Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS)


The etiology and pathogenesis of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) are unknown, and autoimmunity is one of many proposed underlying mechanisms. Human Leukocyte Antigen (HLA) associations are hallmarks of autoimmune disease, and have not been thoroughly investigated in a large ME/CFS patient cohort. We performed high resolution HLA -A, -B, -C, -DRB1, -DQB1 and -DPB1 genotyping by next generation sequencing in 426 adult, Norwegian ME/CFS patients, diagnosed according to the Canadian Consensus Criteria. HLA associations were assessed by comparing to 4511 healthy and ethnically matched controls. Clinical information was collected through questionnaires completed by patients or relatives. We discovered two independent HLA associations, tagged by the alleles HLA-C*07:04 (OR 2.1 [95% CI 1.4–3.1]) and HLA-DQB1*03:03 (OR 1.5 [95% CI 1.1–2.0]). These alleles were carried by 7.7% and 12.7% of ME/CFS patients, respectively. The proportion of individuals carrying one or both of these alleles was 19.2% in the patient group and 12.2% in the control group (OR 1.7 [95% CI 1.3–2.2], pnc = 0.00003). ME/CFS is a complex disease, potentially with a substantial heterogeneity. We report novel HLA associations pointing toward the involvement of the immune system in ME/CFS pathogenesis.


Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a disabling disorder characterized by medically unexplained fatigue, post-exertional malaise and a variety of additional symptoms, such as chronic pain, sleep disturbances and cognitive difficulties. ME/CFS is diagnosed on clinical grounds alone, and different sets of criteria specify the mandatory symptoms as well as recommendations for the exclusion of differential diagnoses1,2,3. The specificity and validity of different diagnostic criteria have been questioned, yet there is no agreement on the level of heterogeneity in ME/CFS, and there is no consensus on how to categorize different subgroups4,5,6,7,8.

The pathogenesis and etiology of ME/CFS are unknown, with several models having been proposed9. One central hypothesis states that autoimmunity is part of the pathophysiology10,11. ME/CFS has been reported to be partly heritable12,13, consistent with a multifactorial etiology dependent on both genetic and environmental factors. This is the prevailing model for a vast number of diseases, including established autoimmune diseases (AID). Several publications report immunological alterations among ME/CFS patients, e.g. changes in natural killer (NK) cell function14,15, cytokine levels16,17, and DNA methylation patterns consistent with immune dysregulation18. Some of these findings have failed to reproduce in other studies, which could be due to differences in methodology, the complexity and heterogeneity of ME/CFS, and lack of power due to small sample sizes19. Resultingly, the autoimmunity hypothesis warrants further evaluation. A characteristic feature of AID is genetic association with certain human leukocyte antigen (HLA) alleles20. Thus, a thorough investigation of HLA associations in ME/CFS is relevant, although HLA associations per se cannot be used as evidence regarding disease etiology21. Studies of HLA associations in CFS have been published, but with great variation in patient inclusion criteria and HLA typing methodology22,23,24,25,26,27,28,29,30. No reproducible, significant associations are evident across these studies. In the largest study, including 110 patients, the strongest significant association was with HLA-DQ3 with an odds ratio (OR) of 1.8 (95% CI 1.2–2.8)22. Associations with HLA alleles DQA1*01, DRB1*13:01 and DQB1*06:02 have also been reported25,27,29. The great majority of these studies include less than 50 patients, and are underpowered for the detection of moderate to weak associations. Hence, in this study we aimed to conduct a comprehensive investigation of HLA associations in a large ME/CFS cohort, applying modern, high resolution HLA typing.


Characterization of patient and control groups

We included 426 ME/CFS patients and 4511 healthy, ethnically matched controls. All patients had been diagnosed in Norway according to the 2003 Canadian Consensus Criteria2, except for four patients where the similarly strict 2010 International Consensus Criteria3 were applied. Demographic and clinical characteristics of patients and controls are shown in Table 1. The mean age at diagnosis for ME/CFS patients was 34.7 years, 82.8% were female, and most patients (45.5%) had a disease duration of 5–10 years, from symptom debut to inclusion. 12.5% of the patients had severe or very severe disease (bedridden). An additional 28.6% had moderate to severe disease (strictly housebound). A total of 41.1% of ME/CFS patients were bed- or housebound, and 86.8% of patients were unable to work full or part time the previous 6 months.

Table 1 Demographic and clinical characteristics of ME/CFS patients and healthy controls.

HLA alleles associated with ME/CFS

In all patients and controls, we obtained 2nd field resolution genotypes of HLA class I genes HLA -A, -B and -C and class II genes HLA -DRB1, -DQB1 and -DPB1. This resolution distinguishes HLA alleles that encode specific HLA proteins. No significant deviations from Hardy-Weinberg equilibrium were noted at any HLA loci, neither in the patient group nor in the control group (Supplementary Table S1). Allele frequencies for all observed HLA Class I and Class II alleles are presented in Supplementary Table S2. Global association tests for each HLA locus (Supplementary Table S3) were significant for HLA-C (p = 0.04) and HLA-DQB1 (p = 0.04). When comparing individual allele frequencies between patients and controls, four HLA risk alleles emerged (Table 2): C*07:04 (OR = 2.1 [95% CI 1.4–3.1], pnc = 0.0001, pc = 0.001), B*57:01 (OR = 1.6 [95% CI 1.2–2.3], pnc = 0.004, pc < 0.05), DQB1*03:03 (OR = 1.5 [95% CI 1.1–2.0], pnc = 0.005, pc < 0.05) and B*44:02 (OR = 1.3 [95% CI 1.0–1.6], pnc = 0.03, pc = n.s.). In order to evaluate the dependency of these associations, we measured the degree of linkage disequilibrium (LD) between the four alleles within the patient group (Fig. 1). Strong LD was observed between C*07:04 and B*44:02 (D’ = 0.90) as well as between B*57:01 and DQB1*03:03 (D’ = 0.69), indicating that these alleles may occur on two distinct haplotypes. The first haplotype, C*07:04 - B*44:02, had an estimated frequency of 3.5% in the patient group and 1.7% in the control group, resulting in an OR of 2.1 (95% CI 1.4–3.1, pnc = 0.0002). The second haplotype, B*57:01 - DQB1*03:03, had an estimated frequency of 3.3% in the patient group and 2.0% in the control group, resulting in an OR of 1.7 (95% CI 1.1–2.5 pnc = 0.01). To further evaluate which allele on each of the two haplotypes represents the most significant association, we performed Svejgaard analyses between C*07:04 and B*44:02 as well as between B*57:01 and DQB1*03:03 (Supplementary Data Sheet S4). None of the two alleles on either of the two haplotypes reached significance when testing their independent association, which is not surprising due to the strong LD mentioned above. We report C*07:04 and DQB1*03:03 as tag alleles for the ME/CFS associations, since these alleles occur at the two loci initially showing global association.

Table 2 HLA alleles showing association (pnc < 0.05) with ME/CFS in 426 patients and 4511 healthy controls.
Figure 1
figure 1

Illustration of Linkage disequilibrium (LD) calculated as D’ values in the patient group between the four alleles with a significant association to ME/CFS. D’ measures the strength of LD between two alleles, and ranges from −1 to 0 for negative LD, and from 0 to 1 for positive LD. The number in each square is the D’ value between the two alleles listed diagonally above the square. Blue colors indicate the strength of LD, with darker colors for stronger LD.

We next wanted to make sure that these HLA associations were not due to gender differences between the cases and controls (82.8% vs 59.8% females, respectively). No significant gender differences were observed between the carrier frequencies of these alleles in either cases (Table 3) or controls (C*07:04 had a carrier frequency of 3.7% in females and 4.0% in males; p = 0.6; DQB1*03:03 had a carrier frequency of 9.0% in females and 8.3% in males; p = 0.4). Furthermore, after stratifying cases and controls according to gender, heterogeneity was rejected (p > 0.5) between the OR values obtained for females only and males only, indicating no gender differences between the HLA associations observed in ME/CFS.

Table 3 Clinical characteristics of 426 ME/CFS patients: in total and stratified according to the presence of either DQB1*03:03 or C*07:04 or both.

The two loci HLA-DRB1 and HLA-DQB1 are physically close, and exhibit particularly strong LD. In our dataset, DQB1*03:03 occurred most frequently with DRB1*07:01. The haplotype DRB1*07:01 - DQB1*03:03 had estimated frequencies of 4.7% and 2.9% in the patient and control group, respectively (OR 1.7 [95% CI 1.2–2.3], pnc = 0.003). Among patients, this haplotype was without exception estimated to carry DQA1*02:01. Genotype data for the HLA-DQA1 locus was only obtained for patients, and was not available for the controls.

There were two alleles with a negative association with ME/CFS, suggesting a potential protection, namely B*08:01 (OR = 0.7 [95% CI 0.6–0.9], pnc = 0.01, pc = n.s.) and DPB1*02:01 (OR = 0.7 [95% CI 0.6–0.9], pnc = 0.02, pc = n.s.) (Table 2). These alleles were not in LD (D’ = −0.29), indicating that the associations are independent. The most frequent B*08:01 haplotype in Norway is the highly conserved so-called autoimmune and ancestral AH8.1 haplotype31 (C*07:01-B*08:01-DRB1*03:01-DQB1*02:01). This haplotype had reduced estimated frequency in the patient group compared to the control group (8.2% vs. 10.3%, OR = 0.8, pnc = 0.06), albeit not significantly.

HLA risk allele carriers and clinical characteristics

The proportion of individuals carrying the allele C*07:04 was 7.7% in the patient group and 3.8% in the control group, while 12.7% of the patients and 8.7% of the controls carried DQB1*03:03 (Supplementary Data Sheet S4). The proportion of individuals carrying one or both of the two alleles was 19.2% in the patient group and 12.2% in the control group (OR 1.7, pnc = 0.00003, 95% CI[1.3–2.2]). Table 3 shows the distribution of clinical characteristics in the patient group, including stratification for C*07:04 and/or DQB1*03:03. Neither gender, initiating events, comorbidity of depression or fibromyalgia, nor AID or ME/CFS among 1st degree relatives were associated with the risk alleles. However, ME/CFS patients carrying one or both of the risk alleles had a significantly higher proportion of comorbid AID (OR = 2.3 [95% CI 1.2–4.3], pnc = 0.01). The frequency of comorbid AID was significantly increased also when stratifying for C*07:04 alone (OR = 2.9 [95% CI 1.2–6.6], pnc = 0.01), but not when stratifying for DQB1*03:03 alone (OR = 1.6 [95% CI 0.8–3.4], pnc = n.s.). These patients, carrying HLA risk alleles, had the following AID, ordered by frequency: Hashimoto’s thyreoiditis/hypothyreosis, psoriasis, rheumatoid arthritis, alopecia areata and Crohn’s disease or ulcerative colitis.


In this project, we performed high resolution HLA genotyping by next generation sequencing (NGS) in 426 adult, Norwegian ME/CFS patients, diagnosed according to the Canadian Consensus Criteria2. There are no previous publications with comprehensive HLA genotyping by NGS in this patient group. We discovered two independent HLA associations, tagged by the alleles HLA-C*07:04 and HLA-DQB1*03:03.

To our knowledge, associations with HLA-C alleles have not previously been studied in ME/CFS. In 1994, Keller et al. performed serologic HLA-DR and DQ typing in 110 patients with Chronic fatigue immune dysfunction syndrome (CFIDS)22. The patients were diagnosed with the Holmes Criteria32, and CFIDS was defined as a subgroup with positive findings in viral reactivation patterns and B- and T-cell tests, indicating post-infectious debut and a certain degree of immune dysfunction. The authors found a significant association (OR = 1.8) with the serotype HLA-DQ3. Serologic HLA typing is of low resolution compared to genetic typing33. HLA-DQ3 corresponds to HLA-DQB1*03 in genetic nomenclature, where DQB1*03:03 is one of the three largest subgroups. Higher resolution HLA-DQB1 typing have been performed in two smaller cohorts (<58 patients), and even though statistically not significant, DQB1*03:03 was observed slightly more frequent among CFS patients, diagnosed with the Fukuda criteria, than among controls23,25. Hence, the findings in existing literature is compatible with the association between ME/CFS and DQB1*03:03 in our material.

HLA-B*08:01 showed reduced frequency in ME/CFS compared to controls in our material. This allele most often occur on the haplotype C*07:01-B*08:01-DRB1*03:01-DQB1*02:01, which was also less prevalent among ME/CFS patients in our material. This ancestral haplotype, AH8.1, is a risk factor for a wide variety of AID, including myasthenia gravis, systemic lupus erythematosus and coeliac disease31, but protective against rheumatoid arthritis34,35. In the existing literature on HLA and CFS, HLA-DRB1 is the locus most frequently studied. In four out of five studies, the frequency of DR3/DRB1*03 was lower in the patient group23,24,25,28, while in the fifth study the frequency was similar in both groups22. Hence, this haplotype seems truly less prevalent among ME/CFS patients.

Some HLA associations previously reported in CFS are not supported by our results27,29. The often cited association with DQA1*01 reported by Smith et al.25 cannot be evaluated in our material since the HLA-DQA1 locus was not genotyped in the control group. In our patient group, DQA1*01 occurred on haplotype with the following DQB1 alleles, ordered by frequency: DQB1*06:02, DQB1*05:01, DQB1*06:03, DQB1*06:04 and DQB1*05:03, and neither of these (pnc > 0.1, Supplementary Table S2), nor all combined (OR = 1.0, pnc = 0.6) were associated with ME/CFS.

The present HLA study in ME/CFS is to our knowledge the largest performed to date (other studies comprise ≤ 110 CFS patients). Our study had 80% power to discover HLA-associations with OR ≥ 1.5 given an allele frequency > 0.05. Interestingly, both C*07:04 and DQB1*03:03 remained significant after Bonferroni multiple test correction. Notably, we performed locus-wise multiple test correction, i.e. correcting for the number of alleles tested at each locus, since alleles at different HLA loci are in strong LD, and therefore do not represent independent tests. The Bonferroni method is considered a strict multiple test correction36, but on the other hand locus-wise correction does not take into account the lack of complete LD between the investigated loci. Taken together, our results need verification in independent cohorts. In general, established HLA associations are reproducible across different populations, but susceptibility loci can also vary between populations20,21,37,38. Therefore, HLA associations in ME/CFS should also be investigated in populations of different ancestry.

Both of the ME/CFS associations observed in our data set were evident at 2nd field resolution (i.e. C*07:04 and DQB1*03:03), which distinguishes alleles encoding amino acid differences. Interestingly, the other C*07 and DQB1*03 alleles were not associated, emphasizing the importance of high resolution HLA genotyping.

We report C*07:04 and DQB1*03:03 as tag alleles for two independent HLA risk associations in ME/CFS, as these alleles are in linkage equilibrium (D’ = 0.06). However, they could still be markers for either one common, or two independently associated, variants outside the loci tested in this study. Alternatively, the associated alleles reported herein could constitute a functional relevance themselves. HLA class I alleles, like C*07:04, could influence disease risk through their interactions with CD8 positive cytotoxic T lymphocytes20,39. Disturbances in CD8 positive T lymphocytes have been reported in ME/CFS, although the results are somewhat conflicting10. Another important function of HLA class I alleles is to serve as ligands for NK cell receptors. Altered numbers and function of NK cells have been reported by several independent researchers in ME/CFS14,15,40. The other associated allele, DQB1*03:03, is an HLA class II allele, which is also interesting in regard to the hypothesis of autoimmunity in ME/CFS. In certain well studied AID, associated HLA class II alleles have been shown to exhibit unique peptide binding properties, as well as HLA-TCR restriction, directly influencing the acquired immune response, e.g. with the production of specific auto-antibodies41. A dysregulated activity of CD4 positive T lymphocytes, the principal cell type interfering with HLA class II alleles, have been discussed as a central mechanism in ME/CFS42. Several studies report increased levels of specific auto-antibodies in ME/CFS patients43, e.g. to neurotransmitter receptors, although most of these lack verification in additional cohorts.

Another interesting question is whether these HLA associations are driven by subgroups of patients, and thereby representing stronger risk alleles. This is a relevant aspect in a complex disease like ME/CFS, where different causal mechanisms may be at play in different subgroups. In our study, ME/CFS patients carrying HLA risk alleles had a significantly higher comorbidity of established AID. We are not aware of any publications reporting associations between C*07:04 or DQB1*03:03 and the AID affecting some of our patients. Therefore, it is unlikely that the associations in our material are driven by HLA associations with already established AID. Familial aggregation is observed for many specific AID, as well as for autoimmunity in general44. 1st degree relatives of ME/CFS patients in our study have a high prevalence of AID (Table 3). These observations could potentially be due to an element of autoimmunity in ME/CFS, or within a subgroup. No other patient characteristics were dominant among the ME/CFS patients with HLA risk alleles, specifically neither self-reported infectious onset nor current disease severity. It can be argued that the lack of validity of self-reported data precludes the detection of possible subgroup identifiers.

In conclusion, we report novel HLA associations in a large cohort of ME/CFS patients fulfilling the Canadian Consensus Criteria, thereby supporting the involvement of the immune system in the ME/CFS pathogenesis.

Materials and Methods

This study is approved by the Norwegian Regional Committees for Medical and Health Research Ethics45. All methods and data handling were performed according to relevant national and institutional regulations and guidelines. All patients gave informed consent. In three cases, written consent was given by a close relative due to the patient being severely ill and unable to sign. A total of 426 adult, Norwegian ME/CFS patients were included. All had been diagnosed in Norway according to the 2003 Canadian Consensus Criteria2, except for four patients where the similarly strict 2010 International Consensus Criteria3 were applied. There were three separate recruitment groups for ME/CFS patients: 214 patients were recruited from recent and ongoing trials with Rituximab46,47 and Cyclophosphamide (Rekeland IG et al., submitted, NCT02444091); 116 patients were recruited from the CFS/ME biobank at Oslo University Hospital; 96 patients were recruited via announcements in patient networks, including patient organizations. Patients from the latter two groups were not included in clinical trials. Duplicates within or between the three recruitment groups were excluded. All patients provided the identity of any 1st, 2nd or 3rd degree relatives with ME/CFS, and we excluded close relatives to ensure that only one patient per extended family was included. Norwegian ethnicity was ensured by evaluation of sur- and family names of all patients, country of birth of parents and grandparents as well as self-perceived ethnicity. The control group consisted of 4511 ethnically matched, healthy individuals drawn from the Norwegian Bone Marrow Donor Registry48. Clinical information was collected for the ME/CFS patients through questionnaires completed by patients or close relatives. The categories applied in this study were gender, age at diagnosis, initiating events, disease duration and severity, comorbidities and family history. Most of the questions were based on the DePaul Symptom Questionnaire49,50. Infection or vaccination as initiating event was self-reported, and in many cases, the time from event to symptom debut was not specified. The disease severity was assessed with self-reported activity level during the previous 6 months, as stated through the DePaul Symptom Questionnaire, question no. 79.

HLA genotyping by next generation sequencing

In 426 ME/CFS patients, we performed high resolution, targeted, next generation sequencing (NGS) of HLA class I genes HLA -A, -B and -C and class II genes HLA -DRB1, -DQB1, -DPB1, -DQA1. Amplification and library preparation were performed with kits from GenDx (Utrecht, The Netherlands) and Illumina (San Diego, USA), 2 ×150 bp paired-end sequencing was performed by The Norwegian Sequencing Centre with Illumina MiSeq Reagent Kit v2 (300-cycles), and HLA genotypes were obtained by analyzing sequencing reads with NGSengine from GenDx, using the IMGT/HLA Database51. The median sequencing depth was above 150 reads per called base. The 4511 healthy Norwegian controls had previously been HLA typed by NGS52. Both patient and control genotypes were analyzed at 2nd field resolution for HLA -A, -B, -C, -DRB1, -DQB1 and -DPB1. HLA alleles can be genotyped at resolution level from 1st field to 4th field. 2nd field resolution distinguishes alleles that encode amino acid differences, i.e. specific HLA proteins, and is therefore of great biological relevance. The genotyping success exceeded 99% in the patient group and 99.9% in the control group for all loci. In the control group, alleles were originally identified at a G group resolution, and certain alleles from the patient group were therefore converted to avoid typing method bias (Supplementary Table S5).

Data analyses were performed in Unphased 3.0.10 and Pypop 0.7.053,54. Assessment of Hardy-Weinberg equilibrium was performed with a chi-square test with a significance level of 0.05. Haplotype frequencies were estimated with an expectation-maximization method for unknown gametic phase. Global associations for each locus were calculated with a likelihood ratio test, with a rare allele frequency threshold of 0.01. Genetic associations were investigated on allelic and haplotypic levels, and ORs with 95% confidence intervals (95% CI) were calculated with Woolf’s formula comprising Haldane’s correction. Risk allele ORs were calculated also with gender stratification, and homogeneity tests were performed with the logit-based estimator. LD calculations and Svejgaard tests were carried out to examine the degree of independence between the associated alleles55. The LD measure D’ was calculated according to the formula D’ = D/Dmax, where Dmax = min [pA(1-pB), (1-pA)pB] for D > 0, Dmax = min [pApB, (1-pA)(1-pB)] for D < 0, D is the standard mathematical definition of LD between alleles A and B56, and p is the frequency of the stated allele. Because of the comparison of multiple allele frequencies we performed locus-wise Bonferroni correction. For each locus, non-corrected p-values were multiplied by the total number of alleles detected at that locus, excluding alleles with a frequency less than 3% in both the control group and the patient group. The significance level was 0.05. Only haplotypes consisting of associated alleles were investigated, and multiple test correction was therefore not applied on the haplotypic level.

Investigation of clinical data

The clinical information was gathered separately for each of the three recruitment groups, controlled in one common database, and exported to SPSS57 for statistical analyses. The patient group was stratified according to the presence of specific HLA alleles, and eight dichotomized clinical variables were assessed with OR calculation by binary logistic regression, and chi-square significance testing.

Restrictions on the availability of material

Individual genotypes of patients are not made available due to Norwegian privacy regulations and laws.


  1. Fukuda, K. et al. The chronic fatigue syndrome: a comprehensive approach to its definition and study. International Chronic Fatigue Syndrome Study Group. Annals of internal medicine 121, 953–959 (1994).

    CAS  Article  PubMed  Google Scholar 

  2. Carruthers, B. M. et al. Myalgic Encephalomyelitis/Chronic Fatigue Syndrome. Journal of Chronic Fatigue Syndrome 11, 7–115, (2003).

    Article  Google Scholar 

  3. Carruthers, B. M. et al. Myalgic encephalomyelitis: International Consensus Criteria. Journal of internal medicine 270, 327–338, (2011).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  4. de Vega, W. C., Erdman, L., Vernon, S. D., Goldenberg, A. & McGowan, P. O. Integration of DNA methylation & health scores identifies subtypes in myalgic encephalomyelitis/chronic fatigue syndrome. Epigenomics 10, 539–557, (2018).

    CAS  Article  PubMed  Google Scholar 

  5. Williams, T. E., Chalder, T., Sharpe, M. & White, P. D. Heterogeneity in chronic fatigue syndrome - empirically defined subgroups from the PACE trial. Psychological medicine 47, 1454–1465, (2017).

    CAS  Article  PubMed  Google Scholar 

  6. Sunnquist, M., Jason, L. A., Nehrke, P. & Goudsmit, E. M. A Comparison of Case Definitions for Myalgic Encephalomyelitis and Chronic Fatigue Syndrome. Journal of chronic diseases and management 2 (2017).

  7. Maclachlan, L. et al. Are current chronic fatigue syndrome criteria diagnosing different disease phenotypes? PloS one 12, e0186885, (2017).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. Jason, L. A., McManimen, S., Sunnquist, M., Newton, J. L. & Strand, E. B. Clinical Criteria Versus a Possible Research Case Definition in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis. Fatigue: biomedicine, health & behavior 5, 89–102, (2017).

    Article  Google Scholar 

  9. Griffith, J. P. & Zarrouf, F. A. A systematic review of chronic fatigue syndrome: don’t assume it’s depression. Primary care companion to the Journal of clinical psychiatry 10, 120–128 (2008).

    Article  PubMed  PubMed Central  Google Scholar 

  10. Sotzny, F. et al. Myalgic Encephalomyelitis/Chronic Fatigue Syndrome - Evidence for an autoimmune disease. Autoimmunity reviews 17, 601–609, (2018).

    CAS  Article  PubMed  Google Scholar 

  11. Blomberg, J., Gottfries, C. G., Elfaitouri, A., Rizwan, M. & Rosen, A. Infection Elicited Autoimmunity and Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: An Explanatory Model. Frontiers in immunology 9, 229, (2018).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. Albright, F., Light, K., Light, A., Bateman, L. & Cannon-Albright, L. A. Evidence for a heritable predisposition to Chronic Fatigue Syndrome. BMC neurology 11, 62, (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  13. Schur, E., Afari, N., Goldberg, J., Buchwald, D. & Sullivan, P. F. Twin analyses of fatigue. Twin research and human genetics: the official journal of the International Society for Twin Studies 10, 729–733, (2007).

    Article  Google Scholar 

  14. Rivas, J. L., Palencia, T., Fernandez, G. & Garcia, M. Association of T and NK Cell Phenotype With the Diagnosis of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS). Frontiers in immunology 9, 1028, (2018).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  15. Curriu, M. et al. Screening NK-, B- and T-cell phenotype and function in patients suffering from Chronic Fatigue Syndrome. Journal of translational medicine 11, 68, (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  16. Montoya, J. G. et al. Cytokine signature associated with disease severity in chronic fatigue syndrome patients. Proceedings of the National Academy of Sciences of the United States of America, (2017).

  17. Hornig, M. et al. Distinct plasma immune signatures in ME/CFS are present early in the course of illness. Science advances 1, (2015).

  18. Trivedi, M. S. et al. Identification of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome-associated DNA methylation patterns. PloS one 13, e0201066, (2018).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  19. Mensah, F. K. F., Bansal, A. S., Ford, B. & Cambridge, G. Chronic fatigue syndrome and the immune system: Where are we now? Neurophysiologie clinique = Clinical neurophysiology 47, 131–138, (2017).

    Article  PubMed  Google Scholar 

  20. Matzaraki, V., Kumar, V., Wijmenga, C. & Zhernakova, A. The MHC locus and genetic susceptibility to autoimmune and infectious diseases. Genome biology 18, 76, (2017).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  21. Lokki, M. L. & Paakkanen, R. The complexity and diversity of major histocompatibility complex challenge disease association studies. Hla 93, 3–15, (2019).

    CAS  Article  PubMed  Google Scholar 

  22. Keller, R. H. et al. Association between HLA class II antigens and the chronic fatigue immune dysfunction syndrome. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America 18(Suppl 1), S154–156 (1994).

    Article  Google Scholar 

  23. Underhill, J. A., Mahalingam, M., Peakman, M. & Wessely, S. Lack of association between HLA genotype and chronic fatigue syndrome. European journal of immunogenetics: official journal of the British Society for Histocompatibility and Immunogenetics 28, 425–428 (2001).

    CAS  Article  Google Scholar 

  24. Helbig, K. J. et al. Variation in immune response genes and chronic Q fever. Concepts: preliminary test with post-Q fever fatigue syndrome. Genes and immunity 4, 82–85, (2003).

    CAS  Article  PubMed  Google Scholar 

  25. Smith, J. et al. Association of chronic fatigue syndrome with human leucocyte antigen class II alleles. Journal of clinical pathology 58, 860–863, (2005).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  26. Ledina, D. et al. Chronic fatigue syndrome after Q fever. Medical science monitor: international medical journal of experimental and clinical research 13, Cs88–92 (2007).

    Google Scholar 

  27. Carlo-Stella, N. et al. Molecular study of receptor for advanced glycation endproduct gene promoter and identification of specific HLA haplotypes possibly involved in chronic fatigue syndrome. International journal of immunopathology and pharmacology 22, 745–754, (2009).

    CAS  Article  PubMed  Google Scholar 

  28. Ortega-Hernandez, O. D. et al. Autoantibodies, polymorphisms in the serotonin pathway, and human leukocyte antigen class II alleles in chronic fatigue syndrome: are they associated with age at onset and specific symptoms? Annals of the New York Academy of Sciences 1173, 589–599, (2009).

    CAS  ADS  Article  PubMed  Google Scholar 

  29. Spitzer, A. R. & Broadman, M. A retrospective review of the sleep characteristics in patients with chronic fatigue syndrome and fibromyalgia. Pain practice: the official journal of World Institute of Pain 10, 294–300, (2010).

    Article  Google Scholar 

  30. Pasi, A. et al. Excess of activating killer cell immunoglobulinlike receptors and lack of HLA-Bw4 ligands: a twoedged weapon in chronic fatigue syndrome. Molecular medicine reports 4, 535–540, (2011).

    CAS  Article  PubMed  Google Scholar 

  31. Candore, G., Lio, D., Colonna Romano, G. & Caruso, C. Pathogenesis of autoimmune diseases associated with 8.1 ancestral haplotype: effect of multiple gene interactions. Autoimmunity reviews 1, 29–35 (2002).

    CAS  Article  PubMed  Google Scholar 

  32. Holmes, G. P. et al. Chronic fatigue syndrome: a working case definition. Annals of internal medicine 108, 387–389 (1988).

    CAS  Article  PubMed  Google Scholar 

  33. Dunckley, H. HLA typing by SSO and SSP methods. Methods in molecular biology (Clifton, N.J.) 882, 9–25, (2012).

    CAS  Article  Google Scholar 

  34. Sandoughi, M., Fazaeli, A., Bardestani, G. & Hashemi, M. Frequency of HLA-DRB1 alleles in rheumatoid arthritis patients in Zahedan, southeast Iran. Annals of Saudi medicine 31, 171–173, (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  35. Viatte, S. & Barton, A. Genetics of rheumatoid arthritis susceptibility, severity, and treatment response. Seminars in immunopathology 39, 395–408, (2017).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  36. Colquhoun, D. An investigation of the false discovery rate and the misinterpretation of p-values. Royal Society open science 1, 140216, (2014).

    ADS  Article  PubMed  PubMed Central  Google Scholar 

  37. Caillat-Zucman, S. Molecular mechanisms of HLA association with autoimmune diseases. Tissue antigens 73, 1–8, (2009).

    CAS  Article  PubMed  Google Scholar 

  38. Zhou, F. et al. Deep sequencing of the MHC region in the Chinese population contributes to studies of complex disease. Nature genetics 48, 740–746, (2016).

    CAS  Article  PubMed  Google Scholar 

  39. Gough, S. C. & Simmonds, M. J. The HLA Region and Autoimmune Disease: Associations and Mechanisms of Action. Current genomics 8, 453–465, (2007).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  40. Brenu, E. W. et al. Longitudinal investigation of natural killer cells and cytokines in chronic fatigue syndrome/myalgic encephalomyelitis. Journal of translational medicine 10, 88, (2012).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. Dendrou, C. A., Petersen, J., Rossjohn, J. & Fugger, L. HLA variation and disease. Nature reviews. Immunology 18, 325–339, (2018).

    CAS  Article  PubMed  Google Scholar 

  42. Brenu, E. W. et al. Immunological abnormalities as potential biomarkers in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis. Journal of translational medicine 9, 81, (2011).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  43. Ortega-Hernandez, O. D. & Shoenfeld, Y. Infection, vaccination, and autoantibodies in chronic fatigue syndrome, cause or coincidence? Annals of the New York Academy of Sciences 1173, 600–609, (2009).

    CAS  ADS  Article  PubMed  Google Scholar 

  44. Cardenas-Roldan, J., Rojas-Villarraga, A. & Anaya, J. M. How do autoimmune diseases cluster in families? A systematic review and meta-analysis. BMC medicine 11, 73, (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  45. Regional committees for medical and health related ethics, (Folder no: 2015/1547)

  46. Fluge, O. et al. Benefit from B-lymphocyte depletion using the anti-CD20 antibody rituximab in chronic fatigue syndrome. A double-blind and placebo-controlled study. PloS one 6, e26358, (2011).

    CAS  ADS  Article  PubMed  PubMed Central  Google Scholar 

  47. Fluge, O. et al. B-Lymphocyte Depletion in Patients With Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: A Randomized, Double-Blind, Placebo-Controlled Trial. Annals of internal medicine, (2019).

  48. The Norwegian bone marrow donor registry, (2018)

  49. Jason, L. A. et al. The development of a revised Canadian myalgic encephalitis chronic fatigue syndrome case definition. Am. J. Biochem. Biotechnol. 6, 120–135 (2010).

    Article  Google Scholar 

  50. Strand, E. B. et al. Comparing the DePaul Symptom Questionnaire with physician assessments: A preliminary study. Fatigue: Biomedicine, Health & Behavior 4, 52–62 (2016).

    Google Scholar 

  51. IPD-IMGT/HLA Database, (2018)

  52. Lande, A., Andersen, I., Egeland, T., Lie, B. A. & Viken, M. K. HLA -A, -C, -B, -DRB1, -DQB1 and -DPB1 allele and haplotype frequencies in 4514 healthy Norwegians. Human immunology 79, 527–529, (2018).

    CAS  Article  PubMed  Google Scholar 

  53. Dudbridge, F. Pedigree disequilibrium tests for multilocus haplotypes. Genetic epidemiology 25, 115–121, (2003).

    Article  PubMed  Google Scholar 

  54. Lancaster, A. K., Single, R. M., Solberg, O. D., Nelson, M. P. & Thomson, G. PyPop update–a software pipeline for large-scale multilocus population genomics. Tissue antigens 69(Suppl 1), 192–197, (2007).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  55. Svejgaard, A. & Ryder, L. P. HLA and disease associations: detecting the strongest association. Tissue antigens 43, 18–27 (1994).

    CAS  Article  PubMed  Google Scholar 

  56. Slatkin, M. Linkage disequilibrium–understanding the evolutionary past and mapping the medical future. Nature reviews. Genetics 9, 477–485, (2008).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  57. IBM Corp. Released 2017. IBM SPSS Statistics for Macintosh, Version 25.0. Armonk, NY: IBM Corp.

Download references


We are very grateful to the patients who participated in the study, and we thank the Norwegian Bone Marrow Donor Registry for access to controls. We would also like to express our gratitude to Kari Sørland and Kristin Risa at Department of Oncology and Medical Physics, Haukeland University Hospital, Wenche Kristiansen and Hilde Haukeland at the CFS/ME Biobank, Oslo University Hospital, and employees at the Norwegian Sequencing Centre, University in Oslo and Oslo University Hospital. This study was funded by the Kavli Trust and the Research Council of Norway. We express our deepest gratitude. The funders did not influence the research or the manuscript in any way.

Author information

Authors and Affiliations



Ø.F., O.M., T.E., O.S., B.L. and M.V. initiated the project and applied for funding. A.L., Ø.F., E.S., D.S., O.M. and M.V. recruited and included patients. A.L., S.F. and M.V. performed sample handling, laboratory work and HLA genotyping. A.L. performed analyses and wrote the first draft of the manuscript. All authors participated in editing the manuscript and approved the final draft.

Corresponding author

Correspondence to Asgeir Lande.

Ethics declarations

Competing interests

A.L., Ø.F., E.S., S.F., D.S., O.M., T.E., B.L. and M.V. all declare that they have no conflict of interest. O.S. is member of EMECC - European M.E. Clinicians Council; otherwise he has no conflict of interest.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Lande, A., Fluge, Ø., Strand, E.B. et al. Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS). Sci Rep 10, 5267 (2020).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing