Abstract
We evaluated the association of dietary fat and protein intake with risk of non-Hodgkin lymphoma (NHL) in a clinic-based study in 603 cases (including 218 chronic lymphocytic leukemia/small lymphocytic lymphoma, 146 follicular lymphoma, and 105 diffuse large B-cell lymphoma) and 1007 frequency-matched controls. Usual diet was assessed with a 128-item food-frequency questionnaire. Unconditional logistic regression was used to estimate ORs and 95% CIs, and polytomous logistic regression was used to assess subtype-specific risks. trans Fatty acid (TFA) intake was positively associated with NHL risk [OR = 1.60 for highest vs. lowest quartile (95% CI = 1.18, 2.15); P-trend = 0.0014], n3 (ω3) fatty acid intake was inversely associated with risk [OR = 0.48 (95% CI = 0.35, 0.65); P-trend < 0.0001], and there was no association with total, animal, plant-based, or saturated fat intake. When examining intake of specific foods, processed meat [OR = 1.37 (95% CI = 1.02, 1.83); P-trend = 0.03], milk containing any fat [OR = 1.47 (95% CI = 1.16, 1.88); P-trend = 0.0025], and high-fat ice cream [OR = 4.03 (95% CI = 2.80, 5.80); P-trend < 0.0001], intakes were positively associated with risk, whereas intakes of fresh fish and total seafood [OR = 0.61 (95% CI = 0.46, 0.80); P-trend = 0.0025] were inversely associated with risk. Overall, there was little evidence for NHL subtype-specific heterogeneity. In conclusion, diets high in TFAs, processed meats, and higher fat dairy products were positively associated with NHL risk, whereas diets high in n3 fatty acids and total seafood were inversely associated with risk.
Introduction
With >65,000 new cases of non-Hodgkin lymphoma (NHL)11 diagnosed in 2010, defining modifiable lifestyle changes that could reduce risk continues to be an important goal. Chronic inflammation secondary to infection, autoimmune disease, and obesity are important themes in NHL risk (1, 2). Several studies have examined dietary fat and protein in NHL risk because macronutrients are believed to interact with and stimulate or suppress the immune system through a number of potential mechanisms (3). In epidemiologic studies, the relationship of dietary fats with cancer risk varies widely, depending on study design and population, cancer type, and which dietary fats are consumed. For NHL, SFA (4–8) and trans fatty acid (TFA) (7) consumption has been positively associated with risk, whereas n3 fatty acid has been reported to have an inverse relationship (9, 10). Previous studies have also found animal protein to be positively associated with NHL risk (4, 8), although other studies found an inverse (5) or no association (6, 7, 11).
The general aim of this study was to evaluate the association of fats and proteins with risk of NHL overall and of common NHL subtypes including chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), diffuse large B-cell lymphoma (DLBCL), and follicular lymphoma (FL), which may have unique etiologies. Because meat and dairy are major contributors to dietary fat and protein intake in the United States, we assessed these foods in association with NHL as well (12). On the basis of previous epidemiologic and mechanistic findings, we hypothesized that animal-based proteins and fats, specifically SFAs and TFAs, would be associated with increased risk of NHL. Plant- and fish-based fats, especially n3 PUFAs, were predicted to be inversely associated with NHL risk.
Participants and Methods
Participants.
The Human Subjects Institutional Review Board at the Mayo Clinic reviewed and approved this study, and all participants provided written informed consent. Full details of the study design are reported elsewhere (13). We offered enrollment starting on September 1, 2002, to all consecutive cases of pathologically confirmed lymphoma (including CLL and Hodgkin lymphoma) with the following criteria: minimum age of 18 y; a resident of Minnesota, Iowa, or Wisconsin at the time of diagnosis; within 9 mo of initial diagnosis at presentation to the Mayo Clinic Rochester; no history of lymphoma, leukemia, or HIV; English-speaking; and able to provide written informed consent. All case materials were reviewed by a Mayo Clinic hematopathologist (W.R.M.) to confirm the diagnosis and to provide a histologic subtype classification according to the WHO Classification of Neoplastic Diseases of the Hematopoietic and Lymphoid Tissues (14). Participants included in this analysis were enrolled into the study from September 1, 2002, through February 29, 2008. Of the 1798 eligible patients identified during this time frame, 1236 (69%) participated, 183 (10%) refused, 39 (2%) were lost to follow-up (i.e., we were unable to contact them after multiple attempts), and 340 (19%) had their eligibility expire (i.e., failed to consent within 9 mo of diagnosis or despite consenting failed to provide all samples and data within 12 mo of diagnosis).
We enrolled clinic-based controls from Mayo Clinic Rochester patients with prescheduled general medical examinations (i.e., not for a diagnostic examination for a specific, active symptom or disease) in the general medicine divisions of the Department of Medicine from September 1, 2002, through February 29, 2008. Controls met inclusion criteria if they had no history of lymphoma, leukemia, or HIV infection and were at least 18 y old; a resident of Minnesota, Iowa, or Wisconsin at the time of appointment; English-speaking; and able to provide written informed consent. We used a computer program that randomly selects participants from eligible patients to frequency match controls to the case distribution on the basis of 5-y age group, sex, and geographic location of residence (8 county groupings based on distance from Rochester, MN, and urban/rural status). The participation rate was 69% or 1315 out of 1899 eligible individuals identified, with 548 (29%) refusals and 36 (2%) that did not provide all samples and data within 12 mo of selection.
Risk factors and dietary assessment.
Information on demographic characteristics, ethnicity, family cancer history, medical history, and selected lifestyle and other factors was collected by using a risk factor questionnaire. Shortly after the beginning of the study in 2003, we added a self-administered FFQ, which was previously used in the National Cancer Institute–Surveillance, Epidemiology, and End Results Interdisciplinary Case-Control Study of NHL (15), and is based on the previously validated Block 1995 Revision of the Health Habits and History Questionnaire (16–18). Completion rates for participants offered an FFQ were 79% (748 of 940) for cases and 87% (1139 of 1315) for controls. FFQ completion by cases varied slightly by age, with a slightly younger average age for those who did not complete the questionnaire (58.3 vs. 61.4 y), but did not differ by sex, educational level, or state of residence; FFQ completion by controls also varied slightly with age, with the mean age of those failing to complete the FFQ being lower (56.4 vs. 61.4 y), but did not differ by sex, educational level, or state of residence.
The FFQ included 103 food items, 25 beverages, cooking doneness for 7 meats, and use of 15 types of supplements. Participants were asked to report “on usual eating habits including foods consumed in a restaurant, as an adult, one year prior to diagnosis or participation, excluding any recent dietary changes.” For each food, participants were asked to indicate their usual portion size (small, medium, or large, with a specific amount provided for medium size) and frequency of consumption (never or <1 time/mo, 1–3 times/mo, 1 time/wk, 2–4 times/wk, 5–6 times/wk, 1 time/d, 2–3 times/d, 4–5 times/d, ≥6 times/d). We also asked how often low-fat or nonfat versions of milk, cottage cheese, hard cheese/cheese spreads, yogurt/frozen yogurt, and ice cream were eaten (never, always, sometimes, or rarely). The daily intakes of foods and nutrients were calculated by multiplying the frequency of consumption of each food by the nutrient content of the specific portion by using the Food Processor SQL nutrition analysis software (version 10.0.0; ESHA Research), under the direction of a dietitian (H.M.O.). This software program provided dietary analysis for >60 nutrients, including carotenoids, fats, and other macro- and micronutrients, and total energy. TFA information was calculated by using the 1989–1993 USDA database. Foods with the highest protein content were nearly all meat and dairy with a few soy products (Supplemental Table 1). Foods with highest fat content included meats, fried foods, condiments (salad dressings), desserts, and dairy (Supplemental Table 1).
Statistical analysis.
Exclusions made for final analysis of dietary data included 19 cases and 3 controls with missing risk factor data, 91 cases and 129 controls who had >10 missing food items or unlikely daily energy intakes (women with <600 or >5000 kcal/d and men with <800 or >6000 kcal/d), and 35 Hodgkin lymphoma cases, resulting in 603 total NHL cases and 1007 controls. Demographic and lifestyle characteristics of cases and controls are reported in Table 1.
Dietary factors of interest were categorized on the basis of the quartile distribution among the controls; dietary factors with less frequent intake were collapsed into fewer categories. We calculated ORs and 95% CIs by using unconditional logistic regression and used the lowest category of intake as the reference. We then performed 1 df trend tests by using the ordinal scoring of the intake quartiles, and significance for an association was declared at P < 0.05. Polytomous logistic regression (19) was used to simultaneously model the comparison between controls and each of 4 NHL subtypes—CLL/SLL, FL, DLBCL, and one group that included all other (less common) NHL subtypes combined. We used a 3 df Wald test to assess heterogeneity in the trend tests of the 4 subtypes, and a P value of <0.05 was considered significant. We reported only the results of the 3 major subtypes in Tables 2–4 because the “all other” group includes a mix of very different lymphoma subtypes, including T-cell lymphomas and miscellaneous B-cell subtypes. All food models were adjusted for the design variables of age, sex, residence, and total energy. Macronutrients were modeled by using the residual method and total energy was included in the model (20, 21).
We evaluated potential confounding by adding educational level (less than high school, high school graduate, some college, college graduate, or graduate/professional school), family history of NHL (first-degree relative with NHL vs. no first-degree relative with NHL), smoking (pack-years), alcohol use (never, former, or current use), and BMI (continuous) to the basic models. Minimally adjusted models including total energy and design variables were reported because other variables did not substantially alter associations. All statistical tests were 2-sided, and all analyses were carried out by using SAS (SAS Institute, Inc.).
Results
Characteristics.
Cases versus controls were well balanced on the design variables of sex (57.4 vs. 53.1% male), age (mean = 60.9 y for cases and 60.1 y for controls), and state of residence (Table 1). There were no striking differences between cases and controls for BMI, educational level, total energy intake, and smoking, whereas controls were slightly more likely to be current drinkers of alcohol (Table 1).
TABLE 1.
Selected baseline characteristics of study participants1
| Variable | Controls (n = 1007) | Cases (n = 603) |
| Sex, n (%) | ||
| Male | 535 (53.1) | 346 (57.4) |
| Age at diagnosis, y | 60.1 ± 13.7 | 60.9 ± 12.3 |
| Age group, n (%) | ||
| ≤40 y | 92 (9.1) | 37 (6.1) |
| 41–50 y | 133 (13.2) | 90 (14.9) |
| 51–55 y | 109 (10.8) | 66 (10.9) |
| 56–60 y | 106 (10.5) | 75 (12.4) |
| 61–65 y | 158 (15.7) | 104 (17.2) |
| 66–70 y | 156 (15.5) | 95 (15.8) |
| 71–75 y | 148 (14.7) | 73 (12.1) |
| ≥76 y | 105 (10.4) | 63 (10.4) |
| State, n (%) | ||
| Iowa | 186 (18.5) | 102 (16.9) |
| Minnesota | 688 (68.3) | 420 (69.7) |
| Wisconsin | 133 (13.2) | 81 (13.4) |
| Smoking, n (%) | ||
| Missing | 1 | 0 |
| Never | 551 (54.8) | 309 (51.2) |
| Former | 392 (39) | 262 (43.4) |
| Current | 63 (6.3) | 32 (5.3) |
| BMI 2 y ago, n (%) | ||
| Missing | 20 | 10 |
| <18.5 kg/m2 | 11 (1.1) | 6 (1) |
| 18.5–24.9 kg/m2 | 320 (32.4) | 172 (29) |
| 25.0–29.9 kg/m2 | 409 (41.4) | 240 (40.5) |
| 30.0–34.9 kg/m2 | 171 (17.3) | 124 (20.9) |
| ≥35.0 kg/m2 | 76 (7.7) | 51 (8.6) |
| Highest educational level, n (%) | ||
| Missing | 3 | 4 |
| Less than high school graduate | 27 (2.7) | 25 (4.2) |
| high school graduate/GED | 223 (22.2) | 117 (19.5) |
| Some college/vocational school | 286 (28.5) | 187 (31.2) |
| College graduate | 199 (19.8) | 130 (21.7) |
| Graduate school | 269 (26.8) | 140 (23.4) |
| Alcohol consumption, n (%) | ||
| Missing | 4 | 0 |
| Never | 94 (9.4) | 59 (9.8) |
| Former | 137 (13.7) | 99 (16.4) |
| Current | 772 (77.0) | 445 (73.8) |
| Total energy intake, kcal/d | 2200 ± 959 | 2200 ± 990 |
Values are means ± SDs or n (%). Percentages do not include missing values. GED, General Educational Development test.
Fat and protein.
Associations of protein and fat consumption with NHL risk are reported in Table 2. Total protein was inversely associated with all NHL, and this association was strongest for CLL/SLL and FL, although DLBCL showed a similar nonsignificant inverse trend (P-trend = 0.06). The total protein association appeared to be driven mainly by animal protein intake for all NHL and for CLL/SLL. Plant-based protein consumption was not associated with NHL risk overall, but there was a significant inverse trend specific for DLBCL, which resulted in a significant heterogeneity test. This association with DLBCL was attenuated when BMI was added to the multivariate model [OR = 0.54 (95% CI = 0.29, 1.01); P-trend = 0.08].
TABLE 2.
Protein and fat intake in association with risk of NHL and NHL subtypes: Mayo Clinic case-control study1
| NHL overall (n = 603) |
Follicular (n = 146) |
CLL/SLL (n = 218) |
DLBCL (n = 105) |
|||||||||||
| Nutrient | Controls | Cases | OR | 95% CI | Cases | OR | 95% CI | Cases | OR | 95% CI | Cases | OR | 95% CI | P-heterogeneity |
| n | n | n | n | n | ||||||||||
| Total protein | ||||||||||||||
| ≤81.2 g/d | 252 | 176 | 1.00 | 44 | 1.00 | 63 | 1.00 | 31 | 1.00 | |||||
| 81.3–92.5 g/d | 252 | 171 | 0.97 | 0.73, 1.27 | 39 | 0.89 | 0.56, 1.42 | 72 | 1.13 | 0.77, 1.66 | 31 | 0.99 | 0.59, 1.69 | |
| 92.6–105 g/d | 252 | 153 | 0.87 | 0.66, 1.16 | 39 | 0.90 | 0.56, 1.43 | 51 | 0.81 | 0.54, 1.22 | 24 | 0.78 | 0.44, 1.36 | |
| >105 g/d | 251 | 103 | 0.56 | 0.41, 0.76 | 24 | 0.54 | 0.32, 0.91 | 32 | 0.48 | 0.30, 0.76 | 19 | 0.59 | 0.32, 1.07 | |
| P-trend | 0.0003 | 0.033 | 0.0008 | 0.06 | 0.49 | |||||||||
| Animal protein | ||||||||||||||
| ≤49.0 g/d | 252 | 164 | 1.00 | 47 | 1.00 | 59 | 1.00 | 29 | 1.00 | |||||
| 49.1–62.1 g/d | 252 | 167 | 0.99 | 0.75, 1.31 | 32 | 0.67 | 0.41, 1.08 | 73 | 1.19 | 0.81, 1.76 | 26 | 0.87 | 0.50, 1.52 | |
| 62.2–76.5 g/d | 252 | 159 | 0.97 | 0.73, 1.28 | 40 | 0.86 | 0.54, 1.35 | 54 | 0.91 | 0.60, 1.37 | 28 | 0.96 | 0.56, 1.67 | |
| >76.5 g/d | 251 | 113 | 0.65 | 0.48, 0.87 | 27 | 0.56 | 0.33, 0.92 | 32 | 0.50 | 0.31, 0.80 | 22 | 0.71 | 0.40, 1.28 | |
| P-trend | 0.0081 | 0.058 | 0.0024 | 0.34 | 0.16 | |||||||||
| Plant-based protein | ||||||||||||||
| ≤21.4 g/d | 252 | 145 | 1.00 | 31 | 1.00 | 43 | 1.00 | 37 | 1.00 | |||||
| 21.5–25.0 g/d | 252 | 157 | 1.11 | 0.83, 1.47 | 35 | 1.13 | 0.68, 1.90 | 60 | 1.44 | 0.93, 2.21 | 25 | 0.69 | 0.40, 1.18 | |
| 25.1–30.1 g/d | 252 | 170 | 1.24 | 0.93, 1.65 | 45 | 1.50 | 0.91, 2.45 | 66 | 1.65 | 1.08, 2.53 | 26 | 0.74 | 0.43, 1.27 | |
| >30.1 g/d | 251 | 131 | 0.96 | 0.71, 1.29 | 35 | 1.18 | 0.70, 1.97 | 49 | 1.22 | 0.78, 1.91 | 17 | 0.49 | 0.27, 0.89 | |
| P-trend | 0.98 | 0.34 | 0.31 | 0.025 | 0.046 | |||||||||
| Total fat | ||||||||||||||
| ≤62.5 g/d | 252 | 123 | 1.00 | 34 | 1.00 | 47 | 1.00 | 20 | 1.00 | |||||
| 62.6–72.0 g/d | 252 | 180 | 1.46 | 1.09, 1.96 | 42 | 1.24 | 0.76, 2.02 | 63 | 1.34 | 0.88, 2.03 | 30 | 1.50 | 0.83, 2.72 | |
| 72.1–81.5 g/d | 252 | 164 | 1.29 | 0.96, 1.73 | 34 | 0.98 | 0.59, 1.63 | 65 | 1.32 | 0.87, 2.01 | 33 | 1.59 | 0.89, 2.86 | |
| >81.5 g/d | 251 | 136 | 1.06 | 0.78, 1.43 | 36 | 1.03 | 0.62, 1.70 | 43 | 0.86 | 0.55, 1.36 | 22 | 1.05 | 0.56, 1.98 | |
| P-trend | 0.98 | 0.85 | 0.56 | 0.84 | 0.77 | |||||||||
| Animal fat | ||||||||||||||
| ≤30.2 g/d | 252 | 130 | 1.00 | 36 | 1.00 | 51 | 1.00 | 17 | 1.00 | |||||
| 30.3–39.6 g/d | 252 | 158 | 1.24 | 0.92, 1.66 | 34 | 0.95 | 0.58, 1.57 | 59 | 1.19 | 0.78, 1.80 | 30 | 1.80 | 0.97, 3.35 | |
| 39.7–50.9 g/d | 252 | 160 | 1.23 | 0.92, 1.64 | 40 | 1.11 | 0.68, 1.80 | 60 | 1.17 | 0.77, 1.77 | 26 | 1.52 | 0.81, 2.88 | |
| >50.9 g/d | 251 | 155 | 1.13 | 0.84, 1.52 | 36 | 0.98 | 0.59, 1.60 | 48 | 0.88 | 0.57, 1.36 | 32 | 1.79 | 0.97, 3.32 | |
| P-trend | 0.47 | 0.91 | 0.58 | 0.13 | 0.32 | |||||||||
| Plant-based fat | ||||||||||||||
| ≤22.0 g/d | 252 | 124 | 1.00 | 34 | 1.00 | 30 | 1.00 | 25 | 1.00 | |||||
| 22.1–29.0 g/d | 252 | 179 | 1.49 | 1.11, 1.99 | 45 | 1.35 | 0.83, 2.17 | 67 | 2.32 | 1.45, 3.70 | 34 | 1.40 | 0.81, 2.42 | |
| 29.1–36.2 g/d | 252 | 163 | 1.41 | 1.05, 1.89 | 35 | 1.07 | 0.65, 1.78 | 61 | 2.22 | 1.38, 3.57 | 23 | 0.98 | 0.54, 1.79 | |
| >36.2 g/d | 251 | 137 | 1.15 | 0.85, 1.55 | 32 | 0.95 | 0.57, 1.60 | 60 | 2.11 | 1.31, 3.39 | 23 | 0.96 | 0.53, 1.73 | |
| P-trend | 0.51 | 0.63 | 0.009 | 0.58 | 0.029 | |||||||||
| SFAs | ||||||||||||||
| ≤20.4 g/d | 252 | 135 | 1.00 | 40 | 1.00 | 52 | 1.00 | 20 | 1.00 | |||||
| 20.5–24.4 g/d | 252 | 172 | 1.26 | 0.94, 1.67 | 38 | 0.95 | 0.59, 1.53 | 66 | 1.24 | 0.83, 1.87 | 27 | 1.33 | 0.73, 2.44 | |
| 24.5–28.4 g/d | 252 | 127 | 0.93 | 0.69, 1.26 | 29 | 0.73 | 0.44, 1.21 | 47 | 0.89 | 0.58, 1.38 | 25 | 1.24 | 0.67, 2.30 | |
| >28.4 g/d | 251 | 169 | 1.19 | 0.89, 1.58 | 39 | 0.95 | 0.59, 1.54 | 53 | 0.95 | 0.62, 1.45 | 33 | 1.57 | 0.87, 2.82 | |
| P-trend | 0.64 | 0.62 | 0.46 | 0.17 | 0.16 | |||||||||
| TFAs | ||||||||||||||
| ≤3.6 g/d | 252 | 110 | 1.00 | 27 | 1.00 | 40 | 1.00 | 17 | 1.00 | |||||
| 3.7–4.4 g/d | 252 | 146 | 1.34 | 0.99, 1.82 | 41 | 1.54 | 0.92, 2.58 | 50 | 1.26 | 0.80, 1.99 | 25 | 1.49 | 0.78, 2.82 | |
| 4.5–5.3 g/d | 252 | 172 | 1.59 | 1.18, 2.14 | 44 | 1.65 | 0.99, 2.75 | 53 | 1.35 | 0.86, 2.11 | 31 | 1.85 | 1.00, 3.43 | |
| >5.3 g/d | 251 | 175 | 1.60 | 1.18, 2.15 | 34 | 1.27 | 0.74, 2.17 | 75 | 1.88 | 1.23, 2.87 | 32 | 1.89 | 1.02, 3.50 | |
| P-trend | 0.0014 | 0.39 | 0.0029 | 0.032 | 0.48 | |||||||||
| PUFAs | ||||||||||||||
| ≤9.4 g/d | 252 | 153 | 1.00 | 38 | 1.00 | 49 | 1.00 | 35 | 1.00 | |||||
| 9.5–11.1 g/d | 252 | 187 | 1.24 | 0.94, 1.64 | 43 | 1.14 | 0.71, 1.83 | 61 | 1.27 | 0.84, 1.93 | 28 | 0.81 | 0.48, 1.38 | |
| 11.2–13.3 g/d | 252 | 157 | 1.03 | 0.77, 1.37 | 40 | 1.04 | 0.65, 1.68 | 60 | 1.23 | 0.81, 1.87 | 26 | 0.74 | 0.43, 1.27 | |
| >13.3 g/d | 251 | 106 | 0.68 | 0.50, 0.93 | 25 | 0.65 | 0.38, 1.11 | 48 | 0.96 | 0.62, 1.49 | 16 | 0.45 | 0.24, 0.84 | |
| P-trend | 0.0089 | 0.12 | 0.83 | 0.011 | 0.11 | |||||||||
| Total n3 fatty acids | ||||||||||||||
| ≤0.9 g/d | 252 | 189 | 1.00 | 45 | 1.00 | 67 | 1.00 | 38 | 1.00 | |||||
| 0.9–1.0 g/d | 252 | 159 | 0.82 | 0.62, 1.08 | 37 | 0.80 | 0.50, 1.28 | 55 | 0.80 | 0.53, 1.19 | 30 | 0.77 | 0.46, 1.28 | |
| 1.0–1.3 g/d | 252 | 161 | 0.82 | 0.62, 1.08 | 43 | 0.93 | 0.59, 1.47 | 54 | 0.77 | 0.51, 1.15 | 21 | 0.53 | 0.30, 0.93 | |
| > 1.3 g/d | 251 | 94 | 0.48 | 0.35, 0.65 | 21 | 0.45 | 0.26, 0.79 | 42 | 0.59 | 0.39, 0.91 | 16 | 0.40 | 0.22, 0.74 | |
| P-trend | <0.0001 | 0.017 | 0.018 | 0.0011 | 0.58 | |||||||||
| Total n6 fatty acids | ||||||||||||||
| ≤7.9 g/d | 252 | 152 | 1.00 | 36 | 1.00 | 51 | 1.00 | 36 | 1.00 | |||||
| 8.0–9.5 g/d | 252 | 186 | 1.24 | 0.94, 1.64 | 45 | 1.26 | 0.79, 2.02 | 55 | 1.09 | 0.72, 1.67 | 30 | 0.84 | 0.50, 1.41 | |
| 9.6–11.4 g/d | 252 | 161 | 1.06 | 0.80, 1.42 | 41 | 1.13 | 0.70, 1.82 | 64 | 1.27 | 0.84, 1.91 | 22 | 0.61 | 0.35, 1.07 | |
| >11.4 g/d | 251 | 104 | 0.68 | 0.50, 0.92 | 24 | 0.66 | 0.38, 1.15 | 48 | 0.93 | 0.61, 1.44 | 17 | 0.47 | 0.26, 0.86 | |
| P-trend | 0.012 | 0.15 | 0.97 | 0.0057 | 0.065 | |||||||||
| Total oleic acids | ||||||||||||||
| ≤18.1 g/d | 252 | 137 | 1.00 | 37 | 1.00 | 49 | 1.00 | 27 | 1.00 | |||||
| 18.2–21.2 g/d | 252 | 158 | 1.14 | 0.85, 1.53 | 38 | 1.03 | 0.63, 1.68 | 51 | 1.03 | 0.67, 1.58 | 29 | 1.07 | 0.61, 1.85 | |
| 21.3–24.6 g/d | 252 | 186 | 1.30 | 0.98, 1.73 | 44 | 1.16 | 0.72, 1.86 | 67 | 1.30 | 0.86, 1.96 | 26 | 0.92 | 0.52, 1.63 | |
| > 24.6 g/d | 251 | 122 | 0.85 | 0.63, 1.15 | 27 | 0.71 | 0.42, 1.20 | 51 | 0.98 | 0.64, 1.52 | 23 | 0.81 | 0.45, 1.46 | |
| P-trend | 0.54 | 0.32 | 0.75 | 0.41 | 0.69 | |||||||||
Unconditional logistic regression multivariable model was adjusted for total energy by using the residual method and the following design variables: age, sex, and residence. CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL, diffuse large B-cell lymphoma; NHL, non-Hodgkin lymphoma; TFA, trans fatty acid.
Neither total fat nor animal fat was significantly associated with NHL or any NHL subtypes. Although the plant-based fat and NHL risk overall were not associated, there was evidence for heterogeneity by subtype (P-heterogeneity = 0.046), with a significant positive association for CLL/SLL only (Table 2). When assessing specific types of fat, SFA intake was not associated with NHL overall or with subtypes (Table 2). TFA intake was positively associated with overall NHL risk and particularly with CLL/SLL and DLBCL risk (Table 2). In contrast, PUFA intake was inversely associated with overall NHL and particularly DLBCL. Among PUFAs, n3 and n6 fatty acids were each inversely associated with NHL risk, although n3 fatty acid intake had the strongest association with overall NHL as well as with each of the major subtypes (Table 2). Oleic acid, an MUFA, was not significantly associated with NHL risk (Table 2).
We next modeled fats and total proteins simultaneously to determine which variables were driving the associations (data not shown). The association of total protein with NHL risk was not attenuated when adjusted for trans, n3, and n6 fatty acid intakes [OR = 0.63 (95% CI = 0.46, 0.87); P-trend = 0.009]. Similarly, the association of n3 was attenuated somewhat but remained significant [OR = 0.63 (95% CI = 0.43, 0.92); P-trend = 0.047], and TFA intake [OR = 1.66 (95% CI = 1.21, 2.29); P-trend = 0.0009] was not attenuated with adjustment for total protein or the other fatty acids. However, after adjustment for total protein, TFAs, and n3 fatty acids, the association of n6 with NHL was attenuated [OR = 0.74 (95% CI = 0.50, 1.10); P-trend = 0.17]. When we adjusted TFAs, n3 and n6 fatty acids, and total protein for total antioxidant capacity from foods, previously reported by our group to be associated with reduced risk of NHL (22), only TFA results were attenuated [OR = 1.40 (95% CI = 1.01, 1.93); P-trend = 0.036] but remained significant, perhaps because of an inverse correlation between antioxidant and TFA consumption.
Foods high in fat and protein.
The relationship of NHL to intakes of commonly consumed foods high in fat and protein from dairy products (Table 3) and meat and eggs (Table 4) was investigated. When assessing total dairy product consumption, there was no significant trend with overall NHL risk or by subtype. There was evidence for heterogeneity in risk by subtype for total milk consumption (P-heterogeneity = 0.018), and total milk intake was significantly associated only with DLBCL. Drinking skim milk had a borderline inverse association with overall NHL (P-trend = 0.06) but had a stronger, significant inverse association with CLL/SLL. Drinking milk with any fat (1%, 2%, or whole milk), on the other hand, was positively associated with NHL risk, with the strongest subtype association in DLBCL. Cottage cheese intake was inversely associated with overall NHL risk, an association that was stronger for CLL/SLL. There was no association between total cheese intake and risk of overall or subtype-specific NHL, nor was there an association after stratification by high- and low- fat cheese consumption (data not shown). Cream in coffee was only marginally associated with FL, and butter was positively associated with overall NHL, although this association was attenuated after adjustment for BMI [OR = 1.21 (95% CI = 0.92, 1.59), P-trend = 0.1]. Ice cream consumption was positively associated with overall NHL risk, and the direction and trend of association were found for all major subtypes. When we excluded participants who mostly ate high-fat ice cream, the association with ice cream was attenuated [OR = 1.35 (95% CI = 0.91, 2.02); P-trend = 0.1], and when we excluded the participants who mostly ate low-fat ice cream, the association strengthened [OR = 4.03 (95% CI = 2.80, 5.80); P-trend < 0.0001], suggesting that the increased risk is related to high-fat ice cream consumption rather than any ice cream consumption. We suspected that this association may have been confounded by BMI, but adjustment for BMI did not attenuate these associations (data not shown). To determine if TFAs in ice cream were driving the association, we adjusted for this variable as well, which also did not attenuate the ice cream association with NHL (data not shown).
TABLE 3.
Dairy product intake in association with risk of NHL and NHL subtypes: Mayo Clinic case-control study1
| NHL overall (n = 603) |
Follicular (n = 146) |
CLL/SLL (n = 218) |
DLBCL (n = 105) |
|||||||||||
| Food2 | Controls | Cases | OR | 95% CI | Cases | OR | 95% CI | Cases | OR | 95% CI | Cases | OR | 95% CI | P-heterogeneity |
| n | n | n | n | n | ||||||||||
| Total dairy | ||||||||||||||
| ≤49.4 servings/mo | 252 | 128 | 1.00 | 40 | 1.00 | 49 | 1.00 | 18 | 1.00 | |||||
| 49.5–83.6 servings/mo | 252 | 179 | 1.41 | 1.04, 1.91 | 37 | 0.99 | 0.59, 1.64 | 73 | 1.48 | 0.97, 2.27 | 33 | 2.01 | 1.07, 3.76 | |
| 83.7–132 servings/mo | 252 | 152 | 1.26 | 0.91, 1.73 | 33 | 0.91 | 0.53, 1.56 | 50 | 1.06 | 0.66, 1.69 | 26 | 1.71 | 0.88, 3.35 | |
| >132 servings/mo | 251 | 143 | 1.12 | 0.79, 1.60 | 36 | 0.98 | 0.55, 1.76 | 46 | 0.88 | 0.52, 1.47 | 28 | 1.83 | 0.89, 3.75 | |
| P-trend | 0.78 | 0.87 | 0.31 | 0.21 | 0.22 | |||||||||
| Total milk | ||||||||||||||
| ≤11.7 servings/mo | 259 | 134 | 1.00 | 35 | 1.00 | 56 | 1.00 | 20 | 1.00 | |||||
| 11.8–30.0 servings/mo | 271 | 181 | 1.35 | 1.02, 1.80 | 45 | 1.33 | 0.82, 2.15 | 77 | 1.38 | 0.93, 2.04 | 26 | 1.34 | 0.72, 2.47 | |
| 30.1–60.9 servings/mo | 229 | 147 | 1.27 | 0.94, 1.72 | 36 | 1.25 | 0.75, 2.08 | 41 | 0.84 | 0.53, 1.32 | 27 | 1.64 | 0.88, 3.05 | |
| >60.9 servings/mo | 248 | 140 | 1.14 | 0.84, 1.55 | 30 | 0.99 | 0.58, 1.70 | 44 | 0.84 | 0.54, 1.32 | 32 | 1.85 | 1.01, 3.40 | |
| P-trend | 0.54 | 0.92 | 0.15 | 0.036 | 0.018 | |||||||||
| Any fat milk | ||||||||||||||
| None | 552 | 284 | 1.00 | 75 | 1.00 | 113 | 1.00 | 46 | 1.00 | |||||
| ≤14.6 servings/mo | 204 | 126 | 1.18 | 0.90, 1.54 | 27 | 0.97 | 0.61, 1.55 | 49 | 1.14 | 0.78, 1.66 | 23 | 1.33 | 0.79, 2.26 | |
| >14.6 servings/mo | 251 | 192 | 1.47 | 1.16, 1.88 | 44 | 1.35 | 0.90, 2.04 | 56 | 1.06 | 0.74, 1.52 | 36 | 1.73 | 1.08, 2.77 | |
| P-trend | 0.0025 | 0.24 | 0.63 | 0.025 | 0.027 | |||||||||
| Skim milk | ||||||||||||||
| ≤1.8 servings/mo | 267 | 194 | 1.00 | 50 | 1.00 | 70 | 1.00 | 32 | 1.00 | |||||
| 1.9–14.7 servings/mo | 244 | 133 | 0.78 | 0.59, 1.04 | 25 | 0.56 | 0.34, 0.94 | 53 | 0.87 | 0.59, 1.30 | 25 | 0.89 | 0.51, 1.55 | |
| 14.8–36.5 servings/mo | 250 | 147 | 0.82 | 0.62, 1.08 | 38 | 0.84 | 0.53, 1.33 | 61 | 0.94 | 0.64, 1.39 | 19 | 0.65 | 0.36, 1.17 | |
| >36.5 servings/mo | 246 | 128 | 0.75 | 0.56, 1.00 | 33 | 0.76 | 0.47, 1.22 | 34 | 0.56 | 0.35, 0.88 | 29 | 1.05 | 0.61, 1.79 | |
| P-trend | 0.06 | 0.42 | 0.034 | 0.84 | 0.66 | |||||||||
| Cottage cheese | ||||||||||||||
| None | 434 | 282 | 1.00 | 67 | 1.00 | 109 | 1.00 | 42 | 1.00 | |||||
| ≤0.9 servings/mo | 99 | 76 | 1.23 | 0.87, 1.72 | 16 | 1.04 | 0.57, 1.88 | 31 | 1.31 | 0.83, 2.08 | 12 | 1.29 | 0.65, 2.54 | |
| 1.0–4.2 servings/mo | 293 | 163 | 0.87 | 0.68, 1.11 | 40 | 0.90 | 0.59, 1.38 | 60 | 0.83 | 0.58, 1.18 | 33 | 1.20 | 0.74, 1.95 | |
| >4.2 servings/mo | 181 | 81 | 0.72 | 0.53, 0.98 | 23 | 0.85 | 0.51, 1.41 | 18 | 0.42 | 0.24, 0.71 | 18 | 1.09 | 0.61, 1.96 | |
| P-trend | 0.031 | 0.47 | 0.0034 | 0.60 | 0.11 | |||||||||
| Total cheese | ||||||||||||||
| ≤0.9 servings/mo | 151 | 269 | 1.00 | 40 | 1.00 | 55 | 1.00 | 24 | 1.00 | |||||
| 1.0–4.2 servings/mo | 159 | 265 | 1.06 | 0.80, 1.41 | 45 | 1.16 | 0.73, 1.85 | 52 | 0.96 | 0.63, 1.46 | 33 | 1.39 | 0.80, 2.43 | |
| 4.3–12.9 servings/mo | 183 | 297 | 1.13 | 0.85, 1.50 | 38 | 0.92 | 0.57, 1.50 | 70 | 1.21 | 0.81, 1.80 | 32 | 1.26 | 0.71, 2.22 | |
| >12.9 servings/mo | 109 | 176 | 1.12 | 0.81, 1.57 | 23 | 0.94 | 0.53, 1.69 | 41 | 1.18 | 0.73, 1.91 | 16 | 1.06 | 0.52, 2.12 | |
| P-trend | 0.39 | 0.65 | 0.30 | 0.84 | 0.62 | |||||||||
| Yogurt | ||||||||||||||
| Never | 278 | 474 | 1.00 | 59 | 1.00 | 110 | 1.00 | 45 | 1.00 | |||||
| ≤0.9 servings/mo | 47 | 63 | 1.27 | 0.84, 1.92 | 17 | 2.22 | 1.21, 4.07 | 10 | 0.68 | 0.33, 1.37 | 11 | 1.83 | 0.89, 3.75 | |
| 1.0–6.5 servings/mo | 149 | 247 | 1.05 | 0.82, 1.36 | 41 | 1.38 | 0.90, 2.12 | 49 | 0.88 | 0.61, 1.28 | 31 | 1.35 | 0.83, 2.20 | |
| >6.5 servings/mo | 128 | 223 | 1.01 | 0.77, 1.33 | 29 | 1.12 | 0.70, 1.81 | 49 | 0.99 | 0.68, 1.45 | 18 | 0.88 | 0.49, 1.57 | |
| P-trend | 0.85 | 0.40 | 0.78 | 0.92 | 0.82 | |||||||||
| Ice cream | ||||||||||||||
| None | 264 | 103 | 1.00 | 28 | 1.00 | 43 | 1.00 | 12 | 1.00 | |||||
| ≤2.1 servings/mo | 267 | 145 | 1.42 | 1.04, 1.93 | 35 | 1.28 | 0.76, 2.18 | 58 | 1.36 | 0.88, 2.10 | 24 | 2 | 0.98, 4.09 | |
| 2.2–6.3 servings/mo | 249 | 148 | 1.60 | 1.17, 2.19 | 39 | 1.65 | 0.97, 2.79 | 51 | 1.31 | 0.83, 2.05 | 26 | 2.5 | 1.22, 5.11 | |
| >6.3 servings/mo | 227 | 206 | 2.45 | 1.80, 3.34 | 44 | 2.03 | 1.20, 3.44 | 66 | 1.84 | 1.18, 2.85 | 43 | 4.64 | 2.34, 9.20 | |
| P-trend | <0.0001 | 0.0056 | 0.012 | <0.0001 | 0.048 | |||||||||
| Butter | ||||||||||||||
| None | 418 | 207 | 1.00 | 42 | 1.00 | 87 | 1.00 | 35 | 1.00 | |||||
| ≤1.8 servings/mo | 103 | 70 | 1.39 | 0.98, 1.98 | 27 | 2.58 | 1.52, 4.40 | 21 | 0.99 | 0.59, 1.68 | 9 | 1.07 | 0.50, 2.30 | |
| 1.9–15.0 servings/mo | 243 | 163 | 1.34 | 1.03, 1.74 | 46 | 1.88 | 1.20, 2.95 | 54 | 1.05 | 0.72, 1.53 | 28 | 1.36 | 0.81, 2.30 | |
| >15.0 servings/mo | 243 | 162 | 1.29 | 0.99, 1.69 | 31 | 1.25 | 0.75, 2.06 | 56 | 1.05 | 0.71, 1.54 | 33 | 1.60 | 0.95, 2.69 | |
| P-trend | 0.034 | 0.20 | 0.77 | 0.06 | 0.37 | |||||||||
| Margarine | ||||||||||||||
| Never | 245 | 446 | 1.00 | 66 | 1.00 | 86 | 1.00 | 41 | 1.00 | |||||
| ≤1.8 servings/mo | 35 | 67 | 0.96 | 0.62, 1.49 | 10 | 1 | 0.49, 2.04 | 11 | 0.86 | 0.43, 1.70 | 3 | 0.49 | 0.15, 1.65 | |
| 1.9–17.1 servings/mo | 178 | 254 | 1.35 | 1.05, 1.73 | 36 | 0.99 | 0.64, 1.53 | 62 | 1.35 | 0.94, 1.94 | 43 | 1.97 | 1.24, 3.11 | |
| >17.1 servings/mo | 144 | 240 | 1.12 | 0.86, 1.47 | 34 | 0.99 | 0.63, 1.55 | 59 | 1.34 | 0.92, 1.95 | 18 | 0.84 | 0.47, 1.50 | |
| P-trend | 0.11 | 0.94 | 0.063 | 0.42 | 0.60 | |||||||||
| Cream in coffee | ||||||||||||||
| Never | 509 | 858 | 1.00 | 115 | 1.00 | 191 | 1.00 | 89 | 1.00 | |||||
| Ever | 93 | 149 | 1.02 | 0.77, 1.36 | 31 | 1.56 | 1.01, 2.41 | 27 | 0.78 | 0.50, 1.21 | 16 | 1.00 | 0.57, 1.76 | |
| P-trend | 0.89 | 0.046 | 0.27 | 1.00 | 0.10 | |||||||||
Unconditional logistic regression multivariable model was adjusted for total energy and the following design variables: age, sex, and residence. CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL, diffuse large B-cell lymphoma; NHL, non-Hodgkin lymphoma.
Serving sizes for each food product are defined in Supplemental Table 2.
TABLE 4.
Meat and egg intake in association with risk of NHL and NHL subtypes: Mayo Clinic case-control study1
| NHL overall (n = 603) |
Follicular (n = 146) |
CLL/SLL (n = 218) |
DLBCL (n = 105) |
|||||||||||
| Food2 | Controls | Cases | OR | 95% CI | Cases | OR | 95% CI | Cases | OR | 95% CI | Cases | OR | 95% CI | P-heterogeneity |
| n | n | n | n | n | ||||||||||
| Total meat | ||||||||||||||
| ≤24.0 servings/mo | 252 | 150 | 1.00 | 39 | 1.00 | 52 | 1.00 | 34 | 1.00 | |||||
| 24.1–39.6 servings/mo | 254 | 154 | 0.96 | 0.71, 1.30 | 41 | 1.10 | 0.67, 1.80 | 53 | 0.97 | 0.62, 1.52 | 26 | 0.66 | 0.37, 1.17 | |
| 39.7–59.4 servings/mo | 250 | 153 | 0.92 | 0.66, 1.27 | 37 | 0.95 | 0.55, 1.65 | 61 | 1.10 | 0.69, 1.75 | 22 | 0.52 | 0.27, 0.98 | |
| >59.4 servings/mo | 251 | 145 | 0.82 | 0.56, 1.18 | 29 | 0.69 | 0.36, 1.33 | 52 | 0.87 | 0.51, 1.50 | 23 | 0.47 | 0.23, 0.96 | |
| P-trend | 0.28 | 0.27 | 0.77 | 0.027 | 0.13 | |||||||||
| Red meat | ||||||||||||||
| ≤19.5 servings/mo | 147 | 252 | 1.00 | 38 | 1.00 | 54 | 1.00 | 30 | 1.00 | |||||
| 19.6–32.4 servings/mo | 145 | 252 | 0.98 | 0.72, 1.32 | 38 | 1.07 | 0.65, 1.77 | 51 | 0.94 | 0.61, 1.46 | 23 | 0.75 | 0.41, 1.35 | |
| 32.4–50.1 servings/mo | 146 | 253 | 0.96 | 0.70, 1.33 | 41 | 1.14 | 0.67, 1.94 | 51 | 0.93 | 0.58, 1.49 | 23 | 0.74 | 0.39, 1.39 | |
| >50.1 servings/mo | 164 | 250 | 1.07 | 0.75, 1.53 | 29 | 0.75 | 0.40, 1.43 | 62 | 1.12 | 0.67, 1.88 | 29 | 0.92 | 0.46, 1.83 | |
| P-trend | 0.77 | 0.55 | 0.72 | 0.76 | 0.57 | |||||||||
| Hamburger | ||||||||||||||
| ≤1.8 servings/mo | 427 | 220 | 1.00 | 58 | 1.00 | 76 | 1.00 | 43 | 1.00 | |||||
| 1.9–2.7 servings/mo | 80 | 50 | 1.24 | 0.83, 1.83 | 11 | 1.04 | 0.52, 2.08 | 15 | 1.08 | 0.59, 1.98 | 9 | 1.13 | 0.53, 2.41 | |
| 2.8–6.3 servings/mo | 283 | 188 | 1.28 | 0.99, 1.66 | 49 | 1.33 | 0.87, 2.04 | 68 | 1.37 | 0.94, 2.00 | 28 | 0.97 | 0.58, 1.62 | |
| >6.3 servings/mo | 217 | 144 | 1.26 | 0.95, 1.68 | 28 | 0.99 | 0.59, 1.65 | 59 | 1.53 | 1.02, 2.30 | 25 | 1.10 | 0.63, 1.92 | |
| P-trend | 0.06 | 0.63 | 0.027 | 0.83 | 0.56 | |||||||||
| Processed meat | ||||||||||||||
| ≤0.9 servings/mo | 323 | 169 | 1.00 | 42 | 1.00 | 65 | 1.00 | 27 | 1.00 | |||||
| 1.0–2.1 servings/mo | 183 | 101 | 1.05 | 0.77, 1.43 | 30 | 1.26 | 0.76, 2.08 | 30 | 0.81 | 0.51, 1.30 | 16 | 1.04 | 0.55, 1.99 | |
| 2.2–6.0 servings/mo | 275 | 173 | 1.21 | 0.92, 1.59 | 34 | 1.00 | 0.61, 1.63 | 63 | 1.15 | 0.78, 1.70 | 36 | 1.61 | 0.94, 2.76 | |
| >6.0 servings/mo | 226 | 159 | 1.37 | 1.02, 1.83 | 40 | 1.47 | 0.89, 2.42 | 60 | 1.35 | 0.89, 2.05 | 26 | 1.45 | 0.80, 2.64 | |
| P-trend | 0.030 | 0.24 | 0.12 | 0.11 | 0.92 | |||||||||
| Fried chicken | ||||||||||||||
| Never | 223 | 411 | 1.00 | 58 | 1.00 | 80 | 1.00 | 40 | 1.00 | |||||
| ≤1.8 servings/mo | 218 | 324 | 1.26 | 0.99, 1.61 | 53 | 1.16 | 0.78, 1.74 | 74 | 1.21 | 0.85, 1.72 | 35 | 1.13 | 0.70, 1.83 | |
| 1.9–2.1 servings/mo | 13 | 22 | 1.03 | 0.51, 2.10 | 4 | 1.2 | 0.40, 3.61 | 4 | 0.87 | 0.29, 2.62 | 1 | 0.44 | 0.06, 3.34 | |
| >2.1 servings/mo | 148 | 250 | 1.04 | 0.79, 1.37 | 31 | 0.88 | 0.54, 1.43 | 60 | 1.19 | 0.81, 1.75 | 29 | 1.17 | 0.69, 1.99 | |
| P-trend | 0.86 | 0.62 | 0.44 | 0.64 | 0.69 | |||||||||
| Other chicken | ||||||||||||||
| ≤1.8 servings/mo | 224 | 339 | 1.00 | 54 | 1.00 | 82 | 1.00 | 44 | 1.00 | |||||
| 1.9–4.2 servings/mo | 166 | 297 | 0.84 | 0.65, 1.08 | 40 | 0.88 | 0.56, 1.37 | 58 | 0.8 | 0.55, 1.17 | 29 | 0.74 | 0.45, 1.23 | |
| 4.3–12.9 servings/mo | 165 | 291 | 0.83 | 0.64, 1.08 | 39 | 0.85 | 0.54, 1.33 | 61 | 0.84 | 0.58, 1.22 | 24 | 0.61 | 0.36, 1.04 | |
| >12.9 servings/mo | 47 | 80 | 0.85 | 0.56, 1.30 | 13 | 1.12 | 0.56, 2.23 | 17 | 0.84 | 0.46, 1.54 | 8 | 0.74 | 0.32, 1.69 | |
| P-trend | 0.19 | 0.80 | 0.38 | 0.11 | 0.70 | |||||||||
| Total seafood | ||||||||||||||
| ≤2.0 servings/mo | 268 | 174 | 1.00 | 41 | 1.00 | 63 | 1.00 | 34 | 1.00 | |||||
| 2.1–4.5 servings/mo | 239 | 165 | 1.05 | 0.79, 1.39 | 45 | 1.25 | 0.79, 1.98 | 55 | 0.97 | 0.65, 1.45 | 32 | 1.04 | 0.62, 1.74 | |
| 4.5–8.4 servings/mo | 256 | 152 | 0.88 | 0.66, 1.18 | 34 | 0.88 | 0.54, 1.45 | 55 | 0.90 | 0.59, 1.35 | 29 | 0.85 | 0.50, 1.46 | |
| >8.4 servings/mo | 244 | 111 | 0.63 | 0.46, 0.85 | 26 | 0.67 | 0.39, 1.16 | 45 | 0.71 | 0.46, 1.10 | 10 | 0.28 | 0.13, 0.60 | |
| P-trend | 0.0028 | 0.097 | 0.12 | 0.0017 | 0.28 | |||||||||
| Fresh fish | ||||||||||||||
| None | 439 | 291 | 1.00 | 71 | 1.00 | 107 | 1.00 | 57 | 1.00 | |||||
| ≤0.9 servings/mo | 77 | 48 | 0.97 | 0.66, 1.45 | 12 | 0.95 | 0.49, 1.84 | 15 | 0.83 | 0.46, 1.52 | 9 | 0.93 | 0.44, 1.98 | |
| 1.0–2.1 servings/mo | 248 | 157 | 0.94 | 0.73, 1.20 | 37 | 0.93 | 0.60, 1.43 | 54 | 0.88 | 0.61, 1.27 | 25 | 0.76 | 0.46, 1.26 | |
| >2.1 servings/mo | 243 | 106 | 0.61 | 0.46, 0.80 | 26 | 0.64 | 0.40, 1.04 | 42 | 0.66 | 0.44, 0.98 | 14 | 0.41 | 0.22, 0.76 | |
| P-trend | 0.0026 | 0.11 | 0.051 | 0.0054 | 0.45 | |||||||||
| Fried fish | ||||||||||||||
| Never | 287 | 473 | 1.00 | 74 | 1.00 | 100 | 1.00 | 54 | 1.00 | |||||
| ≤0.9 servings/mo | 60 | 94 | 1.06 | 0.74, 1.52 | 13 | 0.86 | 0.46, 1.62 | 22 | 1.11 | 0.66, 1.87 | 12 | 1.12 | 0.57, 2.18 | |
| 1.0–1.8 servings/mo | 155 | 267 | 0.93 | 0.72, 1.19 | 41 | 0.99 | 0.65, 1.50 | 53 | 0.91 | 0.63, 1.32 | 24 | 0.76 | 0.46, 1.27 | |
| >1.8 servings/mo | 100 | 173 | 0.9 | 0.67, 1.21 | 18 | 0.67 | 0.38, 1.15 | 43 | 1.11 | 0.74, 1.66 | 15 | 0.71 | 0.39, 1.31 | |
| P-trend | 0.42 | 0.27 | 0.87 | 0.19 | 0.52 | |||||||||
| Canned fish | ||||||||||||||
| Never | 223 | 381 | 1.00 | 54 | 1.00 | 77 | 1.00 | 42 | 1.00 | |||||
| ≤0.9 servings/mo | 85 | 156 | 0.98 | 0.72, 1.35 | 18 | 0.83 | 0.47, 1.46 | 33 | 1.12 | 0.71, 1.76 | 17 | 1.04 | 0.57, 1.90 | |
| 1.0–1.8 servings/mo | 170 | 239 | 1.28 | 0.98, 1.66 | 41 | 1.29 | 0.82, 2.01 | 62 | 1.38 | 0.94, 2.01 | 30 | 1.2 | 0.73, 1.99 | |
| >1.8 servings/mo | 124 | 231 | 0.93 | 0.70, 1.24 | 33 | 1.07 | 0.66, 1.72 | 46 | 1.01 | 0.67, 1.53 | 16 | 0.64 | 0.35, 1.18 | |
| P-trend | 0.79 | 0.50 | 0.56 | 0.36 | 0.63 | |||||||||
| Eggs | ||||||||||||||
| ≤3.6 servings/mo | 423 | 238 | 1.00 | 70 | 1.00 | 79 | 1.00 | 35 | 1.00 | |||||
| 3.7–8.4 servings/mo | 255 | 153 | 1.06 | 0.81, 1.37 | 42 | 1.01 | 0.66, 1.54 | 59 | 1.24 | 0.85, 1.81 | 24 | 1.16 | 0.67, 2.01 | |
| 8.5–25.8 servings/mo | 245 | 163 | 1.16 | 0.89, 1.50 | 29 | 0.72 | 0.45, 1.15 | 63 | 1.36 | 0.93, 1.98 | 34 | 1.70 | 1.02, 2.84 | |
| >25.8 servings/mo | 84 | 48 | 0.94 | 0.63, 1.41 | 5 | 0.34 | 0.13, 0.88 | 17 | 1.01 | 0.56, 1.83 | 12 | 1.71 | 0.83, 3.53 | |
| P-trend | 0.61 | 0.021 | 0.32 | 0.031 | 0.0073 | |||||||||
Unconditional logistic regression multivariable model was adjusted for total energy and the following design variables: age, sex, and residence. CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL, diffuse large B-cell lymphoma; NHL, non-Hodgkin lymphoma
Servings sizes for each food product are defined in Supplemental Table 2.
Whereas total meat consumption was not significantly associated with risk of overall NHL, processed meat was associated with increased risk of NHL. Total seafood consumption or fresh fish consumption alone were both inversely associated with NHL risk. There was no association between egg intake and overall NHL risk, although there was evidence for heterogeneity (P-heterogeneity = 0.0073). Adjustment for BMI did not substantially attenuate any of the meat and NHL associations (data not shown).
Discussion
In this clinic-based study in 603 NHL cases and 1007 controls, protein and PUFA intakes were inversely associated with NHL, with n3 PUFA intake having the strongest inverse association with NHL risk, even after adjustment for total protein and other fatty acids. TFA consumption was strongly positively associated with NHL, especially for CLL/SLL and DLBCL. Processed meat intake was positively associated with NHL, whereas fresh fish and total sea food intakes were inversely associated with NHL risk. Several dairy products were positively associated with NHL risk, most notably high-fat ice cream and any milk containing fat.
Only 2 prior studies to our knowledge have assessed the risk of NHL and TFA intake (7, 23). Although these studies had the advantages of a prospective cohort study design, smaller numbers of cases limited subtype-specific analyses (7, 23). As in the study by Zhang et al. (7), we found a significant positive association between TFA intake and NHL. Laake et al. (23) assessed specific types of TFAs and found that NHL risk was inversely associated with partially hydrogenated vegetable oils but positively associated with partially hydrogenated fish oils and ruminant TFAs. We found the strongest association with TFA intake for CLL/SLL and DLBCL. One mechanism by which TFA consumption might predispose individuals to NHL is through its proinflammatory actions. A recent intervention trial of consumption of industrially produced TFAs found that industrially produced TFA intake for 16 wk led to significantly increased systemic concentrations of the proinflammatory cytokine tumor necrosis factor α (TNF-α) as well as of soluble tumor necrosis factor receptors 1 and 2 (sTNF-R1 and sTNF-R2) (24). TFA intake has significantly decreased in the United States as a result of efforts to increase awareness of its health effects and FDA labeling requirements for TFA content (25), and our findings provide further support for these efforts.
The n3 Fatty acids are believed to have numerous health benefits (26). Previously, Chang et al. (9) reported an association between n3 fatty acid consumption and reduced risk of NHL in a population-based case-control study, particularly of CLL. Our findings support this prior study, although we observed protective effects for CLL, FL, and most strongly for DLBCL. In addition, we found an inverse association between n6 fatty acid consumption and NHL risk, although this association was attenuated after adjustment for other fatty acids and total protein. The n6 fatty acid association is supported by at least 1 prior study that found an inverse association between n6 fatty acid consumption and NHL, especially FL and DLBCL, in a hospital-based case-control study (11). In addition, 2 case-control studies found an inverse association between overall PUFA intake and NHL risk (8, 11). However, other prospective cohort and case-control studies did not observe an association of n3 fatty acid (7, 11), n6 fatty acid (5, 7), or overall PUFA (4, 6) consumption with NHL risk. One proposed mechanism by which n3 fatty acids may be protective against NHL risk is their known antiinflammatory properties (27). In agreement with several prior NHL case-control and cohort studies, we did not find an association between NHL risk and total fat (4, 5, 7–11, 28), animal fat (7, 9, 11), or SFA (9, 11) intake, although a few studies have reported significant positive associations between NHL risk and intake of total fat (6), animal fat (4), and SFAs (4–8).
Sufficient intakes of proteins and different amino acids are necessary for a properly functioning immune system; however, excess intake can impair function as well (29). For example, the incidence of lymphoma in rats fed high-casein (milk protein) diets was elevated compared with that in control-fed rats in 1 study (30). Several case-control and cohort studies also assessed protein intake in relation to NHL risk and found either no association or increased risk with animal and/or total protein (4, 6–8), although 1 study reported a decreased risk associated with protein, particularly animal protein (5). However, animal protein is derived from a broad category of foods with different nutritional compositions, which could explain the inconsistent findings.
We found no association with overall dairy consumption, but we did observe a positive association between consumption of milk containing fat and NHL risk, whereas skim milk trended inversely with NHL risk and was significantly inversely associated with CLL/SLL. We also found a highly significant trend with greater ice cream consumption and NHL risk, even after adjustment for BMI and total energy intake. Importantly, this risk seemed to be specific to those who tended to consume high-fat ice cream. Whereas most case-control and cohort studies assessing milk intake and NHL have not found an association, a few studies reported milk intake to be associated with increased risk of NHL, particularly of DLBCL (8, 31–34). One population-based case-control study found that milk consumption was linked to the translocation t(14, 18) seen in most FL and some DLBCL cases (35), although we observed no risk association for FL. In addition, 1 prior case-control study reported a significant positive association between ice cream consumption and NHL risk (8). It is complicated to link dairy products to health outcomes because they may contain a heterogeneous mix of potentially healthful and harmful components: milk protein (casein) (30), SFAs (4–8), TFAs (until recently), hormones (36), calcium (9), and fat-soluble pesticides and dioxins may be prolymphomagenic (37, 38), whereas vitamin D (11, 39) and conjugated linoleic acid found in dairy foods are thought to have anticancer effects (40).
We also found a significant inverse association between fish and total seafood intake and NHL. This finding is consistent with the n3 fatty acid findings in this study, because this fatty acid is found in abundance in fish, and is also consistent with at least 2 other case-control studies in which total fish, fresh fish, and/or fatty fish intake was significantly inversely associated with all NHLs, and specifically DLBCL and CLL (9, 10). One case-control and 2 prospective cohort studies found no association with any fish intake (4, 8, 41), another case-control study found no association with total seafood intake (31), whereas 1 case-control study found only a suggestive inverse association with fish intake and NHL (5).
There are several limitations to this study. First, differential recall among cases and controls is a concern, although cases were asked to report their usual diet 1 y before diagnosis. Cases may also be more likely to overreport foods considered unhealthy and underreport foods considered to be healthy by the general population in an attempt to explain their illness. FFQs are also subject to measurement error, although our FFQ was fairly comprehensive. The associations reported may also be confounded by other lifestyle factors, although this study did adjust for a number of potential confounders, including BMI and physical activity, which did not alter the results substantially. However, residual and unmeasured confounding (e.g., pesticide exposure) remains a limitation. Our analysis also did not specifically account for replacement foods and macronutrients, such as fruit and vegetables eaten to achieve a calorie intake similar to those who ate high amounts of meats or dairy instead, which could affect the interpretation of the associations reported here. Multiple nutrients and dietary factors were assessed, and we did not account for multiple testing, which means that some of these associations may be due to chance. The cases and controls also came from a limited geographic area and mostly consisted of non-Hispanic white participants, which helped to decrease potential confounding but also reduced the generalizability of these results. This study lacked power to detect associations between dietary factors and rare NHL subtypes.
The strengths of this study included the following: reasonably high response rates, which should limit some biases (e.g., response bias); selection of cases and controls from the same source population (13); and central review of all pathology, along with coding into the WHO system. The FFQ is a well-established and practical approach for assessing usual diet in a large population (42, 43). Moreover, the comprehensive data collected in this study allowed for the adjustment for total energy and other confounders. Adjustment for total energy allowed us to more directly compare 1 food or nutrient with another, regardless of calorie density, and allowed us to test for associations with noncaloric factors in foods that may be driving associations. This study had a sufficient sample size to assess subtype-specific associations for the more common subtypes.
In conclusion, the consumption of proteins and fats appears to influence risk of NHL, and only a few associations varied by subtype. We found a strong positive association between intake of TFAs and NHL risk. Total protein intake was inversely associated with risk of overall NHL. We also saw relatively strong inverse associations with NHL for fish and total seafood and more modest associations for low-fat dairy intakes, potential protein sources. A strong inverse association was observed between n3 fatty acid consumption and risk of NHL, which was similar to that seen for fish and total seafood, a source of this PUFA. Certain dairy products were also associated with increased risk of NHL, but this appeared to be driven by high-fat dairy.
Supplementary Material
Acknowledgments
B.C. and J.R.C. designed the study and wrote the manuscript; M.L., C.A.T., W.R.M., T.G.C., and T.M.H. were responsible for the provision of study materials or patients; H.M.O. and J.R.C. collected and assembled the data; A.H.W., Z.S.F., and S.L.S. analyzed the data; and B.C., H.M.O., Z.S.F., S.L.S., and J.R.C. were responsible for interpreting the data. All authors read and approved the final manuscript.
Footnotes
Abbreviations used: CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; NHL, non-Hodgkin lymphoma; sTNF-R1, soluble tumor necrosis factor receptor 1; sTNF-R2, soluble tumor necrosis factor receptor 2; TFA, trans fatty acid; TNF-α, tumor necrosis factor α.
Literature Cited
- 1.Smedby KE, Baecklund E, Askling J. Malignant lymphomas in autoimmunity and inflammation: a review of risks, risk factors, and lymphoma characteristics. Cancer Epidemiol Biomarkers Prev. 2006;15:2069–77 [DOI] [PubMed] [Google Scholar]
- 2.Larsson SC, Wolk A. Body mass index and risk of non-Hodgkin's and Hodgkin's lymphoma: a meta-analysis of prospective studies. Eur J Cancer. 2011;47:2422–30 [DOI] [PubMed] [Google Scholar]
- 3.Guthrie N, Carroll KK. Specific versus non-specific effects of dietary fat on carcinogenesis. Prog Lipid Res. 1999;38:261–71 [DOI] [PubMed] [Google Scholar]
- 4.Chiu BC, Cerhan JR, Folsom AR, Sellers TA, Kushi LH, Wallace RB, Zheng W, Potter JD. Diet and risk of non-Hodgkin lymphoma in older women. JAMA. 1996;275:1315–21 [DOI] [PubMed] [Google Scholar]
- 5.Cross AJ, Ward MH, Schenk M, Kulldorff M, Cozen W, Davis S, Colt JS, Hartge P, Cerhan JR, Sinha R. Meat and meat-mutagen intake and risk of non-Hodgkin lymphoma: results from a NCI-SEER case-control study. Carcinogenesis. 2006;27:293–7 [DOI] [PubMed] [Google Scholar]
- 6.Purdue MP, Bassani DG, Klar NS, Sloan M, Kreiger N. Dietary factors and risk of non-Hodgkin lymphoma by histologic subtype: a case-control analysis. Cancer Epidemiol Biomarkers Prev. 2004;13:1665–76 [PubMed] [Google Scholar]
- 7.Zhang S, Hunter DJ, Rosner BA, Colditz GA, Fuchs CS, Speizer FE, Willett WC. Dietary fat and protein in relation to risk of non-Hodgkin's lymphoma among women. J Natl Cancer Inst. 1999;91:1751–8 [DOI] [PubMed] [Google Scholar]
- 8.Zheng T, Holford TR, Leaderer B, Zhang Y, Zahm SH, Flynn S, Tallini G, Zhang B, Zhou K, Owens PH, et al. Diet and nutrient intakes and risk of non-Hodgkin's lymphoma in Connecticut women. Am J Epidemiol. 2004;159:454–66 [DOI] [PubMed] [Google Scholar]
- 9.Chang ET, Balter KM, Torrang A, Smedby KE, Melbye M, Sundstrom C, Glimelius B, Adami HO. Nutrient intake and risk of non-Hodgkin's lymphoma. Am J Epidemiol. 2006;164:1222–32 [DOI] [PubMed] [Google Scholar]
- 10.Fritschi L, Ambrosini GL, Kliewer EV, Johnson KC. Dietary fish intake and risk of leukaemia, multiple myeloma, and non-Hodgkin lymphoma. Cancer Epidemiol Biomarkers Prev. 2004;13:532–7 [PubMed] [Google Scholar]
- 11.Polesel J, Talamini R, Montella M, Parpinel M, Dal Maso L, Crispo A, Crovatto M, Spina M, La Vecchia C, Franceschi S. Linoleic acid, vitamin D and other nutrient intakes in the risk of non-Hodgkin lymphoma: an Italian case-control study. Ann Oncol. 2006;17:713–8 [DOI] [PubMed] [Google Scholar]
- 12.Cotton PA, Subar AF, Friday JE, Cook A. Dietary sources of nutrients among US adults, 1994 to 1996. J Am Diet Assoc. 2004;104:921–30 [DOI] [PubMed] [Google Scholar]
- 13.Cerhan JR, Fredericksen ZS, Wang AH, Habermann TM, Kay NE, Macon WR, Cunningham JM, Shanafelt TD, Ansell SM, Call TG, et al. Design and validity of a clinic-based case-control study on the molecular epidemiology of lymphoma. Int J Mol Epidemiol Genet. 2011;2:95–113 [PMC free article] [PubMed] [Google Scholar]
- 14.Jaffe ES, Harris N, Stein H, Vardiman J. World Health Organization classification of tumours, pathology and genetics of tumours of hematopoietic and lymphoid tissues. Lyon (France): IARC Press; 2001.
- 15.Kelemen LE, Cerhan JR, Lim U, Davis S, Cozen W, Schenk M, Colt J, Hartge P, Ward MH. Vegetables, fruit, and antioxidant-related nutrients and risk of non-Hodgkin lymphoma: a National Cancer Institute-Surveillance, Epidemiology, and End Results population-based case-control study. Am J Clin Nutr. 2006;83:1401–10 [DOI] [PubMed] [Google Scholar]
- 16.Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol. 1986;124:453–69 [DOI] [PubMed] [Google Scholar]
- 17.Block G, Coyle LM, Hartman AM, Scoppa SM. Revision of dietary analysis software for the Health Habits and History Questionnaire. Am J Epidemiol. 1994;139:1190–6 [DOI] [PubMed] [Google Scholar]
- 18.Block G, Woods M, Potosky A, Clifford C. Validation of a self-administered diet history questionnaire using multiple diet records. J Clin Epidemiol. 1990;43:1327–35 [DOI] [PubMed] [Google Scholar]
- 19.Hosmer DW, Lemeshow S. Applied logistic regression. Hoboken (NJ): Wiley; 2000.
- 20.Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr. 1997;65(Suppl):1220S–8S; discussion 9S–31S. [DOI] [PubMed]
- 21.Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986;124:17–27 [DOI] [PubMed] [Google Scholar]
- 22.Holtan SG, O'Connor HM, Fredericksen ZS, Liebow M, Thompson CA, Macon WR, Micallef IN, Wang AH, Slager SL, Habermann TM, et al. Food-frequency questionnaire-based estimates of total antioxidant capacity and risk of non-Hodgkin lymphoma. Int J Cancer. 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Laake I, Carlsen MH, Pedersen JI, Weiderpass E, Selmer R, Kirkhus B, Thune I, Veierod MB. Intake of trans fatty acids from partially hydrogenated vegetable and fish oils and ruminant fat in relation to cancer risk. Int J Cancer. 2013 [DOI] [PubMed] [Google Scholar]
- 24.Bendsen NT, Stender S, Szecsi PB, Pedersen SB, Basu S, Hellgren LI, Newman JW, Larsen TM, Haugaard SB, Astrup A. Effect of industrially produced trans fat on markers of systemic inflammation: evidence from a randomized trial in women. J Lipid Res. 2011;52:1821–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Doell D, Folmer D, Lee H, Honigfort M, Carberry S. Updated estimate of trans fat intake by the US population. Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2012;29:861–74 [DOI] [PubMed] [Google Scholar]
- 26.Ruxton CH, Reed SC, Simpson MJ, Millington KJ. The health benefits of omega-3 polyunsaturated fatty acids: a review of the evidence. J Hum Nutr Diet. 2004;17:449–59 [DOI] [PubMed] [Google Scholar]
- 27.Wall R, Ross RP, Fitzgerald GF, Stanton C. Fatty acids from fish: the anti-inflammatory potential of long-chain omega-3 fatty acids. Nutr Rev. 2010;68:280–9 [DOI] [PubMed] [Google Scholar]
- 28.Tsai HT, Cross AJ, Graubard BI, Oken M, Schatzkin A, Caporaso NE. Dietary factors and risk of chronic lymphocytic leukemia and small lymphocytic lymphoma: a pooled analysis of two prospective studies. Cancer Epidemiol Biomarkers Prev. 2010;19:2680–4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Li P, Yin YL, Li D, Kim SW, Wu G. Amino acids and immune function. Br J Nutr. 2007;98:237–52 [DOI] [PubMed] [Google Scholar]
- 30.Ross MH, Bras G. Tumor incidence patterns and nutrition in the rat. J Nutr. 1965;87:245–60 [DOI] [PubMed] [Google Scholar]
- 31.Chang ET, Smedby KE, Zhang SM, Hjalgrim H, Melbye M, Ost A, Glimelius B, Wolk A, Adami HO. Dietary factors and risk of non-hodgkin lymphoma in men and women. Cancer Epidemiol Biomarkers Prev. 2005;14:512–20 [DOI] [PubMed] [Google Scholar]
- 32.Ursin G, Bjelke E, Heuch I, Vollset SE. Milk consumption and cancer incidence: a Norwegian prospective study. Br J Cancer. 1990;61:454–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Franceschi S, Serraino D, Carbone A, Talamini R, La Vecchia C. Dietary factors and non-Hodgkin's lymphoma: a case-control study in the northeastern part of Italy. Nutr Cancer. 1989;12:333–41 [DOI] [PubMed] [Google Scholar]
- 34.Tavani A, Pregnolato A, Negri E, Franceschi S, Serraino D, Carbone A, La Vecchia C. Diet and risk of lymphoid neoplasms and soft tissue sarcomas. Nutr Cancer. 1997;27:256–60 [DOI] [PubMed] [Google Scholar]
- 35.Chiu BC, Dave BJ, Ward MH, Fought AJ, Hou L, Jain S, Gapstur S, Evens AM, Zahm SH, Blair A, et al. Dietary factors and risk of t(14;18)-defined subgroups of non-Hodgkin lymphoma. Cancer Causes Control. 2008;19:859–67 [DOI] [PubMed] [Google Scholar]
- 36.Qin LQ, He K, Xu JY. Milk consumption and circulating insulin-like growth factor-I level: a systematic literature review. Int J Food Sci Nutr. 2009;60Suppl 7:330–40 [DOI] [PubMed] [Google Scholar]
- 37.Baccarelli A, Hirt C, Pesatori AC, Consonni D, Patterson DG, Jr, Bertazzi PA, Dolken G, Landi MT. t(14;18) Translocations in lymphocytes of healthy dioxin-exposed individuals from Seveso, Italy. Carcinogenesis. 2006;27:2001–7 [DOI] [PubMed] [Google Scholar]
- 38.Fritschi L, Benke G, Hughes AM, Kricker A, Turner J, Vajdic CM, Grulich A, Milliken S, Kaldor J, Armstrong BK. Occupational exposure to pesticides and risk of non-Hodgkin's lymphoma. Am J Epidemiol. 2005;162:849–57 [DOI] [PubMed] [Google Scholar]
- 39.Kelly JL, Drake MT, Fredericksen ZS, Asmann YW, Liebow M, Shanafelt TD, Feldman AL, Ansell SM, Macon WR, Herr MM, et al. Early life sun exposure, vitamin D-related gene variants, and risk of non-Hodgkin lymphoma. Cancer Causes Control. 2012;23:1017–29 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Field CJ, Schley PD. Evidence for potential mechanisms for the effect of conjugated linoleic acid on tumor metabolism and immune function: lessons from n–3 fatty acids. Am J Clin Nutr. 2004;79(Suppl):1190S–8S [DOI] [PubMed] [Google Scholar]
- 41.Daniel CR, Sinha R, Park Y, Graubard BI, Hollenbeck AR, Morton LM, Cross AJ. Meat intake is not associated with risk of non-Hodgkin lymphoma in a large prospective cohort of U.S. men and women. J Nutr. 2012;142:1074–80 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, Hennekens CH, Speizer FE. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122:51–65 [DOI] [PubMed] [Google Scholar]
- 43.Willett WC, Hu FB. The food frequency questionnaire. Cancer Epidemiol Biomarkers Prev. 2007;16:182–3 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.