So a new study came out last week that was ALL OVER the internet and Twitterabout how Spanish researchers found a way to predict who will get postpartum depression. I saw the articles everywhere, butI waited to write about it because I wanted to understand exactly what the big new idea was before I covered it here.

In a published study, these Spanish researchers say they've figured out a way topredict 80% of postpartum depression cases, a method they sayhas higher accuracy than any other. As reported at US News & World Report:

"Early diagnosis of postnatal [or, postpartum] depression would make it possible to intervene to prevent it from developing among women at risk," Salvador Tortajada, a researcher at the Polytechnic University of Valencia and lead author of a new study on the methodology, said in a news release from the Scientific Information and News Service in Spain.

Per Psych Central, the researchers used artificial neural networks and extracted a series of risk factors highlighted in previous studies (stuff we already knew about):

• the amount of social support for the mother
• prior psychiatric problems in the family
• emotional changes during the birth
• neuroticism (a tendency to be less emotionally stable and more affected by stress than the average person)
• genetic variations in the serotonin transporter gene (genes with high levels of expression lead to an increased risk of developing the illness)

I read as many of the stories online about this as I could find. They all pretty much parroted the press release from Spain but it was difficult to find much more than that. None of them had a link to the study. None of them explained to me howthis method can or will be implemented in the real world to prevent postpartum depression. None of them explained how, once these women are identified, we can ensure they receive the proper treatment.

Anyway, after doing some digging I finally found the journal in which the study was published online: Methods of Information in Medicine. And then after doing some more diggingI found the study. It's called"Prediction of Postpartum Depression Using Multilayer Perceptrons and Pruning".

Huh? What the heck are multilayer perceptrons?

A multilayer perceptron is a feedfoward artificial neural network model that maps sets of input data onto a set of appropriate output.

Sorry. I don't speak Russian.

An artificial neural network is a mathematical model that tries to simulate the structure or functional aspects of biological neural networks. They can be used to model complex relationships between inputs and outputs or to find patterns in data. Feedforward refers to giving pre-feedback to a person or an organization from which you are expecting feedback.

Wha?

Are you lost? Me too.

Tell me more, you say? I can't. The article costs $25 euros and I just don't have $40 right now to spend on it. Besides which, I doubt I'd be able to understand a darn thing the paper says.

One thing I thought was odd, though:

They say they've discovered two protection factors that reduce the risk of getting postpartum depression– age (the older the woman the lower her chances), and whether or not a woman has worked during pregnancy (which reduces the risk). That seems strange to me, since I've known so many womenwho hadPPD who waited until they were in their 30s or later to have kids, and so many women towho hadPPD who had careers before they had them.