prenatally, it is certain that more work needs to be done. Second, from this table and the preceding review, it is difficult to conclude that prenatal cocaine exposure has wide-ranging effects. There certainly appears to be ample evidence from Spear's lab that prenatal cocaine exposure disrupts early olfactory learning. These findings are in need of independent replication. In fact, the authors have partially replicated a deficit in an odor-aversion task following gestational cocaine administration. The evidence provided by Spear and colleagues (1989a, 1989b), however, indicates that this deficit is relatively transitory and is not found in animals after perhaps 15 days of Furthermore, it does not appear that these cocaine-exposed animals are incapable of learning. In the Goodwin and colleagues (1992) study, where animals were given either 2, 3, or 4 acquisition trials, animals exposed to cocaine prenatally and reared by surrogate untreated mothers did learn the association when given 4 trials. It must be stressed, however, that early deficits that diminish as the animals mature or deficits which can be overcome by repetition can still have long-lasting consequences. Requiring more experience with an association prior to learning that association or having a transitory learning deficit in no way diminishes its potential importance for the organism. On more common tasks of learning, such as passive and shuttle avoidance and maze learning, there is really very little evidence to support the notion that cocaine is acting as a behavioral teratogen. The available evidence would appear to indicate that significant, biologically meaningful deficits on these tasks are not found over a wide range of doses. There are a number of reasons for these failures to find effects. It may be that these tasks are too simple to detect underlying behavioral anomalies. Passive avoidance is a simple task that is easily learned by animals by the time of weaning. Active avoidance is much more difficult, but it too may not place enough challenges on the organism. The work by Heyser and colleagues (1992b) may signal that more complicated tasks, such as reversal of a conditional discrimination, might be necessary to show cocaine's behavioral teratogenic action. However, even in this case the effect was not very substantial, consisting of less than one extra incorrect response relative to controls prior to the first reward. More challenging situations such as successive discriminations (e.g., learning to learn) might lead to bigger differences in performance between cocaine-exposed animals and controls. These types of tasks need to be assessed. Similarly, perhaps cocaine-exposed animals need to be challenged physiologically, either by placing them in extremely stressful situations or assessing their response to other drugs that disrupt normal physiological functioning (see Spear, this volume). Another reason for a failure to find substantial effects on a wide range of behavioral assessments is that cocaine may act as a behavioral teratogen through a number of different mechanisms. It might be a direct toxin to certain developing neurotransmitter systems, it may function indirectly by inducing hypoxia, or both. If the mechanism of action varies in different animals and the effects are not large in any animal, then it would be extremely difficult to detect group differences using the sample sizes normally assessed in these studies. It may be that only a small subset of a group of animals is affected and that group variability must be assessed in addition to alterations in group means. Finally, it may be that cocaine is not a behavi