History

Medical Science Publication No. 4, Volume 1

A STUDY OF COMBAT STRESS IN KOREA, 1952:
PHYSIOLOGIC AND BIOCHEMICAL*

FRED ELMADJIAN,PH. D.
AND
STANLEY W. DAVIS, PH.D.

In late August 1952 a team, consisting of personnel from the Army, Navyand Operations Research Office, went to the Far East to study the physiologic,psychologic and psychiatric aspects of combat stress of infantrymen. Thepreliminary report has already been made (1). This is a presentationof the physiologic and biochemical data compiled and analyzed since thatpreliminary report, with emphasis on the adrenal cortical function andsteroid metabolism. The data include urinary, 17-ketosteroids, Porter-Silberchromogens, sodium, potassium, urea and uric acid.

Subjects

Two principal groups of men in combat were studied: (1) A group of mendesignated as "Able Co.," who experienced an acute combat situationin which they were the lead company in an unsuccessful attempt to regaina hill position. Data to be presented from this group include a pre-combat(A) sample obtained some 2 hours after they were briefed for the attack;another sample (B) some 17 hours after they were returned to the rear (thisgroup includes only five men who were in the original group), and a 4-day(C) and 22-day (D) follow-up after the principal engagement. (2) The secondand smaller group consisted of men who experienced a prolonged and sustainedaction defending the hill position after it was taken, against enemy counterattacksfor 5 days. No precombat (A) data were obtained from this group, whichwas designated as "George Co." However, data will be presentedon (B) some 15 hours after they were relieved and some 10 days (C) afterthe defensive action.

These data are compared with a group of controls, men who were stationedimmediately behind the main line of resistance in blocking position. Dataof a group of psychiatric casualties will also be presented.

*Presented 19 April 1954, to the Course on Recent Advances in Medicine and Surgery, Army Medical Service Graduate School, Walter Reed Army Medical Center, Washington, D. C.

A second aspect of the data to be reported includes a small sample ofeach group, who were selected at random, to whom ACTH was given to testadrenal cortical reserve or responsiveness. Previous to testing the menin the combat area, a group of men in Japan (Camp Omiya) were given theACTH in a test-retest situation to determine reproducibility in the sameindividual of the adrenal reserve of capacity.

Methods

Samples (not including the ACTH test) represent collections over a meantime of about 3 hours. Except for (B) samples of both Able and George Co.,the majority of samples represented before-noon urines. Those receivingACTH were injected with 2 cc. of Wilson's Gel preparation in the afternoonand collections were made over approximately a 15-hour period; the lastvoiding being the following morning about 5 to 6 a. m. See details forcollecting period in tables 1, 2 and 3.

The method for extraction of 17-ketosteroids was that described by Pincus(2). The urine was hydrolyzed and extracted with ether and preparedto the point of Girard separation in Korea. The extracts in test tubeswere flown to the United States and steroid analyses were completed atthe Worcester Laboratories. The Butanol extraction for Porter-Silber chromogensand preparation of the samples for analysis were completed in Korea andsent in test tubes to Dr. Peter Forsham at the Metabolic Institute at theUniversity of California Hospital, for the conduction of the Porter-Silberreaction. The method used was a modification of the Reddy method (3).Methods used for electrolytes, urea and uric acid have been previouslydescribed (1).

Results

Control and Combat Data

In table 1 and in figure 1, we present the mean 17-ketosteroid outputvalues as mg. per hour in the various control and combat groups. In table1 group (c) Able Co. A minus officers has been set up separately for tworeasons: (1) officers contributed urines to the precombat (A) series butnot to the subsequent ones, and (2) the officers exhibited, as a group,an extremely high output level of 17-ketosteroids (17-KS). The average(A) output value for the 5 officers was 1.07±0.187 mg. per hour,whereas the corresponding value for the 15 enlisted men was 0.56±0.06mg. per hour. It is deduced that either these officers at any rate werea group of "high" 17-KS excretors (4) or that emotionaltension attributable to battle anticipation led to especial activationof the pituitary-adrenal system during the pre-combat period.

FIGURE1. Urinary 17-ketosteroidexcretion. From left to right the first (A) represents the mean value ofthe controls; the first column under the second (A) represents the meanof the enlisted men in Able before combat and the second column under thesecond (A) represents the mean of the commissioned and noncommissionedofficers. The third (A) represents the mean of all (enlisted and officers)of men in Able A (see text).

When mg. per hour pre-combat values (c) are compared with post-combatvalues (d) for Able Co., a not quite significantly higher value is obtainedfor the latter period. In contrast the post-combat values (f) for GeorgeCo. are significantly lower than both the Korean control (a) and the post-combat(d) Able Co. values. There are no other statistically significant differencesbetween the various sets of data. It is notable, however, that the psychiatricbattle casualties (h) exhibit the lowest mg. per hour output, indicatingno stimulation, but rather a damping of the 17-KS output.

The Na/K ratios are depicted in figure 2. We note that in the controlgroup we have fairly good agreement between A and A' (test-retest at 1-weekinterval), somewhere around three as a value. However, in observing thedata on Able Co. we find that A of Able has a relatively low ratio, indicatingsome degree of adrenal cortical activity with respect to electrolytes and,furthermore, that it is still low in B. However, in C value we find thatthe value comes back towards normal, in fact a little above normal, andthat this continues in D. With regard to George Co. we find that the ratiois high on the

Table 1. Urinary 17-ketosteroids of CombatInfantrymen in Korea, 1952

Group studied	Date	Time sampled	Number	17-KS
(a) Korean controls	26, 27 Oct	9:00 a. m.-12:30 p. m.	24	0.64±0.04
(b) Able Co. A	13 Oct	3:00 p. m.-5:30 p. m.	20	0.69±0.08
(c) Able Co. A minus officers.	13 Oct	3:00 p.m.-5:30 p.m.	15	0.56±0.06
(d) Able Co. B	15 Oct	12:30 p. m.-4:00 p. m.	20	0.93±0.19
(e) Able Co. C	19 Oct	6:00 a. m.-1:00 p. m.	20	0.64±0.08
(f) George Co. B	21 Oct	1:30 a. m.-6:00 a. m.	10	0.46±0.04
(g) George Co. C	30, 31 Oct	11:00 a. m.-5:00 a. m.	12	0.52±0.03
(h) Psychiatric casualties.	10, 14 Oct	(3) 2:00 a. m.-5:00 a. m. (2) 7:00 a. m.-3:00 p. m.	5	0. 43±0.06
Significances	(c, d) (a, f) (d, f)	t 1. 85 3. 16 2. 43	P <0.1 <0.01 <0.05	>0. 05 ---------- >0. 02

FIGURE2. Urinary sodium potassium (molar) ratio. Numbersat the base of the open columns indicate the number of determinations includedin each group.

average, indicating hypoadrenal activity. Then, some 9 to 10 days laterwe find that the value has returned back towards normal. The psychiatricgroup shows a low Na/K ratio, indicating some degree of adrenal corticalactivity, in fact the same degree of activity as that of Able A or B, butwith a very low 17-KS.

FIGURE3. Urine urea nitrogen excretion.Figures at base of open columns indicate number of determinations includedin each group.

Figure 3 shows the urea nitrogen in terms of mg. per hour. We note thatin the Korean controls A and A' we have fairly decent agreement, namely,within 400 to 500 mg. urea nitrogen per hour. In A of Able there is anincrease; this increase is only of marginal significance. In B we havea real increase in the urea nitrogen, indicating a significant proteincatabolism (almost doubling the urea nitrogen output above that of thenormal controls). In C we see this return of urea nitrogen back towardsits normal value and in D a rather slight increase, but not significant.In the case of the chronic stress situation we find that rather than anitrogen catabolism, the value is in the normal range, and in the post-stressthe C value increases somewhat though not significantly. The psychiatricgroup shows a decreased urea output. This value is quite low and significantlylower than the controls. It should be pointed out that in the B of Able,the increased urea nitrogen excretion is reflected in the blood with avery significantly high urea value; almost 20 mg. per 100 cc. urea nitrogenin the blood.

Figure 4 contains 17-KS values of those individuals of Able on whomwe have both B and C values. We note that there are from B to C some decreasesand some increases in 17-KS. We also note that the decreases are predominant.The data contributing to the large S. E. also are evident in B of Able.We will see what relationship these values of Able B have to the otherindices, such as Na/K, urea and uric acid.

In figure 5 we have a scatter diagram of the Na/K ratio against the17-KS of Able Co. B, indicated by the dots with the circles around themand the X, indicating the George Co. B group. The correlation coefficientof the Able group was -0.6, but was not significant. However, we note thatthe extreme values in Able Co., those who had a low 17-KS, are those whohave a high Na/K ratio and the converse, the low Na/K ratio individuals,are those who have a high 17-KS. We note that in the George Co. (the X's)they are scattered along the low 17-KS and high Na/K ratio area.

Figure 6 shows the urea nitrogen mg. per hour against 17-KS in mg. perhour. We have a very good correlation of +0.74 and a significance of <0.01.Here again we note that the data of George Co. B are clustered around thelow 17-KS and low urea areas.

FIGURE4. 17-ketosteroid changesof individuals of Able Co. on whom both B (17 hours after battle) and C(4-day followup) were obtained.

FIGURE5. Relationship of the Na/Kratio with 17-ketosteroid excretion of both Able Co. B and George Co. B.

We have in figure 7 the same picture as that with the urea when uricacid is plotted against 17-KS. The r here is +0.54, with a significanceof <0.01. In general, we see from these data that Able Co. B shows increasedadrenal cortical activity in the indices examined between 17-KS and Na/Kwith catabolism as shown by the urea and uric acid output. The catabolismis evident also with higher creatinine values. This is not apparent inthe chronic group, namely, George B, where we have no increase in 17-KS;the nitrogen metabolism is of normal range. In the psychiatric group wefind that there is a low 17-KS, but, unlike George B, a low 17-KS withsome degree of electrolyte activity.

ACTH Tests

In table 2 are the 17-KS data on the group of seven soldiers stationedat a rehabilitation center in Japan (Camp Omiya), and subjected to an initial(A) and a repeat (B) ACTH test with urine

FIGURE6. Scatter diagram of urinaryurea nitrogen excretion and the 17-ketosteroids of both Able Co. B andGeorge Co. B.

FIGURE7. Relationship between urinaryuric acid excretion and 17-ketosteroids of both Able Co. B and George Co.B.

collected from the 15 hours following ACTH administration. No pre-injectionsamples were determined. Whether the data are calculated as mg. 17-KS perhour or per 100 mg. creatinine, remarkably high and significant correlationsbetween the data of the two tests are had, indicating that with the ACTHused the 17-KS response is remarkably uniform and highly characteristicfor each individual.

Table 2. Reproducibility of ACTH Test-September16-24, 4 days between A and B*

17-KS, mg./hr.			17-KS mg./100 mg. Creat.
	A	B	A	B
M 7856	0. 52	0. 57	0. 69	0. 70
R 5249	. 21	. 16	. 38	. 67
W 9926	. 62	. 51	. 90	------------------------------
J 1789	. 73	. 73	1. 14	1. 11
M 2878	. 24	. 35	. 55	. 51
M 7056	. 23	. 29	. 30	. 40
P 9139	. 06	. 15	. 10	. 26
r=0.95			r=0.94

*Time of each collection represented urine collected from3 p. m. to 7 a. m. thus including overnight sample. ACTH given about 3p. m.

FIGURE 8. 17-ketosteroidexcretion mg./100 mg. creatinine before and after ACTH. Number at the baseof open bars indicates the number of individuals tested.

Table 3 and figure 8 present the mean data of 17-KS for various subjects in the ACTH test. It should be noted, first of all, that the Korean controls exhibited no increase of 17-KS excretion following ACTH administration. Since the pre-injection urine collections were made between approximately 9:00 a. m. and 12:30 p. m., and the post-injection collection for the following 15 hours which included a period of sleep, this apparent lack of stimulation may in fact involve sufficient stimulation to restore toward the morning level the known decline in 17-KS output values occurring in the afternoon and during sleep (5).

In any event, in contrast to the control subjects, the men of Able Co.tested show increases of 17-KS output following ACTH both in the post-combat(B) period and 10 days later (C). The George Co. subjects, on the otherhand, demonstrate a decline of 17-KS output following ACTH (which is particularlyobvious in the mg. per hour data) during the B period, but 10 days laterduring ACTH an output increase occurs. The psychiatric casualties demonstratea significant increase of 17-KS per 100 mg. creatinine in the post-combatperiod.

Figure 9 contains the Na/K ratio before and after ACTH. We find thaton A and A' on eight individuals we have fairly good agreement in adrenalsensitivity as noted by the Na/K ratio. We find that in B of Able the adrenalresponds to the ACTH even though the presamples already indicate some degreeof adrenal activity. In D sample (C samples were lost) we find that theNa/K ratio is high,

FIGURE9. Urinary sodium/potassium (molar) ratio beforeand after ACTH. Number at the base of the open column represents the numberof individuals included.

Table 3. 17-Ketosteroid/Creatinine ExcretionBefore and After ACTH

	No.	A		B		C
	No.	Before	After	Before	After	Before	After
Korean controls	8	0. 85±0. 05	0. 79±0. 09	----------	----------	---------	----------
Able Co	4	----------	----------	0. 72±0. 09	1. 60±0. 61	0. 68±0. 20	0. 96±0. 07
George Co	4	----------	----------	0. 53±0. 07	0. 43±0. 06	0. 60±0. 06	0. 80±0. 11
Psychiatric casualties	5	----------	----------	0. 76±0. 13	1. 48±0. 13	----------	----------

17-Ketosteroid Excretion Before and After ACTH

	No.	A		B		C
	No.	Before	After	Before	After	Before	After
Korean controls	8	0. 64±0. 05	0. 44±0. 07	----------	----------	----------	----------
Able Co	4	----------	----------	0. 71±0. 11	0. 87±0. 30	0. 49±0. 17	0. 62±0. 21
George Co	4	----------	----------	0. 43±0. 06	0. 27±0. 05	0. 53±0. 04	0. 61±0. 06
Psychiatric casualties	5	----------	----------	0. 43±0. 06	0. 72±0. 32	----------	----------

Note. All after ACTH included night sample: beforeACTH of George Co. B represented collection 1:30 a. m.-5 a. m. ACTH wasgiven at 3 p. m. of same day and collection was concluded next morningat about 6 a. m. Psychiatric casualties after ACTH samples did not representovernight sample but were collected during waking hours.

that it is back to normal in the pre-ACTH samples, and is responsiveto the ACTH by a reduction in the Na/K ratio. Now the interesting featureis that of George Co., where there is a hypoadrenal cortical activity asindicated by the Na/K ratio. There is, however, a response to the electrolytesafter ACTH, though there was no such evidence in the 17-KS. (We will notenext that the 17-KS are about the same as the P-S.) The values in C returnto normal. In B of the psychiatric group we find that though the adrenalsare active there is much more activity in terms of electrolytes after ACTH.

In figure 10 we have the P-S chromogens in terms of mg. per hour inthree separate groups: (1) the Korean control, (2) George Co. B and C,and (3) the psychiatric group. Our Porter-Silber data were scattered becauseonly a portion of the total samples extracted were analyzable. This inall probability is due to the high blank and other defects in this particularmethod used, as well as possible technical errors in preparation for analysis.(At present there is a method described by Nelson and Samuels which doesnot have this difficulty and the high blanks which make readings impossiblehave been obviated and this matter has been corrected.) We notice thatin our Korean controls there are only a few determinations. The first blockindicates the pre-ACTH group and is followed by the post-ACTH. In theseparticular data the post-ACTH group do not include any of the individualsin the pre-ACTH group. There is a clear increase as a result of ACTH injection.In B of George Co. we notice that there are some seven determinations forpre-ACTH and three for post-

FIGURE10. Porter-Silber chromogensbefore and after ACTH (see text).

ACTH. In the post-ACTH data the three are included in the pre-ACTH.We note that the adrenal is non-responsive as measured by the P-S reaction.However, in C we note that there is a response to the ACTH after 9 to 10days, and the four individual determinations are included in the pre-ACTH.In the psychiatric group the two pre-ACTH are not the same individualsas in the post-ACTH. We note here that two features stand out: (1) Thereis in control samples a greater increase in the post-ACTH in the P-S thanthere was in the 17-KS, and (2) the adrenal non-reactivity observed inB of George Co. in the 17-KS is confirmed with the P-S reaction. However,in the psychiatric group there is a difference-we have an increased reactivityto 17-KS after ACTH, but we do not have such a phenomenon in the P-S reaction.The data in this group are admittedly small in number and rather scattered,but we feel that these data are at least internally consistent and indicatea differential steroid excretion.

In figure 11 we have a scatter of all ACTH determinations (samples obtainedin Korea as well as in Japan), where the Na/K ratio is plotted againstthe 17-KS mg. per 100 mg. of creatinine. We observe here a correlationbetween the Na/K ratio and the 17-KS in terms of 100 mg. creatinine, of-0.60 and this is significant to better than 0.001.

FIGURE11. Relation of urinary Na/Kratio with 17-ketosteroids mg./100 mg. creatinine of all post-ACTH samplesof all groups studied in Korea as well as Japan.

In figure 12 we have all the 17-KS determinations on which we have P-S.The scatter diagram represents P-S mg. per hour against 17-KS mg. per hour.We note two features in these data: (1) From about 0.3 mg. of the 17-KSper hour excretion to about 0.7 mg., we find the distribution of thesepoints such that a small increase in 17-KS is related to a greater increasein the P-S. However, we note that after this critical point of somewherebetween 0.7 and 0.8 mg. per hour, the 17-KS continue to increase but theP-S titer decreases. We feel that some fundamental aspect of steroid metabolismis revealing itself here with respect to stress.

FIGURE12. The relation of the 17-ketosteroidand Porter-Silber chromogens determined on same samples.

In figure 13 we plotted those samples which we considered as controls.They represent the Korean controls and C of George Co., and the ACTH inboth these groups. We note that in this chart the solid dot representsall pre-ACTH points and the circled dot indicates post-ACTH test samples.We notice here that the distribution of the points between 0.3 mg. and0.7 mg. is in large portion if not practically all represented by controlvalue, in which case we have small 17-KS increases with very large P-Sincreases (fig. 12). Figure 14 depicts all samples which are consideredas after-combat samples. We note here that the 17-KS values are high whencompared with P-S.

FIGURE13. The distribution of samples considered ascontrols on which both 17-ketosteroids and Porter-Silber chromogens wereobtained.

FIGURE14. Scatter of the combat samples on which both17-ketosteroids and Porter-Silber values were obtained.

Discussion

In the comparison of an acute stress against that of a chronic stresssituation, a large number of indices indicate that the biochemical

status, or say biochemical profile, of a chronic battle situation isquite different from that of the acute. Where, in the acute state thereis an indication of increased steroid output with increased protein catabolism,we find that in the chronic state (as far as the steroid measures are concerned)there is a dulling of the adrenal cortical function and no protein catabolism.However, it should be clear from the electrolyte data observed in the chronicstress situation after ACTH that the adrenal is not completely non-responsive;it is non-responsive as far as 17-KS and P-S are concerned, but it doesrespond to ACTH where some hormone apparently related to the electrolytefunction is stimulated. It may be inferred that whatever hormone isbeing stimulated after ACTH (acting on the electrolytes) in the chronicgroup, these hormones are not the compounds detectable in 17-KS titer orin the P-S titer. In the pre-ACTH psychiatric group we see that 17-KS andthe P-S indicate a lower adrenal cortical function level, but with somedegree of electrolyte action present. Furthermore, when ACTH is given inthe psychiatric group, there is a marked 17-KS response, but the P-Sis not very great, as a matter of fact it is clearly less than normal.The electrolyte response is very sharp. The lowest electrolyte ratios areobserved in the psychiatric group. Reviewing the data, then, each particulargroup, namely, the controls, the acute group, the chronic group and thepsychiatric group, has a different "biochemical profile" andcan be differentiated one from the other (table 4).

Table 4. Summary

Control data		Acute	Chronic	Psychiatric
Urine 17-KS	0.64 mg./hr	(+)	(-)	(-)
P. S. Chrom	0.51 mg./hr	(+)	(-)	(-)
Na/K (molar)	3.0	(-)	(+)	(-)
Urea	450 mg. urea N/hr	(+)	n . . .	(-)
Uric acid	30 mg./hr	(+)	n . . .	n . . .
ACTH 17-KS	(+)	(+ +)	(-)	(+ +)
P. S. Chrom	(+ +)	(-)	(-)	(-)
Na/K	(-)	(-)	(-)	(- -)

As far as the consequences to individuals having these various profilesare concerned: (1) The acute stress pattern is in favor of the individual,especially with respect to the possibility of an added insult to the organismsuch as a physical wound. The organism here has its

adrenals quite responsive. The glands are alert, and in case of anyphysical injury, the organism can very readily handle the result of physicalwounds, such as loss of blood, etc. (2) In the chronic stress situationthis is not the case. The individual's general physiologic condition indicatesthat with an added stress, such as a physical wound, the gland may be unableto handle its role in protecting the organism. The adrenal seems to benon-responsive, especially with regard to C 21 compounds. These are thecompounds (F type) which are very important in resistance and adaptationto stress.

In the psychiatric group we find that there is a low 17-KS output anda low P-S in the pre-ACTH, but in the post-ACTH values there is a markedresponse of 17-KS. This marked increase in 17-KS, with the more or lessfeeble P-S reaction, encourages us to make the following inference: Eitherthe compounds in this particular group are not C 21 in their origin, butC 19, or the C 21 compounds are rapidly converted to C 19. The C 19 ingeneral are mostly androgenic in character and do not have a protectiveeffect in a stress situation. Furthermore, since the steroid excretionis low in the psychiatric group even though the individuals showed markedemotional display and apparent stress, the pituitary possibly has beenblocked and is unable to secrete normal amounts of ACTH. We may make theinference that in all probability increased adrenalin secretion could beresponsible for this blockage. Thorn has shown that a continued infusionof the adrenalin does cause a blockage of the pituitary-adrenal axis. Adrenalinmetabolism in conjunction with adrenal steroid metabolism is an area ofstudy which should be investigated thoroughly, especially with respectto psychiatric breakdown.

As plausible as these points seem, the final proof of these variousstatements will depend on the chromatography of 17-KS, where the individualsteroids will be separated out and quantitatively analyzed. We are gettingto know to a great extent what the precursor of each compound (17-KS) inthe urine is, and in this way can give a quantitative estimate of the varioussteroids presumed to be secreted by the adrenal glands.

A Final Speculation. From the data the following hypothesis maybe offered. The adrenal gland on the first impulse, that is, in the firststages of stress, produces compound F (C 21) in large amounts. (It doesproduce some C 19 as well.) However, as the stress continues, we find thateither the organism ceases to produce C 21 type of compound and producesin majority C 19, or that C 21 compounds are converted to C 19 more rapidly.These phenomena could have purposive explanations. (1) Since C 21 compounds(such as "F") are catabolic, it is clear that if the catabolismcontinued over a long period of time the organism would lose considerablenitrogen and deteriorate into a meta-

bolic disturbance simulating overdosage of "F." However, withthe switch from C 21 to C 19 types (and these compounds are anabolic intheir general metabolic effect) we find here a feed-back mechanism, wherethe excess protein catabolism is now counteracted by the steroids of theC 19 type which are anabolic, whereby the organism now is better able toprotect itself. (2) If "F" were produced in excess over a longperiod of time, it would block the pituitary. The testing of this hypothesiswill depend in great measure on the chromatographic results which shouldbe completed within 6 months.

References

1. A Study of Combat Stress In Korea, 1952. PreliminaryReport. Technical Memorandum ORO-T-41 (FEC).

2. Pincus, G.: The Analysis of Human Urines for SteroidSubstances. J. Clin. Endocrinol. 5: 291,1945.

3. Reddy, W. J., Jenkins, D., and Thorn, G. W.: Estimationof 17-hydroxycorticoids In Urine. Metabolism 1: 511, 1952.

4. Pincus, G., Romanoff, L. P., and Carlo, J.: (Unpublished.)

5. Pincus, G.: A Diurnal Rhythm in The Excretion of UrinaryKetosteroids by Young Men. J. Clin. Endocrinol. 3:195,1943.