In clinical practice, extraintestinal manifestations are challenging conditions that require a multidisciplinary approach and specific tools for diagnosis and follow-up to provide the best patient care. Regardless, screening protocols for extraintestinal manifestations and side effects of treatment are still not standardized in clinical trials. The most adequate approach is to refer patients to an experienced specialist in managing these manifestations [19]. In the present work, after a first evaluation of IBD patients, 30.5% exhibited at least one inflammatory musculoskeletal manifestation. This result is similar to other reports in the literature [2, 20,21,22]. When we analyzed the musculoskeletal complaints of these patients, through anamnesis and clinical exams, 17.8% had inflammatory low back pain, 15.3% had past peripheral arthritis, 2.5% had current arthritis, 26.3% reported a suggestive case of past enthesitis, and 5.3% had current enthesitis. Peripheral arthritis in patients with CD and UC occurs at a frequency that varies between 2.8 and 30% according to many published studies [21, 23,24,25]. Previous investigations demonstrated that the prevalence of inflammatory low back pain on patients with IBD varies between 9 and 30% [26].

Most of the patients who had SpA were women, mostly with CD. Similar to previous studies, SpA characteristics were more frequently reported in female patients with IBD [2].

We found that the prevalence of enteropathic SpA was 13.55%, whereas 10.16% was axial SpA and 4.23% was peripheral SpA, according to the ASAS criteria. We highlight that one patient fulfilled the criteria for both axial and peripheral SpA. The Norwegian IBSEN study followed IBD patients for 20 years and observed a cumulative prevalence of 27.9% of patients with peripheral SpA, 7.7% of patients with axial SpA, and 11.5% of patients with inflammatory low back pain [27].

Throughout the analysis of patients who received a diagnosis of axial SpA, we found that 10 (26.3%) patients had active sacroiliitis on MRI, and most (n = 9) were not using anti-TNF therapy. This suggests that the prevalence is underdiagnosed because the treatment of intestinal manifestations can reduce musculoskeletal manifestations [28, 29].

The prevalence of HLA-B27 was low (7.9%), but similar to that in another study performed in Brazil, that found 5,6% in the IBD cohort [21]. In contrast, HLA B27 were positive in 29% of the IBD patients in a cohort from the outpatients clinic of Ankara University, Turkey [30].

Karreman et al. [23], in a systematic review and meta-analysis, investigated the prevalence and incidence of SpA in patients with IBD. A total of 71 studies were included, and the grouped prevalence of sacroiliitis was 10% (confidence interval of 95% [CI] 8–12%). Geographic area, definition and use of different criteria contributed to the considerable heterogeneity of the results. This same review observed that there were few estimates for enthesitis, with prevalence varying from 1 to 54%. Enthesopathy and enthesitis can have a clinical diagnosis, but US is a highly sensitive and noninvasive tool that can improve the diagnosis. The prevalence of enthesitis in patients with IBD varies from 5 to 10% and is predominant in those with CD [26]. Bandinelli et al. [31] stated that enthesitis is a specific and sometimes isolated signal from IBD associated with SpA, which is constantly underdiagnosed or mistaken for mechanical chronic pain. In fact, in this study, anamnesis with physical exam identified only 1.7% of patients with enthesitis. Therefore, we highlight the importance of US confirming clinically symptomatic and subclinical enthesitis. After US analysis, 42.1% of patients had an MASEI score of ≥ 18, 35.2% of patients had synovitis, and 14.7% had tenosynovitis. Atzeni et al. [3] found that 33.3% of patients with IBD showed an MASEI score of ≥ 18, suggesting SpA. This information aided our differential diagnosis among possible articular and periarticular inflammatory or mechanical etiologies. Considering that the average age of the study population was 45 years old and that 69.5% were overweight or obese, there was an overlap of inflammatory and mechanical pain etiologies. Considering that noninflammatory arthralgia can be clinically diagnosed and might cause disability in patients with IBD, our study identified that, of the 118 interviewed patients, 64.4% complained of axial and/or peripheral mechanical pain and 5.1% complained of diffuse pain. Most investigations about arthropathy in IBD exclude the noninflammatory causes of articular pain [26]. A hospital-based study (ORCHARD) with 1459 patients with IBD and a population-based study (PALM) with 521 patients with IBD described prevalences of 8% and 16%, respectively, of patients clinically diagnosed with mechanical arthralgia.

Hence, adding US to the ASAS classification criteria increased the prevalence from 13 to 23% of the analyzed sample who received an enteropathic SpA diagnosis. Musculoskeletal ultrasonography can represent a valid complementary and easily available imaging technique to support clinical evaluation in the outpatient setting according to many published studies [32,33,34,35,36].

Surprisingly, 21 of 36 patients (58.3%) who reported some inflammatory musculoskeletal manifestation never consulted a rheumatologist. According to Stolwijk [2], this also occurred in almost 50% of patients who reported musculoskeletal disorders. On the other hand, according to Guillo et al. [1], the importance of rheumatologists’ opinions on investigations with patients with IBD and extraintestinal manifestations is clear. There might be many reasons for this. First, gastroenterologists do not always ask patients with IBD about musculoskeletal manifestations related to SpA or do not know exactly which symptoms belong to the spectrum of autoimmune rheumatic diseases. In addition, symptoms related to SpA have a fluctuating characteristic, and for SpA diagnosis, it is not necessary to have all symptoms at the time of diagnosis. Last, a high percentage of patients receive immunosuppressive therapy, including biological therapy, which may also influence SpAs symptoms; therefore, gastroenterologists may think that referral to a rheumatologist would not change the conduct for these patients [2]. Even then, current norms about extraintestinal manifestations of IBD do not include treatment algorithms to aid professionals in the decision-making process [37]. In this case, it is even more important to refer patients with musculoskeletal manifestations to a rheumatologist for joint management.

Finally, considering that there are numerous heterogeneous studies [7, 37,38,39,40] on plasma calprotectin in SpA, and that some investigations include all subtypes of SpA and others are centered only on ankylosing spondylitis, the present work evaluated the characteristics of plasma calprotectin in samples of patients with IBD and articular symptoms. There was no significant statistical correlation among values of plasma calprotectin and inflammatory biomarkers (ESR and CRP) and there was no difference between plasma calprotectin of the studied sample and healthy controls matched by age and sex. Most likely, they included individuals mostly with intestinal and extraintestinal manifestations under control. Future studies should evaluate individuals with moderate to high disease activity. On the other hand, the correlation between plasma and fecal calprotectin was statistically significant with a p value of 0.041. Even though it did not reach significance, it seems that the average plasma calprotectin level is higher in patients with peripheral articular disease. A larger sample size can be used to explore this issue. The literature shows heterogeneous data regarding plasma calprotectin levels (7, 28–31). Some authors report an increase in plasmatic levels of calprotectin in patients with SpA, whereas others have found similar levels between SpA patients and controls [7]. Other investigations have reported lower or similar levels in patients with SpA in comparison to patients with rheumatoid arthritis [7]. Cypers et al. [38] found increased plasmatic levels of calprotectin associated with peripheral involvement in SpA, which could explain the near normality of the observed levels in patients with only axial involvement. However, this observation was not confirmed by De Rycke et al. [40], and more data are needed. Moreover, in another study, plasma calprotectin was normal in ankylosing spondylitis, in contrast with various other inflammatory rheumatic diseases [39]. It was already demonstrated that plasma calprotectin is positively related to CRP, ESR, leukogram and platelets in patients with SpA. Nonetheless, it does not seem to be a reliable biomarker of SpA disease activity, considering that almost no correlation has been found between plasma calprotectin and the BASDAI score of disease activity and BASFI function score [7]. Ometto et al. [7] still reported that plasma calprotectin decreases rapidly after efficient treatment with TNF inhibitors, and in our samples, 39.47% of patients were under therapy with anti-TNF, which could explain the lack of correlation between them and the control group. According to Kalla et al. [41], plasma calprotectin was correlated with current biomarkers such as CRP and fecal calprotectin. In our study, plasma calprotectin showed a significant statistical correlation with fecal calprotectin but did not have a significant correlation with other biomarkers.

Another limitation is the fact that the analyzed sample was a convenience sample; therefore, patients who agreed to participate in the research could be those with a greater chance of having an articular complaint, increasing the frequency of observing musculoskeletal manifestations. It is necessary to consider that the specialist opinion was clearly influenced by the results of articular and enthesis US.

Last, the sensitivity and specificity of the ASAS classification criteria for axial SpA were 83% and 84%, respectively, and the imaging arm has been further scrutinized because evaluation of patients with nonspecific back pain has demonstrated a positive MRI in 20% of patients [42].

Finally, subjective measures and patient-reported outcomes may overestimate disease activity, and thus are unreliable in therapeutic decision making in clinical practice, since they may lead to intensification of, or switching, immunotherapy when it is not necessarily warranted. Clinical data complemented by more objective measures of inflammation such as entheseal sonography can help clarify the diagnosis [35].