The European Journal of Public Health Advance Access originally published online on July 28, 2005
The European Journal of Public Health 2005 15(6):652-656; doi:10.1093/eurpub/cki027
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cancer |
The association between drinking water source and colorectal cancer incidence in Jiashan County of China: a prospective cohort study
Kun Chen1, Weiping Yu1, Xinyan Ma2, Kaiyan Yao1 and Qinting Jiang3
1 Department of Epidemiology, Zhejiang University School of Medicine, Hangzhou, People's Republic of China
2 Cancer Institute of Jiashan-county, 314100 Jiashan Zhejiang, People's Republic of China
3 Department of Biology Chemistry, Hong Kong City University, Hong Kong
Correspondence: Prof. Kun Chen, Department of Epidemiology, Zhejiang University School of Medicine, 353 Yan'an road, Hangzhou Zhejiang 310031, People's Republic of China, tel: +86 571 8721 7190, fax: +86 571 8721 7184, e-mail: ck{at}zjuem.zju.edu.cn
Received January 9, 2004, accepted May 11, 2004
 |
Abstract |
|---|
 Top Abstract Materials and methodsResultsDiscussionReferences |
|---|
Background: The pollution of drinking water, e.g. from rivers and pools, has long been recognized to be associated with an increased risk for colorectal cancer (CRC), but there are few direct prospective cohort studies related to person-years on the relative risks of different sources of drinking water for CRC, hence the reason for our study. Methods: Based on a screening for CRC among residents aged 30 years and over in Jiashan County, Zhejiang Province, China, a total of 64,115 residents were classified into five cohorts by their source of drinking water and followed-up from 1st May 1990 to 1st January 2001. Person-years was calculated for every cohort member and Poisson regression was used to control potential confounding variables including demographic variables and smoking history, and to attain crude and adjusted relative risks based on person-years. Results: A trend was seen toward increasing incidence rates for CRC from the drinking water sources of municipal, river, ditch, mixed water to well in turn as shown by relative risk rates of 29.61, 32.67, 33.45, 40.87 and 58.67 per 100,000 inhabitants. Only the role in risk of well water was significantly different from municipal water (P < 0.05). After the confounding variables were adjusted, the significant risk from well water could be seen for colon cancer, rectal cancer as well as CRC. The relative risks were 1.741 [95% confidence interval (CI) 1.001–3.029], 2.228 (95% CI 1.432–3.466) and 2.022 (95% CI 1.432–2.854), respectively. Conclusion: Drinking well water over a long period was identified as playing a role in the risk for CRC, especially for rectal cancer.
Keywords: cohort study, colorectal cancer, drinking water source, Poisson regression
Colorectal cancer (CRC), including rectal cancer and colon cancer, is one of the most common cancers. It has been estimated that more than half a million people annually become new CRC cases throughout the world. In the USA, CRC is the fourth most common cancer (after lung, prostate and breast) and the second most common cause of cancer death (after lung) (1). Recently the incidence has increased, especially in the developing world with the change towards more western diets. China is a nation with a relatively low incidence of CRC, but in the eastern part of China there has been a higher incidence of CRC reported. Jiashan County has been shown to have the highest incidence of CRC in China, with the mortality rate of 26.3/100,000 per year in males and 18.6/100,000 per year in females (2).
The pollution of drinking water has long been recognized to be associated with an increased risk of cancer in some epidemiological studies (3–5), including several studies concerning chemical and microorganism contamination (6–8). We carried out a prospective cohort study with person-years directly on the relative risks of different sources of drinking water for CRC in Jiashan County, Zhejiang Province, China.
|
Materials and methods |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionReferences |
|---|
Study cohort and baseline exposure
From 1st May 1989 to 30th April 1990, screening for CRC was carried out among 75,842 residents aged 30 years and over in 10 small towns in Jiashan County, Zhejiang Province, China, which is described elsewhere (9). Simultaneously, a face-to-face questionnaire was completed, including demographic data and information concerning drinking water source and smoking history, and other potentially related factors for colorectal cancer. A total of 64,693 participants (85.3%) responded to the questions. Questionnaire data were key-entered twice and processed blind with respect to case/cohort status to avoid bias in coding and interpretation of the data. Subjects who reported a history of cancer in the baseline questionnaire (86 individuals, 0.13%) or who died during the period May 1989 to April 1990 (129 individuals, 0.20%) were excluded. Furthermore, 13 cohort members with inconsistent data were excluded from the analysis. To acquire more information we retained the other 15 subjects with missing variables. The final total of 64,115 residents were classified into five cohorts according to their response to the question ‘What has been your main source of drinking water over your lifetime (having been drunk for the longest time)?’, with the potential answers of: municipal water, private well water, river water, ditch water and mixed water. Mixed water means mainly river water and ditch water having similar drunken years. In fact, the municipal water in Jiashan County is groundwater from below 600–800 m without chlorine disinfection.
Cohort follow-up and outcome data
Cohort members were followed-up for CRC incidence through a tumor reporting system including a rapid reporting system of CRC Registry. The report thoroughness for CRC by the two registrations was documented to be 95.2% (10). Quality control of the two registrations was performed by a search for cancer cases through all the hospitals in the region. Through to December 31, 2000, after 10.6 years and 680,508 person-years of follow-up, 242 cohort members had developed CRC, including 107 (44.2%) cases of colon cancer and 135 (55.8%) cases of rectal cancer. A case-control study based on the cohort population, carried out in July 2002, consequently estimated that the out-migration rate among the cohort population was <1% annually, therefore we regarded the cohort as a fixed cohort, and only considered death as loss to follow-up. A total of 4885 deaths, including 2331 female, 2554 male and one sex missing, in our cohort were ascertained through the Jiashan County Death Registration System. This registration system collected the overall death cases, including the cause of death, in Jiashan Country.
Statistical analysis
The exposure variable of primary interest was drinking water source. Stepwise Poisson regression was used to control the confounding effects of smoking history and the common demographic variables including age, sex, occupation, marital status and education. Age was divided into 10-year classes (<40, 40–49, 50–59, 60–69, 
70); education into illiteracy and literacy; marital status into married and others; occupation into farmer and others: smoking history into four classes (never, ex-smoker, light-smoker, heavy-smoker). Person-years were calculated with every cohort member from 1st May 1990 to end-point. The end-point was considered to be the date of diagnosis of CRC, the death date from other causes or 1st January 2001, whichever came first. Relative risks (RRs) were used with 95% confidence intervals (CIs) to estimate the strength of associations between the exposure and CRC. Poisson regression was used to control potential confounders and to attain crude and adjusted relative risks based on person-years. All these calculation were carried out on STATA Version 7.0.
|
Results |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionReferences |
|---|
The distribution of drinking water source and the putative confounding variables in the cohort at baseline are presented in table 1. Most of the cohort members (62.18%) reported using river water as their main source of drinking water and
11.23%, 9.72%, 6.85% and 10.02%, respectively, answered to drinking municipal groundwater, ditch water, mixed water and well water. Significant at P < 0.001 with Pearson chi-squared, all the distribution of age, sex, occupation, education, marital status and smoking history was imbalance among the five populations according to the five sources of drinking water. Therefore, those potential confounding variables were considered in the following multivariate regression.
|
Table 2 shows the incidence densities of CRC by different sources of drinking water. The incidence densities for colon cancer, rectal cancer and the total CRC were, respectively, 15.72, 19.84 and 35.56 per 100,000 inhabitants. A trend was seen toward increasing incidence densities of CRC from municipal groundwater, river water, ditch water, mixed water to well water in turn as shown by figures of 29.61, 32.67, 33.45, 40.87, 58.67 per 100,000 inhabitants, respectively. When colon cancer and rectal cancer were considered separately, the trend of increasing incidence was similar. Interestingly, people drinking well water have the highest incidence density of CRC, and particularly rectal cancer, which differs significantly from those drinking municipal groundwater who have the lowest CRC incidence density (P < 0.05). The incidence densities of colon cancer were also highest in the well and mixed-water sources groups, but were not significantly different from those of municipal groundwater at the 0.05 level.
|
Crude, age-adjusted and multivariate-adjusted RRs for CRC were next estimated for populations with different drinking water sources, as presented in table 3. Populations served by municipal groundwater sources were chosen as the reference category because the municipal water supply of Jiashan County is groundwater below 600–800 m without chlorine disinfection and because groundwater is considered as unpolluted water (RR = 1.0). The crude, age-adjusted and multivariate-adjusted RRs were similar when related to the same drinking water source. After adjusting for the confounding variables, the risks of well water could be seen to be significantly higher than those of municipal groundwater on all colon cancer, rectal cancer and CRC. The relative risks were 1.741 (95% CI 1.001–3.029), 2.228 (95% CI 1.432–3.466) and 2.022 (95% CI 1.432–2.854), respectively. When other drinking water sources were examined, no significant relationship with CRC was found at the 0.05 signficance level.
|
|
Discussion |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionReferences |
|---|
Previous epidemiological studies on cancer and drinking water source have found that polluted surface water sources, e.g. river water and pond water, are risk factors for CRC with obvious higher incidence density than other drinking water sources (6). In fact, the real association is due to the water quality, i.e. the carcinogens in water and the degree of pollution. Pesticide remains and nitrate pollutants were examined with interest and taken as having positive correlation to CRC (7–11). Recently, drinking water chlorination by-products, e.g. chloroform and carbon tetrachloride, have been reported to be potential carcinogens (6,12,13). Kikuchi (14) found an association between trace element levels in drinking water and the incidence of CRC in Japan, and reported that the incidence of CRC changed when the contents of gold, magnesium, selenium, calcium, phosphorus, sodium and strontium etc. changed. Hardness levels in Taiwan's drinking water were thought to relate to rectal cancer mortality in Yang's study (15).
In our study, compared with municipal groundwater, the incidence density of CRC increased from river water, ditch water, mixed-water to well water in turn. However, well water was the only one where the relative risks had significant statistical meaning. Although systematic error in factors other than drinking water source and the above putative confounding variables cannot be excluded, the observed increased risk of CRC from drinking well water is believable and comprehensible. Under certain circumstances, the quality of well water may be similar to the quality of surface water, and may also similarly easily become polluted. At the same time, well water has a very reduced flow and is far less susceptible to run-off contamination than river and ditch water. Well water is generally more mineralized and contains higher levels of naturally occurring trace elements and radioactive substances, which may be carcinogenic e.g. arsenic and selenium (16,17). It should also be noted that Jiashan County is an agricultural district and, since chemical fertilizer and pesticide were abused in the middle of 20th century, organic phosphorus and organic chlorine in particular, which degrade slowly, may remain in high concentrations in well water, particularly well water at mid-depth.
The finding that populations drinking surface and well water have higher rectal cancer and colon cancer risk than those using municipal groundwater as a drinking source is consistent with results from other investigations (6,18,19). When cancer sites are considered in this study the higher risk of rectal cancer in those drinking well water was statistically significant when considering either one factor analysis, age-adjusted analysis or multivariate analysis. While for colon cancer, the association with well water source was statistically significant only after adjusting for age, sex, occupation, marital status, education and smoking history.
Finally, some limitations of this study need to be discussed when interpreting the results. Some misclassification of exposure may have occurred, since the exposure status for study participants was determined via the questionnaire at the 1989 screening, many members might have subsequently changed their source of drinking water. However, it is likely that, in general any change would have been towards using municipal groundwater, which we considered to be unpolluted water and used as the reference in our study. Hence, the systematic bias is simple and the RR may be underestimated, which could not create a spurious association. One noteworthy problem in the study is the systematic variation in 11 years' follow-up, as was demonstrated in the present study. There may be some under-reporting in CRC Registration and Death Registration, which would influence the outcome of follow-up.
|
Acknowledgments |
|---|
This work was supported by the National Nature Scientific Foundation of China (No: 30170828).
Key points
|
|
References |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionReferences |
|---|
1 Young-In K. Diet, lifestyle, and colorectal cancer: Is hyperinsulinemia the missing link? Nutr Rev 1998;56:275–9.[ISI][Medline]
2 Li LD, Rao KQ. An analysis of cancer incidence and death register data in 1988–1992, in 11 counties, China. Bull Chinese Cancer 2000;9:435–47.
3 Page T, Harris RH, Epstein SS. Drinking water and cancer mortality in Louisiana. Science 1976;193:55–7.
4 Griffith J, Riggan WB. Cancer mortality in US counties with hazardous waste sites and ground water pollution. Arch Environ Health 1989;44:69–74.[ISI][Medline]
5 Gulis G, Fitz O, Wittgruber J, Suchanova G. Colorectal cancer and environmental pollution. Cen Eur J Public Health 1998;6:188–91.
6 Doyle TJ, Zheng W, Cerhan JR et al. The association of drinking water source and chlorination by-products with cancer incidence among postmenopausal women in Iowa: A prospective cohort study. Am J Public Health 1997;87:1168–76.
7 Weyer PJ, Cerhan JR, Kross BC et al. Municipal drinking water nitrate level and cancer risk in older women: the Iowa Women's Health Study. Epidemiology 2001;12:327–38.[CrossRef][ISI][Medline]
8 Van-Leeuwen JA, Walter-Toews D, Abernathy T, Smit B, Shoukri M. Associations between stomach cancer incidence and drinking water contamination with atrazine and nitrate in Ontario (Canada) agroecosystems, 1987–1991. Int J Epidemiol 1999;28:836–40.
9 Chen K, Cai J, Liu XY, Ma XY, Yao KY, Zheng S. Nested case-control study on the risk factors of colorectal cancer. World J Gastroenterol 2003;9:99–103.[ISI][Medline]
10 Ma XY, Yao KY. A investigation of report quality and fail to report rate of cancer register system in Jiaoshan county. Zhejiang J Prevent Med 2002;14:19–20.
11 Gulis G, Czompolyova M, Cerhan JR. An ecologic study of nitrate in municipal drinking water and cancer incidence in Trnava District, Slovakia. Environ Res 2002;88:182–7.[Medline]
12 Brennan RJ, Schiestl RH. Chloroform and carbon tetrachloride induce intrachromosomal recombination and oxidative free radicals in Saccharomyces cerevisiae. Muta Res 1998;397:271–8.
13 Backer LC, Ashley DL, Bonin MA et al. Household exposures to drinking water disinfection by-products: whole blood trihalomethane levels. J Expos Anat Environ Epidemiol. 2000;10:321–6.
14 Kikuchi H, Iwane S, Munakata A, Tamura K, Nakaji S, Sugawara K. Trace element levels in drinking water and the incidence of colorectal cancer. Tohoku J Exp Med 1999;188:217–25.[CrossRef][ISI][Medline]
15 Yang CY, Tsai SS, Lai TC, Hung CF, Chiu HF. Rectal cancer mortality and total hardness levels in Taiwan's drinking water. Environ Res 1999;80:311–6.[Medline]
16 Chiou HY, Chiou ST, Hsu YH et al. Incidence of transitional cell carcinoma and arsenic in drinking water: a follow-up study of 8,102 residents in an arseniasis-endemic area in northeastern Taiwan. Am J Epidemiol 2001;153:411–8.
17 Vinceti M, Nacci G, Rocchi E et al. Mortality in a population with long-term exposure to inorganic selenium via drinking water. J Clin Epidemiol 2000;53:1062–8.[CrossRef][ISI][Medline]
18 Bean JA, Isacson P, Hausler WJ, Kohler J. Drinking water and cancer insidence in iowa. Am J Epidemiol 1982;116:912–22.
19 Gottlieb MS, Carr JK, Morris DT. Cancer and drinking water in louisiana: colon and rectum. Int J Epidemiol 1981;10:117–25.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Osler The life course perspective: a challenge for public health research and prevention Eur J Public Health, June 1, 2006; 16(3): 230 - 230. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||