Association Between Alcohol Consumption and Nocturnal Leg Cramps in Patients Over 60 Years Old: A Case-Control Study

First of all, we would like to thank the Medical Students of the University of Illinois College of Medicine Rockford for their interest in our study. We are honored for the close look the two groups gave to our article and the relevant commentaries they have made.

Regarding the remarks about NLC epidemiology, we suggest to refer to the paper mentioned in our bibliography[1]. To our knowledge, there is no international study about NLC prevalence. We take their point that more statistics from different regions of the world would be valuable to help understand NLC. A Brazilian team is currently working on a new prevalence study based on an updated version of the questionnaire used in our study.

About the age group, as we stated in our previous article1, the limit was chosen according to the literature existing about NLC prevalence.

Bell et al stressed the lack of a scientific justification supporting the hypothesis of an association between NLC and alcohol use. In fact on one hand, some cases of cramps related to alcohol use were described in the literature[2]. On the other hand, we are a team of practicing family physicians in a wine growing region. We are regularly meeting patients that are associating their cramps with their alcohol consumption, especially white wine. NLC are associated with many representations and popular believes, whether for the treatment[3] or the trigger. As data about environmental factors that may trigger cramps are lacking, we found relevant to test this hypothesis supported by both case reports and our clinical experience. In different fields, studies have been made to explore popular believes (birth rates and lunar phases[4-5], weather and pain[6-7] or specifics symptoms[8]...).

In our study we decided to focus only on biomedical factors. In France, collecting data about ethnicity is not current and need to be ethically justified. In our population, the vast majority was Caucasian. But the remark is clearly relevant considering the variation of the repartition of muscles fibers between different ethnic groups[9].

Our choice to recruit people into Family physicians practices is supported by the organization of the French healthcare system. Family physicians are gate keepers and the referent healthcare provider. In this way, people above 60 years old without a referent Family physician are rare. By recruiting into the practices we had more chance to get diversity in our population, and we focus on the population targeted by intervention promoted by Family physicians.

We had to create our screening questionnaire because of the lack of validated instrument. The aim of the questionnaire was to collect biomedical data and clinical information based both on the patient declaration and the medical record. During the collecting phase, we controlled the quality of the process through the confrontation of audio recording of the encounter, content of paper questionnaires in addition to the data entered eCRF.

The large odds ratio is due to the variability of our population and the lack of power. A bigger study would be needed to narrow it. But we are in a Bayesian paradigm, that helps to take a better account of reality and prior data. In Bayesian, the posterior probability is as important as the credibility interval. In our case the PP is 99, 82% which is high.

We compared non-consumers and consumers of alcohol. We decided to consider as non-consumers people who had declare consuming alcohol less than once a week because alcohol consumption in adults over 60 years old in France (especially in our targeted area, the Alsace region) is common. We tried to search for a threshold but could not identify one.

Regarding the modus operandi of alcohol consumption, the remark is pertinent. Binge drinking seems to be more common among a younger population than ours. Our questionnaire was not design to identify specific patterns of consumption. It was not possible to perform sub-groups analysis due to the lack of power of our study.

The Food frequency questionnaire we used is composed of 19 pages with questions and a booklet with pictures to help patients evaluate their consumption of various foods. Regarding alcohol consumption, participants were asked to describe the type of alcohol consumed. This questionnaire was used before in an important international study, EPIC: The European Prospective Investigation into Cancer and Nutrition. It was designed to limit the recall bias. We also tried to limit recall bias by taking time to explain the questionnaire and by offering to the participants enough time to fill it in. Patients presenting known cognitive impairs were not recruited (exclusion criteria).

Regarding the remark about other environmental factors, we also studied physical activity that's the focus of an upcoming article. We are glad that our study may enhance the discussion about alcohol consumption with patients and hope future studies will reinforce our findings.

1. Maisonneuve H, Chambe J, Delacour C, et al. Prevalence of cramps in patients over the age of 60 in primary care: a cross sectional study. BMC Fam Pract. 2016; 17(1): 111.
2. Blank NK. Alcohol ingestion and muscle spasms. JAMA. 1979;241(15):1574-1574
3. Lorenzo M, Schaeffer M, Haller D, Maisonneuve H. Treatment of nocturnal leg cramps by primary care patients over the age of 60. Family Practice 2018; 35(1)1: 29-33
4. Arliss J, Kaplan E, Galvin S. The effect of the lunar cycle on frequency of births and birth complications. Am. J. Obstet Gynecol, 2005 May;192(5):1462-4
5. Margot JL. No evidence of purported lunar effect on hospital admission rates or birth rates. Nurs Res. 2015;64(3):168-73; discussion 173-5.
6. Timmermans EJ, Schaap LA, Herbolsheimer F et al. The Influence of Weather Conditions on Joint Pain in Older People with Osteoarthritis: Results from the European Project on OSteoArthritis. J Rheumatol. 2015 Oct 1;42(10):1885.
7. Jena AB, Olenski AR, Molitor D, Miller N. Association between rainfall and diagnoses of joint or back pain: retrospective claims analysis. BMJ. 2017;359:j5326.
8. Schmidt W, Sarran C, Ronan N, et al. The Weather and Meniere's Disease: A Longitudinal Analysis in the UK. Otol Neurotol. 2017;38(2):225-233.
9. Silva, A, Shen W , Heo M et al. Ethnicity-related skeletal muscle differences across the lifespan. Am. J. Hum. Biol., 2010;22: 76-82.

Competing interests: None declared

The primary purpose of the article was to investigate an association between the prevalence of nocturnal leg cramps and alcohol consumption. Nocturnal leg cramps (NLC) can disrupt sleep and reduce quality of life. Despite affecting more than 46% of French adults over age 60 in 2012, there are limited treatment options for nocturnal leg cramps. If a connection was established between alcohol consumption and NLC, reduction in alcohol consumption might help practitioners reduce the incidence of NLC among their patients. The authors aimed to determine this using a case control study.

Overall the design of the study seemed practical and well executed. During our discussion of this paper, our critique largely centered around the way the data was presented. We felt a number of changes could have been made to make the results and findings more clear.

Our group thought Table 1 should have included a total sample size (n) row in the "Age Group" section underneath the row title "? 80." It would allow the reader to quickly confirm that the number of cases and controls were 70 each as stated in the methods section. Under the heading of medications in Table 1, there were four times as many medications that fell into the "Others" category when compared to the next highest medications category: "Statins." These other medications were not listed in the appendix and should have been to allow the reader to fully rule out medications as a confounder in this study. Our group thought the appendix should have included a copy of the questionnaire used in this study. Time to fill out questionnaires was listed as possibly discouraging participation. Including the questionnaires would have provided the reader with an idea of what type of patient would be more inclined to participate in this study and help the reader decide whether there was any selection bias in this study. We also thought alcohol consumption was not measured accurately, but was only approximated and assumed. Although the authors did ask participants to quantify their weekly consumption of alcohol, they did not ask the participants to discuss how long they had been drinking or what type of beverage they consumed.

One of the main discussion points was an inadequate reporting of the analysis methods used. The authors did not make clear what the variables included in the regression analysis were which made it hard to interpret the findings. We also discussed that the authors declared a "strong association between the global consumption of alcoholic beverages and NLC," but we did not feel as though they had supported this declaration sufficiently. However, in the results section the authors did report a significant odds ratio of 6.5 (95% credibility interval: 1.68-38.05). As mentioned earlier however it is not clear what other variables were included in this analysis so it is difficult to interpret this finding.

Future work should study participants for a longer period of time. We also would have made a more extensive list of inclusion and exclusion criteria to account for more factors that might be related to NLS- like evaluating each participant's weekly level of physical activity. We also would have tried to make the methods and results sections more clear- we felt as though the evidence supporting their conclusion was not as explicit or strong as it could have been.

Based on the reported odds ratio of 6.5 (95% credibility interval: 1.68-38.05), it seems statistically likely that there is an association between regular alcohol consumption and NLC. These results may encourage practitioners to discuss reducing or eliminating alcohol consumption in patients with NLC. Practitioners will still need to communicate that findings reported in this study have only been reported in one study and that the level of evidence is level IV (low-moderate). Our group felt as though the evidence wasn't strong enough to base any clinical recommendations on. This study was well designed and had an interesting outcome, but our group felt changes could have been made to help the reader better rule out confounding factors and biases. These additions would have also offered increased clarity and further substantiated the findings of an already well reasoned and rigorous study.

Competing interests: None declared