Archived publication
This publication is no longer current or has been superseded.
Table 3: Summary of oral reference health standards for arsenic as a non-threshold contaminant, used by different international agencies
Return to where this table appears in the publication
| Jurisdiction | Acceptable risk levela |
Guideline value (μg/L)b |
Risk-specific dosea (μg/kg bw/day) |
Cancer slope factor (per mg/kg bw/day) |
Key studyc |
Critical effectsc |
Basis of valueb |
Reference |
|---|---|---|---|---|---|---|---|---|
New Zealand |
10–5 |
|
0.067 |
0.15 |
WHO (1993) drinking water guideline |
Skin cancer in southwestern Taiwan population |
Fatal skin cancers, slope factor derived from WHO (1993) drinking water guideline (risk of 6 x 10–4 at 10 μg/L), assuming a 7% mortality rate |
MfE and MoH (1997) |
New Zealand drinking water |
10–5 d |
10 |
0.0048e |
2.1 |
US EPA (1988) |
Skin cancer in southwestern Taiwan population |
WHO (1996) |
MoH (2005) |
WHO drinking water |
10–5 |
10 |
0.0012–0.0024f |
4.2–8 |
NRC (2001) |
Internal cancers (lung, bladder) in southwestern Taiwan population |
Practical quantification limit – incidence of bladder and lung cancer in US populations at this concentration range from 12–23 x 10–4, based on maximum likelihood estimates from NRC (2001) |
WHO (2003) |
10–5 d |
10 |
0.0048d |
2.1 |
US EPA (1988) |
Skin cancer in southwestern Taiwan population |
Practical quantification limit – risk of skin cancer at this concentration is 6 × 10–4, which is stated to give rise to estimated lifetime skin cancer risks of 10–4, 10–5, 10–6 for arsenic concentrations of 1.7, 0.17 and 0.017 μg/L, respectively |
WHO (1996) |
|
UK |
10–5 |
|
0.3 |
|
|
Internal cancers |
Based on equivalence with the UK drinking water standard of 10 μg/L assuming 2 L/day is consumed by a 70-kg adult, to avoid disproportionately targeting soil exposures |
EA (2009) |
Canada soil |
10–6 |
|
0.0006 |
1.75 |
Tseng et al (1968) and Tseng (1977), both cited in WHO (2001) |
|
US EPA (1988) |
Environment Canada (1999) |
Canada drinking water |
10–5 –10–6 |
0.3 |
0.0086 |
i |
Morales et al (2000) |
Internal cancers (lung, bladder) in southwestern Taiwan population |
0.3 μg/L was considered to pose negligible risk (95th confidence interval of the lifetime risk is 1.9 x 10–6 to 13.9 x 10–6) Health Canada (2005) concluded that a Poisson model recommended by the US. EPA (2001) and fit by Morales et al (2000) with an external unexposed comparison population is the most appropriate for estimating the cancer risks associated with the ingestion of arsenic in drinking water. Health Canada (2005) adopted assumptions similar to those of the US EPA (2001) regarding the choice of risk metric and the use of a southwestern Taiwanese to Canadian conversion factor |
FPTCDW (2006) |
US EPA – IRIS, R6, R9 |
10–6 |
|
0.00067 |
1.5 |
US EPA (1988) |
Skin cancer in southwestern Taiwan population |
Quadratic-linear multi-stage model that included adjustment for larger water consumption, lower bodyweight and intake from food, in the Taiwanese population |
US EPA (1998) |
US EPA |
|
10 |
0.01–0.045j |
k |
Chen et al (1985; 1988; 1992), Wu et al (1989), all cited in NRC (1999) |
Internal cancers (lung, bladder) in southwestern Taiwan population |
Poisson model with no comparison population, Taiwanese to US conversion, risk estimates from Morales et al (2000) |
US EPA (2001) |
US EPA |
|
|
0.003m |
3.67 |
|
Internal cancers (lung, bladder) in southwestern Taiwan population |
Based on US EPA (2000), proposed rule for changes to the arsenic drinking water guideline |
US EPA (2008) |
a Where the acceptable risk level for a given jurisdiction is not 10–5, the risk-specific dose for a risk of 10–5 is shown in square brackets.
b Where a guideline value is provided, the risk-specific dose has been derived assuming consumption of 2 L/day by a 70-kg adult at the acceptable risk level or actual risk level if appropriate (see d).
c As reported in the reference cited in the reference column.
d Nominal acceptable risk level, actual risk level associated with the guideline value shown in parentheses.
e Calculated from the actual risk level associated with the guideline value.
f Calculated from cited cancer incidence rates (maximum likelihood estimates of bladder and lung cancer in males and females 12–23 per 10,000 population at 10 μg/L in drinking water), ie, 12/10,000.
g Actual risk level associated with the risk-specific dose.
h Range determined from 95% confidence interval for risk at 0.3 μg/L.
i Slope factor is not calculated as no specific risk was provided for the guideline value of 0.3 μg/L.
j Calculated from the risk estimate range presented in Table IIID.2 in US EPA (2001).
k Slope factor is not calculated as no specific risk was provided for the guideline value of 10 μg/L.
m Based on an excess lifetime cancer risk of 10–5.
