Seasonality in Adverse Drug Events: Time-Series Analysis of JADER Using ARIMA/SARIMA and Prophet

Seasonality in Adverse Drug Events: Time-Series Analysis of JADER Using ARIMA/SARIMA and Prophet | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Seasonality in Adverse Drug Events: Time-Series Analysis of JADER Using ARIMA/SARIMA and Prophet Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, Kohei Shiota, Moe Yamashita, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6746128/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract The Japanese Adverse Drug Event Report (JADER) database is a spontaneous reporting system that compiles real-world data on adverse events (AEs) associated with drug use in Japan, including date of occurrence, allowing investigation of temporal trends. This study aimed to identify AEs that exhibit seasonal variation by applying time-series models—the autoregressive integrated moving average (ARIMA), seasonal ARIMA (SARIMA), and Prophet—to monthly AE reports from January 2005 to December 2019. Models were optimized using grid search and time-series cross-validation, and model performance was evaluated using mean absolute error, root mean squared error, and mean absolute percentage error. AEs whose seasonal models outperformed non-seasonal ones were considered likely seasonal. Sixteen AEs, including influenza, Guillain-Barré syndrome, encephalitis, and adverse reactions related to antineoplastic agents or immune checkpoint inhibitors, were identified as exhibiting potential seasonality. Some AEs showed clear peaks in specific months, while others demonstrated discrepancies between observed data and modeled seasonal components. These results indicate that certain AEs may be influenced by seasonal factors such as infectious diseases, climate, or patient behaviors. Identifying these seasonal trends can support proactive risk management, inform medication safety strategies, and assist clinicians, regulators, and researchers in strengthening pharmacovigilance and improving patient outcomes. Health sciences/Health occupations Health sciences/Medical research Seasons Adverse Drug Event Adverse Drug Reaction Reporting System Vaccine ARIMA Prophet Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Spontaneous reporting systems (SRS) play a vital role in pharmacovigilance by collecting real-world data on the adverse events (AEs) associated with medicinal products. In Japan, such data are compiled in the Japanese Adverse Drug Event Report (JADER) database managed by the Pharmaceuticals and Medical Devices Agency (PMDA). This database provides the AE occurrence dates, thus enabling researchers to explore potential temporal trends. Prior studies using JADER and other SRS databases have observed fluctuations in AEs, including dehydration and drug-induced photosensitivity, suggesting possible seasonal factors—whether climatic or behavioral. 1 – 3 These results underscore the importance of examining when AEs spike to better understand and mitigate risks. Although temporal information is crucial, many investigations have relied on monthly AE aggregations without employing advanced time-series methods. This approach can obscure genuine seasonal patterns and fails to adequately account for confounders such as overreporting, underreporting, and missing data. By adopting models that capture more nuanced temporal dynamics, researchers can discover accurate seasonal trends and refine risk management strategies. Statistical tools such as the autoregressive integrated moving average (ARIMA) and its seasonal extension, seasonal ARIMA (SARIMA), are widely used in domains ranging from economics to epidemiology. 4 – 7 ARIMA breaks down time-series data into autoregressive and moving average components, with differencing to handle nonstationarity. SARIMA adds parameters specifically for recurring seasonal fluctuations, making it more suitable when the data exhibit repeated cycles, such as monthly surges in AEs. Facebook introduced a newer approach, the Prophet model. 8 , 9 This model treats a time series as a sum of trend, seasonal, and external event effects and is designed to handle outliers, abrupt changes, and missing values. It also allows the inclusion of domain-specific factors such as holidays or other known events, which can refine forecasts that predict possible shifts in reporting rates particular periods (e.g., influenza season or significant public health announcements). The principal aim of this study was to determine whether the AEs reported in the JADER database display meaningful seasonal patterns by applying the SARIMA and Prophet models. By comparing these methods, we sought to clarify the temporal distribution of AEs and identify the periods of elevated risk. In doing so, we obtained insights that will help clinicians, regulators, and researchers implement more proactive and targeted pharmacovigilance in real-world settings. Materials and Methods Data Source Data compiled from April 2004 to September 2024 in the JADER database were downloaded from the PMDA and Japanese Regulatory Authority websites. 10 All the JADER data have been cleaned and anonymized by the PMDA. In Japan, in the absence of a dedicated vaccine-related AE database, such as Vaccine Adverse Event Reporting System (VAERS), AEs related to vaccines are reported within the JADER database. The database consists of four data tables: 1) patient demographic information (DEMO), 2) drug information (DRUG), 3) AEs (REAC), and 4) primary illness (HIST). Only reports of AE occurrence with dates recorded correctly in the REAC table were analyzed in the present study. AEs with a total count of 100 or more were selected across all months, resulting in a dataset comprising monthly AE counts over a 15-year period from January 2005 to December 2019. Definition of Adverse Events The AEs in the JADER database were defined based on the Medical Dictionary for Regulatory Activities version 27.0. 11 ARIMA and SARIMA Models To examine seasonality in AE reports, we employed both the ARIMA and SARIMA models. The ARIMA model is defined using three parameters: autoregressive order (p), degree of differencing (d), and moving average order (q). Differencing stabilizes the mean of a time series by removing level shifts, thus enabling the modeling of non-stationary data while preserving its linear structure. 4 , 5 ARIMA operates by modeling dependencies between observations in a sequence, making it effective for datasets without strong seasonal components. However, it lacks the ability to handle periodic fluctuations explicitly, which limits its application for data with significant seasonality. SARIMA extends ARIMA by adding the seasonal parameters P, D, Q, and the length of the seasonal cycle s (s = 12 for monthly data), thus allowing us to capture seasonal effects. We determined the optimal ARIMA and SARIMA parameters for each AE using a grid search over p, d, q (0–2) and P, D, Q (0–2), with s = 12. 12 The best parameter combination was selected based on the Akaike Information Criterion. 12 To evaluate the predictive performance, we used five-fold time series split cross-validation while preserving the chronological order to avoid data leakage. 5 , 13 The mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) were calculated. 6 , 14 , 15 For each AE, if the MAE of the SARIMA model was lower than that of the ARIMA model, the AE was considered potentially seasonal. The analyses were conducted using Python and the Statsmodels library for ARIMA and SARIMA modeling. 16 Prophet model Prophet is an open-source software released by Facebook’s Core Data Science team. It works best with time series that display strong seasonal effects and have several seasons of historical data. Prophet is robust to missing data and shifts in trends, and typically handles outliers well. 8 , 9 The strength of the ARIMA model lies in its ability to capture linear dependencies in time-series data. 4 , 5 However, data with complex nonlinear relationships or structural changes falls outside the scope of what a pure ARIMA model can effectively handle. 5 In contrast, Prophet models compensate for this limitation. 5 , 17 Therefore, we constructed both SARIMA and Prophet models. 6 , 7 We employed two variants of Facebook’s Prophet model—one allowing for yearly seasonality (“seasonal Prophet”) and one excluding it (“non-seasonal Prophet”)—to investigate potential seasonal patterns. Seasonal model: yearly_seasonality = True Non-seasonal model: yearly_seasonality = False In both models, the default parameters were utilized with no additional regressors or custom seasonalities. We performed time-series cross-validation using Prophet’s cross_validation function, specifying an initial training period of 1095 days (three years), a period of 365 days (one year), and a horizon of 365 days (one year). We calculated the MAE, RMSE, and MAPE for both models. For each AE, if the seasonal model had a lower MAE than the non-seasonal model, it was identified as potentially exhibiting seasonality. The analyses were performed using Python. 17 Results Data spanning a 15-year period from January 2005 to December 2019 were downloaded from the JADER database and analyzed. The date data were preprocessed, and 390,036 reports were subsequently analyzed, including 727 AEs (Fig. 1 ). The implementation of the seasonal and non-seasonal Prophet models revealed that the MAE of the seasonal model was lower than that of the non-seasonal model for 75 AEs (10.3%) (Supplementary Table S1 ). The implementation of the ARIMA and SARIMA models revealed that the MAE of SARIMA was lower than that of ARIMA for 52 AEs (7.2%) (Supplementary Table S2 ). The seasonal models demonstrated superior performance in both the ARIMA/SARIMA and Prophet analyses for 16 AEs (2.2%), including acute disseminated encephalomyelitis, appetite loss, aspiration, atypical femur fracture, encephalitis, Guillain-Barré syndrome (GBS), hypophysitis, immune-mediated enterocolitis, influenza, intestinal ischemia, lacunar infarction, malaise, paronychia, secondary adrenal insufficiency, transfusion-associated circulatory overload, and type 1 diabetes mellitus. The 16 AEs were forecasted using the Prophet model (Fig. 2 ) and subsequently analyzed with yearly seasonality components (Fig. 3 ). A comparison of the monthly peaks and troughs between the 15-year aggregate report counts and Prophet-derived seasonal components revealed exact temporal alignment only for lacunar infarction (Fig. 4 , Table 1). Seasonal patterns were concordant for appetite loss, influenza, intestinal ischemia, and paronychia. However, a subset of AEs, including aspiration, encephalitis, GBS, hypophysitis, immune-mediated enterocolitis, secondary adrenal insufficiency, and transfusion-associated circulatory overload, demonstrated no concordance between the highest and maximum months or between the lowest and minimum months. Discussion Seasonal Characteristics and Overview Japan has four distinct seasons, and each season exerts unique physiological effects. Tokyo’s meteorological data (2015–2019) show mean monthly temperatures from 5.6°C (January) to 27.3°C (August), 18 and relative humidity ranging from 53.0% (January) to 81.8% (September) (Supplementary Figure S1 ). 19 These seasonal factors and patient behavior may contribute to the seasonality of AE reports. Among the 727 AEs evaluated, only 52 (7.2%) appeared to have a seasonal component and only 16 (2.2%) demonstrated seasonality in both ARIMA/SARIMA and Prophet models. This result means that most AEs do not have a seasonal component—a finding that is not entirely unexpected, and one which reinforces the notion that AEs exhibiting seasonal characteristics are relatively uncommon. According to the results of our analysis using both the Prophet and ARIMA/SARIMA models, 16 AEs displayed seasonality potentially linked to infectious diseases, climatic factors, and patient behavior. The drugs most commonly associated with these AEs were vaccines and antineoplastic agents (Supplementary Table S3 ). Infectious Diseases, Vaccines, and Autoimmune Conditions Among the 16 AEs, influenza, GBS, acute disseminated encephalomyelitis, and encephalitis were frequently associated with influenza vaccines according to the JADER database, peaking from November to January. Although these AEs may be vaccine-related, they may also reflect concurrent infectious triggers. The influenza vaccines approved in Japan are inactivated; therefore, reports of post-vaccination influenza may involve indirect factors. Epidemiological evidence indicates that the older people are more commonly affected (peak incidence of GBS is between 50 and 70 years of age). 20 GBS risk following vaccination is only slightly elevated, 20 – 24 with approximately 1–1.6 cases per million H1N1 vaccinations. 20 Immune checkpoint inhibitors (ICIs) are also suspected drugs for GBS, and ICI-induced cases may differ mechanistically from classical GBS. 23 Seasonal GBS spikes vary internationally and may be linked to prevalent pathogens, including Campylobacter jejuni and influenza A virus. 24 – 27 Acute disseminated encephalomyelitis is a rare demyelinating disease that mostly affects children, with an incidence of 0.2–0.5 per 100,000 individuals. 28 , 29 The average age of onset in pediatric cases is typically between 3 and 7 years, 28 , 30 although the condition can manifest at any age. Several studies have reported a slight male predominance. 28 , 30 A preceding infection has been reported in approximately 75% of patients examined in previous studies 28 , 31 ; although vaccination-associated acute disseminated encephalomyelitis has also been noted, albeit without definitive causal evidence. 28 In JADER, most acute disseminated encephalomyelitis reports involved vaccines, whereas encephalitis was reported with both vaccines and ICIs. Autoimmune encephalitis following vaccination may be mediated by molecular mimicry or immune activation. 32 – 35 The mechanisms of ICI-related encephalitis include the generation of autoantibodies (e.g., anti-N-methyl-D-aspartate receptors). 36 In Asia, the Japanese encephalitis virus remains the leading cause of encephalitis, typically occurring from June to October, whereas in Western countries, the herpes simplex virus predominates as the cause of encephalitis. 37 , 38 Non-infectious AEs and Seasonal Influences Several non-infectious AEs such as aspiration, appetite loss, malaise, atypical femur fracture, lacunar infarction, intestinal ischemia, and paronychia also showed seasonality. In the JADER database, aspiration often involves antipsychotic drugs, possibly due to impaired swallowing. Seasonal affective disorder, a subtype of major depression that occurs from autumn to winter, 39 , 40 might influence dietary habits and mood, further influencing aspiration risk. Appetite loss and malaise present similarly but have been reported with different suspected drugs in JADER. The most suspected drugs for appetite loss include antineoplastic agents, ribavirin, and interferon. For malaise, antineoplastic agents, osteoporosis medications, pregabalin, and valacyclovir are the predominant suspected drugs. Appetite loss peaked in late spring and summer (May–August) and declined in winter (December–February). High temperatures may induce summer heat fatigue, which encompasses fatigue, anorexia, and autonomic dysfunction. 41 – 43 Antineoplastic agents themselves disrupt thermoregulation and contribute to heat- or cold-related health issues. 44 , 45 Seasonal fluctuations in energy intake and stress responses have also been documented, which potentially modulate these AEs. 46 While fractures generally increase during winter, 47 , 48 atypical femur fractures in corticosteroid users peak in April rather than in the colder months, suggesting the possibility of underreporting falls in winter as AEs. Lacunar infarctions appear to be influenced by dehydration from sodium-glucose cotransporter-2 inhibitors, as higher blood viscosity due to volume depletion increases stroke risk. 49 , 50 The incidence of lacunar infarctions peaks in June and September rather than in midsummer, which may indicate that unrecognized summer cases are attributable to ambient heat rather than medication. In addition to dehydration, the risk factors for lacunar infarction include hypertension, diabetes, and hyperlipidemia. 51 Regarding intestinal ischemia and paronychia, the JADER patterns diverge from current epidemiological data. In the United States, acute vascular injury of the intestine peaks in late summer, with more severe outcomes in winter. 52 There is limited literature on paronychia seasonality, 53 although humidity and temperature can favor nail infections. According to the JADER database, both conditions are associated with the use of antineoplastic agents. Paronychia reporting is likely tied to the duration of drug use rather than to temperature shifts; however, a seasonal distribution emerged that may reflect the underlying reporting trends. ICI-induced AEs and Seasonal Factors The seasonality of ICI-induced events, including immune-mediated enterocolitis, hypophysitis, secondary adrenal insufficiency, and type 1 diabetes mellitus, could stem from various mechanisms. Proposed influences include vitamin D fluctuations and photoperiodic changes in endocrine or immune function. 54 – 57 The composition of the gut microbiota, which varies seasonally, may also play a role in enterocolitis. Although direct evidence is limited, these hypotheses highlight the potential of climate and daylight patterns to modulate immune-related AEs. Transfusion-associated Circulatory Overload Transfusion-associated circulatory overload is a transfusion reaction that is especially serious in patients with cardiac comorbidities. 58 Although hospitalizations for heart failure often peak in January, with higher mortality from August to October, 59 the specific factors contributing to the seasonality of transfusion-associated circulatory overload remain unclear. Further research is required to elucidate any direct links. Limitations and Model Considerations This study has inherent limitations related to SRS, such as overreporting, underreporting, missing data, exclusion of healthy subjects resulting in lack of a control population or reference group, and the presence of confounding factors. External factors, such as safety communications and market dynamics, may also influence reporting patterns. The divergence between Prophet forecasts and raw counts often stems from smoothing outliers, short-term fluctuations, and fundamental methodological differences. For AEs such as hypophysitis, immune-mediated enterocolitis, and secondary adrenal insufficiency, the steadily increasing reports of ICIs following the approval of nivolumab in Japan in 2014 make it challenging to separate true seasonal signals from upward reporting trends. Prophet’s flexibility may capture such long-term patterns better by dealing robustly with missing data, outliers, and abrupt changes, thus offering valuable insights into seasonal AE distributions recorded in the JADER database. Moreover, the determination of the seasonality of a time series is highly dependent on the method selected, even when applied to the same database. 60 The ARIMA model assumes that all predictable structures can be captured by trend and autocorrelation terms. When seasonal cycles are subtle or weak, ARIMA may treat them as white noise. In contrast, Prophet offers high model flexibility through additive components, which can lead to overfitting by capturing noise or incidental patterns as meaningful structures in the data. The present study revealed that infectious diseases, AEs related to vaccines, and climatic factors contribute to the seasonal patterns observed in the JADER database. Future research using more granular data, including detailed patient demographics, clinical conditions, and environmental factors, are required to uncover the mechanisms underlying seasonal trends. Conclusion By applying both the ARIMA/SARIMA and Prophet models to the JADER database, we identified several AEs with potential seasonal patterns, many of which were associated with infectious diseases, climate factors, and medication usage, particularly vaccinations. A more nuanced understanding of these seasonal dynamics can bolster risk-management strategies, inform medication safety protocols, and ultimately enhance patient care. This study also demonstrates the utility of combining advanced time-series approaches with domain-specific knowledge, thereby offering a robust framework for professionals seeking reliable predictions of AE occurrence. Anticipating seasonal surges in AE reporting may contribute to improved risk communication, timely clinical interventions, and efficient allocation of healthcare resources. Our findings reinforce the importance of recognizing and analyzing seasonality in pharmacovigilance data and highlight how tailored forecasting methods can capture both long- and short-term trends. By extending these techniques to broader contexts, we hope to guide medical professionals, data analysts, researchers, and clinical application developers in selecting appropriate models for accurate AE forecasting, thus contributing to a more comprehensive understanding of AE seasonality and proactive risk management. Declarations Acknowledgements The authors (s) disclose receipt of the following financial support for this article’s research, authorship, and/or publication: This research was partially supported by the Japan Society for the Promotion of Science KAKENHI grant numbers 25K10049 and 21K06646. The funders had no role in the study design, data collection, analysis, decision to publish, or preparation of this article. Ethics approval Ethical approval was not sought for this study because the study was a database-related observational study which did not directly involve any research subjects. All results were obtained from data openly available online from the PMDA website (www.pmda.go.jp). All data from the JADER database were fully anonymized by the relevant regulatory authority before we accessed them. Our research does not fall within the purview of any of the following laws and guidelines: “Clinical Trials Act (Act No. 16 of April 14, 2017),” “Act on Securing Quality, Efficacy and Safety of Products Including Pharmaceuticals and Medical Devices (Law number: Act No. 145 of 1960, Last Version: Amendment of Act No. 50 of 2015),” “Guideline for good clinical practice E6 (R1), https://www.pmda.go.jp/int-activities/int-harmony/ich/0076.html,” “Ethical guidelines for human genome and gene analysis research, https://www.mhlw.go.jp/general/seido/kousei/i-kenkyu/genome/0504sisin.html,” and “Ethical Guidelines for Medical and Health Research Involving Human Subjects, https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/hokabunya/kenkyujigyou/i-kenkyu/index.html.” Therefore, it is not subject to ethical examination. The study was an observational study without any research subjects. No consent to participate was required due to the retrospective nature of this study. Author Contributions All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Concept and design: Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, and Mitsuhiro Nakamura. Acquisition, analysis, or interpretation of data: Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, and Mitsuhiro Nakamura. Drafting of the manuscript: Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, Kohei Shiota, Moe Yamashita, Nanaka Ichihara, Yuka Nokura, Tomofumi Yamazaki, and Kana Sugishita. Critical review of the manuscript for important intellectual content: Yoko Ino, Kazuhiro Iguchi and Jun Liao. Data Availability Statements All results were obtained from data openly available online from the PMDA website (www.pmda.go.jp). The datasets used and analyzed in the current study are available from the corresponding author upon reasonable request. Competing Interests Statement The authors declare no potential conflicts of interest with respect to the research, authorship, or publication of this article. Funding : This research was partially supported by the Japan Society for the Promotion of Science KAKENHI grant numbers 25K10049 and 21K06646. References Matsumoto, K. et al. A retrospective study of seasonal variation in sodium-glucose co-transporter 2 inhibitor-related adverse events using the Japanese adverse drug event report database. Sci. Rep . 14 , 30072 (2024). Nakao, S. et al. Evaluation of drug-induced photosensitivity using the Japanese Adverse Drug Event Report (JADER) database. Biol. Pharm. Bull . 40 ,2158–2165 (2017). Marrero, O., Hung, E. Y. & Hauben, M. Seasonal and geographic variation in adverse event reporting. Drugs-Real World Outcomes 3 , 297–306 (2016). Box, G. E. P., Jenkins, G. M., Reinsel, G. C. & Ljung, G. M. Time series analysis: forecasting and control. 5th edn, (John Wiley & Sons, 2015). Hyndman, R. J. & Athanasopoulos, G. Forecasting: principles and practice. 3rd edn, (OTexts, 2021) https://otexts.com/fpp3/. Akshay, K.C., Grace, G.H., Gunasekaran, K. & Samikannu, R. Power consumption prediction for electric vehicle charging stations and forecasting income. Sci. Rep. 14 , 6497 (2024). Paik, J., & Ko, K. Seasonal analysis and risk management strategies for credit guarantee funds: a case study from republic of Korea. Stats . 8 , 2 (2025). Taylor, S. J. & Letham, B. Forecasting at scale. Am. Stat. 72 , 37–45 (2018). Facebook. Prophet: Forecasting at Scale. https://facebook.github.io/prophet/ (2017). Pharmaceuticals and Medical Devices Agency (PMDA). Suspected adverse drug reaction information. (in Japanese). https://www.pmda.go.jp/safety/info-services/drugs/adr-info/suspected-adr/0003.html, cited 16 May, 2025. Medical Dictionary for Regulatory Activities. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). https://www.ich.org/page/meddra, cited 16 May, 2025. Kontopoulou, V. I., Panagopoulos, A. D., Kakkos, I. & Matsopoulos, G. K. A review of ARIMA vs. machine learning approaches for time series forecasting in data driven networks. Future Internet , 15 , 255, (2023). Bergmeir, C. & Benítez, J. M. On the use of cross-validation for time series predictor evaluation. Inf. Sci. 191 , 192–213 (2012). Hodson, T. O. Root-mean-square error (RMSE) or mean absolute error (MAE): when to use them or not. Geosci. Model Dev. 15 , 5481–5487 (2022). Kwarteng, S. B. & Andreevich, P. A. Comparative analysis of ARIMA, SARIMA and Prophet model in forecasting. Res. Dev. 5 , 110–120 (2024). Statsmodels. Statsmodels.tsa.statespace.sarimax.SARIMAX. https://www.statsmodels.org/stable/generated/statsmodels.tsa.statespace.sarimax.SARIMAX.html, cited 16 May, 2025. Prophet. Prophet: Quick Start. https://facebook.github.io/prophet/docs/quick_start.html, cited 16 May, 2025. Japan Meteorological Agency. Monthly average temperature for Tokyo (in Japanese). https://www.data.jma.go.jp/obd/stats/etrn/view/monthly_s3.php?prec_no=44&block_no=47662, cited 16 May, 2025. Japan Meteorological Agency. Monthly average relative humidity in Tokyo (in Japanese). https://www.data.jma.go.jp/obd/stats/etrn/view/monthly_s3.php?prec_no=44&block_no=47662&year=&month=&day=&view=a7, cited 16 May, 2025. Bellanti, R. & Rinaldi, S. Guillain–Barré syndrome: a comprehensive review. Eur. J. Neurol. 31 , e16365 (2024). van Doorn, P. A. et al. European academy of neurology/peripheral nerve society guideline on diagnosis and treatment of Guillain–Barré syndrome. J. Peripher. Nerv. Syst. 28 , 535–563 (2023). Salmon, D. A., Dudley, M. Z. & Carleton, B. C. Guillain–Barré syndrome following influenza vaccines affords opportunity to improve vaccine confidence. J. Infect. Dis. 223 , 355–358 (2021). Ding, M. et al. Case report: ICIs-induced Guillain–Barré syndrome recovered from mycophenolate mofetil. Front. Immunol. 14 , 1132692 (2023). Zhai, Q. et al. Guillain–Barré syndrome in northern China: a retrospective analysis of 294 patients from 2015 to 2020. J. Clin. Med. 11 , 6323 (2022). Ginanneschi, F. et al. Clinical features and outcome of the Guillain–Barré syndrome: a single-center 11-year experience. Front. Neurol. 13 , 856091 (2022). Chroni, E. et al. Guillain–Barré syndrome in Greece: seasonality and other clinico-epidemiological features. Eur. J. Neurol. 11 , 383–388 (2004). Finsterer, J. Triggers of Guillain–Barré syndrome: Campylobacter jejuni predominates. Int. J. Mol. Sci. 23 , 14222 (2022). Wolska-Krawczyk, M. Akute disseminierte Enzephalomyelitis [Acute disseminated encephalomyelitis]. Radiologe 62 , 316–321 (2022). (in German) Leake, J. A. D. et al. Acute disseminated encephalomyelitis in childhood: epidemiologic, clinical and laboratory features. Pediatr. Infect. Dis. J. 23 , 756–764 (2004). Cole, J., Evans, E., Mwangi, M. & Mar, S. Acute disseminated encephalomyelitis in children: an updated review based on current diagnostic criteria. Pediatr. Neurol. 100 , 26–34 (2019). Pavone, P. et al. Acute disseminated encephalomyelitis: a long-term prospective study and meta-analysis. Neuropediatrics 41 , 246–255 (2010). Guedes, B. F. et al. Potential autoimmune encephalitis following yellow fever vaccination: a report of three cases. J. Neuroimmunol. 355 , 577548 (2021). Zuhorn, F., Graf, T., Klingebiel, R., Schäbitz, W.-R. & Rogalewski, A. Postvaccinal encephalitis after ChAdOx1 nCov-19. Ann. Neurol. 90 , 506–511 (2021). Wang, H. Anti-NMDA receptor encephalitis and vaccination. Int. J. Mol. Sci. 18 , 193 (2017). Martin, S., Azzouz, B., Morel, A. & Trenque, T. Anti-NMDA receptor encephalitis and vaccination: a disproportionality analysis. Front. Pharmacol. 13 , 940780 (2022). Buckley, M. W. et al. Immune-related encephalitis after immune checkpoint inhibitor therapy. Oncologist 30 , oyae186 (2025). Kumar, R. Understanding and managing acute encephalitis. F1000Res. 9 , 60 (2020). Centers for Disease Control and Prevention (CDC). Japanese encephalitis: data and maps. https://www.cdc.gov/japanese-encephalitis/data-maps/index.html, cited 16 May, 2025. Saeed, S. A. & Bruce, T. J. Seasonal affective disorders. Am. Fam. Physician 57 , 1340–1346 (1998). Donofry, S.D., Roecklein, K.A., Rohan, K.J., Wildes, J.E. & Kamarck, M.L. Prevalence and correlates of binge eating in seasonal affective disorder. Psychiatry Res. 217 , 47–53 (2014). Ebi, K. L. et al. Hot weather and heat extremes: health risks. Lancet 398 , 698–708 (2021). Greaney, J. L., Kenney, W. L. & Alexander, L. M. Sympathetic regulation during thermal stress in human aging and disease. Auton. Neurosci. 196 , 81–90 (2016). Makino, S. et al. Anti-fatigue effects of yogurt fermented with Lactobacillus delbrueckii subsp. bulgaricus OLL1073R-1 in healthy people suffering from summer heat fatigue: a randomized, double-blind, placebo-controlled trial. Nutrients 10 , 798 (2018). Shao, Q. et al. Tumor therapeutic response monitored by telemetric temperature sensing, a preclinical study on immunotherapy and chemotherapy. Sci. Rep. 13 , 7727 (2023). Kokolus, K. M., Hong, C. C. & Repasky, E. A. Feeling too hot or cold after breast cancer: is it just a nuisance or a potentially important prognostic factor? Int. J. Hyperthermia 26 , 662–680 (2010). Fujihira, K., Takahashi, M., Wang, C. & Hayashi, N. Factors explaining seasonal variation in energy intake: a review. Front. Nutr. 10 , 1192223 (2023). Bulajic-Kopjar, M. Seasonal variations in incidence of fractures among elderly people. Inj. Prev. 6 , 16–19 (2000). Bischoff-Ferrari, H. A., Orav, J. E., Barrett, J. A. & Baron, J. A. Effect of seasonality and weather on fracture risk in individuals 65 years and older. Osteoporos. Int. 18 , 1225–1233 (2007). Cortés-Vicente, E. et al. Frequency, risk factors, and prognosis of dehydration in acute stroke. Front. Neurol. 10 , 305 (2019). Song, S. H. et al. Elevated blood viscosity is associated with cerebral small vessel disease in patients with acute ischemic stroke. BMC Neurol. 17 , 20 (2017). Yaghi, S. et al. Lacunar stroke: mechanisms and therapeutic implications. J. Neurol. Neurosurg. Psychiatry 92 , 823–830 (2021). Jehangir, A., Dhital, R., Qureshi, A. & York, E.P. Acute vascular insufficiency of intestine: incidence highest in summer, outcomes worst in winter. Int. J. Colorectal Dis. 34 , 2059–2067 (2019). Khodaei, B., Seyedpour, S., Gholami, B., Garmarudi, G. & Nasimi, M. Seasonal and gender variation in skin disease: a cross-sectional study of 3120 patients at Razi hospital. Int. J. Womens Dermatol. 7 , 799–802 (2021). Knip, M. et al. Environmental triggers and determinants of type 1 diabetes. Diabetes 54 , S125–S136 (2005). Maboshe, W. et al. Low-dose vitamin D₃ supplementation does not affect natural regulatory T cell population but attenuates seasonal changes in T cell-produced IFN-γ: results from the D-SIRe2 randomized controlled trial. Front. Immunol. 12 , 623087 (2021). Grover, S. et al . Vitamin D intake is associated with decreased risk of immune checkpoint inhibitor-induced colitis. Cancer 126 , 3758–3767 (2020). Walker, W. H. II, Walton, J. C., DeVries, A. C. & Nelson, R. J. Circadian rhythm disruption and mental health. Transl. Psychiatry 10 , 28 (2020). Semple, J. W., Rebetz, J. & Kapur, R. Transfusion-associated circulatory overload and transfusion-related acute lung injury. Blood 133 , 1840–1853 (2019). Matsuda, H., Kuragaichi, T. & Sato, Y. Investigating the seasonal variation of heart failure hospitalizations and in-hospital mortality risks in Japan using a nationwide database. J. Cardiol. 83 , 236–242 (2024). Molinaro, A. & DeFalco, F. Empirical assessment of alternative methods for identifying seasonality in observational healthcare data. BMC Med. Res. Methodol. 22 , 182 (2022). Table 1 Table 1 is available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files 20250210Table1.xlsx Table 1 Comparison of monthly seasonal values derived from Prophet analysis and maximum/minimum aggregate counts in reported adverse events. 20250519SupplementaryTableS1.xlsx Supplementary Table S1 Comparative evaluation of MAE, RMSE, and MAPE between seasonal and non-seasonal Prophet models. 20250519SupplementaryTableS2.xlsx Supplementary Table S2 Comparative evaluation of the MAE, RMSE, and MAPE between SARIMA and ARIMA models. 20250519SupplementaryTableS3a.xlsx Supplementary Table S3a Top ten suspected drugs reported for each adverse event. 20250519SupplementaryTableS3b.xlsx Supplementary Table S3b (continued) Top ten suspected drugs reported for each adverse event. 20250519SupplementaryFigureS1.pptx Supplementary Figure S1 Average monthly temperature and relative humidity from 2015 to 2019. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 10 Apr, 2026 Reviews received at journal 03 Nov, 2025 Reviewers agreed at journal 13 Oct, 2025 Reviewers invited by journal 10 Oct, 2025 Editor assigned by journal 07 Oct, 2025 Editor invited by journal 03 Jun, 2025 Submission checks completed at journal 29 May, 2025 First submitted to journal 29 May, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6746128","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":532322896,"identity":"76da4c07-a16d-47e9-b475-7907df8022bb","order_by":0,"name":"Hideyuki Tanaka","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Hideyuki","middleName":"","lastName":"Tanaka","suffix":""},{"id":532322897,"identity":"4077180f-8db5-4cd8-8c29-bdd1d68268df","order_by":1,"name":"Mika Maezawa","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Mika","middleName":"","lastName":"Maezawa","suffix":""},{"id":532322898,"identity":"812c3404-180f-4a1b-b6d4-4186b17f3e89","order_by":2,"name":"Satoshi Nakao","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Satoshi","middleName":"","lastName":"Nakao","suffix":""},{"id":532322899,"identity":"82ac0d19-11fa-41e8-8502-08e36407da65","order_by":3,"name":"Kohei Shiota","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Kohei","middleName":"","lastName":"Shiota","suffix":""},{"id":532322900,"identity":"d8507c9f-8acb-4e32-966f-4d6ed4862d06","order_by":4,"name":"Moe Yamashita","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Moe","middleName":"","lastName":"Yamashita","suffix":""},{"id":532322901,"identity":"1534f48f-57e3-4c3b-900f-c21ba4181bd9","order_by":5,"name":"Nanaka Ichihara","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Nanaka","middleName":"","lastName":"Ichihara","suffix":""},{"id":532322902,"identity":"ddc51184-82a6-4dd7-a89a-138d4dc66d9b","order_by":6,"name":"Yuka Nokura","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Yuka","middleName":"","lastName":"Nokura","suffix":""},{"id":532322903,"identity":"42f864eb-95f3-44fd-8474-e1deefbefd61","order_by":7,"name":"Tomofumi Yamazaki","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Tomofumi","middleName":"","lastName":"Yamazaki","suffix":""},{"id":532322904,"identity":"f00cb6df-ac17-428c-827a-a7702fb7aec9","order_by":8,"name":"Kana Sugishita","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Kana","middleName":"","lastName":"Sugishita","suffix":""},{"id":532322905,"identity":"1bbab96c-eb16-47c8-b6fe-29b6f4b911e5","order_by":9,"name":"Yoko Ino","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Yoko","middleName":"","lastName":"Ino","suffix":""},{"id":532322906,"identity":"cbea60b5-1ef2-4847-85b6-156906d49a7f","order_by":10,"name":"Kazuhiro Iguchi","email":"","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Kazuhiro","middleName":"","lastName":"Iguchi","suffix":""},{"id":532322907,"identity":"a5f36cae-32f6-4418-825e-44fc7f751621","order_by":11,"name":"Jun Liao","email":"","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Jun","middleName":"","lastName":"Liao","suffix":""},{"id":532322908,"identity":"1933d2d8-5ba5-4b15-add4-3781360d71d1","order_by":12,"name":"Mitsuhiro Nakamura","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+ElEQVRIiWNgGAWjYHACZgjF3sBwoAIuyEOMFp4DDAfOkKZFIoGB4QxehVCgO7v5scHPnMP5/DOfXzxwsI0hz+AA88MPDDJ3cGoxu3PMOLF322HLGbdzCkBaig0OsBlLMPA8w63lRoLxAd5thw0YbuckHP7Y9j9xwwEGM6BfDuPRkv754F+gFvmbZxJAtgC1sH8joCXHOBlki8EN9gNQLTyEbMkpNpbdlm5geCaH4cCBcwzFkod5iiUS8PolfbPk223WBnLHjz/+cKCMIY/vePvGDx97cIcYEuAxAJEJ4HhK7DlAjBb2BxAtYPCDKC2jYBSMglEwMgAADAthGNHGNJwAAAAASUVORK5CYII=","orcid":"","institution":"Gifu Pharmaceutical University","correspondingAuthor":true,"prefix":"","firstName":"Mitsuhiro","middleName":"","lastName":"Nakamura","suffix":""}],"badges":[],"createdAt":"2025-05-26 01:23:37","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6746128/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6746128/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94398888,"identity":"ccd0324d-bb75-4056-aa01-60f907f582ea","added_by":"auto","created_at":"2025-10-27 13:57:14","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":255404,"visible":true,"origin":"","legend":"","description":"","filename":"20250529man2.docx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/80a2514aa7a86d46b9ea8b1d.docx"},{"id":94397142,"identity":"32957a43-853b-4dad-b271-f6755071b0be","added_by":"auto","created_at":"2025-10-27 13:56:29","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12114,"visible":true,"origin":"","legend":"","description":"","filename":"20250210Table1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/856908b9782778ba32cfcc21.xlsx"},{"id":94396458,"identity":"88e17d81-609f-406e-947f-7536c63f0c2a","added_by":"auto","created_at":"2025-10-27 13:56:01","extension":"json","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12740,"visible":true,"origin":"","legend":"","description":"","filename":"059ff8df247c4a5eb70a24af05dc2afc.json","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/a426eb13cb2632816f5da969.json"},{"id":94397463,"identity":"1256f457-8e78-4ea2-b9c3-84a1d78fcf27","added_by":"auto","created_at":"2025-10-27 13:56:41","extension":"pptx","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":165434,"visible":true,"origin":"","legend":"","description":"","filename":"20250519SupplementaryFigureS1.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/4e53f681961a2b60d3297a78.pptx"},{"id":94396572,"identity":"7b10c0d8-8905-4834-ba79-0e69651f6681","added_by":"auto","created_at":"2025-10-27 13:56:04","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":23796,"visible":true,"origin":"","legend":"","description":"","filename":"20250519SupplementaryTableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/88b48ab8f3b81395c4a6696e.xlsx"},{"id":94489369,"identity":"9ee6a26a-9389-4f9e-9bd6-fb3a5cf52a48","added_by":"auto","created_at":"2025-10-27 17:04:19","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20621,"visible":true,"origin":"","legend":"","description":"","filename":"20250519SupplementaryTableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/5139b5d54a962f6e3648ce32.xlsx"},{"id":94399180,"identity":"de6e34cd-469d-45fe-bf47-de75df2e038e","added_by":"auto","created_at":"2025-10-27 13:57:23","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12017,"visible":true,"origin":"","legend":"","description":"","filename":"20250519SupplementaryTableS3a.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/d01d65c94adcbf84d09551c7.xlsx"},{"id":94398179,"identity":"ca6688bb-c742-4a72-a684-5657512b1e66","added_by":"auto","created_at":"2025-10-27 13:57:00","extension":"xlsx","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12365,"visible":true,"origin":"","legend":"","description":"","filename":"20250519SupplementaryTableS3b.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/d17ff92a886b30b4aad4d6fa.xlsx"},{"id":94397603,"identity":"4564e344-1222-4af7-abe8-fcfc9d0cf4f5","added_by":"auto","created_at":"2025-10-27 13:56:43","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":111599,"visible":true,"origin":"","legend":"","description":"","filename":"059ff8df247c4a5eb70a24af05dc2afc1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/d1f9ec8075fa3bd9979dab73.xml"},{"id":94399174,"identity":"271cb564-e56e-49e1-a658-fd76166011b5","added_by":"auto","created_at":"2025-10-27 13:57:22","extension":"pptx","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42015,"visible":true,"origin":"","legend":"","description":"","filename":"20250511Figure1.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/cf1e444b3d9090cf45dbd18d.pptx"},{"id":94396456,"identity":"7f79d577-ce8d-460f-82db-4283ae2cf076","added_by":"auto","created_at":"2025-10-27 13:56:01","extension":"pptx","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1951126,"visible":true,"origin":"","legend":"","description":"","filename":"20250511Figure2.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/a6e85676036adea2380724a0.pptx"},{"id":94397250,"identity":"cb948477-f814-4c50-9dc7-2a8882f1e694","added_by":"auto","created_at":"2025-10-27 13:56:34","extension":"pptx","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1565225,"visible":true,"origin":"","legend":"","description":"","filename":"20250511Figure3.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/11f82e5e79d836deae2a7e5d.pptx"},{"id":94395927,"identity":"67512bba-1f29-4b60-8f55-0db5c02d99c3","added_by":"auto","created_at":"2025-10-27 13:55:46","extension":"pptx","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":578679,"visible":true,"origin":"","legend":"","description":"","filename":"20250511Figure4.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/e698e4385aa45a1ecc5461eb.pptx"},{"id":94396575,"identity":"4640d309-f245-4c45-b882-9bc16f5e8709","added_by":"auto","created_at":"2025-10-27 13:56:05","extension":"xml","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":110775,"visible":true,"origin":"","legend":"","description":"","filename":"059ff8df247c4a5eb70a24af05dc2afc1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/36fc78b2ec368b73e8af7c53.xml"},{"id":94396467,"identity":"a1e6302d-f72c-434e-b949-ccb7a0293a8d","added_by":"auto","created_at":"2025-10-27 13:56:01","extension":"html","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":124193,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/6798c451f511da7a74a5f4e6.html"},{"id":94396838,"identity":"2a5e2b56-c522-4989-9b65-8077088429d2","added_by":"auto","created_at":"2025-10-27 13:56:18","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":12785,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart depicting the process of data analysis\u003c/p\u003e","description":"","filename":"20250511Figure11.png","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/6329c0dc7c732f1ae08cf9e3.png"},{"id":94397651,"identity":"0f0afff5-27ac-47a8-9f44-6139926bf3ee","added_by":"auto","created_at":"2025-10-27 13:56:44","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":265991,"visible":true,"origin":"","legend":"\u003cp\u003eForecasting monthly adverse event reports using the Prophet model\u003c/p\u003e","description":"","filename":"20250511Figure21.png","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/eeae9b6d9f206d14e85e40c6.png"},{"id":94397661,"identity":"1917bd0b-86ad-4f6a-8f58-b441a86f8ee9","added_by":"auto","created_at":"2025-10-27 13:56:45","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":200587,"visible":true,"origin":"","legend":"\u003cp\u003eSeasonality curve of the component of monthly adverse event reports in JADER estimated using the Prophet model\u003c/p\u003e","description":"","filename":"20250511Figure31.png","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/092c84c4299ba466b2487793.png"},{"id":94396840,"identity":"5290caa1-d60a-45f7-9cba-0f2e2df0fc21","added_by":"auto","created_at":"2025-10-27 13:56:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":77108,"visible":true,"origin":"","legend":"\u003cp\u003eMonthly distribution of adverse event reports aggregated over a 15-year period (2005–2019)\u003c/p\u003e","description":"","filename":"20250511Figure41.png","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/d0d04d0de7d1d52ddb56b20b.png"},{"id":94491209,"identity":"159516d6-ce01-40c5-ae6a-9b901a93f5bf","added_by":"auto","created_at":"2025-10-27 17:23:52","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1289251,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/1bee4727-f3e5-447a-b6da-a033286ab48c.pdf"},{"id":94398893,"identity":"ac39d7af-1343-4d52-ae5c-4643a6994136","added_by":"auto","created_at":"2025-10-27 13:57:14","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":12114,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 1 \u003c/strong\u003eComparison of monthly seasonal values derived from Prophet analysis and maximum/minimum aggregate counts in reported adverse events.\u003c/p\u003e","description":"","filename":"20250210Table1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/744cc8e760df33ad300143e3.xlsx"},{"id":94398622,"identity":"aab28e84-a0ac-4279-9eec-cec5da1a8878","added_by":"auto","created_at":"2025-10-27 13:57:09","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":23796,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table S1\u003c/strong\u003e Comparative evaluation of MAE, RMSE, and MAPE between seasonal and non-seasonal Prophet models.\u003c/p\u003e","description":"","filename":"20250519SupplementaryTableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/bfaeb7180264ca463d9f58c6.xlsx"},{"id":94397821,"identity":"2a45b195-5efa-4519-9758-c90950adf539","added_by":"auto","created_at":"2025-10-27 13:56:49","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":20621,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table S2 \u003c/strong\u003eComparative evaluation of the MAE, RMSE, and MAPE between SARIMA and ARIMA models.\u003c/p\u003e","description":"","filename":"20250519SupplementaryTableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/dd1159cdab7c9bc735cd31d3.xlsx"},{"id":94397824,"identity":"4639d6ca-19ce-45c5-bfa7-cf2369f63ab6","added_by":"auto","created_at":"2025-10-27 13:56:49","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":12017,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table S3a\u003c/strong\u003e Top ten suspected drugs reported for each adverse event.\u003c/p\u003e","description":"","filename":"20250519SupplementaryTableS3a.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/79408997c945c3549346d4c3.xlsx"},{"id":94395921,"identity":"b09f7dae-8fa7-4efd-9fb1-517aa136d65c","added_by":"auto","created_at":"2025-10-27 13:55:45","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":12365,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table S3b\u003c/strong\u003e (continued) Top ten suspected drugs reported for each adverse event.\u003c/p\u003e","description":"","filename":"20250519SupplementaryTableS3b.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/59b22d0529d6af29d6860b7f.xlsx"},{"id":94398992,"identity":"5d80c00f-a5a9-4144-b058-5fb12a8f7ea3","added_by":"auto","created_at":"2025-10-27 13:57:18","extension":"pptx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":165434,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure S1\u003c/strong\u003e Average monthly temperature and relative humidity from 2015 to 2019.\u003c/p\u003e","description":"","filename":"20250519SupplementaryFigureS1.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6746128/v1/77a678451cab858daf0889eb.pptx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Seasonality in Adverse Drug Events: Time-Series Analysis of JADER Using ARIMA/SARIMA and Prophet","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSpontaneous reporting systems (SRS) play a vital role in pharmacovigilance by collecting real-world data on the adverse events (AEs) associated with medicinal products. In Japan, such data are compiled in the Japanese Adverse Drug Event Report (JADER) database managed by the Pharmaceuticals and Medical Devices Agency (PMDA). This database provides the AE occurrence dates, thus enabling researchers to explore potential temporal trends. Prior studies using JADER and other SRS databases have observed fluctuations in AEs, including dehydration and drug-induced photosensitivity, suggesting possible seasonal factors\u0026mdash;whether climatic or behavioral.\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e These results underscore the importance of examining when AEs spike to better understand and mitigate risks.\u003c/p\u003e\u003cp\u003eAlthough temporal information is crucial, many investigations have relied on monthly AE aggregations without employing advanced time-series methods. This approach can obscure genuine seasonal patterns and fails to adequately account for confounders such as overreporting, underreporting, and missing data. By adopting models that capture more nuanced temporal dynamics, researchers can discover accurate seasonal trends and refine risk management strategies.\u003c/p\u003e\u003cp\u003eStatistical tools such as the autoregressive integrated moving average (ARIMA) and its seasonal extension, seasonal ARIMA (SARIMA), are widely used in domains ranging from economics to epidemiology.\u003csup\u003e\u003cspan additionalcitationids=\"CR5 CR6\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e ARIMA breaks down time-series data into autoregressive and moving average components, with differencing to handle nonstationarity. SARIMA adds parameters specifically for recurring seasonal fluctuations, making it more suitable when the data exhibit repeated cycles, such as monthly surges in AEs.\u003c/p\u003e\u003cp\u003eFacebook introduced a newer approach, the Prophet model.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e This model treats a time series as a sum of trend, seasonal, and external event effects and is designed to handle outliers, abrupt changes, and missing values. It also allows the inclusion of domain-specific factors such as holidays or other known events, which can refine forecasts that predict possible shifts in reporting rates particular periods (e.g., influenza season or significant public health announcements).\u003c/p\u003e\u003cp\u003eThe principal aim of this study was to determine whether the AEs reported in the JADER database display meaningful seasonal patterns by applying the SARIMA and Prophet models. By comparing these methods, we sought to clarify the temporal distribution of AEs and identify the periods of elevated risk. In doing so, we obtained insights that will help clinicians, regulators, and researchers implement more proactive and targeted pharmacovigilance in real-world settings.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eData Source\u003c/h2\u003e\u003cp\u003eData compiled from April 2004 to September 2024 in the JADER database were downloaded from the PMDA and Japanese Regulatory Authority websites.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e All the JADER data have been cleaned and anonymized by the PMDA. In Japan, in the absence of a dedicated vaccine-related AE database, such as Vaccine Adverse Event Reporting System (VAERS), AEs related to vaccines are reported within the JADER database. The database consists of four data tables: 1) patient demographic information (DEMO), 2) drug information (DRUG), 3) AEs (REAC), and 4) primary illness (HIST). Only reports of AE occurrence with dates recorded correctly in the REAC table were analyzed in the present study.\u003c/p\u003e\u003cp\u003eAEs with a total count of 100 or more were selected across all months, resulting in a dataset comprising monthly AE counts over a 15-year period from January 2005 to December 2019.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eDefinition of Adverse Events\u003c/h3\u003e\n\u003cp\u003eThe AEs in the JADER database were defined based on the Medical Dictionary for Regulatory Activities version 27.0.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003ch3\u003eARIMA and SARIMA Models\u003c/h3\u003e\n\u003cp\u003eTo examine seasonality in AE reports, we employed both the ARIMA and SARIMA models. The ARIMA model is defined using three parameters: autoregressive order (p), degree of differencing (d), and moving average order (q). Differencing stabilizes the mean of a time series by removing level shifts, thus enabling the modeling of non-stationary data while preserving its linear structure.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e ARIMA operates by modeling dependencies between observations in a sequence, making it effective for datasets without strong seasonal components. However, it lacks the ability to handle periodic fluctuations explicitly, which limits its application for data with significant seasonality. SARIMA extends ARIMA by adding the seasonal parameters P, D, Q, and the length of the seasonal cycle s (s\u0026thinsp;=\u0026thinsp;12 for monthly data), thus allowing us to capture seasonal effects. We determined the optimal ARIMA and SARIMA parameters for each AE using a grid search over p, d, q (0\u0026ndash;2) and P, D, Q (0\u0026ndash;2), with s\u0026thinsp;=\u0026thinsp;12.\u003csup\u003e12\u003c/sup\u003e The best parameter combination was selected based on the Akaike Information Criterion.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e To evaluate the predictive performance, we used five-fold time series split cross-validation while preserving the chronological order to avoid data leakage.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e The mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) were calculated.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e For each AE, if the MAE of the SARIMA model was lower than that of the ARIMA model, the AE was considered potentially seasonal.\u003c/p\u003e\u003cp\u003eThe analyses were conducted using Python and the Statsmodels library for ARIMA and SARIMA modeling.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003ch3\u003eProphet model\u003c/h3\u003e\n\u003cp\u003eProphet is an open-source software released by Facebook\u0026rsquo;s Core Data Science team. It works best with time series that display strong seasonal effects and have several seasons of historical data. Prophet is robust to missing data and shifts in trends, and typically handles outliers well.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e The strength of the ARIMA model lies in its ability to capture linear dependencies in time-series data.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e However, data with complex nonlinear relationships or structural changes falls outside the scope of what a pure ARIMA model can effectively handle.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e In contrast, Prophet models compensate for this limitation.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e Therefore, we constructed both SARIMA and Prophet models.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eWe employed two variants of Facebook\u0026rsquo;s Prophet model\u0026mdash;one allowing for yearly seasonality (\u0026ldquo;seasonal Prophet\u0026rdquo;) and one excluding it (\u0026ldquo;non-seasonal Prophet\u0026rdquo;)\u0026mdash;to investigate potential seasonal patterns.\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eSeasonal model: yearly_seasonality\u0026thinsp;=\u0026thinsp;True\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eNon-seasonal model: yearly_seasonality\u0026thinsp;=\u0026thinsp;False\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eIn both models, the default parameters were utilized with no additional regressors or custom seasonalities. We performed time-series cross-validation using Prophet\u0026rsquo;s cross_validation function, specifying an initial training period of 1095 days (three years), a period of 365 days (one year), and a horizon of 365 days (one year). We calculated the MAE, RMSE, and MAPE for both models. For each AE, if the seasonal model had a lower MAE than the non-seasonal model, it was identified as potentially exhibiting seasonality. The analyses were performed using Python.\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eData spanning a 15-year period from January 2005 to December 2019 were downloaded from the JADER database and analyzed. The date data were preprocessed, and 390,036 reports were subsequently analyzed, including 727 AEs (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The implementation of the seasonal and non-seasonal Prophet models revealed that the MAE of the seasonal model was lower than that of the non-seasonal model for 75 AEs (10.3%) (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The implementation of the ARIMA and SARIMA models revealed that the MAE of SARIMA was lower than that of ARIMA for 52 AEs (7.2%) (Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The seasonal models demonstrated superior performance in both the ARIMA/SARIMA and Prophet analyses for 16 AEs (2.2%), including acute disseminated encephalomyelitis, appetite loss, aspiration, atypical femur fracture, encephalitis, Guillain-Barr\u0026eacute; syndrome (GBS), hypophysitis, immune-mediated enterocolitis, influenza, intestinal ischemia, lacunar infarction, malaise, paronychia, secondary adrenal insufficiency, transfusion-associated circulatory overload, and type 1 diabetes mellitus. The 16 AEs were forecasted using the Prophet model (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and subsequently analyzed with yearly seasonality components (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). A comparison of the monthly peaks and troughs between the 15-year aggregate report counts and Prophet-derived seasonal components revealed exact temporal alignment only for lacunar infarction (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Table\u0026nbsp;1). Seasonal patterns were concordant for appetite loss, influenza, intestinal ischemia, and paronychia. However, a subset of AEs, including aspiration, encephalitis, GBS, hypophysitis, immune-mediated enterocolitis, secondary adrenal insufficiency, and transfusion-associated circulatory overload, demonstrated no concordance between the highest and maximum months or between the lowest and minimum months.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003eSeasonal Characteristics and Overview\u003c/h2\u003e\u003cp\u003eJapan has four distinct seasons, and each season exerts unique physiological effects. Tokyo\u0026rsquo;s meteorological data (2015\u0026ndash;2019) show mean monthly temperatures from 5.6\u0026deg;C (January) to 27.3\u0026deg;C (August),\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e and relative humidity ranging from 53.0% (January) to 81.8% (September) (Supplementary Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e These seasonal factors and patient behavior may contribute to the seasonality of AE reports. Among the 727 AEs evaluated, only 52 (7.2%) appeared to have a seasonal component and only 16 (2.2%) demonstrated seasonality in both ARIMA/SARIMA and Prophet models. This result means that most AEs do not have a seasonal component\u0026mdash;a finding that is not entirely unexpected, and one which reinforces the notion that AEs exhibiting seasonal characteristics are relatively uncommon. According to the results of our analysis using both the Prophet and ARIMA/SARIMA models, 16 AEs displayed seasonality potentially linked to infectious diseases, climatic factors, and patient behavior. The drugs most commonly associated with these AEs were vaccines and antineoplastic agents (Supplementary Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eInfectious Diseases, Vaccines, and Autoimmune Conditions\u003c/h3\u003e\n\u003cp\u003eAmong the 16 AEs, influenza, GBS, acute disseminated encephalomyelitis, and encephalitis were frequently associated with influenza vaccines according to the JADER database, peaking from November to January. Although these AEs may be vaccine-related, they may also reflect concurrent infectious triggers. The influenza vaccines approved in Japan are inactivated; therefore, reports of post-vaccination influenza may involve indirect factors.\u003c/p\u003e\u003cp\u003eEpidemiological evidence indicates that the older people are more commonly affected (peak incidence of GBS is between 50 and 70 years of age).\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e GBS risk following vaccination is only slightly elevated,\u003csup\u003e\u003cspan additionalcitationids=\"CR21 CR22 CR23\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e with approximately 1\u0026ndash;1.6 cases per million H1N1 vaccinations.\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e Immune checkpoint inhibitors (ICIs) are also suspected drugs for GBS, and ICI-induced cases may differ mechanistically from classical GBS.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e Seasonal GBS spikes vary internationally and may be linked to prevalent pathogens, including \u003cem\u003eCampylobacter jejuni\u003c/em\u003e and influenza A virus.\u003csup\u003e\u003cspan additionalcitationids=\"CR25 CR26\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eAcute disseminated encephalomyelitis is a rare demyelinating disease that mostly affects children, with an incidence of 0.2\u0026ndash;0.5 per 100,000 individuals.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e The average age of onset in pediatric cases is typically between 3 and 7 years,\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e although the condition can manifest at any age. Several studies have reported a slight male predominance.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e A preceding infection has been reported in approximately 75% of patients examined in previous studies\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e; although vaccination-associated acute disseminated encephalomyelitis has also been noted, albeit without definitive causal evidence.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e In JADER, most acute disseminated encephalomyelitis reports involved vaccines, whereas encephalitis was reported with both vaccines and ICIs. Autoimmune encephalitis following vaccination may be mediated by molecular mimicry or immune activation.\u003csup\u003e\u003cspan additionalcitationids=\"CR33 CR34\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e The mechanisms of ICI-related encephalitis include the generation of autoantibodies (e.g., anti-N-methyl-D-aspartate receptors).\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e In Asia, the Japanese encephalitis virus remains the leading cause of encephalitis, typically occurring from June to October, whereas in Western countries, the herpes simplex virus predominates as the cause of encephalitis.\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eNon-infectious AEs and Seasonal Influences\u003c/h2\u003e\u003cp\u003eSeveral non-infectious AEs such as aspiration, appetite loss, malaise, atypical femur fracture, lacunar infarction, intestinal ischemia, and paronychia also showed seasonality. In the JADER database, aspiration often involves antipsychotic drugs, possibly due to impaired swallowing. Seasonal affective disorder, a subtype of major depression that occurs from autumn to winter,\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e might influence dietary habits and mood, further influencing aspiration risk.\u003c/p\u003e\u003cp\u003eAppetite loss and malaise present similarly but have been reported with different suspected drugs in JADER. The most suspected drugs for appetite loss include antineoplastic agents, ribavirin, and interferon. For malaise, antineoplastic agents, osteoporosis medications, pregabalin, and valacyclovir are the predominant suspected drugs. Appetite loss peaked in late spring and summer (May\u0026ndash;August) and declined in winter (December\u0026ndash;February). High temperatures may induce summer heat fatigue, which encompasses fatigue, anorexia, and autonomic dysfunction.\u003csup\u003e\u003cspan additionalcitationids=\"CR42\" citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e Antineoplastic agents themselves disrupt thermoregulation and contribute to heat- or cold-related health issues.\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e Seasonal fluctuations in energy intake and stress responses have also been documented, which potentially modulate these AEs.\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eWhile fractures generally increase during winter,\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e atypical femur fractures in corticosteroid users peak in April rather than in the colder months, suggesting the possibility of underreporting falls in winter as AEs. Lacunar infarctions appear to be influenced by dehydration from sodium-glucose cotransporter-2 inhibitors, as higher blood viscosity due to volume depletion increases stroke risk.\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e The incidence of lacunar infarctions peaks in June and September rather than in midsummer, which may indicate that unrecognized summer cases are attributable to ambient heat rather than medication. In addition to dehydration, the risk factors for lacunar infarction include hypertension, diabetes, and hyperlipidemia.\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eRegarding intestinal ischemia and paronychia, the JADER patterns diverge from current epidemiological data. In the United States, acute vascular injury of the intestine peaks in late summer, with more severe outcomes in winter.\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e There is limited literature on paronychia seasonality,\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e although humidity and temperature can favor nail infections. According to the JADER database, both conditions are associated with the use of antineoplastic agents. Paronychia reporting is likely tied to the duration of drug use rather than to temperature shifts; however, a seasonal distribution emerged that may reflect the underlying reporting trends.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eICI-induced AEs and Seasonal Factors\u003c/h2\u003e\u003cp\u003eThe seasonality of ICI-induced events, including immune-mediated enterocolitis, hypophysitis, secondary adrenal insufficiency, and type 1 diabetes mellitus, could stem from various mechanisms. Proposed influences include vitamin D fluctuations and photoperiodic changes in endocrine or immune function.\u003csup\u003e\u003cspan additionalcitationids=\"CR55 CR56\" citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e The composition of the gut microbiota, which varies seasonally, may also play a role in enterocolitis. Although direct evidence is limited, these hypotheses highlight the potential of climate and daylight patterns to modulate immune-related AEs.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eTransfusion-associated Circulatory Overload\u003c/h2\u003e\u003cp\u003eTransfusion-associated circulatory overload is a transfusion reaction that is especially serious in patients with cardiac comorbidities.\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e Although hospitalizations for heart failure often peak in January, with higher mortality from August to October,\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e the specific factors contributing to the seasonality of transfusion-associated circulatory overload remain unclear. Further research is required to elucidate any direct links.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eLimitations and Model Considerations\u003c/h2\u003e\u003cp\u003eThis study has inherent limitations related to SRS, such as overreporting, underreporting, missing data, exclusion of healthy subjects resulting in lack of a control population or reference group, and the presence of confounding factors. External factors, such as safety communications and market dynamics, may also influence reporting patterns.\u003c/p\u003e\u003cp\u003eThe divergence between Prophet forecasts and raw counts often stems from smoothing outliers, short-term fluctuations, and fundamental methodological differences. For AEs such as hypophysitis, immune-mediated enterocolitis, and secondary adrenal insufficiency, the steadily increasing reports of ICIs following the approval of nivolumab in Japan in 2014 make it challenging to separate true seasonal signals from upward reporting trends. Prophet\u0026rsquo;s flexibility may capture such long-term patterns better by dealing robustly with missing data, outliers, and abrupt changes, thus offering valuable insights into seasonal AE distributions recorded in the JADER database.\u003c/p\u003e\u003cp\u003eMoreover, the determination of the seasonality of a time series is highly dependent on the method selected, even when applied to the same database.\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e The ARIMA model assumes that all predictable structures can be captured by trend and autocorrelation terms. When seasonal cycles are subtle or weak, ARIMA may treat them as white noise. In contrast, Prophet offers high model flexibility through additive components, which can lead to overfitting by capturing noise or incidental patterns as meaningful structures in the data.\u003c/p\u003e\u003cp\u003eThe present study revealed that infectious diseases, AEs related to vaccines, and climatic factors contribute to the seasonal patterns observed in the JADER database. Future research using more granular data, including detailed patient demographics, clinical conditions, and environmental factors, are required to uncover the mechanisms underlying seasonal trends.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eBy applying both the ARIMA/SARIMA and Prophet models to the JADER database, we identified several AEs with potential seasonal patterns, many of which were associated with infectious diseases, climate factors, and medication usage, particularly vaccinations. A more nuanced understanding of these seasonal dynamics can bolster risk-management strategies, inform medication safety protocols, and ultimately enhance patient care. This study also demonstrates the utility of combining advanced time-series approaches with domain-specific knowledge, thereby offering a robust framework for professionals seeking reliable predictions of AE occurrence. Anticipating seasonal surges in AE reporting may contribute to improved risk communication, timely clinical interventions, and efficient allocation of healthcare resources. Our findings reinforce the importance of recognizing and analyzing seasonality in pharmacovigilance data and highlight how tailored forecasting methods can capture both long- and short-term trends. By extending these techniques to broader contexts, we hope to guide medical professionals, data analysts, researchers, and clinical application developers in selecting appropriate models for accurate AE forecasting, thus contributing to a more comprehensive understanding of AE seasonality and proactive risk management.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors (s) disclose receipt of the following financial support for this article\u0026rsquo;s research, authorship, and/or publication: This research was partially supported by the Japan Society for the Promotion of Science KAKENHI grant numbers 25K10049 and 21K06646. The funders had no role in the study design, data collection, analysis, decision to publish, or preparation of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEthical approval was not sought for this study because the study was a database-related observational study which did not directly involve any research subjects. All results were obtained from data openly available online from the PMDA website (www.pmda.go.jp). All data from the JADER database were fully anonymized by the relevant regulatory authority before we accessed them.\u003c/p\u003e\n\u003cp\u003eOur research does not fall within the purview of any of the following laws and guidelines: \u0026ldquo;Clinical Trials Act (Act No. 16 of April 14, 2017),\u0026rdquo; \u0026ldquo;Act on Securing Quality, Efficacy and Safety of Products Including Pharmaceuticals and Medical Devices (Law number: Act No. 145 of 1960, Last Version: Amendment of Act No. 50 of 2015),\u0026rdquo; \u0026ldquo;Guideline for good clinical practice E6 (R1), https://www.pmda.go.jp/int-activities/int-harmony/ich/0076.html,\u0026rdquo; \u0026ldquo;Ethical guidelines for human genome and gene analysis research, https://www.mhlw.go.jp/general/seido/kousei/i-kenkyu/genome/0504sisin.html,\u0026rdquo; and \u0026ldquo;Ethical Guidelines for Medical and Health Research Involving Human Subjects, https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/hokabunya/kenkyujigyou/i-kenkyu/index.html.\u0026rdquo; Therefore, it is not subject to ethical examination. The study was an observational study without any research subjects. No consent to participate was required due to the retrospective nature of this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Concept and design: Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, and Mitsuhiro Nakamura. Acquisition, analysis, or interpretation of data: Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, and Mitsuhiro Nakamura. Drafting of the manuscript: Hideyuki Tanaka, Mika Maezawa, Satoshi Nakao, Kohei Shiota, Moe Yamashita, Nanaka Ichihara, Yuka Nokura, Tomofumi Yamazaki, and Kana Sugishita. Critical review of the manuscript for important intellectual content: Yoko Ino, Kazuhiro Iguchi and Jun Liao.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll results were obtained from data openly available online from the PMDA website (www.pmda.go.jp). The datasets used and analyzed in the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no potential conflicts of interest with respect to the research, authorship, or publication of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e: This research was partially supported by the Japan Society for the Promotion of Science KAKENHI grant numbers 25K10049 and 21K06646.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMatsumoto, K. et al. A retrospective study of seasonal variation in sodium-glucose co-transporter 2 inhibitor-related adverse events using the Japanese adverse drug event report database.\u003cem\u003e Sci. Rep\u003c/em\u003e. \u003cstrong\u003e14\u003c/strong\u003e, 30072 (2024). \u003c/li\u003e\n\u003cli\u003eNakao, S. et al. Evaluation of drug-induced photosensitivity using the Japanese Adverse Drug Event Report (JADER) database. \u003cem\u003eBiol. Pharm. Bull\u003c/em\u003e. \u003cstrong\u003e40\u003c/strong\u003e,2158\u0026ndash;2165 (2017).\u003c/li\u003e\n\u003cli\u003eMarrero, O., Hung, E. Y. \u0026amp; Hauben, M. Seasonal and geographic variation in adverse event reporting. \u003cem\u003eDrugs-Real World Outcomes\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 297\u0026ndash;306 (2016).\u003c/li\u003e\n\u003cli\u003eBox, G. E. P., Jenkins, G. M., Reinsel, G. C. \u0026amp; Ljung, G. M. \u003cem\u003eTime series analysis: forecasting and control.\u003c/em\u003e 5th\u003csup\u003e \u003c/sup\u003eedn, (John Wiley \u0026amp; Sons, 2015).\u003c/li\u003e\n\u003cli\u003eHyndman, R. J. \u0026amp; Athanasopoulos, G. \u003cem\u003eForecasting: principles and practice. \u003c/em\u003e3rd edn, (OTexts, 2021) https://otexts.com/fpp3/. \u003c/li\u003e\n\u003cli\u003eAkshay, K.C., Grace, G.H., Gunasekaran, K. \u0026amp; Samikannu, R. Power consumption prediction for electric vehicle charging stations and forecasting income. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 6497 (2024).\u003c/li\u003e\n\u003cli\u003ePaik, J., \u0026amp; Ko, K. Seasonal analysis and risk management strategies for credit guarantee funds: a case study from republic of Korea. \u003cem\u003eStats\u003c/em\u003e. \u003cstrong\u003e8\u003c/strong\u003e, 2 (2025). \u003c/li\u003e\n\u003cli\u003eTaylor, S. J. \u0026amp; Letham, B. Forecasting at scale. \u003cem\u003eAm. Stat.\u003c/em\u003e \u003cstrong\u003e72\u003c/strong\u003e, 37\u0026ndash;45 (2018).\u003c/li\u003e\n\u003cli\u003eFacebook. Prophet: Forecasting at Scale. https://facebook.github.io/prophet/ (2017).\u003c/li\u003e\n\u003cli\u003ePharmaceuticals and Medical Devices Agency (PMDA). Suspected adverse drug reaction information. (in Japanese). https://www.pmda.go.jp/safety/info-services/drugs/adr-info/suspected-adr/0003.html, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eMedical Dictionary for Regulatory Activities. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). https://www.ich.org/page/meddra, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eKontopoulou, V. I., Panagopoulos, A. D., Kakkos, I. \u0026amp; Matsopoulos, G. K. A review of ARIMA vs. machine learning approaches for time series forecasting in data driven networks. \u003cem\u003eFuture Internet\u003c/em\u003e, \u003cstrong\u003e15\u003c/strong\u003e, 255, (2023).\u003c/li\u003e\n\u003cli\u003eBergmeir, C. \u0026amp; Ben\u0026iacute;tez, J. M. On the use of cross-validation for time series predictor evaluation. \u003cem\u003eInf. Sci.\u003c/em\u003e \u003cstrong\u003e191\u003c/strong\u003e, 192\u0026ndash;213 (2012).\u003c/li\u003e\n\u003cli\u003eHodson, T. O. Root-mean-square error (RMSE) or mean absolute error (MAE): when to use them or not. \u003cem\u003eGeosci. Model Dev.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 5481\u0026ndash;5487 (2022).\u003c/li\u003e\n\u003cli\u003eKwarteng, S. B. \u0026amp; Andreevich, P. A. Comparative analysis of ARIMA, SARIMA and Prophet model in forecasting. \u003cem\u003eRes. Dev.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 110\u0026ndash;120 (2024).\u003c/li\u003e\n\u003cli\u003eStatsmodels. Statsmodels.tsa.statespace.sarimax.SARIMAX. https://www.statsmodels.org/stable/generated/statsmodels.tsa.statespace.sarimax.SARIMAX.html, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eProphet. Prophet: Quick Start. https://facebook.github.io/prophet/docs/quick_start.html, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eJapan Meteorological Agency. Monthly average temperature for Tokyo (in Japanese). https://www.data.jma.go.jp/obd/stats/etrn/view/monthly_s3.php?prec_no=44\u0026amp;block_no=47662, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eJapan Meteorological Agency. Monthly average relative humidity in Tokyo (in Japanese). https://www.data.jma.go.jp/obd/stats/etrn/view/monthly_s3.php?prec_no=44\u0026amp;block_no=47662\u0026amp;year=\u0026amp;month=\u0026amp;day=\u0026amp;view=a7, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eBellanti, R. \u0026amp; Rinaldi, S. Guillain\u0026ndash;Barr\u0026eacute; syndrome: a comprehensive review. \u003cem\u003eEur. J. Neurol.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, e16365 (2024).\u003c/li\u003e\n\u003cli\u003evan Doorn, P. A. et al. European academy of neurology/peripheral nerve society guideline on diagnosis and treatment of Guillain\u0026ndash;Barr\u0026eacute; syndrome. \u003cem\u003eJ. Peripher. Nerv.\u003c/em\u003e \u003cem\u003eSyst.\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 535\u0026ndash;563 (2023).\u003c/li\u003e\n\u003cli\u003eSalmon, D. A., Dudley, M. Z. \u0026amp; Carleton, B. C. Guillain\u0026ndash;Barr\u0026eacute; syndrome following influenza vaccines affords opportunity to improve vaccine confidence. \u003cem\u003eJ. Infect. Dis.\u003c/em\u003e \u003cstrong\u003e223\u003c/strong\u003e, 355\u0026ndash;358 (2021).\u003c/li\u003e\n\u003cli\u003eDing, M. et al. Case report: ICIs-induced Guillain\u0026ndash;Barr\u0026eacute; syndrome recovered from mycophenolate mofetil. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 1132692 (2023).\u003c/li\u003e\n\u003cli\u003eZhai, Q. et al. Guillain\u0026ndash;Barr\u0026eacute; syndrome in northern China: a retrospective analysis of 294 patients from 2015 to 2020. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 6323 (2022).\u003c/li\u003e\n\u003cli\u003eGinanneschi, F. et al. Clinical features and outcome of the Guillain\u0026ndash;Barr\u0026eacute; syndrome: a single-center 11-year experience. \u003cem\u003eFront. Neurol.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 856091 (2022).\u003c/li\u003e\n\u003cli\u003eChroni, E. et al. Guillain\u0026ndash;Barr\u0026eacute; syndrome in Greece: seasonality and other clinico-epidemiological features. \u003cem\u003eEur. J. Neurol.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 383\u0026ndash;388 (2004).\u003c/li\u003e\n\u003cli\u003eFinsterer, J. Triggers of Guillain\u0026ndash;Barr\u0026eacute; syndrome: \u003cem\u003eCampylobacter jejuni\u003c/em\u003e predominates. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 14222 (2022).\u003c/li\u003e\n\u003cli\u003eWolska-Krawczyk, M. Akute disseminierte Enzephalomyelitis [Acute disseminated encephalomyelitis]. \u003cem\u003eRadiologe\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, 316\u0026ndash;321 (2022). (in German)\u003c/li\u003e\n\u003cli\u003eLeake, J. A. D. et al. Acute disseminated encephalomyelitis in childhood: epidemiologic, clinical and laboratory features. \u003cem\u003ePediatr. Infect. Dis. J.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 756\u0026ndash;764 (2004).\u003c/li\u003e\n\u003cli\u003eCole, J., Evans, E., Mwangi, M. \u0026amp; Mar, S. Acute disseminated encephalomyelitis in children: an updated review based on current diagnostic criteria. \u003cem\u003ePediatr. Neurol.\u003c/em\u003e \u003cstrong\u003e100\u003c/strong\u003e, 26\u0026ndash;34 (2019).\u003c/li\u003e\n\u003cli\u003ePavone, P. et al. Acute disseminated encephalomyelitis: a long-term prospective study and meta-analysis. \u003cem\u003eNeuropediatrics\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, 246\u0026ndash;255 (2010).\u003c/li\u003e\n\u003cli\u003eGuedes, B. F. et al. Potential autoimmune encephalitis following yellow fever vaccination: a report of three cases. \u003cem\u003eJ. Neuroimmunol.\u003c/em\u003e \u003cstrong\u003e355\u003c/strong\u003e, 577548 (2021).\u003c/li\u003e\n\u003cli\u003eZuhorn, F., Graf, T., Klingebiel, R., Sch\u0026auml;bitz, W.-R. \u0026amp; Rogalewski, A. Postvaccinal encephalitis after ChAdOx1 nCov-19. \u003cem\u003eAnn. Neurol.\u003c/em\u003e \u003cstrong\u003e90\u003c/strong\u003e, 506\u0026ndash;511 (2021).\u003c/li\u003e\n\u003cli\u003eWang, H. Anti-NMDA receptor encephalitis and vaccination. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 193 (2017).\u003c/li\u003e\n\u003cli\u003eMartin, S., Azzouz, B., Morel, A. \u0026amp; Trenque, T. Anti-NMDA receptor encephalitis and vaccination: a disproportionality analysis. \u003cem\u003eFront. Pharmacol.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 940780 (2022).\u003c/li\u003e\n\u003cli\u003eBuckley, M. W. et al. Immune-related encephalitis after immune checkpoint inhibitor therapy. \u003cem\u003eOncologist\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, oyae186 (2025).\u003c/li\u003e\n\u003cli\u003eKumar, R. Understanding and managing acute encephalitis. \u003cem\u003eF1000Res.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 60 (2020).\u003c/li\u003e\n\u003cli\u003eCenters for Disease Control and Prevention (CDC). Japanese encephalitis: data and maps. https://www.cdc.gov/japanese-encephalitis/data-maps/index.html, cited 16 May, 2025.\u003c/li\u003e\n\u003cli\u003eSaeed, S. A. \u0026amp; Bruce, T. J. Seasonal affective disorders. \u003cem\u003eAm. Fam. Physician\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, 1340\u0026ndash;1346 (1998).\u003c/li\u003e\n\u003cli\u003eDonofry, S.D., Roecklein, K.A., Rohan, K.J., Wildes, J.E. \u0026amp; Kamarck, M.L. Prevalence and correlates of binge eating in seasonal affective disorder. \u003cem\u003ePsychiatry Res.\u003c/em\u003e \u003cstrong\u003e217\u003c/strong\u003e, 47\u0026ndash;53 (2014).\u003c/li\u003e\n\u003cli\u003eEbi, K. L. et al. Hot weather and heat extremes: health risks. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e398\u003c/strong\u003e, 698\u0026ndash;708 (2021).\u003c/li\u003e\n\u003cli\u003eGreaney, J. L., Kenney, W. L. \u0026amp; Alexander, L. M. Sympathetic regulation during thermal stress in human aging and disease. \u003cem\u003eAuton. Neurosci.\u003c/em\u003e \u003cstrong\u003e196\u003c/strong\u003e, 81\u0026ndash;90 (2016).\u003c/li\u003e\n\u003cli\u003eMakino, S. et al. Anti-fatigue effects of yogurt fermented with \u003cem\u003eLactobacillus delbrueckii\u003c/em\u003e subsp. \u003cem\u003ebulgaricus\u003c/em\u003e OLL1073R-1 in healthy people suffering from summer heat fatigue: a randomized, double-blind, placebo-controlled trial. \u003cem\u003eNutrients\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 798 (2018).\u003c/li\u003e\n\u003cli\u003eShao, Q. et al. Tumor therapeutic response monitored by telemetric temperature sensing, a preclinical study on immunotherapy and chemotherapy. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 7727 (2023).\u003c/li\u003e\n\u003cli\u003eKokolus, K. M., Hong, C. C. \u0026amp; Repasky, E. A. Feeling too hot or cold after breast cancer: is it just a nuisance or a potentially important prognostic factor? \u003cem\u003eInt. J. Hyperthermia\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 662\u0026ndash;680 (2010). \u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eFujihira, K., Takahashi, M., Wang, C. \u0026amp; Hayashi, N.\u003c/strong\u003eFactors explaining seasonal variation in energy intake: a review. \u003cem\u003eFront. Nutr.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 1192223 (2023).\u003c/li\u003e\n\u003cli\u003eBulajic-Kopjar, M. Seasonal variations in incidence of fractures among elderly people. \u003cem\u003eInj. Prev.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 16\u0026ndash;19 (2000).\u003c/li\u003e\n\u003cli\u003eBischoff-Ferrari, H. A., Orav, J. E., Barrett, J. A. \u0026amp; Baron, J. A. Effect of seasonality and weather on fracture risk in individuals 65 years and older. \u003cem\u003eOsteoporos. Int.\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 1225\u0026ndash;1233 (2007).\u003c/li\u003e\n\u003cli\u003eCort\u0026eacute;s-Vicente, E. et al. Frequency, risk factors, and prognosis of dehydration in acute stroke. \u003cem\u003eFront. Neurol.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 305 (2019).\u003c/li\u003e\n\u003cli\u003eSong, S. H. et al. Elevated blood viscosity is associated with cerebral small vessel disease in patients with acute ischemic stroke. \u003cem\u003eBMC Neurol.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 20 (2017).\u003c/li\u003e\n\u003cli\u003eYaghi, S. et al. Lacunar stroke: mechanisms and therapeutic implications. \u003cem\u003eJ. Neurol. Neurosurg. Psychiatry\u003c/em\u003e \u003cstrong\u003e92\u003c/strong\u003e, 823\u0026ndash;830 (2021).\u003c/li\u003e\n\u003cli\u003eJehangir, A., Dhital, R., Qureshi, A. \u0026amp; York, E.P. Acute vascular insufficiency of intestine: incidence highest in summer, outcomes worst in winter. \u003cem\u003eInt. J. Colorectal Dis.\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 2059\u0026ndash;2067 (2019).\u003c/li\u003e\n\u003cli\u003eKhodaei, B., Seyedpour, S., Gholami, B., Garmarudi, G. \u0026amp; Nasimi, M. Seasonal and gender variation in skin disease: a cross-sectional study of 3120 patients at Razi hospital. \u003cem\u003eInt. J. Womens Dermatol.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 799\u0026ndash;802 (2021).\u003c/li\u003e\n\u003cli\u003eKnip, M. et al. Environmental triggers and determinants of type 1 diabetes. \u003cem\u003eDiabetes\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, S125\u0026ndash;S136 (2005).\u003c/li\u003e\n\u003cli\u003eMaboshe, W. et al. Low-dose vitamin D₃ supplementation does not affect natural regulatory T cell population but attenuates seasonal changes in T cell-produced IFN-\u0026gamma;: results from the D-SIRe2 randomized controlled trial. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 623087 (2021).\u003c/li\u003e\n\u003cli\u003eGrover, S. et al\u003cem\u003e.\u003c/em\u003e Vitamin D intake is associated with decreased risk of immune checkpoint inhibitor-induced colitis. \u003cem\u003eCancer\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 3758\u0026ndash;3767 (2020).\u003c/li\u003e\n\u003cli\u003eWalker, W. H. II, Walton, J. C., DeVries, A. C. \u0026amp; Nelson, R. J. Circadian rhythm disruption and mental health. \u003cem\u003eTransl. Psychiatry\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 28 (2020).\u003c/li\u003e\n\u003cli\u003eSemple, J. W., Rebetz, J. \u0026amp; Kapur, R. Transfusion-associated circulatory overload and transfusion-related acute lung injury. \u003cem\u003eBlood\u003c/em\u003e \u003cstrong\u003e133\u003c/strong\u003e, 1840\u0026ndash;1853 (2019).\u003c/li\u003e\n\u003cli\u003eMatsuda, H., Kuragaichi, T. \u0026amp; Sato, Y. Investigating the seasonal variation of heart failure hospitalizations and in-hospital mortality risks in Japan using a nationwide database. \u003cem\u003eJ. Cardiol.\u003c/em\u003e \u003cstrong\u003e83\u003c/strong\u003e, 236\u0026ndash;242 (2024).\u003c/li\u003e\n\u003cli\u003eMolinaro, A. \u0026amp; DeFalco, F. Empirical assessment of alternative methods for identifying seasonality in observational healthcare data. \u003cem\u003eBMC Med. Res. Methodol.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 182 (2022).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table 1","content":"\u003cp\u003eTable 1 is available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Seasons, Adverse Drug Event, Adverse Drug Reaction Reporting System, Vaccine, ARIMA, Prophet","lastPublishedDoi":"10.21203/rs.3.rs-6746128/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6746128/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe Japanese Adverse Drug Event Report (JADER) database is a spontaneous reporting system that compiles real-world data on adverse events (AEs) associated with drug use in Japan, including date of occurrence, allowing investigation of temporal trends. This study aimed to identify AEs that exhibit seasonal variation by applying time-series models\u0026mdash;the autoregressive integrated moving average (ARIMA), seasonal ARIMA (SARIMA), and Prophet\u0026mdash;to monthly AE reports from January 2005 to December 2019. Models were optimized using grid search and time-series cross-validation, and model performance was evaluated using mean absolute error, root mean squared error, and mean absolute percentage error. AEs whose seasonal models outperformed non-seasonal ones were considered likely seasonal. Sixteen AEs, including influenza, Guillain-Barr\u0026eacute; syndrome, encephalitis, and adverse reactions related to antineoplastic agents or immune checkpoint inhibitors, were identified as exhibiting potential seasonality. Some AEs showed clear peaks in specific months, while others demonstrated discrepancies between observed data and modeled seasonal components. These results indicate that certain AEs may be influenced by seasonal factors such as infectious diseases, climate, or patient behaviors. Identifying these seasonal trends can support proactive risk management, inform medication safety strategies, and assist clinicians, regulators, and researchers in strengthening pharmacovigilance and improving patient outcomes.\u003c/p\u003e","manuscriptTitle":"Seasonality in Adverse Drug Events: Time-Series Analysis of JADER Using ARIMA/SARIMA and Prophet","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-26 00:37:37","doi":"10.21203/rs.3.rs-6746128/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-10T11:06:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-03T11:17:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"14912468194703995222452264805914784109","date":"2025-10-13T08:48:08+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-10-10T12:49:33+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-07T14:24:22+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-06-03T12:57:47+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-29T21:33:44+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-05-29T21:30:23+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"02569533-1489-417c-af48-f6a63e921c05","owner":[],"postedDate":"October 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":56583235,"name":"Health sciences/Health occupations"},{"id":56583236,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-05-11T05:39:04+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-26 00:37:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6746128","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6746128","identity":"rs-6746128","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}