Fibromyalgia (FM) is a specific disorder characterized by chronic widespread pain that belongs to the central sensitization syndromes spectrum.1 In addition to pain, chronic fatigue and sleep disturbances take part of the core symptoms of the disease.2 Up to 50–70% of patients also exhibit nonspecific gastrointestinal manifestations such as diarrhea, constipation, nausea, dyspepsia, and abdominal distention.2,3 Despite being very frequent, these symptoms are commonly overlooked by doctors and do not receive adequate attention. Frequently these manifestations are referred to as “irritable bowel syndrome (IBS)”.4,5

Recently, some authors have suggested a connection between gluten sensitivity disorders and FM. Indeed, most gastrointestinal symptoms exhibited by patients with FM are also found in patients with celiac disease or gluten sensitivity.6–8 Moreover, there is some evidence that gluten-free diet in this subgroup of patients can produce a slight but significant improvement in FM symptoms.8

Celiac disease (CD) is a chronic autoimmune disorder, primarily affecting the small intestine, caused by a reaction to dietary gluten. CD may manifest without gastrointestinal symptoms; in fact, nearly half of the CD patients diagnosed in adulthood do not have relevant gastrointestinal manifestations. In contrast, they presented with asthenia, arthralgias and myalgias, mental fatigue, insomnia, and depression, which are symptoms that also occurred in nociplastic pain syndromes such as fibromyalgia.9

Non-celiac gluten sensitivity (NCGS) is an emerging entity characterized by gluten-related intestinal and extraintestinal symptoms in patients with negative CD tests. Thus, they do not meet criteria for CD, although their symptoms improve on a gluten-free diet.10,11 Recently, evidence has showed that this entity is even more prevalent than CD, which affects 1% of the population.12 Although it may be associated to intraepithelial lymphocytosis injury to the intestinal, it neither provokes antibody production nor atrophy of the intestinal villi epithelium. Therefore, unlike celiac disease, gluten sensitivity disorder is not an autoimmune disease.12

There is no consistent data on frequency of gastrointestinal symptoms as well as no study performed duodenal biopsies to investigate CD/NCGS in Brazilian FM patients. Thus, we sought to investigate the frequency of CD and NCGS in Brazilian patients with FM treated at the outpatient clinic of a tertiary hospital, as well as to verify the association between gastrointestinal manifestations and classic FM symptoms and their impact on patients’ quality of life.

Between June 2016 and April 2017, 62 FM patients were recruited from the outpatient FM of a single center, Rheumatology Division, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo – HCFMUSP, São Paulo, Brazil. All patients fulfilled the American College of Rheumatology 2016 diagnostic criteria for fibromyalgia.2 Exclusion criteria were patient refusal, previous diagnosis of autoimmune rheumatic disease, inflammatory bowel disease, chronic hepatitis C, vitamin deficiencies, immunodeficiencies, other food intolerances or chronic diseases with inadequate control, such as diabetes and hypo-/hyperthyroidism.

This study was approved by the Local Ethics Committee, and all participants gave written informed consent.

Demographic and clinical characteristics of the patients are shown in Table 1. Ninety-eight per cent of patients were women. Mean age of the complete sample was 54.6±12.5 years.

Any gastrointestinal symptom was observed in 46 (74.2%) individuals. The prevalence of the symptoms, in descending order of frequency, were bloating (35, 56.5%), constipation (34, 54.8%), weight loss/inappetence (21, 33.9%), nausea/vomiting (21, 33.9%) and diarrhea (9, 14.5%). The mean values of the FM indexes were: WPI 8.6±4.5; SS 8.6±2.6; PDS 16.9±6.0 and FIQ 56.5±14.2.

Concerning to antibodies related to celiac disorder, no patient had positivity for tTG or EMA. Of 62 patients, 44 (71%) underwent upper digestive endoscopy. The anatomopathological findings of endoscopy were esophagitis in 10 individuals (22.7%), gastritis in 22 (50%) and duodenitis in 14 (31.8%). Only 2 (%) patients had intraepithelial limphocytosis>25%. No patient had celiac disease, according to Marsh-Oberhüber criteria. Diagnosis of NCGS was made in only one subject (2.3%).

Compared to those with no gastrointestinal symptoms, patients with at least one gastrointestinal manifestation had similar mean values of SS (8.9±2.3 vs. 7.9±3.1, p=0.318), PDS (17.2±5.7 vs 17.7±6.4, p=0.871) and FIQ (58.0±13.5 vs 52.1±15.7, p=0.128). Mean number of painful areas (WPI) were greater in people with gastrointestinal symptoms (9.3±4.6 vs. 6.6±3.9, p=0.027).

Although there was no significant association between presence of gastrointestinal symptoms and the mean value of PDS, when we categorize the scale in tertiles, those participants in the highest tertile (T3: >20) reported more frequently the presence of gastrointestinal symptoms (91.3%), compared with subjects in the lowest tertile (T1: <15) (8.7%) (p=0.037) (Table 2).

In female patients from a Brazilian tertiary hospital, we did not find any case of CD and only woman had NCGS. Nevertheless, almost 75% of the subjects had gastrointestinal symptoms (more than half referring constipation or abdominal distension). Moreover, a positive association between these manifestations and greater central pain amplification was observed.

The study has some strengths, such as the inclusion of a homogeneous sample of patients with long-standing FM from a tertiary hospital and exclusion of patients with conditions that could cause gastrointestinal disturbances. Furthermore, diagnosis of CD/NCGS was accurate since we carried out duodenal biopsy to verify the presence of lymphocytic enteritis. Several prevalence surveys on gastrointestinal disorders in FM did not perform histopathological diagnosis.5,6

In recent years, some evidence has been showed that CD and NCGS have a high prevalence in FM patients,6–8 different from our findings. In a Spanish study with 104 FM patients with associated IBS, the frequency of CD was 6.7%, and 56% of lymphocytic enteritis (Marsh type 1 lesion) was observed in duodenal biopsies.7 Moreover, in CD patients from United States, 9% of them had FM, which is higher than the prevalence of FM in general population.21 In Italy, prevalence of CD in FM was 1%,22 quite similar to our results. These discrepancies may be explained by the fact that the worldwide prevalence of CD varies according to the geographical localization and genetic background of the population. It is more common in people bearing HLA DQ2 and HLA DQ8.23 In Brazil, a previous survey in 94 FM patients found that only one (1.1%) had positive IgG anti-tTG test. This positive subject was submitted to gastrointestinal endoscopy with small bowel biopsy and result was histologically normal (relationship villus-crypt of 3:1 and with 15lymphocytes/100 cells).24 Herein, we observed only one patient with NCGS (2.3%) e none with CD.

Our study confirms and extends findings reported in other populations that also have high prevalence of gastrointestinal complaints in FM patients. Triadafilopoulos et al. showed similar frequency of gastrointestinal problems in patients with FM (73%).25 In a Spanish survey, patients with FM reported gastrointestinal symptoms more often than controls (98% vs 49%).5

Digestive symptoms in patients with FM4 are related to changes in the autonomic nervous system and functional disorders, as well as psychological stress factors. Furthermore, the overlap of FM with some central sensitivity conditions such as irritable bowel syndrome (IBS), migraines, chronic fatigue syndrome, and mood disorders, suggest common etiologic factors.3,4

Furthermore, we demonstrated that individuals with digestive complaints reported an increased intensity cardinal symptom of FM, such as painful areas, fatigue, unrefreshed sleep and dyscognition. Despite this, these individuals did not have worse FIQ values compared to patients without digestive symptoms, suggesting that their presence does not necessarily translate into a greater impairment in patient's life.

This study has some limitations, such as the limited size of the sample. Moreover, 29% of the subjects did not undergo endoscopy and biopsy, which can underestimate the true prevalence of disease related to gluten sensitivity in the population studied. Also, the absence of a healthy control group did not allow us extrapolate study findings to the general population. However, it is well established that FM patients have higher frequency of digestive symptoms, and we sought primarily to assess the association between gastrointestinal complaints and main FM domains. Thus, ultimately, the absence of a control group does not hamper the interpretation of the results related to the primary aim of the study. In addition, patients were selected exclusively from a tertiary hospital and this sample might not be representative of the overall Brazilian FM population. Another limitation comes from potential selection bias since subjects who agree to participate in a study could be more likely to communicate symptoms of any nature and this factor could overestimate the results.

In conclusion, we reported not only a high prevalence of gastrointestinal symptoms in FM patients, but also an association between these manifestations and increased degree of polysymptomatic distress (fatigue, unrefreshed sleep, and cognitive problems). On the other hand, occurrence of gastrointestinal complains was not related to overall impact of FM over different dimensions of the patient's life. Moreover, our results suggest that searching for CD in Brazilian FM patients might not be cost-effective since the frequency of CD/NCGS was very low.

This study was not sponsored by any pharmaceutical companies.

Zanetti CB has no disclosures. Pontes MAG has no disclosures. Moura EGH has no disclosures. Domiciano DS has no disclosures.