Ruminative thinking in older inpatients with major depression

Journal of Affective Disorders 46 (1997) 273–277

Research report

Ruminative thinking in older inpatients with major depression a, a a b Jeffrey M. Lyness *, Yeates Conwell , Deborah A. King , Christopher Cox , Eric a D. Caine a

Department of Psychiatry, University of Rochester School of Medicine and Dentistry, 300 Crittenden Boulevard, Rochester, NY 14642, USA b Department of Biostatistics, University of Rochester School of Medicine and Dentistry, 300 Crittenden Boulevard, Rochester, NY 14642, USA Received 7 January 1997; received in revised form 3 April 1997; accepted 3 April 1997

Abstract Ruminative thinking, the tendency to dwell on particular ideas or themes, can be a prominent part of the phenomenology of major depression, but it rarely has been the focus of empirical research. We attempted to replicate (in adult psychiatric inpatients age $ 50 years with DSM-III-R major depression) the previously published finding that ruminative thinking was associated with melancholia and with psychosis. In our sample, these associations were not present. In addition, we explored the relationships of ruminative thinking to specific areas of thought content (e.g., suicidal ideation, somatic worry), cognitive function and overall functional status; ruminative thinking was not associated with suicidal ideation, but was associated with greater somatic worry and with poorer functional status, although these associations were not independent of overall depressive severity. A substantial proportion of subjects were unable to complete the cognitive measures; ruminative thinking was independently associated with inability to complete these tasks. We conclude that ruminative thinking is a meaningful and common clinical phenomenon among severely depressed older inpatients. Further investigations in inpatients and other populations examining its relationships to other phenomenology, to course and outcome, and to putative underlying mechanisms of depression are warranted.  1997 Elsevier Science B.V. Keywords: Major depression (phenomenology); Ruminative thinking; Functional status

1. Introduction Empirical research regarding the phenomenology of major depression both contributed to, and was stimulated by, DSM-III and its descendants. On *Corresponding author. Tel.: (1-716) 275-6741; fax: (1-716) 273-1082; e-mail: [email protected]

balance the recent DSMs have fostered consideration of clinical psychopathology, but a ‘side effect’ of their criteria-specific diagnostic view has been the tendency to disregard clinical phenomena not mentioned in them. Ruminative thinking in depression is an example of this investigative neglect. As reviewed by Nelson and Mazure (1985), ruminative thinking was defined as ‘the tendency of the patient to dwell

0165-0327 / 97 / $17.00  1997 Elsevier Science B.V. All rights reserved. PII S0165-0327( 97 )00068-2

274

J.M. Lyness et al. / Journal of Affective Disorders 46 (1997) 273 – 277

on one idea to the exclusion of other thoughts.’ Severity of ruminative thinking exists along a spectrum, from mild self-reported preoccupations to fixations on the ruminative idea that totally dominate one’s thoughts and conversations. Descriptions of ruminative thinking in melancholia date back many centuries (Jackson, 1986), but to our knowledge only one research group in the modern era has studied it empirically. Nelson and Mazure (1985), using a sample of psychiatric inpatients with DSM-III major depression, found that ruminative thinking was associated with the melancholic subtype of depression, consistent with their prior work demonstrating an association of ruminative thinking with depressive autonomousness (Nelson et al., 1981) and with delusional depression (Charney and Nelson, 1981). An association of ruminative thinking with melancholia might prove useful clinically, since melancholia itself may predict a different response to somatic and non-somatic therapies as compared with non-melancholic depression (Rush and Weissenburger, 1994). Yet no published studies since 1985 address this issue. Given this background, we attempted to replicate the work of Nelson, Mazure et al. by testing the hypothesis that ruminative thinking in older patients is associated with melancholia and with psychosis. In addition we planned analyses to explore the relationships of ruminative thinking to other clinical realms, hypothesizing that ruminative thinking would be associated with greater suicidal ideation and somatic worry, and greater cognitive and overall functional disability.

2. Methods As the patients in this study overlapped substantially with groups used in previous studies (Lyness et al., 1993a,b, 1995), the recruitment methods will be described only briefly here. All patients of age $ 50 years admitted to the inpatient psychiatric units at Strong Memorial Hospital between November 1990 and August 1993 with a primary diagnosis of DSMIII-R major depression were eligible for inclusion. [N.B.: Subjects judged to have organic mood disorder were excluded. However, patients with significant medical comorbidity not judged to be etiologic

for the depression were included, based on issues of sample generalizability (Lyness et al., 1996).] After complete description of the study to the subjects, their consent to participate was obtained using procedures approved by the University of Rochester School of Medicine and Dentistry’s Research Subjects Review Board. Within seven days of admission, semistructured interviews were administered to the patients by master’s or doctorate level raters trained by our research group. The assessments included the Hamilton Rating Scale for Depression (24-item version) (Ham-D) (Williams, 1988) and the Structured Clinical Interview for DSM-III-R (SCID) (Spitzer et al., 1986). DSM-III-R diagnoses, including diagnostic subtypes and age of depression onset, were assigned by consensus of all investigators and raters after presentation of the data and detailed case histories. Assessments of specific ideation were made using individual items from the Ham-D: the hypochondriasis item (Ham-Hypo) to assess somatic worry (c.f. Lyness et al., 1993b), and the suicide item (Ham-SI) for suicidal ideation. Cognition was assessed by the Mini-Mental State Exam (MMSE) (Folstein et al., 1975) and the Trail-Making B Test (Trails B) (Reitan, 1958). Psychiatric disability, as measured by the Global Assessment of Functioning Scale (GAF) from DSM-III-R, was determined at the consensus conference. Two examiner-rated scales assessed overall functional status, the Instrumental Activities of Daily Living (IADL) and Physical SelfMaintenance Scales (PSMS) (Lawton and Brody, 1969). As per the work of Nelson and Mazure (1985), ruminative thinking was rated as a dichotomous variable (presence or absence of observed ruminative thinking), as derived from their ordinal five point scale. Reliability on this dichotomous variable by our research group personnel was examined by percent agreement, based on observation of semistructured interviews with 7 patients by 2–5 raters per patient (six individual raters total). There was mean 6SD 89.5%613.8% agreement about the presence or absence of observed ruminative thinking among 21 rater pairs. Statistical analyses included t-tests (assuming unequal variances), Chi-square tests, and logistic and multiple linear or logistic regression techniques. For the regressions, a logarithmic transformation was

J.M. Lyness et al. / Journal of Affective Disorders 46 (1997) 273 – 277

used as necessary to make the errors more normally distributed and to stabilize the variance. Statistical significance was defined as p , 0.05.

3. Results Clinical data on the 124 patients enrolled in the study are presented in Table 1. Forty-eight (39% of the total sample) had ruminative thinking. While ruminative thinking was associated with higher score on the Ham-D, it was not significantly associated with melancholia or psychosis. As shown in Table 1, there was not a significant association of ruminative thinking with age or age of onset of depression. Ruminative thinking also was not significantly associated with Ham-SI (X 2 5 2.82, df 5 4, p 5 0.59). However, patients with ruminative thinking were significantly more likely to have somatic worry, as measured by the Ham-Hypo (X 2 5 20.6, df 5 4, p , 0.0001). Turning to functional variables, as shown in Table 1 ruminative thinking was significantly associated with slightly greater functional disability on the PSMS and GAF, although not with the instrumental activities measured by the IADL. Multiple regres-

275

sions were used to determine the association of ruminative thinking with the PSMS and GAF independent of melancholia and overall depressive severity (Ham-D). Ruminative thinking was associated with higher PSMS score independent of melancholia (F 5 3.80, df 5 2,121, one-tailed p 5 0.027) but not when controlled for Ham-D (F 5 0.00, df 5 2,120, one-tailed p 5 0.48). Ruminative thinking’s association with poorer GAF score was not significant when controlled for either melancholia (F 5 2.59, df 5 2,121, one-tailed p 5 0.055) or Ham-D (F 5 0.26, df 5 2,120, one-tailed p 5 0.30). As shown in Table 1, ruminative thinking was not significantly associated with the two cognitive measures, the MMSE or Trails B time. However, despite encouragement by the research raters, 26 patients (21% of the total) were unable to complete the MMSE, and 79 (64%) were unable to complete the Trails B. Reasoning that the ability to complete these cognitive tasks might prove more usefully discriminatory than the scores among those who did complete the tasks, we examined the association of ruminative thinking with inability to complete the MMSE and the Trails B. Ruminative thinking was strongly and significantly associated with inability to complete the MMSE (X 2 5 8.09, df 5 1, p 5 0.004)

Table 1 Ruminative thinkers vs. non-ruminative thinkers on clinical measures Variable

Ruminative thinkers n 5 48 Mean (S.D.)

Non-ruminative thinkers n 5 76 Mean (S.D.)

Age (years)1 Age of onset (years)2 Ham-D 3 IADL 4 PSMS 5 GAF 6 MMSE 4 Trails B time (s)4 Melancholia (present)7 Psychosis (present)8

72.9 (9.1) 58.0 (19.2) 35.5 (7.5) 11.6 (8.6) 4.1 (4.0) 29.7 (6.5) 25.5 (4.3) [n 5 29] 136.5 (58.4) [n 5 10] n 5 35 (73%) n 5 10 (21%)

70.3 (9.9) 57.7 (15.6) 29.0 (7.5) 9.7 (7.3) 2.9 (3.9) 31.8 (6.5) 25.6 (3.3) [n 5 69] 143.7 (90.7) [n 5 35] n 5 49 (65%) n 5 19 (25%)

1

t 5 2 1.45, df 5 106.8, p 5 0.15. t 5 2 0.07, df 5 78.8, p 5 0.95. 3 t 5 2 4.71, df 5 97.9, p 5 0.0001. 4 p 5 NS. 5 t 5 1.72, df 5 97.9, p 5 0.044. 6 t 5 1.78, df 5 99.3, p 5 0.038. 7 X 2 5 0.96, df 5 1, p 5 0.33. 8 X 2 5 0.29, df 5 1, p 5 0.59. 2

276

J.M. Lyness et al. / Journal of Affective Disorders 46 (1997) 273 – 277

and Trails B (X 2 5 16.38, df 5 1, p , 0.0001). These associations remained significant even when multiple regression techniques controlled for melancholia (MMSE: X 2 5 7.91, df 5 1, p 5 0.0049; Trails B: X 2 5 15.25, df 5 1, p 5 0.0001) and Ham-D (MMSE: X 2 5 7.20, df 5 1, p 5 0.0073; Trails B: X 2 5 6.25, df 5 1, p 5 0.0124).

4. Discussion Consideration of our results should be tempered by recognizing that, while the reliability of the ruminative thinking scale among our research group personnel was reasonable, it was not determined solely with patients or raters involved in this specific study. Also, we did not demonstrate the interrater reliability of the other study measures in use by our personnel. In the context of such limitations, our results confirm that ruminative thinking is a clinical phenomenon commonly found in severely depressed inpatients. Yet we were unable to confirm a significant association between ruminative thinking and melancholia. Among the differences between our data and those of Nelson and Mazure (1985) is that our group had a higher rate of ruminative thinking in non-melancholic patients. This may be partly explained by the older age (and associated depressive severity and medical comorbidity) of our group. As well, societal pressures in recent years to decrease utilization of acute hospital services may have led our sample to be more acutely ill, or disabled, than the sample of Nelson and Mazure. This might have affected our results in two ways: if ruminative thinking is associated with functional impairment as is partly supported by our data, then it might contribute to the decision to admit patients, raising its prevalence even among non-melancholic depressives; also, some patients in our severely ill sample had a relative paucity of verbal output, thus decreasing our ability to detect ruminative thinking among the sickest (perhaps especially among melancholic) patients. Ruminative thinking was not associated with psychosis, consistent with the notion that delusions, disturbances of thought content, are separable from the concept of rumination as a thought process

disturbance. Our measures do not allow definitive examination of the relationship of ruminative thinking to thought content, but the data do support the notion that somatic worry may be a focus of rumination. As hypothesized, and consistent with the notion that ruminative thinking is a meaningful part of the phenomenology of severe major depression, ruminative thinking was associated with psychiatric functional impairment, and with impairment in basic physical self-maintenance tasks, although these associations were not separable from those with overall depressive symptom severity. More striking was ruminative thinking’s strong independent association with inability to complete our two tests of cognitive function. This is consistent with, but does not provide direct empirical support for, the speculation that ruminative thinking may contribute to the cognitive deficits of depression by diverting attention from the task at hand (Nelson and Mazure, 1985). Further study of ruminative thinking and neuropsychological function in depression is warranted, although this may prove difficult to examine in more detail if many patients, as in our group, are unable to complete the testing. As well, cognitive psychology perspectives might suggest that ruminative thinkers would be more likely to suffer from comorbid or antecedent anxiety disorders such as panic disorder or generalized anxiety disorder. Other psychological perspectives raise the question of the association between ruminative thinking and obsessive–compulsive personality disorder, or perhaps even obsessive–compulsive disorder. These conditions were too infrequent among our older and severely depressed group for us to study, but such associations warrant study as part of future investigations into the relationships of ruminative thinking to other phenomenology, to course and outcome, and to putative underlying mechanisms of depression in various depressed populations.

Acknowledgements This study was supported in part by grants from the National Institute of Mental Health, including T32 MH18911 and K07 MH01113 (Dr. Lyness) and

J.M. Lyness et al. / Journal of Affective Disorders 46 (1997) 273 – 277

P50 MH40381. We also thank the staff of the Program in Geriatrics and Neuropsychiatry for technical and support services.

References Charney, D.S., Nelson, J.C., 1981. Delusional and nondelusional unipolar depression. Further evidence for distinct subtypes. Am. J. Psychiatry 138, 328–333. Folstein, M.F., Folstein, S.E., McHugh, P.R., 1975. ‘Mini-Mental State’: A practical method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res. 12, 185–198. S.W. Jackson, Melancholia and Depression. From Hippocratic Times to Modern Times, Yale University Press, New Haven, CT, 1986. Lawton, M.P., Brody, E.M., 1969. Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist 9, 179–186. Lyness, J.M., Caine, E.D., Conwell, Y., King, D.A., Cox, C., 1993. Depressive symptoms, medical illness and functional status in depressed psychiatric inpatients. Am. J. Psychiatry 150, 910– 915. Lyness, J.M., King, D.A., Conwell, Y., Cox, C., Caine, E.D., 1993.

277

‘Somatic worry’ and medical illness in depressed inpatients. Am. J. Geriatric Psychiatry 1, 288–295. Lyness, J.M., Cox, C., Curry, J., Conwell, Y., King, D.A., Caine, E.D., 1995. Older age and the underreporting of depressive symptoms. J. Am. Geriatr. Soc. 43, 216–221. Lyness, J.M., Bruce, M.L., Koenig, H.G., Parmelee, P.A., Schulz, R., Lawton, M.P., Reynolds, III C.F., 1996. Depression and medical illness in late life. Report of a symposium. J. Am. Geriatr. Soc. 44, 198–203. Nelson, J.C., Charney, D.S., Quinlan, D.M., 1981. Evaluation of the DSM-III criteria for melancholia. Arch. Gen. Psychiatry 38, 555–559. Nelson, J.C., Mazure, C., 1985. Ruminative thinking. A distinctive sign of melancholia. J. Affect. Disord. 9, 41–46. Reitan, R.M., 1958. Validity of the Trail Making Test as an indication of organic brain damage. Percept. Motor Skills 8, 271–276. Rush, A.J., Weissenburger, J.E., 1994. Melancholic symptom features and DSM-IV. Am. J. Psychiatry 151, 489–498. R.L. Spitzer, J.B.W. Williams, M. Gibbon, Structured Clinical Interview for DSM-III-R (SCID), Biometrics Research, New York, NY, 1986. Williams, J.B.W., 1988. A structured interview guide for the Hamilton depression rating scale. Arch. Gen. Psychiatry 45, 742–747.