This article by Luigi Gagliardi discusses the incidence of white matter injury on head ultrasound across extremely low gestational ages. He was intrigued by Bree Andrews article from Chicago, showing no significant effect of gestational age on developmental outcomes among surviving infants.
The data he discusses shows a relatively constant incidence of white matter injury, despite major changes in mortality.
He notes the systematic review by Gregory Moore and others that shows a modest gradient in moderate and severe impairment at 4 to 8 years of age in former extremely preterm infants.
What he does not note is that, when Moore limited their analysis to severe impairment, there was no difference by gestational age groups. To remind everyone the definition of severe impairment that they used in their review of the published data was
an IQ score more than 3 SDs below the mean, nonambulant cerebral palsy (Gross Motor Functional Classification System, 4-5), no useful vision (worse than 20/200), or no useful hearing despite amplification
Other data are also consistent with this, when you look at large groups, and if you concentrate on scores on developmental screening tests at around 2 years, there is a modest gradient in outcomes by week of gestational age between 22 and 25 weeks. Once you get to an age at which developmental/intellectual evaluation is more stable and more likely to reflect later function, there is little or no evidence of a gestational age effect. Of course there are very few infants less than 23 weeks best-guess gestational age in these data sets.
Being smaller and more immature certainly affects survival, but once you get out of the NICU, there is little evidence that gestational age has a major impact on important functional outcomes.
Neonatal Research 