August 6, 1997
Peter Feng, Ph.D.
Division of Microbiological Studies, OSRS,
CFSAN, FDA, Washington, D.C.
The sprouting of seeds as food for human consumption has been a common practice in many cultures for several centuries. Although little if anything was known about the microbiological flora of sprouts, many of these societies, such as the Chinese, by tradition, seldom consumed raw produce and hence, were able to minimize the incidence of gastroenteritis associated with sprout consumption. In the United States, both alfalfa and bean sprouts have been available in grocery stores and restaurants for quite a number of years. However, with the recent shift in consumer life style towards "healthy living and healthier foods", the consumption of raw sprouts, mostly in salads and sandwiches has also increased in popularity.
Microbiological Flora of Sprouts
Seed sprouting is an easy process that can be done at home or at a commercial scale. All that is required is soaking dried viable seeds to allow for germination and maintain high levels of moisture. Sprouts usually emerge in 2-7 days, depending on seed type. Chemical analyses have shown sprouts to contain proteins, fats, carbohydrates, minerals and vitamins; therefore, it is a nutritious food for the human diet (7). However, it is also a suitable medium for sustaining bacterial growth. Microbiological analyses have shown that alfalfa and mung saan seeds routinely contained high numbers of microbial flora (102-106/g), including coliforms (104/g) and fecal coliforms (102-103/g) (Table 1) and that these organisms appeared to be part of the seed normal flora (12). Furthermore, the sprouting process seems to provide ideal conditions for bacterial growth (2). Hence, the microbial flora of sprouts were oftentimes 2 to 3 logs higher than that observed in seeds (Table 1) Several factors have been identified that contribute to the rapid proliferation of bacteria on sprouts, these include: enzymatic, nutritional and environmental factors (2). Mung beans, alfalfa seeds and soy beans have high level of trypsin inhibitors, which may provide a defense mechanism by which seeds inhibit the trypsin-like enzymes in bacteria. Levels of trypsin inhibitors in seeds decrease during germination, perhaps due to surface leaching, thereby enabling the microbial flora to proliferate (2). Similarly, the nutritional contents of seeds (carbohydrates and amino acids) are relatively low (7); but, these nutrients can increase 10 fold during germination. This provides available substrates for microbial growth. Finally, the high levels of moisture required and the warmth resulting from heat generated from the sprouting process creates a favorable environment for bacterial growth (2). A study by Splittstoesser et al. (17) found that the microbial population of commercial sprouts was very similar to that germinated in the laboratory under aseptic conditions. This study concluded that the bacterial species found on sprouts most likely originated from the seeds rather than due to the sanitary conditions of commercial sprout production (17).
Most microbial analysis of seeds or sprouts have not found the presence of pathogenic bacterial species such as: Salmonella (12,16,18), Bacillus cereus (12,17,18), Staphylococcus aureus (17,18) and Listeria (16). However, a recent study from Thailand found that 8.7% of the 30 bean sprout samples examined, contained Salmonella of various serotypes (5). Another study (2) showed that low levels of Salmonella species seeded into mung bean and alfalfa seeds can increase by as much as 4 to 5 logs in the germinated sprouts (2); hence, the contamination of a few pathogens in seeds can potentially be amplified by the sprouting process to become a microbiological hazard.
Sprout Associated Foodborne Outbreaks
Historically, sprouts have been implicated in a number of foodborne outbreaks worldwide. One of the earlier documented cases occurred in 1973 and was caused by seed sprouting kits imported from Switzerland, that were contaminated with B. cereus. The kits had a mixture of soy seeds from Uganda, cress seeds from Holland and mustard seeds from Denmark (15). Bacteriological analysis showed that the seeds had an Aerobic Plate Count (APC) ranging from 107 to 108/g and contained large numbers of aerobic spore-forming bacteria. Although no precise counts were performed for B. cereus, it was estimated that the organism was present at levels of 105 to 107/g (15). Analysis of unopened sprouting kits showed that the soy seeds were contaminated with a pure culture of B. cereus, while the organism was only a minor part of the flora of the other 2 seeds. The source of B. cereus contaminating the seeds was not determined.
In 1988, beansprouts were implicated in an outbreak of Salmonella saint-paul in the Oxford region of the UK (9). A total of 143 cases were reported and S. saint-paul of the epidemic type was isolated from sprouts obtained from various retail stores as well as from mung bean seeds and other elements on the premises of the producer. In addition, S. virchow was also isolated from sprouts and associated with 7 cases of infection. Beansprouts are germinated mung bean seeds which were imported mainly from Australia and Thailand. During outbreak investigations, analyses of beansprouts from various retail stores showed them to contain a number of Salmonella serotypes (9). This is consistent with a recent report that 8.7% of the beansprout samples analyzed in Thailand carried Salmonella of various serotypes (5).
Salmonella was again implicated in a large outbreak in 1994, causing 282 cases in Sweden and 210 cases in Finland. Both outbreaks were associated with the consumption of alfalfa sprouts made from seeds imported from Australia (14). S. bovismorbificans a serotype very rare in Sweden, was isolated from the implicated sprouts and also from additional sprouts made from the same seed lots. The study reported that the pathogen was not isolated from the seeds, but only from the germinated sprouts. Including this outbreak, a total of 3 outbreaks implicating sprouts contaminated with Salmonella have been reported in Finland in 4 years (14). Alfalfa sprouts was again implicated in an outbreak of S. newport in Denmark in 1995 (1). Although few details about the outbreak were given, the authors reported that the incriminated seed lots contained 0.1 to 0.6 CFU of S. newport/25 g of seeds and no other Salmonella serotypes were detected. Based on the average consumption of 25 to 100 g of sprouts per person, the infectious dosage for S. newport was estimated to be in the range of 5 to 460 cfu (1).
Also in 1995, an international outbreak, again caused by the consumption of alfalfa sprouts contaminated with S. enterica serotype Newport, was reported in Oregon and British Columbia (11). Subtyping analysis showed that the isolates implicated in the Oregon and B.C. cases were indistinguishable from each other and from the isolates implicated in the Danish outbreaks (11); hence, the seeds probably came from the same seed shipper.
Another large international outbreak also occurred in 1995 in Finland and the USA (Arizona, Michigan and 15 other states) caused by alfalfa sprouts contaminated with Salmonella stanley (9). Epidemiological investigations showed that the S. stanley strain isolated from the US and Finnish patients had a unique DNA profile and an antibiotic resistance pattern that was different from other S. stanley strains (9). Evidence also suggested contamination was not due to insanitary production but, rather, the seeds used to produce the sprouts were contaminated with S. stanley (9). Microbiological analysis of seeds and sprouts made from the implicated seed lots failed to show S. stanley. Since contamination of seeds was probably not uniform, and only a single bag of the implicated lot was available for analysis, it was not surprising that the pathogen was not isolated (9). The sprouts that caused the outbreaks in Finland and in the US were made from seeds obtained from the same shipper in the Netherlands, suggesting that the seeds were contaminated prior to shipping. A total of 242 cases were identified in the USA and Finland. Based on the under-reporting rates defined for Salmonella outbreaks, the actual number of cases was probably between 5000-24,000 (9).
More recently, in the world's largest outbreak of E. coli O157:H7 infection which occurred in Japan in 1996, white radish sprouts were implicated epidemiologically for some of the 10,000 cases reported. The pathogen was never found on the seeds. In Japan, in March of 1997, white radish sprouts were again implicated in an outbreak of E. coli O157:H7 that infected 123 persons. This time, O157:H7 was isolated from leftover sprouts in the refrigerator of an infected family; however, exhaustive testing has failed to isolate the pathogen from the same seed lots that were used to produce the implicated sprouts. Finally, in July of 1997, alfalfa sprouts were implicated in foodborne outbreaks of E. coli O157:H7 in Virginia and Michigan. The extent and details of these outbreaks are not yet available.
Effectiveness of Seed Treatment
Since the presence of even a few pathogenic cells on seeds can be amplified by the sprouting process and become a health hazard, it is essential to eliminate these pathogens from seed surfaces. Many have explored the use treatment procedures to reduce the microbial flora of seeds. Previously, treatment of alfalfa seeds with 0.5% sodium hypochlorite solution for 45 min did not prevent the S. bovismorbificans outbreak in Finland in 1994 (14). Jaquette et al. (4) examined the effects of chlorine and heat treatment using alfalfa seeds inoculated with 102-3 cfu of S. stanley/g. Significant reduction in the numbers of viable cells was achieved with up to 1040 ug/ml of chlorine; however, it was not effective in eliminating the pathogen from the seeds. Similarly, treating seeds in hot water (> 54°C) for 10 min was effective in reducing microbial populations; but, it also caused a substantial reduction in viability of seeds (4). These authors concluded that soaking seeds in 2000 to 4000 ug/ml of chlorine appeared to be an effective treatment of seeds; however, it was no guarantee that the sprouts will be free of salmonellae (4). Another study (13), examining disinfection procedures for rice seeds, showed that ethanol, hydrogen peroxide or 1000 ppm of sodium hypochlorite were effective in reducing APC by 2 to 3 logs. Ethanol however, inhibited seed germination. Much better reduction (up to 5 logs) was achieved by soaking the seeds for 5 min in sodium hypochlorite at 60C; but, again they did not totally eliminate the microbial flora. The effectiveness of these chemicals in removing Salmonella was also examined using alfalfa seeds (3). Treatments of seeds in 6% hydrogen peroxide, 80% ethanol or calcium or sodium hypochlorite solutions containing 1800 to 2000 ug/ml of active chlorine, were effective in reducing Salmonella populations by 1000 fold. However, even after 10 min treatment of seeds in these solutions, viable Salmonella were still detected. The author speculated that perhaps Salmonella cells trapped in cracks and crevices on the seed were inaccessible to the chemicals (3). Finally, one study (14) recommended pasteurization of seeds before sprouting or heating or cooking sprouts prior to consumption. No data are available on the effectiveness of pasteurizing seeds. Also, pasteurization will not resolve the problems of microbial contamination during the sprouting process.
REFERENCES
Table 1.
Microbiological load of mung bean and alfalfa seed and sprouts.Seeds (per g) | Sprouts (per g) | REF | |||||
APC | Colif. | Fecal | APC | Colif. | Fecal | ||
bean | NA1 | NA | NA | >106 | 104-5 | 104-5 | 18 |
both | NA | NA | NA | 108 | 106 | 106* | 12 |
alfalfa2 | 103-6 | 104 | 103 | 108 | >104 | 103 | 16 |
both | 106 | NA | NA | 108 | 107 | NA | 17 |
alfalfa2 | NA | NA | NA | 108 | NA | NA | 8 |
bean | NA | NA | NA | 108 | 107 | NA | 6 |
both | 105 | NA | NA | 108 | NA | NA | 2 |
1NA - information not available 2 Retail * mostly Klebsiella |
Table 2. Foodborne outbreaks associated with the
consumption of sprouts.
Date | Sprout | Pathogen | Site | No. | Ref. |
1973 | mixed1 | B. cereus | Texas | 4 | 15 |
1988 | bean | Salmonella saint-paul | UK | 143 | 10 |
1994 | alfalfa | S. bovismordificans | Sweden Finland |
282 210 |
14 |
1995 | alfalfa | S. enterica Newport | Oregon B.C. |
58 70 |
11 |
1995 | alfalfa | S. newport | Denmark | 154 | 1 |
1995 | alfalfa | S. stanley | Arizona Michigan Finland other US |
21 51 114 56 |
9 |
1996 | radish | E. coli O157:H7 | Japan | NA2 | |
1997 | radish | E. coli O157:H7 | Japan | 1233 | |
1997 | alfalfa | E. coli O157:H7 | Virginia Michigan |
NA NA |
|
1Mixed seeds of soy, mustard and cress. 2 NA - information not available 3 Food Chem News, April 14, 1997. |
Home
| Foodborne IllnessHypertext updated j3b/mow 1999-AUG-18