In this interview, we speak to Dr. Gareth Hollands about his latest research that investigated the use of medical imaging to reduce health risks and encourage behavior change.
Please can you introduce yourself, tell us about your background in health risk and behavioral change, and what inspired your latest research?
My name is Gareth Hollands and I have recently started working at University College London, although this research was completed at the University of Cambridge where I worked for the last ten years. I did my PhD from 2006-2010 in health psychology with a focus on the way in which visual images may or may not have the potential to change people’s health-related behavior.
One of my main PhD projects back then was to conduct a Cochrane review on whether feeding back medical imaging scans to individuals can motivate changes in their health-related behaviors. This new paper is in fact an update of that old Cochrane review I did for my PhD, but there’s quite a bit more new evidence since then.
Non-communicable diseases, often linked to behaviors such as smoking, lack of exercise, and a poor diet, are estimated to account for two-thirds of worldwide deaths annually. Why is it therefore important to motivate behavioral change in individuals who regularly partake in risk-related behaviors?
As you say, these kinds of behaviors are hugely important in determining ill health. Most people are aware to some degree that these things aren’t good for them, and we might think this would mean people are highly motivated to make changes, and can easily make them. However, despite people often wishing to behave more healthily, there is a big gap in how we actually behave because our day-to-day ingrained habits, and the often overwhelmingly unhealthy environments that we are constantly exposed to, make this difficult.
As such, health researchers are constantly looking to identify opportunities for interventions to help make changing one’s behavior easier for people, whether at the level of the individual being presented with motivating information – as is the case here - or larger-scale population-level interventions that may involve changing social, economic or physical environments.
Image Credit: Halfpoint/Shutterstock
Why do you believe that the growing use of medical imaging technologies could help motivate behavior change in individuals who regularly partake in risky behavior, and what type of medical imaging technologies could you use to do this?
It is an opportunity to intervene at a time when people are receiving very salient and important information about their health from a clinician, typically when their health is being actively investigated and they are being asked directly to consider following some suggestions about how their health-related behaviors. The review we conducted includes examples from a wide range of imaging technologies and clinical conditions.
The contexts where there were most studies were cardiovascular risk imaging, specifically of arterial, venous, or cardiac health, using either ultrasound or CT scans; UV photography to image UV exposure-related skin damage; and forms of oral photography to image gingivitis, dental plaque, or tooth decay within dental consultations.
In your new meta-analysis, you found that an individual may be more likely to reduce risky behaviors after undergoing an imaging procedure showing visual personalized information about the individual’s own risk of disease. How did you conduct this meta-analysis and what behaviors did the trial report on?
We conducted the systematic review and meta-analysis using standard Cochrane review methods (a standard rigorous methodological approach to reviewing the effects of healthcare interventions), including wide-ranging searches of the existing literature and a systematic process of assessing and statistically analyzing that evidence.
The hope is that in doing so it can provide a clearer picture of the current evidence than just looking at individual research studies separately. The analysis included ten health-related behaviors that can feasibly reduce risks to health, but there was the most evidence available in relation to smoking, medication use, and physical activity.
How can showing individuals their scan results, highlighting the current state of their health, motivate people to change their behavior and live in a healthier way?
The widely understood power of visual images in conveying information is highlighted in two common idioms: “Seeing is believing,” and “A picture is worth a thousand words.” More technically, there is a range of psychological theories that could help us understand this process.
It is thought that concrete images can more readily be comprehended and better engage emotional associations in memory relative to more abstract, conceptual information such as verbal description. Therefore, presenting visual images in this way allows people to actually see evidence of their health status and potentially the direct results of their behavior on their body, whereas just being told about something may not have the same impact because its implications are harder to understand.
The specific mechanisms and processes in relation to medical imaging feedback have not been much studied, however, and there are likely a complex set of factors that may influence this.
During the trials, which risk-related behaviors had the strongest evidence of improvement via visual intervention analysis, and why do you believe this was the case?
The strongest evidence was for smoking reduction. This could be something to do with smoking having strong and well-known links with health outcomes, meaning that it was central to several of these interventions and that people may more readily accept that smoking was contributing to their health status that was under investigation. But it is also likely linked to, and may simply be a case of, there being the largest amount of evidence for this outcome.
Image Credit: Marukosu/Shutterstock
The meta-analysis identified strong evidence for behavioral change in some areas, for example with smoking reduction, but also found that with other behaviors, for example, sun protection and tanning booth use, behaviors were improved by visual interventions, but the results were not statistically significant. Why do you think that individuals seem more motivated to modify their behaviors in some respects but not others?
I do not think we are yet able to confidently conclude that these interventions are likely to be more or less effective for some behaviors than others, because the evidence was often lacking in quality and quantity.
Of course, this could be the case, and we could speculate that some behaviors are likely linked to certain types of feedback, or that some behaviors are more easily achievable and so forth, but it is really too early to say in terms of this review.
Beyond the trials, do you believe that risk-related behavior changes are sustainable for individuals in the long term?
It’s certainly possible. Firstly, it is worth noting that the analyses typically covered extended follow-up periods, with the modal average time point for outcome assessment being 12 months for a number of the outcomes in the review, including smoking, physical activity, and a healthier diet. This highlights the potential for effects elicited by these interventions to be sustained in the medium to long term.
Beyond the data from studies included in this review, there is complementary evidence that risk-reducing behavioral changes such as changes to smoking, physical activity, or diet—whether occurring within intervention contexts or when examined longitudinally in free-living conditions—can be well maintained over time. For example, it has been estimated that for smokers who remain abstinent at 12 months, the annual incidence of relapse after this is only around 10%.
However, of course, this isn’t always the case and the ability to sustain changes in behaviors over the long-term differs as do the environments to which people are exposed to. Many people have very challenging lives, a lack of support and resources, and that combined with the unhealthy environments that surround us, mean making and sustaining changes is often very difficult for people, especially those who experience greater social and material deprivation and poorer health outcomes already.
Image Credit: Moving Moment/Shutterstock
What are the next steps for you and your research?
To continue to produce rigorous evidence on scalable interventions with the potential to inform policy and improve population health, both systematic reviews like this but also primary research in developing and evaluating interventions to change health-related behavior. I would also like to gain wider experience in other topic areas beyond human health, applying some of the same methods to related and similarly important areas such as built environment research and planetary health.
Where can readers find more information?
The review is available on the PLOS Medicine website at: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003920
About Dr. Gareth Hollands
I have recently started working at University College London, although this research was completed at the University of Cambridge where I was a
researcher for the last ten years. Much of my research to date has been on three main research areas. First, interventions that change environmental cues to change behavior, also known as ‘choice architecture’ interventions or ‘nudging’; Second, systematic reviews of interventions to change behavior, and advances in evidence synthesis methods; and Third, the effects of communicating highly salient risk information, such as aversive visual images and genetic risks on people’s behavior.
I have been fortunate to have lots of career highlights including memorable projects and papers that I hope have had and will have a positive impact on research or policy, but the best aspect of my career has been the privilege of working and collaborating with so many thoughtful, inspiring and supportive colleagues on an ongoing basis, both nationally and internationally.
How AI is advancing mammographic density-based breast cancer risk prediction