Effects of caffeine on locomotor activity of horses: determination of the no-effect threshold.

Caffeine is the legal stimulant consumed most extensively by the human world population and may be found eventually in the urine and/or blood of race horses. The fact that caffeine is in foods led us to determine the highest no-effect dose (HNED) of caffeine on the spontaneous locomotor activity of horses and then to quantify this substance in urine until it disappeared. We built two behavioural stalls equipped with juxtaposed photoelectric sensors that emit infrared beams that divide the stall into nine sectors in a 'tic-tac-toe' fashion. Each time a beam was interrupted by a leg of the horse, a pulse was generated; the pulses were counted at 5-min intervals and stored by a microcomputer. Environmental effects were minimized by installing exhaust fans producing white noise that obscured outside sounds. One-way observation windows prevented the animals from seeing outside. The sensors were turned on 45 min before drug administration (saline control or caffeine). The animals were observed for up to 8 h after i.v. administration of 2.0, 2.5, 3.0 or 5.0 mg caffeine kg(-1). The HNED of caffeine for stimulation of the spontaneous locomotor activity of horses was 2.0 mg kg(-1). The quantification of caffeine in urine and plasma samples was done by gradient HPLC with UV detection. The no-effect threshold should not be greater than 2.0 microg caffeine ml(-1) plasma or 5.0 microg caffeine ml(-1) urine.