Antibacterial Honey: in-vitro Activity Against Clinical Isolates of MRSA, VRE, and Other Multiresistant Gram-negative Organisms

Narelle May George; Keith F. Cutting

Disclosure: Mr. Cutting has acted as a paid consultancy for Medihoney and Dermasciences.

The media regularly reminds both the public and healthcare professionals of the dangers infection poses to good health, in particular the difficulties in successfully treating infection caused by multiresistant microorganisms.The development of bacterial resistance to antibiotic therapy is understood to be a natural occurrence. The emergence of resistant strains of bacteria and the ensuing management challenges are compounded by the fact that the development of new antibiotics has decreased in recent years.1 This situation prompts revisiting traditional approaches to infection management and use of those antimicrobials where the emergence of resistant strains has not been demonstrated and is highly unlikely to occur.

More recently, interest in honey as a therapeutic agent has undergone a renaissance. Molan,2 in a review article on honey used as a wound dressing, eloquently presented an array of supportive evidence ranging from case studies to randomized controlled trials that clearly indicates the value of honey in wound care—particularly its antibacterial activity.Molan concludes that the antibacterial activity of honey “rapidly clears infection and protects wounds from becoming infected.”2 This statement brings into sharp belief the antibacterial potency of honey and its value as a therapeutic agent in wound care. This notion is supported in a 2005 report the Australian government commissioned that states “honey has been successfully used on infections not responding to standard antiseptic and antibiotic therapy” and that the full potential of honey will be recognized as the number of antibiotic resistant bacteria increases.3

The antibacterial activity of honey has been related to 4 properties (Figure 1).


Honey appears to offer distinct advantages over “traditional” antibiotic therapy. Nonetheless, it is important to remember that although natural honey from the comb is antibacterial, it is not medical grade and should not be used in wound care. Medical grade honey is filtered, gamma irradiated, and produced under exacting standards of hygiene.

All honeys are not the same and do not possess the same therapeutic advantages; therefore, honey should not be considered as a generic term.6 Medihoney Antibacterial Honey (Medihoney Pty LTD, Richlands, Australia) is a standardized medical honey that is available in many countries including Australia, United Kingdom, Finland, Germany, Austria, and Turkey. It is selected for its antibacterial activity and predominantly sourced from Leptospermum species. Sterility of products is validated against international standards and products are manufactured to meet international quality system requirements. The antibacterial activity of Medihoney is validated for the shelf life of the product, complying with the European Medical Device Directive. The Maori (Polynesian settlers of New Zealand) vernacular name for Leptospermum honey is manuka, the name by which it is more popularly known.

Although the antibacterial activity of honey is recognized, potency varies between types.7 Relevant microbiological data is required in order to better understand the antibacterial activity of specific types of medical honey—particularly its impact on resistant bacterial strains.

Aim. An in-vitro study was initiated in order to gain insight into the antibacterial activity of Medihoney antibacterial honey against a range of multiresistant organisms. Methods A challenge set of 130 clinical isolates with multiple antibiotic resistances was prepared (Figure 2).


A challenge set of 130 clinical isolates with multiple antibiotic resistances was prepared (Figure 2).

The clinical source and antimicrobial phenotype of test strains is shown in the Appendix.


Test strains were nonreplicate, nonclonal clinical isolates that were cultured from diagnostic specimens submitted to a large tertiary referral hospital in Australia over a 14-year period (1990–2004).All organisms were identified using standard methods in accordance with those outlined in the Manual of Clinical Microbiology.8

Antimicrobial susceptibility profiles for all staphylococcal and gram-negative organisms were determined using the automated Vitek system version R07.1 (bioMerieux, Marcy l'Étoile, France). Enterococcal resistance profiles were determined using CLSI agar dilution protocol.9 Vancomycin-resistant phenotypes were confirmed using genotyping and the species identification confirmed using polymerase chain reaction (PCR) of specific Ddl ligases.10 Clonality of test strains was assessed from pulsed field gel electrophoresis profiles obtained using the GenePath system (Bio-Rad Laboratories, Hercules, Calif).

Three ATCC type strains were also tested— Staphylococcus aureus ATCC 25923, Pseudomonas aeruginosa ATCC 27853, and Enterococcus faecalis ATCC 29212.

Plate preparation and inoculum. The antibacterial honey was serially diluted from 1%–20% v/v in Mueller-Hinton agar (BBL211438). Control plates containing only Mueller-Hinton agar were also prepared. Plates were inoculated on the same day as preparation. Test organisms were subcultured from the -70°C freezer onto 5% horse blood Columbia agar base (Oxoid CM 331) andincubated at 35°C in ambient air for 18 h. After this time, a second subculture onto horse blood Columbia agar was performed. After 18 h of incubation at 35°C, 4–5 colonies of each test isolate were inoculated in 2 mL Tryptone Soy Broth (Oxoid CM 129) and incubated at 35°C in ambient air for 2 h. Following incubation, the turbidity of each culture was adjusted to a 0.5 McFarland standard. Ten microliters of the adjusted suspension was then added to 500 uL of sterile physiological saline in a multipoint inoculator.

Mueller-Hinton agar plates containing serial dilutions of Antibacterial honey were inoculated using a multipoint inoculation device that delivered an inoculum per plate of approximately 104 cfu/mL.Agar plates were incubated at 35°C in ambient air for 18 h. After this time, each plate was examined for the presence of bacterial growth. Complete inhibition of bacterial growth was recorded as “no growth.” Media showing partial inhibition (shadowing) or 1–2 colonies of test isolates were reported as “positive for growth.”


All test isolates grew on the control Mueller-Hinton (control) agar plates that contained no antibacterial honey. The MIC data for the 3 ATCC type strains were:4% v/v for S aureus ATCC25923, 8% v/v for P aeruginosa ATCC 27853, and 8% v/v for E faecalis ATCC 29212. All test strains of MRSA were inhibited at a concentration of 4% v/v irrespective of antimicrobial phenotype.

With the gram-negative organisms, 8/10 ESBL producing strains of Escherichia coli (80%) were inhibited at 6% v/v. The remaining 2 strains (20%) were inhibited at 8% v/v. Similarly for Klebsiella species, 11/12 strains (91%) were inhibited at 6% v/v, while the remaining strain was only inhibited at 8% v/v. For Enterobacter species, all 6 test strains (100%) were inhibited at 6% v/v. For Acinetobacter baumannii, concentrations of 8% were required to inhibit 9/11 (82%) of test isolates. All 5 panresistant strains of A baumannii were inhibited at 8% concentrations of the antibacterial honey.

Vancomycin-resistant enterococci, the more resistant VanA/B strains of E faecalis and Enterococcus faecium required 8% for inhibition compared to 6% for the less resistant isolates. Two strains of vancomycin-resistant E faecalis and E faecium were inhibited at 6%, while 4 and 9 strains required a higher concentration of 8% v/v. All 3 strains expressing the VanC phenotype were inhibited at 6% v/v as were 2/3 strains of vancomycin susceptible E faecalis.

Clinical isolates of P aeruginosa were more resistant to the antibacterial honey than other bacterial species. Concentrations of 14% v/v were required to inhibit 17/20 (85%) of test isolates.The remaining 3 strains were inhibited at a lower concentration of 12%.


The above results and comparative MIC50 and MIC90 values for the different organism types are listed in Table 1.


The findings of the present study add to the body of evidence and clearly demonstrate that honey has a valuable therapeutic role to play in wound care, often where modern approaches have failed.2

There are a number of in-vitro studies that have shown the effectiveness of medical honeys on antibiotic resistant organisms, such as MRSA and P aeruginosa.5,11

Although there is variability between MICs reported for similar organisms using nonstandardized honeys with different bacterial potencies, there was generally good correlation between the present findings and those of other researchers11,12 for gram-positive organisms (MRSA, VRE). In the present study, all strains of MRSA, including both resistant phenotypes of MRSA as well as sensitive strains of S aureus, were inhibited at low antibacterial honey concentrations (4% v/v). These data compare favorably with the earlier data from George et al13 comparing medical honeys obtained from New Zealand and Australia Leptospermum sources and those of Cooper et al14 (3% v/v), Blair5 (4.3% v/v), and Allen et al12 (3%–7% v/v).

For VRE, the inhibition profiles of antibacterial honey varied between 6%–8% v/v depending on the species. Favorable comparisons can again be made with 3.8%–10% v/v ranges reported by Cooper et al11 and Allen et al.12

There is a paucity of good comparative data for in-vitro assessment of antibacterial activity of medical honeys against gram-negative organisms other than P aeruginosa. Gram-positive aerobic cocci, (eg, beta-hemolytic streptococci and S aureus) may be considered as primary pathogens often infecting chronic wounds in the earlier stages of wound formation. With delayed healing, wounds are more likely to be colonized by gram-negative coliforms, Pseudomonas species, and anaerobic bacteria. Infections in wounds of longer duration may be considered to be polymicrobial (aerobes and anaerobes). For the multiresistant gram-negative organisms other than P aeruginosa investigated in the present study, the majority of the 3 different species tested were inhibited at antibacterial honey concentrations of 6% v/v with all strains inhibited by 8% v/v. Likewise, antibacterial honey concentrations of 8% v/v were required to inhibit the more resistant Acinetobacter strains. Given the high levels of antibiotic resistance demonstrated in this group of organisms, the low MIC90 values clearly suggest that antibacterial honey has the potential to be an effective alternative antibacterial agent in vivo even with up to a 10- fold dilution of the preparation.

Surprisingly, inhibitory concentrations in the order of 12%–14% v/v for P aeruginosa with antibacterial honey were found. This contrasts with the earlier work of George et al13 where Australian and New Zealand Leptospermum honeys were shown to effectively inhibit the growth of more than 100 clinical strains of this organism at concentrations of 5%–6% v/v. Similar potency against P aeruginosa (4%–9% v/v) has been reported by Cooper et al.14 The findings of the present study highlight the variability of the antibacterial potency of different honeys. Medihoney combines honeys of differing antibacterial actions. The wide range of MICs reported when comparing different honeys against the same class of microorganism illustrate the differences in antibacterial potency that may be encountered between honeys.15 This underlines the value of using a standardized medical grade honey preparation that demonstrates consistent antibacterial activity against a broad range of microorganisms.

In an in-vitro study comparing the antibacterial activity of 13 honeys including 3 commercial honeys of manuka, Medihoney Antibacterial Honey, and Rewarewa against E coli and P aeruginosa, only Medihoney and one beekeeper honey demonstrated inhibition of both organisms at 2.5% wt/v dilution.16

Cooper et al11 investigated the sensitivity of gram-positive cocci to honey. Eighteen strains of MRSA, 7 strains of vancomycin sensitive enterococci isolated from wounds, and 17 strains of vancomycin resistant enterococci were isolated from hospital environmental surfaces. The mean MIC values of manuka and pasture honey (with peroxide activity) were 3.0% and 3.1%, respectively, for MRSA strains. Mean MICs of honey for vancomycin-sensitive enterococci isolated from infected wounds were 4.9% (manuka) and 9.7% (pasture). Similar values were recorded for vancomycin-resistant enterococci— mean MICs for manuka and pasture were 4.6% and 8.3% (v/v), respectively.


The findings of the present study in respect to the antibacterial activity of honey are not unique as they complement previous findings. Although variations in bacterial potency are recognized, the presence of MRSA and other multiresistant bacteria that may be found in wounds causes much concern and leads to an increased consumption of available resources.The results are clear and provide additional in-vitro evidence that Medihoney Antibacterial Honey is an effective antibacterial agent, the implication being that Medihoney provides a valuable opportunity to manage wound infection caused by a range of multiresistant strains of bacteria. In-vitro activity requires corroborative evidence from rigorously conducted studies if claims are to be substantiated. The recent clinical findings of Johnson et al17 and Simon et al18 provide much encouragement that these claims will be confirmed. The Johnson et al17 study also indicates that Medihoney may have an important role to play in infection prophylaxis as it demonstrated that Medihoney was equally effective as mupirocin in the prevention of catheter associated infections. Given that mupirocin can reduce infection rates by at least 7–13 fold,19,20 the prospect that Medihoney will prove to be an effective prophylactic is extremely hopeful.

Back to top