Patients who fit the entry criteria were given a physical examination and a questionnaire about physical activity, tobacco or illegal drug use, occupation, and leg cramps characteristics. To rule out other causes of cramps, we measured serum and urine magnesium, calcium, electrolytes, creatinine, glycemia, albumin, and acid-base status. Magnesuria was measured at the end of each period to evaluate compliance.
Randomization
Patients were enrolled consecutively as they met the entry criteria. We started with a first washout period of 4 weeks. During this period, patients stopped any treatment with magnesium and started taking placebo while recording the number of cramps, their severity and duration, and any sleep disturbance in a daily log. At the end of this period, those who still reported 6 or more cramps were randomized.
Patients randomly received magnesium or placebo in the first drug period, switching to the one not previously used in the second drug period. A second washout period was scheduled between the 2 drug periods Figure 1.
Each pill contained 900 mg of magnesium citrate or matched placebo (same appearance and taste). Patients were instructed to take one pill in the morning and another at bedtime. We determined the amount of magnesium citrate from recommended maximum doses. All the pills were obtained in the same pharmacy and were packaged and dispensed in the same place. Each packet contained 63 pills (56 pills for the 28 days of treatment and 7 to assess compliance) and was labeled with 3 letters: A and B for the drug periods and P for the washout periods. The codes were inside a sealed envelope opened at the end of the analysis.
Outcome Measures
The primary outcome measure was the number of muscle cramps, and the secondary outcomes were the duration and severity of cramps and any sleeping disturbance caused by the cramps.
Each patient had a daily log for each period, in which they recorded the outcome measures. Patients were instructed to record in the log each morning. The number and the severity of cramps were recorded on an analog numerical scale, and their duration was measured in intervals of 0 to 5 minutes, 5 to10 minutes, 10 to 30 minutes, and more than 30 minutes. For the evaluation of sleep disturbance, we used a numerical scale of 1 to 10, where 0 was “no sleep disturbance” and 10 was “could not sleep because of the cramps.”
All patients were contacted by telephone each week to assess compliance and to detect any problem with the treatment. Once a month, supjects were contacted personally to receive the new medication and the new daily log. Patients’ compliance with the treatment was evaluated by counting the number of pills that were left in the package, by phone calls to the participants, and by measurement of urinary magnesium at the end of each period.
Statistical Analysis
We analyzed the primary outcome measures using the Wilcoxon test and the Student t test for dependent variables. We estimated that a sample size of 42 patients was necessary to obtain a difference of at least 25% in the number of cramps between the 2 groups (a = 0.05; power = 90%; expected dropout rate = 15%).
Results
The baseline characteristics of randomized patients are listed in Table 1, and we show the sampling and eligibility process in Figure 2. From March 1996 to March 1997, we interviewed 309 patients. Only 93 met the entry criteria. After the first washout, we randomized 45 eligible patients, and 42 patients finished the study. None of the randomized patients had hypomagnesemia, hyponatremia, or hypokalemia in the initial laboratory determinations.
Twenty patients refused to participate after the first washout period because of diarrhea (4), family problems (5), abdominal pain (5), lack of time (2), complaint of weight gain (2), major surgery (1), and change of medical insurance (1). Twenty-eight patients were not eligible after the first washout period because they did not meet the criteria to continue. Three patients dropped out after randomization because of lack of time (2) or a change of medical insurance (1).
All studied end points are showed in Table 2. The mean number of cramps with placebo was 11.1 (SD ± 7.3) and with magnesium was 11.8 (SD ± 7.6). No statistical differences between placebo and magnesium were found.
We observed no differences in the percentage of common side effects (diarrhea, nausea, vomiting) between magnesium (10.7%) and placebo (10.1%).
We observed a significant decrease in the number of cramps between the first and second treatment months Table 3. This was probably because of a period-effect bias, because regardless of the treatment they received, the patients in the second round of treatment had fewer cramps than those in the first round.