Abstract
Background
Although risk factors for diffuse large B-cell lymphoma (DLBCL) have been suggested, their independent effects, modification by sex, and association with anatomical sites are largely unknown.
Methods
In a pooled analysis of 4667 cases and 22639 controls from 19 studies, we used stepwise logistic regression to identify the most parsimonious multivariate models for DLBCL overall, by sex, and for selected anatomical sites.
Results
DLBCL was associated with B-cell activating autoimmune diseases (odds ratio [OR] = 2.36, 95% confidence interval [CI] = 1.80 to 3.09), hepatitis C virus seropositivity (OR = 2.02, 95% CI = 1.47 to 2.76), family history of non-Hodgkin lymphoma (OR = 1.95, 95% CI = 1.54 to 2.47), higher young adult body mass index (OR = 1.58, 95% CI = 1.12 to 2.23, for 35+ vs 18.5 to 22.4 kg/m2), higher recreational sun exposure (OR = 0.78, 95% CI = 0.69 to 0.89), any atopic disorder (OR = 0.82, 95% CI = 0.76 to 0.89), and higher socioeconomic status (OR = 0.86, 95% CI = 0.79 to 0.94). Additional risk factors for women were occupation as field crop/vegetable farm worker (OR = 1.78, 95% CI = 1.22 to 2.60), hairdresser (OR = 1.65, 95% CI = 1.12 to 2.41), and seamstress/embroider (OR = 1.49, 95% CI = 1.13 to 1.97), low adult body mass index (OR = 0.46, 95% CI = 0.29 to 0.74, for <18.5 vs 18.5 to 22.4 kg/m2), hormone replacement therapy started age at least 50 years (OR = 0.68, 95% CI = 0.52 to 0.88), and oral contraceptive use before 1970 (OR = 0.78, 95% CI = 0.62 to 1.00); and for men were occupation as material handling equipment operator (OR = 1.58, 95% CI = 1.02 to 2.44), lifetime alcohol consumption (OR = 0.57, 95% CI = 0.44 to 0.75, for >400kg vs nondrinker), and previous blood transfusion (OR = 0.69, 95% CI = 0.57 to 0.83). Autoimmune disease, atopy, and family history of non-Hodgkin lymphoma showed similar associations across selected anatomical sites, whereas smoking was associated with central nervous system, testicular and cutaneous DLBCLs; inflammatory bowel disease was associated with gastrointestinal DLBCL; and farming and hair dye use were associated with mediastinal DLBCL.
Conclusion
Our results support a complex and multifactorial etiology for DLBCL with some variation in risk observed by sex and anatomical site.
Diffuse large B-cell lymphoma (DLBCL) is the most common non-Hodgkin lymphoma (NHL) subtype in Western countries (1,2). In Europe and the United States, the age-adjusted incidence rates range from 3.1 to 5.7 per 100000, with a median age at diagnosis in the sixth decade (1,2). DLBCLs are heterogeneous in histology, immunophenotype, and site of presentation (3,4), mostly presenting in the lymph node although extranodal presentations are increasingly recognized (5). Several anatomical sites appear to have distinct epidemiological or clinical characteristics (6–10), but they have generally been too rare to evaluate in individual studies. DLBCLs have also been categorized based on cell of origin (activated B cell, germinal center, and other), which have distinct biological and clinical characteristics (11), although the epidemiological implications of this classification is unknown.
Although medical history, lifestyle, and other risk factors for DLBCL have been published in previous pooled International Lymphoma Epidemiology Consortium (InterLymph) analyses (12–21), there has been no multivariate assessment of factors simultaneously, and limited assessment of risk factors stratified by sex and age or for specific anatomical sites. To advance our understanding of the etiology of DLBCL, we investigated these issues in the most complete and comprehensive pooled analysis to date, which combined 4667 cases and 22639 controls from 19 case-control studies conducted in Europe, North America, and Australia as part of the InterLymph NHL Subtypes Project.
Methods
Detailed methodology, including inclusion and exclusion criteria and pathology review, is provided elsewhere in this issue (22). Contributing studies were approved by local ethics committees, and all participants provided informed consent before interview.
NHL Subtype Ascertainment and Harmonization
Cases were classified according to the World Health Organization classification (23,24) using guidelines from the InterLymph Pathology Working Group (25,26). For Working Formulation cases with no immunophenotyping data, only those from category G (diffuse large cell lymphoma) were assigned to DLBCL (overall reliability of 88.2%), and for those Working Formulation cases with B/T immunophenotyping data, both category G (overall reliability 92.3%) and H (large cell, immunoblastic, B cell; overall reliability 98.9%) were assigned to DLBCL (25). In most studies, primary site of lymphoma was recorded where known, irrespective of disease stage, following Surveillance, Epidemiology, and End Results coding rules (27); primary site of lymphoma was coded as missing when primary site could not be distinguished from biopsy site. We conducted exploratory analyses of risk factors for central nervous system (CNS), gastrointestinal (GI), testis, cutaneous and mediastinal DLBCL; these sites were selected a priori based on known distinct epidemiological or clinical characteristics (6–10).
Risk Factor Ascertainment and Harmonization
Risk factors selected for inclusion in this analysis were medical history, lifestyle, family history, and occupations where data were available from at least four studies. Centralized harmonization of de-identified individual-level data from each study center was a key element of the project. Each exposure variable was harmonized individually, and data were reviewed for consistency among related exposure variables. Details of the data harmonization rules are provided elsewhere in this issue (22).
Statistical Analysis
We first used unconditional logistic regression models to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for risk of DLBCL with each exposure variable, adjusted for age, sex, race/ethnicity, and study (“basic adjusted models”). The statistical significance of each relationship was evaluated by a likelihood ratio test, comparing models with and without the exposure variable of interest, with P values less than .05 identifying putatively influential factors. Individuals with missing data for the exposure variable of interest were excluded. To evaluate effect heterogeneity among the 19 studies, we performed a separate logistic regression within each study and then quantified the variability of the coefficients by the H statistic, adapting the definition by Higgins and Thompson to categorical variables (28).
To consider possible effect modification, we repeated the above logistic regression analyses stratified by sex and age (<60 and 60+ years). We tested for effect heterogeneity by calculating the Wald statistic for the interaction term between each exposure and stratification variable. We also evaluated heterogeneity using forest plots by study design characteristics. Finally, we conducted an exploratory analysis of selected anatomical sites, modeling risk of DLBCL at a specific anatomical site (excluding all other DLBCL cases), adjusted for age, sex, race/ethnicity, and study.
To build a multivariate model for DLBCL, we reviewed the basic adjusted models and then selected the best variable within a given class of related variables to move forward to stepwise regression. A variable was selected after considering the effect size (i.e., stronger OR), exposure prevalence (e.g., choosing either a summary variable or a more common exposure over a rarer one to maximize power), and P value. From this set of selected factors, we evaluated the correlations of these variables as well as the risk estimate for each factor in a series of models that adjusted for other risk factors individually as well as age, race/ethnicity, sex, and study. We then conducted a single logistic regression model including all the selected risk factors, this time including a separate missing category for each variable to ensure that the entire study population was included in the analysis (i.e., not dropped due to missing data). Finally, we conducted a forward stepwise logistic regression with all the selected risk factors, adjusting for age, sex, race/ethnicity, and study; the risk factors remaining in this model were considered to be independent of each other (“final stepwise model”). We conducted these analyses for the entire study population, then separately for men and women. Because of small sample sizes, we did not conduct stepwise logistic regression for DLBCL site-specific analyses. All analyses were conducted using SAS software, version 9.2 (SAS Institute, Inc, Cary, NC).
Results
The pooled study population included 4667 DLBCL cases and 22639 controls from 19 InterLymph studies; full characteristics are provided in Table 1. As has been previously noted for DLBCL (1,2), the percentage of men (55.2%) was higher than that for women (44.8%). Controls for most studies were chosen to approximate the age and sex distribution of all lymphoma cases, leading to some imbalance of the age and sex distribution for the controls used in this analysis. In a sensitivity analysis, we used a subset of controls that were frequency matched by age and sex to the DLBCL cases, which produced similar results to those obtained using the full set of controls (data not shown). Therefore, we retained the full set of controls for our main analyses to increase statistical power.
Table 1.
Characteristics of studies included in the diffuse large B-cell lymphoma analysis, InterLymph NHL Subtypes Project*
| Controls | Cases | Total | |
|---|---|---|---|
| No. (%) | No. (%) | ||
| Total | 22639 (82.9) | 4667 (17.1) | 27306 |
| Study | |||
| North America | |||
| British Columbia | 845 (3.7) | 218 (4.7) | 1063 |
| Iowa/Minnesota | 1245 (5.5) | 112 (2.4) | 1357 |
| Kansas | 948 (4.2) | 27 (0.6) | 975 |
| Los Angeles | 375 (1.7) | 151 (3.2) | 526 |
| Mayo Clinic | 1314 (5.8) | 210 (4.5) | 1524 |
| NCI-SEER | 1055 (4.7) | 413 (8.8) | 1468 |
| Nebraska (newer) | 533 (2.4) | 103 (2.2) | 636 |
| Nebraska (older) | 1432 (6.3) | 94 (2.0) | 1526 |
| UCSF1 | 2402 (10.6) | 509 (10.9) | 2911 |
| UCSF2 | 0 (0.0) | 0 (0.0) | 0 |
| University of Rochester | 139 (0.6) | 32 (0.7) | 171 |
| Yale | 717 (3.2) | 189 (4.0) | 906 |
| Europe | |||
| Engela | 722 (3.2) | 174 (3.7) | 896 |
| EpiLymph | 2460 (10.9) | 516 (11.1) | 2976 |
| Italy multicenter | 1771 (7.8) | 407 (8.7) | 2178 |
| Italy (Aviano-Milan) | 1157 (5.1) | 47 (1.0) | 1204 |
| Italy (Aviano-Naples) | 504 (2.2) | 112 (2.4) | 616 |
| SCALE | 3187 (14.1) | 796 (17.1) | 3983 |
| United Kingdom | 1139 (5.0) | 326 (7.0) | 1465 |
| Australia | |||
| New South Wales | 694 (3.1) | 231 (4.9) | 925 |
| Region | |||
| North America | 11005 (48.6) | 2058 (44.1) | 13063 |
| Northern Europe | 6542 (28.9) | 1624 (34.8) | 8166 |
| Southern Europe | 4398 (19.4) | 754 (16.2) | 5152 |
| Australia | 694 (3.1) | 231 (4.9) | 925 |
| Design | |||
| Population-based | 17389 (76.8) | 3799 (81.4) | 21188 |
| Hospital-based | 5250 (23.2) | 868 (18.6) | 6118 |
| Age | |||
| <30 | 1356 (6.0) | 210 (4.5) | 1566 |
| 30–39 | 2143 (9.5) | 452 (9.7) | 2595 |
| 40–49 | 3090 (13.6) | 609 (13.0) | 3699 |
| 50–59 | 4870 (21.5) | 1128 (24.2) | 5998 |
| 60–69 | 6277 (27.7) | 1351 (28.9) | 7628 |
| 70–79 | 4048 (17.9) | 816 (17.5) | 4864 |
| ≥80 | 839 (3.7) | 84 (1.8) | 923 |
| Missing | 16 (0.1) | 17 (0.4) | 33 |
| Sex | |||
| Men | 13228 (58.4) | 2578 (55.2) | 15806 |
| Women | 9411 (41.6) | 2089 (44.8) | 11500 |
| Race | |||
| White, non-Hispanic | 21145 (93.4) | 4217 (90.4) | 25362 |
| Black | 351 (1.6) | 74 (1.6) | 425 |
| Asian | 321 (1.4) | 111 (2.4) | 432 |
| Hispanic | 334 (1.5) | 101 (2.2) | 435 |
| Other/unknown/missing | 488 (2.2) | 164 (3.5) | 652 |
| Socioeconomic status | |||
| Low | 9266 (40.9) | 1948 (41.7) | 11214 |
| Medium | 6577 (29.1) | 1355 (29.0) | 7932 |
| High | 6386 (28.2) | 1296 (27.8) | 7682 |
| Other/missing | 410 (1.8) | 68 (1.5) | 478 |
| NHL subtype classification | |||
| World Health Organization | 3320 (71.1) | 3320 | |
| Working Formulation | 1347 (28.9) | 1347 | |
* NCI-SEER = National Cancer Institute–Surveillance, Epidemiology, and End Results; NHL = non-Hodgkin lymphoma; SCALE = Scandinavian Lymphoma Etiology Study UCSF = University of California, San Francisco.
Overall and Sex- and Age-Specific Associations From the Basic Adjusted Models
The basic adjusted model results are shown in Supplementary Table 1. The variables selected for incorporation into the final stepwise models (bolded in Supplementary Table 1) included socioeconomic status (SES); B/T-cell activating autoimmune diseases (selected from a group of 18 autoimmune variables, which also included positive associations for Sjögren syndrome, systemic lupus erythematosus, hemolytic anemia, celiac disease, and rheumatoid arthritis); history of any atopic disorder (selected from six atopy variables, which also included inverse associations with allergy, food allergy, asthma, and hay fever); hepatitis C virus (HCV) seropositivity; history of blood transfusion (selected from five transfusion variables); family history of NHL (selected from 15 family history variables, which also included positive associations with family history of any hematologic malignancy and Hodgkin lymphoma); usual and young adult body mass index (BMI) (selected from three anthropometric variables, which also included a positive association with weight); lifetime alcohol use (selected from 16 variables, the majority of which showed inverse associations by current status, type, intensity and duration of use); recreational sun exposure; and occupation as a cleaner/related worker, driver/material handling equipment operator, field crop and vegetable farmer, women’s hairdresser, medical worker, and seamstress/embroiderer. There were no associations with cigarette smoking (from seven variables), physical activity, personal hair dye use (from six variables), or ever having lived or worked on a farm.
We also conducted sex-stratified analyses (including a test for interaction) for all of the basic associations that were reported in Supplementary Table 1 to identify variables to include in the sex-specific stepwise regression models. For women, additional variables were pack-years of smoking, oral contraceptive (OC) use by year started, and age at start of hormone therapy (HT) use. For men, the only additional variable was occupation as a metal processer. Variables selected for the stepwise models with evidence for sex-specific heterogeneity included blood transfusion (P = .046), young adult BMI (P = .023), and lifetime alcohol consumption (P = .0029), all of which showed stronger associations for men than for women (Supplementary Table 1). There was no strong evidence of heterogeneity by age for variables in Supplementary Table 1 (data not shown).
Heterogeneity by Design Variables
In reviewing the H statistic from each of the basic adjusted models, only recreational sun exposure (H = 2.44, 95% CI = 1.11 to 5.34) and adult BMI (H = 2.42, 95% CI = 1.31 to 4.48) showed substantial heterogeneity. In further reviewing forest plots for these two variables as well as other variables selected for the final stepwise models, a majority showed no evidence for heterogeneity by study design (i.e., population-based versus hospital-/clinic-based), World Health Organization versus Working Formulation, location (North America, Northern Europe, Southern Europe, Australia), or specific study. However, there was some suggestion that the associations with adult BMI (Supplementary Figure 1), OC use before 1970 (Supplementary Figure 2), and seamstresses/embroiders (Supplementary Figure 3) were strongest in North American studies; autoimmune disease (Supplementary Figure 4) was strongest in population-based studies; and recreational sun exposure (Supplementary Figure 5) was strongest in Australia and Southern Europe.
Final Stepwise Models
The final stepwise model for all participants (Table 2) included high SES (OR = 0.86); B-cell (OR = 2.36) and both B-/T-cell (OR = 4.86) activating autoimmune diseases; any atopic disorder (OR = 0.82); HCV seropositivity (OR = 2.02); previous blood transfusion (OR = 0.81); family history of NHL (OR = 1.95); young adult BMI (OR = 1.58 for 30+ vs 18.5-22.4 kg/m2); usual adult BMI (OR = 0.58 for 15.0-18.4 vs 18.5-22.5 kg/m2); greater lifetime alcohol consumption (OR = 0.64 for >400kg vs nondrinker); higher recreational sun exposure (OR = 0.78); and occupation as field crop and vegetable farmer (OR = 1.49), seamstress/embroiderer (OR = 1.43), and driver/material handling equipment operator (OR = 1.47). These results were consistent in pairwise adjusted models of all exposure variables, as well as the model simultaneously adjusting for all exposures, further supporting that the effects of these variables were mutually independent.
Table 2.
Results of the final stepwise regression*
| All participants | Sex-specific models | |||||||
|---|---|---|---|---|---|---|---|---|
| Controls | Cases | Men | Women | |||||
| Variable | No. (%) | No. (%) | OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P |
| SES | ||||||||
| Low | 9266 (40.9) | 1948 (41.7) | 1.00 (Referent) | .002 | 1.00 (Referent) | .008 | 1.00 (Referent) | .125 |
| Medium | 6577 (29.1) | 1355 (29.0) | 0.88 (0.81 to 0.95) | — | 0.86 (0.77 to 0.96) | — | 0.92 (0.81 to 1.04) | — |
| High | 6386 (28.2) | 1296 (27.8) | 0.86 (0.79 to 0.94) | — | 0.84 (0.75 to 0.94) | — | 0.95 (0.83 to 1.08) | — |
| Other/missing | 410 (1.8) | 68 (1.5) | 0.92 (0.65 to 1.31) | — | 1.13 (0.73 to 1.77) | — | 0.50 (0.25 to 1.00) | — |
| History of autoimmune disease† | ||||||||
| No autoimmune disease | 19423 (95.9) | 4286 (94.3) | 1.00 (Referent) | <.001 | 1.00 (Referent) | .001 | 1.00 (Referent) | <.001 |
| B-cell activation | 157 (0.8) | 87 (1.9) | 2.36 (1.80 to 3.09) | — | 2.22 (1.36 to 3.61) | — | 2.42 (1.73 to 3.38) | — |
| T-cell activation | 664 (3.3) | 159 (3.5) | 1.03 (0.86 to 1.24) | — | 1.14 (0.89 to 1.45) | — | 0.90 (0.69 to 1.19) | — |
| Both | 15 (0.1) | 14 (0.3) | 4.86 (2.31 to 10.25) | — | 6.20 (1.35 to 28.43) | — | 4.47 (1.84 to 10.84) | — |
| Any atopic disorder‡ | ||||||||
| No | 15601 (68.9) | 3262 (69.9) | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 |
| Yes | 6442 (28.5) | 1315 (28.2) | 0.82 (0.76 to 0.89) | — | 0.84 (0.75 to 0.93) | — | 0.81 (0.72 to 0.91) | — |
| Missing | 596 (2.6) | 90 (1.9) | 1.30 (0.96 to 1.75) | — | 1.04 (0.69 to 1.56) | — | 1.48 (0.92 to 2.38) | — |
| HCV positivity | ||||||||
| No | 6746 (66.8) | 1591 (66.8) | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 | 1.00 (Referent) | .004 |
| Yes | 152 (1.5) | 63 (2.6) | 2.02 (1.47 to 2.76) | — | 2.17 (1.44 to 3.26) | — | 1.98 (1.18 to 3.34) | — |
| Missing | 3194 (31.6) | 728 (30.6) | 0.93 (0.83 to 1.03) | — | 0.81 (0.71 to 0.94) | — | 1.70 (1.40 to 2.08) | — |
| Blood transfusion | ||||||||
| No | 10742 (76.7) | 2654 (81.3) | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 | (Not selected) | — |
| Yes | 1966 (14.0) | 411 (12.6) | 0.81 (0.72 to 0.91) | — | 0.69 (0.57 to 0.83) | — | — | — |
| Missing | 1297 (9.3) | 199 (6.1) | 1.18 (0.93 to 1.49) | — | 1.33 (1.00 to 1.76) | — | — | — |
| Year of first OC use | ||||||||
| No OC use | 2817 (28.8) | 638 (32.7) | (Not eligible) | — | (Not eligible) | — | 1.00 (Referent) | .014 |
| <1970 | 765 (7.8) | 128 (6.6) | — | — | — | — | 0.78 (0.62 to 1.00) | — |
| ≥1970 | 757 (7.7) | 155 (7.9) | — | — | — | — | 1.06 (0.82 to 1.35) | — |
| OC use women, UK date | 19 (0.2) | 6 (0.3) | — | — | — | — | 1.46 (0.55 to 3.86) | — |
| Unknown OC use (women) | 139 (1.4) | 55 (2.8) | — | — | — | — | 2.29 (1.18 to 4.45) | — |
| Men | 5277 (54.0) | 970 (49.7) | — | — | — | — | — | — |
| Age at first HT use | ||||||||
| No HT use | 2399 (28.4) | 582 (33.0) | (Not eligible) | — | (Not eligible) | — | 1.00 (Referent) | .083 |
| <50 y | 584 (6.9) | 137 (7.8) | — | — | — | — | 0.86 (0.68 to 1.09) | — |
| ≥50 y | 569 (6.7) | 97 (5.5) | — | — | — | — | 0.68 (0.52 to 0.88) | — |
| HT use, age unknown | 38 (0.4) | 10 (0.6) | — | — | — | — | 1.05 (0.49 to 2.27) | — |
| Unknown HT use (women) | 279 (3.3) | 66 (3.7) | — | — | — | — | 0.38 (0.22 to 0.64) | — |
| Men | 4584 (54.2) | 870 (49.4) | — | — | — | — | — | — |
| First-degree family history, NHL | ||||||||
| No | 14116 (83.6) | 2758 (80.7) | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 |
| Yes | 278 (1.6) | 110 (3.2) | 1.95 (1.54 to 2.47) | — | 2.07 (1.46 to 2.92) | — | 1.80 (1.30 to 2.50) | — |
| Missing | 2485 (14.7) | 551 (16.1) | 0.86 (0.69 to 1.07) | — | 1.12 (0.80 to 1.57) | — | 0.73 (0.55 to 0.98) | — |
| BMI as a young adult (kg/m2) | ||||||||
| 15 to <18.5 | 384 (2.4) | 64 (1.8) | 0.93 (0.69 to 1.24) | .002 | 1.39 (0.86 to 2.26) | .025 | 0.75 (0.52 to 1.08) | .016 |
| 18.5 to <22.5 | 2804 (17.3) | 517 (14.2) | 1.00 (Referent) | — | 1.00 (Referent) | — | 1.00 (Referent) | — |
| 22.5 to <25 | 1394 (8.6) | 276 (7.6) | 1.11 (0.93 to 1.31) | — | 1.01 (0.80 to 1.27) | — | 1.44 (1.10 to 1.88) | — |
| 25 to <30 | 839 (5.2) | 226 (6.2) | 1.47 (1.22 to 1.77) | — | 1.52 (1.20 to 1.93) | — | 1.50 (1.05 to 2.14) | — |
| 30 to 50 | 173 (1.1) | 54 (1.5) | 1.58 (1.12 to 2.23) | — | 1.63 (1.04 to 2.55) | — | 1.54 (0.88 to 2.69) | — |
| Missing | 10580 (65.4) | 2508 (68.8) | 1.49 (1.24 to 1.79) | — | 1.30 (1.01 to 1.68) | — | 1.66 (1.23 to 2.25) | — |
| Usual adult BMI (kg/m2) | ||||||||
| 15 to <18.5 | 267 (1.6) | 33 (0.9) | 0.58 (0.39 to 0.85) | .042 | (Not selected) | — | 0.46 (0.29 to 0.74) | .007 |
| 18.5 to <22.5 | 3481 (20.3) | 722 (19.7) | 1.00 (Referent) | — | — | — | 1.00 (Referent) | — |
| 22.5 to <25 | 4276 (25.0) | 850 (23.1) | 0.91 (0.81 to 1.03) | — | — | — | 0.89 (0.76 to 1.04) | — |
| 25 to <30 | 6112 (35.7) | 1310 (35.7) | 0.93 (0.83 to 1.04) | — | — | — | 0.92 (0.79 to 1.08) | — |
| 30 to <35 | 1760 (10.3) | 419 (11.4) | 0.95 (0.82 to 1.10) | — | — | — | 0.89 (0.73 to 1.10) | — |
| 35 to 50 | 608 (3.6) | 175 (4.8) | 1.06 (0.86 to 1.30) | — | — | — | 1.14 (0.87 to 1.49) | — |
| Missing | 618 (3.6) | 163 (4.4) | 1.02 (0.81 to 1.28) | — | — | — | 0.90 (0.64 to 1.25) | — |
| Lifetime alcohol consumption | ||||||||
| Nondrinker | 4277 (21.7) | 975 (23.6) | 1.00 (Referent) | <.001 | 1.00 (Referent) | <.001 | (Not selected) | — |
| 1–100 kg | 1444 (7.3) | 305 (7.4) | 0.80 (0.68 to 0.95) | — | 0.69 (0.53 to 0.91) | — | — | — |
| 101–200 kg | 641 (3.3) | 134 (3.2) | 0.79 (0.63 to 0.98) | — | 0.76 (0.56 to 1.02) | — | — | — |
| 201–400 kg | 651 (3.3) | 121 (2.9) | 0.66 (0.53 to 0.83) | — | 0.64 (0.47 to 0.85) | — | — | — |
| >400 kg | 759 (3.9) | 137 (3.3) | 0.64 (0.51 to 0.79) | — | 0.57 (0.44 to 0.75) | — | — | — |
| Drinker, lifetime consumption unknown | 7499 (38.1) | 1362 (33.0) | 0.87 (0.77 to 0.97) | — | 0.75 (0.64 to 0.88) | — | — | — |
| Missing | 4397 (22.4) | 1090 (26.4) | 0.68 (0.59 to 0.80) | — | 0.52 (0.42 to 0.64) | — | — | — |
| Recreational sun exposure (h/week)§ | ||||||||
| Q1 (low) | 2234 (20.6) | 649 (22.7) | 1.00 (Referent) | <.001 | 1.00 (Referent) | .033 | 1.00 (Referent) | .020 |
| Q2 | 2332 (21.6) | 619 (21.6) | 0.90 (0.79 to 1.02) | — | 0.83 (0.68 to 1.00) | — | 0.99 (0.82 to 1.18) | — |
| Q3 | 2159 (20.0) | 512 (17.9) | 0.79 (0.69 to 0.90) | — | 0.77 (0.63 to 0.94) | — | 0.81 (0.67 to 0.99) | — |
| Q4 (high) | 2983 (27.6) | 714 (24.9) | 0.78 (0.69 to 0.89) | — | 0.78 (0.66 to 0.94) | — | 0.79 (0.65 to 0.95) | — |
| Missing | 1111 (10.3) | 369 (12.9) | 0.92 (0.74 to 1.13) | — | 1.07 (0.80 to 1.42) | — | 0.72 (0.49 to 1.05) | — |
| Field crop and vegetable farmer | ||||||||
| No | 10392 (94.8) | 2664 (96.3) | 1.00 (Referent) | .004 | (Not selected) | — | 1.00 (Referent) | .004 |
| Yes | 233 (2.1) | 79 (2.9) | 1.49 (1.14 to 1.95) | — | — | — | 1.78 (1.22 to 2.60) | — |
| Missing | 335 (3.1) | 22 (0.8) | 0.11 (0.02 to 0.56) | — | — | — | — | — |
| Sewer and embroiderer | ||||||||
| No | 11771 (95.4) | 2981 (96.6) | 1.00 (Referent) | .009 | (Not selected) | — | 1.00 (Referent) | .005 |
| Yes | 232 (1.9) | 83 (2.7) | 1.43 (1.10 to 1.87) | — | — | — | 1.49 (1.13 to 1.97) | — |
| Missing | 335 (2.7) | 22 (0.7) | 0.12 (0.02 to 0.62) | — | — | — | — | — |
| Women’s hairdresser | ||||||||
| No | 11357 (96.2) | 2915 (97.7) | 1.00 (Referent) | .011 | (Not selected) | — | 1.00 (Referent) | .013 |
| Yes | 113 (1.0) | 46 (1.5) | 1.61 (1.13 to 2.31) | — | — | — | 1.65 (1.12 to 2.41) | — |
| Missing | 335 (2.8) | 22 (0.7) | 0.13 (0.03 to 0.67) | — | — | — | — | — |
| Driver/material handling equipment operator | ||||||||
| No | 11925 (96.7) | 3031 (98.2) | 1.00 (Referent) | .080 | 1.00 (Referent) | .047 | (Not selected) | — |
| Yes | 78 (0.6) | 33 (1.1) | 1.47 (0.97 to 2.25) | — | 1.58 (1.02 to 2.44) | — | — | — |
| Missing | 335 (2.7) | 22 (0.7) | 0.12 (0.02 to 0.62) | — | 0.93 (0.15 to 5.68) | — | — | — |
* Odds ratio (OR) and 95% confidence interval (CI) adjusted for age, sex, race/ethncity, study, and other factors in the same column. BMI = body mass index; HCV = hepatitis c virus; HT = hormone therapy; NHL = non-Hodgkin lymphoma; OC = oral contraceptive; SES = socioeconomic status.
† Includes self-reported history of specific autoimmune diseases occurring ≥2 years before diagnosis/interview (except the New South Wales study, which did not ascertain date of onset). Autoimmune diseases were classified according to whether they are primarily mediated by B-cell or T-cell responses. B-cell activating diseases included Hashimoto thyroiditis, hemolytic anemia, myasthenia gravis, pernicious anemia, rheumatoid arthritis, Sjögren’s syndrome, and systemic lupus erythematosus. T-cell activating diseases included celiac disease, immune thrombocytopenic purpura, inflammatory bowel disorder (Crohn’s disease, ulcerative colitis), multiple sclerosis, polymyositis or dermatomyositis, psoriasis, sarcoidosis, systemic sclerosis or scleroderma, and type 1 diabetes.
‡ Includes self-reported history of atopic disorders including asthma, eczema, hay fever, or other allergies, excluding drug allergies, occurring ≥2 years before diagnosis/interview (except the New South Wales study, which did not ascertain date of onset).
§ Study specific quartiles among controls.
We next conducted sex-specific stepwise models (Table 2). For both sexes, SES, autoimmune disease B-/T-cell type, atopic disorder, HCV seropositivity, family history of NHL, young adult BMI, and recreational sun exposure were retained in the final models with similar ORs for men and women. For women, OC use before 1970 (OR = 0.78), HT use initiated at age at least 50 years (OR = 0.68), low usual adult BMI (OR = 0.46 for <18.5 vs 18.5-22.4 kg/m2), and occupation as field crop/vegetable farm worker (OR = 1.78), seamstress/embroiderer (OR = 1.49), and women’s hairdresser (OR = 1.65) were also retained in the final model; pack-years of smoking was not retained. For men, blood transfusion (OR = 0.69), lifetime alcohol consumption (OR = 0.57 for >400kg vs nondrinker), and occupation as a driver/material handling equipment operator (OR = 1.58) were retained in the final model; metal processer was not retained.
Site-Specific Results
We next evaluated how the variables used in the final model (Table 2), excluding the occupation and reproductive variables (due to low exposure prevalence), performed for selected anatomical sites that may have distinct epidemiological or clinical characteristics, including CNS (N = 103), GI (N = 323), testis (N = 44), cutaneous (N = 120), and primary mediastinal (N = 91) DLBCL (Table 3). In addition, we reported any additional variables from Supplementary Table 1 that were significant at P value less than .05 for a specific site. Due to small sample sizes, we treated these as exploratory analyses and did not conduct formal heterogeneity tests. Even in this exploratory setting, we observed some notable patterns. CNS DLBCL was inversely associated with atopic disorder (OR = 0.54) and OC use before 1970 (OR = 0.19) and was positively associated with family history of NHL (OR = 4.11) and pack-years of smoking (OR = 1.52 for >35 pack-years versus nonsmoker). GI tract DLBCL was positively associated with B-cell activating autoimmune diseases (OR = 2.78), young adult BMI (OR = 3.96 for 35-50 vs 18.5-22.5 kg/m2), and history of inflammatory bowel disease (OR = 2.70) and was inversely associated with atopic disorder (OR = 0.75) and greater recreational sun exposure (OR = 0.67 for highest versus lowest quartile). For testicular DLBCL, positive associations were observed for B-cell activating autoimmune diseases (OR = 5.96) and pack-years of smoking (OR = 2.72 for >35 pack-years versus nonsmoker). Cutaneous DLBCL was positively associated with B-cell activating autoimmune diseases (OR = 3.80), usual adult BMI (OR = 2.93 for 35-50 vs 18.5-22.5 kg/m2), and pack-years of smoking (OR = 2.47 for >35 pack-years vs nonsmoker). Mediastinal DLBCL was positively associated with family history of NHL (OR = 4.81), having lived on a farm (OR = 3.12), and hair dye use duration (OR = 4.97 for 20+ years versus never). Although none of the risk factors was unambiguously associated with all of these sites, there were consistent trends for family history for all sites, autoimmune disease for all sites except CNS, atopic disorders for all sites except cutaneous, young adult BMI for all sites except cutaneous.
Table 3.
Results for selected sites*
| CNS | Testis | GI | Cutaneous | Mediastinal | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Controls | Cases | Controls | Cases | Controls | Cases | Controls | Cases | Controls | Cases | ||||||||||||
| Variable | No. (%) | No. (%) | OR (95% CI) | P | No. (%) | No. (%) | OR (95% CI) | P | No. (%) | No. (%) | OR (95% CI) | P | No. (%) | No. (%) | OR (95% CI) | P | No. (%) | No. (%) | OR (95% CI) | P | |
| SES | |||||||||||||||||||||
| Low | 6135 (36.1) | 40 (38.8) | 1.00 (Referent) | .62 | 3606 (37.3) | 15 (34.1) | 1.00 (Referent) | .62 | 6190 (36.1) | 142 (44.0) | 1.00 (Referent) | .38 | 6135 (36.1) | 53 (44.2) | 1.00 (Referent) | .63 | 2690 (33.3) | 24 (26.4) | 1.00 (Referent) | .41 | |
| Medium | 5309 (31.2) | 33 (32.0) | 0.96 (0.59 to 1.54) | — | 2848 (29.4) | 14 (31.8) | 1.43 (0.68 to 3.04) | — | 5344 (31.2) | 88 (27.2) | 0.84 (0.64 to 1.11) | — | 5309 (31.2) | 32 (26.7) | 0.83 (0.53 to 1.32) | — | 2505 (31.0) | 28 (30.8) | 1.03 (0.59 to 1.80) | — | |
| High | 5321 (31.3) | 27 (26.2) | 0.78 (0.47 to 1.32) | — | 3205 (33.1) | 15 (34.1) | 1.30 (0.61 to 2.74) | — | 5365 (31.3) | 91 (28.2) | 0.85 (0.64 to 1.13) | — | 5321 (31.3) | 32 (26.7) | 0.81 (0.51 to 1.31) | — | 2870 (35.5) | 39 (42.9) | 1.36 (0.80 to 2.32) | — | |
| History of autoimmune disease† | |||||||||||||||||||||
| No autoimmune disease | 16250 (95.6) | 97 (94.2) | 1.00 (Referent) | .85 | 9276 (95.9) | 41 (93.2) | 1.00 (Referent) | .26 | 16370 (95.5) | 299 (92.6) | 1.00 (Referent) | .002 | 16250 (95.6) | 112 (93.3) | 1.00 (Referent) | .045 | 7754 (95.9) | 86 (94.5) | 1.00 (Referent) | .84 | |
| B-cell activation | 152 (0.9) | 1 (1.0) | 0.76 (0.10 to 5.55) | — | 93 (1.0) | 2 (4.5) | 5.96 (1.33 to 26.66) | — | 153 (0.9) | 7 (2.2) | 2.78 (1.28 to 6.06) | — | 152 (0.9) | 4 (3.3) | 3.80 (1.36 to 10.65) | — | 85 (1.1) | 2 (2.2) | 1.96 (0.45 to 8.53) | — | |
| T-cell activation | 584 (3.4) | 5 (4.9) | 1.36 (0.55 to 3.37) | — | 299 (3.1) | 1 (2.3) | 0.58 (0.07 to 4.52) | — | 602 (3.5) | 14 (4.3) | 1.18 (0.68 to 2.05) | — | 584 (3.4) | 3 (2.5) | 0.69 (0.22 to 2.20) | — | 243 (3.0) | 3 (3.3) | 1.20 (0.37 to 3.92) | — | |
| Both | 15 (0.1) | 0 (0.0) | — | — | 7 (0.1) | 0 (0.0) | — | — | 15 (0.1) | 3 (0.9) | 14.46 (3.91 to 53.51) | — | 15 (0.1) | 1 (0.8) | 11.99 (1.50 to 96.07) | — | 6 (0.1) | 0 (0.0) | — | — | |
| Any atopic disorder‡ | |||||||||||||||||||||
| No | 11181 (65.8) | 75 (72.8) | 1.00 (Referent) | .0080 | 6746 (69.7) | 35 (79.5) | 1.00 (Referent) | .15 | 11310 (66.0) | 235 (72.8) | 1.00 (Referent) | .029 | 11181 (65.8) | 79 (65.8) | 1.00 (Referent) | .30 | 5617 (69.4) | 55 (60.4) | 1.00 (Referent) | .60 | |
| Yes | 5573 (32.8) | 24 (23.3) | 0.54 (0.33 to 0.87) | — | 2878 (29.7) | 9 (20.5) | 0.58 (0.27 to 1.25) | — | 5583 (32.6) | 82 (25.4) | 0.75 (0.57 to 0.98) | — | 5573 (32.8) | 39 (32.5) | 1.24 (0.83 to 1.87) | — | 2428 (30.0) | 34 (37.4) | 0.89 (0.56 to 1.40) | — | |
| HCV positivity | |||||||||||||||||||||
| No | 6746 (66.8) | 38 (67.9) | 1.00 (Referent) | .65 | 2692 (68.3) | 12 (70.6) | 1.00 (Referent) | .62 | 6746 (66.8) | 121 (73.3) | 1.00 (Referent) | .11 | 6746 (66.8) | 28 (59.6) | 1.00 (Referent) | .19 | 1205 (78.3) | 17 (68.0) | 1.00 (Referent) | .76 | |
| Yes | 152 (1.5) | 1 (1.8) | 1.67 (0.21 to 12.98) | — | 64 (1.6) | 0 (0.0) | — | 152 (1.5) | 5 (3.0) | 2.38 (0.91 to 6.22) | — | 152 (1.5) | 2 (4.3) | 3.20 (0.71 to 14.48) | — | 7 (0.5) | 0 (0.0) | — | — | ||
| Blood transfusion | |||||||||||||||||||||
| No | 9899 (82.2) | 64 (80.0) | 1.00 (Referent) | .87 | 6482 (83.5) | 32 (91.4) | 1.00 (Referent) | .28 | 9899 (82.2) | 160 (80.0) | 1.00 (Referent) | .11 | 9899 (82.2) | 64 (86.5) | 1.00 (Referent) | .19 | 6620 (83.3) | 82 (91.1) | 1.00 (Referent) | .099 | |
| Yes | 1743 (14.5) | 13 (16.3) | 1.05 (0.57 to 1.94) | — | 1108 (14.3) | 3 (8.6) | 0.54 (0.16 to 1.80) | — | 1743 (14.5) | 37 (18.5) | 1.37 (0.94 to 1.98) | — | 1743 (14.5) | 8 (10.8) | 0.63 (0.30 to 1.32) | — | 1155 (14.5) | 6 (6.7) | 0.52 (0.22 to 1.21) | — | |
| Year of first OC use | |||||||||||||||||||||
| No OC use | 1853 (23.5) | 19 (39.6) | 1.00 (Referent) | .010 | — | — | — | — | 1853 (23.5) | 29 (26.1) | 1.00 (Referent) | .56 | 1853 (23.5) | 14 (33.3) | 1.00 (Referent) | .19 | 1098 (31.4) | 11 (44.0) | 1.00 (Referent) | .096 | |
| <1970 | 690 (8.7) | 3 (6.3) | 0.19 (0.05 to 0.68) | — | — | — | — | — | 690 (8.7) | 9 (8.1) | 0.82 (0.35 to 1.96) | — | 690 (8.7) | 4 (9.5) | 0.58 (0.16 to 2.04) | — | 402 (11.5) | 0 (0.0) | — | — | |
| ≥1970 | 632 (8.0) | 5 (10.4) | 0.67 (0.20 to 2.24) | — | — | — | — | — | 632 (8.0) | 4 (3.6) | 0.66 (0.19 to 2.24) | — | 632 (8.0) | 0 (0.0) | — | — | 236 (6.8) | 10 (40.0) | 1.92 (0.63 to 5.79) | — | |
| OC use women, UK date | 19 (0.2) | 1 (2.1) | 6.89 (0.80 to 59.52) | — | — | — | — | — | 19 (0.2) | 1 (0.9) | 4.49 (0.52 to 39.00) | — | 19 (0.2) | 0 (0.0) | — | — | 7 (0.2) | 0 (0.0) | — | — | |
| Unknown OC use (women) | 138 (1.7) | 1 (2.1) | — | — | — | — | — | — | 138 (1.7) | 0 (0.0) | — | — | 138 (1.7) | 2 (4.8) | — | — | 2 (0.1) | 0 (0.0) | — | — | |
| Men | 4568 (57.8) | 19 (39.6) | — | — | — | — | — | — | 4568 (57.8) | 68 (61.3) | — | — | 4568 (57.8) | 22 (52.4) | — | — | 1749 (50.1) | 4 (16.0) | — | — | |
| Age at first HT use | |||||||||||||||||||||
| No HT use | 1438 (21.9) | 17 (39.5) | 1.00 (Referent) | .16 | — | — | — | — | 1438 (21.9) | 27 (20.8) | 1.00 (Referent) | .71 | 1438 (21.9) | 9 (19.6) | 1.00 (Referent) | .66 | 1363 (29.4) | 30 (71.4) | 1.00 (Referent) | .31 | |
| <50 y | 488 (7.4) | 6 (14.0) | 0.78 (0.28 to 2.14) | — | — | — | — | — | 488 (7.4) | 11 (8.5) | 1.17 (0.55 to 2.49) | — | 488 (7.4) | 4 (8.7) | 1.00 (0.29 to 3.45) | — | 378 (8.2) | 1 (2.4) | 0.25 (0.03 to 1.90) | — | |
| ≥50 y | 463 (7.0) | 2 (4.7) | 0.25 (0.05 to 1.14) | — | — | — | — | — | 463 (7.0) | 9 (6.9) | 1.08 (0.48 to 2.44) | — | 463 (7.0) | 2 (4.3) | 0.50 (0.10 to 2.47) | — | 343 (7.4) | 1 (2.4) | 0.36 (0.04 to 2.89) | — | |
| HT use, age unknown | 37 (0.6) | 0 (0.0) | — | — | — | — | — | — | 37 (0.6) | 0 (0.0) | — | — | 37 (0.6) | 0 (0.0) | — | — | 29 (0.6) | 0 (0.0) | — | — | |
| Unknown HT use (women) | 278 (4.2) | 1 (2.3) | — | — | — | — | — | — | 278 (4.2) | 0 (0.0) | — | — | 278 (4.2) | 3 (6.5) | — | — | 146 (3.2) | 0 (0.0) | — | — | |
| Men | 3875 (58.9) | 17 (39.5) | — | — | — | — | — | — | 3875 (58.9) | 83 (63.8) | — | — | 3875 (58.9) | 28 (60.9) | — | — | 2374 (51.2) | 10 (23.8) | — | — | |
| First-degree family history, NHL | |||||||||||||||||||||
| No | 9907 (93.2) | 53 (82.8) | 1.00 (Referent) | .014 | 6035 (93.0) | 25 (86.2) | 1.00 (Referent) | .13 | 10053 (80.2) | 168 (70.3) | 1.00 (Referent) | .49 | 9935 (80.1) | 61 (62.2) | 1.00 (Referent) | .11 | 6314 (94.6) | 45 (88.2) | 1.00 (Referent) | .088 | |
| Yes | 218 (2.1) | 5 (7.8) | 4.11 (1.58 to 10.66) | — | 106 (1.6) | 2 (6.9) | — | — | 233 (1.9) | 5 (2.1) | 1.42 (0.55 to 3.63) | — | 231 (1.9) | 4 (4.1) | 2.69 (0.92 to 7.87) | — | 94 (1.4) | 2 (3.9) | 4.81 (1.08 to 21.46) | — | |
| BMI as a young adult (kg/m2) | |||||||||||||||||||||
| 15-<18.5 | 367 (3.8) | 2 (3.4) | 0.70 (0.16 to 3.08) | — | 223 (5.6) | 1 (6.3) | — | — | 366 (3.1) | 9 (4.3) | 2.54 (1.16 to 5.55) | — | 362 (3.5) | 1 (1.5) | — | — | 177 (4.2) | 3 (7.1) | 1.72 (0.45 to 6.59) | — | |
| 18.5-<22.5 | 2610 (27.1) | 16 (27.6) | 1.00 (Referent) | .79 | 1685 (42.2) | 4 (25.0) | 1.00 (Referent) | .07 | 2652 (22.7) | 27 (12.9) | 1.00 (Referent) | .025 | 2618 (25.3) | 15 (22.1) | 1.00 (Referent) | .59 | 1478 (34.7) | 11 (26.2) | 1.00 (Referent) | .41 | |
| 22.5-<25 | 1247 (13.0) | 8 (13.8) | 1.31 (0.54 to 3.13) | — | 778 (19.5) | 0 (0.0) | — | — | 1310 (11.2) | 24 (11.4) | 1.85 (1.04 to 3.28) | — | 1283 (12.4) | 8 (11.8) | 1.17 (0.48 to 2.85) | — | 755 (17.7) | 2 (4.8) | 0.32 (0.07 to 1.50) | — | |
| 25-<30 | 746 (7.8) | 6 (10.3) | 1.74 (0.65 to 4.64) | — | 486 (12.2) | 4 (25.0) | 2.05 (0.50 to 8.42) | — | 762 (6.5) | 10 (4.8) | 1.33 (0.63 to 2.83) | — | 744 (7.2) | 6 (8.8) | 1.46 (0.54 to 3.95) | — | 452 (10.6) | 3 (7.1) | 0.89 (0.24 to 3.33) | — | |
| 30-50 | 147 (1.5) | 1 (1.7) | — | — | 92 (2.3) | 1 (6.3) | — | — | 149 (1.3) | 5 (2.4) | 3.96 (1.46 to 10.75) | — | 142 (1.4) | 2 (2.9) | 2.94 (0.64 to 13.38) | — | 70 (1.6) | 1 (2.4) | 1.31 (0.16 to 10.91) | — | |
| Usual adult BMI (kg/m2) | |||||||||||||||||||||
| 15-<18.5 | 198 (1.4) | 0 (0.0) | — | — | 103 (1.5) | 0 (0.0) | — | — | 199 (1.4) | 1 (0.4) | — | — | 198 (1.4) | 0 (0.0) | — | — | 118 (1.7) | 2 (2.9) | 0.60 (0.13 to 2.63) | — | |
| 18.5-<22.5 | 2797 (19.8) | 17 (19.1) | 1.00 (Referent) | .56 | 1536 (22.5) | 3 (10.0) | 1.00 (Referent) | .93 | 2815 (19.7) | 50 (22.0) | 1.00 (Referent) | .26 | 2797 (19.8) | 17 (21.3) | 1.00 (Referent) | .042 | 1652 (23.5) | 26 (38.2) | 1.00 (Referent) | .49 | |
| 22.5-<25 | 3472 (24.5) | 25 (28.1) | 1.23 (0.65 to 2.30) | — | 1779 (26.0) | 6 (20.0) | 0.91 (0.23 to 3.70) | — | 3494 (24.4) | 54 (23.8) | 0.79 (0.53 to 1.18) | — | 3472 (24.5) | 14 (17.5) | 0.67 (0.33 to 1.37) | — | 1860 (26.5) | 13 (19.1) | 0.53 (0.27 to 1.05) | — | |
| 25-<30 | 5061 (35.7) | 30 (33.7) | 0.98 (0.53 to 1.84) | — | 2346 (34.3) | 16 (53.3) | 1.32 (0.38 to 4.64) | — | 5108 (35.7) | 78 (34.4) | 0.78 (0.54 to 1.14) | — | 5061 (35.7) | 31 (38.8) | 1.08 (0.58 to 2.02) | — | 2326 (33.1) | 20 (29.4) | 0.76 (0.41 to 1.42) | — | |
| 30-<35 | 1512 (10.7) | 8 (9.0) | 0.81 (0.34 to 1.94) | — | 674 (9.9) | 3 (10.0) | 0.79 (0.15 to 4.10) | — | 1539 (10.8) | 32 (14.1) | 1.15 (0.72 to 1.83) | — | 1512 (10.7) | 7 (8.8) | 0.83 (0.33 to 2.08) | — | 672 (9.6) | 4 (5.9) | 0.48 (0.16 to 1.42) | — | |
| 35-50 | 518 (3.7) | 3 (3.4) | 0.82 (0.23 to 2.86) | — | 253 (3.7) | 1 (3.3) | — | — | 540 (3.8) | 6 (2.6) | 0.66 (0.28 to 1.59) | — | 518 (3.7) | 8 (10.0) | 2.93 (1.20 to 7.13) | — | 248 (3.5) | 2 (2.9) | 0.70 (0.16 to 3.06) | — | |
| Lifetime alcohol consumption | |||||||||||||||||||||
| Nondrinker | 2862 (23.3) | 19 (27.1) | 1.00 (Referent) | .51 | 1798 (22.1) | 7 (20.6) | 1.00 (Referent) | .081 | 2938 (18.8) | 55 (19.8) | 1.00 (Referent) | .75 | 2895 (18.7) | 27 (23.7) | 1.00 (Referent) | .33 | 1382 (21.1) | 13 (19.7) | 1.00 (Referent) | .95 | |
| 1–100 kg | 1099 (9.0) | 3 (4.3) | 0.60 (0.15 to 2.37) | — | 731 (9.0) | 0 (0.0) | — | — | 1099 (7.0) | 15 (5.4) | 0.95 (0.48 to 1.88) | — | 1099 (7.1) | 7 (6.1) | 0.79 (0.30 to 2.11) | — | 1076 (16.4) | 12 (18.2) | 1.41 (0.47 to 4.24) | — | |
| 101–200 kg | 576 (4.7) | 2 (2.9) | 0.80 (0.16 to 3.91) | — | 358 (4.4) | 3 (8.8) | 1.59 (0.29 to 8.65) | — | 576 (3.7) | 7 (2.5) | 0.78 (0.33 to 1.87) | — | 576 (3.7) | 5 (4.4) | 1.05 (0.36 to 3.13) | — | 423 (6.5) | 2 (3.0) | 1.50 (0.28 to 7.99) | — | |
| 201–400 kg | 607 (4.9) | 6 (8.6) | 2.29 (0.74 to 7.11) | — | 336 (4.1) | 1 (2.9) | — | — | 607 (3.9) | 6 (2.2) | 0.53 (0.21 to 1.34) | — | 607 (3.9) | 2 (1.8) | 0.40 (0.09 to 1.87) | — | 380 (5.8) | 2 (3.0) | 2.10 (0.39 to 11.36) | — | |
| >400 kg | 735 (6.0) | 4 (5.7) | 1.51 (0.41 to 5.51) | — | 306 (3.8) | 0 (0.0) | — | — | 735 (4.7) | 11 (4.0) | 0.66 (0.30 to 1.45) | — | 735 (4.8) | 1 (0.9) | — | — | 330 (5.0) | 1 (1.5) | — | — | |
| Drinker, unknown use | 4715 (38.4) | 29 (41.4) | 0.94 (0.46 to 1.92) | — | 3847 (47.3) | 18 (52.9) | 0.87 (0.27 to 2.76) | — | 5257 (33.7) | 120 (43.2) | 1.02 (0.68 to 1.54) | — | 5161 (33.4) | 51 (44.7) | 1.04 (0.58 to 1.87) | — | 2268 (34.6) | 34 (51.5) | 0.87 (0.40 to 1.91) | — | |
| Recreational sun exposure (h/week)§ | |||||||||||||||||||||
| Q1 (low) | 2059 (20.4) | 16 (24.2) | 1.00 (Referent) | .53 | 794 (17.8) | 2 (8.7) | 1.00 (Referent) | .32 | 2059 (20.4) | 45 (24.3) | 1.00 (Referent) | .017 | 2059 (20.4) | 11 (20.4) | 1.00 (Referent) | .88 | 969 (18.7) | 19 (23.2) | 1.00 (Referent) | .70 | |
| Q2 | 2160 (21.4) | 12 (18.2) | 0.71 (0.33 to 1.53) | — | 765 (17.2) | 7 (30.4) | 3.07 (0.60 to 15.57) | — | 2160 (21.4) | 49 (26.5) | 1.08 (0.71 to 1.63) | — | 2160 (21.4) | 11 (20.4) | 1.06 (0.46 to 2.48) | — | 937 (18.1) | 14 (17.1) | 0.70 (0.34 to 1.42) | — | |
| Q3 | 1975 (19.6) | 8 (12.1) | 0.54 (0.23 to 1.28) | — | 945 (21.2) | 4 (17.4) | 1.13 (0.20 to 6.45) | — | 1975 (19.6) | 27 (14.6) | 0.57 (0.35 to 0.94) | — | 1975 (19.6) | 8 (14.8) | 0.74 (0.29 to 1.86) | — | 1129 (21.8) | 17 (20.7) | 0.70 (0.35 to 1.39) | — | |
| Q4 (high) | 2799 (27.7) | 14 (21.2) | 0.69 (0.32 to 1.45) | — | 941 (21.1) | 6 (26.1) | 1.34 (0.25 to 7.10) | — | 2799 (27.7) | 44 (23.8) | 0.67 (0.44 to 1.03) | — | 2799 (27.7) | 13 (24.1) | 0.95 (0.42 to 2.16) | — | 1125 (21.8) | 18 (22.0) | 0.74 (0.37 to 1.47) | — | |
| Lifetime cigarette exposure | |||||||||||||||||||||
| Nonsmoker | 6863 (42.5) | 33 (36.3) | 1.00 (Referent) | .024 | 3572 (40.5) | 6 (15.4) | 1.00 (Referent) | .091 | 6926 (42.5) | 130 (43.8) | 1.00 (Referent) | .79 | 6863 (42.5) | 37 (31.9) | 1.00 (Referent) | .015 | 2854 (39.4) | 43 (50.0) | 1.00 (Referent) | .68 | |
| 1–10 pack-years | 2736 (16.9) | 17 (18.7) | 1.51 (0.83 to 2.74) | — | 1531 (17.3) | 4 (10.3) | 1.30 (0.36 to 4.69) | — | 2765 (17.0) | 42 (14.1) | 0.84 (0.59 to 1.19) | — | 2736 (16.9) | 20 (17.2) | 1.44 (0.82 to 2.51) | — | 1316 (18.2) | 19 (22.1) | 0.87 (0.49 to 1.53) | — | |
| 11–20 pack-years | 1739 (10.8) | 3 (3.3) | 0.41 (0.12 to 1.33) | — | 906 (10.3) | 1 (2.6) | — | — | 1755 (10.8) | 30 (10.1) | 0.84 (0.56 to 1.26) | — | 1739 (10.8) | 11 (9.5) | 1.21 (0.61 to 2.41) | — | 747 (10.3) | 6 (7.0) | 0.57 (0.24 to 1.37) | — | |
| 21–35 pack-years | 1754 (10.9) | 17 (18.7) | 2.23 (1.22 to 4.09) | — | 869 (9.8) | 7 (17.9) | 2.35 (0.78 to 7.10) | — | 1769 (10.9) | 34 (11.4) | 0.88 (0.60 to 1.31) | — | 1754 (10.9) | 17 (14.7) | 1.88 (1.04 to 3.39) | — | 684 (9.4) | 7 (8.1) | 0.90 (0.39 to 2.06) | — | |
| >35 pack-years | 1847 (11.4) | 12 (13.2) | 1.52 (0.76 to 3.05) | — | 1064 (12.0) | 14 (35.9) | 2.72 (1.02 to 7.24) | — | 1862 (11.4) | 37 (12.5) | 0.84 (0.57 to 1.24) | — | 1847 (11.4) | 23 (19.8) | 2.47 (1.41 to 4.34) | — | 871 (12.0) | 5 (5.8) | 0.82 (0.31 to 2.17) | — | |
| Smoker, pack-years unknown | 291 (1.8) | 2 (2.2) | 1.54 (0.36 to 6.65) | — | 148 (1.7) | 2 (5.1) | 3.26 (0.61 to 17.44) | — | 292 (1.8) | 8 (2.7) | 1.26 (0.60 to 2.67) | — | 291 (1.8) | 0 (0.0) | — | — | 117 (1.6) | 4 (4.7) | 1.80 (0.59 to 5.46) | — | |
| Ever lived on a farm | |||||||||||||||||||||
| No | 4779 (56.3) | 35 (64.8) | 1.00 (Referent) | .25 | 3125 (66.3) | 14 (73.7) | 1.00 (Referent) | .30 | 4779 (56.3) | 57 (59.4) | 1.00 (Referent) | .94 | 4779 (56.3) | 32 (64.0) | 1.00 (Referent) | .33 | 2825 (67.6) | 19 (67.9) | 1.00 (Referent) | .038 | |
| Yes | 3470 (40.9) | 15 (27.8) | 0.69 (0.36 to 1.31) | — | 1526 (32.4) | 5 (26.3) | 0.58 (0.20 to 1.68) | — | 3470 (40.9) | 36 (37.5) | 0.98 (0.63 to 1.53) | — | 3470 (40.9) | 15 (30.0) | 0.73 (0.38 to 1.39) | — | 1293 (30.9) | 7 (25.0) | 3.12 (1.12 to 8.70) | — | |
| Inflammatory bowel disorder | |||||||||||||||||||||
| No | 14952 (97.5) | 92 (95.8) | 1.00 (Referent) | .11 | 7866 (98.3) | 37 (100) | 1.00 (Referent) | .24 | 15089 (97.5) | 276 (95.5) | 1.00 (Referent) | .0060 | 14952 (97.5) | 102 (96.2) | 1.00 (Referent) | .80 | 6133 (98.4) | 55 (96.5) | 1.00 (Referent) | .61 | |
| Yes | 182 (1.2) | 0 (0.0) | — | — | 131 (1.6) | 0 (0.0) | — | — | 184 (1.2) | 11 (3.8) | 2.70 (1.44 to 5.05) | — | 182 (1.2) | 2 (1.9) | 1.21 (0.29 to 4.96) | — | 93 (1.5) | 1 (1.8) | 1.77 (0.23 to 13.39) | — | |
| Duration of hair dye use | |||||||||||||||||||||
| Never hair dye | 1312 (13.4) | 9 (14.5) | 1.00 (Referent) | .89 | — | — | — | — | 1312 (13.4) | 19 (10.7) | 1.00 (Referent) | .14 | 1260 (13.6) | 14 (17.9) | 1.00 (Referent) | .26 | 571 (14.9) | 11 (26.2) | 1.00 (Referent) | <.001 | |
| 1–8 y | 830 (8.5) | 7 (11.3) | 1.10 (0.38 to 3.19) | — | — | — | — | — | 830 (8.5) | 13 (7.3) | 1.75 (0.76 to 4.05) | — | 756 (8.2) | 4 (5.1) | 1.24 (0.30 to 5.15) | — | 336 (8.7) | 7 (16.7) | 0.58 (0.21 to 1.62) | — | |
| 9–19 y | 608 (6.2) | 4 (6.5) | 0.80 (0.23 to 2.73) | — | — | — | — | — | 608 (6.2) | 3 (1.7) | 0.52 (0.14 to 1.89) | — | 565 (6.1) | 6 (7.7) | 2.33 (0.64 to 8.52) | — | 302 (7.9) | 1 (2.4) | 0.15 (0.02 to 1.22) | — | |
| ≥20 y | 751 (7.7) | 9 (14.5) | 1.38 (0.51 to 3.76) | — | — | — | — | — | 751 (7.7) | 15 (8.5) | 1.87 (0.83 to 4.22) | — | 668 (7.2) | 6 (7.7) | 1.83 (0.50 to 6.64) | — | 375 (9.8) | 11 (26.2) | 4.97 (1.63 to 15.15) | — | |
| Ever use, duration unknown | 518 (5.3) | 4 (6.5) | 1.44 (0.33 to 6.22) | — | — | — | — | — | 518 (5.3) | 13 (7.3) | 1.10 (0.48 to 2.54) | — | 518 (5.6) | 6 (7.7) | 0.41 (0.15 to 1.15) | — | 24 (0.6) | 0 (0.0) | — | — | |
| Men | 5516 (56.4) | 27 (43.5) | — | — | — | — | — | — | 5516 (56.4) | 111 (62.7) | — | — | 5235 (56.6) | 38 (48.7) | — | — | 2011 (52.4) | 11 (26.2) | — | — | |
| Used hair dyes before 1980 | |||||||||||||||||||||
| Never hair dye | 1312 (13.4) | 9 (14.5) | 1.00 (Referent) | .26 | — | — | — | — | 1312 (13.4) | 19 (10.7) | 1.00 (Referent) | .63 | 1260 (13.6) | 14 (17.9) | 1.00 (Referent) | .17 | 571 (14.9) | 11 (26.2) | 1.00 (Referent) | .022 | |
| Ever hair dye use <1980 | 854 (8.7) | 5 (8.1) | 0.61 (0.19 to 1.98) | — | — | — | — | — | 854 (8.7) | 17 (9.6) | 1.77 (0.74 to 4.26) | — | 854 (9.2) | 10 (12.8) | 2.28 (0.68 to 7.66) | — | 594 (15.5) | 10 (23.8) | 2.75 (0.91 to 8.29) | — | |
| Hair dye use only ≥1980 | 853 (8.7) | 5 (8.1) | 0.72 (0.22 to 2.32) | — | — | — | — | — | 853 (8.7) | 9 (5.1) | 1.48 (0.55 to 3.97) | — | 853 (9.2) | 4 (5.1) | 0.93 (0.23 to 3.83) | — | 422 (11.0) | 9 (21.4) | 0.56 (0.22 to 1.45) | — | |
| Ever use, years unknown | 1000 (10.2) | 14 (22.6) | 2.85 (0.82 to 9.95) | — | — | — | — | — | 1000 (10.2) | 18 (10.2) | 1.05 (0.50 to 2.21) | — | 800 (8.6) | 8 (10.3) | 0.50 (0.20 to 1.29) | — | 21 (0.5) | 0 (0.0) | — | — | |
| Missing/unknown | 253 (2.6) | 2 (3.2) | — | — | — | — | — | — | 253 (2.6) | 3 (1.7) | — | — | 253 (2.7) | 4 (5.1) | — | — | 222 (5.8) | 1 (2.4) | — | — | |
| Men | 5516 (56.4) | 27 (43.5) | — | — | — | — | — | — | 5516 (56.4) | 111 (62.7) | — | — | 5235 (56.6) | 38 (48.7) | — | — | 2011 (52.4) | 11 (26.2) | — | — | |
* Odds ratio (OR) and 95% confidence interval (CI) adjusted for age, sex, race/ethnicity, and study. BMI = body mass index; CNS = central nervous system; HCV = hepatits C virus; GI = gastrointestinal; HT = hormone therapy; NHL = non-Hodgkin lymphoma; OC = oral contraceptive; SES = socioeconomic status.
† Includes self-reported history of specific autoimmune diseases occurring ≥2 years before diagnosis/interview (except the New South Wales study, which did not ascertain date of onset). Autoimmune diseases were classified according to whether they are primarily mediated by B-cell or T-cell responses. B-cell activating diseases included Hashimoto thyroiditis, hemolytic anemia, myasthenia gravis, pernicious anemia, rheumatoid arthritis, Sjögren’s syndrome, and systemic lupus erythematosus. T-cell activating diseases included celiac disease, immune thrombocytopenic purpura, inflammatory bowel disorder (Crohn’s disease, ulcerative colitis), multiple sclerosis, polymyositis or dermatomyositis, psoriasis, sarcoidosis, systemic sclerosis or scleroderma, and type 1 diabetes.
‡ Includes self-reported history of atopic disorders including asthma, eczema, hay fever, or other allergies, excluding drug allergies, occurring ≥2 years prior to diagnosis/interview (except the New South Wales study, which did not ascertain date of onset).
§ Study specific quartiles among controls.
Discussion
Our study provides strong evidence that many DLBCL risk factors previously recognized but assessed individually are independent of each other, supporting a complex and multifactorial etiology. Specifically, B-cell activating autoimmune diseases, HCV seropositivity, family history of NHL, and higher young adult BMI were associated with increased DLBCL risk, whereas higher SES, any atopic disorder, and greater recreational sun exposure were associated with decreased risk. Some risk factors were sex specific: OC use before 1970, HT use starting at age at least 50 years, and low usual adult BMI were all inversely associated with risk among women, whereas previous blood transfusion and lifetime alcohol consumption were both inversely associated with risk among men. Occupational associations were largely sex specific.
Several notable findings emerged from this analysis. Perhaps the most prominent was the robust link to risk factors associated with immune function, given the strong and independent associations of B-cell activating autoimmune diseases and HCV seropositivity with DLBCL risk, both of which mechanistically implicate chronic immune stimulation (14,16). In contrast, the immune impact of atopic disorders (with characteristic hypersensitivity reactions and associated release of inflammatory molecules) appears to play a role in decreasing DLBCL risk (18). The independent effects of each of these factors based on our multivariate results support a complex interaction of immune system function and DLBCL risk, which requires further research to understand the underlying mechanisms.
A second notable finding was the positive and independent association of young adult BMI with DLBCL risk, which was the same for both sexes. Although young adult BMI and usual adult BMI were correlated, 51% of the controls moved to a different BMI category (mainly gain) and with our large sample size we were able to model both simultaneously. The consistency of our findings with most previous data from cohort studies (29–34) strongly supports the robustness of excess body weight with DLBCL at younger ages, with implications for primary prevention. Whether BMI impacts DLBCL risk through immune, hormonal, or another mechanism remains to be determined (35). A role for hormonal factors was also supported by the protective effects of both OC and HT use in women, although positive impacts on immune function are also plausible (19).
The inverse association of sunlight with DLBCL risk was independent of physical activity, BMI, and other lifestyle factors including alcohol use. Basic biological and some epidemiological evidence suggests that either vitamin D or UV-induced immune alterations independent of vitamin D could be responsible for the observed inverse association (36–39), although circulating serum vitamin D levels evaluated a median of 5 years before DLBCL diagnosis were not associated with risk in a large pooling project (36). UV radiation induction of regulatory T cells also is plausible but requires further investigation (40,41). Studies using intermediate biomarkers of exposure or effect will be needed to sort out this hypothesis.
An inverse association with transfusion history, significant only for men, will be more fully reported in a separate InterLymph publication; the findings here extend this association as independent of other immune and lifestyle factors. The inverse association is difficult to interpret in light of previous studies, which support a positive association (42), and raises concerns about bias or chance as primary explanations.
Many of the occupations associated with increased DLBCL risk may be related to carcinogenic exposures, but we were unable to pinpoint specific agents; likely candidates include pesticides, solvents, dyes, engine exhausts, wood dust and wood finishing chemicals, and microbes and other biological agents (43–47). Occupational associations were generally sex specific, and this may be due to actual exposure for men and women in the same jobs having different exposure levels or exposure patterns (48,49). A limitation of occupational titles is the high potential for misclassification of exposure status, which can dilute risk estimates.
Our finding of an increased DLBCL risk with family history of NHL was robust even when adjusted for other identified DLBCL risk factors. This is particularly relevant because family history reflects an interaction between shared exposures (both from shared environment and behaviors) and genetic susceptibility. The association between family history of NHL and DLBCL risk was not attenuated, and was even strengthened, by inclusion of environmental and lifestyle factors, suggesting that heritability is an independent risk factor for DLBCL. Further analyses are warranted to determine whether established genetic associations are independent of family history, or whether they help explain in part the association between family history of NHL and DLBCL.
We found little evidence for heterogeneity when we stratified these associations by age at diagnosis (<60 versus 60+ years), noting that 70% of the cases were aged 50–79 years at diagnosis, perhaps limiting our ability to identify age-related differences.
We aimed to explore whether risk factors identified for all DLBCL were common to selected sites (see Table 3), each of which has a somewhat unique descriptive (6–10) but little published risk factor epidemiology. For primary CNS lymphoma, the only established risk factors are congenital and acquired immunodeficiency (particularly HIV disease, which we excluded) (7). Consistent with the only comparable case-control study (50) were our observations of an inverse association with OC use, positive associations with family history of NHL and smoking, and null associations for farming and transfusion history; an inverse association with atopy in our study was null in that study, whereas we did not examine autoimmune disease (only six exposed cases) and did not assess tonsillectomy, reported to reduce risk in the previous study. For GI lymphoma, of which DLBCL is the most common histology (9,51), transplantation, HIV infection, and ulcerative colitis have been suggested as risk factors (52); we excluded HIV patients, whereas the reported association with ulcerative colitis is consistent with our finding for inflammatory bowel disease.
Autoimmune disease, atopy, family history of NHL, and perhaps young adult BMI showed consistent risk patterns across anatomical sites, supporting a shared etiology. Intriguing site-specific patterns included associations for smoking with CNS, testicular and cutaneous DLBCL; inflammatory bowel disease with GI DLBCL; and ever having lived on a farm and hair dye use with mediastinal DLBCL. These findings suggest consideration of unique pathogenic mechanisms such as local inflammatory mechanisms in GI DLBCL and chemical exposure for mediastinal DLBCL. The approximately fourfold increase in CNS and mediastinal DLBCL risk with family history of NHL may suggest that heritability and genetics play even larger roles in these presentations. Nevertheless, caution is needed in interpreting these results due to small number of cases in these analyses, difficulty of diagnosing these rare entities, and the complexity of assigning a primary site for NHL in epidemiological studies. Furthermore, heterogeneity within some of the extranodal sites, particularly mediastinal (which will include true DLBCL but also primary mediastinal B-cell lymphoma), cutaneous (e.g., leg type versus other), and GI tract (e.g., some gastric cases are likely to be associated with Helicobacter pylori infection), could not be assessed.
Strengths specific to this analysis (beyond those of the entire initiative) (22) included the largest sample size of DLBCL cases yet assembled for a risk factor analysis and availability of data on anatomical sites. Specific limitations include the lack of data on DLBCL cell of origin (which may have unique etiologies) (11) and limited power to perform site-specific analyses. We did not adjust our results for multiple comparisons, although most of these exposures had a strong a priori probability. We had few cases above the age of 80 years, although this age group has the highest DLBCL incidence rates. Our analyses excluded HIV-associated DLBCL. Assignment to anatomical sites was limited to those studies that coded this information independent of stage. Beyond HCV, we did not have information on other infections. It will be important for future epidemiological studies to incorporate biological specimens to help elucidate the mechanisms underlying many of these observed associations.
In summary, this large pooled analysis has clarified multiple risk factors for DLBCL and has demonstrated that many previously reported factors act independently on DLBCL risk. These new findings will need to be incorporated into models of DLBCL epidemiology and pathogenesis, and support a complex biology that includes familial, medical, lifestyle, and occupational factors.
Funding
Intramural Research Program of the National Cancer Institute/National Institutes of Health and National Cancer Institute/National Institutes of Health (R01 CA14690, U01 CA118444, and R01 CA92153-S1).
InterLymph annual meetings during 2010–2013 were supported by the Epidemiology and Genomics Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute/National Institutes of Health (2010–2013); Lymphoma Coalition (2010–2013); National Institutes of Health Office of Rare Diseases Research (2010); National Cancer Institute/National Institutes of Health (R13 CA159842 01) (2011); University of Cagliari, Provincial Administration of Cagliari, Banca di Credito Sardo, and Consorzio Industriale Sardo, Italy (2011); Intramural Research Program of the National Cancer Institute/National Institutes of Health (2012); and Faculté de Médecine de Dijon, Institut de Veille Sanitaire, Registre des hémopathies malignes de Côte d’Or, INSERM, Institut National du Cancer, Université de Bourgogne, Groupe Ouest Est d’Etude des Leucémies et Autres Maladies du Sang (GOELAMS), l’Institut Bergonié, The Lymphoma Study Association (LYSA), Registre Régional des Hémopathies de Basse Normandie, and the City of Dijon, France (2013). Meeting space at the 2013 Annual Meeting of the American Association for Cancer Research (AACR) was provided by the Molecular Epidemiology Group (MEG) of the AACR. Pooling of the occupation data was supported by the National Cancer Institute/National Institutes Health (R03 CA125831).
Individual studies were supported by the Canadian Institutes for Health Research (CIHR), Canadian Cancer Society, and Michael Smith Foundation for Health Research (British Columbia); Intramural Research Program of the National Cancer Institute/National Institutes of Health (Iowa/Minnesota); National Cancer Institute/National Institutes of Health (N01-CP-ES-11027) (Kansas); National Cancer Institute/National Institutes of Health (R01 CA50850) (Los Angeles); National Cancer Institute/National Institutes of Health (R01 CA92153 and P50 CA97274), Lymphoma Research Foundation (164738), and the Henry J. Predolin Foundation (Mayo Clinic); Intramural Research Program of the National Cancer Institute/National Institutes of Health and Public Health Service (contracts N01-PC-65064, N01-PC-67008, N01-PC-67009, N01-PC-67010, and N02-PC-71105) (NCI-SEER); National Cancer Institute/National Institutes of Health (R01CA100555 and R03CA132153) and American Institute for Cancer Research (99B083) (Nebraska [newer]); National Cancer Institute/National Institutes of Health (N01-CP-95618) and State of Nebraska Department of Health (LB-506) (Nebraska [older]); National Cancer Institute/National Institutes of Health (R01CA45614, RO1CA154643-01A1, and R01CA104682) (UCSF1); National Cancer Institute/National Institutes of Health (CA143947, CA150037, R01CA087014, R01CA104682, RO1CA122663, and RO1CA154643-01A1) (UCSF2); National Heart Lung and Blood Institute/National Institutes of Health (hematology training grant award T32 HL007152), National Center for Research Resources/National Institutes of Health (UL 1 RR024160), and National Cancer Institute/National Institutes of Health (K23 CA102216 and P50 CA130805) (University of Rochester); National Cancer Institute/National Institutes of Health (CA62006 and CA165923) (Yale); Association pour la Recherche contre le Cancer (ARC), Institut National du Cancer (INCa), Fondation de France, Fondation contre la Leucémie, Agence nationale de sécurité sanitaire de l’alimentation, de l’environnement et du travail (ANSES) (Engela); European Commission (QLK4-CT-2000-00422 and FOOD-CT-2006–023103), Spanish Ministry of Health (CIBERESP, PI11/01810, RCESP C03/09, RTICESP C03/10, and RTIC RD06/0020/0095), Rio Hortega (CM13/00232), Agència de Gestió d’Ajuts Universitaris i de Recerca–Generalitat de Catalunya (Catalonian Government, 2009SGR1465), National Institutes of Health (contract NO1-CO-12400), Italian Ministry of Education, University and Research (PRIN 2007 prot.2007WEJLZB, PRIN 2009 prot. 20092ZELR2), Italian Association for Cancer Research (IG grant 11855/2011), Federal Office for Radiation Protection (StSch4261 and StSch4420), José Carreras Leukemia Foundation (DJCLS-R04/08), German Federal Ministry for Education and Research (BMBF-01-EO-1303), Health Research Board, Ireland and Cancer Research Ireland, and Czech Republic MH CZ - DRO (MMCI, 00209805) (EpiLymph); National Cancer Institute/National Institutes of Health (CA51086), European Community (Europe Against Cancer Programme), and Italian Alliance Against Cancer (Lega Italiana per la Lotta contro i Tumori) (Italy, multicenter); Italian Association for Cancer Research (IG 10068) (Italy, Aviano-Milan); Italian Association for Cancer Research (Italy, Aviano-Naples); Swedish Cancer Society (2009/659), Stockholm County Council (20110209), Strategic Research Program in Epidemiology at Karolinska Institut, Swedish Cancer Society (02 6661), Danish Cancer Research Foundation, Lundbeck Foundation (R19-A2364), Danish Cancer Society (DP 08-155), National Cancer Institute/National Institutes of Health (5R01 CA69669-02), and Plan Denmark (SCALE); Leukaemia & Lymphoma Research (United Kingdom); and Australian National Health and Medical Research Council (ID990920), Cancer Council NSW, and University of Sydney Faculty of Medicine (New South Wales).
Supplementary Material
We thank the following individuals for their substantial contributions to this project: Aaron D. Norman, Dennis P. Robinson, and Priya Ramar (Mayo Clinic College of Medicine) for their work at the InterLymph Data Coordinating Center in organizing, collating, harmonizing, and documenting of the data from the participating studies in the InterLymph Consortium; Michael Spriggs, Peter Hui, and Bill Wheeler (Information Management Services, Inc) for their programming support; and Noelle Richa Siegfried and Emily Smith (RTI International) for project coordination.
References
- 1. Howlader N, Noone AM, Krapcho M, et al. SEER Cancer Statistics Review, 1975–2010. National Cancer Institute; Bethesda, MD, http://seer.cancer.gov/csr/1975_2010/, based on November 2012 SEER data submission, posted to the SEER website April 2013. Accessed 10 May 2013. [Google Scholar]
- 2. Sant M, Allemani C, Tereanu C, et al. ; HAEMACARE Working Group. Incidence of hematologic malignancies in Europe by morphologic subtype: results of the HAEMACARE project. Blood. 2010;116(19):3724–3734 [DOI] [PubMed] [Google Scholar]
- 3. Flowers CR, Sinha R, Vose JM. Improving outcomes for patients with diffuse large B-cell lymphoma. CA Cancer J Clin. 2010;60(6):393–408 [DOI] [PubMed] [Google Scholar]
- 4. Shenoy PJ, Malik N, Nooka A, et al. Racial differences in the presentation and outcomes of diffuse large B-cell lymphoma in the United States. Cancer. 2011;117(11):2530–2540 [DOI] [PubMed] [Google Scholar]
- 5. Møller MB, Pedersen NT, Christensen BE. Diffuse large B-cell lymphoma: clinical implications of extranodal versus nodal presentation–a population-based study of 1575 cases. Br J Haematol. 2004;124(2):151–159 [DOI] [PubMed] [Google Scholar]
- 6. Gundrum JD, Mathiason MA, Moore DB, Go RS. Primary testicular diffuse large B-cell lymphoma: a population-based study on the incidence, natural history, and survival comparison with primary nodal counterpart before and after the introduction of rituximab. J Clin Oncol. 2009;27(31):5227–5232 [DOI] [PubMed] [Google Scholar]
- 7. Schabet M. Epidemiology of primary CNS lymphoma. J Neurooncol. 1999;43(3):199–201 [DOI] [PubMed] [Google Scholar]
- 8. Zinzani PL, Piccaluga PP. Primary mediastinal DLBCL: evolving biologic understanding and therapeutic strategies. Curr Oncol Rep. 2011;13(5):407–415 [DOI] [PubMed] [Google Scholar]
- 9. Stanojevic GZ, Nestorovic MD, Brankovic BR, Stojanovic MP, Jovanovic MM, Radojkovic MD. Primary colorectal lymphoma: an overview. World J Gastrointest Oncol. 2011;3(1):14–18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Markova A, Weinstock MA. Trends in cutaneous lymphoma epidemiology. Clin Lymphoma Myeloma Leuk. 2010;10( suppl 2:S63–S66 [DOI] [PubMed] [Google Scholar]
- 11. Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403(6769):503–511 [DOI] [PubMed] [Google Scholar]
- 12. Morton LM, Hartge P, Holford TR, et al. Cigarette smoking and risk of non-Hodgkin lymphoma: a pooled analysis from the International Lymphoma Epidemiology Consortium (InterLymph). Cancer Epidemiol Biomarkers Prev. 2005;14(4):925–933 [DOI] [PubMed] [Google Scholar]
- 13. Wang SS, Slager SL, Brennan P, et al. Family history of hematopoietic malignancies and risk of non-Hodgkin lymphoma (NHL): a pooled analysis of 10 211 cases and 11 905 controls from the International Lymphoma Epidemiology Consortium (InterLymph). Blood. 2007;109(8):3479–3488 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. de Sanjose S, Benavente Y, Vajdic CM, et al. Hepatitis C and non-Hodgkin lymphoma among 4784 cases and 6269 controls from the International Lymphoma Epidemiology Consortium. Clin Gastroenterol Hepatol. 2008;6(4):451–458 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Kricker A, Armstrong BK, Hughes AM, et al. ; Interlymph Consortium. Personal sun exposure and risk of non Hodgkin lymphoma: a pooled analysis from the Interlymph Consortium. Int J Cancer. 2008;122(1):144–154 [DOI] [PubMed] [Google Scholar]
- 16. Ekström Smedby K, Vajdic CM, Falster M, et al. Autoimmune disorders and risk of non-Hodgkin lymphoma subtypes: a pooled analysis within the InterLymph Consortium. Blood. 2008;111(8):4029–4038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Willett EV, Morton LM, Hartge P, et al. ; Interlymph Consortium. Non-Hodgkin lymphoma and obesity: a pooled analysis from the InterLymph Consortium. Int J Cancer. 2008;122(9):2062–2070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Vajdic CM, Falster MO, de Sanjose S, et al. Atopic disease and risk of non-Hodgkin lymphoma: an InterLymph pooled analysis. Cancer Res. 2009;69(16):6482–6489 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kane EV, Bernstein L, Bracci PM, et al. ; InterLymph Consortium. Postmenopausal hormone therapy and non-Hodgkin lymphoma: a pooled analysis of InterLymph case-control studies. Ann Oncol. 2013;24(2):433–441 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kane EV, Roman E, Becker N, et al. ; InterLymph Consortium. Menstrual and reproductive factors, and hormonal contraception use: associations with non-Hodgkin lymphoma in a pooled analysis of InterLymph case-control studies. Ann Oncol. 2012;23(9):2362–2374 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Zhang Y, Sanjose SD, Bracci PM, et al. Personal use of hair dye and the risk of certain subtypes of non-Hodgkin lymphoma. Am J Epidemiol. 2008;167(11):1321–1331 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Morton LM, Sampson JN, Cerhan JR, et al. Rationale and design of the International Lymphoma Epidemiology Consortium (InterLymph) Non-Hodgkin Lymphoma Subtypes Project. J Natl Cancer Inst Monogr. 2014. (48):1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Jaffe ES, Harris NL, Stein H, Vardiman JW. eds. World Health Organization Classification of Tumours: Pathology and Genetics, Tumours of Hematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001 [Google Scholar]
- 24. Swerdlow S, Campo E, Harris N, et al. eds World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. Lyon, France: IARC Press; 2008 [Google Scholar]
- 25. Morton LM, Turner JJ, Cerhan JR, et al. Proposed classification of lymphoid neoplasms for epidemiologic research from the Pathology Working Group of the International Lymphoma Epidemiology Consortium (InterLymph). Blood. 2007;110(2):695–708 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Turner JJ, Morton LM, Linet MS, et al. InterLymph hierarchical classification of lymphoid neoplasms for epidemiologic research based on the WHO classification (2008): update and future directions. Blood. 2010;116(20):e90–e98 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Johnson CH, Adamo M, Dickie L, et al. 2012 Hematopoietic Manual, Version 2.2. Bethesda, MD: National Cancer Institute, 20892–8316 [Google Scholar]
- 28. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–1558 [DOI] [PubMed] [Google Scholar]
- 29. Bertrand KA, Giovannucci E, Zhang SM, Laden F, Rosner B, Birmann BM. A prospective analysis of body size during childhood, adolescence, and adulthood and risk of non-Hodgkin lymphoma. Cancer Prev Res (Phila). 2013;6(8):864–873 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Maskarinec G, Erber E, Gill J, Cozen W, Kolonel LN. Overweight and obesity at different times in life as risk factors for non-Hodgkin’s lymphoma: the multiethnic cohort. Cancer Epidemiol Biomarkers Prev. 2008;17(1):196–203 [DOI] [PubMed] [Google Scholar]
- 31. Cerhan JR, Janney CA, Vachon CM, et al. Anthropometric characteristics, physical activity, and risk of non-Hodgkin’s lymphoma subtypes and B-cell chronic lymphocytic leukemia: a prospective study. Am J Epidemiol. 2002;156(6):527–535 [DOI] [PubMed] [Google Scholar]
- 32. Patel AV, Diver WR, Teras LR, Birmann BM, Gapstur SM. Body mass index, height and risk of lymphoid neoplasms in a large United States cohort. Leuk Lymphoma. 2013;54(6):1221–1227 [DOI] [PubMed] [Google Scholar]
- 33. Pylypchuk RD, Schouten LJ, Goldbohm RA, Schouten HC, van den Brandt PA. Body mass index, height, and risk of lymphatic malignancies: a prospective cohort study. Am J Epidemiol. 2009;170(3):297–307 [DOI] [PubMed] [Google Scholar]
- 34. Troy JD, Hartge P, Weissfeld JL, et al. Associations between anthropometry, cigarette smoking, alcohol consumption, and non-Hodgkin lymphoma in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Am J Epidemiol. 2010;171(12):1270–1281 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Roberts DL, Dive C, Renehan AG. Biological mechanisms linking obesity and cancer risk: new perspectives. Annu Rev Med. 2010;61:301–316 [DOI] [PubMed] [Google Scholar]
- 36. Purdue MP, Freedman DM, Gapstur SM, et al. Circulating 25-hydroxyvitamin D and risk of non-Hodgkin lymphoma: Cohort Consortium Vitamin D Pooling Project of Rarer Cancers. Am J Epidemiol. 2010;172(1):58–69 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Norval M, Halliday GM. The consequences of UV-induced immunosuppression for human health. Photochem Photobiol. 2011;87(5):965–977 [DOI] [PubMed] [Google Scholar]
- 38. Hart PH, Gorman S, Finlay-Jones JJ. Modulation of the immune system by UV radiation: more than just the effects of vitamin D? Nat Rev Immunol. 2011;11(9):584–596 [DOI] [PubMed] [Google Scholar]
- 39. Gorman S, Scott NM, Tan DH, et al. Acute erythemal ultraviolet radiation causes systemic immunosuppression in the absence of increased 25-hydroxyvitamin D3 levels in male mice. PLoS One. 2012;7(9):e46006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Norval M, McLoone P, Lesiak A, Narbutt J. The effect of chronic ultraviolet radiation on the human immune system. Photochem Photobiol. 2008;84(1):19–28 [DOI] [PubMed] [Google Scholar]
- 41. Schwarz T. 25 years of UV-induced immunosuppression mediated by T cells-from disregarded T suppressor cells to highly respected regulatory T cells. Photochem Photobiol. 2008;84(1):10–18 [DOI] [PubMed] [Google Scholar]
- 42. Castillo JJ, Dalia S, Pascual SK. Association between red blood cell transfusions and development of non-Hodgkin lymphoma: a meta-analysis of observational studies. Blood. 2010;116(16):2897–2907 [DOI] [PubMed] [Google Scholar]
- 43. Becker N, Berger J, Bolm-Audorff U. Asbestos exposure and malignant lymphomas–a review of the epidemiological literature. Int Arch Occup Environ Health. 2001;74(7):459–469 [DOI] [PubMed] [Google Scholar]
- 44. Blair A, Linos A, Stewart PA, et al. Comments on occupational and environmental factors in the origin of non-Hodgkin’s lymphoma. Cancer Res. 1992;52(suppl 19):5501s–5502s [PubMed] [Google Scholar]
- 45. Boffetta P, de Vocht F. Occupation and the risk of non-Hodgkin lymphoma. Cancer Epidemiol Biomarkers Prev. 2007;16(3):369–372 [DOI] [PubMed] [Google Scholar]
- 46. Pearce N, McLean D. Agricultural exposures and non-Hodgkin’s lymphoma. Scand J Work Environ Health. 2005;31( suppl 1:18–25; discussion 15–17. [PubMed] [Google Scholar]
- 47. Neasham D, Sifi A, Nielsen KR, et al. Occupation and risk of lymphoma: a multicentre prospective cohort study (EPIC). Occup Environ Med. 2011;68(1):77–81 [DOI] [PubMed] [Google Scholar]
- 48. Kennedy SM, Koehoorn M. Exposure assessment in epidemiology: does gender matter? Am J Ind Med. 2003;44(6):576–583 [DOI] [PubMed] [Google Scholar]
- 49. Eng A, ‘t Mannetje A, McLean D, Ellison-Loschmann L, Cheng S, Pearce N. Gender differences in occupational exposure patterns. Occup Environ Med. 2011;68(12):888–894 [DOI] [PubMed] [Google Scholar]
- 50. Schiff D, Suman VJ, Yang P, Rocca WA, O’Neill BP. Risk factors for primary central nervous system lymphoma: a case-control study. Cancer. 1998;82(5):975–982 [PubMed] [Google Scholar]
- 51. Howell JM, Auer-Grzesiak I, Zhang J, Andrews CN, Stewart D, Urbanski SJ. Increasing incidence rates, distribution and histological characteristics of primary gastrointestinal non-Hodgkin lymphoma in a North American population. Can J Gastroenterol. 2012;26(7):452–456 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Andrews CN, John Gill M, Urbanski SJ, Stewart D, Perini R, Beck P. Changing epidemiology and risk factors for gastrointestinal non-Hodgkin’s lymphoma in a North American population: population-based study. Am J Gastroenterol. 2008;103(7):1762–1769 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.