A descriptive cross-sectional study was conducted with data between July and August 2025 from 415 patients diagnosed with CHIKF admitted to Lecong Hospital of Shunde, Foshan, China.

Suspected cases of CHIKV were defined as the presence of fever, rash, and arthralgia by the clinician. Laboratory confirmation was performed by real-time RT-PCR. A total of 415 cases were confirmed as acute CHIKF cases based on the presence of CHIKV RNA in the serum samples.

Patients’ sociodemographic information, self-reported comorbidities, and clinical symptoms were collected using case record forms (CRFs) and semistructured health questionnaire. Hematological reports were obtained through the hospital test system including blood test, liver function test, kidney function test and C-reactive protein (CRP) level.

The Kolmogorov-Smirnov test was used to assess the normality of the data. Continuous variables that were normally distributed were compared using t-tests or analysis of variance (ANOVA), while those that were not normally distributed were compared using the Mann-Whitney U test or Kruskal-Wallis test. Nominal categorical variables were compared using the chi-square test or Fisher's exact test, and ordinal categorical variables were compared using the Kruskal-Wallis test. To identify factors independently associated with binary outcomes, multivariate logistic regression analysis was used to screen for risk factors and protective factors, with model fitting performed using the forward stepwise regression method. Variables were added stepwise based on their significance until no more variables met the criteria for entry into the model (P < 0.05). The model outputs included the regression coefficients (b) for each independent variable, along with their standard errors, Z-values, H- values and P-values. The exponentiated regression coefficients (exp(B)) represented the odds ratios (OR), which were used to measure the impact of the independent variables on the dependent variable. For regression analysis, patients were categorized based on the number of cardinal symptoms (fever, arthralgia, rash). "Symptom burden" was defined as a binary outcome, with low burden (0–1 symptom) and high burden (2–3 symptoms). Patients were also categorized based on fever severity, with "fever severity" defined as a binary outcome: mild group (normal temperature and low-grade fever groups) and severe group (moderate and high fever groups). Additionally, patients were categorized based on nucleic acid conversion time, with this time defined as a binary outcome: rapid conversion group (≤ 6 days) and slow conversion group (> 6 days). Correlation analyses were conducted using Pearson's correlation (for normally distributed continuous variables) or Spearman's rank correlation (for non-parametric data). All statistical tests were two-sided, with P < 0.05 considered statistically significant. The analyses were performed using R software (version 4.3.2).

PrimalScheme version 3.2.2(9) was used to design whole-genome targeted amplification primers for CHIKV. Whole-genome sequencing was performed on 92 CHIKV-positive patient samples using the DNBSEQ-G99 sequencing platform (MGI Tech). The reference strain, Réunion Island 80652-1-1-1 (GenBank: PV700165.1), was used for read mapping with BWA version 0.7.18 and SAMtools version 1.20.0 (Danecek et al, 2021; Li and Durbin, 2010). Variant calling and genome assembly were conducted using iVar version 1.4.4 (Grubaugh et al, 2019). A global collection of 1, 990 complete genome sequences of CHIKV (available up to September 1, 2025) was downloaded from NCBI for phylogenetic analysis. Multiple sequence alignment was performed using MAFFT version 7.525 (Rozewicki et al, 2019). The phylogenetic tree was constructed with FastTree version 2.2.0 (Price et al, 2010) and visualized using the iTOL online server (Letunic and Bork, 2024).

A total of 415 laboratory-confirmed cases of acute CHIKV infection were recorded during the study period. Cases were confined to a 20-day outbreak window (July 15 - August 13, 2025) and originated from four administrative divisions in Lecong, Shunde, Foshan: Lecong Community Neighborhood Committee (73 cases, 17.59%), Shuiteng Village Committee (38 cases, 9.16%), Shajiao Community Neighborhood Committee (33 cases, 7.95%), and Pingbu Community Neighborhood Committee (33 cases, 7.95%) (Figure 1A, 1B). The average disease duration was 5.65 days, and discharge dates were predominantly clustered between July 21 and August 13 (Figure 1B). Sociodemographic analysis revealed a slight female predominance (54.22% female, 45.78% male), with the 46–65 years age group representing the largest proportion of cases (130 patients, 31.33%). Most patients (79.28%) had no pre-existing chronic conditions; among those with comorbidities, hypertension (9.89%), chronic gastritis (5.30%), and diabetes (3.61%) were the most common (Table 1).

Asymptomatic infections were the least frequent (1.45%). Regarding the resolution of these main symptoms, the median time for fever resolution was 1 (IQR 0, 2) day, for arthralgia relief was 3 (IQR 2, 4) days, and for rash resolution was 3 (IQR 2, 4) days. A Venn diagram illustrates the distribution of different combinations of main symptoms among the patient population (Figure 1C). Among 415 confirmed acute CHIKF cases, the main clinical manifestations were, in order, arthralgia (83.61%), fever (74.46%), and rash (61.93%) (Table 1). A small number of patients had sporadic symptoms including fatigue (10.12%), oral ulcer (8.92%), conjunctival congestion (3.86%), headache (3.61%), diarrhea (3.37%), and nausea (2.89%). Among these symptoms, arthralgia primarily affected the ankle (42.89%), wrist (32.53%), and finger joints (32.29%); fever mainly presented as low-grade or moderate (< 39°C), with only 7.23% of patients exhibiting high fever (≥ 39°C); and the rash primarily appeared as maculopapular (37.59%) and mainly affected the limbs (50.12%) (Supplementary Table S1).

Among these patients, those presenting with two of the three main symptoms constituted the largest proportion (42.41%). The combination of arthralgia + fever was the most common (24.58%), followed by similar proportions for arthralgia + rash (9.64%) and rash + fever (8.19%). Patients presenting with all three main symptoms were the next largest group (39.52%). Patients with only a single symptom accounted for only 16.63% of the total, with arthralgia being the most common single symptom (9.88%), followed by rash (4.58%) and fever (2.17%).

Based on the distribution characteristics of symptom burden, patients were divided into four groups according to the number of main symptoms present: asymptomatic group (6 individuals), single-symptom group (69 individuals), double-symptom group (176 individuals), and triple-symptom group (164 individuals). Univariate analysis was initially used to assess the distribution characteristics of a series of clinical data, including demographics and hematological indicators, across these symptom groups.

Demographic data presented as frequency/percentage and quartiles (Table 2) showed that age exhibited a significant distribution difference among the symptom groups (H = 36.721, P < 0.001). The asymptomatic group was relatively evenly distributed across age groups, while the single- and double-symptom groups were more common in patients aged 46–65 years. In contrast, the triple-symptom group was distinctly concentrated in patients aged 7–45 years, showing a typical clustered distribution. Although the number of females was slightly higher than males in each symptom group, gender did not show a differential distribution across the groups. While only 20.72% of patients had various chronic medical histories, the distribution of patients with such histories still showed a statistically significant difference among symptom groups (Z = -2.663, P = 0.008). Regarding body temperature, patients in different symptom groups also showed different distribution patterns (H = 176.501, P < 0.001). The vast majority in the single-symptom group had normal temperature, with patients presenting primarily with isolated rash or arthralgia. Only 9 patients had low or moderate fever. In the double-symptom group, the distribution of patients with low and moderate fever was similar and highest, with a small number having high fever. In the triple-symptom group, patients most frequently had moderate fever, followed by low fever, with only a minority having high fever. The duration of fever also showed different distribution patterns across symptom groups. In pairwise comparisons, except for no statistical difference between the asymptomatic and single-symptom groups, all other inter-group comparisons showed significant differences, displaying a trend of increasing fever duration with the number of symptoms (H = 119.413, P < 0.001).

Hematological indicators presented as quartiles (Supplementary Table S2) showed that routine blood tests, liver and kidney function also exhibited different distribution patterns across symptom groups. Leukocytes count, neutrophil (NE) count, platelet (PLT) count, and estimated glomerular filtration rate (eGFR) showed a gradual increasing trend with the increasing number of symptoms, with statistical differences between groups (P < 0.05). In contrast, blood urea nitrogen and creatinine (CR) showed a gradually decreasing trend with increasing symptom number, also with statistical differences between groups (P < 0.05). Interestingly, CRP showed a characteristic trend of first increasing and then decreasing as the number of symptoms increased, also with statistical differences between groups (P < 0.05). Other indicators showed no significant abnormalities.

To further characterize and identify independent risk factors associated with symptom burden, the asymptomatic and single-symptom groups were classified as the lower burden group, and the double-symptom and triple-symptom groups as the higher burden group. Multivariate logistic regression analysis was performed to construct a model. The results found that among demographic data, age (OR = 0.979, 95% CI: 0.960-0.998, P = 0.029), gender (OR = 0.528, 95% CI: 0.289-0.965, P = 0.038), and fever duration (OR = 8.156, 95% CI: 4.544-14.639, P < 0.001) were included in the model (Table 3). Older age (per 1-year increase) and male gender (reference to female) were protective factors, and negatively correlated with high symptom burden, while longer fever duration was a risk factor associated with a higher symptom burden. Among hematological indicators, only eGFR was included in the model (OR = 1.008, 95% CI: 1.000-1.016, P = 0.005). A lower eGFR was a risk factor associated with higher symptom burden (Supplementary Table S3).

Patients were divided into four groups based on fever severity: normal group (106 individuals), low-grade fever group (125 individuals), moderate fever group (154 individuals), and high fever group (30 individuals). Univariate analysis was initially used to assess the distribution characteristics of a series of clinical data, including demographics and hematological indicators, across these fever severity groups.

Demographic data presented as frequency/percentage and quartiles (Table 4) showed that age again showed a significant statistical difference across the fever severity groups (H = 42.584, P < 0.001). The normal temperature group consisted mostly of patients aged 46–65 years. The low-grade and moderate fever groups consisted mostly of patients aged 19–65 years. The high fever group consisted mostly of students aged 7–18 years. The number of symptoms also showed a statistical difference across fever severity groups (H = 147.639, P < 0.001). Patients in the normal temperature group primarily had isolated arthralgia or rash, or both. All febrile patients were typically accompanied by arthralgia or rash, or both. In the low-grade and high fever groups, the numbers of patients with two and three symptoms were similar, while the moderate fever group was characteristically represented by patients with three symptoms. The duration of fever also showed different distribution patterns across the fever severity groups (H = 249.866, P < 0.001). In pairwise comparisons, statistical differences existed only between the low-grade and high fever groups, and between the normal temperature group and all fever groups.

Hematological indicators presented as quartiles (Supplementary Table S4) showed that routine blood tests, liver and kidney function also exhibited different distribution patterns across the fever severity groups. Leukocytes count, NE count, and eGFR showed some increasing trend between the normal temperature group and the fever groups, while blood urea nitrogen and CR showed a decreasing trend between the normal temperature group and the fever groups. However, there were no significant differences among the different fever groups themselves, suggesting that changes in these routine laboratory indicators are likely due to elevated body temperature, disruption of internal homeostasis, and immune system activation.

Similarly, to further characterize and identify independent risk factors associated with symptom severity, the normal temperature and low-grade fever groups were classified as the mild group, and the moderate and high fever groups as the severe group. Multivariate logistic regression analysis was performed to construct a model. The results found that among demographic data, only symptom burden (OR = 10.198, 95% CI: 3.257-31.933, P < 0.001) and fever duration (OR = 2.591, 95% CI: 1.970-3.409, P < 0.001) were included in the model (Table 5). Both multiple symptoms and longer fever duration were risk factors associated with fever severity. Among hematological indicators, only CR (OR = 1.041, 95% CI: 1.014-1.069, P = 0.003) and eGFR (OR = 1.043, 95% CI: 1.025-1.062, P < 0.001) were included in the model (Supplementary Table S5). Lower CR levels and higher eGFR were both risk factors associated with fever severity.

Subsequently, we focused on the time to nucleic acid conversion (negative result). Comparing demographic data across different categories revealed that older patients had a significantly longer disease duration than younger individuals (P < 0.001). There was a weak but significant positive correlation between age and disease duration (R = 0.24, P = 2.6e-05). Similarly, the presence of hypertension was also significantly associated with a prolonged clinical course (P = 0.001). In contrast, no significant differences were observed based on gender (P = 0.111), diabetes (P = 0.085), hyperlipidemia (P = 0.989), or history of chronic gastritis/peptic ulcer (P = 0.164) (Figure 2).

Meanwhile, using the median nucleic acid conversion time of 6 (IQR 5, 6) days, the clinical data were divided into a rapid conversion group (≤6 days) and a slow conversion group (> 6 days). Logistic regression analysis was performed to construct a model. Age (OR = 1.022, 95% CI: 1.012-1.030, P < 0.001), PLT (OR = 0.992, 95% CI: 0.987-0.996, P < 0.001), and eGFR (OR = 0.990, 95% CI: 0.987-0.998, P = 0.046) were included in the regression equation. Older age was an independent risk factor affecting the nucleic acid conversion time, while increased PLT and increased eGFR were independent protective factors affecting the nucleic acid conversion time (Supplementary Table S6-Supplementary Table S7).

In summary, older age and male gender are independent protective factors affecting the number of main clinical symptoms in CHIKF patients, while longer fever duration and lower eGFR are independent risk factors affecting the number of main clinical symptoms. Secondly, greater symptom burden, longer fever duration, lower CR levels, and relatively higher eGFR are independent risk factors affecting the fever severity in CHIKF patients. Thirdly, older age is an independent risk factor affecting the nucleic acid conversion time in CHIKF patients, while relative increases in PLT and eGFR are protective factors.

For genomic characterization of CHIKV, we selected 92 samples having nucleic acid Ct values of less than 30 for targeted whole-genome sequencing, resulting in 92 complete genome sequences. All sequences were deposited in the GenBase database of CNCB (China National Center for Bioinformation) under accession number C_AA119371.1 - C_AA119462.1.

Phylogenetic analysis revealed that all the outbreak strains belonged to the East/Central/South African (ECSA) lineage, specifically clustering within MAL (Figure 3). The closest related strain within the same clade originated from the 2025 CHIKV outbreak on Réunion Island, France (99.96%-100% nucleotide similarity).

Mutation analysis identified between one and six mutations per strain relative to the Réunion Island reference strain 80652-1-1-1 (GenBank: PV700165.1). This included one shared synonymous mutation (C3876T). Two additionally clustered mutations—G7431A and T10430C—were detected in 29.35% (27/92) and 13.04% (12/92) of sequences, respectively (Supplementary Figure S1). Neither mutation resulted in amino acid change.

Importantly, all strains this from outbreak carried two previously reported adaptive mutations, E1-A226V and E2-I211T, both of which have been implicated in enhanced transmission efficiency by Aedes albopictus mosquitoes (Frumence et al, 2025).