0737 Diabetes and cerebral oedema

Slow rehydration after resuscitation in infants presenting with hypertonic dehydration (usually related to gastroenteritis) has been preached and practiced for many years in wards and ‘drip rooms’ around South Africa. Such infants frequently presented with non-ketotic hyperglycaemia, and the question was often asked as to why diabetic infants presenting with a similar metabolic picture in terms of glucose, electrolytes and hyperosmolarity were rehydrated rapidly while the hypertonic gastroenteritis patients were treated slowly to prevent cerebral complications. Much changed some 10 years ago when CT scans showed that cerebral oedema was frequently associated with diabetic ketoacidosis (DKA), with concerns raised that fluid management protocols contributed to the problem. Current research is producing some interesting findings that include the following: Narrowing of the lateral ventricles is evident in >50% of children with DKA and this is frequently associated with mild mental status abnormalities (GCS scores <15), however significant cerebral oedema occurs much less frequently (0.5-2% of DKAs); a low initial pCO2 is significantly associated with the ventricular narrowing; hypokalemia is more common in newly-diagnosed DKA than in established disease; DKA mortality risk is 0.15-0.31%, with cerebral oedema accounting for ⅔ of this mortality; treatment factors are not the cause of the cerebral oedema while risk factors include younger age, lower pH, degree of hypocapnia and severity of the dehydration. DKA is accompanied by increased cerebral blood flow, suggesting a vasogenic process; intervention that includes mannitol, intubation and hyperventilation should be reviewed since the latter in particular may adversely affect the situation. Of particular interest in one study was that even if the target for rehydration and correction of DKA was 48 hours, the mean for >600 patients was 11.6±6.2 hours. Finally, several studies have recently alerted paediatricians to the increasing incidence of hyperglycaemic hyperosmolar non-ketotic dehydration in obese adolescents with early-onset type 2 diabetes – this is obviously something to look for as the prevalence of obesity in adolescents is on the increase worldwide.

Read more:
Indian J Pediatr 2006; 73: 55-60
Pediatr Diabetes 2007; 8: 142-9 and 2006; 7: 75-80
J Pediatr Endocrinol Metab 2007; 20: 5-18