Clinical Outcome, Functional Outcome and Illness Costs
Manic–depressive disorder has its onset in most persons in ado-lescence and young adulthood, between the ages of 15 and 30. However, prepubertal mania and first-onset disease in the ninth decade of life are not unheard of. Once developed, multiple epi-sodes are the rule. A review of the literature indicates that the majority of patients have four or more episodes in a lifetime. Among rapid-cycling patients, the basis for the diagnosis is four or more episodes in a year with an average of more than 50 lifetime episodes. There is no typical pattern to episode recur-rence, with some patients having isolated manic, hypomanic, or depressive episodes, others switching from one pole to the other in linked episodes, and still others switching continually from one pole to the other in quasi-cyclical fashion. However, even among rapid-cycling patients, episodes are rarely periodic. Rather, the pattern is more accurately described by chaotic dynamics.
Episode length typically ranges from 4 to 13 months, with depressive episodes typically longer than manic or hypomanic episodes. Women appear to have more depressive relapses than manic ones, whereas men have a more even distribution. Also women predominate among rapid-cycling patients, representing 70 to 90% in most studies.
The early optimistic view of outcome, which derived primarily from experience in controlled clinical trials, con-trasts with the overall guarded prognosis described by most longitudinal studies in the last three decades, which have been less controlled but more inclusive than formal clinical trials.
Approximately 20 to 40% of patients with manic–depressive disorder do not respond well to lithium, and that proportion may increase to as much as 80% for certain subgroups such as patients who experience rapid-cycling pattern (Dunner and Fieve, 1974; Maj et al., 1989) or mixed manic and depressive episodes (Keller et al., 1986). When assessed 1.5 years after index hospitalization, between 7 and 32% of manic–depressive patients remain chronically ill, depending on the polarity of index episode (Keller et al., 1986). Only 26% of one sample had good outcome after hospitalization for mania, whereas 40% had moderate and 34% had poor outcome (Harrow et al., 1990). The probability of remaining ill at 1, 2, 3 and 4 years after hospitalization for mania was, respectively, 51%, 44%, 33% and 28% (Tohen et al., 1990). Sixty percent of an ambula-tory sample of manic–depressive patients had fair to poor out-come based on a global outcome score after 1-year follow-up (O’Connell et al., 1991).
Relatively little is known regarding clinical outcome in manic–depressive type II patients, although they appear to be at least as impaired in terms of relapse as manic–depressive type I patients. For instance, 70% of manic–depressive II patients fol-lowed up for 5 years experienced multiple relapses, whereas only 11% were episode free (Coryell et al., 1989).
Subsyndromal affective symptoms may remain in up to 13 to 34% (Harrow et al., 1990), and substantial interepisode morbidity may remain despite adequate treatment with lithium. It is not clear whether such interepisode pathology represents incompletely resolved major affective episodes, medication side effects, demoralization due to functional impairment, or a com-bination of these. It should be noted here that side effects are more than a trivial issue, as they may lead to medication dis-continuation in 18 to 53%, a figure that is greater in lower socio-economic classes.
Substantial levels of functional impairment are also character-istic of manic–depressive disorder, even when major clinical in-dices have improved. Tohen and coworkers (1990) found 28% of subjects unemployed after index hospitalization for mania. Bau-wens and associates (1991) found that levels of functional dis-ability correlated both with number of prior episodes and with residual interepisode psychopathology. Five-year follow-up data from the NIMH Collaborative Program on the Psychobiology of Depression (Coryell et al., 1989) provide evidence that levels of impairment in manic–depressive type I and type II disorders are similar. This included similarly fair to very poor work (in 30 and 42% of patients with types I and II, respectively), marital (30 and 23%), social (45 and 45%) and recreational (45 and 48%) function; sense of satisfaction or contentment (57 and 62%); and overall social adjustment (68 and 62%). More recent analysis of that data set has revealed enduring deficits in educational and occupational status at 5 years of follow-up in a mixed group of manic–depressive and unipolar patients, even in those who were recovered for 2 years (Coryell et al., 1993). This led the authors to comment succinctly: “Follow-up studies have usually defined recovery as the absence of symptoms. The present findings show that this convention may result in an overly benign portrayal of outcome”.
Much less is known about which characteristics predict functional deficits in manic–depressive disorder, an issue of some importance in identifying high-risk groups for particular attention. A recent review (Bauer and Whybrow, 2001) indi- cates, surprisingly, that baseline demographic and functional outcome do not predict future functional outcome. However ongoing depressive symptoms, even to a mild degree, are strongly associated with ongoing functional deficits. The di-rection of causability is not clear, however. It is plausible that depressive symptoms render individuals less able to function in work and personal roles. It is equally plausible that unem-ployment, divorce, social isolation and the like can cause or exacerbate depressive symptoms. In fact, both are likely. In any event, careful attention to functional deficits, depressive symp-toms and their interplay is important to optimizing care and hopefully outcome.
Although there are as yet few available data regarding direct and indirect illness costs for manic–depressive disorder, the direct treatment costs of manic–depressive disorder are sub-stantial. Among the major mental disorders, the rate of hospi-talization for manic–depressive disorder is exceeded only by that for schizophrenia. It is also clear that substantial loss of productivity, in addition to personal suffering, may occur in manic–depressive disorder. In mental illness in general, func-tional impairment was responsible for 55% of the costs of non-addictive mental illness in the USA in 1986 (Rice et al., 1990). It is striking that the functional impairment may be responsible for as much as 75% of the costs of affective illness. Specifically in manic–depressive disorder, evidence indicates that costs from lost productivity are substantial as well. For instance, 19% of persons with manic–depressive disorder attempt suicide at some time in their lives (Klerman et al., 1992), thus placing almost one-fifth of persons with manic–depressive disorder at high risk of loss of life through this one cause alone. Without adequate treatment, a person with manic–depressive disorder from age 25 years can expect to lose 14 years of effective major activity (e.g., work, school, family role function) and 9 years of life. The indirect costs of this disorder are also high, because 15% of persons with manic–depressive disorder are unem-ployed for at least five consecutive years and more than 25% of those younger than age 65 years receive disability payments (Klerman, 1992). Therefore, it stands to reason that treatments targeted at reducing functional impairment, as well as clinical outcome, can have a substantial impact on the burden of mental illness costs to society and quality of life for the individual and her or his family.
For clinical usage, clinicians rating scales in general use for depression can also be used for depressive episodes in manic– depressive disorder. Most share in the shortcoming of tending to underrate hypersomnia, hyperphagia and weight gain, “atypi-cal” features that are common in manic–depressive disorder. Among mania scales, the Young Mania Rating Scale (Young et al., 1978) is well validated on outpatients as well as inpatients, and for hypomanic as well as manic episodes. Self-report scales have the advantage of being brief (typically 5 minutes) and amenable to frequent, even daily, usage without undue burden. However, there have been questions about their reliability and validity, particularly in severely ill manic patients, although one instrument, the Internal State Scale (Bauer et al., 1991) has demonstrated reasonable psychometric properties across several replications.