Studies such as this are an enormous undertaking and authors are to be commended for their effort. Assessing actual exposure is typically the weakest component as personal interviews and after-the-fact surveys are known to be unreliable. Water system monitoring records are not intended to represent human exposure. This could sway the subtle associations reported here up or down. At best it can be concluded that additional study should be considered. But even if the associations reported here were much higher or much lower, such results could be misleading given the inherent limitations of the study design. This comment is not intended as a criticism of this work. But recognition of the realities associated with this type of epidemiological study. Learn what we can from it and then move on.
Espejo-Herrera N, Gràcia-Lavedan E, Boldo E, Aragonés N, Pérez-Gómez B, Pollán M, Molina AJ, Fernández T, Martín V, La Vecchia C, Bosetti C, Tavani A, Polesel J, Serraino D, Gómez Acebo I, Altzibar JM, Ardanaz E, Burgui R, Pisa F, Fernández-Tardón G, Tardón A, Peiró R, Navarro C, Castaño-Vinyals G, Moreno V, Righi E, Aggazzotti G, Basagaña X, Nieuwenhuijsen M, Kogevinas M, Villanueva CM. Colorectal cancer risk and nitrate exposure through drinking water and diet. International Journal of Cancer. 2016 Mar 8. doi: 10.1002/ijc.30083.
Ingested nitrate leads to the endogenous synthesis of N-nitroso compounds (NOCs), animal carcinogens with limited human evidence. We aimed to evaluate the risk of colorectal cancer (CRC) associated with nitrate exposure in drinking water and diet. We conducted a case-control study in Spain and Italy during 2008-2013. Hospital-based incident cases and population-based (Spain) or hospital-based (Italy) controls were interviewed on residential history, water consumption since age 18, and dietary information. Long-term waterborne ingested nitrate was derived from routine monitoring records, linked to subjectś residential histories and water consumption habits. Dietary nitrate intake was estimated from food frequency questionnaires and published food composition databases. Odd ratios (OR) were calculated using mixed models with area as random effect, adjusted for CRC risk factors, and other covariables. Generalized additive models (GAMs) were used to analyze exposure-response relationships. Interaction with endogenous nitrosation factors and other covariables was also evaluated. We analyzed 1869 cases and 3530 controls. Average waterborne ingested nitrate ranged from 3.4 to 19.7 mg/day, among areas. OR (95% CIs) of CRC was 1.49 (1.24, 1.78) for >10 vs. ≤5 mg/day, overall. Associations were larger among men vs. women, and among subjects with high red meat intake. GAMs showed increasing exposure-response relationship among men. Animal-derived dietary nitrate was associated with rectal, but not with colon cancer risk. In conclusion, a positive association between CRC risk and waterborne ingested nitrate is suggested, mainly among subgroups with other risk factors. Heterogeneous effects of nitrate from different sources (water, animal and vegetables) warrant further research.