Is Postpartum Psychosis an Autoimmune Disease?

Here's an article for the "everything is connected" file.  Also for "yes, psychiatric disease has biologic underpinnings and is medical illness" file.  Also the "inflammation in the wrong place at the wrong time is super-bad" file.  And it may be of interest to anyone who has had symptoms of autoimmune disease helped by an anti-inflammatory (paleo-type) diet.

Postpartum psychosis is rare and scary.  About 1 in 1000 women become psychotic in the first months after having a baby (though anything up to 12 months after is considered "postpartum" the greatest risk is in the first month).   The most typical presentation is one of manic psychosis, with prominent insomnia, irritability, and delusions of grandeur.  However, some women will also be depressed and be delusional and suicidal, or even with delusions that lead a women to kill her baby.

Not surprisingly, a prior diagnosis of bipolar disorder is the greatest risk factor for developing postpartum psychosis.  However, most women with postpartum psychosis have no history of psychiatric illness at all (1). Often the illness requires hospitalization, and though there are no "consensus treatment guidelines," in almost all cases benzodiazepines (sedative, anti-anxiety meds, such as lorazepam) are used to help stabilize sleep-wake cycles, and in most cases antipsychotics are also used, typically with good effect.  If those aren't helping, lithium is added.

Here's another bit of info about pregnancy.  The fetus is obviously genetically different than mom, so women develop a depressed immune system during pregnancy, in order to protect the growing beastie from mom's antibodies and killer cells.  This is why I was told to studiously avoid unpasturized cheese and raw eggs and deli meat during pregnancy, and why healthy women in the third trimester are much more likely to develop severe complications and die from the flu than women who are not pregnant.

It is well known that in women with a dysfunctional immune system (the autoimmune diseases, such as multiple sclerosis, rheumatoid arthritis, and autoimmune thyroiditis), the autoimmune symptoms are generally greatly ameliorated during pregnancy.  However, this time of relatively low autoimmune symptoms is followed in the post-partum period by a "rebound" with greatly increased symptoms and greater autoantibody titers measured in the serum.

So is post-partum psychosis a symptom of autoimmune disease?  Specifically autoimmune thyroid disease, as thyroid disease (both hyper- and hypothyroidism) is well known to cause psychiatric symptoms, even psychosis?

Well, those societies with socialized medicine were able to gather data in such a way as to start to give us an answer to that question.  In the Netherlands, all the women in a certain area of the country who developed post-partum psychosis and ended up in the hospital were checked for autoimmune thyroid antibodies and thyroid function upon admission to the hospital.  A larger control group of other post-partum women were also checked.  Critically, women who were medicated at admission (particularly with lithium) were excluded from the study, as lithium is known to depress thyroid function.  All women with a previous history of thyroid disease, bipolar disorder, schizophrenia, or psychosis were excluded.  That left a group of 29 women with new-onset post-partum psychosis and 117 controls.

Here is what the researchers found.  5% of post-partum women in the control group had measurable autoimmune thyroid autoantibodies at 4 weeks after delivery, a sign of autoimmune thyroid disease.  This is comparable to surveys of a general population of women in the Netherlands.  None of them had measurable abnormalities in thyroid function or any symptoms.  In contrast, 19% of the post-partum psychosis patients had measurable thyroid autoantibodies at admission (again, prior to receiving any lithium or antipsychotic medication treatment), and half of those women also had measurable thyroid abnormalities.  In the following 9 months, 67% of the postpartum psychosis women with autoimmune thyroid antibodies went on to develop measurable thyroid problems (abnormal TSH or free thyroxine).  None of the control women did.  The odds ratios for these findings were all >2, some as large as 9, which is quite significant, especially considering the size of the sample).

Even though patients with previous bipolar disorder were excluded from this study, the researchers note that the 19% prevalence of autoimmune thyroid antibodies in these psychotic women is similar to the prevalence in women with bipolar disorder (2).  And, to really get your noggins going, twin studies of bipolar disorder show that the presence of autoimmune thyroid antibodies are correlated not only to the illness itself, but to the genetic vulnerability to the illness (3).

The researchers in this study strongly recommended that all women with postpartum psychosis be monitored for thyroperoxidase antibodies and thyroid function abnormalities, and furthermore that all women at high risk for postpartum psychosis be monitored before and throughout pregnancy and the postpartum period.  Though this was a small observational study, the advice seems very reasonable.

And, as always, we find that "post-partum psychosis" like many psychiatric symptoms is the equivalent of a fever - signaling underlying abnormalities, but not always caused by the flu.  Sometimes fevers are caused by different bugs, or cancer, or autoimmune disease.  Differentiating the underlying pathology will go a long way to informing our treatments (and helping in prevention) in the future.

Psychiatrists and other doctors reading this article will be interested in one directed more to healthcare professionals about the same study at the MGH Center for Women's Mental Health blog by Ruta Nonacs, MD PhD.  Thanks to Dr. Trevisan for the link to the blog post!