122
R. Murray et al. / Physiology & Behavior 119 (2013) 115–129
3.2.1. HYPER rats
repeated measures analysis of covariance as described in the Methods
section, significant difference between rat lines was found
Fig. 3 (top) shows preference for water of both stressed and
non-stressed female NS and HYPER rats. When all days on which rats
were offered sucrose as well as water were analyzed by a three factor re-
peated measures analysis of covariance as described for sucrose intake in
the previous section, a significant interaction of rat line × stress/no
stress was found [F(1, 18) = 5.439, p = .032], thereby indicating that
preference for water was affected differently by stress in female HYPER
and NS rats. Fig. 3 shows that this occurred because stress increased pref-
erence for water (i.e., decreased preference for sucrose) in HYPER fe-
males but not in NS females. This significant interaction was supported
by comparison of stress and non-stress conditions within each rat line,
which revealed that stressed HYPER females showed significantly higher
preference for water (and therefore lower preference for sucrose) than
non-stressed HYPER females [F(1, 9) = 16.078, p = .003] whereas
stressed and non-stressed NS females did not show a significant differ-
ence in preference for water [F(1, 8) = .859, p = .381].
When only the first day of each exposure period was analyzed simi-
larly, the interaction of rat line × stress/no stress approached signifi-
cance [F(1, 18) = 3.344, p = .084] thereby indicating that the
preference for water tended to be affected differently by stress in female
HYPER and NS rats. However, comparison of stress and non-stress within
rat lines revealed that while stress tended to increase preference for
water in HYPER females, the difference was not significant [F(1, 9) =
2.895, p = .123], nor was the tendency for stress the decrease prefer-
ence for water in NS females [F(1, 8) = .952, p = .358].
Fig. 3 (bottom) shows preference for water of both stressed and
non-stressed male NS and HYPER rats. When all days in which rats
were offered sucrose as well as water were analyzed by a three factor
repeated measures analysis of covariance as described previously, a sig-
nificant difference between rat lines was found [F(1, 19) = 8.497, p =
.009], thereby indicating HYPER males showed significantly higher pref-
erence for water (and therefore lower preference for sucrose) com-
pared with NS males. The interaction of rat line × stress/no stress
approached significance [F(1, 19) = 4.004, p = .060], thereby indicat-
ing that preference for water tended to be affected differently by stress
in male NS and HYPER rats. Fig. 3 shows that this occurred because
stress increased the preference for water (i.e., decreased the preference
for sucrose) in HYPER males more so than in NS males. This interaction
effect was supported by comparison of stress and non-stress conditions
within each rat line, which revealed that a higher preference for water
(and therefore lower preference for sucrose) in stressed HYPER males
in comparison to non-stressed HYPER males approached significance
[F(1, 9) = 4.234, p = .070], whereas there was no indication of a dif-
ference in preference for water between stressed and non-stressed NS
males [F(1, 9) = .001, p = .973].
When only the first day of each exposure period was analyzed
similarly, a significant difference between rat lines was also found
[F(1, 19) = 11.998, p = .003], thereby indicating that HYPER males
showed higher preference for water overall than did NS males.
More importantly, the interaction of rat line × stress/no stress was
significant [F(1, 19) = 5.405, p = .031] thereby indicating that pref-
erence for water was affected differently by stress in male HYPER and
NS rats. Fig. 3 shows that this occurred because stress increased
preference for water much more so in HYPER males than in NS
males. The interaction effect was supported by comparison of stress
vs. non-stress with each rat line, which revealed that stressed
HYPER males showed significantly higher preference for water than
did non-stressed HYPER males [F(1, 9) = 14.965, p = .004], whereas
preference for water was not significantly different when stressed and
non-stressed NS males were compared [F(1, 9) = 3.106, p = .112].
a
[F(2, 27) = 8.970, p = .001], which indicated that different rat lines
showed significantly different preference for water. Further analyses
comparing pairs of rat lines revealed that preference for water did not
differ when SUS and NS rats were compared [F(1, 17) = .296, p =
.593] whereas RES rats showed significantly higher preference for
water (and therefore lower preference for sucrose) when compared
with either NS or SUS rats [F(1, 19) = 7.293, p = .014 and
F(1, 17) = 9.343, p = .007 respectively]. However, the interaction of
rat line × stress/no stress was not significant [F(2, 27) = .552, p =
.599], which indicated that stress did not affect water preference differ-
ently in the rat lines.
When only the first day of each exposure period was analyzed
similarly, the significant difference between rat lines remained
[F(2, 27) = 10.226, p = .000]. Further analyses comparing pairs of
rat lines revealed that preference for water did not differ when SUS
and NS were compared [F(1, 17) = .002, p = .963], but RES rats
showed significantly higher preference for water when compared
with either NS or SUS rats [F(1, 19) = 12.176, p = .002, and
F(1, 17) = 12.856, p = .002 respectively]. The interaction of rat
line × stress/no stress approached significance [F(2,27) = 3.229,
p = .052], thereby indicating that preference for water tended to be
affected differently by stress in different rat lines. Further analyses
comparing pairs of rat lines revealed that the interaction of rat
line × stress/no stress did not differ when SUS and NS rats were com-
pared [F(1, 17) = .020, p = .890], while this interaction approached
significance when RES rats were compared with either NS or SUS rats
[F(1, 19) = 4.167, p = .055, and F(1, 17) = 3.737, p = .070 respec-
tively]. These results were supported by comparisons of stressed and
non-stressed rats within each rat line, which revealed that preference
for water (and therefore lower preference for sucrose) was distinctly
higher in stressed RES rats than in non-stressed RES rats [F(1, 9) =
9.655, p = .013] while this difference did not reach significance in
either SUS or NS rats [F(1, 7) = 4.292, p = .077 and F(1, 9) =
1.110, p = .320 respectively].
3.2.3. SwHi and SwLo rats
Fig. 4 (bottom) shows the preference for water of both stressed
and non-stressed male SwHi, SwLo, and NS rats. When all days in
which rats were offered sucrose as well as water were analyzed by
a three factor repeated measures analysis of covariance as described
previously, the rat line × stress/no stress interaction was significant
[F(2, 28) = 6.106, p = .006]. When pairs of rat lines were compared
in the same manner, significant rat line × stress/no stress interactions
were found comparing SwHi and NS rats [F(1, 19) = 14.610, p =
.001], and SwHi and SwLo rats [F(1, 18) = 4.956, p = .039], thereby
indicating that stress increased preference for water (i.e., decreased
preference for sucrose) more so in SwHi rats than in NS or SUS rats.
The interaction of rat line × stress/no stress between SwLo and NS
rats was not significant [F(1, 18) = 1.430, p = .247], which indicated
that stress did not affect SwLo and NS rats differently. These results
were supported by the evidence that stressed SwHi rats showed
significantly higher preference for water (and therefore lower
preference for sucrose) than did non-stressed SwHi rats [F(1, 9) =
12.629, p = .006]. In contrast, preference for water did not differ for
stressed and non-stressed SwLo rats [F(1, 8) = .326, p = .583]
while in NS rats, stressed rats tended to show lower preference for
water (i.e., higher preference for sucrose) compared to non-stressed
NS rats [F(1, 9) = 4.687, p = .059.], opposite to the effect of stress
seen in SwHi rats.
When only the first day of each exposure period was analyzed
similarly, a difference between rat lines was found that approached
significance [F(2, 28) = 2.766, p = .080] as did the interaction of
rat line × stress/no stress [F(2, 28) = 2.699, p = .085]. Further anal-
yses comparing pairs of rat lines revealed that SwHi rats showed
3.2.2. SUS and RES rats
Fig. 4 (top) shows the preference for water of stressed and
non-stressed male NS, SUS, and RES rats. When all days in which rats
were offered sucrose as well as water were analyzed by a three factor