Title

An outbreak of food poisoning in Tamil Nadu associated with Yersinia enterocolitica

Creator

Mathew Abraham

Madhukar Pai

Gagandeep Kang

G V Asokan

S R Magesh

Sara Bhattacharji

B S Ramkrishna

Date

September 23, 1997

extracted text

\.
Indian J Med Res 106, November 1997, pp 465-468

1/6 S

An outbreak of food poisoning in Tamil Nadu associated with Yersinia
enterocolitica '
'
• !
.

■

'

■

■

I

Mathew Abraham, Madhukar Pai, Gagandeep Kang*, G.V. Asokan, S.R. Magesh,
Sara Bhattacharji & B.S. Ramakrishna*
Departments of Community Health <& *ICMR Centre for Advanced Research in Enteric Diseases,
Department of Gastrointestinal Sciences, Christian Medical College, Vellore
A-,

)ted September 23, 1997
An outbreak of food poisoning in a Tamil Nadu village, affecting 25 of 48 individuals who participated in
a feast, was investigated. The risk ofad£$eloping illness was associated with consumption of buttermilk

(relative risk 3.8). None of the food items consumed during the feast was available for analysis. Toxin­
producing Y. enterocolitica (serotype 3, biotype 4) was grown from 1 of 11 stool samples from affected
individuals, as well as from a water sample from the source used to dilute the buttermilk. High titres of
antibody of Yersinia were detected in 2 of 12 patients but in neither of the two groups of controls. Toxin
production was noted in buttermilk incubated for 6 h with Y. enterocolitica. This is the first report from
India of a food poisoning outbreak associated with this organism.
Key words Food poisoning - India - outbreak - Yersinia enterbcolilica

Food hygiene and safety receive scant attention in
India despite the fact that several million people, in­
cluding children, are fed every day under State-spon­
sored supplementary feeding programmes such as the
Tamil Nadu Integrated Nutritional Project1. Food poisc^g outbreaks are not uncommon, but are seldom
inWstigated. The present report desc-ibes the investi­
gation of an outbreak of food poisoning which, though
delayed, provided evidence for contamination of water
and buttermilk with a toxin-producing bacterium as the
probable cause..

varying times subsequently 25 (52 %) developed symp­
toms requiring medical attention, and seven had to be
hospitalized.

The outbreak came to the attention of the investiga­
tion team the next day, by which time all remnants of
the meal and water had been discarded. The investiga­
tion was begun on 2nd October 1996, by interviewing
47 of the 48 attendees at the feast (response rate 98 %).
Details of the symptoms and food consumed during the
feast were obtained. Any person who gave a history of
fever and loose stools with or without vomiting or
abdominal cramps was considered a case. Stool sam­
ples were collected the same day from 11 cases. No
blood samples were collected at this time. Five litre
samples of water were collected for analysis from the
borewell which supplied water for the feast (including
buttermilk preparation), as well as from another well

Material & Methods

Outbreak investigation : The outbreak began on 29th
September 1996 in a village in North Arcot District of
Tamil Nadu, India. Villagers (48) partook of a feast
comprising a standard vegetarian south Indian meal. At
465

/
K

I

INDIAN J MED RES, NOVEMBER 1997

•166

located in a field. Two samples of milk were obtained
for analysis, one from the vendor who supplied milk for
the feast, and the other from another source.

Blood samples were collected for serology nearly a
month after the epidemic (October 28th), because by
this time Yersinia enterocolitica was grown from the
stool of one of the affected individuals, as also from the
water. Samples were obtained from 12 feast partici­
pants who fell ill, from 8 who did not develop illness
(controls) and from I 5 residents of the village who did
not participate in the feast (general population). Paired
blood samples were again obtained on 23rd December.
Two futher samples of bore well water were obtainedin November and December.
Laboratory methods : Stool was transported to the
laboratory' on ice, and samples were plated on MacConkey agar, xylose lysine desoxycholate agar, blood
agar and thiosulphate citrate bile salt sucrose agar and
inoculated into selenite F broth which were incubated
overnight at 37°C. Samples were also plated onto
Butzler’s medium and incubated at 42°C under microaerophilic conditions. Selenite F broth was subcul­
tured onto MacConkey agar after 16 h, All plates were
examined at 24 and 48 h after incubation and appropri­
ate biochemical tests were done. Cold enrichment was
done by inoculating 2 g of the original stool sample
with one tube each of phosphate buffered saline (pH
7.2) and selenite F broth and incubated for 6 wk at
4°C2. Weekly subcultures onto MacConkey agar were
made. All media were from Difco Labs and chemicals
from Sigma Chemical Co.
.
Standard techniques2 were used to analyze water
and milk samples, and all the above culture media were
utilized.

'

‘

■'

Biotyping of Yersinia was done by fermentation of
sucrose, raffinose, rhamnose, melibiose and cello;
biose3. Motility of suspected Yersinia colonies was
examined at 25 and 37°C by hanging drop and
Craigie’s tube method2. Isolates biochemically identi­
fied as Y. enterocolitica were tested for fermentation of
salicin and esculin and for production of pyrazinami■ dase3. They were also plated on’Congo Red magnesium

oxalate agar4 and tested for production of enterotoxin
in the suckling mouse assay5. Appropriate positive and
negative controls were used. Serotyping and biotyping
of the two Y. enterocolitica isolates was done by kind
courtesy of Dr Bernard Rowe (PHLS, Colindale, UK).

Y. enterocolitica antigen was prepared6 from a
known pathogenic strain of Y. enterocolitica (Statens
Seruminstitut, Copenhagen, Denmark, serotype 03)
and used for agglutination assay. The assay was done
by standard method and the reaction read after over­
night incubation at 37°C.
The two toxin-producing Y. enterocolitica isolates
were grown in Luria broth (prepared from materials
from Sigma and Difco) overnight and 104 bacteria were
inoculated into 10 ml samples of buttermilk which
were held at 4°C and at room temperature for 6 h. The
buttermilk was centrifuged and the supernatant tested
- fortoxin production in'the suckling mouse assay5.

Escherichia coli isolated from the stool of all indi­
viduals were tested for enteropathogenic serotypes, for
production of heat-labile and heat-stable toxins, Hep-2
cell adherence and for production of shiga-Iike toxins I
and II by enzyme immunoassay7.
Statistical analysis : Epi-Info software (Version 5,
CDC and WHO) was used for data analysis. Relative
risks and 95 per cent confidence intervals were used to
measure the strength of association between develop-^
ment of symptoms and consumption of a particular
food item.

Results
All cases occurred between 29 September and 1
October. The time from consumption of the food to
development of illness ranged from 7 to 49 h (median
16,h). The overall attack rate was 52 per cent, and was
similar for either sex. The affected individuals ranged
in age from 3 to 80 yr, 11 per cent being children. The '
commonest symptoms were fever (96 %), diarrhoea
(92 %), abdominal cramps (92 %), headache (75 %),
rigors (7L°/o) and vomiting (67 %). Bloody stools were
not reported nor, did any.of the affected, individuals
ha 'e an appendicitisdike illness^.There were no deaths*..

/

ABRAHAM et al: YERSINIA FOOD POISONING OUTBREAK

The mean stool frequency was 9 per day, and the mean
duration of diarrhoea was 66 h.
The Table shows the magnitude of association
(measured as relative risk) between consumption of
various food items and illness. The only food item
significantly associated with risk of developing illness
was buttermilk. All those who had consumed butter­
milk fell ill.

Microbiological analysis revealed Y. enterocolitica
in 1 of 11 stool samples of the patients. Y. enterocoli­
tica yjtxs also isolated from the borewell water sample
taken initially. Both isolates exhibited all biochemical
attributes of pathogenic Y. enterocolitica (including
Congo Red binding) and both were positive for toxin
production in the suckling mouse assay. Both isolates
^^.re of serotype 3 and biotype 4. No other enteric
^Rthogens (including enteropathogenic, enterotoxi­
genic, enteroadherent or enterohaemorrhagic E. coll)
were isolated from stool, water or milk samples. Cold
enrichment of the samples yielded Y. enterocolitica
only in the two instances where the organism had al­
ready been detected on direct plating, but not in any of
the other samples. AH water samples had coliform
counts greater than 180/mi. Y. enterocolitica was not
isolated from the second and third samples from the
borewell, or from the other well sampled.
Serology of the first paired samples revealed high
anti-}', enterocolitica antibody titres (1 in 640 and 1 in
5120 dilution) in 2 of 12 patients and borderline titres
(1 in 160) in 1. No significant antibody titres were
detected in any of the 8 controls or 15 general populaa samples. The second paired samples were all nega^Wve for antibody to this organism.
Toxin was detected in inoculated buttermilk kept
for 6 h, both at 4°C and at room temperature.

Discussion

Y. enterocolitica infection may cause intestinal, extraintestinal and reactive (postinfectious) syndromes8,
infection with this organism has been reported sporadi­
cally from India9'11, primarily in patients with diar­
rhoea. The organism usually causes illness by invasion
of the intestine. However, it also produces a heat-stable
enterotoxin resembling the heat-stable enterotoxins

is l

467

Table. Magnitude of association between consumption of food
items and illness
Food item

Relative risk
(RR)
Rice
2.14
Rasam
1.03
Sambar
2.26
Vadai
1.02
Payasam
0.93
Curry (plantain)
1.41
Buttermilk
:3.88*
Drinking water
1.03
* Statistically significant (P < 0.001)

95% confidence
intervals
0.38- 11.95
0.54 - 1 97
0.66-7.73
0.44-2.41

0.51 - 1.70
0.70-2.81
2.13-7.04
0.55 - 1.91

(ST) of E. coli and non-01 Vibrio cholerae. The organ­
ism is a known cause of food poisoning,12 and several
•outbreaks have been reported. The contaminated
source responsible for such epidemics has usually been
some form of milk,13'15 bean sprouts8 or soya tofu16. In
the latter two, spring water or well water were also
implicated. In a study from Ludhiana, rhe organism
was detected in milk and fruit juice during testing of a
number of food samples for pathogens17. The organism
has also been isolated from many environmental water
sources18 where the non-pathogenic variety is more
commonly found8. However, there is a report of an
outbreak due to ingestion of unchlorinated well
water19. There is no report, to date, from India of an
outbreak of food poisoning due to this organism.
Despite several limitations in the investigation of
the outbreak, the circumstantial evidence outlined be­
low leads us to believe that the outbreak was due to
contamination of buttermilk by Y. enterocolitica. Y.
enterocolitica of similar characteristics was isolated
from the stool of a symptomatic individual, as well as
from the water used to dilute the buttermilk. No other
enteric pathogen (including pathogenic E. colt) was
isolated from the stool of symptomatic individuals. The
demonstrable fall in antibody titre in paired samples
from 3 of 12 affected individuals suggests that they had
been exposed to the organism recently. Delay in obtain­
ing the first blood sample may explain why more of the
affected individuals did not have an antibody response
to the organism. It is likely that the Y. enterocolitica

■

ki

468

-

'

-

INDIAN J MED RES, NOVEMBER 1997

present in the water used to dilute the buttermilk prolif­
erated in vitro producing toxin, and that ingestion of
this buttermilk caused the symptoms of food poisoning.
The short incubation period (median incubation period
16 h) and the associated fever and vomiting is consis­
tent with symptomatology clue to ingestion of a pre­
formed toxin. However, it is possible that symptoms in
those affected individuals with longer incubation peri­
ods can be attributed to invasion of bowel mucosa by
the organism. The reason for the well water contamina­
tion was not investigated. However, it is likely that
there was a transient contamination, possibly from pigs
or cattle which are both known to be reservoirs of the
organism8.
Acknowledgment
Authors are grateful to the Chinna-Anaicut village community
for enthusiastic support; to Sh. Maruthamuthu and the CHAD
health education team for organizing the community meetings; and
to Prof. Jayaprakash Muliyil for advice and comments. The ICMR
centre is supported by the Indian Council of Medical Research,
New Delhi and the Wellcome Trust, UK..

References

1. Shckar M, Latham MC. Growth monitoring can and docs
work! An example from the Tamil Nadu Integrated Nutrition
Project in Rural South India. Indian J Pediatr 1992; 59 : 5-15.
2. Collee JG, Duguid JP, Fraser AG, Marmion BP. McCartney’s
Practical microbiology. 13th edn. Edinburgh : Churchill Liv­
ingstone : 1989.

3. Farmer JJ. Kelly MT. Enterobacteriaccae. In : Balows A,
Hausler WJ Jr. Isenberg HD, Shadomy HJ editors Manual of
clinical microbiology, 5th edition. Washington DC : American
Society for Microbiology, 1991; 360-83.

4. Riley G, Toma S. Detection of pathogenic Yersinia entero­
colitica by using Congo Red-magnesium oxalate agar me­
dium. J Clin Microbiol 1989; 27 : 213-4.
5. Dean AG, Ching YC, Williams RG, Harden LB. Test for
Escherichia coli enterotoxin using infant mice : application in
a study of diarrhoea in children in Honolulu. J Infect Dis 1972;
725:407-11.
6. Fowler JM, Brubaker RR. Immunology of Yersiniae, Rose
NR, Friedman A, Fahey JL. editors. In : Manual of clinical
and laboratory immunology. 3rd edition. Washington DC:
American Society for Microbiology, 1989; 371-4.

7. Law D, Hamour AA, Acheson DW, Panigrahi H. Ganguli LA,

Denning DW. Diagnosis of infections with Shiga-like toxin­
producing Escherichia coli by use of enzyme-linked immu­
nosorbent assay for Shiga-like toxins on cultured stool
samples. J Med Microbiol 1994; 47:241-5.
8. Cover TL. Yersinia enterocolitica and Yersinia pseudotuber­
culosis. In : Blaser MJ, Smith PD, Ravdin JI. Greenberg HB.
Gucrrant RL, editors, Infections of the gastrointestinal tract.
New York : Raven Press, 1995 : 811-23.
9. Singh G, Arora NK; Bhan MK, Ghai OP, Dhar S, Shriniwas
DC. Role of Yersinia enterocolitica in childhood diarrhoea.
Indian J Pediatr 1983; 57 : 39-42.
10. Varghese A, Ramachandran VG, Agarwal DS. Isolation and
characterization of Yersinia enterocolitica from paticrtf*cf
gastroenteritis.ym/ian J Med Res 1984, 79 : 35-40.

11. Ram S, Khurana S, Singh R, Sharma S, Vadehra DV. Yersinia
enterocolitica diarrhoea in north India. Indian J Med Res
1987;56:9-13.
12. Bishai WR. Sears .CL. Food poisoning syndromes. Gastroen­
terol Clin North Am 1993; 22 : 579-608.
13. Black RE, Jackson RJ, Tsai T, Mcdvesky M, Shayegani M,
Feeley JC et al. Epidemic Yersinia enterocolitica infection
due to contaminated chocolate milk. N Eng! J Med 1978; 298:
76-9.

■'

14. Shaycgani M, Morse D, DeForge 1, Root T, Parsons LM,
Maupin PS. Microbiology of a major foodborne outbreak of
gastroenteritis caused by Yersinia enterocolitica serogroup 0 :
8. J Clin Microbiol 1983; 7 7 : 35-40.
15. Tacket CO, Narain JP, Sattin R, Lofgren JP, Konigsberg C, Jr
Rendtorff RC et al. A multistate outbreak of infections caused
by Yersinia enterocolitica transmitted by pasteurized milk.
JAMA 1984; 251 : 483-6.
fl

16. Tackct CO, Ballard J, Harris N, Allard J, Nolan C, Quan T et
al. An outbreak of Yersinia enterocolitica infections caused by
contaminated tofu (soybean curd). Am J Epidemiol 1985; 121:
705-11.

17. Ram S, Khurana S, Khurana SB, Vadehra D,V, Sharma S,
Chhina.RS. Microbiological quality and incidence of organ­
isms of public health importance in food and water in
Ludhiana. Indian J Med Res 1996; 103 : 253-8.
18. Javier-Castillo F, Larraz V, Asuncion-Lafarga M, Navarro M,
Gomez-Luz R. Phenotypic characterization and distribution of
Yersinia in human and environmental samples. Microbiologia
1994;77:83-92.'
19. Thompson JS, Gravel MJ. Family outbreak of gastroenteritis
due to Yersinia enlerocolitica serotype 0 : 3 from well water.
Can J Microbiol 1983; 32 : 700-1.
.

'. i

.

*.

■

,.

,,

,

Reprint requests : Dr B.S. Ramakrishna. ICMR Centre for Advanced Research in Enteric Diseases, Department of Gastrointestinal Sciences
Christian Medical College Hospital, Vellore632004
-.1.-'H
„■ ,-it. <
•.
,
u'l,'.;.. ■ /, ■

,