Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Anthropometric Characteristics, Physical Activity, and Risk of Non-Hodgkin’s Lymphoma Subtypes and B-Cell Chronic Lymphocytic Leukemia: A Prospective Study

Anthropometric Characteristics, Physical Activity, and Risk of Non-Hodgkin’s Lymphoma Subtypes... Abstract Anthropometric characteristics, physical activity, and risk of non-Hodgkin’s lymphoma, its subtypes, and B-cell chronic lymphocytic leukemia (B-CLL) were evaluated in a prospective cohort study of 37,931 Iowa women who were aged 55–69 years at baseline in 1986. Through 1998, 261 cases of non-Hodgkin’s lymphoma (137 diffuse, 58 follicular, and 32 small lymphocytic lymphomas) and 63 cases of B-CLL were identified by linkage to the Iowa Cancer Registry. Height, weight, body mass index, waist/hip ratio, and physical activity were not associated with risk of non-Hodgkin’s lymphoma overall or with diffuse or follicular lymphoma in particular. After adjustment for other non-Hodgkin’s lymphoma risk factors, there was an inverse association of baseline body mass index (relative risks (RRs) across quartiles: 1, 0.4, 0.4, 0.3; p trend = 0.03) with risk of small lymphocytic lymphoma. In contrast, for B-CLL there were suggestive positive associations with body mass index at age 50 years (RRs across quartiles: 1, 1.9, 1.5, 2.7; p trend = 0.03) and (more weakly) baseline body mass index (RRs across quartiles: 1, 1.1, 1.6, 1.3; p trend = 0.3). In summary, we found no evidence that height, weight, body mass, or physical activity plays an important role in non-Hodgkin’s lymphoma overall or in diffuse or follicular lymphoma in particular. The opposite associations of body mass index with small lymphocytic lymphoma versus B-CLL may be a chance finding but, if confirmed, would suggest different etiologies for these malignancies. anthropometry; cohort studies; leukemia, lymphocytic, chronic; lymphoma, non-Hodgkin Abbreviations: B-CLL, B-cell chronic lymphocytic leukemia; CI, confidence interval; ICD-O-2, International Classification of Diseases for Oncology, Second Edition; RR, relative risk. Received for publication January 9, 2002; accepted for publication May 2, 2002. In 2001, the American Cancer Society estimated that 56,200 persons would be diagnosed with non-Hodgkin’s lymphoma and that 26,300 would die from it (1). Non-Hodgkin’s lymphoma is now the fifth most commonly diagnosed cancer among both men and women. The incidence of non-Hodgkin’s lymphoma in the United States increased 82.6 percent from 1973 to 1998, a percentage increase exceeded only by melanoma of the skin, prostate cancer, and liver and intrahepatic cancer (2). This increase is part of a longer trend documented since 1950 (3), and improved medical technology, changing systems of diagnostic classification, known risk factors, and the human immunodeficiency virus epidemic can explain only part of this increase (4). B-cell chronic lymphocytic leukemia (B-CLL) is the most common leukemia in the United States (2) and is currently classified together with small lymphocytic lymphoma by the World Health Organization (5). Diets high in fat (6, 7) or meat products (6–9) have been associated with risk of non-Hodgkin’s lymphoma. Several anthropometric measures, including body mass (10, 11) and height (7, 12), have also been suggested as non-Hodgkin’s lymphoma risk factors. Data from epidemiologic, clinical, and animal studies, albeit limited, provide some support for the hypothesis that obesity is associated with impairments of both humoral and cellular immunity (13). Obesity is also associated with insulin resistance and the development of adult-onset diabetes, and adult-onset diabetes has been shown in some studies to be a risk factor for non-Hodgkin’s lymphoma (14–17). Physical inactivity is an important determinant of obesity and appears to be a risk factor for several types of cancer, particularly colon cancer and female reproductive cancers (18, 19). The underlying mechanisms responsible for these associations with physical inactivity are not known but are thought to involve body weight, changes in steroid hormone levels (perhaps caused by changes in menstrual function for women), or modulation of immune function (18–20). The extremely limited data available suggest that there is no association between the level of physical activity and risk of non-Hodgkin’s lymphoma or lymphatic leukemias (21–23). There are no data on either anthropometric characteristics or physical activity and risk of B-CLL specifically. We therefore evaluated whether anthropometric factors and physical activity were associated with non-Hodgkin’s lymphoma or B-CLL in a large prospective cohort of Iowa women. In secondary analyses, we evaluated heterogeneity of the results by the most common subtypes of non-Hodgkin’s lymphoma. MATERIALS AND METHODS The Iowa Women’s Health Study cohort Details on the Iowa Women’s Health Study design specific to non-Hodgkin’s lymphoma and B-CLL have been previously published (24). In 1986, a questionnaire was returned by 41,836 randomly selected women who were aged 55–69 years and had a valid Iowa driver’s license in 1985 (42.7 percent response rate). The rates of non-Hodgkin’s lymphoma (overall and for the major subtypes) in the cohort are nearly identical to the rates from the Iowa Surveillance, Epidemiology, and End Results Cancer Registry (data available on request). Self-reported items on the 1986 baseline questionnaire included demographic data, medical history, and dietary and other lifestyle factors. Participants reported current height and weight and weight at ages 18, 30, 40, and 50 years. The body mass index at baseline and at each respective age was calculated as weight (in kilograms) at the given age divided by baseline height squared (in meters). A paper tape measure and written instructions for having a friend measure circumferences of the waist (1 inch (2.54 cm) above the umbilicus) and hips (maximal protrusion) were enclosed with the questionnaire. The waist/hip ratio was calculated from these measures. This protocol has been shown to be valid (correlation coefficient with measures by trained technician, ≥0.84) and reliable (correlation of measures at two different time periods, ≥0.85) (25). Relative weight at age 12 was ascertained using the following question: “Think back to when you were in sixth grade—or about the age of 12. Would you say at that time, compared to other girls your age, your weight was below average for your age and height, about average for your age and height, or above average for your age and height?” Physical activity was assessed by asking participants three questions about whether they participated in any leisure exercise and, if so, the frequency of moderate and vigorous activities. We did not have estimates of occupational physical activity. Responses to these questions were combined to create a three-level activity score (low, medium, high). This scale is strongly associated with total mortality and cardiovascular disease mortality in this cohort (26). Cohort follow-up Vital status and cancer incidence in the cohort were ascertained through 13 years of continuous follow-up (1986 through 1998). Follow-up questionnaires were mailed in 1987, 1989, 1992, and 1997 to ascertain vital status and address changes. Deaths were also ascertained by annual linkage to a database of Iowa death certificates, supplemented by linkage to the National Death Index for survey nonrespondents and emigrants from Iowa. Outmigration has been estimated at 1 percent per year. Cancer incidence except for nonmelanoma skin cancer was ascertained by annual linkage to the Iowa Cancer Registry, part of the National Cancer Institute’s Surveillance, Epidemiology, and End Results program (2). All participants were linked by a combination of Social Security number, first, last, and maiden name, birth date, and ZIP code. The Iowa Cancer Registry collects cancer data including identifying information, tumor site, morphology, histologic grade, and extent of disease on all persons who were Iowa residents at the time of their diagnosis. All tumor site and morphology data were derived from pathology reports of the diagnosing hematopathologist, and there was no centralized pathology review. Topographic and morphologic data were coded using the International Classification of Diseases for Oncology, Second Edition (ICD-O-2) (27), which for lymphomas was derived from the Working Formulation (28). The histologic subtypes of non-Hodgkin’s lymphoma were grouped according to the method of Groves et al. (29) into the following subtypes based on ICD-O-2 codes: small lymphocytic, follicular, diffuse, high grade, and peripheral T cell. This system is a modification of the Working Formulation for use in epidemiologic studies. Chronic lymphoctyic leukemia was also analyzed but as a distinct endpoint; that is, it was not aggregated with small lymphocytic lymphoma. Data analysis Before data analysis, we excluded women with a self-reported history of cancer or cancer chemotherapy on the 1986 baseline questionnaire (n = 3,902) to provide a cancer-free at-risk cohort. Each woman accumulated person-years of follow-up from the date of receipt of the 1986 baseline questionnaire until the date of non-Hodgkin’s lymphoma (or B-CLL) diagnosis, date of emigration from Iowa, or date of death; if none of these events occurred, person-years were accumulated through December 31, 1998. Two women who were censored at their baseline date were further excluded, leaving a final cohort of 37,932 for analysis. All continuous anthropometric variables were classified into quartiles based on the distribution in the at-risk cohort. Physical activity was grouped in low, moderate, and high levels (26). Relative weight at age 12 years was categorized according to possible responses (below average, average, above average), with the average group serving as the reference category. Age-adjusted and multivariate-adjusted relative risks, along with 95 percent confidence intervals, were calculated as measures of association of the risk factors with non-Hodgkin’s lymphoma incidence and were estimated using Cox proportional hazards regression (30), with the survival modeled as a function of age (31). Other risk factors for non-Hodgkin’s lymphoma in this data set included age, marital status, farm residence (32), transfusion history (33), adult-onset diabetes (33), alcohol use (34), cigarette smoking (35), red meat and fruit intake (6), and use of hormone replacement therapy, and they were evaluated in multivariate models as potential confounders. RESULTS The mean baseline age of the 37,932 women in the at-risk cohort was 61.7 years, and over 99 percent were White. During 449,801 person-years of follow-up (1986–1998), 261 women developed non-Hodgkin’s lymphoma and 63 women developed B-CLL. The mean age at diagnosis for non-Hodgkin’s lymphoma was 70.1 (range, 57–81) years and for B-CLL was 70.7 (range, 60–81) years. Approximately 70 percent of the non-Hodgkin’s lymphomas were nodal, and of the extranodal sites, the stomach, small intestine, and thyroid were the most common sites involved. The most common non-Hodgkin’s lymphoma subtype was diffuse lymphoma (53 percent), followed by follicular lymphoma (22 percent) and small lymphocytic lymphoma (12 percent); only 9 percent of the cases could not be classified into the scheme of Groves et al. (29). The age-adjusted associations of baseline anthropometric characteristics and of physical activity with risk of non-Hodgkin’s lymphoma overall and with selected non-Hodgkin’s lymphoma subtypes are presented in table 1. There was no evidence of an association between height and non-Hodgkin’s lymphoma, B-CLL, or any of the non-Hodgkin’s lymphoma subtypes, with the possible exception of small lymphocytic lymphoma. For small lymphocytic lymphoma, there was an inverse association with height, but none of the point estimates was statistically significant, and there was no evidence of a dose-response relation. There was no evidence of a dose-response relation between baseline weight, body mass index, and waist/hip ratio and risk of all non-Hodgkin’s lymphoma or of diffuse or follicular lymphoma. For small lymphocytic lymphoma and B-CLL, there were also no statistically significant associations although, for weight and body mass index, relative risks were below one for small lymphocytic lymphoma and above one for B-CLL. There was no age-adjusted association between the level of physical activity in 1986 and the risk non-Hodgkin’s lymphoma or B-CLL (table 1), with the possible exception of follicular lymphoma. For follicular lymphoma, there was a suggestive inverse association with the level of physical activity (p trend = 0.1), although this association attenuated after adjustment for other non-Hodgkin’s lymphoma risk factors (relative risk (RR)moderate vs. high activity = 1.4, 95 percent confidence interval (CI): 0.6, 3.0;RRlow vs. high activity = 1.6, 95 percent CI: 0.8, 3.3; p trend = 0.2). The age-adjusted association of historical anthropometric characteristics and lymphoma risk is presented in table 2. For all non-Hodgkin’s lymphoma and for diffuse and follicular non-Hodgkin’s lymphoma, there was little evidence for associations with relative weight at age 12 years and body mass index at ages 18, 30, 40, or 50 years of age. These results did not change after multivariate adjustment (data not shown). For relative weight at age 12 years, there were associations for both small lymphocytic lymphoma and B-CLL. For small lymphocytic lymphoma, women who reported above average weight at age 12 years were at increased risk relative to women who were average weight at age 12 (RR = 2.4, 95 percent CI: 1.0, 5.8), as were women who were below average weight at age 12 (RR = 1.7, 95 percent CI: 0.8, 3.9). In contrast, for B-CLL, compared with women who were of average weight at age 12 years, women who were of below average weight were at a slightly lower risk (RR = 0.7, 95 percent CI: 0.3, 1.3) while women who were of above average weight were at slightly elevated risk (RR = 1.5, 95 percent CI: 0.8, 3.0), and the trend across these categories was statistically significant (p = 0.05). These associations remained unchanged after multivariate adjustment (table 3). For body mass index at age 18, 30, 40, or 50 years of age, there was little evidence of an association with small lymphocytic lymphoma, although at age 40 and at age 50 years, the age-adjusted relative risks were all below one. In contrast, for B-CLL at these ages, there were trends toward increasing risk with higher body mass index, particularly for body mass index at older ages. Multivariate adjustment had little effect on these results, although for small lymphocytic lymphoma, the associations strengthened slightly. Of note, there was generally an opposite pattern in the association of these anthropometric factors and risk of small lymphocytic lymphoma versus B-CLL, particularly for body mass index at age 50 years (table 3). These opposing trends were also apparent, but weaker, for weight and body mass index at baseline. DISCUSSION This is one of the few prospective cohort studies to evaluate the association of anthropometric variables or physical activity with risk of non-Hodgkin’s lymphoma, and this is the only study to our knowledge to systematically assess this in the context of potential confounding by other non-Hodgkin’s lymphoma risk factors (including dietary factors) and to report results for the common subtypes of non-Hodgkin’s lymphoma. This is also the first study to specifically address B-CLL, a malignancy of clonal B cells recently classified with small lymphocytic lymphoma on the basis of a presumed common underlying pathobiology (5). In addition to the prospective design, other strengths of this study include a community setting, comprehensive follow-up of the cohort, use of a Surveillance, Epidemiology, and End Results cancer registry for case ascertainment, an extensive evaluation of anthropometric variables, use of a validated protocol to assess body fat distribution, adjustment for important confounders, and systematic classification of the non-Hodgkin’s lymphoma subtypes. There are also limitations. Most importantly, the sample size (and therefore the precision of our estimates) was limited for the rarer subtypes, particularly small lymphocytic lymphoma. Moreover, multiple comparisons were made, and the few results at p < 0.05 may be due to chance. The initial response to the baseline questionnaire was only 43 percent. However, although this may limit the generalizability of the results, the virtually complete follow-up of the cohort ensures a high internal validity. Furthermore, nonresponders to the baseline survey have also been followed, and although they had slightly higher rates of smoking-related diseases, the association of body mass index with cancer occurrence (all sites and breast, endometrial, colon, and lung individually) was not appreciably affected by nonresponse bias (36). In addition, the anthropometric (37) and physical activity (26) variables have been associated with other morbidity or mortality endpoints in this study, consistent with a majority of the literature. The results of this study, however, may not generalize to males or non-Caucasian populations. Adult weight and body mass We found no evidence of an association of weight or body mass, either at study baseline or during adulthood, with the risk of non-Hodgkin’s lymphoma overall. This is consistent with three other incidence studies (7, 38, 39), a mortality study (40), and a hospital-based case-control study (41). In contrast, a Swedish cohort study (11) found that women with a hospital diagnosis of obesity were at excess risk of developing non-Hodgkin’s lymphoma (43 observed cases; RR = 1.6, 95 percent CI: 1.2, 2.1), although there was a weak inverse association for men (nine observed cases; standardized incidence ratio = 0.7, 95 percent CI: 0.3, 1.3). The only other study reporting an association with obesity was a large population-based case-control study conducted in San Francisco, California (10). Among heterosexuals, there was a strong, graded, positive association between body mass index and the risk of non-Hodgkin’s lymphoma for both men (odds ratio = 2.6 for a body mass index of ≥30 vs. <20, 95 percent CI: 1.2, 5.7) and women (odds ratio = 3.1 for a body mass index of ≥30 vs. <20, 95 percent CI: 1.4, 6.8). The latter associations remained after adjustment for several medical history and lifestyle factors. Associations by non-Hodgkin’s lymphoma subtype were not reported. When we repeated our analyses using the same cutpoints as the latter study, we still found no association (data not reported). In the San Francisco case-control study, 21 percent of the eligible patients died before they could be contacted, and another 10 percent were too ill to be interviewed. Since survival (required to get into the study) could be related to body mass, we excluded all cases in our study who died within 9 months of diagnosis to see if this might explain the discrepancy between our studies. After this exclusion, however, we still found no association between body mass index and the risk of non-Hodgkin’s lymphoma. We clearly had sufficient power to detect an association of the magnitude reported in the San Francisco study. Furthermore, it is unlikely that we were unable to detect an association because of exposure misclassification, as body mass index has been strongly correlated with breast and endometrial cancer incidence (42, 43) and coronary heart disease mortality (37) in this cohort. Although discrepancies remain, the balance of evidence published to date suggests no association of body mass with non-Hodgkin’s lymphoma. To our knowledge, no data have been reported on the association of adult obesity with the risk of non-Hodgkin’s lymphoma subtypes or of B-CLL. There are some published data on the association of body mass index with the risk of all leukemias, with studies showing both weak positive associations (38–40) and no association (11, 44). We found no evidence of an association of weight or body mass index with the risk of diffuse or follicular lymphoma. We did find suggestive inverse associations of weight, body mass index, and body mass index at ages 40 and 50 years with the risk of small lymphocytic lymphoma. In addition, we found suggestive positive associations of body mass index with the risk of B-CLL, particularly body mass indexes in later adulthood. However, most of these results were not statistically significant, and there was generally no clear dose-response trend, although the multivariate results for body mass index at age 50 years were statistically significant for both small lymphocytic lymphoma (inverse association, p trend = 0.04) and B-CLL (positive association, p trend = 0.03). It is somewhat surprising that the associations for small lymphocytic lymphoma and B-CLL are in opposite directions. This is based on the prediction that these two low-grade malignacies of B cells would be predicted to have similar risk factor profiles, because they may represent different clinical spectrums of closely related B-cell subpopulations. Indeed, their biologic and pathologic similarities have led to their classification together in the most recent World Health Organization classification system (5). Because of the small sample sizes in these analyses, however, our findings from this secondary analysis may be due to chance, and thus replication is needed. Height and adolescent obesity Adult height and adolescent obesity in part reflect the influence of nutritional factors during early childhood and adolescence on growth, with taller adult height and greater obesity reflecting more adequate (or excess) availability of food and energy and greater exposure to growth-related hormones. There are few data on the association of adolescent obesity with the risk of non-Hodgkin’s lymphoma and no data for non-Hodgkin’s lymphoma subtypes or B-CLL. The cohort studies of former students of Harvard University and the University of Pennsylvania found no association of obesity in college with the risk of either non-Hodgkin’s lymphoma or lymphatic leukemia (45). In our data, we also found little evidence of an important role for either relative weight at age 12 years or body mass index at age 18 and the risk of non-Hodgkin’s lymphoma overall or for diffuse or follicular non-Hodgkin’s lymphoma. There were, however, positive associations for B-CLL, consistent with the positive associations seen with obesity measures at other ages in this cohort. Relative weight at age 12 years was also associated with small lymphocytic lymphoma, but the increased risk was observed for both being underweight and being overweight relative to peers, although this was statistically significant only for being overweight. The subjective nature of the assessment of weight at age 12 years may have a higher potential for misclassification (both differential and nondifferential), which could bias our results. We found no association of height with the risk of all non-Hodgkin’s lymphoma, non-Hodgkin’s lymphoma subtypes, or B-CLL. There was no association of height with the risk of non-Hodgkin’s lymphoma or leukemia (all) in a cohort study of Icelanders (39), although other cohort studies have reported a positive association with non-Hodgkin’s lymphoma incidence (7), lymphoma mortality (44), and hematopoietic cancer mortality (12). The latter two studies were based on mortality data, had small numbers of events, and lacked adjustment for dietary factors, but the former study using incidence data from the Nurses’ Health Study (7) did not have these limitations and is comparable in design to this study. In the Nurses’ Health Study, women who were ≥68 inches (1.73 m) were at a 2.4-fold increased risk of non-Hodgkin’s lymphoma (95 percent CI: 1.2, 4.7) compared with women who were <62 inches (1.57 m), and there was a clear gradient in risk with increasing height. Neither our study nor the Nurses’ Health Study evaluated the association of height with non-Hodgkin’s lymphoma mortality, but the magnitude of associations found in cohort studies of lymphoma or hematopoietic mortality (12, 44) is compatible with the magnitude of the effect seen in the incidence data from the Nurses’ Health Study. Our incidence data, in contrast, are not compatible with these other studies, and more data based on incidence will be required to clarify the validity of this association. No prior study has published data on the association of height with non-Hodgkin’s lymphoma subtypes or with B-CLL specifically. Body fat distribution To our knowledge, there are no data on body fat distribution and the risk of non-Hodgkin’s lymphoma or B-CLL; we found that this is unlikely to be an important risk factor in the etiology of these malignancies overall or for non-Hodgkin’s lymphoma subtypes. Physical activity Overall, we found no association between the level of leisure-time physical activity and the risk of non-Hodgkin’s lymphoma or B-CLL. A limitation of these data is the lack of occupational physical activity. Although we did observe a suggestive positive association of physical inactivity with the risk of follicular non-Hodgkin’s lymphoma, this association attenuated after multivariate adjustment. A larger study is needed to better characterize this association. Occupational physical activity was not associated with either non-Hodgkin’s lymphoma or leukemia in a registry-based case-control study (other cancers served as controls) (22), and physical activity in college was not associated with the risk of non-Hodgkin’s lymphoma or lymphatic leukemia in two cohort studies of college alumni (21, 23). Physical activity, even modest levels, has been most strongly associated with a reduced risk of colon cancer and cancers of the female reproductive tract (18, 20), and the mechanistic basis for these observations includes control of body size, changes in steroid hormone metabolism, changes in menstrual function (females), or modulation of immune function. As reviewed above, a majority of studies have not found an association of non-Hodgkin’s lymphoma with obesity, suggesting that weight control is unlikely to be an important mechanism. Menstrual characteristics do not appear to be important risk factors for non-Hodgkin’s lymphoma, at least in this cohort (46). Although there are some animal and human data showing immunologic effects of both moderate and strenuous physical activity (18, 19), persons with primary or acquired immunodeficiency do not develop colon and female reproductive tract cancers but, rather, skin cancer, lymphomas, Kaposi′s sarcoma, and other rare or unusual tumors (47). Thus, the overall lack of association of physical activity with the risk of non-Hodgkin’s lymphoma (particularly diffuse lymphoma), a cancer whose development is thought to be correlated with the host immune status, suggests that immunomodulation may not be a major mechanistic basis for the protective association of physical activity with the development of cancer in general. Conclusions We found no evidence that height, weight, body mass, or physical activity plays an important role in the development of non-Hodgkin’s lymphoma overall or of diffuse or follicular non-Hodgkin’s lymphoma in particular. The inverse associations of several body mass index measures with small lymphocytic non-Hodgkin’s lymphoma, but positive associations with B-CLL, are novel. However, the latter associations must be interpreted with caution because of the small sample size, and they clearly require replication. Given the generally null findings from the literature and in this study, as well as weak data for a major role of obesity or physical activity in immune modulation, these factors are unlikely to have a critical role in the etiology of non-Hodgkin’s lymphoma. In addition, the more strongly documented correlates between diet and the risk of non-Hodgkin’s lymphoma (6, 7, 48, 49) indicate that future research will likely be more productive in evaluating the role of specific dietary factors in the etiology of these malignancies. ACKNOWLEDGMENTS The Iowa Women’s Health Study was supported by National Cancer Institute grant R01 CA39741. Dr. Cerhan was supported in part by a National Cancer Institute Preventive Oncology Award (K07 CA64220). The authors thank Ching-Ping Hong for technical assistance and Mary Jo Janisch for assistance in preparing this manuscript. Reprint requests to Dr. James R. Cerhan, Department of Health Sciences Research, Mayo Clinic, 200 First Street SW, Rochester, MN 55905 (e-mail: cerhan.james@mayo.edu). TABLE 1. Age-adjusted relative risks and 95% confidence intervals of non-Hodgkin’s lymphoma, selected non-Hodgkin’s lymphoma subtypes, and B-cell chronic lymphocytic leukemia in relation to baseline anthropometric characteristics and physical activity, Iowa Women’s Health Study, 1986–1998 Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Height (m)                                          <1.58  121,049  75  1      36  1      12  1      13  1      17  1    1.58–1.62  64,782  29  0.7  0.5, 1.1    18  0.9  0.5, 1.6    6  0.9  0.4, 2.5    3  0.4  0.1, 1.5    9  1.0  0.4, 2.2  1.63–1.67  133,575  76  0.9  0.7, 1.3    36  0.9  0.6, 1.5    23  1.8  0.9, 3.6    8  0.6  0.2, 1.4    17  0.9  0.5, 1.8  ≥1.68  130,396  81  1.1  0.8, 1.4    47  1.3  0.8, 2.0    17  1.4  0.7, 2.9    8  0.6  0.2, 1.4    20  1.2  0.6, 2.2                                            Weight (kg)                                          <59.9  108,647  68  1      29  1      15  1      12  1      13  1    59.9–67.1  117,358  60  0.8  0.6, 1.2    32  1.0  0.6, 1.7    17  1.1  0.5, 2.1    6  0.5  0.2, 1.2    19  1.4  0.7, 2.7  67.2–77.1  111,440  58  0.8  0.6, 1.2    37  1.2  0.8, 2.0    7  0.5  0.2, 1.1    6  0.5  0.2, 1.3    14  1.1  0.5, 2.2  ≥77.2  112,356  75  1.1  0.8, 1.5    39  1.3  0.8, 2.1    19  1.2  0.6, 2.4    8  0.7  0.3, 1.6    17  1.3  0.6, 2.7                                            Body mass index (kg/m2)                                          <23.5  109,932  71  1      32  1      14  1      12  1      13  1    23.5–26.1  115,079  51  0.7  0.5, 1.0    24  0.7  0.4, 1.2    16  1.1  0.5, 2.2    5  0.4  0.1, 1.1    14  1.0  0.5, 2.2  26.2–29.7  113,205  66  0.9  0.6, 1.2    43  1.3  0.8, 2.0    9  0.6  0.3, 1.4    8  0.6  0.3, 1.6    19  1.4  0.7, 2.8  ≥29.8  111,586  73  1.0  0.7, 1.4    38  1.2  0.7, 1.9    19  1.3  0.7, 2.6    7  0.6  0.2, 1.5    17  1.3  0.6, 2.6                                            Waist/hip ratio                                          <0.777  113,427  56  1      28  1      12  1      6  1      13  1    0.777–0.831  112,834  54  0.9  0.6, 1.4    28  1.0  0.6, 1.6    12  1.0  0.4, 2.2    5  0.8  0.3, 2.7    19  1.4  0.7, 2.8  0.832–0.892  111,562  83  1.4  1.0, 2.0    45  1.5  1.0, 2.5    18  1.5  0.7, 3.0    11  1.8  0.7, 4.9    16  1.2  0.6, 2.4  ≥0.893  110,067  66  1.1  0.8, 1.6    35  1.2  0.7, 1.9    16  1.3  0.6, 2.7    9  1.5  0.5, 4.2    15  1.1  0.5, 2.2                                            Physical activity index                                          High  111,319  58  1      34  1      10  1      8  1      13  1    Moderate  122,512  66  1.0  0.7, 1.5    34  0.9  0.6, 1.5    15  1.4  0.6, 3.1    8  0.9  0.3, 2.4    22  1.6  0.8, 3.1  Low  206,920  128  1.2  0.9, 1.7    63  1.0  0.7, 1.6    32  1.8  0.9, 3.6    15  1.0  0.4, 2.4    26  1.1  0.6, 2.2  Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Height (m)                                          <1.58  121,049  75  1      36  1      12  1      13  1      17  1    1.58–1.62  64,782  29  0.7  0.5, 1.1    18  0.9  0.5, 1.6    6  0.9  0.4, 2.5    3  0.4  0.1, 1.5    9  1.0  0.4, 2.2  1.63–1.67  133,575  76  0.9  0.7, 1.3    36  0.9  0.6, 1.5    23  1.8  0.9, 3.6    8  0.6  0.2, 1.4    17  0.9  0.5, 1.8  ≥1.68  130,396  81  1.1  0.8, 1.4    47  1.3  0.8, 2.0    17  1.4  0.7, 2.9    8  0.6  0.2, 1.4    20  1.2  0.6, 2.2                                            Weight (kg)                                          <59.9  108,647  68  1      29  1      15  1      12  1      13  1    59.9–67.1  117,358  60  0.8  0.6, 1.2    32  1.0  0.6, 1.7    17  1.1  0.5, 2.1    6  0.5  0.2, 1.2    19  1.4  0.7, 2.7  67.2–77.1  111,440  58  0.8  0.6, 1.2    37  1.2  0.8, 2.0    7  0.5  0.2, 1.1    6  0.5  0.2, 1.3    14  1.1  0.5, 2.2  ≥77.2  112,356  75  1.1  0.8, 1.5    39  1.3  0.8, 2.1    19  1.2  0.6, 2.4    8  0.7  0.3, 1.6    17  1.3  0.6, 2.7                                            Body mass index (kg/m2)                                          <23.5  109,932  71  1      32  1      14  1      12  1      13  1    23.5–26.1  115,079  51  0.7  0.5, 1.0    24  0.7  0.4, 1.2    16  1.1  0.5, 2.2    5  0.4  0.1, 1.1    14  1.0  0.5, 2.2  26.2–29.7  113,205  66  0.9  0.6, 1.2    43  1.3  0.8, 2.0    9  0.6  0.3, 1.4    8  0.6  0.3, 1.6    19  1.4  0.7, 2.8  ≥29.8  111,586  73  1.0  0.7, 1.4    38  1.2  0.7, 1.9    19  1.3  0.7, 2.6    7  0.6  0.2, 1.5    17  1.3  0.6, 2.6                                            Waist/hip ratio                                          <0.777  113,427  56  1      28  1      12  1      6  1      13  1    0.777–0.831  112,834  54  0.9  0.6, 1.4    28  1.0  0.6, 1.6    12  1.0  0.4, 2.2    5  0.8  0.3, 2.7    19  1.4  0.7, 2.8  0.832–0.892  111,562  83  1.4  1.0, 2.0    45  1.5  1.0, 2.5    18  1.5  0.7, 3.0    11  1.8  0.7, 4.9    16  1.2  0.6, 2.4  ≥0.893  110,067  66  1.1  0.8, 1.6    35  1.2  0.7, 1.9    16  1.3  0.6, 2.7    9  1.5  0.5, 4.2    15  1.1  0.5, 2.2                                            Physical activity index                                          High  111,319  58  1      34  1      10  1      8  1      13  1    Moderate  122,512  66  1.0  0.7, 1.5    34  0.9  0.6, 1.5    15  1.4  0.6, 3.1    8  0.9  0.3, 2.4    22  1.6  0.8, 3.1  Low  206,920  128  1.2  0.9, 1.7    63  1.0  0.7, 1.6    32  1.8  0.9, 3.6    15  1.0  0.4, 2.4    26  1.1  0.6, 2.2  * NHL, non-Hodgkin’s lymphoma; B-CLL, B-cell chronic lymphocytic leukemia; RR, relative risk; CI, confidence interval. † All relative risks adjusted for age. View Large TABLE 2. Age-adjusted relative risks and 95% confidence intervals of non-Hodgkin’s lymphoma, selected non-Hodgkin’s lymphoma subtypes, and B-cell chronic lymphocytic leukemia in relation to historical anthropometric characteristics, Iowa Women’s Health Study, 1986–1998 Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Relative weight at age 12 years                                          Below average  98,034  63  1.2  0.9, 1.6    35  1.3  0.9, 1.9    12  1.0  0.5, 1.8    9  1.7  0.8, 3.9    9  0.7  0.3, 1.3  Average  280,767  153  1      78  1      36  1      15  1      40  1    Above average  57,022  38  1.3  0.9, 1.8    19  1.3  0.8, 2.1    9  1.3  0.6, 2.6    7  2.4  1.0, 5.8    12  1.5  0.8, 3.0                                            Body mass index (kg/m2) at age 18 years                                          <19.6  111,383  64  1      27  1      21  1      10        13  1    19.6–21.1  112,679  53  0.8  0.6, 1.2    32  1.2  0.7, 2.0    8  0.4  0.2, 0.9    4  0.4  0.1, 1.3    16  1.2  0.6, 2.6  21.2–23.0  117,534  72  1.1  0.8, 1.5    41  1.5  0.9, 2.4    13  0.6  0.3, 1.2    7  0.7  0.3, 1.8    13  1.0  0.5, 2.1  ≥23.1  106,153  69  1.2  0.8, 1.6    34  1.4  0.8, 2.3    16  0.8  0.4, 1.6    11  1.2  0.5, 2.8    21  1.7  0.9, 3.5                                            Body mass index (kg/m2) at age 30 years                                          <21.1  106,838  58  1      28  1      17  1      8  1      9  1    21.1–22.7  114,610  69  1.1  0.8, 1.6    36  1.2  0.7, 2.0    15  0.8  0.4, 1.6    7  0.8  0.3, 2.2    21  2.2  1.0, 4.7  22.8–24.8  115,176  66  1.0  0.7, 1.5    37  1.2  0.7, 2.0    9  0.5  0.2, 1.1    8  0.9  0.3, 2.5    14  1.4  0.6, 3.3  ≥24.9  105,540  61  1.1  0.7, 1.5    30  1.1  0.7, 1.8    17  1.0  0.5, 2.0    8  1.0  0.4, 2.7    18  2.0  0.9, 4.5                                            Body mass index (kg/m2) at age 40 years                                          <22.1  109,827  61  1      29  1      14  1      10  1      11  1    22.1–23.9  117,838  73  1.1  0.8, 1.6    35  1.1  0.7, 1.8    16  1.1  0.5, 2.2    7  0.6  0.2, 1.7    16  1.3  0.6, 2.9  24.0–26.4  109,119  51  0.8  0.6, 1.2    30  1.0  0.6, 1.7    11  0.8  0.4, 1.7    6  0.6  0.2, 1.7    16  1.4  0.7, 3.1  ≥26.5  105,396  69  1.2  0.8, 1.7    37  1.3  0.8, 2.2    17  1.3  0.6, 2.6    8  0.8  0.3, 2.1    19  1.8  0.9, 3.8                                            Body mass index (kg/m2) at age 50 years                                          <22.9  111,104  67  1      28  1      16  1      11  1      9  1    22.9–25.0  116,883  67  0.9  0.7, 1.3    40  1.3  0.8, 2.2    11  0.6  0.3, 1.4    7  0.6  0.2, 1.6    18  1.9  0.8, 4.2  25.1–28.2  109,549  57  0.9  0.6, 1.2    29  1.0  0.6, 1.7    16  1.0  0.5, 2.0    8  0.7  0.3, 1.8    13  1.4  0.6, 3.4  ≥28.3  111,110  69  1.0  0.7, 1.5    39  1.4  0.9, 2.3    15  0.9  0.5, 1.9    6  0.6  0.2, 1.5    23  2.6  1.2, 5.6  Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Relative weight at age 12 years                                          Below average  98,034  63  1.2  0.9, 1.6    35  1.3  0.9, 1.9    12  1.0  0.5, 1.8    9  1.7  0.8, 3.9    9  0.7  0.3, 1.3  Average  280,767  153  1      78  1      36  1      15  1      40  1    Above average  57,022  38  1.3  0.9, 1.8    19  1.3  0.8, 2.1    9  1.3  0.6, 2.6    7  2.4  1.0, 5.8    12  1.5  0.8, 3.0                                            Body mass index (kg/m2) at age 18 years                                          <19.6  111,383  64  1      27  1      21  1      10        13  1    19.6–21.1  112,679  53  0.8  0.6, 1.2    32  1.2  0.7, 2.0    8  0.4  0.2, 0.9    4  0.4  0.1, 1.3    16  1.2  0.6, 2.6  21.2–23.0  117,534  72  1.1  0.8, 1.5    41  1.5  0.9, 2.4    13  0.6  0.3, 1.2    7  0.7  0.3, 1.8    13  1.0  0.5, 2.1  ≥23.1  106,153  69  1.2  0.8, 1.6    34  1.4  0.8, 2.3    16  0.8  0.4, 1.6    11  1.2  0.5, 2.8    21  1.7  0.9, 3.5                                            Body mass index (kg/m2) at age 30 years                                          <21.1  106,838  58  1      28  1      17  1      8  1      9  1    21.1–22.7  114,610  69  1.1  0.8, 1.6    36  1.2  0.7, 2.0    15  0.8  0.4, 1.6    7  0.8  0.3, 2.2    21  2.2  1.0, 4.7  22.8–24.8  115,176  66  1.0  0.7, 1.5    37  1.2  0.7, 2.0    9  0.5  0.2, 1.1    8  0.9  0.3, 2.5    14  1.4  0.6, 3.3  ≥24.9  105,540  61  1.1  0.7, 1.5    30  1.1  0.7, 1.8    17  1.0  0.5, 2.0    8  1.0  0.4, 2.7    18  2.0  0.9, 4.5                                            Body mass index (kg/m2) at age 40 years                                          <22.1  109,827  61  1      29  1      14  1      10  1      11  1    22.1–23.9  117,838  73  1.1  0.8, 1.6    35  1.1  0.7, 1.8    16  1.1  0.5, 2.2    7  0.6  0.2, 1.7    16  1.3  0.6, 2.9  24.0–26.4  109,119  51  0.8  0.6, 1.2    30  1.0  0.6, 1.7    11  0.8  0.4, 1.7    6  0.6  0.2, 1.7    16  1.4  0.7, 3.1  ≥26.5  105,396  69  1.2  0.8, 1.7    37  1.3  0.8, 2.2    17  1.3  0.6, 2.6    8  0.8  0.3, 2.1    19  1.8  0.9, 3.8                                            Body mass index (kg/m2) at age 50 years                                          <22.9  111,104  67  1      28  1      16  1      11  1      9  1    22.9–25.0  116,883  67  0.9  0.7, 1.3    40  1.3  0.8, 2.2    11  0.6  0.3, 1.4    7  0.6  0.2, 1.6    18  1.9  0.8, 4.2  25.1–28.2  109,549  57  0.9  0.6, 1.2    29  1.0  0.6, 1.7    16  1.0  0.5, 2.0    8  0.7  0.3, 1.8    13  1.4  0.6, 3.4  ≥28.3  111,110  69  1.0  0.7, 1.5    39  1.4  0.9, 2.3    15  0.9  0.5, 1.9    6  0.6  0.2, 1.5    23  2.6  1.2, 5.6  * NHL, non-Hodgkin’s lymphoma; B-CLL, B-cell chronic lymphocytic leukemia; RR, relative risk; CI, confidence interval. † All relative risks adjusted for age. View Large TABLE 3. Multivariate-adjusted relative risks and 95% confidence intervals of small lymphocytic lymphoma and B-cell chronic lymphocytic leukemia in relation to selected anthropometric characteristics, Iowa Women’s Health Study, 1986–1998   Small lymphocytic    B-CLL*  RR*,†  95% CI*    RR  95% CI  Relative weight at age 12 years            Below average  1.5  0.6, 3.8    0.6  0.3, 1.3  Average  1      1    Above average  2.5  1.0, 6.3    1.5  0.8, 2.9    p trend = 0.4    p trend = 0.05  Weight (kg)            <59.9  1      1    59.9–67.1  0.4  0.2, 1.2    1.5  0.7, 3.0  67.2–77.1  0.3  0.1, 0.9    1.2  0.5, 2.6  ≥77.2  0.4  0.1, 1.0    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.7  Body mass index (kg/m2) at age 50 years            <22.9  1      1    22.9–25.0  0.6  0.2, 1.7    1.9  0.8, 4.3  25.1–28.2  0.6  0.2, 1.6    1.5  0.6, 3.5  ≥28.3  0.3  0.1, 0.9    2.7  1.2, 6.0    p trend = 0.04    p trend = 0.03  Body mass index (kg/m2)            <23.5  1      1    23.5–26.1  0.4  0.1, 1.1    1.1  0.5, 2.4  26.2–29.7  0.4  0.2, 1.2    1.6  0.8, 3.3  ≥29.8  0.3  0.1, 0.9    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.4    Small lymphocytic    B-CLL*  RR*,†  95% CI*    RR  95% CI  Relative weight at age 12 years            Below average  1.5  0.6, 3.8    0.6  0.3, 1.3  Average  1      1    Above average  2.5  1.0, 6.3    1.5  0.8, 2.9    p trend = 0.4    p trend = 0.05  Weight (kg)            <59.9  1      1    59.9–67.1  0.4  0.2, 1.2    1.5  0.7, 3.0  67.2–77.1  0.3  0.1, 0.9    1.2  0.5, 2.6  ≥77.2  0.4  0.1, 1.0    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.7  Body mass index (kg/m2) at age 50 years            <22.9  1      1    22.9–25.0  0.6  0.2, 1.7    1.9  0.8, 4.3  25.1–28.2  0.6  0.2, 1.6    1.5  0.6, 3.5  ≥28.3  0.3  0.1, 0.9    2.7  1.2, 6.0    p trend = 0.04    p trend = 0.03  Body mass index (kg/m2)            <23.5  1      1    23.5–26.1  0.4  0.1, 1.1    1.1  0.5, 2.4  26.2–29.7  0.4  0.2, 1.2    1.6  0.8, 3.3  ≥29.8  0.3  0.1, 0.9    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.4  * B-CLL, B-cell chronic lymphocytic leukemia; RR, relative risk; CI, confidence interval. † All relative risks adjusted for age, marital status, farm residence, diabetes, transfusion history, hormone replacement therapy, alcohol consumption, smoking history, red meat consumption, and fruit consumption. View Large References 1. Greenlee RT, Hill-Harmon MB, Murray T, et al. Cancer statistics, 2001. CA Cancer J Clin   2001; 51: 15–36. Google Scholar 2. Ries LG, Eisner MP, Kosary CL, et al. SEER cancer statistics review, 1973–1998. Bethesda, MD: National Cancer Institute, 2001. Google Scholar 3. Devesa SS, Fears T. Non-Hodgkin’s lymphoma time trends: United States and international data. Cancer Res  1992; 52(suppl): 5432s–40s. Google Scholar 4. Hartge P, Devesa SS. Quantification of the impact of known risk factors on time trends in non-Hodgkin’s lymphoma incidence. Cancer Res   1992; 52(suppl): 5566s–9s. Google Scholar 5. Jaffe ES, Harris NL, Stein H, et al. World Health Organization classification of tumours. Pathology and genetics of tumours of haematopoietic and lymphoid tissues. Lyon, France: IARC Press, 2001. Google Scholar 6. Chiu BC, Cerhan JR, Folsom AR, et al. Diet and risk of non-Hodgkin lymphoma in older women. JAMA   1996; 275: 1315–21. Google Scholar 7. Zhang S, Hunter DJ, Rosner BA, et al. Dietary fat and protein in relation to risk of non-Hodgkin’s lymphoma among women. J Natl Cancer Inst  1999; 91: 1751–8. Google Scholar 8. Cunningham AS. Lymphomas and animal-protein consumption. Lancet  1976; 2: 1184–6. Google Scholar 9. Tavani A, Pregnolato A, Negri E, et al. Diet and risk of lymphoid neoplasms and soft tissue sarcomas. Nutr Cancer  1997; 27: 256–60. Google Scholar 10. Holly EA, Lele C, Bracci PM, et al. Case-control study of non-Hodgkin’s lymphoma among women and heterosexual men in the San Francisco Bay Area, California. Am J Epidemiol  1999; 150: 375–89. Google Scholar 11. Wolk A, Gridley G, Svensson M, et al. A prospective study of obesity and cancer risk (Sweden). Cancer Causes Control  2001; 12: 13–21. Google Scholar 12. Davey Smith G, Hart C, Upton M, et al. Height and risk of death among men and women: aetiological implications of associations with cardiorespiratory disease and cancer mortality. J Epidemiol Community Health  2000; 54: 97–103. Google Scholar 13. Stallone DD. The influence of obesity and its treatment on the immune system. Nutr Rev  1994; 52: 37–50. Google Scholar 14. Cartwright RA, McKinney PA, O’Brien C, et al. Non-Hodgkin’s lymphoma: a case-control epidemiologic study in Yorkshire. Leuk Res  1988; 13: 465–72. Google Scholar 15. Cerhan JR, Wallace RB, Folsom AR, et al. Medical history risk factors for non-Hodgkin’s lymphoma in older women. J Natl Cancer Inst  1997; 89: 314–18. Google Scholar 16. Hjalgrim H, Frisch M, Ekbom A, et al. Cancer and diabetes—a follow-up study of two population-based cohorts of diabetic patients. J Intern Med  1997; 241: 471–5. Google Scholar 17. Weiderpass E, Gridley G, Ekbom A, et al. Medical history risk factors for non-Hodgkin’s lymphoma in older women. J Natl Cancer Inst  1997; 89: 816–17. Google Scholar 18. Shephard RJ, Shek PN. Cancer, immune function, and physical activity. Can J Appl Physiol  1995; 20: 1–25. Google Scholar 19. Lee IM. Exercise and physical health: cancer and immune function. Res Q Exerc Sport  1995; 66: 286–91. Google Scholar 20. Friedenreich CM. Physical activity and cancer prevention: from observational to intervention research. Cancer Epidemiol Biomarkers Prev  2001; 10: 287–301. Google Scholar 21. Polednak AP. College athletics, body size, and cancer mortality. Cancer  1976; 38: 382–7. Google Scholar 22. Brownson RC, Chang JC, Davis JR, et al. Physical activity on the job and cancer in Missouri. Am J Public Health  1991; 81: 639–42. Google Scholar 23. Paffenbarger RS Jr, Lee IM, Wing AL. The influence of physical activity on the incidence of site-specific cancers in college alumni. Adv Exp Med Biol  1992; 322: 7–15. Google Scholar 24. Cerhan JR, Wallace RB, Dick F, et al. Blood  transfusion and risk of non-Hodgkin lymphoma subtypes and chronic lymphocytic leukemia. Cancer Epidemiol Biomarkers Prev  2001; 10: 361–8. Google Scholar 25. Kushi LH, Kaye SA, Folsom AR, et al. Accuracy and reliability of self-measurement of body girths. Am J Epidemiol  1988; 128: 740–8. Google Scholar 26. Kushi LH, Fee RM, Folsom AR, et al. Physical activity and mortality in postmenopausal women. JAMA  1997; 277: 1287–92. Google Scholar 27. Percy C, Van Holten V, Muir C. International classification of diseases for oncology. 2nd ed. Geneva, Switzerland: World Health Organization, 1990. Google Scholar 28. Percy C, O’Conor G, Ries L, et al. Non-Hodgkin’s lymphomas: application of the international classification of diseases for oncology (ICD-O) to the Working Formulation. Cancer  1984; 54: 1435–8. Google Scholar 29. Groves FD, Linet MS, Travis LB, et al. Cancer surveillance series: non-Hodgkin’s lymphoma incidence by histologic subtype in the United States from 1978 through 1995. J Natl Cancer Inst  2000; 92: 1240–51. Google Scholar 30. Cox DR. Regression models and life tables (with discussion). J R Stat Soc (B)  1972; 34: 187–220. Google Scholar 31. Korn EL, Graubard BI, Midthune D. Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale. Am J Epidemiol  1997; 145: 72–80. Google Scholar 32. Folsom AR, Zhang S, Sellers TA, et al. Cancer incidence among women living on farms: findings from the Iowa Women’s Health Study. J Occup Environ Med  1996; 38: 1171–6. Google Scholar 33. Cerhan JR. New epidemiologic leads in the etiology of non-Hodgkin lymphoma in older persons: the role of blood transfusion and diet. Biomed Pharmacother  1997; 51: 200–7. Google Scholar 34. Chiu BC, Cerhan JR, Gapstur SM, et al. Alcohol consumption and non-Hodgkin lymphoma in a cohort of older women. Br J Cancer  1999; 80: 1476–82. Google Scholar 35. Parker AS, Cerhan JR, Dick F, et al. Smoking and risk of non-Hodgkin’s lymphoma subtypes in a cohort of older Iowa women. Leuk Lymphoma  2000; 37: 341–9. Google Scholar 36. Bisgard KM, Folsom AR, Hong CP, et al. Mortality and cancer rates in nonrespondents to a prospective cohort study of older women: 5-year follow-up. Am J Epidemiol  1994; 139: 990–1000. Google Scholar 37. Folsom AR, Kushi LH, Anderson KE, et al. Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women’s Health Study. Arch Intern Med  2000; 160: 2117–28. Google Scholar 38. Moller H, Mellemgaard A, Lindvig K, et al. Obesity and cancer risk: a Danish record-linkage study. Eur J Cancer  1994; 3: 344–50. Google Scholar 39. Tulinius H, Sigfusson N, Sigvaldason H, et al. Risk factors for malignant diseases: a cohort study on a population of 22,946 Icelanders. Cancer Epidemiol Biomarkers Prev   1997; 6: 863–73. Google Scholar 40. Lew EA, Garfinkel L. Variations in mortality by weight among 750,000 men and women. J Chronic Dis  1979; 32: 563–76. Google Scholar 41. Franceschi S, Serraino D, Bidoli E, et al. The epidemiology of non-Hodgkin’s lymphoma in the north-east of Italy: a hospital-based case-control study. Leuk Res  1989; 13: 465–72. Google Scholar 42. Folsom AR, Kaye SA, Potter JD, et al. Association of incident carcinoma of the endometrium with body weight and fat distribution in older women: early findings of the Iowa Women’s Health Study. Cancer Res  1989; 49: 6828–31. Google Scholar 43. Folsom AR, Kaye SA, Prineas RJ, et al. Increased incidence of carcinoma of the breast associated with abdominal adiposity in postmenopausal women. Am J Epidemiol  1990; 131: 794–803. Google Scholar 44. Leon DA, Smith GD, Shipley M, et al. Adult height and mortality in London: early life, socioeconomic confounding, or shrinkage? J Epidemiol Community Health  1995; 49: 5–9. Google Scholar 45. Whittemore AS, Paffenbarger RS Jr, Anderson K, et al. Early precursors of site-specific cancers in college men and women. J Natl Cancer Inst  1985; 74: 43–51. Google Scholar 46. Cerhan JR, Habermann TM, Vachon CM, et al. Menstrual and reproductive factors and risk of non-Hodgkin lymphoma: the Iowa Women’s Health Study. Cancer Causes Control  2002; 13: 131–6. Google Scholar 47. Schwartz R. Immunodeficiency, immunosuppression, and susceptibility to neoplasms. J Natl Cancer Inst Monogr  2000;(28): 5–9. Google Scholar 48. Ward MH, Zahm SH, Weisenburger DD, et al. Dietary factors and non-Hodgkin’s lymphoma in Nebraska (United States). Cancer Causes Control  1994; 5: 422–32. Google Scholar 49. Zhang SM, Hunter DJ, Rosner BA, et al. Intakes of fruits, vegetables, and related nutrients and the risk of non-Hodgkin’s lymphoma among women. Cancer Epidemiol Biomarkers Prev  2000; 9: 477–85. Google Scholar http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Epidemiology Oxford University Press

Anthropometric Characteristics, Physical Activity, and Risk of Non-Hodgkin’s Lymphoma Subtypes and B-Cell Chronic Lymphocytic Leukemia: A Prospective Study

Loading next page...
 
/lp/oxford-university-press/anthropometric-characteristics-physical-activity-and-risk-of-non-s60bcyjeuB

References (45)

Publisher
Oxford University Press
ISSN
0002-9262
eISSN
1476-6256
DOI
10.1093/aje/kwf082
Publisher site
See Article on Publisher Site

Abstract

Abstract Anthropometric characteristics, physical activity, and risk of non-Hodgkin’s lymphoma, its subtypes, and B-cell chronic lymphocytic leukemia (B-CLL) were evaluated in a prospective cohort study of 37,931 Iowa women who were aged 55–69 years at baseline in 1986. Through 1998, 261 cases of non-Hodgkin’s lymphoma (137 diffuse, 58 follicular, and 32 small lymphocytic lymphomas) and 63 cases of B-CLL were identified by linkage to the Iowa Cancer Registry. Height, weight, body mass index, waist/hip ratio, and physical activity were not associated with risk of non-Hodgkin’s lymphoma overall or with diffuse or follicular lymphoma in particular. After adjustment for other non-Hodgkin’s lymphoma risk factors, there was an inverse association of baseline body mass index (relative risks (RRs) across quartiles: 1, 0.4, 0.4, 0.3; p trend = 0.03) with risk of small lymphocytic lymphoma. In contrast, for B-CLL there were suggestive positive associations with body mass index at age 50 years (RRs across quartiles: 1, 1.9, 1.5, 2.7; p trend = 0.03) and (more weakly) baseline body mass index (RRs across quartiles: 1, 1.1, 1.6, 1.3; p trend = 0.3). In summary, we found no evidence that height, weight, body mass, or physical activity plays an important role in non-Hodgkin’s lymphoma overall or in diffuse or follicular lymphoma in particular. The opposite associations of body mass index with small lymphocytic lymphoma versus B-CLL may be a chance finding but, if confirmed, would suggest different etiologies for these malignancies. anthropometry; cohort studies; leukemia, lymphocytic, chronic; lymphoma, non-Hodgkin Abbreviations: B-CLL, B-cell chronic lymphocytic leukemia; CI, confidence interval; ICD-O-2, International Classification of Diseases for Oncology, Second Edition; RR, relative risk. Received for publication January 9, 2002; accepted for publication May 2, 2002. In 2001, the American Cancer Society estimated that 56,200 persons would be diagnosed with non-Hodgkin’s lymphoma and that 26,300 would die from it (1). Non-Hodgkin’s lymphoma is now the fifth most commonly diagnosed cancer among both men and women. The incidence of non-Hodgkin’s lymphoma in the United States increased 82.6 percent from 1973 to 1998, a percentage increase exceeded only by melanoma of the skin, prostate cancer, and liver and intrahepatic cancer (2). This increase is part of a longer trend documented since 1950 (3), and improved medical technology, changing systems of diagnostic classification, known risk factors, and the human immunodeficiency virus epidemic can explain only part of this increase (4). B-cell chronic lymphocytic leukemia (B-CLL) is the most common leukemia in the United States (2) and is currently classified together with small lymphocytic lymphoma by the World Health Organization (5). Diets high in fat (6, 7) or meat products (6–9) have been associated with risk of non-Hodgkin’s lymphoma. Several anthropometric measures, including body mass (10, 11) and height (7, 12), have also been suggested as non-Hodgkin’s lymphoma risk factors. Data from epidemiologic, clinical, and animal studies, albeit limited, provide some support for the hypothesis that obesity is associated with impairments of both humoral and cellular immunity (13). Obesity is also associated with insulin resistance and the development of adult-onset diabetes, and adult-onset diabetes has been shown in some studies to be a risk factor for non-Hodgkin’s lymphoma (14–17). Physical inactivity is an important determinant of obesity and appears to be a risk factor for several types of cancer, particularly colon cancer and female reproductive cancers (18, 19). The underlying mechanisms responsible for these associations with physical inactivity are not known but are thought to involve body weight, changes in steroid hormone levels (perhaps caused by changes in menstrual function for women), or modulation of immune function (18–20). The extremely limited data available suggest that there is no association between the level of physical activity and risk of non-Hodgkin’s lymphoma or lymphatic leukemias (21–23). There are no data on either anthropometric characteristics or physical activity and risk of B-CLL specifically. We therefore evaluated whether anthropometric factors and physical activity were associated with non-Hodgkin’s lymphoma or B-CLL in a large prospective cohort of Iowa women. In secondary analyses, we evaluated heterogeneity of the results by the most common subtypes of non-Hodgkin’s lymphoma. MATERIALS AND METHODS The Iowa Women’s Health Study cohort Details on the Iowa Women’s Health Study design specific to non-Hodgkin’s lymphoma and B-CLL have been previously published (24). In 1986, a questionnaire was returned by 41,836 randomly selected women who were aged 55–69 years and had a valid Iowa driver’s license in 1985 (42.7 percent response rate). The rates of non-Hodgkin’s lymphoma (overall and for the major subtypes) in the cohort are nearly identical to the rates from the Iowa Surveillance, Epidemiology, and End Results Cancer Registry (data available on request). Self-reported items on the 1986 baseline questionnaire included demographic data, medical history, and dietary and other lifestyle factors. Participants reported current height and weight and weight at ages 18, 30, 40, and 50 years. The body mass index at baseline and at each respective age was calculated as weight (in kilograms) at the given age divided by baseline height squared (in meters). A paper tape measure and written instructions for having a friend measure circumferences of the waist (1 inch (2.54 cm) above the umbilicus) and hips (maximal protrusion) were enclosed with the questionnaire. The waist/hip ratio was calculated from these measures. This protocol has been shown to be valid (correlation coefficient with measures by trained technician, ≥0.84) and reliable (correlation of measures at two different time periods, ≥0.85) (25). Relative weight at age 12 was ascertained using the following question: “Think back to when you were in sixth grade—or about the age of 12. Would you say at that time, compared to other girls your age, your weight was below average for your age and height, about average for your age and height, or above average for your age and height?” Physical activity was assessed by asking participants three questions about whether they participated in any leisure exercise and, if so, the frequency of moderate and vigorous activities. We did not have estimates of occupational physical activity. Responses to these questions were combined to create a three-level activity score (low, medium, high). This scale is strongly associated with total mortality and cardiovascular disease mortality in this cohort (26). Cohort follow-up Vital status and cancer incidence in the cohort were ascertained through 13 years of continuous follow-up (1986 through 1998). Follow-up questionnaires were mailed in 1987, 1989, 1992, and 1997 to ascertain vital status and address changes. Deaths were also ascertained by annual linkage to a database of Iowa death certificates, supplemented by linkage to the National Death Index for survey nonrespondents and emigrants from Iowa. Outmigration has been estimated at 1 percent per year. Cancer incidence except for nonmelanoma skin cancer was ascertained by annual linkage to the Iowa Cancer Registry, part of the National Cancer Institute’s Surveillance, Epidemiology, and End Results program (2). All participants were linked by a combination of Social Security number, first, last, and maiden name, birth date, and ZIP code. The Iowa Cancer Registry collects cancer data including identifying information, tumor site, morphology, histologic grade, and extent of disease on all persons who were Iowa residents at the time of their diagnosis. All tumor site and morphology data were derived from pathology reports of the diagnosing hematopathologist, and there was no centralized pathology review. Topographic and morphologic data were coded using the International Classification of Diseases for Oncology, Second Edition (ICD-O-2) (27), which for lymphomas was derived from the Working Formulation (28). The histologic subtypes of non-Hodgkin’s lymphoma were grouped according to the method of Groves et al. (29) into the following subtypes based on ICD-O-2 codes: small lymphocytic, follicular, diffuse, high grade, and peripheral T cell. This system is a modification of the Working Formulation for use in epidemiologic studies. Chronic lymphoctyic leukemia was also analyzed but as a distinct endpoint; that is, it was not aggregated with small lymphocytic lymphoma. Data analysis Before data analysis, we excluded women with a self-reported history of cancer or cancer chemotherapy on the 1986 baseline questionnaire (n = 3,902) to provide a cancer-free at-risk cohort. Each woman accumulated person-years of follow-up from the date of receipt of the 1986 baseline questionnaire until the date of non-Hodgkin’s lymphoma (or B-CLL) diagnosis, date of emigration from Iowa, or date of death; if none of these events occurred, person-years were accumulated through December 31, 1998. Two women who were censored at their baseline date were further excluded, leaving a final cohort of 37,932 for analysis. All continuous anthropometric variables were classified into quartiles based on the distribution in the at-risk cohort. Physical activity was grouped in low, moderate, and high levels (26). Relative weight at age 12 years was categorized according to possible responses (below average, average, above average), with the average group serving as the reference category. Age-adjusted and multivariate-adjusted relative risks, along with 95 percent confidence intervals, were calculated as measures of association of the risk factors with non-Hodgkin’s lymphoma incidence and were estimated using Cox proportional hazards regression (30), with the survival modeled as a function of age (31). Other risk factors for non-Hodgkin’s lymphoma in this data set included age, marital status, farm residence (32), transfusion history (33), adult-onset diabetes (33), alcohol use (34), cigarette smoking (35), red meat and fruit intake (6), and use of hormone replacement therapy, and they were evaluated in multivariate models as potential confounders. RESULTS The mean baseline age of the 37,932 women in the at-risk cohort was 61.7 years, and over 99 percent were White. During 449,801 person-years of follow-up (1986–1998), 261 women developed non-Hodgkin’s lymphoma and 63 women developed B-CLL. The mean age at diagnosis for non-Hodgkin’s lymphoma was 70.1 (range, 57–81) years and for B-CLL was 70.7 (range, 60–81) years. Approximately 70 percent of the non-Hodgkin’s lymphomas were nodal, and of the extranodal sites, the stomach, small intestine, and thyroid were the most common sites involved. The most common non-Hodgkin’s lymphoma subtype was diffuse lymphoma (53 percent), followed by follicular lymphoma (22 percent) and small lymphocytic lymphoma (12 percent); only 9 percent of the cases could not be classified into the scheme of Groves et al. (29). The age-adjusted associations of baseline anthropometric characteristics and of physical activity with risk of non-Hodgkin’s lymphoma overall and with selected non-Hodgkin’s lymphoma subtypes are presented in table 1. There was no evidence of an association between height and non-Hodgkin’s lymphoma, B-CLL, or any of the non-Hodgkin’s lymphoma subtypes, with the possible exception of small lymphocytic lymphoma. For small lymphocytic lymphoma, there was an inverse association with height, but none of the point estimates was statistically significant, and there was no evidence of a dose-response relation. There was no evidence of a dose-response relation between baseline weight, body mass index, and waist/hip ratio and risk of all non-Hodgkin’s lymphoma or of diffuse or follicular lymphoma. For small lymphocytic lymphoma and B-CLL, there were also no statistically significant associations although, for weight and body mass index, relative risks were below one for small lymphocytic lymphoma and above one for B-CLL. There was no age-adjusted association between the level of physical activity in 1986 and the risk non-Hodgkin’s lymphoma or B-CLL (table 1), with the possible exception of follicular lymphoma. For follicular lymphoma, there was a suggestive inverse association with the level of physical activity (p trend = 0.1), although this association attenuated after adjustment for other non-Hodgkin’s lymphoma risk factors (relative risk (RR)moderate vs. high activity = 1.4, 95 percent confidence interval (CI): 0.6, 3.0;RRlow vs. high activity = 1.6, 95 percent CI: 0.8, 3.3; p trend = 0.2). The age-adjusted association of historical anthropometric characteristics and lymphoma risk is presented in table 2. For all non-Hodgkin’s lymphoma and for diffuse and follicular non-Hodgkin’s lymphoma, there was little evidence for associations with relative weight at age 12 years and body mass index at ages 18, 30, 40, or 50 years of age. These results did not change after multivariate adjustment (data not shown). For relative weight at age 12 years, there were associations for both small lymphocytic lymphoma and B-CLL. For small lymphocytic lymphoma, women who reported above average weight at age 12 years were at increased risk relative to women who were average weight at age 12 (RR = 2.4, 95 percent CI: 1.0, 5.8), as were women who were below average weight at age 12 (RR = 1.7, 95 percent CI: 0.8, 3.9). In contrast, for B-CLL, compared with women who were of average weight at age 12 years, women who were of below average weight were at a slightly lower risk (RR = 0.7, 95 percent CI: 0.3, 1.3) while women who were of above average weight were at slightly elevated risk (RR = 1.5, 95 percent CI: 0.8, 3.0), and the trend across these categories was statistically significant (p = 0.05). These associations remained unchanged after multivariate adjustment (table 3). For body mass index at age 18, 30, 40, or 50 years of age, there was little evidence of an association with small lymphocytic lymphoma, although at age 40 and at age 50 years, the age-adjusted relative risks were all below one. In contrast, for B-CLL at these ages, there were trends toward increasing risk with higher body mass index, particularly for body mass index at older ages. Multivariate adjustment had little effect on these results, although for small lymphocytic lymphoma, the associations strengthened slightly. Of note, there was generally an opposite pattern in the association of these anthropometric factors and risk of small lymphocytic lymphoma versus B-CLL, particularly for body mass index at age 50 years (table 3). These opposing trends were also apparent, but weaker, for weight and body mass index at baseline. DISCUSSION This is one of the few prospective cohort studies to evaluate the association of anthropometric variables or physical activity with risk of non-Hodgkin’s lymphoma, and this is the only study to our knowledge to systematically assess this in the context of potential confounding by other non-Hodgkin’s lymphoma risk factors (including dietary factors) and to report results for the common subtypes of non-Hodgkin’s lymphoma. This is also the first study to specifically address B-CLL, a malignancy of clonal B cells recently classified with small lymphocytic lymphoma on the basis of a presumed common underlying pathobiology (5). In addition to the prospective design, other strengths of this study include a community setting, comprehensive follow-up of the cohort, use of a Surveillance, Epidemiology, and End Results cancer registry for case ascertainment, an extensive evaluation of anthropometric variables, use of a validated protocol to assess body fat distribution, adjustment for important confounders, and systematic classification of the non-Hodgkin’s lymphoma subtypes. There are also limitations. Most importantly, the sample size (and therefore the precision of our estimates) was limited for the rarer subtypes, particularly small lymphocytic lymphoma. Moreover, multiple comparisons were made, and the few results at p < 0.05 may be due to chance. The initial response to the baseline questionnaire was only 43 percent. However, although this may limit the generalizability of the results, the virtually complete follow-up of the cohort ensures a high internal validity. Furthermore, nonresponders to the baseline survey have also been followed, and although they had slightly higher rates of smoking-related diseases, the association of body mass index with cancer occurrence (all sites and breast, endometrial, colon, and lung individually) was not appreciably affected by nonresponse bias (36). In addition, the anthropometric (37) and physical activity (26) variables have been associated with other morbidity or mortality endpoints in this study, consistent with a majority of the literature. The results of this study, however, may not generalize to males or non-Caucasian populations. Adult weight and body mass We found no evidence of an association of weight or body mass, either at study baseline or during adulthood, with the risk of non-Hodgkin’s lymphoma overall. This is consistent with three other incidence studies (7, 38, 39), a mortality study (40), and a hospital-based case-control study (41). In contrast, a Swedish cohort study (11) found that women with a hospital diagnosis of obesity were at excess risk of developing non-Hodgkin’s lymphoma (43 observed cases; RR = 1.6, 95 percent CI: 1.2, 2.1), although there was a weak inverse association for men (nine observed cases; standardized incidence ratio = 0.7, 95 percent CI: 0.3, 1.3). The only other study reporting an association with obesity was a large population-based case-control study conducted in San Francisco, California (10). Among heterosexuals, there was a strong, graded, positive association between body mass index and the risk of non-Hodgkin’s lymphoma for both men (odds ratio = 2.6 for a body mass index of ≥30 vs. <20, 95 percent CI: 1.2, 5.7) and women (odds ratio = 3.1 for a body mass index of ≥30 vs. <20, 95 percent CI: 1.4, 6.8). The latter associations remained after adjustment for several medical history and lifestyle factors. Associations by non-Hodgkin’s lymphoma subtype were not reported. When we repeated our analyses using the same cutpoints as the latter study, we still found no association (data not reported). In the San Francisco case-control study, 21 percent of the eligible patients died before they could be contacted, and another 10 percent were too ill to be interviewed. Since survival (required to get into the study) could be related to body mass, we excluded all cases in our study who died within 9 months of diagnosis to see if this might explain the discrepancy between our studies. After this exclusion, however, we still found no association between body mass index and the risk of non-Hodgkin’s lymphoma. We clearly had sufficient power to detect an association of the magnitude reported in the San Francisco study. Furthermore, it is unlikely that we were unable to detect an association because of exposure misclassification, as body mass index has been strongly correlated with breast and endometrial cancer incidence (42, 43) and coronary heart disease mortality (37) in this cohort. Although discrepancies remain, the balance of evidence published to date suggests no association of body mass with non-Hodgkin’s lymphoma. To our knowledge, no data have been reported on the association of adult obesity with the risk of non-Hodgkin’s lymphoma subtypes or of B-CLL. There are some published data on the association of body mass index with the risk of all leukemias, with studies showing both weak positive associations (38–40) and no association (11, 44). We found no evidence of an association of weight or body mass index with the risk of diffuse or follicular lymphoma. We did find suggestive inverse associations of weight, body mass index, and body mass index at ages 40 and 50 years with the risk of small lymphocytic lymphoma. In addition, we found suggestive positive associations of body mass index with the risk of B-CLL, particularly body mass indexes in later adulthood. However, most of these results were not statistically significant, and there was generally no clear dose-response trend, although the multivariate results for body mass index at age 50 years were statistically significant for both small lymphocytic lymphoma (inverse association, p trend = 0.04) and B-CLL (positive association, p trend = 0.03). It is somewhat surprising that the associations for small lymphocytic lymphoma and B-CLL are in opposite directions. This is based on the prediction that these two low-grade malignacies of B cells would be predicted to have similar risk factor profiles, because they may represent different clinical spectrums of closely related B-cell subpopulations. Indeed, their biologic and pathologic similarities have led to their classification together in the most recent World Health Organization classification system (5). Because of the small sample sizes in these analyses, however, our findings from this secondary analysis may be due to chance, and thus replication is needed. Height and adolescent obesity Adult height and adolescent obesity in part reflect the influence of nutritional factors during early childhood and adolescence on growth, with taller adult height and greater obesity reflecting more adequate (or excess) availability of food and energy and greater exposure to growth-related hormones. There are few data on the association of adolescent obesity with the risk of non-Hodgkin’s lymphoma and no data for non-Hodgkin’s lymphoma subtypes or B-CLL. The cohort studies of former students of Harvard University and the University of Pennsylvania found no association of obesity in college with the risk of either non-Hodgkin’s lymphoma or lymphatic leukemia (45). In our data, we also found little evidence of an important role for either relative weight at age 12 years or body mass index at age 18 and the risk of non-Hodgkin’s lymphoma overall or for diffuse or follicular non-Hodgkin’s lymphoma. There were, however, positive associations for B-CLL, consistent with the positive associations seen with obesity measures at other ages in this cohort. Relative weight at age 12 years was also associated with small lymphocytic lymphoma, but the increased risk was observed for both being underweight and being overweight relative to peers, although this was statistically significant only for being overweight. The subjective nature of the assessment of weight at age 12 years may have a higher potential for misclassification (both differential and nondifferential), which could bias our results. We found no association of height with the risk of all non-Hodgkin’s lymphoma, non-Hodgkin’s lymphoma subtypes, or B-CLL. There was no association of height with the risk of non-Hodgkin’s lymphoma or leukemia (all) in a cohort study of Icelanders (39), although other cohort studies have reported a positive association with non-Hodgkin’s lymphoma incidence (7), lymphoma mortality (44), and hematopoietic cancer mortality (12). The latter two studies were based on mortality data, had small numbers of events, and lacked adjustment for dietary factors, but the former study using incidence data from the Nurses’ Health Study (7) did not have these limitations and is comparable in design to this study. In the Nurses’ Health Study, women who were ≥68 inches (1.73 m) were at a 2.4-fold increased risk of non-Hodgkin’s lymphoma (95 percent CI: 1.2, 4.7) compared with women who were <62 inches (1.57 m), and there was a clear gradient in risk with increasing height. Neither our study nor the Nurses’ Health Study evaluated the association of height with non-Hodgkin’s lymphoma mortality, but the magnitude of associations found in cohort studies of lymphoma or hematopoietic mortality (12, 44) is compatible with the magnitude of the effect seen in the incidence data from the Nurses’ Health Study. Our incidence data, in contrast, are not compatible with these other studies, and more data based on incidence will be required to clarify the validity of this association. No prior study has published data on the association of height with non-Hodgkin’s lymphoma subtypes or with B-CLL specifically. Body fat distribution To our knowledge, there are no data on body fat distribution and the risk of non-Hodgkin’s lymphoma or B-CLL; we found that this is unlikely to be an important risk factor in the etiology of these malignancies overall or for non-Hodgkin’s lymphoma subtypes. Physical activity Overall, we found no association between the level of leisure-time physical activity and the risk of non-Hodgkin’s lymphoma or B-CLL. A limitation of these data is the lack of occupational physical activity. Although we did observe a suggestive positive association of physical inactivity with the risk of follicular non-Hodgkin’s lymphoma, this association attenuated after multivariate adjustment. A larger study is needed to better characterize this association. Occupational physical activity was not associated with either non-Hodgkin’s lymphoma or leukemia in a registry-based case-control study (other cancers served as controls) (22), and physical activity in college was not associated with the risk of non-Hodgkin’s lymphoma or lymphatic leukemia in two cohort studies of college alumni (21, 23). Physical activity, even modest levels, has been most strongly associated with a reduced risk of colon cancer and cancers of the female reproductive tract (18, 20), and the mechanistic basis for these observations includes control of body size, changes in steroid hormone metabolism, changes in menstrual function (females), or modulation of immune function. As reviewed above, a majority of studies have not found an association of non-Hodgkin’s lymphoma with obesity, suggesting that weight control is unlikely to be an important mechanism. Menstrual characteristics do not appear to be important risk factors for non-Hodgkin’s lymphoma, at least in this cohort (46). Although there are some animal and human data showing immunologic effects of both moderate and strenuous physical activity (18, 19), persons with primary or acquired immunodeficiency do not develop colon and female reproductive tract cancers but, rather, skin cancer, lymphomas, Kaposi′s sarcoma, and other rare or unusual tumors (47). Thus, the overall lack of association of physical activity with the risk of non-Hodgkin’s lymphoma (particularly diffuse lymphoma), a cancer whose development is thought to be correlated with the host immune status, suggests that immunomodulation may not be a major mechanistic basis for the protective association of physical activity with the development of cancer in general. Conclusions We found no evidence that height, weight, body mass, or physical activity plays an important role in the development of non-Hodgkin’s lymphoma overall or of diffuse or follicular non-Hodgkin’s lymphoma in particular. The inverse associations of several body mass index measures with small lymphocytic non-Hodgkin’s lymphoma, but positive associations with B-CLL, are novel. However, the latter associations must be interpreted with caution because of the small sample size, and they clearly require replication. Given the generally null findings from the literature and in this study, as well as weak data for a major role of obesity or physical activity in immune modulation, these factors are unlikely to have a critical role in the etiology of non-Hodgkin’s lymphoma. In addition, the more strongly documented correlates between diet and the risk of non-Hodgkin’s lymphoma (6, 7, 48, 49) indicate that future research will likely be more productive in evaluating the role of specific dietary factors in the etiology of these malignancies. ACKNOWLEDGMENTS The Iowa Women’s Health Study was supported by National Cancer Institute grant R01 CA39741. Dr. Cerhan was supported in part by a National Cancer Institute Preventive Oncology Award (K07 CA64220). The authors thank Ching-Ping Hong for technical assistance and Mary Jo Janisch for assistance in preparing this manuscript. Reprint requests to Dr. James R. Cerhan, Department of Health Sciences Research, Mayo Clinic, 200 First Street SW, Rochester, MN 55905 (e-mail: cerhan.james@mayo.edu). TABLE 1. Age-adjusted relative risks and 95% confidence intervals of non-Hodgkin’s lymphoma, selected non-Hodgkin’s lymphoma subtypes, and B-cell chronic lymphocytic leukemia in relation to baseline anthropometric characteristics and physical activity, Iowa Women’s Health Study, 1986–1998 Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Height (m)                                          <1.58  121,049  75  1      36  1      12  1      13  1      17  1    1.58–1.62  64,782  29  0.7  0.5, 1.1    18  0.9  0.5, 1.6    6  0.9  0.4, 2.5    3  0.4  0.1, 1.5    9  1.0  0.4, 2.2  1.63–1.67  133,575  76  0.9  0.7, 1.3    36  0.9  0.6, 1.5    23  1.8  0.9, 3.6    8  0.6  0.2, 1.4    17  0.9  0.5, 1.8  ≥1.68  130,396  81  1.1  0.8, 1.4    47  1.3  0.8, 2.0    17  1.4  0.7, 2.9    8  0.6  0.2, 1.4    20  1.2  0.6, 2.2                                            Weight (kg)                                          <59.9  108,647  68  1      29  1      15  1      12  1      13  1    59.9–67.1  117,358  60  0.8  0.6, 1.2    32  1.0  0.6, 1.7    17  1.1  0.5, 2.1    6  0.5  0.2, 1.2    19  1.4  0.7, 2.7  67.2–77.1  111,440  58  0.8  0.6, 1.2    37  1.2  0.8, 2.0    7  0.5  0.2, 1.1    6  0.5  0.2, 1.3    14  1.1  0.5, 2.2  ≥77.2  112,356  75  1.1  0.8, 1.5    39  1.3  0.8, 2.1    19  1.2  0.6, 2.4    8  0.7  0.3, 1.6    17  1.3  0.6, 2.7                                            Body mass index (kg/m2)                                          <23.5  109,932  71  1      32  1      14  1      12  1      13  1    23.5–26.1  115,079  51  0.7  0.5, 1.0    24  0.7  0.4, 1.2    16  1.1  0.5, 2.2    5  0.4  0.1, 1.1    14  1.0  0.5, 2.2  26.2–29.7  113,205  66  0.9  0.6, 1.2    43  1.3  0.8, 2.0    9  0.6  0.3, 1.4    8  0.6  0.3, 1.6    19  1.4  0.7, 2.8  ≥29.8  111,586  73  1.0  0.7, 1.4    38  1.2  0.7, 1.9    19  1.3  0.7, 2.6    7  0.6  0.2, 1.5    17  1.3  0.6, 2.6                                            Waist/hip ratio                                          <0.777  113,427  56  1      28  1      12  1      6  1      13  1    0.777–0.831  112,834  54  0.9  0.6, 1.4    28  1.0  0.6, 1.6    12  1.0  0.4, 2.2    5  0.8  0.3, 2.7    19  1.4  0.7, 2.8  0.832–0.892  111,562  83  1.4  1.0, 2.0    45  1.5  1.0, 2.5    18  1.5  0.7, 3.0    11  1.8  0.7, 4.9    16  1.2  0.6, 2.4  ≥0.893  110,067  66  1.1  0.8, 1.6    35  1.2  0.7, 1.9    16  1.3  0.6, 2.7    9  1.5  0.5, 4.2    15  1.1  0.5, 2.2                                            Physical activity index                                          High  111,319  58  1      34  1      10  1      8  1      13  1    Moderate  122,512  66  1.0  0.7, 1.5    34  0.9  0.6, 1.5    15  1.4  0.6, 3.1    8  0.9  0.3, 2.4    22  1.6  0.8, 3.1  Low  206,920  128  1.2  0.9, 1.7    63  1.0  0.7, 1.6    32  1.8  0.9, 3.6    15  1.0  0.4, 2.4    26  1.1  0.6, 2.2  Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Height (m)                                          <1.58  121,049  75  1      36  1      12  1      13  1      17  1    1.58–1.62  64,782  29  0.7  0.5, 1.1    18  0.9  0.5, 1.6    6  0.9  0.4, 2.5    3  0.4  0.1, 1.5    9  1.0  0.4, 2.2  1.63–1.67  133,575  76  0.9  0.7, 1.3    36  0.9  0.6, 1.5    23  1.8  0.9, 3.6    8  0.6  0.2, 1.4    17  0.9  0.5, 1.8  ≥1.68  130,396  81  1.1  0.8, 1.4    47  1.3  0.8, 2.0    17  1.4  0.7, 2.9    8  0.6  0.2, 1.4    20  1.2  0.6, 2.2                                            Weight (kg)                                          <59.9  108,647  68  1      29  1      15  1      12  1      13  1    59.9–67.1  117,358  60  0.8  0.6, 1.2    32  1.0  0.6, 1.7    17  1.1  0.5, 2.1    6  0.5  0.2, 1.2    19  1.4  0.7, 2.7  67.2–77.1  111,440  58  0.8  0.6, 1.2    37  1.2  0.8, 2.0    7  0.5  0.2, 1.1    6  0.5  0.2, 1.3    14  1.1  0.5, 2.2  ≥77.2  112,356  75  1.1  0.8, 1.5    39  1.3  0.8, 2.1    19  1.2  0.6, 2.4    8  0.7  0.3, 1.6    17  1.3  0.6, 2.7                                            Body mass index (kg/m2)                                          <23.5  109,932  71  1      32  1      14  1      12  1      13  1    23.5–26.1  115,079  51  0.7  0.5, 1.0    24  0.7  0.4, 1.2    16  1.1  0.5, 2.2    5  0.4  0.1, 1.1    14  1.0  0.5, 2.2  26.2–29.7  113,205  66  0.9  0.6, 1.2    43  1.3  0.8, 2.0    9  0.6  0.3, 1.4    8  0.6  0.3, 1.6    19  1.4  0.7, 2.8  ≥29.8  111,586  73  1.0  0.7, 1.4    38  1.2  0.7, 1.9    19  1.3  0.7, 2.6    7  0.6  0.2, 1.5    17  1.3  0.6, 2.6                                            Waist/hip ratio                                          <0.777  113,427  56  1      28  1      12  1      6  1      13  1    0.777–0.831  112,834  54  0.9  0.6, 1.4    28  1.0  0.6, 1.6    12  1.0  0.4, 2.2    5  0.8  0.3, 2.7    19  1.4  0.7, 2.8  0.832–0.892  111,562  83  1.4  1.0, 2.0    45  1.5  1.0, 2.5    18  1.5  0.7, 3.0    11  1.8  0.7, 4.9    16  1.2  0.6, 2.4  ≥0.893  110,067  66  1.1  0.8, 1.6    35  1.2  0.7, 1.9    16  1.3  0.6, 2.7    9  1.5  0.5, 4.2    15  1.1  0.5, 2.2                                            Physical activity index                                          High  111,319  58  1      34  1      10  1      8  1      13  1    Moderate  122,512  66  1.0  0.7, 1.5    34  0.9  0.6, 1.5    15  1.4  0.6, 3.1    8  0.9  0.3, 2.4    22  1.6  0.8, 3.1  Low  206,920  128  1.2  0.9, 1.7    63  1.0  0.7, 1.6    32  1.8  0.9, 3.6    15  1.0  0.4, 2.4    26  1.1  0.6, 2.2  * NHL, non-Hodgkin’s lymphoma; B-CLL, B-cell chronic lymphocytic leukemia; RR, relative risk; CI, confidence interval. † All relative risks adjusted for age. View Large TABLE 2. Age-adjusted relative risks and 95% confidence intervals of non-Hodgkin’s lymphoma, selected non-Hodgkin’s lymphoma subtypes, and B-cell chronic lymphocytic leukemia in relation to historical anthropometric characteristics, Iowa Women’s Health Study, 1986–1998 Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Relative weight at age 12 years                                          Below average  98,034  63  1.2  0.9, 1.6    35  1.3  0.9, 1.9    12  1.0  0.5, 1.8    9  1.7  0.8, 3.9    9  0.7  0.3, 1.3  Average  280,767  153  1      78  1      36  1      15  1      40  1    Above average  57,022  38  1.3  0.9, 1.8    19  1.3  0.8, 2.1    9  1.3  0.6, 2.6    7  2.4  1.0, 5.8    12  1.5  0.8, 3.0                                            Body mass index (kg/m2) at age 18 years                                          <19.6  111,383  64  1      27  1      21  1      10        13  1    19.6–21.1  112,679  53  0.8  0.6, 1.2    32  1.2  0.7, 2.0    8  0.4  0.2, 0.9    4  0.4  0.1, 1.3    16  1.2  0.6, 2.6  21.2–23.0  117,534  72  1.1  0.8, 1.5    41  1.5  0.9, 2.4    13  0.6  0.3, 1.2    7  0.7  0.3, 1.8    13  1.0  0.5, 2.1  ≥23.1  106,153  69  1.2  0.8, 1.6    34  1.4  0.8, 2.3    16  0.8  0.4, 1.6    11  1.2  0.5, 2.8    21  1.7  0.9, 3.5                                            Body mass index (kg/m2) at age 30 years                                          <21.1  106,838  58  1      28  1      17  1      8  1      9  1    21.1–22.7  114,610  69  1.1  0.8, 1.6    36  1.2  0.7, 2.0    15  0.8  0.4, 1.6    7  0.8  0.3, 2.2    21  2.2  1.0, 4.7  22.8–24.8  115,176  66  1.0  0.7, 1.5    37  1.2  0.7, 2.0    9  0.5  0.2, 1.1    8  0.9  0.3, 2.5    14  1.4  0.6, 3.3  ≥24.9  105,540  61  1.1  0.7, 1.5    30  1.1  0.7, 1.8    17  1.0  0.5, 2.0    8  1.0  0.4, 2.7    18  2.0  0.9, 4.5                                            Body mass index (kg/m2) at age 40 years                                          <22.1  109,827  61  1      29  1      14  1      10  1      11  1    22.1–23.9  117,838  73  1.1  0.8, 1.6    35  1.1  0.7, 1.8    16  1.1  0.5, 2.2    7  0.6  0.2, 1.7    16  1.3  0.6, 2.9  24.0–26.4  109,119  51  0.8  0.6, 1.2    30  1.0  0.6, 1.7    11  0.8  0.4, 1.7    6  0.6  0.2, 1.7    16  1.4  0.7, 3.1  ≥26.5  105,396  69  1.2  0.8, 1.7    37  1.3  0.8, 2.2    17  1.3  0.6, 2.6    8  0.8  0.3, 2.1    19  1.8  0.9, 3.8                                            Body mass index (kg/m2) at age 50 years                                          <22.9  111,104  67  1      28  1      16  1      11  1      9  1    22.9–25.0  116,883  67  0.9  0.7, 1.3    40  1.3  0.8, 2.2    11  0.6  0.3, 1.4    7  0.6  0.2, 1.6    18  1.9  0.8, 4.2  25.1–28.2  109,549  57  0.9  0.6, 1.2    29  1.0  0.6, 1.7    16  1.0  0.5, 2.0    8  0.7  0.3, 1.8    13  1.4  0.6, 3.4  ≥28.3  111,110  69  1.0  0.7, 1.5    39  1.4  0.9, 2.3    15  0.9  0.5, 1.9    6  0.6  0.2, 1.5    23  2.6  1.2, 5.6  Variable  Person-years  All NHL*    Diffuse    Follicular    Small lymphocytic    B-CLL*  Cases (no.)  RR*,†  95% CI*    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI    Cases (no.)  RR  95% CI  Relative weight at age 12 years                                          Below average  98,034  63  1.2  0.9, 1.6    35  1.3  0.9, 1.9    12  1.0  0.5, 1.8    9  1.7  0.8, 3.9    9  0.7  0.3, 1.3  Average  280,767  153  1      78  1      36  1      15  1      40  1    Above average  57,022  38  1.3  0.9, 1.8    19  1.3  0.8, 2.1    9  1.3  0.6, 2.6    7  2.4  1.0, 5.8    12  1.5  0.8, 3.0                                            Body mass index (kg/m2) at age 18 years                                          <19.6  111,383  64  1      27  1      21  1      10        13  1    19.6–21.1  112,679  53  0.8  0.6, 1.2    32  1.2  0.7, 2.0    8  0.4  0.2, 0.9    4  0.4  0.1, 1.3    16  1.2  0.6, 2.6  21.2–23.0  117,534  72  1.1  0.8, 1.5    41  1.5  0.9, 2.4    13  0.6  0.3, 1.2    7  0.7  0.3, 1.8    13  1.0  0.5, 2.1  ≥23.1  106,153  69  1.2  0.8, 1.6    34  1.4  0.8, 2.3    16  0.8  0.4, 1.6    11  1.2  0.5, 2.8    21  1.7  0.9, 3.5                                            Body mass index (kg/m2) at age 30 years                                          <21.1  106,838  58  1      28  1      17  1      8  1      9  1    21.1–22.7  114,610  69  1.1  0.8, 1.6    36  1.2  0.7, 2.0    15  0.8  0.4, 1.6    7  0.8  0.3, 2.2    21  2.2  1.0, 4.7  22.8–24.8  115,176  66  1.0  0.7, 1.5    37  1.2  0.7, 2.0    9  0.5  0.2, 1.1    8  0.9  0.3, 2.5    14  1.4  0.6, 3.3  ≥24.9  105,540  61  1.1  0.7, 1.5    30  1.1  0.7, 1.8    17  1.0  0.5, 2.0    8  1.0  0.4, 2.7    18  2.0  0.9, 4.5                                            Body mass index (kg/m2) at age 40 years                                          <22.1  109,827  61  1      29  1      14  1      10  1      11  1    22.1–23.9  117,838  73  1.1  0.8, 1.6    35  1.1  0.7, 1.8    16  1.1  0.5, 2.2    7  0.6  0.2, 1.7    16  1.3  0.6, 2.9  24.0–26.4  109,119  51  0.8  0.6, 1.2    30  1.0  0.6, 1.7    11  0.8  0.4, 1.7    6  0.6  0.2, 1.7    16  1.4  0.7, 3.1  ≥26.5  105,396  69  1.2  0.8, 1.7    37  1.3  0.8, 2.2    17  1.3  0.6, 2.6    8  0.8  0.3, 2.1    19  1.8  0.9, 3.8                                            Body mass index (kg/m2) at age 50 years                                          <22.9  111,104  67  1      28  1      16  1      11  1      9  1    22.9–25.0  116,883  67  0.9  0.7, 1.3    40  1.3  0.8, 2.2    11  0.6  0.3, 1.4    7  0.6  0.2, 1.6    18  1.9  0.8, 4.2  25.1–28.2  109,549  57  0.9  0.6, 1.2    29  1.0  0.6, 1.7    16  1.0  0.5, 2.0    8  0.7  0.3, 1.8    13  1.4  0.6, 3.4  ≥28.3  111,110  69  1.0  0.7, 1.5    39  1.4  0.9, 2.3    15  0.9  0.5, 1.9    6  0.6  0.2, 1.5    23  2.6  1.2, 5.6  * NHL, non-Hodgkin’s lymphoma; B-CLL, B-cell chronic lymphocytic leukemia; RR, relative risk; CI, confidence interval. † All relative risks adjusted for age. View Large TABLE 3. Multivariate-adjusted relative risks and 95% confidence intervals of small lymphocytic lymphoma and B-cell chronic lymphocytic leukemia in relation to selected anthropometric characteristics, Iowa Women’s Health Study, 1986–1998   Small lymphocytic    B-CLL*  RR*,†  95% CI*    RR  95% CI  Relative weight at age 12 years            Below average  1.5  0.6, 3.8    0.6  0.3, 1.3  Average  1      1    Above average  2.5  1.0, 6.3    1.5  0.8, 2.9    p trend = 0.4    p trend = 0.05  Weight (kg)            <59.9  1      1    59.9–67.1  0.4  0.2, 1.2    1.5  0.7, 3.0  67.2–77.1  0.3  0.1, 0.9    1.2  0.5, 2.6  ≥77.2  0.4  0.1, 1.0    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.7  Body mass index (kg/m2) at age 50 years            <22.9  1      1    22.9–25.0  0.6  0.2, 1.7    1.9  0.8, 4.3  25.1–28.2  0.6  0.2, 1.6    1.5  0.6, 3.5  ≥28.3  0.3  0.1, 0.9    2.7  1.2, 6.0    p trend = 0.04    p trend = 0.03  Body mass index (kg/m2)            <23.5  1      1    23.5–26.1  0.4  0.1, 1.1    1.1  0.5, 2.4  26.2–29.7  0.4  0.2, 1.2    1.6  0.8, 3.3  ≥29.8  0.3  0.1, 0.9    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.4    Small lymphocytic    B-CLL*  RR*,†  95% CI*    RR  95% CI  Relative weight at age 12 years            Below average  1.5  0.6, 3.8    0.6  0.3, 1.3  Average  1      1    Above average  2.5  1.0, 6.3    1.5  0.8, 2.9    p trend = 0.4    p trend = 0.05  Weight (kg)            <59.9  1      1    59.9–67.1  0.4  0.2, 1.2    1.5  0.7, 3.0  67.2–77.1  0.3  0.1, 0.9    1.2  0.5, 2.6  ≥77.2  0.4  0.1, 1.0    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.7  Body mass index (kg/m2) at age 50 years            <22.9  1      1    22.9–25.0  0.6  0.2, 1.7    1.9  0.8, 4.3  25.1–28.2  0.6  0.2, 1.6    1.5  0.6, 3.5  ≥28.3  0.3  0.1, 0.9    2.7  1.2, 6.0    p trend = 0.04    p trend = 0.03  Body mass index (kg/m2)            <23.5  1      1    23.5–26.1  0.4  0.1, 1.1    1.1  0.5, 2.4  26.2–29.7  0.4  0.2, 1.2    1.6  0.8, 3.3  ≥29.8  0.3  0.1, 0.9    1.3  0.6, 2.8    p trend = 0.03    p trend = 0.4  * B-CLL, B-cell chronic lymphocytic leukemia; RR, relative risk; CI, confidence interval. † All relative risks adjusted for age, marital status, farm residence, diabetes, transfusion history, hormone replacement therapy, alcohol consumption, smoking history, red meat consumption, and fruit consumption. View Large References 1. Greenlee RT, Hill-Harmon MB, Murray T, et al. Cancer statistics, 2001. CA Cancer J Clin   2001; 51: 15–36. Google Scholar 2. Ries LG, Eisner MP, Kosary CL, et al. SEER cancer statistics review, 1973–1998. Bethesda, MD: National Cancer Institute, 2001. Google Scholar 3. Devesa SS, Fears T. Non-Hodgkin’s lymphoma time trends: United States and international data. Cancer Res  1992; 52(suppl): 5432s–40s. Google Scholar 4. Hartge P, Devesa SS. Quantification of the impact of known risk factors on time trends in non-Hodgkin’s lymphoma incidence. Cancer Res   1992; 52(suppl): 5566s–9s. Google Scholar 5. Jaffe ES, Harris NL, Stein H, et al. World Health Organization classification of tumours. Pathology and genetics of tumours of haematopoietic and lymphoid tissues. Lyon, France: IARC Press, 2001. Google Scholar 6. Chiu BC, Cerhan JR, Folsom AR, et al. Diet and risk of non-Hodgkin lymphoma in older women. JAMA   1996; 275: 1315–21. Google Scholar 7. Zhang S, Hunter DJ, Rosner BA, et al. Dietary fat and protein in relation to risk of non-Hodgkin’s lymphoma among women. J Natl Cancer Inst  1999; 91: 1751–8. Google Scholar 8. Cunningham AS. Lymphomas and animal-protein consumption. Lancet  1976; 2: 1184–6. Google Scholar 9. Tavani A, Pregnolato A, Negri E, et al. Diet and risk of lymphoid neoplasms and soft tissue sarcomas. Nutr Cancer  1997; 27: 256–60. Google Scholar 10. Holly EA, Lele C, Bracci PM, et al. Case-control study of non-Hodgkin’s lymphoma among women and heterosexual men in the San Francisco Bay Area, California. Am J Epidemiol  1999; 150: 375–89. Google Scholar 11. Wolk A, Gridley G, Svensson M, et al. A prospective study of obesity and cancer risk (Sweden). Cancer Causes Control  2001; 12: 13–21. Google Scholar 12. Davey Smith G, Hart C, Upton M, et al. Height and risk of death among men and women: aetiological implications of associations with cardiorespiratory disease and cancer mortality. J Epidemiol Community Health  2000; 54: 97–103. Google Scholar 13. Stallone DD. The influence of obesity and its treatment on the immune system. Nutr Rev  1994; 52: 37–50. Google Scholar 14. Cartwright RA, McKinney PA, O’Brien C, et al. Non-Hodgkin’s lymphoma: a case-control epidemiologic study in Yorkshire. Leuk Res  1988; 13: 465–72. Google Scholar 15. Cerhan JR, Wallace RB, Folsom AR, et al. Medical history risk factors for non-Hodgkin’s lymphoma in older women. J Natl Cancer Inst  1997; 89: 314–18. Google Scholar 16. Hjalgrim H, Frisch M, Ekbom A, et al. Cancer and diabetes—a follow-up study of two population-based cohorts of diabetic patients. J Intern Med  1997; 241: 471–5. Google Scholar 17. Weiderpass E, Gridley G, Ekbom A, et al. Medical history risk factors for non-Hodgkin’s lymphoma in older women. J Natl Cancer Inst  1997; 89: 816–17. Google Scholar 18. Shephard RJ, Shek PN. Cancer, immune function, and physical activity. Can J Appl Physiol  1995; 20: 1–25. Google Scholar 19. Lee IM. Exercise and physical health: cancer and immune function. Res Q Exerc Sport  1995; 66: 286–91. Google Scholar 20. Friedenreich CM. Physical activity and cancer prevention: from observational to intervention research. Cancer Epidemiol Biomarkers Prev  2001; 10: 287–301. Google Scholar 21. Polednak AP. College athletics, body size, and cancer mortality. Cancer  1976; 38: 382–7. Google Scholar 22. Brownson RC, Chang JC, Davis JR, et al. Physical activity on the job and cancer in Missouri. Am J Public Health  1991; 81: 639–42. Google Scholar 23. Paffenbarger RS Jr, Lee IM, Wing AL. The influence of physical activity on the incidence of site-specific cancers in college alumni. Adv Exp Med Biol  1992; 322: 7–15. Google Scholar 24. Cerhan JR, Wallace RB, Dick F, et al. Blood  transfusion and risk of non-Hodgkin lymphoma subtypes and chronic lymphocytic leukemia. Cancer Epidemiol Biomarkers Prev  2001; 10: 361–8. Google Scholar 25. Kushi LH, Kaye SA, Folsom AR, et al. Accuracy and reliability of self-measurement of body girths. Am J Epidemiol  1988; 128: 740–8. Google Scholar 26. Kushi LH, Fee RM, Folsom AR, et al. Physical activity and mortality in postmenopausal women. JAMA  1997; 277: 1287–92. Google Scholar 27. Percy C, Van Holten V, Muir C. International classification of diseases for oncology. 2nd ed. Geneva, Switzerland: World Health Organization, 1990. Google Scholar 28. Percy C, O’Conor G, Ries L, et al. Non-Hodgkin’s lymphomas: application of the international classification of diseases for oncology (ICD-O) to the Working Formulation. Cancer  1984; 54: 1435–8. Google Scholar 29. Groves FD, Linet MS, Travis LB, et al. Cancer surveillance series: non-Hodgkin’s lymphoma incidence by histologic subtype in the United States from 1978 through 1995. J Natl Cancer Inst  2000; 92: 1240–51. Google Scholar 30. Cox DR. Regression models and life tables (with discussion). J R Stat Soc (B)  1972; 34: 187–220. Google Scholar 31. Korn EL, Graubard BI, Midthune D. Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale. Am J Epidemiol  1997; 145: 72–80. Google Scholar 32. Folsom AR, Zhang S, Sellers TA, et al. Cancer incidence among women living on farms: findings from the Iowa Women’s Health Study. J Occup Environ Med  1996; 38: 1171–6. Google Scholar 33. Cerhan JR. New epidemiologic leads in the etiology of non-Hodgkin lymphoma in older persons: the role of blood transfusion and diet. Biomed Pharmacother  1997; 51: 200–7. Google Scholar 34. Chiu BC, Cerhan JR, Gapstur SM, et al. Alcohol consumption and non-Hodgkin lymphoma in a cohort of older women. Br J Cancer  1999; 80: 1476–82. Google Scholar 35. Parker AS, Cerhan JR, Dick F, et al. Smoking and risk of non-Hodgkin’s lymphoma subtypes in a cohort of older Iowa women. Leuk Lymphoma  2000; 37: 341–9. Google Scholar 36. Bisgard KM, Folsom AR, Hong CP, et al. Mortality and cancer rates in nonrespondents to a prospective cohort study of older women: 5-year follow-up. Am J Epidemiol  1994; 139: 990–1000. Google Scholar 37. Folsom AR, Kushi LH, Anderson KE, et al. Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women’s Health Study. Arch Intern Med  2000; 160: 2117–28. Google Scholar 38. Moller H, Mellemgaard A, Lindvig K, et al. Obesity and cancer risk: a Danish record-linkage study. Eur J Cancer  1994; 3: 344–50. Google Scholar 39. Tulinius H, Sigfusson N, Sigvaldason H, et al. Risk factors for malignant diseases: a cohort study on a population of 22,946 Icelanders. Cancer Epidemiol Biomarkers Prev   1997; 6: 863–73. Google Scholar 40. Lew EA, Garfinkel L. Variations in mortality by weight among 750,000 men and women. J Chronic Dis  1979; 32: 563–76. Google Scholar 41. Franceschi S, Serraino D, Bidoli E, et al. The epidemiology of non-Hodgkin’s lymphoma in the north-east of Italy: a hospital-based case-control study. Leuk Res  1989; 13: 465–72. Google Scholar 42. Folsom AR, Kaye SA, Potter JD, et al. Association of incident carcinoma of the endometrium with body weight and fat distribution in older women: early findings of the Iowa Women’s Health Study. Cancer Res  1989; 49: 6828–31. Google Scholar 43. Folsom AR, Kaye SA, Prineas RJ, et al. Increased incidence of carcinoma of the breast associated with abdominal adiposity in postmenopausal women. Am J Epidemiol  1990; 131: 794–803. Google Scholar 44. Leon DA, Smith GD, Shipley M, et al. Adult height and mortality in London: early life, socioeconomic confounding, or shrinkage? J Epidemiol Community Health  1995; 49: 5–9. Google Scholar 45. Whittemore AS, Paffenbarger RS Jr, Anderson K, et al. Early precursors of site-specific cancers in college men and women. J Natl Cancer Inst  1985; 74: 43–51. Google Scholar 46. Cerhan JR, Habermann TM, Vachon CM, et al. Menstrual and reproductive factors and risk of non-Hodgkin lymphoma: the Iowa Women’s Health Study. Cancer Causes Control  2002; 13: 131–6. Google Scholar 47. Schwartz R. Immunodeficiency, immunosuppression, and susceptibility to neoplasms. J Natl Cancer Inst Monogr  2000;(28): 5–9. Google Scholar 48. Ward MH, Zahm SH, Weisenburger DD, et al. Dietary factors and non-Hodgkin’s lymphoma in Nebraska (United States). Cancer Causes Control  1994; 5: 422–32. Google Scholar 49. Zhang SM, Hunter DJ, Rosner BA, et al. Intakes of fruits, vegetables, and related nutrients and the risk of non-Hodgkin’s lymphoma among women. Cancer Epidemiol Biomarkers Prev  2000; 9: 477–85. Google Scholar

Journal

American Journal of EpidemiologyOxford University Press

Published: Sep 15, 2002

There are no references for this article.