A longitudinal study was conducted over a I-year period in six selected vil
lages in Petauke and Katete districts in the Eastern Province of Zambia. St
arting in November 1997, 50 animals were sampled at random at each village
every 2 months. The parasitological prevalence of trypanosomosis was determ
ined by the haematocrit centrifugation buffy-coat technique, supplemented w
ith thick and thin Giemsa-stained blood films. Serum samples also were coll
ected for anti-trypanosomal antibody determination by indirect, enzyme-link
ed immunosorbent assay.
Parasitological prevalence was highly variable between villages and between
visits (range: 0-28.6%; median: 3.1%). Seroprevalence was also variable be
tween villages (range: 0-80.8%; median: 50%), but was less variable between
visits. Average annual parasitological prevalence and average annual serop
revalence for each village were highly correlated [R-2 (adjusted for d.f.)
= 0.89, p < 0.01]. Seroprevalence measured on any single visit to a study v
illage was better than parasitological prevalence as a predictor of average
annual parasitological prevalence. (C) 2001 Elsevier Science B,V. All righ
ts reserved.