Electroencephalographic Abnormalities and 5-Year Outcome in First-Episode Psychosis

Abstract

Objective:

To examine the relation of electroencephalographic abnormalities to 5-year outcomes in first-episode psychosis (FEP).

Methods:

Patients (n = 103) had their baseline electroencephalogram (EEG) classified by modified Mayo Clinic criteria. Symptoms and psychosocial functioning were rated after 5 years of treatment.

Results:

Dysrhythmic EEG was associated with persistence in positive and negative symptoms of psychoses and poorer psychosocial functioning at 5-year follow-up, independently of other characteristics, such as duration of untreated illness or premorbid adjustment. A higher percentage of people with comorbid substance use disorder had normal EEG.

Conclusions:

Abnormal baseline EEG in FEP is associated with poorer 5-year symptomatic and functional outcome.

Despite evidence that EEG abnormalities in schizophrenia may be prognostic of treatment outcome,¹ limited research has been carried out. Follow-up of patients with FEP indicate that early EEG findings have prognostic significance for symptoms at 3-year follow-up.^2,3 There are no other follow-up studies concerning the prognostic implications of EEG findings in FEP, and there is no evidence, as yet, demonstrating a relation of EEG findings to long-term psychosocial functioning or comorbid SUD. Our study will extend previous findings in 2 ways. First, we report a longer follow-up (5 years). We will also examine the relation between EEG findings, comorbid SUD, and psychosocial functioning as indexed by the SOFAS and full-time employment or studies.

Method

Sample

One hundred and thirty-two untreated patients were recruited from successive admissions to the PEPP at London Health Sciences Centre, London, Ontario.⁴ All provided informed consent as approved by The University of Western Ontario Ethics Board for Health Science Research.

Measures and Procedures

Patients were diagnosed using the SCID-IV.⁵ Symptoms were assessed using the SAPS⁶ and the SANS.⁷ Level of psychosocial functioning was assessed using SOFAS,⁸ and of weeks of full-time occupation in competitive employment or as a student in the fifth year of follow-up derived from the Life Chart Schedule.⁹ Premorbid adjustment was assessed using the PAS.¹⁰ Total positive and negative symptom scores were derived by global scores for each subscale of the SAPS and SANS, respectively. Further, the symptoms on SAPS and SANS were characterized into the 3 syndromes of reality distortion (delusions and hallucinations), psychomotor poverty (poverty of speech, decreased spontaneous movements, and blunting of affect), and disorganization (poverty of content of speech, inappropriate affect, and formal thought disorder). Symptoms were assessed at baseline and every year thereafter. SOFAS and weeks of occupational functioning were assessed at the 5-year mark only. Baseline comorbidity of SUD was also carried out using the SCID-IV.

Clinical Implications

• Patients with FEP with an abnormal EEG have a poorer symptomatic and functional outcome at 5 years, compared with patients with a normal EEG.

• Patients with FEP and an SUD are more likely to have a normal EEG, compared with patients without comorbidity.

• EEG may be used as a tool for identification of patients who are less likely to respond to treatment early in therapy.

Limitations

• This is an uncontrolled study of prospective patients with FEP.

• EEGs were carried out at baseline only.

• Routine and not quantified EEG assessments were carried out.

EEGs were recorded in accordance with the International 10–20 Electrode Placement System. EEGs were classified according to a modified version of the Mayo Clinic classification of EEG results,¹¹ normal indicating the results are within normal limits for the person’s age and state of alertness; essentially normal indicating 1 or more elements of doubtful normality; and dysrhythmia, which includes 3 levels of varying intensity and frequency of theta–delta or rhythmic activity, including spikes or recorded seizures. A neurologist specializing in EEG and epilepsy, who had no knowledge of the patient’s psychiatric condition, interpreted the EEGs.

Ratings were carried out by well-trained staff with good interrater reliabilities (intraclass correlation of at least 0.80).

Results

Sample

There were 132 patients with an FEP disorder for whom 5-year follow-up data were available. EEGs had been completed shortly after entry into PEPP for 103 patients. There were no significant differences between patients on whom EEGs were available, compared with not available, regarding age, sex, marital status, diagnosis, level of education, initial symptoms, or treatment delay. Patients for whom EEGs were available had poorer social and total premorbid adjustment in childhood and early adolescence. The majority were males (76.7%), single (84.5%), and had high school or a higher level of education (52.4%), and a diagnosis of schizophrenia spectrum disorders (80.7%).

EEG, Symptomatic, and Functional Outcome

Forty-nine of the 103 patients (47.6%) were classified as having a normal EEG, 33 (32%) had an essentially normal EEG, and 21 (20.4%) had a dysrhythmic EEG. EEG results were not significantly related to sex, marital status, age, educational level, duration of untreated illness or psychoses, or premorbid adjustment. There was no significant difference in the distribution of EEG findings between patients with a diagnosis in the schizophrenia, compared with nonschizophrenia, spectrum.

Patients in the EEG categories were compared on level of symptoms at entry and at 5-year follow-up (Table 1). There were no significant differences between the EEG groups in level of initial symptoms. EEG findings were related to symptoms and social and occupational functioning at 5 years. In each, the patients with dysrhythmia on EEG had the poorest outcome.

Table 1.

Relation of EEG status to symptoms and functioning

Symptoms	Normal	Essentially normal	Dysrhythmia	F (df)	P
SAPS global
Initial	10.37	10.36	8.63	2.19 (2/98)	0.12
5-year	1.28	2.83	3.53	6.14 (2/94)	<0.01
SANS global
Initial	12.63	10.63	11.63	1.95 (2/98)	0.14
5-year	5.04	5.33	8.00	3.72 (2/94)	0.02
Reality distortion
Initial	21.95	20.94	16.79	1.78 (2/98)	0.17
5-year	2.06	4.96	7.31	5.17 (2/94)	<0.01
Disorganization
Initial	15.27	14.82	9.61	2.05 (2/97)	0.13
5-year	1.36	2.14	3.12	1.64 (2/88)	0.20
Psychomotor poverty
Initial	13.20	12.00	11.58	1.16 (2/98)	0.32
5-year	8.28	8.96	15.50	4.37 (2/88)	0.01
5-year SOFAS	63.17	64.18	51.26	4.77 (2/91)	0.01
Weeks of full-time occupation	19.67	21.77	7.00	2.99 (2/94)	0.05

Individual comparisons revealed that, for SAPS score at 5 years, both the essentially normal and dysrhythmia groups were more symptomatic than the normal group, but did not differ from each other. Dysrhythmia was associated with more severe negative symptoms (both SANS global scores and psychomotor poverty) than for the other groups. Patients with dysrhythmia also showed higher reality distortion at 5 years. Similarly, dysrhythmic EEG was associated with lower SOFAS scores and fewer weeks of full-time occupation at follow-up than the other groups. Differences between groups on symptoms and functioning remained significant when level of initial symptoms, premorbid adjustment, and treatment delay were entered in an analysis of covariance.

We also examined symptom remission at 5 years. We defined remission as having a score of less than 2 on global ratings of the SAPS and SANS. Level of symptoms had been tracked on a weekly basis during the fifth year, which enabled us to establish whether remission had persisted for at least 6 months. There is a significant relation of EEG findings with remission of positive symptoms (χ² =7.02, df = 2, P < 0.03) and borderline significance for remission of negative symptoms (χ² =5.07, df = 2, P < 0.08) and total symptom remission (χ² = 4.92, df = 2, P < 0.09). In the case of positive symptoms, this reflects a greater percentage of patients with normal EEG being in remission (89.8%) than for those with essentially normal (73.3%) or dysrhythmic EEG (63.2%).

We carried out regression analyses to examine the possible role of symptom remission as a mediator of the relation between EEG and functioning. Given evidence that negative symptoms are especially important in determining psychosocial functioning, we examined remission of positive and negative symptoms separately. With reference to both SOFAS and occupational functioning, entering remission of negative (but not positive) symptoms rendered the relation of abnormal EEG to functioning nonsignificant. These findings are consistent with negative symptoms possibly mediating the relation between EEG abnormality and functioning.

There was a significant relation between presence of an SUD comorbidity and EEG findings (χ² = 11.08, df = 2, P < 0.004). Examination of adjusted residuals indicates that this primarily reflects a larger percentage of people with comorbidity having normal EEG (71.8%) than those without comorbidity (36.7%), and more of those without comorbidity having essentially normal EEG (39.5%, compared with 15.7%, for people without comorbidity).

Discussion

There is evidence that the first 3 to 5 years of treatment of schizophrenia constitutes a critical period in determining longer-term outcome,¹² thus a 5-year follow-up of patients is of particular interest. Our published data³ of 3-year follow-up showed that people with dysrhythmia on EEG showed significantly greater persistence in positive and negative symptoms. Our 5-year outcome data further support this relation. There is also a significant relation of normal EEG with likelihood of meeting criteria for remission of positive symptoms. All these findings are independent of treatment delay or premorbid functioning. Provider ratings of adherence during the first or fifth year of treatment were not significantly related to EEG findings. Another important addition of our findings is the relation between EEG anomalies and functioning at 5 years, as evidenced by both the SOFAS and the number of weeks of full-time occupation in competitive employment or as a student.

More patients with comorbid SUD had a normal EEG. It would suggest that for patients with FEP and an SUD, drug-related and environmental factors probably play a greater role than an underlying brain diathesis seen on EEG. This is further supported by a recent meta-analysis¹³ showing fewer cognitive difficulties in people with comorbid SUD. EEG abnormalities appear to be a marker for a more pernicious form of psychotic illness, perhaps representing treatment refractoriness. The findings of EEG abnormalities in patients considered for clozapine therapy owing to their treatment refractoriness is also consistent with this finding.¹⁴ Clinically, therefore, if patients have an abnormal EEG and are not responding to treatment early in therapy, a more aggressive treatment approach and (or) early introduction of clozapine may be warranted.

We recognize that this is an uncontrolled study, but the consistency of findings across a 5-year span support the robustness of the results. EEG appears to be a relatively simple and inexpensive investigation for identifying patients who are less likely to have a favourable outcome. Replication of these findings in the context of longer-term follow-up studies of FEP is warranted.

Conclusion

An abnormal baseline EEG in patients with FEP is associated with a poorer symptomatic and functional outcome at 5-year follow-up.

Abbreviations

EEG

electroencephalogram

FEP

first-episode psychosis

PAS

Premorbid Adjustment Scale

PEPP

Prevention and Early Intervention Program for Psychoses

SANS

Scale for Assessment of Negative Symptoms

SAPS

Scale for Assessment of Positive Symptoms

SCID-IV

Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition

SOFAS

Social and Occupational Functioning Assessment Scale

SUD

substance use disorder