Factors Affecting CKDu Progression in a CKDu-Endemic Farming Community in Sri Lanka: A Longitudinal Study

Chronic kidney disease of unknown etiology (CKDu) remains a major public health concern in Sri Lanka. We initiated the present study in 2025 to identify the factors that affect disease progression among CKDu patients in an endemic setting.

We enrolled 80 participants from Wilgamuwa (8 women, 72 men), all full-time farmers. For each participant we obtained all serum creatinine results from diagnosis to the present and derived the annualized change in estimated glomerular filtration rate (eGFR) for each interval. We conducted interviews to capture time-varying exposures for each year of disease, including treatment status, work intensity, pesticide application, drinking-water sources and proportions consumed (reverse osmosis [RO], river, well; >50% RO, <50% well), hypertension, and three categories of mentally stressing events: 1) bereavement or family relational conflict, 2) financial burdens such as loans, land disputes, or poverty-related stress, and 3) disease-related stressors. Categorical predictors were dummy coded. Annual eGFR decline (mL/min/1.73 m² per year) was modeled using generalized estimating equations (GEE) with Gaussian errors and an exchangeable working correlation by participant. We report regression coefficients (β) as adjusted mean differences with 95% confidence intervals (CI) and two-sided P values from robust standard errors.

To examine a population-level exposure, we performed a retrospective pre–post analysis anchored to late 2017 when RO water provision scaled in CKDu-endemic areas. We compared the distributions of annual eGFR decline before 2018 versus 2018 and later using a two-sided Mann–Whitney U test and report the U statistic, P value, the Hodges–Lehmann location shift, and the rank-biserial correlation.

In the GEE model, RO use was associated with a more favorable trajectory (β 5.09 mL/min/1.73 m² per year, 95% CI 3.10 to 7.08, P<0.0001). Stopping work was likewise associated with slower decline (β 3.47, 95% CI 1.02 to 5.93, P=0.0056). Mentally stressing event type 2 (financial stress) corresponded to faster decline (β −6.64, 95% CI −11.74 to −1.55, P=0.0106). Pre–post comparisons supported an association between RO water consumption and a reduced rate of kidney function decline. Median annual eGFR decline was −3.45 mL/min/1.73 m² per year (IQR −7.00 to 2.00) before 2018 and −1.00 (IQR −5.00 to 4.00) from 2018 onward (figure 1).The distributions differed (U=22440.5, P=0.0036). The Hodges–Lehmann estimator indicated a +2.00 mL/min/1.73 m² per year shift post intervention, with a rank-biserial correlation of −0.18, consistent with a small shift toward less negative decline.

Among farmers with CKDu in Wilgamuwa, RO water use and reduced occupational load were linked to slower decline of kidney function, while financial stresses were linked to faster decline. The population-level rollout of RO water beginning in late 2017 corresponded to a small but statistically significant improvement in eGFR trajectories at the group level. These findings highlight the potential value of safe drinking-water access and social protection strategies alongside clinical management in CKDu-endemic communities. The cohort remains active, with ongoing recruitment of new participants to enable validation of these associations in a larger sample.