Ankylosing spondylitis (AS) is a chronic inflammatory disorder and a main outcome of rheumatic diseases for which the pathogenesis is not understood clearly (Mauro et al., 2021; Braun and Sieper, 2007). The primary pathophysiology in AS occurs in the subchondral bone and entheses (capsules, ligaments, and tendons) (Schett et al., 2017; Watad et al., 2018; Sieper et al., 2002). Intra-articular signals trigger secondary lesions of synovitis (Watad et al., 2018). Characteristic symptoms of ankylosing spondylitis are loss of spinal mobility and spinal stiffness that can be explained by structural damage and spinal inflammation (Wanders et al., 2005). Another common symptom of AS is chronic back pain that lasts a few months and often accompanied by morning stiffness. Symptoms in patients with AS have an insidious onset and inflammatory characteristics, which usually worsen at rest, improve with activity, are worse at night and in the morning, and can be improved by taking non-steroidal anti-inflammatory drugs (NSAIDs) (Navarro-Compán et al., 2021). A systematic review based on 36 studies showed that the mean prevalence of AS was 23.8 per 10,000 individuals in Europe, 31.9 in North America, 10.2 in Latin America, 16.7 in Asia, and 7.4 in Africa (Dean et al., 2014). Previous studies have shown that AS is more common in young people between 20 and 30 years old, and males are affected approximately twice more than females (Feldtkeller et al., 2003).

Extra-articular manifestations that involve multiple organs and systems are common in patients with AS and vary in incidence and severity (El, 2011). As early as the 1970s, a potential association between AS and several other diseases, such as inflammatory bowel disease (IBD) and psoriasis, was observed (Moll et al., 1974). Uveitis is the most common extra-articular manifestation of AS, with a prevalence of 20%–30% during the disease course, which increases with its duration. In these cases, acute anterior uveitis accounts for 90% (Zeboulon et al., 2008). The involvement of the gastrointestinal tract causes intestinal tissue inflammation and inflammatory bowel disease (De Keyser and Mielants, 2003; Rudwaleit and Baeten, 2006). Bone tissue involvement can increase the risk of osteoporosis and vertebral fractures (Ghozlani et al., 2009; Ulu et al., 2014). Skin involvement may result in psoriasis (Machado et al., 2013), while heart involvement can lead to ischemic heart disease (IHD) (Essers et al., 2016). These complications significantly influence the survival rate of patients with AS.

We noted that some studies have reported cases of AS complicated with other ocular inflammation conditions, including scleritis, keratitis, optic neuritis, and iridocyclitis (Karia et al., 1999; Díaz-Valle et al., 2009; Yoshikawa et al., 1992; Zhao et al., 2015; Lee et al., 2022). This suggests that there may be a potential association between ankylosing spondylitis and ocular inflammation.

Traditional observational studies are susceptible to confounding factors in the exposure environment and prone to reverse causality. Therefore, whether AS is causally linked to ocular inflammatory diseases remains undetermined. Establishing the potential links between AS and ocular inflammatory diseases is crucial for enhancing the prevention of AS complications.

Mendelian randomization (MR) is a statistical approach based on genome-wide association studies (GWASs) used to construct instrumental variables (IVs). When the basic assumptions are met, it can effectively reduce confounding factors and reverse causality to determine the association between exposure and outcome (Davies et al., 2018; O’Donnell and Sabatine, 2018). In this study, we conducted a two-sample MR analysis to explore the causal relationship between AS and ocular inflammatory diseases.

The quality control and screening procedures described above showed that the number of IVs included in the analysis to explore the effect of AS on various ocular inflammations varied from 20 to 24. Each independent IV included in the analysis had an F-value greater than 10, indicating an absence of bias resulting from weak IVs. Supplementary Table S1 provides detailed information.

Our results suggest that AS is associated with an increased risk of uveitis (OR_IVW = 2.825, 95%CI_IVW = 1.709–4.672, and P_IVW < 0.001), iridocyclitis (OR_IVW = 3.806, 95%CI_IVW = 2.809–5.157, and P_IVW < 0.001), scleritis (OR_IVW = 1.738, 95%CI_IVW = 1.190–2.539, and P_IVW < 0.001), and episcleritis (OR_IVW = 5.113, 95%CI_IVW = 2.067–12.645, and P_IVW = 0.004) in the European population. Figure 2 shows the results of the association between AS and ocular inflammatory diseases using IVW, MR–Egger, weighted median, simple mode, and weighted mode methods. The blue dot in the figure represents the OR value, and the blue line segment represents 95%CI. The scatter plot shows the causal relationships between AS and ocular inflammatory diseases using different methods (Figure 3). Each black dot in this figure indicates an SNP, and the lines correspond to the effects of AS on ocular inflammatory diseases using different MR methods.

The results of the MR–Egger intercept indicated that there was no horizontal pleiotropy in the four groups (p > 0.05). The leave-one-out test showed that the overall effect was not driven by the individual SNP (Figure 4). The heterogeneity test showed no heterogeneity between SNPs in the MR analysis of AS on scleritis (Q_MR–Egger = 18.421, p = 0.622; Q_IVW = 19.010, p = 0.645) and episcleritis (Q_MR–Egger = 33.767, p = 0.051; Q_IVW = 34.246, p = 0.062). However, the heterogeneity test showed heterogeneity between SNPs in the MR analysis of AS on uveitis (Q_MR–Egger = 31.730, p = 0.034; Q_IVW = 31.751, p = 0.046) and iridocyclitis (Q_MR–Egger = 44.674, p < 0.001; Q_IVW = 48.244, p < 0.001). We used the random-effects model as a random-effects IVW approach allowed the existence of heterogeneity. Supplementary Figure S1 shows the funnel plots.

No correlation was found between genetically predicted AS and keratitis (OR_IVW = 1.041, 95%CI_IVW = 0.886–1.222, and P_IVW = 0.628) and optic neuritis (OR_IVW = 0.868, 95%CI_IVW = 0.441–1.709, and P_IVW = 0.682).

In this study, we performed a two-sample MR to explore the relationship between AS and ocular inflammatory diseases. We found that AS increased genetically predicted risk for uveitis, iridocyclitis, scleritis, and episcleritis through large-scale GWAS data.

Previous studies on the ocular extra-articular manifestations of AS have primarily focused on the anterior uveitis, which is considered to be the most common extra-articular manifestation of AS (Braun and Sieper, 2007; Zeboulon et al., 2008). Uveitis is an inflammation of the uveal tract of the middle layer of the eye, which includes the iris, ciliary body, and choroid. Inflammation in the anterior layer of the uvea mainly affects the iris and ciliary body, known as iridocyclitis, whereas inflammation in the posterior uvea mainly affects the retina and choroid, also known as chorioretinitis (Selmi, 2014; Tuğal-Tutkun, 2023). An epidemiological study conducted in North America showed an incidence of uveitis of 52.4 per 100,000 person-years and a prevalence rate of 115.3 per 100,000 person-years (Gritz and Wong, 2004). Clinically, uveitis is often acute in onset and characterized by periorbital pain, photophobia, redness, and blurred vision. The prognosis for acute uveitis is generally positive, and satisfactory outcomes can be achieved with timely and effective treatment. However, if left undiagnosed and untreated, it can also turn into chronic progression that may result in a series of severe complications, including glaucoma, macular edema, cataract, iris, and lens adhesion, and even permanent visual loss (Selmi, 2014; McCluskey et al., 2000). Clinicians should remain vigilant for the potential development of uveitis in patients diagnosed with AS. Early diagnosis and treatment of uveitis are of great significance to improve the prognosis of patients and the medical burden on society.

The human major histocompatibility complex (MHC) comprises human leukocyte antigens (HLAs) encoded on chromosome 6 (Bowness, 2015). It was recognized that AS was significantly associated with HLA loci (HLA B27) in pathogenesis and stronger than any other rheumatic disease, as reported decades ago (Brewerton et al., 1973a; Schlosstein et al., 1973). Epidemiological studies have shown that HLA B27 is present in over 90% of patients with AS, and the development of AS carries a risk as high as approximately 5% in individuals who are HLA B27-positive (Khan, 1996; van der Linden et al., 1984). Positivity for HLA B27 can explain 20%–30% of the genetic risk (Braun and Sieper, 2007). In addition to HLA B27, HLA DR1 and HLA B60 may also be associated with AS (Rudwaleit and Höhler, 2001). Possible targets of an autoimmune response in AS are proteoglycan and proteoglycan II (McGonagle et al., 1998). Typical features of AS were presented in animals immunized with proteoglycan (Zhang, 2003).

During the same period, Brewerton et al. observed that around half of the patients with acute anterior uveitis were HLA B27-positive, indicating a significant association between HLA B27 and its pathogenesis. This finding established a genetic link between AS and acute anterior uveitis (Brewerton et al., 1973b). Researchers suggested that HLA B27 may mediate changes in the normal gut microbiota, causing secondary inflammation that leads to the development of AS and uveitis (Hammer et al., 1990; Rosenbaum and Asquith, 2018). In our analysis, we used two-sample MR analysis to confirm that the risk of uveitis and anterior uveitis (iridocyclitis) is significantly associated with AS. Unfortunately, in the reverse MR analysis, we tried but failed to identify sufficient IVs to demonstrate that uveitis increases the risk of AS. This may be due to the insufficient sequencing depth of GWAS data for AS that we used (total number of SNPs = 99,962).

The analysis results showed a significant increase in the genetically predicted risk of scleritis and episcleritis in patients with AS. More attention seems to have been paid to the association between rheumatoid arthritis and scleritis. However, very few studies have explored the risk of scleral inflammation (scleritis and episcleritis) in patients with AS or simply considered scleritis a potential complication of uveitis. The sclera belongs to the outer structure of the ocular wall, accounting for about five–sixth of the entire outer wall. Scleritis is a heterogeneous group of chronic autoimmune diseases centered on the sclera, involving adjacent structures including the cornea and uvea (Okhravi et al., 2005; Lin et al., 2006). Scleritis is characterized by the insidious onset of redness and ocular pain. The severity of scleritis ranges from mild self-limited inflammation in the superficial layer to severe necrosis of the underlying layer, and its severe complications include glaucoma, cataract, uveitis, and corneal perforation (Wakefield et al., 2013; Sainz de la Maza et al., 1994). In recent years, many cases of severe scleritis have been recorded (Raven et al., 2016). The pathogenesis of scleritis is still unclear. It has been found that matrix metalloproteinases (MMPs) are expressed in the sclera tissues affected by necrotizing scleritis and modulated by cytokines, so this tissue necrosis may be partly attributed to MMP activity (Wakefield et al., 2013; Di Girolamo et al., 1997). Apart from that, scleritis and episcleritis should not be confused. Episcleritis is typically characterized by redness, mild ocular discomfort, and pain. In contrast, scleritis is associated with more severe pain and tenderness that is not present in patients with episcleritis (Wakefield et al., 2013; Watson and Hayreh, 1976). Therefore, clinicians should be aware of the risk of scleritis and episcleritis in patients with AS, in addition to uveitis.

Keratitis can be caused by various factors, such as herpes simplex virus infection, fungal infection, and Acanthamoeba infection (Author Anonymous, 1988; Rowe et al., 2013; Brown et al., 2021). In our analysis, we did not identify a potential link between keratitis and AS. However, attention should be paid to the possibility of keratitis secondary to other ocular inflammation. Similarly, we found no potential association between AS and optic neuritis. Current studies suggest that there is a strong association between optic neuritis and multiple sclerosis (MS). About 70% of MS patients present with optic neuritis (Toosy et al., 2014).

Major strengths of this study include the following: we performed MR analysis using large-scale GWAS data, effectively eliminating possible confounding factors in observational studies and using multiple analysis methods to obtain robust results. GWAS data on AS and ocular inflammatory diseases were collected by different consortiums, so there was almost no overlap in the different groups of samples.

There are also some limitations to our study. First, all the samples were obtained from European populations, so the results cannot be effectively generalized to other races or regions. In addition, heterogeneity was identified in certain group analyses. Furthermore, we were unable to elucidate the exact mechanism of ocular inflammation caused by AS. Based on the available statistics, we failed to extract enough IVs in the reverse MR analysis to obtain the possibility of reverse causality between ocular inflammatory diseases and AS, which may be due to the insufficient sequencing depth of GWAS data for AS (total number of SNPs = 99,962). Therefore, the results of this study have diagnostic limitations. Future studies with larger sample size are needed to verify our results.

In summary, the evidence presented in the two-sample MR analysis suggests that there may be a causal relationship between genetically predicted AS and ocular inflammatory diseases. Our results showed that AS increased the genetically predicted risk for uveitis, iridocyclitis, scleritis, and episcleritis. Future studies are needed to elucidate the specific mechanisms. Researchers and clinicians should be aware of the risk of ocular inflammation in patients with AS.