Long-Term Dynamics of Pelvic Inflammatory Disease Attributable to Chlamydia trachomatis in Reproductive Aged Women: A Population-Based Study

Abstract

Background

Pelvic inflammatory disease (PID) caused by Chlamydia trachomatis (C. trachomatis) remains a significant global reproductive health challenge. Understanding its variation across age groups, time periods, and geographic and sociodemographic contexts is essential for effective prevention strategies.

Methods

Estimates of C. trachomatis-related PID among women aged 15–49 years from 1990 to 2021 were analyzed. Temporal trends were quantified using estimated annual percent change (EAPC), and associations with the Sociodemographic Index (SDI) were examined. Future prevalence through 2050 was projected using autoregressive integrated moving average (ARIMA) models.

Results

Between 1990 and 2021, the global number of PID cases attributable to C. trachomatis increased from 143,258 to 232,246 (+62.1%), while prevalence remained relatively stable, rising from 10.71 to 11.92 per 100,000 (EAPC = 0.10). The highest prevalence in 2021 occurred among women aged 30–34 years (17.09 per 100,000), with women aged 25–39 years accounting for most cases. Low-SDI regions had the highest prevalence (16.83 per 100,000), whereas middle-SDI regions showed the fastest increase (EAPC = 0.76). Substantial geographic heterogeneity was observed, with particularly high prevalence in Australasia and Sub-Saharan Africa. Prevalence was nonlinearly associated with SDI (ρ = −0.464). By 2050, the burden is projected to increase most among women aged 30–39 years, reaching 34.49 per 100,000, with low-SDI regions projected to reach 38.51 per 100,000 and the future burden increasingly concentrated in Sub-Saharan Africa.

Conclusions

While global prevalence has remained stable, the absolute number of cases continues to rise, particularly among women of mid-reproductive age and in specific low- and high-SDI regions. These findings highlight the need for context-specific and targeted interventions to address the growing burden of chlamydia-related PID.

Plain Language Summary

Pelvic inflammatory disease is an infection of the female reproductive organs that can cause long-term health problems, including infertility and chronic pelvic pain. Chlamydia trachomatis (C. trachomatis), a common sexually transmitted infection, is one of the major causes of pelvic inflammatory disease. However, how the global burden of this disease has changed over time and how it may develop in the future remain unclear.In this study, we analyzed global data on C. trachomatis-related pelvic inflammatory disease among women aged 15–49 years from 1990 to 2021. We assessed changes in disease burden across different age groups, regions, and levels of social and economic development, and predicted future trends through 2050. We found that the total number of cases increased substantially worldwide, although the overall prevalence remained relatively stable. Women aged 25–39 years experienced the greatest burden, and several regions, particularly in sub-Saharan Africa, showed a higher current and future burden.These findings show that C. trachomatis-related pelvic inflammatory disease remains an important global reproductive health concern. The continued increase in the number of cases highlights the need for targeted prevention strategies, including improved screening, early diagnosis, and timely treatment. Focusing resources on high-risk populations and regions may help reduce future disease burden and protect women’s reproductive health.

Introduction

Pelvic inflammatory disease (PID) is a common and consequential syndrome of the upper female genital tract that can lead to tubal factor infertility, ectopic pregnancy, and chronic pelvic pain.^1,2 Among women of reproductive age, a substantial share of PID arises in the context of urogenital Chlamydia trachomatis (C. trachomatis) infection, which is frequently asymptomatic, easily transmissible, and concentrated in adolescents and young adults.³ Ascending infection can produce endometritis and salpingitis, with inflammatory damage that underpins both acute morbidity and long-term sequelae.⁴ Because PID is polymicrobial, the policy-relevant quantity is not overall PID alone but the portion specifically attributable to C. trachomatis, which more directly links pathogen-focused prevention and treatment to reproductive health outcomes.^5,6

Over the past decade, important advances have reshaped chlamydia control: wider access to nucleic acid amplification testing (including self-collected sampling and dual assays), expedited partner therapy, digital partner notification, and faster treatment pathways.⁷ Point-of-care diagnostics are increasingly available, and community-based screening models have broadened reach in some settings.⁸ Yet progress remains uneven. Screening coverage and partner services vary widely, reinfection is common, and changes in testing technology, coding practices, and care-seeking can blur true epidemiologic trends.^8,9 Co-pathogens such as Neisseria gonorrhoeae and Mycoplasma genitalium complicate clinical attribution, while antimicrobial stewardship considerations underscore the need to target interventions where they will yield the largest reductions in PID and its sequelae.¹⁰

A clearer understanding of prevalence dynamics-how the share of PID attributable to C. trachomatis varies by age within the reproductive window, shifts over calendar time, and differs across settings-can inform the timing and targeting of screening, treatment, and partner management.¹¹ Robust estimation requires harmonized outcome definitions, consistent population denominators, and explicit handling of uncertainty from both exposure distributions and effect-size inputs used to derive population-attributable fractions.^12,13 It also calls for analytic strategies that accommodate overdispersion and between-setting heterogeneity so that observed differences reflect epidemiology rather than measurement artifacts. We included the Sociodemographic Index (SDI) to reflect national development levels, hypothesizing that differences in SDI may underlie global disparities in chlamydia-related PID burden.

In this study, we used internationally comparable, harmonized data with laboratory-based case definitions to quantify the prevalence of PID attributable to urogenital C. trachomatis infection among women aged 15=49 years.¹⁴ Specifically, we examined age-specific patterns, temporal trends from 1990 to 2021, and regional differences in relation to sociodemographic development. By focusing on etiologic attribution rather than overall PID alone, this analysis aims to provide evidence that informs targeted screening, partner management, and reproductive health strategies.

Materials and Methods

Study Design and Setting

We conducted a multinational, population-based analysis of PID attributable to urogenital C. trachomatis infection among women of reproductive age (15–49 years). Annual estimates were compiled for all locations with complete age-sex time series and harmonized to five-year age bands (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, and 45–49 years). The analytic unit was country-age-year, and we additionally summarized results at regional groupings to facilitate cross-setting comparison.

Inputs were drawn from the Global Burden of Disease (GBD) study maintained by the Institute for Health Metrics and Evaluation (GBD Results Tool: https://ghdx.healthdata.org/gbd-results-tool), downloaded on 20 July 2025.^14–16 From the most recent round available at that time, we extracted impairment-specific prevalence where the GBD Estimate type was “Impairment” with Impairment = pelvic inflammatory disease and Cause = chlamydial infection (urogenital C. trachomatis), expressed per 100,000 by age, sex, year, and location.¹⁷(p288)

The GBD data are derived from multiple population-based sources, including hospital inpatient and outpatient records, primary care data, national surveillance systems, household surveys, and published epidemiological studies.^14,18 These heterogeneous data inputs are standardized and harmonized through the DisMod-MR 2.1 Bayesian meta-regression framework to generate internally consistent national estimates of incidence, prevalence, and remission. For countries lacking laboratory diagnostic capacity—such as those relying mainly on syndromic management in parts of Sub-Saharan Africa—GBD does not exclude these settings. Instead, estimates are generated using covariate-based modeling and regional borrowing of strength, incorporating information from neighboring countries and relevant predictors (eg, healthcare access, HIV prevalence, and reproductive health service coverage) to ensure complete global representation.

In the GBD analytical system, impairments such as PID are linked to their etiologic causes through a standardized impairment–cause hierarchy. The burden of PID attributable to C. trachomatis is estimated within the DisMod-MR 2.1 Bayesian meta-regression framework, which ensures internal consistency among epidemiological parameters (incidence, prevalence, remission, and excess mortality).^17,18 Pathogen-specific relative risks and population-attributable fractions (PAFs) derived from systematic reviews and meta-analyses are applied to partition the total PID burden by cause. Accordingly, the extracted prevalence values represent modeled estimates of PID causally attributed to urogenital C. trachomatis, rather than all-cause PID.¹⁵ This approach aligns with the GBD 2021 methodology for quantifying cause-specific impairment prevalence and enables cross-country comparability.

Case Definition

PID was defined as an infectious-inflammatory syndrome of the upper female genital tract-encompassing endometritis, salpingitis, tubo-ovarian abscess, and pelvic peritonitis-that typically presents with lower abdominal or pelvic pain and can lead to tubal factor infertility, ectopic pregnancy, and chronic pelvic pain. Diagnostic standardization across countries was achieved using the International Classification of Diseases (ICD) system, which provides a globally consistent coding framework for aligning case definitions. For coding alignment in hospital/claims analyses, PID mapped to ICD-9 098.1, 098.17, 098.19, 098.2, 098.3, 098.36, 098.37, 098.39, 099.54, 099.56, 614.0–614.9, 615.0–615.1, 615.9 and ICD-10 A54.24, A56.1-A56.11, K67.0-K67.1, N70.00-N70.03, N70.1, N70.11-N70.13, N70.90-N70.93, N71.0-N71.1, N71.9, N73, N73.0-N73.9, N74, N74.2, N74.3, N74.4, N74.8. Chlamydial infection was defined as urogenital C. trachomatis infection confirmed by laboratory methods, primarily nucleic acid amplification tests (NAATs) or culture. Because many infections are asymptomatic yet play a critical role in PID pathogenesis through ascending infection and reinfection risk, both diagnosed and subclinical infections were considered within this etiologic framework. For coding alignment, chlamydial infection mapped to ICD-10 A55-A56.11, K67.0, N74.4 and ICD-9 099.41–099.5.

Temporal Trend Analysis

Age-specific and overall (15–49 years) prevalence rates (per 100,000 women aged 15–49) of PID attributable to urogenital C. trachomatis were taken directly from the Global Burden of Disease modeling pipeline (DisMod-MR 2.1); we did not re-estimate prevalence but summarized temporal change using the Estimated Annual Percent Change (EAPC).^19,20 For each location, we modeled the annual prevalence series P_t with calendar year t as a continuous predictor and fit a log-linear regression as follows:

Here, β is the yearly log change in prevalence; estimation used ordinary least squares, with a population-weighted variant (weights = female population aged 15–49 years) as a robustness check, and t could be centered at the series midpoint to improve numerical stability. Based on the fitted slope β, we defined the EAPC-the average multiplicative percent change per year-by transforming β as:

This transformation expresses the annual change on the percentage scale, facilitating interpretation and comparison across locations. To quantify statistical uncertainty, we derived a 95% confidence interval (CI) for EAPC by propagating the standard error of β through the same transformation:

A 95% confidence interval entirely above 0 indicated a significant increase, entirely below 0 a significant decrease; otherwise the trend was considered stable. When GBD uncertainty draws were available, we repeated the regression for each draw, computed an EAPC per draw, and summarized the distribution (median and 2.5th-97.5^th percentiles) as the point estimate and 95% uncertainty interval. Years with missing values were excluded listwise, and rare zeros were handled by adding a small offset (eg, 0.5 per 100,000) before log transformation; sensitivity checks confirmed negligible impact on estimates.

Projection Modelling

We projected the prevalence of PID attributable to urogenital C. trachomatis among women aged 15–49 years to the year 2050 using time series modeling.^21–23 For each age group (five-year bands from 15–19 to 45–49 years), we extracted the annual prevalence series (1990–2021) and fitted an autoregressive integrated moving average (ARIMA) model using the auto.arima function in the forecast package in R (version 4.3.2). This function selects the optimal model parameters (p,d,q) by minimizing the corrected Akaike information criterion (AICc), while automatically accounting for nonstationarity and autocorrelation. Forecasts were generated for 29 years beyond the last observation (2022–2050), producing point estimates and 95% prediction intervals based on model uncertainty. Predictions were made on the original prevalence scale (per 100,000 population). Historical (observed) and forecasted values were concatenated into a single time series for each age group. For tabular presentation, forecast results were extracted at 10-year intervals, while figures display the full annual trajectories with prediction intervals visualized as shaded bands.

Sociodemographic Index (SDI)

The SDI is a composite summary measure developed within the GBD analytical framework to capture variations in social and economic development across countries and over time. SDI combines three indicators-(1) lag-distributed income per capita, (2) average educational attainment among individuals aged 15 years and older, and (3) total fertility rate among women younger than 25 years. Each component is rescaled to a 0–1 range and equally weighted to produce a composite score, where higher values indicate greater social and economic development.^24,25 In this study, SDI values for each location and year were obtained from the GBD 2021 database. Countries were grouped into five quintiles—low, low-middle, middle, high-middle, and high—according to GBD-defined global thresholds. The association between SDI and the prevalence of PID attributable to C. trachomatis was assessed using Spearman rank correlation, with non-linear patterns examined visually. SDI was also used to stratify temporal trends, age-specific distributions, and future projections to enable cross-setting comparisons.

Statistical Analysis

We assessed the association between the SDI and chlamydia-attributable PID prevalence (per 100,000 women aged 15–49) using Spearman rank correlation across locations for each year, quantifying the strength and direction of the monotonic relationship between these variables. Statistical significance was set at p<0.05. All data cleaning, processing, and visualization were performed in R (version 4.3.2). Scatterplots with fitted trend lines were created using ggplot2, with ggrepel applied to improve label readability by avoiding overlap. Geographical visualizations were produced using the sf package in combination with the rnaturalearth dataset for basemap shapefiles.

Results

Global Burden and Temporal Pattern 1990 to 2021

Globally, the number of women aged 15–49 years diagnosed with PID attributable to chlamydial infection increased from 143,258 (95% UI: 106,144-194777) in 1990 to 232,246 (169,268-318931) in 2021, representing a 62.12% increase in absolute cases. Over the same period, the prevalence rate changed from 10.71 (95% UI 7.94–14.56) to 11.92 per 100,000 women (95% UI 8.69–16.37) (Table 1). The EAPC was 0.10 (95% CI: 0.03–0.18), indicating that despite a considerable rise in absolute case numbers, the overall global rate remained largely stable across the three decades, with population growth likely contributing to the increased case counts.

Table 1.

Trends and Prevalence of Pelvic Inflammatory Disease Attributable to Chlamydial Infection in Women of Reproductive Age, by Global and Regional Estimates, 1990–2021

Feature	Cases_1990	Rates_1990	Cases_2021	Rates_2021	Cases_change	EAPC_CI
Global	143,258 (106144 to 194,777)	10.71 (7.94 to 14.56)	232,246 (169268 to 318,931)	11.92 (8.69 to 16.37)	62.12 (52.38 to 71.75)	0.1 (0.03 to 0.18)
Age group
15–19 years	8067 (4390 to 13,499)	3.16 (1.72 to 5.28)	10,399 (5675 to 17,399)	3.42 (1.87 to 5.73)	28.92 (22.84 to 35.34)	0.06 (−0.04 to 0.16)
20–24 years	19,092 (12580 to 28,150)	7.82 (5.15 to 11.53)	25,441 (16420 to 37,919)	8.66 (5.59 to 12.91)	33.25 (26.13 to 39.9)	−0.06 (−0.2 to 0.08)
25–29 years	31,634 (18910 to 50,373)	14.37 (8.59 to 22.89)	44,834 (25174 to 74,605)	15.41 (8.65 to 25.64)	41.73 (31.26 to 50.7)	−0.1 (−0.22 to 0.01)
30–34 years	31,443 (19670 to 47,474)	16.54 (10.35 to 24.97)	51,097 (31371 to 79,551)	17.09 (10.49 to 26.61)	62.51 (52.27 to 71.22)	−0.06 (−0.09 to −0.02)
35–39 years	26,452 (16524 to 39,220)	15.25 (9.53 to 22.61)	46,478 (27708 to 70,255)	16.73 (9.97 to 25.29)	75.71 (64.38 to 85.46)	0.18 (0.09 to 0.28)
40–44 years	17,959 (10322 to 28,306)	12.81 (7.36 to 20.19)	34,931 (19147 to 55,978)	14.08 (7.72 to 22.56)	94.5 (81.45 to 105.68)	0.23 (0.11 to 0.34)
45–49 years	8611 (5628 to 12,873)	7.57 (4.95 to 11.31)	19,065 (12203 to 29,037)	8.09 (5.18 to 12.32)	121.4 (110.28 to 132.5)	0.15 (0.05 to 0.26)
SDI regions
High SDI	23,048 (16811 to 31,666)	10.17 (7.42 to 13.97)	25,234 (18829 to 33,981)	10.38 (7.74 to 13.97)	9.49 (2.98 to 18.36)	−0.09 (−0.14 to −0.05)
High-middle SDI	21,709 (15563 to 30,220)	7.82 (5.6 to 10.88)	27,637 (20104 to 38,056)	9.06 (6.59 to 12.47)	27.31 (18.83 to 36.82)	0.53 (0.48 to 0.57)
Middle SDI	34,063 (24787 to 47,088)	7.62 (5.54 to 10.53)	62,963 (46521 to 85,910)	10.18 (7.52 to 13.89)	84.84 (72.37 to 99.34)	0.76 (0.7 to 0.83)
Low-middle SDI	39,090 (29364 to 52,199)	14.32 (10.76 to 19.13)	70,112 (50509 to 97,655)	13.85 (9.98 to 19.29)	79.36 (61.63 to 97.62)	−0.41 (−0.53 to −0.3)
Low SDI	25,229 (19403 to 32,994)	22.59 (17.37 to 29.54)	46,159 (32686 to 65,791)	16.83 (11.92 to 23.98)	82.96 (61.31 to 104.53)	−1.45 (−1.64 to −1.26)
Geographical regions
Andean Latin America	1611 (1190 to 2249)	16.99 (12.55 to 23.71)	3283 (2456 to 4408)	18.81 (14.07 to 25.26)	103.73 (81.12 to 126.83)	0.3 (0.25 to 0.35)
Australasia	1521 (1119 to 2082)	28.34 (20.84 to 38.78)	2222 (1629 to 3060)	30.78 (22.57 to 42.39)	46.08 (31.53 to 61.42)	0.01 (−0.07 to 0.09)
Caribbean	969 (714 to 1321)	10.4 (7.66 to 14.17)	1247 (896 to 1712)	10.37 (7.45 to 14.23)	28.66 (17.39 to 40.47)	−0.06 (−0.27 to 0.15)
Central Asia	2250 (1616 to 3149)	13.41 (9.63 to 18.77)	3685 (2655 to 5133)	15.18 (10.94 to 21.16)	63.74 (49.79 to 79.17)	0.31 (0.17 to 0.45)
Central Europe	3158 (2277 to 4382)	10.28 (7.41 to 14.27)	2844 (2145 to 3792)	11.05 (8.33 to 14.73)	−9.92 (−16.4 to −2.42)	0.45 (0.36 to 0.55)
Central Latin America	3975 (2901 to 5456)	9.48 (6.92 to 13.02)	7444 (5462 to 10,102)	10.92 (8.01 to 14.81)	87.26 (72.01 to 103.75)	0.27 (0.18 to 0.35)
Central Sub-Saharan Africa	2309 (1720 to 3066)	18.68 (13.92 to 24.8)	5323 (3743 to 7519)	16.3 (11.46 to 23.03)	130.55 (92.92 to 170.33)	−0.79 (−0.91 to −0.67)
East Asia	17,832 (12592 to 24,703)	5.35 (3.78 to 7.41)	22,989 (16488 to 31,509)	6.95 (4.98 to 9.52)	28.92 (14.29 to 45.47)	0.79 (0.63 to 0.94)
Eastern Europe	6985 (4935 to 9764)	12.63 (8.93 to 17.66)	5747 (4022 to 8081)	11.91 (8.34 to 16.75)	−17.72 (−24.35 to −10.91)	−0.17 (−0.27 to −0.07)
Eastern Sub-Saharan Africa	7707 (5814 to 10,116)	17.86 (13.47 to 23.44)	14,825 (10481 to 21,178)	13.84 (9.79 to 19.77)	92.36 (66.36 to 116.9)	−1.27 (−1.45 to −1.08)
High-income Asia Pacific	8383 (6141 to 11,443)	18.33 (13.43 to 25.02)	7975 (5938 to 10,785)	20.97 (15.61 to 28.35)	−4.86 (−10.58 to 2.4)	0.52 (0.46 to 0.59)
High-income North America	6705 (4781 to 9570)	9.02 (6.43 to 12.87)	5515 (4138 to 7198)	6.56 (4.92 to 8.57)	−17.76 (−28.14 to −2.17)	−1.34 (−1.44 to −1.23)
North Africa and Middle East	5535 (3989 to 7683)	7.08 (5.11 to 9.83)	12,512 (8869 to 17,488)	7.85 (5.57 to 10.97)	126.06 (107.97 to 144.2)	0.58 (0.45 to 0.71)
Oceania	252 (188 to 327)	16.23 (12.11 to 21.07)	328 (224 to 466)	9.45 (6.46 to 13.42)	30.04 (5.59 to 53.41)	−1.06 (−1.4 to −0.72)
South Asia	40,159 (29909 to 53,811)	15.76 (11.73 to 21.11)	73,410 (52211 to 101,309)	14.86 (10.57 to 20.5)	82.8 (59.81 to 107.09)	−0.5 (−0.68 to −0.33)
Southeast Asia	4676 (3336 to 6501)	3.89 (2.78 to 5.41)	7446 (5415 to 10,137)	4.06 (2.96 to 5.53)	59.26 (48.48 to 70.86)	0.08 (0.03 to 0.14)
Southern Latin America	1762 (1267 to 2409)	14.22 (10.22 to 19.44)	2530 (1792 to 3473)	14.52 (10.28 to 19.92)	43.62 (28.84 to 58.72)	−0.04 (−0.09 to 0.01)
Southern Sub-Saharan Africa	2846 (2029 to 3872)	21.41 (15.27 to 29.13)	3808 (2662 to 5454)	17.54 (12.26 to 25.12)	33.82 (14.95 to 52.39)	−1.47 (−1.89 to −1.04)
Tropical Latin America	1797 (1333 to 2375)	4.51 (3.34 to 5.95)	9718 (7234 to 13,148)	16.03 (11.93 to 21.69)	440.75 (404.04 to 478.04)	4.65 (3.31 to 6.01)
Western Europe	5316 (3812 to 7406)	5.56 (3.99 to 7.75)	7996 (5938 to 10,696)	8.58 (6.37 to 11.48)	50.42 (39.37 to 63.45)	1.13 (0.8 to 1.46)
Western Sub-Saharan Africa	17,509 (12818 to 23,616)	40.15 (29.4 to 54.16)	31,397 (21260 to 46,017)	26.19 (17.73 to 38.38)	79.31 (53.6 to 101.94)	−2.08 (−2.35 to −1.8)

Note: Rates are per 100,000 population. The ranges in parentheses for cases and rates represent uncertainty intervals (UI), and those for EAPC represent 95% confidence intervals (CI).

Age Specific Prevalence Across Reproductive Ages

Across reproductive ages, both the number of PID cases attributable to C. trachomatis infection and the corresponding prevalence rates increased with age, peaking in the mid- to late-reproductive period (Figure 1A and B). In 1990, the prevalence rate rose from 3.16 (95% UI: 1.72–5.28) per 100,000 women in the 15–19-year group to 16.54 (10.35–24.97) per 100,000 in the 30–34-year group, before gradually declining to 7.57 (4.95–11.31) per 100,000 in women aged 45–49 years. By 2021, this age pattern persisted, with rates ranging from 3.42 (1.87–5.73) per 100,000 in the youngest group to a peak of 17.09 (10.49–26.61) per 100,000 in the 30–34-year group, followed by a decline to 8.09 (5.18–12.32) per 100,000 in the oldest group (Figure 1A,B and Table 1). The largest proportional increases in case numbers between 1990 and 2021 occurred in women aged 45–49 years (121.4% increase), 40–44 years (94.5%), and 35–39 years (75.71%), whereas younger groups such as those aged 15–19 years and 20–24 years experienced smaller relative increases of 28.92% and 33.25%, respectively. EAPC values were close to zero across all age groups, with modest positive trends observed in the 35–39 years (0.18; 95% CI: 0.09–0.28) and 40–44 years (0.23; 0.11–0.34) groups (Figure 1C and Table 1), suggesting relative stability in prevalence rates despite substantial growth in absolute case numbers.

Figure 1.

Global and regional trends in pelvic inflammatory disease (PID) attributable to Chlamydia trachomatis among women aged 15–49 years, by age group, 1990–2021. (A) Global number of prevalent cases. (B) Global prevalence rate per 100,000 women aged 15–49 years. (C) Estimated annual percentage change (EAPC) in prevalence rate. (D) Contribution of each age group to total cases globally, across five SDI regions, and within 21 GBD geographic regions in 1990 (left) and 2021 (right), based on the Global burden of disease (GBD) regional classification.

The age composition of the global PID burden showed a relatively stable pattern across the study period. Women aged 25–39 years consistently contributed the majority of total cases worldwide, with only minor proportional shifts between 1990 and 2021 (Figure 1D). This persistent dominance of middle reproductive ages in the overall burden underscores the need for targeted prevention and control strategies that focus on these high-impact age groups.

Prevalence by Sociodemographic Index Levels

Across the five SDI categories, substantial variation was observed in both the absolute number and prevalence rate of PID attributable to C. trachomatis infection among women aged 15–49 years (Figure 2A and B). In 1990, the highest prevalence rate was recorded in low SDI regions at 22.59 (95% UI: 17.37–29.54) per 100,000 women, followed by low-middle SDI (14.32, 95% UI: 10.76–19.13), while the lowest rates were seen in middle SDI (7.62, 95% UI: 5.54–10.53) and high-middle SDI (7.82, 95% UI: 5.60–10.88) regions. By 2021, low SDI regions remained the highest at 16.83 (95% UI: 11.92–23.98) per 100,000, despite a marked decline from 1990 levels (EAPC = −1.45, 95% CI: −1.64 to −1.26) (Figure 2C and Table 1). In contrast, middle SDI regions showed the largest proportional increase in prevalence rate, rising to 10.18 (95% UI: 7.52–13.89) per 100,000 in 2021 (EAPC = 0.76, 95% CI: 0.70–0.83), alongside an 84.84% (95% UI: 72.37–99.34) increase in case numbers (Figure 2A,B and Table 1). High-middle SDI regions also experienced a moderate rise in rates (EAPC = 0.53, 95% CI: 0.48–0.57), while high SDI regions remained relatively stable with minimal changes in both prevalence rate (10.38 per 100,000 in 2021) and case numbers (+9.49%). Low-middle SDI regions recorded a small decline in rates (EAPC = −0.41, 95% CI: −0.53 to −0.30) despite a substantial 79.36% increase in absolute case counts (Figure 2C and Table 1). Overall, these patterns highlight a convergence in prevalence rates across most SDI categories over time, driven by declines in low SDI regions and gradual increases in middle and high-middle SDI regions.

Figure 2.

Global, SDI-level, and regional prevalence patterns of pelvic inflammatory disease attributable to chlamydial infection among women aged 15–49 years, 1990–2021. (A) Global and SDI-level number of prevalent cases. (B) Global and SDI-level prevalence rates per 100,000 population. (C) Estimated annual percentage change (EAPC) in prevalence rate by SDI level. (D) Regional prevalence rates in 1990 and 2021. (E) Age-specific prevalence rates in 2021 across SDI levels and regions.

Regional Variation and Change Over Time

Marked geographic heterogeneity was observed in both the absolute burden and prevalence rate of PID attributable to C. trachomatis infection among women aged 15–49 years. In 2021, the highest prevalence rates were reported in Australasia (30.78 per 100,000; 95% UI: 22.57–42.39), Western Sub-Saharan Africa (26.19; 95% UI: 17.73–38.38), and High-income Asia Pacific (20.97; 95% UI: 15.61–28.35). In contrast, the lowest rates occurred in Southeast Asia (4.06; 95% UI: 2.96–5.53), Tropical Latin America (16.03; 95% UI: 11.93–21.69), and East Asia (6.95; 95% UI: 4.98–9.52), with Southeast Asia remaining consistently low over the study period (Figure 2D and Table 1).

From 1990 to 2021, several regions showed substantial shifts in prevalence rate. Tropical Latin America experienced the most pronounced relative increase, with rates rising from 4.51 to 16.03 per 100,000 (EAPC = 4.65; 95% CI: 3.31–6.01), accompanied by a more than fourfold growth in absolute case numbers (+440.75%). In contrast, Western Sub-Saharan Africa and Southern Sub-Saharan Africa exhibited notable declines in prevalence rate despite increases in case counts, largely reflecting rapid population growth. Western Sub-Saharan Africa declined from 40.15 to 26.19 per 100,000 (EAPC = −2.08; 95% CI: −2.35 to −1.80), while Southern Sub-Saharan Africa dropped from 21.41 to 17.54 per 100,000 (EAPC = −1.47; 95% CI: −1.89 to −1.04) (Table 1). The age-specific pattern in 2021 revealed that in most regions, prevalence peaked among women aged 30–34 or 35–39 years, with particularly high rates in Western Sub-Saharan Africa (53.3 per 100,000 in the 30–34-year group) and Central Sub-Saharan Africa (28.3 per 100,000) (Figure 2E). By contrast, East Asia and Southeast Asia consistently displayed lower rates across all reproductive age groups.

Country Level Distribution and Trends

In 2021, marked cross-national variation was observed in the prevalence of PID attributable to C. trachomatis, with New Zealand showing the highest rates (35.66 per 100,000 women), Australia (29.81 per 100,000), and Ghana (28.97 per 100,000; closely followed by Togo 28.83 and Cameroon 28.47). The lowest rates were seen in Iceland (1.87 per 100,000), Malta (2.34 per 100,000), and Portugal (2.49 per 100,000; Netherlands 3.20 and France 3.17 were similarly low) (Figure 3A and Supplementary Table S1). Many countries recorded larger case counts in 2021 than in 1990-for example, Pakistan (+201.06% cases; rate 15.12 per 100,000), Brazil (+452.15%; 16.36 per 100,000), and United Arab Emirates (+472.34%; 8.45 per 100,000)-reflecting population growth and rising rates in several settings (Figure 3B and Supplementary Table S1).

Figure 3.

Country-level prevalence, temporal change, and trends in pelvic inflammatory disease attributable to chlamydial infection among women aged 15–49 years, 1990–2021. (A) Prevalence rate per 100,000 population in 2021. (B) Percentage change in number of prevalent cases from 1990 to 2021. (C) Estimated annual percentage change (EAPC) in prevalence rate over 1990–2021. (D) Prevalence trajectories for the top three and bottom three countries ranked by EAPC.

Trend metrics highlighted divergent trajectories (Figure 3C and D). The steepest increases in prevalence rate were in Brazil (EAPC 4.75), Spain (4.36), and Greece (3.17), with additional rises in Portugal (3.00), Vietnam (1.50), and Japan (0.74). In contrast, sustained declines were evident across parts of Western and Central Africa and some high‑income countries, led by Mali (EAPC −3.94), Burkina Faso (−2.95), and Cameroon (−2.61), and also seen in the United States (−1.47) and South Africa (−1.95) (Figure 3C and D). These patterns underscore substantial national heterogeneity, with pockets of persistent high burden in Australasia and West Africa and rapid upward trajectories in several middle‑income countries.

Association Between SDI and Prevalence

Across development levels, the prevalence of PID attributable to C. trachomatis showed a structured yet non-linear relationship with SDI. In regional time series, rates per 100,000 women were highest at low SDI, reached a trough around mid-SDI, and rose modestly again in some high-SDI settings (ρ = 0.163, P < 0.001; Figure 4A). Illustratively, Western and Central sub-Saharan Africa sat at the high-rate, low-SDI end; East and Southeast Asia clustered in the mid-SDI trough; and Australasia and parts of High-income Asia Pacific showed a slight upturn at higher SDI. In the country cross-section for 2021, the overall association was moderately negative (ρ = −0.464, P < 0.001), with prevalence generally declining as SDI increased (Figure 4B). The broad scatter within SDI bands-including high-SDI outliers (eg, New Zealand, Australia) with comparatively elevated rates and several low-SDI countries with lower-than-expected rates-suggests that factors beyond macro-development (screening coverage, care access, sexual network patterns, and surveillance intensity) shape national burdens. Taken together, SDI captures a strong background gradient, but heterogeneity within bands underscores the need for context-specific prevention and control.

Figure 4.

Association between the Sociodemographic Index (SDI) and prevalence of pelvic inflammatory disease attributable to chlamydial infection among women aged 15–49 years. (A) Non-linear relationship between SDI and prevalence rates across 21 Global burden of disease regions from 1990 to 2021, with Spearman correlation coefficient (ρ) and p-value shown. (B) Cross-sectional association between SDI and prevalence rates in 2021 at the country level, with fitted regression line and 95% confidence interval shading; ρ and p-value from Spearman correlation.

Projections to 2050 by Age and SDI

By age and SDI, projections to 2050 preserve the current life-course pattern but with steeper rises at mid-to-late reproductive ages and clear divergence across development strata. Globally, rates (per 100,000 women) are projected to reach 3.50 (95% forecast interval [FI] 2.93–4.06) at 15–19 years, 8.80 (7.32–10.28) at 20–24, 14.98 (14.03–15.94) at 25–29, 23.57 (16.90–30.24) at 30–34, 34.49 (11.72–57.26) at 35–39, 22.58 (13.77–31.39) at 40–44, and 13.06 (8.06–18.05) at 45–49 by 2050, with the largest absolute gains in the 35–39 and 30–34 groups. Across SDI levels, low-SDI settings remain highest—rising from 16.08 in 2020 to 38.51 (9.72–67.29) by 2050—while high-SDI settings also increase to 30.11 (6.12–54.09). Low-middle SDI reaches 16.94 (10.42–23.46), middle SDI 13.35 (11.86–14.84), and high-middle SDI shows a modest rise to 10.06 (9.30–10.81). Forecast uncertainty widens at the SDI extremes, but the relative ordering (low and high SDI higher; high-middle SDI lower) persists (Figure 5A,B and Supplementary Table S2).

Figure 5.

Predicted prevalence of pelvic inflammatory disease attributable to chlamydial infection among women aged 15–49 years through 2050. (A) Age-specific predicted prevalence rates per 100,000 population from 1990 to 2050 (B) Predicted prevalence rates by Sociodemographic Index (SDI) quintile from 1990 to 2050. (C) Predicted prevalence rates in 2050 by SDI level and GBD region.

By geographical region, projections indicate pronounced heterogeneity. The highest 2050 rates are anticipated in Western Sub-Saharan Africa (69.4 per 100,000 women), Southern Sub-Saharan Africa (62.6 per 100,000), and Central Sub-Saharan Africa (35.8 per 100,000), with elevated values also in Australasia (29.1 per 100,000), High-income Asia Pacific (24.2 per 100,000), and Tropical Latin America (24.6 per 100,000). At the lower end, rates remain modest in Eastern Europe (3.7 per 100,000), Western Europe (6.1 per 100,000), Oceania (8.1 per 100,000), the Caribbean (10.1 per 100,000), and Central Europe (10.4 per 100,000) (Figure 5C and Supplementary Table S2). Overall, these forecasts suggest a continuing concentration of burden in parts of Sub-Saharan Africa, while most European subregions and Oceania are expected to maintain comparatively low levels.

Discussion

This analysis provides a comprehensive assessment of the global burden of PID attributable to C. trachomatis among women of reproductive age, covering trends from 1990 to 2021 and projections to 2050. Although the global age-standardized prevalence rate has changed little over the past three decades, the absolute number of affected women has risen markedly, driven by population growth and demographic shifts. The burden remains concentrated in women aged 30–39 years, who are projected to experience the largest absolute increases by mid-century. These patterns align with evidence from other sexually transmitted infection (STI) studies showing that infection-related sequelae often appear later in reproductive life, reflecting cumulative exposure, delayed diagnosis, and recurrent reinfection.^26–29 Unlike previous GBD 2019–based analyses that reported overall PID or STI trends without specifying etiologic attribution, the present study isolates the burden specifically linked to urogenital C. trachomatis.^30,31 This etiologic distinction provides a more precise understanding of the infection’s contribution to PID across settings and strengthens the interpretability of global and regional trend comparisons.

Marked heterogeneity persists across sociodemographic and geographic contexts. Low-SDI regions continue to experience the highest prevalence rates, reflecting a combination of limited access to timely diagnosis and treatment, under-resourced partner services, and structural barriers to care.³² The observed declines in some low-SDI settings are encouraging, yet their rates remain well above global averages. Middle-SDI regions have shown the most rapid relative increases in prevalence rate, consistent with reports from transitional economies where sexual network connectivity is expanding more rapidly than STI control infrastructure.^32,33 In high-middle SDI settings, rates remain the lowest, which may be due to an equilibrium between accessible healthcare, moderate screening coverage, and relatively contained transmission networks. Notably, some high-SDI regions, such as parts of Australasia and the High-income Asia Pacific, show unexpectedly high prevalence, which mirrors findings from recent programmatic evaluations suggesting that broad testing coverage can reveal persistent transmission pockets in subpopulations with lower health service engagement.³⁴

The association between the Sociodemographic Index and prevalence was non-linear, with a trough at mid-SDI levels and a slight rise again in some high-SDI contexts.³⁵ Similar patterns have been noted in multi-country STI meta-analyses, where high-SDI outliers often correspond to countries with robust surveillance systems that capture more cases, alongside concentrated epidemics in specific groups.³⁶ Within-SDI variability highlights the influence of factors beyond macroeconomic development, including the reach and quality of screening programmes, integration of partner notification, sexual behaviour patterns, and the strength of laboratory and surveillance capacity.³⁶ The causal association between C. trachomatis infection and PID has been firmly established through prospective and meta-analytic evidence.^37,38 However, PID is increasingly recognized as a polymicrobial condition, with contributions from Neisseria gonorrhoeae, Mycoplasma genitalium, and endogenous anaerobes that may vary across settings and over time.^38,39 Recent quantitative syntheses further clarify the proportion of PID cases attributable to C. trachomatis and other pathogens. Goller et al estimated that C. trachomatis and N. gonorrhoeae together accounted for approximately one quarter of PID cases among women attending Australian sexual health clinics.⁴⁰ Similarly, Price et al used multiple methodological approaches to show a consistent global estimate of around 20–30% of PID cases caused by C. trachomatis.⁴¹ In contrast, Mycoplasma genitalium appears to contribute a smaller but non-negligible share.⁴² These studies collectively reinforce the validity of focusing on C. trachomatis–related PID while recognizing its polymicrobial context. Although diagnostic approaches for C. trachomatis and PID have improved substantially during the past three decades, the consistent attribution framework applied here ensures comparability of estimates and facilitates evaluation of long-term global trends.

The persistence of a large burden in mid-reproductive ages suggests that current prevention strategies—often focused on adolescents and young adults—may not fully address the age groups most affected by PID attributable to C. trachomatis.⁴³ Opportunities exist to integrate chlamydia testing into reproductive health services used by women in their 30s, including antenatal care, infertility clinics, and contraceptive consultations.^44,45 Evidence from implementation studies indicates that opportunistic testing in these settings can increase case detection without the need for entirely new service platforms.^43,46 Strengthening partner services remains essential, with approaches such as expedited partner therapy, digital notification systems, and repeat testing after treatment shown to reduce reinfection and subsequent PID risk.

Over the three-decade study period, diagnostic technology, clinical awareness, and healthcare access for C. trachomatis infection have advanced markedly worldwide.^47,48 The transition from culture- and serology-based tests to molecular assays, together with the expansion of sexual and reproductive health services, has improved detection of both infections and related PID.^49,50 These developments likely contributed to part of the temporal variation observed in PID burden, especially in high-resource regions.⁵¹ Nevertheless, the long-term estimates analyzed here reflect standardized and comparable population-level data, providing a robust basis for evaluating global and regional trends over time.

This study benefits from the use of harmonised global data with standardized diagnostic definitions, explicit attribution to C. trachomatis, and consistent methods for quantifying trends and projecting future burden. However, several limitations should be acknowledged. As a secondary population-level analysis based on GBD estimates, the findings reflect ecological associations and cannot be used to infer individual-level risk or causal relationships. The analysis also relies on modelled estimates that may be sensitive to underlying assumptions, particularly in regions lacking diagnostic testing, where data are generated through covariate-based modeling and regional extrapolation. Differences in screening coverage, diagnostic capacity, coding practices, and healthcare-seeking behaviour across countries may introduce measurement and ascertainment bias, while undiagnosed asymptomatic infections may lead to underestimation of the true burden.^18,52,53 The analysis was restricted to women aged 15–49 years, which limits generalizability to girls younger than 15 years, women older than 49 years, and specific clinical populations. Although the GBD framework has been extensively validated across multiple health domains—including maternal mortality, frailty, and cancer—no independent external validation has yet been conducted specifically for pelvic inflammatory disease attributable to C. trachomatis. Therefore, the present estimates should be interpreted with appropriate caution. The forecasts are based on historical patterns and do not account for potential future changes in screening practices, treatment strategies, public health interventions, or sexual behaviour; as such, they should be regarded as scenario-based projections rather than precise predictions.

Conclusion

Overall, the analysis shows that C. trachomatis-attributable PID remains a substantial global reproductive health concern. The stable prevalence rate at the global level masks important regional and age-specific patterns that can guide intervention prioritisation. Tailoring screening and partner services to the most affected age groups, strengthening prevention in high-burden regions, and leveraging new diagnostic and preventive technologies will be critical to reducing the future burden and preventing the long-term sequelae of this preventable condition.

Abbreviations

ARIMA, Autoregressive Integrated Moving Average; CI, Confidence Interval; EAPC, Estimated Annual Percent Change; FI, Forecast Interval; GBD, Global Burden of Disease; PID, Pelvic Inflammatory Disease; SDI, Sociodemographic Index; STI, Sexually Transmitted Infection; UI, Uncertainty Interval.

Data Sharing Statement

The data are available from the Global Burden of Disease Results Tool of the Global Health Data Exchange (http://ghdx.healthdata.org/).

Ethics Approval and Consent to Participate

This study was approved by the Ethics Committee of Yangzhou Women and Children’s Hospital in line with the Declaration of Helsinki (L202603022). As a result of the secondary data analysis in this study, the review board waived informed consent. Additionally, this article does not contain any personal information about patients. All authors are in agreement with the manuscript.

Disclosure

The authors declare no conflicts of interest in this work.