Personalized Depression Treatment
For many people gripped by depression, traditional therapy and medication isn't effective. The individual approach to treatment could be the solution.
Cue is an intervention platform that converts sensor data collected from smartphones into personalised micro-interventions to improve mental health. We analysed the best-fit personalized ML models for each subject using Shapley values to identify their predictors of feature and reveal distinct features that deterministically change mood over time.
Predictors of Mood
Depression is one of the leading causes of mental illness.1 However, only about half of those who have the condition receive treatment1. To improve outcomes, healthcare professionals must be able to recognize and treat patients who are the most likely to benefit from certain treatments.
The ability to tailor depression treatments is one way to do this. Using sensors for mobile phones and an artificial intelligence voice assistant, and other digital tools, researchers at the University of Illinois Chicago (UIC) are working on new ways to determine which patients will benefit from the treatments they receive. With two grants totaling more than $10 million, they will make use of these techniques to determine the biological and behavioral factors that determine the response to antidepressant medication and psychotherapy.
The majority of research to date has focused on sociodemographic and clinical characteristics. These include demographics such as gender, age and education and clinical characteristics such as symptom severity and comorbidities, as well as biological markers.
Few studies have used longitudinal data to predict mood in individuals. They have not taken into account the fact that mood can vary significantly between individuals. Therefore, it is essential to create methods that allow the recognition of the individual differences in mood predictors and the effects of treatment.
The team's new approach uses daily, in-person evaluations of mood and lifestyle variables using a smartphone app called AWARE, a cognitive evaluation with the BiAffect app and electroencephalography -- an imaging technique that monitors brain activity. The team can then develop algorithms to identify patterns of behavior and emotions that are unique to each person.
In addition to these modalities the team created a machine learning algorithm to model the dynamic factors that determine a person's depressed mood. The algorithm combines these personal characteristics into a distinctive "digital phenotype" for each participant.
The digital phenotype was associated with CAT DI scores, a psychometrically validated symptom severity scale. However, the correlation was weak (Pearson's r = 0.08, adjusted BH-adjusted P-value of 3.55 x 10-03) and varied widely among individuals.
Predictors of symptoms
Depression is a leading cause of disability in the world1, however, it is often misdiagnosed and untreated2. In addition, a lack of effective treatments and stigmatization associated with depressive disorders stop many individuals from seeking help.
To allow for individualized treatment in order to provide a more personalized treatment, identifying patterns that can predict symptoms is essential. The current methods for predicting symptoms rely heavily on clinical interviews, which aren't reliable and only reveal a few characteristics that are associated with depression.
Using machine learning to integrate continuous digital behavioral phenotypes that are captured by smartphone sensors and a validated online mental health tracker (the Computerized Adaptive Testing Depression Inventory, CAT-DI) along with other indicators of symptom severity has the potential to improve the accuracy of diagnosis and the effectiveness of treatment for depression. Digital phenotypes permit continuous, high-resolution measurements and capture a wide variety of distinctive behaviors and activity patterns that are difficult to document with interviews.
The study involved University of California Los Angeles students with mild to severe depression symptoms who were enrolled in the Screening and Treatment for Anxiety and Depression program29 developed as part of the UCLA Depression Grand Challenge. Participants were referred to online support or to clinical treatment based on the severity of their depression. Participants with a CAT-DI score of 35 or 65 were assigned online support via the help of a peer coach. those with a score of 75 patients were referred to in-person clinical care for psychotherapy.
At the beginning, participants answered an array of questions regarding their personal demographics and psychosocial characteristics. These included age, sex and education, as well as work and financial status; if they were divorced, partnered or single; the frequency of suicidal ideation, intent, or attempts; and the frequency at that they consumed alcohol. The CAT-DI was used to rate the severity of depression-related symptoms on a scale of 0-100. depression treatment strategies -DI tests were conducted every other week for the participants that received online support, and weekly for those receiving in-person care.
Predictors of the Reaction to Treatment
Research is focusing on personalization of depression treatment. Many studies are focused on finding predictors that can help clinicians identify the most effective drugs for each person. Particularly, pharmacogenetics can identify genetic variants that influence how the body metabolizes antidepressants. This allows doctors to select drugs that are likely to work best for each patient, reducing the time and effort required in trials and errors, while avoid any adverse effects that could otherwise slow progress.
Another option is to create prediction models that combine the clinical data with neural imaging data. These models can be used to determine the best combination of variables that is predictors of a specific outcome, like whether or not a drug will improve symptoms and mood. These models can be used to predict the response of a patient to treatment, allowing doctors to maximize the effectiveness.
A new generation uses machine learning methods such as the supervised and classification algorithms such as regularized logistic regression, and tree-based methods to integrate the effects of several variables to improve the accuracy of predictive. These models have been shown to be useful in predicting the outcome of treatment like the response to antidepressants. These approaches are gaining popularity in psychiatry, and it is likely that they will become the standard for future clinical practice.
The study of depression's underlying mechanisms continues, in addition to predictive models based on ML. Recent research suggests that depression is linked to dysfunctions in specific neural networks. This theory suggests that an individualized treatment for depression will be based on targeted therapies that restore normal function to these circuits.
Internet-based interventions are an effective method to accomplish this. They can offer a more tailored and individualized experience for patients. One study found that an internet-based program improved symptoms and improved quality life for MDD patients. Furthermore, a randomized controlled study of a personalised treatment for depression demonstrated sustained improvement and reduced side effects in a significant number of participants.
Predictors of adverse effects
In the treatment of depression, one of the most difficult aspects is predicting and determining which antidepressant medication will have very little or no side negative effects. Many patients are prescribed various medications before finding a medication that is both effective and well-tolerated. Pharmacogenetics provides an exciting new way to take an effective and precise approach to selecting antidepressant treatments.
There are many predictors that can be used to determine the antidepressant to be prescribed, such as gene variations, phenotypes of the patient such as ethnicity or gender and the presence of comorbidities. However, identifying the most reliable and valid factors that can predict the effectiveness of a particular treatment will probably require randomized controlled trials of significantly larger numbers of participants than those typically enrolled in clinical trials. This is because it may be more difficult to determine the effects of moderators or interactions in trials that only include one episode per participant instead of multiple episodes over a period of time.
Furthermore, the estimation of a patient's response to a specific medication will also likely need to incorporate information regarding symptoms and comorbidities and the patient's prior subjective experience with tolerability and efficacy. Currently, only a few easily measurable sociodemographic variables as well as clinical variables appear to be consistently associated with response to MDD. These include gender, age, race/ethnicity as well as BMI, SES and the presence of alexithymia.
Many challenges remain in the application of pharmacogenetics for depression treatment. First it is necessary to have a clear understanding of the underlying genetic mechanisms is required as well as a clear definition of what is a reliable indicator of treatment response. Ethics like privacy, and the responsible use of genetic information should also be considered. Pharmacogenetics can, in the long run reduce stigma associated with mental health treatment and improve the quality of treatment. But, like any other psychiatric treatment, careful consideration and implementation is essential. In the moment, it's recommended to provide patients with various depression medications that work and encourage them to talk openly with their doctor.
