JOBNAME: corn 25#9 2006 PAGE: 1 OUTPUT: Tuesday November 14 13:09:41 2006
lww/corn/129949/ICO200355
CLINICAL SCIENCE
Effect of Antibacterial Honey on the Ocular Flora in Tear
Deficiency and Meibomian Gland Disease
Julie M. Albietz, PhD* and Lee M. Lenton, FRANZCO, FRACS†
Purpose: To assess for differences in the ocular flora of patients with
dry eye caused by tear deficiency and/or meibomian gland disease and
to assess the effect of antibacterial honey on the ocular flora in these
forms of dry eye.
Methods: In this prospective, open-label pilot study, bacteria isolated
from the eyelid margin and conjunctiva were identified and quantified
before and at 1 and 3 months after initiation of treatment with topical
application of antibacterial honey 3 times daily. Subjects had non-
Sjogren tear deficiency (n = 20), Sjogren syndrome tear deficiency (n =
11), meibomian gland disease (n = 15), and non-Sjogren tear defi-
ciency with meibomian gland disease (n = 20), and there were 18 non–
dry eye subjects.
Results: The total colony-forming units (CFUs) isolated from each
of the dry eye subgroups before antibacterial honey use was signifi-
cantly greater than the total CFU isolated from the non–dry eye
group. Antibacterial honey use significantly reduced total CFUs for
the eyelids and the conjunctiva of dry eye subjects from baseline at
month 1 (eyelids: P = 0.0177, conjunctiva: P = 0.0022) and month 3
(eyelids: P , 0.0001, conjunctiva: P , 0.0001). At month 3, there
were reductions in total CFUs for all dry eye subgroups such that the
CFUs were not significantly different from those of the non–dry eye
group.
Conclusion: From these results, there is sufficient preliminary data
to warrant further study of the effects of antibacterial honey in chronic
ocular surface diseases.
Key Words: dry eye, antibacterial honey, aqueous tear deficiency,
meibomian gland disease
(Cornea 2006;25:1012–1019)
Received for publication June 16, 2005; revision received January 3, 2006;
accepted for publication March 25, 2006.
From the *Institute of Health and Biomedical Innovation, Queensland
University of Technology, Brisbane, Australia; and †Private Practice,
Brisbane, Australia.
Supported in part by Medihoney Pty Ltd, Richlands, Queensland, Australia for
the microbiology, cytology, and immunocytology pathology tests.
Medihoney Pty Ltd did not influence the study design, data collection,
data analysis, or the decision to publish.
Disclaimer: The authors have no financial relationship with Medihoney Pty
Ltd. Pathology testing was outsourced to scientists at an independent
pathology laboratory Sullivan and Nicholaides Pathologists Pty, Ltd.,
Taringa, Queensland.
Reprints: Julie Albietz, River City, Level 2, 401 Milton Rd., Auchenflower,
Australia 4066 (e-mail: julie@darkoptics.com.au).
Copyright Ó 2006 by Lippincott Williams & Wilkins
1012
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
A traditional therapy for a wide range of ailments, honey has
recently been recognized in conventional medicine
because of its antibacterial properties and ability to promote
effective wound healing.1,2 On the ocular surface, honey has
been used in the treatment of corneal edema, blepharitis,
conjunctivitis, and keratitis.3–6 Previous reports have inclu-
ded case series, clinical observations, and animal models, but to
date, there have been no controlled clinical studies in human
eyes to assess the benefits of honey on the ocular surface.
There is some evidence that patients with tear deficiency
and meibomian gland disease have an overgrowth of ocular
flora.7–9 This overgrowth is thought to contribute to the tear
film instability and ocular surface damage7,8 and the increased
risk of contact lens–related keratitis,10 postoperative microbial
keratitis,11 and endophthalmitis12 associated with these
conditions.
The emergence of multiple antibiotic–resistant micro-
organisms has generated renewed interest in natural pro-
ducts with antibacterial activity such as standardized medical
honeys.1–3 MedihoneyTM Antibacterial Honey (Medihoney Pty
Ltd, Richlands, Queensland, Australia) is made from a pro-
prietary mix of honeys, including honeys selected from the
Australian and New Zealand Leptospermum sp. to ensure a
broad spectrum of antibacterial activity.13,14
This study aimed to characterize the ocular flora in
patients with dry eye caused by tear deficiency and/or
meibomian gland disease and assess the effect of topical
application of MedihoneyTM Antibacterial Honey on the
ocular flora in these dry eye groups.
MATERIALS AND METHODS
This report describes a prospective, open-label, pilot
study. All participants received a detailed explanation of the
procedures involved and provided written informed consent.
Written approval for the studies was obtained from the
Queensland University of Technology Biomedical Ethics
Committee (ref. 3306H) and from the Australian Govern-
ment’s Therapeutic Goods Administration under the Clinical
Trial Notification Scheme (trial 2004/31).
Subjects
The study involved 66 subjects with dry eye and 18 non–
dry eye subjects. Patients were recruited from consecutive
patients who attended a private ophthalmology practice in
Brisbane, Australia, and met the inclusion criteria. All subjects
were required to have not worn contact lenses for at least 3
months before entering the study and throughout the study,
Cornea  Volume 25, Number 9, October 2006
JOBNAME: corn 25#9 2006 PAGE: 2 OUTPUT: Tuesday November 14 13:09:42 2006
lww/corn/129949/ICO200355
Cornea  Volume 25, Number 9, October 2006
because contact lens wear can influence the ocular surface
flora.15 Other general subject requirements were: no eye surgery
within the prior 3 months, no active ocular infection, no eye
drops other than nonpreserved lubricants for a least 3 months
before starting the trial and during the trial; no punctum
occlusion for at least 3 months before starting the trial and no
change to punctum occlusion status during the trial; not
pregnant or breast feeding; no change to prescribed systemic
medications that could influence tear secretion, blink function,
or ocular surface health during the trial; and no uncontrolled
systemic disease or significant illness during the trial. Data
collection occurred from late January to late April to avoid
seasonal variations that can influence ocular microbial flora.16
During these warmer months, there is very little change in the
humidity and temperature in Brisbane.
From the guidelines of the National Eye Institute (NEI)
for clinical trials in dry eye,17 the condition was defined
as follows: 1 or more McMonnies Dry Eye Symptom Survey18
primary dry eye symptoms (burn, grittiness, dryness, scratch-
iness, foreign body sensation) experienced ‘‘often’’ or ‘‘con-
stantly’’; fluorescein tear breakup time (FBUT) less than 10
seconds in 1 or both eyes (average of 3 readings); and a com-
bined ocular surface fluorescein and lissamine green staining
score of at least 3 in 1 or both eyes by using the Oxford
method.19 The cornea (fluorescein stain) and the nasal,
temporal, superior, and inferior bulbar conjunctiva (lissamine
green) were each graded separately (a range of 0–5 for each
zone) and combined for a total Oxford score (range of grades
0–25). These staining criteria was more relaxed than NEI
guidelines17 but enabled a broader range of dry eye severity in
the subject population, which was desirable given that this was
the first study examining the ophthalmic use of MedihoneyTM
Antibacterial Honey and was designed to act as a pilot study for
further studies.
Dry eye subjects were subclassified as having non-Sjogren
tear deficiency (NSTD), Sjogren syndrome tear deficiency
(SSTD), meibomian gland disease (MGD), or both non-Sjogren
tear deficiency and meibomian gland disease (NSTD+MGD).
NSTD (n = 20) was defined with a Schirmer 1 test value
(without anesthetic) of no more than 5 mm/5 min.
SSTD (n = 11) was defined as NSTD with an established
diagnosis of Sjogren syndrome by a physician including
symptoms of xerostomia and the presence of at least 1 of the
following serum autoantibodies: antinuclear antibody, rheu-
matoid factor, and Sjogren antibodies class SS-A and SS-B.17
MGD (n = 15) was diagnosed using slit-lamp biomicro-
scopy to assess for features of the disease: posterior eyelid
margin thickening, irregularity, and telangiectasia; meibomian
gland loss, capping, and plugging; and a particulate, cloudy,
yellow or frothy meibomian gland secretion.20
MGD+NSTD (n = 20) subjects met the diagnostic
criteria for both NSTD and MGD.
TABLE 1. Baseline Demographics of Dry Eye Subgroups and Non–Dry Eye Group
Group MGD (n = 15) NSTD (n = 20)
Age (y; mean 6 SD) 59 6 17 64 6 15
Female (%) 60 85
q 2006 Lippincott Williams & Wilkins
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Antibacterial Honey and Ocular Flora
Potential subjects who met the dry eye diagnostic criteria
but had predominant features of other forms of dry eye such as
lid surfacing or blink anomalies, primary epitheliopathies, or
cicatricial disease were excluded from this study.
The non–dry eye group (n = 18) was required bilaterally
to have an absence of McMonnies dry eye symptoms, tear film
stability of more than 10 seconds (average of 3 values), and an
absence of ocular surface staining, Schirmer 1 test (without
anesthetic) of more than 10 mm/5 min, and absence of
meibomian gland anomalies.
Dry eye subgroups and the non–dry eye group were
matched for age but not for sex, given that SSTD and NSTD
affect predominantly women,17,18 and age- but not sex-specific
differences in ocular flora have previously been shown.21
Subject group demographics at baseline are presented in Table 1.
Assessments
This study involved descriptive and analytical compo-
nents. Both the type and number of microbial colonies were
evaluated. Swabs were taken at baseline for all subjects and at
months 1 and 3 for the dry eye subjects. Swabs were taken only
from 1 eye. In the non–dry eye subjects, this was the right eye,
and in the dry eye subjects, it was the eye with the greatest total
ocular surface staining score. One examiner (J.A.) performed
all assessments.
Swabs were performed using sterile cotton-tipped
applicators moistened with saline. With the subject looking
up, the applicator was wiped twice along the conjunctival
fornix from the inner to the outer fornix, and using a separate
swab, twice along the lid margin from the inner to the outer
canthus. The swabbing was performed using a firm, even pres-
sure. The conjunctiva was swabbed before the lids to avoid spill-
age from the eyelid margin into the conjunctival sac. Swabs
were placed immediately in transport media and processed
within 6 hours. Processing involved placing swabs in 0.5 mL
of sterile saline and rotating 10 times by the rolling swab
handle between the forefinger and the thumb. Using aseptic
technique, 50 mL of swab-saline solution (post–washing of the
swab) was used to inoculate each culture plate. The following
media were inoculated: horse blood agar (isolation of aerobic
bacteria), chocolate agar (fastidious aerobic and capnophilic
bacteria), MacConkey agar/gram-positive Cocci agar (coliform,
staphylococcal, streptococcal selective media), and anaerobic
agar (Propionibacterium sp.). All culture plates (except
anaerobic agar) were incubated for 48 hours and examined
after 24- and 48-hour incubation. Anaerobic culture plates
were incubated and examined after 72-hour incubation. In all
cases, incubation occurred at 35°C.
The number of colony-forming units (CFUs) was
determined for each culture sample. Samples with ‘‘too many
colonies to count’’ or those with confluent growth were defined
to have 101 CFUs. Specific CFUs were determined for those
SSTD (n = 11) MGD+NSTD (n = 20) Non–Dry Eye (n = 18)
59 6 14 55 6 16 57 6 17
91 55 67
1013
JOBNAME: corn 25#9 2006 PAGE: 3 OUTPUT: Tuesday November 14 13:09:44 2006
lww/corn/129949/ICO200355
Albietz and Lenton
with no more than 100 CFUs. Attempts were made to identify
all the different colonies to at least the genus level. The
microbiologist performing the cultures and sensitivities was
masked to the underlying diagnosis and treatment status of the
subject. The total CFUs for the lid swabs and the total CFUs
for the conjunctival swabs were determined for each subject at
the various time points by adding the CFUs for all types of
bacteria isolated. For example, if a subject at baseline had a lid
swab that cultured 69 CFUs of Staphylococcus aureus, 7 CFUs
of coagulase-negative staphylococcus, and more than 100
CFUs of Coliform sp., the subject’s total lid CFUs at baseline
was estimated as 69 + 7 + 101 = 177.
Study Treatments
After baseline assessments, the dry eye subjects were
instructed to apply Medihoney ophthalmic honey 3 times daily
to the inferior conjunctival fornix and to the lower lid margins
and immediately close the eyes for several minutes. Medihoney
ophthalmic honey contained 100% Medihoney antibacterial
honey. The product was sterilized by gamma irradiation to
destroy spore-forming organisms that may be present in the
honey without loss of antibacterial activity.22 The MedihoneyTM
Antibacterial Honey was subjected to quality-control proce-
dures to ensure a high level of antibacterial activity against
S. aureus (American Type Culture Collection 9144) that is at
least equivalent to an 18% solution of the reference standard
antiseptic, phenol. Treatment with the honey was for 3 months,
and subjects were not permitted to alter any additional topical
treatments for their dry eye condition from 1 month before the
baseline visit. For example, if daily lid hygiene procedures
and/or an oral flaxseed oil supplement were part of the subject’s
dry eye treatment regimen before the baseline visit, they were
required to continue these practices for the duration of the
study. Dry eye subjects were permitted to use nonpreserved unit
dose artificial tears, as required, for the duration for the study.
Non–dry eye subjects did not use the honey.
Analysis
CFU data were not normally distributed. Comparisons
between total baseline CFUs for the dry eye subgroups and
non–dry eye group were made using Mann–Whitney U tests.
The paired sign test or the Wilcoxon signed rank test was used
to test changes from baseline to month 1 and baseline to month
3 for the total CFUs of the treated dry eye subjects. The P value
for the paired sign test was quoted if significant, indicating the
data distribution was asymmetric; otherwise, the Wilcoxon
signed rank P value was used. The Bonferroni adjustment was
used to account for multiple comparisons (baseline vs. month
1 and baseline vs. month 3) in eyes treated with the honey.
Medians and ranges were recorded in addition to means and
SDs, given that the CFU data were not normally distributed.
The Statview software program (version 5.0.1; SAS Institute,
Cary, NC) was used to perform the analyses. Differences
were considered significant when P , 0.05, except where
Bonferroni adjustments were required, and differences were
considered significant at P , 0.025.
1014
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Cornea  Volume 25, Number 9, October 2006
RESULTS
Participant Flow and Follow-Up
A total of 84 subjects were recruited (66 dry eye subjects
and 18 non–dry eye subjects). All subjects had baseline lid and
conjunctival swabs taken, and all dry eye subjects commenced
the MedihoneyTM Antibacterial Honey , but only 55 dry eye
subjects remained in the study at the month 1 assessment stage
and 49 dry eye subjects remained in the study at the month 3
assessment stage. The 17 discontinuations during the trial were
because of excessive stinging after MedihoneyTM application
(5 of 17), poor compliance with treatment or follow-up
schedule (6 of 17), unrelated personal reasons (3 of 17), and
loss to follow-up (3 of 17). Subject numbers in the dry eye
groups at various time-points are given in Tables 2 and 3.
Baseline Comparison of Ocular Flora in Dry Eye
Subgroups With Non–Dry Eye Group
Quantitative Differences in CFU
All dry eye subgroups had significantly higher total
CFUs than the non–dry eye group at baseline for both the lids
and conjunctiva (Tables 2 and 3). The Sjogren syndrome group
had significantly greater total lid CFU than the other dry eye
subgroups (P = 0.0495). There were no other significant
differences between dry eye subgroups in total lid or
conjunctival CFUs.
Qualitative Differences
Coagulase-negative Staphylococcus was the most
frequently isolated bacteria from the lids and the conjunc-
tiva in the non–dry eye group and each dry eye subgroup.
Corynebacterium sp. was the second most frequently isolated
bacteria from the lids and conjunctiva in all dry eye subgroups
but was isolated from only 1 subject in the non–dry eye group.
Propionibacterium sp. was the third most commonly isolated
bacteria in the dry eye subjects, followed by S. aureus and
coliforms. In the non–dry eye group, Propionibacterium sp.
was isolated in some subjects, but S. aureus and coliforms
were not isolated. Other species of bacteria were occasionally
isolated from the lids and/or conjunctiva. There were no
significant differences between dry eye subgroups in terms of
species and CFUs of bacteria isolated (Tables 4 and 5).
Effect of Medihoney on Total CFUs
Effect on Lids
For the dry eye subjects as a collective group, there were
significant reductions in the total CFUs from baseline at month
1 (P = 0.0177) and month 3 (P , 0.0001). Although there was
a reduction in total CFUs for each dry eye group between
baseline and month 1 (Table 2), these differences were not
significant. At month 3, there were significant reductions in
total lid CFUs compared with baseline for the MGD (P =
0.0150) and MGD+NSTD (P = 0.0159) subgroups. The reduc-
tions in total CFUs at month 3 compared with baseline for the
NSTD (P = 0.0258) and SSTD (P = 0.0287) subgroups were
not significant, given that the Bonferroni adjustment required
P , 0.025 for significance. By month 3, all dry eye subgroups
q 2006 Lippincott Williams & Wilkins
JOBNAME: corn 25#9 2006 PAGE: 4 OUTPUT: Tuesday November 14 13:09:46 2006
lww/corn/129949/ICO200355

Cornea  Volume 25, Number 9, October 2006 Antibacterial Honey and Ocular Flora

TABLE 2. Eyelid Data: Total Colony Counts for Dry Eye Subgroups and Non–Dry Eye Group at Baseline and at Months
1 and 3 After Starting Topical MedihoneyTM Antibacterial Honey
Time Point Total CFUs* Dry Eye† MGD NSTD SSTD MGD+NSTD Non–Dry Eye
Baseline Mean 6 SD 106 6 82 95 6 66 92 6 82 161 6 101 96 6 72 12 6 18
Median 101 101 84 109 101 3
Range 0–400 2–200 1–300 61–400 6–232 0–47
n 66 15 20 11 20 18
P‡ ,0.0001 0.0002 ,0.0001 ,0.0001 ,0.0001 —
1/12 Mean 6 SD 72 6 55 57 6 48 82 6 54 63 6 68 77 6 56 —
Median 95 68 98 23 100 —
Range 1–200 2–119 2–180 1–200 2–195 —
n 55 14 16 10 15 —
P ,0.0001 0.0023 0.0001 0.0088 0.0001 —
3/12 Mean 6 SD 37 6 48 40 6 49 29 6 42 29 6 47 37 6 46 —
Median 7 9 5 8 7 —
Range 0–183 0–150 0–132 0–183 0–104 —
n 49 12 14 10 13 —
P 0.0887 0.1806 0.2215 0.3116 0.1538 —
*Total CFU, sum of colony-forming units of all bacteria cultured for that subject.
†Dry eye, all dry eye subjects.
‡Based on a comparison of total CFUs between dry eye group at that time point and non–dry eye baseline data. P , 0.025 for significance given Bonferroni adjustments were required.

had total lid colony counts not significantly different from the
non–dry eye group (Table 2).
Effect on Conjunctiva
Total CFUs for the conjunctiva in dry eye subjects as a
collective group was reduced significantly at month 1 (P ,
0.0001) and month 3 (P , 0.0001) compared with the total
CFUs at baseline. As for the eyelid data, the total conjunctival
CFU for the each dry eye subgroup reduced between baseline
and month 1 (Table 3), but these individual subgroup
differences were not significant. At month 3, there were
significant reductions in the total CFU compared with baseline
for the NSTD (P = 0.0184), SSTD (P = 0.0125), and
MGD+NSTD (P = 0.0058) subgroups. The MGD total CFUs
also reduced at month 3, but the P value of 0.412 was not
significant, given the required Bonferroni adjustment. By month
3, all dry eye groups had total conjunctival colony counts not
significantly different from the non–dry eye group (Table 3).
Effect of Medihoney on the Most Commonly Isolated
Lid and Conjunctival Bacteria
In the 49 dry eye subjects who completed the 3-month
treatment trial, the effect of the honey on the CFUs of the most
commonly isolated bacterial species was examined. Only
species or groups of bacteria isolated from the lids and/or
conjunctiva at baseline in 10 or more subjects were analyzed

TABLE 3. Conjunctival Data: Total Colony Counts for Dry Eye Subgroups and Non–Dry Eye Group at Baseline and at
Months 1 and 3 After Starting Topical MedihoneyTM Antibacterial Honey
Time Point Total CFU Dry Eye MGD NSTD SSTD MGD+NSTD Non–Dry Eye
Baseline Mean 6 SD 79 6 71 64 6 60 74 6 70 111 6 67 78 6 63 7 6 23
Median 74 48 74 99 94 1
Range 0–300 0–191 0–200 4–300 0–200 0–100
n 66 15 20 11 20 18
P ,0.0001 ,0.0001 0.008 ,0.0001 ,0.0001 —
1/12 Mean 6 SD 45 6 50 28 6 43 58 6 57 33 6 44 54 6 50 —
Median CFU 22 7 39 6 32 —
Range 0–200 0–108 0–200 0–109 0–161 —
n 55 14 16 10 15 —
P 0.008 0.0377 0.005 0.0829 0.0002 —
3/12 Mean 6 SD 19 6 38 21 6 48 18 6 35 25 6 27 14 6 27 —
Median CFU 4 2 4 6 5 —
Range 0–150 0–150 0–100 0–128 0–100 —
n 49 12 14 10 13 —
P 0.0675 0.6852 0.1175 0.1175 0.984 —
See Table 2 for explanation.

q 2006 Lippincott Williams & Wilkins 1015

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
JOBNAME: corn 25#9 2006 PAGE: 5 OUTPUT: Tuesday November 14 13:09:48 2006
lww/corn/129949/ICO200355

Albietz and Lenton Cornea  Volume 25, Number 9, October 2006

TABLE 4. Eyelid Bacteria Isolated in Dry Eye Groups and Non–Dry Eye Group at Baseline
No. Subjects (%) With Organism Isolated
Bacteria Isolated NSTD (N = 20) SSTD (N = 11) MGD (N = 15) MGD+NSTD (N = 20) Non–Dry Eye Group (N = 18)
Coagulase-negative Staphylococcus 18 (90) 8 (73) 13 (87) 19 (95) 15 (84)
Corynebacterium sp. 12 (60) 7 (64) 4 (27) 7 (35) 1 (6)
Propionibacterium sp. 7 (35) 1 (9) 2 (13) 6 (30) 4 (22)
Staphylococcus aureus 3 (15) 3 (27) 4 (27) 4 (20) 1 (6)
Coliforms 2 (10) 4 (36) 2 (13) 4 (20) 0 (0)
Streptococcus viridans 2 (10) 0 (0) 0 (0) 0 (0) 0 (0)
Proteus sp. 0 (0) 0 (0) 0 (0) 2 (10) 0 (0)
Pseudomonas sp. 0 (0) 1 (9) 0 (0) 1 (5) 0 (0)
Serratiamarcescens 0 (0) 0 (0) 0 (0) 1 (5) 1 (6)
Micrococcus sp. 0 (0) 0 (0) 0 (0) 1 (5) 0 (0)
Acinetobacter sp. 0 (0) 0 (0) 1 (7) 0 (0) 0 (0)
None 0 (0) 0 (0) 0 (0) 0 (0) 2 (11)

(Table 6). There were significant reductions in total lid and
conjunctival CFU at months 1 and 3 for coagulase-negative
Staphylococcus and Corynebacterium sp. A significant re-
duction in CFU of Propionibacterium sp. on the lids occurred
from baseline to month 1 but not from baseline to month 3 and
not in the conjunctiva at either the month 1 or month 3 time
points. CFUs for S. aureus were significantly reduced from
baseline on the lids at month 3 but not month 1 and not in the
conjunctiva at either the month 1 or month 3 time points.
Coliforms were significantly reduced from baseline on the lids
and the conjunctiva at month 3 but not month 1.
DISCUSSION
The baseline assessment of ocular flora in normal (non–
dry eye group) and dry eye subjects produced findings
consistent with those of previous studies,23–26 with the predom-
inant bacteria isolated from the lids and conjunctiva in both
normal and dry eye subjects being coagulase-negative staphy-
lococci, followed by diphtheroids (Corynebacterium sp.,
Propionibacterium sp.) and S. aureus. The dry eye subjects
predominantly cultured commensal bacteria, but the total
colony counts in all dry eye groups were significantly greater
than those in the non–dry eye group. The exception was the
coliforms (including strains such as Escherichia, Klebsiella,
Enterobacter) that are not regarded as commensal ocular
flora23 and were isolated in 9% to 50% of subjects in the dry
eye subgroups but not in the normal group. There are several
possible reasons to explain the overgrowth of commensal
ocular flora in tear-deficient and meibomian disease dry eye
subgroups and the presence of noncommensal bacteria. First,
eyes with tear deficiency27 and eyes with meibomian gland
disease28 have reduced levels of antibacterial proteins in the
tear film, which would normally limit populations of micro-
organisms found on the eye and assist in protecting the ocular
surface from opportunistic pathogens.29 In addition, the
stagnation of meibomian gland secretions in meibomian
gland disease may encourage proliferation of indigenous lid
bacteria on the eyelids.7
There were no significant between-group differences in
the predominant types of bacteria isolated from tear-deficient
and/or meibomian gland disease dry eye subgroups. Possible
explanations for this may lie in the strong association of
meibomian gland disease with both Sjogren and non-Sjogren
tear deficiency30,31 and the common pathogenic progresses in
the development of chronic dry eye disease.32 Patients with
Sjogren syndrome are more deficient in tear film antibacterial
proteins than patients with non-Sjogren dry eye,33 and this may

TABLE 5. Conjunctival Bacteria Isolated in Dry Eye Groups and Non–Dry Eye Group at Baseline
No. of Subjects (%) With Organism Isolated
Bacteria Isolated NSTD (N = 20) SSTD (N = 11) MGD (N = 15) MGD+NSTD (N = 20) Non–Dry Eye Group (N = 18)
Coagulase-negative Staphylococcus 12 (60) 3 (27) 10 (67) 18 (90) 7 (41)
Corynebacterium sp. 9 (45) 8 (72) 1 (7) 8 (40) 1 (6)
Propionibacterium sp. 4 (20) 0 (0) 2 (13) 7 (35) 3 (18)
Staphylococcus aureus 0 (0) 0 (0) 3 (20) 1 (5) 1 (6)
None 2 (10) 0 (0) 2 (13) 1 (5) 5 (29)
Coliforms 3 (15) 5 (45) 2 (13) 3 (15) 0 (0)
Streptococcus viridans 2 (10) 0 (0) 0 (0) 0 (0) 0 (0)
Haemophilus sp. 1 (5) 0 (0) 0 (0) 0 (0) 0 (0)
Proteus sp. 0 (0) 0 (0) 0 (0) 2 (10) 0 (0)
Pseudomonas sp. 0 (0) 1 (9) 0 (0) 1 (5) 0 (0)
Serratiamarcescens 0 (0) 0 (0) 0 (0) 1 (5) 1 (6)

1016 q 2006 Lippincott Williams & Wilkins

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
JOBNAME: corn 25#9 2006 PAGE: 6 OUTPUT: Tuesday November 14 13:09:50 2006
lww/corn/129949/ICO200355

Cornea  Volume 25, Number 9, October 2006 Antibacterial Honey and Ocular Flora

TABLE 6. Effect of Medihoney on Eyelid and Conjunctival Isolates in Dry Eye Subjects
Common Normal Baseline 1/12 3/12 P (Baseline P (Baseline
Ocular Flora Site CFU Data CFUs CFUs CFUs vs. 1/12) vs. 3/12)
Coagulase-negative Staphylococcus Lid Mean 6 SD 40 6 40 26 6 34 17 6 31 0.0213 0.0009
Median 19 6 2
Range 0–100 0–100 0–100
Conjunctiva Mean 6 SD 35 6 40 14 6 28 5 6 16 0.0005 0.0003
Median 3 3 0
Range 0–100 0–100 0–88
Corynebacterium sp. Lid Mean 6 SD 41 6 43 16 6 34 6 6 20 0.0005 0.0001
Median 19 0 0
Range 0–100 0–68 0–100
Conjunctiva Mean 6 SD 20 6 31 4 6 14 4 6 18 0.0146 0.0001
Median 4 6 6
Range 0–100 0–100 0–100
Propionibacterium sp. Lid Mean 6 SD 8 6 15 266 266 0.0768 0.1413
Median 0 0 0
Range 0–100 0–32 0–49
Conjunctiva Mean 6 SD 467 265 3 6 10 0.3116 0.1413
Median 6 0 6
Range 0–31 0–26 0–100
Staphylococcus aureus Lid Mean 6 SD 30 6 43 35 6 45 18 6 38 0.6658 0.0215
Median 0 3 0
Range 0–100 0–100 0–94
Conjunctiva Mean 6 SD 10 6 30 17 6 32 8 6 24 0.1094 0.7349
Median 4 6 6
Range 0–100 0–100 0–100
Coliforms Lid Mean 6 SD 38 6 48 50 6 50 5 6 23 0.4240 0.0156
Median 6 44 0
Range 0–100 0–100 0–100
Conjunctiva Mean 6 SD 38 6 48 55 6 49 8 6 25 0.6702 0.0104
Median 0 42 0
Range 0–100 0–100 0–100
See Table 2 for explanation.

explain why subjects with autoimmune dry eye group have
higher total CFUs on the lids and the conjunctiva than the
other dry eye groups.
In this study, topical application of MedihoneyTM
Antibacterial Honey for 3 months reduced the total bacterial
CFUs in the lids and the conjunctiva of dry eye subjects to
non–dry eye CFU levels. The antibacterial effects of honey
have also recently been shown in an animal model. Al-Waili5
used natural unprocessed honey of multifloral origin on rat
conjunctivas after inoculation of conjunctival abrasions with
a variety of bacterial pathogens (S. aureus, E. coli, Proteus sp.,
P. aeruginosa, or Klebsiella). In the controlled study, cultures
taken from eyes treated with the honey at 4-times-a-day dosing
yielded no bacterial growth at least as rapidly (or sometimes
more rapidly) than eyes treated with ofloxacin or chloram-
phenicol at 4-times-a-day dosing.5
The antibacterial effect of honey could have 2 main
potential advantages for chronic dry eye patients. First, it
reduces the potential susceptibility of the dry eye patient to
bacterial conjunctivitis and keratitis. Second, it may assist in
improving the clinical signs and symptoms of dry eye disease
by reducing the production of certain bacterial lipases. These
bacterial lipases are thought to hydrolyze meibomian gland
lipids, releasing free fatty acids, which are toxic to the ocular
surface epithelia and may destabilize the tear film, thus further
exacerbating the dry eye disease.7,34 In particular, coagulase
negative Staphylococcus, P. acnes, and S. aureus have been
implicated in the production of these lipolytic enzymes.34
MedihoneyTM Antibacterial Honey use was able to produce
significant reductions in CFUs of coagulase-negative Staph-
ylococcus. The CFUs of P. acnes did not change significantly
with MedihoneyTM Antibacterial Honey use, possibly because
this species was isolated only in very low numbers (Table 6).
The CFUs of S. aureus were reduced on the eyelid but not in
the conjunctiva. The effectiveness of MedihoneyTM Antibac-
terial Honey against Propionibacterium sp. and S. aureus on
the ocular flora therefore requires further study, given their
proposed role in the pathogenesis of chronic meibomian gland
disease.
Coliforms do not form part of the commensal ocular
flora and yet were isolated in appreciable frequencies in a
significant number of dry eye subjects.23 Coliforms are

q 2006 Lippincott Williams & Wilkins 1017

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
JOBNAME: corn 25#9 2006 PAGE: 7 OUTPUT: Tuesday November 14 13:09:51 2006
lww/corn/129949/ICO200355
Albietz and Lenton
gram-negative rod-shaped facultative anaerobic bacteria and
are normal flora in the bowel and pathogenic in the urinary
tract.23 Coliform species are occasionally isolated in acute
bacterial ocular infections,35 and their presence in the ocular
flora would indicate an imbalance in the normal flora.
Therefore, the reduction in coliforms with MedihoneyTM
Antibacterial Honey use in the dry eye subjects suggests
a return to a more normal ocular flora.
The subjects of this study were part of a larger subject
group in which the effect of MedihoneyTM Antibacterial
Honey on tear film and ocular surface parameters was
examined. In this larger study, dry eye subjects using the
MedihoneyTM Antibacterial Honey 3 times daily for a 3-month
period had significant improvements in tear film and ocular
surface parameters such as a ocular surface staining, tear film
stability, goblet cell density, and conjunctival inflammation
(unpublished data). These improvements in tear film and
ocular surface parameters may be caused, at least in part, by
the reduction in ocular flora produced by MedihoneyTM
Antibacterial Honey use. Reductions in total CFUs with oral
minocycline therapy have been associated with improvements
in clinical symptoms and signs in subjects with eyelid
disease.26 Symptoms and signs returned after cessation of the
minocycline, which was also accompanied by an increase in
bacterial CFUs.26 Although we did not reassess colony counts
after cessation of the MedihoneyTM in this pilot study,
anecdotal reports from patients and our own clinical
observations indicate a deterioration in symptoms and clinical
signs after cessation of the product, which may have been
caused by an increase in bacterial colony counts.
In terms of the antibacterial activity of honey, there are
4 main factors involved. First, honey has a low water activity
(0.56 to 0.62), much too low for the survival of vegetative
bacterial cells.36 Second, honey has a low pH, generally
between 3.2 and 4.5.36 Third, a major factor is the action of the
glucose oxidase enzyme present in diluted honey that results in
the slow release of hydrogen peroxide at a very low but con-
tinuous level (;1–2 mmol/L for antibacterial honeys).36,37 The
glucose oxidase enzyme has been found to be practically
inactive in full-strength honey, and it gives rise to hydrogen
peroxide only when the honey is diluted.37 Hydrogen peroxide
is an effective ocular antimicrobial agent even at very low
concentrations; it is used in contact lens disinfection systems
and as a preservative in some ocular medication medications.38
Althogh the MedihoneyTM Antibacterial Honey applied to the
eyelid and the conjunctiva was 100% undiluted, it would have
been diluted immediately by the tear film on contact with the
eye. The duration of the antibacterial effects of the hydrogen
peroxide on the ocular surface has yet to be determined.
Finally, the floral source a honey will greatly affect its
antibacterial activity.36,39,40 Some honeys are up to 100 times
more active than others.39–42
Large differences exist in the antibacterial activity levels
of honeys from the same floral source.39,40
Manuka (Leptospermum scoparium, honey from New
Zealand) and related Leptospermum sp. honey from Australia,
which are contained in MedihoneyTM Antibacterial Honey, have
a high level of antibacterial activity that cannot be accounted for
in terms of either the high sugar content, low pH, or hydrogen
1018
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Cornea  Volume 25, Number 9, October 2006
peroxide production.41,42 Antibacterial honeys also seem to
stimulate monocytes in cell culture to release cytokines, tumor
necrosis factor-a, interleukin-1, and interleukin-6, which
activate the immune response to infection.43
Both topical and systemic antibiotics can significantly
reduce the CFUs in normal eyes26,44 and in eyes with chronic
lid disease,24 but these effects are only short term, and the
CFUs eventually return to pretreatment levels.44 Dry eye condi-
tions such as tear deficiency and meibomian gland disease are
chronic conditions for which there is no cure. Repeated dosing
with topical or systemic antibiotics to control lid bacteria is
expected to create resistance problems, and there are additional
side effects to consider with repeated or long-term antibiotic
dosing such as ocular surface toxicity with topical antibiotics
and systemic side effects of oral agents. Ta et al26 determined
that 4 in 10 patients with chronic eyelid disease cultured
coagulase-negative Staphylococcus strains resistant to oral
tetracyclines. Another study determined that patients with
features of dry eye conditions such as eyelid disease and
conjunctival inflammation and those with autoimmune con-
ditions such as Sjogren syndrome were at increased risk of
culturing conjunctival flora multiresistant to 5 or more
antibiotics commonly used in the treatment of eye infections.45
Therefore, given these antibiotic resistance issues, there is
merit in further exploring the potential antibacterial roles of
standardized antibacterial honey as an antibiotic-sparing agent
in the management of ocular surface disease. Possible
development of resistance with ocular use of MedihoneyTM
Antibacterial Honey will need to be examined in further
research.
Because this study was a pilot study, we did not inves-
tigate the duration of the antibacterial effects of MedihoneyTM
Antibacterial Honey continued once treatment was ceased.
This will need to be determined if MedihoneyTM is to be
considered as part of antimicrobial prophylaxis program before
ocular surgery or contact lens wear in patients with dry eye
problems. The 3-month period for reduction in bacterial colony
counts to normal levels will also require further study if the
product is to be considered an antimicrobial agent in the
treatment of more acute superficial ocular infections such as
blepharitis and conjunctivitis. Evidence from an animal model
of conjunctivitis, however, suggests that antibacterial honey is
at least as effective as topical broad-spectrum antibiotics in the
management of acute bacterial conjunctivitis.3 We effectively
used MedihoneyTM Antibacterial Honey in the treatment of
acute keratitis, conjunctivitis, and canaliculitis, but a controlled
trial would be required to verify the efficacy and limitations of
the product in the management of more acute ocular infections.
Also, although MedihoneyTM was effective in reducing overall
bacterial counts to non–dry eye levels and in reducing colony
counts of commensal ocular flora, other noncommensal
bacteria (eg, Pseudomonas sp., Serraciamarcesans, Proteus
sp.) were isolated in too few subjects in this study to determine
if MedihoneyTM was effective in vivo against these more
potentially pathogenic bacteria. Nevertheless, this pilot study
has shown that MedihoneyTM is effective in reducing the
proliferation of bacteria on the eyelids and conjunctiva in tear-
deficient dry eye and meibomian gland disease. There is
sufficient evidence to further explore the antibacterial effects of
q 2006 Lippincott Williams & Wilkins
JOBNAME: corn 25#9 2006 PAGE: 8 OUTPUT: Tuesday November 14 13:09:52 2006
lww/corn/129949/ICO200355
Cornea  Volume 25, Number 9, October 2006
MedihoneyTM Antibacterial Honey in a randomized controlled
clinical trial.
ACKNOWLEDGMENTS
Assistance with statistical analysis was provided by Kate
Downs from Datapharm Australia, Pty Ltd.
REFERENCES
1. Gethin G. Is there enough clinical evidence to use honey to manage
wounds? J Wound Care. 2004;13:275–278.
2. Molan P. The role of honey in the management of wounds. J Wound Care.
1999;8:415–418.
3. Al-Waili NS. Investigating the antimicrobial activity of natural honey and
its effects on the pathogenic bacterial infections of surgical wounds and
conjunctiva. J Med Food. 2004;7:210–222.
4. Mansour AM. Corneal oedema treated with honey. Clin Experiment
Ophthalmol. 2002;30:149–150.
5. Mozherenkov VP, Miniaeva TG. The use of products from bee raising in
ophthalmology and otorhinolaryngology (Russian). Med Sestra. 1991;50:
47–51.
6. Popescu MP, Palos E, Popescu F. Efficacy of combined biological therapy
with bee products in localized palpebral and conjunctival eye lesions with
reference to the clinico-functional changes. (Russian). Rev Oftalmol.
1985;29:53–61.
7. Dougherty JM, McCulley JP. Bacterial lipases and chronic blepharitis.
Invest Ophthalmol Vis Sci. 1986;27:486–491.
8. Bowman RW, Dougherty JM, McCulley JP. Chronic blepharitis and dry
eyes. Int Ophthalmol Clin. 1987;27:27–35.
9. Groden LR, Murphy B, Rodnite J, et al. Lid flora in blepharitis. Cornea.
1991;10:50–53.
10. Houang E, Lam D, Fan D, et al. Microbial keratitis in Hong Kong:
relationship to climate, environment and contact-lens disinfection. Trans
R Soc Trop Med Hyg. 2001;95:361–367.
11. Huang SC, Wu SC, Wu WC, et al. Microbial keratitis—a late
complication of penetrating keratoplasty. Trans R Soc Trop Med Hyg.
2000;94:315–317.
12. Scott IU, Flynn HW Jr, Feuer W, et al. Endophthalmitis associated with
microbial keratitis. Ophthalmology. 1996;103:1864–1870.
13. Blair S, Carter DA. Honey and drug resistant pathogens. Paper presented
at: Joint Scientific Meeting of the Australian Society for Microbiology;
July 2000; Cairns, Queensland, Australia.
14. George N, Sturgess R, Foagagali J, et al. Honey Vantimicrobial agents au
naturel. Paper presented at: Australian Infection Control Association
Third Biennial Conference; June 2004; Hobart, Australia.
15. Fleiszig SM, Efron N. Microbial flora in eyes of current and former
contact lens wearers. J Clin Microbiol. 1992;30:1156–1161.
16. Rubio EF. Climatic influence on conjunctival bacteria of patients
undergoing cataract surgery. Eye. 2004;18:778–784.
17. Lemp MA. Report on the National Eye Institute/Industry Workshop on
Clinical Trials in Dry Eyes. CLAO J. 1995;21:221–232.
18. McMonnies CW. Key questions in a dry eye history. J Am Optom Assoc.
1986;57:512–517.
19. Bron AJ. The Doyne Lecture: reflections on the tears. Eye. 1997;11:583–602.
20. Hykin PG, Bron AJ. Age related morphological changes in the lid margin
and meibomian gland anatomy. Cornea. 1992;11:334–342.
21. Thiel HJ, Schumacher U. Normal flora of the human conjunctiva:
examination of 135 persons of various ages. (German). Klin Monatsbl
Augenheilkd. 1994;205:348–357.
q 2006 Lippincott Williams & Wilkins
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Antibacterial Honey and Ocular Flora
22. Postmes TAE, van den Bogaard, Hazen M. The sterilization of honey with
cobalt 60 gamma radiation: a study of honey spiked with spores of Clostri-
dium botulinum and Bacillus subtilis. Experimenta. 1995;51:986–989.
23. Granato PA. Pathogenic and indigenous microorganisms of humans. In:
Murray P, ed. Manual of Clinical Microbiology. Vol. 1, 6th ed. American
Society for Microbiology, Washington, D.C., 1995:14–15.
24. Groden LR, Murphy B, Rodnite J, et al. Lid flora in blepharitis. Cornea.
1991;10:50–53.
25. Gritz DC, Scott TJ, Sedó SF. Ocular flora of patients with AIDS compared
with those of HIV-negative patients. Cornea. 1997;16:400–405.
26. Ta CN, Shine WE, McCulley JP, et al. Effects of minocycline on the ocular
flora of patients with acne rosacea or seborrheic blepharitis. Cornea. 2003;
22:545–548.
27. Seal DV, McGill JI, Mackie IA, et al. Bacteriology and tear protein profiles
of the dry eye. Br J Ophthalmol. 1986;70:122–125.
28. Kiratli H, Irkec M, Orhan M. Tear lactoferrin levels in chronic meibomitis
associated with acne rosacea. Eur J Ophthalmol. 2000;10:11–14.
29. Seamone CD, Jackson WB. Immunology of the external eye. In: Tasman
W, Jaeger EA, eds. Clinical Ophthalmology. Philadelphia, PA: Lippincott-
Raven; 1997:1–52.
30. Pflugfelder SC, Tseng SC, Sanabria O, et al. Evaluation of subjective
assessments and objective diagnostic tests for diagnosing tear-film dis-
orders known to cause ocular irritation. Cornea. 1998;17:38–56.
31. Shimazaki J, Goto E, Ono M, et al. Meibomian gland dysfunction
in patients with Sjogren syndrome. Ophthalmology. 1998;105:1485–
1488.
32. Stern ME, Beurerman RW, Fox RI, et al. The pathology of dry eye: the
interaction between the ocular surface and lacrimal glands. Cornea. 1998;
17:584–589.
33. Ohashi Y, Ishida R, Kojima T, et al. Abnormal protein profiles in tears
with dry eye syndrome. Am J Ophthalmol. 2003;136:291–299.
34. McCulley JP, Dougherty JM. Bacterial aspects of chronic blepharitis.
Trans Ophthalmol Soc U K. 1986;105:314–318.
35. Mahajan VM. Acute bacterial infections of the eye: their aetiology and
treatment. Br J Ophthalmol. 1983;67:191–194.
36. Molan PC. The antibacterial activity of honey. 1. The nature of the
antibacterial activity. Bee World. 1992;73:5–28.
37. Bang LM, Buntting C, Molan P. The effect of dilution on the rate of
hydrogen peroxide production in honey and its implications for wound
healing. J Alternat Complement Med. 2003;9:267–273.
38. Riley MV, Wilson G. Topical hydrogen peroxide and the safety of ocular
tissues. CLAO J. 1993;19:186–190.
39. Molan PC. The antibacterial activity of honey. 2. Variation in the potency
of the antibacterial activity. Bee World. 1992;73:59–76.
40. Wilkinson JM, Cavanagh HMA. Antibacterial activity of 13 honeys
against Escherichia coli and Pseudomonas aeruginosa. J Med Food.
2005;8:100–103.
41. Davis C. Antimicrobial Properties of Australian Honeys—Final Report. A
report by the Rural Industries Research and Development Corporation,
Centre for Food Technology; Brisbane, Australia, July 1997.
42. Allen KL, Molan PC, Reid GM. A survey of the antibacterial activity of
some New Zealand honeys. J Pharm Pharmacol. 1991;42:817–822.
43. Tonks AJ, Cooper RA, Jones KP, et al. Honey stimulates inflam-
matory cytokine production from monocytes. Cytokine. 2003;21:242–
247.
44. Ermis SS, Aktepe OC, Inan UU, et al. Effect of topical dexamethasone and
ciprofloxacin on bacterial flora of healthy conjunctiva. Eye. 2004;18:
249–252.
45. Mino de Kaspar H, Shriver EM, Nguyen EV, et al. Risk factors for
antibiotic-resistant conjunctival bacterial flora in patient undergoing
intraocular surgery. Graefes Arch Clin Exp Ophthalmol. 2003;241:730–733.
1019