The phrase "late-life schizophrenia" refers to older patients who first manifested schizophrenia in early adulthood, as well as to those whose symptoms first appeared in later life (
6). According to the Epidemiologic Catchment Area (ECA) studies (
7), the lifetime prevalence of schizophrenia among people age 65 years and older is only .3 percent, and the one-year prevalence is .2 percent.
We believe that several methodologic aspects of the ECA studies may have caused an underestimation of the prevalence of schizophrenia among the elderly population. These studies used
DSM-III criteria (
8), which did not permit the diagnosis of late-onset schizophrenia and did not include negative symptoms, and they undersampled areas with higher rates of mentally ill elderly persons such as public housing (
9). The reliance on retrospective self-reports in the ECA studies also has obvious limitations for older patients with schizophrenia. Thus the true current prevalence of schizophrenia in the elderly population is unknown, but it is probably higher than the ECA figures suggest.
Early-onset schizophrenia
Kraepelin's original conception of dementia praecox was that this disorder was typified by onset in adolescence or young adulthood, with a progressive decline in functioning thereafter (
10). Consistent with this view, Davidson and associates (
11) reported that more than 60 percent of their sample of chronically institutionalized patients with schizophrenia had diagnosable dementia. However, chronic institutionalization is atypical for most elderly patients with schizophrenia and is experienced by only the most severely ill sector of the population with schizophrenia (
12). The decline pictured by Kraepelin appears to characterize the outcome for about 20 percent of patients, while another 20 to 30 percent experience marked improvement or even recovery (
13,
14,
15).
A related issue is the notion of "schizophrenic burnout." It involves a deficit state, with a disappearance of positive symptoms, such as delusions and hallucinations, and an increase in negative symptoms, such as social withdrawal, affective flattening, and avolition. This pattern may describe the course seen in a small minority of patients, but a majority of older patients with schizophrenia continue to have both positive and negative symptoms, and another minority have substantial improvement in both types of symptoms (
16,
17,
18).
The reasons for the recovery seen in some patients in later life are not fully clear. One possibility is changing role expectations—patients who were unable to meet the challenges of young adulthood may have less difficulty meeting role expectations associated with later life, which are typically less challenging. However, we do not believe that this is the main reason for recovery. Neurobiological changes associated with later life that may foster a decrease in psychotic symptoms have been documented; an example is reduction in dopamine activity (
19).
The rates of full remission remain uncertain, but there are clear examples of people with schizophrenia who show remarkable recovery after many years of chronic dysfunction. A recently publicized example of recovery is that of John Nash, a Nobel-prize-winning mathematician who developed schizophrenia in his late twenties (
18). Dr. Nash's symptoms remitted after decades of severe illness, whereupon he returned to a productive academic career in mathematics.
Patient characteristics that are associated with better outcome include good premorbid adjustment, acute onset of symptoms, short duration of an ongoing episode, and a paucity of negative symptoms (
20,
21). According to Wyatt (
22), early intervention with antipsychotic medications during patients' first psychotic episode may also lead to a better long-term outcome.
Late-onset schizophrenia
The historical tendency to view schizophrenia as having an onset restricted to adolescence or young adulthood was formalized in criteria put forth by Feighner and colleagues (
23) and in
DSM-III criteria (
8), which did not permit diagnosis of schizophrenia if symptoms developed after age 40 or 45, respectively. However, a number of reports challenged this notion (
24,
25,
26,
27,
28).
DSM-III-R (
29) permitted diagnosis of schizophrenia late-onset type after age 45, and
DSM-IV (
30) includes no age-of-onset restriction for diagnosis of schizophrenia.
It has been suggested that late-onset schizophrenia is a neurodegenerative condition (
31), results from acquired brain lesions (
32), or is secondary to sensory deficits and interpersonal isolation (
33,
34,
35). Such theories suggest that patients with late-onset schizophrenia have relatively normal brain and psychosocial functioning before the onset of these triggering changes. Our studies challenge these views.
In investigations conducted at the Clinical Research Center for the Study of Late-Life Psychoses at the University of California, San Diego, patients with late-onset schizophrenia (those whose age at onset of schizophrenia was 45 years or older) were compared with patients of a similar age who had earlier onset of schizophrenia (
16). Similarities between the groups included the severity of positive symptoms, chronicity of course, level of uncorrected sensory impairment, family history of schizophrenia, early childhood maladjustment, increased mortality relative to the general population, qualitative response to neuroleptic medications, overall pattern of neurocognitive impairment, and presence or absence of gross structural brain abnormalities seen via neuroimaging. Similar to early-onset patients, patients with late-onset schizophrenia also had an elevated number of minor physical anomalies compared with normal subjects, suggesting subtle aberrations in their early development (
36).
Several differences between early- and late-onset patients were also observed in these studies (
16). Patients with late-onset schizophrenia were more likely to be women and to have the paranoid subtype of the disorder. They had less severe negative symptoms, better premorbid functioning in early adulthood, and less impairment in the specific neurocognitive areas of learning and abstraction or cognitive flexibility, and they required lower dosages of neuroleptic medications for management of their psychotic symptoms. Quantitative analyses of MRI scans in a subsample of patients showed larger thalami in the late-onset group than in the early-onset group (
37).
The similarities observed between patients with early- and late-onset schizophrenia suggest that the latter is "true" schizophrenia, in that the two groups were similar in terms of apparent genetic risk, positive symptoms, treatment response, and course. On the other hand, the differences between these groups suggest that neurobiologic differences might mediate the onset of schizophrenic symptoms (
38,
39,
40). Elucidation of such mediating factors may suggest novel intervention or prevention strategies.