Friday, October 23, 2015

Why is the prevalence of major depression not changing?

The availability of estimates of major depression prevalence (the proportion of population with a major depressive episode in the past year) from multiple Canadian surveys has made it possible to assess whether the prevalence of major depression is changing or not. The answer is "no" (click here for more details), or here. In a more recent study, we sought to determine why. Ahead of time, it was possible to say that there are several explanations due to the fact that prevalence, roughly speaking, is the incidence (rate of occurrence of new case) multiplied by the mean duration. With increasing treatment of depression, one might hope that episode incidence would decrease, since long-term use of antidepressants should reduce recurrence rates. Alternatively, treatment should shorten the duration of episodes. In this recent study, we looked at both things. But, neither one changed. Apparently, it is difficult to discern the effects of antidepressant medications on population health. Either the current efforts of the health system are failing to affect the burden of depression or other factors (e.g. increasing risk factors for depression) are at play that may have obscured this impact.

Sunday, July 19, 2015

Is depression and allergy or infection?

Depression has been a difficult disease to figure out scientifically. The frustration with a lack of progress leads to a natural tendency to want to hit a home run. Perhaps this is something like a football team that falls behind in the score, and starts throwing "long bombs" as a wishful strategy to pull out a victory. This mindset is currently colouring the interpretation of recent findings of a correlation between depression and various inflammatory indicators. Very frequently, we see non-critical representations of these findings as a possible revolution in our understanding of depression, here is a recent example. The idea that some of the behavioural and psychological manifestations of depression may be mediated by inflammatory messengers is decades old and was once discussed as a relationship between these medications and the syndrome of sickness behaviour seen in inflammatory states. There are, of course, many important implications of this finding, including the explanation that they provide for the negative impact of depression on the course of some inflammatory diseases (such as coronary artery disease) and the sometimes puzzling tendency for the strength of association of depression with medical illnesses not to associate closely with the degree of threat associated with those illnesses, e.g. see here. However, the overselling of this idea often seems to act against the aim of developing a comprehensive understanding of this condition - an overselling that, to me, seems driven by an excessive need to crow-bar knowledge of this condition into a more familiar disease model, an idea that I discuss in more detail here.

Saturday, June 27, 2015

Are antidepressants ineffective?

This question became quite controversial a few years ago when several authors made the assertion that these medications didn't work any better than placebo. This of course became a lightning rod for critics of pharmacological treatments. A few years later, I came across a very nice summary of the issue in a blog called Science-based Medicine. In retrospect, the answer seems to be the one that has now become widely accepted: that the medications do work, but not as well as might have been believed in the past. A nice thing about this contribution is that is is brief and easy to read, in distinction to some of the polemics that have appeared on this topic.

Friday, June 19, 2015

Suicidal thoughts in a chronic medical condition

A recent study looked at the predictors of suicidal thoughts in people with multiple sclerosis, which is a neurological condition characterized by inflammatory lesions occurring at different times and in different locations in the central nervous system. This work was summarized in a blog called "Everyday Health" and also in the Psychiatric Times. The frequency of suicidal ideation increased with age, which may reflect the increasing burden of such conditions as people get older. Also, the suicidal ideation was more severe in those with very difficult symptoms  to manage, such as bladder and bowel symptoms or difficulty swallowing. But, it is not just a situation of having to deal with difficult problems. The ways in which people coped with problems was also important. For example, those with emotionally focussed coping strategies (looking for ways to avoid painful emotions) had more suicidal ideation that those that used solution-oriented or problem solving strategies. A link the study can be found here.

Monday, June 8, 2015

Is Cognitive Behavioural Therapy Becoming Less Effective?

A recent meta-analysis examined the efficacy of CBT for unipolar major depression, finding a trend towards lower effectiveness over the past few decades. A copy of the paper may be obtained here. The authors hypothesized that diminishing effectiveness may be related to one of two factors: (1) as the therapy has become wildly popular, therapists may be diverging from the manualized protocols for CBT, in other words not administering the therapy to the same high standards as earlier clinical trials did, or (2) as more experience has been gained with the therapy, some of its mystique may have worn off - leading to lower placebo response in the active treatment arms. On the latter point, it is important to remember that in a trial the subset of participants receiving CBT has outcomes that are determined both the specific impact of the treatment and non-specific factors related to the therapy. I wonder if there might not be another explanation. In past decades, few people outside of the mental health world were even aware of CBT and often even the first few, psychoeducationally oriented, CBT sessions were real eye openers for them. The basic concepts of CBT sometimes led to a big change in peoples' ways of looking at the world. In this day and age, however, everyone has heard of CBT, many of read books about it or encountered similar ideas about the relationship between thoughts and emotions in popular media. Skillfully delivered CBT has a lot to offer people who only have a cursory knowledge of it, of those who have tried to do it themselves with books or websites devoted to CBT. However, nowadays, some proportion of the gains may already have been made before people enter therapy.

Monday, May 18, 2015

Understanding depression?

These are just a few musings about depression research. As a researcher, I am very interested in understanding the direction that the field of major depression is going. There has been an important change with the emergence of RDoCs - which is a series of dimensional symptom ratings that are intended to replace or complement (depending on who you talk to) the broad diagnostic category of Major Depression. I am at a conference today and it strikes me that there has been a large impact of this idea in the research community. It has become popular to say that there are 100s or 1000s of forms of depression, and that by characterizing these carefully and individually, we will be moving towards "precision medicine" approaches. I find this fascinating. It is certainly true that all illness categories are a kind of short-hand describing patterns of pathophysiology (and this case psychology and social interaction as well) that cannot be exactly the same in every person afflicted. This heterogeneity is important and it makes sense (to me) to try to parse it out. This could lead to therapeutic gains. However, what might be even better would be insights that help to integrate and simplify this diversity, providing perhaps a simpler explanation for the diversity of experiences and clinical and neurobiological findings - reflecting the scientific parsimony of Occam's razor. The former approach seems to be currently in vogue, whereas the latter approach appears to have declined over the past few years. However, parsimony always wins in the end.

Monday, April 20, 2015

Childhood Adversity and Major Depression

A very consistent result in the epidemiological literature is the association between childhood adversity and the subsequent occurrence of major depression during adulthood. More recently, this concept has been expanded with studies reporting associations between childhood adversity and a large number of different adult health outcomes (e.g. cardiovascular disease). A weakness in this literature is that most of the studies arise from retrospective assessment of childhood adversities - which is problematic since peoples' perceptions of their childhoods could easily be influenced by their adult health status. We recently sought to get around this problem by linking a childhood survey to data collected later, during adulthood. There were two outcomes that stood out as being strongly related to childhood adversity: depression and smoking. These are, of course, related outcomes since smoking is strongly related to depression in adolescents, who may smoke to self-medicate and depressed adolescents may also be at much higher risk of becoming nicotine dependent. These findings, summarized here, suggest that depression may be the "lynch pin" linking childhood adversities to later health problems through intermediate mechanisms such as smoking. This may provide an opportunity (management of depression) to prevent many of the later-life health problems that seem to be associated with adverse childhood events.

Thursday, April 16, 2015

Prevalence of Bipolar I and II Disorder in Canada

There have only been two national surveys of mental health in Canada. One of these was conducted in 2002 (the CCHS 1.2) and a second one in 2012 (CCHS-MH). The first survey provided an estimate of bipolar disorder prevalence, see here. This prevalence was higher than expected, so a modified instrument that distinguished the two main types of Bipolar Disorder (I and II) in the 2012 survey. The prevalence has been revisited by Keltie McDonald, an MSc student in epidemiology at the University of Calgary. She found what seemed to be a more plausible prevalence, about 0.9% (traditionally, it has been believed that about 1% of the general population have Bipolar Disorder). However, there some very substantial inconsistencies also observed, e.g. few people with bipolar disorders were taking lithium, and few who reported that they had been diagnosed with Bipolar Disorder were detected by the version of the interview used in the survey. This paper raises some questions about the accuracy of research diagnostic interviews to accurately assess this condition. The result is not actually very surprising. The epidemiological diagnostic interviews used in this type of survey follow inflexible scripts and are well known not to perform well for all diagnoses. It is likely that the best way to study this condition will need to use data from other sources, such as anonymous, aggregate statistics from the health system, e.g. see another paper by Bulloch et al.  

Thursday, April 9, 2015

Stigma against depression in Canada

A recent paper by Dr. Heather Stuart, of Queens University has addressed the question of whether there is stigma against depression among the general public in Canada. She used data from a large national "Rapid Response" survey conducted by Statistics Canada. The survey focussed on people who had sought treatment for mental health - Sadly, a large proportion reported that they believed that others stigmatized depression (about half), whereas one third reported that they themselves had encountered stigma. The full test of the paper is available here.

Models and Metaphors for Major Depression

Since major depression is a condition that is not yet fully understood, there is a lot of thinking about it that is shaped by general ideas rather than facts. In the medical literature, there are several ideas about depression that seem to be shaped by analogies to other conditions. The ways in which better understood conditions work, seems often to shape the ways in which people strive to understand more mysterious conditions such as major depression. In a recent paper, I have sought to describe eight such models. Since they are ideas that are usually based more on analogy than solid evidence, it is perhaps fair to call them "metaphors" rather than models or theories. Here is my list of eight such metaphors:

(1)  as chemical imbalances (e.g., a presumed or theoretical imbalance of normally balanced neurotransmission in the brain).

(2) as degenerative conditions (e.g., a brain disease characterised by atrophy of specified brain structures).

(3) as toxicological syndromes (a result of exposure to a noxious psychological environment, e.g. "toxic stress").

(4) Injuries (e.g., externally induced brain damage related to damaging stressors).

(5) deficiency states (e.g., a serotonin deficiency).

(6) an obsolete category (e.g., similar to obsolete terms such as 'consumption' or 'dropsy').

(7) a medical mysteries (e.g., a condition poised for a paradigm-shifting breakthrough).

(8) or evolutionary vestiges (residual components of once adaptive mechanisms have become maladaptive in modern environments).

Each of these possibilities may have elements of truth within them and may ultimately blossom into full-fledged theories capable of effectively structuring knowledge. However, for now it is important to remember that the biomedical "pedigrees" of such ideas may provide them with more weight than they really deserve. 

Wednesday, March 4, 2015

Emergence of Mood Disorders in Adolescence

There is solidifying consensus that the earliest manifestations of mood disorders occur often during childhood and adolescence. This kind of statement is most strongly supported for the highly recurrent forms of mood disorder such as bipolar disorder or patterns of melancholic depression. What implications these early-onset disorders may have for clinical care and public health are yet to be determined. A common assertion is that screening should be used for early detection, but the mere fact that these disorders have an early onset does not in itself support the use of screening. Many of th early symptoms of these disorders are likely to lack specificity, something that is not apparent in data from studies focusing on high risk or ultra-high risk cohorts. Most people with bipolar disorders, for example, may have had non-specific signs such as mood swings, irritability, substance abuse or behavioural problems, but this does not mean that a large proportion of people with those signs will go on to develop bipolar disorder. The interesting scenario that arises as a result of these issues was the focus of a Google "hangout" discussion sponsored by the Journal Evidence Based Mental Health. The discussion has been recorded and is available by clicking here.

Tuesday, February 24, 2015

Epidemiology for Canadian Students

I have been teaching the Fundamentals of Epidemiology more or less continuously at the University of Calgary since the early 1990s. I have always been frustrated by the lack of a text book that emphasizes (or even mentions) the main sources of data that students of epidemiology in Canada will use during their careers. For this reason, I have written an introductory textbook that I hope Canadian students will find useful. For more information, click here. In addition to emphasizing (not exclusively) the Canadian context, there are a couple of features that have been emphasized. One is that each chapter is divided into two parts. The first part provides basic content and the second, called "thinking deeper" delves a little deeper into a more advanced topic related to that chapter. In this way, I have attempted to make the book appealing both for undergraduate students and graduate students. Second, rather than reserving the topic of bias, which is a difficult topic for complex parameters such as odds ratios or risk ratios, this topic is addressed in chapters focusing on simpler parameters, such as prevalence. I hope that this will make some of the most difficult topics for introductory students (selection bias in analytical studies, differential versus non-differential misclassification bias) more accessible since they can first master the concepts in a simpler form.

Saturday, February 7, 2015

Updated epidemiological information on depression in Canada - good news or bad?

In the January 2015 issue of the Canadian Journal of Psychiatry we published some updated information on major depression epidemiology in Canada. These new estimates come from a survey called the CCHS-MH, which employed an adaptation of the World Health Organization World Mental Health Composite International Diagnostic Interview and had a sample of n = 25 113. Previously, national estimates were only available from a 2002 survey (the CCHS 1.2) or from studies using a short-form version of the diagnostic interview. Among survey respondents with past-year MDD, 63.1% had sought treatment and 33.1% were taking an antidepressant (AD). This report confirms prior results indicating that antidepressant use has stopped increasing in Canada, however, an encouraging sign is that treatment-seeking appears to have increased. There are frequent assertions that only a minority of people with major depression seek treatment, but this is no longer the case in Canada. It is likely that there have been increases in treatments other than antidepressant medication, such as psychotherapy or second-generation antipsychotic medications, however, additional studies are needed to confirm these hypotheses. The 2012 survey also provided an initial opportunity to examine comorbidity between major depression and alcohol abuse and dependence. The earlier national mental health survey (conducted by Statistics Canada in 2002) was unable to assess abuse and dependence due to the measurements instrument selected for that survey. In the 2012 survey 4.8% had past-year alcohol abuse and 4.5% had alcohol dependence. In the same issue, we apply some quantitative data synthesis techniques to pool estimates of major depression prevalence across many surveys that have looked at this in Canada's national population. We used a meta-regression technique to determine whether there was an increase in major depression prevalence, finding the slope not to significantly differ from zero, indicating no-change in prevalence. The abstract for the paper is available here. Is this good news or bad news? It does combat the notion that there is an epidemic of depression in Canada, e.g. see here, which is simply not true. On the other hand, with the increases in treatment that have occurred since the 1990s, one might have hoped for a decreased prevalence, which has also not occurred.