Physiology&Behavior,Vol. 49, pp. 245-250. ©Pergamon Press plc, 1991. Printed in the U.S.A.
0031-9384/91 $3.00 + .00
Sex Differences in Animal Tests of Anxiety A M A N D A L. J O H N S T O N .1 A N D S A N D R A E. F I L E t 2
*Department of Psychiatry and Behavioral Sciences, The Medical University of South Carolina, Charleston, SC 29425 and "pPsychopharmacology Research Unit, UDMS Division of Pharmacology Guy's Hospital, London SE1 9RT, England R e c e i v e d 14 M a y 1990
JOHNSTON, A. L. AND S. E. FILE. Sex differences in animal tests of anxiety. PHYSIOL BEHAV 49(2) 245-250, 1991.--To explore further the meaning of sexually dimorphic behavior in the open-field test, male and female hooded Lister rats were tested in three tests of anxiety. In the social interaction test, the social interaction scores of the female rats were lower and did not increase as readily following familiarization to the apparatus as those of the male rats. In the elevated plus-maze test, female rats showed a reduced aversion to the open arms compared to male rats; and in a modified Vogel conflict test, the punished licking rates of the female rats were lower than those of the male rats. It is concluded that the behavior of male and female rats differs in these tests, but that firm conclusions concerning sex differences in anxiety levels cannot be made because all three tests did not lead to predictions which were in the same direction. It is also suggested that cautious interpretation is necessary because these tests may measure different variables in male and female rats and they may not be valid tests of anxiety for female rats. Sex differences
Anxiety
Social interaction test
Elevated plus-maze test
Vogel conflict test
much faster than females it would not have been possible to match body weights in older animals. The social interaction test has previously been used to test male and female adolescent (days 31-45) rats (12). At this age there were no sex differences and both sexes showed the same increased social interaction in a familiar, as opposed to unfamiliar, test arena. The levels of social interaction were much higher than those seen in adult males, e.g., at day 65 (10), and both sexes showed a high incidence of rough and tumble play fighting. The sex differences in the open field appear around puberty with the males showing a postpubertal decrease in activity, but the females retaining their prepubertal levels (4,16). It will therefore be interesting to see whether adult females retain high levels of social interaction. However, it is important to be cautious about any extension of the social interaction test to females since it is possible that social investigation of like-sex partners serves different functions in males and females. The social interaction test was extensively validated in male rats and this validation cannot be assumed necessarily to apply also to females.
THERE axe sex differences in a wide variety of nonreproductive behaviors in rodents [for review see (3)]. In particular, it is well documented that adult females show greater ambulatory and rearing activity and defecate less than males in the open-field test [see (1, 5, 6, 16)]. The higher level of motor activity and lower defecation rate seen in females has been taken to indicate that they exhibit less emotionality (anxiety, fearfulness) than males (13,14). However, this interpretation has been severely criticized because there are many physiological differences such as body weight or food intake that could account for differences in ambulation and defecation (1,2). In addition, measures from the open-field test are believed by many to be nonspecific and complex, reflecting aspects of many behaviors including exploration, locomotor activity and anxiety (1,9). The purpose of these experiments was to explore further the meaning of the sexually dimorphic behavior in the open-field test by testing male and female rats in more specific animal tests of anxiety. The tests chosen were the social interaction test (8), in which anxiety is measured by the decrease in social interaction between pairs of rats placed in a novel enclosure; the elevated plus-maze test (17), in which anxiety is measured by the aversion of rats for the open elevated arms of the maze; and a modified Vogel punished drinking test (18), in which anxiety is measured by shock-induced inhibition of licking. Young adult rats were tested and in order to equate body weights in the two sexes the males were tested 1 week younger than the females. The rats were therefore tested at days 63 (males) and 70 (females), which is at least 3 weeks past adolescence and certainly after sexual puberty. Since male rats gain weight so
METHOD
Animals In Experiments 1 and 2, separate groups of equivalent weight (176-224 g) male (n = 16) and female (n = 16) hooded Lister rats (Olac Ltd., Bicester) were used. The females were 70 days old and the males 63 days old at the time of testing. In Experiment 1 the males weighed 199.2---2.3 g and the females 1 9 7 . 6+- 1.5 g and in Experiment 2 the males weighed 201.5---3.2 g and the fe-
tPresent address: Agouron Pharmaceuticals, La Jolla. 2Requests for reprints should be addressed to Sandra E. File.
245
246
males 2 0 3 . 0 _ 1.7 g. The rats were housed under dim lighting conditions (light level averaged 35 radiometric lux), in a room with a 12-h light:12-h dark cycle (lights on: 0700 h) and, except as otherwise indicated, were allowed free access to food and water. The rats were singly housed 5 days before the first social interaction test and remained in these conditions for all the testing procedures.
Apparatus The social interaction test arena was a wooden box with a solid floor, 60 x 60 × 35 cm. The arena was dimly lit (the illuminance was 60 radiometric lux). Interruption of infrared photobeams, positioned 4.5 cm above the floor, gave an automated measure of locomotor activity. A camera was mounted vertically above the arena and the social behavior of the rats was observed on a monitor in an adjacent room. An observer recorded the duration of behaviors on a keyboard that fed directly into a microcomputer. The plus-maze consisted of two open arms (50 x 10 cm) and two enclosed arms, of the same size, with 40-cm high walls. The apparatus was arranged so that the arms of the same type were opposite each other. There was a central square of 10 cm and the light level at this point was 119 radiometric lux. The apparatus was wooden and elevated to a height of 50 cm. The Vogel apparatus consisted of a rectangular box 27 x 19.5 × 18 cm, with a metal grid floor, through which scrambled shocks were delivered (0.175 mA, 0.5 s). At one end of the box the rat had access to a stainless steel drinking spout. Licks were counted by the clicks made by a ball-bearing in the spout and were recorded directly by a microcomputer which also controlled shock presentation.
Procedure
JOHNSTON AND FILE
periment were tested only once in the plus-maze. The Vogel test. Seven days after testing on the elevated plus-maze training for the Vogel test commenced. Each rat was deprived of water for 24 h and then placed in the test chamber for 15 rain to allow familiarization to the apparatus. During this time the rat had free access to the water in the chamber. It was then returned to its home cage and water deprived for a further 24 h. The next day each rat was given a 6-min test. The rat was allowed 1 rain of unpunished licking, then for the next 5 min every 21st lick was punished with a foot-shock. The number of punished and unpunished licks were recorded. Rats were tested between 1300 and 1800 h in a randomized order.
Experiment 2. The social interaction test. In Experiment 2 the social interaction test was used to investigate possible sex differences in response to increasing the familiarity of the test. The procedure was identical to that used in Experiment 1, except that in order to investigate increasing levels of familiarity the rats were tested on two additional occasions. Therefore, on the two successive days following the second social interaction test, all pairs of rats received a third (familiar × 2) and fourth (familiar × 3) trial in the arena.
Statistics The data were analyzed by one-way analyses of variance (ANOVA) with sex as the independent factor. For the social interaction test, a split-plot A N O V A (with familiarity as the within group factor) was used to analyze the time spent in social interaction and the locomotor activity scores. For the elevated plusmaze, ANOVA was performed on the percentage of arm entries made onto the open arms, the percentage of time spent on the open arms and the total number of arm entries. The number of unpunished and punished licks recorded in the Vogel apparatus were also analyzed using ANOVA.
Experiment 1. The social interaction test. Rats were paired within their sexes on the basis of weight. Each pair of rats was tested in an unfamiliar and a familiar condition. On day one of testing, pairs of rats were placed for 7.5 min in the test arena with which they were unfamiliar. On the following day, the rats were placed individually in the arena for 7.5 min to familiarize them with the apparatus. On the third day the same pairs of rats were given a second 7.5-min test in the now familiar arena. On both test days, the time spent by the rats in active social interaction (including sniffing, following, grooming, kicking, boxing, biting, wrestling, crawling under or over the partner) was scored by an observer, who was blind to the sex of the rats. The arena was cleaned after each trial. Rats were tested between 0800 and 1200 h in a randomized order. Immediately following the second social interaction test, one member of each pair was tested on the elevated plus-maze. The elevated plus-maze. Each rat was placed in the center of the maze, facing one of the enclosed arms. During a 5-min test period, the behavior of the rat was recorded by an observer sitting 1 m from the center of the maze. The following measures were recorded: the number of entries onto, and the time spent on, (a) open and (b) enclosed arms. The maze was cleaned after each trial. These values were then converted into three measures: the percentage of arm entries made onto the open arms and the percentage of time spent on the open arms (believed to inversely reflect levels of anxiety); and the total number of ann entries (believed to reflect general activity). Although it has been found that repeated testing of male undrugged rats does not change the scores in the plus-maze (17), repeated testing does reduce the effects of benzodiazepine drugs (11). Therefore, the rats in this ex-
RESULTS
Experiment 1 The social interaction test. Figure 1 shows that male rats spent more time in social interaction than female rats, F ( 1 , 1 4 ) = 23.2, p < 0 . 0 0 0 5 . Both male and female rats spent more time in social interaction in the familiar condition than in the unfamiliar condition [familiarity factor, F ( 1 , 1 4 ) = 4 2 . 1 , p < 0 . 0 0 0 1 , see Fig. 1], but the scores of the male rats showed a more pronounced increase with the increase in familiarity than those of the female rats [this was indicated by a significant sex x familiarity interaction, F ( 1 , 1 4 ) = 6 . 9 , p < 0 . 0 5 ] . In this experiment there was no incidence of aggression in either the male or female rats. Female rats displayed higher locomotor activity scores than male rats in this test [see Fig. 1; F ( 1 , 1 4 ) = 3 0 . 9 , p
0031-9384/91 $3.00 + .00
Sex Differences in Animal Tests of Anxiety A M A N D A L. J O H N S T O N .1 A N D S A N D R A E. F I L E t 2
*Department of Psychiatry and Behavioral Sciences, The Medical University of South Carolina, Charleston, SC 29425 and "pPsychopharmacology Research Unit, UDMS Division of Pharmacology Guy's Hospital, London SE1 9RT, England R e c e i v e d 14 M a y 1990
JOHNSTON, A. L. AND S. E. FILE. Sex differences in animal tests of anxiety. PHYSIOL BEHAV 49(2) 245-250, 1991.--To explore further the meaning of sexually dimorphic behavior in the open-field test, male and female hooded Lister rats were tested in three tests of anxiety. In the social interaction test, the social interaction scores of the female rats were lower and did not increase as readily following familiarization to the apparatus as those of the male rats. In the elevated plus-maze test, female rats showed a reduced aversion to the open arms compared to male rats; and in a modified Vogel conflict test, the punished licking rates of the female rats were lower than those of the male rats. It is concluded that the behavior of male and female rats differs in these tests, but that firm conclusions concerning sex differences in anxiety levels cannot be made because all three tests did not lead to predictions which were in the same direction. It is also suggested that cautious interpretation is necessary because these tests may measure different variables in male and female rats and they may not be valid tests of anxiety for female rats. Sex differences
Anxiety
Social interaction test
Elevated plus-maze test
Vogel conflict test
much faster than females it would not have been possible to match body weights in older animals. The social interaction test has previously been used to test male and female adolescent (days 31-45) rats (12). At this age there were no sex differences and both sexes showed the same increased social interaction in a familiar, as opposed to unfamiliar, test arena. The levels of social interaction were much higher than those seen in adult males, e.g., at day 65 (10), and both sexes showed a high incidence of rough and tumble play fighting. The sex differences in the open field appear around puberty with the males showing a postpubertal decrease in activity, but the females retaining their prepubertal levels (4,16). It will therefore be interesting to see whether adult females retain high levels of social interaction. However, it is important to be cautious about any extension of the social interaction test to females since it is possible that social investigation of like-sex partners serves different functions in males and females. The social interaction test was extensively validated in male rats and this validation cannot be assumed necessarily to apply also to females.
THERE axe sex differences in a wide variety of nonreproductive behaviors in rodents [for review see (3)]. In particular, it is well documented that adult females show greater ambulatory and rearing activity and defecate less than males in the open-field test [see (1, 5, 6, 16)]. The higher level of motor activity and lower defecation rate seen in females has been taken to indicate that they exhibit less emotionality (anxiety, fearfulness) than males (13,14). However, this interpretation has been severely criticized because there are many physiological differences such as body weight or food intake that could account for differences in ambulation and defecation (1,2). In addition, measures from the open-field test are believed by many to be nonspecific and complex, reflecting aspects of many behaviors including exploration, locomotor activity and anxiety (1,9). The purpose of these experiments was to explore further the meaning of the sexually dimorphic behavior in the open-field test by testing male and female rats in more specific animal tests of anxiety. The tests chosen were the social interaction test (8), in which anxiety is measured by the decrease in social interaction between pairs of rats placed in a novel enclosure; the elevated plus-maze test (17), in which anxiety is measured by the aversion of rats for the open elevated arms of the maze; and a modified Vogel punished drinking test (18), in which anxiety is measured by shock-induced inhibition of licking. Young adult rats were tested and in order to equate body weights in the two sexes the males were tested 1 week younger than the females. The rats were therefore tested at days 63 (males) and 70 (females), which is at least 3 weeks past adolescence and certainly after sexual puberty. Since male rats gain weight so
METHOD
Animals In Experiments 1 and 2, separate groups of equivalent weight (176-224 g) male (n = 16) and female (n = 16) hooded Lister rats (Olac Ltd., Bicester) were used. The females were 70 days old and the males 63 days old at the time of testing. In Experiment 1 the males weighed 199.2---2.3 g and the females 1 9 7 . 6+- 1.5 g and in Experiment 2 the males weighed 201.5---3.2 g and the fe-
tPresent address: Agouron Pharmaceuticals, La Jolla. 2Requests for reprints should be addressed to Sandra E. File.
245
246
males 2 0 3 . 0 _ 1.7 g. The rats were housed under dim lighting conditions (light level averaged 35 radiometric lux), in a room with a 12-h light:12-h dark cycle (lights on: 0700 h) and, except as otherwise indicated, were allowed free access to food and water. The rats were singly housed 5 days before the first social interaction test and remained in these conditions for all the testing procedures.
Apparatus The social interaction test arena was a wooden box with a solid floor, 60 x 60 × 35 cm. The arena was dimly lit (the illuminance was 60 radiometric lux). Interruption of infrared photobeams, positioned 4.5 cm above the floor, gave an automated measure of locomotor activity. A camera was mounted vertically above the arena and the social behavior of the rats was observed on a monitor in an adjacent room. An observer recorded the duration of behaviors on a keyboard that fed directly into a microcomputer. The plus-maze consisted of two open arms (50 x 10 cm) and two enclosed arms, of the same size, with 40-cm high walls. The apparatus was arranged so that the arms of the same type were opposite each other. There was a central square of 10 cm and the light level at this point was 119 radiometric lux. The apparatus was wooden and elevated to a height of 50 cm. The Vogel apparatus consisted of a rectangular box 27 x 19.5 × 18 cm, with a metal grid floor, through which scrambled shocks were delivered (0.175 mA, 0.5 s). At one end of the box the rat had access to a stainless steel drinking spout. Licks were counted by the clicks made by a ball-bearing in the spout and were recorded directly by a microcomputer which also controlled shock presentation.
Procedure
JOHNSTON AND FILE
periment were tested only once in the plus-maze. The Vogel test. Seven days after testing on the elevated plus-maze training for the Vogel test commenced. Each rat was deprived of water for 24 h and then placed in the test chamber for 15 rain to allow familiarization to the apparatus. During this time the rat had free access to the water in the chamber. It was then returned to its home cage and water deprived for a further 24 h. The next day each rat was given a 6-min test. The rat was allowed 1 rain of unpunished licking, then for the next 5 min every 21st lick was punished with a foot-shock. The number of punished and unpunished licks were recorded. Rats were tested between 1300 and 1800 h in a randomized order.
Experiment 2. The social interaction test. In Experiment 2 the social interaction test was used to investigate possible sex differences in response to increasing the familiarity of the test. The procedure was identical to that used in Experiment 1, except that in order to investigate increasing levels of familiarity the rats were tested on two additional occasions. Therefore, on the two successive days following the second social interaction test, all pairs of rats received a third (familiar × 2) and fourth (familiar × 3) trial in the arena.
Statistics The data were analyzed by one-way analyses of variance (ANOVA) with sex as the independent factor. For the social interaction test, a split-plot A N O V A (with familiarity as the within group factor) was used to analyze the time spent in social interaction and the locomotor activity scores. For the elevated plusmaze, ANOVA was performed on the percentage of arm entries made onto the open arms, the percentage of time spent on the open arms and the total number of arm entries. The number of unpunished and punished licks recorded in the Vogel apparatus were also analyzed using ANOVA.
Experiment 1. The social interaction test. Rats were paired within their sexes on the basis of weight. Each pair of rats was tested in an unfamiliar and a familiar condition. On day one of testing, pairs of rats were placed for 7.5 min in the test arena with which they were unfamiliar. On the following day, the rats were placed individually in the arena for 7.5 min to familiarize them with the apparatus. On the third day the same pairs of rats were given a second 7.5-min test in the now familiar arena. On both test days, the time spent by the rats in active social interaction (including sniffing, following, grooming, kicking, boxing, biting, wrestling, crawling under or over the partner) was scored by an observer, who was blind to the sex of the rats. The arena was cleaned after each trial. Rats were tested between 0800 and 1200 h in a randomized order. Immediately following the second social interaction test, one member of each pair was tested on the elevated plus-maze. The elevated plus-maze. Each rat was placed in the center of the maze, facing one of the enclosed arms. During a 5-min test period, the behavior of the rat was recorded by an observer sitting 1 m from the center of the maze. The following measures were recorded: the number of entries onto, and the time spent on, (a) open and (b) enclosed arms. The maze was cleaned after each trial. These values were then converted into three measures: the percentage of arm entries made onto the open arms and the percentage of time spent on the open arms (believed to inversely reflect levels of anxiety); and the total number of ann entries (believed to reflect general activity). Although it has been found that repeated testing of male undrugged rats does not change the scores in the plus-maze (17), repeated testing does reduce the effects of benzodiazepine drugs (11). Therefore, the rats in this ex-
RESULTS
Experiment 1 The social interaction test. Figure 1 shows that male rats spent more time in social interaction than female rats, F ( 1 , 1 4 ) = 23.2, p < 0 . 0 0 0 5 . Both male and female rats spent more time in social interaction in the familiar condition than in the unfamiliar condition [familiarity factor, F ( 1 , 1 4 ) = 4 2 . 1 , p < 0 . 0 0 0 1 , see Fig. 1], but the scores of the male rats showed a more pronounced increase with the increase in familiarity than those of the female rats [this was indicated by a significant sex x familiarity interaction, F ( 1 , 1 4 ) = 6 . 9 , p < 0 . 0 5 ] . In this experiment there was no incidence of aggression in either the male or female rats. Female rats displayed higher locomotor activity scores than male rats in this test [see Fig. 1; F ( 1 , 1 4 ) = 3 0 . 9 , p
