CKDu: kidney failure in young workers who have none of the usual causes

Start here: what CKDu is

Almost everywhere in the world, when a doctor finds that someone's kidneys are failing, the cause is one of two things: diabetes or high blood pressure. Those two conditions damage the kidney's filters slowly, over decades, and they account for most of the global burden of chronic kidney disease ^{[1] [2]}.

Along the Pacific coast of Central America, and in farming districts of Sri Lanka and India, a different pattern turns up. Men in their twenties and thirties — cane cutters, brickmakers, construction and salt-pan workers — arrive at clinics with kidneys that are already half destroyed, and they have neither diabetes nor hypertension. Their urine contains little or no protein, which is the signature of the diabetic and hypertensive kidney. Their kidneys on ultrasound are small and shrunken. Under the microscope the damage is not in the filters but in the plumbing downstream: the tubules and the tissue between them.

That entity has collected an unhelpful pile of names — chronic kidney disease of unknown or undetermined etiology (CKDu), chronic kidney disease of non-traditional cause (CKDnt), chronic interstitial nephritis in agricultural communities (CINAC), and, for the Central American epidemic specifically, Mesoamerican nephropathy (MeN) ^{[3] [4] [5]}. The names differ because the arguments about cause differ, and the inconsistent terminology is itself now recognised as an obstacle to research ^[6]. This review uses CKDu for the general entity and Mesoamerican nephropathy where the source is specifically about Central America.

Who gets it, and where. The epidemiology is unusually specific, and that specificity is the main clue. The disease concentrates in hot lowlands, in men, in occupations involving heavy manual labour in heat — and, crucially, not in people living in the same communities who do not do that work. Population surveys found high male-to-female ratios in high-risk lowlands but small sex differences within occupational categories, and low prevalence among non-workers; it appears in sugarcane and other high-intensity agriculture, and in non-agricultural heavy manual work in hot environments, but not among subsistence farmers ^[7]. In the Mesoamerican Nephropathy Occupational Study, which recruited 569 workers with no known diabetes, hypertension or kidney disease from sugarcane, corn, plantain, brickmaking and road construction, baseline CKD prevalence was already 7.4% — age-standardised at 14.1% in Salvadoran sugarcane, 11.6% in Salvadoran corn and 8.1% in Nicaraguan brickmaking ^[8].

The South Asian epidemics look related but not identical. In Sri Lanka's North Central Province, age-standardised CKDu prevalence was 12.9% in males and 16.9% in females, with the severe stages far more frequent in men ^[9]. In Uddanam, Andhra Pradesh, 18.2% of a 2,210-person community sample had CKD, and in 73% of those cases diabetes, long-standing hypertension and significant proteinuria were all absent ^[10]. A 2025 meta-analysis across CKDnt-endemic tropical coastal regions put the pooled odds ratio for CKD in agricultural versus non-agricultural populations at 2.12 ^[11].

Why it matters. These are young, working, otherwise healthy men, and the disease kills them. In El Salvador, end-stage renal disease has been the leading cause of hospital deaths in adults and the second cause of death in men ^[12]. In Costa Rica's Guanacaste province, age-adjusted CKD mortality among men rose from 4.4 to 38.5 per 100,000 between 1970 and 2012, against 3.6 to 8.4 in the rest of the country — a spatiotemporal pattern that roughly followed sugarcane expansion into hot, dry lowlands with manual harvesting ^[13]. The Mesoamerican death toll runs into the tens of thousands ^[7]; one 2016 review put it at more than 20,000 ^[14]. In Uddanam, verbal-autopsy work has since found CKD to be the leading cause of death outright ^[15].

Why it is in the news. Because of what the leading explanation implies. The best-supported hypothesis is that the disease is caused by the job in the heat — repeated days of strenuous labour at high wet-bulb temperature, with too little water and too little rest, each day inflicting a small kidney injury that does not fully heal before the next one. If that is right, then a warming climate is directly expanding the population at risk, and CKDu becomes a strong candidate for the first chronic disease whose epidemiology is being visibly rewritten by climate change. The literature says this in as many words: the entity "may be one of the first major diseases attributed to climate change and the greenhouse effect" ^[14], and continued global warming "will increase the number of people at risk for dangerous heat exposure and kidney disease" ^[16].

Its counterpart. This review is the "unknown cause" half of a pair. The chronic kidney disease review covers the traditional-cause disease: the diabetic and hypertensive kidney, the KDIGO staging grid, and the drug classes that slow decline. Almost everything in that review is defined by having a cause you can name and a medicine that targets it. CKDu is the mirror image — a disease defined by the absence of an identifiable cause, and one for which, as a consequence, there is no targeted drug at all. What it has instead is a preventable exposure.

Three pillars follow — measurements, management, and progress — with a cumulative-injury model between the first two that explains why a hazard too small to notice in one shift becomes kidney failure in a decade.

Pillar 1: measurements and diagnosis

A diagnosis made by exclusion

There is no test for CKDu. There is a test for reduced kidney function — the estimated glomerular filtration rate, eGFR, computed from serum creatinine — and then there is the work of ruling out everything else. A case is CKDu when eGFR is persistently low, protein in the urine is absent or minimal, and diabetes, hypertension, glomerular disease and other known causes have been excluded ^{[4] [17]}.

That is a fragile definition, and the field knows it. A 2025 systematic review of 60 studies found that only 30% used a combination of pre-specified eGFR and proteinuria cut-offs; 19 studies classified all CKD without an identifiable cause as CKDu regardless of proteinuria; and the criteria used to exclude diabetes- and hypertension-related kidney disease were stated in fewer than half, with wide variability. The authors' conclusion is blunt: such heterogeneity causes misclassification and erroneous estimation of disease burden, and makes comparisons between studies difficult ^[18]. A European consensus statement has since proposed a formal definition and reporting standard for chronic kidney disease of unexplained cause, noting that it is a diagnosis of exclusion with an estimated global prevalence of at least 16–20% among CKD patients ^[19].

Contrast this with the traditional-cause disease. Diabetic kidney disease announces itself with albuminuria and a known antecedent; hypertensive nephrosclerosis comes with years of recorded blood pressure. CKDu announces itself with nothing. Its patients feel well until they do not, and by the time creatinine is abnormal a large fraction of nephron mass is gone.

The finding that broke the case open: kidney function falls across a work shift

The single most important measurement in this field is also the simplest. Draw blood from a cane cutter before the shift and after it, and compute eGFR both times.

In 189 sugarcane cutters aged 18–49 in three regions of El Salvador, mean workday temperature was 34–36 °C before noon and 39–42 °C at midday. Across the shift, urine specific gravity, urine osmolality and urinary creatinine rose, urinary pH fell, and serum creatinine, uric acid and urea nitrogen all increased — the signature of a kidney working under a fluid deficit. Pre-shift serum uric acid was remarkably high, and pre-shift eGFR was already below 60 mL/min in 14% of the male workers ^[20]. In 105 healthy Guatemalan sugarcane workers, the mean cross-shift decline in eGFR was 21.8% in a single day, with increasing wet-bulb globe temperature, high uric acid, decreased urine pH, urinary leukocyte esterase and serum hyperosmolality all identified as risk factors ^[21].

A 21.8% drop in filtration in one working day is not a subtle finding. It is subclinical — the worker goes home, drinks, sleeps, and creatinine largely recovers — which is precisely why it went unnoticed for so long, and precisely why it matters.

And across a harvest

Extend the same logic from a shift to a season. Cross-harvest studies draw blood before the harvest and at the end of it, five or six months later, and ask whether the recovery was complete.

It is not. Over a five-month El Salvador harvest, eGFR fell by 5.3 mL/min/1.73 m² in workers without an intervention and 3.4 in workers receiving one ^[22]. In a Nicaraguan cohort stratified by job, cross-harvest incident kidney injury — defined as a serum creatinine rise of at least 0.30 mg/dL or 1.5 times baseline — occurred in 2% of the low-to-moderate workload group and 27% of the very-high-workload burned-cane cutters ^[23]. In a three-year prospective study of 569 young male workers in Nicaragua and El Salvador, mean annual eGFR decline was −4.3 mL/min/1.73 m² per year, with a subgroup of very young men (mean age 23) losing −12.5 per year ^[24].

Cross-harvest creatinine fluctuations have been shown to forecast subsequent kidney-function decline ^[25], and point-of-care biomarkers have been assessed for predicting the same trajectory ^[26].

What is actually injured: tubules, not filters

Serum creatinine is a crude instrument, and it rises with dehydration and muscular work as well as with kidney injury — a real confounder in this population. So investigators have gone looking for markers specific to the tubule.

Among sugarcane workers with a cross-harvest creatinine rise of at least 0.3 mg/dL compared with workers whose creatinine stayed stable, urinary monocyte chemoattractant protein-1 showed a strongly differential cross-harvest trend, alongside a panel including kidney injury molecule-1, calbindin, glutathione S-transferase-π, clusterin and interleukin-18 ^[27]. Cross-shift changes in acute-kidney-injury biomarkers have been measured directly in cane farmers and cutters ^[28], and exposome and metabolome analyses have been used to look for predictors of injury in the same population ^{[29] [30]}.

The biopsy tells the same story from the tissue side. In 19 male Nicaraguan sugarcane workers with suspected Mesoamerican nephropathy and a mean eGFR of 57, biopsies showed glomerulosclerosis averaging 38%, glomerular hypertrophy and signs of chronic glomerular ischaemia, with mild-to-moderate tubulointerstitial damage; 47% had low plasma sodium and 21% low plasma potassium, and on follow-up eGFR continued to fall by a mean of 4.4 mL/min/1.73 m² per year even though most patients had stopped working in sugarcane ^[31]. An earlier series had established the clinical and pathological characterisation ^[32], and acute tubulointerstitial nephritis has been proposed as the early lesion in the genesis of the disease ^[33].

A systematic review of 13 biopsy studies across the Sri Lankan and Mesoamerican epidemics found interstitial fibrosis to be the predominant lesion in every one, with no characteristic histopathological signature other than a proximal-tubular lysosomal inclusion body claimed to indicate a toxic aetiology; the three pathogenetic mechanisms proposed were repeated acute insults leading to scarring, low-grade chronic insults producing non-inflammatory fibrosis, and combined tubulointerstitial and glomerular injury ^[34]. A direct Sri Lanka–Mesoamerica biopsy comparison found many similarities in the biochemical and morphological profile, supporting a common aetiology, but with a more mixed morphology, more interstitial inflammation and more vascular change in the Sri Lankan patients ^{[35] [36]}.

The pre-shift finding, and the healthy-worker trap

Two measurement problems deserve naming, because both make the epidemic look smaller than it is.

The first is that in a 2-year prospective cohort of apparently healthy men aged 18–30 in nine rural Nicaraguan communities, growth mixture modelling found three subpopulations: 81% remained stable, starting at 116 mL/min/1.73 m² and losing 0.6 per year; 9.5% experienced rapid decline despite normal baseline function, starting at 112 and losing 18.2 per year; and 9.5% already had baseline dysfunction at 58 ^[37]. The rapid decliners were indistinguishable from everyone else at baseline. Outdoor and agricultural work, and lack of shade during work breaks, were the factors associated with rapid decline.

The second is the healthy-worker selection effect. In the Nicaraguan workload cohort, 32% of workers who dropped out mid-harvest reported kidney injury as the reason for leaving ^[23]. A study that measures only the workers still present at end-of-harvest systematically misses the ones the harvest has already broken.

Centerpiece: a simple simulatable model of cumulative harvest injury

The measurements above describe an injury too small to notice on any given day and a disease that takes about a decade to arrive. The model that connects them is deliberately the simplest thing that can carry the argument.

Let a worker's filtration rate fall by a fixed background amount every year — the slope a healthy young adult shows anyway — plus a fixed additional amount for every harvest season worked in heat:

eGFR(n) = eGFR₀ − n · (s_background + s_harvest)

where n counts harvest seasons, one per year. Four of the five numbers this needs are directly quotable. The starting point eGFR₀ = 116 mL/min/1.73 m² and the background slope s_background = 0.6 per year are the stable young-male subpopulation from the Nicaraguan cohort ^[37]. The per-harvest term s_harvest is 5.3 mL/min/1.73 m² without an intervention and 3.4 with one, from the El Salvador water-rest-shade trial ^[22]. The rapid-decline arm, 18.2 per year, is again from the Nicaraguan cohort ^[37]. The threshold drawn across the figure, eGFR = 60, is the cut-point the field uses in the same populations — the one already crossed by 14% of working cane cutters before their shift began ^[20].

What the model explains. Three things, and each is a policy argument.

First, why nobody noticed for thirty years. The per-harvest loss is 5.3 mL/min/1.73 m² — an amount that would be invisible in any single worker, indistinguishable from measurement noise, and fully consistent with feeling perfectly well. It only becomes a disease by repetition. This is the same structural argument the occupational lung diseases make: in silicosis and black lung the injuring quantity is cumulative dust dose, the product of concentration and time, and the disease appears decades after the exposures that caused it. CKDu is the kidney's version, with heat and workload in place of dust.

Second, why the disease is the job and not ageing. The background trajectory — a young man in the same community who does not do the work — never approaches the threshold in a forty-year working life. The entire difference between a healthy retirement and dialysis at 35 is the harvest term.

Third, why prevention is worth so much. Moving s_harvest from 5.3 to 3.4 is not a marginal gain; it moves the threshold crossing from 9.5 harvests to 14, which for a man who starts cutting at 18 is the difference between kidney failure at 28 and kidney failure at 32 — and the real-world programmes described in Pillar 3 do considerably better than the trial's 3.4.

What the model deliberately does not do. It does not claim to predict any individual. It is linear where the real disease is not: kidney function trajectories in these cohorts are noisy, show partial recovery between seasons, and change slope as disease advances. It also treats a five-month harvest loss as an annual quantity, which assumes no net recovery in the off-season — an assumption the work-recovery-cycle literature is actively testing ^[39]. The claim being made is about the ordering of the scenarios and the size of the gaps, which is exactly the part that is grounded.

Pillar 2: management, and why the real lever sits upstream

The honest position on treatment

There is no cure for CKDu, and no drug directed at its cause, because its cause is contested. Kidney tissue that has been replaced by fibrosis does not come back. What exists is the general management of chronic kidney disease — the measures set out in the CKD review — applied to a population that is younger, poorer, and further from care than the patients those measures were developed for.

That last point is not a footnote. The burden of Mesoamerican nephropathy has been described specifically in terms of diagnosis and treatment inside a complex web of socioeconomics, context and stigma ^[40]. Dialysis and transplantation are the endpoint for a disease that concentrates in agricultural workers in low- and middle-income settings, and access to kidney replacement therapy is exactly what those settings lack ^{[41] [42] [43]}. A man who loses his kidneys at 35 also loses the job that was keeping his family fed, which is one reason workers hide symptoms and stay in the field.

The follow-up data make the limits plain. In the Nicaraguan biopsy cohort, eGFR continued to fall by 4.4 mL/min/1.73 m² per year after most patients had stopped working in sugarcane ^[31]. Removing the exposure late does not undo the damage.

Rehydration is not automatically protective — and can be harmful

One management point deserves emphasis because the intuitive answer is wrong. If the mechanism is recurrent dehydration, then drinking more should help, and it does — but what you drink matters.

In rats subjected to recurrent heat-induced dehydration, rehydration with an 11% fructose-glucose solution matching the composition of a typical soft drink produced greater dehydration than plain water, despite a larger total fluid intake: higher plasma and urinary osmolarity, higher copeptin, and worse renal injury, with activation of aldose reductase and fructokinase. Water sweetened with non-caloric stevia had the opposite effect ^[44]. The authors' framing is that this "emphasizes the danger of drinking soft drink-like beverages as an attempt to rehydrate following dehydration."

Sugary drinks are cheap, available and marketed at exactly this workforce. The literature has argued the point sharply enough that one paper is titled around whether the problem is dehydration or rehydration ^[45].

The lever that actually works: water, rest and shade

Because there is no drug, the intervention literature is an occupational-hygiene literature, and its central object is the water-rest-shade (WRS) programme adapted from the US Occupational Safety and Health Administration's campaign of the same name.

Phase 1 in El Salvador introduced WRS midway through a six-month harvest for a 60-person cutting group: water in individual backpacks, mobile shaded rest areas, scheduled rest periods, plus ergonomically improved machetes. Wet-bulb globe temperature exceeded 26 °C from 09:00 onwards, averaging a maximum of 29.3 °C around 13:00. Self-reported water consumption rose 25%, heat-stress and dehydration symptoms fell — and daily production rose from 5.1 to a high of 7.3 tons per person per day, a larger increase than in other cutting groups at the same company ^[46].

The kidney outcomes followed. In the paired trial, cross-shift eGFR decrease was smaller in the intervention group after the programme began, and the over-harvest decline appeared to halt after its introduction: −3.4 mL/min/1.73 m² in the group receiving it against −5.3 in the group that did not ^[22]. Those two numbers are the model's harvest term.

The productivity finding has since been replicated at scale, which matters because "it will cost output" is the standing objection. Across five harvest seasons at a large Nicaraguan mill, covering 749 seed cutters and 535 burned-cane cutters, productivity fell about 3% per °C of wet-bulb globe temperature — and rose roughly 19% in seed cutters and 9% in burned-cane cutters as the rest-shade-hydration programme scaled up, despite the programme reducing the total time allocated for work each day ^[47]. Rest is not lost production; heat is.

Pillar 3: progress — the argument, the interventions, and the climate

The etiology debate, stated fairly

Two camps. Both are represented in the literature by explicit position pieces, and the honest summary is that one has accumulated much more support without formally closing the case.

Heat stress and workload. The case is built from the epidemiological specificity, the dose-response, and the intervention response. Cross-shift and cross-harvest declines track heat and workload; the disease appears in heavy manual work in hot environments whether or not agriculture is involved, and not in subsistence farmers in the same regions; studies of pesticides and infectious risk factors have been largely negative; and non-occupational risk factors do not explain the observed patterns ^[7]. A 2024 review lays out the Bradford Hill-style argument directly: associations between heat exposure and kidney damage are "strong, consistent, and specific, occur after acute and chronic exposure, display dose-effect and dose-response relationships, are plausible, and coherent," whereas support for alternative causes is weak ^[16].

The proposed mechanism is specific and testable. Recurrent dehydration raises plasma osmolarity, which activates the polyol pathway inside the proximal tubule, generating endogenous fructose that is then metabolised by fructokinase — a pathway that itself causes injury. Wild-type mice given delayed rehydration after recurrent heat exposure developed renal injury with elevated creatinine, increased urinary NGAL, proximal tubular injury, inflammation and fibrosis, together with increased renal cortical sorbitol and fructose; fructokinase-knockout mice given the identical protocol were protected, and access to water during the dehydration period also protected ^[48]. A parallel strand implicates uric acid: strenuous work concentrates urate in the urine, and repeated uricosuria with urate crystal formation has been proposed as a chronic tumour-lysis-like injury ^{[49] [14]}, with the urate transporters that set serum and urinary urate levels well characterised ^{[50] [51] [52]}. Rising serum sodium and osmolarity have been shown to be independent risk factors for developing CKD in a five-year cohort ^[53].

Agrochemicals and metals. The competing position is that a nephrotoxin is the primary cause and heat is a co-factor. Its strongest evidence is pathological: almost all CINAC patients show a proximal tubular lesion with enlarged dysmorphic lysosomes containing electron-dense aggregates, features also seen in known toxin-induced nephropathies such as analgesic and aristolochic nephropathy ^[54]. Its strongest epidemiological argument is a list of things heat does not explain: that Sri Lankan incidence rose alongside mechanisation of paddy farming in the 1990s, that CINAC is absent from hotter northern Sri Lanka, Cuba and Myanmar where agrochemicals are sparsely used, that the endemic areas form a mosaic rather than a temperature gradient, and that the disease appears in women, children and adolescents who do not do the harsh work ^[4].

The Sri Lankan data carry that argument. Urinary cadmium was significantly higher in CKDu cases than in matched controls, with a significant dose-effect relationship between urine cadmium concentration and CKDu stage, and food from the endemic area contained cadmium and lead above reference levels ^[9]. Phosphate fertiliser has been identified as a main source of arsenic in affected areas ^[55], hard water and nephrotoxic metals implicated in combination with glyphosate ^{[56] [57]}, and drinking well water and occupational herbicide exposure associated with disease in a case-control design ^{[58] [59] [60] [61]}. Heavy-metal associations have since been reported from central India as well ^[62].

Against that, the systematic epidemiological review of pesticides and CKDu is genuinely equivocal: of 21 analytical studies, 13 reported at least one positive association, but the exposure measurement was usually unspecific and unquantified, most designs were cross-sectional, and the four studies with the strongest designs each implicated a different pesticide — with three of them conducted in areas without CKDu epidemics. The conclusion is that existing studies provide scarce evidence for an association with the regional epidemics, but that a role for nephrotoxic agrochemicals cannot be conclusively discarded, and that nobody has yet studied the interaction between pesticides and heat stress ^[63].

Where that leaves it. The 2025 tropical-coastal meta-analysis is the most useful arbiter currently available, because it puts the competing factors on one scale. Pooled odds ratios: drinking well water 2.75, malaria 2.64, low water intake 2.06, water source 1.50, agrochemicals 1.50, heat exposure 1.46, alcohol 1.27. Latitude and temperature were significant moderators, with a 1 °C increase in the epidemic region associated with an 8% increase in CKD risk. The authors' conclusion is that CKDnt is multifactorial — driven by heat exposure, infectious disease, physically demanding work without adequate hydration, water contamination and agrochemical exposure together ^[11]. Notably that malaria odds ratio is not a stray finding; leptospirosis and other infections have been proposed as contributors in their own right ^{[64] [65] [66] [67]}, and malaria has its own well-documented renal complications. Genetic susceptibility is also under study ^{[68] [69]}.

The interventions, and the best result in the field

The strongest evidence that heat stress is causal is that removing it works.

The Nicaraguan programme is the one to know. Following findings from the 2017–18 harvest, recommendations enhancing the rest schedule and improving access to hydration and shade were issued before 2018–19, and the actual work conditions were then observed rather than assumed. Cross-harvest incident kidney injury was 70% lower in the second harvest than the first — with no such improvement among the seed-cutter groups where implementation was less successful ^[70]. The distinction between designing an intervention and implementing it is the paper's explicit point.

Continuing that programme produced the most striking numbers in this literature. Across 1,044 workers and 1,938 person-harvests over four seasons of progressively enhanced rest-shade-hydration-hygiene, among burned-cane cutters — the job group with the highest workload and the worst outcomes at the start — incident kidney injury fell from 21% in harvest 1 to 1% in harvest 4; clinically diagnosed acute kidney injury fell from 20 to 8 per 1,000 worker-months; end-harvest leukocyturia fell from 26% to 1%; and the cross-harvest rise in C-reactive protein went from a median 1.75-fold increase to none at all ^[38]. That is the right panel of the figure above. The authors read it as supporting a causal relationship between occupational heat stress, kidney injury and CKDnt — which is the correct reading, since an intervention that removes only heat exposure should not fix a disease caused by something else.

Clinical acute kidney injury at the same plantation tracked the weather directly: each 1 °C of wet-bulb globe temperature was associated with an 18% higher rate of clinically diagnosed AKI on the same day ^[71].

The field is not uncritical of itself. A 2026 scoping review of interventions to prevent CKDu in working populations found only eight qualifying studies — four graded weak, three moderate, one strong. Most addressed only one of eight intervention-quality domains, adoption; few addressed acceptability, appropriateness, cost, feasibility, fidelity or penetration; and none addressed sustainability ^[72]. Promising evidence, thin evidence base. A systems-based approach to prevention in Latin American workers has been proposed in response ^[73].

Heat standards, and the measurement that would make them enforceable

Occupational heat limits already exist. NIOSH and ACGIH publish recommended exposure limits specifying the maximum combination of environmental heat, measured as wet-bulb globe temperature, and metabolic heat from workload, with lower limits for unacclimatised workers. Reviewing 25 outdoor occupational heat-related illnesses investigated by OSHA between 2011 and 2016, CDC found that heat stress exceeded those exposure limits in all 14 fatalities and in eight of 11 non-fatal cases; where WBGT is unavailable, a heat-index screening threshold of 29.4 °C was suggested ^[74]. The limits work. They are simply not applied to the workforces in this review.

Making them applicable is partly a measurement problem, and that problem is being solved. WBGT can be calculated from ordinary meteorological data as a climate-change assessment tool ^[75], estimated from remote and low-cost sources ^[76], and is now published on a global grid as an extreme-heat risk indicator ^[77]. WBGT has been revisited critically after sixty years of use ^[78] and compared against alternative thermal indices ^{[79] [80] [81] [82]}. Heat exposure in sugarcane harvesters has been characterised directly in Costa Rica ^[83], El Salvador and Nicaragua ^{[84] [85]}, and Thailand ^[86]; the workload itself has been described as "an excessively strenuous occupation" ^[87].

The climate argument

This is where CKDu stops being a regional occupational problem.

Physical work capacity is a measurable function of heat, and the relationship is now quantified across the full range of conditions climate change will produce: using WBGT, work capacity falls 10% at mild heat stress (WBGT 18 °C) and 78% in the most extreme conditions (WBGT 40 °C) ^[88]. A systematic review of climate change and occupational health found a significant proportion of studies concerned Mesoamerican nephropathy specifically, and reported that a day exceeding 32 °C can reduce daily labour supply in exposed sectors by up to 14% ^[89]. Escalating summer heat exposure has been framed as a future threat even to the European workforce ^[90], and heat-stress profiles have been documented for Indian workplaces ^{[91] [92] [93]}, Gulf construction ^{[94] [95]}, and Hong Kong rebar work ^[96].

The kidney-specific consequences are already being reported outside the original epidemic zones: heat-stress nephropathy in CKDu hotspots of Odisha ^[97], kidney effects among salt-pan workers ^[98], heatwave exposure associated with rapid kidney-function decline in a Chinese cohort ^[99], ambient heat and kidney function in CKD patients from a post-hoc trial analysis ^[100], a dehydration-associated kidney failure case series from China ^[101], and an integrative review of heat stress and kidney health among agricultural workers in the United States ^[102]. Heat stress and kidney injury is now being written up as a general nephrology concern amid climate change rather than a tropical curiosity ^[103].

Two further exposures travel with the same work and deserve mention: pre-harvest sugarcane burning, whose health burden on workers and nearby communities is under active study ^{[104] [105]}, and the gendered pattern of exposure, with women in these settings carrying both work and off-work heat and particulate exposures that the male-focused literature has largely missed ^{[106] [107] [108] [109]}.

What would settle it

The International Society of Nephrology's International Consortium of CKDu Collaborators met in February 2025 to name the obstacles. Three themes: differences in CKDu burden between regions are poorly described and the inconsistent terminology makes it worse; some proposed primary causes are also progression factors in CKD of all causes, so distinguishing disease onset from progression is genuinely hard; and the field needs alternatives to self-report for capturing exposures ^[6]. That third point is the practical one — wearable and environmental exposure measurement, of the kind the wristband chemical-burden work is piloting ^[107], is what would let the heat and toxin hypotheses finally be tested against each other in the same workers.

Meanwhile the preventive answer does not depend on winning the argument. Rest, shade, water and a workload ceiling remove the injury that is measurable today, at no cost to output ^{[47] [38]}, and preventing kidney disease before it starts is where the broader nephrology field is also heading ^{[110] [111]}.