Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Honey for wounds

Case series in venous leg ulcers [2]
Case report of a diabetic foot ulcer [3]

In 2001 Bandolier online featured a systematic review [1] of honey and wounds, together with a follow up including two more randomised trials. These have prompted more than their fair share of attention through emails and other contacts, so a brief re-visit makes sense. Disappointingly, there is little more of note to report.

Honey has been used to treat infected wounds and burns. There is logic to this because it is hyperosmolar, and because it contains specific antimicrobial substances. Trials have been randomised, but usually of low reporting quality otherwise, and predominantly from a single researcher. Healing of burns at seven days was higher with honey (48%) than active controls (16%, Figure 1), with an NNT of 3.1 (2.5 to 4.2).

Figure 1: Healing of burn wounds at seven days with honey compared with active controls

Case series in venous leg ulcers [2]

A case series without controls recruited 40 patients with predominantly venous leg ulcers which had failed to respond to a minimum of 12 weeks of appropriate treatment with compression bandaging following appropriate guidelines. There were also sensible exclusions, like concomitant use of antibiotics, ulcers larger than 18 x 8 cm, an ulcer present for more than two years, or an inability of carers to change dressings.

Treatment was with medihoney (an Australian antibacterial honey) spread to a depth of about 3 mm and to the shape of the ulcer, with compression bandage, though how often wasn't stated. Pain, healing, wound size, and odour were monitored over a further 12 weeks.

Table 1 shows some of the results. There were 13 withdrawals, eight because of increased pain or worsening ulcer. On the other hand, 20 patients who had pain initially had their pain much reduced. Pain appeared to be related to the size of the leg ulcer.

Table 1: Results of a case series of 40 patients with predominantly venous leg ulcers, unhealed after 12 weeks of compression bandaging using standard guidelines

Age 54-96 years, 90% venous ulcers, average size 10 sq cm, range 0.2 to 61 sq cm after 12 weeks of compression therapy according to local and national guidelines
Increasing ulcer pain (6), poor health (3), deteriorating ulcer (2), death (1), not treated to protocol (1)
Pain decreased
From average initial of discomfort (moderate pain) to almost no pain, and average wound area from 8 to 4 sq cm
No pain initially
Average wound area decreased from 3 to 2.4 sq cm
Pain increased (including dropouts)
Pain increased from slight/discomfort to discomfort/distressing. Average wound area increased from 12 to 13 sq cm

Of the 27 patients who remained in the study, seven had wounds healed by 12 weeks, and 20 had a reduction in wound size. Six, though, had also received oral antibiotics. Odour was present in 26 patients, and odour scores dropped during the study, from an average mild/moderate odour to average no odour/mild odour. Only seven of 40 patients were less satisfied with the honey than with previous treatment.

Case report of a diabetic foot ulcer [3]

Here a 79-year old man had diabetic heel and forefoot ulcers, which were treated over 14 months, including five hospital admissions and four operations (with loss of several toes). The ulcers measured 8 x 5 and 3 x 3 cm, and were infected with MRSA, VRE (vancomycin-resistant Enterococcus), and Pseudomonas. The cost of his ulcer treatment over the 14 months was US$390,000.

Refusing amputation, the patient was discharged home. The wounds were smeared with ordinary honey from a supermarket, and then wrapped. No oral antibiotics were used.

Within two weeks granulation tissue appeared, and within a year the ulcers were healed. Two years later the man was ambulatory with a walker.


The importance of case reports in relation to adverse events with drug has recently been eloquently emphasised [4]. Case reports with treatments can also be compelling. Here we have two forms of anecdotal evidence, which can be placed in the scales.

The case series was a non-comparative trial on patients who had failed to respond to conventional therapy, but where a majority appeared to respond well to honey. After 12 weeks of honey they had reduced pain, reduced odour, and ulcers that began to heal or had healed. We cannot know whether 24 weeks of compression therapy would have done as well as 12 weeks of compression plus 12 weeks of compression and honey, and therein is the weakness of a case series.

Perhaps the case report is more convincing. This patient had everything stacked against him. A long period of ineffective treatment, ulcers infected with some very nasty resistant bugs, and the best medical advice urging amputation. Honey helped within a couple of weeks. Here causation seems plausible.

We have biological plausibility that honey should be effective. It comes from knowledge that its hyperosmolar nature and antibiotics should help. Randomised trials show that it does help, even though those trials had weaknesses. Case series and reports add to evidence.

What is disappointing is that in the four years or so since the systematic review, no major randomised trials have been published to help us decide who to treat, with what, when, and for what condition. This is curious when there is an apparently relatively simple, low-tech solution to some frightening, costly problems. Our health service and academic institutions seem too sclerotic to organise effective clinical research.


  1. OA Moore et al. Systematic review of the use of honey as a wound dressing. BMC Complementary and Alternative Medicine 2001 1: 2. (
  2. CE Dunford, R Hanano. Acceptability to patients of a honey dressing for non-healing venous leg ulcers. Journal of Wound Care 2004 13: 1-7.
  3. JJ Eddy, MD Gideonson. Topical honey for diabetic foot ulcers. Journal of Family Practice 2005 54: 533-536.
  4. JK Aronson. Unity from diversity: the evidential use of anecdotal reports of adverse drug reactions and interactions. Journal of Evaluation in Clinical Practice 2005 11: 195-208.

previous or next story