6 Quality Assurance
The basic LC-HRMS approach is the same in both labs, namely creating
a database of authentic samples and a second database consisting
of all kinds of syrups that may be used for adulteration and eventually
statistical evaluation with the goal to identify markers that occur in syrup
but not in authentic honey.
Both labs underline that for this approach a database of authentic
honeys for comparison is absolutely essential as honey is a highly diverse
matrix. This fact must be taken into account to ensure that markers are
actually due to adulteration instead being e.g. a botanical marker.
Both labs share the same approach and used a Sciex X500R QToF
HRMS (figure 1) for this trial. However, the implementation was done completely
independent in the labs. In particular, the samples used for creation
of honey and syrup databases were entirely different in both labs. The
chromatographic conditions and mass spectrometric settings were also not
standardized. Furthermore, the identified markers were not harmonized,
so each lab was using its own markers. Each lab provided 20 anonymized
samples of global origins (table 1) for the interlaboratory comparison.
The results were in agreement for 36 samples (90%) for which both
laboratories arrived at the same conclusion. Both laboratories detected
an adulteration in 31 of the samples. In 5 cases no adulteration was
detected in both laboratories. For the four remaining cases results differed
between the labs. The reasons for these differences could be:
• The syrup used for adulteration was available only in the syrup database
of one lab
• Sensitivity for the respective syrup marker differed between the labs
due to different method setup and was below the limit of detection
in one lab
• Internal thresholds (“bee feed allowance”) used to distinguish positive
and negative findings are set at different levels
• Statistical evaluation was done in a different way resulting in different
In spite of these potential differences in the method itself, this laboratory
comparison shows, that the same approach, namely the identification
of markers by comparison of syrups with authentic honeys, leads mostly
to the same conclusion, although implementation was done differently.
In addition to the results of LC-HRMS analysis the results of NMR are
also provided for 26 out of the 40 samples. Twelve samples that were
judged adulterated according to LC-HRMS were confirmed adulterated
by NMR. Results of additional three samples that were judged adulterated
by one lab using LC-HRMS were also confirmed using NMR. In ten
samples LC-HRMS was able to detect an adulteration that NMR was not
able to detect. The reason for that is most likely the aforementioned difference
in sensitivity between both methods to trace syrup addition. One
sample was confirmed authentic by NMR and LC-HRMS.
As can be seen, in general there is also a good agreement between
LC-HRMS and NMR. Differences between both methods seem to be
caused by the difference in sensitivity. While NMR is not able for those
samples to detect deviations in the major compounds in honey, looking at
the minor compounds using LC-HRMS still reveals the presence of syrup.
It was mentioned a few times that LC-HRMS is very sensitive. It has to
be sensitive as the concentrations of the markers in pure syrup are already
very low. In order to be able to still detect those markers with only a few
percent of syrup added to honey, a high sensitivity is necessary.
However, it is not in our interest to find traces of syrup additions in honey.
We are aware that residual syrup from feeding the bees, e.g. during winter
time, might be relocated in the hive which potentially leads to a positive
finding. These residues of bee feed can hardly be regarded as food fraud,
as an integral part of food fraud consists of deception with the intent to
maximize profit. But the beekeeper feeds the bees to keep them alive. Only
the addition of several percent of syrup to honey is of economic interest.
The term commonly used in the USA is “economically motivated adulteration”
(EMA) and it is not to be mixed up with intentional adulteration
(IA). While the addition of syrup to honey for illicitly increasing profit is a
case of EMA, residues of substances that were intentionally introduced at
some stage of production (e.g. antibiotics) would be a case of IA.
In conclusion, this first interlaboratory comparison for honey authenticity
using LC-HRMS shows, that it is possible to reach a high level of
agreement of results when the same approach is applied, even if implementation
and datasets used for building databases are very different.
However, it is necessary to conduct further interlaboratory comparisons
with a greater number of labs in order to demonstrate a general comparability
among different labs utilizing this technology.
In 2008 Arne Duebecke joined Quality Services International
(QSI) laboratories in Bremen, Germany. He
worked on pyrrolizidine, tropane, quinolizidine and ergot
alkaloids in honey and other bee products as well as
plant materials, including tea (Camellia sinensis), herbs
and spices. During that period he took different roles in
project management, research and development and
quality management. In 2013 he moved the focus to the
verification of authenticity and detection of adulteration
of foodstuffs. The methods employed comprised enzymatic
approaches, SIRA, HPLC, LC-MS, HRMS, NMR and
multivariate statistics. In 2017 Arne took the lead of the
newly founded Tentamus Center for Food Fraud (TCF²).
Additionally, he was recently appointed business unit
manager for medical devices at QSI. He regularly presents
latest QSI and TCF² research at a variety of international
conferences and also prepared a number of publications
for trade journals as well as peer-reviewed journals. He is
active member of the AOAC Food Authenticity & Fraud
Panel and the committee on food authenticity of DIN
(German Institute for Standardization, Berlin).
Managing director of FoodQS GmbH – an accredited
laboratory specialized in analytics of honey and bee
Member of different national and international
working groups and more than 20 years of experience
in analytics of bee products.
Currently focused on developing new methods for
the identification of honey adulterations in order to
maintain the good reputation of honey worldwide.