Background: The etiology of non-Hodgkin lymphoma (NHL) is poorly understood, especially regarding modifiable risk factors that could inform prevention messages. Obesity and diet are potentially modifiable risk factors that may contribute to lymphomagenesis by influencing inflammation or carcinogen exposure. Prospective data on obesity, diet, and NHL risk are limited and inconsistent. After 14 years of follow-up in the Nurses' Health Study (NHS), we observed significant associations of fat and vegetable intake with NHL risk. We expanded these analyses to further examine the roles of obesity and diet in NHL etiology in the NHS and Health Professionals Follow-up Study (HPFS) cohorts after substantially longer follow-up.
Methods: The NHS was established in 1976, when 121,700 female nurses aged 30–55 years completed a self-administered questionnaire on lifestyle behaviors and medical history. The HPFS comprises 51,529 male dentists and other health professionals aged 40–75 years at baseline in 1986. Biennial questionnaires are sent to participants to update risk factor information and identify new diagnoses and have included a validated semiquantitative food frequency questionnaire (FFQ) to assess diet every 2 to 4 years since 1980 in the NHS and 1986 in the HPFS. The present analyses included 91,227 women who completed the 1980 NHS-FFQ and 47,542 men who completed the 1986 HPFS-FFQ and had no prior history of cancer. We confirmed 966 incident diagnoses of NHL among women through 2008 and 566 cases among men through 2006. Cox proportional hazards models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for risk of NHL associated with current and young adult (NHS, age 18; HPFS, age 21) body mass index (BMI) and selected dietary factors (i.e., total fat, animal fat, vegetable fat, saturated fat, trans fat, fruits, and vegetables). We examined updated cumulative average food and energy-adjusted nutrient intakes. Multivariable models controlled for potential confounding by age, race, BMI, height, smoking history, physical activity, and total energy intake.
Results: Young adult BMI was positively associated with NHL risk in both women (p-trend=0.02) and men (p-trend=0.002) in multivariable models. Women with BMI ≥30 kg/m2 at age 18 (vs. <21 kg/m2) had a 19% higher risk of NHL (HR: 1.19; 95% CI: 0.74–1.91). Men with BMI ≥30 kg/m2 at age 21 (vs. <23 kg/m2) had a 63% higher NHL risk (HR: 1.63; 95% CI: 0.92–2.88). Current BMI was also associated with NHL in men (HRBMI ≥30: 1.48; 95% CI: 1.05–2.09) but not in women (HRBMI ≥30: 0.88; 95% CI: 0.70, 1.12). Neither total nor specific dietary fats were significantly associated with NHL; however, women in the top quintile (Q5) of trans fat intake (vs. Q1) had a non-significantly increased NHL risk (HR: 1.19; 95% CI: 0.95–1.51). Men and women who consumed ≥4 servings of vegetables/day (vs. <2 servings/day) had a slightly lower risk of NHL (HRmen: 0.87, 95% CI: 0.67–1.13; HRwomen: 0.84, 95% CI: 0.67–1.04). Total fruit intake was not associated with NHL.
Conclusions: These preliminary analyses in two large prospective cohorts support the hypothesis that obesity in young adulthood predicts increased risk of all NHL in both men and women. We also observed a suggested positive association of NHL risk with trans fat intake in women and a modest inverse association with vegetable intake in men and women, consistent with our earlier reports. It is plausible that there is etiologic heterogeneity across the >30 NHL histologic subtypes. Ongoing analyses to examine the relationship of obesity and diet with major NHL histologic subtypes in the NHS and HPFS may yield further insights into their roles in NHL etiology and prevention.
Citation Information: Cancer Prev Res 2011;4(10 Suppl):A73.