Analysis of bacterial foodborne disease outbreaks in China between 1994 and 2005


  • Editor: Willem van Leeuwen

Correspondence: Jun-Wen Li, Institute of Environment and Health, No.1, Dali Road, Tianjin 300050, China. Tel.: +86 22 84655345; fax: +86 22 23328809; e-mail:


To gain an understanding of the outbreaks of bacterial foodborne diseases and the subsequent health impact, we reviewed 2447 papers from journals published in China that reported 1082 bacterial foodborne disease cases occurring between 1994 and 2005. Among the 1082 outbreaks of bacterial foodborne disease for which the etiology was determined, Vibrio parahaemolyticus caused the most outbreaks, followed by Salmonella, and Clostridium botulinum led to the most deaths. Most of the outbreaks occurred between May and October, except for Clostridium botulinum, which mainly occurred in January and February. In littoral provinces, Vibrio parahaemolyticus caused the most events, whereas in inland provinces, the largest percentage of events was caused by Salmonella. This review provides a background and analysis of Chinese foodborne disease caused by bacteria. We hope that this review can be compared to reviews from other regions of the world, in an attempt to prevent future outbreaks from occurring.


Foodborne disease, or food poisoning, is defined as acute or subacute noninfectious diseases caused by biological or chemical agents that enter the body through the ingestion of food (Anon, 1994). Food poisoning can also be caused by bacteria, fungi, chemicals, animals and plants. Between 1994 and 2005, there were 12 687 events of foodborne diseases reported by Chinese authorities, resulting in 289 380 persons becoming ill and 2297 dying. Among the events for which the etiology was determined, microbial pathogens were responsible for 4515 events, resulting in 146 852 persons becoming ill and 349 dying. Bacterial foodborne diseases accounted for more than 28.25% of all foodborne disease events and more than 42.75% of cases annually (Chinese Health Annual edition committee, 1995–2005; Zhang, 2002). However, no one has synthesized and analyzed the data from all of these events.

It is thus timely and appropriate to review the Chinese literature on bacterial foodborne diseases to gain a better insight into the epidemiology and socioeconomic impact in China. We undertook a systematic review and analysis of bacterial foodborne diseases by collecting information from 2447 published articles in the public journals in China. This paper will not only provide a background on bacterial foodborne diseases in China, but may benefit other countries struggling against their own indigenous foodborne diseases to improve prevention measures.

Materials and methods

We first searched for reviews about Chinese foodborne diseases published in English. A literature search of MEDLINE was conducted (keywords ‘China’ and ‘foodborne disease’ or ‘food poisoning’) to identify published reports of foodborne diseases in China.

We then searched for reports published in Chinese and found official reports as well as events reported in public journals (all were published in Chinese). The data from official reports between 1995 and 2005 comes from the Chinese Health Annual. Data from the year 1994 comes from the Zhang (2002). The majority of public journal data sources come from the China National Knowledge Infrastructure China Academic Journals Full-text database (, the Vip Information database ( and the Wanfang database ( The collection period was from January 1995 to July 2006. The correlative information was isolated from the articles and entered into a Microsoft excel spreadsheet and reviewed independently by two reviewers. Analyses included etiology, the number of events, number of cases, number of deaths, the concourse, date, food vehicle, region and reason. Data were summarized in Microsoft excel and statistical analysis was performed using spss10.0 programs.


Literature search

As of July 2006, only two reviews of foodborne diseases in China existed in MEDLINE, with both being published in Chinese (Liu et al., 2004, 2006).

In the Chinese public journals database, we entered the key word ‘food poisoning’ and ‘foodborne diseases’, and limited the search window to January 1994 to July 2006. Of the 4854 articles about foodborne diseases, we collected 2308 articles that investigated and analyzed single outbreaks of foodborne diseases. An additional 93 reports were collected from the Vip Information and Wanfang Data databases and 46 articles were collected from the correlative journals in library. As a result, we collected 2447 articles about foodborne disease outbreaks. According to the date and region, 37 events were deleted because of repetition, leaving us with 2410 events. By etiology analysis, 1082 events were considered to bacterial, whereas the source of the remainder was attributed to chemicals, plants, animals, epiphyte, a combination of sources or unknown reasons.

General features

This review identifies 1082 foodborne disease outbreaks in China by bacterial contamination during the 12-year period between 1994 and 2005. This resulted in 57 612 foodborne illnesses, and 51 foodborne deaths reported in public journals (Table 1). Etiologic agents were investigated and indicated that Vibrio parahaemolyticus caused the greatest number of events (19.50%), followed by Salmonella (16.73%), Bacillus cereus (13.40%), Proteus (11.46%) and Staphylococcus aureus (7.76%). Of the 57 612 cases, Salmonella was the most identified agent, accounting for 22.16% of illnesses. Vibrio parahaemolyticus accounted for 18.73% of illnesses, followed by Proteus (11.56%), a mixture of bacteria (11.2%) and B. cereus (9.97%). Clostridium botulinum caused the most deaths (62.75%) and the highest mortality rate (32 deaths in 254 cases).

Table 1.   Proportion of 1082 bacterial foodborne disease events
Etiologic agentEventsCasesDeaths
Vibrio parahaemolyticus21119.5010 79018.7300
Salmonella18116.7312 76922.1647.84
Bacillus cereus14513.4057449.97611.76
Mixture bacteria898.23645411.2000
Staphylococcus aureus847.7630555.3000
Escherichia coli726.6537406.4911.96
Other Vibrio353.2312822.2300
Clostridium botulinum302.772540.443262.75
Citrobactcr aerogenes60.551810.3100
Plesiomonas shigelloides40.372270.3900
Other Staphylococcus30.28830.1400
Pseudomonas cocovenenans30.28140.0223.92
Enterobacter cloacae20.18620.1100
Clostridium perfringens10.092170.3800
Alcaligenes faecalis10.0950.0100
Pseudomonas aeruginosa10.09160.0300
Unknown bacteria312.8719523.3911.96
Total108210057 61210051100

Subtypes and details about the six major causes of foodborne diseases

Salmonella can be broken down into six subgroups, serogroups A–F. Among 181 Salmonella foodborne disease events, serogroup D accounted for the majority (45.30%), followed by serogroup B (22.65%), serogroup C (7.73%), serogroup E (5.52%) and serogroup F (1.10%). There were no data about serogroup A and an unknown serogroup Salmonella accounted for 17.68%. Of Salmonella events for which the species was determined, Salmonella enteritidis caused the most events (28.73%) followed by Salmonella typhimurium (14.36%). All other species caused very small percentages of the events (Table 2).

Table 2.   Percentage of 181 Salmonella foodborne disease events
Serogroup B4122.65373429.241
 S. typhimurium2614.36163912.841
 S. derby42.212912.280
 S. taint-paul42.21142611.170
 S. agona21.10260.200
 S. paratyphi B10.55190.150
 Unknown serogroup B42.213332.610
Serogroup C147.737856.150
 S. newport42.211831.430
 S. paratyphi C21.101180.920
 S. bovismorbificans21.10520.410
 S. cholerae-suis21.101441.130
 S. braenderup10.551170.920
 S. bovismorbificans10.55300.230
 S. rissen10.551240.970
 Unknown serogroup C10.55170.130
Serogroup D8245.30624448.903
 S. enteritidis5228.73433633.960
 S. dublin105.527535.903
 S. blegdam84.421691.320
 S. typhi63.313472.720
 S. onarimon10.5560.050
 Unknown serogroup D52.766334.960
Serogroup E105.524563.570
 S. weltevreden31.661741.360
 S. uganda21.101080.850
 S. anatum21.101481.160
 S. senftenberg10.55100.080
 S. butantan10.55100.080
 S. london10.5560.050
Serogroup F21.10750.590
 Unknown Salmonella3217.68147511.550
Total181100.0012 769100.004

Proteus mirabilis accounted for 49.19% of Proteus foodborne diseases followed by an unknown Proteus (29.03%) and Proteus vulgaris (15.32%). Other types of Proteus seemed irrelevant in these events (Table 3).

Table 3.   The state of 124 Proteus foodborne disease events
P. mirabilis6149.19299244.93
P. vulgaris1915.3292113.83
P. morganii54.034226.34
P. inconstans21.61220.33
Unknown Proteus3729.84230234.57

Of the events of Escherichia coli foodborne disease, enteropathic E. coli (EPEC) was the most common pathogen (70.83%); other types of E. coli only occupied a small portion. There was no data about enterohemorrhagic E. coli (EHEC)-related infections (Table 4).

Table 4.   The state of Escherichia coli foodborne disease events
Escherichia coliEventsCases
  1. EPEC, enteropathic E. coli; EIEC, enteroinvasive E. coli; ETEC, enterotoxigenic E. coli; ESIEC, entero-Shiga-like invasive E. coli.

Unknown E.coli11.39370.99

Data characterizing Clostridium botulinum are shown in Table 5. These 30 outbreaks resulted in the hospitalization of 254 people and the deaths of 32 people. The majority of the cases (26.67%) were caused by toxin type B, followed by toxin type A (23.33%) and toxin type E (16.67%). Thirty percent were caused by an unknown toxin type.

Table 5.   The state of Clostridium botulinum foodborne disease events
Clostridium botulinumEventsCasesDeaths
Type A723.334316.93721.88
Type B826.678633.86721.88
Type E516.672710.631031.25
Mixture of type A and type B13.3331.1800.00
Unknown type930.009537.40825.00

Among Vibrio events not caused by Vibrio parahaemolyticus, 35 events were caused by other Vibrio. Vibro alginolyticus caused 45.71% of the events (649 cases) and Vibro mimicus caused 14.29% of the events (39 cases).

Shigella flexneri caused 41.67% of the events (580 cases), Shigella sonnei caused 25.00% of the events (614 cases), Shigella dysenteriae caused 12.50% of the events (148 cases), Shigella boydii caused 8.33% of the events (21 cases), and a mixture of Shigella flexneri and Shigella sonnei caused 4.17% of the events (92 cases; Table 6).

Table 6.   The state of Shigella foodborne disease events
S. flexneri1041.6758038.23
S. sonnei625.0061440.47
S. dysenteriae312.501489.76
S. boydii28.33211.38
Mixture of S. flexneri and S. sonnei14.17926.06
Unknown Shigella28.33624.09

Effects of regional distribution

Distribution of pathogens in inland provinces was different from that in littoral provinces. In littoral provinces, Vibrio parahaemolyticus caused 23.04% of events, probably because Vibrio parahaemolyticus exists mainly in aquatic environments. Salmonella (14.36%), Proteus (12.35%), a mixture agent (10.35%) and B. cereus (10.02%) also were prevalent in littoral provinces. In contrast, inland provinces experienced more Salmonella (20.09%), B. cereus (18.32%), Vibrio parahaemolyticus (14.13%), Proteus (10.6%) and Staphylococcus aureus (8.39%) The breakdown of foodborne diseases as a factor of geographical location is shown in Table 7.

Table 7.   Regional distribution of 1082 bacterial foodborne disease events
Etiologic agentLittoral provinceInland provinceUnknown
Vibrio parahaemolyticus13823.04600120.826414.13378414.02930100556.08
Mixture bacteria6210.35436815.15255.5219327.1626.671548.59
Bacillus cereus6010.0221637.508318.32349812.9626.67834.63
Staphylococcus aureus447.3516195.62388.3913675.0626.67693.85
Escherichia coli345.6811854.11367.9525359.3926.67201.12
Other Vibrio274.5110023.4861.322660.9926.67140.78
Clostridium botulinum50.83270.09235.082090.7726.67181
Other bacteria416.8425908.99275.9616526.1226.67241.34
Total59910028 82410045310026 996100301001792100

Effects of time of year

Bacterial foodborne diseases often thrive in summer and autumn because of the rise in temperature. In warmer weather, the rate of bacterial multiplication accelerates, increasing the possibility bacterial foodborne disease outbreaks. As shown in Fig. 1, as the temperature became warmer between March and May, the number of events increased from 29 to a startling 123. The incidents continued to increase, peaking during the hottest months (August and September). With the coming of winter, the number of incidents dropped sharply and remained at low levels for the following 6 months. This seasonal variation rule was applicable to most of the bacteria. However, Clostridium botulinum outbreaks were an exception, with the highest incidence rate being in Spring.

Figure 1.

 Distribution of 1078 bacterial foodborne diseases during the calendar year.

Effects of concourse

Of the 1082 bacterial foodborne disease events, 31.05% (336 events) occurred in food services units and 26.25% (284 events) took place in the home. Bacterial diseases tended to be more prevalent in specific concourses (Table 8). The bacteria that were most common in food services units were Vibrio parahaemolyticus (55.45%; 117 events), Proteus (33.06%; 41 events) and E. coli (47.22%; 34 events). The bacteria that were most common in the home were Salmonella (50.83%; 92 events), Staphylococcus aureus (36.90%; 31 events) and C. botulinum (86.67%; 26 events). Finally, bacteria that were most common in the schools were B. cereus (35.86%; 52 events) and Shigella (58.33%; 26 events).

Table 8.   Concourse distribution of 1082 bacterial foodborne disease events
Etiologic agentFood
service units
Vibrio parahaemolyticus1172818205104153211
Bacillus cereus142452242033023145
Mixture bacteria3412221131102389
Staphylococcus aureus103121591222184
Escherichia coli341511312020472
Other Vibrio1942130320135
Unknown bacteria9512310000131
Clostridium botulinum0260003000130
Citrobacter aerogenes40010000016
Plesiomonas shigelloides21001000004
Other Staphylococcus00110000013
Pseudomonas cocovenenans03000000003
Enterobacter cloacae20000000002
Clostridium perfringens10000000001
Pseudomonas aeruginosa00100000001
Alcaligenes faecalis00000010001


Bacterial pathogens are widely responsible for many foodborne diseases. Between 1994 and 2003 in China, 37% (2938/7929) of foodborne disease events and 52% (110 415/213 951) of cases were caused by microbial pathogens (Chinese Health Annual edition committee, 1995–2005). In a report made by a national foodborne disease surveillance system of China, it was shown that bacterial etiology accounted for 38.5% (2221/5770) of all events and 50.9% (82 888/162 955) of cases as a part of foodborne diseases between 1992 and 2001 (Liu et al., 2004). In these reports, the bacterial pathogen was often barely discussed and seldom analyzed. Thus, in the present review, the bacterial pathogen has its own epidemiologic focus and analysis.

The surveillance of foodborne disease is difficult because it is nearly impossible to obtain absolute data on foodborne diseases in China. We obtained the data for this review from published journals to study the epidemic regulations of bacterial foodborne disease in China. The limitations concerning the quantity and quality of data presented in this report should be recognized.

First, illnesses attributed to foodborne diseases are likely an underestimation because not all events are recognized or reported and most sporadic cases of foodborne disease illness are not detected. Second, the data collected from journals are affected by many factors, i.e. the quality of the surveillance system, economic levels and quantity of journals. Therefore, the data are not sufficient to show the dynamic regulation, allowing us to examine only a static state. We expect the study has a role in providing background information and prevention and control efforts of bacterial foodborne disease. Third, prompt epidemiologic investigation helps prevent additional cases and can identify new risk factors for intoxication.

The outbreaks described in this report illustrate several important points about bacterial foodborne diseases. During 1994–2005, 1082 bacterial foodborne disease events were reported, resulting in 57 612 illnesses and 51 deaths. Of the 1082 outbreaks, V. parahaemolyticus and Salmonella were the most frequently identified agents, involving 36.23% of events and 40.89% of cases. It is reported that Salmonella is the most important agent associated with foodborne diseases, followed by V. parahaemolyticus (Meng, 1990). However, there has been a trend for the latter to overtake Salmonella as the most prevalent bacteria. The highest mortality rate was attributed to C. botulinum, which caused 62.75% of the 51 deaths. We have not yet found the reports of bacterial foodborne diseases caused by E. coli O157 : H7 in China. It is possible that the E. coli O157 : H7 pathogen had not been determined during the outbreaks. Our findings are comparable with those of Liu et al. (2004), who analyzed the reports of foodborne disease outbreaks in 13 provinces covered by the National Foodborne Disease Surveillance System during 1992–2001. Vibrio parahaemolyticus (31.1%) accounted for the largest number of outbreaks and cases.

The results in this review are very different from results gathered from developed countries, namely the USA, UK and Australia. Mead et al. (1999) estimated that foodborne diseases cause c. 76 million illnesses, 325 000 hospitalizations, and 5000 deaths in the USA each year. Three pathogens, Salmonella, Listeria, and Toxoplasma, are responsible for 1500 deaths each year (Mead et al., 1999). Olsen et al. (2000) reported that between 1993 and 1997, a total of 2751 outbreaks of foodborne diseases were reported in the USA. Salmonella enteritidis accounted for the largest number of outbreaks, cases, and deaths (Olsen et al., 2000). Lynch et al. (2006) reported that among bacterial pathogens, S. enteritidis accounted for the largest number of outbreaks and outbreak-related cases. Listeria monocytogenes accounted for the majority of deaths of any pathogen (Lynch et al., 2006). Wheeler et al. (1999) reported that 781 cases were identified in the community cohort, giving an incidence of 19.4/100 person years in the UK. The ratio of cases in the community to cases reaching national surveillance was lower for bacterial pathogens (Salmonella 3.2 : 1 and Campylobacter 7.6 : 1) than for viruses (rotavirus 35 : 1 and small round structured viruses 1562 : 1) (Wheeler et al., 1999). In 2004, 118 outbreaks of foodborne diseases were reported in Australia, affecting 2076 persons, of whom 5.6% were hospitalized and two people died. Foods prepared in restaurants and catering settings caused the most outbreaks and the most common sporadic diseases were Campylobacteriosis (15 640 cases) and Salmonellosis (7842 cases) (OzFoodNet Working Group, 2005).

Based on our present analysis, the most practical and effective approach for the control of bacterial foodborne disease in China would be the implementation of a well-defined education campaign, combined with a systematic food quality monitor supported by governmental assistance in enforcing the HACCP principles, followed by efforts to monitor the success of the outcome. If such a strategy could be implemented and rigorously controlled by public health authorities, it should be possible to effectively control, or at least substantially reduce, the prevalence and intensity of bacterial foodborne diseases in many parts of China.

Conflicts of interest

We declare that we have no conflicts of interest.