A case-control study of nutrient status and invasive cervical cancer. I. Dietary indicators.

A study of 748 cases and 1,411 hospital and community controls in four Latin American countries evaluated the association between certain elements of diet and invasive cervical cancer. Subjects were interviewed about their adult consumption of 58 food items, including the major sources of putative protective agents (vitamin A, carotenoids, vitamin C, and folacin) as well as other behavioral and medical characteristics related to cervical cancer. Participation rates were above 95% for both cases and controls. After adjustment for age, study site, sexual and reproductive behavior, socioeconomic status, screening practices, and detection of human papillomavirus 16/18 by filter in situ hybridization, a slightly lower risk was observed for the highest quartiles of consumption of fruit and fruit juices, while no reductions in risk were associated with vegetables, foods of animal origin, complex carbohydrates, legumes, or folacin-rich foods. When nutrient indices were derived, significant trends of decreasing risk were observed for vitamin C (adjusted odds ratio (OR) = 0.69 for the highest vs. the lowest quartile; p for trend = 0.003), beta-carotene (OR = 0.68; p = 0.02), and other carotenoids (OR = 0.61; p = 0.003). Inclusion of vitamin C and beta-carotene in the same model attenuated the association with beta-carotene, while the association with vitamin C remained unchanged. The results are consistent with those of other investigations and provide support for a protective effect of vitamin C, carotenoids, and other substances found in the same fruits and vegetables against the development of invasive cervical cancer. However, the fact that the associations were driven by relation in two of the study sites and among women of higher socioeconomic status leaves open the possibility of selection bias or effects of unidentified aspects of dietary patterns.