Epidemiological characteristics and socioeconomic factors of sexually transmitted infections in China during 2002–2021

Abstract

Background

Sexually transmitted infections (STIs), including HIV, gonorrhea, and syphilis, pose significant public health challenges globally. In China, rising STI incidences are associated with rapid urbanization and changing sexual behaviors. This study aims to analyze the trends and socioeconomic factors of these three STIs in China.

Methods

Incidence data for HIV, gonorrhea, and syphilis from 31 provinces (2002–2021) were obtained from the China Hygiene and Health Statistics Yearbook. JoinPoint regression was used to assess trends, hotspot analysis to identify spatial clusters, global principal component analysis (GPCA) to reduce data dimensionality, and a panel fixed-effects model (FEM) to identify influential factors.

Results

The annual average percentage changes (AAPCs) in incidence were 20.68% for HIV, - 2.37% for gonorrhea, and 11.55% for syphilis. Hotspot analysis revealed persistent HIV clusters in western provinces, gonorrhea clusters in the eastern coastal and parts of central provinces, and expanding syphilis clusters from western and central provinces to the entire country. GPCA extracted three components, explaining 83% of the overall variance. PA1 includes factors such as nighttime light data (NLD) and population density (PD). PA2 includes passenger volume (PAX), the working-age population aged 15–64 (P_15-64) and the number of healthcare institutions (HCIs). PA3 encompasses GDP per capita (pGDP) and the sex ratio (SR). FEM analysis showed that socioeconomic factors had varying impacts on the three STIs. HIV was negatively associated with PA1 (β = - 1.049, P < 0.05) and positively associated with PA2 (β = 0.614, P < 0.001), and negatively associated with PA3 (β = - 0.721, P < 0.001). Gonorrhea was positively associated with PA1 (β = 16.005, P < 0.001) and PA2 had strong positive correlations in most regions, while PA3 was not significant nationally. Syphilis had a negative association with PA3 at the national level (β = - 5.517, P < 0.001).

Conclusion

This study demonstrates temporal and spatial variations in STI epidemiology in China from 2002 to 2021, reflecting socioeconomic development disparities. The findings highlight the need for geographically and socioeconomically tailored interventions to improve STI control strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-22515-8.

Introduction

Sexually transmitted infection (STI) is a major public health issue worldwide. Health organizations, such as the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC), often prioritize three infections—human immunodeficiency virus (HIV) infection, gonorrhea, and syphilis—due to their high incidence and the severe consequences of untreated cases, which affect both individual and public health [1–5]. In recent years, with the globalization of the economy and social advances, incidence of STIs continue to rise [5, 6]. Reports from the WHO on STIs outline global trends in the incidence of STIs, indicating that over one million people aged 15 to 49 acquire a curable STIs every day worldwide [7]. Furthermore, data from the CDC in the United States reveal the latest surveillance on STIs, showing increasing incidence trends both in the US and globally [8]. Additionally, research by Rowley et al. (2019) also provides epidemiological data on STIs for the year 2016, emphasizing changes observed in recent years [9]. Although many governments have launched major initiatives to prevent and control STIs [7], they remain a major, pressing public health problem worldwide [10].

Over the past thirty years, China has faced significant challenges in preventing and controlling STIs, as factors such as the acceleration of urbanization, frequent population mobility, and changes in sexual attitudes within the country have resulted in high incidence of these infections. The rapid economic development and social transformation in China have contributed to increased sexual risk behaviors, particularly among young adults and migrant populations [11]. To address this situation, the Chinese government has issued a series of policies to strengthen the prevention and control of STIs nationwide. The establishment of a national STI surveillance system in 1995 has significantly improved the monitoring and reporting of STI cases across the country [12]. Despite these efforts, STIs remain a significant public health concern in China, causing an enormous disease burden. The incidence of HIV, syphilis, and gonorrhea continues to be a major challenge, particularly in certain high-risk populations and regions [13].

HIV, gonorrhea, and syphilis exhibit high incidence rates and are legally notifiable diseases in China, with reliable long-term data available [11]. Conducting an in-depth analysis of these STIs provides a solid foundation for understanding the trends and developing prevention strategies for sexually transmitted infections in China. At the same time, we found that since the 2010 s, a resurgence of syphilis has been observed globally, with increasing rates reported in various regions such as Europe [14], the United States [15], and Japan [16]. This confirms that these trends are part of a worldwide pattern rather than being confined to China, thereby highlighting the need for contextualizing national data within this broader international framework to better understand and address the drivers of this public health challenge [17, 18].

Previous research has highlighted the multifaceted nature of STIs, demonstrating that their incidence and spread are shaped by a myriad of social, demographic, and behavioral factors [19]. The present study aimed to describe the epidemiological characteristics and spatiotemporal distribution of three selected main STIs (HIV, gonorrhea, and syphilis) in China. Additionally, it explored the strength of the association between influencing factors and these infections. Data on HIV, gonorrhea, and syphilis incidence from 31 Chinese provinces, autonomous regions, and municipalities during 2002–2021 were sourced from the China Hygiene and Health Statistics Yearbook, which compiles official statistics from healthcare institutions and surveillance systems. Additional healthcare infrastructure, demographic, and economic indicators were extracted from the China Statistical Yearbook and provincial statistical yearbooks. The national STI surveillance system, established in 1995, has improved comprehensive reporting systems, ensuring systematic data collection and analysis. Over the years, improvements in reporting systems, including electronic health records and online platforms, have expanded coverage and enhanced surveillance infrastructure. Eastern provinces have seen significant enhancements, while western and central regions have gradually improved. Despite initial challenges, ongoing efforts have improved data quality and coverage [20].

The JoinPoint regression model is widely used for analyzing public health trends, enabling the detection of significant changes in incidence rates over time through breakpoints and the estimation of APC and AAPC [21]. While powerful, it requires careful interpretation of segmented trends to avoid misidentification of breakpoints. GPCA, an extension of PCA for time-series data, effectively captures variability and patterns, reducing data dimensionality to highlight key factors. However, GPCA may overlook non-linear relationships and interactions between variables [22]. FEM robustly analyzes panel data by accounting for both individual and time effects, controlling for unobserved heterogeneity [23]. Yet, it assumes constant effects of explanatory variables over time, which may not always hold true. By leveraging the strengths of these models and addressing their limitations, our study offers a rigorous analysis of STI epidemiology in China, ensuring a nuanced understanding of the temporal and spatial variations in STI incidence [6].

Methods

Study design

This study employs a comprehensive approach to understanding the STI landscape in China by analyzing epidemiological trends of HIV, gonorrhea, and syphilis using incidence data from 2002 to 2021. The analysis units are the Chinese provinces, and the data are stratified by both time and region to identify temporal trends and regional variations. The analysis process begins with trend analysis using the JoinPoint regression model to identify significant changes in STI incidence over time and to calculate APC and AAPC, taking into account regional variations and temporal trends. Annual data on influencing factors, including economic, demographic, and hygiene determinants, were collected for each province from 2002 to 2021.GPCA is then used to reduce data dimensionality and highlight key factors, while panel FEM accounts for both individual and time effects in the data, evaluating the impact of these factors on different regions of China. Additionally, spatiotemporal distribution and hotspot analysis are performed to identify high and low incidence clusters. This approach ensures a thorough examination of both temporal and spatial patterns in STI epidemiology in China, while leveraging the strengths of each analytical method (Fig. 1).

Fig. 1.

Flowchart of the study design and analytical approach. Abbreviation: AAPC, annual average percentage changes; APC, average percentage changes; pGDP, GDP per capita; UR, urbanization rate; UUR, urban unemployment rate; PAX, passenger volume; MOU, mobile phone year-end users; NLD, nighttime light data; PD, population density; SR, sex radio; P_15–64, working-age population aged 15–64; FFY, proportion of illiterate and semi-illiterate population aged 15 and above; HCIs, healthcare institutions; Beds, beds in healthcare institutions; HCP, healthcare personnel. Note: Scree Plot, Graph showing eigenvalues to determine the number of principal components to retain; F-Test, Statistical test assessing the overall significance of a regression model; Hausman Test, Test to choose between fixed effects and random effects models in panel data

Data source

Data on STIs (HIV, gonorrhea, syphilis) from 31 provinces, autonomous regions, and municipalities in China during 2002–2021 were obtained from the China Hygiene and Health Statistics Yearbook [20]. Data on various indicators, including the number of beds in healthcare institutions (Beds), healthcare personnel (HCP), healthcare institutions (HCIs), GDP per capita (pGDP), population density (PD), urbanization rate (UR), nighttime light data (NLD), working-age population aged 15–64 (P_15–64), Mobile phone year-end users (MOU), passenger volume(PAX), Urban unemployment rate (UUR), Sex radio (SR), Proportion of illiterate and semi-illiterate population aged 15 and above (FFY).were extracted from the China Statistical Yearbook [24] and the statistical yearbooks of various provinces covering the study period.

JoinPoint model

The JoinPoint regression model was used to establish segmented regression based on the temporal characteristics of disease distribution. This involved identifying several connecting points to divide the study period into different intervals. Trend fitting and optimization were conducted for each interval to evaluate the changes in STIs rates across different intervals. The main outcome indicators included the APC and AAPC. Applying the JoinPoint model to fit the incidence of STIs in each province of China from 2002 to 2021, segmented trends of STIs across provinces were derived.

Spatiotemporal distribution and hot spot analysis

We conducted spatiotemporal distribution and hot spot analyses using ArcGIS Pro 2.5 to examine the distribution and clustering patterns of HIV, gonorrhea, and syphilis in China from 2002 to 2021. For spatiotemporal distribution analysis, we created choropleth maps to visualize the incidence of each STI across provinces, allowing for the identification of high and low incidence areas. To detect statistically significant spatial clusters, we performed hot spot analysis using the Getis-Ord Gi* statistic. This analysis was conducted for five time points (2002, 2007, 2012, 2017, and 2021) to capture temporal changes in clustering patterns. The resulting maps highlighted hot spots (high-high clusters) and cold spots (low-low clusters) for each STI, revealing the spatiotemporal shifts in disease concentration.

Time-series global principal component analysis

GPCA is an extended principal component analysis method designed for dealing with time-varying data series, capturing variability and patterns effectively. It considers the variance of both the overall dataset and each time point. In consideration of the influencing factors of STIs, including the economic and demographic variables and the healthcare levels in each province, 13 indicators and their corresponding panel data were selected for GPCA. Eigenvalues were computed, and the cumulative variance contribution ratio was derived using the Kaiser–Meyer–Olkin (KMO) and Bartlett’s sphericity tests. Factors were determined based on the principle that the cumulative variance contribution rate should be greater than or equal to 80%. After utilizing a scree plot to determine the number of principal components and conducting factor analysis, we identified the principal components by eliminating unqualified influencing factors. This process resulted in the determination of the specific influencing factors contained within each principal component. Analyzing the factor scores facilitates the identification of the underlying structural determinants contributing to the development of HIV, gonorrhea, and syphilis across different provinces.

Panel fixed effects model

For our analysis, we employed FEM, a panel data analysis method that considers both individual and time dimensions to study variable relationships while controlling for individual fixed effects. We applied this method to panel data from 31 Chinese provinces spanning 2002 to 2021. Following the economic zones in the statistical system and classification standards, we divided the country into four regions: the East (including Beijing, Tianjin, Hebei, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and Hainan), the Center (including Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan), the West (Inner Mongolia Autonomous Region, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang), and the Northeast (Liaoning, Jilin, and Heilongjiang). Our methodology involved first testing the panel model form using the F-statistic, then applying the Hausman test to choose between fixed or random effect models. Subsequently, we conducted panel fixed effect regression, utilizing the main factors derived from GPCA to regress with the three types of STIs separately, producing panel regression results for the whole country and the four regions. All GPCA analysis and panel fixed-effects modeling were performed using R software (version 4.3.3).

Results

Trends of HIV, gonorrhea, and syphilis in China

From 2002 to 2021, significant changes have taken place in the epidemiological trends of HIV, gonorrhea, and syphilis incidence in China. The HIV incidence experienced a steady increase from 0.06 in 2002 to a peak of 5.10 in 2019, before declining slightly to 4.27 in 2021. The gonorrhea incidence initially increased from 13.28 in 2002 to a peak of 17.34 in 2004, followed by a general decline to a low of 6.82 in 2012. Thereafter, it exhibited some fluctuations, reaching 9.07 in 2021. The syphilis incidence saw a consistent rise from 4.67 in 2002 to 38.37 in 2019. There was a slight decrease in 2020 to 33.08, but it rebounded to 34.05 in 2021 (Fig. 2).

Fig. 2.

Epidemiological trends and AAPC diagram from JoinPoint model of sexually transmitted infections

The trends in the incidence of HIV, gonorrhea, and syphilis in China from 2002 to 2021 demonstrate significant variations, as illustrated in Fig. 2, Figure S1 and Table S1. HIV incidence exhibited a pronounced overall AAPC of 20.68%. The APC increased by 34.04% from 2002 to 2013, followed by a more moderate increase of 4.46% from 2013 to 2021. Geographically, the AAPC of HIV increased across all provinces, with Chongqing showing the highest rise (39.99%), while Anhui reported the smallest AAPC at 14.36%. Gonorrhea incidence generally demonstrated a decreasing trend with an overall AAPC of − 2.37%. The incidence displayed four distinct trends with three breakpoints: an increase with an APC of 10.78% during 2002–2004, followed by a decrease with an APC of 10.92% during 2004–2012, a second increase with an APC of 8.06% from 2012 to 2017, and a minor decline of 3.00% from 2017 to 2021. Province-specific AAPC varied notably, with Hainan showing the highest upward trend (AAPC = 6.54%), while Shanghai exhibited the most significant decrease (AAPC = − 10.08%). Syphilis incidence exhibited an overall AAPC of 11.55%, with four distinct phases: a rapid increase with an APC of 31.33% during 2002–2007, followed by a moderate increase of 19.30% from 2007 to 2010, a slower increase of 3.21% from 2010 to 2019, and a decline of − 4.85% from 2019 to 2021. Geographically, all provinces experienced rising AAPC for syphilis, with Guizhou reporting the largest increase (AAPC = 25.21%), while Shanghai showed the lowest significant increase at 1.4060%. These findings underscore the dynamic nature of STI epidemiology in China, influenced by regional factors and evolving over time across different provinces.

The proportional distribution of STIs in China underwent significant changes from 2002 to 2021. Nationally, HIV’s share decreased from 33.31% to 9.01%, while gonorrhea saw a dramatic reduction from 73.74% to 19.14%. Conversely, syphilis exhibited a substantial increase, rising from 25.93% to 71.85% of total STIs. Regional variations were notable: in 2021, Guangxi reported the highest HIV proportion (24.33%), Zhejiang led in gonorrhea (38.49%), and Qinghai had the highest syphilis share (91.54%). Conversely, Tibet had the lowest HIV proportion (1.46%), Xinjiang the lowest gonorrhea (5.08%), and Guangxi the lowest syphilis (48.74%). These shifts underscore a changing STIs landscape across China, with syphilis emerging as the predominant STIs in most regions, while HIV and gonorrhea show varying trends (Figure S1).

The spatiotemporal distribution and hot spot analysis of STIs

The spatial and temporal distribution of HIV, gonorrhea, and syphilis in China shows distinct regional patterns (Figure S3). In 2002, HIV incidence was generally low across the country, with slight concentrations in western provinces. By 2007, incidence had increased slightly in some western regions. This trend continued through 2017 and 2021, with western provinces, especially Yunnan, maintaining the highest rates. For gonorrhea, the highest rates in 2002 were in eastern coastal provinces like Zhejiang and Jiangsu. This trend persisted through 2017, though rates began to decline by 2021 but remained relatively high in these areas. Syphilis showed significant changes over time, with higher incidence in 2002 scattered across provinces but concentrating in the east and northeast regions. By 2007, rates intensified, particularly in western and northeastern provinces, and this trend continued through 2021, spreading to central and eastern provinces.

The hotspot clusters of STIs from 2002 to 2021 reveal marked regional variations. For HIV, hotspots in 2002 were in the western provinces, especially Guangxi and Yunnan, and remained in the western and central regions through 2021, with Yunnan showing strong hotspot activity. Gonorrhea hotspots were concentrated in eastern coastal provinces in 2002, such as Zhejiang and Jiangsu, and persisted through 2021 with decreasing intensity. Syphilis hotspots were initially scattered in 2002, with notable activity in western provinces, expanding significantly by 2007 and becoming even more widespread by 2021, indicating a persistent increase in these regions (Fig. 3).

Fig. 3.

Local hotspot clusters of the annual incidence of sexually transmitted infection in 31 provinces in China in 2002, 2007, 2012, 2017 and 2021. Abbreviated note in the picture: BJ, Beijing; TJ, Tianjin; HB, Hebei; SX, Shanxi; NMG, Inner Mongolia; LN, Liaoning; JL, Jilin; HLJ, Heilongjiang; SH, Shanghai; JS, Jiangsu; ZJ, Zhejiang; AH, Anhui; FJ, Fujian; JX, Jiangxi; SD, Shandong; HN, Henan**; HB*, Hubei; HN, Hunan; GD, Guangdong; GX, Guangxi; HN*, Hainan; CQ, Chongqing; SC, Sichuan; GZ, Guizhou; YN, Yunnan; XZ, Tibet; SX*, Shaanxi; GS, Gansu; QH, Qinghai; NX, Ningxia; XJ, Xinjiang

Factor extraction via GPCA

After processing the collected variables, we calculated the Variance Inflation Factor (VIF) and used stepwise regression to remove variables with VIF ≥ 5. The remaining variables were PD, pGDP, HCIs, NLD, PAX, P15 - 64, SR, FFY, and UUR. Subsequently, we incorporated all influencing factors into the scree plot to aid in determining the number of principal components or factors to retain. After a comprehensive evaluation, we concluded that three principal components should be selected for our analysis. The principal component analysis (PCA) was conducted to evaluate factors potentially influencing STIs in China. The KMO test (0.59) and Bartlett’s test of sphericity (significant with P < 0.001) confirmed data suitability for PCA. Three principal components with a cumulative contribution rate of 80% were selected. The factor model showed the correlations between the factors, where PA1 comprises factors such as PD and NLD; PA2 includes PAX, P15 - 64, and HCIs; and PA3 encompasses pGDP and SR. Simultaneously, we conducted a principal component analysis, and the correlation coefficients among the various influencing factors are presented in Fig. 4a, providing a clearer depiction of the relationships between the factors and the incidence across different regions (Table S2, Fig. 4).

Fig. 4.

The correlations between the factors. Note: ^* indicates a significance level of ≤ 0.05; ^** indicates a significance level of ≤ 0.01; ^*** indicates a significance level of ≤ 0.001. Abbreviation: PA1 includes factors such as nighttime light data (NLD) and population density (PD). PA2 includes passenger volume (PAX), the working-age population aged 15–64 (P_15–64) and the number of healthcare institutions (HCIs). PA3 encompasses GDP per capita (pGDP) and the sex ratio (SR)

Factor analysis of economic, hygiene, and demographic factors across provinces in China revealed notable disparities. For PA1, which comprises factors such as PD and NLD, Shanghai has the highest factor score at 4.184, while Tibet has the lowest score at − 0.765. For PA2, which includes PAX, P_15–64, and HCIs, Guangdong has the highest factor score at 2.175, whereas Tibet has the lowest score at − 1.439. For PA3, which encompasses pGDP and SR, Beijing has the highest factor score at 1.562, while Henan has the lowest score at − 0.507. These results provide insights into the varying levels of healthcare infrastructure, economic activity, and urban development across different provinces in China (Table 1).

Table 1.

Factor analysis of PA1, PA2 and PA3 by province in China

Province	Factor score of PA1	Province	Factor score of PA2	Province	Factor score of PA3
Shanghai	4.184	Guangdong	2.175	Beijing	1.562
Tianjin	1.827	Shandong	1.909	Fujian	0.665
Beijing	1.030	Henan	1.780	Zhejiang	0.582
Jiangsu	0.903	Sichuan	1.405	Jiangsu	0.535
Shandong	0.611	Jiangsu	1.313	Inner Mongolia	0.438
Guangdong	0.345	Hebei	1.012	Guangdong	0.221
Zhejiang	0.324	Hunan	0.811	Chongqing	0.148
Henan	0.288	Hubei	0.640	Tianjin	0.129
Hebei	0.119	Anhui	0.612	Hubei	0.106
Anhui	0.022	Zhejiang	0.448	Xinjiang	0.072
Hainan	− 0.149	Yunnan	0.118	Shaanxi	0.062
Liaoning	− 0.164	Guangxi	0.106	Liaoning	0.014
Shanxi	− 0.170	Liaoning	0.100	Qinghai	− 0.062
Fujian	− 0.288	Jiangxi	0.000	Ningxia	− 0.098
Hubei	− 0.354	Heilongjiang	− 0.046	Hunan	− 0.103
Chongqing	− 0.356	Fujian	− 0.145	Sichuan	− 0.117
Guangxi	− 0.394	Shaanxi	− 0.157	Tibet	− 0.124
Ningxia	− 0.394	Shanxi	− 0.285	Shanghai	− 0.131
Hunan	− 0.397	Guizhou	− 0.349	Jilin	− 0.145
Jiangxi	− 0.398	Chongqing	− 0.492	Shandong	− 0.152
Guizhou	− 0.399	Jilin	− 0.499	Jiangxi	− 0.197
Shaanxi	− 0.399	Inner Mongolia	− 0.554	Hainan	− 0.200
Jilin	− 0.477	Gansu	− 0.631	Shanxi	− 0.213
Yunnan	− 0.507	Xinjiang	− 0.731	Heilongjiang	− 0.231
Heilongjiang	− 0.545	Beijing	− 0.847	Yunnan	− 0.292
Sichuan	− 0.562	Shanghai	− 1.010	Anhui	− 0.358
Gansu	− 0.601	Tianjin	− 1.236	Hebei	− 0.363
Xinjiang	− 0.757	Hainan	− 1.308	Gansu	− 0.384
Qinghai	− 0.758	Qinghai	− 1.345	Guangxi	− 0.390
Tibet	− 0.765	Ningxia	− 1.352	Guizhou	− 0.468
Inner Mongolia	− 0.820	Tibet	− 1.439	Henan	− 0.507

Abbreviations: PA1 includes factors such as nighttime light data (NLD) and population density (PD). PA2 includes passenger volume (PAX), the working-age population aged 15–64 (P_15–64) and the number of healthcare institutions (HCIs). PA3 encompasses GDP per capita (pGDP) and the sex ratio (SR)

Identifying the determinants from a panel fixed-effects regression

A fixed-effects panel data regression model was utilized to assess the impact of two principal components on STIs across various regions in China. For HIV, the results show strong negative correlations with PA1 at the national level (β = − 1.049, P < 0.05) and in the eastern region (β = − 0.880, P < 0.001). PA2 shows a significant positive correlation at the national level (β = 0.614, P < 0.001) and in the central region (β = 0.478, P < 0.001). PA3 demonstrates consistent negative correlations with HIV at all levels, including the national level (β = − 0.721, P < 0.001), suggesting that higher economic status and urbanization rates are associated with lower HIV prevalence. For gonorrhea, PA1 exhibits a significant positive correlation at the national level (β = 16.005, P < 0.001) and in the eastern (β = 20.192, P < 0.001) and western regions (β = 20.304, P < 0.001). PA2 shows significant negative correlations in the central region (β = − 1.897, P < 0.001) and positive correlations in the northeastern region (β = 1.462). PA3 has a significant positive correlation in the northeastern region (β = 3.291, P < 0.001). For syphilis, PA1 shows significant negative correlations in the central (β = − 49.625, P < 0.001) and western regions (β = − 81.014, P < 0.001). PA2 demonstrates strong positive correlations in the northeastern (β = 18.435, P < 0.001) and western regions (β = 11.270, P < 0.001). PA3 has significant negative correlations at all levels, including the national level (β = − 5.517, P < 0.001), suggesting that higher GDP per capita and urbanization rates are associated with lower syphilis rates (Table 2).

Table 2.

Panel fixed effects model correlation coefficients for sexually transmitted infection across national and regional levels

	National	Eastern	Central	Western	Northeastern
N	620	200	120	240	60
HIV
PA1	− 1.049a (0.577)	− 0.880c (0.198)	− 4.083c (0.664)	− 24.715c (2.776)	− 5.644c (1.280)
PA2	0.614c (0.206)	0.042 (0.070)	0.478c (0.155)	1.589b (0.736)	− 1.255c (0.362)
PA3	− 0.721c (0.169)	− 0.369c (0.064)	− 0.513c (0.177)	− 0.513c (0.363)	− 1.544c (0.264)
Gonorrhea
PA1	16.005c (1.482)	20.192c (2.860)	1.309 (1.156)	20.304c (4.138)	12.917c (4.391)
PA2	− 1.356b (0.529)	− 1.264 (1.010)	− 1.897c (0.270)	− 1.676 (1.097)	1.462 (1.243)
PA3	0.178 (0.434)	− 1.235 (0.929)	0.253 (0.308)	1.487c (0.541)	3.291c (0.904)
Syphilis
PA1	− 3.368 (3.454)	− 13.406c (3.825)	− 49.625c (5.746)	− 81.014c (3.825)	− 38.276c (11.287)
PA2	9.408c (1.233)	6.314c (1.351)	6.706c (1.341)	11.270c (3.650	18.435c (3.194)
PA3	− 5.517c (1.012)	0.902 (1.243)	− 4.816c (1.530)	− 11.827c (1.798)	− 9.384c (2.324)

The general formula for panel regression model, assuming a fixed-effects or random-effects model based on tests like Hausman test, can be represented as: $Y_{\mathit{it}}={\alpha }_i+{\beta }_1{X}_{1it}+{\beta }_2{X}_{2it}+{\beta }_3{X}_{3it}{+\epsilon }_{\mathit{it}}$

$Y_{\mathit{it}}$ is the dependent variable for entity (i) at time (t)

${\alpha }_i$ is a fixed effect for province (i)

$X_{1it}$ to $X_{3it}$ are the independent variables for entity (i) at time (t), representing PA1 includes factors such as nighttime light data (NLD) and population density (PD). PA2 includes passenger volume (PAX), the working-age population aged 15-64 (P15-64) and the number of healthcare institutions (HCIs). PA3 encompasses GDP per capita (pGDP) and the sex ratio (SR)

${\beta }_1\sim {\beta }_3$ are the coefficients of the independent variables

${\epsilon }_{\mathit{it}}$ is the error term

^aindicates a significance level of ≤ 0.05

^bindicates a significance level of ≤ 0.01

^cindicates a significance level of ≤ 0.001

Discussion

The findings of this study underscore the complex and multifaceted nature of STIs in China, revealing significant temporal trends and spatial variations in the incidence of HIV, gonorrhea, and syphilis from 2002 to 2021. These results align with previous research highlighting the dynamic epidemiology of STIs and the influence of various social, behavioral, and healthcare factors on their transmission and incidence [5, 25, 26]. While some progress has been made in controlling certain STIs, the persistent high incidence of others, particularly syphilis, indicate ongoing challenges in STI prevention and control efforts. The regional variations in epidemiological trends and the differential impacts of economic factor and healthcare factors underscore the need for tailored, location-specific interventions [27, 28]. Moreover, the emergence of new hotspots and the shifting proportional distribution of these diseases over time highlight the importance of adaptive and responsive public health strategies [1, 29].

Over the past 20 years, the incidence of HIV demonstrated a pronounced overall increase with an AAPC of 20.68%, reaching its peak in 2019 before slightly declining. In contrast, gonorrhea showed a general decreasing trend (AAPC = − 2.37%) with fluctuations, while syphilis exhibited a substantial increase (AAPC = 11.55%) throughout the study period. Notably, by 2021, syphilis had emerged as the predominant STI, accounting for 71.85% of total STIs, up from 25.93% in 2002. The analysis of spatial and temporal distribution revealed distinct regional patterns: HIV high-incidence clusters were found in northwestern and southwestern regions, gonorrhea incidence was pronounced in southeastern coastal areas and western China, and syphilis transmission was concentrated in western, eastern, and central China. These hotspot clusters from 2002 to 2021 demonstrated marked regional variations. The availability and accessibility of STI diagnosis and treatment services significantly impact these regional patterns. Variations in healthcare infrastructure, public health outreach, and resource allocation can lead to differences in detection and reported incidence, influencing the observed regional patterns in STI hotspots. The GPCA identified three key principal components, PA1, PA2, and PA3, influencing the incidence of STIs, which encompass economic factors, healthcare infrastructure, and demographic factors. Additionally, the interrelationships among these factors were calculated, enabling a more accurate prediction of the complex relationships between the influencing factors and the incidence of STIs. The FEM data regression model revealed complex interactions between these factors and STI incidence, with varying effects across different regions and diseases. It is crucial to acknowledge that these data rely on individuals seeking testing and care to detect and document cases. Furthermore, the observed trends in STI incidence during 2020 and 2021 were greatly influenced by the COVID- 19 pandemic. The pandemic led to widespread reductions in STI testing, changes in sexual behavior due to public health interventions such as lockdowns, and subsequent variations in STI transmission dynamics. This phenomenon was documented in studies across multiple regions, including Italy, the United States, Australia, and Spain, underscoring the importance of considering these factors in understanding STI epidemiology during this period [30–33].

Our findings corroborate and extend several previous studies while also offering novel insights into the dynamic landscape of STIs in China. The observed overall increase in HIV incidence aligns with the spatial–temporal analysis conducted by Huang et al. (2022), which examined HIV/AIDS trends in mainland China from 2007 to 2017 [6]. However, our study expands this observation to a more extensive temporal framework and identifies similar trend variations. Our investigation identified distinct temporal inflection points in gonorrhea incidence, largely corroborating the observations made by Wang et al. (2023) in their analysis of incidence variations from 2005 to 2021 [29]. Our study, however, extends the temporal scope by commencing from 2002, thereby encompassing a broader timeframe. Furthermore, we offer a more comprehensive spatiotemporal analysis of disease trends, enhancing the depth and breadth of understanding in this domain. The marked escalation in syphilis incidence substantiates the findings of Tucker et al. (2010), who chronicled the resurgence of syphilis in China [34]. Our research presents an updated and more granular analysis of this trend, revealing its sustained acceleration through 2021. The spatial patterns elucidated in our study partially correspond with the research of Yang et al. (2012), who identified comparable geographic distributions of HIV in China [35]. However, our analysis provides a more holistic view by incorporating gonorrhea and syphilis, thereby uncovering distinct spatial patterns for each STI. The impact of demographic, economic, and healthcare factors on the incidence of STIs has been extensively examined in existing literature. These studies have highlighted the multifaceted nature of these influences, demonstrating how various social determinants and healthcare infrastructure can significantly affect the spread and prevalence of STIs in different populations [36–38].

The principal strength of this study resides in its comprehensive spatiotemporal analysis of three major STIs in China over a two-decade period. The implementation of advanced statistical methodologies, including principal component analysis and fixed-effects panel data regression, facilitates a nuanced understanding of the multifaceted factors influencing STI incidence. The integration of demographic factors, economic factors and healthcare infrastructure, into the GPCA and FEM analytical framework provides a multidimensional perspective on the factors influencing the occurrence and development of STIs. This approach allows for the formulation of targeted interventions tailored to different scenarios, thereby more effectively reducing the incidence of STIs. The extensive geographical coverage, encompassing all mainland China provinces, enables a thorough examination of regional variations in STI incidence and associated factors. Furthermore, our research design’s capacity to capture long-term trends while accounting for short-term fluctuations represents a significant methodological advancement in the field.

This study provides valuable insights into the spatiotemporal trends and regional variations in STI incidence in China, but its ecological design has several limitations. Aggregated provincial data preclude examination of individual risk factors and behaviors, potentially leading to an ecological fallacy. Additionally, this design masks within-province variations and underestimates the true STI burden due to underreporting and undiagnosed cases, particularly for HIV and syphilis, where asymptomatic infections are prevalent [39]. While the study identifies key factors influencing STI incidence, it does not fully account for behavioral and social determinants that play crucial roles in STI transmission [40]. Meanwhile, although the FEM controls for time-invariant factors, it does not establish causality. The results should be interpreted as associations rather than causal relationships. Future studies should employ longitudinal designs or instrumental variable approaches to better assess causal relationships. Furthermore, the focus on province-level data may obscure important within-province variations, especially in large and diverse provinces [41]. Finally, the analysis does not consider potential interactions between the three STIs, which could influence their respective transmission dynamics [42]. Therefore, the findings may not fully capture the complexities of STI transmission dynamics, limiting their utility for targeted interventions. Future research should include individual-level data to improve public health strategies.

Our analysis reveals higher STI rates in provinces with better healthcare resources, likely due to more robust surveillance systems and higher reporting accuracy [4, 43]. However, this could also indicate actual incidence increases, possibly driven by increased risk behaviors or population mobility. To differentiate between improved detection and true burden, future research should compare regional testing rates with reported incidence rates, helping to determine whether the higher rates are primarily due to better detection or underlying factors [29, 44].

The findings of this study have several important policy implications for STI prevention and control in China. To effectively address the distinct spatial patterns observed for each STI, geographically tailored prevention and control strategies are essential. Specifically, expanding HIV testing programs to increase the availability and accessibility of testing in high-incidence regions, such as through mobile testing units, community-based testing, and integration into routine healthcare services, is crucial [45]. Additionally, comprehensive syphilis screening and treatment programs should be implemented, particularly in areas with high prevalence, including prenatal screening for pregnant women, contact tracing, and partner notification services to interrupt transmission chains [46]. For high-risk populations, such as sex workers, men who have sex with men (MSM), and intravenous drug users, harm reduction strategies like needle exchange programs, condom distribution, and peer-led education initiatives should be developed and implemented [47]. Strengthening healthcare infrastructure is also vital, which can be achieved by enhancing the quality and accessibility of STI-specific services, including testing, treatment, and prevention programs. This includes training healthcare providers, ensuring adequate supplies of diagnostic tools and medications, and reducing barriers to care, such as cost and stigma [27]. Integrating STI services into primary care settings will further ensure that these services are accessible to a broader population, providing STI education, counseling, and promoting regular screening [48]. Furthermore, conducting further research to evaluate the effectiveness of these interventions at both the population and individual levels, and establishing robust monitoring and evaluation systems to track the progress and outcomes of STI prevention and control programs, will help identify areas for improvement and ensure effective resource use [49, 50]. By implementing these specific and targeted interventions, policymakers can more effectively address the unique challenges posed by STIs in different regions of China. Implications for Further Research: Our findings suggest that demographic, hygiene, and economic factors influence the incidence of STIs. This highlights the need for further studies to investigate how climate change adaptation strategies can be systematically integrated into STI control efforts. Such integration may involve evidence-based modifications to the timing and intensity of prevention measures to address the societally driven changes in STI transmission. While our ecological design provides a valuable starting point, more detailed studies at the individual and social levels are necessary to inform such strategies [51]. Integrated Approach: The varying effects of factors across different STIs and regions call for an integrated approach to STI control. Policies should promote collaboration between different health sectors and incorporate multi-disease strategies rather than focusing on individual STIs in isolation [52]. Enhanced Surveillance: The dynamic nature of STI epidemiology revealed in this study emphasizes the importance of robust and responsive surveillance systems. Policies should support the development and maintenance of real-time monitoring systems to detect and respond to emerging hotspots and changing trends [4]. Research priorities: The study’s findings highlight several areas for future research, including investigating the factors behind the decline in Gonorrhea incidence, exploring the reasons for the dramatic rise in syphilis cases, and examining the complex interactions between economic factors and STI transmission [53].

Conclusion

The study identifies significant trends in HIV, gonorrhea, and syphilis incidence in China from 2002 to 2021, with HIV and syphilis rates generally increasing and gonorrhea declining. Regional disparities were evident, with variations across provinces. Socioeconomic factors influenced STI incidence, underscoring the need for tailored prevention strategies. Given the ecological design, findings should be interpreted with caution and suggest areas for further investigation. This analysis offers valuable insights for policymakers to develop more effective STI control and prevention strategies.

Abbreviations

AAPC

Annual average percentage changes

APC

Average percentage changes

Beds

Beds in healthcare institutions

CDC

Centers for Disease Control and Prevention

FEM

Fixed-effects model

FFY

Proportion of illiterate and semi-illiterate population aged 15 and above

GDP

Gross domestic product

GPCA

Global principal component analysis

HCP

Healthcare personnel

HCIs

Healthcare institutions

HIV

Human immunodeficiency virus

KMO

Kaiser–Meyer–Olkin

MOU

Mobile phone year-end users

MSM

Men who have sex with men

P_15-64

Number of people aged 15–64

NLD

Nighttime light data

PAX

Passenger volume

PD

Population density

STI

Sexually transmitted infection

SR

Sex radio

UUR

Urban unemployment rate

UR

Urbanization rate

VIF

Variance inflation factor

WHO

World Health Organization

Authors’ contributions

GQ, BZ and Y-HJ conceived and designed the research. J-TS, TJ, C-YL, M-YZ, and X-YZ collected the data for analysis. J-TS, TJ, C-YL, M-YZ, X-YZ, XZ, carried out the statistical analysis. J-TS,TJ and C-YL wrote the first draft of the manuscript. GQ, Y-HJ and BZ made the key revision. All authors contributed to the scientific discussions and approved the final draft. The corresponding authors attest that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Funding

BZ and GQ acknowledge support from the Jiangsu Provincial Research Hospital (grant no. YJXYY202204-YSB09 and YJXYY202204-YSA03). Y-HJ acknowledges support from the Science and Technology Project of Nantong City (grant no. MS22022086). X-YZ acknowledges support from the 2023 Preventive Medicine Research Project of Jiangsu Commission of Health (grant no. Ym2023079). XZ acknowledges support from the Ministry of Science and Technology of China (grant no. 2022YFC2304901). The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding authors had full access to all study data and materials and had final responsibility for the decision to submit for publication.

Data availability

Data are provided in the article or uploaded as online supplemental information. All study datasets analyzed for this manuscript are available from the corresponding author, Prof. Gang Qin, upon reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

Gang Qin is an editorial board member of BMC Public Health. The remaining authors have no conflicts of interest to declare.