Aspects of personality of soldiers presenting to an endoscopy clinic

ASPECTS

OF PERSONALITY OF SOLDIERS TO AN ENDOSCOPY CLINIC

K.C.M.

WILSON,

R. WHITEOAK.

M.

PRESENTING

DEWEY and J.P. WATSON

Abstract-Eysenck Personality Questionnaires were completed hy 11consecutive series of 60 soldiers attending a military endoscopy clinic. Ewzh individual was pair-matched with :I non-endoscoped soldier from a medical ward ;md ;t soldier from the hospital alcohol trwtment unit. Alcohol dependence and related disabilities were measured wing the MAST. CAGE and SADQ questionnaires and Ilver function tests and MCV estimate\ were performed. Alcohol questionnaire score\ and laboratory te\t\ were analqsed by discriminant function analysis and by applying ‘cut-off’ scores of the questionnaire\. Analyses of the perwnality dimensions were carried out by analy,ses of variance and were examined in relation to sample. alcohol que\tionn;Gre scores and endoscoplc tmdings. The soldier\ from the alcohol treatment unit had the highest neuroticism score. followed hy the wldiers attending the endoscopq clinic. Of the endoscoped soldiers. those diagnosed ;IS non-ulcer ciy\prptics had higher neuroticism scores than soldiers with ulcers. This finding was not associated wth increawd prevalence of alcohol dependence and related disabilities.

INTRODUCTION AIM of this study is to examine the relationship between personality dimensions as described by Eysenck and the presentation of patients with gastrointestinal symptoms to an endoscopy clinic. It is now recognised that a third to a half of outpatients in gastroenterological departments are classified as having functional disorders [I. 21. Johnson et rrl. [3] has recently demonstrated a close association between functional abdominal disorders (irritable colon) and psychological and social problems. More specifically. Alexander [4] has studied the relationship between personality and peptic disease and postulated that ulcer sufferers exhibited a dependency need and a wish to be loved and cared for. Apart from Wolf and Wolff’s study [5] demonstrating the blanching effect of specific emotions on the gastric mucosa, most early studies were of a descriptive or psychoanalytic nature. In 1964, Kanter and Hazelton [6] studied 35 duodenal ulcer patients and demonstrated a tendency for these patients to have higher neuroticism and lower extraversion scores than a control sample as measured by the Maudsley Personality Inventory. Viscum [7] supported the finding of increased neuroticism in similar groups of patients. A more recent study [8] has shown that duodenal ulcer patients had higher introversion and psychoticism scores (on the EPQ) than controls. Gastric ulcer patients had higher psychoticism and trait anxiety scores than controls and higher neuroticism. psychoticism and hostility scores than the duodenal ulcer group.

THE

British Military Ho\pitnl. Royal Army Medwl Corps. Correspondence should he :iddrcs\ed to Dr K.C.M. Wilwn, Liverpool Hospital. Prescot Street. Liverpool. U.K.

Department

of P\ychoperiatric\.

Royal

86

K. C. M.

WII.SON

PI 01.

Roberts [9] examined a group of patients presenting with gastrointestinal symptoms in the absence of organic pathology (functional dyspepsia) and found them to have an abnormally high degree of neuroticism compared with asymptomatic, disease-free controls. Similar results amongst alcohol abusers who had dyspeptic symptoms were found when compared with non-dyspeptic abusers. Roberts suggests that personality influences the expression of dyspepsia. even in those cases where there is recognisable organic pathology. In a large study of American service personnel in training, Sandberg and Blidding [IO] drew attention to the psychological differences between patients with gastrointestinal symptoms presenting to the doctor and non-presenting symptomatic patients. The authors suggested that soldiers presenting with their symptoms do so as a result of psychological ‘decompensation’ under stress and subsequent increased alcohol ingestion. This study examines the relationship between upper gastrointestinal disease, aspects of personality dimensions and alcohol dependency and related disabilities. The study was carried out in a Military Hospital on serving male soldiers attending for the assessment and follow-up of gastrointestinal disease and related symptoms.

METHOD A consecutive series of60 soldiers attending an endoscopy clinic was examined. Each soldier was pairmatched with a non-endoscoped inpatient from the hospital medical wards and an inpatient from the alcohol treatment unit so that comparisons between drinking habits could be made in relation to the criteria for admission to the alcohol treatment unit. The age of each individual was matched to within 4 yr of the endoscoped patient with whom he was paired. Rank was matched directly but in a minority of cases had to be matched to broader subgrouping\ (senior officers, junior officers, senior non-commissioned officers, junior non-commissioned officers and privates). Duration of service was defined by the total amount of time served by the individual and was matched to the nearest year. The domestic situation of each individual w’as taken into account so as to differentiate between those soldiers living in communal facilities (barracks) and those living in married accommodation. All personnel in the study were male and stationed in the U.K. Each subject was also matched with his pair in terms of the number of cigarettes smoked each day to the nearest ten. The project was described to the subject and verbal consent was gatncd prtor to data collection. The by the patient. He EPQ and an incorporated alcohol questionnaire. the CAGE [I I]. were completed then completed the Severity of Alcohol Dependency Questionnaire [ 121 and was given a structured interview based on the MAST [13]. The data collection took approximately 35 mitt and wa\ carried out by the same researcher in each case. Blood was taken for the measurement of aspartate transaminase. alanine transaminase. y-glutamyl transferase and the mean corpuscular volume. Written consent was asked for at the end of the interview and questionnaire completion. Exclusion criteria included the withdrawal of consent and the exclusion of blood results known to be contaminated by concurrent non-alcohol-related illness or medication. Each patient was interviewed within the first week of admission to the hospital and patients with a history of previous admissions to the alcohol treatment unit were excluded from the study. Endoscopy was carried out on the endoscopy clinic sample by the clinical team responsible for the management of the patient. Diagnosis was made macroscopically. Biopsy and microscopic examination were not indicated in any of the sample cases. The findings were classified according to evidence of acute or healing ulcer disease, inflammatory disease (both suh categorized by the site). and those symptomatic patients with no organic pathology evident on macroscopic examinatton. In the statistical analyses. the variables screening for alcohol dependence and related dtsabilities were subjected to discriminant function analyses. using Wilk’s A as criterion in order that the three samples of patients could be compared. The results of the MAST. CAGE and SADQ were also analysed with reference to ‘cut-off scores’ standardised on civilian populations. The Eysenck Personality Questionnaire scores were analysed by analyses of variance.

Personality

of endoscoped

87

soldiers

RESULTS

From a consecutive series of 60 subjects interviewed from the endoscoped population, 55 complete sets of results were obtained. Of the 60 non-endoscoped control patients from the medical wards, three withdrew consent, seven had incomplete laboratory investigations and three sets of blood investigations were excluded from the study due to non-alcoholic liver disease known to affect the results. In the matched Index sample from the alcohol treatment unit, 58 subjects were interviewed, two patients refused consent and ten cases had incomplete laboratory investigations. One hundred and forty-eight patients had complete sets of results and were included in the discriminant function analyses. TABLE 1.-S

I A IMICAI

ANALYSES

(Function I) ATU sample vs medicaliendoscopy sample Eigen value Canonical correlation Wilk’s h Y?

Alcohol

O.YY2 0.706 0.489 101.87

(Function 2) Endscopy sample vs medical sample 0.0268 0. I62 0.974 3.7673

dependency

A stepwise discriminant function analysis was employed using Wilk’s A as criterion. After six steps, all variables except alanine transaminase had been included: From the scatter plot diagram two functions were identified: function 1 concerned the relationship between the alcohol treatment unit sample and the other two samples and function 2 concerned the relationship between the endoscoped sample and the medical inpatient control sample. In the analyses of the data concerning function 1 the eigen value and canonical correlation were high and Wilk’s A low, indicating that the sample from the alcohol treatment unit could be differentiated by these tests from the other two samples. In the analyses of function 2. the endoscoped sample could not be differentiated from the non-endoscoped medical inpatient control sample in terms of fulfilling admission criteria to the alcohol treatment unit. When the cut-off scores were applied to the MAST, CAGE and SADQ similar proportions of subjects in the endoscoped and non-endoscoped medical control samples were classified as ‘alcoholics’. Substantially fewer scored above cut-off than in the sample from the alcohol treatment unit. From the discriminant function analyses and the application of recognised cut-off scores of the instruments that we used, the endoscoped sample did not differ from the non-endoscoped medical patient sample in terms of the prevalence of dependency, related disabilities or in the number of patients that fulfilled the criteria to be admitted to a military alcohol treatment unit. However, both these samples were differentiated from the alcohol treatment unit sample by the discriminant function analyses, the number of individuals recognised as being ‘alcoholics’ and the severity of dependency.

XX

K. C. M. WII.~OU et t/l. T:INLL II.--PKOBf.L11

DRINKERS

CAGE Score = >2

AhD

SFVk.Rf

f \r OF DEPENfIt\.(.Y

Instruments MAST Score = >i

ARRAhlCiED

BY SAhlf’LF

SADQ 15-30

Score < I5

3Of

Sample EndOSCOpq ,I = 60

20 (33’%,)

Eysenck Personality The independent analysis of variance means.

30

53

(XX’%, )

(50%)

Questionnuire unulyses dimensions represented in the EPQ were followed by planned pair-wise comparison

T~\BI F Ill.-MLAI\S

OF EPQ

DIMFNSIONS

REf’RtSFNlED

5

7

( X% )

(GO)

subjected between

to an group

RY GROUP

Group Dimension N

P

ti

L

Endoscon\

Medical

Y.2 All wmple means differ slgnlficantly. F = X.60 with Z and IX0 degrees of freedom. ~1 = 0.0005

11.3

3 0 3.1 No significant diftrrence hetwceen wmpb F’ = 0.23 with 2 and IX0 degree\ of freedom. p = (1.7

ATU

13.6

: h

I I.5 11.3 13.7 AI‘U sample ditfers significantly from the endovxped and medical wmplo. F = I.Y5 with 2 and IX0 degrees of freedom. ,J = 0.00X 7.x 7.0 No significant difference between urnpIe\. b = 1.75 with 2 and IX0 degree% of freedom. /> = 0.2

7.5

Significant differences between all groups were demonstrated for the neuroticism scores 0, = 0.0005). Patients from the alcohol treatment unit had the highest. followed by the endoscoped sample, with the non-endoscoped medical inpatient sample having the lowest neuroticism score. Other significant differences were found in the extraversion score. the lowest value being in the patients from the alcohol treatment unit, significantly lower than both the endoscoped and medical inpatient samples (p = 0.008). Further analysis of the endoscoped sample was carried out to examine the relationship between endoscopic findings. alcohol ingestion and personality subgroups defined by endoscopic dimensions. Using an analyhis of variance, diagnosis were found to differ significantly on neuroticism (Table IV). No other subgroup differences were found.

Personality TABLE W-MEAN

of endoscoped

PERSONALIT\ DlMENSlONSC'oRESCiROUPEDRY

Diagnoses

II

Neuroticism

No pathologyiasymptomatic Functional dyspepsia Duodenal ulcer Duodenitis Gastritis Oesophagitis F value Significance

2 II 35 8 2 2

7.0 12.0 Y.X IS.2 16.5 15.0 2.57 0.37

Psychoticism I.5 2.7 2.8 3.4 6.5 5.0 1.20

‘.

8’)

soldiers

0.32

ENDOSCOPIC‘C'RIIERIA

Extraversion Il.5 1.5.4 13.Y 12.5 IX.0 13.0 0.7

Lie 7.(100 Y.X 7.x 6.4 5.0 1.5

0.64

I .6 I).IX

‘Significance at p < 0.0.5. In is recognised that the presence of non-erosive gastro-duodenitis is not an exclusion criterion for the diagnosis of functional dyspepsia [14-161. Patients with non-ulcerative inflammatory disease of the upper gastrointestinal tract were combined with the patients suffering from functional dyspepsia and contrasted with patients diagnosed as ulcer sufferers. A posteriori comparisons were made between the combined group of functional dyspeptics and those patients with diagnosed ulcer disease (Table V). TABLE V.-POOLED

Contrast Functional

vs ulcer

VARIANCE

ESI~IA-~~

Value

T value

- 13. I4

4.Y

T nrobabillt\ O.OM

The functional dyspeptic group was found to have a significantly greater neuroticism score than the ulcer group of patients and this was not correlated with an increase in alcohol dependency and related disabilities as measured by the instruments used in this study. DISCUSSION

The aim of this study has been to examine the relationship between symptoms of upper gastrointestinal pathology and the neuroticism dimension as described by Eysenck. The study also examines the influence of alcohol ingestion. Excessive alcohol ingestion has been shown to influence the degree of neuroticism [ 171 and may have a causative significance in peptic disease [IS]. By including a reference sample from the alcohol treatment unit we have been able to analyse the data concerning the alcohol-related problems in two ways: discriminant function analyses of the three samples showed that the nonendoscoped medical inpatient sample and the endoscoped sample were indistinguishable in terms of fulfiling the criteria to be admitted to the alcohol treatment unit. Using the instruments’ ‘cut-off scores’, 30% of the patients were CAGE positive and 50% were MAST positive in both the endoscoped and medical inpatient samples. Both these samples had a significantly lower prevalence of ‘alcoholics’ as defined by these questionnaires when compared with the alcohol treatment unit sample.

YO

K. C. M. WII so;v <‘I t/l.

Analyses of variance followed by pair-wise comparisons between the group means demonstrated that population samples from the alcohol treatment unit, the endoscopy clinic and non-endoscoped medical inpatients differed significantly in neuroticism scores. The alcohol treatment unit sample had the highest value, followed by the endoscoped sample, which was significantly greater than the nonendoscoped medical sample which approximated to the mean neuroticism score of non-hospitalised soldiers [ 191. A posteriori comparisons were carried out and the relationship between endoscopic diagnoses and neuroticism scores were examined. An increased neuroticism score was found in the functional dyspeptic group of patients when compared with those patients with ulcer disease. The increased neuroticism found in the endoscoped sample can not be explained by an increased prevalence of alcohol dependency or related disabilities. Within this sample of patients we have identified a group of patients presenting with functional dyspepsia who have higher neuroticism scores than patients with diagnosed ulcer disease. This diagnostic subgroup of patients does not have an increased prevalence of alcohol-related problems when compared with those with ulcer disease. We have demonstrated that there is an association between functional dyspepsia and neuroticism. confirming the work of Roberts [‘?‘I.We have also shown that this relationship cannot be explained by intcrmediatory variables of excessive alcohol ingestion, smoking habit and social status as has been found in other functional gastrointestinal disorders [3]. Functional dyspeptics represent a large proportion of the patients attending the clinic. The authors suggest that further research into the relationship between aspects of personality and presentation to the endoscopy clinic is indicated with the specific aim of developing therapeutic strategies designed to help this group of patients.

REFERENCES H,ZR\EY FRD. S.&I 111SY. RI_,\II AE. Organic and functional diwrder\ in X00 ga\trot‘ntcrolog) outpatients. Lnnwr IYX3: i: 637-631. FtRGLWN A. SIRRJS W. EASI~OOD MA. Frequency of functional g;r~tro-intenstinal disordrrs. Ltrrw/ lY77; i: 613-613. JOIIN~N 0. .I.~c~ortu~~ BJ, I’oRr)r OH. Aswclatlons betwen symptom\ of irritable colon and psychological and social conditions and lifestyle. Hr Mu/ .I; 292: lh33~16.15. AL~XANII~K F. The influence ot p\ychogcnic factors upon g~~\trointestin~~l di\turbancc\: general principles. objectives and preliminary results. f’\,whotrmr/ Quurf 1031; 3: 5lO-53Y. WOLF S. WOLFF HG. Hurmrr? (;crstr-ic F~rncrior~. Nell York: Oxford University Prtz\\. IY-l3. KAN,IER VB. HA/~I ION JE. An attempt to mensurc wm6 asprct of perwmilitv in \oune men with

TENNAN? C, GOIILSION K. L~~l)(i~lllr)nr(.Kr

P. Psychological correlate\ of gastric and duodrnal P,s,w120/ Mwl IYX6: 16: 3hS-371. ROHEWS DM. The psychogenic basis of dyspepsia in the ahscncc ot organic dlsraw. J fi Arrtiy r?ld c‘orp 1978; 124: 13-1X. SANDRERG B. BL.IIIIN(; A. Duodenal ulcer in al-my tralnee5 durinp basic mllitar! training. J Psyclmso,i~ Rrr 1076; 20: 6 l-7-1. ulcer dlwaw.

Personality

of endoscoped

91

soldiers

18.

HALL P. The CAGE questionnaire, validation of a new alcoholism Am J Psychiat; 131: 1121-I 123. ST-OCKWEUT. MURPHY D. HODC;SONR. The Severity of Alcohol Dependency Questionnaire. Its use. reliability and validity. Br J Addict 1983; 78: 145-155. SELZER M. The Michigan Alcoholism Screening Test. The quest for a new diagnostic instrument. Am J Pphiat 1971; 127: 89-94. GREEULAW R. SHEAHAN DG. DF Lt (‘A V. et ~1. Gastroduodenitis. A broader concept of peptic ulcer disease. Dig Dis Sci 1080; 25: 660-672. DE LUCA VA. WINNAN GG. SHEAHAN DG. er crl. Is gastro-duodenitis part of the spectrum of peptic ulcer disease? J c/in Gastroent 1081; 3: 17-22. TALLY NJ. PIPER DW. The association between non ulcer dyspepsia and other gastrointestinal disorders. Seem J Gastroenr 1985; 20: SY&YOO. MULLAN MJ. GURLING HMG, OPPENHEIMBE, MURRAY RM. The relationship between alcoholism and neurosis; evidence from a twin study. Rr J P.~ychiar 1986; 148: 4351141. BURRRIDCEEJ, LEWIS RD. HALS-VEDCH. Alcohol and the gastrointestinal tract. Mrti C/in North

I’).

Am 1984; 68: 77-s’). EYSENCK HJ. EYSENCK S. Munual

II.

MA~FIELD D, M~LECNI G.

screening

12. 13. 13. IS. 16. 17.

Sevenoaks.

instrument.

Kent:

Hodder

of the

and Stoughton.

Eysenck

1975.

Prrsonuliiy

Questionnaire

(Junior

and Adult).