Analgesic Effect of Low-Level Laser Therapy before Heel Lance for Pain Management in Healthy Term Neonates: A Randomized Controlled Trial

Neonates frequently undergo painful procedures such as heel lance, which can elicit immediate physiological and behavioral stress responses and may have longer-term neurodevelopmental implications. Pain assessment in neonates is challenging due to nonverbal responses, necessitating multidimensional tools (e.g., NFCS, NPASS). Current non-pharmacologic strategies like breastfeeding and oral sweet solutions are effective but not always feasible (e.g., maternal inability to breastfeed, uncertainty around repeated sucrose in preterm infants). Low-level laser therapy (LLLT), a noninvasive photobiomodulation technique, can reduce pain via neural conduction blockade, endorphin release, and anti-inflammatory effects. Prior neonatal LLLT evidence is scarce and inconclusive, with one negative trial using laser acupuncture at a distant acupoint (Yintang). The authors hypothesized that local LLLT pretreatment at the heel lance site would provide analgesia comparable to expressed breast milk (BM) feeding by reducing nociceptive input, thereby attenuating behavioral and physiological pain responses during heel lance in healthy term neonates.

Guidelines from the American Academy of Pediatrics and Cochrane reviews support breastfeeding and sweet solutions (sucrose/glucose) to reduce procedural pain in neonates, with minimal adverse effects. However, barriers exist to routine breastfeeding during procedures, and concerns remain about repeated sucrose exposure. Evidence on LLLT for neonatal procedural pain is limited. One RCT applying 0.3 J laser at the Yintang acupoint found LLLT less effective than oral sucrose; conversely, acupressure near the heel (BL60, KI3) reduced pain, suggesting proximity to the nociceptive site may matter. In adults and animal models, LLLT reduces pain through modulation of peripheral and central nociception, inhibition of substance P, enhancement of endogenous opioids, and preconditioning effects, but optimal parameters (power, wavelength, frequency, duration) are debated. Salivary cortisol is a validated neonatal stress biomarker post-procedure; salivary α-amylase is a stress marker in older populations but less consistent in neonates due to gland maturation.

Design: Prospective, single-center, open-label, randomized controlled trial conducted at the newborn nursery of Kaohsiung Chang Gung Memorial Hospital (Taiwan) from August 2017 to October 2018. Trial registration: NCT03268148. IRB approval: 201700474A3C501 (19 July 2017). Sample size: Based on prior effects of glucose on pain scores, power 90%, alpha 0.05, 10% dropout, total required n=130 (G*Power 3.1.9.2). Participants: Healthy term neonates (37–42 weeks GA), Apgar ≥7 at 5 min, no illness/abnormal exam; exclusions included Apgar <7 at 5 min, birth injuries, major malformations, severe congenital anomalies, cyanotic heart disease, severe infections, drug withdrawal, prior analgesic/sedative treatment, need for NICU. Additional dropout criteria: crying >5 min pre-procedure, need for a second heel lance, or parental withdrawal. Only the first heel lance was included; neonates requiring early or repeat heel lances were excluded. Randomization and blinding: 1:1 allocation (LLLT vs BM) via Excel-generated block randomization (block size 2) by an independent researcher using sealed envelopes. Open-label interventions; outcome assessors (three investigators) were blinded and evaluated videos. Setting and procedures: Infants studied in a quiet room, fed within 3 h but not within 30 min prior. Placed on infant warmer; continuous monitoring (SureSigns VM4) of heart rate, respiratory rate, SpO2, and BP (pre and 3 min post). Two synchronized cameras recorded monitor and infant behavior from baseline through the procedure (~8–10 min). Saliva samples collected at baseline, immediately post heel lance, and 20 min post. Parents were not present. Interventions: - LLLT group: After baseline saliva, LLLT applied locally to the heel lance site using a LaserPen (RJ-laser, Waldkirch, Germany): power 150 mW, wavelength 810 nm, Nogier frequency E 4672 Hz, duration 20 s (energy 1.5 J), applied by the same physician. Heel lance performed by the same research nurse thereafter. - Breast milk (BM) group: After baseline saliva, 5 mL expressed BM administered via syringe over 2 min. Heel lance performed by the same research nurse. Outcomes: Primary behavioral outcomes during heel lance: latency to first cry, duration of crying, squeeze time (from puncture to completion of five capillary tubes), and pain scores using NPASS (0–10) and NFCS (0–8) assessed at baseline and 1, 2, 3 min post-lance from video review by two blinded observers with adjudication by a third. Secondary physiological outcomes: heart rate, respiratory rate, SpO2 (averaged over 1 min baseline and 3 min post), systolic/diastolic BP (baseline and 3 min post). Salivary biomarkers: cortisol (Salimetrics EIA) and α-amylase (Salimetrics kinetic assay) per manufacturer protocols; samples refrigerated at 4°C then frozen at −20°C within 4 h; analyzed in duplicate. Safety: Adverse events (vomiting, local redness, swelling) monitored post-intervention. Statistics: SPSS 22.0. Between-group comparisons via chi-square (categorical) and independent t-tests (continuous). Within-group salivary biomarker comparisons via paired t-tests. Two-sided p<0.05 considered significant.

Participants: 142 assessed; 125 randomized (LLLT n=62, BM n=63). Exclusions post-randomization: LLLT 2 withdrew (early discharge), LLLT 5 and BM 4 required second heel lance (insufficient blood). Analyzed: LLLT n=55, BM n=59. Groups were similar at baseline: gestational age 39.31±1.02 vs 39.10±1.18 weeks (p=0.299); birth weight 3130.91±368.15 vs 3122.63±370.61 g (p=0.905); male 40.0% vs 49.2%; cesarean 23.6% vs 25.4%; Apgar 1 min 9±0 both; 5 min 10±0 both. Primary behavioral outcomes: - Latency to first cry (s): 5.61±12.09 (LLLT) vs 5.10±7.27 (BM), p=0.795. - Duration of crying (s): 97.07±62.16 (LLLT) vs 128.08±108.16 (BM), p=0.062 (trend toward shorter with LLLT). - Squeeze time (s): 95.07±40.16 (LLLT) vs 122.51±95.98 (BM), p=0.047 (significantly shorter with LLLT). - Pain scores (NPASS, NFCS): no significant between-group differences at baseline or 1, 2, 3 min post-lance. Secondary physiological outcomes: No significant between-group differences pre- vs post-lance for heart rate, respiratory rate, SpO2, SBP, or DBP. Salivary biomarkers: One BM specimen insufficient; four LLLT specimens insufficient for α-amylase (cortisol only). Between groups, no significant differences at baseline, immediately, or 20 min post. Within-group changes: - BM group cortisol increased at 20 min vs baseline (p=0.006) and vs immediate post (p<0.001). - BM group α-amylase decreased immediately post vs baseline (p<0.001), then increased at 20 min vs immediate (p<0.001). - LLLT group showed no significant within-group changes in cortisol or α-amylase at measured time points. Safety: No visible discomfort during LLLT; no adverse events (vomiting, local redness, swelling) reported.

Applying LLLT locally to the heel before lance resulted in a significantly shorter squeeze time and behavioral/physiological responses comparable to expressed BM feeding, supporting the hypothesis that local photobiomodulation can provide procedural analgesia equivalent to BM in healthy term neonates. The lack of differences in NPASS/NFCS and vital signs suggests similar overall pain experience between groups, while shorter squeeze time may reflect reduced nociception or improved tolerance during blood collection with LLLT. Biomarker findings indicate greater variability and a delayed cortisol rise in the BM group, whereas LLLT maintained more stable cortisol and α-amylase levels, potentially reflecting attenuated stress responses, though between-group differences were not significant. Compared to prior negative LLLT trials targeting distant acupoints (e.g., Yintang), this study employed local irradiation at the nociceptive site with different parameters (810 nm, 150 mW, 20 s, 1.5 J; Nogier E 4672 Hz), which may better modulate peripheral nerve conduction and local nociceptive pathways. Practical barriers often limit breastfeeding during early-morning heel lances; expressed BM is commonly used locally and was an ethical, feasible comparator. Overall, findings support LLLT as a noninvasive, painless alternative that could be integrated into neonatal procedural pain protocols, with potential advantages in workflow (shorter procedure time) and biomarker stability.

In healthy term neonates undergoing heel lance, low-level laser therapy administered at the lance site before the procedure achieved analgesic effects comparable to expressed breast milk feeding, with the added benefit of a shorter procedural (squeeze) time and less variability in salivary stress biomarkers. LLLT appears safe and feasible and should be considered as an analgesic option for heel lance in clinical practice. Future research should include multicenter, blinded trials comparing LLLT to other standard interventions (e.g., breastfeeding, sucrose), dose–response optimization of laser parameters, evaluation in preterm and medically fragile infants, and assessment of long-term neurodevelopmental outcomes and implementation strategies.

Single-center study limits generalizability. Open-label design may introduce performance bias, although outcome assessors were blinded. Potential confounding from minimal probe pressure during LLLT was mitigated by instructing a light touch but cannot be entirely excluded. Choice of expressed breast milk (rather than breastfeeding or sucrose) as comparator may differ from some standards of care, though justified by local practice and ethical considerations. Some salivary samples were insufficient for analysis, reducing biomarker sample size.