Risk of non-Hodgkin lymphoma and nitrate and nitrite from drinking water and diet.

INTRODUCTION

Nitrate and nitrite are precursors in the in vivo formation of N-nitroso compounds, potent animal carcinogens.

METHODS

We conducted a population-based case-control study of non-Hodgkin lymphoma (NHL) in 1998 to 2000 in Iowa, Detroit, Seattle, and Los Angeles. Because nitrate levels were elevated in many drinking water supplies in Iowa, but not in the other study centers, we evaluated water nitrate levels and risk of NHL in Iowa only. Monitoring data for public supplies were linked to water source histories from 1960 onward. Nitrate was measured at interview homes with private wells. We limited most analyses to those with nitrate estimates for > 70% of their person-years since 1960 (181 cases, 142 controls). For those in the diet arm of the study (458 cases, 383 controls from 4 centers) and for Iowa participants in both the diet and drinking water analyses, we estimated dietary nitrate and nitrite intake using a 117-item food-frequency questionnaire that included foods high in nitrate and nitrite. Odds ratios and 95% confidence intervals were calculated using logistic regression, adjusting for the study matching factors, education, and caloric intake (diet analyses only).

RESULTS

We found no overall association with the highest quartile of average drinking water nitrate (> 2.90 mg/L nitrate-N: odds ratios = 1.2; 95% confidence interval = 0.6-2.2) or with years > or = 5 mg/L (10+ years: 1.4; 0.7-2.9). We observed no evidence of an interaction between drinking water nitrate exposure and either vitamin C or red meat intake, an inhibitor and precursor, respectively, of N-nitroso compound formation. Among those in the diet arm, dietary nitrate was inversely associated with risk of NHL (highest quartile: 0.54; 0.34-0.86). Dietary nitrite intake was associated with increasing risk (highest quartile: 3.1; 1.7-5.5) largely due to intakes of bread and cereal sources of nitrite.

CONCLUSION

Average drinking water nitrate levels below 3 mg/L were not associated with NHL risk. Our study had limited power to evaluate higher levels that deserve further study.