Sugar and Pain Relief in Newborns: A Mini Review

Sugar and Pain Relief in Newborns: A Mini Review

I am not a physician, so it did come as something of a surprise when I discovered it is common practice to give newborn babies sugar for pain relief. I am a medical scientist though, and I don’t know of a single way in which sucrose could possibly result in pain relief, unless you’re one of those people who eats to fill the void in your life. That is rather different to physical pain though. If there is a biochemical pathway in which sucrose, taken orally or otherwise does provide pain relief, then I haven’t heard of it, save for it having been suggested that sweet tasting solutions possibly mediate analgesia via endogenous opioid release (Smith et al., 1990). However, after a brief review of the literature, I will argue what really needs to be considered is the effect (or lack of any) sucrose has on pain activation in the brain.

Firstly, the number of studies cited in which the main, if not the only marker of pain relief is ‘crying time’ is worrying to say the least (Blass and Shah, 1995 ; Haouari et al., 1995 ; Abad et al., 1996 ; Skogsdal et al., 1997 ; Isik et al., 2000 ; Bilgen et al., 2001). Isik et al., 2000 also looked at heart rate and found no significant difference between the control, sucrose or glucose groups, and so based their conclusion that sucrose reduces pain on crying time alone. Contrary to this Bilgen et al., 2001 did find significant changes in heart rate.

Abad et al., 1996 also looked at oxygen saturation, respiratory rate and heart rate. They found no significant changes in either oxygen saturation or respiratory rate, and the supposed “significant change” in heart rate was based on n values for the control and sucrose solution groups of 12 and 8 respectively. Haouari et al., 1995 had an n value of 15 for each group. How large should an n value be before the Central Limit Theorem kicks in? The general rule of thumb is at least n = 30. The higher the n value the better, because the probability of a random deviation of a particular size (from the population average) decreases with an increase in the sample size. The n value should also be kept the same, or as close as possible between groups, otherwise it can cause confounding.

Unlike many previous studies, Skogsdal et al., 1997 actually had an acceptable n value of 30 for each group. They compared glucose solutions and breast milk, measuring crying time and heart rate; finding that it returned to baseline faster at higher concentrations of glucose than did breast milk or the controls.

In short, many studies on this subject to date have inadequate sample sizes and a methodology which is fundamentally undermined by the systematic error of assuming that a reduction in crying time or other behavioural changes are adequate markers of pain relief. In fact, we know that although pain stimulation generally evokes cortical and behavioural changes that pain can be processed at the cortical level without producing behavioural changes, and cortical pain responses were recorded in infants who did not display a change in facial expression (Slater et al., 2008). All this means that a low pain score based on behavioural changes alone is not enough to indicate pain relief.

What needs to be considered is where sucrose acts in the brain. A study that has looked at the effect of sucrose on pain relief by measuring brain activity found no significant difference in nociceptive brain activity after noxious heel lance between the sucrose and control group (Slater et al., 2010). There were some methodological problems with the study by Slater et al., 2010, but it remains the case that even though previous studies have shown sucrose decreases various infant pain profile scores, this fact is irrelevant when taking into account such scoring systems do not reflect nociceptive activity in the brain of newborns.

In rats given oral sucrose it has been shown that decreased nociceptive behaviour persists after midbrain transection, which suggests the forebrain is not needed for this effect (Anseloni et al., 2005). Therefore, sucrose could inhibit facial motor activity by mediating brainstem inhibition of behaviour, despite strong pain activation occurring in the forebrain. This means it’s possible to feel pain even though you can’t express it, which is just the sort of thing you’d miss if the focus of your study is crying time or heart rate instead of pain activation in the brain.

This has not been an extensive review. It is always recommended to follow up references and seek out others. Based on what I have read so far, I remain unconvinced sucrose is actually able to provide pain relief in newborns, rather it seems to simply make it appear as though it does by suppressing their ability to express pain. Always remember: correlation does not imply causation.

This post originally appeared on Jonathan’s former blog The Carapace in 2011. The blog was decommissioned in 2012. Since then a Cochrane review in 2013 has come out in support of sucrose for analgesia in newborns, but it too relied on studies that focused on crying time and heart rate instead of pain activation in the brain.

[Header: Creative Commons licensed Flickr photo by carowallis1]