Online Submissions: http://www.journaltcm.
com J Tradit Chin Med 2013 February 15; 33(1): 119-124
info@journaltcm.com
Microbial and heavy metal contamination in commonly consumed
traditional Chinese herbal medicines
Adelinesuyien Ting, Yiingyng Chow, Weishang Tan
aa
Adelinesuyien Ting, Yiingyng Chow, Weishang Tan,
School of Science, Monash University Sunway Campus, Ja-
lan Lagoon Selatan, Bandar Sunway, 46150 Petaling Jaya, Se-
langor, Malaysia
Supported by Monash University Sunway Campus
Correspondence to: Prof. Adelinesuyien Ting, School of
Science, Monash University Sunway Campus, Jalan Lagoon
Selatan, 46150 Bandar Sunway, Selangor Darul Ehsan, Malay-
sia. adelsuyien@yahoo.com; adeline.ting@monash.edu
Telephone: +86-603-5514 6105
Accepted: October 6, 2012
Abstract
OBJECTIVE: The increasing popularity and wide-
spread use of traditional Chinese herbs as alterna-
tive medicine have sparked an interest in under-
standing their biosafety, especially in decoctions
that are consumed. This study aimed to assess the
level of microbial and heavy metal contamination
in commonly consumed herbal medicine in Malay-
sia and the effects of boiling on these contamina-
tion levels.
METHODS: Four commonly consumed Chinese
herbal medicine in Malaysia-"Eight Treasure Herbal
Tea", "Herbal Tea", Xiyangshen (Radix Panacis Quin-
quefolii) and Dangshen (Radix Codonopsis) were
evaluated in this study. Herbal medicines were pre-
pared as boiled and non-boiled decoctions, and
their microbial enumeration and heavy metal de-
tection were conducted with plate assay and atom-
ic absorption spectroscopy, respectively.
RESULTS: Findings revealed that herbal medicines
generally had 6 log10cfu/mL microbial cells and that
boiling had significantly reduced microbial contam-
JTCM | www. journaltcm. com
ISSN 0255-2922
2013 JTCM. All rights reserved.
EXPERIMENTAL STUDY
inants, where no Bacillus spp., Staphylococcus spp.
and Clostridium spp. were recovered. Heavy metals
such as Mn, Cu, Cd, Pb, Fe and Zn were also detect-
ed from all the samples, generally in low concentra-
tions (<1 mg/L) except for Mn (18.545 mg/L). All de-
coctions (after boiling) have reduced concentra-
tions of Cu, while others were not significantly dif-
ferent. Comparisons between samples with single
and multi-herbs suggest level of microbial and met-
al contamination is not influenced by number of
herbs in sample.
CONCLUSION: Herbal medicines generally have mi-
crobial and heavy metal contaminants. However,
the boiling process to generate decoctions was
able to successfully reduce the number of microbes
and Cu, ensuring safety of herbal medicines for con-
sumption.
2013 JTCM. All rights reserved.
Key words: Decoction processing; Drug contamina-
tion; Heavy metal poisoning, nervous system; Mi-
crobial consortia; Drugs, Chinese herbal
INTRODUCTION
The use of Chinese herbal medicines (CHMs) as a nat-
ural remedy is a common practice among the Chinese
communities. This practice has slowly transcended to
other cultures especially with the introduction of Tradi-
tional Chinese Medicine (TCM) as alternative medi-
cine. The World Health Organization (WHO) report-
ed that 70% to 80% of the world population uses herb-
al medicine for primary healthcare.1 In Singapore, a
country dominated by Chinese communities, 89% of
the children have consumed CHMs in the first 30
months of their life.2 The high demand for CHMs is
119 February 15, 2013 | volume 33 | Issue 1 |
 Ting ASY et al. / Experimental Study

attributed to the generalized perception that CHMs sample 1 from shop B. The composition of herbs in
are natural, thus safer compared to synthetic drugs.3 In each sample is provided in Table 1A, B, C and D. For
addition, CHMs are reportedly able to cure diseases each sample, individual herbs were cut into pieces mea-
which are resistant to synthetic drugs such as the use of suring 1 cm 1 cm, pooled together and subsequently
Zemaphyte, a combination of 10 different CHMs, for distributed equally for the microbial assay and heavy
atopic eczema.4 CHMs are also known to be easily ab- metal determination.
sorbed by the human body.5 Table 1A Latin and Chinese Pinyin names of herbal plants in
Although the popularity of CHMs grew among the "Eight Treasure Herbal Tea" [CHM1 (A and B)]
consumers, regulations on the biosafety, production Latin name Chinese Pinyin name
and preparation of CHMs is very lacking.6 Biosafety
testing on CHMs to evaluate microbial and metal con- Spica Prunellae Vulgaris12 Xiakucao
tamination levels are mostly absent from the monitor- Herba Lophatheri13 Danzhuye
ing and regulatory exercise. In fact, tests are only con-
ducted when toxicity cases are reported. Although mi- Rhizoma Imperatae12 Baimaogen
crobial contamination can be reduced by boiling, some Flos Bombacis 14
Mumianhua
bacterial endospores and fungal spores still persist, caus-
ing health hazards.7 There were reports that Rhodiola Herba Plantaginis12 Cheqiancao
root extracts in a well-known herbal product, was con- Folium Mori12 Sangye
taminated by Bacillus subtilis, causing liver damage in
consumers.8 Aflatoxin from Aspergillu ssp., commonly Saccharumofficinarum L. 13
Ganzhegan
detected in CHMs such as Yuejebaohe Wan and Feier
Flos Dendranthematis2 Juhua
Pian,9 is another example of common hazard in the use
fo CHMs. It is therefore crucial to determine and un- Table 1B Latin and Chinese Pinyin names of herbal plant
derstand the level of microbial contamination in parts used for "Herbal Tea" [CHM2 (A and B)]
CHMs. Chinese
Latin name
Heavy metal contamination in CHMs is also prevalent Pinyin name
and has been found in CHMs such as Radix codonopsis Herba Schizonepetate Tenuifolia15 Jingjie
and R. angelicae Sinensis, contaminated with lead and Radix Saposhnikoviae16 Fangfeng
arsenic, respectively.10 In addition, Mn, Cu, Cd, Fe,
and Zn are also detected in a screening test done using Notopeteryginm Incisum15 Jianghuo
2080 CHMs samples, with 42 of these samples having Radix Angelicaepubescentis 15
Duhuo
levels exceeding the permissible legal limits.2 Unlike mi-
crobial contamination, heavy metal residues will persist Rhizoma Chuanxiong17 Chuanxiong
in the CHMs even after boiling, causing heavy metal
Radix Bupleuri Chinensis15 Chaihu
poisoning.11 Therefore, it is crucial to determine the lev-
els of heavy metals in the CHMs before decoction to Radix Peucedani13 Qianhu
evaluate biosafety of CHMs.
Scutellaria Baicalensis Georgi.15 Huangjing
In our study, we aimed to evaluate the microbial and
heavy metal levels in four of the most common CHMs Radix Ilicis Asprellae18 Gangmeigen
consumed by the local Chinese community in Malay- Fructus Aurantii Submaturus15 Zhike
sia. These CHMs were recommended by practitioners
of CHM in Malaysia. As microbial and heavy metal Poria15 Fuling
levels may differ prior to and after boiling, we evaluat- Radix Glycyrrhizae15 Gancao
ed this as well to determine the impact of boiling and
Table 1C Latin and Chinese Pinyin names of herbal plant
the biosafety of decoctions for consumption. parts used for "Xiyangshen" [CHM3 (A and B)]
Latin name Chinese Pinyin name
MATERIALS AND METHODS Radix Panacis Quinquefolii19 Xiyangshen
"Eight treasure herbal tea" (CHM1), "Herbal tea" Table 1D Latin name and Chinese common names of herbal
(CHM2), "Xiyangshen root" (CHM3) and "Dangshen plant parts used for "Dangshen root" [CHM4 (A and B)]
root" (CHM4) were bought randomly from two shops. Latin name Chinese Pinyin name
One was designated as shop A (Taman Midah, Cheras,
Kuala Lumpur, Malaysia) and the other as shop B (Ta- Radix Codonopsis13 Dangshen
man Dahlia, Cheras, Kuala Lumpur, Malaysia). To re- Determination of microbial contaminants in CHMs
spect privacy, both shops have requested for anonymi- For sample CHM1A, 1g of the sample was mixed with
ty. The samples were then labelled accordingly, i.e. 9mL sterile distilled water in a 50 mL Schott bottle.
CHM1A to indicate herbal medicine sample 1 from The Schott bottle was then agitated on a rotary shaker
shop A whereas CHM1B to indicate herbal medicine (250 rpm) for 1 h, at room temperature (27 3 ).

JTCM | www. journaltcm. com 120 February 15, 2013 | volume 33 | Issue 1 |
 Ting ASY et al. / Experimental Study
The mixture was then serially diluted (10-1 to 10-6). Ali-
quots (0.1 mL) from dilution factors 10-2, 10-4 and 10-6
were spread-plated onto the following agar plates; Nu-
trient agar (NA), HiCromeTM Bacillus Agar, Mannitol-
Salt Agar (MSA), Reinforced Clostridial Agar (RCA),
and Potato Dextrose Agar (PDA). Triplicates were pre-
pared. All agar plates (except RCA) were incubated in
aerobic condition at room temperature (273) for
24 h. The RCA plates were incubated in an anaerobic
jar at room temperature (273) for 24 h. Negative
controls using 0.1mL of sterilized distilled water were
spread-plated on all agar plates. Observation was made
after 4 days of incubation. Procedure was repeated for
samples CHM2A, CHM3A, CHM4A and all samples
for shop B. For the effect of boiling on microbial con-
taminants in CHMs, the samples were boiled at 100
for 1 h and described procedures repeated.
Characterization of microbial contaminants
The microbial contaminants were characterized based
on their cultural and morphological morphologies on
the selective and differential agar used. Fungal contami-
nants were identified based on their spores structures
stained with Lactophenol Cotton Blue.
Quantification of heavy metals from CHMs
For sample CHM1A, 10 g of the sample was added in-
to 200 mL of sterile distilled water in a 500 mL Schott
bottle. Triplicates were prepared. The bottles were then
agitated on a rotary shaker (250 rpm) for 1h, at room
temperature (27 3 ) to allow the samples to mix
well. The mixture was then filtered using the Aspirator
A-3S EYELA vacuum pump. The collected herbal solu-
tions were then analyzed using PerkinElmer AAnalyst
100 atomic absorption spectroscopy (AAS) for their
manganese (Mn), lead (Pb), copper (Cu), cadmium
(Cd), zinc (Zn) and iron (Fe) contents. Operational pa-
rameters are summarized in Table 2. Since the AAS
maximum detection limit is 20 ppm, a series of stan-
dard solution was prepared (4, 8, 12, 16 and 20 ppm).
The absorbance readings obtained were then calculated
using the equations generated from the standard curves
for each metal. For assay of boiled decoction, the herb-
al samples were boiled at 100 for 1 h prior to heavy
metal quantification using similar procedures. Proce-
dure was also repeated for analysis of samples from
shop B.
Statistical analysis
The data was analysed with Statistical Package for the
Social Sciences (SPSS), version 18.0. T-test, one way
analysis of variance (ANOVA) and Tukey's studentized
range test were applied accordingly. Difference was con-
sidered significant at P<0.05.
RESULTS
Prior to boiling, microbial contamination was present
in all samples except CHM3A. These samples had be-
JTCM | www. journaltcm. com
Table 2 AAS parameters and furnace temperature program
for determination of heavy metals
Lamp current Wavelength Slit
Element
(mA) (nm) (nm)
Mn 20 279.5 0.2
Pb 10 283.3 0.7
Cu 15 324.8 0.7
Cd 4 228,8 0.7
Fe 20 248.3 0.2
Zn 15 213.9 0.7
Notes: AAS: atomic absorption spectroscopy; Mn: manganese;
Pb: lead; Cu: copper; Cd: cadmium; Zn: zinc; Fe: iron.
tween 4 to 6 log10cfu/mL microbial cells detected
which were reduced to 0 log10cfu/mL after boiling (Fig-
ure 1). This observation was consistent for all samples
except sample CHM4B enumerated on NA. For this
sample, 4.9 log10cfu/mL of bacterial cells were recov-
ered from boiled decoctions, although none were the
endospore forming Bacillus and Clostridium, as well as
Staphylococcus spp. (Figure 1). Comparison among the
three genera showed Bacillus spp. as the most prevalent
bacteria, detected from 7 of the 8 samples. The least
common bacteria recovered was the Clostridium spp.,
with only 2 samples (CHM3B, CHM4B) having these
bacteria, which coincidently was from the same shop
(Figure 1). Staphylocoocus spp. was also less prevalent
compared to Bacillus spp., detected only in samples
CHM1 and CHM2, obtained from both shops A and
B. Fungal contaminants were also prevalent in 6 of the
8 tested samples. This study also showed that a sample
of CHM was able to have many types of contamina-
tion. We also observed that there was no significant dif-
ference in number of contaminants recovered from
samples with single herbal species (CHM3, CHM4)
compared to CHMs with multiple herbal species
(CHM1, CHM2) on all agar tested. However, we noted
with interest that Clostridium contamination was ob-
served in two of the four samples obtained from shop B,
with both these samples having only a single type of herb.
The use of BHicrome agar, RCA and MSA were
helpful to determine the contaminants based on their
ability to grow on these media. Two species of Bacil-
lus were detected as a result of colony pigmentation
on BHiCrome agar. B. subtilis formed dark green,
flat colonies; while B. coagulans formed pink, small
raised colonies. Two common species of Staphylococ-
cus were also recovered from the samples; S. aureus
and S. epidermidis, which showed typical pink and
yellow colonies on MSA, respectively. The fungal iso-
late was identified as Aspergillu ssp. based on the for-
mation of conidia.
Of the 6 metals assessed, Mn had the highest concen-
tration detected from the CHMs samples with
1.394-18.545 mg/L, while Cd had the least amount de-
121 February 15, 2013 | volume 33 | Issue 1 |
 Ting ASY et al. / Experimental Study

A B

C D

Figure 1 Microbial contaminants (log10cfu/mL) recovered from CHM sampleson, (A) Nutrient Agar, (B) Potato Dextrose Agar, (C) Hi-
Crome Bacillus Agar, (D) Reinforced Clostridial Agar, and (E) Mannitol Salt Agar
CHM: Chinese herbal medicine; CHM1A:Eight treasure herbal teasample from shop A; CHM1B:Eight treasure herbal teasam-
ple from shop B; CHM2A: "Herbal tea" sample from shop A; CHM2B: "Herbal tea" sample from shop B; CHM3A: Xiyangshen root
sample from shop A; CHM3B: Xiyangshen root sample from shop B; CHM4A: Dangshen root sample from shop A; CHM4B: Dangsh-
en root sample from shop B. The effect of boiling on the microbial contaminants was evaluated within the same sample using
paired-sample t-test (aindicates significant difference between before and after boiling within the same sample at P0.05). Bars in-
dicate standard error.
tected (0.105-.314 mg/L). Other metals such as Cu, common,20 as microbes can be easily introduced any-
Pb, Fe, and Zn, were all <3 mg/L (Figure 2). All time during the cultivation and preparation of CHMs.
these metals were found in all samples tested, with Thus to prevent microbial contaminants, CHMs
CHM4A generally having higher concentrations of should be handled under strict regulations and pack-
Mn, Cu, Cd, Pb, and Zn; while sample CHM1B aged using clean packaging materials.21 In our study,
had the least concentration of metals compared to Bacillus spp. was the most commonly recovered con-
the rest of the samples. After boiling, concentrations
taminant compared to Clostridium and Staphylococcus
of heavy metals varied according to the samples. We
observed that Cu was the only metal which concen- spp. This agrees with literatures.20 Our focus on Bacil-
tration was significantly reduced in all boiled decoc- lus was because this endospore-forming bacterium was
tions. Other metals such as Cd, Mn, Zn and Fe, known to produce pre-formed toxins which could
were either reduced or increased in metal concentra- cause foodborne illness.22,23 Interestingly, no Bacillus
tions after boiling. Pb concentration was elevated af- spp. were recovered from the boiled decoctions in this
ter boiling; however the values were not significantly study, suggesting that boiled decoctions were generally
different from non-boiled samples. We also noted icines with reportedly 48% of American ginseng root
that there were no significant differences in concen- samples harbouring 105cfu of Aspergillus/g sample21 and
tration of heavy metals between samples with single was detected in seeds of American ginseng as well.25As-
herbal species (CHM3, CHM4) compared to multi- pergillus can continue to spread during storage if the
ple herbal species (CHM1, CHM2). CHMs are not dried properly,22 thus it is crucial to pro-
cess CHMs carefully.
DISCUSSION For heavy metal contamination in CHMs, our resultof-
The discovery of microbial contamination in CHMs is fered the novelty of documenting heavy metal concen-

JTCM | www. journaltcm. com 122 February 15, 2013 | volume 33 | Issue 1 |
 Ting ASY et al. / Experimental Study

A B

C D

E F

Figure 2 Concentrations of, (A) Cd, (B) Cu, (C) Zn, (D) Mn, (E) Pb and (F) Fe, detected in each sample before and after boiling.
CHM: Chinese herbal medicine; CHM1A:Eight treasure herbal teasample from shop A; CHM1B:Eight treasure herbal teasam-
ple from shop B; CHM2A: "Herbal tea" sample from shop A; CHM2B: "Herbal tea" sample from shop B; CHM3A: Xiyangshen root
sample from shop A; CHM3B: Xiyangshen root sample from shop B; CHM4A: Dangshen root sample from shop A; CHM4B: Dangsh-
en root sample from shop B. The effect of boiling on the metal concentration was evaluated within the same sample using
paired-sample t-test (aindicates significant difference between before and after boiling within the same sample at P0.05). Bars in-
dicate standard error.
trations in boiled decoctions, as existing literatures non-boiled samples were 1.028 and 0.750 mg/L, well
mainly discussed metal levels in non-boiled CHMs.26,27 within the permissible level of 10 mg/L.28 However, we
We found that Pb concentrations in boiled and detected higher concentrations of Cd from boiled de-

JTCM | www. journaltcm. com 123 February 15, 2013 | volume 33 | Issue 1 |
 Ting ASY et al. / Experimental Study
coctions in samples CHM3A and CHM4A, exceeding
slightly the 0.3 mg/L permissible level.28 The higher
concentrations of Cd in herbal medicine appeared to
be a common finding. High Cd levels were also detect-
ed in 79 samples of various herbal medicines in Italy
(up to 0.75 mg/L),29 and in ginseng purchased from
the US, Europe and Asia.30 We postulate that this
could be associated with the fact that most plants natu-
rally had active Cd uptake via roots which then re-
mained in the plant tissues.31 We do not have permissi-
ble levels of Cu, Zn, Mn and Fe as they are not yet
identified as major health hazards in CHMs. Neverthe-
less, these trace elements are now increasingly found in
the environment as a result of heavy usage of pesticides
(Cu-based pesticides)7 leading to groundwater contami-
nation,32 and the overuse of Mn-based fungicides.33
These heavy metals must be monitored especially since
Mn and Fe have potential to increase in concentrations
in boiled decoctions as determined in this study. These
metals must be monitored to reduce the risk of bioac-
cumulation of metals in our body upon consumption.
ACKNOWLEDGEMENTS
The authors extend their gratitude to Monash Universi-
ty Sunway Campus for the financial assistance and fa-
cilities to conduct the study.
REFERENCES
1 Akerele O. Nature's medicinal bounty: don't throw it
away. World Health Forum 1993;14(4): 390-395.
2 Koh H, Woo S. Chinese proprietary medicine in Singa-
pore-Regulatory control of toxic heavy metals and unde-
clared drugs. J Drug Safety 2000; 23(5): 351-362.
3 Ernst E. Herbal medicines: balancing benefits and risks.
Novartis Found Symp 2007; 282: 154-172.
4 Atherton DJ, Sheehan MP, Rustin MH, et al. Treatment
of atopic eczema with Traditional Chinese Medicinal
plants. Pediat Dermatol 1992; 9(4):373-375.
5 Liang YZ, Xie PS, Chan K. Quality control of herbal med-
icines. J Chromatography 2004; 812(1-2): 53-70.
6 Saad B, Azaizeh H, Hijleh GA, et al. Safety of traditional
Arab herbal medicine. J Safety Med Plants 2006; 3(4):
433-439.
7 Ochei J, Kolhatkar A. Bacteriology, Medical Laboratory
Science: Theory and Practice. India: Tata McGraw-Hill
2000; 547-548.
8 Stickel F, Droz S, Patsenker E, et al. Severe hepatotoxicity
following ingestion of Herbalife nutritional supplements
contaminated with Bacillus subtilis. J Hepatology 2009; 50
(1): 111-117.
9 Lu ZY, Ye JW, Yuan WP, et al. Determination and analysis
of the alfatoxin in seven Traditional Chinese Medicines.
Chin Arch Tradit Chin Med 2001; 5: 527-528.
10 Zhang HF, Zhao CJ, Ni N. Determination of heavy met-
al elements in five beneficial traditional Chinese medi-
cines. J Shenyang Pharmac Univ 2003; (1):8-11.
11 Ernst E. Toxic heavy metals and undeclared drugs in
Asian herbal medicines. Trends Pharmacol Sc 2002; 23(3):
136-139.
JTCM | www. journaltcm. com
12 Chauhan NS. Medicinal and aromatic plants of Himachal
Pradesh. 1st ed. New Delhi: Indus Publishing,1999: 129.
13 Hou JP, Jin YY. The healing power of Chinese herbs and
medicinal recipes. New York: Haworth Press, 2005:
330-498.
14 Watson RR, Preedy VR. Botanical medicine in clinical
practice. United Kingdom: CAB, 2008: 52-63.
15 Hempen CH, Fisher T. A meteria medica for Chinese
medicine. China: Churchill Livingstone Elsevier, 2009:
15-73.
16 Watson RR. Complementary and alternative therapies in
the aging population. United States of America: Academic
Press, 2009: 256.
17 Watson RR, Preedy VR. Botanical medicine in clinical
practice. United Kingdom: CABI, 2008: 112-120.
18 Hu SY. Food plants of China. Hong Kong: The Chinese
University Press, 2005: 230-248.
19 Watson RR, Preedy VR. Botanical medicine in clinical
practice. United Kingdom: CABI, 2008: 52-63.
20 Raman P, Patino LC, Nair NG. Evaluation of metal and
microbial contamination in botanical supplements. J Ag-
ricul Food Chem 2004; 52(26): 7822-7827.
21 Tournas VH, Katsoudas E, Miracco EJ. Moulds, yeasts
and aerobic plate counts in ginseng supplements. Int J
Food Microbiol 2006; 108(2): 178-181.
22 McGuire M, Beerman KA. Nutritional sciences: from fun-
damentals to food. 3rd ed. Unites States of America:
Brooks Cole; 2012: 215-217.
23 Wearing J. Bacteria: Staphylococcus, Streptococcus, Clos-
tridium, and other bacteria. Canada: Crabtree Publishing
Company, 2010.
24 Sonenshein AL, Hoch JA, Losick R. Bacillus subtilis and
its closest relatives: from genes to cells. Washington: ASM
Press, 2001: 128-129.
25 Zhang GZ, Zhang SF. Fungal detection of American gin-
seng seeds from Beijing and northeast area in China.
Zhong Guo Zhong Yao Za Zhi 2002; (27): 658-661.
26 Ang HH. Lead contamination in Eugenia dyeriana herbal
preparations from different commercial sources in Malay-
sia. Food Chem Toxicol 2008; 46(6): 1969-1975.
27 Caldas ED, Machado LL. 2004. Cadmium, mercury and
lead in medicinal herbs in Brazil. Food Chem Toxicol
2004; 42(4): 599-603.
28 Harris ESJ, Cao SG, Littlefield BA, et al. Heavy metal
and pesticide content in commonly prescribed individual
raw Chinese Herbal Medicines. Sc Total Environ 2011;
409(20): 4297-4305.
29 Pasquale AD, Paino E, Pasquale RD, et al. Contamina-
tion by heavy metals in drugs from different commercial
sources. Pharmacol Res 1993; 27(1): 9-10.
30 Khan IA, Allgood J, Walker LA, et al. Determination of
heavy metals and pesticides in ginseng products. J AOAC
Int 2001; 84(3): 936-939.
31 Kumar A. Environmental Contamination bioreclamation.
1st ed. New Delhi: APH Publisher, 2004: 91-92.
32 Kirschmann JD. Nutrition Almanac. 6th ed. New York:
McGraw-Hill, 2006: 75-83.
33 Howe P, Malcolm HM, Dobson S. Manganese and its
compounds: environmental aspects. World Health Organi-
zation; 2004(63): 62-63.
124 February 15, 2013 | volume 33 | Issue 1 |