Chronic kidney disease in India: Incidence, prevalence, mortality, DALYs, and State-level disparities, 1990–2023

Study design and data source

We synthesised this report abiding with the GATHER guidelines.¹³ This was a population-based observational study using the GBD database.¹⁴ We analysed the burden of CKD across all 28 Indian States and 3 Union Territories, with minor Union Territories aggregated into a single analytical unit (‘other Union Territories’), in accordance with GBD reporting conventions. Data were obtained from the Institute for health metrics and evaluation (IHME) Global health data exchange (GHDx) and GBD results tool.¹⁵ The data retrieval focused on four primary metrics: age-standardised incidence, prevalence, mortality rates, and DALY, reported per 100,000 populations from 1990 till 2023. SDI, and population at State level was also retrieved for 2023. The data downloaded from GBD is available publicly as open source: https://doi.org/10.5281/zenodo.17432524 .

Case definition, GBD data source, and GBD estimation framework

CKD was defined according to GBD Cause Level 3 classification and ICD-10 codes N18-N19. This definition was applied consistently across the entire study period (1990–2023) through retrospective harmonisation within the GBD analytical framework. GBD estimates are derived from a synthesis of multiple data sources. This included vital registration systems, sample registration system (SRS), verbal autopsy studies, population-based surveys, hospital discharge data, and disease registries. Data availability varies by State and time period, and the gaps are addressed using Bayesian meta-regression, DisMod-MR 2.1 to generate comparable estimates across geography and time. Cause-specific mortality is estimated using the cause of death ensemble model (CODEm),¹⁶ and DALYs are calculated as the sum of years of life lost due to premature mortality and years lived with disability, based on GBD disability weights.¹⁷

Outcome measures

• (i)

  Current burden estimation: We reported age-standardised rates (ASRs) of incidence, prevalence, mortality, and DALYs for each State from last available year in GBD data base, i.e. 2023.

• (ii)

  Temporal trend analysis: Long-term temporal changes were assessed using estimated annual percentage change (EAPC). A log-linear regression was fitted for each State: ln ASR_t=α+ βt=EAPC= e^{{ β}}–1×100 where t is calendar year. The 95% uncertainty interval (UI) was derived from the standard error of β. States were classified into four epidemiological clusters based on whether the ASR in 2023 was above or below the national mean (high-burden vs. low-burden) and whether the EAPC was positive or negative (increasing vs. decreasing trend).

• (iii)

  Trajectory modelling: For each State and metric, temporal trajectories from 1990 to 2023 were visualised using locally estimated scatterplot smoothing (LOESS) with 95% UIs. State-specific panels were arranged alphabetically to permit comparative interpretation. National trajectories were overlaid as a benchmark in each panel.

• (iv)

  Socio-demographic index (SDI) and cross-sectional analysis: The SDI, a composite measure of lag-distributed income, average educational attainment, and fertility rate in women under 25 years, scaled from 0 to 1, was used as a proxy for development. SDI values for 2023 were extracted for all Indian States. The relationship between SDI and each CKD metric was assessed using LOESS regression with 95% confidence shading.

• (v)

  Inequality assessment: Inter-State inequality was quantified using the population-weighted Gini coefficient and quintile ratio (mean ASR in the top 20% vs. bottom 20% of States ranked by SDI). Socioeconomic inequality was further evaluated using Lorenz concentration curves and concentration indices. States were ranked by SDI from lowest to highest, and cumulative proportions of population were plotted against cumulative proportions of outcome counts. The concentration index was calculated as: CI=2×AUC−1 where AUC is the area under the concentration curve. A negative CI indicates disproportionate burden among low-SDI States (pro-poor inequality). Bootstrap resampling (1000 iterations) was used to generate 95% UIs for the concentration curve.

• (vi)

  Uncertainty reporting: All GBD estimates include 95% UIs generated from posterior simulations. Uncertainty was propagated through all analyses, including trajectory plots, EAPC estimates, and concentration indices. No separate sensitivity analyses were conducted because GBD metrics inherently incorporate uncertainty through probabilistic modelling, which has been extensively validated.

Statistical analysis

All analyses were performed using Python version 3.11 (pandas, numpy, statsmodels, and matplotlib libraries). LOESS smoothing and bootstrap procedures were conducted using locally weighted regression functions.

Current burden of CKD

The national age-standardised incidence rate ranged from 226.4 per 100,000 (95% UI: 210.6-492.0) in Rajasthan to 316.4 (295.8-684.6) in Haryana, with most States clustering between 230 and 270 per 100,000 ( Table I). Prevalence exceeded 10,000 per 100,000 population in every State, ranging from 10,452 (UI: 9,802-22,467) in Uttar Pradesh to 12,539 (11,678-26,730) in Haryana, indicating a high chronic disease pool across the nation. Mortality rates ranged three-fold, from 6.71 deaths per 100,000 (3.80-21.72) in Rajasthan to 19.84 (13.92-54.84) in Telangana, demonstrating significant geographic disparities in fatal outcomes. DALY rates, reflecting combined morbidity and mortality, ranged from 312.2 (226.3-858.7) in Rajasthan to 666.4 (515.3-1704.4) in Telangana, with States such as Punjab, Other Union Territories, and Chhattisgarh also exhibiting consistently elevated DALY burdens.

Temporal trends

From 1990 to 2023, age-standardised CKD rates declined nationally, although the magnitude and direction of change varied by State and metric ( Table II, Fig. 1). The incidence rate declined at a national EAPC of -0.45% (95% UI -0.50 to -0.40). Most States showed decreasing trends, while some demonstrated stagnation or slight increases. Prevalence declined more modestly (EAPC -0.22%), reflecting extended patient survival and rising absolute case counts driven by population ageing. Mortality exhibited a steeper decline, with a national EAPC of -0.90%, consistent with improvements in early detection and treatment access. DALYs declined at the fastest rate, (EAPC -1.10%), indicating gains in both longevity and quality of life. States were categorised into four epidemiological clusters based on burden (relative to national ASR) and trend (positive or negative EAPC): (i) High-burden, decreasing trend (HB-D) e.g., Telangana, Punjab, Goa; (ii) High-burden, increasing trend (HB-I) e.g., Uttar Pradesh, Odisha; (iii) Low-burden, decreasing trend (LB-D) e.g., Kerala, Himachal Pradesh; (iv) Low-burden, increasing trend (LB-I).

Close

Fig. 1. State-wise CKD data for India in 2023. EAPC % calculated by % change from 1990-2023. Black circles represent Top 5 and blue circles shows Bottom 5 States. For Incidence, the highest was in Bihar and the lowest in Jharkhand. For Prevalence, the highest was in Haryana and Tamil Nadu while the lowest in Jharkhand (-0.36%). For Mortality, and DALY, the highest was in Uttar Pradesh and lowest in Tamil Nadu.

Export to PPT

Socio demographic index association

In cross-sectional analysis of 2023 State-level data ( Fig. 2, Supplementary Table), a non-linear association was observed in all. ASR incidence increased with rising SDI, with a clear upward inflection beyond an SDI of 0.65, rising from approximately 240 per 100,000 in low-SDI States to >280 per 100,000 in the highest SDI States. In contrast, prevalence demonstrated an inverted U-shaped trajectory, increasing from low-SDI to mid-SDI States (peaking around 11,500-12,000 per 100,000) before stabilising or marginally declining in high-SDI States, indicating greater disease detection and improved survival rather than uncontrolled progression. ASR mortality showed a weak negative association with SDI, declining from ∼14-16 deaths per 100,000 in low-SDI States to <10 per 100,000 in higher-SDI States, though variability remained high across middle-SDI regions. This pattern was mirrored in DALY rates, which decreased from >600 DALYs per 100,000 in low-SDI States to <400 per 100,000 in the most developed States. Notably, LOESS smoothing curves demonstrated widening uncertainty bands at mid-range SDI values, indicating heterogeneous health system performance and unequal translation of socioeconomic progress into improved kidney health outcomes.

Close

Fig. 2. State-wise CKD and SDI data for India in 2023, where black circles represent the top 5 States, blue circles the bottom 5 States, and grey circles represent other States. For incidence and prevalence, Haryana ranked highest, while the lowest burden was observed in States including Madhya Pradesh, Rajasthan, and Karnataka. For mortality and DALY, Telangana ranked highest, whereas the lowest burden was mainly seen in Rajasthan, Madhya Pradesh, Arunachal Pradesh, and Maharashtra.

Export to PPT

State-wise trajectory analysis

Thirty-one-panel trajectory figures (Supplementary Figs. 1-4) demonstrated distinct temporal signatures. High-SDI States such as Kerala and Tamil Nadu exhibited steady declines in incidence and mortality from the early 2000s onward, with narrow uncertainty intervals. Conversely, low-SDI States such as Uttar Pradesh and Jharkhand showed persistently high rates with minimal reduction over time. DALY trajectories showed the most dramatic divergence. States including Telangana, Punjab, and Jammu and Kashmir had sustained high burden across the entire study period despite declining trends indicating persistent unmet clinical need.

Geographic inequality

ll four metrics exhibited statistically significant inter-State inequality (Supplementary Fig. 5). Population-weighted Gini coefficients were 0.21 for incidence, 0.27 for prevalence, 0.38 for mortality, and 0.41 for DALYs, indicating greatest disparity in fatal and disabling outcomes. Concentration curves demonstrated: (i) Incidence: slightly pro-poor concentration (CI=-0.06), indicating higher incidence among lower-SDI States. (ii) Prevalence: near-equality (CI=-0.013; bootstrap 95% UI including zero), indicating minimal socioeconomic disparity. (iii) Mortality: significantly pro-poor (CI=-0.12), indicating excess deaths in disadvantaged States. (iv) DALYs: strongest inequality (CI=-0.24), reflecting compounded mortality and disability among low-SDI populations. Notably, bootstrap uncertainty bands did not cross the line of equality for mortality or DALYs, confirming robust pro-poor inequality.

Analysis of CKD aetiology distribution across 31 Indian States showed substantial metric-dependent heterogeneity (Supplementary Figs. 6-9). CKD-Other/unspecified causes dominated incidence (66.4%) and prevalence (67.9%) but contributed proportionally less to mortality (38.4%). Hypertension showed inverse patterns: low prevalence (6.4%) but high mortality (21.3%), indicating elevated case-fatality. Diabetes mellitus type 2 maintained consistent contributions across metrics (18-24%). Type 1 diabetes, though rare in incidence (0.27%) and prevalence (0.37%), accounted for 5.8% of deaths. Geographic variation was reported, with Telangana showing highest diabetes-related CKD mortality and West Bengal demonstrating elevated CKD-glomerulonephritis burden.