Chronic arsenicism and cancer in Inner Mongolia

Consequences of well-water levels greater than 50 m g/l

Z.D. Luo, Y.M. Zhang, L. Ma, G.Y. Zhang, X. He, R. Wilson. D.M.Byrd,
J.G. Griffiths, S. Lai, L. He, K. Grumski, and S.H. Lamm


Inner Mongolia is an autonomous region of China south of the Gobi desert and north and west of the great wall of China. (Figure 1). Its' capital was established as Huhhot in 1954. The Huhhot region lies in a triangular, segmented and sunken lake basin in Inner Mongolia, China, south of the Daqing (Great Green) Mountains and along the northern coast of the Yellow River (Figure 2). This region comprises approximately 4800 km2. The subterranean waters occur in a Q4 earth stratum and have a naturallv occurring high concentration of arsenicfrom the local rock. High levels of arsenic (mainlv As+3) and of fluoride are found in the water. as are low levels of dissolved oxygen. sulfate. nitrate, and selenium, and traces of iron and manganese. The aquifers provide a reducing envi ronment. rich in organic substances, in which arsenic easily moves and accumulates. The western Huhhot basin is an agricultural area, raising wheat. Millet, corn, green beets, potatoes and sunflowers. Arsenical pesticides have not been used. No factories, mines or other industries discharge arsenic into the local air, water or soil. Examination of the surface soils, air, fish and crops have not found arsenic levels above the Chinese standards. The smoking habits in Huhhot resemble those of the general Chinese culture.

Cases of chronic arsenicism were first recognized clinically at Zhi Ji Liang village in the Huhhot area of Inner Mongolia, when the Anti-Epidemic Station of Huhhot investigated requests for dermatological medications in May 1990 (Luo Zhendong et al., 1993). Although national water qualitv studies in 1984 had demonstrated the presence of elevated levels of arsenic. no prior clinical diagnoses of chronic arsenicism had been made. Clinical studies were conducted on the residents of three villages in the western area of the Huhhot Basin. The source of well water was identified for study participants. The well-water arsenic levels were measured, and the participants' exposure characteristics were determined. Arsenic levels varied among the wells, with 426 wells providing water for a 36OOkm2 area. Epidemiological analyses were conducted on the findings for those residents of the three villages who, for at least six months, had used the wells with arsenic levels exceeding the national standard of 5O m g/l. Studv design, medical and environmental data collection and assessment, and data analysis have been conducted bv the staff of the Huhhot Sanitation and Anti-Epidemic Station. This presentation is the work of the Inner Mongolia Cooperative Arsenic Project (IMCAP).


The water quality analysis studies used a standard calorimetric method (GB5750-85. silver diethyldithiocarbamate) (Fan Cheng-wan et al.. 1993: Zhang Yumin et al.. 1994). This method has been validated over this concentration range using an atomic absorption method. The residents' drinking water sources varied and were from shallow well water. deep well water. And surface water. Shallow wells. 4-10m deep, are either operated as manual or big-mouth wells. Shallow wells mav extend to 30m deep. Large-mouth shallow public wells were dug in the 1950s for agricultural purposes. Deep wells are usually 90-400m deep with iron or steel pipes.although some have cement columns that are 50-60m deep. Deep wells have provided water to a large part of the communities since the 1980s. The surface wells are located near the Yellow River and the People's Canal and contain water from these sources. Overall arsenic concentrations have been examined from 355 water sources in 62 villages of 15 counties - 305 shallow wells, 33 deep wells, and 17 surface sources. Other water quality parameters investigated included dissolved oxygen, fluoride, selenium, nitrate, iron and manganese.

Table 1 Diagnostic criteria for the classification of chronic arsenicism

1) At least 6 months exposure to arsenic levels of greater than 50m g/l

2) Hyperkeratosis on the palms and soles, hyperpigmentation and hypopigmentation on the trunk

3) Neuropathy and/or multiple pre-cancerous skin lesions

4) Hair arsenic levels greater than 520m g/g and/or urine arsenic levels greater than 88m g/l

Table 2 Dermatologic severity grade of chronic arsenicism I

Low grade skin lesion

a. No obvious hyperkeratosis of the palms and soles.

b. Maculopapular lesions (barely palpable spots) that resemble rice grains.

c. Limited pigmentary changes on trunk, pale or light brown color.

II. Middle grade skin lesion

a. Obvious hyperkeratosis of the palms and soles.

b. Palpable or wart-like bumps over large area, separated.

c. Coarse skin with moderate sized spots of pigmentation that are less pronounced in trunk and limbs and are distributed in web-like form. III. High grade skin lesion a. Very obvious hyperkeratosis of the palms and soles. Warty with hollow centers. b. Coalesced coarse hvperpigmented skin over large areas. and areas of bleeding. ulcers. or neoplasm. etc. c. Brown or black pigmentation. May show piebaldism

Clinical investigation revealed the presence of many persons with hvperkeratosis (marked thickening of the outer layer of the skin) and hyper/hvpopigmentation (mottled or multi-color skin). Diagnostic criteria of chronic arsenicism included evidence of elevated exposure (>5Om g/l for at least six months), clinical findings, and possibly evidence of body burden of arsenic (Table 1). Severity of clinical findings were categorized o a three-point scale that graded the observed skin thickening and pigmentation changes (Table 2). Figures 3 to 8 demonstrate the typical dermal findings. Biopsies were taken of neoplastic tissues, whether appearing malignant or benign. Biological specimens were obtained to measure the arsenic content of hair and the arsenic concentration of urine. Case definitions were based on diagnostic criteria     (Table 3).

Table 3 Case definition of chronic arsenicism

1. Criteria 1 plus criteria 2 Grade II for both hvperkeratosis and hyper/hypopigmentation.

i.e., Arsenic exposure >50 m g/l and >6 months duration plus hyperkeratosis and hyper/hypopigmentation Grade II Alternatively:

2. Criteria 1 plus criteria 2 Grade I for both hvperkeratosis and hyper/hypopigmentation plus criteria 4.

i.e.,Arsenicexposure >50 m g/l and >6 months duration and hair arsenic >520m g/g and/or urine arsenic >88 m g/l plus hyperkeratosis and hyper/hypopigmentation Grade I. Full population medical health surveys were conducted in the three villages found to have cases of chronic arsenicism. The medical surveys included both brigades later found to have elevated arsenic levels (>50 m g/1) and brigades without elevated arsenic levels. Most analyses in this preliminary report were limited to individuals exposed to >50 m g/l, since such exposure was included in the original case definition. Nonetheless, the clinical data had been collected for nearlv all (95.7%) of the residents of the three villages studied, whether their well-water arsenic level exceeded 50 m g/l or not. Future reports will extend the analysis to residents exposed to well-water arsenic levels up to 50m g/l.


3.1 Arsenic

High arsenic concentrations occurred in shallow and deep wells but not from surface sources (Table 4). The arsenic-contaminated deep aquifer was found to have a widespread and continuous distribution, forming a large flat area that is located in the central Huhhot Basin (>1000 km2). In contrast, the shallow wells with high arsenic levels were located along the fringe of the basin depression, facing the mountain and the alluvial plain. These wells had a distribution that was discontinuous and limited.

Table 4 Arsenic concentrations in the western part of the Huhhot Basin by well type

Well Type    Number of wells tested    Concentration Range               % above
                                                                         50m g/l
                                                                          ( m g/l)

Shallow                      305                                 <10-1860                              20.7
Deep                            33                                 <10-360                                54.6
Surface                         17                                 <10-20                                    0

Total                           355                                 <10-1860                              22.3

The arsenic concentrations in shallow wells, ranging up to 1860 m g/l, were greater than the national standard of 50 m g/l in 63 of the 305 samples (20.7%). The arsenic concentrations in deep wells, ranging up to 360 m g/l, were greater than the national standard in 18 of 33 wells (54.6%). The arsenic concentrations in surface waters, ranging up to 2O m g/1, were not elevated above the national standards in any of 17 tested sources. Generally, deep wells were motor-pumped, shallow wells were manually operated, and large-mouthed wells used buckets.

Arsenic valence determinations generally showed a predominance of As+3 (52-75%), both for shallow wells and for deep wells. Overall, arsenic concentrations above the national standard were measured from 81 of 355 water sources (22.3%). Similar arsenic concentrations occur in the alluvial and lake plains of the Yellow River basin and the Big Black River basin, indicating a large area of excessive arsenic in the drinking water.

3.2 Non-arsenic water quality results

Numerous parameters of water quality were assessed in addition to the arsenic concentration.

(a) Dissolved oxygen

Levels ranged from 120-340 m g/l in 8 hand-pumped samples. was 130 m g/l in a large-mouth sample, and was undetectable in an artesian well sample. Dissolved oxygen levels of 350-1300 m g/l generally indicate an oxygenated environment. Dissolved oxygen levels of 0-350 m g/l generally indicate a reduced environment. Thus, both the shallow and deep aquifers appeared to be reducing environments

(b) Fluoride

Fifty-three of 385 samples from deep and shallow wells (14%) exceeded the national standard for fluoride (1 mg/1). Twenty-one of the 53 samples (40%) exceeded 2 mg/I. Five wells had excessive levels of both fluoride and arsenic.

(c) Selenium

The highest level for selenium in 18 specimens was 0.0014 mg/l, which is approximately 1% of the national standard of 0.1 mg/l.

(d) Nitrate

Specimens from three deep wells showed no detectable nitrate. Specimens from 12 shallow wells showed nitrite levels between 2 and 12 m g/l in four (33 %), less than 2 m g/l in five (42 %), and no nitrite was detected in the three remaining samples (25%).

(e) Iron

Detectable traces of iron were found in 7 out of 10 deep well samples (70%) with two samples (20%) greater than 0.3 mg/l. Seventy out of 125 shallow well samples (56%) had detectable traces of iron, seven (6%) of which were greater than 0.3 mg/l.

(f) Manganese

Twenty-five of 27 shallow ground water samples (93%) had detectable manganese, of which three, (11%) exceeded the national standard of 0.1 mg/l.

3.3 Epidemiological results

The three villages had a population of 3329 residents, of whom 3185 were medically examined (95.7%, Table 5.5). Tie Men Geng had 280 residents, of whom 228 were examined (81.4%). Brigade # 4 of Zhi Jhi Liang had 271 residents, all of whom were examined (100%), and Brigade # 2 of Black River Village had 333 residents, all of whom were examined (100%). The remaining brigades of Zhi Jhi Liang had 1054 residents, of whom 962 were examined (91.3%), and the remaining brigades of Black River Village had 1391 residents, all of whom were examined (100%).

Table 5 Cases of chronic arsenicism in three studied villages

Village                  Population       Number        Cases        Examined%         Prevalence %

Tie Men Geng                280                 228                 42               81.4                    18.4

Zhi Jhi Liang                1326               1233                28                92.9                     2.3

   Brigade 4                    271                 271                28              100                      10.3

   Brigades 1, 2, 3, 5     1054                962                   0*              91.3                       --

Black River (Hei He)    1724              1724                 98              100                        5.7

  Brigade 2                      333                333                 98              100                     29.4

  Brigades 1, 3, 4, 5       1391             1391                   0*            100                          --

Areas w. As >50mg/l    884                832               168               94.1                    20.1

Areas w. As <50mg/l   2445             2353                 --                 96.2                        --

Total population            3329             3185              168                95.7                        --

*Arsenic level in weil-water did not exceed 50 m g/L.

A total of 168 cases of chronic arsenicism were diagnosed among 832 examined residents from the areas of the three villages where the well-water arsenic levels exceeded 50 m g/l. Forty-two cases of chronic arsenicism were diagnosed among the examined residents of Tie Men Geng (18. among the examined residents of Zhi Ji Liang Brigade #4 (10.3%), and 98 among the residents of Black River Village Brigade #2 (29.4%). The residents of the other four brigades of Zhi Ji Liang and of Black River Village did not meet the exposure requirements for case definition. Males exhibited signs of chronic arsenicism significantly more often than did females (chi square = 7.59: P < 0.05). No cases of blackfoot disease were observed. Ages of individuals with chronic arsenicism ranged from 5 to 76 years; prevalence increased with age from about 10% prior to age 20 to about 40% at 60 (+) years of age (Table 6).

Table  6      Chronic arsenicism by age for persons in Tie Men Geng and Zhi Jhi Liang Brigade 4 with exposure known to be >50 m g/l As

Age                                            Prevalence (%)                                R elative prevalence

5-19 years                                             11                                                      1.0

20-39 years                                           24                                                      2.2

40-59 years                                           32                                                      2.9

60 (+) years                                           39                                                      3.5

P = 0.0001 for trend.

The relationships between chronic arsenic poisoning and arsenic concentrations in drinking water were investigated in villages Tie Men Geng and Zhi Ji Liang. The incidence rate of chronic arsenic poisoning increased with increasing arsenic concentration. Arsenic exposures were determined for 945 of the examined people. Exposure levels were less than 50m g/l for 624 people and 50 m g/l or greater for 321 people. Seventy cases of chronic arsenicism were identified among these 321 people (21.8%), with the prevalence rate increasing with increasing exposure levels (Table 7)

Table 7 Chronic arsenicism and arsenic concentration in drinking-water for persons in Tie Men Geng and Zhi Jhi Liang Brigade 4 with exposure known to be >50m g/l As

As Concentration (mg/L)        Examined Individuals        Cases                  Prevalence %

50-199                                                  241                            26                              10.8

200-399                                                 34                             10                              29.4

400-649                                                 18                             11                              61.1

650-950                                                 28                             23                              82.1

50-950                                                  321                            70                              21.8

Table 8   Skin disease prevalence by drinking-water arsenic levels for cases in Tie Men Geng and Zhi Jhi Liang Brigade 4 with exposure known to be >50m g/l As

Arsenic Concentration      Hyperkeratosis   Hyperpigmentation   Severe Skin Lesion
      (m g/1)                           (%)                (%)                (%)

50-199                                         27                                16                                   13

200-349                                       12                                  9                                     9

350-499                                       60                                50                                   20

500-649                                       63                                50                                   50

650-799                                       91                                55                                    82

800-950                                       86                                 71                                   86

The prevalence of hyperkeratosis, hyperpigmentation, and severe skin lesions also increased with increasing level of arsenic exposure in villages Tie Men Geng and Zhi Jhi Liang (Table 8). The prevalence rate appears to increase at exposures of 350 or greater. The severe skin lesions were found to include squamous cell carcinomas, basal cell carcinomas, and Bowen's carcinomas. Full diagnostic ascertainment is not yet complete.


Residents of the western Huhhot basin have ingested drinking waters markedly contaminated by arsenic for many years. The wells from the deep aquifer were more likely (55 %) to be contaminated (arsenic level >50 m g/l), but some wells from the shallow aquifer had higher levels of contamination (up to 1860 m g/l). These waters were also high in fluoride, low in selenium and low in dissolved oxygen.

This study has shown that the prevalence of chronic arsenicism increased both with age (a marker of duration of exposure for a non-mobile, chronically exposed population) and with level of exposure (when clinical observations of the examined residents were stratified by exposure level). The prevalence of hyperkeratosis, hyperpigmentation, and of severe skin lesion increased for exposures of 350 m g/l or greater. Although Table 7 shows that 26 cases occurred among persons exposed to arsenic concentrations of 50-199 m g/l, no cases occurred at exposures below 120 m g/l in this group. The data in this current analysis do not permit a determination of whether signs of chronic arsenicism have been observed among persons exposed to levels of <50 m g/l, since exposure to greater than 50 m g/l was a criterion for inclusion in this analysis. However, the clinical records which have already been collected on the 2353 residents whose well-water arsenic level did not exceed 50 m g/l may contain the answer to that question. These records have yet to be analysed.

Chronic arsenicism is a systemic health problem. The study of long-term exposure to high levels of arsenic in drinking water merits further studv. The primary risk factor is probably the rate and duration of arsenic ingestion. The exposed population in Inner Mongolia presents an opportunity to study the range of effects and pathologies of arsenic intoxication. Given that the population is not mobile and that a large percentage of the population is exposed to water containing more than 50 m g/l arsenic, the opportunitv exists here to assess the health risks at various exposure levels and to design intervention programs. The opportunity also exists to assess other dietarv, environmental, or social factors which may influence its appearance. High fluoride, low selenium, and the nature and effect of humic acid derivatives may be primary candidates as influential factors in disease appearance. The current examination of data for those exposed to >50 m g/l arsenic can be extended to the data already collected on those in the surveillance villages exposed to waters with arsenic levels <50 m g/l Examination of biological samples (biopsies) may provide a method for identifying the mechanisms of arsenic toxicity and distinguishing the carcinogenic effects on the skin from those of sunlight. An opportunity is present to answer scientific questions about the dose relationships of adverse health effects from arsenic, including cancer (Byrd et al., 1996). In particular, an opportunity is present to develop an engineering solution to the treatment of a water system characterized by numerous widespread individual wells rather than a central municipal water supply. The detailed investigation of the arsenic problem in Inner Mongolia may provide answers for members of both the Inner Mongolian and international communities.


Byrd, D.M., Roegner, M.L., Griffiths, J.C. et al. (1996) Carcinogenic risks of inorganic arsenic in perspective. Int. Arch. Occup. Environ. Health, 68, 484-94.

Fan Chengwan, Naren Gaowa, Zhang Yumin et al. (1993) Analysis for arsenical water and approach for reason of rich arsenic in Western Huhhot Basin. Environment and Health, 10(2), 56-8.

Luo Zhendong, Zhang Yumin, Ma Liang et al. (1993) Epidemiological survey of chronic arsenic poisoning at Tie Mengeng and Zhi Jiliang villages in Inner Mongolia. Chinese Public Health, 9(8), 347-8.

Zhang Yumin, Ma Liang, Luo Zhengdong et al. (1994) Water quality analysis of arsenic-enriched groundwater in the large area of Western Huhhot Basin. Rural Eco-Environment, 10(1), 59-61.

Figures can be viewed HERE

Figure 1 Map of Inner Mongolia (from Land of Genghis Khan, by David Lai. University of Victoria. BC. p. 6.1995) showing Huhhot, south of the Da Qing Shan (Daqing mountains) and north of the Great Wall of China.

Figure 2 Map of Huhhot Region (From Huhhot government 1994).

Figure 3 Hyperkeratosis of the hands (Grade II)

Figure 4 Hyperkeratosis of the feet and hands (Grade II)

Figure 5 'Raindrop' hypopigmentation of the chest wall (includes male nipple) (Grade II)

Figure 6 Wart-like hyperkeratosis of the palms with obvious hyperpigmentation (Grade III).

Figure 7 Squamous cell carcinomas on hand, arm, and chest (Grade III).

Figure 8 Bowen=s carcinoma of the hand (Grade III)