Introduction Testicular cancer accounts for <2% of malignant neoplasms in men but is the most common tumor in young adult men aged 15-44 years in the United States.[1] The only well-established risk factor for testicular cancer is cryptorchidism (nondescent of the testes into the scrotum at birth) (reviewed in Reference 2). Although many other potential risk factors have been evaluated in numerous studies, no consistent associations have emerged (reviewed in Reference 2). The influence of diet on testicular cancer risk has not been extensively explored, but a few studies have found associations with dietary fats, milk, red meat, fruits, and green vegetables. An ecological study of per capita fat consumption and testicular cancer rates[3] and a case-control study have been suggestive of increased testicular cancer risk with a high-fat diet.[4] Diets high in red meat and milk and low in fruits and green vegetables have also been associated with elevated risk of testicular cancer.[4-6] Although testicular cancers can be divided clinically into slow-growing seminomas, rapidly growing nonseminomas, and mixed germ cell tumors (those with seminomatous and nonseminomatous elements), only one study has analyzed diet in relation to histology.[4] To evaluate the association of dietary factors with risk of testicular cancer, we conducted a hospital-based case-control study at The University of Texas M. D. Anderson Cancer Center. Specifically, we examined whether testicular cancer risk increased with a high dietary intake of fats, meat, and milk and low consumption of fruits and vegetables. We also assessed whether the dietary factors associated with risk varied by testicular cancer histopathology. Methods We identified men with testicular cancer who registered at M. D. Anderson Cancer Center (Houston, TX) between January 1990 and October 1996 through the M. D. Anderson Tumor Registry and the M. D. Anderson Genitourinary Oncology Clinic. Potential cases were defined as men who were alive during the data collection phase of the investigation, were between the ages of 18 and 50 years at the time of diagnosis, and lived in Texas, Louisiana, Arkansas, or Oklahoma. The age restriction was chosen because testicular cancer occurs most often between 18 and 50 years, and we restricted the region to the south-central United States so that the cases and controls would be from a defined geographic area. All men diagnosed with testicular cancer were eligible for inclusion, regardless of their ethnicity, tumor stage, and histology. To identify controls, we asked each case to nominate healthy friends who had never had cancer and were of the same ethnicity, age (±5 yr), and state of residence as the case. We chose friend controls, because the cases arose from an undefined population, which is a common concern with hospital-based case-control studies. We assumed friends of cases, had they developed testicular cancer, would have been more likely to have come to M. D. Anderson Cancer Center than a population-based control group. Cases and controls completed self-administered questionnaires eliciting information on demographics; lifestyle habits; medical history, including history of cryptorchidism, family history of cancer, body size, and shape; and diet. To assess diet, we used a modified and revised version of the National Cancer Institute's (NCI's) Health Habits and History Questionnaire (HHHQ), which contained 152 foods and beverages[7,8] and has been validated in a range of populations.[9,10] The time period assessed by the questionnaire was the year before cancer diagnosis for cases and the previous year for controls. From the information we calculated food consumption and nutrient intake by using DietSys (version 4.0), the nutrient analysis program developed for the NCI's HHHQ.[11] To adjust for total energy intake, all dietary factors were analyzed per 1,000 kcal dietary intake according to the nutrient density adjustment method described by Willett,[12] and we also used total daily calories in multivariable modeling. We contacted by mail 335 eligible men diagnosed with testicular cancer, and 101 (30.1%) agreed to participate. We also approached 155 more men in the outpatient Genitourinary Oncology Clinic, of whom 86 (55.5%) agreed to participate. Of these 187 men, 25 were excluded because their HHHQs were not completed and two were excluded because they reported eating too many foods daily, leaving 160 cases for analysis. We contacted 202 healthy potential controls by mail. Of these, 148 men (73.3%) returned the questionnaires. Eleven controls were excluded because they did not complete the HHHQ and one was excluded for consuming too many food items per day, leaving 136 healthy controls for analysis. Because there were multiple cases who would be excluded in the analysis of matched data because they did not have a friend control, we evaluated the effect of dissolving the match on our crude and adjusted results. To do so, we compared the unconditional and conditional logistic regression point estimates for all cases and controls, regardless of matching. Because the point estimates in both of these analyses were essentially the same (data not shown), we presented the results with the matching dissolved. For descriptive analyses, we tested the differences in the means and distributions of cases and controls by using Student's t-test and the X2 test. Using unconditional logistic regression modeling, we adjusted for the potential confounders of age, education, income, ethnicity, history of cryptorchidism, and total daily caloric intake. Age, years of education, and total daily calories were entered as continuous variables, income as a scaled variable, and ethnicity and history of cryptorchidism as categorical variables. The quartile cut points used for dietary intake were based on the consumption distribution among the controls. Multivariable trend tests were calculated by entering the dietary component as a scaled variable in the model. The model building strategy used stepwise backward elimination, and the models were compared by using the likelihood ratio test.[13] All significance tests were two-sided and the a was set at 0.05. Results In Table 1 we present selected demographic characteristics for cases and controls grouped by histopathological type. There were 82 men diagnosed with nonseminoma tumors, 46 men diagnosed with pure seminomas, and 32 men diagnosed with mixed germ cell tumors. The ethnic distribution of men with nonseminoma, seminoma, and mixed germ cell tumors was generally similar to that of controls. The controls tended to be older, had more education, and had higher incomes than the cases, although not all these factors were statistically significant across all histological groups. Although we tried to match on age, the cases tended to identify friend controls that were older than themselves. For this reason, in our analyses we consistently controlled for age as a confounder. History of cryptorchidism was significantly more common in all histological groups than in the controls. As shown in Table 2, there were many differences in dietary intake between controls and cases (all histopathological types combined). The cases had significantly higher daily total energy, total fat, saturated fat, and meat intake than controls (p = 0.0001, 0.0002, 0.0001, and 0.006, respectively). Although lower in cases than controls, milk consumption (including whole, 2%, and skim) was not significantly different. Dietary fiber intake was significantly lower in cases than controls (p = 0.006), as was fruit consumption (p = 0.05). Consumption of vegetables and dark green and deep yellow fruits and vegetables was lower in cases than controls, but the difference was of borderline statistical significance (p = 0.09 and 0.08, respectively). Dietary intake further varied by the cases' histology. Men diagnosed with nonseminoma testicular cancer consumed significantly more fats and meats and significantly less dietary fiber and vegetables than controls. The men with seminomas had significantly higher intakes of fats and meat than controls but fiber, fruits, and vegetable intake similar to controls. Men diagnosed with mixed germ cell tumors consumed more fats and less fruit than controls. Multivariable logistic regression analyses of individual dietary components (comparing lowest with highest quartiles of consumption) adjusting for age, education, income, ethnicity, cryptorchidism, and total daily calories. Increasing total fat, saturated fat, meat, and cholesterol consumption were associated with increasing risk of nonseminoma testicular cancer, with odds ratios (ORs) of 6.3, 2.3, 5.3, and 4.6 for the highest quartiles, respectively. An inverse relation was observed for calcium and dietary fiber intake and nonseminoma cancer risk and no consistent trend was observed for milk, fruit, vegetable, or dark green and deep yellow fruit and vegetable consumption. Risk for seminoma testicular cancer increased with increasing intake of meat and total cholesterol (the ORs for the highest vs. the lowest quartiles were 3.6 and 3.1, respectively). Increased total and saturated fat intakes were marginally associated with increased seminoma risk (the ORs for the highest vs. the lowest quartiles were 1.9 and 2.1, respectively). Higher total fat consumption was nearly significantly related to increased mixed germ cell tumor risk (the OR for the highest vs. the lowest quartile was 4.2). When multivariable regression analysis with backward elimination was used to assess the independent effects of multiple dietary components, only saturated fat remained in the model for nonseminoma testicular cancer (Table 3). In multivariable regression analyses for seminoma and mixed germ cell tumor risk, the only dietary variables that remained in each of the models were cholesterol and total fat, respectively (Table 3). Discussion There are few published studies of the relation between diet and testicular cancer. The results of our case-control analysis provide evidence that dietary habits, particularly high total fat and saturated fat intake, may increase risk of testicular cancer. We also found an association with increased meat intake and increased seminoma risk. This agrees with an investigation in Uruguayan men in which higher red meat consumption and total fat intake were associated with increased risk of seminomas.[4] In that study, no associations were observed for nonseminoma tumors, but the results were based on a limited number of cases (30 seminoma and 28 nonseminoma cancers).[4] We found an inverse association between dietary fiber and testicular cancer risk that has not been previously reported. Our findings are weakly supportive of previous studies in which higher intakes of fruits and vegetables were associated with decreased risk of testicular cancer. Gallagher and co-workers[6] conducted a large Canadian population-based study and reported that green vegetable consumption had a protective relative risk of 0.5 (95% confidence interval = 0.3-0.9) in the highest compared with the lowest quartile. In a case-control study in England, Davies and colleagues[5] collected data on fresh fruit and vegetable consumption in adolescence. Although cases tended to have eaten fewer oranges, apples, and vegetable and fruit salads than the population controls, the difference was not statistically significant.[5] The weak associations found in our study with fruit and vegetables were not consistent across testicular cancer histologies in the multivariable analysis, despite reports that these foods reduce the risk of many cancers (reviewed in Reference 14). We did not find an association between milk intake and testicular cancer risk, which is in contrast to the results of Davies and colleagues,[5] who found that for every increased quarter pint of milk consumed per day during adolescence the relative risk was 1.39 (95% confidence interval = 1.19-21.9). However, the results of the British study should be viewed somewhat cautiously, inasmuch as it was unclear whether consumption of milk itself or some component in the milk (such as saturated fat or calcium) increased risk. In our study, we found dietary calcium intake to have a significant inverse dose-response gradient with risk of nonseminoma testicular cancer. To our knowledge, no other previous study has reported an association with calcium intake and testicular cancer. The increased risk of testicular cancer found with higher intake of dietary fat is consistent with many reports for cancers at other hormonally influenced sites such as breast, prostate, colon, and ovary.[15-18] However, not all studies have confirmed the associations with dietary fat,[19] and the relationship between nutrition and cancer probably involves more than one nutrient or component in the diet. The association with dietary fat should be interpreted cautiously, because spurious associations with cancer in case-control studies have been reported (reviewed in Reference 12). We observed a higher reported total daily caloric intake in the cases than in the controls. To possibly explain this, we examined weight, body mass index, and crude estimators of physical activity (in adolescence, age 20, and adulthood) and found them to be similar between cases and controls (data not shown). There has been considerable controversy over whether increased total caloric intake itself increases cancer risk or whether a dietary component (such as saturated fat) is the factor that increases risk. For example, Slattery and co-workers[19] reported that colon cancer risk was not associated with dietary fat, protein, or carbohydrate intake after adjustment for total energy intake. Certainly, the finding that caloric restriction in mice reduces the incidence of sporadic and induced tumors supports the hypothesis that higher caloric intake increases cancer risk.[20] However, to conclude that higher saturated fat intake is associated with increased risk of testicular tumors or to speculate on possible biologic mechanisms would be premature. Although we report many interesting associations, there are several limitations in our study. Because cases reported diet for the year previous to diagnosis and controls reported diet for the previous year, we were concerned that the discrepancy in reporting years could affect our results. We therefore introduced a term for year of diagnosis into the model to evaluate the impact on the point estimates, but that had little effect. We also split the case group by stratifying on the year of the case's diagnosis (i.e., 1990-93 and 1994-96) and compared these results with those of our analysis for all cases and controls combined; again there were no differences in the point estimates. Of greater concern is the potential for bias in dietary recall inherent in case-control studies, although Willett[12] suggests that diet can be adequately recalled for up to 10 years with acceptable levels of misclassification. In comparisons of prospectively and retrospectively collected diet data, Wilkens and co-workers[21] reported that fat intakes were overreported by colon cancer cases compared with controls but not by prostate or breast cancer cases. However, a prospectively designed study for a rare tumor (such as testicular cancer) would be inefficient. It is therefore impossible to determine the extent and direction of recall bias among the testicular cancer patients and controls, although this remains a possible explanation for our observations. We used friends of cases as controls, because we thought population-based controls would not reflect the population from which the cases arose, despite the problem that friends might be too "similar" to cases. We do not believe "over-matching" occurred, because our controls were significantly older and better educated than our cases. However, this implies that the associations found between diet and testicular cancer may be spurious because of the inclusion of more highly educated controls. For example, the more educated controls may be more conscious of fat intake and methods to reduce dietary fat consumption. Also, participation in this study was less than ideal, particularly for cases. The effect of the low response on the associations found with diet is difficult to predict but remains a concern in this investigation. There were several limitations to our study, but it did reveal specific dietary components that were associated with testicular cancer risk. Because high total and saturated fat and meat intakes were found to be consistently and significantly associated with risk of testicular cancer, more dietary studies addressing this issue are needed. Cholesterol intake, although likely correlated with meat consumption, should also be examined. Future research should include an evaluation of dietary calcium and fiber, inasmuch as we found higher intakes reduced risk. Clarification of the role of genetic and early life exposures in the etiology of testicular cancer and the identification of other risk factors (including diet) could provide further insights into this cancer that occurs at a relatively young age. Acknowledgments and Notes The authors thank all the men who participated in the study, Dr. Maureen Goode (Dept. of Scientific Publications, The University of Texas M. D. Anderson Cancer Center) for editing the manuscript, and Dr. Philip Beckett (Texas Children's Hospital, Baylor College of Medicine) for helpful comments. This work was supported by The University of Texas M. D. Anderson Cancer Center Education Program in Cancer Prevention, which is supported by National Cancer Institute Grant R25-CA-57730. Nutr & Cancer, 3/03 Testicular Carcinoma & Dietary Issues Remember we are NOT Doctors and have NO medical training. This site is like an Encylopedia - there are many pages, many links on many topics. Support our work with any size DONATION - see left side of any page - for how to donate. You can help raise awareness of CAM.