Dietary factors and risk of non-Hodgkin lymphoma by histologic subtype: a case-control analysis.

There is speculation that etiologic heterogeneity exists among tumors classified as non-Hodgkin lymphoma (NHL), although it is not known whether diet-related associations vary between tumor subgroups. We analyzed data on 1,642 NHL cases and 5,039 controls aged 20 to 74 years from a population-based case-control study conducted in eight Canadian provinces to explore associations between dietary factors and NHL by histologic subtype. Dietary information was collected using a 69-item food frequency questionnaire. Tumors were categorized into histologic subtypes using the contents of pathology reports from the original histopathologic review of diagnostic material. Odds ratios (OR) relating consumption of dietary factors (divided into three categories) to each NHL subtype (diffuse, follicular, small lymphocytic, high grade, peripheral T cell, and unspecified lymphomas) were calculated using polytomous logistic regression. We found an increased risk of NHL with high (versus low) intake of processed meat (OR, 1.49), cheese (OR, 1.38), eggs (OR, 1.49), and dessert foods (OR, 1.24). Positive associations with NHL were also found for high consumption of total fat (OR, 1.28), saturated fat (OR, 1.29), and monounsaturated fat (OR, 1.27). Associations for consumption of some vegetables and fats were found to differ between lymphoma subtypes. Given the large number of diet/subtype comparisons done, however, the possibility that this heterogeneity arose by chance cannot be ruled out. In conclusion, these findings generally do not support the existence of etiologic heterogeneity between histologic subtypes of NHL in their associations with components of dietary intake.