Rethinking the 'Cause' of Depression

Stephen M. Strakowski, MD


June 23, 2017

Hello. I'm Stephen M. Strakowski. I'm the founding chair of the new Department of Psychiatry at the new Dell Medical School at the University of Texas in Austin. Today I want to talk about depression and follow up on some implied discussion in a previous recording that we had a number of months ago.

Major Depression Overview

Major depression is among one of the most common ailments, and it may be the most common major medical problem once we eliminate viruses and so forth. About 7% of people are affected with depression annually, and there is somewhere around a 17% lifetime risk to perhaps as high as 1 in every 5 people.[1] Most people have either dealt with depression themselves or have family members or friends who've struggled with depression.

There are a number of studies that don't always agree, but in some studies it looks like depression is the leading cause of disability worldwide.[2] If not the leading cause, it's certainly one of the top two or three causes. There are higher rates in women, although there is some debate about that; it may be that it's just missed more often in men. It seems to be increased in people living in poverty and in stressful environments.

The cause of depression is unknown. It's so common for depression to be thought of as a heterogeneous condition, and so that means that there are a number of likely etiologies. Because of this, our research group published a paper in 2013 posing a consideration of whether depression should be thought of as a nonspecific response to brain injury rather than a specific condition per se.[3]

As this is a controversial discussion, I thought it would be interesting to talk about it today. I look forward to receiving comments from those of you who listen to these broadcasts.

Depression Criteria

The symptoms of major depression are fairly well known. They include depressed mood and anhedonia. Anhedonia is most likely the more specific diagnostic symptom of the two. Symptoms include changes in sleep and appetite, changes in psychomotor behavior—either agitation or retardation—fatigue, feelings of worthlessness or excessive guilt, impaired concentration, and suicidal thinking.[4,5]

Many of these—for example, the low mood or anhedonia—are very, very common in the general population, perhaps as high as 30% (for low mood) or more, and aren't always or even typically associated with the depressive syndrome.[6] In fact, the symptoms can be reliably diagnosed reasonably well. In the DSM-5 Field Trial, the kappa measure of reliability tended to be about 0.6 (0.8 or higher is very good and 0.1 or 0.2 is very weak[7]). So, researchers were able to recognize these symptoms in a person when two separate raters conducted the evaluation on two separate occasions.

However, the syndrome reliability of depression itself—when you accumulate the symptoms we have just talked about, whereby the patient has at least five symptoms and at least one is depressed mood or anhedonia—was poor. It ranged across sites from as low as a kappa 0.13, which is poor, to a high of a kappa 0.42, or only fair. This suggests that we are not always good at recognizing this condition,[7] perhaps due to the representation that depression is somewhat nonspecific.

The ICD-10 Field Trial was not particularly better.[8] Kappa values ranged from as low as 0.09, which is very low and very poor, to a 0.40, virtually the same range as seen in the DSM-5 trials. From a syndrome and symptom constellation construct, even though the symptoms are recognizable, there isn't good reliability among different clinicians. This, I think, contributes to some of the confusion around treating depression.

Depression and Comorbidity

A second common problem with depression is that it is perhaps the most common comorbid condition in all of medicine. If you look at Table 1, it is the most common comorbidity in psychiatry.

Table 1. The Most Common Comorbidity in Psychiatry: Major Depression

Bipolar I disorder 90%
Generalized anxiety disorder 67%
Panic disorder 50%
Schizophrenia 50%
Alcohol use disorder 40%+
Drug use disorder 40%+
Post-traumatic stress disorder 37%
Borderline personality disorder 30%+
Obsessive-compulsive disorder 30%
ADHD 25%

Strakowski SM, Adler CM, DelBello MP. Curr Psychiatry Rep. 2013;15:386-402.

Bipolar I disorder is defined by the occurrence of mania, and yet 90% of people with this condition will experience depression.[3] I have posited that the depression is simply a nonspecific response to the underlying neuropathology that causes mania, which is a much more predictive and specific syndrome. It is also much more reliably diagnosed.

In generalized anxiety disorder (GAD), depression occurs in almost 70% of people.[3] Over half of people with panic disorder or schizophrenia develop depression. Over 40% or more with alcohol and drug abuse develop it too. As you can see, depression is much, much more common in the presence of another psychiatric condition than it is in the general population. Again, this suggests that the underlying pathology causing the other psychiatric condition may be putting people at risk for depression, so depression is representing a nonspecific response to another brain pathology.

Similarly, in other types of medicine, as you can see in Table 2, depression is very, very common.[3] I actually believe that any condition that affects the brain increases the likelihood of depression. That includes medication and medication side effects. Any medication that's crossing the blood-brain barrier appears to increase the risk for major depression. That's often a risk that isn't considered when prescribing medications. If blood pressure medication is causing depression, that risk to the patient's health may be higher than, say, mild hypertension.

Table 2. Depression Comorbidity in Medicine?

Alzheimer's disease 50%
Multiple sclerosis 50%
Migraine 47%
Parkinson's disease 40%
COPD 40%
Epilepsy 35%
Cardiovascular disease 35%
Diabetes 33%
Stroke 30%+
Chronic renal disease 30%
Cancer 30%

Strakowski SM, Adler CM, DelBello MP. Curr Psychiatry Rep. 2013;15:386-402.

In Alzheimer's disease, multiple sclerosis, and migraine, the rates of depression tend to be as high as 50%.[3] In stroke, it ranges as low as 30% depending upon where the stroke occurred. But the rates of depression are very, very high following a prefrontal stroke. Even cardiovascular disease, renal disease, cancer and diabetes, which aren't thought to directly affect the brain, all have increased rates of depression. The mechanism here may be through inflammatory processes or other nonspecific general responses to illness. But again, this supports the notion that depression can often be a nonspecific brain response to another underlying medical pathology.

Genetic Risk and Depression

Additionally, major depression, it turns out, is one of our least heritable psychiatric conditions. Most of the major conditions we diagnose are very heritable. For example, bipolar disorder has a heritability of about 85%. Alcohol abuse, which is thought of as being strictly under behavior control, in fact has a genetic risk factor of over 60%. In contrast, major depression has a heritability risk somewhere between 31% and 40%.[9]

This risk increases in people who have very well-characterized, recurrent major depressive illness. These cases may fall under the classic construct of manic-depressive illness in which there are recurrent affective illnesses that don't just include mania. In those patients, the genetic liability approaches 70%.[10] The general run-of-the-mill depressed patient has a family history risk of only about 30%-40%.[9] This suggests, again, that there may be many, many more environmental factors affecting depression than in other major psychiatric conditions. Returning to bipolar disorder, 71% of the genetic liability for mania is independent from the risk for depression, again suggesting that depression may be a secondary component of bipolar I illness.[10]

Finally, in families of people with other primary psychiatric conditions, there are higher rates of depression. Families with bipolar disorder, OCD, Parkinson's, migraine, PTSD, and personality disorders have higher rates of depression, often higher than the rates of the primary genetic illness.[3] Again, this suggest that various types of brain pathology or even inherited risks for brain pathology increase the risk for a nonspecific depressive response.

Treatment Options in Depression

Major depression, as you know, responds to a number of nonspecific treatments. The SSRIs are relatively specific, but in any given patient, we don't know whether the patient will respond to an SSRI, SNRI, or bupropion, the latter of which has a different mechanism. In fact, the most effective antidepressant we have, electroconvulsive therapy, is among the least specific treatments we have. It suggests, again, that different types of depression perhaps are responding to different types of interventions. Unfortunately, at this point, we are not good at identifying which intervention is the most likely to be effective.

A number of new treatments, such as ketamine, offer a brand-new window into how we might manage depression. This is opening up the idea that depression is a heterogeneous and often nonspecific construct within the setting of other neuropathology. In fact, the placebo response in depression tends to be very high, which is part of the struggle that there has been in identifying good and specific antidepressant therapies.[11]

The psychotherapies that work in depression are also quite variable. The evidence-based ones like cognitive-behavioral therapy or certain types of interpersonal therapies work by very different mechanisms, and they don't work the same way in everyone. The nonspecificity and heterogeneity of the condition probably also affects the selection of psychotherapies.

There are many proposed mechanisms for what is underlying depression—monoamine hypotheses, GABA hypotheses, inflammatory hypotheses, and the HPA axis hypotheses among them. All of these may be correct in that they all may be components of this heterogeneous, nonspecific response to a variety of underlying brain pathologies.[3]

Future Considerations in Depression

With that in mind, we have an interesting construct in the way that we have started thinking about the underlying potential cause of depression, recognizing that we often aren't going to know it. If the patient has a major stress component, for example, simply providing medications without relieving the stress is likely to be ineffective.

If there is an underlying primary condition, such as bipolar disorder, OCD, or alcohol abuse, failing to recognize and treat the primary condition is probably going to make it much more difficult to manage the depression. In fact, managing the underlying condition may be what's necessary and may be all that's necessary to manage the depression.

What this requires of us is to be thoughtful and careful when we are assessing depression, to look for potential associations with other conditions. For example, do we really need to add this statin or antihypertensive or other drug that crosses the blood-brain barrier and increases the risk for depression? Or if someone is on a medication and is depressed, should we think about stopping the medication?

Very thoughtful, systematic treatment trials—recognizing what the potential contributors to this possibly nonspecific response are and remembering medication side-effects—are very important to treating depression.

Within all of this, I actually do believe that there is a genetic depressive condition that is likely a recurrent illness that we just haven't completely defined yet, so we're not able to separate it from the more general heterogeneous collection of depressive syndromes. As genetics, imaging, and other technologies progress, this is likely to change.

I look forward to hearing your comments and thoughts about this. If you'd like to read more about our thoughts on this topic, here is a link to our 2013 paper. I hope that we can continue to think about it collaboratively and cleverly to help manage people and improve their lives as much as possible.

Thank you for watching today. I appreciate it.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.