Hostname: page-component-cd9895bd7-dk4vv Total loading time: 0 Render date: 2024-12-25T15:51:25.379Z Has data issue: false hasContentIssue false

Les limites de validité des scores dimensionnels de la Hopkins Symptom Checklist : à propos d’une analyse en composantes principales sur 457 femmes

Published online by Cambridge University Press:  28 April 2020

P. Cialdella
Affiliation:
Service du Dr Escolier, hôpital Saint-Jean-de-Dieu, 290, route de Vienne, 69008Lyon
F. Munoz
Affiliation:
Unité INSERM 265 sur les composantes physiques, psychiques et sociales de la santé, 151, cours A.-Thomas, 69008Lyon, France
N. Mamelle
Affiliation:
Unité INSERM 265 sur les composantes physiques, psychiques et sociales de la santé, 151, cours A.-Thomas, 69008Lyon, France
Get access

Résumé

La Hopkins Symptom Checklist-58 ou HSCL-58 (Derogatis et al., 1974) est un autoquestionnaire de 58 items largement employé au cours des essais de psychotropes et en épidémiologie psychiatrique pour l’évaluation des troubles névrotiques et affectifs. Cet instrument permet des scores sur 5 dimensions définies sur des bases théoriques et empiriques (analyses factorielles), dont la validité repose sur la stabilité de la structure factorielle au sein de différents échantillons. Or, la plupart des analyses factorielles de la HSCL-58 ont concerné des groupes de patients recrutés dans des centres de soin (groupes cliniques). Il importait donc de vérifier la stabilité factorielle sur des sujets névrotiques provenant d’une population tout-venant (groupes subcliniques).

Au cours d’une enquête épidémiologique sur les facteurs de risque de prématurité (Mamelle et al., 1987), menée dans 4 maternités de Lyon, 1 643 femmes enceintes de 6 mois avaient rempli un questionnaire psychopathologique comprenant 45 items extraits de la HSCL (la plupart des items manquants par rapport à la HSCL-58 n’appartenaient à aucune des 5 dimensions). Les évaluations d’avant-grossesse ont été utilisées pour définir un groupe ≪subclinique≫, en sélectionnant les femmes qui avaient obtenu les scores totaux à la HSCL (45 items) les plus élevés, ce score procurant une estimation de gravité névrotique. Le seuil de gravité ne pouvant être qu’arbitraire, nous avons décidé de retenir un ratio nombre de sujets/nombre de variables égal à 10. Au total, 457 femmes (27,8% de l’échantillon initial) ont été considérées dans l’analyse. Pour nous assurer de la proximité de ce groupe ≪subclinique≫ d’avec les groupes ≪cliniques≫, nous avons comparé les notes moyennes des items d’anxiété et de dépression de notre échantillon avant et après sélection, avec celles de 3 groupes décrits par Derogatis et al. (1974) : 2 groupes cliniques, un de patients névrotiques anxieux, un autre de déprimés névrotiques, et un troisième groupe de sujets normaux, représentatif de la population d’Oackland. Une analyse en composantes principales avec rotation varimax assortie d’une méthode de choix du nombre de facteurs décrite par Comrey (1978) a été utilisée. Les notes moyennes des items d’anxiété et de dépression de notre groupe ≪subclinique≫ de 457 femmes se sont révélées proches de celles des névrotiques anxieux de Derogatis, mais plus faibles que celles des déprimés névrotiques, et plus élevées que celles des sujets normaux (Tableau I). Une solution à 4 facteurs est apparue la meilleure: ≪vulnérabilité≫, ≪somatisation≫, ≪tension≫, et ≪troubles cognitifs≫ (Tableau II). Deux facteurs (somatisation et troubles cognitifs) reproduisent de près la structure attendue, le facteur vulnérabilité étant proche d’un facteur dépression, mais les items de sensitivité et d’anxiété n’ont pas formé les facteurs espérés.

Nos résultats confirment donc les données de la littérature concernant la stabilité relative des dimensions somatisation et obsession (ou plutôt troubles cognitifs), et l’instabilité des dimensions anxiété et sensitivité, le cas de la dépression étant intermédiaire. Il semble, en conclusion, que les scores dimensionnels de la HSCL-58 ne présentent pas une validité suffisante pour être employés dans les groupes subcliniques.

Summary

Summary

The Hopkins Symptom Checklist-58 or HSCL-58 (Derogatis et al., 1974) is a 58-item autoquestionnaire widely used in the field of psychopharmacological trials and in psychiatric epidemiology for the assessment of neurotic and affective disorders. This instrument allows scorings along five dimensions, defined on theoretical as well as empirical (factor analyses) bases, the validity of such scores resting on the factorial invariance of the HSCL across different samples. However, most of the previous factor analytic studies of the HSCL-58 have been carried out in samples drawn from treatment facilities (clinical samples). Therefore, it was important to verify the factorial invariance among neurotic subjects drawn from unselected populations (sub-clinical samples).

During an epidemiological study on the risk factors of premature birth (Mamelle et al., 1987) conducted in 4 maternity hospitals in Lyon, 1643 women at the sixth month of pregnancy had completed a psychopathological questionnaire composed of 45 items from the HSCL-58 (most of the items lacking from the 58-item form did not belong to any of the 5 principal dimensions). The pre-pregnancy assessments were used in order to define a “sub-clinical” sample, by selecting the women who had obtained the highest total scores to the HSCL (45 items), this score being used as an index of neurotic suffering. The severity threshold being obviously arbitrary, we decided to retain a ratio number of subjects/number of items equal to 10. Finally, 457 women (27,8% of the initial sample) were considered for the factor analysis. In order to verify the similarity of this sub-clinical sample with clinical ones, we compared the Anxiety and Depression item mean scores of our sample, before and after selection, with those of three samples described in Derogatis et al. (1974): 2 clinical samples, one of anxious neurotics and another of neurotic depressed patients, and a third sample of normal subjects representative of the population of Oakland. A Principal Component Analysis with Varimax rotation and a method of selection for the number of factors described by Comrey (1978) were performed.

The Depression and Anxiety item mean scores of our “sub-clinical” sample were found to be close enough to those of Derogatis’ anxious neurotics, but lower than those of Derogatis’ neurotic depressed patients and higher than those of normal subjects (Table I). A 4-factor solution seemed to be the best: “Vulnerability”, “Somatization ”, “Tension “ and “Cognitive Disturbances” (Table II). The Somatization as well as the Cognitive Disturbances factors reproduced adequately the expected structure whi/e the Vulnerability factor was close to a Depression factor, but the Sensitivity and Anxiety intems did not constitute the expected factors.

Our results confirm previous data from the literature about the relative stability of the Somatization and Obsession-Compulsion (which would be better termed Cognitive Disturbances) dimensions, and the unstability of Anxiety and Sensitivity dimensions, while the case of the Depression dimension remains intermediate. In conclusion, it seems that the dimensional scores of the HSCL-58 do not present sufficient validity for their use in sub-clinical groups.

Type
Article original
Copyright
Copyright © European Psychiatric Association 1989

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Références

Angst, J., Dobler-Mikola, A. & Binder, J. (1984) The Zurich study : a prospective epidemiological study of depressive, neurotic and psychosomatic syndromes. I. Problem, methodology. Eur. Arch. Psychiatr. Neurol. Sci. 234, 1320CrossRefGoogle ScholarPubMed
Comrey, A.L. (1978) Common methodological problems in factor analytic studies. J. Consult. Clin. Psychol. 46, 648659CrossRefGoogle Scholar
Cyr, J.J., McKenna-Foley, J.M. & Peacock, E. (1985) Factor structure of the SCL-90-R : Is there One? J. Pers. Assess 49, 6, 571577CrossRefGoogle Scholar
De Bonis, M. (1974) Content analysis of 27 anxiety inventories and rating scales. In: Psychological measurements in psychopharmacology. Mod. Prob. Pharmacopsychiatry. 7, 221237CrossRefGoogle Scholar
Derogatis, L.R., Lipman, R.S., Covi, L., Rickels, K. & Uhlenhuth, E.H. (1970) Dimensions of outpatient neurotic pathology: comparison of a clinical versus an empirical assessment. J. Consult. Clin. Psychol. 34, 164171CrossRefGoogle Scholar
Derogatis, L.R., Lipman, R.S., Covi, L. & Rickels, K. (1971) Neurotic symptom dimensions as perceived by psychiatrists and patients of various social classes. Arch. Gen. Psychiatry 24, 454464CrossRefGoogle ScholarPubMed
Derogatis, L.R., Lipman, R.S., Covi, L. & Rickels, K. (1972) Factorial invariance of symptom dimensions in anxious and depressive neuroses. Arch. Gen. Psychiatry 27, 659665CrossRefGoogle ScholarPubMed
Derogatis, L.R., Lipman, R.S. & Covi, L. (1973) SCL-90: An outpatient psychiatrie rating scale-preliminary report. Psychopharmacol. Bull. 9, 1328Google Scholar
Derogatis, L.R., Lipman, R.S., Rickels, K., Uhlenhuth, E.H. & Covi, L. (1974) The Hopkins symptom Checklist (HSCL): A measure of primary symptom dimensions. In : Psychological measurements in psychopharmacology. Mod. Prob. Pharmacopsychiatry 7, 79110CrossRefGoogle Scholar
Derogatis, L.R. & Cleary, P.A. (1977) Factorial invariance across gender for the primary symptom dimensions of the SCL-90. Br. J. Soc. Clin. Psychol. 16, 347356CrossRefGoogle ScholarPubMed
Gerin, P. (1983) Traduction de la HSCL-58 (unpublished paper)Google Scholar
Goldberg, D. & Simpson, N. (1986) The diagnosis of anxiety in primary care settings. In: Drug Treatment of Neurotic Disorders: Focus on Alprazolam. (Lader, M.H. & Davies, H.C., eds.) Churchill Livingstone, Edinburgh, pp. 7683Google Scholar
Guelfi, J.D., Barthelet, G., Lancrenon, S. & Fermanian, J. (1984) Structure factorielle de la HSCL sur un echantillon de patients anxio-dépressifs français. Ann. Med. Psychol. 142, 6, 889898Google Scholar
Harman, H.H. (1976) Modem Factor Analysis. University of Chicago Press, ChicagoGoogle Scholar
Kahn, R.J., McNair, D.M., Lipman, R.S., Covi, L., Rickels, K., Dawning, R., Fisher, S. & Frankenchaler, L.M. (1986) Imipramine and chlordiarepoxide in depressive and anxiety disorders. IL Efficacy in anxious outpatients. Arch. Gen. Psychiatry 43, 7985CrossRefGoogle Scholar
Lipman, R.S., Rickels, K., Covi, L., Derogatis, L.R. & Uhlenhuth, E.H. (1969) Factors of symptom distress: doctor ratings of neurotic outpatients. Arch. Gen. Psychiatry 21, 328338CrossRefGoogle ScholarPubMed
Lipman, R.S., Covi, L., Rickels, K., McNair, D.M., Downing, R., Kahn, R.J., Lasseter, V.K. & Faden, V. (1986) Imipramine and chlordiarepoxide in depressive and anxiety disorders. I. Efficacy in depressed outpatients. Arch. Gen. Psychiatry 43, 6877CrossRefGoogle ScholarPubMed
Mamelle, N., Gerin, P., Measson, A., Munoz, F. & Collet, P. (1987) Assessment of psychological modifications during pregnancy : contributions of Derogatis Symptom cheklist (SCL-90-R). J. Psychosom. Obstet. Gynaecol. 7, 3950CrossRefGoogle Scholar
Mattsson, N.B., Williams, H.V., Rickels, K., Lipman, R.S. & Uhlenhuth, E.H. (1969) Dimensions of symptom distress in anxious neurotic outpatients. Psychopharmacol. Bull. 5, 1931Google ScholarPubMed
Myers, J.K., Weissmann, M.M., Tischler, G.L., Holzer, C.E., Leaf, P.J., Orvaschel, H., Anthony, J.C., Boyd, J.H., Burke, J.D., Kramer, M. & Stoltzman, R. (1984) Six-Month prevalence of psychiatrie disorders in 3 communities. Arch. Gen. Psychiatry 41, 959967CrossRefGoogle Scholar
Mullaney, J.A. (1984) The relationship between anxiety and depression : a review of some principal component analytic studies. J. Affective Disord. 7, 139147CrossRefGoogle ScholarPubMed
Nunnaly, J.C. (1978) Psychometric Theory. McGraw-Hill, New YorkGoogle Scholar
Roth, M., Gurney, C., Garside, R. & Kerr, T. (1972) The relationship between anxiety States and depressive illnesses. Br. J. Psychiatry 121, 147167CrossRefGoogle ScholarPubMed
Uhlenhuth, E.H., Balter, M.B., Mellinger, G.D., Cisin, I.H. & Clinthorne, J. (1983) Symptom Checklist Syndromes in the general population. Arch. Gen. Psychiatry 40, 11671173CrossRefGoogle ScholarPubMed
Walkey, F.H. & McCormick LA. (1985) Multiple replication of factor structure : a logical solution for a number of factors problem. Multivar. Behav. Res. 20, 5767CrossRefGoogle ScholarPubMed
Weissman, M.M. & Klerman, G.L. (1977) Sex differences and the epidemiology of depression. Arch. Gen. Psychiatry 34, 98111CrossRefGoogle ScholarPubMed
Williams, H.V., Lipman, R.S., Rickels, K., Covi, L., Uhlenhuth, E.H. & Mattsson, N.B. (1968) Replication of symptom distress factors in anxious neurotic patients. Multivar. Behav. Res. 3, 199212CrossRefGoogle Scholar
Zwick, W.R. & Velicer, W.F. (1986) Comparison of five rules for determining the number of components to retain. Psychol. Bull. 99, 3, 432442CrossRefGoogle Scholar
Submit a response

Comments

No Comments have been published for this article.