Supakatisant 2015.

Study characteristics
Methods	2‐arm randomised controlled trial.
Participants	Setting: antenatal care clinic at the Department of Obstetrics and Gynecology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand. Inclusion criteria: pregnant women with leg cramps (defined as: sudden tonic or clonic involuntary contraction of the gastrocnemius muscle associated with severe pain). 14–34 weeks of gestation, having pregnancy‐induced leg cramps at least twice a week. Exclusion criteria: other medical disease, concurrent obstetrics complication, other prescriptions for leg cramps, history of magnesium allergy, pregnant women with multifetal gestation, subsequently developed pregnancy‐induced hypertension and preterm labour treated with tocolytic agent.
Interventions	Experimental intervention: oral magnesium bisglycinate chelate (100 mg magnesium), 1 tablet, 3 times a day with meals, for 4 weeks. 43 women randomised (data for 41). Control: placebo, 1 tablet, 3 times a day with meals, for 4 weeks. 43 women randomised (data for 39).
Outcomes	50% reduction of number of leg cramps, 50% reduction of pain score of leg cramps. Side effects.
Notes	ISRCTN0389660. HW contacted the authors on 23/3/15 to request additional data on frequency and intensity of leg cramps after treatment, and side effects ‐ no response received. Dates of study: between June 2010 and August 2011 Funding sources: Grant for Development of New Faculty Staff, Chulalongkorn University. Declarations of interest: the authors declare that they have no conflicts of interest.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random number table using a block‐of‐4 technique, generated by co‐investigator who did not have patient contact.
Allocation concealment (selection bias)	Low risk	Sequentially‐numbered opaque plastic containers of identical size, shape and colour tablets.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Both healthcare providers and women were masked to treatment assignment.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Women self‐reported outcomes. The treatment assignment was not revealed until data collection was completed.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Three women left the study because of personal reasons and other 3 women were lost to follow‐up. 86 women were included in the intention‐to‐treat analysis by a ‘worst‐case’ scenario.
Selective reporting (reporting bias)	Low risk	Assessed from study protocol and published report, all prespecified outcomes reported.
Other bias	Unclear risk	Baseline characteristics appear similar between groups, although possibly placebo group had less frequent but more severe leg cramps.