Retrospectacle: A Neuroscience Blog

Some interesting research is being reported in the mainstream media: that buckwheat honey can reduce coughing and soothe sore throats as well as over-the-counter cough syrup (dextromethorphan), in children. ABC News specifically reported that, in comparison to baseline levels, coughing children who received honey slept better and coughed less that those that received cough syrup or nothing at all. I decided to go to the journal article, published in Archives of Pediatrics and Adolescent Medicine, and read about the results myself.

The authors’ objective was to compare honey, honey-flavored dextromethorpham (DXM), and no treatment on coughing and sleep behavior in children with colds. DXM is an antitussive and common ingredient in cold medications. However at high doses (high above therapeutic doses) can be dissociative hallucinogen as well as cause a type of brain damage called Olney’s lesions. So, it seems quite reasonable to test whether a drug-free alternative might work as well as DXM syrup for children.

(Continued below the fold…)

Parents of 105 children with comparable symptoms of upper respiratory infections and aged 2-18 were administered a survey before the experimental trials begun. Then honey, honey-flavored DXM, or no treatment was administered to the children on the next night, with the parents subsequently reporting on the effects using the same survey. See a clip of the survey below, view entire survey here.

i-ae9153dc209adcc243b19764950e88ff-honey 2.bmp

Children received the treatment in a “partially double-blinded fashion” in an opaque 10 ml syringe placed in a brown paper bag. The no-treatment group was not blinded, but the honey and DXM groups were. The treatments were given to the children by the parents 30 minutes before bedtime, and the surveys were conducted over the phone the next morning.

Parents’ subjective self-reports indicated that the group of children which received honey fared as well or slightly better than those who received the DXM cough syrup or nothing. (Results indicated below, lower score indicate more desireable outcome.) Significant differences existed in the child sleep, child’s cough, and parent sleep. Although it is worth noting that even the group which received no treatment fared better than the first day, suggesting that the symptoms naturally improved over time even in the abscence of intervention.

i-4e5f4329fe47a3f384c5567b7c476130-honey 3.bmp

These results are not entirely unexpected. Honey is a viscous liquid which coats the throat and reduces the feeling of irritation. It also has some anti-microbial properties and is an anti-oxidant. Also, it tastes great, as opposed to many cough syrups which have a less-than-pleasing taste. This might make children more apt to take it without a fuss. There were a few downsides to honey: some hyperactivity and insomnia in a few children, but it was minor. The subjective nature of the survey should also be considered as less than ideal– the parents, while reported to be blinded to the condition of the treatment, might have actually have known via the child’s reaction to the medication.

Honey as an alternative to DXM cough syrup may be important not just in children. DXM should not be taken if a person is also taking SSRIs, MAOis, CNS depressants (alcohol, antihistimines, etc), and a host of other medications, as it could result in a fatal drug reaction. Honey would provide a safe replacement if this paper’s conclusions as to its effectiveness were duplicated in other studies (which were perhaps a bit more methodologically rigorous.)

Reference:
Paul et al. 2007. Effect of Honey, Dextromethorphan, and No Treatment on Nocturnal Cough and Sleep Quality for Coughing Children and Their Parents. Arch Pediatric Adoles Med. 161, 1140-1147. PDF

Comments

#1 Rosie Redfield
December 5, 2007

I’m skeptical.

Honey will (briefly) coat the part of the throat that swallows liquids, but I ‘m pretty sure it won’t directly contact the parts of the respiratory tract whose irritation causes coughing. (Think of how a sore throat feels different from a cough.) And although concentrated honey can kill bacteria, the levels of honey that would be circulating in the bloodstream after swallowing 10 ml are very unlikely to have any effect on bacteria or viruses in the respiratory tract.

The obvious next experiment should compare honey with sugar-water.
#2 user
December 5, 2007

remember, don’t feed infants honey or they will die of botulism!!!!!11
#3 Shelley Batts
December 5, 2007

Well I don’t think its THAT clearcut (honey = death). But you are right that it is not advisable to give honey to infants ( less than 2 years old) due to the risk of botulism.
#4 Shelley Batts
December 5, 2007

Hi Rosie, thanks for dropping by. I too was skeptical, actually I read the article pretty posed to tear it apart. The methods had flaws, but the results they got were significant. I found that in itself interesting, but I too would like to see more support for their conclusion in more rigorous trials that were less subjective. Another issue I had was that the DXM syrup was honey flavored and mixed to be of similar sweetness to honey, however as a former child I can attest that I know cough syrup when i taste it (yuck). The faces alone made by children who have to take cough syrup might clue in the parents as to what their child was receiving, which might in turn influence the later report.
#5 rev_matt_y
December 5, 2007

My 4 year old had a horrible sore throat and cough last night. If she has what I had last week it’s going to be a lot worse tonight tonight, so I’ll try the honey approach as at least unlikely to make things worse.
#6 Gelfling
December 5, 2007

It occurs that children often cry when they get cough syrup. This causes extra postnasal drip, which then would exacerbate the cough.

I wonder if they allowed for that as one of their variables?

Only a thought, meant to point out how difficult it is to neutralise all variables in the real world, such as the “yuck” face mentioned above.
#7 rehana
December 5, 2007

Gelfling: that would still make the cough syrup a less effective treatment.

Shelley: If the kids made faces when they got cough syrup, wouldn’t the parents be more likely to think it was working, not less?
#8 SabrinaW
December 5, 2007

While normally a placebo effect is seen as a hindrance to medical research, I think in this case a psychological boost from using honey vs. a cough syrup is not necessarily bad because cough syrup does not actually cure illness, but merely assuages symptoms. So if the child prefers to take the honey, the parent feels better giving it, and the results are comparable, I think that is an acceptable result.

Of course, the cough syrup manufacturers won’t be happy.
#9 Bryce M
December 5, 2007

Great, we find this out just when Colony Collapse Disorder is affecting honeybees!
#10 Shelley Batts
December 5, 2007

If the kids made faces when they got cough syrup, wouldn’t the parents be more likely to think it was working, not less?

The reason for blinding is to preclude bias, so the effects of knowledge re: cough syrup or honey would depend on the implicit bias of the parent, and also what they were told about the nature of the study. If the parent was biased against cough syrup, which perhaps many are given the risks at high doses (and the instant popularity of this study), they might be influenced negatively in regards to its effects of their child.
#11 SabrinaW
December 5, 2007

Great, we find this out just when Colony Collapse Disorder is affecting honeybees!

While I am reluctant to use studies to fuel social agendas, I hope that this study plus the wound treatment study will help raise awareness about the impending disaster of losing the honeybee (and it would be a disaster to agriculture).
#12 Lisa Briand
December 5, 2007

I was recently in Britain for a conference and came down with a cold. When I went to the pharmacy for cough medicine I was very disappointed to find that the cough medicine there did not contain DXM but rather the main active ingredient was honey. Being a typical american, I scoffed at this however now I realize that perhaps the Brits are on to something!
#13 Ian
December 5, 2007

Interesting study, and fun if true. What caught my eye was the way the authors analyzed the response scale data (cough frequency, severity, etc.) to determine the effect of honey – they should have taken into account the night-to-night change in the no-treatment condition, but for some reason they didn’t.
The study design includes a “no treatment” condition where parents receive an empty syringe; it’s appropriate to have a “background” condition like this for comparison. And, in fact, the response scale data show that the children were coughing less on night #2, even when they had not received honey or DM. That’s good to know, even if we don’t know what caused that improvement.
But then the researchers ignore this night-to-night difference in the no-treatment group when analyzing the honey and DM groups. What they should have done, minimally, was to subtract that no-treatment night-to-night difference from the treatment group night-to-night differences – that would give the effect of honey/DM *over and above* the extraneous improvement captured by the no-treatement group. (For example, using data reported in the results section, night1 to night2 difference in cough frequency was .92 for the no-treatment group, 1.89 for honey, and 1.39 for DM. Subtracting the no-treatment dif from the honey dif cuts the honey effect in half, to .97, and reduces it even further for the other responses.)
So, would the night-to-night differences for the honey group still be significant after holding the background improvement rate constant? Unfortunately, the study doesn’t report the F- and t-values and/or variability needed to figure this out exactly. For me, this one’s filed under “needs more research or clarification.”
#14 jvarisco
December 5, 2007

My mom always used to have me talk honey with some ginger in it when I had a cough, and it worked pretty well. If honey by itself is effective that’s even better.
#15 #1 Dinosaur
December 5, 2007

“There were a few downsides to honey which seemed to be related to sugar intake: some hyperactivity and insomnia in a few children, but it was minor.” (emphasis added)

As a scientist, why are you continuing to promulgate the myth (by not labeling it as such) about sugar and hyperactivity? It’s been shown over and over that sugar intake has no direct effect on children’s behavior. (Many more references to be found simply by googling “Sugar and Hyperactivity)
#16 Abby Normal
December 6, 2007

When I was a kid I used to get tea with lots of lemon and honey for a cough. I’d get it for a sore throat too. But first I’d have to gargle salt water. Seemed to work pretty well most the time. (Not that that validates anything. I’m just sharing.)

When I was older and living on my own I continued the traditional treatment, but added warm bourbon. I don’t know if they were any more or less effective. But they certainly left me feeling pretty good.
#17 Shelley Batts
December 6, 2007

why are you continuing to promulgate the myth (by not labeling it as such) about sugar and hyperactivity

Dinosaur, I guess you learn something new everyday. ‘My bad!’ Thanks for pointing out those studies, as I was not aware of them.
#18 RPM
December 6, 2007

FYI:

Funding/Support: This work was supported by an unrestricted research grant from the National Honey Board, an industry-funded agency of the US Department of Agriculture.

Also, isn’t there literature suggesting you shouldn’t give young children cough syrup? Aren’t there some nasty side-effects?
#19 Shelley Batts
December 6, 2007

Interesting funding situation. While it doesn’t *negate* the results, it seems to heighten the need for independent, more rigorous methods to test the question. As for not giving young children cough syrup (older than 2 yr old), I’m not aware of any specific studies pointing to nasty side effects at the recommended doses (I mentioned two at higher doses thoguh– hallucinogen and Olney’s lesions), are you?
#20 David Colquhoun
December 6, 2007

Aha only just discovered that someone else read the whole paper. I’ve also done an analysis of it at http://dcscience.net/?p=209
My conclusion is that it’s all spin, plus a small effect as a result of the ‘no treatment’ group not being blind. The sad thing is that the Penn State media person told me that the first author had approved their inaccurate press release.
#21 Shelley Batts
December 6, 2007

The press releases contained quite a lot of spin, however the paper itself did report significant differences between the ‘nothing’ group and the honey group, but not the DMX and the honey group, so *thats* confusing. Plus the authors themselves mention that they had found in a previous study that DMX didn’t reduce symptoms of coughing etc, and if there’s no statistical difference between honey and DMX in this study, then what are we to glean from that?

I think that the idea is intriguing but would like to see it repeated under more neutral and stringent (ie, actually blinded conditions) auspices before *I* would publish it. Not to mention approve a press release that states things so definitively.
#22 Nathaniel
December 7, 2007

DMX has been shown to be largely ineffective. I would argue that the majority of the effects of cough syrup are caused by the coating of the throat and the sedative effect of the medications therein. The graphs clearly show that honey is more effective than the DMX though. Even if the study had it’s flaws the results speak for themselves. Perhaps they should also test high fructose corn syrup and honey laced with DMX… just to make sure what’s actually causing them to get better.

Just goes to show that simple home remedies can work. And can work better than something prescribed by a doctor. That being said… doctors are still needed. The majority of homeopathic medicine is a practice of mind over body (they believe it works and their brain promotes the healing). However, it seems that some remedies actually do have at least some degree of a basis in actual results.
#23 tgibbs
December 8, 2007

The press releases contained quite a lot of spin, however the paper itself did report significant differences between the ‘nothing’ group and the honey group, but not the DMX and the honey group, so *thats* confusing. Plus the authors themselves mention that they had found in a previous study that DMX didn’t reduce symptoms of coughing etc, and if there’s no statistical difference between honey and DMX in this study, then what are we to glean from that?

The effect size seems very small, so it is possible that they just missed a small but significant effect between DMX and no treatment. You often see this in significance testing, in which A is significantly different from C, but A and B are not significantly different and B and C are not significantly different. Remember that “not significantly different” is not the same as “not different.”

Or (as David Colquhoun points out in his analysis), maybe they simply “got lucky.” P = 0.04 is significant by convention, but does not inspire a huge degree of confidence. In any case, it looks very likely that the effect of honey is simply a placebo effect–which is not necessarily a bad thing from a parent’s point of view. Witness the miraculous analgesic effect of a Band-Aid (preferably decorated with cartoon characters).

As an aside, the book that I loan out to students seeking to learn more about statistics is my tattered copy of David Colquhoun’s “Lectures in Biostatistics” (with stern admonitions to return it, as it is sadly out of print). There are other books that provide more detail about how to do statistics, but it is the best book that I know for learning how to think about statistics.
#24 Luna_the_cat
December 10, 2007

An old deep-south recipe for cough medicine was to cut a large onion in half and boil it in honey until it became transparent, then strain and use the honey. I wonder now if it was just an issue of making people feel like they were doing something special to make the honey into “medicine”.