Dose-Response Relationship Between Prevalence of Cerebrovascular Disease and Ingested Inorganic Arsenic

Arsenic is a ubiquitous element widely distributed in the environment. It is mainly transported in the environment by water. Humans are exposed to inorganic and organic arsenic through environmental, medicinal, and occupational sources. Inorganic arsenic has been used for the treatment of psoriasis, leukemia, and bronchial asthma and as a tonic. Both inorganic and organic forms of arsenic are present in food with different contents. However, organic forms are much less toxic than inorganic ones. Industry workers involved in smelting copper, gold, and lead ores, producing and applying pesticides, dyes, and arsenic-containing pigment, and manufacturing glass, semiconductors, and various pharmaceutical substances may have a high exposure to airborne arsenic. However, the main source of arsenic exposure for the general population is through ingestion of high-arsenic drinking water.¹² The safety level for arsenic in drinking water set by the US Environmental Protection Agency was 0.05 mg/L. In the United States, it has been estimated that ≈350 000 people may drink water containing more than this level of arsenic.³

Arsenic has been well documented as one of the major risk factors for blackfoot disease (BFD), a unique peripheral vascular disease identified in the endemic area on the southwestern coast of Taiwan, where residents had used high-arsenic artesian well water for more than 50 years. Clinically, the disease starts with numbness or coldness and ends with gangrene and spontaneous amputation of one or more affected extremities.⁴⁵ The pathological types of BFD include arteriosclerosis obliterans (70%) and thromboangiitis obliterans (30%), which develops from severe underlying systemic arteriosclerosis.⁶ An increased risk of peripheral vascular disease as a result of drinking water containing inorganic arsenic has also been observed in Mexico, Chile, and the Xinjiang province of China.⁷⁸⁹ The association between peripheral vascular lesions and chronic arsenic exposure through inhalation of airborne arsenic in copper smelter workers or consumption of contaminated wine in vintners in Moselle has also been described in previous studies.¹⁰¹¹

Both environmental and occupational exposure to inorganic arsenic have been shown to produce an increased but not statistically significant mortality from cerebrovascular disease (CVD). An excess mortality from CVD was associated with environmental exposure to inorganic arsenic through drinking water among BFD patients and residents of Taiwan.⁵¹² Occupational exposure to inorganic arsenic for workers in copper smelters and in pesticide manufacturing has also been related to increased mortality from CVD.¹³¹⁴¹⁵ However, most previous studies were either ecological correlation studies or occupational cohort studies. The former may have the problem of ecological fallacy, and the latter may be subject to the limitations of multiple exposure to various airborne chemicals and the healthy-worker effect, which may underestimate the arsenic-induced risk due to the fact that the severely ill are ordinarily excluded from employment. We performed this study to investigate the dose-response relationship between the prevalence of CVD and inorganic arsenic ingested through drinking well water among residents of Lanyang Basin in Taiwan.

Materials and Methods

A total of 18 villages in the four townships of Lanyang Basin located on the northeastern coast of Taiwan were included in the present study. These areas included Yukuang, Baiyun, Kuangwu, and Yutien in Chiaohsi Township; Kunting, Junghsiau, Meichern, Jishyang, Mehu, Shinnan, and Konglau in Chuangwei Township; Huhsing, Litzer, and Shiawei in Wuchieh Township; and Wuyuan, Jenchu, Puchern, and Sanchi in Tungshan Township (Fig 1). Because of the abundance of underground water, residents in Lanyang Basin had used shallow well water (<40 m deep) since the 1940s (>50 years). Although the implementation of a tap-water system was begun in the study area in the 1990s, some residents still drink well water. The water in the wells in Lanyang Basin was found to have an arsenic content ranging from undetectable to 3.59 mg/L, with a wide variation in median arsenic concentrations ranging from undetectable to 0.14 mg/L in various villages. The variation in arsenic levels in well water in Lanyang Basin was much more striking than the arsenic level in artesian well water of the BFD-endemic area.¹⁶ The main source of exposure of local residents in both areas to inorganic arsenic was through drinking well water.

Names and addresses of all adult residents in the study area were abstracted from household records kept in local household-registration offices where sociodemographic characteristics including sex, birth date, marital status, education, migration, and occupation of all members of every household are registered and annually updated. The selection of study subjects from the household-registration system was effective and efficient because of the completeness and accuracy of the registration information. A total of 8102 residents, including 4056 men and 4046 women, who agreed to participate were interviewed at home from October 1991 through September 1994. The standardized personal interview based on a structured questionnaire was conducted by four public health nurses who were well trained in interview technique and questionnaire details. Information obtained from the interview included history of well-water consumption, residential history, sociodemographic characteristics, history of cigarette smoking, history of alcohol consumption, physical activities, history of sunlight exposure, and personal and family histories of hypertension, diabetes, CVD, heart disease, and cancer. The history of cigarette smoking that was obtained included age at which the subject started smoking, average number of cigarettes smoked per day, and age at which smoking was stopped. Information concerning the age at which habitual alcohol consumption was begun, average quantity of alcohol consumption per day, and age at which alcohol consumption was stopped was also obtained. Physical activity level at work was evaluated on the basis of the type of job and hours worked per day.

A detailed history of residential village water consumption, including water source and duration of consumption, obtained from the questionnaire interview was used to derive cumulative arsenic exposure from drinking well water. A total of 3901 well-water samples (one sample from each household) were collected during home interviews, immediately acidified with hydrochloric acid, and then stored at −20°C until subsequent assay. Hydride generation combined with flame atomic absorption spectrometry was used to determine arsenic concentration in these samples. The arsenic exposure level of each study subject from drinking well water was derived from the arsenic concentration in well water of the household. The cumulative arsenic exposure from drinking well water in milligrams per liter times number of years the water was consumed (mg/L×year) for each study subject was calculated as the sum of products derived by multiplying the arsenic concentration in well water (in milligrams per liter) by the duration of drinking well water (in years) during consecutive periods of living in different villages, ie, Σ(Ci×Di), where Ci is the arsenic level in well water of the residence where a given study subject lived during period i and Di is the duration of drinking well water during the same period i. In other words, this cumulative index equates the level of arsenic in well water with the duration of drinking the water. Both cumulative arsenic exposure from drinking well water and average arsenic concentration in drinking water were available only for those subjects who had a complete history of arsenic exposure from drinking well water throughout their lifetime. For a given subject, these two arsenic exposure indices were classified as unknown if the arsenic level in the well water of any residence throughout the subject’s lifetime was not available.

Cases of CVD were identified from home-visit personal interview and ascertained by hospital medical records according to the World Health Organization criteria.¹⁷ During the interview, blood pressure was also measured according to the standard protocol recommended by the World Health Organization. Both systolic and diastolic blood pressures were measured three times with a mercury sphygmomanometer after the subject had rested for 20 minutes or longer. The average of these three measurements was used for analysis. Diagnostic criteria for hypertension was an average systolic blood pressure ≥160 mm Hg, an average diastolic blood pressure ≥95 mm Hg, or a history of hypertension regularly treated with antihypertensive drugs. Diabetes mellitus patients were defined as those who had a history of diabetes mellitus and were regularly treated with sulfonylurea or insulin.

In the analysis of the association between indices of long-term inorganic arsenic exposure and CVD prevalence, age- and sex-adjusted odds ratios and their 95% confidence intervals (CIs) were derived from multiple logistic regression models.¹⁸ Logistic regression analyses were also used to estimate the multivariate-adjusted odds ratios and their 95% CIs.

Results

Among 8102 study subjects, there were 1830 (22.6%), 2721 (33.6%), 2121 (26.2%), and 1630 (20.1%) in the age groups of 40 through 49, 50 through 59, 60 through 69, and ≥70 years old, respectively. A total of 3309 (40.9%) of subjects were illiterate, 4149 (51.3%) had an educational level of elementary school, and only 625 (7.7%) had an educational level of junior high school or above. The subjects were mainly farmers or fishermen.

A total of 139 CVD patients, including 95 with cerebral infarction, were identified and confirmed in this study. Table 1 shows age- and sex-specific prevalence of CVD. CVD prevalence increased with age in both men and women. Men had a higher CVD prevalence than women for all age groups except the 50- through 59-year-olds. The age-adjusted CVD prevalence was higher in men than in women. The age-specific CVD prevalences by inorganic arsenic level in well water are shown in Fig 2. Subjects ≥50 years old who drank well water with an arsenic concentration ≥300 μg/L had higher CVD prevalences than those who drank water with arsenic levels of 50 through 299 and <50 μg/L, respectively.

Table 2 shows no significant associations between CVD prevalence and cigarette smoking or alcohol consumption. The age- and sex-adjusted odds ratios of being affected with CVD for those who had smoked cigarettes <40 and ≥40 years were 0.84 and 1.40, respectively, compared with nonsmokers. Compared with subjects who had no alcohol drinking habit, the age- and sex-adjusted odds ratios of being affected with CVD were 1.46 and 1.47, respectively, for those who had an alcohol drinking habit for <40 and ≥40 years.

As shown in Table 3, there were significant dose-response relationships between CVD and cerebral infarction prevalence and arsenic content in well water after adjustment for age and sex. The prevalences of CVD and cerebral infarction were found to increase significantly with cumulative arsenic exposure from <0.1 to >5.0 mg/L×year. The biological gradient between the prevalence of cerebral infarction and arsenic exposure indices was much more prominent than that for CVD prevalence. Patients affected with hypertension had a significantly higher prevalence of CVD and cerebral infarction than nonhypertensive residents after adjustment for age and sex; increased prevalences of CVD and cerebral infarction were found for diabetic patients, but the association was significant for CVD only.

Table 4 shows multivariate-adjusted odds ratios of being affected with CVD and cerebral infarction for subjects with hypertension and diabetes mellitus and provides arsenic exposure indices analyzed by multiple logistic regression analysis. As shown in each model I for analysis of CVD and cerebral infarction prevalence, there were statistically significant dose-response relationships between the prevalences of CVD and cerebral infarction and arsenic content in well water after adjustment for multiple risk factors. As also shown in each model II for CVD and cerebral infarction prevalence, significant dose-response relationships were observed between cumulative arsenic exposure and prevalence of CVD and cerebral infarction after adjustment for age, sex, cigarette smoking, alcohol consumption, hypertension, and diabetes mellitus. Hypertensive patients still had a significantly higher prevalence of CVD and cerebral infarction than nonhypertensive patients, but the associations became statistically nonsignificant between the status of diabetes mellitus and the prevalence of CVD and cerebral infarction.

Discussion

CVD was the leading cause of death from 1963 to 1981 and has become the second-leading cause of death since 1982 in Taiwan.¹⁹ Although age-adjusted mortality from CVD has declined since 1967 in Taiwan, mortality from CVD increased with age, and CVD remained predominantly a disease of the elderly.^2021222324 The age-specific incidence rates of CVD in Taiwan were higher than those in the United Kingdom and the United States, similar to those in Japan, and lower than those in six cities of mainland China.^252627282930 The age-specific and age-adjusted incidences of CVD in the general population in Lanyang Basin were similar to those of the general population in Taiwan.¹⁷²⁵ Cerebral hemorrhage is reported to be more common in Oriental than in white people.²⁵²⁹³⁰ Cerebral infarction has a higher incidence and a lower fatality than cerebral hemorrhage in Taiwan.¹⁷²⁴ As observed in the present study, the percentages of cerebral infarction and cerebral hemorrhage were 68.3% (95/139) and 15.8% (22/139), respectively, among CVD patients. CVD prevalence was observed to increase significantly with age in the present study, which is consistent with findings of previous studies. The age effect may be related to the elevated cumulative exposure to environmental factors, including arsenic, in well water and/or degenerative changes resulting from aging.

Generally speaking, men had a higher mortality rate and a similar incidence of CVD compared with women in Taiwan. In the present study, men had a higher CVD prevalence than women, but the gender difference was not statistically significant after adjustment for age, arsenic exposure indices, hypertension, diabetes mellitus, cigarette smoking, and alcohol consumption. Cigarette smoking and alcohol consumption have been documented as risk factors for CVD.^31323334 However, positive but statistically nonsignificant associations with CVD were observed for cigarette smoking and alcohol consumption in the present study.

Hypertension is a well-documented risk factor for CVD. As shown in the present study, hypertension was also associated with the prevalence of CVD and cerebral infarction after adjustment for age, sex, arsenic exposure indices, cigarette smoking, and alcohol consumption. This finding is similar to those reported in previous studies.³⁵³⁶³⁷ Diabetes mellitus has also been reported as a CVD risk factor.³⁸³⁹ In the present study, no statistically significant associations between CVD and cerebral infarction prevalence and diabetes mellitus were observed.

The atherogenic effects of arsenic have been well documented. Occupational exposure to inorganic arsenic through inhalation of polluted air from copper smelting and pesticide manufacturing was found to be associated with a moderate but not statistically significant excess mortality from ischemic heart disease and CVD.¹³¹⁴¹⁵ Inorganic arsenic ingested through drinking water has been related to the development of peripheral vascular disease and ischemic heart disease among arsenic-exposed residents in Taiwan,^4512404142 Chile,⁸ and Mexico⁷ and among Moselle (Germany) vintners exposed to inorganic arsenic through contaminated wine.¹¹ However, the association between CVD and environmental exposure to arsenic has never been elucidated in previous studies. A statistically significant dose-response relationship between the prevalence of CVD and cerebral infarction and inorganic arsenic exposure through drinking well water was observed in the present study. Moreover, the association remained statistically significant after adjustment for other risk factors including age, sex, hypertension, diabetes mellitus, cigarette smoking, and alcohol consumption. On the basis of the striking dose-response relationship, it may be implied that CVD and cerebral infarction may be induced by long-term exposure to inorganic arsenic ingested through drinking well water.

Physical and chemical characteristics of well water, such as pH value and levels of arsenic, sodium, calcium, magnesium, manganese, iron, mercury, chromium, lead, nitrite and nitrate nitrogen, fluoride, and bicarbonate, have been studied intensively in the Lanyang Basin.⁴³ Arsenic level was found to be the only item that was significantly higher than the maximum allowable limit in ≈30% of well water in the study area. Arsenic is thus the main chemical in the water responsible for the increased prevalence of CVD. The mechanism of arsenic-induced CVD is yet to be elucidated. Inorganic arsenic may increase CVD risk through its effects on atherosclerosis directly or on CVD risk factors, including hypertension and diabetes mellitus. The dose-response relationship between long-term exposure to ingested inorganic arsenic and risk of hypertension and diabetes mellitus has been reported in recent studies.⁴⁴⁴⁵ Additional studies are required to examine whether there are direct effects of inorganic arsenic on the atherogenic process through interference with lipid metabolism, creation of endothelial injuries, or induction of monoclonal expansion of smooth muscle cells. Epidemiological studies have shown that arsenic and vinyl chloride monomer seem to have a specific capability to cause various vascular lesions, including angiosarcomas and atherosclerotic plaques. These observations suggest that somatic mutation and cell proliferation may play a role in the pathogenesis of atherosclerotic plaque.⁴⁶ The hypothesis that monoclonal expansion of smooth muscle cells is a key process for atherosclerosis induced by inorganic arsenic is suggested by evidence of the dual effect of arsenic on carcinogenicity and atherogenicity observed in our previous studies.⁵⁴⁷

The chemical properties of arsenic are similar to those of nitrogen and phosphorus, which are important elements of DNA, RNA, and protein. Arsenate may hinder the normal functions of some enzymes that are regulated by the process of phosphorylation and dephosphorylation through disruption of the formation of ATP from ADP and orthophosphate. Arsenide is known to react strongly with sulfhydryl groups of proteins; it may interfere with the normal biochemical functions of proteins that are regulated by the formation of −S−S− bonds involving the cysteine side chains in the proteins.¹² Whether arsenic may induce CVD through its interference with the structural or functional proteins involved in atherosclerosis requires further investigation.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes