by Diana Khanna
As of 2021, it is estimated that over 44.9% of the world's population owns and uses a smartphone. Additionally, it was projected that over one billion people would be using wearable devices by 2022, as the technology becomes more accessible and affordable for people around the world. As this trend is expected to continue with more people adopting these devices, researchers are exploring the science of ‘remote sensing’- using sensors placed in daily life settings to gather data- to track mental health conditions.
Digital technology can passively and continuously monitor a wide range of physiological and behavioural data. Physiological data that can be monitored include motion and heart rate, while behavioural data that can be monitored include sociability, sleep and wake cycles, cognition, activity, and movement. By continuously monitoring an individual's physiological and behavioural patterns, digital technology can provide valuable insights into changes in mood and behaviour that may identify signals of depression. This can potentially help identify individuals who may struggle with or be at risk of developing depression and help them access prompt treatment, which can lead to improved treatment outcomes and better prognosis.
A group of researchers conducted a systematic review of 51 peer-reviewed studies that used remote sensing technologies to measure depressive symptoms. The studies were published in both medical and computer science literature. The researchers summarised the findings of these studies and identified any potential biases in the methods used or limitations that might affect the generalisability of the findings. A review of this type contributes to the understanding of the current state of research and identify any gaps or inconsistencies in the literature that needs to be addressed in future research.
Certain objective digital data have been linked with depressed mood. Location-based data from GPS can be used to determine the duration of homestay or amount of time spent at home which been consistently linked with behaviours like social withdrawal observed in depression. Applications that track physical activity, such as time spent engaging in movements and step counts, can provide insights into an individual's level of physical activity. Studies included in this review found that higher levels of physical activity are negatively correlated with depression. Although it is not yet clear whether these behaviours cause depression or are caused by depression.
Sleep patterns are another commonly studied feature using remote sensing technologies. This is typically done using accelerometers, light, and heart rate sensors that are embedded in wearable devices such as smartwatches or fitness trackers. These sensors can provide detailed information on an individual's sleep patterns, such as sleep duration, quality, and disruptions, which were analysed to identify characteristics that may be related to depressive states. Individuals with a low sleep stability, which includes a lack of consistency in the timing, duration, and quality of sleep, more time spent in bed, and late sleeping hours tend to have higher depression scores. The relationship between digital data and depressed mood was mainly found in studies with longer follow-up periods, compared to studies with shorter time intervals between measurements.
Studies used Bluetooth and microphone sensors to measure the social proximity of individuals to one another, in other words, sociability. In small sample studies it was reported that Bluetooth sensing data is a better indicator of an individual's depressive state than simply counting the number of incoming and outgoing messages and calls. However, in larger and diverse samples, low number and duration of calls made and received were also found to be indicative of depressive mood.
Circadian rhythm or the body’s 24-hour sleep-wake cycle plays a key role in regulating sleep, digestion, hormone levels, and other bodily functions. Monitoring the circadian activities includes movement patterns such as hour-based activity levels, consistency of work hours, and timing of the peak activity levels among others. Even though disruptions in circadian rhythms, like variations in physical activity levels throughout the day, have been thought to affect depression, the digital indicators were not significantly associated in most of the studies.
The authors of the review point out that there is a lack of standardisation in studies examining remote sensing to monitor depression. The variability in the studies is due to factors such as the timespans, sample size, diversity of the sample, how digital features are constructed and defined, and how depression is assessed. This can make it difficult to replicate and generalize the findings of these studies. They recommend increasing the sample sizes (larger sample size recommended) and diversity as some wearable devices may work better on lighter skin tones and the male sex. Additionally, the validity and reliability of sensors and the methods used to assess depression may also be called into question if not reported in the study. Similarly, the assessment of depression should be performed using validated questionnaires. Therefore, it is important to note that further research is needed to fully understand the association between the digital indicators and depression.
Photo by Anthony Tran / Unsplash
Commentaires