At standard laboratory ambient temperatures (T-a) of 20-24 degreesC, periph
eral injections of lipopolysaccharide (LPS) reliably produce fever in young
rats. In contrast, old rats may show a blunted fever, no fever, or even hy
pothermia after LPS. In the present study we hypothesized that old rats mig
ht use behavioral thermoregulation to help them develop a fever. Young and
old rats were implanted with temperature transmitters. At least 1 wk postop
eratively they were placed in a thermally graded alleyway (T-a 10-40 degree
sC). On the third and sixth day they were taken out of the gradient, placed
at an T-a of 23 degreesC, injected intraperitoneally with LPS or saline, a
nd left at 23 degreesC for 3 h. At the end of that time, all young rats had
become febrile, whereas the old rats had not. When the rats were replaced
in the thermal gradient, the young animals continued to develop a fever tha
t was similar to fever in young rats left at 23 degreesC. The old animals c
hose significantly warmer positions in the thermal gradient than did the yo
ung animals and only then became febrile. Although there was a tendency for
the young rats to prefer higher T-a after LPS than after saline, these dif
ferences were not significant. However, the differences in the old rats wer
e significant. These results suggest that the LPS had increased the thermal
set point in the old rats, but they could develop febrile responses only a
t the warm T-a they selected.