Inhibition and facilitation of eating by electric shock—II: Shock level, shock schedule and strain of rats

Inhibition and facilitation of eating by electric shock—II: Shock level, shock schedule and strain of rats

Journalof Psbcheromatlc Resedrch1963 Vol 7 pp 21510223 Pergamon Press Ltd INHIBITION AND FACILITATION ELECTRIC SHOCK-II: SHOCK SCHEDULE AND STRATN ...

592KB Sizes 1 Downloads 26 Views

Journalof Psbcheromatlc

Resedrch1963 Vol 7 pp 21510223 Pergamon

Press Ltd

INHIBITION AND FACILITATION ELECTRIC SHOCK-II: SHOCK SCHEDULE AND STRATN GRAHAM

M

ST~RRITT”

and

(RecerLlerl22

Jdy

KENNETH

Prmted

III Northern

Irelaxl

OF EATING BY LEVEL, SHOCK OF RATS SmmmGt

1963)

and obestty have been produced m the laboratory rat only by such radical techniques as producmg hypothalannc lesions [I, 3, 61, or sublectmg the rat to contmuous electric shock which rt can turn off only by consummg a rich hqutd diet [9] These methods, although useful for other purposes, appear to contribute ltttle toward an understandmg of the chmcal problems of overeating and obesity The latter condttlons typically mvolve neither lesions nor direct envrronmental pressures to consume rich foods In contrast, some expertmental research with animals has mdicated that eatmg ehmmates or greatly reduces a number of the phystological and behavioral evidences of fear [4, 21. If eating does reduce fear, then stressful sttmulatton should be able to induce overeating m the absence of lesions or direct coercion to eat. One worker [7, 81 has reported “compulsive” overeating, without lesions or coercton, when rats were given brief shocks during dally free eating periods. In a similar expertment with the addition of an unshocked control group [5], it was found that overeating occurred during shock, while between shocks eating was inhibited, with the net result that shocked Ss showed reliably lower total food intake than unshocked controls Since faclhtatlon of eating was seen during shock, it appeared worthwhile to study this procedure further, to determine whether total food Intake might be htgher in shocked Ss than m unshocked controls with other intensities or schedules of shock or m another strain of rats. OVEREATING

EXPERIMENT

1

The specific alms of this experiment were to explore a wider range of shock mtensltles, and to compare two strams of rats, as variables mfluencmg total food Intake and eatmg durmg shock In addltlon, the period of food deprlvatlon with shock durmg dally free eatmg periods was extended from 4 days, as used m previous studies, to 8 days, to determme whether greater faclhtatlon of eatmg by shock would occur with longer apphcatlon of the deprlvatlon and shock condltlons Method The Ss were 25 Sprague-Dawley albmo rats and 25 Long-Evans hooded rats, obtamed from Rockland Farms, New York All were naive females approximately 90 days old when started m the experiment The No Shock Group was started at a later time, after all shocked Ss had completed their part m the experiment * Umverslty of Colorado Medlcal Center, Denver, 1 Denver Umverslty, Denver, Colorado, U S A 215

Colorado,

U S A

216

GRAHAM

The appdrdtus

conslsted

M STERRITT and KENNETH SHEMBERG

of d box

5:

In

2: 11 m

c 20 in

high, with d gla\s front, gdivanlzed ot 4, In dldmeter s s rods spdced t In between centres Alternate bals were connected to opposite poles of a model 228 Applegate constant current stimulator (C J Applegate Co, Boulder, Colorado) through d telegraph key by which t dehvered shock, at the start of each mmute A clear plastic tube sloping 45- to the vertical held rf large supply of food pellets This tube emptied Into a cup 2 In I d i A In deep The food pellets were 4 mm x 3 3 mm i: 45 mg, manufactured by the Noyes Co , Lancaster, N H Ss had free access to this supply of pellets% all times when I; the apparatus The four shock-levels used, 0 1. 0 2. 0 4 and 0 8 mA. were nreset with d 200 kQ reslstor acres\ the grid floor A sloping data board v&h a tabular data sheet and stopwatch located Just outslde the window dt the pellet-tube end of the cage facllrtated recording of pellet-taking sheet metal

sides, a Llear plastic

hd dnd a grrd floor

made

Procedure Ss remamed m home cages wtth Rockland Lab Chow pellets (Teklad Corp , Monmouth, III ) and water ad lib for 3 days, with dally weighing to estabhsh “orlgmal” body weight Over the followmg 5 days, Ss were reduced to 80 per cent of orlgmal weight by hrmtmg feeding to approximately 10 g of wet mash made from Purma Lab mash chow for rats (Ralston Purina Co , St LOUIS, MO ) and water This comprised the entire dally feeding which was given In feeder cages located In the \ame room that contdmed the expertmental apparatus, away from the home cage room Feedmg occurred at approximately the same time as Ss went Into the apparatus In the later phases of the experiment Water was provided ad Irb In home cages throughout the experiment On the 6th day each S spent approximately 15 mm in the apparatus wIthout shock This WdS d famlharlzatlon period Every S explored the apparatus and ate 30 pellets on this day

Deprived phase On the 8 days followmg famlharlzation each S went Into the apparatus for 21 mm, with shock on durmg the first 5 set of each mm Shock was never given to the No-Shock Group Shock remamed off for the first mm of each day and pellets were not counted during this ttme, when the S was assumed to be “warming up ” After leavmg the apparatus, the S went Into the feeder cage for 20 mm, where vufficlent wet mash was given to bring Ss back to 80 per cent of orlgmal body weight Sated phase For the followmg 4 days the same procedure wds followed. except that Ss now hdd ad lib access to Rockland lab food pellets in home cages, and were further sated with wet mash url lrb during the 30 mm Just before going Into the apparatus

Ss were weighed dally for 10 days followmg the last day of the sated phase During Ss had a generous supply of Rockland lab pellets and water m home cages at all times

thlb perrod

Results RcwltJ

111the depnceclphase

Number of pellets taken durmg the first 5 set of every mln, when Ss were shocked. dc well as total number of pellets taken and per cent taken during the first 5 set are shown m Table 1 In FIN. 1 eatmg IS shown as a function of 5 set Intervals, totalled over all minutes nnd all 8 days The two strains of rats did not differ slgmficantly 111any measures For this reason the strains were comduring either the deprived or the sated phase blned m Fig 1 and m all subsequent analyses Gross mequahtles of variance between the various shock groups and slz‘ible numbers of zero eating scores at higher shock levels argued for the use of nonparamctrlc statistics The Kruskal-Wallace test indicated slgnlfcant overall dlfYerence\ between the various lntensltles of shock, III measures of total eating (p -C 0 001). In absolute

Inhlbltlon TABLE

1 --NUMBERS

and faclhtatlon

of edtmg



NO shock ~-_.~__. (a) Pellets taken III first 5 set (sum for 20 mm/day, all days) -(b) Pellets taken m 20 mm (sum for all days) ___~__ x

lOO*

/

* 8 3 per cent would

shock-

OF PELLETS TAKEN IN FIRST S-SEC INTERVALS 20 MIN IN EACH GROUP Deprwed

(a)/(b)

by electric

02 IllA

01 mA

Alblno Hooded

277 269

307 302

Albino Hooded

3641 3398

3215 2950

Albino Hooded

7 6 J 9

9 5 10 2

be expected

phase-8

3841 531 1821 3258

04 IllA

08 mA

II6 159

IO 6

823 977

173 175 -~

02 ,------

L

I

Group. ,

ma

Group

0 8

ma

Group

I

I

.

O* 5

FIG 1 Total numbers with albmo and hooded

.

I

IOO-

I

-.. t

days

Group

ma

o4

I

IV

for all 5 set mtervals

.

~ /--

IAKEIU

58 34

on the basrs of equal probdblhtles

No Shock

TOTAL

217

Sated phase -4

days

21 1 ( I4 I I63 1163

AND

II

f

-

..a_.

.

--.__a___

I

I

I

10 15 2025 30 354045 5 SEC INTERVALS

.

t

_.I

I

I

50 55

of pellets eaten m each 5-set Interval m the deprived phase, rats combmed (N = 10 per group) *“O” IS the shock Interval for all but the No Shock Group

amounts of eating durmg the first 5 set @ < 0 OOI), and m per cent taken durmg the first 5 set (p < O*OOl), all probablhtles reported m this paper are two-tad). Total eating None of the shocked groups ate greater total amounts than were eaten by the No Shock Group, m fact, the group given 0 1 mA shock was the only shocked group that did not eat slgmficantly less than the unshocked group (No Shock Group vs 0 2 mA Shock Group, 1) < 0 05, 0 4 mA Shock Group, p < 0 001, O-8 mA Shock Group, p < 0 001, except as specrfically noted, all 2-group comparlsons are based on the Mann-Whitney U Test) Eatmg durrzng shock The 0 2 mA Shock Group ate greater numbers of pellets durmg the tmle Intervals when shock was on than were eaten m the same time lntervals by the No Shock Group (p < 0 05) The groups given 0 4 and 0 8 mA of shock ate slgmficantly fewer pellets durmg shock than were eaten durmg the same time Intervals by the unshocked group (No Shock Group vs 0 4 mA Shock Group, p < 0 05, 0 8 mA Shock Group, p < 0 001).

218

GRAHAM M

STERRKT

and KENNETHSHEMBERG

Percentage eaten durrng Jhock A larger percentage of eating occurred durmg the time that shock was on m the 0 2 mA Shock Group, as compared to the No Shock Group (JJ < O*OOl), while the 0 8 mA Shock Group did a slgmficantly lower proportlon of then total eating during the time that shock was on, by comparison to the No Shock Group (p < 0 001) Indmdual dlfSerencer Slgmficant overall differences between variances are mdlcated by Bartlett’s test (Variances of total eatmg, p < 0 001, variances of eating durmg shock, p < 0 001) F-tests indicate that m total eating the group given 0 1 mA of shock was more variable (p < 0 05) while the group given O-8 mA of shock was less variable [p < 0 001) than the No Shock Group. The other shocked groups did not differ from the No Shock Group m variance Inspection of the data of the group given 0 1 mA of shock indicates that some Ss overate m response to shock while other Ss given the same intensity of shock underate, by comparison to the range of eating seen m the No Shock Group The variance of eating during the first 5 set within the No Shock Group IS exceeded by that m the 0 1, 0 2 and O-4 mA Shock Groups (m each case, the F-test IS slgmficant at p < 0 01) All of the Ss of the group given 0.8 mA of shock ate very httle during shock, leading to significantly less varlablhty m the O-8 mA Shock Group than m the No Shock Group (F-test, p < 0 001). Day l-4 Z’J Day 5-8 The deprived condltlon was continued over 8 days to determine whether the effects of shock on eating would increase or decrease over time. Comparisons of the first 4 days with the second 4 days of the deprived phase, in total amounts eaten, are presented m Fig. 2. Total amounts eaten increased slgmficantly from the first to the second 4 days wlthm all groups (t-tests based on within-subjects difference scores, every p -=z0 01 except for the 0 8 mA Shock Group where p < 0 05). The 0 1 and O-2 mA shock groups did not differ from the No Shock Group m amount of mcrease of total eating from the firrt to the second 4 days, while the groups given 0.4 and 0.8 mA of shock

No

114

Shock

Group

0 4 .

ma

Group

0 8

ma

Group

5-8 DAYS

FK

2

Total numbers of pellets eaten on the first 4 days and on the second 4 days of the deprived phase

inhlhltlon

and faclhtatlon of eatmg by electric shock-11

219

both showed stgmficantly less gain m eatmg than appeared m the No Shock Group (No Shock Group vs 0 4 mA group, p < O-05, O-8 mA group, p < 0 01) These findmgs reveal that the mhtbttton of total eatmg by shock remained and, m the case of the groups given stronger shock, became more pronounced from the first to the second 4-day perrod, relative to eating by the No Shock Group Under none of the shock condtttons did the proportcon of eating that occurred during shock change stgmficantly from the first to the second 4-day period, 1.e. the facthtatlon of eating during shock, at moderate levels of shock, relative to total eating, became nerther greater nor less as a result of the longer apphcatlon of the experimental conditions. Results during the statedphase As m the deprived phase, the albino and hooded strams did not differ u-r any measures, so their scores were combmed for analysis The Kruskal-Wallace test for overall differences between means reveals srgruficant effects associated with shock level m total eatmg (p < 0.001) and m eating during the first 5 set (JJ < O*OOl) None of the Ss grven 0 8 mA of shock ate anything during Consequently, shock in the sated phase, and few ate any pellets between shocks percentages taken during shock are not meanmgful m this group. The other 4 groups show no srgmficant overall differences m percentage taken during shock. None of the shocked groups ate greater total amounts than the unshocked group ate The groups given higher intensities of shock ate srgnlficantly less than the No Shock Group (No Shock Group vs. 0.4 mA Shock Group, p < 0 01; No Shock Group vs. 0.8 mA Shock Group, p < O*OOl). The great between-subjects vartabtlrty of total eatmg among moderately shocked Ss and the tendency of some shocked Ss to overeat, relative to the range of total The subject eating m the No Shock Group, did not persrst mto the sated phase that ate the greatest amount m the sated phase was m the No Shock Group. None of the shocked groups ate a reliably greater number or percentage of pellets during the shock intervals than were eaten m the same time mtervals by the No Shock Group Body wezght gain Body wetght gain was calculated from the 3-day period Just before the first day of pre-experimental deprivation until the tenth day after the end of the sated phase None of the groups differed slgmficantly from each other m this measure EXPERIMENT

2

It would seem that if shocked Ss overeat while shock IS on, one way to cause them to eat greater fotal amounts than controls would be to increase the proportlon of time that shock IS on The purpose of experiment 2 was to determine If rats given shock more frequently or for longer durations ~111m fact eat more than unshocked controls An addItIona goal was to determine whether apenodlc, unpredictable sequences of shock-on, shock-off Intervals would have the same effects on eatmg as the umformly perlodlc shock schedules studled previously Method Subjects Ss were 63 female Long-Evans hooded rats, approximately ment, obtamed from Rockland Farms Jnc , New York

90 days old at the start of the expert-

220

GRAHAM

M STERKITT and KENNFTH SHEMUERG

Apparatus wds the sdme ds *as used m Experiment 1 All of the shocked the shock mtenslty which had ylelded the greatest amountsof overeatmgdurmg

groups received 0 2 mA, shock m Expernnent 1

The procedure wc\s the same as m Experunent 1 except that Ss were run for 4 days under the food-deprived condltlon and 4 ddys sated, rather than 8 deprived and 4 sated (the results of Experlment 1 had mdlcated that the extra 4 days of the deprived condltlon drd not yield stronger effects of shock on eatmg) Ss were dlvlded mto seven groups One group was given one shock Interval of 5 set duration at the start of each 60 set m the apparatus (5-60 group) Another group received 5 set of shock at the start of every 30-set Interval (5-30 group) The third and fourth groups were given 10 set of shock at the start of every 60 set (lo-60 group) or every 30 set (lo-30 group) These were the perlochc shock schedule groups Two groups received aperIodIc schedules of shock, with 5 and 10 set durations dehvered at unpredictable mtervals In one group (A-100 group) the total duration of shock was 100 set/20 mln (the same total duration as has given the 5-60 group) In the other group (A-200) the total durdtton of shock was 200 sec,l20 mm (the same as the lo-60 and 5-30 groups recclved) In addltlon to the 6 shocked groups there was an unshocked group Nme Ss per group were present m the deprived phase Schedulmg problems led to a reduction to 6 per group m the sated phase

Table 2 shows the rates of eating found under each of the shock schedules deprived and 117the sated phase TARLI- 2 -

RATF

OF EATING, EXPRESSED AS PELLETS TAKEN

Shock NO

Pcrlodlc

m the

PER 100 SEC

schedule

schedules

AperIodIc

schedules

A-100 .~ 8 17

A-200

3hoch

Edtmg durmg shock (durmg first 5 set m No Shock Group) Eating between shocks (durmg last 55 set m No Shock Group) Total eatmg

Deprived

5 72

5-60 _~ 1547

Sated

2 33

1779

712

450

357

8 29

7 48

Deprived

5 68

499

3 49

3 37

3 90

3 50

2 88

Sated

2 78

366

257

151

094

I 45

I 29

5 69 2 74

5 86 483

4 23 333

4 28 201

5 92 1 82

3 88 2 02

3 78 2 32

Deprived Sated

lo-60 797

5-30 lo-30 _~.~~~ 889 99X

8 35

Four one-way analyses ot variance were carried out, one based on total eating, deprived phase, one on eating between shocks, deprived phase, and the other two on The measure total and between-shock eating, respectively, m the sated phase employed m all four analyses was the rate of eatmg, expressed as pellets taken/l00 set (which yielded convenient whole-number scores) The overall analyses of variance indicated that differences between groups were slgnlficant in the two deprived phase measures (total eating, p i 0 05, eating between shocks, p -, 0 01) but not 111the sated phase measures With respect to eatmg dwmg shock, Bartlett’s test Indicated slgmficant differences between groups m the varlablhty of eating m the deprived phase and m the sated phase (In each case p / 0.02, eating during the first 5 set was used m the case of

Inhlbltlon and faclhtatlon

of eating by electric shock-11

221

the No Shock Group). Scatterplots showed that means and variances of eatmg during shock were correlated. Therefore the nonparametrlc Kruskall-Wallace test was applied to this measure, yleldmg the finding that m either phase, differences between groups m rate of eating during shock could reasonably be attributed to chance. Total

eating

Shocked Groups vs. No Shock Group

With regard to the mam question of this expenment, orthogonal comparisons Indicated that none of the shocked groups, mdlvldually or m any combmatlon, ate greater total amounts than were eaten by the No Shock Group. Penodcc vs. Apenodlc

shock schedules

Orthogonal comparisons of deprived--phase measures of total eating reveal that the aperlodlc shock groups, collectively, ate less by comparison to the No Shock Group (p < 0 025) and by comparison to the combined perlodlc shock groups (p < 0 OS). Measures of eating between shocks ylelded the same findings (apenodlc shock groups vs No Shock Group, p < 0 001, aperiodic vs. periodic shock groups, p < 0 01) Regular shock schedules-

efects

of duration andfrequency

of shocks

A separate 2 x 2 analysis of variance was carried out with the deprived phase total eating scores of the 4 regular shock schedule groups, yleldmg a slgmficant effect only m the interaction of duration and frequency of shocks Orthogonal comparisons indicated that the group given the least total shock-time per 20 mm (5-60 Group) and the one given the greatest total shock-time (IO-30 Group) ate more than was eaten by the 2 groups (5-30 and 10-60) given the mlddhng amounts of shock time (Orthogonal comparison of 5-30 and IO-60 groups vs 5-60 and IO-30 Groups, p < 0.05). Inspection of Table 2 leads to the conclusion that this complex effect cannot be slmphlied by analysmg it mto durmg-shocks and between-shocks components-the same trends appear, though less rehably, m eating durmg shock and between shocks as m the total eatmg scores Although effects m the sated phase were not statlstlcally reliable, the between groups trends are more easily interpreted m the sated phase, the unrehable trend was toward less eatmg during and between shocks as the total amount of shock time Increased, whether shock time was lengthened by increasing the frequency of shocks or their duration. It IS of interest that the shock schedule which ylelded the greatest amounts of eating (5-60 schedule) IS the one which was employed m all of the prior studies of this phenomenon, none of the “new” schedules, which had been expected to yield higher rates of eating, did so by comparison to the 5-60 schedule and this schedule again falled to elicit a rehably higher net rate of eating by comparison to the unshocked control group Eatmg durmg Ahock us eating between shocks

A difference score was computed for each S by subtracting the rate of eatmg between shocks from the rate durmg shock. This was done separately for the deprived and sated phases 4

222

GRAHAM M. STERRITT and KENNETH SHEMBERG

In the deprived phase every one of the shocked groups ate at a rehably greater rate durmg shock than between shocks (tests based on difference scores wlthm the A-100 group, p < 0 05; wlthm all other shocked groups, p < 0 01). In the sated phase only the 5-60 group ate rehably more during shock than between shocks, although the trend was m this dIrectIon m all of the other shocked groups as well SUMMARY

AND

CONCLUSIONS

Prior research has indicated that eating reduces fear and ehmmates behavioral signs of pam. Overeating was predicted on this basis m rats that were permitted to eat In a sltuatlon that afforded no effective means of escapmg from painful stIrnull. It was found that subjects given moderately intense shock ate more than controls, but only during shock, eating less than unshocked controls between shocks. At no mtenslty and with no schedule of shock, wlthm the range studled here, did shocked groups eat greater total amounts than were eaten by unshocked control groups At higher shock mtensltles, or when shock was delivered on an unpredictable, aperlodlc schedule, shocked groups consistently underate by comparison to unshocked control groups It can be concluded that shocking hungry rats durmg dally free eating periods IS not a promlsmg method for producing predictable overeating among groups of rats Some zndzazduals apparently do overeat, while others undereat, at the same rather mild shock level This iindmg opens the posslblhty that the method employed here may be useful m studies of the role of genetic factors, early experiences, drugs, etc , as determinants of mdlvldual differences m the dlrectlon of the effects which stressful stlmulatlon has on eating. It remains to be demonstrated, however, that overeating ehclted m this way will persist long enough to produce overweight. Acknowledgements-This research was supported by a grant (M-2776) from the National Instztute Mental Health The authors are indebted to Dr Donald Stdson for statzstical consultation

of

REFERENCES BROBECK J R , TEPPERMAN J

and LONG C N H Experimental hypothalamzc hyperphagla zn the albmo rat Yale J Bzol Med 15, 831-853 (1943) FARBER I E Response fixation under anxzety and nonanxzety condltlons J Exp Psycho/ 38, 111-131 (1948) MILLER N E , BAILEY C J and STEVENSON J A. F Decreased “hunger” but Increased food Intake resulting from hypothalamic lessons Sczence 112, 256-259 (1950) PAVLOV I P Condztzoned Rejexe, Translated by G V Anrep, Oxford Umverslty Press, Humphrey Mzlford (1927) STERRITT G M Inhzbztlon and faclhtatlon of eatmg by electric shock J Comp Physzol Psychof 55, 226-229 (1962) TEITELBAUM P Sensory control of hypothalamic hyperhagta J Camp Physzol Psycho/ 48, 156-163 (1955) ULLMAN A D The experlmental productzon and analysis of a “compulsive eatzng symptom” m rats J Comp Phvyzot Prythol 44, 575-581 (1951) ULLMAN A D Three factors involved m producmg “compulszve eatmg” m rats J Camp Phyrzol Psychol 45, 490496 (1952) WILLIAM? D R and TCITELBAUM P Control of drmkmg behavzor by means of dn operdntcondztlonmg technique Sczence 124, 1294-1296 (1956)