Preterm birth is associated with painful procedures during the neonatal intensive care unit (NICU) stay. Full-term newborns can also experience pain, following surgery. These procedures can have long-lasting consequences. It has been shown that children born preterm show pain responses and cardiac alterations. This study aimed to explore the heart rate reactivity to pain in 107 subjects born either preterm or full-term who were between 7 and 25 years old at testing. We also evaluated the effect of pain experienced at birth, as represented by a longer NICU stay, time under ventilation, and surgery at birth. Participants were asked to immerse their right forearm in 10°C water for 2 minutes. Electrocardiograms were recorded at baseline and during the immersion procedure. Full-term subjects showed a stable increase in heart rate throughout the procedure, whereas preterm ones showed a strong increase at the beginning, which decreased over time. Also, preterm and full-term subjects who experienced pain at birth showed higher resting heart rate, stronger sympathetic activity, and lower cardiac vagal activity. Our study demonstrated a long-term impact of a long NICU stay and surgery at birth on cardiac autonomic activity. This could lead to impaired reactions to pain or stress in later life.
In the last decades, advances in knowledge and improvements in treatment options have increased survival rates for preterm infants [
Only a few studies evaluated heart rate changes in reaction to a painful procedure and have yielded contradictory results [
One hundred seven French-speaking subjects aged between 7 and 25 years (mean:
Medical birth files were revised after experimental pain testing by two of the authors (Mélanie Morin and Louis Couturier) to determine the extent of painful procedures experienced at birth for preterm subjects and the type of surgery for full-term participants. Since the information on the exact number of painful procedures experienced at birth was not always available in the birth files, which resulted in missing data for some subjects, we evaluated the number of days in the NICU and under mechanical ventilation. The correlation was high between the number of days spent in the NICU and the number of painful procedures (heel sticks, tracheal suctioning, venipuncture, etc.) (see Figure
Correlation between the number of days spent in the NICU and the number of painful procedures.
Data was not available for all subjects (
A cold water bath was used as a test stimulus to cause a prolonged pain sensation. This experimental procedure is commonly used in children to assess pain [
Electrocardiogram (ECG) activity was recorded using a standard 3-lead montage sampled at a frequency of 1000 Hz with the Powerlab system and Chart software (AD Instruments, Colorado Springs, CO). ECGs were recorded for two minutes prior to testing (baseline) and during the CPT (immersion). Instantaneous RR intervals were obtained from the ECG waveform with a peak detection algorithm to detect successive R-waves. HR variability analyses in the frequency domain were done subsequently. Fast Fourier transforms were used to calculate the power spectral density of HR oscillations (window length: 1024). The FFT method can provide a noisy estimator of the power spectrum. This is why we used it in combination with Welch’s method. This allows us to achieve a consistent estimator by averaging periodogram from overlapping intervals [
Demographic information was collected prior to testing for each participant. Pediatric ECG electrodes were first installed and baseline ECG was obtained as the participants were asked to relax. Participants were then asked to immerse their right arm in cold water, while ECG was recorded.
Descriptive statistics are presented as mean ± standard deviation (SD). After checking for assumptions of normality and homogeneity of variances, Kruskal-Wallis tests and
There were no significant group differences (preterm versus full-term) in resting heart rate (
Unstructured covariance structure for percentage change in HR.
df |
|
|
|
---|---|---|---|
Time | 1 | 0.15 | 0.697 |
Group | 1 | 1.07 | 0.303 |
Time × group | 1 | 6.92 | 0.010 |
Time × time | 1 | 0.35 | 0.558 |
Time × time × group | 1 | 5.76 | 0.018 |
Age | 1 | 6.10 | 0.015 |
Average percentage change in heart rate from baseline (SEM) during the CPT for 15 s intervals.
There was a different quadratic relationship between the percentage change and time in full-term compared to preterm subjects. The full-term group shows an increase of their percent change in HR over time followed by a decreased (concave function) while the preterm group shows an opposite effect. They show a high increase at the beginning that instantly decreases (convex function).
Further analyses were made to verify whether pain experienced at birth had an impact on mean percentage change in HR. The subjects were divided into 4 groups (full-term, full-term with surgery, low-pain preterm, and high-pain preterm) according to their type of birth and the length of stay in the NICU and number of days under mechanical ventilation, as proxies for the pain suffered at birth. There were significant differences for GA, birth weight, 5 min APGAR, days under mechanical ventilation, and days spent in NICU between the groups (Kruskal-Wallis test: all
Birth characteristics (median (25–75)).
Full-term ( |
Full-term surgery ( |
Low-pain preterm ( |
High-pain preterm ( |
| |
---|---|---|---|---|---|
Age at testing | 15 (11–17) | 15 (13–17) | 14 (10–17) | 16 (10–19) | 0.756 |
Gestational age (weeks) | — | 40 (39-40) | 32 (31–33) | 28 (27–31) | <0.001 |
Birth weight (g) | — | 3380 (2862–3935) | 1602.5 (1405–2081) | 1220 (920–1350) | <0.001 |
5 min APGAR | — | 9 (9-9) | 8 (7–9) | 8 (6–9) | 0.005 |
Days under mechanical ventilation | — | 12 (3–15) | 0 (0-1) | 28 (6–44) | <0.001 |
Days spent in NICU | — | 20 (14–33) | 35 (18–46) | 77 (51–102) | <0.001 |
A
Heart rate variability (mean ± SD).
Full-term ( |
Full-term surgery ( |
Low-pain preterm ( |
High-pain preterm ( |
|
---|---|---|---|---|
Baseline | ||||
Heart rate | 78.2 (12.5) | 84.5 (12.7) | 80.2 (14.3) | 84.5 (15.1) |
LF (nu) | 46.9 (19.9) | 61.3 (17.5) | 49.8 (19.9) | 58.8 (17.4) |
HF (nu) | 43.8 (18.5) | 27.7 (10.8) | 42.5 (19.0) | 32.7 (14.5) |
LF/HF* | 1.2 (0.4–1.8) | 2.0 (1.4–3.2) | 1.4 (1.4–2.6) | 1.9 (1.3–3.8) |
|
||||
% change in HR | 9.1 (8.8) | 2.9 (9.6) | 8.8 (9.9) | 3.9 (9.0) |
ANCOVA results.
HR (bpm) | LF (nu) | HF (nu) | ln (LF/HF ratio) | % change in HR | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
df |
|
|
df |
|
|
df |
|
|
df |
|
|
df |
|
|
|
Type of birth | 1 | 0.135 | 0.715 | 1 | <0.001 | 0.985 | 1 | 0.223 | 0.638 | 1 | 0.131 | 0.718 | 1 | 0.039 | 0.844 |
Pain at birth | 1 | 4.275 |
|
1 | 6.414 |
|
1 | 9.466 |
|
1 | 8.448 |
|
1 | 7.856 |
|
Type of birth × pain | 1 | 0.059 | 0.808 | 1 | 0.492 | 0.485 | 1 | 0.672 | 0.414 | 1 | 0.756 | 0.387 | 1 | 0.041 | 0.841 |
Age | 1 | 19.971 |
|
1 | 16.675 |
|
1 | 4.492 |
|
1 | 9.711 |
|
1 | 5.033 |
|
A
A
It has previously been suggested that numerous painful interventions in the neonatal period could compromise the cardiovascular autonomic responses typically observed during painful stimulations in later life [
Despite a comparable percentage change of HR in the first 15 seconds of the CPT, preterm and full-term subjects differed in the evolution of their HR response. Full-term subjects had a mean increase of
Similar results are found when comparing mean percentage change in HR. In fact, high-pain preterm subjects show a lower mean percentage change in their HR. In addition, we found that full-term subjects with surgery also showed a lower mean percentage change in HR. These findings suggest that having suffered intense pain at birth regardless of gestational age alters heart rate changes to experimental pain. To our knowledge, no other study has clearly exposed this relationship. Previous studies only compared preterm to full-term subjects. Preterm infants who experienced a higher number of painful procedures following birth showed less autonomic reactivity [
A higher baseline HR was observed in subjects who underwent painful procedures at birth, when age at testing was entered as a covariate in the analysis. This difference is not found if we compare full-term to preterm birth, suggesting that the pain suffered is the cause of the altered HR. This is consistent with previous studies that found that the high-pain preterm group showed a higher resting HR, suggesting a state of chronic arousal [
We also found important autonomic reactivity differences in our groups. In fact, full-term surgery and high-pain preterm participants showed a stronger sympathetic activity (LF) and a lower cardiac vagal activity (HF) at rest, suggesting again a state of chronic arousal. Our study confirmed a sympathovagal imbalance in subjects who experienced pain at birth (either born preterm or full-term). This sympathovagal imbalance has also been found in subjects with migraine and alterations in autonomic nervous system function and is thought to contribute to the development of symptoms in these patients [
However, some limitations should be taken into account concerning our study. First, not all subjects were able to complete the entire 2-minute CPT procedure (47%). A minimum recording of two minutes of the heart rate has been shown to be necessary to perform a short-term HRV analysis [
Clinically, our results have important implications because they demonstrate an impact of the length of stay in the NICU and the number of days under mechanical ventilation (proxies used to evaluate the number of painful and stressful procedures experienced at birth) on resting and stressed autonomic cardiac activity years after the stay in the NICU. Subjects with prior pain exposure, either born preterm or full-term, showed a higher baseline HR, a stronger sympathetic activity, and a lower cardiac vagal activity even at adulthood. This study adds to the prevailing literature in the sense that the impact on heart rate comes from pain experienced at birth (represented by a longer NICU stay and time under mechanical ventilation in preterm or surgery in full-term subjects), independently of the birth status (preterm versus full-term), and is still found in young adulthood. This could lead to impaired reactions to pain or stress in later life. It is important to consider the long-term impact of pain at birth, even in the presence of some analgesia in full-term neonates who underwent surgery. We showed that the regulation of the cardiac autonomic system is still altered years after all procedural pain is over. Conclusions of our study show the importance of providing better analgesic care for the neonates and limiting as much as possible painful medical procedures during this critical time.
None of the authors has any potential conflict of interests to declare regarding the publication of this paper.
Drs. Sylvie Lafrenaye and Serge Marchand are supported members of the Centre de Recherche Clinique Étienne-Le Bel of the CHUS. Dr. Serge Marchand holds grants from the Canadian Institute for Health Research (CIHR). The authors would like to thank the statistician of the Centre de Recherche Clinique, Marie-Pierre Garant, for her help with the analysis. Melanie Morin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.