Early psychological research suggested that animals, including humans, performed best when experiencing a certain amount of anxiety—not too much, but not too little. Despite flaws in the theory, it remained a fixture in popular consciousness; today, many people believe that anxiety helps them perform, which can lead it to become ingrained as a habit.
More recent research suggests that treating anxiety using habit-reduction techniques can help people break free from anxiety loops—and may be more effective than medication or CBT.
Anxiety is no stranger in modern times. Nearly 1 in 3 individuals will suffer from an anxiety disorder in their lifetime, and since the COVID-19 pandemic began, the U.S. Census Bureau reported that adults were more than 3x as likely to screen positive for an anxiety disorder in 2020 compared to 2019. And this is just anxiety disorders. Collectively, we share common brain mechanisms that take information from the past and try to project it into the future to predict everything from stock trends to whether we should dine at a restaurant again to our moods (“affective forecasting”). Yet, our brains need precedent and accurate information to do this successfully. I need not remind us how unpredictable this past year has been as we continue to ride waves of uncertainty ranging from COVID-19 (and its variants) to the economy to schools, not knowing whether they will sink our individual boats of mental well-being. We've also been turning to eating, drinking, and distraction to bail out the water with each new swell. On top of this, we started 2020 with a bit of water already in the boat: the idea that we need some level of anxiety as an adrenaline boost to get things done and perform well in life. Back in 1908, when the field of psychology was in its infancy, two animal behavior researchers at Harvard, Robert Yerkes and John Dodson, published a paper entitled “The relation of strength of stimulus to rapidity of habit formation.” In this manuscript, they described an interesting observation: that Japanese dancing mice learned a task more efficiently when they received a moderate shock as a negative reinforcer as compared to a mild or severe shock. Does Anxiety Help Us Perform? They concluded that animals needed some level of arousal—but not too much—to learn best. This paper was only referenced ten times over the next half-century, yet in four of the citing articles, these findings were described as a psychological law (now perhaps irrevocably imprinted on the internet as the “Yerkes-Dodson Law” or “Yerkes-Dodson Curve”) In a paper published in 1955, without evidence, the German-born British psychologist Hans Eysenck suggested that the Yerkes-Dodson Law could hold true for anxiety: He speculated that increased arousal might improve a subject’s task performance. Two years later (1957) one of Eysenck’s former graduate students, P. L. Broadhurst, published a paper boldly entitled “Emotionality and the Yerkes-Dodson law.” In it, he reported that holding a rat’s head underwater (i.e. air deprivation) for increasing amounts of time—which he described as a measure of “intensity of imposed motivation”—increased the rat’s swimming speed up to a point where it then slightly dropped off again. Using the terms “motivation,” “arousal,” and “anxiety” interchangeably, he boldly concluded, “it is clear from these results that the Yerkes-Dodson Law may be taken as confirmed.” From dancing mice to drowning rats, the anxiety-performance inverted U-shaped curve was psychologized into existence: a little anxiety is good for performance, a lot of anxiety, not so much.
Is the Yerkes-Dodson Effect Real? Fast forward half a century, a review of the psychological literature pertaining to stress and work performance found that only 4 percent of the papers supported the inverted U-shaped curve, whereas 46 percent found a “negative linear” relationship; any level of stress inhibits performance —the more the stress the worse the performance. Despite these clear differences—and likely enabled by the ease of information spread on the internet—the overgeneralized Yerkes-Dodson Law became folklore, evidenced by seeming exponential growth in the number of citations (fewer than 10 in 1990, fewer than 100 in 2000, and greater than 1,000 a decade later). So when we get stressed or anxious, our brains have a nice rationalization (backed up by a quick internet search) that it’s OK—in fact, we need it to check off our to-do list or perform well in a meeting. With this, anxiety becomes even harder to treat.
In the 1980s, when the first Selective Serotonin Reuptake Inhibitor (SSRI) was developed, all eyes were on safe(r) medications for depression and anxiety than older classes of drugs, especially ones that didn’t have addictive potentials like benzodiazepines (e.g. Valium), that were so ubiquitous that they showed up in prose, poetry, and music: Mother needs something today to calm her down And though she’s not really ill There's a little yellow pill She goes running for the shelter of a mother’s little helper. —The Rolling Stones
How Worrying Becomes a Habitual Behavior Yet overlooked was a somewhat radical idea about anxiety: that just like any behavior that we do over and over, being worried and anxious can become a habit. At the same time that others were singing the Stones or proclaiming the miracle of Prozac, Thomas Borkovec, a researcher at Penn State, was quietly studying anxiety. Anxiety is defined as, “a feeling of worry, nervousness, or unease, typically about an imminent event or something with an uncertain outcome”; Borkovec honed in on the element of worry, and found that similar to any habit that is developed as a response to a negative stressor (termed negative reinforcement), it too could drive cycles of anxiety. Negative reinforcement requires three elements, a trigger, a behavior, and a reward. For example, if someone is stressed (trigger), eats some cake or chocolate (behavior), and feels better (reward), they set down the habit of stress eating.
Borkovec and colleagues suggested that anxiety can trigger worry as a mental behavior. Worry, in turn, distracts from the worse-feeling feeling of anxiety, or gives us a sense of control. This is our prefrontal cortex taking a lack of information (uncertainty) and thinking of all the worst-case scenarios that could happen—if we can’t control what’s happening, we can at least worry about it, which feels like we’re doing something. Are Anxiety Treatments Effective at Breaking Worry Loops? Yet, despite the concept of anxiety/worry habit loops gathering evidence in research labs, it never made the “bench to bedside” leap into therapeutic development. In today’s teched-up medical system, one would reason that with access to the human genome (and epigenetics), as well as neuroimaging and medications aimed at specific neurotransmitters, we’d have excellent targeted treatments for anxiety. Yet, gold-standard medications—SSRIs—yield a number needed to treat of 5.15, which means that if I give five patients in my clinic the proverbial Prozac, the odds are that only one will significantly improve. For those that prefer to try cognitive behavioral therapy (CBT), the current gold standard for anxiety, they might be pleased to find that it yields medium to large effect sizes in general. Yet the odds are no better than chance (approximately 50 percent) that they will respond to treatment.
Our approaches may still a bit stuck in the 70s: As rational beings, we still approach anxiety with the “just stop it” mentality, thinking that we can think our way out of stress and worry. We all wish we could will our way out of feeling anxious or worried. We haven’t been looking beyond the thinking parts of our brains (the prefrontal cortex), which ironically, as the youngest member of the brain club from an evolutionary perspective, are the first bits to go offline when we’re stressed or anxious. How Can We Break the Anxiety Habit? Around 2016, when my lab was studying how best to help people quit unhealthy habits such as smoking or overeating (which are largely driven by negative reinforcement), someone testing one of the programs I had developed mentioned that her big trigger for eating was anxiety. She asked if I could develop a program for anxiety, which also forced me to come to terms with my roughly 1 in 5 hit rate with anxiety treatment in my outpatient clinic. So I looked into it. Discovering Borkovec’s work for myself, I had a lightbulb moment: My lab’s work targeting habit loops for smoking had yielded 5x better quit rates than the gold-standard treatment, and a study at UCSF of our app-based treatment for overeating was underway (and would go on to show a 40 percent reduction in craving-related eating); maybe we could apply these methods to anxiety. We developed an app-based treatment that targeted anxiety habit loops; my lab started testing it. In a study of anxious physicians, we found that they showed a 57 percent drop in their own anxiety after three months of using the program. People with Generalized Anxiety Disorder (GAD) showed a drop of 67 percent in anxiety after two months of training [this paper is currently under review]. And with these data in hand, we could calculate the number needed to treat: 1.6. Fewer than two people needed to use the treatment to see a significant reduction in anxiety. These results made sense—by targeting and even tapping into the much older (and stronger) brain systems of habit formation, one can learn to work with anxiety and worry, without falling back to outdated notions of control. Perhaps now we can just stop our “just stop it” mentalities and learn to lean into uncertainty instead of worrying so much about the future. This article was adapted from Unwinding Anxiety: New Sciences Shows How to Break the Cycles of Worry and Fear to Heal Your Mind.
About the author
Judson Brewer, MD, Ph.D. is an addiction psychiatrist, neuroscientist and New York Times best-selling author (Unwinding Anxiety). He is the Director of Research and Innovation at the Mindfulness Center and associate professor in psychiatry at the School of Medicine at Brown University. He is also the executive medical director of behavioral health at Sharecare Inc.