Abstract
Determining whether SARS-CoV-2 exhibits seasonality like other respiratory viruses is critical for public health planning. We evaluated whether COVID-19 rates follow a seasonal pattern using time series models. We used time series decomposition to extract the annual seasonal component of COVID-19 case, hospitalization, and mortality rates from March 2020 through December 2022 for the United States and Europe. Models were adjusted for a country-specific stringency index to account for confounding by various interventions. Despite year-round disease activity, we identified seasonal spikes in COVID-19 from approximately November through April for all outcomes and in all countries. Our results support employing annual preventative measures against SARS-CoV-2, such as administering seasonal booster vaccines in a similar timeframe as those in place for influenza. Whether certain high-risk individuals may need more than one COVID-19 vaccine booster dose each year will depend on factors like vaccine durability against severe illness and levels of year-round disease activity.
Similar content being viewed by others
Introduction
The Coronavirus Disease 2019 (COVID-19) pandemic has caused unprecedented worldwide morbidity, mortality, and social and economic disruption1. Globally, waves of infection, hospitalization, and mortality have paralleled the emergence of new variants of concern which have shown increased transmissibility2 and improved ability to evade vaccine- and infection-induced immunity3. Vaccination strategies to date have struggled to keep pace, and booster doses have been deployed to bolster protection against infection and symptomatic disease and maintain peak protection against severe disease throughout the pandemic4,5,6,7,8.
Many respiratory viruses show distinct seasonal patterns and result in waves of illness during the winter months9,10. These patterns are likely caused by a combination of host, pathogen, and environmental factors, including increased indoor activity and seasonal weather fluctuations known to impact viral stability outside the host9,11,12,13. Further, containment measures and pandemic virus variants have been shown to impact epidemic curves across various countries and could shift seasonal patterns14,15.
To date, there is still discussion about whether SARS-CoV-2 currently follows, or will follow in the future, similar seasonal patterns to other respiratory viruses16,17,18. Determining whether SARS-CoV-2 exhibits seasonal spikes in disease activity is critical for public health planning, including informing vaccination policy about the optimal timing for deploying additional doses. To help answer this urgent public health question, we used time series models to evaluate the seasonal patterns of COVID-19 cases, hospitalizations, and mortality in the United States and Europe with the objective to determine if COVID-19 follows typical respiratory virus seasonal patterns.
Methods
Primary analysis
We followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for our ecological study. Daily rates of COVID-19 cases, hospitalizations, and mortality per million population by country for the United States, all countries in the European Union, and the United Kingdom were obtained from the public-use Our World in Data (OWID) GitHub repository19,20. OWID sources case and mortality data from Johns Hopkins University21 and hospitalization data from various official country-specific sources18. Data were included from 01 Mar 2020 through 31 Dec 2022, as available by country. If daily data were missing or unavailable, data were substituted from that country’s corresponding weekly data.
Statistical analysis
We used a Prophet time series model to decompose daily country-specific time series rates of COVID-19 cases, hospitalizations, and mortality separately and adjusted for the country stringency index in an additive, linear time series22,23. In these models, we decomposed the observed data into a trend component and weekly and annual seasonal components and included the country-specific stringency index as a regressor to account for potential confounding by nonpharmaceutical interventions. The stringency index was developed by the Oxford Coronavirus Government Response Tracker project and is a composite of nine metrics including school closures, workplace closures, cancellation of public events, restrictions on public gatherings, closures of public transport, stay-at-home requirements, public information campaigns, restrictions on mobility and access to services, and international travel restrictions24. Uncertainty interval widths for computed point estimates were set to 95%, with 2000 uncertainty samples used to compute the intervals. Bayesian sampling was performed with 2000 Markov Chain Monte Carlo samples to obtain uncertainty intervals for seasonality estimates. In our models, we focused our analysis on the annual seasonal component only, which models the annual seasonal changes in observed rates after any long-term trends. The estimated rate is the sum of the long-term trend, weekly, and seasonal components, as well as the estimated effect of the country-specific stringency index. Positive values for the seasonal component indicate rates during that time of year are higher than during other parts of the year, while negative values indicate decreased rates during that season.
In addition to extracting the decomposed trend, annual seasonal component, weekly seasonal component, and the estimated effect of the stringency index, the monthly median value of the annual seasonal component was computed and displayed in a country-level heatmap to depict the months driving annual seasonal increases more clearly.
Modeling validation
To validate our modeling approach, we applied the same Prophet model (excluding the stringency index regressor) to pre-pandemic country-level influenza data from the World Health Organization25. This approach was used to assess whether the model would predict well-documented annual seasonal patterns for influenza and if those patterns matched those we documented for SARS-CoV-2. Influenza positivity rates were defined by computing the percent positive specimens of all respiratory specimens processed from 04 Oct 2009 through 19 Dec 2021 using the same countries (the United States, European Union countries, and the United Kingdom). Malta was not evaluated, as influenza data were not available from the World Health Organization.
Human subjects protection
All data used in this study were publicly available and de-identified and therefore did not constitute human subjects research per 45 CFR 46.102. Due to this, the study did not require IRB registration or review.
R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria) was used for all analyses.
Results
Collapsed across all years of our study period, our model showed distinct annual seasonality consistent across COVID-19 case (Fig. 1a), hospitalization (Fig. 1b), and mortality (Fig. 1c) rates from approximately November through April. The additional impact of the annual seasonal component for cases was most substantial in January through March, where there was an additional 848 COVID-19 cases per 1,000,000 population. The annual seasonal component for hospitalizations indicated a clear differentiation between seasons, with up to 75 additional hospitalizations per 1,000,000 from November through April. COVID-19-associated mortality showed a similar trend in the annual seasonal component, with an additional two deaths per 1,000,000 population due to the annual seasonal effect primarily occurring from November through February. These trends were consistent in all countries evaluated.
Total rates of COVID-19 cases, hospitalizations, and mortality, as well as influenza positivity, collapsed across all countries by month and year can be seen in Supplementary Figs. s1–s4. Country-specific time-series annual seasonal components not aggregated by month, along with other individual decomposed components (trend, weekly seasonality, and stringency index), can be seen in Supplementary Figs. s5–s8 (cases), Supplementary Figs. s9–s12 (hospitalizations), and Supplementary Figs. s13–s16 (mortality). Most of the positive annual seasonality values were within the typical respiratory viral season in the Northern Hemisphere and are indicated with a purple highlight in the annual seasonal component Supplementary Figs. s5, s9, and s13.
Our analyses using the exact model specifications described for rates of COVID-19 (excluding the stringency index regressor) also documented the well-established annual seasonality of pre-pandemic influenza between December and April over twelve US influenza seasons (Fig. 2), consistent with current knowledge regarding annual influenza patterns in the Northern Hemisphere temperate regions26.
Heatmap of the influenza percent positivity median annual seasonal component by month, Oct 2009 through Dec 2021. Data Source: World Health Organization25.
Discussion
Although SARS-CoV-2 continues to cause disease throughout the year, we identified seasonal spikes in COVID-19 cases, hospitalizations, and mortality from November through April across all years of the pandemic to date in the United States and Europe, a finding that is consistent with the typical months of seasonal respiratory virus epidemics in the northern hemisphere27. Our results indicate seasonal spikes are consistent with seasonal patterns seen for influenza27, respiratory syncytial virus (RSV)28, and other coronaviruses29, and are compatible with mathematical simulations of COVID-19 activity16,17.
There are many possible reasons for the seasonality of respiratory viruses, including climate-related changes in viral transmissibility, modified host factors (e.g., waning of infection- or vaccine-induced immunity), and changes in human behavior during the winter months9,18. Regardless of the mechanisms, knowledge of pathogen seasonality is imperative for instituting targeted interventions to lessen the impact when the burden on our healthcare infrastructure is the greatest. Accordingly, our findings have important vaccine policy implications. Additional doses of COVID-19 vaccines or modified versions of the vaccines administered before the winter months will likely have the most significant public health impact on the COVID-19 burden. This is analogous to providing influenza vaccine before peak flu activity each year to mitigate the largest spikes in disease burden. Despite evidence that protection provided by current mRNA COVID-19 vaccines wanes significantly against omicron infection and symptomatic disease after only 3–4 months, even after a booster5,8,30, this short-term added protection could still provide meaningful defense against SARS-CoV-2 infection if deployed just before seasonal waves which last 3–4 months on average.
Because SARS-CoV-2 appears to be more transmissible than influenza and other seasonal respiratory viruses, it seems likely that year-round SARS-CoV-2 activity will remain elevated compared to other pathogens31. COVID-19 continues to cause substantial morbidity and mortality throughout the year, including outside of the traditional viral respiratory season. In addition, rapid evolution of new variants or subvariants could impact seasonal patterns. Our data showed smaller waves of COVID-19 in the summer months, which were likely driven by new variants that emerged during this time period over the course of the pandemic (i.e., the delta variant in summer 2021 and the omicron subvariant BA.4/5 in summer 2022)32. Novel variants or subvariants that exhibit enhanced immune escape or transmissibility or any other property that increases viral fitness could alter seasonal patterns or cause an off-season outbreak33. Thus, additional COVID-19 vaccine booster doses may be needed at a frequency greater than once annually for certain high-risk individuals. This determination will be a careful balance between epidemiological, benefit-risk, and programmatic considerations (including concerns regarding “booster fatigue”34,35) moving forward and will likely depend primarily on COVID-19 vaccine durability against severe illness and levels of year-round disease activity. For this reason, continued surveillance of real-time vaccine performance and the emergence of new variants remains critical.
It should be noted that other viral pathogens also have, on occasion, followed atypical seasonal patterns. For example, the 2009 H1N1 influenza pandemic began in the spring of 2009 toward the end of the typical influenza season in the Northern Hemisphere36. A pandemic was declared on June 11, 2009 and cases peaked in July. It was not until the second autumn wave that disease patterns became more aligned with the typical influenza season36. In addition, pandemic containment measures can impact seasonal trends. For example, during the COVID-19 pandemic, both influenza and respiratory syncytial virus transmission followed atypical patterns37. Further, at the time of writing, China is currently experiencing a large wave of COVID-19 that likely corresponds with the lifting of country-wide lockdown measures38 Vaccination could also shift the seasonality of respiratory viruses, however, this has not occurred for influenza, the only other respiratory virus for which vaccination is available and uptake is high.
Our methodology also detected the annual seasonality of influenza virus in the same countries, corresponding to known annual seasonal patterns of influenza27,28, underscoring the utility of the methodology we used for detecting seasonal patterns in common respiratory viruses. Regardless, our results have at least five limitations. First, we could not account for the potential underreporting of cases, which may have a large effect more recently with increases in at-home SARS-CoV-2 testing that may not be reported39. Finding similar results for COVID-19 hospitalizations and deaths, which are less likely to be under-reported, however, was reassuring. Second, statistical modeling may not fully reflect the intricacies of preventing transmissible infectious diseases, such as the impact of COVID-19 vaccination, waning immunity, or changes in testing, nonpharmaceutical interventions, or healthcare-seeking behavior over time. Third, although the pandemic is in its third year, the longitudinal data available for modeling was limited compared to other common seasonal viruses. Because of this, similar models created in the future may illustrate different outputs given variable prevention behaviors, vaccines and vaccine uptake, and novel SARS-CoV-2 variants. Fourth, our findings are not generalizable beyond the United States and Europe. More research is needed to understand if the same annual seasonal patterns in SARS-CoV-2 activity are seen in the Southern Hemisphere or Asia–Pacific regions. Finally, with SARS-CoV-2, there is always the potential for new variants to emerge that could meaningfully escape prior vaccine- or infection-induced immunity and cause significant epidemics outside of regular seasonal patterns identified thus far in the pandemic. Therefore, the public health community should continue to plan and maintain the capability for sufficient response in the event of this possibility.
Conclusion
Our study suggests that COVID-19 activity and associated hospitalization and mortality in the United States and Europe peak during the traditional winter viral respiratory season despite continual transmission throughout the year. Thus, employing annual protective measures against SARS-CoV-2 such as administering seasonal booster vaccines or other non-pharmaceutical interventions for the general population in a similar timeframe as those already in place for influenza prevention (i.e., beginning in early autumn) is a prudent strategy to stay ahead of likely forthcoming seasonal waves of COVID-19. However, whether certain high-risk individuals may need more than one booster dose each year will depend on factors like vaccine durability against severe illness and levels of year-round disease activity. Additional confirmatory studies are needed, including those conducted in the Southern Hemisphere and other regions outside the United States and Europe.
Data availability
All data utilized in this study are open and available to the public, as referenced in the manuscript. Reasonable requests for programming scripts can be made to the corresponding author.
References
World Health Organization. COVID-19 Weekly Epidemiological Update. 1–10 (2022).
Centers for Disease Control and Prevention. COVID-19: Omicron Variant: What You Need to Know, https://www.cdc.gov/coronavirus/2019-ncov/variants/omicron-variant.html.
Thorne, L. G. et al. Evolution of enhanced innate immune evasion by SARS-CoV-2. Nature 602, 487–495. https://doi.org/10.1038/s41586-021-04352-y (2022).
Bar-On, Y. M. et al. Protection by a fourth dose of BNT162b2 against omicron in Israel. N. Engl. J. Med. https://doi.org/10.1056/NEJMoa2201570 (2022).
Ferdinands, J. M. et al. Waning 2-dose and 3-dose effectiveness of mRNA vaccines against COVID-19-associated emergency department and urgent care encounters and hospitalizations among adults during periods of delta and omicron variant predominance—VISION network, 10 states, August 2021–January 2022. MMWR Morb. Mortal Wkly. Rep. 71, 255–263. https://doi.org/10.15585/mmwr.mm7107e2 (2022).
Tartof, S. Y. et al. Effectiveness of a third dose of BNT162b2 mRNA COVID-19 vaccine in a large US health system: A retrospective cohort study. Lancet Reg. Health Am. https://doi.org/10.1016/j.lana.2022.100198 (2022).
Thompson, M. G. et al. Effectiveness of a third dose of mRNA vaccines against COVID-19-associated emergency department and urgent care encounters and hospitalizations among adults during periods of delta and omicron variant predominance—VISION Network, 10 States, August 2021–January 2022. MMWR Morb. Mortal Wkly. Rep. 71, 139–145. https://doi.org/10.15585/mmwr.mm7104e3 (2022).
Tartof, S. Y. et al. Immunocompromise and durability of BNT162b2 vaccine against severe outcomes due to omicron and delta variants. Lancet Respir. Med. https://doi.org/10.1016/S2213-2600(22)00170-9 (2022).
Moriyama, M., Hugentobler, W. J. & Iwasaki, A. Seasonality of respiratory viral infections. Annu. Rev. Virol. 7, 83–101. https://doi.org/10.1146/annurev-virology-012420-022445 (2020).
Hawkes, M. T. et al. Seasonality of respiratory viruses at northern latitudes. JAMA Netw. Open 4, e2124650. https://doi.org/10.1001/jamanetworkopen.2021.24650 (2021).
Biryukov, J. et al. Increasing temperature and relative humidity accelerates inactivation of SARS-CoV-2 on surfaces. mSphere https://doi.org/10.1128/mSphere.00441-20 (2020).
Grohskopf, L. A. et al. Prevention and control of seasonal influenza with vaccines: Recommendations of the advisory committee on immunization practices, United States, 2021–22 influenza season. MMWR Recomm. Rep. 70, 1–28. https://doi.org/10.15585/mmwr.rr7005a1 (2021).
Lowen, A. C. & Steel, J. Roles of humidity and temperature in shaping influenza seasonality. J. Virol. 88, 7692–7695. https://doi.org/10.1128/JVI.03544-13 (2014).
Cascini, F. et al. How health systems approached respiratory viral pandemics over time: A systematic review. BMJ Glob. Health https://doi.org/10.1136/bmjgh-2020-003677 (2020).
Cascini, F. et al. A cross-country comparison of Covid-19 containment measures and their effects on the epidemic curves. BMC Public Health 22, 1765. https://doi.org/10.1186/s12889-022-14088-7 (2022).
Liu, X. et al. The role of seasonality in the spread of COVID-19 pandemic. Environ. Res. 195, 110874. https://doi.org/10.1016/j.envres.2021.110874 (2021).
Merow, C. & Urban, M. C. Seasonality and uncertainty in global COVID-19 growth rates. Proc. Natl. Acad. Sci. U. S. A. 117, 27456–27464. https://doi.org/10.1073/pnas.2008590117 (2020).
Choi, Y. W., Tuel, A. & Eltahir, E. A. B. On the environmental determinants of COVID-19 seasonality. Geohealth 5, e2021GH000413. https://doi.org/10.1029/2021GH000413 (2021).
Our World in Data. Data on COVID-19 (coronavirus) by Our World in Data, https://covid.ourworldindata.org/data/owid-covid-data.csv (2022).
Hannah Ritchie, D. B. E. M. L. R.-G. C. A. C. G. E. O.-O. J. H. B. M. & Roser, M. Coronavirus Pandemic (COVID-19). Our World in Data (2020).
University, J. H. Coronavirus Resource Center, https://coronavirus.jhu.edu/map.html (2022).
Taylor, S. & Letham, B. Forecasting at scale. Peer J. Preprints 5, e31903192. https://doi.org/10.7287/peerj.preprints.3190v2 (2017).
prophet: Automatic Forecasting Procedure v. 1.0 (2021).
Hale, T. et al. A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker). Nat. Hum. Behav. 5, 529–538. https://doi.org/10.1038/s41562-021-01079-8 (2021).
World Health Organization. World Health Organization FLUMART Outputs, https://apps.who.int/flumart/Default?ReportNo=12 (2022).
Centers for Disease Control and Prevention. Influenza (flu): Flu Season, https://www.cdc.gov/flu/about/season/flu-season.htm.
Lofgren, E., Fefferman, N. H., Naumov, Y. N., Gorski, J. & Naumova, E. N. Influenza seasonality: Underlying causes and modeling theories. J. Virol. 81, 5429–5436. https://doi.org/10.1128/JVI.01680-06 (2007).
Chadha, M. et al. Human respiratory syncytial virus and influenza seasonality patterns-Early findings from the WHO global respiratory syncytial virus surveillance. Influenza Other Respir. Viruses 14, 638–646. https://doi.org/10.1111/irv.12726 (2020).
Nichols, G. L. et al. Coronavirus seasonality, respiratory infections and weather. BMC Infect. Dis. 21, 1101. https://doi.org/10.1186/s12879-021-06785-2 (2021).
Tartof, S. Y. et al. BNT162b2 (Pfizer–Biontech) mRNA COVID-19 vaccine against omicron-related hospital and emergency department admission in a large US health system: A test-negative design. SSRN Pre Print (2022).
Centers for Disease Control and Prevention. Influenza (Flu): Similarities and Differences betwen Flu and COVID-19, https://www.cdc.gov/flu/symptoms/flu-vs-covid19.htm (2022).
Wiemken, T. & Clarke, J. SARS-CoV-2 Variant Tracker, https://surveillance.shinyapps.io/variants/.
Wang, Q. et al. Antibody evasion by SARS-CoV-2 Omicron subvariants BA.2.12.1, BA.4, & BA.5. Nature (2022). https://doi.org/10.1038/s41586-022-05053-w
University of Minnesota Center for Infectious Diseases Research and Policy. CDC advisers discuss future of COVID-19 booster shots, https://www.cidrap.umn.edu/news-perspective/2022/04/cdc-advisers-discuss-future-covid-19-booster-shots (2022).
Oliver, S. Framework for future doses of COVID-19 vaccine doses and next steps, https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2022-04-20/07-COVID-Oliver-508.pdf (2022).
Centers for Disease Control and Prevention. 2009 H1N1 Pandemic Timeline, https://www.cdc.gov/flu/pandemic-resources/2009-pandemic-timeline.html.
World Health Organization. Joint statement—Influenza season epidemic kicks off early in Europe as concerns over RSV rise and COVID-19 is still a threat. https://www.who.int/europe/news/item/01-12-2022-joint-statement---influenza-season-epidemic-kicks-off-early-in-europe-as-concerns-over-rsv-rise-and-covid-19-is-still-a-threat.
World Health Organization. China, https://covid19.who.int/region/wpro/country/cn. (2023).
Rader, B. et al. Use of at-home COVID-19 tests—United States, August 23, 2021-March 12, 2022. MMWR Morb. Mortal Wkly. Rep. 71, 489–494. https://doi.org/10.15585/mmwr.mm7113e1 (2022).
Funding
This study was sponsored by Pfizer.
Author information
Authors and Affiliations
Contributions
Dr. T.L.W. had full access to all the data in the study and took responsibility for the integrity and data analysis accuracy. All authors had full access to all data and accepted responsibility for submitting it for publication. Concept and design: all authors. Accessed and verified data and analysis: T.L.W., F.K. Interpretation of data: all authors. Drafting of the manuscript: T.L.W., J.M.M. Critical revision of the manuscript for important intellectual content: all authors. Obtained funding: N/A. Administrative, technical, or material support: all authors. Supervision: J.M.M.
Corresponding authors
Ethics declarations
Competing interests
TLW, LP, JMM, LJ, JN, and FK are employees and shareholders of Pfizer Inc.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Wiemken, T.L., Khan, F., Puzniak, L. et al. Seasonal trends in COVID-19 cases, hospitalizations, and mortality in the United States and Europe. Sci Rep 13, 3886 (2023). https://doi.org/10.1038/s41598-023-31057-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-31057-1
- Springer Nature Limited
This article is cited by
-
COVID-19 pandemic in Taiz Governorate, Yemen, between 2020 and 2023
BMC Infectious Diseases (2024)
-
An intranasal combination vaccine induces systemic and mucosal immunity against COVID-19 and influenza
npj Vaccines (2024)
-
Urban Wastewater-Based Surveillance of SARS-CoV-2 Virus: A Two-Year Study Conducted in City of Patras, Greece
Food and Environmental Virology (2024)
-
Strong biological correlations as a cause of autonomous oscillations in epidemics
Journal of Mathematical Biology (2023)
-
The Structure of Pandemic Vulnerability: Housing Wealth, Residential Segregation, and COVID-19 Mortality
Population Research and Policy Review (2023)