Chronic Pain and Joint Hypermobility: A Brief Diagnostic Review for Clinicians and the Potential Application of Infrared Thermography in Screening Hypermobile Inflamed Joints

Hippocrates was the first to describe joints with excessive mobility in a tribe located south of Russia. However, the term “hypermobility” was first used only in the 19th century to explain marfanoid joint characteristics [ 1 , 2 ]. Joint Hypermobility (JH) is a term given to joints whose range of motion exceeds what is considered normal for that joint, taking into account the individual’s age, sex, and ethnicity [ 3 ], commonly referred to as “loose joints,” “double-jointed,” or even used synonymously with “loose ligaments” [ 4 ]. It is a relatively common condition and can confer greater flexibility abilities to the joints of healthy individuals, acrobats, gymnasts, and dancers, and should not be considered a pathological condition or disease in these cases [ 5 - 7 ]. Some studies conducted with school-age children suggested a prevalence of JH in the general population ranging from 30% to 50% [ 8 , 9 ], reaching up to 64.6% in the age group accumulated up to 15 years old and declining over time, possibly affecting about 20% of the adult population [ 10 - 12 ].

The hypermobile syndromes, whether hEDs or HSD, constitute a group of complex conditions with multifactorial origins that remain poorly understood from both biomechanical and genomic-molecular perspectives [ 45 ]. These conditions have undergone several conceptual, semiological, and diagnostic revisions over the past decades. However, it was only in 2017 that hEDS, HSD, and benign hypermobility itself were objectively distinguished, creating a diagnostic logic for each condition based on the most current evidence [ 80 ]. Nonetheless, despite efforts to create a model that sheds light on diagnosis and, consequently, the implementation of more precise therapies and guidance, several variables related to joint hypermobility still remain underdiagnosed or inadequately diagnosed [ 43 ]. Halverson et al. used the term “diagnostic odyssey” to refer to the extensive journey that hEDS patients must undergo before receiving a definitive diagnosis, with approximately 10 misdiagnoses (on average) along the way. Indeed, previous studies have shown that hEDS patients take an average of 5 years to be correctly diagnosed [ 43 , 81 ]. In a survey conducted by Donohue et al. (2018), for example, it was observed that hEDS patients are frequently misdiagnosed as having fibromyalgia. This is likely due to the similarity of signs and symptoms between this disease and those manifested in hEDS and HSD, both in terms of musculoskeletal pain and the inherent dysautonomic states [ 82 ]. This underscores the complexity of the syndromic entanglement that can make the definitive diagnosis of conditions related to joint hypermobility a significant challenge. Both HSD and hEDS affect their carriers in a multidimensional and varied manner, encompassing physical (biological) and psychological (emotional) aspects [ 83 ]. Notably, musculoskeletal conditions, especially those related to joint instability phenomena (sprains, dislocations, and subluxations, for example), and chronic pain stand out among the others, causing significant losses to affected individuals [ 83 , 84 ]. However, headache, dizziness, persistent fatigue, and gastrointestinal symptoms, along with mental disorders such as anxiety, depression, persistent feelings of distress, and catastrophizing, are also commonly present in the daily lives of these patients [ 84 ]. History and clinical examination remain fundamental for the diagnosis of syndromic conditions related to hypermobility and cannot be underestimated. Additionally, questionnaires contribute to standardizing the diagnosis of hypermobile joints, a key point for the diagnostic hypothesis [ 85 ]. Their aim is to objectively identify JH, determine the extent of the disease, and help exclude other connective tissue diseases characterized by JH [ 82 ]. This is particularly important considering a possible connection between syndromes related to JH and some rheumatologic diseases, as noted by Rodgers et al. (2017) in cases of hEDS [ 47 ]. With the popularization and increased accessibility of infrared imaging devices, it is possible that the use of this technology could become an ally in joint assessment of individuals with syndromes related to hypermobile joints, especially regarding joint inflammatory processes. Heat, redness, swelling, and pain are cardinal clinical features of joint inflammation [ 86 ]. However, clinicians face a challenge in unequivocally diagnosing inflamed joints when there is inflammatory arthralgia in a characteristically subclinical presentation [ 87 ]. The inflammatory response is typically attributed to elevated levels of pro-inflammatory cytokines, such as tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) [ 88 ], individually or synergistically involved in many other pathological processes [ 89 , 90 ]. Dysfunctions in microvascular homeostasis are primary components in these processes [ 91 , 92 ] and may be crucial for assessing subclinical inflammatory signs. IRT would be a tool for addressing these abnormalities by early and indirectly detecting microvascular dysfunctions before typical clinical inflammatory manifestations occur [ 93 , 94 ]. In light of the case series presented here, some authors [ 63 , 95 , 96 ] have described, for example, that the thermal distribution of normal knee surfaces corresponds to a central hypo-radiant zone (relative to the patellar contour) with a progressive increase in infrared radiation towards its periphery – but this is not a universal rule. According to Marziani et al. (2023) [ 97 ], the normal anterior knee has been previously described by other authors [ 51 , 98 ] as exhibiting symmetry in the image of both knees, with an oval isothermal area showing a lower temperature corresponding to the patella – possibly due to the fact that the skin in this region is cooler over tendons and bones than over muscles [ 93 ]. The temperature of the skin over the knee may indicate the presence of an inflammatory process within the joint. This process is marked by heightened vascularity of the synovial membrane and can be readily observed in regions where there is no underlying bone [ 97 ]. Inflammation-generated heat distorts the thermal gradient, and the normal thermal pattern is lost [ 63 ]. Comparing the Pain Maps, infrared images, and comparative thermal distribution graphs, it is possible to observe that there may not always be a direct relationship between the presence of pain and potentially inflammatory mechanisms observable through IRT. In Figure 3 , the participant (a) reports pain in both knees. The regions of interest (ROI), centered on the patella, suggest apparent points of asymmetry in infrared radiation on the Rainbow Hi Contrast (rainbowHC) color scale, with greater asymmetry on the left compared to the contralateral knee. However, according to Ammer’s study (2012) [ 68 ], based on 566 thermal images from healthy individuals, the average temperature of the normal (asymptomatic) anterior knee was 29.5 ± 1.6°C (95% CI: 28.5 to 30.5°C). Thus, the ROIs of participant (a) in this example appear thermometrically normal. However, upon evaluating the temperature distribution ( Figure 3 ), we can observe an asymmetry regarding the extreme temperatures between the right and left sides, with the presence of more extreme temperatures on the left ( Table 2 ). Additionally, the thermal matrix distribution graph is more leptokurtic on the right in relation to the contralateral knee, following a non-parametric distribution (Shapiro-Wilk: p<0.001) with a statistically significant difference more relevant on the left (SDright = 0.4; SDleft = 0.61; Mann-Whitney: p = 0.001). Salisbury et al. (1983) [ 63 ], following series of individuals with “arthritic” knees under anti-inflammatory therapy, observed a gradual transition from an irregular thermal gradient with marked thermal outliers in inflamed joints to a leptokurtic distribution, ie, increasingly narrow and symmetrical, typically observed in “normal” knees. It is worth noting that in their study, thermograms were collected from knees laterally under acclimatization at 20ºC. In this case, although both knees present a loss of the so-called normal thermal grid, the left side appears to be more affected from a thermological point of view than the right. On the other hand, participant (b) indicates the right knee as painful, with both joints showing an apparent loss of the gradual transition of the thermal gradient – suggestive of dysfunction in thermological studies – with no statistically relevant thermal differences between them (SDright = 0.27; SDleft = 0.31; Shapiro-Wilk: p<0.001; Mann-Whitney: p = 0.7), with both graphs showing a wide (graphical) distribution of temperatures in an approximately symmetrical manner visually. However, even though visual symmetry is present, the left knee has a greater variation in temperatures compared to the contralateral knee. Participant (c) reports the left knee as painful again, with both knees showing a presumed loss of the normal thermal gradient. However, in this case, a more leptokurtic distribution is observed on the right, along with a statistically significant increase in thermal distribution to the left (SDright = 0.37; SDleft = 0.87; Shapiro-Wilk: p<0.001; Mann-Whitney: p = 0.043). The participant who does not present any symptoms in the knees (d), included here as a supposedly healthy control without joint hypermobility, showed a statistically significant difference between the temperature sets – higher on the right (SDright = 0.27; SDleft = 0.24; Shapiro-Wilk: p<0.001; Mann-Whitney: p<0.001) – and presented a more leptokurtic curve on the left. All these results may reinforce the potential influence of musculoskeletal pain mechanisms of neuropathic origin previously reported and common to individuals with joint hypermobility, which may be the source of pain without a necessary thermographic correspondence, as well as the subclinical inflammatory processes that can compromise the clinical assessment of joints, even those that are asymptomatic. Indeed, Fernandez et al. (2022) [ 99 ] in a retrospective study with 79 volunteers with hEDS and HSD (71% with hEDS) reported dysfunction of small fibers in 70% of those evaluated, as well as a structural deficit based on intraepidermal nerve fiber density in 78% of cases. Small fiber neuropathies consist of dysfunction of A-delta and C fibers that transmit thermal, nociceptive, and autonomic information and have already been suggested in previous cases in individuals with hEDS/HSD [ 100 - 102 ]. Consequently, frequent burning sensations, hypoesthesia, and allodynia [ 103 ] are commonly reported by individuals with these conditions. It is worth noting that, although there is considerable evidence in favor of the use of IRT in the analysis of arthritic joints [ 104 - 108 ], its use in clinical practice still does not seem to be well established, especially due to the persistence of some discrepancies among the various findings and respective conclusions [ 109 , 110 ]. However, several studies have verified objective temperature differences between diseased and healthy joints, especially when performed under ROI cooling (dynamic/functional thermography), creating greater enhancement between areas of higher and lower infrared radiation [ 111 , 112 ].

Conditions associated with hypermobile joints, whether the hEDS or HSD, are multifactorial entities and are often confused with other diseases present in various other conditions. Not infrequently, this leads affected individuals to a long search for a diagnosis that justifies such heterogeneous signs and symptoms and often not even being included in the differential diagnosis for painful musculoskeletal syndromes. The development of diagnostic criteria segments these diseases more clearly and contributes both to therapeutic and scientific efforts. IRT, although lacking consensus for its use in clinical practice, can be an ally in identifying inflamed or dysfunctional joints. However, even though showing evidence that is capable of identifying dysfunctional joint patterns, further studies are needed to correlate a dysfunctional or arthritic joint with painful symptoms. Once this goal is achieved IRT may enable professionals involved in the care of these individuals to develop therapeutic strategies that take into account the neuropathic and nociceptive components of chronic musculoskeletal pain and, perhaps, achieve better therapeutic outcomes.