Background Alcohol consumption has been associated with reduced risk for a number of diseases. A recent study has shown reduced levels of symptom reporting among drinkers compared to non-drinkers in patients with fibromyalgia (FM), a condition characterised by chronic widespread pain (CWP).

Objectives To determine whether levels of alcohol consumption were associated with the prevalence of CWP in a population survey. To determine among those with CWP, whether consumption was associated with pain-related disability.

Methods The MUSICIAN study surveyed patients registered at general practices in the UK. Information collected included pain and alcohol consumption. Respondents were categorized according to whether they met the ACR 1990 definition of CWP and those with CWP were further categorized by presence of highly disabling pain, as measured by a Chronic Pain Grade (CPG) of 3 or 4. The study determined whether they had ever drunk alcohol regularly and, if so, how much they currently drank on average per week. Information was available on gender and potential confounders of the relationship (age, body mass index, employment status, smoking). For each level of alcohol consumption (units/week: lifetime non-drinkers [referent]; 0-5; 6-10; 11-12; 21-35; 35+) prevalence of CWP was calculated, as well as the proportion with disabling pain in those with CWP. The association between alcohol consumption and CWP/pain-related disability was examined using logistic regression. Analyses were stratified by sex and adjusted for confounders; adjusted odds ratios (aORs) and 95% confidence interval (CI) are reported.

Results 13587 respondents provided data required for analysis (mean age 55 years, 56.8% female). In males, CWP prevalence decreased with increasing alcohol consumption from 16.7% in non-drinkers to 11.8% in those drinking 21 to 35 units/week of alcohol (aOR 0.69, 95% CI 0.52-0.92), and was greatest in those drinking more than 35 units/week (20.1%, aOR 0.93, 0.63-1.37). In females, the prevalence of CWP amongst non-drinkers was 21.0% and this decreased with higher consumption to 14.8% in those drinking 11 to 20 units/week (aOR 0.76, 0.61-0.94), before increasing again in those drinking 21 to 35 units (18.3%, aOR 0.90, 0.61-1.34) and more than 35 (20.6%, aOR 0.70, 0.29-1.74). 2060 respondents reported CWP (mean age 56 years, 64.2% female). In males with CWP, disabling pain decreased from 45.5% in non-drinkers to 15.5% in those drinking 21 to 35 units/week (aOR 0.30, 0.14-0.65), and was greatest in those drinking more than 35 units (55.0%, aOR 1.09, 0.45-2.67). In females with CWP, disabling pain decreased from 47.5% in non-drinkers to 17.5% in those drinking 11 to 20 units/week (aOR 0.35, 0.20-0.59), then rose in those drinking 21 to 35 units (26.5%, aOR 0.44, 0.18-1.05) and more than 35 (28.6%, aOR 0.16, 0.02-1.28).

Conclusions Moderate alcohol consumption was associated with lower CWP prevalence, and strongly associated with lower levels of disability in those with CWP. A potential biological mechanism is alcohol's agonist effects on the neurotransmitter γ-aminobutyric acid (GABA), and disruptions to GABA pain inhibitory pathways have been suggested in persons with FM. Further investigation of the mechanism for these associations is required, specifically whether the excess of highly disabling pain in lifetime non-drinkers with CWP can be explained by other lifestyle or psychosocial factors.

Acknowledgements The following are members of the MUSICIAN study team: Gary Macfarlane (Principal Investigator), John McBeth (Investigator), Deborah Symmons (Investigator), Karina Lovell (investigator), Philip Keeley (Investigator), Phil Hannaford (Investigator), Chrysa Gkazinou (Trial manager), Marcus Beasley (Research Assistant), Elizabeth Jones (PhD student), Gordon Prescott (Statistician), and Steve Woby (Investigator).

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.2503