Predicting Outcome in Schizophrenia

Aug 01, 2006

Can sociodemographic and clinical variables predict outcome in cases of schizophrenia? Results from studies related to prognostic variables for schizophrenia have yielded interesting yet inconsistent results.

August 2006, Vol. XXIII, No. 9

Can sociodemographic and clinical variables predict outcome in cases of schizophrenia? Attempts to answer this question have been geared toward helping clinicians, patients, and family members have more informed expectations for the long-term prognosis. Results from studies related to prognostic variables for schizophrenia have yielded interesting yet inconsistent results.

In this article, we review the plethora of variables implicated as having prognostic significance in schizophrenia and briefly discuss possible reasons why the literature has not yielded consistent conclusions. Since presentation at intake may have important implications for treatment course, we also review findings from our recent prospective longitudinal study focusing on predicting outcome based on presentation at intake.1 The study addressed the effects of sex, education, and duration of illness; as well as positive, negative, affective, and cognitive symptoms on the level of long-term functioning.

Early reviews of prognosis in psychotic disorders from researchers such as Stephens2 and Vaillant3,4 suggested that positive long-term outcome was related to variables such as severity of the disorder at onset, prominent affective features, good premorbid functioning, being married, having a precipitating event, being confused, or having a family history of manic depressive disorder. Poor outcome was related to insidious onset, exhibiting an asocial premorbid personality, being withdrawn, never being married, or having a family history of schizophrenia. Surprisingly, Valliant's study results4 also led to suggestions that the course of psychosis is not reliably predictable at onset.

Similarly, there are several other inconsistencies in outcome studies investigating the role of sex, age, and symptoms, as well as various other clinical variables (Table). The reasons for such inconsistencies may have included sampling that mixed patients with new onset of schizophrenia and patients who had long-term psychosis, a lack of consideration of the stage of disorder (early course vs late course), and an inability to incorporate prospective randomized study design for the natural course of the illness.

Our recent investigation attempted to minimize some of these concerns in order to further clarify the relationship between clinical and demographic variables and functional outcome.1 We used a prospective design to examine the relationship between symptoms at intake and later functional outcome in first-episode (FE) patients and previously treated (PT) patients with schizophrenia. Patients were selected over a 20-year period; they were rigorously characterized using historical records, underwent physical and psychiatric evaluation by a research psychiatrist, were rated on clinical scales, and were interviewed by research associates using standardized diagnostic measures. These procedures were followed by consensus conferences that established a patient sample with no confounding diagnoses (eg, neurologic/medical illness or traumatic injury) or behaviors (eg, drug abuse). Patients were followed up for an average of 3 years. We also examined the contributions of the patients' sex, education, and duration of disorder to the functional outcome. The findings are discussed below, as well as the results of other investigations of prognostic variables in schizophrenia.


Some studies have found that illness course differs between men and women;5-7 some have found no differences between the sexes in this regard. 8-10 Our study found that male and female patients differed in the degree to which initial symptoms were correlated with later functioning.1 PT women were the only group in whom early levels of functioning did not determine quantity of work. Quantity of work reflects the number of hours or days per week of work as distinguished in the scale from quality of work, which reflects the competence of the individual at that job. There was a higher quantity of work among PT women, even with lower initial levels of functioning, leaving less capacity for incremental improvement relative to other groups.

The sex of the participant also influenced the relationship between depressive symptoms at intake and later quantity of work-higher levels of depression resulted in decreased quantity of work only in women.1 This may reflect a greater range of affective expression among women with schizophrenia, as well as a higher quantity of work relative to men, allowing greater variability in both dimensions.

We also found that men had lower levels of functioning than women at follow-up.1 Men were 3 times less likely to have meaningful social relationships than women were. This may be because men exhibit more socially unfavorable behaviors, which contribute to poorer social course, while women have a higher tendency towards cooperativeness and compliance, which may positively influence social course.11 In addition, estrogen may serve a functionally protective role by causing schizophrenia to manifest at a later age in women than in men, affording women greater opportunities to develop social and occupational skills.5,11 Thus, the sex of the patient appears to play an important role both in long-term social and occupational relationships.


Being married has been found to coincide with better outcome in patients with schizophrenia.12-14 Higher premorbid functioning may lead to the ability to enter into a long-term relationship. Alternatively, being married may be a surrogate marker for later onset, allowing greater social and support network development prior to illness. In addition, marriage may bode well for outcome because it is less socially acceptable to abandon a sick spouse than it is to avoid initiating a relationship with a psychotic individual.


Having obtained a higher level of education has also been correlated with better outcome in patients with schizophrenia.15 As noted above, this may be related to higher premorbid functioning. Furthermore, higher education may yield greater employment possibilities after illness onset, as it does in the general population. Our study found that the impact of education was influenced by treatment status. Although FE patients showed a positive association between level of education and future quantity of work, this was not apparent for PT patients.1 This suggests that the increased employability benefit of education among FE patients is lost as the illness progresses.

Age at onset

Better prognosis has been associated with older age at onset and possibly greater life experience or development prior to illness.16 Yet, some have found no relationship between age at onset and outcome,14 while others have suggested that the older the person is at onset, the poorer the outcome.

Sociodemographic  Mixed findings
Clinical  Predicts more positive outcome


Premorbid levels of functioning A higher level of premorbid functioning has been associated with a better course, particularly if there is greater social contact with friends and family.17 Our study results also demonstrated that the level of functioning at intake was among the most important prognostic variables.

Treatment status

Studies investigating differences between FE and PT patients suggest that although both groups improve in positive symptomology at 6-month and 1- and 2-year follow-up, only PT patients show a reduction in negative symptoms.18-20 Our findings indicated that symptoms at intake have distinct patterns of prognostic significance for functional outcome in PT patients compared with FE patients.1

Duration of illness

Better prognosis has also been associated with a shorter duration of untreated psychosis.12,14,16,21 Such findings reflect the possibility that psychotic symptoms are biomarkers for unchecked pathologic brain-damaging processes, or “toxic psychosis,” leaving the brain and the patient in a more disabled state. Alternatively, the duration and number of psychotic episodes may be correlated with poor outcome because of increasing social isolation and disengagement from a large number of people who witness the patient's psychotic episodes. Similarly, periods of untreated psychosis may yield longer and more frequent periods of unemployment with resulting gaps in the patient's work history, further complicating the ability to obtain subsequent employment.

Research, including our own, has found that longer duration of illness at intake has a negative impact on overall symptoms at follow-up. Longer duration of illness is associated with more severe positive symptoms at 2 years, poorer social and occupational functioning prior to presentation, and poorer outcome at 1 and 3 years in FE and PT patients.20,22 These results are also consistent with studies showing that the lag between onset and treatment is an important determinant of the time to remission and the level of remission in schizophrenia.6

Symptom severity

Some studies report that symptom severity at intake correlates with symptom- based and quality-of-life outcomes in FE patients;8,23 however, others have failed to find such relationships.24 For instance, severity of positive symptoms at intake has been correlated with poor outcome at 1 year and better long-term outcome at 23 years, but there was no correlation found at 2 to 8 years.23,25 Negative symptoms have been correlated with poor functioning at intake and at 2 years, but not at 1-year follow-up.23,26 Depressive symptoms have been correlated with 1-year outcome, yet others suggest that depressive symptoms predict early remission while flat affect predicts longer episodes and shorter remission.18,23

We concluded in our study that symptoms at intake have important patterns of prognostic significance.1Lower levels of positive, negative, and depressive symptoms at intake predicted a higher overall level of functioning at followup. This was reflected in the fuller lives and better occupational and social interactions of the patients we studied. Thus, patients who present with higher levels of positive, negative, or depressive symptoms at intake may be expected to have a poorer prognosis than those with lower levels of symptoms. This is particularly relevant since the type of symptoms at intake appears to be less important than the intensity of symptoms in predicting the later level of functioning.1 An encouraging aspect of our previous findings was that, over time, patients had increased levels of functioning and decreased levels of depression and negative and positive symptoms.

Surprisingly, the class of antipsychotic medication used was not associated with functional outcomes. Specifically, there was no higher level of functioning with newer antipsychotic medications than with older ones.1 This is consistent with several recent studies suggesting that there is no distinction in terms of efficacy between older and newer antipsychotics; rather, they appear to have equal efficacy across several domains.27-29


Several recent studies suggest that specific cognitive domains are associated with long-term functioning.30,31 However, another study failed to reach similar conclusions after controlling for covarying factors.1 This discrepancy may result from using a composite cognitive score in the latter study rather than assessing individual domains. Interpreting the contribution of cognitive measures in schizophrenia has also been complicated by the use of different batteries and measures in various centers and studies. Current efforts to create a more standardized approach to cognitive assessments will likely clarify the contribution of this domain to functional outcome.


Analyses that incorporate the specific contributions of sociodemographic (eg, sex, age at onset, education, marriage) and clinical variables (eg, premorbid levels of functioning, treatment status, duration of illness, symptom severity, cognition) are useful in assessing functional prognosis as well as symptom progression. Such comprehensive evaluations can assist clinicians, patients, and family members in developing more informed expectations for the long-term prognosis based on presentation at intake.

Farzin Irani is currently completing her doctoral training in neuropsychology at Drexel University in Pennsylvania. She is in the division of neuropsychiatry at the University of Pennsylvania Health Science Center.

Steven Siegel is assistant professor in the division of neuropsychiatry, and director at the Stanley Center for Experimental Therapeutics in Psychiatry in the department of psychiatry, University of Pennsylvania.

Both authors report that they have no conflicts of interest to disclose.


1. Siegel SJ, Irani F, Brensinger CM, et al. Prognostic variables at intake and long-term level of function in schizophrenia. Am J Psychiatry. 2006;163:433-441.
2. Stephens JH. Long-term course and prognosis in schizophrenia. Semin Psychiatry. 1970;2:464-485.
3. Vaillant GE. Prospective prediction of schizophrenic remission. Arch Gen Psychiatry. 1964;11:509-518.
4. Vaillant GE. A 10-year followup of remitting schizophrenics. Schizophr Bull. 1978;4:78-85.
5. Hambrecht M, Maurer K, Hafner H, Sartorius N. Transnational stability of gender differences in schizophrenia? An analysis based on the WHO study on determinants of outcome of severe mental disorders. Eur Arch Psychiatry Clin Neurosci. 1992;242:6-12.
6. Loebel AD, Lieberman JA, Alvir JM, et al. Duration of psychosis and outcome in first-episode schizophrenia. Am J Psychiatry. 1992;149:1183-1188.
7. Szymanski S, Lieberman JA, Alvir JM, et al. Gender differences in onset of illness, treatment response, course, and biologic indexes in first-episode schizophrenic patients. Am J Psychiatry. 1995;152:698-703.
8. Browne S, Clarke M, Gervin M, et al. Determinants of quality of life at first presentation with schizophrenia. Br J Psychiatry. 2000;176:173-176.
9. Ho BC, Andreasen NC, Flaum M, et al. Untreated initial psychosis: its relation to quality of life and symptom remission in first-episode schizophrenia. Am J Psychiatry. 2000;157:808-815.
10. Kendler KS, Walsh D. Gender and schizophrenia. Results of an epidemiologically-based family study. Br J Psychiatry. 1995;167:184-192.
11. Hafner H. Gender differences in schizophrenia. Psychoneuroendocrinology. 2003;28(suppl 2):17-54.
12. Harrison G, Croudace T, Mason P, et al. Predicting the long-term outcome of schizophrenia. Psychol Med. 1996;26:697-705.
13. Leff J, Sartorius N, Jablensky A, et al. The International Pilot Study of Schizophrenia: five-year followup findings. Psychol Med. 1992;22:131-145.
14. Wiersma D, Nienhuis FJ, Slooff CJ, Giel R. Natural course of schizophrenic disorders: a 15-year followup of a Dutch incidence cohort. Schizophr Bull. 1998;24:75-85.
15. Geddes J, Mercer G, Frith CD, et al. Prediction of outcome following a first episode of schizophrenia. A follow-up study of Northwick Park first episode study subjects. Br J Psychiatry. 1994;165:664-668.
16. Loebel AD, Lieberman JA, Alvir JM, et al. Duration of psychosis and outcome in first-episode schizophrenia. Am J Psychiatry. 1992;149:1183-1188.
17. Carpenter WT Jr, Strauss JS. The prediction of outcome in schizophrenia. IV. Eleven-year follow-up of the Washington IPSS cohort. J Nerv Ment Dis. 1991;179:517-525.
18. Eaton WW, Thara R, Federman E, Tien A. Remission and relapse in schizophrenia: the Madras Longitudinal Study. J Nerv Ment Dis. 1998;186:357-363.
19. Gupta S, Andreasen NC, Arndt S, et al. The Iowa Longitudinal Study of recent onset psychosis: oneyear follow-up of first episode patients. Schizophr Res. 1997;23:1-13.
20. Szymanski SR, Cannon TD, Gallacher F, et al. Course of treatment response in first-episode and chronic schizophrenia. Am J Psychiatry. 1996;153:519-525.
21. Drake RJ, Haley CJ, Akhtar S, Lewis SW. Causes and consequences of duration of untreated psychosis in schizophrenia. Br J Psychiatry. 2000;177:511-515.
22. Larsen TK, McGlashan TH, Moe LC. First-episode schizophrenia: I. Early course parameters. Schizophr Bull. 1996;22:241-256.
23. Robinson DG, Woerner MG, Alvir JM, et al. Predictors of treatment response from a first episode of schizophrenia or schizoaffective disorder. Am J Psychiatry. 1999;156:544-549.
24. Lastra I, Vazquez-Barquero JL, Herrera Castanedo S, et al. The classification of first episode schizophrenia: a cluster-analytical approach. Acta Psychiatr Scand. 2000;102:26-31.
25. Jonsson H, Nyman AK. Prediction of outcome in schizophrenia. Acta Psychiatr Scand. 1984;69:274-291.
26. Addington J. Group treatment for smoking cessation among persons with schizophrenia. Psychiatr Serv. 1998;49:925-928.
27. Geddes J, Freemantle N, Harrison P, Bebbingon P. Atypical antipsychotics in the treatment of schizophrenia: systematic overview and meta-regression analysis. BMJ. 2000;321:1371-1376.
28. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med. 2005;353:1209-1223.
29. Rosenheck R, Perlick D, Bingham S, et al. Effectiveness and cost of olanzapine and haloperidol in the treatment of schizophrenia: a randomized controlled trial. JAMA. 2003;290:2693-2702.
30. Harvey PD, Bertisch H, Friedman JI, et al. The course of functional decline in geriatric patients with schizophrenia: cognitive-functional and clinical symptoms as determinants of change. Am J Geriatr Psychiatry. 2003;11:610-619.
31. Green MF. What are the functional consequences of neurocognitive deficits in schizophrenia? Am J Psychiatry. 1996;153:321-330.