Worry is one of the most common ways that anxiety shows up in life. It fills our minds with nerve-wracking thoughts about the future. What if I forget what to say during my speech? How will I make friends in the new city I moved to? What will happen to my family if I receive harrowing news from my next doctor’s visit?
For many people, these kind of worrisome thoughts are uncomfortable, but they’re mostly fleeting and little more than a nuisance. For others, they can become all-consuming and debilitatingly painful. Chronic, intense worry of this kind is one of the key symptoms of generalised anxiety disorder (GAD) – among the most frequently diagnosed mental health problems. Although there are established treatments – such as mindfulness meditation and cognitive behavioural therapy – for GAD and other conditions involving out-of-control worry, they are far from perfect. They don’t work for everyone and sometimes their benefits fade over time.
While some researchers seek to better understand worry by studying neurochemistry, and others look to answers in people’s upbringing or past traumas, I believe it’s vital to study the inner workings of the mind – specifically those basic mental processes that cause worries to arise too easily and persist for too long.
Much of this comes down to how we plan for the future. When you think of planning, such as planning a meal or planning a holiday, it might seem like a benign mental activity that has little to do with pathological worry. However, planning can go wrong in various ways and scientists and clinicians have long suggested that worrying and planning are intimately linked.
For instance, more than 30 years ago, the British clinical scientist Andrew Mathews proposed that worry had evolved to help us cope with future danger – to plan and problem-solve how to avert imagined threats from materialising. He documented how these processes had gone awry in his patients with generalised anxiety – they engaged in planning that was clearly unproductive. For example, some of them got stuck repeatedly imagining succumbing to an impending threat.
My work, which I outlined more formally in a recent paper in Trends in Cognitive Sciences, builds on these earlier ideas and suggests that people suffering from chronic worry engage in biased and/or excessive planning. Biased planning involves inaccuracies in the planning process, which can lead us to imagine events in unrealistic ways that are incongruent with how they actually tend to transpire. Excessive planning involves planning at unhelpful times (such as when a threat is far in the future) or beyond the point when it’s useful. If you’ve ever found yourself catastrophising about the future, or tossing and turning in bed about a challenge that’s months way, then you will be able to relate to these descriptions.
The promise of thinking about worry in this way is that understanding someone’s particular planning biases and excesses will point to ways to correct them, which will help them worry less and experience reduced anxiety.
Although Mathews and others continued to put forward their hypothesis that anxiety is essentially a dysfunctional form of planning for dangerous outcomes (known as the ‘planning-anxiety hypothesis’), until recently the scientific tools simply weren’t available to test it. Planning is difficult to study scientifically because planning is a private mental process and may, in many instances, even elude our own consciousness. For example, when people plan how to take a series of actions in abstract planning tasks (such as in a game of chess), brain imaging research shows that their hippocampus ‘pre-plays’ sequences of actions at speeds they couldn’t possibly be conscious of (such as four actions in 150 milliseconds). So, although it’s clear we have some access to actions we evaluate during planning, the richer process of simulating possible action sequences, and deciding between them, seems to hide in the subconscious.
New neuroimaging methods can be used to decode the sequences of planning actions
This has implications for modern psychotherapy too. Although therapy can be beneficial, therapists are limited in their capacity to gauge how planning goes awry. The therapy process relies either on thoughts we’re aware of, or thoughts not in our awareness but that are detectable in our behaviour (for example, I might not be aware I’m angry, but my short and sarcastic remarks betray my true feelings). Although I have no doubt an expert therapist could infer that their client is affected by certain planning biases (for example, they might detect a patient’s exaggerated desire for control), it would be unrealistic to expect them to be able to detect certain planning biases underlying worry – especially those operating at a non-conscious and computational level. For this level of understanding, we need cutting-edge research methods that reveal the dynamics of how the brain simulates action sequences (that is, how it simulates planned future actions, such as how you will respond to questions in an upcoming job interview).
Contemporary research is using two approaches to find out more about the way people simulate the future during mental planning. First is to create planning tasks in a laboratory setting, such as maze tasks or chess problems, for which we know all the ‘rules of the game’ – all possible events, actions, transitions between events, and what situations are most valuable. In such laboratory tasks, the constraints allow scientists to use computational analyses to infer the way people have simulated future actions based on their decision-making patterns. The second approach is to go more directly to the source of planning by analysing brain activity. What’s really exciting is that new neuroimaging methods can now be used to decode the sequences of actions someone is planning. So, for example, if someone is planning how to exit a maze, we can use neural decoding to predict in advance the exact sequence of action the maze navigator will take before they act.
How does all this new science about planning ultimately help us to explain anxiety? The framework I’ve developed links the latest science of planning to the effort to uncover precise planning biases in anxious thought. My framework breaks down planning into two component parts: (1) when to start and stop planning; and (2) how to form a model of future scenarios we’re unsure about.
Let’s begin with when we start and stop planning, which can be used to explain why some people worry too often. To decide when to begin planning, your brain carries out a cost-benefit analysis, called ‘meta-control’, based on the most efficient use of its limited computational resources. Planning is exhausting, so the benefits of engaging planning must outweigh its inherent cost to effortfully simulate many action sequences. Seen within this cost-benefit analysis, you can think of chronic worriers as potentially having a bias that overvalues the benefits of planning, and undervalues its costs. To return to the example of a job interview, a chronic worrier might plan for a job interview while lying in bed months ahead of when the interview is scheduled, even though it would be far more advantageous to delay planning altogether to get to sleep. One of my research goals is to use computational models to measure planning biases like these that corrupt meta-control.
After planning begins, it must stop at some point (or we’ve entered some horrifying Stephen King novel). This is another situation in which your brain must perform a cost-benefit analysis, in this case to evaluate whether your current plan is good enough or whether the chance of creating a better plan is high enough that it justifies the cost of engaging in further planning. For people with chronic worry, they might fail to reach this ‘good enough’ threshold to stop planning. This could be because their simulation of future outcomes is overly catastrophic, so they get stuck searching for actions that can effectively avert the intense threats. Alternatively, they might have a very high threshold for what it means to find a ‘good enough plan’. For instance, a chronic worrier facing a job interview might implicitly believe they need to form a plan that has a 95 per cent chance of success (in this case, getting hired). Such a high threshold for success might be near-impossible to internally simulate. As a consequence, they get stuck in a protracted worry episode. Another goal of my research is to develop new methods to measure these unrealistic thresholds reliably.
A chronic worrier might use their limited resources on simulating unhelpful details of the interview
The capacity to control when we initiate or terminate planning critically depends on what scientists call a ‘model of the world’, which forms the second main part of my framework. Essentially, this is the imagined set of actions and events relevant to your plan – and it’s another aspect of planning that can go wrong in pathological worry. When chronic worriers form a model of the world for a particular scenario, such as a job interview, they may be biased in which actions and events they focus on, and which they ignore.
For example, assuming you don’t have a worry problem, you might plan ahead for which interview questions you might be asked and how to answer them, but you wouldn’t go so far as to simulate how you will open and close the door to the interview room. You realise implicitly that ignoring those details in your model of the job interview won’t harm your ability to form a good plan. In contrast, a chronic worrier might use their limited resources to focus on simulating minute and unhelpful details of the interview. For instance, they might think about how exactly they will move and breathe in the interview to avoid sounding awkward, even though this won’t help them succeed. By spending all their mental resources on these fine details, they forego thinking through other actions more fundamental to their goal of getting hired, such as how to answer difficult questions well. This is another target of my research – to uncover the mental processes that lead to unhelpful mental models of future events.
When the science of planning becomes more precise, I’m anticipating that we’ll be able to identify a person’s specific planning biases, and that will present an opportunity to think of ways to causally alter those biases. One approach could be to adapt the kind of planning tasks that we’re currently using in the laboratory. For example, imagine we discovered that someone initiated planning too often because they undervalued the cost of planning and that this was provoking excessive worry. One possibility is that we could use incentives in a laboratory planning task to train the person to perform a more accurate cost-benefit analysis of when to initiate and when to delay planning. Another approach could be focused on helping improve the mental models people use for simulating future challenges.
These are just a couple of examples of the many possible applications of this line of research. Anxiety and worry are the bane of so many people’s lives. I hope that over the next decade we’ll witness real progress, and the planning-anxiety framework will inspire new and effective treatments for chronic worriers.
