- Email: [email protected]
Prenatal protein malnutrition affects the social interactions of juvenile rats
Physiology & Behavior, Vol. 60, No. 1, 197-201, 1996 Copyright © 1996 Elsevier Science Inc. Printed in the USA. All rights reserved 0031-9384/96 $15.00 + .00
SSDI 0031-9384(95)02236-8
ELSEVIER
Prenatal Protein Malnutrition Affects the Social Interactions of Juvenile Rats S. S. ALMEIDA,* J. TONKISS'~ 1 A N D J. R. G A L L E R t
*Laborat6rio de Nutrig~o e Comportamento, FFCLRP e N~cleo de Neuroci~ncias e Comportamento da Universidade de S[to Paulo, 14040-901 Ribeirfto Preto, SP, Brasil, and "]'Centerfor Behavioral Development and Mental Retardation, M923, Boston University School of Medicine, 80 E. Concord St., Boston, MA 02118 USA Received 3 August 1995 ALMEIDA, S. S., J. TONKISS AND J. R. GALLER. Prenatalprotein malnutrition affects the social interactions of juvenile rats. PHYSIOL BEHAV 60(1) 197-201, 1996.--The effects of prenatal protein malnutrition on juvenile social behavior was investigated in male and female rats. Animals were provided with 25% (control) or 6% (low protein) casein diets before and during pregnancy. After birth eight pups in each litter (six males and two females) were fostered to lactating control mothers. After weaning (21 days of age) all animals received a lab chow diet until behavioral testing began at 45 days of age. To assess social interaction, pairs of rats of the same gender, consisting of one malnourished and one control rat, were placed in a familiar rectangular arena on 3 consecutive days. Playful social behavior (pin), nonplayful social behaviors (anogenital sniff, walk-over, side-mount, and allogroom), and nonsocial behavior (rear) were recorded in 10-min sessions. Prenatal malnutrition significantly decreased both playful and nonplayful social behaviors, and increased nonsocial rearing. No significant gender differences were observed. The finding that early social behavior is altered by prenatal malnutrition opens the possibility that such changes may play an important role in determining some of the later behavioral differences described in the adult animal. Prenatal malnutrition Social behavior Walk-over Nonsocial rearing Rat
Social play
SOCIAL play has a high incentive value (4,9) and young rats will readily learn instrumental responses for the opportunity to engage others in bouts of play (13,17). Although an immediate goal of play is difficult to identify, the critical importance of this behavior is evident in the finding that play deprivation leads to significant alterations of behavior in adult life, such as lower social interaction and aggression (20,27). Current theories propose that play may be a forerunner of adult aggression (27,28), and it is through this amicable rough-and-tumble play that young rats learn their place in the social hierarchy without sustaining serious injury (15). Alternatively, play may serve to equip individuals with some basic skills and strategies essential for a variety of behaviors that are expressed in adulthood (23). Malnutrition experienced at times encompassing the suckling period has been shown to significantly impact the social milieux in rats and mice, affecting both mother-infant [reviewed in (7)] and interpup interactions [reviewed in (22)]. Although studies of social play have not been performed explicitly in malnourished animals, Frafikovfi (5) described various aspects of social interac-
Pinning
Social grooming
tion in infant and juvenile rats experiencing protein-calorie malnutrition from birth. The results can be summarized as indicaring an overall reduction in social interaction with a significant increase in aggressive behavior. For example, the latency for the first approach of one rat to another was significantly increased in pairs of malnourished rats relative to that observed in wellnourished pairs, whereas the number of attacks was significantly higher. Findings of increased aggression (10,19,31,34,38) and increased social responsiveness (34,37,39) in postnatally malnourished animals, when tested as adults, may result from significant alterations in the early social environment. Unlike postnatal malnutrition, malnutrition limited to the prenatal period has been shown to produce only minimal changes in mother-pup interactions in rats (8,33). However, the effect of such treatment on other aspects of the early social environment (i.e., interactions between pups during infancy and prior to adolescence) has not been examined. Thus, the aim of the present investigation was to examine social play (pins) and other social interactions (walk-overs, allogrooms, side mounts, and anogenital
1 To whom requests for reprints should be addressed.
197
198
ALMEIDA, TONKISS AND GALLER
sniffs) between prenatally malnourished and well-nourished control rats at around the peak age for social play (45 days). Pinning behavior is a well-established measure of social play in rats (15,16), and is known to be significantly affected by prenatal insults [e.g., (41)]. The remaining measures have previously been used to assess social interactions in adult rats, and have been shown to be disrupted by postnatal malnutrition (34,37,39,40). METHOD
Nutritional Treatment Five weeks prior to mating, nulliparous female rats (Sprague-Dawley VAF plus; Charles River Laboratories, Kingston, MA) were obtained and allowed ad lib access to one of two isocaloric diets (Teklad, Madison, WI). The diets were formulated to be of adequate protein (25% casein) or low-protein (6% casein) content [detailed description given in (8)]. Males obtained from the same source were acclimated to the experimental diets of the females for 1 week prior to mating. Each male was introduced into a cage of two to three females receiving the same diet treatment. Mating occurred over a 10-day period. The presence of sperm in a vaginal smear detected whether mating had occurred. One week prior to the projected delivery date the females were caged individually in polycarbonate breeding cages (51 X 41 X 21 cm; Lab Products Inc., Maywood, NJ). Following parturition, litters were culled to eight pups (six males and two females) and fostered as whole litters to well-nourished mothers that had given birth no more than 24 h previously. Pups were selected for fostering based upon their proximity to the litter mean body weight for each sex. Litters born to dams fed the 6% casein diet and cross-fostered to lactating dams given 25% casein diet were designated as 6-25 (prenatally malnourished; n = 8). Litters born to dams fed the 25% casein diet and fostered to other lactating dams fed 25% casein were designated as 25-25 (controis; n = 8). A more detailed description of the nutritional, mating and fostering procedures is given in a previous article (32). The rats were weighed at birth and on the day of behavioral testing using an electronic balance (Ohaus, GT 4000). At weaning (21 days of age) rats of the same sex and nutritional treatment were caged in pairs and given ad lib access to rat chow (PMI foods Inc., formula 5001).
there were eight rats per treatment group all derived from different litters. These rats were housed singly in wire hanging cages until the experiment began at 45 days of age. Single housing was used because social isolation has been reported to increase social play (12,15,16). To familiarize the animals to the experimental conditions, all rats were placed individually in the experimental cage for a 5-min session on each of 5 consecutive days. No data were recorded during these familiarization sessions. Test sessions of 10 rain were then given on each of 3 consecutive days and consisted of placing a pair of rats of the same sex (one 6-25 and one 25-25 rat) in the arena. The rats of a pair were selected so as not to differ in body weight by more than 10 g. For identification purposes one rat in the pair was marked with black ink on the top of the head. The mark was randomly varied according to the dietary treatment of the rat. Videotapes were analyzed by two experimenters blind to the nutritional condition of the rat, and the following variables (34,35) were recorded: Pin: one animal lies on its back while the other stands over and in contact with it. Walk-over: the animal walks over the head or back of another. Allogroom: the licking or nibbling of the fur of one animal by the other. Side mount: one animal approaches the other from the side and places its forepaws on the other's dorsal surface. Anogenital sniff: the sniff of the anogenital region of one rat by the other. Rear: standing on the hind legs in a nonsocial context.
Statistical Analyses The body weights of the nutritional treatment groups were compared by two-way ANOVA (treatment x sex). Separate analyses were applied for the birth weights and the weights at testing because of vastly dissimilar variances at the two ages. Behavioral data were analyzed by a three-way ANOVA (sex X treatment × session) with treatment and session as repeated measures (i.e., the "pair" was used as the unit for analysis allowing two levels of treatment). Post hoc analyses were conducted by Newman-Keuls tests. RESULTS
Body Weight Housing Conditions The animals were maintained in a room with temperature of 73 ___3°F and 45-55% humidity. A reverse 12-h night (0800-2000 h)/12-h day (2000-0800 h) light cycle accommodated behavioral observations to the active waking period of the rat. Continuous dim illumination was provided by red fluorescent lighting during the dark portion of the cycle.
Apparatus The testing arena consisted of an acrylic plastic cage measuring 27 X 25 x 32 cm, with 2 cm of Sani-Chips (P. J. Murphy, Montville, NJ) covering the floor. Two red lights (60 W) provided the only illumination in the experimental room and were mounted 80 cm above the arena. Experimental sessions were recorded by a vertically mounted videocamera (Panasonic, model WV-BL200), linked to a monitor and a video cassette recorder in an adjacent room.
Procedure On postnatal day 33, one male and one female rat per litter were selected at random for behavioral testing. Thus, for each sex
The mean body weights of the two treatment groups at birth and at testing (day 45) are given in Table 1. At birth, the 6-25 rats showed significant body weight deficits compared with well-nourished control rats, F(1, 28) = 6.31, p < 0.02 (15% for males and 7% for females). No effect of sex and no significant treatment X sex interaction were indicated by ANOVA. At testing, body weight differences between the groups had narrowed but the treatment effect remained statistically significant, F(1, 28) = 5.88, p < 0.05. Male rats weighed more than the females,
TABLE 1 MEAN + SEM BODY WEIGHTS Well-Nourished 25-25 Group
Prenatally Malnourished 6-25 Group
Deficit
8) = 8)
5.86 + 0.22 5.55 ::t: 0.22
5.00 + 0.33 5.17 -t- 0.21
15% 7%
8) = 8)
249 5 : 2 186 :t: 4
237 + 4 180:1:4
Age Birth Male ( n = Female (n Testing Male ( n = Female ( n
5% 3%
MALNUTRITION AND SOCIAL BEHAVIOR
E
~
199
,z' 1 5 -
104
•
1
2
10
*
3
~=
0
1
2
3
45
SESSION
SESSION
FIG. 1. Mean+ SEM of pins, walk-overs, allogrooms, and rears for prenatally malnourished (6-25 group, open bars) and well-nourished control rats (25-25 group, filled bars) for each of three sessions. * p < 0.05, **p < 0.01 compared to 25-25 (Newman-Keuls). No significant sex differences were indicated, hence the data have been combined for illustrative purposes.
F(1, 28) = 255.91, p < 0.001, but there was no significant treatment × sex interaction.
Behavioral Measures Interobserver reliabilities were calculated on 12 randomly selected subjects and correlation values (r s) of _> 0.90 were obtained for each of the behaviors measured. Rear had the lowest reliability (r = 0.90) and pin had the highest ( r = 0.96). Pinning. Figure 1 (upper left) illustrates the mean number of times the 6-25 rats pinned the 25-25 rats and vice versa over the three test sessions. Prenatal malnutrition affected pinning behavior as indicated by a significant effect of treatment, F(1, 14) = 9.36, p < 0.01. A treatment by sessions interaction, F(2, 28) = 4.95, p < 0.05, and a main effect of session, F(2, 2 8 ) = 5.15, p < 0.05, were also indicated by the ANOVA. Post hoc analyses showed that there was no significant difference in the number of pins exhibited by 6-25 and 25-25 rats on session 1. However, the incidence of pinning increased in the 25-25 group over the three sessions whereas the incidence of pinning in the 6-25 group remained fairly constant. Consequently, 25-25 rats made significantly more pins than the 6-25 rats on sessions 2 and 3 ( p < 0.05). No significant effect of sex and no significant interactions involving sex were indicated. Walk-over Figure 1 (upper right) illustrates the mean incidence of walk over for 6-25 and 25-25 rats over the three test sessions. Prenatal malnutrition affected walk-over behavior as indicated by a significant effect of treatment, F(1, 14) = 14.27, p < 0.01. A treatment by session interaction, F(2, 2 8 ) = 4.30, p < 0.05, and a main effect of session, F(2, 28) = 3.96, p < 0.05, were also indicated by the ANOVA. Post hoc analysis showed that the pattern of results was strikingly similar to that observed for pinning. Although there was no significant difference in the number of walk-overs exhibited by 6-25 and 25-25 rats on session 1, the incidence of walk-over increased in the 25-25
group whereas the incidence of walk-over in the 6-25 group remained fairly constant over sessions. Consequently, 25-25 rats made significantly more walk-overs than the 6-25 rats on sessions 2 and 3 ( p < 0.01). No significant effect of sex and no significant interactions involving sex were indicated. Allogroom. Figure 1 (lower left) illustrates the mean incidence of allogroom for 6-25 and 25-25 rats over the three test sessions. ANOVA indicated a significant effect of treatment, F(1, 14) = 8.00, p < 0.05, due to the 25-25 rats performing more allogrooming than the 6-25 rats. The incidence of allogrooming decreased across sessions, F(2, 2 8 ) = 5.26, p < 0.05, but no significant treatment by session interaction was indicated. The main effect of sex and interactions involving sex were not significant. Side mount. ANOVA indicated a significant effect of session, F(2, 28) = 4.13, p < 0.05, due to a decrease in the incidence of this behavior across sessions, but there was no treatment effect or treatment by session interaction. Similarly, there was no significant effect of sex and no significant interactions involving sex (data not shown). Anogenital sniff. Anogenital sniff was not affected by any of the factors studied (data not shown). Rear. Figure 1 (lower right) illustrates the mean incidence of rearing in 6-25 and 25-25 rats over the three test sessions. ANOVA showed a significant treatment by session interaction, F(2, 2 8 ) = 4.11, p < 0.03. The main effects of treatment and session were not significant. Similar to pinning and walk-over, the two treatment groups did not differ on session 1. However, over sessions the incidence of rearing increased in malnourished animals and decreased in control animals, leading to significantly greater levels of rearing in the prenatally malnourished rats on sessions 2 and 3 compared with the controls ( p < 0.05). There was no significant effect of sex and no significant interactions involving sex. Relationships between behaviors--factor analysis. The apparent similarity of the findings for pinning and walk-over (Fig. 1) suggested a possible link between these two behaviors. Hence, a principal components factor analysis of the correlations among the six measures was carried out. The data were combined over sessions such that there were 96 observations on the 32 rats. The first principal component (unrotated) had an eigenvalue of 1.86 accounting for 31% of the variance. Correlations between individual behaviors loading on the first factor were: walk-over = 0.773, side mount = 0.756, and allogroom = 0.690, with pin loading only slightly (0.396). Rearing and anogenital sniff were not correlated and rearing loaded on separate subsequent components, confirming the nonsocial nature of this behavior. DISCUSSION
The present study demonstrated that prenatal malnutrition reduced the incidence of various social behaviors (pin, walk-over, and allogroom) relative to that of well-nourished animals. In contrast, nonsocial exploratory behavior (rear) was increased by malnutrition. None of these differences were present on the first test session; rather, they emerged with repeated testing of the same dyads over subsequent test days. In the case of pinning behavior, which has been interpreted as "dominance behavior" (1,15,21,41), these data are consistent with previous reports showing that dominance patterns emerge over repeated pairings of the same individuals (16). Thus, the finding that control animals showed a greater incidence of pins than prenatally malnourished animals, accounting on the average for 70% of total pins occurring by session 3, would suggest that adolescent wellnourished animals were more likely to assume dominance. Grooming of the partner has also been linked to social dominance
200
ALMEIDA, TONKISS AND GALLER
(24). Consequently, the observation of greater allogrooming of the prenatally malnourished rats by the control animals supports the interpretation of greater submissiveness in the malnourished rats. The absence of differences between males and females in the incidence of pinning agrees with some previously reported results (29), but it is in contradiction with other studies in which a lower number of pins was observed in females rats (11,14,18,26,36,41). It is possible that the predominance of males in the culled litters may have influenced the social development of the females prior to weaning. Because males constituted 75% of the litter, male-female interactions would have been more likely to occur than if each sex had been represented in equal numbers. The lower frequency of walk-over and allogroom in malnourished rats suggests that prenatal malnutrition has a significant effect on nonplayful social interactions as well as those considered to be playful. These two behaviors, which were highly correlated in the present study (as shown by factor analysis), were included by Panksepp and Beatty (16) in a general behavioral category that they labeled "social solicitations." If these behaviors do represent an attempt of one animal to elicit a social response from the other, the lower level of these behaviors in prenatally malnourished rats may reflect a decrease in the ability of these animals to initiate and maintain social contact. Interestingly, the direction of these results is exactly opposite to that derived from malnutrition studies in rats or mice in which the period of malnutrition encompassed the suckling period. When tested in adulthood, such previously malnourished animals have been reported as more socially reactive and more aggressive than controls (34,37-40). The only previous study of social interactions in malnourished juvenile rats examined subjects malnourished postnatally and they were malnourished at the time of testing (5). Such concurrently malnourished subjects generally showed lower levels of social interaction but a greater incidence of attacks. Clearly, there are differences between social interactions in adolescence and in adulthood (e.g., in adolescent rats there is a high component of "play-fighting," which contrasts with real fighting in adult rats) [see (17) for a review, (29)]. Consequently, both playful and nonplayful social behaviors are likely to be present in different degrees in juvenile as compared to adult animals. Nonetheless, data support the conclusion that juvenile fighting may predict adult aggressiveness (27), which suggests that at least some of the differences between the findings on agonistic behavior from the present study (pinning) and previous studies (fighting) may lie with the period of malnutrition rather than the difference in the age of testing. Rearing on the hind legs was scored as a nonsocial behavior,
and the independence of this behavior from the others was confirmed by factor analysis. It has been suggested that rearing expresses the vertical component of exploration when it occurs in an empty cage (6), but the significance of rearing in the present context is uncertain. The finding that prenatally malnourished rats developed higher levels of rearing than controls on sessions 2 and 3, with lower levels of social interaction, suggests that rearing may have represented a form of exploration related to escape from the social situation. The presence of greater rearing activity in the prenatally malnourished rats also signifies that differences in social behavior cannot be attributed to generally lower activity in the malnourished animals. Although there was a significant weight advantage to the control animals at the time of testing, it was small (Table 1). Previously reported results have shown that a small body weight difference is not a determining factor in social behavior (30,36,37) and even weight advantages as large as 5-15% do not predict dominance (17). Thus, changes in playful and nonplayful behaviors following prenatal malnutrition are unlikely to be attributed to differences in body weight. In conclusion, if the hypothesis is true that early social behavior is relevant to the development of adult behavior (3,11,15,25), then the well-nourished control rats could be described as more adapted to withstand the constant pressure imposed by their environment in adulthood. However, the extent to which adult behavior can be determined by social behavior early in life remains to be investigated, and additional work related to the effects of malnutrition on both juvenile and later social behavior is needed. The observation that chronically malnourished children exhibit less social involvement, less happy affect, and greater timidity in free play compared with similar children provided with high nutritional supplementation over the first 4 years (2) suggests the importance of such studies. ACKNOWLEDGEMENTS This work was supported by NIH grants HD22539 and DA07934 (J.R.G.), grants 90/3474-0 from Funda~o de Amparo ~ Pesquisa do Estado de S~o Paulo, and 400015/91-9 (S.S.A.) and 530060/93-0 (S.S.A.) from Conselho Nacional de Desenvolvimento Cientifico e Tecnol6gico. S. S. Almeida was the recipient of a Postdoctoral Research Fellowship from Funda~.5ode Amparo h Pesquisa do Estado de S§o Paulo (FAPESP, proc. No. 93/1538-9) and a Research Fellowship from the Conselho Nacional de Desenvolvimento Cientifico e Tecnol6gico (CNPq, proc. No. 30000/91-1). During the preparation of this work S. S. Almeida was a Visiting Research Scholar at the Center for Behavioral Development and Mental Retardation, Boston University School of Medicine. We would like to thank Dr. Robert Harrison for statistical advice and Eva Sabo and Pratibha Singh for technical assistance.
REFERENCES 1. Baenninger, L. P. Comparison of behavioural development in socially isolated and grouped rats. Anim. Behav. 15:312-323; 1967. 2. Barrett, D. E.; Radke-Yarrow, M. Effects of nutritional supplementation on children's responses to novel, frustrating, and competitive situations. Am J. Clin. Nutr. 42:102-120; 1985. 3. Bekoff, M. The development of social interaction, play and rectacommunication in mammals: an ethological perspective. Q. Rev. Biol. 47:412-434; 1972. 4. Calcagnetti, D. J.; Schechter, M. D. Place conditioning reveals the rewarding aspect of social interaction in juvenile rats. Physiol. Behav. 51:667-672; 1992. 5. Fra~kov~, S. Effect of protein-calorie malnutrition on the development of social behavior in rats. Dev. Psychobiol. 6:33-43; 1973. 6. Frafikovfi, S.; Barnes, R. H. Influence of early malnutrition in early life on exploratory behavior of rats. J. Nutr. 96:477-484; 1968. 7. Galler, J. R.; Ricciuti, H. N.; Crawford, M. A.; Kucharski, L. T. The
8. 9. 10. 11. 12. 13.
role of the mother-infant interaction in nutritional disorders. In: Galler, J. R., ed. Nutrition and behavior. New York: Plenum Press; 1984:269-304. Galler, J. R.; Tonkiss, J. Prenatal protein malnutrition and maternal behavior in Sprague-Dawley rats. J. Nutr. 121:762-769; 1991. Humphreys, A. P.; Einon, D. F. Play as a reinforcer for maze-learning in juvenile rats. Anita. Behav. 29:259-270; 1981. Levitsky, D. A.; Barnes, R. H. Nutritional and environmental interactions in the behavioral development of the rat: Long-term effects. Science 176:68-71; 1972. Meaney, M. J.; Stewart, J. A descriptive study of social development in the rat (Rattus norvergicus). Anita. Behav. 29:34-45; 1981. Niesink, R. J. M.; Van Ree, J. M. Involvement of opioid and dopaminergic systems in isolation-induced pinning and social grooming of young rats. Neuropharmacology 28:411-418; 1983. Normansell, L.; Panksepp, J. Effects of morphine and naloxone on
MALNUTRITION AND SOCIAL B E H A V I O R
play-rewarded spatial discrimination in juvenile rats. Dev. Psychobiol. 23:75-83; 1990. 14. Olioff, M.; Stewart, J. Sex differences in the play behavior of prepubescent rats. Physiol. Behav. 20:113-115; 1978. 15~ Panksepp, J. The ontogeny of play in rats. Dev. Psychobiol. 14:327332; 1981. 16. Panksepp, J.; Beatty, W. W. Social deprivation and play in rats. Behav. Neural Biol. 30:197-206; 1980. 17. Panksepp, J.; Siviy, S.; Normansell, L. The psychobiology of play: Theoretical and methodological perspectives. Neurosci. Biobehav. Rev. 8:465-492; 1984. 18. Poole T. B.; Fish, J. An investigation of individual, age, and sexual differences in the play of Rattus norvergicus (Mammalia: Rodentia). Proc. Zool. Soc. Lond. 179:249-260; 1976, 19. Randt, C. T.; Blizard, D. A.; Friedman, E. Early life undernutrition and aggression in two mouse strains. Dev. Psychobiol. 8:275-279; 1975. 20. Rozenzweig, M. R. Effects of environment on development of brain and behavior. In: Tobach, E.; Aronson, L. R.; Shaw, E., eds. The biopsychology of development. New York: Academic Press; 1971:303-342. 21. Small, W. S. Notes on the psychic development of the young white rat. Am. J. Psychol. 11:80-100; 1899. 22. Smart, J. L. Undernutrition and aggression. In: Brain, P. F.; Benton, D., eds. Multidisciplinary approaches to aggression research. Amsterdam: Elsiever/North Holland Biomedical Press; 1981:179-191. 23. Smith, P. K. Does play matter? Functional and evolutionary aspects of animal and human play. Behav. Brain Sci. 5:139-184; 1982. 24. Spigel, I. M.; Trivett, S. Grooming behavior and competitive dominance in the albino rat. J. Comp. Physiol. Psychol. 78:409-411; 1972. 25. Takahashi, L. K. Postweaning environmental and social factors influencing the onset of expression of agonistic behavior in norway rats. Behav. Proc. 12:237-260; 1986. 26. Takahashi, L. K.; Lore, R. K. Play fighting and the development of agonistic behavior in male and female rats. Aggress. Behav. 9:217227; 1983. 27. Taylor, G. T. Fighting in juvenile rats and the ontogeny of agonistic behavior. J. Comp. Physiol. Psychol. 94:953-961; 1980. 28. Thor, D. H.; Flannelly, K. J. Age of intruder and territorial-elicited aggression in male Long-Evans rats. Behav. Biol. 17:237-241; 1976
201
29. Thor, D. H.; Holloway, W. R. Social Play in juvenile rats: A decade of methodological and experimental research. Neurosci. Biobehav. Rev. 8:455-464; 1984. 30. Thor, D. H.; Holloway, W. R. Social play soliciting by male and female juvenile rats: Effects of neonatal androgenization and sex of cagemate. Behav. Neurosci. 100:275-279; 1986. 31. Tikal, K.; Bene~ov~, O.; Frahkovfi, S. The effects of pyrithioxine and pyridoxine on individual behavior, social interactions, and learning in rats malnourished in early postnatal life. Psychopharmacologia 46:325-332; 1976. 32~ Tonkiss, J.; Galler, J. R. Prenatal protein malnutrition and working memory performance in adult rats. Behav. Brain Res. 40:95-107; 1990. 33. Tonkiss, J.; Shumsky, J. S.; Shultz, P. L.; Almeida, S. S.; Galler, J. R. Prenatal cocaine but not prenatal malnutrition affects foster mother-pup interactions in rats. Neurotoxicol. Teratol. 17:601-608; 1995. 34. Tonkiss, J.; Smart, J. L. Interactive effects of genotype and early life undernutrition on the development of behavior in rats. Dev. Psychobiol. 16:287-301; 1983. 35. Vanderschuren, L. J. M. J.; Niesink, R. J. M.; Spruijt, B. M.; Van Ree, J. M. Effects of morphine on diffferent aspects of social play in juvenile rats. Psychopharmacology (Berlin) 117:225-231; 1995. 36. Ward, I. L.; Stehm, K. E. Prenatal stress feminizes juvenile play patterns in male rats. Physiol. Behav. 50:601-605; 1991. 37. Whatson, T. S.; Smart, J. L. Social behavior of rats following pre and postnatal undemutrition. Physiol. Behav. 20:749-753; 1978. 38. Whatson, T. S.; Smart, J. L.; Dobbing, J. Social interactions among adult male rats after early undernutrition. Br. J. Nutr. 32:413-419; 1974. 39. Whatson, T. S.; Smart, J. L.; Dobbing, J. Dominance relationship among previously undernourished and well fed male rats. Physiol. Behav. 14:425-429; 1975. 40. Whatson, T. S.; Smart, J. L.; Dobbing, J. Undernutrition in early life: Lasting effects on activity and social behavior of male and female rats. Dev. Psychobiol. 9:529-538; 1976. 41. Wood, R. D.; Bannoura, M. D.; Johanson, 1. B. Prenatal cocaine exposure: Effects on play behavior in the juvenile rat. Neurotoxicol. Teratol. 16:139-144; 1994.
SSDI 0031-9384(95)02236-8
ELSEVIER
Prenatal Protein Malnutrition Affects the Social Interactions of Juvenile Rats S. S. ALMEIDA,* J. TONKISS'~ 1 A N D J. R. G A L L E R t
*Laborat6rio de Nutrig~o e Comportamento, FFCLRP e N~cleo de Neuroci~ncias e Comportamento da Universidade de S[to Paulo, 14040-901 Ribeirfto Preto, SP, Brasil, and "]'Centerfor Behavioral Development and Mental Retardation, M923, Boston University School of Medicine, 80 E. Concord St., Boston, MA 02118 USA Received 3 August 1995 ALMEIDA, S. S., J. TONKISS AND J. R. GALLER. Prenatalprotein malnutrition affects the social interactions of juvenile rats. PHYSIOL BEHAV 60(1) 197-201, 1996.--The effects of prenatal protein malnutrition on juvenile social behavior was investigated in male and female rats. Animals were provided with 25% (control) or 6% (low protein) casein diets before and during pregnancy. After birth eight pups in each litter (six males and two females) were fostered to lactating control mothers. After weaning (21 days of age) all animals received a lab chow diet until behavioral testing began at 45 days of age. To assess social interaction, pairs of rats of the same gender, consisting of one malnourished and one control rat, were placed in a familiar rectangular arena on 3 consecutive days. Playful social behavior (pin), nonplayful social behaviors (anogenital sniff, walk-over, side-mount, and allogroom), and nonsocial behavior (rear) were recorded in 10-min sessions. Prenatal malnutrition significantly decreased both playful and nonplayful social behaviors, and increased nonsocial rearing. No significant gender differences were observed. The finding that early social behavior is altered by prenatal malnutrition opens the possibility that such changes may play an important role in determining some of the later behavioral differences described in the adult animal. Prenatal malnutrition Social behavior Walk-over Nonsocial rearing Rat
Social play
SOCIAL play has a high incentive value (4,9) and young rats will readily learn instrumental responses for the opportunity to engage others in bouts of play (13,17). Although an immediate goal of play is difficult to identify, the critical importance of this behavior is evident in the finding that play deprivation leads to significant alterations of behavior in adult life, such as lower social interaction and aggression (20,27). Current theories propose that play may be a forerunner of adult aggression (27,28), and it is through this amicable rough-and-tumble play that young rats learn their place in the social hierarchy without sustaining serious injury (15). Alternatively, play may serve to equip individuals with some basic skills and strategies essential for a variety of behaviors that are expressed in adulthood (23). Malnutrition experienced at times encompassing the suckling period has been shown to significantly impact the social milieux in rats and mice, affecting both mother-infant [reviewed in (7)] and interpup interactions [reviewed in (22)]. Although studies of social play have not been performed explicitly in malnourished animals, Frafikovfi (5) described various aspects of social interac-
Pinning
Social grooming
tion in infant and juvenile rats experiencing protein-calorie malnutrition from birth. The results can be summarized as indicaring an overall reduction in social interaction with a significant increase in aggressive behavior. For example, the latency for the first approach of one rat to another was significantly increased in pairs of malnourished rats relative to that observed in wellnourished pairs, whereas the number of attacks was significantly higher. Findings of increased aggression (10,19,31,34,38) and increased social responsiveness (34,37,39) in postnatally malnourished animals, when tested as adults, may result from significant alterations in the early social environment. Unlike postnatal malnutrition, malnutrition limited to the prenatal period has been shown to produce only minimal changes in mother-pup interactions in rats (8,33). However, the effect of such treatment on other aspects of the early social environment (i.e., interactions between pups during infancy and prior to adolescence) has not been examined. Thus, the aim of the present investigation was to examine social play (pins) and other social interactions (walk-overs, allogrooms, side mounts, and anogenital
1 To whom requests for reprints should be addressed.
197
198
ALMEIDA, TONKISS AND GALLER
sniffs) between prenatally malnourished and well-nourished control rats at around the peak age for social play (45 days). Pinning behavior is a well-established measure of social play in rats (15,16), and is known to be significantly affected by prenatal insults [e.g., (41)]. The remaining measures have previously been used to assess social interactions in adult rats, and have been shown to be disrupted by postnatal malnutrition (34,37,39,40). METHOD
Nutritional Treatment Five weeks prior to mating, nulliparous female rats (Sprague-Dawley VAF plus; Charles River Laboratories, Kingston, MA) were obtained and allowed ad lib access to one of two isocaloric diets (Teklad, Madison, WI). The diets were formulated to be of adequate protein (25% casein) or low-protein (6% casein) content [detailed description given in (8)]. Males obtained from the same source were acclimated to the experimental diets of the females for 1 week prior to mating. Each male was introduced into a cage of two to three females receiving the same diet treatment. Mating occurred over a 10-day period. The presence of sperm in a vaginal smear detected whether mating had occurred. One week prior to the projected delivery date the females were caged individually in polycarbonate breeding cages (51 X 41 X 21 cm; Lab Products Inc., Maywood, NJ). Following parturition, litters were culled to eight pups (six males and two females) and fostered as whole litters to well-nourished mothers that had given birth no more than 24 h previously. Pups were selected for fostering based upon their proximity to the litter mean body weight for each sex. Litters born to dams fed the 6% casein diet and cross-fostered to lactating dams given 25% casein diet were designated as 6-25 (prenatally malnourished; n = 8). Litters born to dams fed the 25% casein diet and fostered to other lactating dams fed 25% casein were designated as 25-25 (controis; n = 8). A more detailed description of the nutritional, mating and fostering procedures is given in a previous article (32). The rats were weighed at birth and on the day of behavioral testing using an electronic balance (Ohaus, GT 4000). At weaning (21 days of age) rats of the same sex and nutritional treatment were caged in pairs and given ad lib access to rat chow (PMI foods Inc., formula 5001).
there were eight rats per treatment group all derived from different litters. These rats were housed singly in wire hanging cages until the experiment began at 45 days of age. Single housing was used because social isolation has been reported to increase social play (12,15,16). To familiarize the animals to the experimental conditions, all rats were placed individually in the experimental cage for a 5-min session on each of 5 consecutive days. No data were recorded during these familiarization sessions. Test sessions of 10 rain were then given on each of 3 consecutive days and consisted of placing a pair of rats of the same sex (one 6-25 and one 25-25 rat) in the arena. The rats of a pair were selected so as not to differ in body weight by more than 10 g. For identification purposes one rat in the pair was marked with black ink on the top of the head. The mark was randomly varied according to the dietary treatment of the rat. Videotapes were analyzed by two experimenters blind to the nutritional condition of the rat, and the following variables (34,35) were recorded: Pin: one animal lies on its back while the other stands over and in contact with it. Walk-over: the animal walks over the head or back of another. Allogroom: the licking or nibbling of the fur of one animal by the other. Side mount: one animal approaches the other from the side and places its forepaws on the other's dorsal surface. Anogenital sniff: the sniff of the anogenital region of one rat by the other. Rear: standing on the hind legs in a nonsocial context.
Statistical Analyses The body weights of the nutritional treatment groups were compared by two-way ANOVA (treatment x sex). Separate analyses were applied for the birth weights and the weights at testing because of vastly dissimilar variances at the two ages. Behavioral data were analyzed by a three-way ANOVA (sex X treatment × session) with treatment and session as repeated measures (i.e., the "pair" was used as the unit for analysis allowing two levels of treatment). Post hoc analyses were conducted by Newman-Keuls tests. RESULTS
Body Weight Housing Conditions The animals were maintained in a room with temperature of 73 ___3°F and 45-55% humidity. A reverse 12-h night (0800-2000 h)/12-h day (2000-0800 h) light cycle accommodated behavioral observations to the active waking period of the rat. Continuous dim illumination was provided by red fluorescent lighting during the dark portion of the cycle.
Apparatus The testing arena consisted of an acrylic plastic cage measuring 27 X 25 x 32 cm, with 2 cm of Sani-Chips (P. J. Murphy, Montville, NJ) covering the floor. Two red lights (60 W) provided the only illumination in the experimental room and were mounted 80 cm above the arena. Experimental sessions were recorded by a vertically mounted videocamera (Panasonic, model WV-BL200), linked to a monitor and a video cassette recorder in an adjacent room.
Procedure On postnatal day 33, one male and one female rat per litter were selected at random for behavioral testing. Thus, for each sex
The mean body weights of the two treatment groups at birth and at testing (day 45) are given in Table 1. At birth, the 6-25 rats showed significant body weight deficits compared with well-nourished control rats, F(1, 28) = 6.31, p < 0.02 (15% for males and 7% for females). No effect of sex and no significant treatment X sex interaction were indicated by ANOVA. At testing, body weight differences between the groups had narrowed but the treatment effect remained statistically significant, F(1, 28) = 5.88, p < 0.05. Male rats weighed more than the females,
TABLE 1 MEAN + SEM BODY WEIGHTS Well-Nourished 25-25 Group
Prenatally Malnourished 6-25 Group
Deficit
8) = 8)
5.86 + 0.22 5.55 ::t: 0.22
5.00 + 0.33 5.17 -t- 0.21
15% 7%
8) = 8)
249 5 : 2 186 :t: 4
237 + 4 180:1:4
Age Birth Male ( n = Female (n Testing Male ( n = Female ( n
5% 3%
MALNUTRITION AND SOCIAL BEHAVIOR
E
~
199
,z' 1 5 -
104
•
1
2
10
*
3
~=
0
1
2
3
45
SESSION
SESSION
FIG. 1. Mean+ SEM of pins, walk-overs, allogrooms, and rears for prenatally malnourished (6-25 group, open bars) and well-nourished control rats (25-25 group, filled bars) for each of three sessions. * p < 0.05, **p < 0.01 compared to 25-25 (Newman-Keuls). No significant sex differences were indicated, hence the data have been combined for illustrative purposes.
F(1, 28) = 255.91, p < 0.001, but there was no significant treatment × sex interaction.
Behavioral Measures Interobserver reliabilities were calculated on 12 randomly selected subjects and correlation values (r s) of _> 0.90 were obtained for each of the behaviors measured. Rear had the lowest reliability (r = 0.90) and pin had the highest ( r = 0.96). Pinning. Figure 1 (upper left) illustrates the mean number of times the 6-25 rats pinned the 25-25 rats and vice versa over the three test sessions. Prenatal malnutrition affected pinning behavior as indicated by a significant effect of treatment, F(1, 14) = 9.36, p < 0.01. A treatment by sessions interaction, F(2, 28) = 4.95, p < 0.05, and a main effect of session, F(2, 2 8 ) = 5.15, p < 0.05, were also indicated by the ANOVA. Post hoc analyses showed that there was no significant difference in the number of pins exhibited by 6-25 and 25-25 rats on session 1. However, the incidence of pinning increased in the 25-25 group over the three sessions whereas the incidence of pinning in the 6-25 group remained fairly constant. Consequently, 25-25 rats made significantly more pins than the 6-25 rats on sessions 2 and 3 ( p < 0.05). No significant effect of sex and no significant interactions involving sex were indicated. Walk-over Figure 1 (upper right) illustrates the mean incidence of walk over for 6-25 and 25-25 rats over the three test sessions. Prenatal malnutrition affected walk-over behavior as indicated by a significant effect of treatment, F(1, 14) = 14.27, p < 0.01. A treatment by session interaction, F(2, 2 8 ) = 4.30, p < 0.05, and a main effect of session, F(2, 28) = 3.96, p < 0.05, were also indicated by the ANOVA. Post hoc analysis showed that the pattern of results was strikingly similar to that observed for pinning. Although there was no significant difference in the number of walk-overs exhibited by 6-25 and 25-25 rats on session 1, the incidence of walk-over increased in the 25-25
group whereas the incidence of walk-over in the 6-25 group remained fairly constant over sessions. Consequently, 25-25 rats made significantly more walk-overs than the 6-25 rats on sessions 2 and 3 ( p < 0.01). No significant effect of sex and no significant interactions involving sex were indicated. Allogroom. Figure 1 (lower left) illustrates the mean incidence of allogroom for 6-25 and 25-25 rats over the three test sessions. ANOVA indicated a significant effect of treatment, F(1, 14) = 8.00, p < 0.05, due to the 25-25 rats performing more allogrooming than the 6-25 rats. The incidence of allogrooming decreased across sessions, F(2, 2 8 ) = 5.26, p < 0.05, but no significant treatment by session interaction was indicated. The main effect of sex and interactions involving sex were not significant. Side mount. ANOVA indicated a significant effect of session, F(2, 28) = 4.13, p < 0.05, due to a decrease in the incidence of this behavior across sessions, but there was no treatment effect or treatment by session interaction. Similarly, there was no significant effect of sex and no significant interactions involving sex (data not shown). Anogenital sniff. Anogenital sniff was not affected by any of the factors studied (data not shown). Rear. Figure 1 (lower right) illustrates the mean incidence of rearing in 6-25 and 25-25 rats over the three test sessions. ANOVA showed a significant treatment by session interaction, F(2, 2 8 ) = 4.11, p < 0.03. The main effects of treatment and session were not significant. Similar to pinning and walk-over, the two treatment groups did not differ on session 1. However, over sessions the incidence of rearing increased in malnourished animals and decreased in control animals, leading to significantly greater levels of rearing in the prenatally malnourished rats on sessions 2 and 3 compared with the controls ( p < 0.05). There was no significant effect of sex and no significant interactions involving sex. Relationships between behaviors--factor analysis. The apparent similarity of the findings for pinning and walk-over (Fig. 1) suggested a possible link between these two behaviors. Hence, a principal components factor analysis of the correlations among the six measures was carried out. The data were combined over sessions such that there were 96 observations on the 32 rats. The first principal component (unrotated) had an eigenvalue of 1.86 accounting for 31% of the variance. Correlations between individual behaviors loading on the first factor were: walk-over = 0.773, side mount = 0.756, and allogroom = 0.690, with pin loading only slightly (0.396). Rearing and anogenital sniff were not correlated and rearing loaded on separate subsequent components, confirming the nonsocial nature of this behavior. DISCUSSION
The present study demonstrated that prenatal malnutrition reduced the incidence of various social behaviors (pin, walk-over, and allogroom) relative to that of well-nourished animals. In contrast, nonsocial exploratory behavior (rear) was increased by malnutrition. None of these differences were present on the first test session; rather, they emerged with repeated testing of the same dyads over subsequent test days. In the case of pinning behavior, which has been interpreted as "dominance behavior" (1,15,21,41), these data are consistent with previous reports showing that dominance patterns emerge over repeated pairings of the same individuals (16). Thus, the finding that control animals showed a greater incidence of pins than prenatally malnourished animals, accounting on the average for 70% of total pins occurring by session 3, would suggest that adolescent wellnourished animals were more likely to assume dominance. Grooming of the partner has also been linked to social dominance
200
ALMEIDA, TONKISS AND GALLER
(24). Consequently, the observation of greater allogrooming of the prenatally malnourished rats by the control animals supports the interpretation of greater submissiveness in the malnourished rats. The absence of differences between males and females in the incidence of pinning agrees with some previously reported results (29), but it is in contradiction with other studies in which a lower number of pins was observed in females rats (11,14,18,26,36,41). It is possible that the predominance of males in the culled litters may have influenced the social development of the females prior to weaning. Because males constituted 75% of the litter, male-female interactions would have been more likely to occur than if each sex had been represented in equal numbers. The lower frequency of walk-over and allogroom in malnourished rats suggests that prenatal malnutrition has a significant effect on nonplayful social interactions as well as those considered to be playful. These two behaviors, which were highly correlated in the present study (as shown by factor analysis), were included by Panksepp and Beatty (16) in a general behavioral category that they labeled "social solicitations." If these behaviors do represent an attempt of one animal to elicit a social response from the other, the lower level of these behaviors in prenatally malnourished rats may reflect a decrease in the ability of these animals to initiate and maintain social contact. Interestingly, the direction of these results is exactly opposite to that derived from malnutrition studies in rats or mice in which the period of malnutrition encompassed the suckling period. When tested in adulthood, such previously malnourished animals have been reported as more socially reactive and more aggressive than controls (34,37-40). The only previous study of social interactions in malnourished juvenile rats examined subjects malnourished postnatally and they were malnourished at the time of testing (5). Such concurrently malnourished subjects generally showed lower levels of social interaction but a greater incidence of attacks. Clearly, there are differences between social interactions in adolescence and in adulthood (e.g., in adolescent rats there is a high component of "play-fighting," which contrasts with real fighting in adult rats) [see (17) for a review, (29)]. Consequently, both playful and nonplayful social behaviors are likely to be present in different degrees in juvenile as compared to adult animals. Nonetheless, data support the conclusion that juvenile fighting may predict adult aggressiveness (27), which suggests that at least some of the differences between the findings on agonistic behavior from the present study (pinning) and previous studies (fighting) may lie with the period of malnutrition rather than the difference in the age of testing. Rearing on the hind legs was scored as a nonsocial behavior,
and the independence of this behavior from the others was confirmed by factor analysis. It has been suggested that rearing expresses the vertical component of exploration when it occurs in an empty cage (6), but the significance of rearing in the present context is uncertain. The finding that prenatally malnourished rats developed higher levels of rearing than controls on sessions 2 and 3, with lower levels of social interaction, suggests that rearing may have represented a form of exploration related to escape from the social situation. The presence of greater rearing activity in the prenatally malnourished rats also signifies that differences in social behavior cannot be attributed to generally lower activity in the malnourished animals. Although there was a significant weight advantage to the control animals at the time of testing, it was small (Table 1). Previously reported results have shown that a small body weight difference is not a determining factor in social behavior (30,36,37) and even weight advantages as large as 5-15% do not predict dominance (17). Thus, changes in playful and nonplayful behaviors following prenatal malnutrition are unlikely to be attributed to differences in body weight. In conclusion, if the hypothesis is true that early social behavior is relevant to the development of adult behavior (3,11,15,25), then the well-nourished control rats could be described as more adapted to withstand the constant pressure imposed by their environment in adulthood. However, the extent to which adult behavior can be determined by social behavior early in life remains to be investigated, and additional work related to the effects of malnutrition on both juvenile and later social behavior is needed. The observation that chronically malnourished children exhibit less social involvement, less happy affect, and greater timidity in free play compared with similar children provided with high nutritional supplementation over the first 4 years (2) suggests the importance of such studies. ACKNOWLEDGEMENTS This work was supported by NIH grants HD22539 and DA07934 (J.R.G.), grants 90/3474-0 from Funda~o de Amparo ~ Pesquisa do Estado de S~o Paulo, and 400015/91-9 (S.S.A.) and 530060/93-0 (S.S.A.) from Conselho Nacional de Desenvolvimento Cientifico e Tecnol6gico. S. S. Almeida was the recipient of a Postdoctoral Research Fellowship from Funda~.5ode Amparo h Pesquisa do Estado de S§o Paulo (FAPESP, proc. No. 93/1538-9) and a Research Fellowship from the Conselho Nacional de Desenvolvimento Cientifico e Tecnol6gico (CNPq, proc. No. 30000/91-1). During the preparation of this work S. S. Almeida was a Visiting Research Scholar at the Center for Behavioral Development and Mental Retardation, Boston University School of Medicine. We would like to thank Dr. Robert Harrison for statistical advice and Eva Sabo and Pratibha Singh for technical assistance.
REFERENCES 1. Baenninger, L. P. Comparison of behavioural development in socially isolated and grouped rats. Anim. Behav. 15:312-323; 1967. 2. Barrett, D. E.; Radke-Yarrow, M. Effects of nutritional supplementation on children's responses to novel, frustrating, and competitive situations. Am J. Clin. Nutr. 42:102-120; 1985. 3. Bekoff, M. The development of social interaction, play and rectacommunication in mammals: an ethological perspective. Q. Rev. Biol. 47:412-434; 1972. 4. Calcagnetti, D. J.; Schechter, M. D. Place conditioning reveals the rewarding aspect of social interaction in juvenile rats. Physiol. Behav. 51:667-672; 1992. 5. Fra~kov~, S. Effect of protein-calorie malnutrition on the development of social behavior in rats. Dev. Psychobiol. 6:33-43; 1973. 6. Frafikovfi, S.; Barnes, R. H. Influence of early malnutrition in early life on exploratory behavior of rats. J. Nutr. 96:477-484; 1968. 7. Galler, J. R.; Ricciuti, H. N.; Crawford, M. A.; Kucharski, L. T. The
8. 9. 10. 11. 12. 13.
role of the mother-infant interaction in nutritional disorders. In: Galler, J. R., ed. Nutrition and behavior. New York: Plenum Press; 1984:269-304. Galler, J. R.; Tonkiss, J. Prenatal protein malnutrition and maternal behavior in Sprague-Dawley rats. J. Nutr. 121:762-769; 1991. Humphreys, A. P.; Einon, D. F. Play as a reinforcer for maze-learning in juvenile rats. Anita. Behav. 29:259-270; 1981. Levitsky, D. A.; Barnes, R. H. Nutritional and environmental interactions in the behavioral development of the rat: Long-term effects. Science 176:68-71; 1972. Meaney, M. J.; Stewart, J. A descriptive study of social development in the rat (Rattus norvergicus). Anita. Behav. 29:34-45; 1981. Niesink, R. J. M.; Van Ree, J. M. Involvement of opioid and dopaminergic systems in isolation-induced pinning and social grooming of young rats. Neuropharmacology 28:411-418; 1983. Normansell, L.; Panksepp, J. Effects of morphine and naloxone on
MALNUTRITION AND SOCIAL B E H A V I O R
play-rewarded spatial discrimination in juvenile rats. Dev. Psychobiol. 23:75-83; 1990. 14. Olioff, M.; Stewart, J. Sex differences in the play behavior of prepubescent rats. Physiol. Behav. 20:113-115; 1978. 15~ Panksepp, J. The ontogeny of play in rats. Dev. Psychobiol. 14:327332; 1981. 16. Panksepp, J.; Beatty, W. W. Social deprivation and play in rats. Behav. Neural Biol. 30:197-206; 1980. 17. Panksepp, J.; Siviy, S.; Normansell, L. The psychobiology of play: Theoretical and methodological perspectives. Neurosci. Biobehav. Rev. 8:465-492; 1984. 18. Poole T. B.; Fish, J. An investigation of individual, age, and sexual differences in the play of Rattus norvergicus (Mammalia: Rodentia). Proc. Zool. Soc. Lond. 179:249-260; 1976, 19. Randt, C. T.; Blizard, D. A.; Friedman, E. Early life undernutrition and aggression in two mouse strains. Dev. Psychobiol. 8:275-279; 1975. 20. Rozenzweig, M. R. Effects of environment on development of brain and behavior. In: Tobach, E.; Aronson, L. R.; Shaw, E., eds. The biopsychology of development. New York: Academic Press; 1971:303-342. 21. Small, W. S. Notes on the psychic development of the young white rat. Am. J. Psychol. 11:80-100; 1899. 22. Smart, J. L. Undernutrition and aggression. In: Brain, P. F.; Benton, D., eds. Multidisciplinary approaches to aggression research. Amsterdam: Elsiever/North Holland Biomedical Press; 1981:179-191. 23. Smith, P. K. Does play matter? Functional and evolutionary aspects of animal and human play. Behav. Brain Sci. 5:139-184; 1982. 24. Spigel, I. M.; Trivett, S. Grooming behavior and competitive dominance in the albino rat. J. Comp. Physiol. Psychol. 78:409-411; 1972. 25. Takahashi, L. K. Postweaning environmental and social factors influencing the onset of expression of agonistic behavior in norway rats. Behav. Proc. 12:237-260; 1986. 26. Takahashi, L. K.; Lore, R. K. Play fighting and the development of agonistic behavior in male and female rats. Aggress. Behav. 9:217227; 1983. 27. Taylor, G. T. Fighting in juvenile rats and the ontogeny of agonistic behavior. J. Comp. Physiol. Psychol. 94:953-961; 1980. 28. Thor, D. H.; Flannelly, K. J. Age of intruder and territorial-elicited aggression in male Long-Evans rats. Behav. Biol. 17:237-241; 1976
201
29. Thor, D. H.; Holloway, W. R. Social Play in juvenile rats: A decade of methodological and experimental research. Neurosci. Biobehav. Rev. 8:455-464; 1984. 30. Thor, D. H.; Holloway, W. R. Social play soliciting by male and female juvenile rats: Effects of neonatal androgenization and sex of cagemate. Behav. Neurosci. 100:275-279; 1986. 31. Tikal, K.; Bene~ov~, O.; Frahkovfi, S. The effects of pyrithioxine and pyridoxine on individual behavior, social interactions, and learning in rats malnourished in early postnatal life. Psychopharmacologia 46:325-332; 1976. 32~ Tonkiss, J.; Galler, J. R. Prenatal protein malnutrition and working memory performance in adult rats. Behav. Brain Res. 40:95-107; 1990. 33. Tonkiss, J.; Shumsky, J. S.; Shultz, P. L.; Almeida, S. S.; Galler, J. R. Prenatal cocaine but not prenatal malnutrition affects foster mother-pup interactions in rats. Neurotoxicol. Teratol. 17:601-608; 1995. 34. Tonkiss, J.; Smart, J. L. Interactive effects of genotype and early life undernutrition on the development of behavior in rats. Dev. Psychobiol. 16:287-301; 1983. 35. Vanderschuren, L. J. M. J.; Niesink, R. J. M.; Spruijt, B. M.; Van Ree, J. M. Effects of morphine on diffferent aspects of social play in juvenile rats. Psychopharmacology (Berlin) 117:225-231; 1995. 36. Ward, I. L.; Stehm, K. E. Prenatal stress feminizes juvenile play patterns in male rats. Physiol. Behav. 50:601-605; 1991. 37. Whatson, T. S.; Smart, J. L. Social behavior of rats following pre and postnatal undemutrition. Physiol. Behav. 20:749-753; 1978. 38. Whatson, T. S.; Smart, J. L.; Dobbing, J. Social interactions among adult male rats after early undernutrition. Br. J. Nutr. 32:413-419; 1974. 39. Whatson, T. S.; Smart, J. L.; Dobbing, J. Dominance relationship among previously undernourished and well fed male rats. Physiol. Behav. 14:425-429; 1975. 40. Whatson, T. S.; Smart, J. L.; Dobbing, J. Undernutrition in early life: Lasting effects on activity and social behavior of male and female rats. Dev. Psychobiol. 9:529-538; 1976. 41. Wood, R. D.; Bannoura, M. D.; Johanson, 1. B. Prenatal cocaine exposure: Effects on play behavior in the juvenile rat. Neurotoxicol. Teratol. 16:139-144; 1994.