This experiment was performed to study the effects on femoral bone of endur
ance training performed during the 3 months before orchidectomy in rats whi
ch were then killed 90 days later. A total of 70 male Wistar rats were used
at 8 weeks old. One day 0 of the experiment, 10 rats were killed by cervic
al dislocation and used as first controls. Among the 60 others, 30 were sel
ected for treadmill running (60% maximal oxygen uptake, 1 h . day(-1), 6 da
ys . week(-1) for 90 days). The 30 other rats remained at rest. On day 90,
10 exercised (IE) and resting (IR) rats were killed and used as intermediar
y controls. Among the 20 other animals of each group, 10 were surgically ca
strated (CXE, CXR) or 10 sham-operated (SHE, SHR) and killed on day 180. On
day 90 femoral failure load (three-point bending test) was greater in IE t
han in IR. Simultaneously, the deoxypyridinolinuria was lower in IE than in
IR. On day 180, femoral bones were thinner in CXR than in CXE. The lowest
values for trabecular bone are in the distal femoral metaphysis were measur
ed in CXE and CXR rats, but the value measured in CXE was no different from
that measured in SHR. Simultaneously total femoral bone density was lower
in CXR than in SHE, while no difference concerning femoral metaphyseal dens
ity was observed between CXE and SHR. These results confirmed that enduranc
e running increased femoral bone growth and modelling and femoral trabecula
r area, and thereby peak bone mass, in 8-month-old male rats. In resting an
imals, castrated after the training period, androgen deficiency decreased f
emoral density, mineral content and trabecular area. This decrease was not
observed in castrated but previously exercised rats. Thus, by increasing pe
ak bone mass, it was considered that endurance training may have a preventi
ve effect against orchidectomy-induced bone loss.