One method of understanding the workings of the black box brain is to observe what happens when it departs from normal baseline, desired behavior.  For example, what can postpartum depression tell us about the brain?  Do the treatments work and if so, what do those treatments also tell us about the brain?  How do we diagnose or even define postpartum depression and what does that tell us about the brain?

Let us break it down and start with depression.  The standard Diagnostic and Statistical Manual of Mental Disorders, edition 4 (DSM-IV) states that a unipolar, major depressive episode (i.e. depression) exhibits itself in a severely depressed mood for at least 2 weeks.  Starting from here, we already have a complication.  General health practitioners have more discrete and objective diagnostic guidelines: they diagnose strep throat for example not by observing a sore throat lasting more than 2 weeks but rather by the presence of any detectable streptococcal bacteria in the throat.  In contrast, whether a patient is diagnosed as suffering from depression depends on whether the patient expects to be depressed.  A person mourning or grieving is not depressed, though a person mourning or grieving for too long or for no apparent loss is.  Hidden but implied in that diagnostic manual description is the phrase, “abnormal, unexpected, and persistently depressed mood not commensurate with recent experienced events in the patient’s primary culture.”  That is a lot of subjective words.

Treatment depends on the diagnosis and causal model.  Since the starting diagnosis is already complicated, the treatment issue becomes more difficult.  Again, general medicine has the objective advantage: if strep throat is caused by the presence of streptococcal bacteria, eradicating the bacteria cures the disease.  Originally, the causal model for any mental disorder – depression did not have its own specific classification until fairly recently – included evil spirits.  Hundreds of years ago, treating depression did not so much involve physicians as it did exorcists.  By about the 19^th century with the industrial revolution in full swing, the causal model shifted to include a focus on the imbalance of environment.  Treatment during this period involved removing the patient from the chaotic environment causing the imbalance to a peaceful retreat called a sanitorium.  By about the mid-20^th century, the causal model shifted again with the discovery of the first anti-depressants: tricyclics and monoamine oxidase inhibitors (MAOIs).  Oddly enough, it was the discovery of the treatments that triggered the causal model shift rather than the other way around. 

A monoamine is a type of neurotransmitter that includes a single (the “mono-“ prefix) amino acid (the “-amine” suffix).  A prime example is dopamine.  Dopamine is a feel-good transmitter and also involved in reward based learning.  This is the transmitter through which highly addicative and illegal drugs such as cocaine work.  Oxidization is the breaking down process, so a monoamine oxidase (MAO) is a compound that destroys monoamines, including dopamine.  Finally, an MAO inhibitor (MAOI) prevents the MAO from destroying the dopamine, allowing the dopamine to continue acting as a feel-good transmitter.  Tricyclics and later, more advanced selective serotonin reuptake inhibitors (SSRIs) work under similar principles, but with slightly different targets.  Discovering that these drugs can stop the compounds that cause depression includes the discovery that some compounds can cause depression.

But again, since the underlying diagnosis comes from a subjective definition and not the presence of a foreign contaminant – everyone has MAOs in their body, but only some have depression – the practicing physician is forced to prescribe drugs and check if it stops depression.  If it did, then the patient was depressed and now is not.  If it did not, then the patient was not depressed, at least not in the way that can be treated with drugs.  The treatment confirms or pehaps even decides the diagnosis.  This is how the treatment triggered the causal model shift rather than the other way around.  Reversing the normal order in this manner introduces many more risks because the underlying causal model is less well understood.  The treatments are more likely to have unintended consequences. 

The state of affairs with depression is messy and not at all objective or clear cut.  This is a function not of our professional physician training but of the incomplete model of the brain.  While this is fine to acknowledge in theory, what can be done in practice?  We can do what any scientist would do: focus on a subset, analyze the data, and form a model of the specific phenomenon that may hopefully shed light on the phenomenon in general.  Postpartum depression is one such specific subset.

Postpartum depression affects 5-25% of all new mothers.  It can also affect new fathers.  Symptoms include, but are not limited to:

• Easily frustrated
• Feeling inadequate
• Low self-esteem
• Social withdrawal
• Sadness
• Guilt
• Overwhelmed feeling
• Exhaustion

The first half of the symptoms suggest personality traits.  Someone easily frustrated faces more difficulties in performing tasks, to the point where the cost of continuing to completion exceeds available resources.  Repeated incomplete tasks makes one feel inadequate and withdrawn with low self-esteem.  Being isolated out of a sense of failure allows the constant refrain of failure to echo in the thoughts, supporting a self-sustaining recurrent feedback loop.  The second half of the symptoms would aptly describe the state of affairs at this point.  Placing this chain of events into a simplistic work flow in parallel with the dopamine traces would look like this:

The dopaminergic reward based learning mechanism is well documented – dopamine levels spike above baseline upon anticipation of a reward, stay level on receipt of anticipated reward, and crash on unexpected miss of a reward.  Since dopamine is a feel-good transmitter, the spikes feel good (e.g. evening before an expected big birthday party with presents) and the crashes feel bad (e.g. everyone forgot the birthday and no presents).  In terms of postpartum depression, the goal anticipation may be that the new mother and father expect to be the best, ideal parents for their child.  The infant then cries for food, cries for company, cries for entertainment, cries for diaper changes, cries for exercise, and cries for the the sake of hearing crying.  The goal apparently is missed since few parents expect ideal parenting to result in crying.  But the final key is that infant calls are naturally pitched to prevent parents and caregivers from adjusting their goals away from being the best, ideal parents.  Infant cries are very difficult to ignore – perhaps it is evolution.  But that results in a lot of dopamine crashes.

There is internal feedback – the sense of failure from missing the goal due to infant crying increases doubt that increases the cost of accomplishing tasks that slows down task accomplishment that causes more missed goals.  There is external feedback – the infant is crying and constantly reminding the parents of apparently missed goals. 

An MAOI, for example, would remove those dopamine crashes and extend those dopamine spikes.  It would reduce the bad feelings, reduce the doubt, reduce the cost of accomplishing tasks, and allow the afflicted parent to function effectively.  But besides the somewhat discomforting concept of a pill that removes doubt, the dopaminergic reward based learning system requires dopamine spikes and crashes to learn.  Removing this action implies it would impair learning. 

Addressing postpartum depression within this dopamine system entails adjustments through learning.  In the birthday party example, the subject learns that sometimes people forget to organize large parties with many presents.  The subject may need to remind others of the date.  Whether this is done directly or subtly depends on the culture and the art form.  Regarding postpartum depression, perhaps this involves setting up the initial goal before the infant’s birth to include crying as a part of life.  Or perhaps it involves enlisting support of some form such that the parents feel comfortable enough to adjust their goals and directions in the face of unknown doubts and very loud crying.  This may involve support groups or cognitive behavioral therapy.  The specifics are beyond the scope of this post as it may require the input of many more experienced and knowledgable experts.  However, by understanding a different causal model of postpartum depression, parents may be able to address potential issues by first working with their brain and learning mechanisms while retaining the final safety net of overriding them via anti-depressants only as necessary.