Trend dynamics of mumps prevalence among children aged 1–9 years in Changzhou, 2005–2023: a joinpoint regression and age-period-cohort analysis

Abstract

Background

Despite the inclusion of one dose of measles-mumps-rubella (MMR) vaccination in the Expanded Programme on Immunization (EPI), mumps cases and outbreaks continue to occur. The objective of this study is to explore the epidemiologic pattern of mumps in Changzhou from 2005 to 2023 and to predict the future trend of mumps.

Methods

A Joinpoint regression model was employed to ascertain the temporal trends in mumps prevalence. To interpret the age, period, and cohort effects on mumps prevalence, an age-period-cohort (APC) analysis was employed. Bayesian age-period-cohort (BAPC) models were utilized to project the future prevalence rate from 2024 to 2033.

Results

One join-point was identified, with an upward trend from 2005 to 2012 (APC = 13.880) and a downward trend from 2012 to 2023 (APC = -8.262). The peak age-specific prevalence rate occurred at five to six years of age. From 2011 to 2013, the prevalence of mumps was found to be significantly elevated. In the cohort born between 2005 and 2010, the risk of mumps remained stable, while Subsequent birth cohorts exhibited fluctuating and increasing trends. The BAPC model predicted a gradual decline in prevalence rate over the next 10 years with the two-dose MMR vaccine.

Conclusions

The age-specific prevalence rate of mumps among children aged 1–9 years in Changzhou has not declined significantly, and the efficacy of the single-dose MMR vaccine is limited. This necessitates the timely optimization of the immunization strategy.

Introduction

Mumps is a common childhood infectious disease caused by infection with the mumps virus (MuV). The mumps virus is primarily transmitted through respiratory droplets or direct contact with respiratory secretions [1, 2]. The clinical manifestations of the disease are characterized by pain and swelling of the parotid gland. However, it can also involve various other tissues and organs, including testicular inflammation in post-pubertal males, mastitis and tubulitis in post-pubertal females, pancreatitis, and serious complications such as encephalitis, meningitis, myocarditis, and nephritis [1, 3, 4]. Although mumps is a benign condition that typically results in complete recovery within a few weeks of infection, long-term sequelae such as seizures, cranial nerve palsy, hydrocephalus, and deafness may occur [5, 6].

Mumps is a vaccine-preventable disease; therefore, immunization with mumps-containing vaccines (MuCV) is considered to be one of the main means of preventing mumps infections [7]. MuCV were originally produced as monovalent vaccines, but nowadays they are primarily combination vaccines, including the measles-mumps vaccine (MM), the measles-mumps-rubella vaccine (MMR), and the measles-mumps-rubella-varicella vaccine (MMRV) [6]. To date, 122 (63%) of the 194 member countries of the World Health Organization (WHO) have incorporated at least one dose of MuCV into their national immunization programs [3, 4]. Globally, the first dose of MuCV is typically administered at 12 to 18 months of age, with the second dose occurring between 2 and 6 years of age [8]. The recommended timing of the second dose of MuCV in each country is typically influenced by the prevalence of mumps. There is considerable variation in the timing of the second dose, with approximately one-third (48/145) of countries administering the second dose at 10–23 months of age [9]. In 1996, China initiated the use of the MMR vaccine, primarily the S79 strain mumps vaccine, which was derived from the Jeryl-Lynn strain in the United States. In 2008, China incorporated the MMR vaccine into the Expanded Programme on Immunization (EPI), providing one complimentary dose of MMR vaccine to children at 18 months of age. In June 2020, the combined measles-rubella vaccine, administered at eight months of age, was adjusted to MMR vaccine. School-age children received a total of two doses of MMR vaccine at eight and 18 months of age, respectively [10].

Mumps has long coexisted with a prevalence rate of over 10 per 100,000 in China, with localised outbreaks occurring annually [11], resulting in a severe disease burden. It is of significant interest to ascertain whether the incorporation of MuCV into the EPI would disrupt the natural pattern of mumps epidemics and effectively control the scale of the disease. The objective of this study is to estimate the effects of age, period, and cohort on mumps prevalence and predict the trend of mumps prevalence over the next 10 years. The findings will facilitate the identification of the epidemiological pattern of mumps in Changzhou, the improvement of public health, and the rational allocation of medical resources. Furthermore, the results will provide a foundation for the refinement and updating of MMR immunization strategies.

Methods

Data sources

The age-specific prevalence rate of mumps in children aged 1–9 years from 2005 to 2023 was obtained from the China Information System for Disease Control and Prevention, along with population information for each year. The standard population was derived from China’s Seventh Census data, which represents the most recent and comprehensive demographic statistics available in China, reflecting the latest changes in population structure (https://www.stats.gov.cn/sj/pcsj/). The demographic data for Changzhou for the Subsequent decade was replaced by the average population from 2005 to 2023.

Calculation of evaluation indicators

Two evaluation indicators, the age-standardized incidence rate (ASR) and the estimated annual percentage change (EAPC), were calculated to assess the trends in mumps prevalence, respectively [12–14]. Given the disparate demographic profiles (e.g., age and sex compositions) among the study population, the direct method was employed for the standardization. In this method, refers to the age-specific rates, refers to the age subgroup, and refers to the number of persons in the age subgroup in the standard population,

The EAPC is a quantitative measure of the trends in ASR over a specific time interval [15]. A regression line was fitted to the natural logarithm of the rates, specifically , where and = calendar year. The EAPC is calculated as the regression coefficient (β) minus 1, which is then multiplied by 100. If both the EAPC and its 95% confidence interval (CI) were greater than 0, the ASR was considered to have an upward trend. Conversely, if both the EAPC and its 95% CI were less than 0, the ASR was deemed to have a downward trend. Otherwise, the ASR was considered to be stable over time.

Statistical analysis

Joinpoint regression analysis

The Joinpoint regression model is, in fact, a segmented linear regression model, which is a collection of linear statistical models that evaluates the trend of mumps prevalence over time by identifying statistically significant turning points. The model identified the joinpoints by means of the least squares method (the sum of the squared residuals between the estimated and actual values) and selected the optimal number of joinpoints by means of a permutation test, thus avoiding the typical linear-based non-objectivity of trend analysis [16, 17]. By fitting a regression to the natural logarithm of the prevalence rate of mumps in different segments, it was possible to derive the annual percentage change (APC) and its 95%CI for each period [18]. The APC is the estimated annual percentage rate of change from one joinpoint to the next. It is primarily used to determine the direction and magnitude of the trend [19]. When APC is greater than zero, it indicates that the prevalence rate increased by year. Conversely, when APC is less than zero, it indicates that the prevalence rate decreased by year. The test of significance was conducted using a Monte Carlo permutation method, with statistical significance set at P < 0.05 [20].

Age-period-cohort model

The APC model, a parametric statistical model based on the Poisson distribution [21], is a commonly utilized tool in sociology and epidemiology. It decomposes changes in mumps prevalence into three dimensions: age, period and cohort. This allows for the analysis of the influence of each factor on the trend of changes in mumps prevalence. Holford proposed the APC model in 1983, which can be expressed as follows: (λ represents age group prevalence, µ represents the intercept, α, β, and γ represent age, period, and cohort bias, respectively, and ε represents random error) [22]. The complete linear correlation between age, period, and cohort precludes the possibility of obtaining unique estimates of the parameters, also known as the non-identification problem [23]. However, estimable functions of the APC parameters have recently emerged as a useful approach for obtaining relatively efficient estimates [21]. The study focused on the following estimable parameters and functions: net drift, local drifts, fitted longitudinal age-specific rates in reference cohort, period rate ratios and cohort rate ratios. Moreover, in accordance with the specifications of the modeling requirements, the age and period intervals must be identical. The study employed one-year intervals spanning from the year 2005 to 2023, with 19 periods (2005–2023), 9 age groups (1–9 years) and 27 birth cohorts (1996–2022). The center age group (5 years), calendar year (2014), and birth cohort (2009) were used as references in this study. Parametric hypotheses were tested using the Waldχ² test with a test level of α = 0.05.

Bayesian age-period-cohort models

BAPC have become a widely utilized methodology for the analysis and projection of age-specific cancer prevalence and mortality, as they are capable of circumventing the necessity for robust parametric assumptions concerning future values of period and cohort effects [24]. The advantages of the BAPC prediction model are as follows: (1) The BAPC assumes that adjacent time effects are similar and attributed to the prior probability distribution. This approach serves to weaken the effects of age, period, and cohort. It also avoids the identifiability issues associated with the APC model. Consequently, the method produces more reasonable and reliable estimates and predictions, as evidenced by the findings of studies [25, 26]. (2) The use of the Integrated Nested Laplace Approximation (INLA) technique to approximate the marginal posterior distribution enables the avoidance of the complex convergence problem that is characteristic of Markov-chain Monte Carlo (MCMC) methods [24, 27]. (3) The generation of standardized prediction rates and age-specific prediction rate results can provide predicted values at different percentiles, thus enabling a more comprehensive assessment of the burden of disease.

Statistical software and others

Three-dimensional Surface plots of mumps prevalence from 2005 to 2023 were generated using the griddata interpolation function of Matlab software (version r2024a, MathWorks, Natick, MA, USA). Joinpoint analysis was conducted using the Joinpoint Regression 4.9 software (Statistical Research and Applications Branch, National Cancer Institute, USA). The APC model was established through the utilisation of an online web-based tool (https://analysistools.cancer.gov/apc/) provided by the National Cancer Institute, which employs the R software as a foundation. Descriptive statistics, Spearman’s correlation analysis, BAPC predictive modeling and plotting were conducted using R4.4.0 software (R core team), with the “ggplot2,” “INLA,” and “BAPC” packages being employed primarily.

Results

The prevalence of mumps in Changzhou

The total number of mumps cases among children aged 1–9 years in Changzhou from 2005 to 2023 was 6,766, of which 4,299 were male and 2,467 female, representing 63.54% and 36.46% respectively. The prevalence of mumps exhibited a cyclical fluctuation, with an annual average prevalence rate of 95.903/100,000 (see Fig. 1). The crude prevalence rate decreased from 69.037 per 100,000 in 2005 to 45.646 per 100,000, while the standardized incidence rate decreased from 81.644 per 100,000 in 2005 to 44.999 per 100,000. The prevalence by gender was consistent with the total, with all 2023 mumps prevalence rates lower than in 2005. Table 1 presents the EAPC and 95%CIs for mumps prevalence for the years 2005–2023. The data indicated a downward trend in mumps prevalence for the overall population, as well as for males and females, over the past two decades. Nevertheless, the decline in prevalence was not statistically significant. With regard to age group, an increasing trend was observed in the younger age group (1–4 years), while a decreasing trend was observed in the older age group (5–9 years).

Fig. 1.

Three-dimensional surface plots of mumps prevalence

Table 1.

Prevalence rates for children aged 1–9 years, 2005–2023

Characteristics	Average number of cases	Average prevalence rate	2005		2023		2005–2023
			Incident cases	Rate per 100,000(ASR)	Incident cases	Rate per 100,000(ASR)	EAPC(95%CI)
Total	357.421	95.903	236	69.037(81.644)	188	45.646(44.999)	−1.940(−4.951 to 1.167)
Male	227.053	116.301	165	92.722(110.983)	116	53.638(53.324)	−2.822(−5.646 to 0.086)
Female	130.368	73.564	71	43.320(49.580)	72	36.809(35.833)	−0.373(−0.373 to 3.225)
Age group
1	7.474	19.813	2	4.297	5	17.2939	5.154(−3.270 to 14.312)
2	14.526	36.935	10	20.695	11	32.1619	2.199(−3.149 to 7.843)
3	34.842	90.252	17	34.269	18	43.6639	6.575(1.978 to 11.380)
4	49.263	127.901	15	55.494	37	83.9554	4.762(0.162 to 9.574)
5	51.632	142.388	42	157.901	25	46.7045	−0.657(−4.735 to 3.595)
6	55.263	154.793	39	134.790	39	70.1843	−1.904(−4.737 to 1.013)
7	50.579	114.818	46	146.464	18	36.5742	−5.039(−8.774 to −1.151)
8	51.316	115.627	36	97.808	24	45.2105	−6.358(−11.079 to −1.386)
9	41.211	124.512	29	62.234	11	21.1287	−6.810(−12.131 to −1.167)

The correlation between EAPC and age was analyzed in Fig. 2, which revealed that EAPC increased with age up to three years of age, followed by a gradual decrease in EAPC with age from three to nine years of age. In the correlation analysis, EAPC was found to be significantly negatively correlated with age in the total population (r_s = −0.917, P = 0.001), as well as in the male population (r_s = −0.850, P = 0.006). However, in the female population, the correlation was not statistically significant, although it was negatively correlated with age (r_s = −0.55, P = 0.133).

Fig. 2.

Correlation analysis chart of the EAPC and age

Analysis of the prevalence trend of mumps

The Joinpoint regression model was employed to identify the trend in the prevalence of mumps across distinct temporal periods, along with the estimation of the APC and its associated 95%CI (Table 2). After taking the natural logarithm of the prevalence rates for the total population and for males and females, the Shapiro-Wilk test indicated that the prevalence rates were normally distributed (W_total = 0.975, W_male = 0.978, W_female = 0.953, with P values all greater than 0.05). A statistically significant turning point was observed in 2012 among both the total population and the different sexes. The highest crude prevalence rates were observed in 2012 (194.984/100,000 in the total population, 222.915/100,000 in males, and 163.085/100,000 in females). In the total population, an initial increase in prevalence rates was observed, although this trend was not statistically significant (APC_2005−2012 = 13.880, 95% CI: −0.130 to 29.855). This was subsequently followed by a statistically significant downward trend (APC_2012−2023 = −8.262, 95% CI: −13.607 to −2.586). The prevalence rates for both sexes also showed a tendency to increase and then decrease, with a greater magnitude of change observed in females than in males.

Table 2.

APC values for different time periods for children aged 1–9 years, 2005–2023

Characteristics	2005–2012	2012–2023	SSE	MSE
Total	13.880(−0.130 to 29.855)	−8.262(−13.607 to −2.586)	477.372	31.825
Male	10.673(−2.993 to 26.264)	−7.976(−13.639 to −1.942)	319.893	21.326
Female	19.713(4.442 to 37.215)	−8.634(−13.670 to −3.304)	172.266	11.684

SSE Sum of Squared Errors; MSE Mean Squared Error

In addition, we analyzed the trend of crude mumps prevalence rates in various age Subgroups. The majority of age groups exhibited a single joinpoint, with prevalence rates reaching a peak in 2012 and subsequently declining significantly. A few age groups (7- and 8-year-olds) demonstrated an overall declining trend without a joinpoint, as illustrated in Fig. 3.

Fig. 3.

Results of Joinpoint regression model for mumps prevalence

Analysis of the APC model

The results of the APC model test and the age, period, and cohort effects are presented in Table 3; Fig. 4, respectively. As can be observed, the net drift values were not statistically significant in the total population and in the male and female populations (net drift of −0.492, −0.952, and 0.084, respectively). This suggests that the overall temporal trend in prevalence did not decrease significantly (P > 0.05). In contrast, the age, period, and cohort deviations, as well as the RR values, were found to be statistically significant after controlling for age and period as well as cohort effects, respectively. The chi-square results indicated that the local drifts were not equal to the net drifts (P < 0.05). The trend of each age group differed significantly. Absolute values of local drift greater than 1% were considered to indicate a Substantial change, with the prevalence rising annually in children aged 1–4 years (local drift greater than 1%) and declining annually in children aged 6–9 years (local drift less than − 1%). The trends for each age group by gender were also consistent with the overall population.

Table 3.

APC model test for children aged 1–9 years, 2005–2023

Hypothesis Tests	Total			Male			Female
	χ2	df	P	χ2	df	P	χ2	df	P
Net Drift = 0	0.612	1	0.434	2.072	1	0.150	0.011	1	0.917
All Age Deviations = 0	188.794	7	0.000	153.702	7	0.000	131.496	7	0.000
All Period Deviations = 0	269.498	17	0.000	213.657	17	0.000	222.447	17	0.000
All Cohort Deviations = 0	133.214	25	0.000	126.489	25	0.000	83.722	25	0.000
All Period RR = 1	289.463	18	0.000	236.310	18	0.000	230.921	18	0.000
All Cohort RR = 1	148.413	26	0.000	153.598	26	0.000	84.389	26	0.000
All Local Drifts = Net Drift	99.663	9	0.000	91.178	9	0.000	61.298	9	0.000

Fig. 4.

Parameter estimates of age, period, and cohort effects on mumps prevalence among children aged 1–9 years in Changzhou, 2005–2023. (A-C) Longitudinal age curve, period RR, cohort RR, local drift and their corresponding 95% CIs for the total population. (E-H) Longitudinal age curve, period RR, cohort RR, local drift and their corresponding 95% CIs for the male population. (I-L) Longitudinal age curve, period RR, cohort RR, local drift and their corresponding 95% CIs for the female population

The Longitudinal Age Curve demonstrated the age effect. After controlling for period and cohort effects, the peak prevalence rate was mainly centered at 5–6 years of age, with the highest prevalence rate at 6 years of age (rate = 100.801/100,000, 95% CI: 77.889 to 130.453/100,000). The age effect on the prevalence of mumps exhibited a clear trend of increasing from 1 to 6 years of age, followed by a slight decline and then a resurgence at 9 years of age, which was comparable to the prevalence observed at 3 years of age. When stratified by sex, the age effect was observed to approximate the total population. Peak prevalence rate was noted in males at the age of 6 years (rate = 129.88/100,000, 95% CI: 98.266 to 171.667/100,000) and in females at the age of 5 years (rate = 71.483/100,000, 95% CI: 52.974 to 96.459/100,000).

Period effects were primarily observed in Period RR, with a reference period of 2014. After controlling for age and cohort effects, the prevalence of mumps exhibits cyclical fluctuations, with a prevalence cycle of approximately two to three years. The prevalence of mumps reached its peak between 2011 and 2013, with the highest prevalence observed in 2012 (RR = 2.182, 95% CI: 1.659 to 2.87). After that, the prevalence rate gradually decreased following the year 2020. The period effect, when examined by gender, was found to be consistent with the overall trend.

The cohort effect was most evident in the cohort relative risk (Cohort RR), with the reference cohort being 2009. A higher risk of morbidity was observed for the 1996–2004 birth cohort in comparison to the 2009 cohort. A more stable risk of morbidity was observed in the 2005–2010 birth cohort. Since 2011, however, the risk of morbidity in the Subsequent birth cohort has been on the rise. The cohort effect in the 2019–2023 period exhibited considerable instability, with wide confidence intervals. The observed cohort effect stratified by sex was consistent with the overall findings.

Predicted trends of mumps prevalence in 2024–2033

The predicted prevalence rates of mumps per 100,000 for the total population and for males and females were calculated using the BAPC prediction method. The projected trajectory of mumps prevalence from 2024 to 2033 is shown in Fig. 5. In general, the age-standardized prevalence rates were projected to decline over the following 10-year period. The overall standardized incidence rate of mumps in Changzhou was projected to decline to 4.947/100,000 in 2033, with a rate of 12.225/100,000 in males and 0.991/100,000 in females.

Fig. 5.

BAPC model prediction results for mumps prevalence. Note: The observations are presented as dots with the predicted distributions between the 5% and 95% quartiles. The shaded bands indicate the prediction intervals in 10% increments. The predicted mean is depicted as a solid black line, and the vertical dashed line indicates the point at which the prediction begins

Discussions

The prevention and control of mumps epidemics has been an emerging public health issue. The MMR vaccination strategy has been shown to be effective in previous studies, e.g. the average annual prevalence of mumps in the United States was 100/100,000 in 1967 and was reduced to 0.1/100,000 in 1993 [28], in Denmark from 726/100,000 in 1979 to 1/100,000 in 1995 [5], and in France from 859/100,000 in 1986 to 9/100,000 in 2011 [29]. In recent years, however, a growing number of studies have found a renewed upward trend in the reported incidence of mumps. Since 2006, multiple outbreaks of mumps have been reported in the United States [30, 31], resulting in the largest outbreak of mumps in the 21 st century with 6,584 cases reported, mainly in vaccinated college students aged 18–24 years, 84% of whom had previously received two doses of MMR [32]; another large multi-state outbreak in 2009 and 2010 reported more than 3,500 cases, 89% of whom had previously received two doses of MuCV [31]. Developed countries such as Canada, Australia and the United Kingdom have also reported sporadic outbreaks of mumps [4]. The waning of immunity has been identified as the primary cause of these outbreaks. In regions where two doses of MMR have been administered extensively, it is recommended to consider the incorporation of a third dose into the local immunization programme.

In Changzhou, the MMR vaccination strategy was 1 dose of voluntary vaccination at own expense until 2008. From 2008 to 2020, Changzhou implemented 1 dose of MMR vaccination (at 18 months) and from 2020, the MMR vaccination strategy was changed to 2 doses (at 8 months and 18 months). In this study, we analysed the trend of mumps prevalence among children aged 1–9 years in Changzhou from 2005 to 2023 and found that the prevalence of mumps did not appear to have decreased significantly in the last 20 years and that the EAPC, although < 0, was not statistically significant. The joinpoint model was further used to identify trends in mumps prevalence over time, and only from 2012 did the prevalence of mumps decrease significantly. This may be due to the national mumps epidemic in 2011–2012 [33], which led to a reduction in the Susceptible population and a corresponding decrease in the Subsequent mumps prevalence, on the one hand, and the 1-dose MMR vaccination, which may have provided some protection and maintained a low level of mumps prevalence. Notably, the inclusion of the MMR in routine immunization revealed an increasing trend in the younger age group (1–4 years) and a decreasing trend in the older age group (5–9 years) through the results of the age-specific EAPC and correlation analyses, which is a rather interesting finding. Firstly, the observed increase in prevalence among younger children may be associated with delayed administration of the first MMR. When the single-dose MMR strategy was first introduced, a high proportion of children received the vaccine promptly, with the vast majority completing their MMR vaccination within one month of reaching the recommended age for the first dose. However, with the increase in the number of self-paid vaccines available for 18-month-olds in recent years, and some parents’ lack of understanding of MMR. Vaccine hesitancy regarding MMR has become increasingly prominent [34], resulting in a significant decline in timely vaccination rates and an extended periods without immunity protection among younger age groups. Consequently, the prevalence rate in this population exhibited an upward trend, underscoring the significance of prompt vaccination in establishing a robust immunization barrier. Secondly, the decline in prevalence rates among older age groups can be attributed to various factors. Following its incorporation into the national immunization programme in 2008, the MMR had not yet been widely administered to children aged 5–9 years, resulting in a high prevalence rate of mumps in this age group. Concurrently, China experienced a nationwide mumps epidemic from 2011 to 2012 [33]. We supposed that children born during this epidemic period (aged 8–9 years in 2020) would exhibit lower susceptibility than pre-epidemic birth cohorts, likely attributable to the combined effects of natural infection (including asymptomatic cases) during the outbreak and vaccination. In addition, the outbreak of COVID-19 after 2020 disrupted the prevalence patterns of MuV among school-aged children (5–9 years old), especially the widespread application of non-pharmacological interventions reduced the risk of MuV transmission.

The prevalence of mumps can be influenced by age, period and birth cohort, either individually or in combination. The age effect reflects the risk of disease in individuals of different ages; the period effect refers to the impact of innovations in diagnostic and treatment techniques, changes in MMR vaccination strategies, and disease epidemiological patterns on disease incidence; and the birth cohort effect represents the interaction between the age and period effects, and is the influence of the socio-economic environment, historical events, and lifestyle after birth on mumps prevalence. Regarding age effects, the peak prevalence rate in children aged 1–9 years was mainly concentrated in 5–6 years, which is consistent with available findings and may be related to children’s school enrolment, increased exposure to the external environment and insufficient persistence of vaccine immunity [4, 35, 36]. The prevalence is also high in the younger age groups, Suggesting that we need to target younger children for enhanced Surveillance, while focusing on improving MMR vaccine coverage and timeliness in this age group. In terms of period effect, mumps had an epidemic cycle of about 2 years. The high prevalence of mumps in 2011–2012 was mainly due to the fact that the Susceptible population was concentrated in the 5–9 year age group, the MMR vaccination strategy was only adjusted in 2008, and the updated strategy did not yet have a high coverage of 5–9 year olds, resulting in a mumps outbreak. Since 2020, both the trend and magnitude of mumps prevalence have decreased, primarily due to changes in public health behaviours during the COVID-19 pandemic, with non-pharmacological interventions such as mask use and school closures reducing the risk of mumps exposure [37]. Second, the adaptation and updating of the two-dose MMR strategy has also played an important role in increasing protective antibody levels in Susceptible populations to prevent MuV infection.Regarding birth cohort effects, the risk of birth cohort morbidity fluctuated upward during the period of the 1-dose MMR strategy (2008–2020). We considered that the protective effect of a single dose of MMR was limited and that the attenuated immune response induced by the vaccine was insufficient to provide durable protection. The RR values continued to increase after the change in vaccination strategy in 2020, but with wide confidence intervals (P > 0.05), Suggesting that we need to further strengthen disease Surveillance to determine whether the 2-dose MMR strategy can break the natural epidemic cycle of mumps and maintain low prevalence.

Predicting prevalence helps understand disease burden and informs decisions about health resource allocation. We used the BAPC model to predict 10-year prevalence trends. The age-standardized rates were relatively optimistic, indicating a downward trend. The two-dose MMR strategy was found to be instrumental in controlling the prevalence of mumps in specific populations aged 1–9 years and reducing the burden of disease. In light of the elevated risk of breakthrough infections attributable to waning vaccine-induced immunity, persistent epidemic resurgence was observed among older age groups [4, 38].In regions with sustained high coverage of two-dose MMR vaccination, the introduction of a third MMR dose may help to elicit enhanced population-level immunity, thereby effectively curbing mumps resurgence.

This study also has some limitations. First, the study was conducted only among children aged 1–9 years in Changzhou, a relatively underrepresented population whose results cannot be easily extrapolated. Second, unlike the GBD database, the future population size of Changzhou was not available. The results of the predictive model were based on the assumption that the population size would remain constant over the next decade, but a decline in the prevalence rate can still be observed. Finally, the APC model was analyzed at the population level. As such, it may be subject to the ecological fallacy.

Conclusion

Mumps prevalence among children aged 1–9 years in Changzhou did not show a significant decrease from 2005 to 2023, the protective effect of a 1-dose MMR immunization strategy is limited. Predictive modeling results Suggest that 2 doses of MMR are effective in reducing mumps epidemics. There is a need for continued strengthening of mumps surveillance in the future, especially for younger children for whom timely and comprehensive immunization services are essential.

Abbreviations

APC

Age-period-cohort

APC

Annual percentage change

ASR

age-standardized incidence rate

BAPC

Bayesian age-period-cohort

CI

confidence interval

EAPC

estimated annual percentage change

EPI

Expanded Programme on Immunization

INLA

Integrated Nested Laplace Approximation

RR

Relative risk

MCMC

Markov-chain Monte Carlo

MM

measles-mumps vaccine

MMR

Measles-mumps-rubella vaccine

MuV

Mumps virus

MuCV

Mumps-containing vaccines

MMRV

measles-mumps-rubella-varicella vaccine

WHO

World Health Organization

Author contributions

Conceptualization, X.L. and C.H.; methodology, Q.X., C.H. and D.W.; validation, D.W. and C.H.; formal analysis, D.W., S.W. and S.X.; investigation, S.W., C.T., S.X., P.Z., H.G., J.L. and D.W.; resources, X.L. and J.L.; data curation, P.Z., C.T., Q.X., C.H., D.W. and S.X.; writing—original draft preparation, Q.X., C.H., D.W. and S.X.; writing—review and editing, X.L., H.G. and J.L.; supervision, X.D.; project administration, X.L.; funding acquisition, X.D., C.T. and Q.X. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Open Research Fund Program of Changzhou Institute for Advanced Study of Public Health, Nanjing Medical University (No.CPHN202301), Changzhou Medical Young Talent Science and Technology Project (Lcyx2024006), Changzhou Health Green Seedling Talent Plan, Changzhou Science and Technology Foundation (CJ20220237), Young Talent Project (CZQM2023026) and Changzhou Science and Technology Foundation (CJ20245050). Funders had no role in the study design, data collection, interpretation and write-up of the manuscript.

Data availability

The study data are available for academic purposes upon reasonable request. To gain access, please contact the primary corresponding author Xufeng Lv (czcdc@163.com) for further details.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.