Risk preference is governed by different hypothalamus–habenula circuits, study finds

To survive in complex and unpredictable environments, humans and other animals need to learn to predict the consequences of their actions, so that they can reduce risks and maximize their gains. In humans, this risk evaluation process is known to be highly sophisticated and is applied daily to a wide range of situations beyond life-and-death scenarios.

Past neuroscience and psychology research found that when animals are in situations where they can choose between a safe or risky option, they tend to exhibit consistent behavioral patterns. In other words, they tend to either preferentially pick the riskier (i.e., more uncertain) options or the safer ones.

While this tendency to preferentially choose either risky or safe options is well-documented, the neural underpinnings of these different patterns in decision-making remain poorly understood. Researchers at University of Zurich and ETH Zurich recently set out to investigate how the risk preferences of mice are encoded in their brains. Their findings are published in Nature Neuroscience.

“Confronted with choosing between certain (safe) and uncertain (risky) options, animals show strong preference for either option consistently across extended time periods,” wrote Dominik Groos, Anna Maria Reuss and their colleagues in their paper. “How such risk preference is encoded in the brain remains elusive. A candidate region is the lateral habenula (LHb), which is prominently involved in value-guided behavior.”

The LHb is a small structure known to play a role in the processing of aversive (i.e., unfavorable) outcomes, thus contributing to reward-based learning. Groos, Reuss and their colleagues set out to explore the possibility that neurons in this region represent the risk preferences of mice.

To do this, they used a simple task in which mice could choose between two options, specifically two waterspouts that delivered different amounts of sucrose water when they licked them. One of the waterspouts always delivered 5 µl of sucrose water and was thus a safe option, while the other was riskier, as it delivered 17 µl of sucrose water in 25% of cases and 1 µl 75% of the time.

While the mice engaged in this task, the researchers used a combination of imaging and optogenetic techniques to gather information about what circuits in their brain were activated. The methods they used included two-photon calcium imaging, whole-brain anatomical tracing and optogenetic tools.

“Using a balanced two-alternative choice task and longitudinal two-photon calcium imaging in mice, we identify risk-preference-selective activity in LHb neurons reflecting individual risk preference before action selection,” wrote Groos, Reuss and their colleagues.

“Using whole-brain anatomical tracing, multi-fiber photometry and projection-specific and cell-type-specific optogenetics, we find glutamatergic LHb projections from the medial (MH) but not lateral (LH) hypothalamus providing behavior-relevant synaptic input before action selection. Optogenetic stimulation of MH→LHb axons evoked excitatory and inhibitory postsynaptic responses, whereas LH→LHb projections were excitatory.”

Overall, the findings gathered by the researchers suggest that the risk preference of mice during reward-based decision-making tasks is supported by functionally different hypothalamus–habenula circuits. Further studies could examine the circuits they observed further or try to determine if they also shape the risk preferences of other animal species.

If extended to humans, these results could also contribute to the study of mental health disorders characterized by unhelpful risk-taking behavior, such as substance use disorder, compulsive gambling and bipolar disorder. This could potentially help to identify new therapeutic targets for reducing the risk-taking tendencies of individuals diagnosed with these disorders.

© 2025 Science X Network