Can we train pain-related attention?

Spoiler alert: This blogpost contains null findings. 

Paying too much attention to pain is bad. At least, that’s the idea for people who have chronic benign pain, where the pain no longer provides much useful information about the state of the body. People with chronic pain often report being frequently interrupted by pain, meaning that they’re not able to focus on other important, valued things in their lives. But it’s not just attending to the pain itself that’s bad. Habitually attending to cues for pain – things that might predict a future experience of pain, or that remind us of pain, also increases interruption in daily life, and may increase the salience of pain.

So, here’s the question: can we use training to help people pay less attention to pain cues? And if so, will this reduce their pain and increase their quality of life? There are a handful of trials that have asked this question for adults with chronic pain. These studies have typically used a computerized training program called Attention Bias Modification, or ‘ABM’ for short. In ABM, individuals view things that are pain-related and non-pain-related on a computer screen. The pain-related information might be a word to do with pain (e.g., ‘ouch’ or ‘sharp’), or a photograph of someone in pain. Individuals don’t need to respond to the words or photographs themselves, but instead are required to find a little shape that replaces one of them and to press a button as quickly as possible to say where (or what) that shape is. Speed is key; participants are given warnings if they are too slow. The idea is that individuals with chronic pain will habitually attend to the pain-related information over the non-pain information; they will show an ‘attention bias’. But, by making the little shape always replace the non-pain information, they will learn over time to refocus their attention on the non-pain information (and thus away from the pain-related information) in order to get better and faster at the task. The hope is that the new attention control skills will transfer to real-life pain cues.

Findings for ABM in adult chronic pain are mixed, but there is some preliminary data to suggest that it may help improve pain and pain-related functioning. We wondered: could ABM be useful for teenagers with chronic pain? Adolescence is an interesting time for all things pain- and brain-related. For one, chronic pain often first emerges during adolescence, making it a potentially useful window for early intervention. For another, teenage brains are more ‘plastic’ than adult brains. They are still undergoing structural and functional reorganization, especially in prefrontal areas that are involved in attention control. Given this increased malleability of the teen brain, especially in areas to do with attention, we wondered if adolescents might be more receptive to ABM training.

The short answer is: they’re not. At least, within the context of our randomized, placebo-controlled trial, we saw no significant differences between young people with chronic pain (N=66) receiving ABM, placebo training, or no training on any outcomes. Those outcomes included pain symptoms as well as pain-related physical and psychological functioning. Why didn’t it work? One reason might be that our adolescents didn’t (on average) show an attention bias towards pain in the first place. You could argue that we should have only recruited adolescents if they showed an attention bias at baseline. But there are two issues with this argument: first, when we look only at the adolescents who did have an attention bias at baseline, they didn’t do any better than those who showed no bias. Second, the task we use to measure attention bias (the ‘dot-probe’) isn’t super reliable. In fact, data suggest that the reliability of the pain dot-probe task is about zero. This is probably at least partly because attention is dynamic. But it means that if an individual completes the task on a different day, or even later the same day, they are likely to get a different score. So, preselecting patients based on a single dot-probe score didn’t seem like the most sensible approach. There are other reasons too, like perhaps we didn’t give adolescents enough training to make any real changes, or perhaps the stimuli weren’t relevant or salient enough, or perhaps the training simply isn’t powerful enough to change the way we process something as attention-grabbing as pain. These reasons remain untested for now, and may well apply to adult studies as well, laying the groundwork for research going forward.

Our trial also provided some novel insights which we think could guide future studies. First of all, our null findings didn’t stop us from publishing the data in the top journal in the field. In fact, the reviewers were receptive to the findings, all of whom emphasized the importance of publishing null results. But the reviewers were also cautious that we shouldn’t throw the baby out with the bathwater. Importantly, we were underpowered to detect changes with small effect sizes, and recent research outside of pain suggests that if ABM has any benefit, it is likely to be of small effect. More importantly, this research area is at an early stage. There have been some promising, albeit mixed, findings in the adult chronic pain literature, and our study is the first and only trial with young people with chronic pain. Researchers who study anxiety disorders are finding novel, interesting ways to better train attention, and applying these to chronic pain samples could be especially fruitful.

We also learned a few things about conducting trials of novel interventions for young people with pain. Computerized ABM is unmistakably different from the types of psychological interventions commonly used to treat pediatric chronic pain, which typically involve face-to-face clinical contact. ABM training has low face validity. But young people and their families were receptive to the training, to going on the computer a couple of times a week, and to completing all of this training at home. We had high retention in the study, and adolescents completed the training at a high level. This offers hope for similar interventions that use computer-based training with this population, even for programs that have relatively low face validity and may tackle more implicit pain experiences. These types of computerized training interventions are unlikely to replace conventional psychological and multidisciplinary treatment, but could be useful adjuncts for patients to complete at home in between treatment sessions at the clinic.

About Lauren Heathcote

Lauren is an experimental psychologist with a keen interest in all-things-research. She did her PhD at the University of Oxford (where they like to call a PhD ‘DPhil’), looking at the role of cognitive biases in adolescent non-cancer pain. She is now a postdoc in the anesthesiology department at Stanford University, where she is exploring the role of threat perception in adolescent chronic and post-cancer pain.

In her spare time, Lauren likes to climb Californian mountains, eat sushi, and salsa dance into the wee hours.


Heathcote, L.C., Jacobs, K., Van Ryckeghem, D.M.L, Fisher, E., Eccleston, C., Fox, E., & Lau, J.Y.F. (in press). Attention bias modification training for adolescents with chronic pain: a randomized placebo-controlled trial. Pain.