To examine the hypothesis that episodes of severe hypoglycaemia may cause cumulative cognitive impairment. 100 Type 1 (insulin-dependent) diabetic patients were examined. Their age range was 25-52 years, and the onset of diabetes had occurred after the age of 19 years. Patients with evidence of organic brain disease, including cerebrovascular disease, were excluded. A questionnaire was used to assess the number, frequency and severity of hypoglycaemic episodes experienced during treatment with insulin and the accuracy of this retrospective information was verified from general practice and hospital case-notes. A detailed neuropsychological assessment was undertaken, including tests of pre-morbid and present IQ (Wechsler-Revised), memory and information-processing speed. Significant correlations were observed between the frequency of severe hypoglycaemia and the magnitude of intellectual decline, Performance IQ, inspection time and reaction time (patients with the more frequent hypoglycaemia had poorer performance). Two sub-groups of patients were identified on the basis of their experience of severe hypoglycaemia, and were balanced for pre-morbid IQ, age and duration of diabetes. One sub-group (n = 23) had never experienced severe hypoglycaemia (Group A), whilst the other sub-group (n = 24) had suffered at least five episodes of severe hypoglycaemia (Group B). Group B had greater intellectual impairment than Group A, and Group B also had a significantly slower mean reaction time and higher reaction time variance when compared with Group A.

The IQ scores (WAIS-R) of 100 patients with insulin-treated diabetes (aged 25-52 yr) were compared with those of 100 healthy control subjects who were matched to the diabetic patients for sex, age, education, and social class. The diabetic group had lower WAIS-R performance and verbal IQ scores than the control group (P = 0.017 and P = 0.033, respectively) after controlling for premorbid IQ. The extent of the difference was modest, representing approximately 33% of an SD in IQ. When frequency of severe hypoglycemia was controlled for the difference in performance IQ between the diabetic patient group and the control group was abolished, whereas the difference between the groups in verbal IQ persisted. It is hypothesised that cumulative severe hypoglycemia might be the major factor in the slight performance IQ differences between diabetic patients and control subjects. The origin of the verbal IQ differences, although obscure, might be related to the social impact of the disorder.

Ten pre-senile Alzheimer's patients, 11 patients with Korsakoff's psychosis and 11 age- and pre-morbid intelligence-matched controls were given a test of Inspection Time, which estimates the efficiency of visual encoding or iconic memory. Alzheimer's patients had impaired Inspection Time while the Korsakoff group performed very similarly to the controls. Inspection Time performance correlated significantly with psychometric tests of cognitive ability and with clinical tests of cognitive ability (Mini Mental State Examination and Cambridge Mental Disorders of the Elderly Examination). The early stage of information processing measured by the Inspection Time procedure appears to be damaged by the Alzheimer's disease process, and to impose a rate-limiting effect on a wide variety of mental tests.