In this study, we conducted a modeling study on NIDs, which were categorized into four groups based on their primary mode of transmission. To evaluate the impact of PHSMs on the dynamics of these 24 NIDs, an ensemble model was employed to comparatively analyze the transmission characteristics across three distinct periods: PHSMs period I, PHSMs period II, and the epidemic period. Furthermore, we aimed to evaluate the impact of PHSMs on the dynamics of these 24 NIDs. Our findings revealed distinct seasonal patterns in the incidence of different NID categories, exhibiting variations across the studied periods. Comparative analysis of predicted and actual data identified 8 NIDs susceptible to PHSMs, including HFMD, dengue fever, Japanese encephalitis, malaria, pertussis, scarlet fever, mumps, and rubella (Fig. 5J, 5K). Notably, rubella and malaria exhibited a close-to-elimination incidence across the three periods (Fig. 5C, 5J). The decline in malaria cases was primarily attributed to stringent travel restrictions, while the decrease in rubella can be attributed to the increased coverage of MR (measles and rubella) and MMR (measles, mumps, and rubella) vaccine17, 18.
Cross-correlation analysis revealed the highest relative correlation coefficients for HFMD, dengue fever, malaria, scarlet fever, and mumps occurring within the same month, suggesting these diseases were immediately impacted following the implementation of PHSMs. This can be attributed to the immediate effectiveness of PHSMs in controlling the spread of these diseases, which are primarily transmitted through direct contact and respiratory droplets. The quick response of these diseases to PHSMs may also be related to their relatively short incubation periods, which allows for a rapid reflection of changes in transmission dynamics following the implementation of PHSMs. However, Japanese Encephalitis, categorized under susceptible NIDs, may demonstrate a larger RR due to its low monthly incidence (Fig. 4K) and the potential instability of the model, where minor modifications can induce significant RR fluctuations. Despite Rubella having a brief incubation period, the actual case count approaches nullity due to the synergistic effect of vaccines and PHSMs (Fig. 4X), rendering its RR relatively insensitive to variations in PHSMs.
The seasonal variations in NIDs observed before the pandemic can primarily be attributed to the interplay between transmission models and behavioral patterns19, 20. Respiratory diseases, such as mumps, exhibited higher prevalence during the winter21, 22. This phenomenon can be ascribed to increased indoor gatherings in enclosed spaces without physical distancing, facilitating virus transmission23. Additionally, the low humidity and temperature during winter were crucial factors increasing vulnerability to upper respiratory tract infections24. Some studies have suggested that seasonal variations can impact immune responses, potentially rendering individuals more susceptible to infections during specific times of the year25, 26.
The substantial decrease of 24 NIDs during PHSMs period I (Fig. 1B) can be attributed to a confluence of factors. Stringent PHSMs enacted by the Chinese government served as an effective barrier to interpersonal contact and disease transmission. Parallelly, the advent of the Omicron BA.2 pandemic significantly heightened public awareness and vigilance towards infectious diseases, prompting the widespread adoption of protective behaviors. The reduction in NIDs observed during this period can be partially attributed to the decreased mobility of infected individuals, coupled with an increased propensity to wear masks. Additionally, compliance with PHSMs extended even to susceptible individuals, even in the absence of government enforcement. This collective adherence significantly contributed to the observed decrease in the transmission of respiratory viruses, underscoring the effectiveness of PHSMs in the containment of infectious diseases.
However, the impact of PHSMs varied across diseases with different modes of transmission. Respiratory infectious diseases, notably HFMD and mumps, displayed a pronounced susceptibility to PHSMs27. Conversely, the impact on bloodborne and sexually transmitted diseases was more limited, consistent with previous research findings10, 11, 28. In the case of most respiratory diseases, the implementation of PHSMs significantly curtailed their transmission (Fig. 5E), reaffirming previous studies on the efficacy of PHSMs in controlling these diseases10, 28. This reduction can be attributed to the primary transmission route of these diseases, i.e., respiratory droplets, which can be effectively managed by measures such as mask-wearing, physical distancing, and improved ventilation. In contrast, bloodborne and sexually transmitted diseases, primarily transmitted through direct contact with infected bodily fluids, may not be as effectively mitigated by these measures28. It is important to note, however, that while the median of RR for AIDS, syphilis, HCV, and HBV were generally less than 1, indicating a reduced impact, this does not imply that these diseases were entirely unaffected (Fig. 5E). Interestingly, the incidence of gonorrhea increased during 2021 (Fig. 3P), potentially due to its relatively short incubation period (1–9 days)29, which resulted in quick responses to PHSMs. In contrast, diseases such as AIDS, syphilis, HCV, and HBV, which typically exhibit longer incubation periods (over a month)30, may experience a substantial delay between symptom onset and reporting.
Travel restrictions have a profound impact on the transmission dynamics of zoonotic infectious diseases, particularly those with short incubation periods and high susceptibility to imported cases, such as dengue fever and malaria31. These diseases are primarily transmitted by mosquitoes. The winter season, which creates an inhospitable environment for mosquitoes32, naturally suppresses the spread of these diseases throughout most regions of mainland China. During the study periods, stringent international travel restrictions, combined with the synchronization between the isolation period and the incubation period for imported cases of dengue fever and malaria, frequently facilitated the detection of these cases during quarantine. Consequently, local outbreaks of these diseases were markedly decreased during PHSMs periods. However, a contrasting pattern was observed in the case of brucellosis, another zoonotic infectious disease. The monthly incidence of brucellosis recorded a significant surge during PHSMs periods. This divergence can potentially be ascribed to the enforcement of various PHSMs which may have inadvertently impeded farmers' timely access to veterinary services, thereby escalating the risk of brucellosis transmission among livestock. Furthermore, the epidemic may have precipitated food shortages in rural areas, spurring an increase in home breeding or poultry farming. This surge in close human-animal interactions, particularly with brucellosis-infected animals, may have inadvertently amplified the risk of human infection.
Our study has several limitations. Firstly, we focused on 24 NIDs instead of all infectious disease, particularly leaving a gap in the study of respiratory diseases. While influenza was excluded due to its reliance on sentinel surveillance33, this does not undermine the overall conclusion of the study. Respiratory diseases continue to exhibit high incidence and are notably susceptible to PHSMs. The selection of 24 high-incidence NIDs for this study is indicative of the broader trends observed across all NIDs. Additionally, the reported incidence may be influenced by reporting biases, delays between reporting and disease onset, and data quality concerns15. These factors potentially affect the accuracy of the predictions generated by our model. Thirdly, the association between model selection and disease types remains nebulous and may be influenced by seasonality, warranting adjustments tailored to individual disease characteristics. Finally, inherent limitations of time-series models must be acknowledged. These include the inability to capture sudden, unexpected events and the direct impact of intervention measures on disease transmission. These models also rely on assumptions about data being stationary and following certain trends. Therefore, when comparing our results with those from other studies, it is essential to consider these limitations with due diligence.
Our research revealed that the implementation of PHSMs in response to various SARS-CoV-2 variants can significantly impact the transmission dynamics of most infectious diseases. Intriguingly, the relaxation of all PHSMs by the Chinese government did not trigger a significant resurgence in NIDs. Instead, widespread self-isolation practices adopted by the population after infection with the Omicron BA.2 variant restrained the transmission of other infectious diseases. This period of restricted transmission was, however, followed by a more extensive spread of diseases such as influenza and mycoplasma pneumoniae34, 35. Our findings emphasize that while PHSMs can provide an effective short-term solution in controlling the spread of infectious diseases, their long-term application may inadvertently lead to a decrease in population immunity, thus creating a potential environment conducive to large-scale outbreaks. For a more sustainable and long-term management of infectious disease transmission, it is imperative to prioritize the development and widespread implementation of effective vaccines, similar to the successful approach observed with the MMR vaccine36.