Abstract
Background: Respiratory infectious diseases represent the highest incidence of infectious diseases globally, caused by a diverse array of pathogens. Current monitoring often focuses narrowly on influenza and coronaviruses. This study aimed to establish the pathogen spectrum of local acute respiratory infections in Nanjing and investigate pathogen co-infection dynamics. Time series models were employed to predict disease outbreaks and serve as tools for surveillance and response strategy formulation.
Methods: From June 2023 to February 2024, 2,046 influenza-like illness (ILI) samples were collected weekly from the community in Xuanwu District, Nanjing. We established a respiratory pathogen spectrum and analyzed age distribution and symptom counts. We compared demographics, symptoms, and viral loads between co-infections and single infections. An autoregressive integrated moving average (ARIMA) model was constructed to predict future incidence.
Results: The total detection rate was 53.57% (1096/2046). Dominant pathogens included influenza A (24.00%), influenza B (10.95%), and HCoV (4.64%). Specific pathogens showed statistically significant associations with age and symptom counts. Mixed infections occurred in 6.11% of cases, though no significant difference in age or symptom count was found between co-infection and simple infection. The optimal ARIMA model (0,1,4)(0,0,0) achieved an R² of 0.930, indicating strong predictive performance.
Conclusions: The respiratory pathogen spectrum in Nanjing is complex, with distinct age-specific patterns. Co-infection dynamics did not correlate with patient age or gender. The ARIMA model provides a reliable estimate for future incidence trends, which plateaued following the winter peak.
Introduction
The landscape of respiratory infectious diseases underwent a dramatic shift following the implementation and subsequent lifting of non-pharmaceutical interventions (NPIs) during the COVID-19 pandemic. Measures such as masking and social distancing created an "immunity debt," leading to a significant resurgence of common pathogens once restrictions were eased. While global surveillance often focuses narrowly on Influenza and SARS-CoV-2, a broader understanding of the pathogen spectrum is essential for effective public health response in the post-NPI era. This study characterizes the dynamics of 16 respiratory pathogens in Nanjing, China, following the cessation of "Zero-COVID" policies.
Methods
This prospective study was conducted in the Xuanwu District of Nanjing from June 2023 to February 2024. Weekly throat swabs were collected from 2,046 community members presenting with influenza-like illness (ILI) and fever (≥ 37°C). Samples were analyzed using multiplex fluorescence quantitative PCR to detect 16 pathogens, including influenza strains, coronaviruses (HCoV), rhinoviruses (HRV), and bacteria such as S. pneumoniae. For Enteroviruses (EV) and Rhinoviruses, additional sequencing was performed for genotype identification. Finally, an ARIMA time-series approach was used to capture trends and seasonality to predict future infection rates based on historical weekly data.
Results
Demographic and Clinical Profile
The study population spanned a wide age range, with a mean age of 36.8 ± 20.6 years. The majority of participants fell within the 14–59 age group (67.12%). Clinical symptoms were diverse, with cough being the most universal (79.60%), followed by sore throat (58.29%) and fatigue (46.99%).
Pathogen Prevalence and Age Distribution
The overall pathogen detection rate was a staggering 53.57%, significantly higher than the ~11% observed in similar studies during the height of COVID-19 restrictions. Influenza A (specifically the H3N2 subtype) was the primary driver of community fever, followed by Influenza B and seasonal Coronaviruses.
Pathogen distribution varied significantly by age group. M. pneumoniae and Adenovirus (HADV) were significantly more common in children under 14, while the 14–59 age group bore the brunt of Influenza A, Influenza B, and HCoV infections.
Co-infection and Modeling
Mixed infections were detected in 6.11% of cases (125/2046). The most frequent combination involved Influenza A and Rhinovirus. Interestingly, co-infection did not significantly correlate with age, gender, or a higher number of symptoms, suggesting that "double sickness" does not necessarily translate to increased clinical severity in a community setting.
The ARIMA (0,1,4)(0,0,0) model proved remarkably robust. With an R² of 0.930, the model's predictions for late 2023 and early 2024 closely matched actual case counts, successfully capturing the winter peak and the subsequent plateau.
Discussion
The resurgence of respiratory pathogens in Nanjing underscores the impact of the "immunity debt" following the lifting of pandemic restrictions. The dominance of Influenza A (H3N2) and the age-specific niches of pathogens like M. pneumoniae provide critical data for local clinical diagnostics. Furthermore, the robust performance of the ARIMA model highlights its potential as a high-precision tool for early warning systems and hospital resource planning. While limited by its nine-month duration and single-district focus, this study establishes a vital baseline for long-term respiratory disease surveillance in the post-pandemic era.
Summary updated