Magnesium is one of the main electrolytes in the human body and is involved in multiple mechanisms necessary for the correct functioning of the body.

The normal range for magnesium levels in serum is 0.7–1.03mmol/L. Hypomagnesaemia (level <0.7mmol/L)1 can manifest with different symptoms, such as weakness, cramps or heart arrythmias, although it is usually asymptomatic if the levels are greater than 0.5mmol/L.1 Some genetic disorders are associated with increased renal excretion of magnesium.2

We present the case of a boy aged 9 years in follow-up in the department of paediatric endocrinology due to colloid cysts in the thyroid gland and transient elevation of thyroid-stimulating hormone (TSH) (peak level, 11.67mU/L; normal range, 0.60–4.84mU/L) with a normal level of free thyroxine (FT4) normal and negative thyroid peroxidase (TPO) antibodies. Thyroid function normalised and the cysts stabilised, but the patient exhibited persistent hypomagnesaemia during the follow-up (levels ranging between 0.53–0.57mmol/L).

The relevant findings of the personal history were the diagnosis of attention-deficit hyperactivity disorder (ADHD) managed with methylphenidate, pectus excavatum and malposition of teeth.

The patient had never experienced headaches, dizzy spells, muscle weakness or cramps.

There was no family history of interest.

In the physical examination, the patient was not obese (body mass index: 15.68kg/m2, 20th percentile, z=−0.6) and neurodevelopment was adequate for age.

In the laboratory tests performed to identify the aetiology of hypomagnesaemia, the levels of albumin, parathyroid hormone (PTH), vitamin D, calcium and creatinine (Cr) were normal (estimated glomerular filtration rate, 169.9mL/min/1.73m2). The fractional excretion of magnesium was elevated (8%; normal range, 2%–4%), as was the urine calcium level (Ca/Cr 0.32mg/mg; normal range, <0.2mg/mg) which suggested that the cause of hypomagnesaemia was renal.

An abdominal ultrasound ruled out renal and urinary tract anomalies, such as nephrocalcinosis.

Genetic testing, consisting in exome sequencing of 9 genes associated with hypomagnesaemia, identified a de novo, heterozygous likely pathogenic mutation in exon 1 of gene CNNM2 (c.1310G>A; p.(Gly437Glu)) associated with autosomal dominant renal hypomagnesaemia type 6 (HOMG 6; OMIM 613882).

Initially, given the absence of symptoms, treatment was limited to a diet rich in magnesium, but since the low levels of magnesium persisted (0.57mmol/L), supplementation with magnesium tablets (300mg/day) was initiated.

The kidneys are the main regulators of the concentration of magnesium in blood. Certain proteins, like cyclin M2 (CNNM2) facilitate the excretion of magnesium in the kidney. This protein is encoded by gene CNNM2, and changes in this gene can cause HOMG6 or primary hypomagnesemia, seizures, and impaired intellectual development 1 (HOMGSMR1; OMIM 616418). In addition, some variants at the CNNM2 locus have been associated with neuropsychiatric disorders, intellectual disability and language disorders.2

In the variant identified in our patient, the amino acid change p.(Gly437Glu) occurs in a domain that is important for the function of the protein (Bateman domain, amino acids 429–578). It has been hypothesised that the activity of protein CNNM is regulated by conformational changes in this domain associated with the binding of Mg2+ and ATP.3 The replacement of a glycine, a small hydrophobic amino acid, by glutamic acid, a polar amino acid with a negative charge, causes an essential change in the protein. Furthermore, in silico tools used for pathogenicity prediction (PolyPhen2, Sift, Mutation Taster, etc.) predict a pathogenic effect.

Hypomagnesaemia was present in the patient before he started treatment with methylphenidate, so this drug did not seem to be the cause.

Most described individuals with changes in the CNNM2 genes present severe neurologic symptoms associated with hypomagnesaemia.4 In our patient, ADHD is the only neurologic manifestation present to date. Petrakis et al.5 also described a case of HOMG6 and ADHD and discussed the possibility of a milder neurologic phenotype in HOMG6. However, given the high prevalence of ADHD in the paediatric population, it is not possible to rule out a chance association with hypomagnesaemia.

Our patient did not have nephrocalcinosis, a feature described in other individuals with hypomagnesaemia and disorders of phosphate and calcium metabolism secondary to changes in the CNNM6 gene.6

As was the case of our patient, most changes in the CNNM2 gene are sporadic.

To date, changes in this gene have not been associated with thyroid disorders, and we did not find any reports in the literature of variants causing both hypomagnesaemia and thyroid disorders.

Changes in the CNNM2 gene are infrequent, so we thought it would be relevant to report this case. Hypomagnesaemia, which may be oligosymptomatic, and disorders like ADHD could be manifestations of the same disease.