Twelve horses were infected with various doses of Trichinella spiralis
and then tested for infection using direct (artificial digestion) and
indirect (enzyme immunoassay) methods. Horses became infected in a do
se-dependent manner. Larvae accumulated preferentially in the tongue,
followed by the masseter, neck, supraspinatus, trapezius, and diaphrag
m. At lower infection levels, the tongue harbored several times more p
arasites than were found in other tissues. The sensitivity of artifici
al digestion methods for detecting infections was directly related to
sample size. One-gram samples were not reliable for detecting infectio
n levels of <3 larvae per g (LPG). In sample sizes of 5 or 10 g the te
chnique allowed infections as low as 1 LPG to be detected. The enzyme
immunoassay (EIA) detected all infected horses; the times following in
fection at which horses became seropositive varied in a dose-dependent
manner, but 11 of 12 horses were positive in the EIA. by 4 weeks post
inoculation. One horse, with a larval density in the tongue of 0.39 LP
G, did not become seropositive until 7 weeks postinoculation. The resu
lts suggest that artificial digestion of horse carcasses for trichinae
should concentrate on tissue samples from the tongue or masseter musc
les. Sample sizes should be a minimum of 5 g using pooled-sample diges
tion methods to assure detection of all infections which might pose a
human health risk. The EIA is a potential substitute for artificial di
gestion methods and could also be useful for antemortem testing and fo
r epidemiological studies.