Although it’s well known that neurogenesis or the production of new neurons occurs throughout the lifespan, there are only a few select areas of the brain that continue to do so. Examples include the subventricular zone underlying the neocortex that supplies new neurons for the olfactory bulb, as well as the dentate gyrus of the hippocampus, a structure associated with memory.
In the prenatal period, however, neurons supplying the neocortex are produced up to 29 weeks of gestation in humans [1]. (Pregnancy lasts approximately 40 weeks.) Malik et al. (2013) have found that the hypoxic (low oxygen) environment of the womb is vital for the continued proper development of the brain. Not only did they find that neuron production was reduced in preterm infants, but they also found that in an animal model of preterm delivery, normal neurogenesis could be partly resumed by the application of an hypoxic mimic, dimethyloxallyl glycine.
Stages of neurogenesis in the cerebral cortex. Borrowed from here.
Premature birth is a considerable risk factor for a variety of neurodevelopmental conditions, including autism. For example, Limperopoulos et al. (2008) reported that 26% of the preemies in their study displayed measurable symptoms of autism in infancy, although it’s unknown how many of these infants would have gone on to receive an autism diagnosis in childhood as this was not a longitudinal study. However, they did find that younger gestational age was an important factor in increased risk for positive scores on the Child Behavior Checklist and Vineland Scales.
Another study which used the M-CHAT in toddlers to assess autism symptomology found that the presence of other disabilities in the extreme preterm population (<26 weeks gestation) could lead to diagnostic confusion. In fact, large percentages of those with severe motor, cognitive, hearing, or vision impairments screened positive on the M-CHAT, prompting the authors to advise caution in applying the autism label to this population without more thorough testing [2]. However, another study by Johnson et al. (2010) found that extreme prematurity was indeed a risk factor for autism, with 8% of their sample eventually receiving an autism spectrum diagnosis by later childhood. Thus, this last study is probably a better estimate of true autism risk in extreme preterm children. But it’s also apparent that these preterm kids are at high risk for a variety of neurodevelopmental and medical conditions.
One of the more interesting points to note, however, is that extreme preterm delivery overlaps the period in which neurogenesis is still occurring in the developing brain, suggesting that disturbances to the production of neurons may play an important role in autism risk. In support of this, premature birth often coincides with an increased risk for haemorrhage (bleeding) within the germinal zone (neural stem cells). Del Bigio (2011) reports that haemorrhage in preterm infants is accompanied by decreased neuronal proliferation.
However, from our own recent work, we know that genes involved in autism are likely to influence multiple stages of a neuron’s development. Therefore, it is quite reasonable to suspect that not only is neurogenesis perturbed in extreme prematurity, but maturation of already-born neurons is probably also affected. How this ultimately leads to autism in only a subset of cases is still a partial mystery. Hopefully future research will allow us to better understand when, why, and how autism occurs and specifically what prematurity does to tip that balance.
Science Over a Cuppa 
I am retiring from service to the state of North Carolina. For the last 9.5 years I served as a psychologist with the Children’s Developmental Services Agency (CDSA) which is part of the Department of Health and Human Services (DHHS). A large part of my job involved evaluation children who had not reached their third birthday. Many of them were born prematurely (less than 38 weeks gestation). I diagnosed a lot of autism. It was difficult for me to apply that diagnosis to children under 18 months old although parents often knew or suspected that symptoms were present in their children as early as 6 months old. Post-term babies (gestation of more than 42 weeks became an interest when I saw a baby with confirmed gestation of a little more than 43 weeks. The child had poor muscle tone. He was born with meconium stain. My impression was that there would be more developmental problems as he got older. You might want to look into post-term births. One study indicated that more children might be post-term than pre-term. Post-term children develop behavior problems according to that study. I also happen to be seeing a 35 year old female patient that was delivered at 44 weeks gestation. In the vernacular, “she is a mess.” A nurse associate of mine at the CDSA tells me that post-term births are extremely rare now. Hmmm
I have also suspected the same, though not being a clinician I’m not in a position to do much else but wonder and read articles. However, the idea of striking an ideal gestational age “balance” makes sense. As one example, the switch of GABA from excitatory to inhibitory perinatally is probably well-timed, and shifts in either direction could disturb that balance. Thanks for your input, Steven.
Dear Emily
This is an excellent blog about the link between premature birth and autism, as a person with autism myself whom had a premature birth, cerebral asphyxia (leading to left hemisphere brain injury – which resulted in visual agnosias and receptive aphasia and other conditions) and a placental abruption. Thank you for this informative blog I shall share this. 🙂
Kindest regards
Paul
Thanks, Paul, glad you like it. Hopefully it’ll be helpful to others. 🙂
The Johnson study is important since it is the only population based study of pre-term births associated with later autism diagnosis. The authors retrieved the birth records of all pre-term births in the UK and Ireland born in 1995 and followed up the children more than a decade later. Studies that have selected all poor outcomes associated with specific unfavorable pre peri or neonatal events or factors have consistently reported that around 7% of those with any developmental problem result in an ASD diagnosis at follow-up.
See table2 of my paper:
Click to access 1379116999.pdf
I have proposed that the 7% conundrum can be explained by the approximate prevalence of familial broad autism phenotype in the general population and that the genetic influences underlying familial BAP is a background genetic effect that is always reliant on other genetic, environmental and epigenetic events that can lead to an ASD diagnosis.
Yeah, didn’t you mention something to this effect in relation to lyme disease or something of the like? I remember, I still have the paper and intend on reading it. But it was a very striking observation, one I’ll definitely be keeping in mind in future research.
Pingback: Prematurity, Brain Bleeds, & Autism | Science Over a Cuppa·