2019 Volume 25 Issue 2 Pages 265-275
Five hundred and sixteen food samples retailed in Japan were examined for viable bacterial counts (VBC), coliforms, and Cronobacter spp. Cronobacter spp. was detected in 12.4% (64/516) of the total food samples, comprising 15.4% (35/228) in agricultural products, 14.7% (28/190) in livestock products, 1.5% (1/65) in marine products, and 0% in infant foods (0/33). Cut and packed vegetables for cooking, flours, and beef offal showed the high prevalence of Cronobacter spp., whereas raw milk, marine products, and infant foods showed the lower prevalence. High correlations were observed between the levels of VBC and the prevalence of Cronobacter spp., and the levels of coliform and the prevalence of Cronobacter spp. in some types of food sample. Cronobacter spp. was seen in bovine feces (37.5%, 3/8), soil (16.7%, 4/24) and composts (10.0%, 1/10). These results suggested that the bovine feces might be one of the potent natural habitats of Cronobacter spp.
Cronobacter spp., former Enterobacter sakazakii, are Gram-negative, non-spore forming rods, belonging to Enterobacteriaceae family (Farmer et al., 1980; Iversen et al., 2007; Iversen et al., 2008; Forsythe, 2010).
Cronobacter infection is a rare disease, and approximately 4–6 cases per year occur in the USAi). Premature infants and low birth weight infants are recognized as high-risk groups; however, other age groups could also be infected with the pathogen. Cronobacter causes infantile sepsis, necrotizing enteritis and meningitis (Mutjens et al., 1983; Simmons et al., 1989; van Acker et al., 2001).
Powdered infant formula (PIF) is considered one of the most suspicious vehicles involved in the transmission of Cronobacter infections to infants (Drudy et al., 2006). Sporadic cases and outbreaks of Cronobacter infection, probably caused by PIF, were reported from the EU countries, the USA, and other countries (Drudy et al., 2006). A few cases of Cronobacter infection in low birth weight infants were also reported from Japan, although their routes of infection remained unclear (Ibuki et al., 2009; Teramoto et al., 2010).
In Japan, Cronobacter spp. including E. sakazakii contamination has been reported in PIF and other foods; however, the information is very limited. Oonaka et al. reported the prevalence of E. sakazakii in domestic and imported PIF in Japan, and showed that the prevalence was 6.6% (4/61) and 5.7% (5/88) in the domestic and imported PIF, respectively (Oonaka et al., 2010). Ogihara et al. examined the contamination of Cronobacter spp. in dried foods, and reported that spices, herbs and dried vegetables were highly contaminated with Cronobacter spp. (Ogihara et al., 2014) . Ueda reported that the prevalence of Cronobacter spp. was higher in fresh vegetables compared with dried foods and soil samples from agricultural fields were sometimes contaminated with this bacrterial species (Ueda, 2017). However, the contamination risks and routes are not completely revealed as described above, and more information about Cronobacter spp. habitats is needed.
In this study, we surveyed the prevalence and levels of contamination of Cronobacter spp. and other bacteria in various retail food samples and farm-associated environments, and aimed to clarify the natural habitats of the microorganisms.
Samples Two hundred and twenty eight agricultural products were collected and the level of bacterial contamination was determined. In detail, 50 whole vegetables (including 14 leafy vegetables, 10 root vegetables, 16 sprouts, and 10 mushrooms), 100 cut and packed vegetables for salad (including 15 single type of vegetable, 83 mixed salad vegetables), 30 cut and packed vegetables for cooking (including 23 vegetables for stir-frying, 7 vegetables for simmering), and 48 flours (including 6 buckwheat, 6 wheat, 5 rice, 2 corn, 2 oat, 2 rye, and 25 other types of flours) were collected and examined. All the samples except raw milk samples were purchased in Tokyo and Kanagawa Prefectures during April 2013 and March 2017. Raw milk samples were obtained from farm in Kanagawa Prefecture during the same period. Regarding livestock products, 190 samples were examined: 60 raw minced/cut meat (including 8 beef, 31 pork, 15 chicken, and 6 mix of beef and pork meat), 68 beef offal (including 27 stomachs, 10 small intestines, 14 large intestines, 3 rectums, 8 hearts, and 6 livers), 26 pork offal (including 4 stomachs, 2 small intestines, 5 large intestines, 7 hearts, 4 livers, 3 uteri, and 1 lung), 6 chicken offal (including 3 gizzards, 2 livers, and 1 heart), and 30 raw milk samples. Sixty five marine products, consisting of 8 whole raw fish (including 2 horse mackerel, 2 flatfish, 4 other fish), 8 sashimi (including 2 tuna, and 6 other fish), and 49 marinated fish for cooking (including 7 sardine, 7 salmon, 5 marlin, 5 mackerel, 2 saury, 2 yellowtail, and 21 other fish) and 33 infant foods were also examined. All samples of whole vegetables, cut and packed vegetables for salad consisted with single vegetable, cut and packed vegetables for cooking, rice, beef offal except 2 liver , pork and chicken offal, raw milk and infant foods were produced in Japan. About 2/3 of cut and packed vegetables for salad consisted with more than 2 types of vegetables were domestic foods, and the rest were mixed with domestic and imported vegetables. Flour samples and marine products were domestic and imported foods. From farm-associated environments, 24 soil and 10 compost, and 8 bovine feces samples were examined in this study. The soil samples were collected from farms in Tokyo and Kanagawa Prefectures from March 2015 to March 2016. The compost samples and bovine feces were collected in Kanagawa Prefecture during the same periods.
Isolation of viable bacterial counts (VBC) and coliforms Food samples were cut into small pieces with sterilized scissors. Ten grams or 25 g of each sample was weighed and suspended in 9 volumes of sterilized buffered peptone water (BPW; Oxoid Ltd., Basingstoke, Hampshire, UK), and homogenized for 1 min using a stomacher (Stomacher-400, Seward Medical, London, UK).
To determine the VBC, first, four 250 µL aliquots of initial suspension were plated onto 4 Plate Counts Agar (PCA; Difco, Detroit, MI) plates, as a result, 1 mL of initial suspension was used in total. Then, a part of the initial suspension was serially diluted with BPW, and 100 µL of each dilution was plated onto PCA plates. After incubation at 37 °C±1 °C for 48 h, the number of colonies was determined. All dilutions were plated in duplicate, and the averages of colony number were used as results.
To determine the number of coliforms, the samples prepared above were plated onto of XMG plates (XMG; NISSUI PHARMACEUTICAL CO., LTD., Tokyo, Japan) in the same method as VBC. After 24 h incubation at 37 °C±1 °C, the number of typical colonies with red color was calculated.
Isolation of Cronobacter spp. by qualitative and quantitative methods Detection of Cronobacter spp. was performed according to the method of ISO/TS 22964:2006 (ISO, 2006). Briefly, 10 g or 25 g of each sample was suspended in 9 volumes of sterilized BPW, and homogenized for 1 minute using a stomacher. The homogenate was incubated at 37 °C±1 °C for 18 h for pre-enrichment. After incubation, 0.1 mL of the pre-enrichment culture was inoculated into 10 mL of mLST-vancomycin medium (LVM; Oxoid), followed by incubation at 44 °C±1 °C for 24 h. One loop of incubated LVM was streaked onto Chromocult Enterobacter sakazakii agar plate (ESA; Merck KGaA, Darmstadt, Germany) and incubated at 44 °C±1 °C for 24 h. Five typical colonies with blue-green color were selected and steaked onto Tryptic soy agar plates (TSA; Difco). After incubation at 25 °C±1 °C for 48 h, the colonies were examined for the confirmation test.
To enumerate Cronobacter spp., a part of initial suspension and dilution prepared for VBC and coliforms were plated onto ESA plates in the same manner as VBC plates. After incubation at 44 °C±1 °C for 24 h, the number of typical colonies was counted.
For confirmation of Cronobacter spp. according to ISO/TS 22964:2006, 5 typical colonies on ESA agar plates were used. In case that the presumptive colonies were less than 5, all of the typical colonies were examined.
Stastical analysis Relationship between VBC and Cronobacter spp. contamination and between coliform counts and Cronobacter spp. contamination were statistically compared. The samples in each food category, such as vegetables, flours, and meat and offal, were divided as “higher” (approximately higher-half) and “lower” (approximately lower-half) contamination groups based on the VBC or coliform counts. The levels of “higher” and “lower” contamination were different among the food category. And the numbers of Cronobacter spp. positive samples in each group were compared by using 2x2 Fisher's exact test.
VBC of 516 food samples ranged from less than the detection limit (< 101 CFU/g) to 108 CFU/g (Table 1). The prevalence of VBC in whole vegetables, cut and packed vegetables for salad and cooking, raw meat, and offal was relatively high, and the prevalence of VBC in flours, raw milk, marine products, and infant foods was relatively low. Especially, half of the infant food samples showed VBC below the detection limit. Regarding coliforms, which are used as an indicator of fecal contamination, their prevalence was similar to that of VBC; bacterial contamination levels in raw milk and infant foods were low (Table 2).
Sample type | Number of samples | Level of contamination (CFU/g) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
<101 | 101 | 102 | 103 | 104 | 105 | 106 | 107 | 108 | |||
Agricultural products | |||||||||||
Whole vegetables | 50 | 1 | 8 | 9 | 12 | 19 | 1 | ||||
Cut and packed vegetables for salad | 100 | 9 | 35 | 35 | 18 | 3 | |||||
Cut and packed vegetables for cooking | 30 | 2 | 15 | 13 | |||||||
Flours | 48 | 5 | 6 | 19 | 8 | 9 | 1 | ||||
Total | 228 | 5 | 6 | 19 | 18 | 52 | 47 | 45 | 35 | 1 | |
Livestock products | |||||||||||
Raw minced meat | 60 | 2 | 4 | 18 | 20 | 14 | 1 | 1 | |||
Beef offal | 68 | 5 | 24 | 20 | 18 | 1 | |||||
Pork offal | 26 | 9 | 11 | 6 | |||||||
Chicken offal | 6 | 3 | 2 | 1 | |||||||
Raw milk | 30 | 6 | 13 | 4 | 4 | 3 | |||||
Total | 190 | 6 | 13 | 6 | 16 | 56 | 52 | 38 | 2 | 1 | |
Marine products | |||||||||||
Raw fish | 8 | 1 | 3 | 3 | 1 | ||||||
Sashimi | 8 | 1 | 3 | 2 | 2 | ||||||
Marinated fish for grill | 49 | 2 | 1 | 4 | 12 | 16 | 13 | 1 | |||
Total | 65 | 3 | 1 | 8 | 17 | 21 | 14 | 1 | |||
Others | |||||||||||
Infant foods | 33 | 16 | 13 | 3 | 1 | ||||||
Total | 33 | 16 | 13 | 3 | 1 |
Sample type | Number of samples | Level of contamination (CFU/g) | ||||||||
---|---|---|---|---|---|---|---|---|---|---|
<101 | 101 | 102 | 103 | 104 | 105 | 106 | 107 | |||
Agricultural products | ||||||||||
Whole vegetables | 50 | 7 | 3 | 5 | 13 | 9 | 13 | |||
Cut and packed vegetables for salad | 100 | 1 | 6 | 10 | 29 | 39 | 13 | 2 | ||
Cut and packed vegetables for cooking | 30 | 1 | 2 | 9 | 17 | 1 | ||||
Flours | 48 | 33 | 6 | 2 | 5 | 2 | ||||
Total | 228 | 41 | 12 | 15 | 40 | 56 | 31 | 32 | 1 | |
Livestock products | ||||||||||
Raw minced meat | 60 | 1 | 5 | 11 | 23 | 14 | 5 | 1 | ||
Beef offal | 68 | 6 | 12 | 18 | 15 | 11 | 6 | |||
Pork offal | 26 | 1 | 1 | 8 | 13 | 3 | ||||
Chicken offal | 6 | 2 | 2 | 1 | 1 | |||||
Raw milk | 30 | 29 | 1 | |||||||
Total | 190 | 37 | 21 | 31 | 47 | 39 | 14 | 1 | ||
Marine products | ||||||||||
Raw fish | 8 | 3 | 1 | 1 | 2 | 1 | ||||
Sashimi | 8 | 6 | 1 | 1 | ||||||
Marinated fish for grill | 49 | 21 | 9 | 11 | 4 | 2 | 2 | |||
Total | 65 | 30 | 10 | 13 | 7 | 2 | 3 | |||
Others | ||||||||||
Infant foods | 33 | 33 | ||||||||
Total | 33 | 33 |
Agricultural products were highly contaminated with Cronobacter spp. (Table 3). In whole vegetables, 2 out of 10 (20%) root vegetables and 7 out of 16 (43.8%) sprouts were contaminated with Cronobacter spp. On the other hand, 14 leafy vegetables and 10 mushrooms were shown negative to this bacterial species by the qualitative method. Cut and packed vegetables for salad were mainly consisted with lettuce and cabbage, and showed a middle level of contamination. Cut and packed vegetables for cooking were mainly consisted with sprouts, and showed high prevalence (26.7%). In flours, Cronobacter spp. were detected in 4 buckwheat, 1 wheat, 1 rice, and 4 other flour samples (1 wheat bran, 1 seed mix, 1 cereal, and 1 potato powders) samples. Regarding livestock products, the prevalence of Cronobacter spp. was relatively high, except for raw milk. Especially, the prevalence of Cronobacter spp. in beef minced meat, gastrointestinal tract of cattle and pigs was above 25%, although the contamination levels were below the detection limit by the quantitative method. The prevalence of Cronobacter spp. in marine products was low, and negative for all infant food samples tested in this study. The prevalences of Cronobacter spp. were not different between domestic and imported products in most food categories, except for buckwheat (data not shown). Buckwheat samples showed the positive results in 4 out of 6 samples (66.7%), and 3 out of 4 positive samples were imported (data not shown).
Category | Number of samples | Qualitative analysis | |
---|---|---|---|
Sample type Detail |
Number of positives | Percentage ofpositives | |
Agricultural products | 228 | 35 | 15.4 |
Whole vegetables | 50 | 9 | 18.0 |
leafy vegetables | 14 | 0 | 0.0 |
root vegetables | 10 | 2 | 20.0 |
sprouts | 16 | 7 | 43.8 |
mushrooms | 10 | 0 | 0.0 |
Cut and packed vegetables for salad | 100 | 8 | 8.0 |
single | 15 | 0 | 0.0 |
mixed | 85 | 8 | 9.4 |
Cut and packed vegetables for cooking | 30 | 8 | 26.7 |
for stir frying | 23 | 6 | 26.1 |
for simmering | 7 | 2 | 28.6 |
Flours | 48 | 10 | 20.8 |
buckwheat | 6 | 4 | 66.7 |
wheat | 6 | 1 | 16.7 |
rice | 5 | 1 | 20.0 |
corn | 2 | 0 | 0.0 |
oat | 2 | 0 | 0.0 |
rye | 2 | 0 | 0.0 |
others | 25 | 4 | 16.0 |
Livestock products | 190 | 28 | 14.7 |
Raw minced/cut meat | 60 | 6 | 10.0 |
beef minced/cut meat | 8 | 2 | 25.0 |
pork minced/cut meat | 31 | 2 | 6.5 |
chicken minced/cut meat | 15 | 1 | 6.7 |
beef and pork minced meat | 6 | 1 | 16.7 |
Beef offal | 68 | 17 | 25.0 |
stomach | 27 | 7 | 25.9 |
rumen (1st compartment) | 5 | 1 | 20.0 |
reticulum (2nd compartment) | 8 | 2 | 25.0 |
omasum (3rd compartment) | 3 | 0 | 0.0 |
abomasum (4th compartment) | 11 | 4 | 36.4 |
small intestine | 10 | 3 | 30.0 |
large intestine | 14 | 6 | 42.9 |
rectum | 3 | 0 | 0.0 |
heart | 8 | 0 | 0.0 |
liver | 6 | 1 | 16.7 |
Pork offal | 26 | 4 | 15.4 |
stomach | 4 | 2 | 50.0 |
small intestine | 2 | 0 | 0.0 |
large intestine | 5 | 1 | 20.0 |
heart | 7 | 1 | 14.3 |
liver | 4 | 0 | 0.0 |
uterus | 3 | 0 | 0.0 |
lung | 1 | 0 | 0.0 |
Chicken offal | 6 | 1 | 16.7 |
gizzard | 3 | 0 | 0.0 |
liver | 2 | 1 | 50.0 |
heart | 1 | 0 | 0.0 |
Raw milk | 30 | 0 | 0.0 |
Marine products | 65 | 1 | 1.5 |
Raw fish | 8 | 0 | 0.0 |
horse mackerel | 2 | 0 | 0.0 |
flatfish | 2 | 0 | 0.0 |
others | 4 | 0 | 0.0 |
Sashimi | 8 | 0 | 0.0 |
tuna | 2 | 0 | 0.0 |
others | 6 | 0 | 0.0 |
Marinated fish for cooking | 49 | 1 | 2.0 |
sardine | 7 | 1 | 14.3 |
salmon | 7 | 0 | 0.0 |
marlin | 5 | 0 | 0.0 |
mackerel | 5 | 0 | 0.0 |
saury | 2 | 0 | 0.0 |
yellowtail | 2 | 0 | 0.0 |
others | 21 | 0 | 0.0 |
Others | 33 | 0 | 0.0 |
Infant foods | 33 | 0 | 0.0 |
Total | 516 | 64 | 12.4 |
By the quantitative method, most of the samples (501/516) were below the detection limit (Table 4). The CFUs ranged from 101 to 104 CFU/g, were detected only in 18 samples of agricultural products, 2 samples of livestock products, and 1 sample of marine products.
Category | Quantitative analysis | |||||
---|---|---|---|---|---|---|
Sample type | Number of samples | Level of contamination (CFU/g) | ||||
<101 | 101 | 102 | 103 | 104 | ||
Agricultural products | 228 | 211 | 7 | 7 | 2 | 1 |
Whole vegetables | 50 | 41 | 4 | 4 | 1 | |
Cut and packed vegetables for salad | 100 | 98 | 2 | |||
Cut and packed vegetables for cooking | 30 | 26 | 1 | 2 | 1 | |
Flours | 48 | 46 | 1 | 1 | ||
Livestock products | 190 | 188 | 2 | |||
Raw minced meat | 60 | 59 | 1 | |||
Beef offal | 68 | 68 | ||||
Pork offal | 26 | 25 | 1 | |||
Chicken offal | 6 | 6 | ||||
Raw milk | 30 | 30 | ||||
Marine products | 65 | 64 | 1 | |||
Raw fish | 8 | 8 | ||||
Sashimi | 8 | 8 | ||||
Marinated fish for cooking | 49 | 48 | 1 | |||
Others | 33 | 33 | ||||
Infant foods | 33 | 33 | ||||
Total | 516 | 496 | 10 | 7 | 2 | 1 |
The relationship between VBC and Cronobacter spp. contamination was shown in Table 5.1. The levels of VBC seemed to correlate with the percentages of Cronobacter spp. positive. For example, as to the vegetables, including whole vegetables, cut and packed vegetables for salad, and cut and packed vegetables for cooking, the samples with higher (≧106 CFU/g) VBC showed significantly (p < 0.01) more Cronobacter spp. contamination compared with those with lower (<106 CFU/g) VBC (Table 5.2). As to the flours, the samples with higher (≧103 CFU/g) VBC showed significantly (p < 0.01) more Cronobacter spp. contamination compared with those with lower (<103 CFU/g) VBC, and as to the meat and offal, including raw minced meat, beef, pork and chicken offal, the samples with higher (≧105 CFU/g) VBC showed significantly (p < 0.05) more Cronobacter spp. contamination compared with those with lower (<105 CFU/g) VBC (Table 5.2).
Viable bacterial counts (CFU/g) | No. of samples | No. of Cronobacter positive | % of Cronobacter positive | p value by Fisher's exact test |
---|---|---|---|---|
Vegetables | ||||
≥106 | 81 | 20 | 24.7% | 0.0002 |
<106 | 99 | 5 | 5.1% | |
Flours | ||||
≥103 | 18 | 9 | 50.0% | 0.0002 |
<103 | 30 | 1 | 3.3% | |
Meat and offal | ||||
≥105 | 93 | 22 | 23.7% | 0.0197 |
<105 | 67 | 6 | 9.0% |
The relationship between coliform counts and Cronobacter spp. contamination was shown in Table 6.1. The levels of coliform counts also seemed to correlate with the percentages of Cronobacter spp. positive. For example, as to the vegetables, the samples with higher (≧105 CFU/g) coliform counts showed significantly (p < 0.01) more Cronobacter spp. contamination compared with those with lower (<105 CFU/g) coliform counts (Table 6.2). As to the flours, the samples with higher (≧103 CFU/g) coliform counts showed significantly (p < 0.01) more Cronobacter spp. contamination compared with those with lower (<103 CFU/g) coliform counts, and as to the meat and offal, the samples with higher (≧104 CFU/g) coliform counts showed not significant but nearly significant (p = 0.0510) more Cronobacter spp. contamination compared with those with lower (<105 CFU/g) coliform counts (Table 6.2).
Coliform counts (CFU/g) | No. of samples | No. of Cronobacter positive | % of Cronobacter positive | p value by Fisher's exact test |
---|---|---|---|---|
Vegetables | ||||
≥105 | 64 | 19 | 29.7% | 0.00001 |
<105 | 116 | 6 | 5.2% | |
Flours | ||||
≥103 | 7 | 5 | 71.4% | 0.0025 |
<103 | 41 | 5 | 12.2% | |
Meat and offal | ||||
≥104 | 54 | 14 | 25.9% | 0.0510 |
<104 | 106 | 14 | 13.2% |
The prevalence of Cronobacter spp. from the soil of farms, composts and bovine feces was shown in Table 7. Cronobacter spp. contamination was detected in soils, composts and bovine feces, although the levels of contamination were below the detection limit. Especially, bovine feces showed a prevalence (37.5%) of Cronobacter spp. higher as compared to that of the others.
Sample type | No. of samples | No. of Cronobacter positive | % of Cronobacter positive |
---|---|---|---|
Soil | 24 | 4 | 16.7 |
Composts | 10 | 1 | 10.0 |
Bovine feces | 8 | 3 | 37.5 |
Total | 42 | 8 | 19.0 |
Despite many studies of the prevalence of Cronobacter spp. in many foods and environments, its natural habitats remain unknown. In this study, the prevalence of Cronobacter spp. in varieties of foods and farm-associated environments was investigated for determination of the contamination level of foods in Japan, and investigation of natural habitats of Cronobacter spp.
Previous studies showed that sprouts were often contaminated with this bacterial species. Berthold-Pluta et al. suggested that sprouts are one of natural habitat of Cronobacter spp. (Berthold-Pluta et al., 2017). Baumgartner et al.(2009) detected Cronobacter in 14 out of 23 samples comprising sprouts, fresh herbs and salad. Among them, 7 positive samples were sprouts (Baumgartner et al., 2009). Kim et al. (2009) reported the prevalence of Cronobacter spp. in mixed sprout and radish sprout to be 13.3% and 0%, respectively (Kim et al., 2009). Hochel et al. showed that 1 out of 9 wheat sprouts was positive to Cronobacter spp. (Hochel et al., 2012). In our study, sprouts also showed a very high prevalence rate (43.8%) of Cronobacter spp. Cut and packed vegetables for cooking, which mainly contained bean sprouts, were also highly contaminated by this bacterial species.
The prevalence of Cronobacter spp. in buckwheat flour was higher than that in other types of flour. However, there was only one report concerning the detection of Cronobacter spp. in buckwheat flour (Li et al., 2014). Further studies are needed to evaluate the risk of contamination of Cronobacter spp. in buckwheat flour. From these, the prevalence rates of Cronobacter spp. in retail foods in Japan are at similar levels as that found in the other countries.
Previous studies have reported the detection of Cronobacter spp. in ground beef, for example, Kandhai et al. (2010) detected Cronobacter spp. in 2 out of 141 samples in the Netherlands (Kandhai et al., 2010) , and Mohammed et al. detected the organism in 8 out of 50 samples in Egypt (Mohammed et al., 2015). Cronobacter spp. were also reported to be detected in feed, feces and carcasses of cattle (Molloy, 2009; McAuley, 2014), and soil samples from farm, flower beds and sandboxes (Ueda, 2017). This study revealed that Cronobacter spp. was significantly isolated from the intestines of cattle and pigs, feces of cattle, soil of farms and composts. These findings suggest that one of the natural habitats of Cronobacter spp. might be the gastrointestinal tract of livestock. Actually, 3 out of 8 feces samples from cow were positive with Cronobacter spp. in our preliminaly research. Further study is needed to investigate the relationship between feces of livestock, soil, and other environmental samples in the agricultural fields.
Furthermore, higher VBC and coliform counts correlated with the percentages of Cronobacter spp. contamination. It is considered that viable bacteria and coliforms might be an index for Cronobacter spp. contamination in some types of food.
Acknowledgements This work was supported by Ministry of Health, Labour and Welfare under Health Labour Sciences Research Grant (No. H23-syokuhin-ippan-010). This work was also partially supported by the Japan Society for the Promotion of Science (JSPS) under Grant-in-Aid for Scientific Research (C) (KAKENHI) Grant No. 15K07733.