Is honey an effective remedy for symptoms of upper respiratory tract infections?

What is the effect of honey on upper respiratory tract infection (URTI) symptoms?

URTI is an umbrella term used to describe infections of the upper respiratory tract (which includes the nose, sinuses, and throat) that commonly involve symptoms like coughing, a sore throat, nasal congestion, runny nose, and sneezing.

Most URTIs are caused by viruses, as opposed to bacteria, which means that treatment with antibiotics is ineffective. As such, usual care for URTIs focuses primarily on temporary symptom relief through the use of medications like decongestants, antihistamines, analgesics, and antitussives (i.e., cough suppressants). Beyond these conventional therapies, alternative and complementary^[1] therapies are also being explored. The results of one review article evaluating some of these therapies’ efficacy for one major URTI, the common cold, is laid out in Figure 1. Another such therapy is honey.

Reference: Allan et al. CMAJ. 2014 Feb.^[4]

Although honey has a long history of use as a home remedy to treat symptoms of URTIs, there is currently little scientific evidence to support its use. While a 2018 Cochrane systematic review^[5] found some evidence that honey may relieve cough in children, the authors of the review rated the available evidence as low to moderate quality, and concluded that there was no strong evidence for or against using honey. However, the effect of honey on other URTI symptoms and in adult populations was not examined. As such, the goal of the study under review was to conduct a meta-analysis of the effects of honey on URTI symptoms in people of all age groups.

While usual care for URTIs focuses primarily on temporary symptom relief through the use of medications, alternative and complementary therapies are also being explored. One such therapy is honey. The study under review is a meta-analysis that investigated the effects of honey on upper respiratory tract infection (URTI) symptoms in people of all age groups.

The authors conducted a systematic review and meta-analysis of randomized controlled trials that looked at the effects of honey on URTI symptoms in people of all age groups with clinically or laboratory-diagnosed URTI symptoms. The honey could be of any type, and administered in any way, alone or together with other treatments, and the control group could be placebo, no treatment, or usual therapy.

Ultimately, 12 trials were included in the meta-analysis, of which eight were in children or adolescents, and four were in adults. Eight trials used pure honey, three trials used syrups containing honey and plant complexes or herb extracts, one trial used honey mixed with milk, and one trial used honey mixed with coffee. The reason these trials add up to 13 is that one trial included two treatments: a honey-only treatment and a honey plus coffee treatment. The types of comparators included placebo and usual care, the latter including diphenhydramine, dextromethorphan, carbocysteine, combinations of different medications, and, strangely, coffee.

Assessed outcomes included cough frequency and severity, and combined symptom score. Separate meta-analyses were performed for different comparators (i.e., placebo and usual care). Outcome data were summarized in a random-effects model.

The researchers tested heterogeneity using the I² statistic, and assessed the risk of bias of the included trials with the Cochrane Collaboration’s risk of bias assessment tool for RCTs, where bias was assessed as high, low, or unclear over six domains: selection, performance, detection, attrition, reporting, and other. Funnel plots to assess publication bias could not be used due to the small number of studies. The meta analysis was preregistered, and followed PRISMA guidelines.

This was a systematic review and meta-analysis of 12 randomized controlled trials that examined the effects of honey (of any type, and administered in any way, alone or together with other treatments) as compared to placebo, no treatment, or usual care, on URTI symptoms in people of all age groups.

When compared to usual care, honey improved cough frequency (n=8, SMD=-0.36), cough severity (n=5, SMD=-0.44), and combined symptom scores (n=3, MD=-3.96), with low heterogeneity detected (0% for cough frequency and combined symptom scores). Trials included in these analyses had a variable or high risk of bias.

In the analysis comparing honey to placebo, honey did not have a statistically significant improvement on combined symptom score, although it was close (n=2, SMD=-0.63, 95% CI: -1.44 to 0.18). There were high levels of heterogeneity detected between the two trials as well, but both had a low risk of bias. These possibly positive results are summarized in Figure 2.

All of the results above (i.e., for honey vs. usual care, and honey vs. placebo) remained largely unchanged in the analyses that excluded interventions that used syrups, and honey mixed with milk or coffee.

In the one adult-only analysis (n=4) that could be performed, honey was not better than usual care for improving cough frequency.

The findings of this meta-analysis indicate that honey improves cough frequency, cough severity, and combined symptom scores when compared to usual care. However, most trials included in these analyses were of variable or high risk of bias. Also, only a handful of trials were included in the analyses, with four trials performed in adults, and just two comparing honey to placebo.

Some strengths of this meta-analysis are that it was preregistered and PRISMA-compliant, it employed a comprehensive search strategy without language or date restrictions, and that it assessed risk of bias with the Cochrane Collaboration’s risk of bias assessment tool. The fact that the authors used a random-effects model (which is more appropriate for trials that aren’t direct replications of one another) to summarize the data, even in analyses in which heterogeneity^[6] was low, was also a plus. There are, however, some limitations worth addressing.

First, eight of the 12 trials included in the meta-analysis were assessed as being of unclear or high risk of bias in at least three out of the six domains of the risk of bias assessment tool. Specifically, in the individual analyses comparing honey to usual care, there was roughly an equal mix of trials of a low and high risk of bias for the outcome of cough frequency, while for cough severity and combined symptom score, most trials were of a high risk of bias. This reduces confidence in these results. On the other hand, in the analysis comparing honey to placebo for combined symptom score, the trials included were assessed as being of low risk of bias.

With the above said, the risk of bias in some of the trials may have been incorrectly assessed. For example, although two^[7] trials by the same research group were assessed as having a low risk of bias for random sequence generation, (i.e., the process of randomly allocating participants to a group) these trials did not sufficiently describe the randomization process. This means that they should have been assessed as having an unclear risk of bias for this domain, and suggests there should be some skepticism about the overall results of the risk of bias analysis.

Second, most of the analyses only included a small number of trials. Specifically, in the analyses comparing honey to usual care, there were five trials for cough severity, and only three trials for combined symptom score. Moreover, just two trials were included in the analysis comparing honey to placebo, while four trials were included in the analysis restricted to adults. Overall, the small number of trials in most of the analyses decreases confidence in the results.

Third, and relevant to what was said above: although the researchers intended to use funnel plots to assess publication bias, due to the small number of trials included in the meta-analysis, they ended up not doing so. Considering that trials with positive findings are more likely^[8] to be submitted and published than trials with negative or null findings, the absence of a publication bias assessment reduces confidence in the results of the meta-analysis.

Fourth, the relatively broad eligibility criteria resulted in considerable variability in some aspects of the trials. For example, the treatments included several different types of honey, combinations of honey with other ingredients, and administration methods. Also, there were several different comparators in the “usual care” analyses, with coffee strangely considered as a type of usual care. Despite this, the results remained largely unchanged in the analyses that excluded the interventions that did not use pure honey, which is reassuring, as is the finding that there were low or zero levels of heterogeneity in most of the analyses. However, it should be noted that it is difficult to get a sufficient estimation of heterogeneity in meta-analyses with a small number of trials, such as the one under review. This, together with the considerable variability in some methodological aspects of the trials, suggests some skepticism is warranted related to the levels of heterogeneity detected.

Although the researchers tried to include trials with a variety of URTI symptoms, most trials only measured cough frequency and severity. Even the combined symptom score outcome only combined cough-related measures. This means that, although the initial intention of the researchers was to assess the effect of honey on several URTI symptoms, the results of the meta-analysis only relate to coughing, and cannot be generalized to all URTI symptoms.

Finally, it’s worth pointing out that, while the results using placebo as a comparator suggest that honey is not effective for improving URTI symptoms, it could be that honey “works” because of its texture (i.e., by forming a soothing mechanical barrier on the throat lining), and the placebos worked equally well because of their similarity in texture to honey. However, this is merely speculation. Also, only two trials were included in the honey vs. placebo analysis, which decreases confidence in these results.

The results of this meta-analysis suggest that, compared to usual therapy, honey improves cough frequency and severity in children and adolescents with URTIs. However, these results are mostly based on a handful of trials, most of which are at a high risk of bias.

This meta-analysis suggests that, in children and adolescents with clinically or laboratory-diagnosed URTI symptoms, honey, as compared to usual care, improves cough frequency and severity. However, these results are based on only a handful of trials, most of which were found to be at a high risk of bias, and, as such, should be considered with skepticism.