Discussion
We performed a nationwide cross-sectional correlational study featuring 5794 GBS cases from the prepandemic (2017–2019) and pandemic (2020–2021) periods. Specifically, we examined the demographic distribution and seasonality of GBS, compared the real-world incidence of GBS with the forecasted trajectory based on the prepandemic incidence data and evaluated the temporal association between the time-series trends of common infectious diseases and GBS. This study is the first to integrate temporal changes in the incidence of GBS and common respiratory and gastrointestinal infectious diseases during the pandemic era. The key findings of this study are as follows: (1) in general, older adults and men showed a higher incidence of GBS, although differences in certain age groups and interruption of seasonality were detected during the pandemic; (2) the actual incidence of GBS was lower than the forecasted estimates in 2020, and a statistically significant decrease in common respiratory and gastrointestinal infectious diseases was temporally associated with the downward incidence trend; and (3) the actual GBS incidence in 2021 was higher than the forecasted estimates in 2021, although the infectious disease landscape did not change significantly from that in 2020.
First, with regard to demographic factors, a higher incidence of GBS was detected in older adults and men during the COVID-19 pandemic, which is consistent with the results of prepandemic large-scale epidemiological studies.18 19 Notably, compared with the prepandemic incidence, the GBS incidence during the pandemic was lower than expected across most age groups below 60 years, which may be attributable to socioeconomic and behavioural changes during the pandemic. Specifically, age groups 30–39, 40–49 and 50–59 years comprised nearly 65% of the economically active age groups in 2020–2021 and were responsible for the most intimate workplace human–human interactions. We postulated that policy measures such as lockdowns, increase in at-home workforce and NPIs collectively influenced the incidence of community-based infections that originally spread via contact, droplets or airborne modes of transmission, and these subsequently contributed to a decline in postinfectious GBS. Similarly, younger age groups (<10 years old and ages 10–19 years, which represent the school-age population) underwent a drastic transition from regular schooling to online schooling during the early to-mid pandemic period, preventing their excessive exposure to many community pathogens. Second, a brief interruption in the seasonal pattern of GBS was observed during the pandemic. Previous longitudinal epidemiological studies on GBS incidence in South Korea have suggested a peak incidence in the late spring to summer months.10 However, in our analysis, we observed a double peak during the prepandemic era: from late spring to summer and in the winter months. Although the late spring to summer peak was maintained during the pandemic period, likely due to a lesser decrease in gastrointestinal infections (especially from bacterial pathogens), the winter peak diminished in 2021, likely due to a substantial decrease in the outbreak of respiratory infections, which mostly tended to have a peak incidence during the winter in South Korea (figure 2A and table 1).
The incidence of GBS was significantly lower in 2020 than the forecasted estimates based on the prepandemic (2017–2019) incidence data. We hypothesised that this was due to a decrease in non-COVID-19 infectious triggers during the COVID-19 pandemic. A potential link between SARS-CoV-2 infection and GBS has been suggested in a systematic review, where sevenfold to eightfold increased prevalence of GBS has been reported in patients with COVID-19, compared with the general population rates.20 At a population level, however, previous epidemiological studies have indicated a rather decreased incidence of GBS overall during the pandemic.21 Additionally, a study based on the diagnoses of alleged infectious aetiology showed that respiratory and gastrointestinal infectious diseases have significantly decreased during the pandemic period.12 Therefore, we inferred that the change in the incidence of non-COVID-19 infectious disease may have exhibited a close temporal association with the changes in GBS incidence during the pandemic months.
The incidence of common non-COVID-19 respiratory and gastrointestinal infections that were evaluated from sentinel surveillance data were significantly lower than expected during the pandemic era. This finding may be attributable to the effect of NPIs that were implemented at a national level from as early as 6 May 2020, to mitigate SARS-CoV-2 spread22 23; mask wearing may have effectively decreased respiratory pathogen spread23; and hand hygiene may have prevented the spread of gastrointestinal pathogens.24 The lockdowns prevented social gatherings and eating out, thus limiting the spread of communicable diseases.25 Notably, there was a significant decrease in almost all respiratory sentinel surveillance pathogens from 2020 to 2021. In contrast, only a few gastrointestinal pathogens showed a significant decrease during the same period. With regard to the time-series correlation analyses for infectious diseases, respiratory pathogens, human bocavirus and parainfluenza virus showed a significant positive correlation with the incidence of GBS, whereas only Campylobacter spp. showed a significant positive correlation for gastrointestinal pathogens. In 2020–2021, the respiratory pathogens lost statistical significance in their correlation, but a host of other gastrointestinal pathogens such as Salmonella, pathogenic E. coli, Y. enterocolitica, adenovirus and Campylobacter showed a positive temporal association, reflecting their increased explanatory power over the incidence of GBS during the pandemic. We hypothesised that these gastrointestinal pathogens were the main drivers of postinfectious GBS in 2020 and 2021. The loss of significance in respiratory infection may probably be due to a significant decrease in respiratory infectious disease, as discussed earlier; this may have acted as a force that shifted the GBS incidence curve further down the y-axis (figure 2B). Moreover, the significantly lower incidence of GBS in 2020 from March to May, which are traditionally the peak incidence months, added to the magnitude of the downward trend that was presumably caused by the decrease in respiratory and gastrointestinal infections. The same interpretation can be implemented to explain the seasonal pattern of GBS during the pandemic, for which there was a significantly lower incidence during the spring and winter months than expected. Of interest, SARS-CoV-2 infection during the pandemic period showed no significant temporal association with the incidence of GBS in our study. Although an increased risk of GBS may be suspected in patients with COVID-19, this may not have translated into the whole population, where non-COVID-19 infections serve as the major driving force behind GBS onset.
In 2021, a general decrease in non-COVID 19 infectious diseases may have persisted; however, since February 2021, a new variable to the equation was introduced: the SARS-CoV-2 vaccine. During this period, the common respiratory and gastrointestinal infection landscape did not change significantly from 2020 (table 3), but GBS incidence increased above the forecasted estimates. This finding holds special value, because from June 2021 to November 2021, COVID-19 incidence was relatively stable (ranging from a few hundreds to thousands of new confirmed cases per day), and NPIs were strictly executed across the nation. The only novel factor consistently introduced to the public during this period was the SARS-CoV-2 vaccine: starting from February 2021, vaccines from four different manufacturers were approved in South Korea: ChAdOx1-S/nCoV-19 (Oxford-Astrazeneca), Ad.26. COV2.S (Janssen), BNT162b2 (Pfizer-BioNTech) and mRNA-1273 (Moderna) vaccines. Most studies have concluded that, in the two recent decades, there has been no or only minutely increased risk of GBS following influenza vaccination.26–28 A potential link between GBS and vaccination against measles, mumps, rubella (MMR), hepatitis B, diphtheria, tetanus, pertussis (DTP) and polio vaccines was also suggested, but the strength of association has been negligible.29–31 The only case in which a strong association was supported by evidence was swine influenza vaccination in New Jersey in 1976.8 During the pandemic period, large population-based, database-driven studies have confirmed an increased risk of GBS with viral vector-based SARS-CoV-2 vaccines.32 33 We hypothesised that the relative increase in GBS incidence during this period may have been temporarily associated with a mass vaccine rollout (r=0.55, p=0.006). Notably, older age groups were injected primarily with viral vector-based vaccines in the early vaccine rollout in South Korea, and the increased GBS incidence from June to September 2021 coincided with the increase in vaccination in the 60–69 and 70–79 year age groups (online supplemental figure 1). Nevertheless, this hypothesis needs to be validated in future studies and should be interpreted with caution, as the benefit of the SARS-CoV-2 vaccine outweighs the risk of rare complications. We are currently investigating the contribution of the vaccine mechanism, dose and homogeneity of serial SARS-CoV-2 vaccinations to GBS incidence in 2021.
Our study had several limitations. First, this was a cross-sectional study, so only an association can be suggested. However, gastrointestinal and respiratory tract infections account for a large portion of the total infectious disease spectrum and are the widely accepted triggers of GBS.34 Therefore, the temporal association observed in our study may not be just be a coincidental finding. Second, in seeking the fluctuation in the incidence of infectious diseases, we only used sentinel surveillance data collected from predesignated hospitals mandated to report specified pathogens. Nevertheless, we believe that pathogen-specific data collection and analysis has an advantage over diagnosis code-based data collection and analysis in terms of reliance and clarity, as some diagnoses allegedly related to infection may also be due to non-infectious causes (eg, acute pharyngitis and gastroenteritis of unspecified origin). Furthermore, the use of specific pathogens considers the seasonality of each pathogen, which is inevitably neglected in the diagnosis-oriented approach. Third, our correlation analysis was based on the hypothesis that most mucosa-associated infections can potentially cause GBS. There is a varying degree of evidence for each listed infectious disease having an association with GBS, some with more extensive research than others. However, we also hoped to ensure that the potential pathogens insufficiently covered by the previous literature were not overlooked. Lastly, no human behavioural factors were considered; reluctance to visit the hospital in mild cases of these infections may also be a behavioural factor contributing to under-reporting bias.
In conclusion, the COVID-19 pandemic continues to exert influence on socioeconomic and epidemiological grounds, thereby impacting public health. Remarkably, the pandemic is responsible for the changes in the epidemiology of traditional infectious immune triggers (ie, non-COVID-19 infections) and the introduction of novel immune stimuli (ie, SARS-CoV-2 vaccines) to the general public. Henceforth, it would be prudent to remain vigilant and carefully monitor the evolution of autoimmune phenomena in the long term. GBS is one of many rare examples of these conditions, and the same analogy may be implicated in diseases that share a similar molecular mimicry-based pathophysiology. We believe that tracking down changes in these potential immune triggers and evaluating the temporal association of these attributable risk factors with disease incidence can be used as supportive evidence of coherence in causation analysis and can improve the accuracy of expectations of future epidemiological outcomes.
