Journal Logo

Research Paper

The cortical response to a noxious procedure changes over time in preterm infants

Bembich, Stefano; Marrazzo, Francesca; Barini, Alice; Ravalico, Paola; Cont, Gabriele; Demarini, Sergio*

Author Information
doi: 10.1097/j.pain.0000000000000605

1. Introduction

The second half of gestation is a period of very rapid development and fundamental neuroplastic changes for the human brain.30 Preterm newborns spend a significant portion of this time in a neonatal intensive care unit (NICU). During their hospital stay, preterm infants undergo many diagnostic and therapeutic procedures, which are associated with varying degrees of invasiveness and, presumably, pain. All this takes place in a period when, physiologically, external stimulations should be minimal40 and programming of stress systems seems to take place.22,53

Newborns exhibit nociceptive event-related behavioural and physiological reactions after a stimulus felt as painful by adults. Additionally, near-infrared spectroscopy (NIRS) studies have consistently found activation of the neonatal somatosensory,7,8,46 motor,8 and frontal36,37 cortex, after a noxious stimulation, in both term and preterm newborns. A functional magnetic resonance imaging study also showed overlapping brain activity between adult and infant, in association with a nociceptive stimulus.19 Such findings support the hypothesis that nociception may be a conscious experience in neonates and may include both sensory-discriminative and affective components, as in adults.

Studies in both animal and human neonates showed that repeated noxious stimulation may be associated with short-term sensitization to successive nociceptive stimuli.17,29 Repeated noxious stimulation is linked to both hypersensitivity, as shown by spinal withdrawal reflexes,5,6,13 and hyperalgesia, as evidenced by behavioural responses (grimacing/crying).50 In research using electroencephalography (EEG), preterm infants, at term equivalent age, showed a higher nociceptive event-related neuronal activity in the brain, when compared with healthy age-matched term infants.48

Repeated noxious stimulation in the neonatal period may be associated with habituation16 and with long-term mechanical and thermal hyposensitivity.42,54 Moreover, the spinal withdrawal reflex to repeated mechanical stimulation decreases with increasing age in the neonatal period.5 Therefore, stressful stimulations seems to exert some influence on central nervous system development, possibly inducing modified neuroplasticity.40

Pain-related stress in the neonatal period may also have long-term consequences on the preterm brain, as repetitive noxious stimulations may be linked to altered brain development.40 A greater number of stressful procedures are associated with a reduced width in the frontal cortex and parietal cortex and a poorer functional connectivity between temporal lobes, at term equivalent age.49 Moreover, the number of noxious procedures (heel prick for blood sampling) seems to be related to reduced white matter and subcortical gray matter volume.11 Additionally, neonatal pain has been associated with: (1) a slower development of the corticospinal tract55; (2) reduced cortical thickness in pre/post central and frontal brain regions, in 7-year-old children39; (3) altered functional brain activity, as assessed by EEG in school-age children.14

The aim of this study was to monitor early changes over time in cortical activation associated with a standard noxious stimulus (heel prick) in preterm infants. We tested the hypothesis that, during NICU stay, both advancing postmenstrual age (PMA) and increasing number of heel pricks would affect the magnitude of nociceptive event-related cortical activation.

2. Materials and methods

2.1. Participants

The study was performed at the Neonatal Intensive Care Unit of the Institute for Maternal and Child Health IRCCS “Burlo Garofolo” (Trieste, Italy). We recruited 16 preterm newborns, 10 females and 6 males, with a gestational age between 27 and 32 weeks. All infants were clinically stable and had a normal brain ultrasound. No participant required respiratory support at the time of NIRS assessment, and none was given sedatives or analgesics. All infants were on oral caffeine and vitamin D. More detailed clinical data are reported in Tables 1 and 2. The Independent Committee for Bioethics of our hospital approved the research, in accordance with the Declaration of the World Medical Association. Informed consent was previously obtained from parents, after a full technical and procedural explanation regarding the study.

Table 1.:
Clinical data at observation.
Table 2.:
Frequency of potentially painful procedures,10 other than heel prick, performed on participants before entering (postmenstrual age range: 27-32 weeks) or during the study (postmenstrual age range: 29-36 weeks).

2.2. Multichannel near-infrared spectroscopy recording

Multichannel NIRS allows noninvasive measurement of cerebral hemodynamics by continuously monitoring changes in oxyhaemoglobin (HbO2), deoxyhaemoglobin, and total haemoglobin.26 HbO2 is an estimate of cerebral blood flow and, when increased, it is an indirect measure of brain activation.33 This is based on neurovascular coupling: increased regional neuronal activity increases oxygen consumption and is accompanied by local vasodilatation, increased blood flow, and oxygenation. Each NIRS channel comprises a pair of source and detection fibres (or optodes), which measure the absorption of near-infrared (NIR) light by oxygenated and deoxygenated haemoglobin at the surface of the cerebral cortex.52 By multiplying the number of channels (or registering sites), multichannel NIRS represents an improvement in spatial resolution, compared with 1- or 2-channel NIRS devices, while maintaining a very high temporal resolution.32

We used the Hitachi ETG-100 OT device (Hitachi Medical Corporation, Tokyo, Japan), which records simultaneously from 24 channels arising by 18 light emitters and detectors of 1 mm in diameter, placed on the scalp. The device is a continuous wave NIRS system (CW-NIRS) that does not allow absolute measurements of cerebral oxygenation and hemodynamics.41 It emits NIR light at 2 wavelengths, 780 and 830 nm, and the reflected light is sampled every 100 milliseconds. Changes in the concentration of haemoglobins are estimated using the modified Beer–Lambert law.52 Hemodynamic variations are reported as mM·mm, which is the product of the haemoglobin concentration changes expressed in millimolar and the optical path length expressed in millimetres.

In this study, the optodes were arranged in two 3 × 3 patterns and were positioned on the left and right sides of the head by means of a custom-made fiber holder, providing 12 channels per hemisphere. The holder is made of soft material (neoprene) and is equipped with elastic bands aimed to fix the holder on the head. The distance between adjacent emitters and detectors was set at 2 cm, to allow optimal contact of all the optodes with the scalp even in the smallest heads. This distance ensures good sensitivity to hemodynamic responses in the newborn,51 and the total power output per illuminating fiber was maintained at 0.7 mW/mmq. To have a reliable localization of registered cortical hemodynamics between participants, fiber holders were placed with reference to the international 10–20 EEG placement system.25 In particular, the central optode of the inferior channel row of each holder was placed over T3 on the left side and over T4 on the right side, maintaining in both cases the central channel column of holders on the virtual line joining T3 with C3 (left) and T4 with C4 (right). This fibers' placement allowed us to assess hemodynamic changes mainly in the parietal, temporal, and posterior frontal areas of each hemisphere (Fig. 1).

Figure 1.:
Schematic representation of optical fibers placed on both sides of the newborn's head (red dots: light emitters; blue dots: light detectors; numbered squares: channels). Ten to twenty electroencephalography system reference points are also indicated.

2.3. Procedure

To evaluate possible change in the magnitude of nociceptive event-related cortical activity with advancing PMA, we assessed cortical activation during a heel-prick procedure once a week in all babies. All the neonates were tested in their incubator or crib. Each infant had at least 3 consecutive recordings. Blood tests were always requested by the attending physician for medical reasons (not for research purposes). Swaddling, as nonpharmacologic analgesia, was used in all infants, beginning 2 minutes before the heel-prick procedure.12

The NIRS probe was positioned over the infant's scalp. The adequacy of the NIRS signal was tested once the infant appeared stable. There was a waiting period for the newborn to adapt to the equipment and, finally, the adequacy of the NIR signal was checked. Then, the baby was swaddled and, after 2 minutes, the heel-prick procedure began. The nurse held the infant's foot for the duration of the study. First, baseline data were collected for 10 seconds, then a nonnoxious tactile (control) stimulus was performed, and, after a 25-second waiting period, the noxious stimulus was performed.

The nonnoxious stimulation was a tactile heel-disinfectant procedure (the heel was gently wiped with a disinfectant cloth for 5 seconds). The noxious stimulus was a clinically required heel prick. The heel prick was performed using a semiautomated lancet with a scalpel depth of 0.85 m (Quikheel Preemie Lancet, Becton Dickinson AG, Basel, Switzerland). The heel was squeezed neither during the heel-prick stimulation nor during the 25 seconds after stimulation.31

The times when nonnoxious and noxious stimuli had to be performed were signalled to the nurse by the investigator collecting NIRS data, who manually marked such events. Near-infrared spectroscopy optodes were removed from the infant once the study was completed.

2.4. Data analysis

We analyzed increased HbO2, as an estimate of cortical activation, during the heel-prick procedure. This parameter was chosen because it is best detected by the NIRS device we used,18 and it allowed us to assess significant neonatal nociceptive event-related cortical activity in previous studies.8–10 Possible components of HbO2 detection related to slow cerebral blood flow fluctuations or to heartbeat noise were removed by bandpass filtering NIR signals between 0.02 and 1 Hz.10 To prevent movement artifacts, a filter was used to remove detections with rapid changes in haemoglobin concentration (signal variations >0.1 mM·mm over 2 consecutive samples).10 We also visually inspected the signals recorded in each channel for all participants, to detect low signal-to-noise ratios due to bad transmission of NIR light, eg, due to the loss of contact of the optodes with the scalp. Channels with a suboptimal signal were excluded from statistical analysis.

Significantly activated channels during heel prick were identified by the Student t test. Because there was only one direction of interest to test statistical significance (HbO2 increase), a 1-tailed test was chosen.10 For each participant's NIRS detection and for every channel, a baseline was calculated as the mean change in HbO2 concentration over the 10 seconds before disinfection. The hemodynamic response associated with noxious stimulation was calculated as the mean change in HbO2 concentration over the 25 seconds after the actual prick in the heel. All the data collected in the whole sample, throughout the entire study (65 heel pricks), were included in statistical analysis. To identify nociceptive event-related cortical activity, we performed 1-tailed paired t tests comparing, for every channel, HbO2 mean concentration changes during baseline and during the hemodynamic response after the noxious stimulation. We also analyzed the hemodynamic response associated with heel disinfection (tactile stimulation), as a nonnoxious stimulus to which compare results obtained with heel prick. The hemodynamic response associated with the nonnoxious stimulus was again calculated, for each participant's NIRS detection and for every channel, as the mean change in HbO2 concentration over the 25 seconds after heel disinfection by a soft cotton pad. To identify a possible tactile-related cortical activity, we separately performed 1-tailed paired t tests comparing, for every channel, HbO2 mean concentration changes during baseline and during the hemodynamic response after heel disinfection. A false discovery rate (FDR) approach was used with statistical analysis of both stimulations to control type I error in multiple testing situations (q = 0.05).44

Then, by multiple regression analysis, we studied the effect of: (1) the increasing PMA (cortical maturation) and (2) the increasing number of heel pricks throughout NICU stay (experience with a specific noxious stimulation) at each NIRS assessment, as independent variables (or regressors), on HbO2 weekly variations in activated cortical areas, as the dependent variable. Statistical analyses were performed by SPSS version 13.0 for Windows (SPSS Inc, Chicago, IL).

3. Results

3.1. Spatial distribution of cortical activation after a noxious heel prick

In association with a nociceptive stimulus, we observed a significant bilateral activation of the newborn's frontal cortex. Four channels passed the FDR threshold (P < FDR 0.05). Two were located on the left posterior frontal cortex (channel 3 [t(64) = −3.012; P = 0.002] and channel 8 [t(63) = −3.273; P = 0.001]), and the other 2 were located on the right posterior frontal cortex (channel 17 [t(64) = −2.619; P = 0.0055] and channel 22 [t(62) = −3.056; P = 0.0015]) (Fig. 2). Supplemental Figure 1 describes further paired t test analysis, specifically focused on the somatosensory cortex (see Figure, Supplemental Digital Content 1, available online at

Figure 2.:
Haemoglobin time course variation of channels significantly activated in premature infants, in association with repeated heel pricks, and their localization on the cortex (red: oxyhaemoglobin trend; blue: deoxyhaemoglobin trend). Time, reported in seconds, is on the x-axis (range: 0–25 seconds; the green vertical line shows the onset of the prick). Haemoglobin (Hb) variation, reported as millimolar per millimeter (mM·mm) of the optical path length, is on the y-axis (range: −0.5/+0.5 mM·mm).

As a post hoc analysis, we also tested for the laterality of newborn's cortical response, with respect to the left (N = 27) or right (N = 38) side pricked. Contralateral stimulation to the left frontal cortex significantly activated both channel 3 (t(37) = −3.569; P = 0.0005) and channel 8 (t(37) = −2.675; P = 0.0055). Ipsilateral stimulation to the left frontal cortex significantly activated channel 8 (t(25) = −1.912; P = 0.034), but not channel 3 (t(26) = −0.671; P = 0.25). Contralateral stimulation to the right frontal cortex significantly activated both channel 17 (t(26) = −2.009; P = 0.028) and channel 22 (t(25) = −2.911; P = 0.0035). Ipsilateral stimulation to the right frontal cortex resulted in an activation around significance in channel 17 (t(37) = −1.673; P = 0.05), whereas channel 22 did not show any significant activation (t(36) = −1.366; P = 0.09).

Figure 3 shows an example of NIRS signal collected across channels 3, 8, 17, and 22 in an individual infant, during a heel prick, at 29 and 35 PMA weeks. Supplemental Figure 2 describes paired t test analysis performed on total haemoglobin and deoxyhaemeglobin (see Figure, Supplemental Digital Content 2, available online at Supplemental Figure 3 reports within subjects analysis performed on HbO2 mean concentration in channels 3, 8, 17, and 22, during heel prick, over the first 3 monitored noxious events (see Figure, Supplemental Digital Content 3, available online at

Figure 3.:
Example of activated channels in the same infant at 29 postmenstrual age (PMA) weeks and at 35 PMA weeks (red: oxyhaemoglobin trend; blue: deoxyhaemoglobin trend). A cortical activation, contralateral to the side of the prick (right heel), can be seen already at 29 PMA weeks, but it is more evident at 35 weeks. At 29 PMA weeks, a cortical activation, ipsilateral to the side of heel prick, can also be observed. Cortical activation in channel 8 at 35 PMA weeks seems dampened, if compared with the same channel at 29 PMA weeks. Time, reported in seconds, is on the x-axis (range: 0-25 seconds; the green vertical line shows the onset of the prick). Haemoglobin (Hb) variation, reported as millimolar per millimeter (mM·mm) of the optical path length, is on the y-axis (range: −0.5/+0.5 mM·mm).

3.2. Nonnoxious tactile stimulation

In association with nonnoxious stimulation (heel disinfection), no channel survived the FDR threshold (P < FDR 0.05). Thus, no significant cortical activation was observed after performing the nonnoxious stimulation.

3.3. Effect of increasing postmenstrual age and repeated heel pricks on cortical activation associated with noxious stimulation, during neonatal intensive care unit stay

We performed multiple regression analysis, to study the effect of both PMA and the number of heel pricks on HbO2 weekly variations in the 4 posterior frontal areas significantly activated by the noxious stimulation. The results were as follows: in channel 3 (left hemisphere): R2 = 0.03 (F(2,62) = 0.789; P = 0.46); in channel 8 (left hemisphere): R2 = 0.14 (F(2,61) = 4.899; P = 0.011); in channel 17 (right hemisphere): R2 = 0.19 (F(2,62) = 7.066; P = 0.002); in channel 22 (right hemisphere): R2 = 0.02 (F(2,60) = 0.747; P = 0.48).

With regard to the effect of each single regressor, we observed a significant and positive effect of PMA on HbO2 weekly variations in the left posterior frontal cortex (channel 8 [t(62) = 2.898; P = 0.005]) and in the right posterior frontal cortex (channel 17 [t(63) = 2.725; P = 0.008]). In association with a noxious stimulation, cortical activation progressively increased in such regions, between the 29th and 36th weeks (Figs. 4 and 5). No effects of PMA were observed on HbO2 weekly variations in channel 3 (left posterior frontal cortex) and channel 22 (right posterior frontal cortex).

Figure 4.:
Scatter plot showing the significant positive effect of postmenstrual age on oxyhaemoglobin variation observed in channel 8/left posterior frontal cortex (β = 0.498; P = 0.006). Postmenstrual age (PMA) is on the x-axis (range: 29–36 weeks). Oxyhaemoglobin variations (HbO2), reported as millimolar per millimeter (mM·mm) of the optical path length, is on the y-axis (range: −0.5/+1.0 mM·mm).
Figure 5.:
Scatter plot showing the significant positive effect of postmenstrual age on oxyhaemoglobin variation observed in channel 17/right posterior frontal cortex (β = 0.383; P = 0.007). Postmenstrual age (PMA) is on the x-axis (range: 29–36 weeks). Oxyhaemoglobin variations (HbO2), reported as millimolar per millimeter (mM·mm) of the optical path length, is on the y-axis (range: −0.5/+1.0 mM·mm).

With regard to the effect of the number of heel pricks, we observed a significant and negative effect of an increasing number of heel pricks on HbO2 weekly variations, again in the left posterior frontal cortex (channel 8 [t(62) = −2.591; P = 0.012]) and in the right posterior frontal cortex channel (17 [t(63) = −3.661; P = 0.001]). In association with a heel prick, cortical activation progressively decreased in such regions, as the number of heel pricks increased during NICU stay (Figs. 6 and 7). The number of heel pricks had no effect on HbO2 weekly variations in channel 3 and channel 22.

Figure 6.:
Scatter plot showing the significant negative effect of postmenstrual age on oxyhaemoglobin variation observed in channel 8/left posterior frontal cortex (β = −0.354; P = 0.017). The number of previous heel pricks is on the x-axis (range: 0–50 events). Oxyhaemoglobin variations (HbO2), reported as millimolar per millimeter (mM·mm) of the optical path length, is on the y-axis (range: −0.5/+1.0 mM·mm).
Figure 7.:
Scatter plot showing the significant negative effect of postmenstrual age on oxyhaemoglobin variation observed in channel 17/right posterior frontal cortex (β = −0.509; P = 0.0005). The number of previous heel pricks is on the x-axis (range: 0–50 weeks). Oxyhaemoglobin variations (HbO2), reported as millimolar per millimeter (mM·mm) of the optical path length, is on the y-axis (range: −0.5/+1.0 mM·mm).

We performed post hoc analysis to evaluate the effect of increasing PMA and increasing number of heel pricks on HbO2 weekly variations in the overall posterior left and right frontal cortex. Such variations were estimated by averaging, respectively, HbO2 weekly variations in channels 3 and 8 (left side) and in channels 17 and 22 (right side). The effect on the overall posterior frontal areas was similar to that on single channels. The multiple regression analysis results on HbO2 weekly variations were as follows: R2 = 0.13 (F(2,61) = 4.645; P = 0.013) in the left posterior frontal cortex and R2 = 0.14 (F(2,60) = 4.865; P = 0.011) in the right posterior frontal cortex. Increasing PMA showed a significant, positive effect on HbO2 weekly variations in both the left (t(62) = 2.910; P = 0.005) and the right posterior frontal cortex (t(61) = 2.207; P = 0.031). The increasing number of heel pricks showed a significant, negative effect on HbO2 weekly variations in both the left (t(62) = −2.368; P = 0.021) and the right posterior frontal cortex (t(61) = −3.046; P = 0.003). Table 3 summarizes the results of all the regression analysis performed in this study.

Table 3.:
Multiple regression analysis results.*

4. Discussion

Our first finding was that, after a noxious stimulus, preterm newborns showed an activation of the posterior frontal cortex, involving both superior and inferior regions. Such activation seemed to be contralateral to the side of the heel prick, confirming previous studies performed in both newborns36,37 and adults.24 We also found a weaker ipsilateral cortical response, particularly in the left frontal cortex. Previous studies on premature newborns found both contralateral46 and bilateral7 cortical response to a noxious stimulation. Such discrepancies could be due to methodological differences among studies.45 An alternative explanation might be an ongoing maturational process of nociceptive pathways lateralization in the preterm brain. On the contrary, we did not observe significant cortical responses to a nonnoxious stimulation (heel disinfection). Bartocci et al.7 used the disinfection of the hand as a nonnoxious stimulation and found an activation of the somatosensory cortex. However, the intensity of such nonnoxious stimulation cannot be standardized between experiments. Therefore, the 2 studies might have differed in the intensity of the nonnoxious stimulation.

Most neonatal studies, evaluating cortical responses to noxious stimulations, found an activation of the somatosensory cortex (eg, Refs. 7,46). In our study, such activation was found only when channels in the somatosensory cortex were averaged together (see Supplementary Digital Content 1, available online at This result could be due to different NIRS devices (2 channel vs multichannel NIRS). Alternatively, one could speculate on a different maturational timing of cortical areas: the posterior frontal cortex may exhibit a more definite and localized processing of nociceptive stimuli earlier than the somatosensory cortex. More research is needed on this topic.

Our second finding was that, in the same frontal areas, we detected a progressively stronger nociceptive event-related activation with increasing PMA, between the 29th and 36th week of gestation. Similar findings were reported in a cross-sectional study, performed by 2-channel NIRS.46 In that study, cortical activity related to nociceptive events increased with PMA, whereas cortical response latency decreased. A similar correlation was reported between latency to facial pain expression and PMA: the former decreases as the latter increases.47 Unlike previous reports, we prospectively studied the variation of cortical responses to a noxious stimulation in the same infants over time. We speculate that the changing response might be due to endogenous cortical maturation processes, such as growth of cortico-cortical connecting fibers and increasing cerebral gyration.30

Our third finding was that, in the same cortical regions, the magnitude of nociceptive event-related cortical activation decreased, as the number of noxious events increased. Differing from previous studies, we observed this process longitudinally from its onset, during the recurrent noxious procedures experienced in the NICU. Previous studies on pain in preterm newborns during NICU stay found evidence of both increased sensitivity and decreased reactions to subsequent noxious stimulations.

With regard to increased sensitivity to nociceptive events, repeated noxious stimulation in newborns has been linked to both hypersensitivity5,6,13 and hyperalgesia.50 Preterm infants born before 28 weeks of gestation show increased autonomic and behavioural responses to noxious procedures with increasing postnatal age.38 From a neurophysiological standpoint, preterm infants, assessed at term equivalent age by EEG, show a higher nociceptive event-related neuronal activity, when compared with term infants.48 The authors speculated that preterm infants may exhibit experience-dependent changes in cortical nociception processing. An altered response may be not limited to hospital stay because long-lasting phenomena have been reported. A higher sensitivity to mechanical touch on the heel injured by previous heel pricks was observed in premature infants throughout the first year of life.1 Very preterm newborns showed a positive correlation between the length of NICU stay and pain expression intensity during immunizations, in the first year of life.2 Among school-aged children, those who experienced a NICU stay demonstrated a higher perceptual sensitization to heat stimulation than controls.23

With regard to decreased reactions to noxious stimulations, after 4 weeks in the NICU, preterm newborns showed lessened behavioral responses to subsequent noxious procedures, such as heel pricks. Pain response seemed to be negatively correlated with the number of previous noxious procedures.27 Cardiac autonomic reactivity, during a finger lance blood sampling, was less intense in 4 month-old NICU graduates.35 At 18 months of age, a lower sensitivity to minor traumas was reported in former preterm infants with a birthweight <1000 g.21 Eleven-year-old children, born at 26 gestational weeks or less, showed a decreased sensitivity to thermal, but not to mechanical stimuli.54

In the neonatal rat brain, repeated exposure to noxious events may increase apoptosis and cell death.3,15 In human neonates, at term equivalent age, stressful procedures have been associated with a reduced width in the frontal and parietal cortex.49 We speculate that such structural effects on the neonatal cortex may be associated with the decrease we observed in nociceptive event–related activation of the frontal cortex, as the number of heel pricks increased. Whether decreased cortical responses to noxious stimuli are entirely due to the number of procedures in the NICU or due to other factors (eg, different brain development in preterm infants) cannot be established at present.

Alternatively, our findings may indicate that the premature brain is not only passively subjected to stimulations, but may also try to adapt to them. A higher cumulative procedural pain exposure is associated with lower cortisol responses to stress at 32 PMA weeks.20 We may speculate that the preterm brain may also build adaptive responses to repeated stressful events, as already observed in the stress hormone system.20

In interpreting our results, one should also consider that not everything that happens in the NICU is painful. Several attempts have been made to help infants to cope with a stressful environment. Among those, Kangaroo Mother Care (KMC) is routinely practiced in our NICU.4 Kangaroo Mother Care has been showed to decrease infant stress responses during NICU stay34 and to promote functional development, as assessed by EEG, in preterm infants at term equivalent age.28 Additionally, as assessed by magnetic resonance imaging in adolescence, KMC seems to promote motor cortex maturation in former preterm infants.43 Taking into account these results, the nociceptive response we measured might have been modulated by the counteracting effect of KMC.

Scatter plots showing the effect of PMA and of previous heel pricks on HbO2 variation in the posterior frontal cortex seem to indicate a differential effect. The positive effect of PMA on cortical activation seems to be greater in the left frontal cortex, whereas the negative effect of previous heel pricks on cortical activation seems to be greater in the right frontal cortex. Thus, PMA and previous heel pricks may differentially affect cortical areas processing nociception in the preterm brain. However, further research is needed to confirm this finding and to elucidate its possible clinical significance.

Our study has some limitations. (1) NIRS, the functional neuroimaging technique we used, can only measure relative changes in HbO2 at the cortical surface,52 not absolute values. Subcortical structures, which play an important role in pain processing in the adult,24 cannot be studied. (2) NIRS has a lower spatial resolution than that of functional magnetic resonance imaging. However, it can be used in the real NICU world, studying infants in their cribs and/or incubators. (3) Our sample size was small and our results should be replicated in larger studies. However, we prospectively studied the same preterm infants and our analysis included 65 distinct events (heel pricks). (4) The heel-prick event was marked manually during NIRS data collection. This may have introduced the risk of a human error, increasing the signal variability in terms of latency to event-related cortical response. However, although EEG detects an electrical cortical response that occurs in milliseconds, NIRS measures a cortical hemodynamic response that takes place over several seconds after stimulation, especially in newborns.33 Such a physiologically long-time course should have limited the variability in the latency to cortical response, due to manual event marking. (5) Our infants experienced noxious procedures other than heel prick, both before and during the study period. The procedure distribution differed over time, due to the evolution of newborns' clinical conditions. Although experience with other noxious procedures may have influenced the newborn's cortical response, we speculate that the localized activation we found in the cortex should be mainly associated with a localized stimulation (the heel prick).

To conclude, we found that, in preterm infants, repeated noxious, and presumably painful, stimulations were associated with a significant activation of the posterior frontal cortex. The magnitude of such cortical activation increased with advancing PMA, between 29 and 36 weeks. In the same regions, however, nociceptive event-related cortical activation decreased as the number of noxious events increased. We speculate that these findings may be related to both endogenous cortical maturation and neuroplasticity processes, such as synaptogenesis, synaptic pruning and ongoing apoptosis for fine-tuning cortical connections, induced by experience with a repeated noxious event in the NICU. Given the possible long-term consequences of such minor painful procedures, interventions aimed at protecting the developing brain, such as effective analgesia, should be actively promoted.

Conflict of interest statement

The authors have no conflicts of interest to declare.

This study was presented, in part, at the 2011 Pediatric Academic Societies' Annual Meeting (Denver, April 30-May 3, 2011), and then published, as an abstract, in the congress proceedings.

This study has been financed by the grant No. 50/11 of the Institute for Maternal and Child Health IRCCS “Burlo Garofolo” (Trieste, Italy).

Appendix A Supplemental Digital Content

Supplemental Digital Content associated with this article can be found online at


[1]. Abdulkader HM, Freer Y, Garry EM, Fleetwood-Walker SM, McIntosh N. Prematurity and neonatal noxious events exert lasting effects on infant pain behaviour. Early Hum Dev 2008;84:351–5.
[2]. Allegaert K, Devlieger H, Bulckaert D, Naulaers G, Casaer P, Tibboel D. Variability in pain expression characteristics in former preterm infants. J Perinat Med 2005;33:442–8.
[3]. Anand KJ, Garg S, Rovnaghi CR, Narsinghani U, Bhutta AT, Hall RW. Ketamine reduces the cell death following inflammatory pain in newborn rat brain. Pediatr Res 2007;62:283–90.
[4]. Anderson GC, Marks EA, Wahlberg V. Kangaroo care for premature infants. Am J Nurs 1986;86:807–9.
[5]. Andrews K, Fitzgerald M. Cutaneous flexion reflex in human neonates: a quantitative study of threshold and stimulus-response characteristics after single and repeated stimuli. Dev Med Child Neurol 1999;41:696–703.
[6]. Andrews K, Fitzgerald M. Wound sensitivity as a measure of analgesic effects following surgery in human neonates and infants. PAIN 2002;99:185–95.
[7]. Bartocci M, Bergqvist LL, Lagercrantz H, Anand KJS. Pain activates cortical areas in the preterm newborn brain. PAIN 2006;122:109–17.
[8]. Bembich S, Brovedani P, Cont G, Travan L, Grassi V, Demarini S. Pain activates a defined area of the somatosensory and motor cortex in newborn infants. Acta Paediatr 2015;104:e530–3.
[9]. Bembich S, Cont G, Baldassi G, Bua J, Demarini S. Maternal holding vs oral glucose as non-pharmacologic analgesia in newborns: a functional neuroimaging study. JAMA Pediatr 2015;169:284–5.
[10]. Bembich S, Davanzo R, Brovedani P, Clarici A, Massaccesi S, Demarini S. Functional neuroimaging of breastfeeding analgesia by multichannel near-infrared spectroscopy. Neonatology 2013;104:255–9.
[11]. Brummelte S, Grunau RE, Chau V, Poskitt KJ, Brant R, Vinall J, Gover A, Synnes AR, Miller SP. Procedural pain and brain development in premature newborns. Ann Neurol 2012;71:385–96.
[12]. Cignacco E, Hamers JPH, Stoffel L, van Lingen RA, Gesser P, McDougall J, Nelle M. The efficacy of non-pharmacological interventions in the management of procedural pain in preterm and term neonates. A systematic review. Eur J Pain 2007;11:139–52.
[13]. Cornelissen L, Fabrizi L, Patten D, Worley A, Meek J, Boyd S, Slater R, Fitzgerald M. Postnatal temporal, spatial and modality tuning of nociceptive cutaneous flexion reflexes in human infants. PLoS One 2013;8:e76470.
[14]. Doesburg SM, Chau CM, Cheung TP, Moiseev A, Ribary U, Herdman AT, Miller SP, Cepeda IL, Synnes A, Grunau RE. Neonatal pain-related stress, functional cortical activity and visual-perceptual abilities in school-age children born at extremely low gestational age. PAIN 2013;154:1946–52.
[15]. Dührsen L, Simons SH, Dzietko M, Genz K, Benedix I, Boos V, Sifringer M, Tibboel D, Felderhoff-Mueser U. Effects of repetitive exposure to pain and morphine treatment on the neonatal rat brain. Neonatology 2013;103:35–43.
[16]. Fitzgerald M. Pain in infancy: some unanswered questions. Pain Rev 1995;2:77–91.
[17]. Fitzgerald M, Millard C, McIntosh N. Cutaneous hypersensitivity following peripheral tissue damage in newborn infants and its reversal with topical anesthesia. PAIN 1989;39:31–6.
[18]. Gervain J, Mehler J, Werker JF, Nelson CA, Csibra G, Lloyd-Fox S, Shukla M, Aslin RN. Near-infrared spectroscopy: a report from the McDonnell infant methodology consortium. Dev Cogn Neurosci 2011;1:22–46.
[19]. Goksan S, Hartley C, Emery F, Cockrill N, Poorun R, Moultrie F, Rogers R, Campbell J, Sanders M, Adams E, Clare S, Jenkinson M, Tracey I, Slater R. fMRI reveals neural activity overlap between adult and infant pain. Elife 2015;4:e06356.
[20]. Grunau RE, Holsti L, Haley DW, Oberlander T, Weinberg J, Solimano A, Whitfield MF, Fitzgerald C, Yu W. Neonatal procedural pain exposure predicts lower cortisol and behavioral reactivity in preterm infants in the NICU. PAIN 2005;113:293–300.
[21]. Grunau RV, Whitfield MF, Petrie JH. Pain sensitivity and temperament in extremely low-birthweight premature toddlers and preterm and full-term controls. PAIN 1994;58:341–6.
[22]. Grunau RE, Whitfield MF, Petrie-Thomas J, Synnes AR, Cepeda IL, Keidar A, Rogers M, Mackay M, Hubber-Richard P, Johannesen D. Neonatal pain, parenting stress and interaction, in relation to cognitive and motor development at 8 and 18 months in preterm infants. PAIN 2009;143:138–46.
[23]. Hermann C, Hohmeister J, Demirakça S, Zohsel K, Flor H. Long-term alteration of pain sensitivity in school-aged children with early pain experiences. PAIN 2006;125:278–85.
[24]. Ingvar M. Pain and functional imaging. Philos Trans R Soc Lond B Biol Sci 1999;354:1347–58.
[25]. Jasper HH. The ten–twenty electrode system of the International Federation. Electroencephalogr Clin Neurophysiol 1958;10:367–80.
[26]. Jöbsis FF. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977;198:1264–7.
[27]. Johnston CC, Stevens BJ. Experience in a neonatal intensive care unit affects pain response. Pediatrics 1996;98:925–30.
[28]. Kaffashi F, Scher MS, Ludington-Hoe S, Loparo KA. An analysis of the kangaroo care intervention using neonatal EEG complexity: a preliminary study. Clin Neurophysiol 2013;124:238–46.
[29]. Knaepen L, Patijn J, van Kleef M, Mulder M, Tibboel D, Joosten EAJ. Neonatal repetitive needle pricking: plasticity of the spinal nociceptive circuit and extended postoperative pain in later life. Dev Neurobiol 2013;73:85–97.
[30]. Kostović I, Jovanov-Milošević N. The development of cerebral connections during the first 20–45 weeks' gestation. Semin Fetal Neonatal Med 2006;11:415–22.
[31]. Lago P, Garetti E, Merazzi D, Pieragostini L, Ancora G, Pirelli A, Bellieni CV. Guidelines for procedural pain in the newborn. Acta Paediatr 2009;98:932–9.
[32]. Maki A, Yamashita Y, Ito Y, Watanabe F, Mayanagi Y, Koizumi H. Spatial and temporal analysis of human motor activity using non-invasive NIR topography. Med Phys 1995;22:1997–2005.
[33]. Meek J. Basic principles of optical imaging and application to the study of infant development. Dev Sci 2002;5:371–80.
[34]. Mörelius E, Theodorsson E, Nelson N. Salivary cortisol and mood and pain profiles during skin-to skin care for an unselected group of mothers and infants in neonatal intensive care. Pediatrics 2005;116:1105–13.
[35]. Oberlander TF, Grunau RE, Whitfield MF, Fitzgerald C, Pitfield S, Saul JP. Biobehavioral pain responses in former extremely low birth weight infants at 4 months' corrected age. Pediatrics 2000;105:e6.
[36]. Ozawa M, Kanda K, Hirata M, Kusakawa I, Suzuki C. Effect of gender and hand laterality on pain processing in human neonates. Early Hum Dev 2011;87:45–8.
[37]. Ozawa M, Kanda K, Hirata M, Kusakawa I, Suzuki C. Influence of repeated painful procedures on prefrontal cortical pain responses in newborns. Acta Paediatr 2011;100:198–203.
[38]. Porter FL, Wolf CM, Miller JP. Procedural pain in newborn infants: the influence of intensity and development. Pediatrics 1999;104:e13–22.
[39]. Ranger M, Chau CM, Garg A, Woodward TS, Beg MF, Bjornson B, Poskitt K, Fitzpatrick K, Synnes AR, Miller SP, Grunau RE. Neonatal pain-related stress predicts cortical thickness at age 7 years in children born very preterm. PLoS One 2013;8:e76702.
[40]. Ranger M, Grunau RE. Early repetitive pain in preterm infants in relation to the developing brain. Pain Manag 2014;4:57–67.
[41]. Roche-Labarbe N, Carp SA, Surova A, Patel M, Boas DA, Grant PE, Franceschini MA. Noninvasive optical measures of CBV, StO2, CBF index, and rCMRO2 in human premature neonates' brains in the first six weeks of life. Hum Brain Mapp 2010;31:341–52.
[42]. Schmelzle-Lubiecki BM, Campbell KAA, Howard RH, Franck L, Fitzgerald M. Long-term consequences of early infant injury and trauma upon somatosensory processing. Eur J Pain 2007;11:799–809.
[43]. Schneider C, Charpak N, Ruiz-Peláez JG, Tessier R. Cerebral motor function in very premature-at-birth adolescents: a brain stimulation exploration of kangaroo mother care effects. Acta Paediatr 2012;101:1045–53.
[44]. Singh AK, Dan I. Explore the false discovery rate in multichannel NIRS. Neuroimage 2006;33:542–9.
[45]. Slater R, Boyd S, Meek J, Fitzgerald M. Cortical pain responses in the infant brain. PAIN 2006;123:332.
[46]. Slater R, Cantarella A, Galella S, Worley A, Boyd S, Meek J, Fitzgerald M. Cortical pain responses in human infants. J Neurosci 2006;26:3662–6.
[47]. Slater R, Cantarella A, Yoxen J, Patten D, Potts H, Meek J, Fitzgerald M. Latency to facial expression change following noxious stimulation in infants is dependent on postmenstrual age. PAIN 2009;146:177–82.
[48]. Slater R, Fabrizi L, Worley A, Meek J, Boyd S, Fitzgerald M. Premature infants display increased noxious-evoked neuronal activity in the brain compared to healthy age-matched term-born infants. Neuroimage 2010;52:583–9.
[49]. Smith GC, Gutovich J, Smyser C, Pineda R, Newnham C, Tjoeng TH, Vavasseur C, Wallendorf M, Neil J, Inder T. Neonatal intensive care unit stress is associated with brain development in preterm infants. Ann Neurol 2011;70:541–9.
[50]. Taddio A, Shah V, Gilbert-MacLeod C, Katz J. Conditioning and hyperalgesia in newborns exposed to repeated heel lances. JAMA 2002;288:857–61.
[51]. Taga G, Homae F, Watanabe H. Effects of source-detector distance of near infrared spectroscopy on the measurement of the cortical hemodynamic response in infants. Neuroimage 2007;38:452–60.
[52]. Villringer A, Chance B. Non-invasive optical spectroscopy and imaging of human brain function. Trends Neurosci 1997;20:435–42.
[53]. Volpe JJ. The encephalopathy of prematurity—brain injury and impaired brain development inextricably intertwined. Semin Pediatr Neurol 2009;16:167–78.
[54]. Walker SM, Franck LS, Fitzgerald M, Myles J, Stocks J, Marlow N. Long-term impact of neonatal intensive care and surgery on somatosensory perception in children born extremely preterm. PAIN 2009;141:79–87.
[55]. Zwicker JG, Grunau RE, Adams E, Chau V, Brant R, Poskitt KJ, Synnes A, Miller SP. Score for neonatal acute physiology-II and neonatal pain predict corticospinal tract development in premature newborns. Pediatr Neurol 2013;48:123–9.e1.

Preterm newborn; Repeated noxious events; Cortical response

Supplemental Digital Content

© 2016 International Association for the Study of Pain