Food Control 47 (2015) 291e297

Contents lists available at ScienceDirect

Food Control
journal homepage: www.elsevier.com/locate/foodcont

Enterobacteriaceae contamination in chocolate processing

M.S. Nascimento*, E.M. Reolon, A.R.B. Santos, V.E. Moreira, N. Silva
~o Paulo, 13070-178, Brazil
Department of Microbiology, Institute of Food Technology, Av. Brasil, 2880, Campinas, Sa

a r t i c l e i n f o a b s t r a c t

Article history: A survey was conducted on two Brazilian chocolate manufacturing companies (A and B) to investigate/
Received 20 December 2013 determine Enterobacteriaceae, coliforms and Salmonella contamination. Samples of different chocolate
Received in revised form types, processed cocoa-based ingredients, manufacturing environment, and workers' hand surfaces were
11 July 2014
analyzed. Salmonella and Escherichia coli were not detected. However, in 25.8% of the chocolate samples,
Accepted 11 July 2014
Enterobacteriaceae were isolated while total coliforms were detected in 13.3%. Among the processed
Available online 19 July 2014
cocoa-based ingredients, two samples of cocoa liquor showed contamination by Enterobacteriaceae and
total coliforms, and in one sample thermotolerant coliforms were also recovered. For equipment and
Keywords:
Cocoa
utensils, Enterobacteriaceae were isolated in around 25.4% of the samples, whereas total coliforms were
Salmonella detected in 10.2%. Moreover, in company A 14.3% of manufacturing environment samples were
Coliforms contaminated by thermotolerant coliforms. The food handlers from both companies showed a high
Escherichia coli percentage of Enterobacteriaceae contamination on their hands. The results showed that the
Chocolate manufacturing environment including food handlers was considered the most likely Enterobacteriaceae
contamination source of the finished product. These data also highlight Enterobacteriaceae as a good
hygiene indicator for the chocolate industry.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction Nascimento, Brum, Pena, Berto, & Efraim, 2012), this process cannot
be considered as an effective control step.
Outbreaks of Salmonella associated with consumption of choc- According to ICMSF (2005, 2011), cocoa roasting is the main step
olate have been reported over the last 40 years and despite tech- in cocoa-chocolate processing responsible for the reduction of
nological development, continue to occur today (D'Aoust, 1977; Salmonella. Nonetheless, depending on the initial load and the
Werber et al., 2005). Although Salmonella cannot grow in this process parameters, cocoa roasting cannot ensure a significant
kind of product, it can remain viable for a long period of time decrease of the Salmonella count. Nascimento et al. (2012) observed
(Tamminga, Beumer, Kampelmacher, & van Leusden, 1976) which is a heat resistance of this pathogen during cocoa hot air roasting,
due to the low water activity (0.3e0.5) and high fat content (>20%) with D110
C ranging from 4.8 to 8.9 min and D140
C of ca. 2.5 min.
present in chocolate. These characteristics also provide a great heat However, thermal resistance of Salmonella can be reduced by
resistance to Salmonella (D'Aoust, 1977; Podolak, Enache, Stone, increasing the moisture of the product (Izurieta & Komitopoulou,
Black, & Elliot, 2010). Furthermore, the high fat content protects 2012).
the bacteria against gastric acidity resulting in an extremely low Due to great heat resistance and hence the absence of an
infectious dose. In Canada a contamination level of 0.043 MPN/g effective kill step during the chocolate manufacturing, the micro-
was found in chocolate samples involved in an outbreak caused by biological quality of the raw materials and environment plays an
S. Nima (Hockin et al., 1989). important role in the safety of the end product. Cocoa products are
In chocolate making, the conching process that uses tempera- not the only ingredients that may introduce Salmonella into choc-
tures of 50e80
C (Beckett, 2009) might contribute to reducing the olate, but they have been implicated as the potential source of some
initial microbial load of the chocolate mass. However, due to the outbreaks (Craven, Mackel, Baine, Barker, & Gangarosa, 1975;
heat resistance of Salmonella, previously reported in the literature G€astrin et al., 1972). Cordier (1994) points out raw cocoa beans as
(Goepfert & Biggie, 1968; Krapf & Gantenbein-Demarchi, 2010; a major source of Salmonella contamination in unclean zones
(storage and cleaning rooms or raw material handling zones). In a
previous study, we detected Salmonella in 3% and Escherichia coli in
* Corresponding author. Tel.: þ55 19 3743 1819; fax: þ55 19 37431799. 48% of dried cocoa beans stored on farms (Nascimento et al., 2010).
E-mail address: [email protected] (M.S. Nascimento). In samples from cocoa processing industries, 7% of cocoa liquor had

http://dx.doi.org/10.1016/j.foodcont.2014.07.020
0956-7135/© 2014 Elsevier Ltd. All rights reserved.
292 M.S. Nascimento et al. / Food Control 47 (2015) 291e297

total coliforms (Nascimento et al., 2011). Apart from cocoa, other out according to the method described by Midura and Bryant
ingredients such as milk powder, nuts and egg derivatives have (2001) on product contact surfaces (equipment and utensils) at
already been implicated in Salmonella outbreaks (Bell & Kyriakides, different processing steps, e.g., mixing/milling, conching,
2002). Furthermore, the processing environment has also been tempering, molding, cooling and packaging. From each product
pointed out as the contamination source of finished products (CDR- contact surface, five areas of 10 cm2 (total of 50 cm2) were chosen
HPA 2006; Kapperud et al., 1990). randomly and sampled using a swab. After that, the swab was
Enterobacteriaceae and coliforms have been widely used as homogenized in 50 ml of Phosphate Buffer (PB) and the solution
universal hygiene indicators for products and environment. Iden- obtained was used for microbiologic determinations. In the case of
tifying possible entry points for contamination through Enter- smaller utensils and workers' hands, the entire surface was
obacteriaceae enumaration enables the establishment of effective sampled. Collections of the air inside the plant were carried out by
measures to prevent or at least minimize the contamination risk of the sedimentation method, using plate count agar (Rice, Baird,
final product for Salmonella. When quality control programs have Eaton, & Clesceri, 2012). Swab samples from the environment and
already been implemented, the monitoring of Enterobacteriaceae hand surfaces were maintained below 4
C until the time of
can be used to predict deviations and defects in the hygiene control analysis.
measures (ICMSF, 2011). For the chocolate industry, ICMSF (2011)
recommends the determination of Enterobacteriaceae for in- 2.2. Microbiological analysis
process samples (residues from product contact surfaces), pro-
cessing environment and end products. However, Enter- For quantitative analysis of Enterobacteriaceae, pour plate
obacteriaceae and/or total coliforms investigation cannot always method with overlay was carried out using Violet Red Bile Agar
replace the direct monitoring of Salmonella, since low levels do not supplemented with Glucose (VRBG, Difco, Sparks, USA), with in-
guarantee the absence of this pathogen (Cordier, 2008). cubation at 35
C/24 h (Kornacki & Johnson, 2001). Most Probable
Although Salmonella has been recognized as a potential hazard Number (MPN) method was used to determine total and thermo-
for the chocolate industry (Cordier, 1994; ICMSF 2011), there are no tolerant coliforms, and E. coli. The test was carried out in pre-
Brazilian survey reports and few published international data on sumptive Tryptose Lauryl Sulfate (Difco). After incubation at 35
C/
contamination of Salmonella or other Enterobacteriaceae in end 24-48 h, confirmation of total coliforms was performed in Brilliant
products or processing environments (Barrera, Blanco, & Agut, Green Bile broth (Difco) with incubation at 35
C/24-48 h and
2001; CDR-HPA 2006; Craven et al., 1975; Torres-Vitela, Escartin, thermotolerant coliforms in EC broth (Difco) with incubation at
& Castillo, 1995). For this reason, a wide-ranging project to evaluate 45.5
C/24 h. Confirmation of E. coli was carried out by biochemical
microbial contamination throughout the Brazilian production chain tests (indole, Voges-Proskauer, methyl red and citrate) (Kornacki &
of cocoa and chocolate was carried out (Nascimento et al., 2010; Johnson, 2001). For aerobic mesophilic plate count agar (PCA, Difco)
2011). The aim of this part of the study was to investigate the was used with incubation at 35
C/48 h (Morton, 2001).
presence of Enterobacteriaceae, coliforms, E. coli and Salmonella in Analysis for Salmonella was performed according to the Food
the processing environment, cocoa-based ingredients and finished and Drug Administration method (Andrews & Hammack, 2007).
products in two Brazilian chocolate manufacturing companies. The pre-enrichment of 25 g of cocoa liquor and chocolate was
carried out in 225 ml of 10% reconstituted skimmed milk (Nestle,
2. Material and methods Brazil) supplemented with brilliant green solution; for product
contact surfaces and workers' hands, 25 ml of the swab solution
2.1. Sampling was added to 225 ml of Buffered Peptone Water (BPW, Difco). After
incubation at 35
C for 18e24 h, 0.1 ml of each portion was added to
Two different chocolate manufacturing companies (A and B), 10 ml of Rappaport-Vassiliadis Modified broth (Difco) and 1.0 ml to
located in Sa ~o Paulo State, Brazil, were evaluated for Enter- 10 ml Tetrathionate broth (Difco). The enrichment broths were
obacteriaceae, coliforms and Salmonella contamination. Three visits incubated for 24 h at 42
C and 35
C, respectively. After that,
were conducted in each processing plant over a twelve-month cultures were streaked on Xylose Lysine Deoxycholate agar (XLD,
period. Samples of chocolate (milk chocolate, n ¼ 32; dark choco- Difco), Hecktoen Enteric agar (HE, Difco) and Bismuth Sulfite agar
late, n ¼ 30; white chocolate, n ¼ 30; light chocolate, n ¼ 13; 70% (BS, Difco) and the plates were incubated at 35
C for 24e48 h.
cocoa chocolate, n ¼ 15), processed cocoa-based ingredients (cocoa Presumptive Salmonella colonies were confirmed by biochemical
butter, n ¼ 120; cocoa liquor, n ¼ 90), environment (equipment and (triple sugar iron, lysine decarboxylase, Voges-Proskauer, urease,
utensils, n ¼ 59), and hand surface of food handlers (n ¼ 6) were indole, and b-galactosidase) and serological tests. The result was
analyzed (Table 1). In addition, the quality of air inside the facilities expressed as presence or absence in 25 g, cm2 or ml.
was determined by analysis of total aerobic mesophilic count in 18
samples. 2.3. Statistical analysis
For chocolate and processed cocoa-based ingredients used in
the manufacturing of the sampled chocolate, 300-g samples from A correlation test was performed in Excel for hygiene indicator
the same lot were collected. Environmental sampling was carried microorganism counts.

Table 1
Sampling plan carried out in two chocolate processing companies.

Company Sampling

Finished product Cocoa-based ingredients Environment

Milk Dark White Light 70% Cocoa Cocoa Cocoa Equipment Utensils Air Workers'
chocolate chocolate chocolate chocolate chocolate liquor butter hands

A 15 15 15 e 15 45 60 17 11 7 3
B 17 15 15 13 e 45 60 18 13 11 3
Total 32 30 30 13 15 90 120 35 24 18 6
 M.S. Nascimento et al. / Food Control 47 (2015) 291e297 293

Table 2
Analysis results of chocolates and their respective cocoa-based ingredients from company A.

Product Contamination parameter Enterobacteriaceae Total coliforms Thermotolerant

(log cfu/g)a (log MPN/g)b coliforms
(log MPN/g)b

Milk chocolate Positive samples 6 (40%) 6 (40%) 0

Counts in positive samples (mean) 1.0 to 1.5 (1.2) 0.6 to 1.0 (0.8) <0.5
Cocoa liquor Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Dark chocolate Positive samples 5 (33.3%) 2 (13.3%) 0
Counts in positive samples (mean) 1.0 to 1.7 (1.3) 0.6 to 1.0 (0.8) <0.5
Cocoa liquor Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
70 % cocoa chocolate Positive samples 1 (6.7%) 1 (6.7%) 0
Counts in positive samples (mean) 2.4 0.6 <0.5
Cocoa liquor Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
White chocolate Positive samples 5 (33.3%) 1 (6.7%) 0
Counts in positive samples (mean) 1.0 to 1.3 (1.1) 0.6 <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Total chocolate Positive samples 17 (28.3%) 10 (16.7%) 0
Counts in positive samples (mean) 1.0 to 2.4 (1.5) 0.6 to 1.0 (0.8) <0.5
Total cocoa-based ingredient Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
a
Colony form unit, detection limit 1.0 log cfu/g.
b
Most probable number, detection limit 0.5 log MPN/g.

3. Results 0.30 to 3.99 log cfu/cm2 or ml. Salmonella and E. coli were not
detected. Enterobacteriaceae were isolated from two samples of
3.1. Chocolate equipment and six of utensils. In four of these samples, the counts
exceeded 2 log cfu/cm2 or ml. In addition, total coliforms were
Salmonella, E. coli and thermotolerant coliforms were not observed in five samples (1 from equipment and 4 from utensils).
detected in any of the 120 chocolate samples. In company A, six out Four samples of utensils also showed thermotolerant coliforms
of 15 samples of milk chocolate, five of dark chocolate and five of (Table 4). The mixer was the equipment in which the highest
white chocolate showed Enterobacteriaceae count, ranging from Enterobacteriaceae count (2.3 log cfu/cm2) was detected, whereas
1.00 to 1.70 log cfu/g. In one sample of 70% cocoa chocolate a count for utensils spatulas and chocolate molds had the highest
of 2.36 log cfu/g was observed. Total coliforms were isolated from contamination.
six samples of milk chocolate, two of dark chocolate, one of white In company B, aerobic mesophilic microorganisms were recov-
chocolate and one of 70% cocoa chocolate (Table 2). In company B, ered in 24 out of 31 samples, with counts between 0.30 and
Enterobacteriaceae was detected in nine samples (52.9%) of milk 2.18 log cfu/cm2 or ml. Salmonella, E. coli and thermotolerant co-
chocolate, one (7.1%) of dark chocolate and two (13.3%) of white liforms were not detected. In seven environmental samples (4 from
chocolate, with counts between 1.0 and 1.8 log cfu/g. One out of 13 equipment and 3 from utensils) Enterobacteriaceae were isolated;
light chocolate samples showed a count of 2.27 log cfu/g. In regard the highest count 1.65 log cfu/cm2 was observed in one sample
to total coliforms, the same number of contaminated samples (2) from a chocolate mold. On the other hand, total coliforms were
was shown for milk, white and light chocolate (Table 3). recovered in only one sample (spatula) (Table 5).

3.2. Cocoa processed ingredients 3.4. Air

Salmonella and E. coli were absent in all analyzed samples. No In the monitoring of air quality of the processing environment,
samples from company A showed contamination (Table 2). In the average counts for aerobic mesophilic were 0.28 log cfu/plate
company B two samples of cocoa liquor e one used in the pro- for company A and 0.74 log cfu/plate for company B, respectively.
cessing of milk chocolate and the other in light chocolate - Enter-
obacteriaceae (1.30 and 2.22 log cfu/g) and total coliforms (1.36 and
3.5. Food handlers
0.96 log MPN/g) were recovered (Table 3). Furthermore, the sample
used as an ingredient for milk chocolate also showed thermoto-
As to workers' hands, aerobic mesophilic count ranged from 1.15
lerant coliforms (0.96 log MPN/g). to 4.05 log cfu/ml, and Salmonella and E. coli were not detected.
Four out of six food handlers (two from each company) showed
3.3. Processing environment Enterobacteriaceae on their hands, with counts between 0.01 and
1.26 log cfu/ml. In addition, total coliforms were isolated from the
In company A, 24 out of 28 samples of the equipment and hand surface of three workers. Thermotolerant coliforms were also
utensils showed aerobic mesophilic microorganisms, ranging from recovered from one of the food handlers of company A (Table 6).
294 M.S. Nascimento et al. / Food Control 47 (2015) 291e297

Table 3
Analysis results of chocolates and their respective cocoa-based ingredients from company B.

Product Contamination parameter Enterobacteriaceae Total coliforms Thermotolerant

(log cfu/g)a (log MPN/g)b coliforms
(log MPN/g)b

Milk chocolate Positive samples 9 (52.9%) 2 (11.8%) 0

Counts in positive samples (mean) 1.0 to 1.7 (1.3) 0.6 to 1.0 (0.8) <0.5
Cocoa liquor Positive samples 1 (5.9%) 1 (5.9%) 1 (5.9%)
Counts in positive samples (mean) 1.3 1.4 1.0
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Dark chocolate Positive samples 1 (6.7%) 0 0
Counts in positive samples (mean) 1.0 <0.5 <0.5
Cocoa liquor Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Light chocolate Positive samples 2 (15.4%) 2 (15.4%) 0
Counts in positive samples (mean) 1.0 to 2.3 (1.6) 0.6 to 2.0 (1.3) <0.5
Cocoa liquor Positive samples 1 (7.7%) 1 (7.7%) 0
Counts in positive samples (mean) 2.2 1.0 <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
White chocolate Positive samples 2 (13.3%) 2 (13.3%) 0
Counts in positive samples (mean) 1.5 to 1.8 (1.7) 0.6 to 1.6 (1.1) <0.5
Cocoa butter Positive samples 0 0 0
Counts in positive samples (mean) <0.5 <0.5 <0.5
Total chocolate Positive samples 14 (23.3%) 6 (10%) 0
Counts in positive samples (mean) 1.0 to 2.3 (1.7) 0.6 to 2.0 (1.4) <0.5
Total cocoa raw material Positive samples 2 (1.9%) 2 (1.9%) 1 (1%)
Counts in positive samples (mean) 1.3 to 2.2 (1.8) 1.0 to 1.4 (1.2) 1.0
a
Colony form unit, detection limit 1.0 log cfu/g.
b
Most probable number, detection limit 0.5 log MPN/g.

4. Discussion Enterobacteriaceae count higher than the M level e one of 70% co-
coa chocolate from company A (Table 2) and one of light chocolate
This is the first study carried out in the Brazilian chocolate in- from company B (Table 3). However, despite the low number of
dustry that has investigated Enterobacteriaceae contamination samples, this information requires attention by both companies
throughout the chocolate processing line, from cocoa processed since it may indicate a specific failure in the GHP program.
ingredients to end products. Salmonella and E. coli were not Although cocoa raw materials have been considered a possible
detected in any of the 395 samples analyzed. However, the pres- source of Salmonella (Cordier, 1994; ICMSF, 2000), in the present
ence of both microorganisms in chocolate has already been re- study cocoa-based ingredients showed a low percentage of
ported in literature. Torres-Vitela et al. (1995) isolated Salmonella contamination by the bacteria groups investigated (<1%). Based on
from two samples of wrapped chocolate. In Spain, Barrera et al. the results, cocoa liquor could be indicated as one of the likely
(2001) detected E. coli in 7% of the analyzed samples, with counts contamination sources of only one chocolate milk sample from
of 0.6 and 1.1 log MPN/g. company B (Table 3). In a previous study, a slightly higher per-
Besides Salmonella and E. coli, thermotolerant coliforms were centage of Enterobacteriaceae contamination (7%) was found in the
not detected in any chocolate sample either (Tables 2 and 3). same kind of product collected directly from the cocoa processor
Therefore, all analyzed chocolate samples were in accordance with industry (Nascimento et al., 2011). Neither ANVISA (2001) nor
the microbiological criteria established by the Brazilian National ICMSF (2011) establish Enterobacteriaceae limits for cocoa liquor or
Health Surveillance Agency (ANVISA), which determine the cocoa butter. However, if the ICMSF's guiding values for the end
absence of Salmonella in 25 g of product and the thermotolerant product are used as a means of comparing the results, the detection
coliform count up to 1 log cfu/g (ANVISA, 2001). of Enterobacteriaceae count above 2 log cfu/g, even in one sample,
In regard to the other microorganisms investigated, company A would suggest a possible failure in the supplier's quality assurance
showed a higher contamination rate of Enterobacteriaceae and co- system. As only the cocoa raw materials were evaluated, it is not
liforms when compared to company B; 28.3% versus 23.3% and possible to rule out the possibility of some other ingredient playing
16.7% versus 10%, respectively. A previous study carried out by an important role as a contamination vehicle of the end product.
Torres-Vitela et al. (1995) detected coliforms in ca. 30% of the According to Beckett (2009), hygiene processes and working
chocolate marketed in Mexico. The distribution of contamination practices are essential to prevent Salmonella cross-contamination
among the chocolate types was significantly different between the during chocolate making. Based on the results, the manufacturing
companies. In company A, the Enterobacteriaceae contamination environment was considered the most likely Enterobacteriaceae
rate was not very different when comparing milk (6/15), dark (5/15) contamination source of the end product. Company A showed the
and white chocolate (5/15), whereas in company B milk chocolate highest contamination level of all hygiene indicators evaluated. The
showed the highest occurrence of Enterobacteriaceae (60%). The contamination was detected more frequently in utensils, except for
Enterobacteriaceae guiding values of ICMSF (2011) for finished two types of equipment from the unclean zone (mixing and milling
chocolate products are 1 log cfu/g as the limit separating acceptable machines). Seven samples from this company showed Enter-
from marginally acceptable (m) and 2 log cfu/g as marginally obacteriaceae contamination >1 log cfu/cm2 and in four the counts
acceptable from unacceptable (M). Only two samples showed were >2 log cfu/cm2 (Table 4). In Brazil, there are not any
 M.S. Nascimento et al. / Food Control 47 (2015) 291e297 295

Table 4
Analysis results of equipment and utensils from company A.

Sample Number of samples Determination

Aerobic mesophilic count Enterobacteriaceae Total coliforms Thermotolerant coliforms

(cfu/cm2 or utensil) (log cfu/cm2 or utensil) (log MPN/cm2 or utensil) (log MPN/cm2 or utensil)

Mixer 1 3.0 2.3 0.4 <0.5

Milling machine 2 2.2 1.9 <0.5 <0.5
1.3 <0 <0.5 <0.5
Wheeled tank 1 1.5 <0 <0.5 <0.5
Conching strainer 1 2.4 <0 <0.5 <0.5
Conching outlet valve 1 0.3 <0 <0.5 <0.5
Tank before Tempering 1 2.0 <0 <0.5 <0.5
Tempering 1 1.6 <0 <0.5 <0.5
Tank before filling 1 <0 <0 <0.5 <0.5
Filler nozzle 1 2.0 <0 <0.5 <0.5
Cooling tunnel 1 1.2 <0 <0.5 <0.5
Coating machine 2 1.3 <0 <0.5 <0.5
0.9 <0 <0.5 <0.5
Cutting machine 1 <0 <0 <0.5 <0.5
Scales 1 0.6 <0 <0.5 <0.5
Wrapping machine 2 1.0 <0 <0.5 <0.5
1.9 <0 <0.5 <0.5
Transport box 1 1.3 1.5 <0.5 <0.5
Mold 5 3.5 2.9 0.6 0.4
1.5 <0 <0.5 <0.5
3.5 2.9 1.4 1.4
<0 <0 <0.5 <0.5
1.0 <0 <0.5 <0.5
Spatula 2 4.0 2.4 <0.5 <0.5
2.1 0.6 0.4 0.4
Transport tray 2 3.3 0.7 0.4 0.4
3.1 <0 <0.5 <0.5
Butter paper 1 <0 <0 <0.5 <0.5

Positive samples 28 24 (85.7%) 8 (28.6%) 5 (17.9%) 4 (14.3%)

Counts in positive samples (mean) 1.9 1.9 0.3 0.1

*Colony form unit, detection limit 0.1 log cfu/cm2 or utensil. **Most probable number, detection limit 0.5 log MPN/cm2 or utensil.

microbiological criteria established for the chocolate processing identified as a cause of Salmonella outbreaks linked to consumption
environment. Due to the characteristics of some raw materials and of chocolate (Cordier, 1994). Craven et al. (1975) recovered Salmo-
the process employed, a low level of Enterobacteriaceae contami- nella from six out of 286 environmental samples e four samples
nation is expected. However, detection of this group above the from the bean processing room and two from the molding plant.
limits recommended by ICMSF (2011), 2e3 log cfu/cm2 for pro- The air contamination was very low in both companies, and
cessing environments and 1 log cfu/cm2 for in-process samples, there was no significant difference between the unclean and clean
may indicate failures in the hygiene control measures. It should be zones. In the literature there is not any report on the recommen-
noted that in the present study, on many occasions, the environ- dation level for contamination of the inside air in the food
mental sampling was performed during the operation time on manufacturing environment. Nevertheless, the microbial counts
product contact surfaces, i.e., these samples can be classified as in- detected in both plants can be considered acceptable when
process samples. In company B, the contamination was detected at compared to the criterion established for laboratories, which is
a lower level and in a more dispersed form throughout the pro- 1.5 log cfu/plate (Rice et al., 2012).
cessing line. Only one sample of chocolate mold showed Enter- Enterobacteriaceae is classified as a transient microorganism
obacteriaceae count >1 log cfu/cm2 (Table 5). In company A the group on the skin (ICMSF, 1988). Hence, the detection of it on hands
hygiene procedures include dry cleaning for the processing line and indicates inadequate hygienic habits or a possible contamination of
wet cleaning for the utensils, while in company B dry cleaning was hands during the work. Usually, in large chocolate processing
carried out daily while once a week wet cleaning was performed. plants, due to automation of the production line, there is almost no
The contamination observed in the manufacturing environment of human intervention. However, in small or medium ones, such as
company B and in the utensils and molds in company A may be those surveyed in this study, frequent product handling can be
related to the practice of wet cleaning, since it is well known that observed. This information associated with the Enterobacteriaceae
water provides favorable conditions for microbial growth. Besides, detection rate (66.7%) obtained on workers' hands from both plants
even though quality control programs such as HACCP and GMP (Table 6) point out the food handlers as a potential contamination
have been implemented in both surveyed companies, there were vehicle for the end product in companies A and B. According to
no robust hygiene barriers from unclean to clean zones; traffic Scott et al. (2009) improper handling practices and employee traffic
patterns for both workers and tools between these areas were not patterns may be one of the causes responsible for the introduction
uncommon. It might also contribute to spreading contamination in of Salmonella into the low-moisture food processing environment.
the processing environment. Furthermore, particularly in the case In addition, in this study aerobic mesophilic count, Enter-
of company A, the findings of the processing environment require obacteriaceae and coliforms were tested to determine the degree of
attention, since a high level of Enterobacteriaceae suggests an correlation between them when used as hygiene indicators in the
increased risk for the presence of Salmonella. In addition, deviations chocolate processing environment. A good correlation was not
in hygiene programs and good manufacturing practices have been observed between aerobic mesophilic and Enterobacteriaceae
296 M.S. Nascimento et al. / Food Control 47 (2015) 291e297

Table 5
Analysis results of equipment and utensils from company B.

Sample Number of samples Determination

Aerobic mesophilic count Enterobacteriaceae Total coliforms Thermotolerant coliforms

(log cfu/cm2 or utensil)a (log cfu/cm2 or utensil)a (log MPN/cm2 or utensil)b (log MPN/cm2 or utensil)b

Wheeled tank before conching 1 <0 <0 <0.5 <0.5

Conching outlet valve 1 1.7 0.01 <0.5 <0.5
Wheeled tank tube connection 2 <0 <0 <0.5 <0.5
1.7 0.5 <0.5 <0.5
Tray transport cart 1 1.7 <0 <0.5 <0.5
Tempering machine 1 1.3 <0 <0.5 <0.5
Tempering tank outlet valve 1 <0 <0 <0.5 <0.5
Tank before filling 1 1.1 <0 <0.5 <0.5
Filling machine 3 1.2 1.0 <0.5 <0.5
0.3 <0 <0.5 <0.5
1.3 <0 <0.5 <0.5
Vibrator table 1 1.8 0.3 <0.5 <0.5
Cooling tunnel 2 2.1 <0 <0.5 <0.5
0.7 <0 <0.5 <0.5
Unmoulding table 2 1.0 <0 <0.5 <0.5
1.7 <0 <0.5 <0.5
Scales 1 0.6 <0 <0.5 <0.5
Wrapping machine 2 0.5 <0 <0.5 <0.5
Storage Box before wrapping 1 <0 <0 <0.5 <0.5
Packing paper 2 0.7 <0 <0.5 <0.5
1.0 0.5 <0.5 <0.5
Transport Box 2 <0 <0 <0.5 <0.5
<0 <0 <0.5 <0.5
Mold 4 1.2 <0 <0.5 <0.5
2.2 1.7 <0.5 <0.5
0.3 <0 <0.5 <0.5
0.5 <0 <0.5 <0.5
Spatula 2 0.5 0.3 0.4 <0.5
1.9 <0 <0.5 <0.5
Knife 1 <0 <0 <0.5 <0.5
Tray 1 0.5 <0 <0.5 <0.5
Positive samples 31 24 (77.4%) 7 (22.6%) 1 (3.2%) 0
Counts in positive samples (mean) 1.1 0.6 0.4 e
a
Colony form unit, detection limit 0.1 log cfu/cm2 or utensil.
b
Most probable number, detection limit 0.5 log MPN/cm2 or utensil.

counts (r2 ¼ 0.43). Enterobacteriaceae were isolated in only 31.2% of In conclusion, this study examined a considerable number of
the samples that showed aerobic mesophilic microorganisms. Due samples throughout chocolate manufacturing. Although the pro-
to the high percentage of recovery (81.4%) and low correlation with cessing environment showed a contamination level above those
Enterobacteriaceae, the aerobic mesophilic count cannot be suggested by ICMSF, the microbiological quality of 98% of the
considered a good hygiene indicator for the chocolate processing chocolate samples evaluated was satisfactory. Nevertheless, the
environment. From all samples analyzed (end product, raw mate- results highlight the need to reinforce preventive control measures,
rial and environment), 52 presented Enterobacteriaceae and 27 especially hygiene and working practices, in chocolate processing
coliforms. Coliforms were the only microorganisms isolated in plants to obtain a safe product for consumption.
three chocolate samples from company A. It is worth noting that in
these samples the contamination level recovered was very low Acknowledgments
(0.55 log MPN/g). Furthermore, a weak correlation was observed
between Enterobacteriaceae and total coliforms (r2 ¼ 0.32). There- The authors wish to thank Fundaça~o de Amparo a Pesquisa do
fore, for the conditions and samples tested in this study Enter- ~o Paulo (FAPESP) process 2007/50519-1 and Conselho
Estado de Sa
obacteriaceae seemed to be a better hygiene indicator for the Nacional de Desenvolvimento Científico e Tecnolo  gico (CNPq)
chocolate industry. process 578400/200/8-1 for financial support.

Table 6
Analysis results of food workers' hand surface from companies A and B.

Microorganism group Company A Company B

Positive samples (%) Counts in positive samples (mean) Positive samples (%) Counts in positive samples (mean)

Aerobic plate count (log cfu/ml)a 3 (100%) 1.9 to 3.6 (2.5) 3 (100%) 1.1 to 4.0 (2.4)
Enterobacteriaceae (log cfu/ml)a 2 (66.7%) 1.0 to 1.3 (1.2) 2 (66.7%) 0.01 to 0.7 (0.4)
Total coliforms 2 (66.7%) 0.2 to 0.4 (0.3) 1 (33.3%) 0.04
(log MPN/ml)b
Thermotolerant coliforms 1 (33.3%) 0.4 0 <0.5
(log MPN/ml)b
a
Colony form unit, detection limit 0.01 log cfu/ml of hand surface swab.
b
Most probable number, detection limit 0.5 log MPN/ml of hand surface swab.
 M.S. Nascimento et al. / Food Control 47 (2015) 291e297 297

References Izurieta, W. P., & Komitopoulou, E. (2012). Effect of moisture on Salmonella spp heat
resistance in cocoa and hazelnut shells. Food Research International, 45,
^ncia Nacional de Vigila ^ncia Sanita
ria e ANVISA. (2001). Resoluça ~o RDC n
12, de 1087e1092.
Age
Kapperud, G., Gustavsen, S., Hellesnes, I., Hansen, A. H., Lassen, J., Hirn, J., et al.
02 de janeiro de 2001. Aprova regulamento t ecnico sobre padro ~es microbiolo gicos
rio Oficial, Brasília, 10 jan. 2001. Seça~o 1. (1990). Outbreak of Salmonella typhimurium infection trace to contaminated
para alimentos. Dia
chocolate and caused by a strain lacking the 60-megadalton virulence plasmid.
Andrews, W. H., & Hammack, T. S. (2007). Salmonella. In Food and drug adminis-
Journal of Clinical Microbiology, 28, 2597e2601.
tration, bacteriological analytical manual. online http://www.cfsan.fda.gov/
Kornacki, J. L., & Johnson, J. L. (2001). Enterobacteriaceae, coliforms and Escherichia
ebam/bam-5.html.
lisis microbiologico de tabletas de coli as quality and safety indicatiors. In F. P. Downes, & K. Ito (Eds.), Standard
Barrera, M. C., Blanco, M. J., & Agut, M. (2001). Ana
methods for the microbiological examination of foods (4th ed.) (pp. 69e82).
chocolate y cacao em polvo. Alimentaria, 38(326), 71e74.
Washington: American Public Health Association (APHA).
Beckett, S. T. (2009). Industrial chocolate manufacture and use (4th ed.). London:
Krapf, T., & Gantenbein-Demarchi, C. (2010). Thermal inactivation of Salmonella spp
Wiley-Blackwell.
during conching. LWT e Food Science and Technology, 43, 720e723.
Bell, C., & Kyriakides, A. (2002). Salmonella e A practical approach to the organism
Midura, T. F., & Bryant, R. G. (2001). Sampling plants, sample collection, shipment,
and its control in foods. Malden: Blackwell Science.
and preparation for analysis. In F. P. Downes, & K. Ito (Eds.), Standard methods
CDR-HPA, Health Protection Agency e UK. (2006). National increase in human
for the microbiological examination of foods (4th ed.) (pp. 13e23). Washington:
Salmonella Montevideo infections in England and Wales: March to June 2006.
American Public Health Association (APHA).
CDR: The Communicable Disease Report, Weekly, 16(26), 1e2.
Morton, R. D. (2001). Aerobic plate count. In F. P. Downes, & K. Ito (Eds.), Standard
Cordier, J. L. (1994). HACCP in the chocolate industry. Food Control, 5, 171e175.
methods for the microbiological examination of foods (4th ed.) (pp. 63e67).
Cordier, J. L. (2008). Production of powdered infant formula and microbiological
Washington: American Public Health Association (APHA).
control measures. In J. M. Faber, & S. Forsythe (Eds.), Enterobacter sakazakii (pp.
Nascimento, M. S., Brum, D., Pena, P. O., Berto, M. I., & Efraim, P. (2012). Inactivation
145e185). Washington, DC: ASM Press.
of Salmonella during cocoa roasting and chocolate conching. International
Craven, P. C., Mackel, D. C., Baine, W. B., Barker, W. H., & Gangarosa, E. J. (1975).
Journal of Food Microbiology, 159, 225e229.
International outbreak of Salmonella Eastbourne infection traced to contami-
Nascimento, M. S., Reolon, E. M., Santos, A. R. B., Moreira, V. E., Silva, I. F., & Silva, N.
nated chocolate. Lancet, 1(7910), 788e792.
(2011). Enterobacteriaceae in processed cocoa products. Revista Instituto Adolfo
D'Aoust, J. Y. (1977). Salmonella and the chocolate industry a review. Journal of Food
Lutz, 70, 81e85.
Protection, 40(10), 718e727.
Nascimento, M. S., Silva, I. F., Silva, J. C., Reolon, E. M., Santos, A. R. B., & Silva, N.
G€astrin, B., Kampe, A., Nystrom, K. G., Oden-Johanson, B., Wessel, G., & Zetterberg, B.
(2010). Enteropathogens in cocoa pre-processing. Food Control, 21, 408e411.
(1972). Salmonella durham epidemic caused by contaminated cocoa. Lakartid-
Podolak, R., Enache, H., Stone, W., Black, D. G., & Elliot, P. (2010). Sources and risk
ningen, 69(46), 5335e5338.
factors for contamination, survival, persistence, and heat resistance of Salmo-
Goepfert, J. M., & Biggie, R. A. (1968). Heat resistance of Salmonella Typhimurium and
nella in low e moisture foods. Journal of Food Protection, 73, 1919e1936.
Salmonella Senftenberg 775W in milk chocolate. Applied Microbiology, 16,
Rice, E. W., Baird, R. B., Eaton, A. D., & Clesceri, L. S. (2012). Standard methods for the
1939e1940.
examination of water and wastewater (22nd ed.). Washington, DC: American
Hockin, J. C., D'Aoust, J. Y., Bowering, D., Jessop, J. H., Khanna, B., Lior, H., et al. (1989).
Public Health Association, American Water Works Association & Water Envi-
An international outbreak of Salmonella nima from imported chocolate. Journal
ronment Federation (Part 9000).
of Food Protection, 52, 51e54.
Scott, V. N., Chen, Y., Freier, T. A., Kuehm, J., Moorman, M., Meyer, J., et al. (2009).
International Commission on Microbiological Specifications for Foods e ICMSF.
Control of Salmonella in low-moisture foods. I: minimizing entry of Salmonella
(1988). Microrganisms in foods 6 e Microbiological ecology of food commodities.
into a processing facility. Food Protection Trends, 29, 342e353.
Gaithersburg, Maryland: Aspen Publishers, Inc.
Tamminga, S. K., Beumer, R. R., Kampelmacher, E. H., & van Leusden, F. M. (1976).
International Commission on Microbiological Specifications for Foods e ICMSF.
Survival of Salmonella Eastbourne and Salmonella Typhimurium in chocolate.
(2000). Microrganisms in foods 4 e Application of the hazard analysis critical
Journal of Hygiene, 76, 41e47.
control point (HACCP) system to ensure microbiological safety and quality. Oxford:
Torres-Vitela, R., Escartin, E. F., & Castillo, A. (1995). Risk of Salmonellosis associated
Blackwell Scientific Publications.
with consumption of chocolate in Mexico. Journal of Food Protection, 58(5),
International Commission on Microbiological Specifications for Foods e ICMSF.
478e481.
(2005). Cocoa, chocolate, and confectionery. In Microrganisms in foods 6-
Werber, D., Dreesman, J., Feil, F., Van Treeck, U., Fell, G., Ethelberg, S., et al. (2005).
Microbiological ecology of food commodities (2nd ed.) (pp. 467e479). New
International outbreak of Salmonella Oranienburg due to German chocolate.
York: Kluwer Academic/Plenum Publishers.
BMC Infectious Diseases, 5, 7e17.
International Commission on Microbbiological Specifications for Foods e ICMSF.
(2011). Microrganisms in foods 8 e Use of data for assessing process control and
product acceptance. New York: Springer.