Intended for healthcare professionals


The analgesic effect of sucrose in full term infants: a randomised controlled trial

BMJ 1995; 310 doi: (Published 10 June 1995) Cite this as: BMJ 1995;310:1498
  1. Nora Haouari, visiting clinical fellowa,
  2. Christopher Wood, lecturera,
  3. Gillian Griffiths, research nursea,
  4. Malcolm Levene, professor of paediatricsa
  1. a University Division of Paediatrics and Child Health, D Floor, Clarendon Wing, Leeds General Infirmary, Leeds LS2 9NS
  1. Correspondence to: Professor Levene.
  • Accepted 11 April 1995


Objective: To evaluate the effects of different sucrose concentrations on measures of neonatal pain.

Design: Randomised, double blind, placebo controlled trial of sterile water (control) or one of three solutions of sucrose—namely, 12.5%, 25%, and 50% wt/vol.

Setting: Postnatal ward.

Patients: 60 healthy infants of gestational age 37-42 weeks and postnatal age 1-6 days randomised to receive 2 ml of one of the four solutions on to the tongue two minutes before heel prick sampling for serum bilirubin concentrations.

Main outcome measure: Duration of crying over the first three minutes after heel prick.

Results: There was a significant reduction in overall crying time and heart rate after three minutes in the babies given 50% sucrose as compared with controls. This was maximal one minute after heel prick in the 50% sucrose group and became statistically significant in the 25% sucrose group at two minutes. There was a significant trend for a reduction in crying time with increasing concentrations of sucrose over the first three minutes.

Conclusion: Concentrated sucrose solution seems to reduce crying and the autonomic effects of a painful procedure in healthy normal babies. Sucrose may be a useful and safe analgesic for minor procedures in neonates.

Key messages

  • Key messages

  • Little is done to minimise the discomfort of these procedures

  • Placing 2 ml of a 25% or 50% sucrose solution on the tongue before heel prick significantly reduces crying time

  • There is a dose-response effect in the reduction of crying with increasing concentrations of sucrose

  • Sucrose on the tongue may be a useful and safe form of analgesia in newborn infants


The ability of neonates to perceive and react to pain has been much debated in recent years. We know that most of the anatomical pathways and neurotransmitter function necessary for pain perception are fully or nearly fully developed in the neonatal period.1 2 Yet many people are still reluctant to believe that pain felt by neonates may be as severe as that felt by older children or adults.

All neonates born in developed countries are subjected to painful procedures. Heel pricks for Guthrie tests are near universal and intramuscular vitamin K injections and routine circumcision are still common in some countries. Blood samples for serum bilirubin and blood sugar estimations are taken when clinically indicated, and if a neonate is sick frequent blood sampling, venepuncture, and more severe procedures which cause tissue injury are common. We estimate that in our hospital every baby has at least one heel prick procedure, 15% of babies have two to five blood samples taken or intramuscular injections, and 2% have five or more such procedures.

We assessed the use of sucrose to reduce pain in neonates subjected to routine blood sampling by heel prick.

Patients and methods

Healthy full term infants who required heel prick sampling for serum bilirubin estimations were recruited from the postnatal wards of this hospital over six months. Babies with Apgar scores of less than 7 at one minute or who had received naloxone were excluded. All heel prick samples were drawn by a single, experienced nurse (GG) using a sterile lance. The babies responses were observed by NH and CW.

Infants were taken to a warm, quiet nursery for blood sampling and were fully clothed apart from the foot used for sampling. Before skin preparation a pulse oximeter was applied to the baby's hand to measure changes in oxygen saturation and heart rate during the study. Parents could be present if they wished but did not speak to or touch the baby during the procedure. We calculated that to achieve 80% power to show a 50% reduction in crying time (P=0.05) 15 babies would be required in each group.

Babies were allocated at random to receive one of four solutions on to the tongue: sterile water (control) or 12.5%, 25%, or 50% sucrose (wt/vol). Randomisation was by means of pre-prepared solutions in coded bottles. Investigators were blind to the nature of the solutions throughout. Stratification was not used. The test solution (2 ml) was given by syringe into the baby's mouth over less than one minute. The heel was then exposed, cleaned with a sterile swab, and held gently. Two minutes after the test solution the heel was lanced and gently squeezed.

The baby's behavioural state3 was recorded immediately before lancing. Four facial expressions reported to occur during an acute painful stimulus4 were also recorded on a 0-4 scale. Crying during sampling and during the three minutes after sampling (recovery phase) was recorded on audio tape and later analysed blindly for the duration of crying. Crying time was defined as the number of seconds that the baby cried within the first three minutes and the duration of the first cry was the duration of continuous crying before a quiet interval of five seconds.

The study was approved by the hospital ethics committee and all parents gave informed consent. Analysis of non-parametric data was by the Mann-Whitney U test (Minitab) or a trend test.5 Changes in heart rate were expressed as a percentage of resting heart rate. Parametric statistical analysis was by Student's t test.


Sixty babies were studied. There were no withdrawals after randomisation. Details of the babies in each group are shown in table I. There were no significant differences between the groups in sex, gestational age, birth weight, postnatal age, mode of delivery, behavioural state at rest, or time since last feed.


Details of neonates studied

View this table:

Figure 1 shows the duration of crying in the first three minutes after heel prick in each treatment group. The median crying time and median duration of first cry are shown in table II. There was a significant reduction in both the total crying time and the duration of first cry in the 50% sucrose group as compared with the controls (P=0.02). Over the range of sucrose doses studied there was a significant trend towards a reduction in crying time with greater concentrations of sucrose (P=0.007). There was a similar trend in the reduction of the duration of first cry with increasing concentrations of sucrose (P=0.004).


Total time crying in first three minutes after heel prick in 60 babies given sterile water (controls) or 12.5%, 25%, or 50% sucrose. Points are individual values. Horizontal lines are median values. Shaded areas are interquartile ranges. (Test for trend across groups: P=0.007)


Crying time and change in heart rate after heel prick

View this table:

The median time crying in each one minute epoch after heel prick is shown in figure 2 and table II. There was a significant reduction in crying time at the end of the first minute in the 50% sucrose group as compared with the controls (P<0.02), and in the second minute the duration of cry was significantly less in the 25% (P=0.02) and 50% (P=0.003) sucrose groups. By the end of the third minute crying was reduced in all four groups, with no significant differences between them.


Time crying during each one minute epoch after heel prick in infants given sterile water (controls) or 12.5%, 25%, or 50% oral sucrose. Points are median values for each group

There were no significant differences in oxygen saturation values before, during, or after sampling. Changes in heart rate are shown in table II. There was a significant difference in heart rate three minutes after lancing (P=0.02) in the controls as compared with the 50% sucrose group. There were no differences between babies given less concentrated solutions of sucrose. There were no significant differences in facial responses between the four groups.


Using a lancet to draw blood is a common procedure and is recognised to cause pain in patients able to report it. We have shown that there is a dose-response effect of increasing concentrations of sucrose, resulting in a reduction in crying time in healthy full term infants. This effect is evident by the end of the first minute after heel prick and with a 50% sucrose solution seems to be maximal by the end of the second minute. Sucrose not only reduces the duration of crying but also reduces tachycardia induced by a painful procedure.

A fairly high proportion of infants in this study were born by caesarean section (27%). These babies spent longer in hospital with their convalescing mothers and were therefore more likely still to be in hospital when jaundice was noted. The proportions of babies born by caesarean section were not significantly different among the groups and we think it unlikely that anaesthetic agents given to the mothers would have affected their babies responses to painful stimuli.

Sucrose increases pain latency in animals after a standard noxious procedure. Blass et al found that in 10 day old rats intraoral infusions of 3.5%, 7.5%, and 11.5% sucrose all increased latency before each animal lifted its paw off a hotplate.6 This effect was completely reversed by naltrexone. Blass and Hoffmeyer subsequently showed that 2 ml of 12% intraoral sucrose significantly reduced the duration of cry in newborn babies subjected to heel prick or circumcision.7 Our group repeated that study with 7.5% sucrose and found no difference in the duration of crying.8 However, we remained intrigued by the reports of Blass et al and undertook this study with different sucrose concentrations. We conclude that in our first study we used a sucrose solution too weak to induce a measurable analgesic effect.

Non-sucrose sweet substances such as saccharin also seem to increase pain latency in animals.9 Other substances such as milk and in particular milk fat seem to have a similar effect through the endogenous opiate pathways.10 We do not know whether simply cuddling an infant after heel prick is as effective in reducing crying as 50% sucrose. We are undertaking further clinical studies to evaluate the effect of sucrose and other substances in pain prevention in immature babies and older children.

This study was supported by the Special Trustees of the Leeds General Infirmary.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.