Background and aims Sucrose and breast milk during painful procedures are reported to decrease pain behavioural expression in neonates. Recent data showed a persistent cortical pain response while using the sucrose during a painful procedure.

To compare the efficacy of sucrose versus breast milk for specific-pain brain activity relief during a painful procedure in neonates.

Methods Randomised, prospective, controlled study. Each term newborn was randomly assigned to sucrose or breastfeeding group at day 3 during a systematic venipuncture. Change in the total haemoglobin concentration in the controlateral somatosensory cortex (Near Infra-red Spectroscoy, NIRS) was assessed 10 seconds before and after the venipuncture. Neonatal Facial Coding System (NFCS) was assessed 2 min before and at the time of the venipuncture. Groups were compared using Wilcoxon test for the variations in NIRS and Chi-square test for the NFCS scores.

Results 113 newborns were included (sucrose: 56, breastfeeding: 57) with a mean (sd) of 39.3 weeks (0.9) for gestational age and 3370 g (478) for birth weight. 103 were analysed for the NIRS (sucrose: 55, breastfeeding: 48). Median (quartiles) of total haemoglobin concentration change was -8.5 µmol/L (-34.5; 12.5) for sucrose group and 12.3 µmol/L (-23.4; 39.3) for breastfeeding group with no statistical difference (p = 0.06). NFCS scores were significantly different with 46.8% with a painful score in the breastfeeding versus 26.8% in the sucrose (p = 0.03).

Conclusions No difference were found between sucrose and breastfeeding on specific-pain brain activity during a venipuncture in term newborns. A discordance was revealed between NFCS scores and NIRS analysis.