Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Quinine for nocturnal leg cramps

Analysing visits to the Bandolier Internet site showed that quinine and nocturnal leg cramps ( Bandolier 12 ) was one of the most frequently viewed pages. So we sought any more recent information since a meta-analysis published in 1995. A second meta-analysis [1] was published in 1998, and was interesting because it included unpublished material and provided some empirical evidence of publication bias.


Searching using three computerised databases was up to July 1997. Unpublished data was found through examining an FDA report, enquiries to British and German regulatory authorities, and pharmaceutical companies.

For inclusion studies had to be randomised and double blind, and to be in ambulatory patients. Information was abstracted on age and sex of patients, treatment duration, outcome measures, adverse effects and washout periods. The main efficacy outcome was the reduction in nocturnal leg cramps in a four-week period, severity of cramps, and their duration.


There were four published studies with 73 patients and three unpublished studies with 336 patients. The inclusion criteria meant that patients had to have more than two cramps a week. The dose of quinine was between 200 and 500 mg quinine, and the treatment period was one to four weeks (two weeks or more in six of the seven studies). All studies gave the number of cramps, six described their severity, and only one their duration.

With placebo the number of cramps in a four-week period (by extrapolation if studies were shorter) was 17 (about four a week) for all studies, 21 in the published studies and 16 in the unpublished studies (Table 1). Quinine reduced the number of cramps in a four-week period by 4, 9 and 3 respectively (Table 1). There was also evidence that the severity of the cramps was reduced.

Table 1: Results of published and unpublished randomised studies of quinine for nocturnal leg cramps

  Number of: Over a four-week period (95% CI)
  Trials Patients Cramps on placebo Cramps avoided with quinine
Published 4 73 20.6 (14.9 to 26.6) 8.8 (4.2 to 13.5)
Unpublished 3 336 16.3 (14.7 to 17.9) 2.5 (0.2 to 5.6)
All 7 409 17.1 (15.4 to 18.8) 3.6 (2.2 to 5.1)

Published studies showed a greater effect than unpublished studies (Table 1). Larger studies showed lower and more consistent effects than small ones (Figure 1).

Figure 1: Individual trials of quinine (filled circles unpublished, open circles published)

Tinnitus was the only adverse effect that occurred with significantly higher frequency when subjects took quinine rather than placebo (8 of 397 patients).


This is interesting stuff from a number of aspects. Firstly, it confirms that quinine is effective for nocturnal cramps. For someone who has about four cramps a week, quinine can be expected to reduce this to three a week.

There was clear publication bias. Unpublished trials had a smaller size of effect than published ones. But the published trials were small, and much of the extra effect came from two trials with 27 patients in one and eight in the other. The lesson may not be about publication bias, but the possibility of small trials, however well conducted, giving the wrong answer because of the random play of chance.


  1. M Man-Son-Hing & G Wells. Quinine for nocturnal leg cramps. A meta-analysis including unpublished data. Journal of General Internal Medicine 1998 13: 600-606.
previous or next story in this issue