To the Editor: The recent article by Roberts et al1 requires further comment. In anorexia nervosa, hypercalcaemia is extremely unusual because patients are likely to be under- or malnourished, with consequent hypocalcaemia and hypovitaminosis D, rather than the opposite, as implied by the authors. In one of their references,2 hypercalcaemia is only briefly included, and the mechanism is not discussed or substantiated. Another of their references3 does not include hypercalcaemia at all as a metabolic disturbance. In addition, hypercalcaemia in itself is not a diagnosis, and indicates a significant underlying pathophysiological disturbance whose differential diagnoses need to be carefully dissected.

Where there are problems with the interpretation of calcium homoeostasis, such as renal impairment and hypoalbuminaemia, ionised calcium should be measured because it is the physiologically active agent. In the first patient described by Roberts et al, although the investigations are incomplete, primary hyperparathyroidism (PHPT) needs to be carefully considered, as the parathyroid hormone (PTH) level is not normal in the setting of hypercalcaemia. The normal physiological response would dictate that the PTH level should be low to suppressed. Nephrocalcinosis and renal impairment then fit snugly into the diagnosis of PHPT.4 Phosphate level, expected to be low in this condition, might have been masked by exogenous phosphate supplement.

For Patient 2, stool electrolyte analysis could further support the presence of laxative misuse and aid in the interpretation of urinary results. Faecal fluid in this situation should be high in sodium, potassium and calcium concentrations. The low urinary sodium and calcium levels are therefore appropriate, and indicate relatively intact tubular function. The diagnosis of hypercalcaemia is thus difficult without a PTH measurement and in the presence of gastrointestinal confounders, even if all investigations were available. Nevertheless, familial benign hypocalciuric hypercalcaemia (FBHH) is a strong probability given the low urinary calcium excretion, especially before iatrogenic manipulation of calcium and phosphate homoeostasis. Nephrocalcinosis can theoretically occur in FBHH.5 It is important that the diagnosis is made in both cases, so that a familial study can be carried out if indicated given the patients’ age. One may also wonder if the intermittent hypercalcaemia contributed to or aggravated the psychiatric disturbance in both patients.

In reply: Disturbances of calcium metabolism in anorexia nervosa are complex, particularly with associated renal insufficiency. Although hypocalcaemia is observed in patients with anorexia nervosa, our article attempted to highlight nephrocalcinosis and hypercalcaemia.

We agree that hypercalcaemia is not in itself a diagnosis, and identifying the underlying pathology is essential. Ingestion of vitamin D preparation remains a likely explanation for the hypercalcaemia observed in Patient 1; however, primary hyperparathyroidism was considered as a possible differential diagnosis. Patient 1 had two normal parathyroid hormone tests in the setting of hypercalcaemia and renal impairment. This is consistent with secondary hyperparathyroidism and vitamin D ingestion as documented. The coexistence of primary hyperparathyroidism cannot be excluded. Ionised calcium may be a useful measure if the patient had hypoalbuminaemia, and this would be our normal practice.

Patient 2 had ionised calcium measured twice (one result high, one low), but these added little to the case description and message of the article. Faecal electrolytes were not measured in Patient 2. However, we acknowledge the potential utility of this investigation when interpreting electrolyte disorders. We also agree that familial or genetic conditions should be considered if clinically appropriate.