Developmental trajectory of movement-related cortical oscillations during active sleep in a cross-sectional cohort of pre-term and full-term human infants

Abstract

In neonatal animal models, isolated limb movements during active sleep provide input to immature somatomotor cortex necessary for its development and are somatotopically encoded by alpha-beta oscillations as late as the equivalent of human full-term. Limb movements elicit similar neural patterns in very pre-term human infants (average 30 corrected gestational weeks), suggesting an analogous role in humans, but it is unknown until when they subserve this function. In a cohort of 19 neonates (31-42 corrected gestational weeks) we showed that isolated hand movements during active sleep continue to induce these same somatotopically distributed oscillations well into the perinatal period, but that these oscillations decline towards full-term and fully disappear at 41 corrected gestational weeks (equivalent to the end of gestation). We also showed that these highly localised alpha-beta oscillations are associated with an increase in delta oscillations which extends to the frontal area and does not decline with age. These results suggest that isolated limb movements during active sleep could have an important role in experience-dependent somatomotor development up until normal birth in humans.

Figure 1

Examples of movement-related localised EEG changes. EEG recording during isolated right hand movement (yellow vertical line) in infants with corrected gestational age 34 weeks (left panels) and 39 weeks (right panels). Both infants are in active sleep, with continuous EEG activity as expected in this sleep state; in these short examples rapid eye movements and irregular respiratory rate and depth are only evident in the older infant. The arrow points to the localised increase in alpha-beta oscillations post-movement over the contralateral central region (C3). Only EEG recordings overlying frontal, central and central-parietal regions are shown for illustrative purposes.

Figure 2

The amount of isolated right hand movements did not change with corrected gestational age. Amount of isolated right hand movements per 30 minutes of active sleep for each of 19 infants.

Figure 3

Isolated right hand movements were associated with a localised increase in alpha-beta power. Left: Median (and 95% CI) alpha-beta power pre and post movement at the contralateral central-parietal (CP3) and central electrode (C3). Right: Topographical heat map of the median change in alpha-beta power.

Figure 4

Isolated right hand movements were associated with a localised increase in delta power. Left: Median (and 95% CI) delta power pre and post movement at the contralateral frontal (F3) and central-parietal electrode (CP3). Right: Topographical heat map of the median change in delta power.

Figure 5

Movement-related increase in contralateral central alpha-beta activity declined with age. Lower panel: Alpha-beta power changes at C3 electrode associated with each movement from 19 infants with best linear fit to the data (solid line, on a logarithmic scale). Top panel: Topographical heat map of the median change in alpha-beta power for younger (left: 31–35 weeks CGA) and older infants (right: 36–42 weeks CGA).