Dear Editor,

We read with great interest the recent report in Epidemiology & Infection by Demirdag and colleagues showing that COVID-19 non-pharmaceutical interventions (NPIs) disrupted the usual epidemiology of respiratory viruses, with heterogeneous post-relaxation trajectories across pathogens and settings [Reference Bedir Demirdag1]. These observations raise the question of whether other vaccine-preventable infections experienced comparable pandemic-era declines and post-reopening rebounds.

Mumps is an acute respiratory infection characterized by parotid swelling and pain, and may lead to complications such as meningitis, orchitis, and pancreatitis [Reference Lisha2]. In China, reported mumps cases have generally declined, particularly after large-scale vaccination began in 2008 and during periods of stringent COVID-19 control. We have collected monthly data on the number of mumps cases in China over the past 20 years and analysed the domestic incidence during the COVID-19 epidemic from the end of 2019 to the end of 2022, to investigate the changes in mumps epidemiological data under the intervention measures and the trend of the incidence after ‘re-opening’.

We collected monthly notification numbers of mumps cases from January 2010 to July 2023 from the National Health Commission of China. We incorporated the Oxford Stringency Index (SI) to quantify COVID-19 policy stringency and fitted a SARIMA-intervention model [Reference Hale3]. The model assessed the association between SI and monthly notifications, and predicted trajectories after reopening (December 2022). We also built a counterfactual SARIMA using pre-pandemic data to estimate expected notifications from December 2019 in the absence of COVID-19. Finally, we compared the actual notification incidence up to July 2023, SARIMA-intervention model prediction results, and counterfactual model prediction results to analyse the association between mumps and the strictness of COVID-19 policy interventions, as well as the feasibility of the SARIMA-intervention model to predict the mumps incidence in the post-pandemic period.

Mumps showed bimodal seasonality with an overall downward trend, with baseline shifts around 2012, 2017, and 2020 (Figure 1). In 2017, annual notifications rose unusually, with peaks increasing from 20096 to 33458. In 2020, the peak fell from 37913 to a historic low of 15211, followed by a gradual decline. The counterfactual SARIMA (1,0,1)(0,1,1)[12] tracked 2017–2020 patterns, indicating a slight upward trend. After the COVID-19 onset, observed notifications declined rapidly, diverging from counterfactual forecasts. By April 2020, observed values were 81% below counterfactual predictions (Figure 2), reaching 89% below in December 2022. Wilcoxon rank-sum tests showed significantly lower monthly notifications in 2020–2022 versus 2010–2019 (P < 0.001 each year). In the intervention model (1,1,0)(1,1,0)[12], the SI coefficient was β = −17.41 (95% CI −41.75 to 6.93; P > 0.05), indicating no statistically significant association with notifications. Notably, intervention-model predictions aligned with observed notifications during the pandemic and showed no immediate rebound after reopening, consistent with post-pandemic observations.

Figure 1. Observed mumps notifications versus SARIMA-intervention predictions and pre-pandemic counterfactual forecasts (train: 2010/2001–2019/2011 for counterfactual; forecast: from 2019/2012 to 2024/2012; observed: to 2023/2007).

Figure 2. Per cent deviation of observed notifications from counterfactual forecasts, 2020–2022: (counterfactual − observed) / counterfactual × 100%.

Our results did not replicate Y. Liu’s results on the incidence of tuberculosis. Although notifications differed significantly between pre-pandemic and pandemic periods, evidence was insufficient to link declines to SI. Neither observed nor predicted notifications rebounded after lifting restrictions, contrary to expectations. Given respiratory transmission, one might expect notifications to rise with increased population mixing; thus, we anticipated convergence towards the counterfactual after early 2023, which was not observed. China implemented a free MMR dose for children aged 18–24 months in 2008 [Reference Chengyu4], which may have contributed to subsequent reductions. However, the abnormally increased incidence in 2017 may be explained by the insufficient protection of MMR against the prevalence of new F/G genotype strains or the 5- to 7-year epidemic cycle of mumps itself [Reference Wang5]. All the above demonstrate the differences in epidemiological characteristics between mumps and TB.

Several factors may explain the lack of rebound: (i) expansion to a two-dose mumps vaccine schedule since 2020 [Reference Chengyu4], which will be able to provide long-term immunity and keep incidence at a low level; (ii) limited post-reopening observation window (8 months); (iii) unchanged reporting practices; (iv) pathogen-specific transmission dynamics distinct from TB; and (v) sustained personal protective behaviours.

In summary, mumps notifications in China did not rebound immediately after reopening, contrasting with reports for TB. Continued vaccination coverage and surveillance are warranted, and these findings may inform post-pandemic control strategies.