Absence of Cognitive Impairment at Long-Term Follow-Up in Adolescents Treated With ECT for Severe Mood Disorder

Several studies have shown that ECT is a safe and effective treatment for adolescents with intractable mood disorder, in particular when patients exhibit catatonic or psychotic symptoms (1). To our knowledge, only one study (2) has systematically assessed cognitive efficiency in children given ECT. In that study of 16 children treated with ECT, cognitive efficiency was significantly reduced after an ECT series but returned to pretreatment level between 5 and 27 months later. We did not find any other study assessing the long-term effects of ECT on cognitive functions, especially memory, in young subjects. This was the aim of the present study. We focused on memory functioning, using several tests to assess anterograde memory abilities.

METHOD

We searched, from chart review, for all adolescents treated with ECT for mood disorder before 19 years of age in five psychiatry departments in Paris. At least 1 year had to have elapsed since the last ECT. During the period 1987–1996, 20 patients were found to meet those criteria. Among them, one had died from suicide, one could not be traced, and six refused to participate. The remaining 12 subjects were approached for evaluation. One was psychotic and too ill to complete the evaluation, and one was not able to because of mental retardation. Thus, 10 patients (four men and six women) who received bilateral ECT could be included in the study. On the basis of chart review, we chose 10 psychiatric comparison subjects who had never been given ECT but were individually matched with the ECT subjects for sex, age, date and place of hospitalization, and DSM-III-R diagnosis (table 1).

After informed consent was obtained, all subjects were given a battery of clinical and cognitive evaluations. Current clinical state was rated on the 21-item Hamilton Depression Rating Scale and the Brief Psychiatric Rating scale (BPRS). Cognitive testing included the Mini-Mental State (3), the attention section of the Wechsler Memory Scale—Revised (4), and the California Verbal Learning Test (5) (French translation: B. Deweer), which assesses anterograde memory and verbal learning. Each subject from the ECT group was asked for spontaneous comments on his or her experience with ECT, including possible reports of subjective memory impairment, and completed Squire’s Subjective Memory Questionnaire (6) (French translation: I. Amado-Boccara), which includes 18 items, each rated from –4 to 4 and summed to produce a total score.

Statistical analysis was performed with SAS software (SAS, Inc., Cary, N.C.). Case-control comparisons for measures at the index episode and at follow-up used the Wilcoxon sign rank test. Nonparametric Spearman correlations between measures were computed within the entire study group. Two-sided p values are given.

RESULTS

The characteristics of the subjects in both groups at the index episode are shown in table 1. It was not possible to match each ECT subject to a comparison subject with exactly the same subtype of mood disorder or the same degree of clinical severity of the episode. One patient in the ECT group had comorbid mosaic Down’s syndrome and an IQ of 77. The comparison subject to which she was matched had a similar IQ, 79.

The mean duration of the follow-up period was 5.2 years (SD=2.6). The mean time from last ECT course was 3.5 years (SD=1.7), as two patients had received a second course of ECT. All patients were in remission when interviewed, except one who still had mild manic symptoms (BPRS score=30). All subjects except one were still receiving outpatient psychiatric care. Most were treated with psychotropic medications, and the treatment regimens of the two groups were comparable. No patient was receiving benzodiazepines.

The two groups did not differ on any objective measure of cognitive functioning. The mean total scores on the Mini-Mental State for the two groups were similar and indicated normal functioning for this age range: 28.7 (SD=1.7) for the ECT group and 29.3 (SD=0.7) for the comparison group (sign rank test, value=–3.5, p=0.50) (3). Similarly, the two groups did not differ significantly on any of the six scores for the attention section of the Wechsler Memory Scale—Revised (total score: mean=93.5, SD=13.2, and 93.4, SD=10.8, respectively) (sign rank test, value=1.5, p=0.91). The two groups also had similar scores on all California Verbal Learning Test measures, including general learning performance and learning characteristics; for instance, the mean scores for total recall were 57.3 (SD=7.6) and 55.7 (SD=7.5), respectively (sign rank test, value=3.0, p=0.70).

Finally, each group’s scores on the California Verbal Learning Test measures were compared to normative scores obtained by Deweer et al. (unpublished data, 1996) for a community sample of 48 normal subjects aged 19 to 29 years. The pair of patients with low IQs was not compared to this community sample. Scores from the remaining nine subjects in each patient group were similar to results from the normative sample, except the scores for short-delay free recall and short-delay cued recall, which were significantly lower than normal in the comparison group, 10.7 (norm, 12.5) and 11.9 (norm, 13.4), respectively (sign rank test, value=–20.0, p=0.04). There was no single difference between the ECT group and the normal community group.

For all patients combined (N=20), none of the cognitive scores was correlated to age or number of years of education. Greater psychopathology was significantly associated with poorer cognitive performance (BPRS and Mini-Mental State: r_S=–0.53, p=0.04; BPRS and California Verbal Learning Test total recall: r_S=–0.72, p=0.02; Hamilton depression scale and Mini-Mental State: r_S=–0.33, p=0.15; Hamilton depression scale and California Verbal Learning Test total recall: r_S=–0.41, p=0.07). Poorer attention tended to be associated with poorer cognitive performance (Wechsler attention and Mini-Mental State: r_S=0.31, p=0.18; Wechsler attention and California Verbal Learning Test total recall: r_S=0.34, p=0.14). Similar correlations were also found within each group separately.

Six patients in the ECT group reported having had subjective memory impairment immediately after the ECT course. At follow-up, scores on Squire’s Subjective Memory Questionnaire were very close to 0, indicating that the patients judged their current memory to be neither worse nor better than before ECT (mean total score=–1.9, SD=6.4). Only two patients had negative scores: the first one was still complaining of memory impairment 23 months after completion of ECT (total score=–18), and the second (total score=–2) was the most recently treated (17 months from ECT).

DISCUSSION

There has been little information on the long-term cognitive consequences of ECT use in adolescents. Our results show that a small group of young patients given ECT for psychotic mood disorder could not be distinguished years later from carefully matched comparison subjects, with similar diagnoses but no ECT treatment, on objective measures of cognitive function. The patients who received ECT did not exhibit measurable anterograde memory deficit an average of 3.5 years after treatment. This conclusion should be tempered by the limitations of the study. Given the rarity of ECT use for adolescents (1), the small number of subjects may have limited the likelihood of establishing group differences because of lack of statistical power; we cannot exclude that the occurrence of long-term memory impairment might be a rare but possible event. Another limitation concerns the lack of assessment of retrograde amnesia in our study, due to our failure to find appropriate instruments for a study without a prospective design. Further studies are needed to confirm our results.

Received March 10, 1999; revisions received June 30 and Aug. 4, 1999; accepted Sept. 3, 1999. From the Department of Child and Adolescent Psychiatry, the Centre National de la Recherche Scientifique Unit矍ixte de Recherche 7593, the Department of Neuropsychology, and the Department of Psychiatry, Groupe Hospitalier Piti笓alpè³²iç±¥; the Department of Biostatistics, H󯨴al Saint Louis, Paris; and the Department of Adolescent Psychiatry, Institut Mutualiste Montsouris, Paris. Address reprint requests to Dr. Cohen, Department of Child and Adolescent Psychiatry, Groupe Hospitalier Piti笓alpè³²iç±¥, 47 Boulevard de l’H󯨴al, 75013 Paris, France; [email protected] (e-mail). Supported by SmithKline Beecham France and by grants to Dr. Taieb from the Foundation Lundbeck France and from the Soci赩 d’Etudes et de Soins pour les Enfants Paralysè± et Polymalformè±® The authors thank the following for their support: Prof. J.P. Oli矡nd Drs. I. Amado-Boccara, B. Benchetrit, W. DeCarvalho, B. Deweer, M. Herv窠L. Meignan, H. Lida-Pulik, and M.L. Paillç±¥-Martinot.

TABLE 1

Enlarge table