FOODBORNE PATHOGENS AND DISEASE
Volume 5, Number 2, 2008
ª Mary Ann Liebert, Inc.
DOI: 10.1089=fpd.2007.0071
Seroprevalence of Trichinella, Toxoplasma,
and Salmonella in Antimicrobial-Free and
Conventional Swine Production Systems
Wondwossen A. Gebreyes,1 Peter B. Bahnson,2 Julie A. Funk,3 James McKean,4
and Prapas Patchanee1
Abstract
There has been a growing niche-market, outdoor, antimicrobial-free (ABF) swine production system in the
last few years prompted by consumers’ demand for a more ‘‘natural’’ pork product. The impact of such
production systems on reemergence of current and historically signi?cant swine-associated pathogens has
not been determined. The objectives of the current study were to determine and compare Salmonella,
Toxoplasma, and Trichinella seropositivity in two swine production systems: outdoor ABF and intensive
indoor production systems. These three foodborne pathogens represent those with the highest importance
for pork consumption. A total of 675 serum samples from three participating states, Wisconsin, North
Carolina, and Ohio, were investigated. We found signi?cantly higher seroprevalence of Salmonella and
Toxoplasma from ABF herds (54% and 7%, respectively) than conventional (39% and 1%, respectively) ( p ¼
0.001). Two pigs, both from ABF herds, were found to be seropositive for Trichinella. The results from this
preliminary study suggest that all three pathogens were more commonly present in pigs that were reared
in an ABF, outdoor, niche-market type of environment than the conventional, indoor-reared herds though
there were some geographical variation in Salmonella. This warrants a robust epidemiologic study to
determine the role of various risk factors in the two production systems that may lead to persistence of
bacterial (Salmonella) pathogens and reemergence of parasites (such as Trichinella) of historical signi?cance.
Introduction
biosecurity measures, the prevalence of these
pathogens also declined.
Swine are implicated as one of the impor- Pigsandundercookedporkareconsideredan
tant contributors of foodborne infections in
important source of human infection with Tox-
humans. Apart from the common bacterial
oplasma (Dubey et al., 1992; Davies et al., 1998;
pathogens, such as Salmonella, zoonotic parasitic
Gamble et al., 1999). Prevention and eradication
infestations including Toxoplasma gondii and
of these diseases in humans is a critical public
Trichinella spiralis are known to have historical
health goal of the pork industry, veterinarians,
signi?cance in swine production. As the pork
and public health authorities. Mead et al. (1999)
industry shifted to a more intensive indoor pro-
estimated the number of foodborne-related ill-
duction system complemented with stringent
nesses caused by Toxoplasma gondii was more
1College of Veterinary Medicine, Department of Veterinary Preventive Medicine, The Ohio State University, Columbus, Ohio.
2School of Veterinary Medicine, Department of Medical Sciences, University of Wisconsin—Madison, Madison, Wisconsin.
3College of Veterinary Medicine, Department of Clinical Sciences and 4College of Veterinary Medicine, Department of Veterinary
Diagnostic and Production Animal Medicine, Iowa State University, Ames, Iowa.
199

---

200
GEBREYES ET AL.
than 100,000 cases in the United States alone.
with good management practices are suf?ciently
In addition, Toxoplasma ranks in the top-three
adequate to eradicate T. spirallis in swine herds.
leading causes of mortality by foodborne path-
Of greater concern, there has been a paucity of
ogens, predominantly being associated with
information on the signi?cance of the rising
fetuses and immunocompromised persons.
outdoor niche-market production systems on
Toxoplasma gondii is commonly reported in
the rate of T. gondii infection in pigs. This study
pigs. A report on the national seroprevalence
was aimed to investigate and compare the ser-
of T. gondii in commercial slaughter pigs be-
oprevalence of Salmonella, Toxoplasma gondii,
tween 1983 and 1984 with 11,842 serum sam-
and Trichinella spiralis in pigs reared in two
ples demonstrated about 23% of ?nishing pigs
types of production systems; niche-market, out-
throughout the United States were serological
door, and antimicrobial-free (ABF) and inten-
positive by the agglutination test (Dubey et al.,
sive indoor (conventional) reared herds in three
1991). Sows had a twofold higher seropreva-
pig-producing states: North Carolina, Ohio, and
lence (46%) as compared to ?nisher swine (Du-
Wisconsin.
bey et al., 1991). Based on the seroprevalence
survey of pigs including 17 major pig produc-
Methods
ing states in the United States in 1995 and 2000,
the proportion of seropositive breeding females
Sample collection sites
and ?nisher pigs was dramatically reduced
Two types of swine production systems were
as compared to the previous decade (Patton
included in this study: intensive indoor (con-
et al., 2002).
ventional) and outdoor ABF systems. Under the
During 1997–2001, an outbreak with 33 tri-
conventional system of pig production, antimi-
chinellosis cases (and no deaths) was reported
crobials were added to the feed for growth
(Roy et al., 2003). While there are many reports
promotion and were also used for therapeutic
in the United States indicated decline in hu-
purposes. In the extensive ABF pig production
man trichinellosis related of undercooked pork
system, pigs were reared in open ?elds in a
consumptions, occurrence of cases related with
barricaded area and had free access to the en-
wild game meat consumption remains a con-
vironment including soil and water. In addition,
cern (Gamble et al., 1999; Roy et al., 2003). Tri-
no antimicrobials were used for any purpose
chinella infection from consumption of pork
other than injections (as needed) for the sows at
products in the Western World is minimal in
the farrowing units. Sick pigs that were given
this era of modern swine production in inten-
antimicrobials for treatment, in the ABF units,
sive production units. On the other hand, in the
were immediately removed from the herd and
last few years, the numbers of niche-market,
kept in a different hoop barn or pen enclosure
outdoor pig production units are on the rise.
and marketed as conventional and excluded
Pigs in these facilities are raised with full or
from this study. A total of 675 serum samples
partial outdoor access on dirt with open access
from the three participating states—Wisconsin
to soil, vegetation, and wild fauna. The role of
(n ¼ 351), North Carolina (n ¼ 201), and Ohio
these emerging production systems on the
(n ¼ 123)—were submitted to the Iowa State
potential reemergence of Trichinella and other
University Diagnostic Laboratory for analysis.
important parasitic diseases has not been in-
After removing missing and inconsistent (spe-
vestigated.
cimens that were not con?rmed as from con-
In contrast to the large survey study of Tox-
ventional or ABF) data, 616 total data points
oplasma gondii antibody in pigs, the infection rate
were further analyzed (Table 1).
for Trichinella spp. in pigs in the United States is
not well determined. Davies and his colleagues
Serology
reported only 1 out of 2183 (0.057%) pigs reared
in indoor production systems in the state of
Serological analysis of the parasitic (Toxo-
North Carolina was seropositive for T. spirallis
plasma gondii and Trichinella spiralis) and bacterial
(Davies et al., 1998). The authors concluded that
(Salmonella) pathogens using ELISA was con-
modern swine production systems in concert
ducted at the Iowa State University Diagnostic

---

ANTIMICROBIAL-FREE AND CONVENTIONAL SWINE
201
Table 1. Number of Samples in the Three
test where appropriate. A multivariable logistic
Participating States by Production System
regression was also used to determine the poten-
Production system
tial association of each of the two parasitic path-
ogens on Salmonella seroprevalence (outcome).
State
ABFa
Conventional
Total
Wisconsin
245
80
325
Results
North Carolina
74
127
201
Ohio
5
85
90
Overall the proportion of seropositive ani-
Total
324
292
616
mals was 47% (291 of 616), 4.1% (25 of the 616
aABF, antimicrobial-free system.
specimens), and 0.3% (2 of the 616 samples) for
Salmonella, Toxoplasma, and Trichinella, respec-
tively (Table 2). The highest seroprevalence of
Center. The standard protocols were followed to
Salmonella was detected from Wisconsin (59%
detect antibody response against parasitic and
with 95%CI: 54–64) followed by North Carolina
Salmonella infection. Brie?y, diluted test serum
(34% with 95%CI: 27–40) and Ohio (34% with
samples including negative and positive controls
95%CI: 24–44) (Table 2). When comparing the
were incubated in the test well to form antigen–
two production systems, we found a signi?-
antibody complexes. After washing, the anti-
cantly higher seroprevalence of Salmonella from
swine IgG peroxidase conjugate was then added
ABF herds (54%) than conventional (39%)
and incubated to allow the anti-swine IgG per-
(w2
oxidase to bind to the antigen–antibody com-
df¼1, p ¼ 0.001) (Table 2).
Seroprevalence of Toxoplasma was signi?-
plex. After a ?nal wash, a chromogen substrate
cantly higher among samples from ABF (6.8%,
was added to all wells and the plate was read
n ¼ 22) than conventional (1.1%, n ¼ 3) (w2
on a plate reader at 450 nm. The cut-off OD
df¼1,
p ¼ 0.001) showing a similar trend as the Sal-
used to determine positivity was value of $ 0.2
monella seroprevalence (Table 2). Two Trichinella
and 0.3 for Toxoplasma gondii and Trichinella
seropositive specimens (one each from Wis-
spiralis, respectively, while the positive cut-off of
consin and North Carolina) were found. Both of
OD% ¼ 25 for detection antibody was evidence
these Trichinella positive specimens originated
of Salmonella infection. Seroprevalence of each of
from ABF herds. A retesting of both of these
the three pathogens was determined and com-
positive specimens after 9 months of the initial
parison between niche-market ABF and conven-
testing resulted in deterioration of the specimen
tionally reared productions were done.
resulting in lower OD reading (?vefold lower in
both cases); one from 1.476 to 0.342 (and 0.374
Statistical analysis
with third repeat), remaining still within the
SPSSÃ’ software package version 15.0 (SPSS
positive range and the other from 0.387 to 0.074
Inc., Chicago, IL) was used to carry out by
(and 0.096 with third repeat), which is consid-
Pearson’s chi-square analysis or Fisher’s exact
ered negative.
Table 2.
Seroprevalence of Salmonella, Toxoplasma, and Trichinella in Conventional
and ABF Swine Production Systems
Total tested (n ¼ 616)
Production system
No. positive=no. samples
Pathogen (state)
(% prevalence)
95% CI
ABFa
Conventional
p value
Salmonella
291=616 (47.2)
43–51
176=324 (54.3%)
115=292 (39.4%)
0.001
Wisconsin
192=325 (59.1)
54–64
150=245 (61.2%)
42=80 (52.5%)
0.1
Ohio
31=90 (34.4)
24–44
5=5 (100%)
26=85 (30.6%)
0.001b
North Carolina
68=201 (33.8)
27–40
21=74 (28.4%)
47=127 (37%)
0.2
Toxoplasma
25=616 (4.1)
2–6
22=324 (6.8%)
3=292 (1.1%)
0.001
Trichinella
2=616 (0.3)
0–1
2=324 (0.34%)
0=292 (0%)
0.2b
aABF, antimicrobial-free system.
bFisher’s exact test.

---

202
GEBREYES ET AL.
While at least one sample from North Caro-
oprevalence of Toxoplasma among pigs raised in
lina and Wisconsin was positive for all of the
an ABF environment than those reared in con-
three pathogens, none of the samples from Ohio
ventional settings. Even though we have not
were positive for Toxoplasma and Trichinella. We
done a detailed risk factor analysis in the current
found no signi?cant association between Sal-
study, the extensive (outdoor) nature of the ABF
monella seropositivity and either of the two
production system is conducive for exposure
parasitic pathogens (Toxoplasma and Trichinella).
of the pigs to various known risk factors for
There was some evidence of an association be-
Toxoplasma infestations such as cats (the de?ni-
tween seropositivity for Toxoplasma and Trichi-
tive host) and other species that can be harbor-
nella (Fisher’s exact test, p ¼ 0.08). This may be
ing cysts in their musculature as compared to
due to the fact that both were more commonly
the indoor conventional production system.
detected from ABF herds but none from con-
Although the prevalence for Trichinella was very
ventional ones. Multivariable logistic regression
low, it was still higher than our expectations (as
results also supported the univariate ?ndings
well as U.S. Department of Agriculture prev-
that there is no association between the occur-
alence estimates [Dubey et al., 1992]). The fact
rence of the two parasitic pathogens and shed-
that both positive animals were identi?ed
ding of Salmonella.
among ABF, outdoor-reared swine highlights
the potential foodborne risks of these produc-
tion systems. These pigs are reared in an open-
Discussion
air environment with potential exposure to wild
Three globally important zoonotic foodborne
and other domestic fauna, which are the po-
pathogens including Salmonella, Trichinella, and
tential sources of this parasite. In addition, the
Toxoplasma were studied. Serological analysis of
outdoor environment with open access to soil,
these pathogens was conducted on samples col-
vegetation, and moisture allows for viable en-
lected from two distinct pig production systems:
vironments for Toxoplasma oocysts. The similar
ABF, outdoor-reared, niche-market type and
trend of higher seroprevalence of Toxoplasma
conventional (indoor-reared). The results from
and Trichinella in outdoor-reared swine was re-
this preliminary study suggest that Salmonella
ported in the Netherlands (van der Giessen et al.,
and Toxoplasma were more commonly present in
2007). The frequency distribution of Toxoplasma
pigs that were reared in ABF, outdoor, niche-
seroprevalence reported in this study ranged
market type of environment than the conven-
from 0.38% in intensive production to 5.62%
tional, indoor-reared herds. This was found to
in outdoor-reared system which was relatively
be consistent in all three states—North Carolina,
in concordance with our ?ndings (Table 2).
Ohio, and Wisconsin—that were included in the
According to the National Animal Health
current study.
Monitoring System’s seroprevalence survey re-
The high seroprevalence of Salmonella in all
ported by Gamble et al. (1999), only 0.16% of a
three states has not been a surprise since the
total 3048 lactating sows’ sera and 0.013% of a
?ndings indicate prior history of Salmonella ex-
total of 7987 gestating=?nishing pigs’ sera were
posure as compared to current infections when
found to be positive for Trichinella spp. anti-
prevalence based on culture and isolation
bodies. However, the Trichinella spp. seroposi-
method is used. Consistent with our previous
tive prevalence may vary by the different
reports, the prevalence of Salmonella was higher
locations, production types, and management
among ABF herds than conventional ones
styles; for instance, 15 out of 4078 (0.4%) of se-
(Gebreyes et al., 2006). A recent report on the
rum samples from 156 farms in the northeastern
Salmonella infection demonstrated only 13.2%
United States were found positive to Trichinella
seroprevalence in 90 Canadian swine farms
spp. antibodies (Gamble and Bush, 1999).
which was much lower when compared to 41%
Some evidence of an association between
seoprevalence under conventional setting in our
seropositivity for Toxoplasma and Trichinella
study (Table 2) (Rajic et al., 2007).
( p ¼ 0.08) in this study may be due to the fact
Another most interesting ?nding in the cur-
that both were more commonly detected from
rent study was the signi?cantly higher ser-
ABF herds but none (for Trichinella) or rare (for

---

ANTIMICROBIAL-FREE AND CONVENTIONAL SWINE
203
Toxoplasma) from conventional herds. The ?nd-
multidrug-resistant Salmonella in antimicrobial-free and
ing in this preliminary study warrants the need
conventional pig production. J. Food Prot. 2006;69:743–
for a robust epidemiologic study to determine
748.
Mead PS, Slutsker L, Dietz V, McCaig LF, Bresee JS, Sha-
the role of various production-associated risk
piro C, Grif?n PM, and Tauxe RV. Food-related illness
factors in the two production systems on the
and death in the United States. Emerg. Infect. Dis. 1999;
safety and wholesomeness of pork products,
5:607–625.
particularly on the persistence of bacterial (Sal-
Patton S, Faulkner C, Anderson A, Smedley K, and Bush E.
monella) and potential reemergence of parasitic
Toxoplasma gondii infection in sows and market-weight
(Trichinella and Toxoplasma) pathogens.
pigs in the United States and its potential impact on
consumer demand for pork. 00-130. Available at http:==
www.pork.org=PorkScience=Documents=00-130%20-
Acknowledgments
Patton-UofTenn.pdf. Accessed 11 November 2006. Des
Moines, IA; National Pork Board, 2002.
This project was funded by a grant from the
Rajic A, Chow EY, Wu JT, Deckert AE, Reid-Smith R,
National Pork Board (NPB-04-108).
Manninen K, Dewey CE, Fleury M, and McEwen SA.
Salmonella infections in ninety Alberta swine ?nishing
References
farms: serological prevalence, correlation between cul-
ture and serology, and risk factors for infection. Food-
Davies PR, Morrow WEM, Deen J, Gamble HR, and Patton
borne Pathog. Dis. 2007;4:169–177.
S. Seroprevalence of Toxoplasma gondii and Trichinella
Roy SL, Lopez AS, and Schantz PM. Trichinellosis
spiralis in ?nishing swine raised in different production
surveillance—United States, 1997–2001. MMWR Sur-
systems in North Carolina, USA. Prev. Vet. Med. 1998;
veill. Summ. 2003;52:1–8.
36:67–76.
van der Giessen J, Fonville M, Bouwknegt M, Langelaar M,
Dubey JP, Gamble HR, Rodrigues AO, and Thulliez P.
and Vollema A. Seroprevalence of Trichinella spiralis
Prevalence of antibodies to Toxoplasma gondii and Tri-
and Toxoplasma gondii in pigs from different housing
chinella spiralis in 509 pigs from 31 farms in Oahu, Hawaii.
systems in the Netherlands. Vet. Parasitol. 2007;148:371–
Vet. Parasitol. 1992;43:57–63.
374.
Dubey JP, Leighty JC, Beal VC, Anderson WR, Andrews
CD, and Thulliez P. National seroprevalence of Toxo-
plasma gondii in pigs. J. Parasitol. 1991;77:517–521.
Gamble HR, Brady RC, Bulaga LL, Berthoud CL, Smith
Address reprint requests to:
WG, Detweiler LA, Miller LE, and Lautner EA. Pre-
Wondwossen A. Gebreyes, D.V.M., Ph.D.,
valence and risk association for Trichinella infection in
DACVPM
domestic pigs in the northeastern United States. Vet.
Parasitol. 1999;82:59–69.
College of Veterinary Medicine
Gamble HR and Bush E. Seroprevalence of Trichinella in-
Department of Veterinary Preventive Medicine
fection in domestic swine based on the National Animal
The Ohio State University
Health Monitoring System’s 1990 and 1995 swine sur-
1920 Coffey Rd.
veys. Vet. Parasitol. 1999;80:303–310.
Columbus, OH 43210
Gebreyes WA, Thakur S, and Morrow WE. Comparison of
prevalence, antimicrobial resistance, and occurrence of
E-mail: Gebreyes.1@osu.edu

---

---