Bell's palsy (BP), defined as acute peripheral facial nerve palsy of unknown cause, is the most common acute mono-neuropathy (1). This disease affects 11–40 people per 100,000 annually (2). Typically, BP results in a partial or complete inability to move the affected side of the facial muscles. There is no race, geographic, or gender preference (3). The risk factors include diabetes, hypertension, and pregnancy (3, 4). Although the facial paresis or paralysis usually resolves within several months, permanent symptoms may occur and affect the quality of life of the patient (5). Indeed facial paralysis was significantly associated with increased depression, anxiety, and a worse quality of life (6). The cause of BP remains unknown, and many events, such as ischemia, viral infection, acute cold exposure, and anatomic causes, have been proposed as causes of BP (7). Investigators proposed that immune alterations related to a previous viral infection have an important role in the etiopathogenesis of BP (8). The reasons for this include alterations in lymphocyte subsets in the peripheral blood during the acute stage of the disease (9, 10). Additionally, lymphocytes from patients with BP were stimulated in the presence of neuritogenic basic protein from the peripheral nerve myelin sheath (11). A similar pattern of response was also demonstrated in patients with Guillain–Barre syndrome (12, 13), an autoimmune demyelination neuritis triggered by a preceding infection (14). However, serological evidence and histopathology of facial nerves support the infectious nature of BP. Herpes simplex viruses 1 and 2 and varicella-zoster viruses are the most common viruses associated with BP (15). Other infections, such as Epstein–Barr virus, cytomegalovirus, and human immunodeficiency virus, have also been reported to be associated with BP (15–17). Human papillomaviruses (HPV) are small double-stranded DNA viruses known as risk factors for multiple cancers (18–20) and are also associated with autoimmune diseases, such as systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) (21–23). The shared pathogenesis between these autoimmune diseases and BP is based on a viral infection and a subsequent autoimmune reaction (24, 25). These findings suggest a possible association between HPV infection and BP. However, there is no epidemiologic evidence to support this hypothesis. Thus, we conducted this original nationwide population-based cohort study to assess the incidence rate and risk of BP in people with a history of HPV infection when compared with those without HPV infection.

In this population study, people with HPV infection had a significantly increased risk of BP compared with the sex-, age-, and index date-matched non-HPV group. After adjusting for sex, age, and co-morbidities, the HPV group had a 25% increased risk of developing BP compared with the non-HPV group. The consistency of a positive association from different sensitivity analyses validated our findings. In our stratified analysis, we found a prominent interaction effect between age and HPV infection on the risk of new-onset BP. Among participants aged 50 years and older, having an HPV infection associated with a higher risk of developing BP than those without an HPV infection. We also observed significant interactions between HPV infection and some comorbidities (hypertension and COPD). The simultaneous presence of HPV infection and hypertension or COPD revealed a higher aHR than HPV infection alone.

Our study showed consistent findings in reporting the epidemiology of BP with previous studies. In Table 2, the incidence rate was similar in women and men, which was compatible with a previous study in the UK (28). Our study showed that patients aged above 30 years were more susceptible to BP. This result is in agreement with a previous study that showed that BP was more common in mid- and late-life, with the highest incidence in the fifth decade of life (29). We reported that hypertension increased the risk of developing BP by 50%, which is in accord with a previous case–control study that enrolled 201 patients with BP aged over 40 years or older and reported a 4.5-fold increase in the odds of BP being associated with hypertension (30).

The underlying mechanism by which an HPV infection increases the risk of developing BP remains unclear. However, an HPV infection was associated with autoimmune diseases, such as SLE, RA, and psoriasis (31, 32). A previous study showed that an HPV infection triggered SLE via molecular mimicry (22) because the HPV protein has a high and widespread similarity to several human proteins (21, 33). Similarly, molecular mimicry plays an important role in the immunopathogenic mechanisms of Guillain–Barré syndrome and related diseases (14). An antecedent infection evokes an immune response, which, in turn, cross-reacts with peripheral nerve components due to the sharing of cross-reactive epitopes. The most common identified triggering agents are Campylobacter jejuni, Mycoplasma pneumoniae, Cytomegalovirus, and Epstein–Barr virus. All of these pathogens have carbohydrate sequences in common with peripheral nerve tissues (34). Therefore, molecular mimicry might explain how an HPV infection increases the risk of BP based on the high similarity between Guillain–Barré syndrome and BP. There is currently a lack of evidence for HPV protein sequences shared with human proteins which are associated with the peripheral nervous system. Myelin basic protein, an important antigen for T cells and CNS demyelinating disease, was reported to have a region with a structural similarity to human papillomavirus peptides (35). Furthermore, inflammatory cerebrospinal fluid findings are present in BP, suggesting that it is a generalized central nervous affliction rather than a disease of the peripheral nervous system (36).

Several limitations should be considered when interpreting our findings. First, the epidemiological evaluation of a true HPV infection is challenging because many cases are not clinically recognized. Using retrospective ICD-9-based methods to select study groups of HPV infection may have a selection bias because we can only identify symptomatic patients who required medical attention. A large proportion of HPV infections may be asymptomatic; therefore, individuals who were “clinically unrecognized” could be ignored and may have been included in the comparison group. However, if HPV is associated causally with subsequent BP, misclassification would bias the estimated adjusted HRs toward the null. Therefore, for a better ascertainment of HPV exposure, we excluded patients who were diagnosed with HPV without associated treatment procedures for warts. The purpose of accruing only those patients who received associated treatment into the HPV group was to validate the diagnosis of the physician and to avoid a misclassification bias. However, some patients with an HPV infection with low-grade cervical lesions not requiring treatment might have been enrolled into the control group; similarly, such misclassification may bias our results to null, and thus our findings are acceptable. Second, the ICD-9 codes range from cutaneous benign infections to cervical and anal lesions. Grouping conditions included mucosotropic (Alpha-papillomavirus) non-carcinogenic HPVs (genital warts) to those caused by carcinogenic HPVs as well as to those in a different genus (Beta- or Gamma-papillomavirus) (plantar warts), and this is an inherent limitation of NHIRD because the ICD codes are broad but cannot provide detailed information for biological subgroup analysis. Similarly, methods using ICD-9 codes to identify cases lack information on HPV infection disease severity, which might also lead to an estimated bias because the baseline severity also affects the outcome. However, the methods applied in this study could be a proxy for HPV infection. This is because patients who request medical aid are symptomatic related to pathogenic HPV infections (either cutaneous type or mucosal type or both). These pathogenic effects of HPV infections should be prevented or treated. To overcome these limitations, further studies linking NHIRD to other disease-specified databases are needed. Third, this was a mono-country evaluation, and therefore our findings may not be applicable to non-Asian ethnic groups. Considering possible ethnic and geographical differences in the incidence and serotypes of HPV, further studies should be conducted in other ethnic groups. Fourth, HPV vaccination may also alter the immune system of the host. There are no data regarding HPV vaccination in NHIRD because HPV vaccination is a self-paid service, not covered by the NHI, and is therefore not included in the NHIRD.

In conclusion, this population-based cohort study demonstrated a higher risk of BP in patients with a previous symptomatic HPV infection, and the association was consistent in three different models. The risk was more prominent in the elderly, those with hypertension, or those with COPD. Further studies are required to clarify the underlying biological mechanisms of these associations. HPV vaccination could be another important issue for preventing BP in patients with relevant demographic characteristics and co-morbidities.

The data analyzed in this study is subject to the following licenses/restrictions: the data source used for this study was the claims data from Taiwan's National Health Insurance Research Database (NHIRD). Analysts had full access to all data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Data are available from the National Health Insurance Research Database (NHIRD) published by Taiwan National Health Insurance (NHI) Bureau. Due to legal restrictions imposed by the government of Taiwan in relation to the “Personal Information Protection Act,” data cannot be made publicly available. Requests for data can be sent as a formal proposal to the NHIRD (http://nhird.nhri.org.tw). Requests to access these datasets should be directed to http://nhird.nhri.org.tw.

The studies involving human participants were reviewed and approved by The Institutional Review Board of the China Medical University in Taiwan approved this study [CMUH-104-REC2-115-(AR4)]. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Acquisition of data and statistical analysis were conducted by H-TY. Study conception and design were carried out by K-YL, M-CC, JW, Y-MH, RC, and H-TY. Analysis and interpretation of data were performed by K-YL, M-CC, Y-MH, RC, JW, and H-TY. K-YL, M-CC, and Y-MH contributed to writing (original draft preparation). Y-MH, M-CC, and JW contributed to writing (review and editing). All the authors were involved in drafting the article or revising it and approved the final version to be published.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.