09/04/2026
This week I have had a disproportionate number of calls and referrals from my OB/GYN colleagues regarding postpartum depression and postpartum psychosis.
The questions are almost always the same three:
How do you know?
How fast can it happen?
What is the best treatment approach?
Here is what I actually do.
Every pregnant patient I identify as being at elevated psychiatric risk comes into my office before delivery, and she does not come alone.
I want her partner, her mother, or whoever is going to be in a meaningful caregiving role once the baby arrives.
I educate all of them on the early signs of postpartum depression, postpartum mania, and postpartum psychosis.
Not just the mother.
The people who will be watching her.
That partner education is one of the most important parts of the entire process, and one of the parts I see missed most often.
Social withdrawal.
Difficulty bonding.
Emotional flatness.
Irritability.
Behavior that simply seems unlike her.
Families can misinterpret those things as rejection, indifference, ingratitude, or failure.
Sometimes they respond by pushing harder.
“You should be happy.”
“Why aren’t you enjoying this?”
“You need to bond with the baby.”
Shame then enters the picture at exactly the point when disclosure matters most.
Those are not character failures.
They may be early warning signs.
The correct response is not to correct her.
It is to call the clinician.
If depressive symptoms begin during the third trimester and she meets treatment criteria, I also begin thinking about authorization before delivery. Zuranolone criteria commonly include onset during the third trimester or within four weeks postpartum.
I would much rather deal with an insurance company while a patient is stable than start fighting for access while she is acutely ill.
Then there is postpartum psychosis.
People hear “abrupt onset” and assume it means there was no warning.
Those are not the same thing.
Early changes can include insomnia, restlessness, irritability, mood fluctuation, agitation, unusual behavior, and sometimes confusion.
The syndrome can then escalate very quickly.
And in a substantial portion of first-onset cases, the postpartum psychotic episode is also the first recognized presentation of bipolar disorder.
No prior diagnosis.
Sometimes no obvious family history.
Nothing on a routine risk screen that would have told you exactly what was coming.
Which is why observation matters so much.
One of the highest-yield questions I ask is incredibly simple:
When the baby sleeps, can she sleep?
I ask her.
And when possible, I ask her partner separately.
There is a major clinical difference between:
“She is exhausted and wants to sleep, but the baby keeps waking her.”
and:
“The baby is sleeping, but she is not.”
Or:
“She has slept three hours in two days and says she does not feel tired.”
Sleep deprivation is expected with a newborn.
Reduced need for sleep is not.
And reduced need for sleep is often noticed by somebody else before the patient recognizes it as a symptom.
That question costs nothing.
It may also buy you the most valuable thing in postpartum psychiatry:
time.
In Part 2, I am going to explain why the postpartum period is such a uniquely vulnerable neurobiological window: what happens to progesterone, allopregnanolone, estrogen, GABA-A signaling, glutamate, prolactin, the reproductive axis, sleep, thyroid function, iron status, and why those changes matter clinically.
