For people with major depressive disorder, low moods are just one aspect of the disease. People can also feel their thinking skills such as concentration, attention and memory are also affected. Previous studies indicate that nearly half of people in primary care for depression continued to have thinking difficulties even when their depression improved. This can have lasting effects on their day-to-day life, employment, and chances of long-lasting recovery.
Despite this, few psychiatrists have reported regularly assessing thinking skills to help treat or monitor their patients. Asking patients to self-report their symptoms can be unreliable, due to inaccurate recall. Patients may be monitored in clinics or laboratories, but these are not real-life settings and so can also skew results.
“Thinking difficulties seem to occur before a depressive episode, but experiencing cognitive difficulties over a period of time also seems to predict an episode,” said Professor Til Wykes, a clinical psychologist at King’s College London in the UK. “It is very likely that the cognitive problems will also get worse before an episode,” she said.
Using wearable technologies and smartphone apps may overcome these problems, as two recent studies funded by the Innovative Medicines Initiative (IMI) project RADAR-CNS show. The project used wearable devices and smartphone technology to find new ways of monitoring major depressive disorders, epilepsy, and multiple sclerosis.
App tasks
In one study published in the journal Psychological Medicine, researchers used an app called THINC-it® to monitor over 500 participants diagnosed with depression in the UK, the Netherlands, and Spain. The study aimed to test if warning signs of a depression episode could be found, which in future could help prevent them having major effects on a person’s life.
“The idea is that instead of diagnosing and then treating someone, which is what most people do, we could have a predict and prevent programme that could be more helpful,” said Prof. Wykes.
The app allowed participants to self-report any difficulties in their thinking skills every six weeks. They rated difficulties with organisation, concentration, and forgetfulness on a scale from 0 (never) to 4 (often).
The app also measured participants’ performance through in-app tasks, also every six weeks. These tasks assessed their attention, working memory, processing speed, and attention switching. Both measurements were taken for up to two years and then used to calculate how long difficulties in performance lasted.
Prof. Wykes and co-authors found that those who reported persistent thinking difficulties (more than 75% of the time surveyed) were also reporting higher levels of depression compared to those who had less persistent thinking difficulties. Interestingly, some thinking difficulties correlated with certain effects of depressive episodes. For example, those who had difficulties with their processing speed were found to have worse symptoms of depression.
The study confirms that using smartphones to self-report and monitor thinking difficulties could help monitor and manage depression.
Tech acceptance
The second study, published in the Journal of Affective Disorders, assessed which factors would most likely lead to adoption of mHealth technology to manage depression.
Researchers asked 171 people with a history of depression to choose their preferred technology from a series with four factors: privacy, clinical support, established benefit and the device’s ability to detect their symptoms.
The researchers used a method known as ‘mixed logit models’ to find what was most likely to affect adoption. Instead of simply asking people what they preferred, the study’s roughly 170 participants were asked to choose between technologies that were strong in some of the four factors, though weak in others. Through several comparisons, certain factors would continually win out over others, indicating that these were the preferred factors. This method lets researchers account for random differences of taste among the participants.
The researchers found that overall, participants valued ‘accuracy’ as the most important factor, with only modest compromises willing to be made to increase other factors such as privacy. ‘Established benefit’ was the least valued attribute, meaning participants were happy with technology that had possible but uncertain benefits.
However, the paper mentions that different groups within the study had different priorities. For example, those who were younger, less accepting of technology, or were of a specific nationality were more willing to compromise on some accuracy in exchange for more privacy and clinical support.
Nonetheless, together these papers show that these self-reported assessment apps can provide accurate real-world data that are also accepted by the people they are supposed to monitor.
Prof. Wykes is now looking to do further research to see how using mobile apps can be best used to predict the onset of depressive episodes, though she notes that any algorithms behind these apps will need to be both highly controlled and involve input from the people they are supposed to help.
“I don't think you can provide a prediction algorithm without having clinicians and people with depression involved in the decision making, even if only in a supportive way,” she said. “I think the way forward in therapy is to make sure the ‘end users’ are involved at every single stage, including in choosing the supervision of the algorithms.”
RADAR-CNS is supported by the Innovative Medicines Initiative, a partnership between the European Union and the European pharmaceutical industry.