The National Adult Reading Test as a measure of premorbid IQ in schizophrenia

Objectives. To investigate the validity of the NART as an estimate of premorbid IQ in schizophrenia. Design. A within‐in participants, follow‐back design was adopted. Methods. A sample of adults with schizophrenia who had presented to psychiatric services and had a measure of IQ routinely taken during childhood were traced and subject to follow‐up WAIS‐R and NART IQ assessment (N = 24). Measures of current IQ and NART estimated premorbid IQ were compared with the measure of IQ taken ‘premorbidly’, i.e. in childhood. Results. There were no significant differences between childhood and adult measures of IQ. However there were significant differences between these two indices and NART estimated IQ, particularly where IQ deviated from general population means. The Vocabulary subtest of the WAIS‐R performed better as an estimate of both premorbid and current IQ in the sample. Conclusion. Use of a word‐reading test such as the NART to predict past levels of intellectual function should proceed with caution, particularly where IQ does not fall in the ‘average category’. Use of more than one index of prior level of function is recommended.