Trends in pediatric rheumatic heart disease in the Indonesia–Malaysia–Thailand Growth Triangle economic sub-region (Indonesia, Malaysia, and Thailand), 1990–2023: A Global burden of disease study Authors: Muhammad Iqhrammullah Published: 2026 ABSTRACT ABSTRACTBackground:The extent to which socio-demographic development translates into reductions in rheumatic heart disease (RHD) burden in the Indonesia–Malaysia–Thailand Growth Triangle (IMT-GT) Economic Sub-Region have not been assessed. This study aimed to assess temporal trends, socio-demographic gradients, and development-adjusted performance of pediatric RHD from 1990 to 2023.Methods:Data from the Global Burden of Disease Study 2023 for Indonesia, Malaysia, and Thailand were analyzed. Age-specific incidence, mortality, and disability-adjusted life years (DALYs) were estimated by the sex. Temporal trends were assessed using generalized additive models to estimate average annual percentage changes (AAPC). The association between DALY rates and the socio-demographic index (SDI) was evaluated using elasticity analysis and performance relative to SDI-based expectations was quantified as the gap between observed and expected DALY rates.Results:Between 1990 and 2023, the burden of RHD declined across the IMT-GT sub-region, with substantial cross-country variation. AAPCs indicated modest declines in DALYs and mortality in all countries, whereas incidence showed no consistent long-term decrease. Higher SDI was significantly associated with lower DALY rates across all pediatric age groups (P < 0.001). SDI-benchmarked comparisons indicate that Indonesia has not achieved the level of DALY reduction expected for its development status, whereas Thailand has exceeded these expectations and Malaysia has remained on the predicted trajectory.Conclusions:The burden of pediatric RHD has declined in the IMT-GT Economic Sub-region over the past three decades, driven primarily by reductions in mortality. Sociodemographic development was strongly associated with lower DALY rates, but substantial cross-country differences in development-adjusted performance persist. INTRODUCTION Rheumatic heart disease (RHD) is a preventable sequelae of acute rheumatic fever that continues to cause substantial disability and premature mortality among children and adolescents in low- and middle-income countries despite the availability of effective prevention strategies.[12] In Southeast Asia, the burden remains considerable, particularly in settings undergoing rapid socio-economic transition.[3] The Indonesia–Malaysia–Thailand Growth Triangle (IMT-GT) Economic Sub-region provides a unique context in which geographically contiguous countries have experienced marked improvements in income, education, and health system coverage over the past three decades, yet differ in demographic profiles, primary healthcare organization, and access to preventive services.[4] These contrasts offer an opportunity to examine how socio-demographic progress translates into reductions in RHD burden and to identify the gaps between expected and observed health outcomes within a shared regional development framework. The Global Burden of Disease (GBD) Study provides comparable estimates of RHD incidence, mortality, and disability-adjusted life years (DALYs) across time and settings, allowing the assessment of both temporal trends and the extent to which sociodemographic development translates into health gains.[5] The sociodemographic index (SDI), a composite of income, education, and fertility, enables the benchmarking of observed burden against expected levels and distinguishes the structural development from health system performance.[6] Nonetheless, the burden of this disease in pediatric populations is rarely examined,[78] despite childhood and early adolescence being critical periods for disease onset and prevention.[9] Assessing incidence, mortality, and DALYs among this population, therefore, provides a more complete understanding of the mechanisms driving changes in burden. This study aimed to quantify long-term trends and evaluated SDI-benchmarked performance of pediatric RHD in the IMT-GT Economic Sub-region from 1990 to 2023. METHODS Study design and data source This retrospective ecological study used estimates from the GBD 2023, produced by the Institute for Health Metrics and Evaluation, to assess the burden of pediatric RHD in the IMT-GT Economic Sub-Region from 1990 to 2023. The present study did not directly collect primary RHD data. Instead, it analyzed the final GBD 2023 estimates for RHD incidence, deaths, and DALYs by country, year, sex, and pediatric age group. Data were extracted using the Global Health Data Exchange (GHDx) query tool. According to the GBD 2023 Data Coverage for Rheumatic heart disease, the available source metadata for RHD in the study settings included vital registration and vital registration–sample verbal autopsy data.[1011] Measures and stratification The study analyzed incidence, mortality, and DALYs for RHD, following the strategy of previous reports.[1213] Estimates were reported as rates per 100,000 population with 95% uncertainty intervals (UIs). Analyses were stratified by country (Indonesia, Malaysia, and Thailand), sex (male, female, and both), and pediatric age groups (<5 years, 5–9 years, and 10–14 years). Percentage change between 1990 and 2023 was calculated to quantify long-term shifts in burden. Global Burden of Disease 2023 modeling framework GBD 2023 used separate approaches to estimate deaths and nonfatal outcomes. Cause-specific deaths from RHD were estimated using the Cause of Death Ensemble Model (CODEm), which compares several statistical models and covariates before selecting the model or combination of models with the best predictive performance.[1014] Non- fatal outcomes, including RHD incidence and prevalence, were estimated using DisMod-MR 2.1, a Bayesian meta-regression model that links incidence, prevalence, remission, and mortality. The present study used the final GBD 2023 estimates and did not directly collect or reclassify primary RHD data. Before producing final estimates, GBD processed these data by addressing incomplete age and sex information, redistributing nonspecific or implausible cause-of-death codes, assessing source completeness, and excluding sources with insufficient completeness or excessive nonspecific coding. Uncertainty was propagated through draw-level simulations, and 95% uncertainty intervals were defined using the 2.5th and 97.5th percentiles.[13] Sociodemographic index Sociodemographic development was assessed using the SDI, a composite indicator of lag-distributed income per capita, mean educational attainment among individuals aged ≥15 years, and total fertility rate under age 25 years. SDI estimates were obtained from the GBD covariate database for the corresponding locations and years. Statistical analysis All analyses were conducted using R software (version 2025.09.1 Build 401). GAMs were fitted using the mgcv package, and data processing and visualization were performed using tidyverse packages. Generalized additive models Nonlinear temporal trends in incidence, mortality, and DALYs were assessed using generalized additive models (GAMs) with a smoothing spline for year and fitted by restricted maximum likelihood. Model fit was evaluated using the effective degrees of freedom (edf), F statistic, adjusted R2, and deviance explained. Inflection points were identified from the changes in the sign of the first derivative of the fitted spline. Average annual percentage change (AAPC) was derived from the first derivative of the GAM-fitted values on the logarithmic scale. The mean derivative was multiplied by 100 to obtain AAPC, and the 2.5th and 97.5th percentiles were used to calculate the corresponding 95% interval. The AAPC calculation could be expressed as (Equation 1): where Yt denotes the estimated rate at year t. The 95% interval was obtained from the 2.5th and 97.5th percentiles of the annual percent change values. Socio-demographic index elasticity and development-adjusted performance The association between DALY rates and SDI was quantified using a log–log linear regression model, and the regression coefficient was interpreted as SDI elasticity. Expected DALY rates for a given SDI level were estimated using a GAM and the performance gap was calculated as the difference between observed and expected DALY rates. Positive values indicate a higher-than-expected burden relative to development level, whereas negative values indicate a lower-than-expected burden. Expected DALY rates for a given SDI level were estimated using a generalized additive model. The performance gap was calculated as where positive values indicate a higher-than-expected burden relative to the level of development and negative values indicate a lower-than-expected burden. Ethics statement This study used publicly available, anonymized, and aggregated data from the GBD Study 2023. As no individual-level identifiable information was accessed, the analysis was exempt from ethical review and the requirement for informed consent in accordance with institutional and international research guidelines. RESULTS Burden of rheumatic heart disease Between 1990 and 2023, substantial reductions in DALY and mortality rates were observed across all pediatric age groups, whereas incidence changed only modestly [Figure 1]. In children aged 10–14 years, the DALY rate declined by 53.2% in Indonesia (from 168.88 to 79.02 per 100,000), with deaths decreasing by 56.0% (2.05–0.90 per 100,000), while incidence showed a minimal reduction (−3.0%). A steeper decline was observed in those aged 5–9 years, where DALYs fell by 75.3% and deaths by 78.8%. Across sexes, females generally had slightly higher DALY and mortality rates than males, but the temporal trajectories were parallel. The age gradient persisted throughout the study period, with the highest burden in children aged 10–14 years, followed by those aged 5–9 years and <5 years. A summary of sex-specific DALY, mortality, and incidence rates in 2023 and their percentage change from 1990 for children aged 10–14 years in the IMT-GT Economic Sub-Region is presented in Table 1. Temporal trends in age-specific DALY, incidence, and mortality rates of rheumatic heart disease in the IMT-GT Economic Sub-Region (Indonesia, Malaysia, and Thailand), 1990-2023. Panels represent DALY rates (a-c), incidence rates (d-f), and death rates (g-i). From left to right, columns represent Indonesia, Malaysia, and Thailand. Lines represent point estimates and shaded areas indicate 95% uncertainty intervals. Estimates are presented for three pediatric age groups (<5 years, 5–9 years, and 10–14 years) Table 1Burden of rheumatic heart disease in 2023 among children aged 10–14 years in the Indonesia-Malaysia-Thailand growth triangle sub-regionCountry and sexDALYDeathsIncidence Val (95% CI)Change (%)*Val (95% CI)Change (%)*Val (95% CI)Change (%)*Indonesia Male72.26 (37.31 to 124.01)−43.80.81 (0.36 to 1.48)−46.930.17 (15.49 to 52.13)−4.7 Female86.18 (37.97 to 159.34)−59.21.00 (0.40 to 1.93)−61.627.62 (14.19 to 49.51)−1.1 Both79.02 (48.92 to 119.65)−53.20.90 (0.52 to 1.43)−5628.93 (15.01 to 50.49)−3Malaysia Male18.57 (10.38 to 31.08)−79.40.17 (0.08 to 0.34)−84.112.08 (5.80 to 21.30)−31.1 Female24.26 (12.18 to 40.02)−75.30.24 (0.10 to 0.46)−79.211.13 (5.31 to 19.83)−29.5 Both21.27 (14.06 to 30.65)−77.40.20 (0.12 to 0.33)−81.711.63 (5.84 to 20.34)−30.3Thailand Male28.72 (17.44 to 46.98)−42.20.27 (0.13 to 0.49)−49.824.37 (11.75 to 43.79)−0.7 Female24.90 (14.63 to 36.56)−64.90.23 (0.11 to 0.38)−72.220.87 (10.33 to 38.25)−1.8 Both26.86 (18.38 to 38.18)−79.40.25 (0.15 to 0.37)−63.122.66 (11.14 to 40.93)−1.2*Change (%) indicates the relative difference between 1990 and 2023. Values are presented as rate per 100,000 population (95% uncertainty interval). CI: Confidence interval, DALY: Disability-adjusted life year Burden of rheumatic heart disease in 2023 among children aged 10–14 years in the Indonesia-Malaysia-Thailand growth triangle sub-region *Change (%) indicates the relative difference between 1990 and 2023. Values are presented as rate per 100,000 population (95% uncertainty interval). CI: Confidence interval, DALY: Disability-adjusted life year The reduction in RHD burden occurred unevenly across three periods, as presented in Figure 2. Indonesia showed relatively slow early progress in 1990–2000 and 2001–2010, followed by a stronger decline in 2011–2023. Meanwhile, Malaysia had the strongest earlier reductions, while Thailand showed weaker improvement in the most recent period. For DALY rates, Indonesia declined by 14.2%, 13.0%, and 35.0% across the three periods; Malaysia declined by 38.2%, 43.2%, and 31.8%; and Thailand declined by 31.0%, 28.6%, and only 7.2%, respectively. Similar patterns were observed for death rates, where the most recent period showed substantial reductions in Malaysia and Indonesia, with declines of 39.2% and 37.8%, respectively, but only a modest reduction in Thailand by 9.6%. In contrast, incidence rates showed relatively stable but small negative changes in Indonesia and Thailand, generally around 1%–2% across periods. Malaysia showed a more fluctuating incidence pattern, shifting from a marked change of 30.1% in 2001–2010 to 10.9% in 2011–2023. Period-specific changes in rheumatic heart disease burden (DALYs, deaths, and incidence) across Indonesia, Malaysia, and Thailand for the periods 1990–2000, 2001–2010, and 2011-2023. Red points indicate the rate at the start year of each period, while green points indicate the rate at the end year. Horizontal lines represent the magnitude and direction of change between the start and end years Nonlinear temporal trends Generalized additive models showed highly significant nonlinear temporal trends for DALY, mortality, and incidence rates in all three countries (all P < 0.001), with excellent model fit (adjusted R2 = 0.981–0.998) [Table 2]. Declines in DALY and mortality rates were observed in Indonesia, Malaysia, and Thailand, although the magnitude and timing of changes differed. Malaysia demonstrated the steepest average annual reductions in DALYs (AAPC −4.83%, 95% UI −11.93 to 3.71) and deaths (−5.55%, −14.59 to 1.65), with major inflection points occurring after 2011. Indonesia and Thailand showed more modest declines in DALYs (−2.34% and −2.47% per year, respectively) and mortality (−2.53% and −3.08% per year), with earlier turning points in the late 1990s and early 2000s. In contrast, incidence rates exhibited minimal long-term change in all countries, with AAPC estimates close to zero and multiple inflection points. The higher edf values for incidence (8.73–8.92) compared with DALYs and mortality reflect more complex and irregular temporal patterns. The timing of major inflection points was largely consistent between DALY and mortality trends within each country. Table 2Nonlinear temporal trends in disability-adjusted life year, mortality, and incidence rates of rheumatic heart disease among children aged 10–14 years in the Indonesia-Malaysia-Thailand growth triangle economic sub-regionBurden estimatesedf F Adjusted - R2Year of inflectionsAAPC (95% UI)Indonesia DALYs8.0981258.510.9971998, 2001−2.34 (−6.65 to 0.38) Deaths8.0911236.440.9971998, 2001−2.53 (−7.48 to 0.41) Incidence8.832186.680.9811995, 2005, 2009, 2014, 2022−0.09 (−0.60 to 0.28)Malaysia DALYs8.2521561.340.9982012, 2013, 2021−4.83 (−11.93 to 3.71) Deaths8.2641538.660.9982011, 2014, 2022−5.55 (−14.59 to 1.65) Incidence8.731107.270.9971992, 1997, 2001, 2018−1.20 (−12.26 to 15.74)Thailand DALYs8.562460.290.9921991, 1999, 2002, 2020−2.47 (−8.28 to 2.64) Deaths8.558450.40.9921991, 1999, 2002, 2020−3.08 (−9.83 to 3.52) Incidence8.918285.540.9871996, 1999, 2004, 2013, 2021−0.04 (−0.73 to 0.46)All models were statistically significant (P<0.001). DALYs: Disability-adjusted life years, AAPC: Average annual percentage changes, UI: Uncertainty intervals, edf: Effective degrees of freedom Nonlinear temporal trends in disability-adjusted life year, mortality, and incidence rates of rheumatic heart disease among children aged 10–14 years in the Indonesia-Malaysia-Thailand growth triangle economic sub-region All models were statistically significant (P<0.001). DALYs: Disability-adjusted life years, AAPC: Average annual percentage changes, UI: Uncertainty intervals, edf: Effective degrees of freedom Associations with Sociodemographic development DALY rates declined nonlinearly with increasing sociodemographic development across all pediatric age groups [Figure 3]. The inverse association between SDI and DALY rates was strongest in children aged <5 years (elasticity − 6.37), followed by those aged 5–9 years (−4.79) and 10–14 years (−3.92), and was statistically significant in all models (P < 0.001) [Table 3]. These findings indicate that a 1% increase in SDI was associated with a 3.9%–6.4% reduction in DALY rates, with the greatest sensitivity to development observed in the youngest age group. The effective degrees of freedom decreased with age (edf 2.81 in < 5 years to 1.71 in 10–14 years), suggesting that the SDI–burden relationship became more linear in older children. Across the SDI range, Malaysia consistently occupied the highest SDI levels and the lowest DALY rates, whereas Indonesia showed persistently higher DALY rates at comparable SDI levels. Association between sociodemographic development and DALY rates of rheumatic heart disease in the IMT-GT Economic Sub-Region, 1990–2023. Observed age-specific DALY rates plotted against the Socio-demographic Index (SDI) for children aged <5 years (a), 5–9 years (b), and 10–14 years (c). The corresponding SDI-based expected DALY rates for children aged <5 years (d), 5–9 years (e), and 10–14 years (f) estimated using generalized additive models (black lines). The performance gap, defined as the difference between observed and expected DALY rates for children aged <5 years (g), 5–9 years (h), and 10–14 years (i). Positive values indicate a higher-than-expected burden relative to the level of development, whereas negative values indicate a lower-than-expected burden Table 3Elasticity of disability-adjusted life years rates in relation to sociodemographic development and mean performance gap from expected sociodemographic index based levels for rheumatic heart disease in the Indonesia-Malaysia-Thailand growth triangle economic sub-regionAge group (years)edfElasticityElasticity PGap from expected SDI-based DALY IndonesiaMalaysiaThailand<52.81−6.37<0.0018.43−0.96−7.475–92.79−4.79<0.00117.53−0.18−17.3510–141.71−3.92<0.00136.932.16−39.09All models were statistically significant (P<0.001). SDI: Sociodemographic index, DALY: Disability-adjusted life year, edf: Effective degrees of freedom Elasticity of disability-adjusted life years rates in relation to sociodemographic development and mean performance gap from expected sociodemographic index based levels for rheumatic heart disease in the Indonesia-Malaysia-Thailand growth triangle economic sub-region All models were statistically significant (P<0.001). SDI: Sociodemographic index, DALY: Disability-adjusted life year, edf: Effective degrees of freedom Performance gap Substantial differences were observed between countries in their performance relative to SDI-based expectations [Figure 3 and Table 3]. Indonesia consistently exhibited a higher-than-expected burden across all age groups, with the gap widening markedly with age (mean gap: 8.43 in < 5 years, 17.53 in 5–9 years, and 36.93 per 100,000 in 10–14 years). In contrast, Thailand showed a lower-than-expected burden throughout the study period, with increasingly negative gaps in older children (−7.47, −17.35, and − 39.09 per 100,000, respectively). Malaysia remained close to the expected SDI-predicted levels, with only minimal deviations (−0.96, −0.18, and 2.16 per 100,000). Based on the temporal trajectories, the aforementioned performance patterns were observed to be established early and persisted over time. DISCUSSION Based on the comprehensive assessment of long-term trends and development-adjusted performance of pediatric RHD in the IMT-GT, three integrated findings emerge. First, the overall burden declined in all countries, but through different epidemiological pathways. Second, these reductions were not accompanied by comparable declines in incidence. The incidence remained relatively stable in Indonesia and Thailand and fluctuated in Malaysia, suggesting that improvements in survival and disease management have outpaced progress in preventing new RHD occurrence. Third, although higher socio-demographic development was strongly associated with lower DALY rates, countries differed substantially in their ability to translate development into effective disease control. Indonesia consistently showed higher-than-expected DALY rates, Thailand showed lower-than-expected rates, and Malaysia remained close to SDI-predicted levels. This divergence suggests that advances in clinical care and survival have outpaced progress in preventing new cases, in line with previous reports.[1516] In the present study, the absence of a significant AAPC for the incidence in Malaysia indicates that the decline occurred through the periods of fluctuation. In Indonesia and Thailand, incidence remained essentially stable over time despite clear reductions in mortality. These findings highlight the persistence of upstream determinants of acute rheumatic fever, including overcrowding, delayed treatment of streptococcal pharyngitis, and unequal access to primary healthcare. As suggested previously, developing countries continue to face major gaps in primary prevention due to socioeconomic and health system constraints, while secondary prevention has progressed more rapidly through expanded screening and implementation of long-term penicillin prophylaxis.[17] As a result, RHD is increasingly transitioning from a fatal disease to a chronic condition, with more children surviving into later ages with established disease.[1819] The strong negative association between the SDI and DALY rates across all pediatric age groups, herein, confirms the central role of social and economic development. Indonesia consistently exhibited higher than expected DALY rates for its level of development, with the gap widening with age, indicating missed opportunities to convert sociodemographic gains into effective prevention and long-term disease control. In contrast, Thailand achieved lower-than-expected DALY rates across all age groups, suggesting more efficient delivery of preventive and clinical services. Malaysia’s observed burden closely matched SDI-based expectations, reflecting performance consistent with its development level. Thailand’s better-than-expected performance may reflect long-term investments in primary healthcare infrastructure and universal health coverage. Thailand’s primary healthcare system evolved from selective child survival interventions in the 1980s into a more comprehensive system supporting universal health coverage in the 2000s.[20] These health system features may have improved early detection, continuity of care, referral pathways, and access to secondary prophylaxis for children with suspected or established RHD.[20] For example, a school-based study in urban Thailand found the significant decline RHD prevalence between 1986 and 2006, alongside marked improvement in socioeconomic indicators.[21] Further, the present study found that the association between SDI and DALY rates in children aged <5 years is the strongest, indicating the greatest impact of development on early-life risk and access to primary prevention. The attenuation of this relationship in older age groups reflects the cumulative nature of RHD, where burden increasingly depends on historical exposure, delayed diagnosis, and long-term survival. The progressively widening performance gap with age, particularly in Indonesia, further suggests that RHD control failures are cumulative over time. This pattern aligns with recent evidence showing that mortality declines are mainly driven by better treatment of clinical RHD, while incidence reduction depends on structural socioeconomic improvements.[22] These findings have important implications for regional health policy within the IMT-GT framework. Although IMT-GT cooperation has not traditionally prioritized health,[2324] these findings highlight the need to incorporate RHD control into the agenda and to ensure that economic gains are effectively translated into measurable reductions in disease burden. Achieving sustained reductions in incidence in Indonesia will require a pragmatic expansion of primary prevention – particularly low-threshold antibiotic treatment for streptococcal infections – supported by reliable delivery systems for long-term prophylaxis and broader improvements in access to care.[1822] In addition, this opens opportunities for regional pharmaceutical collaboration to stabilize essential medicine supply, supporting long-term secondary prophylaxis delivery.[24] At the same time, the experience of Thailand demonstrates that it is possible to achieve outcomes better than expected for a given level of development through effective health-system organization.[2526] This study is limited by the fact that the GBD study-derived estimates depend on the availability and quality of underlying data and modeling assumptions, particularly in areas with limited surveillance. Due to the nature of the data source in GBD, subclinical and undiagnosed RHD cases may be underestimated compared with estimates derived from direct community or school-based echocardiographic screening. Moreover, the ecological design does not allow causal inference between socio-demographic development and changes in disease burden. In addition, SDI is a composite indicator and does not capture specific health system characteristics such as coverage of primary care, adherence to secondary prophylaxis, or access to cardiac surgery. Finally, subnational variations were not examined and may mask important within-country inequalities. CONCLUSION Pediatric RHD in the IMT-GT sub-region has declined over the past three decades, mainly through reductions in mortality and DALYs. The pattern of improvement differed across countries, with slower early progress but stronger recent decline in Indonesia, stronger earlier reductions in Malaysia, and weaker recent improvement in Thailand. Development-adjusted analyses revealed substantial differences in how socioeconomic gains were translated into health outcomes, with persistent underperformance in Indonesia, better-than-expected outcomes in Thailand, and Malaysia remaining close to SDI-predicted levels. Strengthening primary prevention, early diagnosis, and long-term prophylaxis are essential to achieve further and more equitable reductions in pediatric RHD burden. Conflicts of interest There are no conflicts of interest.