Research sheds light on eating disorders, compulsive gambling and drug abuse

Apr 23 2006

Researchers at Harvard Medical School (HMS) report in the April 23 issue of Nature that they have identified neurons that encode the values that subjects assign to different items. The activity of these neurons might facilitate the process of decision-making that occurs when someone chooses between different goods.

"We have long known that different neurons in various parts of the brain respond to separate attributes, such as quantity, color, and taste. But when we make a choice, for example: between different foods, we combine all these attributes--we assign a value to each available item," says Camillo Padoa-Schioppa, PhD, HMS research fellow in neurobiology and lead author of the paper. "The neurons we have identified encode the value individuals assign to the available items when they make choices based on subjective preferences, a behavior called 'economic choice.'"

Everyday examples of economic choice include choosing between working and earning more or enjoying more leisure time, or choosing to invest in bonds or in stocks. Such choices have long been studied by economists and psychologists. In particular, research in behavioral economics shows that in numerous circumstances, peoples' choices violate the criteria of economic rationality. This motivates a currently growing interest for the neural bases of economic choice--an emerging field called "neuroeconomics." In general, it is believed that economic choice involves assigning values to available options. However, the underlying brain mechanisms are not well understood.

In the study, Padoa-Schioppa and John Assad, PhD, HMS associate professor of neurobiology, found a population of neurons located in the orbitofrontal cortex (OFC) that assigns values to different goods on a common value scale. Assigning values on a common scale allows comparing goods, like apples and oranges, that otherwise lack a natural basis for comparison.

Previous work in other laboratories showed that lesions to the OFC can result in choice deficits such as eating disorders, compulsive gambling, and abnormal social behavior. Moreover, the OFC is part of brain circuitry implicated in drug abuse, which can also be thought of as a choice deficit. Padoa-Schioppa and Assad's results establish a more direct link between the activity of OFC and the mental valuation process underlying choice behavior.

"The activity of these neurons reflects the value subjects assign to the available goods when they make choices," says Padoa-Schioppa.

"A concrete possibility is that various choice deficits may result from an impaired or dysfunctional activity of this population, though this hypothesis remains to be tested."

In his experiment, Padoa-Schioppa had macaque monkeys choose between two types of juice offered in different amounts. In some trials, the monkey chose between one drop of grape juice (which monkeys prefer) and one drop of apple juice. In other trials, the monkey chose between one drop of grape juice and two drops of apple juice, etc. Behaviorally, Padoa-Schioppa observed a trade-off between juice type and juice quantity. The monkey might choose grape juice when one or two drops of apple juice were available. However, the monkey might be indifferent between the two juices when offered one drop of grape juice versus three drops of apple juice, and might always choose the apple juice if four or more drops were available. This indicates that the value the monkey assigns to one drop of grape juice is roughly equal to the value the monkey assigns to three drops of apple juice.

On the basis of such choice pattern, Padoa-Schioppa could correlate the activity of neurons in the OFC directly with the value assigned to the two juices. For example, the activity of one particular neuron could be low when the monkey chose one drop of grape juice or when it chose three drops of apple juice (low value). The activity of the same neuron might be medium-high when the monkey chose two drops of grape juice or when it chose six drops of apple juice (medium-high value). And the activity of the same neuron might be high when the monkey chose three drops of grape juice or when it chose 10 drops of apple juice (high value). This means that the activity of such OFC neurons encodes the value chosen by the monkey independently of the physical characteristics of the juice--its taste, quantity, etc.

Padoa-Schioppa also found that other neurons in the OFC encode the value of only one of the two juices offered to the monkey. Some neurons encoded the value of the grape juice, while other neurons encoded the value of the apple juice. "The monkey's choice may be based on the activity of these neurons," says Padoa-Schioppa.

An important aspect of the results is that neurons in the OFC encode the economic value of offered and chosen goods independently of the particular way the goods are offered to the monkey, and independently of the specific action the animal uses to signal its choice. For example, OFC neurons encoding the value of apples and oranges would do so regardless of whether the apple is offered on the right and the orange on the left, or vice versa.

"This result has broad implications for possible psychological models of economic choice," says Padoa-Schioppa. "It suggests that economic choice is, at its essence, a choice between goods as opposed to a choice between actions--such as reaching to the right to take the apple or to the left to take the orange."

This work follows a paper published in February in the journal Cognition by Padoa-Schioppa with collaborators Lucia Jandolo and Elisabetta Visalberghi of the Italian National Research Council in Rome. That behavioral study found that monkeys indeed make choices by assigning values to available foods. In the present study, Padoa-Schioppa and Assad identify the neurons in the OFC that correlate such values.