Introduction

Neonates, and specially premature newborns, had long been considered devoid of sensory perception and even of the ability to feel pain [1]. But, in recent decades, a growing body of evidence, based on numerous studies on the perceptual and cognitive abilities of fetuses and newborns, has demonstrated that sensory systems are relatively mature at (term) birth, at both peripheral and central levels (for reviews: e.g. [2], [3], [4], [5], [6]). There are different possible explanations for this early undervaluation: one main aspect is that neonates are non-verbal individuals, which means that the initial assessment of their sensory abilities was made by adults caretakers that did not necessarily share the same sensory world; moreover, the immaturity of neonates may have been overestimated, by supposing a priori limited sensory and cognitive skills; and at the same time, there were technological limitations that prevented the direct observation of fetal behavior and the study of infant brain function.

Data on perinatal chemosensory and auditory perception are abundant, while the extent of newborns’ tactile sensory capacities has been poorly investigated [6] despite the fact that receptors and neural pathways associated with tactile perception are long known to be the first to develop in utero [7]. As soon as the 7^th week of gestation, the different types of skin sensory receptors gradually develop to be finally present on the whole skin by the 20^th week [8], [9], while the connections between the spinal cord and the brain are functional only by the 20^th-24^th week [10]. Passive tactile stimuli (e.g. pressures on the maternal abdomen) can already be perceived by the fetus [11], [12], while premature and full-term newborns are known to perceive and react not only to nociceptive stimuli (e.g. heel lance; e.g. [13], [14], [15], [16]), but also to lighter stimulations such as abdominal or foot tactile stimulations of only a few grams (e.g. [17], [18], [19]). Pre-term and full-term infants can also perceive, through touch, objects’ characteristics like shape, texture and weight, thus demonstrating well-developed active touch abilities (e.g. [20], [21], [22], [23], [24], [25]). The question of the sensitivity to passive touch (i.e. sensitivity to a stimulation imposed on the individual’s skin) is of particular importance for infants born prematurely (before 37 weeks of gestation) who need intensive medical care and develop in sensory environments (Neonatal Intensive Care Units) dramatically different from those in which full-term infants mature and which consist in numerous sensory (including tactile and sometimes noxious, e.g. repeated blood sampling) stimulations during a critical period of brain development [26]. In pre-terms, there are discrepancies between studies as some indicate a lower reactivity of premature newborns to tactile stimulations as compared to their full-term counterparts [17], [18], while others point out a higher sensitivity [19], [27], [28], [29], [30] which continues at least throughout the first year [31]. This discrepancy may be due to different factors, especially relative to the location (e.g. abdomen, thigh, leg, forearm) and/or intensity (e.g. from 0.6 to 26 g, soft stroke) of the tactile stimulation, as well as the measurements of the subject’s responses. In particular, some authors have examined only the local reflex following a tactile stimulation (e.g. [19], [27], [28]) while others have taken into account less localized behavioral responses (e.g. [32]). Differences between premature and full-term infants may exist in terms of specificity and localization of the responses, as demonstrated by EEG and EMG recordings [29], [33]. It has been argued that premature infants may lack the inhibitory circuits or the specialization of required brain areas to respond selectively to relevant tactile stimuli at the appropriate location, showing more extensive responses than full-term infants even for non-nociceptive stimuli [19], [33]. Another category of newborns, the early-term (born between 37 and 38 weeks of gestation), are of interest. Although data are relatively scarce, there is a growing body of evidence that there are risks associated with early-term births such as increased morbidity and lower later cognitive abilities (e.g. [34], [35], [36], [37], [38], [39]), which has led to the recent reassessment of their “status of full-term” ([34], [40]). Moreover, neural specialization for tactile perception seems to present a critical period of strong brain restructuration around 37 weeks post-conception [33], which makes this precise time of birth of special interest for testing tactile sensitivity.

In the present study, we tested the hypothesis that infants born either pre-term, early-term or at term differ in their tactile sensitivity to passive touch at term-equivalent age. The reasons why they could differ may be differences in terms of maturation and/or sensory experiences (e.g. exposure to enhanced sensory stimulations in neonatal intensive care units). If the hypothesis that the specificity and selectivity of tactile responses are related to gestational age, we would expect full-term newborns to be less responsive to a light (0.008 g) mechanical stimulus and to have more localized responses. In addition, we hypothesized that adults, who have fully developed inhibitory systems, may not perceive such a light, non-“ecologically-pertinent” stimulation. We therefore compared their tactile sensitivity of that of the infants, using the same experimental device. The results of such investigation are of primary importance as they can give a novel view on how routine practices may be perceived and also on how the newborns’ behaviors may inform us about their perceptions. This work is also important in understanding development of tactile processing in general.

Methods

2) Study participants

Twenty-nine infants and thirty young adults were included in the present study.

a) Infant participants and groups.

Twenty-nine infants, born either preterm (25 to 36 weeks of gestation; N = 10), early-term (37 to 38 weeks of gestation; N = 9) or full-term (39 to 41 weeks of gestation; N = 10) were recruited from the neonatal intensive care units (NICU) and maternity wards of two university hospitals (15 infants from Rennes, 14 infants from Brest) that are known to share the same general practices: on both sites, the NICU followed the developmental care guidelines of the Newborn Individualized Developmental Care and Assessment Program (NIDCAP), which recommend in particular that lights and sounds be reduced as much as possible and the infants‘ activity rhythm be respected. These guidelines encourage parents to visit whenever they want (day or night) and for as long as they wish, and to perform skin-to-skin and routine care when present. To be included in the study, infants had to be without clinical (i.e. respiratory and cardiac distress) and/or major congenital complications, suspected infections, neurological and sensory perception disorders, and/or under analgesic or sedative treatment.

The characteristics of the three groups of infants are detailed in Table 1:

1. PRE-TERM Group consisted of ten preterm infants (50% females) born at 31.2 (± 3.4) ( ± SD) gestational weeks with a mean weight at birth of 1427.5 (± 459.9) g.
2. EARLY-TERM Group was composed of nine early-term newborns (33% females) born at 37.9 (± 0.6) gestational weeks with a mean weight at birth of 2605.6 (± 405.1) g.
3. FULL-TERM group consisted of ten full-term newborns (70% females) born at 40.7 (± 0.5) gestational weeks with a mean weight at birth of 3441.3 (± 451.8) g.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of the populations at birth and at time of testing.

https://doi.org/10.1371/journal.pone.0229270.t001

b) Adult participants.

Thirty adult participants (50% females; aged from 21 to 36 years old) were recruited. They had all been full-term newborns and did not present any cerebral lesion or known important sensory perception disorders at birth and at the time of the study.

Results

Responses of pre-term, early-term and full-term infants to the light tactile versus sham stimulation (S1 Table)

In total, 18 of the 29 term-aged infants (62.1%) reacted to the tactile stimulation whereas only 8 (27.6%) showed a behavioral change during the sham stimulation (McNemar’s test, N = 29, X² = 5.56, df = 1, P = 0.018) (Table 2, Fig 1). In addition, the mean intensity of the response was higher when exposed to the tactile (X±ES = 1.86±0.36) than the sham stimulation (X±ES = 0.55±0.23; Table 2). The most frequent reaction (14 out of the 18 reacting infants, Chi square Test, N = 18, X² = 4.5, p = 0.034) was a non-localized response, i.e. a movement of a body part other than the hand touched, to the localized tactile stimulation. No effects of subject’s sex (girls versus boys: Fisher Test, N = 29, df = 1, p>0.1) or side of presentation (right versus left hand: Fisher Test, N = 29 df = 1, p>0.1) could be evidenced for any of the infant groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Percentage of preterms, early-terms and full-terms, as well as adults reacting respectively to the tactile and sham stimulations.

Statistical analyses: McNemar’s tests: * P<0.05, ** P<0.01; Fisher tests, # # P<0.01; # # # P<0.001.

https://doi.org/10.1371/journal.pone.0229270.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Responses of infants (pre-term, early-term and full-term) and adults to both types (tactile versus sham) of stimulations.

https://doi.org/10.1371/journal.pone.0229270.t002

There were clear differences between the three groups of infants as most of the responses to the tactile stimulation were presented by the preterm infants (Table 2, Fig 1). Indeed, almost all pre-term infants (9/10) reacted to the tactile stimulation, whereas only two showed slight behavioral changes during the sham stimulation (McNemar’s test, N = 10, X² = 7, df = 1, p = 0.008). In addition, they reacted more strongly to the tactile stimulation (intensity of response: X±ES = 2.20±0.53) than to the sham one (X±ES = 0.20±0.13; Wilcoxon test: N = 10, p = 0.013) (Table 2, Fig 2). Results were more mitigated in full-term infants as even if seven of the ten full-term newborns reacted to the tactile stimulation, four reacted also to the sham one (N = 10, X² = 1.3, df = 1, P = 1). Early-term infants appeared in overall poorly reactive: the behavior of only two out of the nine early-term infants changed following the tactile stimulation and the sham test (N = 9, X² = 0, df = 1, P = 1). Besides, both early-term (Wilcoxon test: N = 9, p = 1) and full-term (N = 10, p = 0.135) infants did not exhibit a significant increased reaction to the tactile stimulation as compared to the sham one (Table 2, Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Intensity of the responses of infants (pre-term, early-term and full-term) to both types (sham versus tactile) of stimulations.

Statistical analyses: Sham versus Tactile: Wicoxon test: * p<0.05, NS p>0.1; Comparison of infant groups: Kruskal-Wallis and subsequent Mann-Whitney U-test: * p<0.05, *** p<0.005.

https://doi.org/10.1371/journal.pone.0229270.g002

These differences among the three infant groups were further evidenced when comparing the intensity of reactions to both types of stimulation: while no difference was evidenced between pre-term, early-term and full-term infants for the sham stimulation (Kruskal-Wallis: X² = 1.55, p = 0.46), clear differences were highlighted for the tactile stimulation (X² = 9.14, p = 0.01), with both pre-term and full-term infants showing increased reactions as compared to their early-term counterparts (Fig 2).

Discussion

The results of this pilot study revealed high differences in terms of tactile sensitivity in preterm, early-term and full-term infants at term-equivalent age. Thus, premature infants showed a higher tactile sensitivity as compared to infants born near or at term, confirming earlier reports [19],[30]. Thus, less full-term infants reacted to the very light mechanical stimulation, while almost none of the early-term infants did. This study highlights for the first time that early-term births could be associated also with altered sensory processing despite the fact that those early-term neonates had a weight and Apgar scores at birth in the normal range. In accordance with our hypothesis, adults showed no or almost none evidence of perception of this stimulation.

There are two non-exclusive hypotheses about the processes involved in the differences observed between the groups of newborns:

1) The higher sensitivity observed in the preterm infants could be related to postnatal experiences. At the time of testing, the infants had already spent a long time (from 20 to 88 days) in the neonatal care unit (NICU) where they were submitted to numerous and various procedures, corresponding to a wide range of cutaneous sensory stimuli, either light (e.g. diaper changes) or strong/noxious (e.g. venipuncture, intubation) [48], [49], Previous works have in particular highlighted that the degree of reaction to harmful treatments in preterm infants was related to the time spent in hospital after birth and the number of experienced noxious procedures [50], [51]. As tactile stimulation of preterm infants appears to induce sensitization rather than habituation [52], part of our results could be explained by the accumulation of aversive postnatal experiences. It has also been argued that the developing organisms’ sensory system is adapted to the precise stage of their development [53]. Young preterm infants are still at a “fetal” stage when they are supposed to be still stimulated by light tactile stimulations from the womb walls. This could explain why the softer stimulation of kangaroo care (consisting in a skin-to-skin contact between mother and infant which has to start as soon as possible and be maintained as long as possible) can yield better outcomes than a non-“calibrated” massage [54]. The external world, especially neonatal intensive care units (NICU) where preterm newborns are kept, may also over-stimulate their “prenatal” sensory system [55].

2) Preterm infants may also show a differential maturation due to an earlier stage of development of the sensory system at birth. At the skin level, preterm newborns’ epidermal barrier is thinner, more permeable to exogenous materials and more prone to damage than in full term newborns [56] and the density of skin mechanoreceptors maybe higher in preterm newborns (mechanoreceptor density at the 20^th week of gestation is known to be greater than in adult [8], [9], [57]). However, central processes are certainly involved too as there are cortical responses to a light tactile stimulation in preterms (aged of 36 gestational weeks) [58] and the volume of grey matter shows a clear linear increase until 40 weeks of gestation [59], [60] with a critical period of strong brain restructuration between 35–37 weeks. According to Fabrizi et al. [40], preterm infants, who may have lacked this particular stage of brain maturation, react similarly to nociceptive versus non-nociceptive stimulations. Intriguingly, the early-term newborns included in this study differed from the other groups. Their status has been recently revaluated [34], [40] and it is considered that their birth at the precise time of this critical period may explain some of their particularities. Further studies should be performed on a larger sample to confirm the hyporeactivity of early-term newborns, and if so, to understand why they are less sensitive or responsive. The mechanisms underlying such effects are to date largely unknown: only a few hypothesis have been put forward including brain development alterations as mentioned above, physiological immaturity or lower glucose levels [36].

Both maturational and experiential factors are certainly at stake as specialization of neural circuits is also dependent on experience [61]. Further studies, especially longitudinal studies in preterm and early-term infants over the first weeks of life, are required to disentangle the roles of maturation and postnatal experiences respectively. In order to better understand the extent of variations in tactile sensitivity of pre-term and early-term infants in comparison to full-terms, it would also be of high interest to investigate reactions to a larger range of tactile stimulations, by using von Frey filaments of increasing density. More studies of the central and peripheral systems involved in tactile sensitivity would also help understanding the processes. But in any case, the higher sensitivity shown by most preterm and some full-term infants must be taken into account in neonatal care. This characteristic may persist until later stages of development as parents report that children born preterm show a long-term tactile over-sensitivity and reluctance to being touched [62].

Very few adults perceived the 0.008 grams stimulation and the commonly accepted threshold is around 0.06 grams [63]. These findings emphasize the different sensory worlds of neonates and how adults may misinterpret a newborns’ non-localized responses to a localized tactile stimulation. This makes it hard for adult caregivers to assess the newborn’s perception, and in turn make an appropriate response that avoids over-sensitizing the extremely sensitive premature newborn. The supposed lack of inhibitory circuits that allow to restrict responses to pertinent stimuli (e.g. not aversive) and localization [19], [27] prevents the transformation from an excessive non-organized answer (non-localized or generalized movement to even light stimulations) to a control organized one, regulating and preventing exaggerated answers to non-pertinent stimulations such as this very light stimulus (for a review see: [64]). It is essential for caretakers to be able to recognize such reactions which differ from the usual well known indicators (faces, crying, physiological changes: [33], [49]).

Overall, not much is known about tactile processing in early development and these results contribute to the understanding of the development and processing of tactile function. They open new lines of thought about tactile sensitivity of newborns and may be of significant importance for obstetricians, midwives and child psychiatrists.

Supporting information

Acknowledgments

We are grateful to the hospital staff and especially Doctor A. de La Pintière, members of the nursing staff, S. Le Gall and I. Paire, NIDCAP program team members S. Bleunven, I. Olivard, B. Kerleroux, N. Ratynski, the research nurse S. Roudaut, for their help during the research. We also thank all the participants and their parents very grateful, for volunteering to be part of this study. We thank also A. Cloarec and C. Symes for correcting the English. The study resulted from collaborations within the interdisciplinary Group of Scientific Interest GIS “Cerveau-Comportement-Société”.