Background

Despite several attempts to understand the differences in the prognosis of recovery after a whiplash trauma, the biological mechanisms remain unexplained. The main symptom in patients with a whiplash-associated disorder (WAD) is long-lasting pain [1]. Initial pain sensations after the trauma seem to be the best prognostic factors for the outcome in patients suffering from WAD [2,3,4,5].

It has been suggested that genetic predisposition might play a role in the pathogenesis of WAD. In the field of pain genetics, the catechol-O-methyltransferase (COMT) gene is one of the most well-studied ones [6]. It has been suggested that polymorphisms within the gene regulate pain perception in healthy subjects [7], as well as in patients with fibromyalgia [8, 9], lower back pain [9, 10] and also acute pain after motor vehicle accidents [11]. Tammimäki and Männistö’s review of the literature indicated that diverging results from published studies call for further investigations, especially in the field of musculoskeletal pain [12].

The COMT gene regulates formation of both soluble and membrane-bound protein isoforms. It contains many single nucleotide polymorphisms (SNPs), some of which have been investigated concerning their role in pain regulation. The COMT Val158Met SNP, rs4680, is the most frequently investigated variant [9]. The Met/Met genotype of the SNP rs4680 has been associated with several mental diseases [13], as well as fibromyalgia and widespread pain [14]. In addition, the importance of haplotypes has been demonstrated, especially regarding the haplotype that includes rs6269, coding for the soluble form, and rs4633, rs4680 and rs4818, coding for both the soluble and membrane-bound forms. Three common haplotypes based on linkage disequilibrium and pain-related data have been identified [15]. The low pain sensitivity (LPS) haplotype codes for the highest levels of the COMT enzyme activity and has been suggested to mediate for high tolerance of musculoskeletal pain. The average pain sensitivity (APS) haplotype codes for lower enzyme activity and average pain tolerance. The high pain sensitivity (HPS) haplotype is known to code for the lowest levels of the enzyme and the lowest tolerance of pain [15, 16].

While the COMT gene contains functional polymorphisms that have been found to influence human pain [17,18,19], its role in WAD is not known. Because of the relationship between initial pain sensation and recovery in WAD patients [20], it is plausible to hypothesize that COMT variations could influence pain regulation and clinical outcome in this patient population.

The primary aim of this study was to evaluate whether or not self-reported non-recovery, including pain, was related to COMT genotype in patients with WAD. The secondary aim was to investigate whether or not background factors, including mental health, were related to haplotypes and to non-recovery.

Methods

Participants

Patients were enrolled between September 2002 and January 2004. Potentially eligible patients had sustained a whiplash injury less than 24 h before arrival at the emergency department (ED). Exclusion criteria were WAD IV (fractures), age < 15 years, inability to read and understand the Swedish language and impaired cognitive function making the self-report impossible. The medical treatment was administered according to the routine protocols at the hospital. The patients were informed about the study and those who agreed to participate were invited to an out-patient visit for inclusion. The patients were originally part of a randomized controlled trial in which standard treatment was compared with an intervention [21].

A blood sample for the genetic analysis was taken and the participants were asked to fill in questionnaires regarding socio-demographic information, including age (<24, 25–65, >66), gender, educational level (high school or less or university), working status, marital status and ethnic background (Caucasian, Caucasian-Hispanic, Asian, Middle Eastern or African). They were asked to rate their health-related quality of life (HRQoL) the week before the accident (i.e., a retrospective rating) according to the Short Form-36 (SF-36) Health Survey [22]. Visual analogue scales (VAS) were used for ratings of pain and mental status (anxiety and depression) the week before the whiplash injury occurred (i.e., a retrospective rating) and at the time of inclusion, with 0 indicating minimum and 100 maximum symptoms. The patients also filled in the Hospital Anxiety and Depression (HAD) [23] and the Posttraumatic Stress Disorder (PTSD-10) questionnaires [24]. A score of >10 indicated a state of anxiety or depression on the HAD score [25] and a score of >5 indicated PTSD according to PTSD-10 [26]. These self- reported data did not, however, fulfil the criteria for the medical conditions per se.

Genetic analysis

Blood was used to isolate DNA using the QIAamp DNA MiniKit, according to the manufacturer’s protocol (http://www.qiagen.com/). The selected polymorphisms in the COMT gene were: rs6269, rs4633, rs4818 and rs4680. The COMT rs4680 (Val158Met) polymorphism was genotyped using the TaqMan® Drug Metabolism Genotyping Assay, according to the manufacturer’s protocol [27]. This assay consists of sequence-specific forward and reverse primers and two TaqMan® MGB probes, each labelled with a FAM™ or VIC® reporter dye at the 5’end and a non-fluorescent quencher at the 3’end of each one. Polymerase chain reaction (PCR) was performed in a 5-μl reaction mixture containing 1 × TaqMan®Universal PCR Master Mix 2.5 μl; 40× TaqMan®Drug Metabolism Genotyping Assay Mix 0.25 μl and 3–20 ng Genomic DNA diluted in H2O (Applied Biosystem®). An ABI PRISM®7900HT Sequence Detection System was used to perform allele discrimination PCR reactions according to the following thermal cycler conditions: initial step of 10 min at 95 °C, followed by 50 cycles of denaturation at 15 s, 92 °C and annealing at 90 s, 60 °C. The PCR products were analysed and allele sizes were determined using SDS 2.2 (Applied Biosystem®). On the other hand, a fluorescence-based competitive allele-specific PCR (KASPar) assay (KBioscience®), based on a public genome sequence (www.ensembl.org/), was used to analyse the SNPs rs6269, rs4633 and rs4818. Alleles were determined using SNPviewer2®. χ2 test or Fisher’s test was used to investigate deviations from Hardy-Weinberg Equilibrium (HWE). The EM algorithm was computed using SNP & Variation Suite 7 (Golden Helix) to estimate linkage disequilibrium (r2) and haplotype blocks. Three common haplotypes were found, as in Diatchenko et al. [15]. LPS was defined as homozygous for the G_C_G_G haplotype (GCGG/GCGG). APS was defined as homozygous for the A_T_C_A haplotype (ATCA/ATCA) and heterozygous for ACCG/GCGG or GCCG/ATCA. HPS was defined as heterozygous for ATCA or ACCG.

An additional categorization of genotypes was determined according to the presence of at least one LPS haplotype [11].

Outcome measures

The primary outcome measure was self-reported recovery at 12 months, assessed by means of the single question: ‘Do you feel recovered after the injury?’ (Yes/No). Secondary outcome measures were the SF-36 responses (physical and mental score) and pain and mental status measured by visual analogue scales (VAS pain and VAS mental).

Statistical methods

Nominal variables were tested with chi-square test. The Kruskal-Wallis and Anova were used for comparisons measured on interval and ratio scale variables. The results were regarded as significant if p was less than 0.05, two-tailed. The p-values were presented without adjustment for multiple comparisons. Logistic regression was used to identify variables associated with being risk factors for non-recovery at 12 months. First, crude associations for each selected variable were studied in univariable models. Secondly, a multivariable model was used to adjust the associations. The results were regarded as significant at p < 0.05. The associations are presented as odds ratios (ORs) with 95% confidence intervals [28].

The statistical analysis was performed using SPSS version 20.0 (SPSS, Inc., Chicago, Il. USA).

Ethics

The study was approved by the Regional Research Ethics Committee Stockholm, Sweden and all patients gave their written informed consent before inclusion (Dnr-240/01).

Results

A total of 133 patients were enrolled in our study, but the outcome analysis was based on 128 patients because five lacked 12-month follow-up data. The distribution of genotypes and haplotypes are presented in Table 1. Most of the patients had the APS haplotype (102 out of 133).

Table 1 Distribution of genotypes and haplotypes for the study population (n = 133)
Full size table

As shown in Tables 2, 56% of the patients were females; the mean age was 40 years and most of the patients were of Caucasian origin, had a secondary school education, were working and were non-smokers. There were no differences regarding these background factors and the distribution of LPS, APS and HPS haplotypes (Table 2). Mental health data at inclusion showed that 20.2% of the patients had self-reported symptoms of anxiety and 8.7% had signs of PTSD before the whiplash injury, but there were no significant differences between the different haplotypes and these factors as also shown in Table 2.

Table 2 Background data at baseline (n = 133). There were no significant differences in any of the comparisons between the variables
Full size table

No significant differences could be detected either between the distribution of haplotypes and the VAS and SF-36 results before (retrospective rating) or after the accident or at the 12-month follow-up (Fig. 1), (Table 3).

Fig. 1
figure 1

Association between haplotypes reported before the accident, after the accident and at 12- month follow-up and mean levels of pain

Full size image
Table 3 Pain and mental status ratings on Visual Analogue Scale (VAS) and HRQoL according to SF-36 in relation to haplotype. No significant differences between the haplotype groups
Full size table

Fifty-six patients (43.8%) reported non-recovery at 12 months, with no difference in the recovery rate between the different COMT haplotypes (LPS 42.1%, APS 43.9% and HPS 45.5%, respectively).

Regression analyses with adjustments for confounding factors (Table 4) showed no significant differences regarding the COMT haplotypes in relation to recovery, but high levels of initial pain according to VAS (adjusted OR, 10.4 (3.0–36.6) and anxiety according to HAD ratings (adjusted OR, 3.5 (1.1–11.6) were associated with non-recovery at 12 months.

Table 4 Association between the predictor variables and self-reported non-recovery at 12 months after the injury. All predictor variables were measured at baseline. Odds ratios and 95% confidence intervals (CIs)
Full size table

Discussion

This study, investigating the association between COMT haplotypes and recovery after whiplash injuries, could not detect any differences in the recovery rate between the three COMT haplotypes. Nor were any significant differences detected between COMT haplotypes regarding the secondary outcome measures, i.e., level of pain and quality of life according to the SF-36 at 12 months. However, the multivariate analysis showed that high levels of pain and anxiety after the whiplash injury were significantly associated with non-recovery.

In a meta-analysis by Tammimäki et al. of pain and psychological symptoms after motor vehicle collisions, an association was found between different COMT haplotypes and musculosceletal pain [12]. Furthermore, they noted that several other factors, such as pain stimulus, chronicity of pain, sex, sampling and sample size, were identified as contributors to differences in their results. McLean and co-workers [11] reported on pain and psychological symptoms after motor vehicle collisions and found a connection with the pain sensitive COMT genotype. In our study, pain and anxiety affected the outcome, but were not connected to the COMT haplotypes. It is possible that the finding that 20% of the respondents reported symptoms of anxiety and 9% had signs of PTSD prior to the accident affected the results. It has previously been shown that anxiety and PTSD symptoms can moderate pain sensations [29, 30].

Another explanation for the findings in this study, as compared to previous research, could be the differences in ethnicity between the studies. Mclean et al. reported differences in pain perception in a European-American population [11], Xiang et al. could not detect any differences in pain perception in a Chinese population [31], while, in the present study, almost 90% of the participants consists of Caucasians. One limitation of this study is the lack of a control group. The distribution of COMT haplotypes varies between different settings and geographic locations [8, 32] and may affect the outcome. Furthermore, gender differences are also relevant in this field of research [33]. Indeed, several studies point towards a complex COMT-by-gender interaction effect [34, 35]. However, in our study, no difference could be detected between the sexes, probably due to a small sample size.

Nor could our study demonstrate any differences in pain ratings between patients with the most common COMT genotypes. This finding is in contrast to the findings by Bortsov and co-workers, who found differences between genotypes 6 weeks after a motor vehicle collision [36]. One reason might be the different methods of COMT haplotyping and subgrouping compared to that cohort. Grouping into three haplotypes (i.e., LPS; APS; HPS) was done as suggested by Diatchenko and colleagues [15]. Another reason for the contradictory findings might be differences in self-reporting methodology (verbal vs. VAS) and the timing for rating pain. In our study, the pain ratings were done within two weeks after the accident, a time point at which the level of pain still might not have been stabilized and the frequency of high pain ratings were high and, consequently, could lead to a poorer outcome [4, 37].

Pain recovery comprises a complex interplay between several systems, from sensory to immune to brain circuits [38]. Research on genetic susceptibility to pain has identified several COMT SNPs as possible candidates; however, more studies are needed to define the possible causality of this relationship.

Conclusion

In summary, no relationship between the outcome regarding reported non-recovery or pain levels and COMT haplotypes in patients with acute whiplash injuries could be detected in this material. Nor did we find associations between psychological factors and the COMT haplotypes. High levels of initial pain and anxiety were associated with non-recovery, thereby confirming previously published literature.