We report more accelerated neurophysiological development for neonates who received skin-to-skin contact (SSC) over an eight-week period using both linear and complexity analyses of EEG-sleep behaviors, compared with two non-SSC cohorts. These behaviors are physiologic surrogates for multiple interconnecting neuronal pathways throughout the neuroaxis within brainstem, diencephalon and cortex which subserve state regulation (Steriade 2006
) (Datta and MacLean 2007
) (McCarley 2007
). Five specific neuronal pathways include the ponto-medullary to basal-frontal pathway subserving respiratory activity, the pedunculo-pontine/geniculocalacarine pathways identified with REM behavior, the ascending reticular activating pathway subserving arousals, the corticothalamic pathways subserving quiet sleep (NREM) expression and the cortico-cortical pathways representing spectral beta power and complexity. Our study results suggest that the immature brain responds differently to the developmental intervention of skin-to-skin contact when compared with groups not receiving this form of developmental care. The present findings reinforce our previous report of more organized sleep behavior after a single SSC session at 32 weeks PMA (Ludington-Hoe et al. 2006
) and support previously published findings that a different form of developmental intervention, Neonatal Individualized Developmental Care and Assessment Program (NIDCAP), which also documented accelerated brain maturation (Als et al. 2004
), based on EEG and neuroimaging measures. The beneficial results of neonatal developmental practices are also supported by the experimental evidence that maternal care epigenetically programs stress responses in offspring with effects on adult behavior. (Szyf et al 2007
We previously reported both advanced and delayed expressions of brain function in a healthy preterm cohort at term compared to a full-term cohort, based on a seven-item physiologic dysmaturity index comprised of seven EEG-sleep behaviors (Scher 1997a
) (Scher 1997b
). This dysmaturity index was derived after winnowing down from thirty-four to seven physiologic measures of EEG-sleep that best differentiated between two cohorts. We also reported that increased complexity was positively associated with brain maturation (Scher 2005
) (Janjarasjitt 2008
). Combined linear and complexity measures can better reflect the brain's physiologic adaptive response to the conditions of prematurity. This biological adaptive process was previously defined as ontogenetic adaptation (Oppenheim 1981
), and is now integrated into the contemporary concept of developmental neural plasticity (Cicchetti and Blender 2006
). The dysmaturity index, as defined by our research group, represents a physiologic marker of the brain's adaptation to the environmental influences of extrauterine life for a healthy preterm cohort who did not suffer postnatal illnesses (Scher1997a) (Scher1997b). Our findings of altered sleep organization and maturation after the developmental intervention of SSC, further extends this concept of adaptation of a preterm cohort to the therapeutic intervention of one type of developmental care. Given the long convalescent hospitalization of most preterm infants, attention to the maintenance of state homeostasis using various forms of developmental intervention has been emphasized (Bertelle et al. 2007
) as a strategy of neuroprotection. SSC appears to accelerate EEG-sleep state organization and maturation as a non-pharmacologic neuroprotective intervention when compared with two non-SSC cohorts. The prolonged benefits of these non-pharmacologic neuroprotective techniques will need to be more systematically studied in different categories of at-risk infants after discharge to the home setting. Recent functional MRI studies of mother's responses to their infant's facial cues emphasize the positive plasticity changes that are documented in the caregiver as a result of an enriched interaction with their infant (Strathern et al 2008). Recommended neuroprotective strategies that are beneficial in group studies, within the Neonatal Intensive Care Unit and at home, need to be tailored for the individual child relative to his/her specific strengths and vulnerabilities.
Our finding of more accelerated (i.e. more complex) right hemispheric maturation for the SSC cohort compared to both non-SSC controlled cohorts is supported by past studies which demonstrated right hemispheric dominance of the immature brain (Chiron et al. 1997
, Schore 2001
). We speculate that more advanced cortico-cortical connections exist in early life preferentially in the right rather than the left hemisphere, because of the greater responsiveness of the right hemisphere to sensory stimulation. This right hemispheric response to sensory input has been previously demonstrated based on the neonate's response to painful stimuli (Fernandez et al. 2003
). Despite a general paucity in the immature brain of cortical lateralization for many functions, the right hemisphere in the neonate demonstrates that more dominant cortical functions are already expressed for spatial sensory representation and arousal in response to positive or noxious stimuli.
To our knowledge, our study is the first to report the impact of SSC on neonatal neurophysiologic maturation over multiple-weeks, using both linear and complexity analyses. Comparisons among SSC studies remain difficult because different methods to access sleep organization were used, different methods of developmental intervention were applied, and environmental variables such as light, sound, and tactile stimulation were not uniformly reported. For example, studies reported no differences in sleep organization with changes in sleep input (Becker et al 1993
, Ariagno el al 1997
, Hellstrom-Westas etal 2001, Brandon et al 2001, Westrup et al 2002
, Mirmiran et al 2003).
Our combined use of linear and complexity analysis methods emphasizes the potential advantage of combining multiple analytic strategies to assess in brain maturation involving multiple interconnected neuronal pathways after neuroprotective interventions. While frequency analyses can best characterize the strength of a regional source by power spectra, complexity analysis methods can interpret physiological behaviors as a measure of the state of a dynamical system. Our research group previously demonstrated how another complexity measure, correlation dimension (CD), can characterize functional brain maturation in preterm populations (Scher et al. 2005
) (Janjarasjitt et al. 2008
). CD increases in neonates with maturation during both active and quiet sleep (Scher et al. 2005
) (Pereda et al. 2006
) (Janjarasjitt et al. 2008
). Non-linear methods also have been used to assess interdependences between EEG signals within different brain regions in both neonates (Pereda et al. 2001
). De la Cruz et al (2007) recently reported that both linear and non-linear interdependences coexist in the term infant while exclusively non-linear relationships exist in the preterm infant.
We recognize that our study cohort was small. Repeated assessments with a larger sample size are required, including equal numbers of subjects of each gender and with similar ranges of birthweights. Moreover, our study did not examine sleep measures when the infants were not receiving SSC. Therefore, the degree to which sleep changes carryover to non-SSC periods is unknown. However, our results are provocative because of the highly significant results in the same direction, supporting the conclusion that accelerated neurophysiologic maturation may occur in the SSC cohort. This is also supported by our discriminant analyses. Secondly, no contemporaneous controls were obtained. Historical controls certainly introduce bias, although the recording conditions for both SSC and non-SSC cohorts were the same in both study protocols based on feeding, recording parameters and environmental aspects in similar level III neonatal intensive care units. Also, this is only a preliminary study in preparation for more expanded testing and analyses with contemporaneous control cohorts. Thirdly, the neurophysiologist scoring the pilot studies was not blinded to the knowledge that SSC had been the pilot group at term age. Fourthly, EEG-sleep segments were scored without knowledge of pretest or test conditions. Lastly, neonatal cohorts with varying severities of illness who receive SSC were not tested since altered or injured brain circuitries may respond differently to developmental care intervention.