Concept · medicine
Chronic kidney disease
6 studiesEvidence last moved Sep 24, 2026
Chronic kidney disease (CKD) is lasting damage or reduced filtering by the kidneys, usually defined by a low estimated glomerular filtration rate (eGFR) or protein (albumin) leaking into the urine. This page draws on cross-sectional surveys of biomarkers and cadmium, a validation study of billing codes, a Mendelian randomization study of uric acid, and a primary-care cohort of people with stage 3 CKD.
CKD is common, often silent, and heavily studied with observational data, so it is a good place to learn how the way kidney function is measured changes what a study finds. It also shows why an association between a blood marker and CKD does not mean the marker causes it.
Studies
6
Findings
5
6 supporting · 0 challenging · 4 qualifying citations
Open tensions
2
Latest change
Concept page published
Chronic kidney disease
Currently
What we know
- Moderate CKD almost always comes with other illnesses, and the load predicts mortality.
- Cystatin C picks up links with inflammation at milder levels of kidney function than creatinine does.
- Codes rarely label someone falsely, but miss about two in three people with reduced kidney function.
- Cadmium is a plausible kidney toxicant, but it predicts individual cases poorly.
- The uric acid-CKD link seen in ordinary studies is probably not causal.
Largest unresolved question
Urinary cadmium behaves differently across settings: in NHANES it was linked to albuminuria but not to low-eGFR CKD (OR 0.70, not significant), while in Sri Lanka it rose with CKD stage. The populations, exposure levels and case definitions (eGFR-based versus albuminuria-based, with an eGFR equation not validated in South Asians) differ, so this is a difference of scope rather than a direct conflict.
Common misconceptions
A biomarker that is higher in people with CKD is a cause of CKD.
Urate is associated with CKD observationally, yet genetically higher urate did not affect kidney function. Several markers, such as TNF-alpha receptor 1 and cadmium, may rise partly because damaged kidneys clear or handle them differently.
Database studies using diagnosis codes capture most people with CKD.
In Ontario, codes caught only about a third of older adults with eGFR below 45, and missed women and people over 80 more often, which can bias comparisons between groups.
Stage 3 CKD is mainly a kidney problem managed on its own.
In the Derby cohort almost everyone had other chronic conditions and many took ten or more medicines; comorbidity count independently predicted death, mostly cardiovascular.
Related
Claim ledger
What the evidence shows
Drawn from 6 studies in this library. Mix labels say which citation roles are present; they are not a strength score. Supports means evidence for a finding; Challenges means evidence against a stated position; Qualifies marks scope.
Moderate CKD almost always comes with other illnesses, and the load predicts mortality.
In a primary-care cohort of 1,741 people with stage 3 CKD, only 4% had no other listed chronic condition, 59% took five or more medicines, and having three or more comorbidities was associated with about three times the risk of death over 3.6 years (HR 3.15).
- How often does mild kidney disease come with other illnesses?— No comparison group without CKD, mostly white volunteers; association, not causation.
Cystatin C picks up links with inflammation at milder levels of kidney function than creatinine does.
How kidney function is measured changes the picture: in 6,814 MESA adults, cystatin C correlated with inflammatory and clotting markers (such as CRP, IL-6 and fibrinogen) even in people without CKD, whereas creatinine-based eGFR tracked most markers only once CKD was present.
- Is mildly reduced kidney function linked to inflammation?— Cross-sectional with no measured GFR, so cystatin C might relate to inflammation for reasons other than filtration.
Codes rarely label someone falsely, but miss about two in three people with reduced kidney function.
Administrative billing codes are specific but insensitive for CKD: in 123,499 Ontario adults aged 66+, codes had 96.9% specificity but only 32.7% sensitivity against eGFR below 45, with lower sensitivity in women (25.7%) than men (43.7%).
Cadmium is a plausible kidney toxicant, but it predicts individual cases poorly.
Cadmium exposure is associated with kidney damage in two very different populations: in US NHANES adults, blood cadmium above 1 mcg/L went with higher odds of CKD (OR 1.48) and albuminuria (OR 1.41); in endemic Sri Lankan districts, urinary cadmium was higher in CKD of uncertain cause than in controls and rose with disease stage.
- Is everyday low-level cadmium exposure linked to kidney damage?— Cross-sectional, not adjusted for smoking, diabetes or hypertension; ROC AUC about 0.6.
- What is causing a mysterious kidney disease in rural Sri Lanka?— Cross-sectional; reduced kidney function could alter cadmium handling; AUC about 0.68.
Study Role Design N Population Outcome Is everyday low-level cadmium exposure linked to kidney damage? Supports Cross-sectionalSecondary analysis of four pooled NHANES cycles (1999-2006) in US adults; blood and urinary cadmium compared with eGFR-defined CKD and albuminuria using survey-weighted logistic regression adjusted for age, sex, race/ethnicity and BMI, plus ROC curves. N=5426 · 5426 adults aged 20 or over from the NHANES urine subsample with both cadmium and kidney measures; pregnant or menstruating women excluded. Non-institutionalised US civilian adults aged 20 years and over in NHANES 1999-2006. Chronic kidney disease (eGFR below 60 mL/min/1.73 m2) and albuminuria (raised urinary albumin-to-creatinine ratio). What is causing a mysterious kidney disease in rural Sri Lanka? Supports Cross-sectionalRandom household prevalence survey in endemic districts plus comparison of urine, blood, hair and nail toxicant levels in CKDu cases versus endemic and non-endemic controls, alongside environmental sampling (2010-2012) 6698 people aged 15-70 invited to the prevalence survey (74% responded); 733 CKDu cases identified; metal analyses on subsets (e.g. urine metals in 495 cases, 250 non-endemic controls; hair/nails in 80 cases vs 48 controls; pesticides in 57 cases vs 39 controls) Adults aged 15-70 without diagnosed diabetes in three endemic districts of Sri Lanka's farming regions, plus controls from a non-endemic district (Hambantota) Age-standardised CKDu prevalence; risk factors; concentrations of cadmium, arsenic, lead, selenium and pesticide residues in biological and environmental samples The uric acid-CKD link seen in ordinary studies is probably not causal.
Mendelian randomization using 26 urate-raising variants found no causal effect of serum urate on eGFR or CKD, although the same variants raised gout risk and observational analysis of the same cohorts showed 1.48 times the odds of CKD per 1 mg/dl urate.
Debates
Tensions and limits
Some items are genuine disagreements on the same question. Others mark different assays, populations, or outcomes.
Urinary cadmium behaves differently across settings: in NHANES it was linked to albuminuria but not to low-eGFR CKD (OR 0.70, not significant), while in Sri Lanka it rose with CKD stage. The populations, exposure levels and case definitions (eGFR-based versus albuminuria-based, with an eGFR equation not validated in South Asians) differ, so this is a difference of scope rather than a direct conflict.
Urinary cadmium behaves differently across settings: in NHANES it was linked to albuminuria but not to low-eGFR CKD (OR 0.70, not significant), while in Sri Lanka it rose with CKD stage. The populations, exposure levels and case definitions (eGFR-based versus albuminuria-based, with an eGFR equation not validated in South Asians) differ, so this is a difference of scope rather than a direct conflict.
- Is everyday low-level cadmium exposure linked to kidney damage?
- What is causing a mysterious kidney disease in rural Sri Lanka?
Study Role Design N Population Outcome Is everyday low-level cadmium exposure linked to kidney damage? Supports Cross-sectionalSecondary analysis of four pooled NHANES cycles (1999-2006) in US adults; blood and urinary cadmium compared with eGFR-defined CKD and albuminuria using survey-weighted logistic regression adjusted for age, sex, race/ethnicity and BMI, plus ROC curves. N=5426 · 5426 adults aged 20 or over from the NHANES urine subsample with both cadmium and kidney measures; pregnant or menstruating women excluded. Non-institutionalised US civilian adults aged 20 years and over in NHANES 1999-2006. Chronic kidney disease (eGFR below 60 mL/min/1.73 m2) and albuminuria (raised urinary albumin-to-creatinine ratio). What is causing a mysterious kidney disease in rural Sri Lanka? Supports Cross-sectionalRandom household prevalence survey in endemic districts plus comparison of urine, blood, hair and nail toxicant levels in CKDu cases versus endemic and non-endemic controls, alongside environmental sampling (2010-2012) 6698 people aged 15-70 invited to the prevalence survey (74% responded); 733 CKDu cases identified; metal analyses on subsets (e.g. urine metals in 495 cases, 250 non-endemic controls; hair/nails in 80 cases vs 48 controls; pesticides in 57 cases vs 39 controls) Adults aged 15-70 without diagnosed diabetes in three endemic districts of Sri Lanka's farming regions, plus controls from a non-endemic district (Hambantota) Age-standardised CKDu prevalence; risk factors; concentrations of cadmium, arsenic, lead, selenium and pesticide residues in biological and environmental samples
PaperFren reads this as a limit on how far one study travels — different assays, populations, or outcomes — not a forced fight between papers.
Observational data link higher urate to CKD, but genetic (Mendelian randomization) evidence does not support a causal effect. Both designs were applied to overlapping cohorts; the conflict is between an association and a causal estimate.
Observational data link higher urate to CKD, but genetic (Mendelian randomization) evidence does not support a causal effect. Both designs were applied to overlapping cohorts; the conflict is between an association and a causal estimate.
Qualified studies asking the same question reach different answers. The disagreement is listed, not scored.
Timeline
How understanding moved
Study years are when the paper was published. Evidence edits are dated changes to this page's claims. Explanations are when PaperFren added a Discovery — not a claim that the science happened that day.
2026
Concept page published
Chronic kidney disease
Change log
What changed
Dated edits to this page's evidence: studies added or removed from a claim, claims added or withdrawn, and new explanations tagged here. Rewordings are not listed.
- Concept page published
Papers
6 studies in this library bear on Chronic kidney disease, ordered by citations.
- What is causing a mysterious kidney disease in rural Sri Lanka?
Chronic kidney disease of unknown cause was far more common in endemic Sri Lankan districts than earlier estimates suggested, and patients had higher urinary cadmium that rose with disease stage.
- Is everyday low-level cadmium exposure linked to kidney damage?
In ordinary US adults, higher cadmium levels well below occupational safety limits went along with more albumin leaking into the urine, though urinary cadmium was not linked to reduced kidney filtration.
- How often does mild kidney disease come with other illnesses?
Almost everyone with moderate chronic kidney disease in primary care had other chronic conditions, and having more of them was linked to a higher risk of dying.
- Can billing codes tell us who has chronic kidney disease?
Hospital and doctor billing codes rarely flag kidney disease falsely, but they miss about two out of three older people whose blood tests show reduced kidney function.
- Is mildly reduced kidney function linked to inflammation?
As kidney function declines, markers of inflammation and blood clotting rise steadily, and this is visible even in people without chronic kidney disease when kidney function is measured with cystatin C.
- Does high blood uric acid actually cause kidney disease?
People whose genes give them lifelong higher uric acid do not have worse kidney function, suggesting the familiar link between uric acid and kidney disease is not causal.
Compare studies
Select 2–10 studies. Design and N are labels, not a ranking.
Nothing selected yet.
Questions
What is still open
Urinary cadmium behaves differently across settings: in NHANES it was linked to albuminuria but not to low-eGFR CKD (OR 0.70, not significant), while in Sri Lanka it rose with CKD stage. The populations, exposure levels and case definitions (eGFR-based versus albuminuria-based, with an eGFR equation not validated in South Asians) differ, so this is a difference of scope rather than a direct conflict.
Observational data link higher urate to CKD, but genetic (Mendelian randomization) evidence does not support a causal effect. Both designs were applied to overlapping cohorts; the conflict is between an association and a causal estimate.
Ask PaperFren about Chronic kidney disease
Study this conceptflashcards and short-answer questions
Why can Mendelian randomization reach a different conclusion about urate and CKD than an ordinary observational analysis?
Observational analyses compare people who differ in urate but also in many other ways, and low kidney function can itself raise urate, so the association (OR 1.48 per mg/dl) may reflect confounding or reverse causation. Mendelian randomization uses urate-raising variants assigned at conception, which are less tied to these factors. In the 2019 study these variants raised gout risk, proving they worked, yet had no effect on eGFR or CKD despite over 99% power. Its limits are that it estimates lifelong urate effects in European-ancestry people and relies on assumptions about pleiotropy.
A researcher wants to estimate CKD prevalence by sex from billing codes. What problem does the evidence predict?
The Ontario validation study found codes had high specificity (96.9%) but low sensitivity (32.7%), and sensitivity was much lower in women (25.7%) than men (43.7%). Prevalence would be underestimated overall and more so in women, making a sex difference appear larger than it is. Linking to laboratory eGFR would be more accurate, noting the study itself used a single eGFR and could not capture albuminuria-defined CKD.