New Year, New Me: The Neuroscience of Habit Formation

Post by Deborah Joye

What is habit formation?

As 2018 winds to a close, many of us begin looking to 2019 with the intention of making changes in the new year. Whether it’s eating healthy, learning a new skill, or getting better sleep, the underlying goal of most New Year’s Resolutions is to build new, life-changing habits. What can neuroscience teach us about forming long-lasting habits more easily? In essence, habit formation involves learned associations between an event and a behavioral response. Before we develop automatic associations (habits), we begin with purposeful, goal-directed behavior. Action-outcome associations are goal-directed behaviors in which an individual performs some action, and, if the outcome is rewarding, the behavior is reinforced. So, if I eat a treat and my mood improves, I will be more likely to eat that treat again. With training and repetition, this behavior becomes automatic and the association becomes an ‘outcome insensitive’ stimulus-response association. Once this behavior is automatic, I might eat that treat and feel no different, or even feel a little bit sick, but that wouldn’t necessarily stop me from eating it again. This is because the stimulus of seeing the treat now leads to the automatic response of eating it. Once we’ve performed a particular action sequence enough times with a similar response, the brain tries to free up processing space by saving an automatic stimulus-response association that can be triggered with almost no thinking.

What’s happening in the brain?

During the process of learning goal-directed associations, connections between the cortex (responsible for higher-level cognition, like thinking and planning) and basal ganglia (important for selecting a movement for a particular situation) change their activity, reflecting the switch to more automatic associations. A signal arises during the early learning process (before the behavior is automatic) in a basal ganglia region known as the dorsolateral striatum (DSL), which ‘chunks’ the task-related events together so that the brain sees the whole task from beginning to end as one event. Neurons related to the task fire at the beginning or end of the task (or both) while neurons, unrelated to the task, are quiet. Thus, the entire task is represented as a single event within the DSL. During the shift from trial-and-error learning to a more consistent task response, the strength of this ‘chunked’ representation increases. This representation appears to remain stable as long as the routine is performed and at least partially reinforced (rewarded). Other cells in the DSL have different roles in habit formation; some cells don’t respond during the task at all but respond right after, representing ‘outcome feedback’. When the task is first being learned, some DSL neurons respond to correct performance while other cells signal when the task is performed incorrectly. As the response to the task becomes more automatic, the number of cells that appear to participate in error-signaling (signalling for incorrect performance) gets smaller and smaller, and the cells responding to correct performance increase. This lack of error-signaling for well-formed habits could be the reason why our brains are less sensitive to the outcomes of habitual actions and the reason why habits are so difficult to change.

The DSL is not the only brain region that forms a chunked representation of learned tasks. A version of the chunking pattern also develops in a region of the prefrontal cortex known as the infralimbic cortex. In contrast to the DSL pattern, the chunking pattern in the infralimbic cortex develops later in the learning process as the response to the task becomes consistent and outcome-insensitive. The infralimbic pattern is also different from the DSL pattern because it’s sensitive to changes in the task that require changes in behavior, such as changing which action is needed to receive a reward. The infralimbic pattern decays rapidly when the rules of the task change and it re-emerges when an alternative routine takes shape. Overall, different regions of the brain may function in parallel to promote habit formation – an infralimbic response (cortical-associative-limbic circuit), and a DSL response (basal ganglia).

What’s new?

It was previously thought that the brain circuitry underlying goal-directed actions and habitual actions were competing with one another for dominance in the brain. The idea was that all actions begin as goal-directed, and then the habitual action system takes over and inhibits the goal-directed connections, freeing up brain processing for other things. However, more recent evidence suggests that the two systems can actually work together. For example, goal-directed action circuitry may be needed to initiate a given routine, but habitual automaticity can result in completion of a complex set of behaviors that the brain has learned to see as one unit (see ‘chunking’ in the DSL and infralimbic cortex described above). A goal-directed action such as entering the bathroom to get ready for bed can result in a habitual sequence of actions, resulting in using the bathroom, brushing your teeth, and washing your face. The complementary actions of the goal-oriented and habit systems may be why we begin driving to work, when we are intending to drive somewhere else, or why we intend to drink black coffee and end up pouring in cream and sugar– the goal-oriented action is overtaken by a habitual routine. Research has also shown that context cues (for example, cues from our environment) play an important role in habit formation. When people are trained on a sequential task (performing step 1, then step 2, then step 3), repeated practice resulted in fast reporting of the next step when primed with the prior step. When people are particularly fast at reporting the next step (interpreted as strong habits), their habits were likely to persist, even when they intentionally wanted to add, remove, or change one of the steps. The influence of environmental cues on habitual action can also be seen, for example, in individuals who maintain sobriety while in a controlled environment such as a rehab facility, but struggle to remain sober and relapse once immersed in the environment in which they formerly used drugs. 

Why does it matter?

Understanding how the brain represents goal-directed and habitual actions is integral to understanding pathologies such as addiction and compulsive disorders (binge-eating, obsessive-compulsive disorder), and for grappling with the complex emotions surrounding every day habits. Addiction disorders represent a situation wherein the drugs of abuse have hijacked the person’s reward system, forming very powerful habitual associations. Understanding the deep connections that the brain forms between a given stimulus, the perceived reward, and the related environment can lead to better, longer-lasting treatment for addiction as well as decreased societal stigma surrounding addiction. The same can be said for eating disorders such as binge-eating, where food has become associated with significant reward regardless of hunger status, resulting in disordered eating and pathological weight gain. Finally, understanding how the brain forms habitual connections is critical for addressing the emotional component of bad habits. For many of us, there can be a stigma attached to bad habits if they are thought to reflect a personal flaw or failing (laziness, selfishness, etc.). It is useful to know how we can improve our behavior to create better habits. But, perhaps, it is more important  to realize which aspects of habit formation are under our personal control.

What can I do about it?

What can we do to change our habits or to create new ones? Research has shown that people with good self-control are not necessarily exerting a high level of effort to maintain their good habits. Instead, people with strong self-control have weak habits for unhealthy behaviors and strong habits for healthy behaviors. One way to reinforce this is to redesign our environments to actively avoid circumstances in which bad habits arise. Research investigating eating habits in college students show that their studying significantly improves when smartphones are hidden and that eating improves when junk foods are hidden or even just placed out of arm’s reach. Even placing a bowl of fruit in a prominent space in the kitchen can result not only in eating more fruit, but also in an identity shift (“I am a healthy eater”), which can help cement new habitual action. This corresponds nicely with the idea of ‘habit discontinuity’, which involves a significant redesign of your environment, such as moving to an entirely new location. The absence of the normal habit cues makes implementing new habits much easier, since cognitive resources are not being spent on resisting the temptation of old cues. Another idea is “temptation bundling,” which involves combining something you like to do habitually with something you don’t like to do (for example, watching a certain TV show that you love only while you’re at the gym). This gives a rewarding value to something that previously was not perceived as rewarding in an effort to bootstrap better habits. Finally, it is crucial to consider the impact of stress on habit formation. Since habit learning is enhanced during and immediately after a stressful situation or environment, awareness of how we cope with stress is critical. It is possible that, by consciously exerting effort to choose good habits during times of stress, we may reap the benefits of more rapidly-developed and long-lasting change in our habitual actions in general. This is something to keep in mind as many of us visit with our extended families during the holiday season.

Smith, K. S., & Graybiel, A. M. (2016). Habit formation. Dialogues in Clinical Neuroscience, 18(1), 33–43. http://doi.org/10.1111/clr.12458_111

Carden, L., & Wood, W. (2018). Habit Formation and Change. Current Opinion in Behavioral Sciences, 20, 117–122. http://doi.org/https://doi.org/10.1016/j.cobeha.2017.12.009

Robbins, T. W., & Costa, R. M. (2017). Habits. Current Biology, 27(22), R1200–R1206. http://doi.org/10.1016/j.cub.2017.09.060