It is fairly common knowledge that the uterine environment affects fetal development; if you don’t believe that, you have clearly never tried to order a coffee or have a sip of wine in public while pregnant. Doing so is enough to elicit dirty looks and even nasty reprimands from complete strangers.
But it’s not just chemicals. Historical analyses indicate that waves of neurodevelopmental disorders occur after viral and bacterial pandemics. Studies in mice suggest that it is maternal inflammation, rather than a direct infection, that elicits these disorders; when pregnant mice are given proinflammatory molecules without any infectious agent, their pups exhibit altered behaviors. But the implications for human health haven’t been clear.
Now, a team has some evidence of a direct connection between inflammation in humans and changes in their offspring.
Interleukin-6 (IL-6), an inflammatory signaling molecule, is considered to be a marker for overall levels of inflammation. So a bunch of researchers (10 of them) decided to measure IL-6 levels in a bunch of pregnant women (84 of them) to see if they would correlate with any changes on the architecture of their babies’ brains or the behavior of their toddlers.
First, the scientists took blood from the women throughout their pregnancies and measured IL-6 levels. Once born, their babies were allowed to get used to life outside of the womb for a month. When they hit four weeks of age, they were put into a functional MRI machine that measured the degree of connectivity within and between 264 of their brain regions, organized into 10 functional networks. This connectivity can provide a sense of brain organization, and there’s a growing body of evidence that this organization goes amiss in those with neurodevelopmental and psychiatric disorders.
Then, the researchers used data from 68 of the mother-baby pairs to train a machine learning algorithm to correlate brain connectivity to maternal IL-6 levels. Based on what the machine learned, it was able to predict (retro-dict?) the IL-6 levels of the remaining 16 mothers based on the maps of their babies’ brains.
Connectivity within the network in charge of detecting salience was the most tightly linked to maternal IL-6 levels. Salience is a fancy way of saying relevance; this network is important for helping a brain isolate and focus on only relevant stimuli from an environment filled with noise. Connections between higher-order systems and subcortical regions were also altered. These connections are thought to be important for emotional and cognitive development.
Next, the researchers looked more closely at the specific brain regions altered in children who had been exposed to elevated IL-6 levels. They found that the specific regions overlapped with regions previously identified as associated with working memory in a meta-analysis of more than 900 functional MRI studies.
Working memory is the ability to hold items, like phone numbers, in your mind for a long enough time to work with them. It is an essential part of executive function, our ability to consider our own actions. It is also notably dysfunctional in autism spectrum disorder, ADHD, and schizophrenia.
Although the scientists did not look for overlaps with regions involved in all other behaviors and cognitive abilities, they did look at some, like language and negative emotionality, and did not find an overlap. This suggests—but as the authors themselves point out repeatedly, does not demonstrate—that IL-6 might specifically affect working memory but not other cognitive or emotional abilities.
Conveniently, working memory can be measured at two years of age. So, the researchers waited and, once these babies turned two, they measured their working memory. This was done by showing them eight different covered cups on a spinning tray, having the tots put stickers into six of the cups, putting the tray under the table, and spinning it. The test was to see if the kids could find their stickers now that the cups were in new places.
With this metric in hand, they again used their machine learning model to see if a mother’s prenatal IL-6 levels could predict her toddlers’ score on the memory test. It did, with higher IL-6 levels correlating with lower scores. IL-6 levels from the third trimester, when the fetal brain is undergoing its most intense maturation, were most predictive. Based on these predictive abilities, the researchers concluded that maternal inflammation (at least as represented by IL-6) adversely affects the connectivity among brain regions involved in working memory and, in fact, adversely affects working memory itself.
Obviously, there are limitations to this study. The babies were sleeping when their brain connections were measured. The regions where connectivity were diminished overlapped with those known from previous studies to be involved in working memory, but those previous studies were all done in adults.
And as the authors of the study themselves stress over and over again in their paper, this is not a demonstration that IL-6 specifically affects only working memory or that maternal inflammation causes psychiatric disorders. There are a lot of environmental and genetic factors impacting the fetal brain and a lot of behaviors that each factor can affect.
This is hardly the first evidence that changes in the womb can affect the developing baby’s brain, with links to later cognitive abilities and emotional equilibrium, as well as the development of later psychiatric disorders. Cigarette smoke does it, just like inflammation. But maternal inflammation is not only caused by infections. Inflammation can be caused by obesity, by malnutrition, by poverty, and by stress. All of these things affect mothers, and those effects are manifested and magnified in their babies’ brains and later functioning.
, 2018. DOI: 10.1038/s41593-018-0128-y (About DOIs).