In the first chapter, the definitions, frequency, causes and consequences of clinical
hypoglycaemia are discussed. Previous research on the cognitive effects of acute
hypoglycaemia is reviewed in the second chapter. Four studies of the effects of blood
glucose manipulation on cognitive function are then described.
In Study 1, the effect of acute hypoglycaemia on memory consolidation was studied in
non-diabetic subjects. Functional brain imaging was used to identify regions with
altered metabolism following hypoglycaemia. No effect of hypoglycaemia was seen on
memory or brain imaging results, suggesting that acute hypoglycaemia does not affect
previously-formed memories, though other possibilities are discussed.
In study 2, the effects of acute hypoglycaemia on learning and recall were studied, and a
novel test of prospective memory, mimicking memory in real life, was developed.
Subjects had type 1 diabetes and either normal or impaired awareness of hypoglycaemia.
Both learning and recall were impaired during hypoglycaemia. Prospective memory was
impaired to a similar degree, reinforcing the ecological validity of laboratory cognitive
function testing. No statistically significant differences between normal- and impaired-awareness subjects were seen, but this may have been due to lack of statistical power.
A previous study had unexpectedly reported that high-level cognitive functions were
unaffected by hypoglycaemia. Study 3 re-addressed this issue using a more difficult
cognitive test to exclude the possibility of a ceding effect, and impairment by
hypoglycaemia was confirmed. A separate task, intended to indicate which of speed and
accuracy of cognitive processing is primarily affected, yielded no significant results.
In study 4, the symptoms of hyperglycaemia reported by 400 people with insulin-treated
diabetes were examined using factor analysis. Four categories were identified, and
labelled as 'osmotic', 'neurological', 'mental agitation' and 'malaise'. Substantial overlap
with hypoglycaemia symptoms was noted. The mean blood glucose level for symptom
onset was 15 mmol/1, with a tendency for less intense symptoms and a higher symptom
threshold in older people and those with impaired hypoglycaemia awareness.