Mania Compared With Unipolar Depression in Old Age

Transcription

1 Mania Compared With Unipolar Depression in Old Age Kenneth I. Shulman, M.D., S.M., Mauricio Tohen, M.D., Dr.P.H., Andrew Satlin, M.D., Gopmath Mallya, M.D., and Douglas Kalunian, M.D. Objective: The goal of this study was to clarify the meaning and importance of mania in old age. Method: The authors conducted a retrospective study ofso elderly patients consecutively admitted to a private mental hospital with an index episode ofmania. As a comparison group, they used 50 age- and sex-matched patients with unipolar depression. They reviewed the charts of the I 00 patients for family history, clinical course, and neurological disorders. Outcome was determined by contacting patients, families, physicians, institutional settings, and vital statistics records. Survival analysis compared mortality rates. Results: The manic patients had a greater familial predisposition to affective disorder and were younger at first psychiatric hospitalization. For the 20 manic patients whose first affective episode was depression, an average of 1 5 years elapsed before mania became manifest. Eighteen of the manic patients, compared with only four of the depressed patients, had neurological disorders. The manic patients had a significantly higher mortality rate than the depressed patients; by the end of the follow-up, 25 of the manic patients, compared with 1 0 of the depressed patients, had died. Conclusions: Mania appears to have a poorer prognosis and to be a more severe form ofaffective illness than unipolar depression. The 1 8 manic patients with neurologi cal disorders seemed to have secondary mania. Subtle cerebral changes due to aging may have been responsible for the conversion to mania in the 20 patients who experienced a long latency from f irst depression to onset ofmania. The low frequency ofearly-onset mania in this study group highlights the need to differentiate early- from late-onset mania. (Am J Psychiatry 1992; 149: ) T he incidence of first admissions for mania tends to increase with advancing age (1, 2), yet the meaning and importance of mania in old age remains unclear. In contrast, the prevalence of bipolar disorder in the community decreases with age from 1.4% in young adults to 0.4% in those older than 65 years (3). For many elderly manic patients, a latency period averaging as long as 17 years separates the first episode of depression and the onset of mania (4-6). Furthermore, numerous depressive episodes often occur during the latency period. Even among young patients with unipolar depression, Akiskal et al. (7) found that 20% experienced a manic episode within a 6-year follow-up. Therefore, Presented at the 144th annual meeting of the American Psychiatric Association, New Orleans, May , Received Jan. 25, 1991; revision received July 22, 1991; accepted Aug. 29, From McLean Hospital, Belmont, Mass., and the Department of Psychiatry, Harvard Medical School, and the Department of Epidemiology, Haryard School of Public Health, Boston. Address reprint requests to Dr. Shulman, Department of Psychiatry, Sunnybrook Health Science Centre, 2075 Bayview Ave., Toronto, Ont. M4N 3MS, Canada. Supported in part by a grant from the Ontario Mental Health Foundation (Dr. Shulman) and by grant AG from the National Institute on Aging (Dr. Satlin). Copyright C) 1992 American Psychiatric Association. the concept of a unipolar-bipolar dichotomy of affective disorder (8) has been challenged by spectrum and threshold theories that consider mania a more severe form of affective disorder (9, 10). The average age at onset of affective disorder has been reported to be in the late 40s or SOs in elderly manic patients. Indeed, few elderly subjects with bipolar disorder have suffered from mania before the age of 30 (5). Age at onset may distinguish subgroups of manic patients with different clinical and biological features (11). In both mania and depression in old age, a good deal of attention has been directed at the role of cerebral-organic factors. Retrospective studies have found an association between coarse brain disorders and mania in old age (4, 5, 12). In a recent prospective study, Broadhead and Jacoby (6) used CT scans to study elderly as well as young patients with mania. They concluded that organic factors may play an important role in the genesis of affective disorder in old age. Based in part on the earlier findings of Shulman and Post (4), we hypothesized that, compared with depression in old age, mania in old age 1) is a more severe form of affective disorder, 2) is associated with a preponderance of neurological disorders, and 3) has a poorer prognosis. Usually, depression in old age has Am J Psychiatry 149:3, March

2 MANIA AND DEPRESSION IN OLD AGE been studied separately from mania (13, 14). To our knowledge, this is the first reported study comparing elderly manic patients with age- and sex-matched patients with unipolar depression. METHOD Using medical record data, we conducted a chart review of all patients discharged from a private mental hospital in to identify SO consecutive admissions of patients 65 years old or older who met DSM- III-R criteria for a manic episode. Since the outcome data were recorded in 1 990, the minimum follow-up was 3 years and the maximum was 10 years. Subjects with medical problems or cerebral-organic disorders were not excluded. For each year s group of manic patients, a comparison group of age- and sex-matched patients who met DSM-III-R criteria for major depression were selected. In this way, SO elderly patients with mania and SO elderly patients with major depression were included in the study. To select a homogeneous group of depressed patients, patients suffering from bipolar depression at index episode were excluded. Information was recorded regarding demographic data, family history of affective disorder, and previous psychiatric and neurological history. All geriatric patients at the hospital were routinely assessed by a neurologist. We recorded the presence of a neurological disorder only when the evidence was clearly documented by the clinical consultation note or by a radiologic report. When there was any uncertainty as to the presence or absence of a neurological disorder, consensus was reached by the two raters (K.I.S. and G.M.). To assess the reliability of the information collected from medical records regarding family history of affective disorder, type of index episode, and presence of preexisting neurological disorder, we compared the ratings of 12 randomly selected patients with blind ratings of the same patients made by one of us (M.T.). For all three variables there was full agreement on the 12 patients. The number and type of subsequent psychiatric hospitalizations were determined by contacting patients, relatives, physicians, and appropriate institutional settings. The dates of any hospitalization, institutionalization, or death occurring during the follow-up period were recorded. Institutionalization was defined as any placement in a supervised residential setting, including rest homes, homes for the aged, nursing homes, or chronic care settings. Treatment issues were not controlled in this naturalistic study. To estimate the effects of risk factors simultaneously, logistic regression models (15) and the computerized Statistical Analysis System (SAS) (16) were used. We obtained adjusted odds ratios with 95% confidence intervals. Survivalcurves based on the Kaplan-Meier method (17) were used to estimate time to psychiatric rehospitalization, institutionalization, or death. We obtained adjusted hazard ratios (1 8) and their 95% confidence intervals, controlling for other variables, by using the SAS (19). RESULTS Comparison of Patients With Mania and Depression Each group was composed of 35 (70%) women and 15 (30%) men. The mean age at index admission of the women in the mania group was 70 years (SD=4.S, range=65-83); the mean age of the women in the depression group was also 70 (SD=4.1, range=65-79). The mean age at index admission of the men in the mania group was 74 years (SD=7.0, range=65-85); the mean age of the men in the depression group was 72 (SD=4.7, range=66-79). No significant differences existed between the subgroups of manic and depressed patients. There was no significant difference in length of follow-up between the mania and depression groups. The mean follow-up period was 5.6 years (SD=2.S, range=3-10). Outcome information regarding mortality was obtained for all 100 patients either by direct contact with patients and families or by referring to the official vital statistics records. Eighteen (36%) of the manic patients had evidence of neurological disorders, compared with four (8%) of the depressed patients. The difference was statistically significant after adjusting for age, sex, and previous episodes (adjusted odds ratio=8.0, 95% confidence interval= , p=o.oo1 ). Table 1 lists the type of neurological disorders evidenced by the 1 8 manic patients. Because of the retrospective nature of the study we did not attempt to establish an exact temporal relationship between first episode of mania and cerebral-organic disease. The manic patients had a significantly higher mortality rate than the depressed patients (adjusted hazard ratio=2.4, 95% confidence interval= , p=o.o2) after controlling for age, sex, previous episodes, and neurological disorder. At the end of the follow-up, 25 (50%) of the manic patients were dead compared with 10 (20%) ofthe depressed patients. Figure 1 shows that the probability of remaining alive after S years of follow-up was approximately 90% for patients with unipolar major depression and 65% for patients with mania. By 10 years the probability of remaining alive for each group was 75% and 30%, respectively. With the use of survival curves we can estimate the probability of remaining alive for 10 years after being alive for S years; for manic patients this probability was 54% (0.35 divided by 0.65) and for depressed patients it was 82% (0.70 divided by 0.85). Only one death by suicide was reported-that of a patient with unipolar depression-in the entire study group. The depressed patients had a shorter time to institutionalization (adjusted hazard ratio=3.6, 95% confidence interval=1.1-s.4, p=o.o2) after controlling for age, sex, previous episodes, and neurological disorder. At follow-up, 18 (37%) of the 49 depressed patients were institutionalized compared with 13 (30%) of 44 manic patients. (Patients who died while institutionalized were included in this analysis.) Twenty-six (52%) of the manic patients compared 342 AmJ Psychiatry 149:3, March 1992

3 SHULMAN, TOHEN, SATLIN, ET AL. TABLE 1. Demographic and Clinical Data for 18 Elderly Patients With Mania and Neurological Disorders at Index Hospitalization Age at Onset Age at First Family History of Affective Manic of Affective Disorder Episode Patient Sex Disorder (years) (years) Neurological Disorder 1 M Negative Chronic alcoholism, peripheral neuropathy 2 F Positive Frontal lobotomy, bilateral encephalomalacia 3 F Negative SO SO Chronic alcoholism, dementia 4 F Negative Chronic alcoholism, seizures/delirium S M Negative Chronic alcoholism, blackouts/delirium 6 M Negative Cerebral contusions 7 F Positive Parkinson s disease 8 F Negative Right cerebral infarct 9 10 F F Negative Negative Multiple cerebral infarcts Encephalopathy/neuroleptic malignant syndrome 1 1 M Negative Closed head injury 12 M Negative Recurrent cerebral contusions (boxing), lacunar infarct 13 F Positive Left cerebral hemorrhage 20 years before onset of affective disorder 14 F Positive Cerebral infarct 15 M Positive Dementia 16 M Negative Embolic cerebral infarct, mild dementia 1 7 F Negative Right capsular necrosis, cerebral vasculitis (rheumatoid) I 8 F Positive Parkinson s disease with 16 (32%) of the depressed patients had a positive family history of affective disorder in a first-degree relative (odds ratio=2.3, 95% confidence interval= , p=o.o4). The manic patients were significantly younger at first psychiatric hospitalization than the depressed patients (two-tailed t test=2.3, df=98, p=o.o2). The mean age at first psychiatric hospitalization of the depressed patients was 62.7 years (SD=1 1.0, range=34-79). For the manic patients the mean age at first psychiatric hospitalization was 55.5 years (SD=18.8, range=19-84). FIGURE 1. Cumulative Probability of Death for 50 Elderly Patients With Mania and 50 Elderly Patients With Unipolar Depression at Index Hospital Admission Manic Patients Twenty-three (46%) of the manic patients were rehospitalized in a psychiatric unit at least once during the follow-up period. Having a neurological disorder at the time ofthe index admission was a predictor of subsequent psychiatric hospitalization (odds ratio=4.6, 95% confidence interval= , p=o.os) after adjusting for sex, age, previous episodes, and age at first manic episode. Neurological disorder also predicted institutionalization (adjusted odds ratio=7.7, 95% confidence interval= , p=o.o3) after adjusting for age, sex, and previous episodes. Twelve (24%) of the manic patients were institutionalized during the follow-up period. Unlike Shulman and Post (4), we found no significant sex differences between the manic and depressed patients with respect to neurological disorders. In addition, an age at onset of more than 55 years predicted a shorter time in remission until first psychiatric rehospitalization (adjusted hazard ratio=3.9, 95% confidence interval= ) after controlling for age, sex, neurological disorder, and previous episodes. Six (33%) of the 1 8 neurologically impaired manic patients had a positive family history of affective disorder in first-degree relatives (table 1 ), compared with 26 (52%) of all so manic patients (x2=4.6,df=1, p=o.o3) VEAIATSK Numb.r Twenty (40%) of the manic patients experienced depression as their first affective episode at a mean age of 47.5 years (SD=18.3, range=19-84). The mean latency from first depression to first mania in these 20 patients was 14.9 years (SD=S.1, range=1-48). There was a significant difference in mean age at first mania between men (mean=s8.7, SD=16.6, range=19-83) and women (mean=68.7, SD=1 3.2, range=42-85) (t=2.0s, df=47, p=o.o4). The clinical course of six manic patients was characterized by manic episodes but not depressive episodes before the index admission. Depressed Patients Not surprisingly, older age predicted higher mortality in the depressed patients (adjusted hazard ratio=1.2, 95% confidence interval= , p=o.o2) after con- of Am J Psychiatry 1 49:3, March

4 MANIA AND DEPRESSION IN OLD AGE trolling for sex, previous episodes, and neurological disorder. Also, depressed patients who had been institutionalized before the index episode had a higher mortality (adjusted odds ratio=17.2, 95% confidence interval= , p=o.os) after controlling for age and sex. In addition, nonmarried status predicted institutionalization for the depressed patients (adjusted odds ratio=4.7, 95% confidence interval= , p= 0.02) after controlling for age and sex. DISCUSSION This study supports a number of hypotheses and findings of earlier studies of elderly manic patients (4-6). In particular, the prevalence of cerebral-organic disorder was found to be significantly higher in manic patients (36%) than depressed patients (8%). Neurological disorders were heterogeneous in nature and nonspecific with regard to brain location. As in previous studies, the prevalence of Alzheimer-type dementia was no higher than the expected proportion in the general population. Alcoholrelated disorders accounted for four of the 1 8 cases of neurological impairment in the manic patients, and cerebrovascular disease was evident in six of the 1 8 patients. Despite the uncertain relationship of neurological disorder to the onset of mania, a unique pattern appears to be evident in these 1 8 patients. The age at first mania and age at onset of affective disorder were the same or very close temporally in these patients. They seemed to fit the category of patients with secondary mania described by Krauthammer and Klerman (20), associated with a relatively lower rate of genetic predisposition to affective disorder. This is in marked contrast to the long latency from first depression to manic episode for many of the elderly manic patients who did not have evidence of neurological disorder. Manic patients experienced a poor outcome, suffering a significantly higher mortality rate. Half of the manic patients were dead after almost 6 years of followup, compared with 20% of the depressed patients. This high mortality rate has not been reported in follow-up studies of patients with early-onset mania (21). The higher mortality rate among manic patients may account for the relatively shorter time to institutionalization for depressed patients. For elderly manic patients whose first psychiatric hospitalization was for depression, an average of 15 years elapsed before mania became manifest. This is suggestive of a conversion that may be due to factors different from those operating in patients with obvious neurological disorders. It should be emphasized that this retrospective study used coarse measures of cerebral-organic disorder. Therefore, we may have underestimated the true role of cerebral dysfunction, especially subtle cerebrovascular pathology. With newer imaging techniques such as magnetic resonance imaging, positron emission tomography, and single photon emission computed tomography, this issue can be studied more accurately. Genetic loading may be more of a factor in elderly manic patients than in elderly depressed patients. Although genetic factors are generally considered to play a less prominent role in old age (22), mania has been associated with a higher prevalence of positive family history than has unipolar depression (23). We found a higher proportion of manic patients with a positive family history of affective disorder (52%) than has been reported in other studies (5, 6, 12). However, we applied the same assessment measures to the depression comparison group and found that only 32% had a positive family history. The manic patients had an earlier age at onset of first psychiatric hospitalization than the depressed patients. An earlier age at onset is associated with a positive family history (5, 6). Even in old age, manic patients appear to be genetically predisposed to affective disorder but perhaps less so than young manic patients (23). Overall, age at onset of mania was late in life (mean= 55.5 years), a finding that is consistent with other studies of old age (4, 5). The average age at onset of earlyonset mania found in other studies was 29 years (23, 24). Indeed, only three of our SO elderly manic patients had suffered from mania when they were younger than 30. A study of university clinic outpatients (25) compared elderly patients with young depressed patients and found a higher prevalence of mania in the younger patients-7.3% versus 1.4%. Does the young patient with bipolar disorder burn out (26) or is there a high mortality rate? This raises questions about fundamental differences between early-onset and late-onset mania. Our data are consistent with the spectrum theory of Akiskal (9) and the threshold theory of Tsuang et al. ( 1 0), which consider mania a more severe form of affective disorder. Controlled studies of the treatment of mania, particularly with lithium carbonate, are needed to explore the hypothesis that mania in old age represents a manifestation of greater CNS vulnerability (27). As a group, the SO elderly manic patients in this study had a genetic predisposition to affective disorder that became manifest relatively late in life. This applies especially to the manic patients without coarse brain disorders. With growing interest in elderly manic patients, greater numbers of patients will be available for study. Furthermore, the advent of new technologies (28, 29) and advances in our conceptualization of affective disorders should help to clarify the special meaning and importance of mania in old age. We have highlighted two distinct subgroups of patients with mania in old age: 1) patients whose first episode of affective disorder is depression and who have a long latency period before they experience their first episode of mania and 2) those suffering from secondary mania associated with neurological disorders, whose mania is often their first affective episode or follows closely after their first episode of depression. Further studies of mania in the elderly should focus on its distinct nature compared with depression and its fundamental difference from mania that begins early in life. 344 Am J Psychiatry 149:3, March 1992

5 SHULMAN, TOHEN, SATLIN, ET AL. REFERENCES I. Spicer CC, Hare EH, Slater E: Neurotic and psychotic forms of depressive illness: evidence from age-incidence in a national samplc. Br J Psychiatry I 973; 123: Eagles JM, Whalley U: Ageing and affective disorders: the age at first onset of affective disorders in Scotland, I Br J Psychiatry 1985; 147: Weissman MM, Leaf PJ, Tischler GL, Blazer DG, Karno M, Bruce ML, Florio LP: Affective disorders in five United States communities. Psychol Med 1988; 18: Shulman K, Post F: Bipolar affective disorder in old age. Br J Psychiatry 1980; 136:26-32 S. Stone K: Mania in the elderly. Br J Psychiatry 1989; 155: Broadhead J, Jacoby R: Mania in old age: a first prospective study. Int J Geriatr Psychiatry 1990; 5: Akiskal H, Walker P, Puzantian VR, King D, Rosenthal TL, Dranon M: Bipolar outcome in the course of depressive illness. J Affective Disord 1983; 5: Leonhard K, Korffl, Schulz H: Die Temperamente in den Familien der monopolaren und bipolaren phasischen Psychosen. Psychiatr Neurol 1962; 143: Akiskal H: Diagnosis and classification of affective disorders: new insights from clinical and laboratory approaches. Psychiatr Developments 1983; 2: Tsuang MT, Faraone SV, Fleming JA: Familial transmission of major affective disorders: is there evidence supporting the distinction between unipolar and bipolar disorders? Br J Psychiatry 1985; 146: Shulman K: The influence of age and ageing on manic disorder. IntJ Geriatr Psychiatry 1989; 4: Glasser M, Rabins P: Mania in the elderly. Age Ageing 1984; 13: Murphy E: The prognosis ofdepression in old age. BrJ Psychiatry 1983; 142: Baldwin RC, Jolley DJ: The prognosis of depression in old age. BrJ Psychiatry 1986; 149: Rothman KJ: Modern Epidemiology. Boston, Little, Brown, Harrell FE: The logist procedure, in SUGI Supplemental Library User s Guide, Version 5. Edited by Hastings RP. Cary, NC, SAS Institute, Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure Time Data. New York, John Wiley & Sons, I 980 I 8. Cox DR: Regression models and life tables. J R Stat Soc Series B 1972; 34: Harrell FE: The PHGLM procedure, in SUGI Supplemental Library User s Guide, Version 5. Edited by Hastings RP. Cary, NC, SAS Institute, Krauthammer C, Klerman GL: Secondary mania: manic syndromes associated with antecedent physical illness or drugs. Arch Gen Psychiatry 1978; 35: Tohen M, Waternaux CM, Tsuang MT: Outcome in mania: a four year prospective follow-up of 75 patients utilizing survival analysis. Arch Gen Psychiatry 1 990; 47: Mendlewicz 5, Fieve R, Rainer J, Fleiss JL: Manic depressive illness: a comparative study of patients with and without a family history. Br J Psychiatry 1972; 120: Winokur G: Mania and depression: family studies and genetics in relation to treatment, in Psychopharmacology: A Generation of Progress. Edited by Lipton MA, DiMascio A, Killam KF. New York, Raven Press, I Goodwin FK, Jamison KR: Manic-Depressive Illness. New York, Oxford University Press, I Muserti L, Perugi G, Soriani A, Rossi VM, Cassano GB, Akiskal HS: Depression before and after age 65: a re-examination. Br J Psychiatry 1989; 155: Winokur G: The Iowa 500: heterogeneity and course in manic depressive illness (bipolar). Compr Psychiatry 1975; 16: Himmelhoch JM,Neil JF, May SJ, Fuchs CZ, Licata SM: Age, dementia, dyskinesias, and lithium response. Am J Psychiatry 1980; 137: SchwartzJM, Baxter LRJr, MazziottaJC, Gerner RH, Philps MC: The differential diagnosis ofdepression: relevance ofpositron emission tomography studies of cerebral glucose metabolism to the bipolar-unipolar dichotomy. JAMA 1987; 258: O Connell RA, Van Heertum RL, Billick SB, Holt AR, Gonzalez A, Notardonato H, Luck D, King LN: Single photon emission computed tomography (SPECT) with [123I]IMP in the differential diagnosis of psychiatric disorders. J Neuropsychiatry I 989; 1: Am J Psychiatry 1 49:3, March