Objectives The aim of this study was to investigate the association between the catechol-O-methyltransferase (COMT) ‘pain sensitivity’ haplotypes and chronic widespread pain (CWP) in two distinct cohorts.
Methods Cases of CWP and controls free of pain were selected from two population-based studies: the Epidemiology of Functional Disorders study (EPIFUND) (UK) and the European Male Ageing Study (European). The number of cases and controls were 164 and 172, and 204 and 935, respectively. Identical American College of Rheumatology criteria were used in both studies to ascertain CWP status. The EPIFUND study had three time points and cases were classified as subjects with CWP at two or three time points and controls as those free of pain at all three time points. Four single nucleotide polymorphisms (SNP): rs6269, rs4633, rs4818 and rs4680 (V158M) were genotyped using Sequenom technology. Allele and genotype frequencies were compared and haplotype analysis was conducted using PLINK software.
Results No differences in allele or genotype frequencies for any of the four SNP were observed between cases and controls for either cohort. Haplotype analysis also showed no difference in the frequency of haplotypes between cases and controls.
Conclusions There was no evidence of association between the COMT ‘pain sensitivity’ haplotypes and CWP in two population-based cohorts.
Statistics from Altmetric.com
The effect of a non-synonymous single nucleotide polymorphism (SNP), rs4680 (V158M), in the gene coding for the enzyme catechol-O-methyltransferase (COMT) has been widely investigated in genetic association studies of chronic widespread pain (CWP). COMT degrades catecholamines, such as dopamine and epinephrine and catechol-containing drugs. The V158M SNP has been shown to affect the thermostability of the enzyme resulting in reduced activity1 and to influence pain sensitivity by means of β-endorphins.2 An overrepresentation of the MV and MM genotypes in patients with fibromyalgia compared with healthy controls has been reported3 4 but also refuted.5 The sample size for each of these studies was relatively small, with a maximum of 91 subjects in the case or control group. A larger population-based study observed no association between V158M and CWP6; however, that study did not use American College of Rheumatology criteria to define CWP and controls could have had acute pain.
A previous study of common SNP in the COMT gene identified three common haplotypes, which accounted for 96% of all haplotypes observed.7 These haplotypes discriminated between low pain sensitivity (LPS), average pain sensitivity (APS) and high pain sensitivity (HPS) and predicted the risk of developing temporomandibular disorder (TMD).7,–,9 The haplotypes also affect COMT activity, with individuals possessing the LPS and HPS haplotypes having high and low activity, respectively. Both the LPS and HPS haplotypes posses the val version of rs4680, and although this is the only SNP in the haplotype that has a purported function it does not explain the differences in pain sensitivity observed.7 Differences in the secondary protein structure of the messenger RNA of the haplotypes have also been reported emphasising the importance of understanding how SNP interact in haplotype form.10 11 One study found the HPS haplotype to be associated with fibromyalgia and symptom severity in Spanish subjects.12 This study was small, the reported association with fibromyalgia has not been confirmed elsewhere, and it is not clear whether the reported association was directly related to CWP or other symptoms associated with fibromyalgia. The aim of the current analysis was to investigate whether the proposed COMT pain sensitivity haplotypes were associated with CWP in two distinct population-based cohorts.
We used data from the Epidemiology of Functional Disorders (EPIFUND) study, a prospective cohort of individuals from the north-west of England,13 and the European Male Ageing Study (EMAS).14 In both studies, pain status was ascertained from data collected in postal surveys that used identical schedules of questions. Subjects were asked to shade on a body manikin any area they had experienced pain for 1 day or longer in the past month. The American College of Rheumatology definition, as used in the criteria for fibromyalgia, was used to classify CWP: pain in contralateral quadrants of the body and the axial skeleton that has been experienced for 3 months or longer.
DNA samples were collected with informed consent from 1189 subjects in EPIFUND (buccal cells) and from 2636 subjects in EMAS (whole blood). A 90% success rate for samples and SNP was used. Following DNA quality checks, 994 and 2285 subjects who had complete pain information from EPIFUND and EMAS cohorts, respectively, formed the population from which cases and controls were identified. In EPIFUND, pain status was assessed at three time points. Subjects who reported CWP at two or three time points were classified as cases (n=164, 66% female, median age 52.6 years; 95% CI 50.3 to 53.9) and subjects who were pain free (reported no pain lasting 1 day or longer in the past month) at all three time points were classified as controls (n=172, 57% female, median age 48.5 years; 95% CI 46.4 to 51.6) Ethical approval was received from the local research ethics committee. EMAS is a cohort study in which men were recruited from the population registers of eight European cities (Florence, Leuven, Lodz, Malmo, Manchester, Santiago de Compostela, Szeged and Tartu). Ethical approval for this analysis was received from seven centres. CWP cases (n=204, median age 59.1 years; 95% CI 56.5to 61.9) and pain-free controls (n=935, median age 59.5 years; 95% CI 58.2 to 60.7) were identified.
The four SNP that comprise the COMT ‘pain sensitivity’ haplotypes (rs6269, rs4633, rs4818 and rs4680) were genotyped using Sequenom technology. Haplotype analysis was carried out in PLINK15 using an expectation–maximisation algorithm to infer haplotypes and estimate population frequencies of the three common haplotypes formed from the four SNP: GCGG (LPS), ATCA (APS) and ACCG (HPS). This provides a Wald statistic p value for the frequency of each haplotype between cases and controls compared with the other haplotypes combined and provides an overall, or omnibus, p value, which assesses the differences in frequency across all the haplotypes. The analysis in EMAS subjects was adjusted for study centre as minor allele frequency (MAF) (p<0.001) and CWP prevalence (p<0.001) differ across centres. With 80% power and 5% type I error, the haplotype analysis in EPIFUND was powered to detect an allelic OR of 2.0 when comparing the HPS haplotype with others and EMAS was powered to detect an allelic OR of 1.7.
All SNP were in Hardy–Weinberg equilibrium (p>0.01) in both cohorts. The linkage disequilibrium between the four SNP in this haplotype block was very similar for the two cohorts (figure 1). Similar frequencies of each haplotype were observed in the two studies (table 1) and there were no significant differences in the frequency between cases and controls in either study. The frequency of the HPS haplotype (ACCG) was identical in cases and controls in EPIFUND (9%); in EMAS the frequency was 12% in cases compared with 9% in controls; however, this difference was not significant. The omnibus test statistic showed that there was no difference between cases and controls across all haplotypes in both studies. Comparison of the frequency of alleles and genotypes between cases and controls for the V158M (rs4680) SNP showed no significant difference between cases and controls in either study (table 2). There was no association between CWP and the three other COMT SNP genotyped (data not shown). The single SNP analysis in EPIFUND was powered to detect an allelic OR of 1.6 and in EMAS an allelic OR of 1.4, based on a MAF of 0.40 or greater (MAF range from 0.4–0.5)
There are a number of issues to consider when interpreting these results. Previous studies have tended to be among women with fibromyalgia and the relationship with COMT may be sex specific. However, we found no evidence for SNP–gender interaction in the EPIFUND cohort and SNP–CWP associations remained non-significant when we excluded men from the EPIFUND cohort (data not shown).
Selecting pain-free controls maximised the likelihood of observing true associations. The inclusion of subjects without CWP as controls would have introduced heterogeneity into the control group as some non-CWP individuals would have pain (eg, TMD) that may be associated with the genes of interest. Another way to treat those subjects was to classify them as an ‘intermediate’ group that had some pain but not CWP. An association between COMT and the intermediate group compared with the pain-free controls would suggest that any association with COMT was with pain per se, and not specifically with CWP. We have conducted this analysis and found no association.
Subjects of self-reported non-European ancestry were excluded from the analysis of the EMAS cohort and there was no evidence of between-centre heterogeneity (tested by SNP–centre interaction terms). Ethnicity information was not available in the EPIFUND study, although 90% are thought to be white British.16 Therefore, although it must remain a possibility, it is unlikely that population stratification would have occurred.
This is the first study to investigate the frequency of the COMT pain sensitivity haplotypes7 in subjects with CWP. Our findings suggest that they do not play a role in CWP susceptibility. A recent small study (n=58) of patients with shoulder pain receiving surgical pain relief found no differences in COMT haplotype frequencies among patients when compared with ‘healthy’ subjects.17 The COMT ‘pain sensitivity’ haplotypes may be associated with the psychological state,17 18 which affects an individual's interpretation of pain, although the relationship is unclear.
The data surrounding COMT is equivocal. There are several putative explanations for this, including false positive associations with fibromyalgia and other pain disorders. Alternatively, the genetics of CWP may differ from that of other pain syndromes. While CWP is the cardinal feature of fibromyalgia, a diagnosis of fibromyalgia requires the presence of other key symptoms including fatigue, and studies of fibromyalgia and CWP may not be directly comparable. It is also possible that pain severity or other non-pain factors drive the associations reported between pain and COMT in pain disorders. Phenotypic heterogeneity in CWP may be greater than in disorders such as fibromyalgia and TMD, and the current study may be underpowered to detect the effect of the pain sensitivity haplotypes. Establishing the true pain–COMT relationship will help move the understanding of the genetics of pain forward by determining a role of COMT in pain and directing the focus of attention to other potentially important areas of genetic interest.
The authors wish to thank all of the primary care practices and participants in the EPIFUND study, the EPIFUND study team and Arthritis Research UK lab staff for assistance with DNA extraction. The authors thank the men who participated in the seven countries and the research/nursing staff in the seven centres of the EMAS study used in the current analysis: C Pott (Manchester), E Wouters (Leuven), M del Mar Fernandez (Santiago de Compostela), M Jedrzejowska (Lodz), H-M Tabo (Tartu) and A Heredi (Szeged) for their data collection, and C Moseley (Manchester) for data entry and project coordination. DV and SB are senior clinical investigators of the Fund for Scientific Research-Flanders, Belgium (F W O-Vlaanderen). SB is holder of the Leuven University Chair in Gerontology and Geriatrics.
The first two authors contributed equally to this work;
↵* nee Limer.
Members of the EMAS Group Gianni Forti, Luisa Petrone, Antonio Cilotti (Florence); Dirk Vanderschueren, Steven Boonen, Herman Borghs (Leuven); Krzysztof Kula, Jolanta Slowikowska-Hilczer, Renata Walczak-Jedrzejowska (Lodz); Ilpo Huhtaniemi (London); Aleksander Giwercman (Malmö); Frederick Wu, Alan Silman, Terence O'Neill, Joseph Finn, Philip Steer, Abdelouahid Tajar, David Lee, Stephen Pye (Manchester); Felipe Casanueva, Marta Ocampo, Mary Lage (Santiago); György Bartfai, Imre Földesi, Imre Fejes (Szeged); Margus Punab, Paul Korrovitz (Tartu); and Min Jiang (Turku). For additional information regarding EMAS, contact Frederick Wu, MD, Department of Endocrinology, Manchester Royal Infirmary, Manchester, UK.
Funding This work was supported by Arthritis Research UK, Chesterfield, UK, grant no 17552. The European Male Ageing Study (EMAS) is funded by the Commission of the European Communities Fifth Framework Programme ‘Quality of life and management of living resources’ grant QLK6-CT-2001-00258 and is supported by funding from Arthritis Research UK.
Competing interests None.
Ethics approval EMAS ethical approval for the study was obtained in each of the centres in accordance with local practice and requirements. For EPIFUND ethical approval was obtained from Manchester LREC.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.