Psoriasis is a chronic, immune-mediated, inflammatory skin disease characterized by the rapid proliferation of skin cells, resulting in the formation of distinct red scaly patches on the skin. Psoriasis vulgaris is the most common clinical variant (1, 2). Previous researches have suggested that autoreactive T cells contribute to the pathogenesis of psoriasis, which indicates that psoriasis may be autoimmune in nature (3–6). Additionally, psoriasis patients have been found to have a higher prevalence of various immune-mediated disorders, including autoimmune bullous diseases. (AIBDs) (7–9).

Bullous pemphigoid (BP) is a chronic autoimmune subepidermal blistering disease characterized clinically by urticarial pruritus and tense blisters on an erythematous background (10). It is associated with autoantibodies which directed against two hemidesmosomal proteins: BP180 (BP antigen2, collagen XVII) and BP230 (BP antigen 1) in the basement membrane zone of the epidermis (11). The specific mechanisms by which BP can be induced are unknown, although the majority are linked to factors such as genetics, drug intake, viral infections, physical agents, and diet that may locally disrupt the skin basement membrane zone (BMZ) (12, 13). In recent years, increasing researches have indicated a link between psoriasis and BP (14–16), especially psoriasis vulgaris (8). However, since potential confounding factors cannot be fully evaluated or controlled in observational studies, the causality of the link between psoriasis vulgaris and BP remains unknown. In order to have a deeper understanding of their pathogenesis and timely identify the patients at risk, it is crucial to evaluate the causal relationships between psoriasis vulgaris and BP.

Mendelian randomization (MR) is a typical method of drawing conclusions about causal relationships between exposures and outcomes by using genetic variations [i.e. single nucleotide polymorphisms (SNPs)] as instrumental variables (IVs) (17, 18). Since the genetic variations have formed before born, MR estimates are not affected by reverse causation or confounding factors, overcoming the limitations of observational investigations (19). In order to examine the causal relationship between psoriasis vulgaris and BP, a two-sample bidirectional MR analysis was carried out using publicly available summary statistics from genome-wide association studies (GWAS) data.

The present study, to the best of our knowledge, was the first MR analysis study to evaluate the bidirectional causal relationships between psoriasis vulgaris and BP. It’s interesting to note that psoriasis vulgaris was significantly correlated with a higher risk of BP, employing genetic instruments from the top most accessible GWAS data. However, no evidence was discovered by reverse MR to support the role of BP in the risk of psoriasis vulgaris.

In recent years, several cases and clinical studies have demonstrated the potential relationship between psoriasis vulgaris and BP (8, 14, 15). In one study that examined 145 cases of coexisting autoimmune bullous diseases (AIBDs) and psoriasis, the most prevalent AIBD in patients with psoriasis was BP (63.4%) and the most prevalent type of psoriasis was psoriasis vulgaris (14), indicating the potential association between psoriasis vulgaris and BP. Another study also demonstrated that patients with BP had a higher prevalence of psoriasis than those without BP (5.2% vs. 1.2%; OR, 4.4; 95% CI, 2.2-8.9; P<0.0001) and psoriasis vulgaris was the most common type (8). Furthermore, psoriasis was developed before BP in 93.3% of the patients who had both conditions (8). According to a literature review containing 40 case reports (7), psoriasis predated BP in most of the instances, with an average interval of 20 years (range: 1–60 years). The occurrence of psoriasis prior to BP in most cases indicated the potential impact of psoriasis on developing BP. However, since the results of abovementioned observational studies might be influenced by uncontrolled residual confounding factors such as comorbidities, lifestyle, and socioeconomic levels (8, 27, 28), the relationship between psoriasis and BP as well as the direction of the relationship remained undetermined. Thus, in the present study, we employed an MR approach, which is free of common confounding biases existing in observational studies (29), to determine the causal relationship between psoriasis vulgaris and BP. Overall, the result of our MR analysis was consistent with the abovementioned observational studies, and further demonstrated the causal effects of psoriasis vulgaris on BP.

However, the risk of psoriasis after BP remains unclear. A survey of 104,669 Japanese patients with psoriasis revealed that treating bullous diseases with corticosteroids was likely to trigger an immediate development of pustular psoriasis. In contrast, evidence is scarce for psoriasis vulgaris developing after BP in previous research (30). Our MR analyses did not detect a pronounced role of BP in the risk of psoriasis vulgaris either. Combined with previous research, we revealed an interesting phenomenon that psoriasis vulgaris causally increases the subsequent BP risk, yet a significant effect of BP on psoriasis vulgaris development was not found. The reason for the phenomenon remained unclear.

Previous basic research initially investigated the coexisting mechanism of BP and psoriasis. Multiple investigations have shown similar dysregulated cytokine expression and inflammatory responses in the immune systems of psoriasis and bullous pemphigoid patients (27, 31, 32). For instance, IL-1 was found necessary for the development of psoriatic lesions such as epidermal proliferation and differentiation (31), and the level of IL-1b was found to be higher in blister fluid than in BP serum (33). The shared cytokines may play a part in the coexistence of psoriasis and AIBD. Furthermore, the keratinocytes present in both psoriasis and BP can produce neutrophil chemokines, such as (IL)-8, resulting in a common histological feature of neutrophil infiltration, which further leads to the releasing of various metalloproteinases, including disintegrant and metalloproteinases ADAM9, ADAM10, and ADAM17. These enzymes contribute to extensive degradation of matrix proteins and subsequent exposure of self-epitopes at the dermal-epidermal junction (27). In addition, prior research has demonstrated that both involved and uninvolved psoriatic lesions in patients with psoriasis vulgaris exhibit disruptions to laminin 1 and laminin a1 inside the basement membrane zone (BMZ) (14, 34). The study hypothesized that the antigenicity of basement membranes in patients with psoriasis began to decrease at a younger age, which promotes the generation of antibasement membrane autoantibodies and the development of BP earlier in life (14). These partially explained our findings that psoriasis vulgaris causally increases the risk of BP development, possibly due to the neutrophil infiltration or other systemic inflammation of psoriasis vulgaris that may cause exposure of self-epitopes at the dermal-epidermal junction and further induce BP.

However, despite the similar cytokines found in the pathogenesis of psoriasis and BP, some cytokines of the two diseases were quite different, indicating different pathogenesis. While the IL-23/T-helper (Th)17 inflammatory pathway, which involves tumor necrosis factor (TNF)-alpha, IL-23, and IL-17, is critically involved in psoriasis pathogenesis (35, 36), the Th-2 pathway, primarily involving IL-4, IL-5, IL-6, IL-10, IL-8, IL-18, and IL-31, has been identified as the essential driver of BP antibody production (37–40). The complexity may partially explain why BP cannot increase psoriasis vulgaris in turn. Psoriasis with BP may offer an intriguing representation of the intricacy of the cutaneous inflammatory networks (41).

Partially due to the complexities of cutaneous inflammatory networks, some of the treatments for psoriasis vulgaris or BP are quite selective (42). On one hand, several anecdotal cases of autoimmune bullous diseases were reported to be caused by biologics targeting important cytokines in psoriasis (43–45). Besides, while UV radiation has shown efficacy in treating psoriasis, they were found to exacerbate or induce BP (46). On the other hand, although dupilumab, a dual inhibitor of IL-4 and IL-13 signaling, has been used successfully in BP, it is generally known to cause psoriasis (47). The contradictory nature of above treatment suggested the complex relationship between the two diseases. Thus, despite the causal effects of psoriasis vulgaris on BP and the insignificant effect of BP on psoriasis vulgaris found in this study, psoriasis should be treated with caution to prevent the onset of BP, and vice versa. Fortunately, some previous findings suggested that tofacitinib could be a safe and useful treatment for people with combined psoriasis and BP, which shed light on the treatment guidance for both comorbid disorders (48, 49). To realize precise prevention and early diagnosis, the exact path mechanism underlying the causal relationship is warranted to be investigated deeply.

This MR study investigated the causal relationship between psoriasis vulgaris and BP and possessed several strengths over previous studies. First, due to the MR study design, this study is free of many biases inherent in observational studies, such as confounding, reverse causality, and so on (29). Second, since the primary IVW method is prone to horizontal pleiotropy, additional robust MR analytical methods were applied in the sensitivity analysis. Furthermore, outlier SNPs identified in the MR-PRESSO outlier test were excluded in the final analysis, which reduced the possibility of horizontal pleiotropy and enhanced the reliability of our results (50).

Nevertheless, the present study was not free from limitations. First of all, although various methods were utilized in the sensitivity analyses to enhance the robustness of results, the difference in the significance of the results among different MR methods made the conclusion obscure, especially for the effect of psoriasis vulgaris on BP. However, the OR results by all the employed methods were directionally consistent and the OR values were strong, indicating the existence of the causal effects of psoriasis on BP ( Table 1 ). Secondly, the data source is limited in two ways: 1) the sample size of BP is relatively small; and 2) only psoriasis vulgaris in relation to BP was investigated, while other forms of psoriasis remain unclear. So, further GWAS data and studies are needed. Thirdly, the GWAS datasets employed in our research were based on European. So our results may not be fully representative of the whole population. Moreover, it is worth noting that population stratification, a common confounding factor, may bring about false–positive genetic associations, which result in biased IVs for MR (21, 50). Therefore, a larger multi-racial GWAS will boost the efficiency of future MR studies to clarify the actual relationship between psoriasis vulgaris and BP, and we should carefully utilize our conclusion in racially and ethnically diverse populations. Fourthly, the potential pleiotropy effect in our study cannot be cleared off, although the MR-Egger intercept in our study indicated little horizontal pleiotropism and we took steps to identify and exclude outlier variants (51). Lastly, in the majority of cases, genetic variation may only account for a tiny percentage of exposures or traits (52, 53). The listed limitations suggested that the risk of developing psoriasis vulgaris in patients with BP might remain to be confirmed, and vice versa. Although we found no evidence for the causal effects of BP on psoriasis vulgaris, further studies are needed to finally clarify the subject.

Based on our bidirectional two-sample MR analysis, genetically predicted psoriasis vulgaris is causally associated with an increased risk of developing BP. Foremost, our research serves as a reminder to clinicians that in daily clinical practices, heightened attention should be given to the emergence of symptoms such as bullous tense blisters, urticarial plaques, or erythema in patients with psoriasis vulgaris, enabling early diagnosis and timely treatment of BP. The biological mechanisms behind this causality are complex and needed to be investigated in further studies, in order to decrease the occurrence of new-onset BP in psoriasis vulgaris patients as well as determine the investment in screening concomitant conditions in both directions.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

AZ: Writing – review & editing, Writing – original draft, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. ZY: Writing – review & editing, Supervision, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. TH: Writing – review & editing, Project administration, Methodology, Formal analysis. MW: Writing – review & editing, Supervision, Resources, Project administration, Conceptualization.