- Email: [email protected]
Reductions in maternal food and water intake account for prenatal stress effects on neurobehavioral development in B6D2F2 mice
Physiology & Behavior, Vol. 44, pp. 781-786. Copyright©PergamonPress pie, 1988. Printedin the U.S.A.
0031-9384/88$3.00 + .00
Reductions in Maternal Food and Water Intake Account for Prenatal Stress Effects on Neurobehavioral Development in B6D2F2 Mice G. R. W A R D * A N D P. E. W A I N W R I G H T t 1
*Department o f Psychology and t D e p a r t m e n t o f Health Studies University o f Waterloo, Waterloo, Ontario, N 2 L 3G1 Canada R e c e i v e d 24 M a y 1988 WARD, G. R. AND P. E. WAINWRIGHT. Reductions in maternalfood and water intake account for prenatal stress on neurobehavioral development in B6D2F2 mice. PHYSIOL BEHAV 44(6)781-786, 1988.--This study evaluated the role of stress-induced reductions in food intake in pregnant B6D2FI mice in the production of developmental abnormalities in the offspring. One group of dams underwent one hour of physical restraint stress twice daily from days 12 to 17 of gestation. A second group was not restrained but, during this period, each dam was pair-fed to a weight-matched partner in the stressed group. A third group was left undisturbed and allowed unlimited access to lab chow and water. The restraint stress procedure reduced the average daily food and water intake in the dams, resulting in lower maternal weights, with the unstressed pair-fed group being affected more severely than the stressed group. The offspring of both the stressed and pair-fed dams were lighter than the offspring of the dams fed ad lib on day 20 postconception, and exhibited lower brain weights on day 32 postconception. Neurobehavioral development, as assessed on a battery of sensorimotor tests on day 32 postconception, was retarded in the prenatally stressed pups in comparison to the ad lib pups, while the pair-fed pups did not differ from either group. There were no differences among any of the groups on day 50 body or brain weight, body or tail length, anogenital distance, or locomotion, rearing and defecation in the open field. These results suggest that reductions in maternal food intake account for at least some of the deficits commonly attributed to prenatal stress and that, when the stressing procedure is known to reduce food consumption, pair-fed control groups must be included before any interpretation of the nature of prenatal stress effects can be made. Prenatal stress Restraint Mice Anogenital distance
Neurobehavioral development Food and water intake
BOTH prenatal stress and prenatal undernutrition in laboratory rodents are reported to produce abnormalities such as body weight deficits (2, 13, 20), increased incidence of cleft palate (7, 10, 21), increased frequency of lordosis in males (15, 18, 20, 27), and elimination of the sex difference in the size of the SDN-POA in rats (2,3). These similarities have important implications for the field of prenatal stress research, since common experimental stressing procedures such as crowding (5), electric shock (6), random noise (1), physical restraint (6,17), and injections of ACTH and synthetic glucocorticoids (1,22) have all been found to produce significant reductions in food intake. Procedures such as physical restraint usually entail removal of food and water during the treatment period. Kinsley and Svare (17) compared food intake in stressed pregnant mice with that of an unstressed control group which had its food and water re-
Brain weight
Open field
Body length
moved for equivalent periods of time. Their results showed that the food intake of the stressed animals was still significantly reduced relative to that of this food restricted control group. Therefore, the decreased duration of food availability was not the cause of the reduced daffy food and water intake. Rather, as noted by the authors, restraint stress had a prolonged effect on the ingestive behaviors and body weight of the pregnant dams. Therefore, could a decrease in maternal nutritional status account for some of the neurobehavioral defects which are reported to result from prenatal stress? We investigated this hypothesis by applying restraint stress to pregnant mice, and incorporating a pair-feeding procedure to control for the effects of reduced food intake in order to differentiate between those effects which were due directly to prenatal stress and those which were due only to reduced nutritional status in the pregnant dam. Besides the
1Requests for reprints should be addressed to P. E. Wainwright.
781
782
WARD AND W A I N W R I G H T t20 T
I
~
m Ioo
/ 42-
~g
f,
........~' % \
//,/
/~
40-
/
.....~,-'"'
//
~'/"
// /
3m-
~ r
~
..~4"
5O
12
13
14
15
16
17
18
GESTATION DAY 12
FIG. 1. Effect of restraint stress on days 12-17 of gestation on food and water intake in B6D2F1 mice. Values are group means presented as percentage of ad lib consumption. (x) Food intake, ((3) fluid intake.
common morphological measures of brain and body weight, body length and anogenital distance, we included a measure of neurobehavioral development using a test battery previously shown to be sensitive to the retarding effects of prenatal protein malnutrition on behavioral development in this animal model (26), Because prenatal stress is commonly believed to reduce activity in offspring (4), we also included locomotor activity in the open field at weaning. METHOD
Subjects All B6D2F1 mice used in this study were obtained at four weeks of age from Charles River Breeding Laboratories, St. Constant, Quebec. Three to five animals were housed together in groups o f the same sex, with free access to laboratory chow (Purina Mills Inc. No. 5001, St. Louis, MO) and tap water until the time o f mating in this study. All females had given birth to one litter previously and the litters had been weaned from 6 to 10 weeks before the dams were mated again in the present study, at which time they were from 35 to 40 weeks of age. They were maintained under a reversed 12 hour dark:light cycle (dark 0800-2000 hr) in standard opaque plastic mouse cages (29x18x13 cm) with Beta Chip hardwood bedding and several sheets of toilet tissue for nesting material.
Procedure At approximately 0900 hr, a male was placed into each cage of females and removed 6 hours later. At that time, any female exhibiting a vaginal copulatory plug was weighed, placed into a separate cage, and maintained on lab chow and tap water ad lib. The day of mating was referred to as day 0, and all pre- and postnatal timing referred to as " d a y s postconception." On day 12, each dam was weighed again,
13
14
1,,5
16
17
GESTATION DAY
FIG. 2. Effects of restraint stress and pair-feeding on days 12-17 of gestation on maternal weight in B6D2F1 mice. Values are group means; Days 14-17, ST > PF; Days 16--17, AL > ST > PF;p<0.05. (×) Stress, ((3) pair-fed, (E3) ad lib.
transferred to a clean cage, and assigned to a particular treatment group on the basis of body weight on days 0 and 12. The stressed (ST) group was given access to food and water ad lib and underwent two daily one-hour sessions of restraint stress (commencing at 1030 hr and 1730 hr) from day 12 to day 17. The restraint procedure was adapted from Rozenweig and Blaustein (21), and consisted of enclosing the mouse within a folded wire mesh screen (20×20 cm). Half the screen was folded over the mouse and the two halves were stapled together so as to restrict all gross movement. The restrained mouse was placed on four layers of toilet tissue and covered with a folded cloth towel to help prevent heat loss. The temperature beneath the towel but away from the surface of the mouse was found to range from 29.0°C to 29.5°C, while the temperature inside the folded screen, on the body surface of the mouse, was found to range from 34.5°C to 36.5"C after one hour of restraint. We have found (manuscript submitted) that this procedure leads to significant elevations in plasma corticosterone levels in pregnant mice relative to pair-fed controls (1,093+_71 vs. 550+_36 /zg/100 ml, respectively).The pair-fed (PF) group was fed the amount of food and water per gram body weight which their weight-matched ST group member consumed on the corresponding day of gestation, and were left undisturbed during the stressing session except that they had their food and water removed for that hour. The pair-feeding was carried out each day from days 12 to 18, and food and water were provided ad lib on day 19. The ad lib fed (AL) group was provided with food and water ad lib and left undisturbed throughout the day. The food was weighed on a triple beam balance to the nearest 0.1 gram while,the water was measured to the nearest 0.1 ml, and all feeding was done between 0830 hr and 0900 hr. The pregnant dams were weighed on day 12, 14, 16 and 17 andqeft undisturbed until day 20, when
FOOD I N T A K E AND P R E N A T A L STRESS
783
TABLE 1 EFFECTS OF PRENATALRESTRAINTSTRESS AND PAIR-FEEDING ON PREWEANINGPUP DEVELOPMENT* Stress
Pair-Fed
Ad Libitum
Day 20 (Bi~h) 1.62f [0.05l (14)
1.597 [0.04] (13)
1.87, [0.08] (14)
Body Weight(g)
6.95 [0.18] (14)
7.24 [0.19] (13)
7.43 [0.22] (14)
Brain Weight (g)
0.3867f [0.0o45] (14)
0.3854f [o.oo31] (13)
0.4032* [0.o040] (14)
30.69f [0.18] (14)
30.90 [0.21] (13)
31.24, [0.14] (14)
Body Weight (g) Day 32
Developmental Age (days)
*Data are presented as means with [SEM], and (n)=number of litters. Means with different superscripts are significantly different (p<0.05).
the pups were weighed and transferred with their mothers to clean cages. The litters were culled on day 20 to eight pups (four males and four females where possible). On day 32, the pups were again weighed and two males and two females were selected randomly to undergo assessment of behavioral development. The test battery used was developed by Wahlsten (24,25) and has been described in detail in an earlier paper (26). Briefly, each pup was tested on twelve tests measuring such developmental landmarks as righting, cliff aversion, grasping, climbing, visual placing, eye opening, and auditory startle. The mean score for all tests was entered into a regression equation, derived previously from the scores of untreated B6D2F2 offspring, and which predicted age from behavioral score, thereby allowing quantitative assessment of treatment effects in days. After the behavioral testing, one pup of each sex was given an overdose of sodium pentobarbital and peffused intracardially with 1(1% buffered formalin. The brain was then extracted and trimmed by severing the brain stem at the level of the loons and removing the olfactory bulbs. The brain was stored in 10?b buffered formalin and weighed two weeks later. On day 50, each pup was again weighed and one male and one female were selected randomly for measures of open field activity. The enclosure measured 50x50 cm and the floor was divided into squares 5x5 cm. Activity was measured by the number of squares crossed during a 4 minute period, and the number of times the mouse reared on its hind legs was recorded during this same period, as was the number of fecal boll Before each animal was tested, the bottom and sides of the enclosure were wiped clean with alcohol and a clean sheet of white paper was placed on the bottom. Open field testing was done under red light, and all testing throughout the experiment was Conducted with the experimenter unaware of the groups to which the mice belonged. After open field testing, the two mice were killed and their brains processed as described above.
Data Analysis The data were analysed using the general linear model (GLM) provided by the Statistical Analysis System (SAS) to do analysis of variance (ANOVA) and covariance (ANCOVA). A repeated measures analysis was done on the data of maternal food and water intake and body weight. Individual group comparisons were carried out using t-tests, and the litter mean score was used as the unit of analysis for all offspring data. Preliminary analysis of the data included sex as a variable and, in the absence of a significant sex by treatment interaction, further analyses were conducted on the data collapsed across sex. Alpha was set at 0.05. RESULTS
Maternal Variables Food and water intake. Figure 1 shows the daily food and water intake in the ST dams, expressed as a percentage of that of the AL dams, from days 12 to 18 of gestation. Repeated measures analysis showed that the amount consumed increased in both groups over days of gestation, F(6,156)=22.28, p<0.0001, and F(6,154)=8.61, p<0.0001, for food and water, respectively. The stress treatment reduced both food, F(1,26)=62.14, p<0.0001, and water consumption, F(1,26)=11.01, p<0.005, and individual group comparisons for each day of treatment showed the effect to be significant throughout the first 5 days of the treatment for food comsumption, and throughout the first 4 days for water consumption. While water consumption did not differ between the groups after this time, food consumption in the ST dams increased on the 7th day (the day after the cessation of the stress treatment). This treatment by day interaction was significant for both food, F(6,156)= 13.84, p<0.0001, and water, F(6,154)=4.53, p<0.0005, consumption. Body weight. The data on maternal body weight from gestation days 12 to 17 are shown in Fig. 2. The data confirm that the groups were matched on body weight on day 12. Repeated measures analysis of variance showed that all groups gained weight throughout the treatment period, F(3,116)=323.72, p<0.0001. There were significant group differences over the days of treatment, F(2,39)=150.86, p<0.0001, and individual analyses on each day showed significant group differences on every day on which the dams were weighed after day 12 (i.e., days 14, 16 and 17). The group by day interaction was also significant, F(6,116)=22.51, p<0.0001, and individual group comparisons showed the PF dams to be lighter than both the ST and the AL dams on all three days, while the ST dams were significantly lighter than the A L dams on days 16 and 17.
Pup Variables The data on preweaning animals are presented in Table 1 and those at weaning (day 50) in Table 2. Body weight. As shown in Fig. 3, there were significant group differences on day 20 body weight, F(2,38)=6.41, p<0.004, and, while the ST and PF pups did not differ from each other, both were lighter than the AL pups. There were no differences in body weight among any of the groups on day 32 or day 50. Brain weight. As shown in Fig. 4, there were significant group differences on day 32 brain weight, F(2,38)=6.29, p <0.004, with both the ST and PF pups having lighter brains than those of the AL pups. Once again, the ST and PF groups did not differ from each other. There were no significant differences in brain weight among the groups on day 50.
784
WARD AND W A I N W R I G H T
TABLE 2 EFFECTS OF PRENATAL RESTRAINT STRESS AND PAIR-FEED|NG ON PUP DEVELOPMENT AT WEANING* Stress
Pair-Fed
Ad Lib
Body Weight (g)
18.41 [0.421 (14)
18.80 [0.53] (13)
18.95 [0.35] (14)
Brain Weight (g)
0.4289 [0.0034] (14)
0.4295 [0.0048] (13)
0.4372 [0.0061] (14)
Females
80.76 [0.96] (12)
80.83 [1.44] (12)
80.82 [0.89] (11)
Males
82.87 [1.16] (15)
83.94 [1.30] (12)
84.38 [0.95] (13)
Body Length (mm)
....
a..¢~
e$
STRrSS
Anogenital Distance (ram)
PA|R--FrD
A D LIB
TREATMENT
Females
5.16 [0.12] (12)
5.36 [0.20] (12)
4.98 [0.11] (11)
Males
11.80 [0.34] (15)
12.30 [0.38] (12)
12.09 [0.20] (13)
91.54 [10.33] (13)
81.82 [11.83] (11)
76.54 [10.30] (12)
Open Field Rearing
19.96 [2.68] (13)
22.91 [3.93] (11)
20.00 [3.21] (12)
Defecation Score (number of boli)
2.31 [0.48] (13)
1.55 [0.57] (11)
1.67 [0.53] (12)
Open Field Locomotion
•-o1
*Data are collapsed across sex except where noted and presented as means with [SEM], and (n)=number of litters.
Behavioral development. Figure 5 shows the data on day 32 behavioral development. Although the overall analysis of variance was not significant, F(2,38)=2.41, p<0.131, the individual group comparisons showed that the ST offspring were retarded by 0.5 days relative to the A L offspring (p <0.05), but neither differed from the PF offspring. Open field behavior and morphology on day 50. Table 2 shows the data on activity, rearing and the number of boli in the open field, and body length and anogenital distance on day 50. As expected, anogenital distance, covaried for body length, was larger in males than in females, but there were no significant group differences on any of these variables. DISCUSSION The results of this study confirm those of Kinsley and Svare (17), in that restraint stress in pregnant mice reduced their food and water intake and body weight gain. In addition, the data show that the degree of undernutrition resulting from such reductions was sufficient to produce deficits
FIG. 3. Effect of prenatal restraint stress and pair-feeding on pup body weight on day 20 postconception. Values are group means+ SEM; ST = PF < AL, p<0.05.
on offspring birth weight and on brain weight on day 32 postconception. In fact, the undernutrition caused by the reduced intake was the only condition associated with effects on the offspring. Although prenatally stressed offspring did show retarded behavioral development relative to the controls fed ad lib, they did not differ from the pair-fed controls on any measure in this experiment. The consistency of the reductions in food intake observed in this study with those reported by Kinsley and Svare is interesting in light of the methodological differences between the two experiments. The present study used restraint stress only for two daily 60-minute periods, while the earlier study supplemented restraint with concurrent heat and illumination for three daily 30-minute periods. Furthermore, the stressing procedure was c a r d e d out during the dark period in this study as opposed to the light period in the prior experiment. Therefore, although mice exhibit different patterns of feeding in the light and dark period, stress applied during either time appears to affect ingestive behavior, suggesting that this is a relatively robust effect. Nevertheless, the magnitude of the effect did differ between the studies, being less pronounced in this study compared to the earlier one. While Kinsley and Svare found food intake to be reduced by more than 50% for the first three days of the treatment, the reductions in this study ranged from 25% to less than 40% over the first three days. This discrepancy may be due to some of the methodological differences described above. However, it should be noted that the degree of undernutrition in this study was still sufficient to produce significant decreases in body weight gain in the pregnant dams. Interestingly, although both undernourished groups consumed similar amounts of food and water each day, the stressed clams were able to maintain their body weights during the first two days of the treatment whereas the unstressed, pair-fed dams actually lost weight during the same period of time. We have
FOOD INTAKE AND PRENATAL STRESS
785 IBI . l '
O
|.-&
:ii!::~i:!:i
!!i~i!ii!ii! ::i i::?-
i:! : - -
. - - m STRESS PAIR--FED
:i __2
AD UB
TREATMENT
W~ v
I
IO.II - - " m l m m ~ - STRESS
~ w PAIR--FED
. " AD LIB
'
I
TREATMENT
FIG. 4. Effect of prenatal restraint stress and pair-feeding on pup
FIG. 5. Effect of prenatal restraint stress and pair-feeding on pup
brain weight on day 32 postconception. Values are group means+SEM; ST = PF < AL, p<0.05.
behavioral developmenton day 32 postconception. Values are group means+SEM; ST < AL, p<0.05.
found that the stressing procedure alters the daily pattern of consumption of a liquid diet (manuscript submitted) such that the animals feed for longer periods throughout the day. As has been suggested by other authors (6), stress could also be causing metabolic changes in the animals. This should be addressed by future work. The decreased pup body weight on day 20 in both the ST and PF groups in this study is consistent with findings of studies of prenatally stressed end-term-fetuses (29) or newborn pups (2, 12, 16, 17, 20, 23). The fact that these pups exhibited catch-up growth is also consistent with a number of other reports (11, 15, 19). On the other hand, very few studies have reported effects on brain weight following prenatal stress. One study (28) has reported decreased brain weight in rats at 14 and 21 days after birth, but not at 8 and 40 days. Another study (3) reported no differences in brain weight in prenatally-stressed rats at 175 days of age. The results obtained in the current study suggest that prenatal stress, acting by way of reduced nutritional status, may lead to decreased brain weight in the early postnatal period but not at later ages. Several studies (8, 9, 14) have reported delayed behavioral and motor development following prenatal stress. Had the experimental design in the present study been similar to that used in these studies, the fact that the ST pups were significantly retarded in relation to the AL pups would have lent support to the earlier findings. However, the fact that the ST pups did not differ from the PF pups on this measure suggests that most, if not all, of the present findings were due to stress-induced maternal undernutrition. The fact that the growth deficiency present at birth was not apparent at later ages suggests that lactation was not seriously impaired in the previously treated dams. Since fos-
tering was not carded out in this study, the relative contribution of pre- and postnatal factors on development cannot be assessed. However, we have conducted further studies (manuscript submitted) which extend some of the findings reported here to offspring which were fostered after birth to untreated dams. These data support the current work in that undernutrition resulted in lighter brains on day 32, while there was no effect of prenatal stress. Behavioral development was not affected by either treatment in these fostered pups. In summary, these f'mdings suggest that the effects on brain and behavioral development in offspring which are generally attributed to prenatal stress may be, to a great extent, mediated by undernutrition due to stress-induced changes in the ingestive behavior of the pregnant dams. Furthermore, they imply that undernutrition should be taken into account in any attempt to identify the actual mechanism of prenatal stress effects, particularly if it has been shown that the stressing procedure reduced food intake in pregnant dams. Certainly, if hormonal changes are hypothesized to play a major role, it is important that the investigators show, through the inclusion of a pair-fed control group, that the effects cannot be explained solely by stressinduced maternal undernutrition. ACKNOWLEDGEMENTS The authors thank Dawn McCutcheon of the Health Studies Department for her substantial contribution in the conduct of this research. Kathryn Biota of the Psychology Department also provided technical assistance and Doris Winfield made helpful comments on the writing of this manuscript. This research was supported by a Natural Science and Engineering Council of Canada Grant A7617 awarded to P. Wainwright.
786
WARD AND WAINWRIGHT REFERENCES
1. Alario, P.; Gamallo, A. ; Beato, M. J.; Trancho, G. Body weight gain, food intake and adrenal development in chronic noise stressed rats. Physiol. Behav. 40:29-32; 1987. 2. Anderson, D. K.; Rhees, R. W.; Fleming, D. E. Effects of prenatal stress on differentiation of the sexually dimorphic nucleus of the preoptic area (SDN-POA) of the rat brain. Brain Res. 332:113-118; 1985. 3. Anderson, D. K.; Fleming, D. E.; Rhees, R. W.; Kinghorn, E. Relationships between sexual activity, plasma testosterone, and the volume of the sexually dimorphic nucleus of the preoptic area in prenatally stressed and non-stressed rats. Brain Res. 370:1-10; 1986. 4. Archer, J. E.; Blackman, D. C. Prenatal psychological stress and offspring behavior in rats and mice. Dev. Psychobiol. 4:193-248; 1971. 5. Armario, A.; Ortiz, R.; Balasch, J. Effect of crowding on some physiological and behavioral variables in adult male rats. Physiol. Behav. 32:35--37; 1984. 6. Armario, A.; Hildago, J.; Giralt, M. Evidence that the pituitary-adrenal axis does not cross-adapt to stressors: Comparison to other physiological variables. Neuroendocrinology 47:263-267; 1988. 7. Barlow, S. M.; McElhattan, P. R.; Sullivan, F. M. The relation between maternal restraint and food deprivation, plasma corticosterone, and induction of cleft palate in the offspring of mice. Teratology 12:97-104; 1975. 8. Barlow, S. M.; Knight, A. F.; Sullivan, F. M. Delay in postnatal growth and development of offspring produced by maternal restraint stress during pregnancy in the rat. Teratology 18:211218; 1978. 9. Barlow, S. M.; Knight, A. F.; Sullivan, F. M. Prevention by diazepam of adverse effects of maternal restraint stress on postnatal development and learning in the rat Teratology 19:105-110; 1979. 10. Barlow, S. M.; Knight, A. F.; Sullivan, F. M. Diazepaminduced cleft palate in the mouse: The role of endogenous maternal corticosterone. Teratology 21:149-155; 1980. 11. Beckhardt, S.; Ward, I. L. Reproductive functioning in the prenatally stressed femal rat. Dev. Psychobiol. 16:111-118; 1983. 12. Dahlof, L. ; Hard, E.; Larsson, K. Influence of maternal stress on the development of the fetal genital system. Physiol. Behav. 20:193-195; 1978. 13. Dunlap, J. L.; Zadina, J. E.; Gougis, G. Prenatal stress interacts with prepubertal social isolation to reduce male copulatory behaviour. Physiol. Behav. 21:873-875; 1978.
14. Fride, E.; Soreq, H.; Weinstock, M. Are the effects of gestational stress on motor development and cerebellar cholinesterase activity mediated prenatally? Int. J. Dev. Neurosci. 4:407-413; 1986. 15. Herrenkohl, L. R.; Whitney, J. B. Effects of prepartal stress on postpartal nursing behavior, litter development and adult sexual behavior. Physiol. Behav. 17:1019-1021; 1976. 16. Joffe, J. M. Modification of prenatal stress effects in rats by dexamethasone and adrenocorticotrophin. Physiol. Behav. 19:601-606; 1977. 17. Kinsley, C.; Svare, B. Prenatal stress effects: Are they mediated by reductions in maternal food and water intake and body weight gain? Physiol. Behav. 37:191-193; 1986. 18. Meisel, R. L.; Dohanich, G. P.; Ward, I. L. Effects of prenatal stress on avoidance acquisition, open-field performance and lordotic behavior in male rats. Physiol. Behav. 22:527-530: 1979. 19. Peters, D. A. V. Prenatal stress: Effects on biogenic amine and plasma corticosterone levels. Pharmacol. Biochem. Behav. 17:721-725; 1982. 20. Rhees, R. W.; Fleming, D. E. Effects of malnutrition, maternal stress, or ACTH injections during pregnancy on sexual behavior of male offspring. Physiol. Behav. 27:879-882; 1981. 21. Rosenzweig, S.; Blaustein, F. M. Cleft palate in A]J mice resulting from restraint and deprivation of food and water. Teratology 3:47-52; 1970. 22. Simpson, C. W.; Dicara, L. V.; Wolf, G. Glucocorticoid anorexia in rats. Pharmacol. Biochem. Behav. 2:19-25; 1974. 23. Smith, D. J.; Joffe, J. M.; Heseltine, G. F. D. Modification of prenatal stress effects by adrenalectomy, dexamethasone and chlorpromazine. Physiol. Behav. 15:461-469; 1975. 24. Wahlsten, D. A. A developmental time scale for postnatal changes in brain and behavior on B6D2F~ mice. Brain Res. 72:251-264; 1974. 25. Wahlsten, D. Genetic variation in the development of mouse brain and behavior: Evidence from the middle postnatal period. Dev. Psychobiol. 8:371-380; 1975. 26. Wainwright, P.; Russell, V. A quantitative assessment of the effects of malnutrition on behavioral development in laboratory mice. Behav. Neural Biol. 39:140-144; 1983. 27. Ward, I. L. Prenatal stress feminizes and demasculinizes the behavior of males. Science 175:82-84; 1972. 28. Wiggins, R. C.; Gottesfeld, Z. Restraint stress during late pregnancy in rats elicits early hypermyelination in the offspring. Metab. Brain Dis. 1:197-203; 1986. 29. Wilke, D. L.; Tseu, S. R.; Rhees, R. W.; Fleming, D. E. Effects of environmental stress or ACTH treatment during pregnancy on maternal and fetal plasma androstenedione in the rat. Horm. Behav. 16:293-303; 1982.
0031-9384/88$3.00 + .00
Reductions in Maternal Food and Water Intake Account for Prenatal Stress Effects on Neurobehavioral Development in B6D2F2 Mice G. R. W A R D * A N D P. E. W A I N W R I G H T t 1
*Department o f Psychology and t D e p a r t m e n t o f Health Studies University o f Waterloo, Waterloo, Ontario, N 2 L 3G1 Canada R e c e i v e d 24 M a y 1988 WARD, G. R. AND P. E. WAINWRIGHT. Reductions in maternalfood and water intake account for prenatal stress on neurobehavioral development in B6D2F2 mice. PHYSIOL BEHAV 44(6)781-786, 1988.--This study evaluated the role of stress-induced reductions in food intake in pregnant B6D2FI mice in the production of developmental abnormalities in the offspring. One group of dams underwent one hour of physical restraint stress twice daily from days 12 to 17 of gestation. A second group was not restrained but, during this period, each dam was pair-fed to a weight-matched partner in the stressed group. A third group was left undisturbed and allowed unlimited access to lab chow and water. The restraint stress procedure reduced the average daily food and water intake in the dams, resulting in lower maternal weights, with the unstressed pair-fed group being affected more severely than the stressed group. The offspring of both the stressed and pair-fed dams were lighter than the offspring of the dams fed ad lib on day 20 postconception, and exhibited lower brain weights on day 32 postconception. Neurobehavioral development, as assessed on a battery of sensorimotor tests on day 32 postconception, was retarded in the prenatally stressed pups in comparison to the ad lib pups, while the pair-fed pups did not differ from either group. There were no differences among any of the groups on day 50 body or brain weight, body or tail length, anogenital distance, or locomotion, rearing and defecation in the open field. These results suggest that reductions in maternal food intake account for at least some of the deficits commonly attributed to prenatal stress and that, when the stressing procedure is known to reduce food consumption, pair-fed control groups must be included before any interpretation of the nature of prenatal stress effects can be made. Prenatal stress Restraint Mice Anogenital distance
Neurobehavioral development Food and water intake
BOTH prenatal stress and prenatal undernutrition in laboratory rodents are reported to produce abnormalities such as body weight deficits (2, 13, 20), increased incidence of cleft palate (7, 10, 21), increased frequency of lordosis in males (15, 18, 20, 27), and elimination of the sex difference in the size of the SDN-POA in rats (2,3). These similarities have important implications for the field of prenatal stress research, since common experimental stressing procedures such as crowding (5), electric shock (6), random noise (1), physical restraint (6,17), and injections of ACTH and synthetic glucocorticoids (1,22) have all been found to produce significant reductions in food intake. Procedures such as physical restraint usually entail removal of food and water during the treatment period. Kinsley and Svare (17) compared food intake in stressed pregnant mice with that of an unstressed control group which had its food and water re-
Brain weight
Open field
Body length
moved for equivalent periods of time. Their results showed that the food intake of the stressed animals was still significantly reduced relative to that of this food restricted control group. Therefore, the decreased duration of food availability was not the cause of the reduced daffy food and water intake. Rather, as noted by the authors, restraint stress had a prolonged effect on the ingestive behaviors and body weight of the pregnant dams. Therefore, could a decrease in maternal nutritional status account for some of the neurobehavioral defects which are reported to result from prenatal stress? We investigated this hypothesis by applying restraint stress to pregnant mice, and incorporating a pair-feeding procedure to control for the effects of reduced food intake in order to differentiate between those effects which were due directly to prenatal stress and those which were due only to reduced nutritional status in the pregnant dam. Besides the
1Requests for reprints should be addressed to P. E. Wainwright.
781
782
WARD AND W A I N W R I G H T t20 T
I
~
m Ioo
/ 42-
~g
f,
........~' % \
//,/
/~
40-
/
.....~,-'"'
//
~'/"
// /
3m-
~ r
~
..~4"
5O
12
13
14
15
16
17
18
GESTATION DAY 12
FIG. 1. Effect of restraint stress on days 12-17 of gestation on food and water intake in B6D2F1 mice. Values are group means presented as percentage of ad lib consumption. (x) Food intake, ((3) fluid intake.
common morphological measures of brain and body weight, body length and anogenital distance, we included a measure of neurobehavioral development using a test battery previously shown to be sensitive to the retarding effects of prenatal protein malnutrition on behavioral development in this animal model (26), Because prenatal stress is commonly believed to reduce activity in offspring (4), we also included locomotor activity in the open field at weaning. METHOD
Subjects All B6D2F1 mice used in this study were obtained at four weeks of age from Charles River Breeding Laboratories, St. Constant, Quebec. Three to five animals were housed together in groups o f the same sex, with free access to laboratory chow (Purina Mills Inc. No. 5001, St. Louis, MO) and tap water until the time o f mating in this study. All females had given birth to one litter previously and the litters had been weaned from 6 to 10 weeks before the dams were mated again in the present study, at which time they were from 35 to 40 weeks of age. They were maintained under a reversed 12 hour dark:light cycle (dark 0800-2000 hr) in standard opaque plastic mouse cages (29x18x13 cm) with Beta Chip hardwood bedding and several sheets of toilet tissue for nesting material.
Procedure At approximately 0900 hr, a male was placed into each cage of females and removed 6 hours later. At that time, any female exhibiting a vaginal copulatory plug was weighed, placed into a separate cage, and maintained on lab chow and tap water ad lib. The day of mating was referred to as day 0, and all pre- and postnatal timing referred to as " d a y s postconception." On day 12, each dam was weighed again,
13
14
1,,5
16
17
GESTATION DAY
FIG. 2. Effects of restraint stress and pair-feeding on days 12-17 of gestation on maternal weight in B6D2F1 mice. Values are group means; Days 14-17, ST > PF; Days 16--17, AL > ST > PF;p<0.05. (×) Stress, ((3) pair-fed, (E3) ad lib.
transferred to a clean cage, and assigned to a particular treatment group on the basis of body weight on days 0 and 12. The stressed (ST) group was given access to food and water ad lib and underwent two daily one-hour sessions of restraint stress (commencing at 1030 hr and 1730 hr) from day 12 to day 17. The restraint procedure was adapted from Rozenweig and Blaustein (21), and consisted of enclosing the mouse within a folded wire mesh screen (20×20 cm). Half the screen was folded over the mouse and the two halves were stapled together so as to restrict all gross movement. The restrained mouse was placed on four layers of toilet tissue and covered with a folded cloth towel to help prevent heat loss. The temperature beneath the towel but away from the surface of the mouse was found to range from 29.0°C to 29.5°C, while the temperature inside the folded screen, on the body surface of the mouse, was found to range from 34.5°C to 36.5"C after one hour of restraint. We have found (manuscript submitted) that this procedure leads to significant elevations in plasma corticosterone levels in pregnant mice relative to pair-fed controls (1,093+_71 vs. 550+_36 /zg/100 ml, respectively).The pair-fed (PF) group was fed the amount of food and water per gram body weight which their weight-matched ST group member consumed on the corresponding day of gestation, and were left undisturbed during the stressing session except that they had their food and water removed for that hour. The pair-feeding was carried out each day from days 12 to 18, and food and water were provided ad lib on day 19. The ad lib fed (AL) group was provided with food and water ad lib and left undisturbed throughout the day. The food was weighed on a triple beam balance to the nearest 0.1 gram while,the water was measured to the nearest 0.1 ml, and all feeding was done between 0830 hr and 0900 hr. The pregnant dams were weighed on day 12, 14, 16 and 17 andqeft undisturbed until day 20, when
FOOD I N T A K E AND P R E N A T A L STRESS
783
TABLE 1 EFFECTS OF PRENATALRESTRAINTSTRESS AND PAIR-FEEDING ON PREWEANINGPUP DEVELOPMENT* Stress
Pair-Fed
Ad Libitum
Day 20 (Bi~h) 1.62f [0.05l (14)
1.597 [0.04] (13)
1.87, [0.08] (14)
Body Weight(g)
6.95 [0.18] (14)
7.24 [0.19] (13)
7.43 [0.22] (14)
Brain Weight (g)
0.3867f [0.0o45] (14)
0.3854f [o.oo31] (13)
0.4032* [0.o040] (14)
30.69f [0.18] (14)
30.90 [0.21] (13)
31.24, [0.14] (14)
Body Weight (g) Day 32
Developmental Age (days)
*Data are presented as means with [SEM], and (n)=number of litters. Means with different superscripts are significantly different (p<0.05).
the pups were weighed and transferred with their mothers to clean cages. The litters were culled on day 20 to eight pups (four males and four females where possible). On day 32, the pups were again weighed and two males and two females were selected randomly to undergo assessment of behavioral development. The test battery used was developed by Wahlsten (24,25) and has been described in detail in an earlier paper (26). Briefly, each pup was tested on twelve tests measuring such developmental landmarks as righting, cliff aversion, grasping, climbing, visual placing, eye opening, and auditory startle. The mean score for all tests was entered into a regression equation, derived previously from the scores of untreated B6D2F2 offspring, and which predicted age from behavioral score, thereby allowing quantitative assessment of treatment effects in days. After the behavioral testing, one pup of each sex was given an overdose of sodium pentobarbital and peffused intracardially with 1(1% buffered formalin. The brain was then extracted and trimmed by severing the brain stem at the level of the loons and removing the olfactory bulbs. The brain was stored in 10?b buffered formalin and weighed two weeks later. On day 50, each pup was again weighed and one male and one female were selected randomly for measures of open field activity. The enclosure measured 50x50 cm and the floor was divided into squares 5x5 cm. Activity was measured by the number of squares crossed during a 4 minute period, and the number of times the mouse reared on its hind legs was recorded during this same period, as was the number of fecal boll Before each animal was tested, the bottom and sides of the enclosure were wiped clean with alcohol and a clean sheet of white paper was placed on the bottom. Open field testing was done under red light, and all testing throughout the experiment was Conducted with the experimenter unaware of the groups to which the mice belonged. After open field testing, the two mice were killed and their brains processed as described above.
Data Analysis The data were analysed using the general linear model (GLM) provided by the Statistical Analysis System (SAS) to do analysis of variance (ANOVA) and covariance (ANCOVA). A repeated measures analysis was done on the data of maternal food and water intake and body weight. Individual group comparisons were carried out using t-tests, and the litter mean score was used as the unit of analysis for all offspring data. Preliminary analysis of the data included sex as a variable and, in the absence of a significant sex by treatment interaction, further analyses were conducted on the data collapsed across sex. Alpha was set at 0.05. RESULTS
Maternal Variables Food and water intake. Figure 1 shows the daily food and water intake in the ST dams, expressed as a percentage of that of the AL dams, from days 12 to 18 of gestation. Repeated measures analysis showed that the amount consumed increased in both groups over days of gestation, F(6,156)=22.28, p<0.0001, and F(6,154)=8.61, p<0.0001, for food and water, respectively. The stress treatment reduced both food, F(1,26)=62.14, p<0.0001, and water consumption, F(1,26)=11.01, p<0.005, and individual group comparisons for each day of treatment showed the effect to be significant throughout the first 5 days of the treatment for food comsumption, and throughout the first 4 days for water consumption. While water consumption did not differ between the groups after this time, food consumption in the ST dams increased on the 7th day (the day after the cessation of the stress treatment). This treatment by day interaction was significant for both food, F(6,156)= 13.84, p<0.0001, and water, F(6,154)=4.53, p<0.0005, consumption. Body weight. The data on maternal body weight from gestation days 12 to 17 are shown in Fig. 2. The data confirm that the groups were matched on body weight on day 12. Repeated measures analysis of variance showed that all groups gained weight throughout the treatment period, F(3,116)=323.72, p<0.0001. There were significant group differences over the days of treatment, F(2,39)=150.86, p<0.0001, and individual analyses on each day showed significant group differences on every day on which the dams were weighed after day 12 (i.e., days 14, 16 and 17). The group by day interaction was also significant, F(6,116)=22.51, p<0.0001, and individual group comparisons showed the PF dams to be lighter than both the ST and the AL dams on all three days, while the ST dams were significantly lighter than the A L dams on days 16 and 17.
Pup Variables The data on preweaning animals are presented in Table 1 and those at weaning (day 50) in Table 2. Body weight. As shown in Fig. 3, there were significant group differences on day 20 body weight, F(2,38)=6.41, p<0.004, and, while the ST and PF pups did not differ from each other, both were lighter than the AL pups. There were no differences in body weight among any of the groups on day 32 or day 50. Brain weight. As shown in Fig. 4, there were significant group differences on day 32 brain weight, F(2,38)=6.29, p <0.004, with both the ST and PF pups having lighter brains than those of the AL pups. Once again, the ST and PF groups did not differ from each other. There were no significant differences in brain weight among the groups on day 50.
784
WARD AND W A I N W R I G H T
TABLE 2 EFFECTS OF PRENATAL RESTRAINT STRESS AND PAIR-FEED|NG ON PUP DEVELOPMENT AT WEANING* Stress
Pair-Fed
Ad Lib
Body Weight (g)
18.41 [0.421 (14)
18.80 [0.53] (13)
18.95 [0.35] (14)
Brain Weight (g)
0.4289 [0.0034] (14)
0.4295 [0.0048] (13)
0.4372 [0.0061] (14)
Females
80.76 [0.96] (12)
80.83 [1.44] (12)
80.82 [0.89] (11)
Males
82.87 [1.16] (15)
83.94 [1.30] (12)
84.38 [0.95] (13)
Body Length (mm)
....
a..¢~
e$
STRrSS
Anogenital Distance (ram)
PA|R--FrD
A D LIB
TREATMENT
Females
5.16 [0.12] (12)
5.36 [0.20] (12)
4.98 [0.11] (11)
Males
11.80 [0.34] (15)
12.30 [0.38] (12)
12.09 [0.20] (13)
91.54 [10.33] (13)
81.82 [11.83] (11)
76.54 [10.30] (12)
Open Field Rearing
19.96 [2.68] (13)
22.91 [3.93] (11)
20.00 [3.21] (12)
Defecation Score (number of boli)
2.31 [0.48] (13)
1.55 [0.57] (11)
1.67 [0.53] (12)
Open Field Locomotion
•-o1
*Data are collapsed across sex except where noted and presented as means with [SEM], and (n)=number of litters.
Behavioral development. Figure 5 shows the data on day 32 behavioral development. Although the overall analysis of variance was not significant, F(2,38)=2.41, p<0.131, the individual group comparisons showed that the ST offspring were retarded by 0.5 days relative to the A L offspring (p <0.05), but neither differed from the PF offspring. Open field behavior and morphology on day 50. Table 2 shows the data on activity, rearing and the number of boli in the open field, and body length and anogenital distance on day 50. As expected, anogenital distance, covaried for body length, was larger in males than in females, but there were no significant group differences on any of these variables. DISCUSSION The results of this study confirm those of Kinsley and Svare (17), in that restraint stress in pregnant mice reduced their food and water intake and body weight gain. In addition, the data show that the degree of undernutrition resulting from such reductions was sufficient to produce deficits
FIG. 3. Effect of prenatal restraint stress and pair-feeding on pup body weight on day 20 postconception. Values are group means+ SEM; ST = PF < AL, p<0.05.
on offspring birth weight and on brain weight on day 32 postconception. In fact, the undernutrition caused by the reduced intake was the only condition associated with effects on the offspring. Although prenatally stressed offspring did show retarded behavioral development relative to the controls fed ad lib, they did not differ from the pair-fed controls on any measure in this experiment. The consistency of the reductions in food intake observed in this study with those reported by Kinsley and Svare is interesting in light of the methodological differences between the two experiments. The present study used restraint stress only for two daily 60-minute periods, while the earlier study supplemented restraint with concurrent heat and illumination for three daily 30-minute periods. Furthermore, the stressing procedure was c a r d e d out during the dark period in this study as opposed to the light period in the prior experiment. Therefore, although mice exhibit different patterns of feeding in the light and dark period, stress applied during either time appears to affect ingestive behavior, suggesting that this is a relatively robust effect. Nevertheless, the magnitude of the effect did differ between the studies, being less pronounced in this study compared to the earlier one. While Kinsley and Svare found food intake to be reduced by more than 50% for the first three days of the treatment, the reductions in this study ranged from 25% to less than 40% over the first three days. This discrepancy may be due to some of the methodological differences described above. However, it should be noted that the degree of undernutrition in this study was still sufficient to produce significant decreases in body weight gain in the pregnant dams. Interestingly, although both undernourished groups consumed similar amounts of food and water each day, the stressed clams were able to maintain their body weights during the first two days of the treatment whereas the unstressed, pair-fed dams actually lost weight during the same period of time. We have
FOOD INTAKE AND PRENATAL STRESS
785 IBI . l '
O
|.-&
:ii!::~i:!:i
!!i~i!ii!ii! ::i i::?-
i:! : - -
. - - m STRESS PAIR--FED
:i __2
AD UB
TREATMENT
W~ v
I
IO.II - - " m l m m ~ - STRESS
~ w PAIR--FED
. " AD LIB
'
I
TREATMENT
FIG. 4. Effect of prenatal restraint stress and pair-feeding on pup
FIG. 5. Effect of prenatal restraint stress and pair-feeding on pup
brain weight on day 32 postconception. Values are group means+SEM; ST = PF < AL, p<0.05.
behavioral developmenton day 32 postconception. Values are group means+SEM; ST < AL, p<0.05.
found that the stressing procedure alters the daily pattern of consumption of a liquid diet (manuscript submitted) such that the animals feed for longer periods throughout the day. As has been suggested by other authors (6), stress could also be causing metabolic changes in the animals. This should be addressed by future work. The decreased pup body weight on day 20 in both the ST and PF groups in this study is consistent with findings of studies of prenatally stressed end-term-fetuses (29) or newborn pups (2, 12, 16, 17, 20, 23). The fact that these pups exhibited catch-up growth is also consistent with a number of other reports (11, 15, 19). On the other hand, very few studies have reported effects on brain weight following prenatal stress. One study (28) has reported decreased brain weight in rats at 14 and 21 days after birth, but not at 8 and 40 days. Another study (3) reported no differences in brain weight in prenatally-stressed rats at 175 days of age. The results obtained in the current study suggest that prenatal stress, acting by way of reduced nutritional status, may lead to decreased brain weight in the early postnatal period but not at later ages. Several studies (8, 9, 14) have reported delayed behavioral and motor development following prenatal stress. Had the experimental design in the present study been similar to that used in these studies, the fact that the ST pups were significantly retarded in relation to the AL pups would have lent support to the earlier findings. However, the fact that the ST pups did not differ from the PF pups on this measure suggests that most, if not all, of the present findings were due to stress-induced maternal undernutrition. The fact that the growth deficiency present at birth was not apparent at later ages suggests that lactation was not seriously impaired in the previously treated dams. Since fos-
tering was not carded out in this study, the relative contribution of pre- and postnatal factors on development cannot be assessed. However, we have conducted further studies (manuscript submitted) which extend some of the findings reported here to offspring which were fostered after birth to untreated dams. These data support the current work in that undernutrition resulted in lighter brains on day 32, while there was no effect of prenatal stress. Behavioral development was not affected by either treatment in these fostered pups. In summary, these f'mdings suggest that the effects on brain and behavioral development in offspring which are generally attributed to prenatal stress may be, to a great extent, mediated by undernutrition due to stress-induced changes in the ingestive behavior of the pregnant dams. Furthermore, they imply that undernutrition should be taken into account in any attempt to identify the actual mechanism of prenatal stress effects, particularly if it has been shown that the stressing procedure reduced food intake in pregnant dams. Certainly, if hormonal changes are hypothesized to play a major role, it is important that the investigators show, through the inclusion of a pair-fed control group, that the effects cannot be explained solely by stressinduced maternal undernutrition. ACKNOWLEDGEMENTS The authors thank Dawn McCutcheon of the Health Studies Department for her substantial contribution in the conduct of this research. Kathryn Biota of the Psychology Department also provided technical assistance and Doris Winfield made helpful comments on the writing of this manuscript. This research was supported by a Natural Science and Engineering Council of Canada Grant A7617 awarded to P. Wainwright.
786
WARD AND WAINWRIGHT REFERENCES
1. Alario, P.; Gamallo, A. ; Beato, M. J.; Trancho, G. Body weight gain, food intake and adrenal development in chronic noise stressed rats. Physiol. Behav. 40:29-32; 1987. 2. Anderson, D. K.; Rhees, R. W.; Fleming, D. E. Effects of prenatal stress on differentiation of the sexually dimorphic nucleus of the preoptic area (SDN-POA) of the rat brain. Brain Res. 332:113-118; 1985. 3. Anderson, D. K.; Fleming, D. E.; Rhees, R. W.; Kinghorn, E. Relationships between sexual activity, plasma testosterone, and the volume of the sexually dimorphic nucleus of the preoptic area in prenatally stressed and non-stressed rats. Brain Res. 370:1-10; 1986. 4. Archer, J. E.; Blackman, D. C. Prenatal psychological stress and offspring behavior in rats and mice. Dev. Psychobiol. 4:193-248; 1971. 5. Armario, A.; Ortiz, R.; Balasch, J. Effect of crowding on some physiological and behavioral variables in adult male rats. Physiol. Behav. 32:35--37; 1984. 6. Armario, A.; Hildago, J.; Giralt, M. Evidence that the pituitary-adrenal axis does not cross-adapt to stressors: Comparison to other physiological variables. Neuroendocrinology 47:263-267; 1988. 7. Barlow, S. M.; McElhattan, P. R.; Sullivan, F. M. The relation between maternal restraint and food deprivation, plasma corticosterone, and induction of cleft palate in the offspring of mice. Teratology 12:97-104; 1975. 8. Barlow, S. M.; Knight, A. F.; Sullivan, F. M. Delay in postnatal growth and development of offspring produced by maternal restraint stress during pregnancy in the rat. Teratology 18:211218; 1978. 9. Barlow, S. M.; Knight, A. F.; Sullivan, F. M. Prevention by diazepam of adverse effects of maternal restraint stress on postnatal development and learning in the rat Teratology 19:105-110; 1979. 10. Barlow, S. M.; Knight, A. F.; Sullivan, F. M. Diazepaminduced cleft palate in the mouse: The role of endogenous maternal corticosterone. Teratology 21:149-155; 1980. 11. Beckhardt, S.; Ward, I. L. Reproductive functioning in the prenatally stressed femal rat. Dev. Psychobiol. 16:111-118; 1983. 12. Dahlof, L. ; Hard, E.; Larsson, K. Influence of maternal stress on the development of the fetal genital system. Physiol. Behav. 20:193-195; 1978. 13. Dunlap, J. L.; Zadina, J. E.; Gougis, G. Prenatal stress interacts with prepubertal social isolation to reduce male copulatory behaviour. Physiol. Behav. 21:873-875; 1978.
14. Fride, E.; Soreq, H.; Weinstock, M. Are the effects of gestational stress on motor development and cerebellar cholinesterase activity mediated prenatally? Int. J. Dev. Neurosci. 4:407-413; 1986. 15. Herrenkohl, L. R.; Whitney, J. B. Effects of prepartal stress on postpartal nursing behavior, litter development and adult sexual behavior. Physiol. Behav. 17:1019-1021; 1976. 16. Joffe, J. M. Modification of prenatal stress effects in rats by dexamethasone and adrenocorticotrophin. Physiol. Behav. 19:601-606; 1977. 17. Kinsley, C.; Svare, B. Prenatal stress effects: Are they mediated by reductions in maternal food and water intake and body weight gain? Physiol. Behav. 37:191-193; 1986. 18. Meisel, R. L.; Dohanich, G. P.; Ward, I. L. Effects of prenatal stress on avoidance acquisition, open-field performance and lordotic behavior in male rats. Physiol. Behav. 22:527-530: 1979. 19. Peters, D. A. V. Prenatal stress: Effects on biogenic amine and plasma corticosterone levels. Pharmacol. Biochem. Behav. 17:721-725; 1982. 20. Rhees, R. W.; Fleming, D. E. Effects of malnutrition, maternal stress, or ACTH injections during pregnancy on sexual behavior of male offspring. Physiol. Behav. 27:879-882; 1981. 21. Rosenzweig, S.; Blaustein, F. M. Cleft palate in A]J mice resulting from restraint and deprivation of food and water. Teratology 3:47-52; 1970. 22. Simpson, C. W.; Dicara, L. V.; Wolf, G. Glucocorticoid anorexia in rats. Pharmacol. Biochem. Behav. 2:19-25; 1974. 23. Smith, D. J.; Joffe, J. M.; Heseltine, G. F. D. Modification of prenatal stress effects by adrenalectomy, dexamethasone and chlorpromazine. Physiol. Behav. 15:461-469; 1975. 24. Wahlsten, D. A. A developmental time scale for postnatal changes in brain and behavior on B6D2F~ mice. Brain Res. 72:251-264; 1974. 25. Wahlsten, D. Genetic variation in the development of mouse brain and behavior: Evidence from the middle postnatal period. Dev. Psychobiol. 8:371-380; 1975. 26. Wainwright, P.; Russell, V. A quantitative assessment of the effects of malnutrition on behavioral development in laboratory mice. Behav. Neural Biol. 39:140-144; 1983. 27. Ward, I. L. Prenatal stress feminizes and demasculinizes the behavior of males. Science 175:82-84; 1972. 28. Wiggins, R. C.; Gottesfeld, Z. Restraint stress during late pregnancy in rats elicits early hypermyelination in the offspring. Metab. Brain Dis. 1:197-203; 1986. 29. Wilke, D. L.; Tseu, S. R.; Rhees, R. W.; Fleming, D. E. Effects of environmental stress or ACTH treatment during pregnancy on maternal and fetal plasma androstenedione in the rat. Horm. Behav. 16:293-303; 1982.