In short, hypomagnesemia is associated with hypokalemia and hypocalcemia, and the clinical features closely resemble the features of an extremeyl low calcium- with tetany, seizures and postive Trousseau and Chvostek signs. You have either failed to absorb enough magnesium somehow, or (more likely) there is some sort of magnesium-wasting illness, with diarrhoea and renal tubular loss being the usual culprits. In epidemiological terms, 99% of all hypomagnesemia seen in hospital will be the consequence of loop diuretics, but many other "classical" causes would need to be mentioned in the exam answer, including hyperaldosteronism, hyperparathyroidism, alcoholism and pancreatitis.
Question 16 from the first paper of 2007 was the only ever question in the recent CICM history which asked about magnesium disturbance. The candidates were invited to "write a short note on hypomagnesemia". As far as literature goes, there is not much out there. Oh's Manual dedicates a paragraph each to both high and low magnesium levels, with a Blue Box (93.7, p. 957) describing the causes of hypomagnesemia. The contents of this box is presented below in a virtually unchanged state (apart from correcting the spelling of cyclosporine). Beyond Oh's, one may wish to explore the 1999 JASN article by Zalman Agus.
A better way to organise the list of causes would be by pathophysiological disturbance, as below.
Sequestration and chelation
Unclear association with low magnesium
Associated biochemical abnormalities
This list of clinical features has been constructed from the brief blurb in The Manual, as well as the excellent 1999 Agus article. Agus reports on some heroic early studies, such as the voluntary magnesium depletion experiments performed by Maurice Shils in 1969 (he somehow convinced several NG-fed head and neck cancer patients to receive a magnesium-depleted diet, "prepared in the research kitchen and composed of refined nutrients which were purchased in bulk and analyzed prior to use"). The volunteers developed features which would be misteken for hypocalcemia if you didn't know their blood results: tetany, positive Chvostek and Trousseau signs, and generalized convulsions were observed. Given that hypocalcemia typically develops at the same time,
For those people who (like me) can never remember which is which;
This would not have ever aroused any interest in most people, but hypomagnesemia as a cause of seizures has come up as a highly specific answer in Question 17.2 from the second paper of 2016. There, the college presented us with scenario of a post-ictal Crohns patient with an ileostomy. Magnesium level was 0.28 mmol/L, suggesting that it was the culprit. This would not be the first time it was associated with seizures: case reports have been published (eg. Fernández-Rodríguez et al, 2007) and the seizures disappeared when the magneisum was corrected.