An Outbreak of Salmonellosis Associated with Eating Alfalfa Sprouts

An Outbreak of Salmonellosis Associated with Eating Alfalfa Sprouts, Lexington, Virginia

Elizabeth Barrett, D.M.D., M.S.P.H, Cindy Chaos M.P .H.,October 16, 1996


On February 20, 1996, a physician who provides medical care to two universities in Lexington, Virginia contacted the Lexington/Rockbridge Health Department to report 14 laboratory confirmed cases of salmonellosis among students at the two facilities. The health department reported the situation to Cindy Chaos, M.P.H., district epidemiologist for the Central Shenandoah Health District. The next day, Ms. Chaos travelled to Lexington to interview the ten ill students from University A. Examination of their food histories showed that the students ate almost all their meals at the university cafeteria and further suggested that the food vehicle might have been alfalfa sprouts. Of the ten students, five definitely and one probably ate sprouts at the university cafeteria during the three days prior to becoming ill. Although complete food histories had not yet been obtained from the students from University B, the apparent link among these four students was that during the three days prior to becoming ill, all had eaten at a local restaurant that routinely serves sprouts. No interactions were found between the ill students’ at University A and the ill students at University B. Isolates from stool specimens from the”14 students were sent to the Division of Consolidated Laboratory Services (DCLS) where they were serotyped as Salmonella stanley. To determine the cause of this 1996 outbreak, a case-control study was conducted at University A, focusing on foods served by the university cafeteria.



During June 1995, the Centers for Disease Control and Prevention (CDC) had notified the states of a significant national increase in S. stanley infections.Twenty-four states, including Virginia, had reported more than the usual number of S. stanley isolates during 1995. An outbreak of S. stanley infections had also occurred in Finland in March 1995. Case-control studies conducted in Arizona, Michigan and Finland implicated alfalfa sprout consumption as theonly exposure associated with the infections.The U.S. and Finland outbreak strain had the same antimicrobial sensitivity pattern and pulsed field gel electrophoresis pattern (PFGE) that was distinct from S. stanleyisolates not linked to the international outbreak.1


Traceback investigations led to a Dutch shipper, who supplied alfalfa seed to the United States and Finland.In the U.S., the seed was imported by a Kentucky-based company and then distributed to sprouting farms nationwide; sprouting takes about five days under moist, warm conditions. Alfalfa sprouts were then sold to grocery stores and restaurants. The CDC implicated specific seed lots in the U.S. outbreaks, but S. stanley was not isolated from any seeds or sprouts collected. By about August 1995, it appeared that the implicated seed lots had been sold and used. The1996 outbreak described in this report, therefore, occurred well past the international S. stanleyoutbreak.


University A is an all male military institute with a student body of 1,200. The cafeteria is open from about 7 a.m. to 7 p.m. Breakfast includes a buffet line, a make-your-own waffle station, and a fruit bar. Lunch and dinner include a deli bar, pasta station, grill line, salad bar, dessert bar and ice cream station. In addition, a taco bar and a potato bar are available at lunchtime. The waffles, salad bar and deli bar are available between meals as well. University B is a coeducational liberal arts college (1,950 students) with grounds adjourning University A. Illness appeared to be unrelated to the University B cafeteria since at least two of the four ill

University B students had no meals from the cafeteria during the three days prior to becoming ill.There was no overlap of academic activities or university-served meals between the two schools.



Epidemiologic Investigation

Surveillance for additional cases was conducted at the emergency room and laboratory of Stonewall Jackson Hospital in Lexington. . In addition, letters to physicians were put in their hospital mailboxes alerting them of the outbreak and asking that they test and report anyone presenting with symptoms compatible with salmonellosis. On March 5, Ms. Chaos and Dr. Elizabeth Barrett from the Office of Epidemiology visited University A and examined medical records for other students who had a diarrheal illness during February.


On March 14, a self-administered standardized questionnaire was given to University A students to obtain information about illness characteristics and foods eaten at the university cafeteria. A case was defined as laboratory confirmed S. stanley infection or gastroenteritis consisting of diarrhea (> 3 episodes of diarrhea in a 24 hour period) with two or more of the following symptoms: nausea, vomiting, abdominal cramps, fever, chills or headache, in a University A student during February 1996. Two sets of controls were selected: roommate-   controls and clinic-controls. Roommate-controls were persons who shared a barrack room with a case but did not themselves meet the case definition for illness. One to four roommate-controls were identified for each case (each barrack room is shared by two to five students). Clinic-­controls were defined as students who presented to the university clinic during March who, when questioned by the clinic staff, gave no history of gastrointestinal illness during February. III students were asked to try to remember foods eaten during the week before becoming ill. Controls were asked to try to recall foods eaten during the week of February 11 to February 17, 1996..


A three-day food history was administered to the ill students at University B but further study of these students was not done.


Environmental Investigation

No sprouts from the University A cafeteria were available for testing. On February 20, an environmental health specialist from the Lexington/Rockbridge Department of Health visited the restaurant where the University B students had eaten. Sprouts were obtained for testing about one week later. Since sprouts are very perishable it is very unlikely that they were from the same batch as those consumed by the students.


A traceback of sprouts purchased by University A and the restaurant was conducted by the Virginia Department of Agriculture and Consumer Services (VDACS) and the CDC.


Laboratory Investigation

Isolates from stool specimens from the ten University A and the four University B students were sent to CDC for antimicrobial sensitivity testing and PFGE analysis.


Stool specimens were requested from all students identified by record review as having presented to the University A clinic during February with a complaint of diarrhea. Specimens were also requested from roommates of ill students who reported having had symptoms compatible with Salmonella infection during February but had not sought medical attention.


Seeds and sprouts from the Virginia sprouting farm identified in the traceback were submitted to DCLS for testing.




Epidemiologic Investigation

No additional cases were reported from the local hospital or area physicians. The record review at the University A medical facility found 14 additional students who had presented during February complaining of diarrhea. In addition, five roommates of ill persons reported that they had experienced diarrhea during February.


When the criteria for defining a case were applied, a total of 26 University A students met the case definition for illness, 14 of whom had laboratory confirmed S. stanley infection. All 26 reported having diarrhea that lasted from 1 to 10 days (median=4 days). The other predominant symptoms were abdominal pain or cramps (85%), headache (65%) nausea (60%), chills (54%), and fever (42%), with temperatures ranging from 99.0°F to 103.0°F. Only two (9%) reported vomiting. Nine (35%) were hospitalized at the University A medical facility for one to four days; 13 missed one to four days of school or their usual activities because of their illness. Onset of illness ranged from February 7 to February 23, 1996 (Figure 1). The median duration of illness was four days (range 1-11 days).


Compared to the roommate-controls, cases were 14.9 times as likely to have eaten alfalfa sprouts at the University A cafeteria (14/23 vs. 4/41, 95% Confidence Interval {C.I.} 3.3-70.1). Comparing cases with clinic-controls, 14/23 cases recalled eating alfalfa sprouts compared to 1/23 controls (Odds Ratio {O.R.} = 34.2, 95% C.I. 3.6-821.6).


Only one other food item was statistically significantly associated with illness. Cases were more likely than roommate-controls to have eaten lettuce from the deli bar (O.R. = 6.3, 95% C.I. 1.1-45.7). However, this lettuce consumption was not statistically significantly associated with illness when the analysis was done with clinic-controls. Furthermore, although the lettuce served on the deli bar was the same as that served on the salad bar, only the deli bar lettuce was statistically significantly associated with illness.


Interviews were completed for three of the four ill University B students. During the three days prior to becoming ill, two recalled having sprouts at the same Lexington restaurant. The third had eaten at the restaurant but did not recall having sprouts; the fourth student could not be reached for interview. In addition, there was one University A student with laboratory confirmed salmonellosis who did not recall eating sprouts but had eaten a sandwich from this restaurant. The onsets of illness for these students are also shown in Figure 1.


Environmental Investigation

The traceback of sprouts purchased by University A and the Lexington restaurant showed that both had received sprouts from the two local produce distributors. The sprouts were grown at two farms, one in Virginia and one in Ohio. Although both companies had been using primarily Australian seed since the spring of 1995, CDC records showed that in March 1995 the Ohio company had received one of the seed lots implicated in the international outbreak and the Virginia company had received one of the implicated seed lots in April 1995.


Laboratory Investigation

The 14 isolates sent to CDC from stool specimens from University A and University B students all had the same antimicrobial sensitivity pattern whichwas identical to the unique and unusual pattern identified by CDC as the 1995 outbreak strain (resistance to trimethoprim- , sulfamethoxazole, sulfisoxazole, streptomycin, tetracycline and kanamycin; susceptibility to ampicillin). PFGE testing on four isolates from University A students and two from University B students showed them to beidentical to the 1995 outbreak strain.


All sprouts and seeds collected were negative for Salmonella.




The results of this epidemiologic investigation indicated that an outbreak of salmonellosis due to alfalfa sprout consumption occurred in Lexington during February 1996. The laboratory findings support this conclusion.Specifically, isolates from stool specimens collected from ill students were identical to the strain of S. stanleyidentified by CDC as the cause of an international outbreak during 1995.


The traceback showed that seed from lots implicated in the 1995 outbreaks had been distributed that same year to both growing farms that provided sprouts to University A and the restaurant. Although during early 1996 both companies claimed to be using only seed from Australia, it appears that some leftover seed must have been used by one of the growers, causing this outbreak in Lexington.


Apparently, only a small amount of seed need be contaminated with Salmonella in order for illness to occur. Salmonellais very hardy, survives on the dry seed for extended periods of time, and then multiplies extensively during warm, moist sprouting conditions. Laboratory experiments showed that S.stanley counts increased at between 3.3 and 6.0 logs per gram during sprouting of inoculated seed.1The effectiveness of washing the contaminated sprouts prior to eating is questionable.


In November 1995, a nationwide outbreak of S. newportoccurred; once again alfalfa sprouts from Europe were implicated. In addition, the Food and Drug Administration (FDA) has isolated Salmonella species from seed imported from Australia, indicating that a potential public health risk may exist with all alfalfa seed.2Methods for disinfecting seed prior to sprouting are being studied and disseminated to sprout growers by FDA and CDC.



1.         Mahon BE et al., An international outbreak of Salmonella infections caused by alfalfa

sprouts grown from contaminated seed. Presented at the Epidemic Intelligence Service Conference, Centers for Disease Control and Prevention, Atlanta, April 24, 1996.

2.         Maddon l, Memo: Microbiological safety of alfalfa sprouts, Food and Drug Administration, March 1, 1996.



We very much appreciate the help of everyone who assisted with this investigation: The physician and nurses at University A; Martin Coggins, Lexington/Rockbridge Health Department; Sarah Henderson, Mary Mismus, Marguerite Fussell, June Dvorak, CLA, Dave. Peery, Daksha Patel, M.T., Division of Consolidated Laboratory Services; Foodborne and Diarrheal Diseases Laboratory Section, Centers for Disease Control and Prevention.


Elizabeth Barrett, D.M.D., M.S.P.H                          Cindy Chaos M.P .H.

Epidemiologist                                                           District Epidemiologist

Office of Epidemiology                                                 Shenandoah Health District

Virginia Department of Health                                      Virginia Department of Health

October 16, 1996