The Case Presentation can be found on page 1234. DISCUSSION Severe hypokalemia is an unusual metabolic abnormality in children. Apart from a detailed clinical history and meticulous physical examination central to the diagnostic approach of a child with hypokalemia is the evaluation of the concurrent acid ⁄base status, the presence or absence of hypertension and of increased or normal urine potassium. In the presence of normal blood pressure and without urinary potassium losses, extrarenal causes of hypokalemia should be investigated. If hypokalemia is associated with hypertension, Cushing’s syndrome, renovascular diseases or causes of excess renin and mineralocorticoids should be sought (1,2). In infants, the combination of hypertension and hypokalemia is particularly suggestive of congenital adrenal hyperplasia due to CYP11B1 (11b-hydroxylase) deficiency (3). The lack of hypertension in our patient in addition to the normal values of serum 11-deoxycortisol, 11-deoxycorticosterone, dehydroepiandrosterone, dehydroepiandrosterone sulphate, androstenedione and testosterone ruled out congenital adrenal hyperplasia due to 11b-hydroxylase deficiency. In normotensive children, like our patient, the most common causes of hypokalemia include pyloric stenosis, protracted diarrhoea, skin losses and potassium-losing tubulopathies, such as Bartter and Gitelman syndromes (4). In children with compatible personal history, diuretics, amphotericin B and laxative abuse should also be considered in the differential diagnosis of hypokalemia. In the infant presented, pyloric stenosis and protracted diarrhoea were initially considered as the cause of severe hypokalemia with associated low urine potassium, but the history was not consistent with gastrointestinal losses, since the infant had mild non-projectile vomiting for 2 days only, without any associated diarrhoea. Bartter andGitelman syndromes were initially included in the differential diagnosis, but were unlikely by the lack of frank metabolic alkalosis, the consistently low urine potassium on several measurements and the lack of hypomagnesemia, a consistent feature of Gitelman syndrome and a frequent feature of Bartter syndrome type III (4–6). The negative history, the normal serum glucose and the nonconsistent history alsomade other extrarenal causes of hypokalemia, such as the use of insulin and of b2-agonists, highly unlikely. Absence of concurrent hypochloremia and metabolic alkalosis and the high serum chloride minimized the possibility of cystic fibrosis and congenital chloridorrhea (7). Finally, the low urine potassium along with the arterial pH of 7.45 excluded type I distal renal tubular acidosis, a condition that can be associated with hypokalemia in the context of metabolic acidosis due to defective urine acid secretion (8). When the dietary intake of potassium is reduced, depletion of potassiumandhypokalemia result because the renal excretionof thiscation fails todecreasepromptly.Eventually,when hypokalemia is established, renal potassium excretion is minimized, like in our patient, who at presentation had both severehypokalemiaandminimalurinepotassiumlosses. When severe, hypokalemia can induce rhabdomyolysis, but this was not seen in our patient who had a normal serum CPK (9). After meticulous history taking with emphasis on dietary intake, the mother admitted that the infant was exclusively fed with a mixture of water, flour and butter for at least 15 days due to his refusal to take the formula. Apparently, this is not an uncommon practice for infants with intolerance to breast milk or formula in rural areas of Bulgaria, the home Acta Paediatrica ISSN 0803–5253
Read full abstract
May 20, 2010
Biff F Palmer 
Open Access
