Psychology and Neuroscience of Affective Disorders Lecture PDF

Module: Psychology and Neuroscience of Affective Disorders Week 1: Introduction to affective disorders Topic 3: Emerging focuses in affective disorders (Part 1 of 4) Lecture transcript Dr Rebecca Strawbridge Department: Centre for Affective Disorders Slide 5 In this session, I firstly hope to inform...

Module: Psychology and Neuroscience of Affective Disorders Week 1: Introduction to affective disorders Topic 3: Emerging focuses in affective disorders (Part 1 of 4) Lecture transcript Dr Rebecca Strawbridge Department: Centre for Affective Disorders Slide 5 In this session, I firstly hope to inform you of major topics that are being researched in 21st century academic examination of affective disorders. Broadly, these aim to do two things which correspond to the second and third learning outcomes. For the second, I'll provide a conceptualization of research avenues attempting to get us closer to a comprehensive, mechanistic understanding of affective disorders, and then how research is moving towards translational implications that will benefit people affected by affective disorders. Slide 6 To do this, I'll first talk about recent calls to action in researching unipolar depression and bipolar disorders. And then how the 21st century is providing us with new tools to progress the field, and then hot topics in advancing towards understanding mechanisms of illness, followed by hot topics in relation to new treatments, and finally to talk through hot topics surrounding other approaches to improving treatment outcomes for people living with affective disorders. © King’s College London 1 Slide 7 The next few slides evidence that there is both the need and an actioning to progress evidence and research related to affective disorders. And for this to specifically reach patients in order to benefit them and reduce the overall burden of these costly conditions. For unipolar depression, there was a call to action in The Lancet in February 2022. This outlined a need for united action between healthcare professionals, academics, patients, and service infrastructures, and indeed the whole of society. Slide 8 Here are the key messages from this call to action. You might wish to take a moment to ingest these points listed here by pausing this video. They include the high prevalence and poor detection of depression in health services, the heterogeneity of depression between different people suffering from it, and relatedly its uniqueness and complexity, the importance of culture and prevention of illness, and improving treatment specifically in early detection of illness, ensuring an understanding of each patient's needs between themselves and clinician, that's formulation. Collaborative care between the different individuals involved. Following stepped care pathways in accordance with evidence based guidelines, and making sure that care is also person centred and long term. Slide 9 We see something quite similar for bipolar disorders. Although the call to action reference here is from two years earlier, similarities exist at least in terms of collaborative care, illness complexity, severity, and uniqueness and the need for improved diagnosis. Additionally, there is a specific key message about lack of quality care available in low and middle income countries. Again, it may be useful to pause the video here to go through these key messages. Those related to burden are slightly different from unipolar depression, focusing more on the high heritability and premature mortality of bipolar disorders, which appear greater in bipolar than unipolar affective disorders, albeit with a lower prevalence of people experiencing bipolar conditions. The other key message focus here is specifically related to real world care, not following best practice guidelines in terms of under using some effective therapies, particularly lithium and overuse of other potentially problematic treatments, notably antidepressants. Between bipolar and unipolar affective disorders, many of the emerging focuses or hot topics are common and will be discussed together throughout this session. Slide 10 But here we see some light at the end of the tunnel. Essentially, here is some evidence that action is being taken with an example of depression overall as focused on by academics in China, who found firstly that publications on the topic of depression have been increasing every year for at least the last decade and secondly, presenting a theme cloud of hot topics as demonstrated by clustering of keywords from these papers. These include, as I will go on to discuss, understanding the pathogenesis and mechanisms of depression, biological factors, and management, including both prevention and acute treatment. 2 © King’s College London Slide 11 In this section on 21st century players, I go into three different recent phenomena that are being used across the spectrum of affective disorders research. So first, of course, technology is being used to progress understanding in this as all domains of life. The opportunities that smartphones, smart watches, and similar devices bring have stimulated research into data collection using various newly accessible measures. Activity monitoring, actigraphy, or actimetry, uses accelerometers to track the amount of physical activity and also assessments of sleep both in quality, amount, and timings. Passive data from smartphones includes time and duration, using various applications and information about calls and texts. Relevant applications from both actimetry and passive phone data have included detection of relapse in people with bipolar. So increases in activity before or during manic relapse alongside reduction in sleep, often increased or certainly erratic phone use or the opposite and depression. Passive monitoring could also use data from microphones related to speech and calls, but this is less research due to concerns around privacy. But speech and text data is being applied, for example, to assist with diagnostic accuracy and discrimination between people with different diagnosis. However, this is mainly from specific assessment data where patients have explicitly consented for use, and this would therefore come under active data. Slide 12 Other types of technologically assisted active monitoring include prominently ecological momentary assessment, or EMA, where participants in studies are prompted several times per day to provide snapshot data often related to moods. And this can be used to provide an accurate description of longitudinal fluctuations in mood, with additional data collected being used to identify, for example, risk factors for mood changes. A variety of smartphone applications are being developed for various types of data collection, even to the extent of tailoring to individual studies. Slide 13 A common feature across these types of technology enhanced data collection is the potential or ability to obtain a large amount of deep data at a high frequency, which can then result in well powered, sensitive, and high quality information gathering and analysis. Slide 14 Machine learning is increasingly being used as we now have the computing power to be able to process a lot of data in complex analyses. Machine learning is a type of artificial intelligence where algorithms are applied to data to make predictions. These algorithms can be quite complex in attempting to delineate relationships between many variables. But you can think of these like a big regression analysis where, for example, variables predicting an outcome can be ascertained from data input. There is a lot of excitement about the potential of machine learning, but we need to remember two key points, especially, when interpreting, for example, headlines saying machine learning can predict X or Y. Firstly, a lot of data is needed to properly train a model, and we often don't have this. Although big data is a topic you will often hear about and attempts to create harmonised large data sets or ramping up. Secondly, overfitting can occur. In this context overfitting refers to training the model tightly, 3 © King’s College London so it fits the data being input really well, but in such a way that it will never translate to work on a new set of data entered in the future. It's obviously critical that any exciting looking finding can be reliably replicated in subsequent studies. But we haven't yet seen this in the field of depression, and we will talk more about inconsistent findings later. Slide 15 These approaches are being used in a variety of ways for various possible uses. These will be expanded on later, but three key ways are, firstly in diagnostic or indeed, subgroup differentiation usually cross sectional. Secondly, in predicting outcomes, good or bad to treatment or lack thereof, and thirdly, in picking up relapse or risk for illness in people who are currently well. Slide 16 This is an image showing how machine learning could translate to useful findings in an area such as affective disorders. Let's say we wanted to see if we could identify factors which could indicate that someone presenting with depression had an unrecognised bipolar disorder. We would firstly gather a lot of data from a lot of patients, that's A. The data we put into a model would either be based on the available data that we already have, or existing knowledge of characteristics differing between unipolar and bipolar disorders or a combination that's B. An initial set of data would then be used to train the model, essentially looking at the factors distinguishing between bipolar and unipolar diagnoses, that C, and from this the results of the model emerge, that's D. The model is then tested for how well it performs on a new set of data that E and once this has been validated, theoretically, it can be used, for example, to actually make early diagnosis of bipolar from depressed groups in practice. Of course, though, there is needed comprehensive replication and refinement between E and F. Slide 17 Moving on to 21st century social subjects, we've already talked about smartphones a little bit, but specifically related to the rise in use of social media sites, there have, of course, been well publicised findings and views related to the relationship between social media and mental health, especially related to suicide. An example, which is often not discussed, is work showing that for some individuals, for example, people with social anxiety or in a community with severe mental illnesses, social media has provided social benefits and networks of other people with similar experiences and provides a platform for essentially peer support as well as public support and awareness. See here the article by Naslund et al. The study by Yu et al. is a good example of using research to monitor evidence and publicise benefits. The authors used monitoring of text online, so big data, to identify and demonstrate changing attitudes related to stigma and social support related to depression. Slide 18 Conversely, the potential harms of social media include, especially for young people, research into problematic smartphone use, and some even referring to this as an addiction disorder, although many addiction academics consider this premature. But the relationship between time using social media sites and young people has been linked to increased symptoms of 4 © King’s College London depression. The harms in this area can also be tools for intervention. See here the article from 2020 by Lopez Castroman et al describing an Internet based intervention to help prevent suicides. This is all work that is still relatively in its infancy, and we can expect to expand. Slide 19 Finally, we have to mention the pandemic when on this topic. Of course, there were widespread concerns in 2020 raised about the potential impacts of lockdown, isolation and COVID-19 illness on affective disorders. The field is currently moving so fast that it's hard to keep up, but early systematic reviews suggested that despite high population levels of psychological distress, the prevalence of diagnosable mental illnesses may not have changed much. This is subject to change and may be outdated even by the time you're watching this lecture. The first systematic review to assess rates of clinically significant depression and anxiety after having a COVID-19 infection identified rates that were similar to the general population at this time suggesting preliminarily that there may not be affective sequelae of the virus itself. That's the review by Bournistrova et al. Two other systematic reviews looking at the rates of mental illness in the general population have found high rates of distress, especially during lockdowns. But studies that carefully diagnosed affective disorders didn't find much difference either from before the pandemic or when compared to previous pandemics. It might be that these findings are partly due to a change in nature rather than extent of illness, and this could be changing over time, for example, within lockdowns, social anxiety may have decreased, but health anxiety increased. Slide 20 When looking at people with existing bipolar disorders one scoping review has so far identified a lack of clarity surrounding firstly, the impact on health care service use and provision, and, secondly, the well being of people living with bipolar with some mixed findings coming from different studies. The authors did suggest that the evidence pointed towards people with bipolar being more careful about virus safety than other populations. This may be due to anxiety levels, which are frequently high in this population, and/or the high comorbidity with physical health conditions that may increase clinical vulnerability. Slide 21 We need to make sure that people can access adequate care regardless of economic and cultural factors. This is still an issue even for an illness as common as depression in high income countries. But the challenges are more pronounced in lower income areas, and these challenges are compounded by a lack of cultural consideration in widely available, evidence based interventions. Interventions that are straightforward to deliver by, for example, lesser trained individuals but still show effectiveness are being prioritised globally. Psycho education is a good example, as evidenced in the referenced paper by Francisco Colom who highlights that psycho education is feasible to deliver in low resource settings because it doesn't require a lot of training or clinical expertise to deliver it effectively. It's best delivered in groups which reduces the cost per patient. 5 © King’s College London Slide 22 Globally, in the last decade or so the West has become more aware of the importance of culture and equanimity of care for people from diverse cultures. This is therefore a bit of a hot topic in depression. Cultural adaptation of interventions is on the rise, see this nice meta analysis from Hall et al. A really good example, which received a lot of attention, was the Friendship Bench, which started in Zimbabwe and is being applied to different cultures now. The Friendship Bench is an adapted problem solving therapy that can be delivered quickly and cheaply. It is also properly evidenced. The first cluster randomised trial included over 500 people with common mental disorders, and those receiving the intervention did a lot better than those receiving a control intervention, which included standard care plus general information education and support. The benefits that they received included depression severity and symptoms of mental health more broadly. 6 © King’s College London