In this study, we used bidirectional MR analysis to screen three specific groups of arthritis for associations with the volume of specific cerebral cortical regions: arthrosis promotes an increase in the volume of the Left Crus I Cerebellum region; Gout promotes a decrease in the volume of the Left Frontal Operculum Cortex region; and Gout promotes an increase in the volume of the Left Precentral Gyrus region. Correspondingly, reverse MR analysis yielded no positive results. To the best of our knowledge, this is the first study to apply Mendelian randomization to explore arthritis with specific cerebral cortical volumes on a large scale.
Whether peripheral inflammatory diseases such as arthritis can affect the central nervous system has been a hot topic in immunological and neuroscientific research, and the positive results of our study confirm to some extent the possibility of an association.Given the important role of pro-inflammatory cytokines in the pathogenesis of arthritis, this effect may be largely achieved through pro-inflammatory cytokines.When peripheral inflammation, such as arthritis, occurs in the body, a large number of cytokines are activated. As the main inflammatory mediators, these circulating cytokines not only participate in inducing and maintaining the pathological process of arthritis but also act on the CNS through various pathways[13].One of the classic pathways is propagation through neurons. Studies have shown that IL-17, IL-1β, TNF-α, etc. can act on peripheral nociceptive sensory neurons, thereby affecting the release of cytokines in the spinal cord and ultimately inducing central pain sensitization[14]. In turn, therapeutic neutralization of these cytokines can reduce pain in patients[15].The vagus nerve, as a potential second nerve propagation route, has also been observed to be activated by TNF and IL-1β under certain conditions[16].In addition to neurotransmission,volume diffusion allows these cytokines to enter the brain through periventricular organs, which lie outside the blood-brain barrier[17].Furthermore, circulating cytokines can also directly enter the CNS through the blood-brain barrier. In this case, direct disruption of the blood-brain barrier may occur, with cytokines infiltrating into the brain through the injury[18].For instance, Park et al. observed disruption of the blood-brain barrier in a collagen-induced arthritis (CIA) model[19].Cytokines have also been reported to be actively transported into the brain through cytokine transporters at the blood-brain barrier[20].
The multiple pathways of action of peripheral inflammation on the brain ultimately lead to the production of centrally derived pro-inflammatory cytokines by microglia in the brain. These cytokines can slowly diffuse in the brain and are accompanied by the rapid activation of some specific neural pathways[21].Cytokines act differently in the brain depending on their receptor distribution and localization. For example, the granule cells of the dental gyrus, the pyramidal cells of the hippocampus, and the anterior pituitary gland are all known to bind to IL-1 receptors[22].
In view of this, our experimental results may imply the existence of a concentrated distribution of specific cytokine receptors in Left Crus I Cerebellum, Left Frontal Operculum Cortex, and Left Precentral Gyrus.
A common form of arthritic pain is knee osteoarthritis (KOA), resulting primarily from degenerative changes in articular cartilage and secondary bone-building development. Millions of people worldwide suffer from chronic pain, stiffness, and gait abnormalities caused by this condition. Motor deficits and emotional and cognitive deficits can result from chronic KOA pain[23]. Previous neuroimaging studies have shown that osteoarthritis can affect the structure and function of the brain, as well as the plasticity of the cortex.For example, Barroso et al. reported that patients with KOA had lower gray matter (GM) volumes in the precentral cortex than healthy controls (HC)[24], and Lewis et al. found that patients with KOA had decreased GM volumes in the bilateral amygdala, the nucleus ambiguus, and the ipsilateral primary somatosensory cortex compared with HC[25].In our results, Arthrosis was found to promote increased volume in the Left Crus I Cerebellum area, whereas in another previous study on the effects of KOA on brain function, it was found that KOA could lead to abnormalities in the left cerebellum function of patients, as evidenced by low fALFF measurements in this brain region of the patients compared to HC (fALFF score is the ratio of the power of the low-frequency component to the power of the full-frequency component)[26].Our findings are consistent with those of previous studies, which suggest that arthritis may contribute to functional abnormalities in the Left Crus I Cerebellum region by increasing its volume.Based upon the specificity of its function, the Left Crus I Cerebellum is the most prominent and laterally expanded region of the human cerebellum[27]. In addition to limb motor control and behavioral cognition, this region is also involved in sensorimotor adaptations to pain[28]. We hypothesized that the emotional experience of chronic pain in arthritis patients may be associated with changes in this brain region.
Gout is another common type of inflammatory arthritis. A disorder of uric acid metabolism leads to an elevated blood uric acid concentration, which triggers localized inflammation due to the deposition of uric acid crystals around the joints. Joints affected by this condition are markedly inflamed and painful[29].It has been traditionally believed that the central nervous system (CNS) is rarely involved in gout. Nevertheless, recent studies have indicated that hyperuricemia may have some beneficial effects on the CNS when it comes to the development of gout, which is particularly evident in neurodegenerative and psychiatric disorders.For example, it has been shown that hyperuricemia is associated with a lower risk of Alzheimer's disease[30]. Black CN et al. found that the severity and duration of symptoms of major depression or anxiety disorders were negatively correlated with uric acid levels[31].Additionally, it has been reported that gout affects the volume of specific brain regions as well. In a Mendelian randomization study of gout and brain volume, researchers found that genetically predicted gout was significantly associated with gray matter volume in the whole brain, and genetically predicted hyperuricemia was also significantly correlated with gray matter volume in several regions, including the cerebellum, midbrain, pons, and brainstem[32]. Compared with this previous study, our study provides a more detailed delineation of brain regions, further revealing the correlation between gout and the more subtle brain regions of the Left Frontal Operculum Cortex and Left Precentral Gyrus.Furthermore,Yang et al., in an imaging study using MRI technology to detect cortical volume in gout patients, found that the cortical thickness in the left upper frontal lobe region of gout patients was thinner than that of the healthy control group[33]. This finding is similar to the genetic prediction in our study that gout can cause a decrease in the volume of the Left Frontal Operculum Cortex. And this strongly supports our experimental results.In fact, both the Left Frontal Operculum Cortex and the Left Precentral Gyrus have extremely complex functions, and when focusing on their functions and pathways related to peripheral inflammation and pain, we found that there are reports that the left frontal-amygdala pathway is involved in endogenous sexual pain inhibition[34], while the Left Precentral Gyrus has been reported, along with other subcortical regions, to be involved in modulating pain and inducing pain synchronization[35]. Based on this, we speculate that gout may induce volume changes in the Left Frontal Operculum Cortex and Left Precentral Gyrus, thereby affecting the pain perception function of gout patients.Given the previously reported potential protective effects of gout and hyperuricemia on some neuropsychiatric diseases, Left Frontal Operculum Cortex and Left Precentral Gyrus may also be potential targets for the effects of uric acid in the CNS.
Recent studies suggest that some central nervous system diseases may be genetically linked to arthritis. Researchers have found 15 single nucleotide polymorphisms (SNPs) associated with RA and frontotemporal dementia (FTD) when comparing genome-wide association studies (GWAS) in neurodegenerative diseases and chronic immune-mediated diseases.Notably, most of these SNPs are found on chromosome 6 in the human leukocyte antigen (HLA) region[36].Felsky et al. also demonstrated a correlation between polygenic risk for RA and microglia density in the brains of older adults[37].The Mendelian randomization study is a genetics-based analysis, and our findings reveal a genetic association between specific arthritis and specific cerebral cortical regions, which may provide ideas for further exploration of the common genetic susceptibility that exists between arthritis and CNS disorders.
Our study has the following advantages: A primary advantage of MR analysis is that it has a larger sample size than traditional neuroimaging methods, as well as the ability to effectively avoid interference from reverse causality and potential confounders. Additionally, multiple sensitivity analyses were conducted in this study to ensure robustness. In comparison with previous studies, we have also delineated the cortical regions in a more detailed manner, revealing the relationship between arthritis and certain fine structures in the brain in a more precise manner.
Nevertheless, there are some limitations to this study. As a first point, the data used in this study were drawn from Europeans, and the findings may not be generalizable globally. Further, the human brain is capable of performing complex neurobiological functions by interacting synergistically with multiple brain regions. Our findings only indicated a correlation between arthritis and a single brain region, which cannot be interpreted as reflecting the presence of specific brain pathways and networks. Finally, despite the fact that reverse MR did not yield any positive results, this does not necessarily suggest that the CNS is not able to influence peripheral inflammation in an inverse manner. An example would be the vagus nerve, which could influence the onset and progression of arthritis by modulating the production of TNF and other proinflammatory cytokines that contribute to the suppression of immune responses in the CNS[38].