ABSTRACT
Influenza poses a significant medical burden and has the potential to cause global pandemics. To better understand the epidemiological characteristics of influenza reinfections and identify factors associated with an increased risk of reinfection, we analyzed influenza data from 2010 to 2023 obtained from the National Notifiable Infectious Disease Reporting Information System in Chongqing, China. The Andersen-Gill model was used to evaluate the association between demographic characteristics and the risk of reinfection. A total of 676,811.47 person-years were observed, with a median time to reinfection of 9.40 months. The log-rank test revealed significant differences in reinfection probability by sex, age, occupation, and residence (p < .001). Independent risk factors for influenza reinfection included male sex (adjusted hazard ratio [aHR]: 1.026, 95% confidence interval [CI]: 1.004–1.049), age ≤ 5 years (aHR: 4.645, 95% CI: 4.292–5.027), 6–18 years (aHR: 3.891, 95% CI: 3.574–4.235), age 46–64 years (aHR: 1.336, 95% CI: 1.158–1.541), and age ≥ 65 years (aHR: 2.946, 95% CI: 2.533–3.426), urban residency (aHR: 1.353, 95% CI: 1.315–1.391), and preschool-aged children (aHR: 2.103, 95% CI: 1.830–2.416). Influenza reinfection in Chongqing is relatively common and significantly associated with several key factors, including male sex, age ≤ 18 or > 45 years, preschool-aged children, and urban residency. These findings underscore the importance of tailored public health strategies, particularly promoting influenza vaccination among at-risk groups, to protect vulnerable populations from reinfection. Future studies are needed to refine influenza prevention measures for at-risk individuals, especially those susceptible to reinfection.
KEYWORDS: Epidemiology, influenza virus, reinfection, recurrence, repeated infection
Introduction
Acute respiratory infections (ARIs) remain a leading cause of acute illness globally, imposing a significant burden on public health systems, especially among vulnerable populations.1 ARIs are caused by diverse pathogens and range from mild, self-limiting conditions to severe, life-threatening diseases, particularly affecting young children, the elderly, and immunocompromised individuals.2,3 Among these pathogens, the influenza virus is a major contributor to the global burden of ARIs.4 Influenza primarily affects the respiratory system and continues to impose a substantial medical burden, with its potential to cause global pandemics remaining a significant concern.5 Influenza causes over 5 million hospitalizations globally each year, accounting for 14.1% of adult respiratory diseases hospitalizations 4. As a respiratory infectious disease caused by human influenza viruses, it significantly contributes to global mortality and morbidity and incurs considerable economic costs.6 A five-year analysis of influenza surveillance data from 82 Chinese cities identified a total of 3,735,934 influenza cases, representing 72.71% of all cases, with estimates indicating 300,000–650,000 respiratory deaths annually attributable to seasonal influenza.7
Epidemiological studies on influenza reinfections are limited and estimated reinfection rates among humans vary significantly, ranging from 2% to 32%.8–10 Price et al.8 reported a cumulative reinfection rate of less than 1% after one year, increasing to 4.6% after five years and 9.6% after a decade, based on a 13-year surveillance dataset from Queensland, Australia. In contrast, Wang et al.9 observed a cumulative incidence of 2% at six months, 4% at one year, 5% at two years, and 7% after 4.9 years in Guangxi, China, over a seven-year period. Notably, Camacho et al.10 reported an exceptionally high reinfection rate of 32% on Tristan da Cunha Island within 59 days during a two-wave influenza A/H3N2 epidemic. These heterogeneous estimates highlight the need for updated data in influenza reinfection research. However, epidemiological studies on reinfections remain limited.
Chongqing, a centrally-administered municipality in southwestern China with an area of over 82,400 km2 and a population of more than 34 million, has a substantial proportion of residents aged 60 years or older (21.87%).11 Despite the initiation of influenza surveillance in 2005, understanding of influenza reinfection rates and characteristics in Chongqing remain limited. This study aims to delineate the epidemiological features of influenza reinfections and identify factors associated with an increased risk of reinfection.
Material and methods
Data source of influenza cases
Data for this study included all influenza cases in Chongqing from 2010 to 2023 recorded in the National Notifiable Infectious Disease Reporting Information System (NIDRIS) database. The database mandates that all newly diagnosed influenza cases be reported within 24 hours after diagnosis, along with demographic data, including patient names, identity card numbers, the names of guardians (for minors under the age of 18), sex, date of birth, age, residential address, diagnosis classification (clinically diagnosed or laboratory-confirmed), dates of illness onset and diagnosis, and virus serotype for laboratory-confirmed cases (if documented by the physician).
Case definitions
Clinically diagnosed cases and laboratory-confirmed cases were defined according to the influenza diagnosis and treatment protocol established by the National Health Commission of the People’s Republic of China.12 A clinically diagnosed case was defined as a patient exhibiting symptoms such as cough, fever, sore throat, and dry throat, along with a clinician-confirmed influenza-related epidemiological history. A laboratory-confirmed case required not only the manifestation of these clinical symptoms but also a positive result from at least one of the following etiological tests: (a) real-time fluorescence-based quantitative polymerase chain reaction (PCR) or rapid multiplex PCR for influenza virus nucleic acid; (b) a rapid antigen test using either a colloidal gold assay or immunofluorescence; (c) a fourfold or greater increase in influenza virus-specific immunoglobulin G (IgG) antibody titers, or (d) successful isolation of the influenza virus. The study was approved by the Institutional Review Board of the Chongqing Center for Disease Control and Prevention.
Exclusion criteria were as follows: (1) suspected diagnoses; (2) duplicate or misreported medical records; (3) cases lacking essential identification data (including but not limited to patient names and identity cards numbers), which prevented reliable identity verification; and (4) reinfection episodes occurring within 14 days of the preceding event.
Identification criteria for reinfection case
Given the duration of viral shedding following the initial infection,13 a reinfection case was defined as a subsequent influenza notification for the same patient occurring more than 14 days after a previous notification, consistent with other studies.8 All cases were reported by doctors, with only a small proportion being confirmed by laboratory detection.
The inclusion criteria for identifying reinfection cases in this study were as follows: (a) matching patient names with identification documents; (b) the presence of at least two consistent data points among the patient’s phone number, current address, and the name of a parent or guardian (if available); (c) a diagnosis interval of greater than 14 days between two influenza episodes within the period from 2010 to 2023.
Laboratory-confirmed influenza cases with specified viral typing were categorized as: (1) untyped, (2) influenza A (including all subtypes), (3) influenza B (both Victoria and Yamagata lineages), or (4) A and B coinfection.
Statistical analysis
Baseline characteristics for continuous data were presented as mean ± standard deviation (SD) or median [interquartile range (IQR)] for normally and non-normally distributed continuous data, respectively. Comparisons of primary infection characteristics between individuals with and without reinfection were conducted using the chi-square test or Fisher’s exact test for categorical variables and the t-test or Mann-Whitney U test for continuous variables. Patients without a reinfection by December 31, 2023, were right-censored at their last follow-up date. Person-years of follow-up were calculated from the date of the first influenza diagnosis to the date of reinfection or December 31, 2023, whichever occurred first. Kaplan-Meier estimators were used to determine the cumulative incidence of reinfection, and the log-rank test was used to assess the proportion of influenza reinfections over time after the first infection; pairwise comparisons were conducted if necessary.
Our analysis of influenza reinfection was conducted using the Andersen-Gill model, a robust statistical approach specifically designed for recurrent event analysis that accounts for within-subject correlation of multiple events.14 Hazard ratios (HRs) and their corresponding 95% confidence intervals (CIs) were estimated to explore the characteristics associated with influenza reinfection in both univariate and multivariate analyses.
Classification of influenza reinfection patterns
Patients with consecutive infections involving specific combinations of influenza virus types were included. They were categorized into nine groups based on the virus types of their two consecutive infections:
A – A: reinfection with influenza A after an initial A infection;
A&B – A&B: subsequent coinfection with A and B following an initial A&B coinfection;
B – B: reinfection with influenza B after a prior B infection;
A – A&B: coinfection with A and B following an initial A infection;
A – B: reinfection with B after an A infection;
A&B – A: reinfection with A after an A&B coinfection;
A&B – B: reinfection with B after an A&B coinfection;
B – A: reinfection with A after a B infection;
B – A&B: coinfection with A and B after a B infection.
Patients were classified as either homotypic, involving the same virus type in both infections (groups a, b, and c), or heterotypic, involving different virus types across infections (groups d through i).
Sensitivity analyses
We conducted sensitivity analyses to assess the robustness of our primary findings, consistently adjusting for predefined covariates (age, sex, occupation, residence, and influenza virus type) in all analyses. Four approaches were implemented:
1. Temporal stability assessment through 2023 data exclusion
To account for the atypical influenza activity in 2023 (potentially due to the relaxation of COVID-19 measures and detection bias,15 we excluded data from 2023 and assessed the consistency of reinfection incidence and risk factors.
2. Diagnostic technology-adapted stratified analysis
We stratified analyses into two diagnostic eras to evaluate potential detection bias resulting from evolving diagnostic technologies:
Era 1 (2010–2015): Dominated by viral culture and immunofluorescence assays.
Era 2 (2016–2023): Characterized by widespread adoption of RT-PCR.
This stratification allowed us to assess whether changes in diagnostic methods affected the reinfection rate estimates.
3. Multidimensional confounder adjustment
In addition to the predefined covariates, we incorporated four factors related to transmission dynamics and healthcare resources – population density, PM2.5 levels, health facility density, and per capita disposable income – into the multivariate model to assess their impact on reinfection risk.
4. Modeling framework robustness verification
We compared results obtained from the Andersen-Gill model with those from a standard Cox proportional hazards model to evaluate the impact of model selection on risk estimates. A two-tailed P-value <.05 was considered statistically significant for all analyses. All analyses were conducted using R software, version 4.0.2 (The R Foundation, Boston, MA). The main model fitting was conducted using the ‘survival’ packages.16
Results
Patient demographic characteristics
From January 2010 to December 2023, our study identified 696,343 influenza cases. After screening, 521,818 cases from 489,526 individuals were included (Figure 1). The age at first episode ranged from 17 to 90 years, with a median of 8.00 years (IQR: 6.00–13.00). Of the study population, 87.56% were minors. Of the 489,526 individuals, 260,318 (53.18%) were male, and 229,208 (46.82%) were female; 297,954 (60.87%) were students, and 404,991 (82.73%) lived in urban areas. Overall, 380,755 (77.78%) cases were laboratory-confirmed cases, while 108,771 (22.22%) were clinically diagnosed. Among the laboratory-confirmed cases, influenza A was the most prevalent (38.15%), followed by untyped influenza (32.04%), influenza B (5.88%), and coinfection with influenza A/B (1.71%). Figure 2 displays the seasonal patterns of influenza virus activity in Chongqing from 2010 to 2023. Although influenza viruses circulate year-round, peak activity is generally observed during fall and winter, sometimes extending into spring. A marked increase in influenza incidence was observed after 2018, with a particularly sharp rise in 2023, likely influenced by policy adjustments in the post-COVID-19 era. These temporal patterns were consistently observed across total influenza cases (including both clinically diagnosed and laboratory-confirmed cases) and laboratory-confirmed cases with identified subtypes.
Figure 1.
Flow diagram of selection process for patients.
Figure 2.
Temporal patterns of influenza virus activity and reinfection in Chongqing, China, 2010–2023.
a. Total influenza cases (clinical and laboratory-confirmed).
b. Influenza reinfection cases (clinical and laboratory-confirmed).
c. Laboratory-confirmed influenza cases with identified subtypes.
d. Laboratory-confirmed reinfection cases with identified subtypes.
Table 1 compares the demographic characteristics between patients with and without reinfection. Those with reinfection tended to be younger, predominantly male, and more likely to have laboratory-confirmed diagnoses (p < .001).
Table 1.
Sociodemographic characteristics of individuals with and without influenza reinfection in Chongqing, China, 2010–2023.
| Characteristic | Total | Patients with 1 episode, N (%) | Patients with ≥2 pisodes(reinfection), N (%) | Statistics | P-value |
|---|---|---|---|---|---|
| Total | 489,526 | 459,603 | 29,923 | – | – |
| Sex | |||||
| Male | 260318 | 243967 (93.72) | 16351 (6.28) | 27.445 | <.001 |
| Female | 229208 | 215636 (94.08) | 13572 (5.92) | ||
| Age at first episode (years) | |||||
| 0–5 | 121267 | 104889 (86.49) | 16378 (13.51) | 15957.618 | <.001 |
| 6–18 | 307365 | 295145 (96.02) | 12220 (3.98) | ||
| 19–45 | 48589 | 47866 (98.51) | 723 (1.49) | ||
| 46– 64 | 8253 | 7986 (96.76) | 267 (3.24) | ||
| ≥65 | 4052 | 3717 (91.73) | 335 (8.27) | ||
| Median (IQR*) | 8.00 (6.00–13.00) | 9.00 (6.00–14.00) | 5.00 (3.00–7.00) | 80.545 | <.001 |
| Occupation | |||||
| Farmers | 3075 | 2662 (86.57) | 413 (13.43) | 77777.602 | <.001 |
| Other | 174968 | 163092 (93.21) | 11876 (6.79) | ||
| Preschool-aged children | 9542 | 2599 (27.24) | 6943 (72.76) | ||
| Students | 297954 | 287350 (96.44) | 10604 (3.56) | ||
| Medical staff | 3987 | 3900 (97.82) | 87 (2.18) | ||
| Residence | |||||
| Urban | 404991 | 381080 (94.09) | 23911 (5.91) | 177.760 | <.001 |
| Rural | 84535 | 78523 (92.89) | 6012 (7.11) | ||
| Primary infection with influenza virus type | |||||
| Clinically diagnosed cases | 108771 | 99691 (91.65) | 9080 (8.35) | 1393.189 | <.001 |
| Laboratory-confirmed cases | |||||
| Untyped influenza virus | 156827 | 147993 (94.37) | 8834 (5.63) | ||
| Influenza A | 186778 | 177300 (94.93) | 9478 (5.07) | ||
| Influenza B | 28763 | 26785 (93.12) | 1978 (6.88) | ||
| Coinfection of influenza A and B | 8387 | 7834 (93.41) | 553 (6.59) |
*Interquartile Range, IQR; No, Number.
Characteristics of reinfection patients
Over the 14-year observation period, the study population accumulated 676,811.47 person-years of data. The vast majority (93.89%) experienced only a single influenza episode, while a small proportion (6.11%; 29923 individuals) had multiple reinfections, ranging from 1 to 8 episodes. Among the reinfection group, 93.56% (27,997 individuals) experienced one reinfection, 5.43% (1,626 individuals) experienced two, and 1.00% (300 individuals) experienced three or more.
The intervals between infections varied widely, ranging from 0.50 to 149.13 months, with a median interval of 9.40 months and a mean of 18.61 months. The median time between the first and subsequent reinfections decreased progressively: 19.04 months (first to second reinfection), 13.68 months (second to third), 11.83 months (third to fourth), 9.02 months (fourth to fifth), 9.25 months (fifth to sixth), 7.06 months (sixth to seventh), 9.01 months (seventh to eighth), and 9.08 months (eighth to ninth).
In the population experiencing reinfection, statistically significant differences were observed in sex, age, occupation, and the distribution of influenza virus subtypes across varying reinfection counts, as detailed in Table 2.
Table 2.
Characteristics of influenza reinfection in Chongqing, China, 2010–2023.
| Characteristic | Total | Frequency of reinfection episode, N (%) |
Statistics | P-value | |||
|---|---|---|---|---|---|---|---|
| 0 | 1 | 2 | ≥3 | ||||
| Total | 489526 | 459603 (93.89) | 27997 (5.72) | 1626 (0.33) | 300 (0.06) | – | – |
| Sex | |||||||
| Male | 260318 | 243967 (93.72) | 15313 (5.88) | 885 (0.34) | 153 (0.06) | 29.171 | <.001 |
| Female | 229208 | 215636 (94.08) | 12684 (5.53) | 741 (0.32) | 147 (0.06) | ||
| Age groups (years) | |||||||
| 0–5 | 121267 | 104889 (86.49) | 14862 (12.26) | 1245 (1.03) | 271 (0.22) | – | 0.005* |
| 6–18 | 307365 | 295145 (96.02) | 11856 (3.86) | 341 (0.11) | 23 (0.01) | ||
| 19–45 | 48589 | 47866 (98.51) | 708 (1.46) | 13 (0.03) | 2 (0.00) | ||
| 46 –64 | 8253 | 7986 (96.77) | 256 (3.10) | 10 (0.12) | 1 (0.01) | ||
| ≥65 | 4052 | 3717 (91.73) | 315 (7.77) | 17 (0.42) | 3 (0.07) | ||
| Mean (SD) | 11.67 ± 11.00 | 11.96 ± 11.00 | 7.30 ± 10.03 | 5.01 ± 8.88 | 3.72 ± 8.86 | 5445.213 | <.001 |
| Occupation | |||||||
| Farmers | 3075 | 2662 (86.57) | 390 (12.68) | 19 (0.62) | 4 (0.13) | – | 0.005* |
| Other | 174968 | 163092 (93.21) | 11064 (6.32) | 701 (0.40) | 111 (0.06) | ||
| Preschool-aged children | 9542 | 2599 (27.24) | 6141 (64.36) | 635 (6.66) | 167 (1.75) | ||
| Students | 297954 | 287350 (96.44) | 10318 (3.46) | 269 (0.09) | 17 (0.01) | ||
| Medical staff | 3987 | 3900 (97.82) | 84 (2.11) | 2 (0.05) | 1 (0.03) | ||
| Residence | |||||||
| Urban | 404991 | 381080 (94.10) | 22430 (5.54) | 1242 (0.31) | 239 (0.06) | 192.754 | <.001 |
| Rural | 84535 | 78523 (92.89) | 5567 (6.59) | 384 (0.45) | 61 (0.07) | ||
| Primary infection with influenza virus type | |||||||
| Clinically diagnosed cases | 108771 | 99691 (91.65) | 8212 (7.55) | 692 (0.64) | 176 (0.16) | 1768.618 | <.001 |
| Laboratory-confirmed cases | |||||||
| Untyped influenza virus | 156827 | 147993 (94.37) | 8398 (5.36) | 403 (0.26) | 33 (0.02) | ||
| Influenza A | 186778 | 177300 (94.93) | 9080 (4.86) | 338 (0.18) | 60 (0.03) | ||
| Influenza B | 28763 | 26785 (93.12) | 1798 (6.25) | 159 (0.55) | 21 (0.07) | ||
| Coinfection of influenza A and B | 8387 | 7834 (93.41) | 509 (6.07) | 34 (0.41) | 10 (0.12) | ||
*Fisher’s exact test was used for statistical analysis; No., Number.
Incidence of influenza reinfection
From 2010 to 2023, the cumulative incidence of influenza reinfection increased progressively: 0.70% at 6 months, 5.40% at 12 months, 9.30% at 24 months, 11.70% at 36 months, 17.40% at 48 months, 21.50% at 60 months, and peaked at 23.90% after 120 months (Figure 3a). Females consistently exhibited higher reinfection probabilities than males at each time point: 0.80% vs. 0.70% at 6 months, 5.30% vs. 5.50% at 12 months, 9.00% vs. 9.50% at 24 months, 11.40% vs. 12.00% at 36 months, 16.90% vs. 17.80% at 48 months, 20.80% vs. 22.10% at 60 months, and 23.20% vs. 24.50% at 120 months (Figure 3b). Children under the age of 5 showed the highest reinfection probabilities, reaching 44.00% after 120 months (Figure 3c). Reinfection probabilities in urban areas were consistently higher than those in rural areas at all measured time points: 0.70% vs. 0.80% at 6 months and 25.90% vs. 19.70% at 120 months (Figure 3d). Children in home care had significantly higher reinfection probabilities, with aprobability of 90.22% at 120 months, compared to 8.60% among medical staff (Figure 3e). Laboratory-confirmed cases with mixed A and B infections exhibited the highest reinfection probability, escalating to 43.10% by 60 months. Notably, the log-rank test revealed significant differences in reinfection probabilities by sex, age groups, occupation, and residence (p < .001).
Figure 3.
Kaplan-Meier survival analysis of influenza reinfection following primary infection episodes in Chongqing, China, 2010–2023.
a. Cumulative probability of reinfection for all patients.
b. Cumulative probability of reinfection by sex.
c. Cumulative probability of reinfection by age group.
d. Cumulative probability of reinfection by residence type.
e. Cumulative probability of reinfection by occupational group.
Predictors of influenza reinfection
The univariate Cox regression analysis revealed significant associations between influenza reinfection and several factors, including age group, occupation, place of residence, and the type of influenza virus in the initial infection. Multivariate Cox regression analysis further identified independent risk factors for reinfection, including male sex (adjusted hazard ratio [aHR]: 1.026, 95% CI: 1.004–1.049) and specific age groups: ≤ 5 years (aHR: 4.645, 95% CI: 4.292–5.027), 6–18 years (aHR: 3.891, 95% CI: 3.574–4.235), 46–64 years (aHR: 1.336, 95% CI: 1.158–1.541), and ≥65 years (aHR: 2.946, 95% CI: 2.533–3.426).
Additionally, residing in an urban area (aHR: 1.353, 95% CI: 1.315–1.391) and being a preschool-aged child (aHR: 2.103, 95% CI: 1.830–2.416) were associated with an increased risk of reinfection. In contrast, students (aHR: 0.185, 95% CI: 0.160–0.211) and medical staff (aHR: 0.363, 95% CI: 0.282–0.467) exhibited a significantly reduced risk. Primary infections with influenza A (aHR: 1.619, 95% CI: 1.571–1.667), coinfection with influenza A and B (aHR: 1.878, 95% CI: 1.733–2.035), and influenza B (aHR: 1.540, 95% CI: 1.472–1.612) were identified as significant independent risk factors for reinfection. Table 3 summarizes the univariate and multivariate analyses of influenza reinfection.
Table 3.
Univariate and multivariate analyses of factors associated with influenza reinfection in Chongqing, China, 2010–2023.
| Factors | Univariate analysis |
Multiple analysis |
||||
|---|---|---|---|---|---|---|
| HR | 95%CI | P-value | aHR | 95%CI | P-value | |
| Sex(ref: female) | ||||||
| Male | 1.043 | (1.019,1.067) | <.001 | 1.026 | (1.004,1.049) | 0.021 |
| Age group(ref: 19–45 year) | ||||||
| 0–5 | 6.970 | (6.472,7.505) | <.001 | 4.645 | (4.292,5.027) | <.001 |
| 6–18 | 2.255 | (2.093,2.429) | <.001 | 3.891 | (3.574,4.235) | <.001 |
| 46–64 | 1.725 | (1.501,1.982) | <.001 | 1.336 | (1.158,1.541) | <.001 |
| ≥65 | 4.436 | (3.897,5.048) | <.001 | 2.946 | (2.533,3.426) | <.001 |
| Occupation(ref: farmers) | ||||||
| Other | 1.099 | (0.994,1.215) | 0.065 | 0.348 | (0.303,0.399) | <.001 |
| Preschool-aged children | 7.262 | (6.568,8.029) | <.001 | 2.103 | (1.830,2.416) | <.001 |
| Students | 0.563 | (0.510,0.623) | <.001 | 0.185 | (0.160,0.211) | <.001 |
| Medical Staff | 0.292 | (0.231,0.369) | <.001 | 0.363 | (0.282,0.467) | <.001 |
| Residence types(ref:rural) | ||||||
| Urban | 1.332 | (1.295,1.371) | <.001 | 1.353 | (1.315,1.391) | <.001 |
| Primary infection with influenza virus type(ref: clinically diagnosed cases) | ||||||
| Untyped influenza virus | 1.320 | (1.281,1.360) | <.001 | 1.461 | (1.419,1.504) | <.001 |
| Influenza A | 1.460 | (1.417,1.505) | <.001 | 1.619 | (1.571,1.667) | <.001 |
| Influenza B | 1.466 | (1.398,1.537) | <.001 | 1.540 | (1.472,1.612) | <.001 |
| Coinfection of influenza A and B | 1.698 | (1.565,1.841) | <.001 | 1.878 | (1.733,2.035) | <.001 |
HR: adjusted hazard ratio; CI: confidence interval.
Characteristics of influenza reinfection based on identifiable virus types
We conducted a subgroup analysis of influenza reinfections in 29,923 patients, identifying 8,588 cases with identifiable virus types in both primary and subsequent infections. Homotypic reinfection occurred in 4,579 cases (53.32%), with a mean interval of 14.55 months, while heterotypic reinfection was observed in 4,009 cases (46.68%), a mean interval of 16.26 months between episodes. The interval between reinfections significantly varied across groups (p < .001), ranging from 0.50 to 80.90 months. Patients with homotypic reinfections had an mean age of 7.17 years, significantly lower than the 7.55 years for those with heterotypic reinfections (p < .001).
Homotypic reinfections were significantly more frequent in children under 5 years of age, accounting for 65.85% of these cases, and were most commonly observed during the winter season, constituting 45.79% of total reinfections. In contrast, heterotypic reinfections peaked among individuals aged 6 to 18 years, with winter also being the predominant season for these cases, and were especially common among students and urban residents (p < .001). A detailed summary of these findings is provided in Supplementary Table S1.
Sensitivity analyses
We analyzed data from 2010 to 2023 using a standard Cox proportional hazards model. The results aligned with those presented in Table 3 and are detailed in Supplementary Table S2. After excluding 2023 data, the cumulative probability of influenza reinfection increased steadily over time, from 1.49% at 6 months to 23.88% at 120 months. The 6-month reinfection probability was lower when excluding 2023 data (0.73%) compared to the complete 2010–2023 dataset (1.49%), but cumulative probabilities at other time points remained stable (Supplementary Table S4; see Supplementary Table S2 for multivariate analysis results). Stratified analyses by time period (pre-2016 [2010–2015] and post-2016 [2016–2023]) revealed that the cumulative probability of reinfection was lower in the pre-2016 period compared to the full 2010–2023 dataset, while post-2016 probabilities were similar to those of the full dataset (Supplementary Table S3). Including socioeconomic and environmental factors – such as population density, PM2.5 levels, healthcare facility density, and per capita disposable income – had minimal impact on the results, further supporting the robustness of our findings (Supplementary Table S5).
Discussion
Influenza reinfections are common throughout the lifespan.17–19 In our study, over 6% of patients experienced reinfection, with the risk increasing as the interval since the previous episode lengthened, consistent with previous findings.8,9 Compared to a study from Guangxi, China,9 our research reported a higher overall reinfection rate, except during the first six months. This discrepancy likely arises from differences in reinfection criteria: we defined reinfections as episodes separated by at least 14 days, while Wang et al.9 used a 35-day interval. Our 14-day threshold aligns with the documented duration of viral shedding13,20 and is supported by other studies.8
Reinfection episodes in our study began to rise significantly from 2018, with minor peaks in 2019 and 2022 and a record high peak in 2023, mirroring the patterns of initial influenza cases. The consistent trend between initial infections and reinfections suggests that an increase in infections is likely to drive a corresponding rise in reinfections. The surge in influenza cases observed in 2019 and 2023 aligns with reports from other countries and regions within China.9,21,22 One possible explanation is the emergence of a variant with a three-amino-acid deletion, which was antigenically distinct from the vaccine strain used during the 2018–2019 season. Additionally, the sharp rise in influenza cases in 2023 may be linked to the relaxation of national control measures following the COVID-19 pandemic.23 In sensitivity analyses excluding 2023 data, the cumulative reinfection rates remained largely consistent, except for a notable difference at 6 months (0.73% vs. 1.49%), further underscoring the robustness of our findings.
Furthermore, our analysis indicated that individuals aged 19–45 years were less likely to experience influenza reinfection compared to other age groups. Reinfection rates were highest among children ≤5 years, followed by those aged 6–18 years and individuals ≥65 years. This age-related susceptibility pattern aligns with the established understanding that children’s developing immune systems and older adults’ waning immunity make them more prone to recurrent infections, as supported by previous studies.8,9,24 Specifically, children’s immature immune systems increase their susceptibility to infections.25 Adults’ immunological responses, as indicated by modeling studies, provide longer-lasting protection following exposure to the same influenza strain.26 Similarly, adults over 45 years of age exhibit a higher likelihood of reinfection compared to those aged 18–45 years, consistent with the well-documented effects of immunosenescence on influenza immunity.27,28 However, this population is the least likely to seek medical care for symptoms related to acute respiratory infections.29 Future studies should collect region-specific data on healthcare-seeking behavior and testing rates to better distinguish true epidemiological differences from potential detection bias.
Our study further revealed that residing in urban areas was significantly associated with an increased risk of influenza reinfection. The availability of advanced medical facilities in urban areas may lead to more frequent influenza testing, potentially increasing reinfection detection rates due to detection bias.30,31 Healthcare-seeking behavior may also influence reinfection risk. For example, Dai et al.29 reported that only 50.92% of patients with acute respiratory infections sought treatment at healthcare facilities in rural areas.
Our study also found that the mean interval to influenza reinfection was 18.61 months after the initial infection, which decreased to 15.51 months when both reinfections involved identified virus types, a finding consistent with other studies in China.9 More than half of the reinfections were caused by the same virus type, with shorter intervals typically observed in consecutive homotypic infections, suggesting waning immunity.8 The marked increase in genetic diversity and antigenic drift likely contributed to the observed reinfections and the shortening of intervals.32
This study has several strengths and limitations that should be acknowledged. A key strength is the extensive observation period and large sample size, which enabled detailed documentation of infection cases. Our in-depth analysis of reinfection characteristics in both the general influenza population and specific subtypes provides valuable insights into these patterns. The use of the Andersen-Gill model, which is more suitable than the standard Cox model for analyzing recurrent influenza infections, ensures that within-individual correlations are appropriately addressed.14 To address potential biases, we conducted sensitivity analyses by excluding 2023 data, stratifying by diagnostic era (pre-2016 and post-2016), adjusting for socioeconomic and environmental factors, and comparing model results. The minimal variation observed supports the robustness of our findings. However, the post-pandemic period may reflect a distinct immunological context that warrants further investigation.
However, this study has limitations inherent to record-based data. The lack of individual-level vaccination history limits our ability to assess vaccine-induced immunity’s role in modifying age- and occupation-related reinfection risks, potentially overestimating risks in unvaccinated populations and underestimation vaccination’s protective effects.33 Vaccination significantly reduces influenza-related hospitalizations in high-risk groups, particularly among the elderly and children.34 Future studies should incorporate vaccination data to improve reinfection risk assessments. Changes in China’s healthcare system, such as expanded universal healthcare coverage, increased healthcare-seeking behavior, and improved infectious disease reporting, may have enhanced case detection and reporting over time. These factors could potentially influence the observed temporal trends in reinfection rates.35 Furthermore, our findings may underestimate the true incidence of influenza reinfection, as reinfections often present with milder symptoms, reducing the likelihood of individuals seeking medical attention or testing.8,36,37 The definition of reinfections relied on a 14-day threshold between positive tests, which, although aligned with typical viral shedding periods, may not fully distinguish true reinfections and prolonged viral shedding or reactivation.38 Future studies should establish a more accurate threshold by integrating clinical symptom profiles, serological testing, and time-based criteria. Finally, data limitations prevented competing risk analyses accounting for mortality and other censoring events, which could be addressed in future studies through linkage with mortality registries.
Conclusions
In conclusion, our 14-year study highlights a significant burden of influenza reinfection in Chongqing, particularly among males and individuals aged ≤ 18 or >45 years. These findings underscore the importance of tailored public health strategies, including promoting influenza vaccination among at-risk groups, to protect vulnerable populations from reinfection. Future studies are needed to refine influenza prevention measures for at-risk individuals, especially those susceptible to reinfection.
Supplementary Material
Biographies
Jiang Long, MD, is Deputy Director of Chongqing Center for Disease Control and Prevention, China. His research interests include vaccine evaluation and infectious disease surveillance. He has published more than 20 peer-reviewed papers, including over 10 SCI papers. Dr. Long serves as a visiting professor at Chongqing Three Gorges Medical College and a postgraduate tutor at Southwest University. He was awarded the 2024 Public Health Talent Development Support Program by the National CDC.
Li Qi, PhD, is Director of the Institute of Infectious Diseases at Chongqing Center for Disease Control and Prevention, China, and a supervisor for graduate students at Chongqing Medical University. Her research focuses on vaccine strategies and infectious disease epidemiology. She has published over 60 papers, including 30 + SCI papers as first or corresponding author. Dr. Qi received postdoctoral training at China CDC and was a visiting scholar at U.S. CDC.
Funding Statement
This study was supported by Medical scientific research project of Chongqing Health Commission in 2024 [Grant No. 2024WSJK070], Natural Science Foundation of China [Grant No. 12371503], the Chongqing Science and Technology Bureau [Grant No. CSTC2021jscxgksb-N0005 and cstc2024ycjh-bgzxm0224], 2023 Key Disciplines On Infectious Disease Control and Prevention in Chongqing, and China Preventive Medicine Association [Grant no. CPMA2024CRBFK]. The analyses and interpretations of the data and the final content of the manuscript were produced independently of the financial sponsors.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Author contributor
All authors contributed to the study conception and design. Ju Wang performed the initial analysis and drafted the manuscript. Ju Wang, Yu Xiong, Baisong Li, and Xiaoqing Fu contributed to data curation and critically reviewed the manuscript. The methodology was developed by Ju Wang, Xiaoqing Fu, Jule Yang, and Dayong Xiao. Ju Wang, Jiang Long, and Li Qi supervised the project and obtained funding. All authors approved the final version of the manuscript.
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Ethical approval
Ethical approval for this study was obtained from the institutional review board of the Chongqing Center for Disease Control and Prevention (Record number: 2021026).
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2025.2497576
References
- 1.Daffa MB, Khansa F, Ghifari MH, Zaskia SR, Jung Y.. Global prevalence of natural disaster-induced acute respiratory infections: a systematic review and meta-analysis of 290,380 participants. Narra X. 2024;2(3):e181. doi: 10.52225/narrax.v2i3.181. [DOI] [Google Scholar]
- 2.Wang X, Li Y, Shi T, Bont LJ, Chu HY, Zar HJ, Wahi-Singh B, Ma Y, Cong B, Sharland E, et al. Global disease burden of and risk factors for acute lower respiratory infections caused by respiratory syncytial virus in preterm infants and young children in 2019: a systematic review and meta-analysis of aggregated and individual participant data. Lancet (Lond, Engl). 2024;403(10433):1241–10. doi: 10.1016/S0140-6736(24)00138-7. [DOI] [PubMed] [Google Scholar]
- 3.MacLean AJ, Richmond N, Koneva L, Attar M, Medina CAP, Thornton EE, Gomes AC, El-Turabi A, Bachmann MF, Rijal P, et al. Secondary influenza challenge triggers resident memory B cell migration and rapid relocation to boost antibody secretion at infected sites. Immunity. 2022;55(4):718–733.e8. doi: 10.1016/j.immuni.2022.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lafond KE, Porter RM, Whaley MJ, Suizan Z, Ran Z, Aleem MA, Thapa B, Sar B, Proschle VS, Peng Z, et al. Global burden of influenza-associated lower respiratory tract infections and hospitalizations among adults: a systematic review and meta-analysis. PLOS Medicine. 2021;18(3):e1003550. doi: 10.1371/journal.pmed.1003550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Saunders-Hastings PR, Krewski D. Reviewing the history of pandemic influenza: understanding patterns of emergence and transmission. Pathog (Basel, Switz). 2016;5(4):66. doi: 10.3390/pathogens5040066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zhang R, Lai KY, Liu W, Liu Y, Ma X, Webster C, Luo L, Sarkar C. Associations between short-term exposure to ambient air pollution and influenza: an individual-level case-crossover study in Guangzhou, China. Environ Health Perspectives. 2023;131(12):127009. doi: 10.1289/EHP12145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yu LJ, Li XL, Wang YH, Zhang HY, Ruan SM, Jiang BG, Xu Q, Sun YS, Wang LP, Liu W, et al. Short-term exposure to ambient air pollution and influenza: a multicity study in China. Environ Health Perspectives. 2023;131(12):127010. doi: 10.1289/EHP12146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Price O, Birrell FA, Mifsud EJ, Sullivan SG. Epidemiology of repeat influenza infection in Queensland, Australia, 2005–2017. Epidemiol Infect. 2022;150:e144. doi: 10.1017/S0950268822001157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wang J, Jiang L, Xu Y, He W, Zhang C, Bi F, Tan Y, Ning C. Epidemiology of influenza virus reinfection in Guangxi, China: a retrospective analysis of a nine-year influenza surveillance data: characteristics of influenza virus reinfection. Int J Infect Dis: IJID: Off Publ Int Soc Infect Dis. 2022;120:135–141. doi: 10.1016/j.ijid.2022.04.045. [DOI] [PubMed] [Google Scholar]
- 10.Camacho A, Ballesteros S, Graham AL, Carrat F, Ratmann O, Cazelles B. Explaining rapid reinfections in multiple-wave influenza outbreaks: Tristan da Cunha 1971 epidemic as a case study. Proc Biol Sci. 2011;278(1725):3635–3643. doi: 10.1098/rspb.2011.0300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Statistics CMBO, Chongqing NSOI. Chongqing statistical yearbook 2023. Beijing: China Statistics Press; 2024. [Google Scholar]
- 12.Zhong NS, Li YM, Yang ZF, Wang C, Liu YN, Li XW, Shu YL, Wang GF, Gao ZC, Deng GH, et al. Chinese guidelines for diagnosis and treatment of influenza (2011). J Thorac Dis. 2011;3(4):274–289. doi: 10.3978/j.issn.2072-1439.2011.10.01. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fielding JE, Kelly HA, Mercer GN, Glass K. Systematic review of influenza A(H1N1)pdm09 virus shedding: duration is affected by severity, but not age. Influenza Resp Viruses. 2014;8(2):142–150. doi: 10.1111/irv.12216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Andersen P, Gill R. Cox’s regression model for counting processes: a large sample study. Ann Statist. 1982;10(4):10. doi: 10.1214/aos/1176345976. [DOI] [Google Scholar]
- 15.Wang Q, Jia M, Jiang M, Liu W, Yang J, Dai P, Sun Y, Qian J, Yang W, Feng L. Seesaw effect between COVID-19 and influenza from 2020 to 2023 in world health organization regions: correlation analysis. JMIR Public Health Surveillance. 2023;9:e44970. doi: 10.2196/44970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Therneau TM, Grambsch PM. CRAN version 3.8-3. Comprehensive R Archive Network (CRAN); 2024. https://cran.r-project.org/web/packages/survival/index.html. [Google Scholar]
- 17.Patel MM, York IA, Monto AS, Thompson MG, Fry AM. Immune-mediated attenuation of influenza illness after infection: opportunities and challenges. Lancet Microbe. 2021;2(12):e715–e725. doi: 10.1016/S2666-5247(21)00180-4. [DOI] [PubMed] [Google Scholar]
- 18.Arevalo CP, Le Sage V, Bolton MJ, Eilola T, Jones JE, Kormuth KA, Nturibi E, Balmaseda A, Gordon A, Lakdawala SS, et al. Original antigenic sin priming of influenza virus hemagglutinin stalk antibodies. Proc Natl Acad Sci USA. 2020;117(29):17221–17227. doi: 10.1073/pnas.1920321117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Agustian D, Mutyara K, Murad C, Uyeki TM, Kartasasmita CB, Simoes EA. Epidemiology and population-based incidence of influenza in two communities, Bandung district, West Java, Indonesia, 2008–2011. Narra J. 2024;4(3):e981. doi: 10.52225/narra.v4i3.981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Carrat F, Vergu E, Ferguson NM, Lemaitre M, Cauchemez S, Leach S, Valleron AJ. Time lines of infection and disease in human influenza: a review of volunteer challenge studies. Am J Epidemiol. 2008;167(7):775–785. doi: 10.1093/aje/kwm375. [DOI] [PubMed] [Google Scholar]
- 21.Sullivan SG, Arriola CS, Bocacao J, Burgos P, Bustos P, Carville KS, Cheng AC, Chilver MB, Cohen C, Deng YM, et al. Heterogeneity in influenza seasonality and vaccine effectiveness in Australia, Chile, New Zealand and South Africa: early estimates of the 2019 influenza season. Eurosurveillance. 2019;24(45). doi: 10.2807/1560-7917.ES.2019.24.45.1900645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jia W, Zhang X, Sun R, Li P, Zhen X, Li Y, Wang D, Li C, Song C. Changes in the epidemiological characteristics of influenza in children in Zhengzhou, China, in the post-COVID-19 era. BMC Public Health. 2024;24(1):1938. doi: 10.1186/s12889-024-19460-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhao X, Li M, Haihambo N, Jin J, Zeng Y, Qiu J, Guo M, Zhu Y, Li Z, Liu J, et al. Changes in temporal properties of notifiable infectious disease epidemics in China during the COVID-19 pandemic: population-based surveillance study. JMIR Public Health Surveill. 2022;8(6):e35343. doi: 10.2196/35343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Cohen C, Kleynhans J, Moyes J, McMorrow ML, Treurnicht FK, Hellferscee O, Mathunjwa A, von Gottberg A, Wolter N, Martinson NA, et al. Asymptomatic transmission and high community burden of seasonal influenza in an urban and a rural community in South Africa, 2017–18 (PHIRST): a population cohort study. Lancet Global Health. 2021;9(6):e863–e874. doi: 10.1016/S2214-109X(21)00141-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sullivan SG, Price OH, Regan AK. Burden, effectiveness and safety of influenza vaccines in elderly, paediatric and pregnant populations. Ther Adv Vaccines Immunother. 2019;7:2515135519826481. doi: 10.1177/2515135519826481. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ranjeva S, Subramanian R, Fang VJ, Leung GM, Ip DKM, Perera R, Peiris JSM, Cowling BJ, Cobey S. Age-specific differences in the dynamics of protective immunity to influenza. Nat Commun. 2019;10(1):1660. doi: 10.1038/s41467-019-09652-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.McElhaney JE, Verschoor CP, Andrew MK, Haynes L, Kuchel GA, Pawelec G. The immune response to influenza in older humans: beyond immune senescence. Immun Ageing. 2020;17(1):10. doi: 10.1186/s12979-020-00181-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Allen JD, Ross T. M. mRNA vaccines encoding computationally optimized hemagglutinin elicit protective antibodies against future antigenically drifted H1N1 and H3N2 influenza viruses isolated between 2018-2020. Front Immunol. 2024;15:1334670. doi: 10.3389/fimmu.2024.1334670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Dai P, Qi L, Jia M, Li T, Ran H, Jiang M, Tang W, Yan C, Yang W, Ren Y, et al. Healthcare-seeking behaviours of patients with acute respiratory infection: a cross-sectional survey in a rural area of southwest China. BMJ Open. 2024;14(2):e077224. doi: 10.1136/bmjopen-2023-077224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sundaram N, Schaetti C, Purohit V, Kudale A, Weiss MG. Cultural epidemiology of pandemic influenza in urban and rural Pune, India: a cross-sectional, mixed-methods study. BMJ Open. 2014;4(12):e006350. doi: 10.1136/bmjopen-2014-006350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhang X, Dupre ME, Qiu L, Zhou W, Zhao Y, Gu D. Urban-rural differences in the association between access to healthcare and health outcomes among older adults in China. BMC Geriatr. 2017;17(1):151. doi: 10.1186/s12877-017-0538-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Kato-Miyashita S, Sakai-Tagawa Y, Yamashita M, Iwatsuki-Horimoto K, Ito M, Tokita A, Hagiwara H, Izumida N, Nishino T, Wada N, et al. Antigenic variants of influenza B viruses isolated in Japan during the 2017-2018 and 2018-2019 influenza seasons. Influenza Other Respir Viruses. 2020;14(3):311–319. doi: 10.1111/irv.12713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Doyle JD, Chung JR, Kim SS, Gaglani M, Raiyani C, Zimmerman RK, Nowalk MP, Jackson ML, Jackson LA, Monto AS, et al. Interim estimates of 2018–19 seasonal influenza vaccine effectiveness — United States, February 2019. MMWR Morb Mortal Wkly Rep. 2019;68(6):135–139. doi: 10.15585/mmwr.mm6806a2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kalligeros M, Shehadeh F, Mylona EK, Dapaah-Afriyie C, van Aalst R, Chit A, Mylonakis E. Influenza vaccine effectiveness against influenza-associated hospitalization in children: a systematic review and meta-analysis. Vaccine. 2020;38(14):2893–2903. doi: 10.1016/j.vaccine.2020.02.049. [DOI] [PubMed] [Google Scholar]
- 35.Yip W, Fu H, Chen AT, Zhai T, Jian W, Xu R, Pan J, Hu M, Zhou Z, Chen Q, et al. 10 years of health-care reform in China: progress and gaps in universal health coverage. Lancet (Lond, Engl). 2019;394(10204):1192–1204. doi: 10.1016/S0140-6736(19)32136-1. [DOI] [PubMed] [Google Scholar]
- 36.Sonoguchi T, Sakoh M, Kunita N, Satsuta K, Noriki H, Fukumi H. Reinfection with influenza a (H2N2, H3N2, and H1N1) viruses in soldiers and students in Japan. The J Infect Dis. 1986;153(1):33–40. doi: 10.1093/infdis/153.1.33. [DOI] [PubMed] [Google Scholar]
- 37.Davies JR, Grilli EA, Smith AJ. Influenza A: infection and reinfection. J Hyg (Lond). 1984;92(1):125–127. doi: 10.1017/S002217240006410X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.van der Vries E, Stittelaar KJ, van Amerongen G, Veldhuis Kroeze EJ, de Waal L, Fraaij PL, Meesters RJ, Luider TM, van der Nagel B, Koch B, et al. Prolonged influenza virus shedding and emergence of antiviral resistance in immunocompromised patients and ferrets. PLOS Pathogens. 2013;9(5):e1003343. doi: 10.1371/journal.ppat.1003343. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.