Skip to main content
. 2024 Jan 18;22(1):e8488. doi: 10.2903/j.efsa.2024.8488

TABLE 31.

Key epidemiological studies on renal effects and As exposure.

Reference study population design Outcome definition Population size (n) case/control Arsenic exposure Results Additional information/confounders

Meliker et al. (2007)

Population in six counties in Michigan, USA, ecological study

 Mortality from kidney disease

Population in the six counties 740,000 and

1293 deaths 1979–1997. Compared to the entire state

 Population‐weighted w‐As in the six counties 7.6 μg/L and 1.3 for the rest of Michigan SMR 1.28 (1.15–1.42) in males and 1.38 (1.25–1.52) in females Adjusted for age and race by design

Hsueh et al. (2009)

Taiwan

Case control study

 CKD (eGFR < 60 over at least 3 months) from hospital files 125 cases and 229 controls from hospital pool of patients with eGFR

u‐tiAs, μg/g creatinine (cases/controls)

≤ 11.78 (19/75)

11.78–20.74 (30/78)

> 20.74 (76/76)

OR

1.0

1.41 (0.62–3.19)

4.34 (1.94–9.69)

 Adjusted for age, sex, education, ethnicity, smoking, coffee, analgesics, hypertension, diabetes from interviews

Zheng et al. (2015)

USA (Strong Heart Study)

Cohort study

 Incident CKD based on eGFR < 60, dialysis or kidney transplant 3119 individuals free of CKD at baseline. 502 incident cases

u‐tiAs, μg/g creatinine (cases/non‐cases)

≤ 5.7 (109/663)

5.8–9.7 (110/671)

9.7–15.6 (128/656)

≥ 15.6 (155/627)

HR by quartiles

1

1.1 (0.8–1.4)

1.2 (0.9–1.7)

1.6 (1.2–2.2)

 Adjusted for age, gender, location (state), education, smoking status, BMI, hypertension medication, SBP and baseline eGFR. Also, cross‐sectional analyses were performed, showing a positive association (reverse causation)

Hsu, Hsieh, et al. (2017)

Taiwan

Cohort study

 Incident CKD from National Health Insurance Registry over 14 years 6093 individuals free of CKD at baseline, 447 incident cases

w‐As (μg/L) (individuals/cases)

≤ 10 (2029/132)

10.1–49.9 (1850/140)

50–149.9 (1281/101)

≥ 150 (933/74)

HR

1.0

1.15 (0.91–1.46)

1.36 (1.05–1.76)

1.35 (1.02–1.80)

 Adjusted for age, sex, BMI, education, smoking, alcohol, regular analgesic use, hypertension, diabetes, dyslipidaemia.

Cheng et al. (2018)

Cohort study Taiwan

Cohort study

 Incident End Stage Renal Disease (ESRD = CKD requiring dialysis) in members of National Health Insurance Registry over 12 years, 1998–2010 362,505 individuals, born before 1958 and free of ESRD at baseline. 5442 incident cases of ESRD

w‐As (μg/L) aggregated by 323 townships

< 50

≥ 50

HR

1.0

1.12 (1.06–1.19)

Adjusted for age, sex, income, urbanisation

level, hypertension, hyperlipidaemia, coronary artery disease, diabetes, anaemia, congestive heart failure. Many individuals in the category ≥ 50 μg/L had very high w‐As, but the authors state in the text that the risk was similar in a category 50–349 μg/L and ≥ 350 μg/L without presenting details
Chen, Parvez, et al. (2011), Bangladesh/mean 37 years, cohort for follow‐up, cross‐sectional for baseline proteinuria (HEALS) Proteinuria Proteinuria yes/no 1030/9130

w‐As (μg/L)

Q1: 0.1–7

Q2:8–39

Q3: 40–91

Q4: 92–179

Q5: 180–864

Cohort: HR (95% CI):

1.00 (ref),

0.84 (0.70–1.01),

0.79 (0.65–1.07),

0.85 (0.70–1.04),

0.84 (0.69–1.06),

p: NS.

Cross sectional: OR (95% CI):

1.00 (ref),

1.01 (0.79–1.31),

1.33 (1.04–1.70),

1.54 (1.22–1.96),

1.65 (1.31–2.09),

p < 0.01

 Adjusted for urinary creatinine, age, gender, BMI, cigarette smoking status, education length, SBP, DBP, diabetes status. Also, data for u‐As changes over time
Huang et al. (2009), Korea/unknown age, cross sectional Kidney biomarkers of (tubular) toxicity: ß2‐microglobulin (β2 MG), N‐acetyl‐b‐D‐glucosaminidase (NAG) activity in urine Total 290

u‐tiAs μg/g (50th percentile)

< 3.94 versus > 3.94

High versus low u‐As mean (SD):

β2 MG: 74.3 (152) versus 78.3(171), p: NS;

NAG: 1.82 (1.91) versus 1.66 (2.36), p: NS.

No significant correlation for u‐As and β2‐MG or NAG

 No data for adjustments
McClintock et al. (2014), Bangladesh/18–75 years, cross‐sectional (prevalence of hematuria at baseline) and cohort (incidence of hematuria at follow‐up) (HEALS) Haematuria (dipstick test)

Cross‐sectional: 1189/6654;

cohort 5362/949

w‐As (mean 99 μg/L),

0.1–3

3–25

25–66

66–142

142–949.1

per SD increase

Cross sectional, OR (95%):

1.00 (Ref)

1.07 (0.87–1.31)

0.95 (0.76–1.17)

1.30 (1.06–1.59)

1.66 (1.37–2.02)

1.20 (1.13–1.27) p < 0.01

Cohort, HR (95%):

1.00 (Ref)

1.10 (0.88–1.37)

0.96 (0.77–1.20)

1.05 (0.84–1.31)

1.34 (1.09–1.65)

1.10 (1.04–1.16),

p < 0.01

 Adjusted for age, BMI, cigarette smoking status, education length, SBP, DBP, occupational dye exposure and change in u‐As before incident visit

Peters et al. (2014), Bangladesh/mean 36 years,

cross‐sectional (HEALS)

 Renal function using plasma cystatin C and calculated the estimated glomerular filtration rate (eGFR) Total 478, CKD stages 1–5: 95 w‐As (mean 96.5 μg/L)

eGFR prediction and log u‐As B(SE): 22.55 (1.44), p: NS, r‐square: 29.

eGFR prediction and log water‐As B(SE): 20.67 (0.56), p: NS, r‐square: 29.2

 Adjusted for log(age), sex, current smoking, log (u‐creatinine), recruitment year. Also data for urinary methylated metabolites
Peters, Hall, et al. (2015), Peters, Liu, et al. (2015), Bangladesh/30–65 years, cross‐sectional (FOX) Renal function using plasma cystatin C and calculated the estimated glomerular filtration rate (eGFR)

Total 374

eGFR< 90 versus > 90: 153/222

w‐As (mean 138 μg/L),

w‐As (mean 138 μg/L)

Nean change (95% CI) in eGFR (mL/min/1.73 m2) for a 10% increase in w‐As: −0.03 (−0.12, 0.07), p: NS; u‐As: −0.16 (−0.37, 0.04), p: NS

GFR < 90 versus > 90: mean (SD)

w‐ As μg/L: 1legend 40.67 (129.4) versus 135.9 (120.1), p: NS;

u‐As μg/L: 248.1 (215.3) versus 219.4 (196.4), p: NS

 Adjusted for log (age), sex, log (BMI), ever smoking. Also, data for glutathione

Abbreviations: As, arsenic; BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; ESRD, end stage renal disease; FOX, Folate and Oxidative Stress; GFR, glomerular filtration rate; HEALS, Health Effect of Arsenic Longitudinal Study; HR, hazard ratio; MG, macroglobulin; n, number; NAG, N‐acetyl‐b‐D‐glucosaminidase; NS, not significant; OR, odds ratio; Q, quantile; ref, reference; SBP, systolic blood pressure; SD, standard deviation; SE, standard error; SMR, standardised mortality ratio; tiAs, total inorganic arsenic; USA, United States of America; u‐As, urinary arsenic; u‐tiAs, total urinary inorganic arsenic (sum of iAs and its methylated metabolites MMA and DMA); vs., versus; w‐As, water arsenic.