A Singapore research team has developed an innovative method for identifying signs of depression that utilises data obtained from wearable devices.
Researchers from the Nanyang Technological University, Singapore (NTU Singapore) have designed a predictive computer programme that employs measurements from wearable devices to determine signs of depression, expertly predicting if individuals are at high or low risk. In a trial comprising data of depressed and healthy participants, the novel programme achieved 80% accuracy in identifying those at risk of depression.
The World Health Organization (WHO) estimates that around 260 million people are affected worldwide by depression. Half of these cases are believed to be undiagnosed and untreated. The COVID-19 pandemic has significantly exacerbated mental health issues, which is reinforced by a study from Singapore’s Institute of Mental Health that identified a rise in depression was linked to the pandemic.
Employing wearable devices
Wearable devices, such as activity trackers, are used by nearly a billion people globally, rising from 722 million in 2019. The NTU researchers saw an opportunity to use these devices to monitor mental health, developing a machine learning-powered computer programme called the Ycogni model to screen for signs of depression. Ycogni achieves this by analysing an individual’s physical activity, sleep patterns, and circadian rhythms from data obtained by wearable devices that measure heart rate, step count, energy expenditure, and sleep data.
The researchers developed the Ycogni model in a study involving 290 working adults in Singapore; the participants wore Fitbit Charge 2 devices for two weeks and completed two health surveys that screened for signs of depression at the start and end of the study.
The average age of the participants was 33 years, and the sample closely mirrored the ethnic population of Singapore. The participants wore the trackers at all times, only removing them to shower or charge the device.
Professor Josip Car, Director at the Centre for Population Health Sciences at NTU’s Lee Kong Chian School of Medicine (LKCMedicine), who co-led the study, said: “Our study successfully showed that we could harness sensor data from wearables to aid in detecting the risk of developing depression in individuals. By tapping on our machine learning programme, as well as the increasing popularity of wearable devices, it could one day be used for timely and unobtrusive depression screening.”
Associate Professor Georgios Christopoulos, from NTU’s Nanyang Business School, who co-led the study, commented: “This is a study that, we hope, can set up the basis for using wearable technology to help individuals, researchers, mental health practitioners and policymakers to improve mental well-being. But on a more generic and futuristic application, we believe that such signals could be integrated with Smart Buildings or even Smart Cities initiatives: imagine a hospital or a military unit that could use these signals to identify people-at-risk.”
Pinpointing signs of depression
In addition to being able to determine if an individual was at a higher risk of becoming depressed, the team were able to associate specific patterns of the participant’s behaviours to signs of depression, including feelings of helplessness and hopelessness, change in appetite or weight, and loss of interest in daily activities.
Subsequent analysis revealed that people with more varied heart rates between 2 am and 4 am, and between 4 am and 6 am were more prone to severe depressive symptoms. This confirms the findings of earlier studies that changes in heart rate while sleeping may be a valid physiological marker of depression.
Furthermore, the study also attributed irregular sleeping patterns, such as varying waking times and bedtimes, to a higher tendency of depressive symptoms. The researchers stated that despite weekday rhythms being predominantly determined by work routine, the ability to follow this routine differentiates between depressed and healthy individuals, with healthy people displaying a greater regularity in the time they went to sleep and woke up.
Professor Car said: “We look forward to expanding on our research to include other vital signs in the detection of depression risk, such as skin temperature. Fine-tuning our programme could help in facilitating early, unobtrusive, continuous, and cost-effective detection of depression in the general population.”
Associate professor Christopoulos commented: “Our team will also be working on expanding to other types of psychological status, such as mental fatigue, which seems to be an alarming problem nowadays. Wearables can also be part of a feedback system that could support therapists to better evaluate the psychological status of their patients – for instance, improvements in sleep quality.”
The research team is now looking to investigate the impacts of smartphone usage on signs of depression and the risk of developing depression by enhancing their model with smartphone data, including duration and frequency of use and reliance on social media.
