Printer Friendly

Obesity, body composition, and risk of renal cell cancer: a population-based case-control study.

INTRODUCTION

Kidney and renal pelvis cancers account for nearly 4 percent of all new cancer cases in the United States, with 54,390 cases projected for the year 2008. (1,2) Over the past thirty years, kidney cancer incidence rates have nearly doubled--from 7.1 per 100,000 in 1975 to 13.3 per 100,000 in 2005. (1) Renal cell carcinoma (RCC) accounts for approximately 85 percent of all renal tumors. (3) Because genetic predispositions explain only 2 percent of RCC cases (3), incidence trends are likely due to environmental or lifestyle factors. The prevalence of obesity in particular has risen at a corresponding rate in the United States, from 13 percent of the population in 1960-1962 to 32 percent in 2001-2004 (4), implicating the role of population-level changes in body composition on trends in RCC incidence.

The positive association between obesity and risk of RCC has been established in the literature and recognized by the International Agency for Research on Cancer. (5,6) Initially, increased risks were observed only among women in studies using body mass index (BMI) as an obesity indicator. (7-10) A quantitative review of 22 cohort and case-control studies subsequently found equally strong associations between men and women, with a summary relative risk of 1.07 (95 percent CI: 1.05--1.09) per unit increase in BMI. (5)

Most studies of RCC and body composition have used BMI as an obesity indicator. (5) While BMI functions as an efficient measure of body composition in population samples, it remains an indirect indicator of body fat, which is presumably the etiologic factor in cancer development. To date, measures of body fat distribution in studies of RCC have been limited to waist circumference and waist-to-hip ratio (WHR). (11,12) We are not aware of any RCC studies using skinfold thickness as an adjunct measure of body fat distribution, nor of studies that assess body fat composition. Skinfold measurement is considered reliable for estimating peripheral adiposity (13), the influence of which can be compared against measures of central adiposity. Body fat composition as determined by bioelectrical impedance analysis (BIA) can complement BMI and body fat distribution measures. Recent studies using BIA have found increased risk of cancers of the breast (14,15), esophagus16, and rectum (17) with increasing body fat.

The present paper reports findings from a population-based case-control study of RCC in Florida and Georgia, evaluating the role of body composition using anthropometric and self-report measures. Study factors include height, weight, BMI, body fat composition, and body fat distribution measured at time of interview, as well as self-reported past weight. Measures of body fat distribution include hip and waist circumference, and skinfold thickness at the abdomen, triceps, suprailiac crest, and thigh.

SUBJECTS AND METHODS

Study Population

Ethical approval for this study was obtained from the institutional review boards of the University of Florida, the Florida Department of Health, Georgia State University, and Emory University Hospital. Cases with incident, histologically confirmed RCC were identified from records in three participating hospitals in North Florida and the Florida Cancer Data System registry. All white and African-American cases aged 20 years or older and diagnosed between January 1, 2000 and December 31, 2004 were considered for inclusion. Cases were excluded if their cancer was diagnosed in a transplant kidney or if they did not reside in Florida or Georgia. Among 459 living, eligible cases who could be contacted, 316 (69 percent) agreed to participate and were interviewed by trained personnel. An additional 15 cases were included from urology offices in North and Central Florida, and four from Emory University Hospital in Atlanta, producing a sample of 335 cases.

A sample of population controls was identified using random-digit dialing (RDD)18, frequency-matched to cases by age (+/5 years), sex and race. Until an adequate sample of cases had been acquired, frequencies from Surveillance, Epidemiology, and End Results (SEER) data were used for matching. (19) During the final year of the study, random sampling of telephone lists specific to African-American households was used to supplement RDD and ensure an adequate sample of African-American controls. Respondents were eligible as controls if they met the matching criteria and reported no history of kidney disease. Among 801 eligible respondents who could be contacted by telephone, 337 (42 percent) participated as controls.

Measurements

Participants were interviewed in-person by personnel trained in epidemiologic interviewing, dietary assessment and anthropometry. Data was collected using a structured questionnaire, which assessed medical, occupational, and family histories, lifetime use of cigarettes, beverages and artificial sweeteners, and lifetime exposure to radiation, pesticides and environmental tobacco smoke. For cases, age at diagnosis was known and used as a reference for assessing all exposure histories, minimizing potential exposure misclassification. A 70-item Block food frequency questionnaire (20) was used to assess past nutritional intake for a single one-year period.

At time of interview, weight in kilograms and height in centimeters were taken without shoes. Weight was measured using a TANITA BF-350 Body Composition Analyzer/Scale (TANITA Corporation of America, Arlington Heights, IL.), and height using a GPM anthropometer (Siber Hegner, Zurich, Switzerland). Self-reported weight at age 20, age 40, age 60, and maximum lifetime weight were collected as part of the epidemiologic interview. Body mass index at time of interview was calculated by dividing weight in kilograms by height in meters squared. Classification of BMI used four groups derived from the obesity classification system of the World Health Organization (21): Normal = 18.5-24.9; Overweight = 2529.9; Obese =30-39.9; Extremely obese = > 40.

Body fat distribution was assessed at time of interview by waist and hip circumference in centimeters using a GPM measuring tape (Siber Hegner, Zurich, Switzerland) and skinfold thickness in millimeters at the triceps, abdomen, suprailiac crest and thigh using skinfold calipers (Holtain Ltd., Wales, UK). Three separate caliper readings were taken at each skinfold site, and an average of the three was calculated. Caliper measurements had an upper limit of 45 mm. Two additional skinfold measures were computed for "core" subcutaneous fat (abdomen + supraliliac) and at all four sites combined. To compare the influence of central and peripheral fat distribution, variables were calculated representing the percentage of each individual site within the overall measure. Body fat percentage at time of interview was measured using two methods: (1) foot-to-foot bioelectrical impedance (TANITA); and (2) four-site skinfold thickness measurements. Bioelectrical impedance could not be used on subjects with pacemakers or internal electronic devices, representing 6 percent of cases and 4 percent of controls. Sex-specific formulas for the estimation of body fat by skinfolds were taken from Golding et al. (1989). (22)

Statistical Methodology

Analyses of demographic factors between cases and controls employed the Pearson chi-square test for independence. Measures of relative risk were estimated by the odds ratio (OR) and 95 percent confidence interval (CI), using unconditional logistic regression. Potential confounders in the association between RCC and body composition included medical history, tobacco use, daily caloric intake, and daily fat intake. The influence of these factors was tested in models adjusted for age, race, smoking (pack-years), and BMI and stratified by sex. Subsequent models for the influence of body composition factors on RCC were controlled for age, sex, race, smoking (pack-years), education, daily caloric intake, and daily fat intake. Tests for trend employed the Wald chi-square statistic, computed for continuous variables within adjusted models. All tests for statistical significance were two-sided and evaluated a [+ or -] = 0.05. Statistical analyses were performed using SPSS 14.0 software (SPSS Inc., Chicago, IL.).

RESULTS

Demographic information was complete for all 335 cases and 337 controls (Table 1), except for one case who declined to report educational status. Cases and controls did not differ significantly with respect to sex, race, or annual household income. Although controls tended to be better educated than cases, this difference was not statistically significant (P = 0.16). The mean age among cases at time of interview was 66 years, while the mean age among controls was 62 years a difference that was statistically significant (P = 0.001). Cases were interviewed between 0.4 years and 6.3 years following diagnosis, with a mean diagnosis-to-study interval of 3.1 years.

Table 2 provides adjusted odds ratios for selected risk factors for RCC, stratified by sex. Notably, high levels of daily caloric and fat intake were positively associated with RCC among women, but not men. Conversely, pack-years of tobacco use produced a significant association with RCC for men (P = 0.008) but not for women. These three potential confounders were chosen as covariates in subsequent adjusted models.

Table 3 provides t-tests for the statistical difference in means between cases and controls on continuous body composition measures. Few measures of past weight collected by self-report were significantly different between cases and controls. Differences were generally stronger for relative than for absolute measures of obesity. The greatest differences were observed for measures of body fat percentage, whether by BIA or skinfold thickness. Body fat percentage calculated from skinfold thickness was consistently lower than that calculated from BIA, likely due to the proportion of subjects whose skinfold thickness exceeded the upper limit on one or more caliper measurements (22 percent). Mean body fat percentage by BIA was 35.6 percent among cases and 33.6 percent among controls (P = 0.006). Cases also had significantly greater mean values than controls for maximum lifetime BMI (P = 0.026), BMI at time of interview (P = 0.033), waist (cm) at time of interview (P = 0.031), WHR (P = 0.002), and skinfold thickness (mm) measured at the suprailiac crest (P = 0.020), thigh (P = 0.034), core (P = 0.018), and all four sites combined (P = 0.015).

Adjusted odds ratios for height, weight, and relative weight are presented on Table 4. No significant differences between cases and controls were observed for measured height in centimeters or self-reported weight in kilograms at ages 20, 40, or 60 (results not shown). Measures of BMI were significantly associated with RCC among men, but not women. Men in the category of "extreme obesity" (BMI > 40) at time of interview were 3.5 times more likely to have RCC than those in a healthy BMI range. Maximum lifetime BMI was also significantly associated with RCC among men, with an odds ratio of 7.3 for extreme obesity compared with healthy BMI. These results should be carefully interpreted, given the wide confidence intervals that result from low cell counts in certain BMI categories.

Table 5 presents adjusted odds ratios for RCC by sex for body fat percentage at time of interview using both BIA and skinfold thickness methods. Body fat percentage by BIA produced significant tests for trend for both men (P = 0.042) and women (P = 0.032), although when the variable was treated as categorical, significant risk increases were not observed for men. Body fat percentage by skinfold thickness was significantly associated with RCC in men, with those in the highest quartile having a risk of RCC 2.5 times greater than those in the lowest quartile. Among women, body fat percentage by BIA proved to be the better predictor of RCC, with those in the highest quartile being nearly twice as likely to have RCC as those in the lowest quartile. Within the entire sample, having a body fat percentage of 40 percent or greater increased risk for RCC by 73 percent (OR = 1.73; 95% CI: 1.16, 2.57) (not shown in table). Men and women have different standards for what is considered a healthy range of body fat percentage (22), with higher allowed values for women than men. However, we considered a 40 percent threshold adequately high for observing increased risks in both sexs in non-stratified analyses. Correspondingly, when this analysis was stratified by sex, this threshold imposed a greater risk on men (OR = 2.92) than women (OR = 1.44).

Table 6 provides adjusted odds ratios for RCC by sex for selected body fat distribution measures. Among men, being in the highest quartile of waist circumference distribution produced a marginally significant 60 percent risk increase. Waist-to-hip ratio proved a more sensitive predictor of RCC risk among men, with those in the highest quartile being twice as likely to have RCC as those in the lowest quartile. When treated as a dichotomous variable for the entire sample (results not in table), WHR of 1.00 or greater increased risk of RCC by 60 percent (OR = 1.60; 95% CI: 1.03, 2.49). Most skinfold thickness measurements were not associated with RCC for either sex, whether treated as absolute or relative to the total of all four sites. Only results for suprailiac crest skinfolds are shown on the table. Abdominal skinfold thickness alone was not associated with RCC in men or women. However, among men, those in the highest quartile of suprailiac crest skinfold thickness had an odds ratio of 2.1 compared with those in the lowest quartile (P = 0.007). Core skinfold thickness (abdomen + suprailiac crest) was also associated with RCC among men (P = 0.036), although the association was weaker than for suprailiac crest alone--an expected result when combining a significant factor with a non-significant factor in the same measure.

All tests for the influence of lifetime weight gain on risk of RCC produced non-significant results, whether treated as absolute or relative weight. Gain in BMI from age 20 to maximum lifetime weight produced a marginally significant linear trend among men (P = 0.085), although when the variable was treated as categorical no significant associations were observed. To test whether possible cancer-related weight loss among cases was producing underestimated associations between RCC and body composition, we also assessed the influence of weight and BMI loss from maximum lifetime weight to time of interview. No significant differences were observed between cases and controls for either absolute or relative weight loss.

Lastly, race-stratified analyses were performed on body composition measures to determine whether RCC risk increases varied between whites and African-Americans. Among whites, significant RCC risk increases were observed for being obese (BMI > 30) at time of interview (OR = 1.7; 95% CI: 1.0, 2.7) and at maximum lifetime weight (OR = 2.1; 95% CI: 1.1, 4.0), and for being in the highest quartiles of body fat percentage (OR = 2.3; 95% CI: 1.2, 4.2) and core skinfold thickness (OR = 1.8; 95% CI: 1.1, 3.1). The sample of African-American participants (n = 144) was not large enough to detect associations between RCC and most body composition factors. However, among African-Americans, those in the highest quartile of waist-to-hip circumference had a marginally significant odds ratio of 3.0 (95% CI: 0.9, 10.1) compared with those in the lowest quartile, and a significant linear trend (P = 0.033).

DISCUSSION

The present study found recent BMI to be significantly associated with RCC among men, but not women--a finding that conflicts with other studies reporting significant risk increases for both sexs. (5,11) Most case-control studies of RCC have assessed BMI using self-reported height and weight, rather than anthropometric measures. (8-10, 23-31) Among these studies, reported odds ratios for the highest quartile of recent BMI range from 1.2 to 2.7 among women and from 1.3 to 3.3 among men. Recently, a population-based case-control study in Canada assessed odds ratios of RCC for clinical obesity (defined as BMI > 30), using the same standards as those in the present paper. (31) Use of this reference system makes findings more comparable among studies and facilitates the application of research findings to the practice of health promotion.

Because measures of current or recent BMI are proximal to diagnosis among cases, they are generally considered as indicators of past body composition, which is likely the more relevant factor. The present study found no associations between RCC and self-reported past obesity (with the exception of maximum lifetime weight and BMI), while the literature reports inconsistent findings on past weight (8,10,24,27,32,33) and weight gain. (24,28,30) However, two studies have found that BMI at age 20 was associated with increased risk for RCC among both women and men. (8,24) Cohort studies have also shown that past levels of obesity 34-37 and change in BMI since age 20 (38) are associated with risk for RCC.

Research has suggested a number of potential mechanisms explaining the association between obesity and risk of RCC. These include: (1) metabolic syndrome39; (2) hormonal influences (28); (3) associations between obesity and hypertension (40); (4) use of amphetamines or diuretics (30); (5) dietary factors (24,40); and (6) lipid peroxidation. (41) Of these, investigations into metabolic syndrome and diet have produced promising leads in understanding how obesity relates to RCC etiology. Metabolic syndrome results when excess weight increases circulating levels of free fatty acids and peptide hormones, leading to insulin resistance and compensatory hyperinsulinaemia. (39) Carcinogenesis may result because insulin and insulin-like growth factor 1 promote cellular proliferation and inhibit apoptosis in many tissue types. Increased WHR is believed to be associated with these changes (12), and waist circumference may function as a reliable indicator of metabolic changes that is less invasive than sampling blood serum. This study found positive associations between WHR and RCC, particularly among men and African-Americans. Central subcutaneous fat, as measured by skinfold thickness at the suprailiac crest, was also associated with RCC among men in this study. This represents the first known investigation of associations between skinfold thickness and RCC.

Among women, this study's results for dietary factors were stronger than results for body composition factors. Controlling for BMI, daily caloric and fat intake were both significantly associated with RCC in women. Elevated cholesterol and lipid levels that are associated with obesity may support tumor development and progression by an inhibitory effect on immune cells. (40) Asal et al. (1988) suggested that diets high in fat lead to arterionephrosclerosis, which in combination with cholesterol deposits in the renal tubules may enhance the development of renal cancer. (24) Further research is warranted to determine whether dietary factors explain links between obesity and RCC, or whether dietary and body composition factors independently increase risk for RCC, allowing for expected interactions between diet and obesity.

A number of limitations should be acknowledged when interpreting the results of this study. Among cases, the average time from diagnosis to follow-up was 3 years, leading to potential survivor bias in which excluded deceased cases had significantly different rates of exposure than those included in the sample. Yu et al. (1991) reported that obese patients diagnosed with RCC may have a better prognosis than non-obese RCC patients. (42) Interviewed cases in the present study may therefore have been more obese than those who did not survive long enough to participate, resulting in over-estimates of population risk ratios. The correspondence of our findings with those of prospective studies that avoid such biases (34-38), suggests that survivor bias was not likely. This study's low response rates for both cases (69 percent) and controls (42 percent) suggest that selection biases may have been present. While the low participation rate among controls is of concern, it should be noted that differences in education and income--cited among those factors typically differing as a result of this participation bias (43)--were not significantly different between cases and controls. Despite efforts to frequency match controls by age (+/- 5 years), controls were significantly younger than cases. This discrepancy occurred as a result of obtaining initial matching frequencies from SEER data, which represent national cancer statistics, while the case sample for this study was limited geographically to North Florida and Georgia.

It has been found that obese subjects are more likely to under-report while underweight subjects are more likely to over-report their body weight. (44) This may lead some investigators using self-report measures of obesity to report no association with renal cell cancer, when in fact a real association exists (Type II error). Evidence for the association between obesity and RCC should therefore be considered stronger when obesity indicators are based on anthropometric measures. In the present study, anthropometric measures of obesity revealed generally stronger associations than self-report measures, which are subject to recall errors. However, it should be noted that skinfold thickness measurements are subject to high inter-rater variability. (22) Bioelectric impedance analysis was considered the more reliable measure of body fat percentage.

This study adds to present epidemiologic findings that high BMI is a risk factor for renal cell cancer, although the lack of association among women conflicts with reports in the literature. It is likely that BMI functions as an indirect indicator of some other obesity-related variable, such as body fat percentage or body fat distribution, both of which were assessed in the present study. Body fat percentage proved to be a significant predictor of RCC among both men and women, while WHR and suprailiac skinfold thickness predicted RCC among men. Our findings provide tentative evidence for metabolic syndrome and diet as potential mechanisms explaining the link between obesity and RCC. Subsequent studies of body composition and RCC would benefit from a more comprehensive anthropometric assessment of obesity, evaluating the relative impact of absolute and relative weight, body fat percentage, and body fat distribution.

Acknowledgements

This research was supported by a Research Scholar grant from American Cancer Society (TURSG-02-068-01-PBP). Additionally, the authors would like to thank the research assistants who participated in this project for their effort in soliciting cases and controls and conducting interviews.

Conflict of interest: None

REFERENCES

(1.) Ries LAG, Melbert D, Krapcho M, Stinchcomb DG, Howlander N, Horner MJ, et al (eds). SEER Cancer Statistics Review 1975-2005. Bethesda, MD: National Cancer Institute. http://seer.cancer.gov/csr/1975_2005/, Accessed June 2, 2008.

(2.) American Cancer Society: Cancer Facts and Figures 2008. Atlanta, GA: American Cancer Society. 2008.

(3.) Lipworth L, Tarone RE, McLaughlin JK. The epidemiology of renal cell carcinoma. J Urol 2006;176: 2353-2358.

(4.) National Center for Health Statistics: Health, United States, 2006. Hyattsville, MD: National Center for Health Statistics, 2006.

(5.) Bergstrom A, Hsieh C-C, Lindblad P, Lu CM, Cook NR, Wolk A. Obesity and renal cell cancer--a quantitative review. Br J Cancer 2001;85: 984-990.

(6.) International Agency for Research on Cancer. IARC Handbooks of Cancer Prevention. Lyon, France: IARC, 2002.

(7.) Wynder EL, Mabuchi K, Whitmore, WF Jr. Epidemiology of Adenocarcinoma of the Kidney. J Natl Cancer Inst 1974;53:1619-1634.

(8.) Yu M, Mack TM, Hanisch R, Cicioni C, Henderson BE. Cigarette smoking, obesity, diuretic use, and coffee consumption as risk factors for renal cell carcinoma. J Natl Cancer Inst 1986;77:351-356.

(9.) McLaughlin JK, Mandel JS, Blot WJ, Schuman LM, Mehl ES, Fraumeni JF Jr.A population-based case-control study of renal cell carcinoma. J Natl Cancer Inst 1984;72:275-284.

(10.) Kreiger N, Marrett LD, Dodds L, Hilditch S, Darlington GA. Risk factors for renal cell carcinoma: results of a population-based case-control study. Cancer Causes Control 1993;4:101-110.

(11.) Pischon T, Lahmann PH, Boeing H, Friedenreich C, Norat T, Tjinneland A, et al. Body size and risk of renal cell carcinoma in the European Prospective Investigation into Cancer and Nutrition (EPIC). Int J Cancer 2006;118:728-738.

(12.) Prineas RJ, Folsom AR, Zhang ZM, Sellers TA, Potter J. Nutrition and Other Risk Factors for Renal Cell Carcinoma in Postmenopausal Women. Epidemiology 1997;8: 31-36.

(13.) Ketel IJG, Volman MNM, Seidell JC, Stehouwer CD, Twisk JW, Lambalk CB. Superiority of skinfold measurements and wait over waist-to-hip ratio for determination of body fat distribution in a population-based cohort of Caucasian Dutch adults. Eur J Endocrinol 2007;156:655-661.

(14.) MacInnis RJ, English DR, Gertig DM, Hopper JL, Giles GG. Body Size and Composition and Risk of Postmenopausal Breast Cancer. Cancer Epidemiol Biomarkers Prev 2004;13:2117-2125.

(15.) Krebs EE, Taylor BC, Cauley JA, Stone KL, Bowman PJ, Ensrud KE. Measures of Adiposity and Risk of Breast Cancer in Older Postmenopausal Women. J Am Geriatr Soc 2006;54: 63-69.

(16.) MacInnis RJ, English DR, Hopper JL, Giles GG. Body size and composition and the risk of gastric and oesophageal adenocarcinoma. Int J Cancer 2006;118: 2628-2631.

(17.) MacInnis RJ, English DR, Haydon AM, Hopper JL, Gertig DM, Giles GG. Body size and composition and risk of rectal cancer (Australia). Cancer Causes Control 2006;17:1291-1297.

(18.) Waksberg J: Sampling Methods for Random Digit Dialing. J Am Stat Assoc 1978;73:40-46.

(19.) National Cancer Institute: SEER*Stat Database: Incidence--SEER 17 Regs Public-Use, Nov 2005 Sub (2000-2003). Bethesda, MD: National Cancer Institute, 2006.

(20.) Block G, Woods M, Potosky A, Clifford C. Validation of a Self-Administered Diet History Questionnaire Using Multiple Diet Records. J Clin Epidemiol 1990;43:1327-1335.

(21.) World Health Organization: Global Database on Body Mass Index. Geneva, Switzerland: World Health Organization, 2006.

(22.) Golding LA, Myers CR, Sinning WE (eds). Y's Way to Physical Fitness: The Complete Guide to Fitness Testing and Instruction. Champaign, IL: Human Kinetics Publishers, Inc, 1989.

(23.) Goodman MT, Morgenstern H, Wynder EL. A case-control study of factors affecting the development of renal cell cancer. Am J Epidemiol 1986;124: 926941.

(24.) Asal N, Risser D, Kadamani S, Geyer JR, Lee ET, Cherng N. Risk factors in renal cell carcinoma: Methodology, demographics tobacco, beverage use and obesity. Cancer Detect Prev 1988;11: 359-377.

(25.) Maclure M, Willet W. A case-control study of diet and risk of renal adenocarcinoma. Epidemiology 1990;1: 430-440.

(26.) McCredie M, Stewart JH. Risk factors for kidney cancer in New South Wales, Australia. II. Urologic disease, hypertension, obesity, and hormonal factors. Cancer Causes Control 1992;3:323-331.

(27.) Benhamou S, Lenfant MH, Ory-Paoletti C, Flamant R. Risk factors for renal cell carcinoma in a French case-control study. Int J Cancer 1993;55: 32-36.

(28.) Lindblad P, Wolk A, Bergstrom R, Persson I, Adami HO. The Role of Obesity and Weight Fluctuations in the Etiology of Renal Cell Cancer: A Population-Based Case-Control Study. Cancer Epidemiol Biomarkers Prev 1994;3:631-639.

(29.) Muscat JE, Hoffmann D, Wynder EL. The epidemiology of renal cell carcinoma. Cancer 1995;75: 2552-2557.

(30.) Chow W-H, McLaughlin JK, Mandel JS, Wacholder S, Niwa S, Fraumeni JF Jr. Obesity and Risk of Renal Cell Cancer. Cancer Epidemiol Biomarkers Prev 1996;5:17-21.

(31.) Pan SY, DesMeules M, Morrison H, Wen SW, Canadian Cancer Registries Epidemiology Research Group. Obesity, High Energy Intake, Lack of Physical Activity, and the Risk of Kidney Cancer. Cancer Epidemiol Biomarkers Prev 2006;15:2453-2460.

(32.) McLaughlin JK, Gao Y-T, Gao R-N, Zheng W, Ji B-T, Blot WJ, et al. Risk factors for renal-cell cancer in Shanghai, China. Int J Cancer 1992;52: 562565.

(33.) Chiu BC-H, Gapstur SM, Chow WH, Kirby KA, Lynch CF, Cantor KP. Body mass index, physical activity and risk of renal cell carcinoma. Int J Obes (Lond) 2006;30: 940-947.

(34.) Chow W-H, Gridley G, Fraumeni JF, Jarvholm B. Obesity, hypertension and the risk of kidney cancer in men. N Engl J Med 2000;343:1305-1311.

(35.) Bjirge T, Tretli S, Engeland A. Relation of Height and Body Mass Index to Renal Cell Carcinoma in Two Million Norwegian Men and Women. Am J Epidemiol 2004;160:1168-1176.

(36.) Flaherty K, Fuchs CS, Colditz GA, Stampfer MJ, Speizer FE, Willett WC, et al. A prospective study of body mass index, hypertension, and smoking and the risk of renal cell carcinoma (US). Cancer Causes Control 2005;16:10991106.

(37.) Samanic C, Chow WH, Gridley G, Jarvholm B, Fraumeni JF Jr. Relation of body mass index to cancer risk in 362,552 Swedish men. Cancer Causes Control 2006;17:901-909.

(38.) Van Dijk BAC, Schouten LJ, Kiemeney L, Goldbohm RA, van den Brandt PA. Relation of height, body mass, energy intake and physical activity to risk of renal cell carcinoma: Results from the Netherlands cohort study. Am J Epidemiol 2004;160:1159-1167.

(39.) Calle EE, Kaaks R. Overweight, Obesity and Cancer: Epidemiological Evidence and Proposed Mechanisms. Nat Rev Cancer 2004;4:579-591.

(40.) Amling CL. The association between obesity and the progression of prostate and renal cell carcinoma. Urol Oncol 2004;22:478-484.

(41.) Gago-Dominguez M, Castelao J. Lipid Peroxidation and RCC: further supportive evidence and new mechanistic insight. Free Radic Biol Med 2006;40: 721-733.

(42.) Yu ML, Asal NR, Geyer JR. Later recurrence and longer survival among obese patients with renal cell carcinoma. Cancer 1991;68:1648-1655.

(43.) Austin H, Hill HA, Flanders D, Greenberg RS. Limitations in the Application of Case-Control Methodology. Epidemiol Rev 1994;16: 65-76.

(44.) Kuskowska-Wolk A, Karlsson P, Stolt M, Rossner S. The predictive validity of body mass index based on self-reported weight and height. Int J Obes (Lond) 1989;13:441-453.

Nabih R. Asal, Ryan P. Theis, Suzanne M. Dolwick Grieb, Deborah Burr, Dan Benardot, Tariq Siddiqui Department of Epidemiology and Biostatistics, College of Public Health and Health Professions, University of Florida, Gainesville, Florida, USA

Corresponding author: Nabih R. Asal, Ph.D. College of Public Health and Health Professions, University of Florida, P.O. Box 100231, Gainesville, FL 32610-0231, Phone: (352) 273-5363, Fax: (352) 263-5365,

E-mail: asal@phhp.ufl.edu
TABLE 1--Demographic characteristics of renal cell carcinoma cases and
population controls, Florida and Georgia, 2003-2006

 Cases (n = 335) Controls (n = 337)

 No. % No. %

Age at interview
20-39 years 3 0.9 20 5.9
40-59 years 88 26.3 102 30.3
60-79 years 203 60.6 179 53.1
80 + years 41 12.2 36 10.7

Sex
Female 154 46.0 161 47.8
Male 181 54.0 176 52.2

Race
White 262 78.2 266 79.4
African-American 73 21.8 71 21.1

Education
Less than high school 41 12.3 28 8.3
High school diploma 184 55.1 183 54.3
Bachelor's or higher 109 32.6 126 37.4

Annual household income
Less than $15,000 38 11.3 45 13.4
$15,000-$24,999 60 17.9 52 15.4
$25,000-$49,999 107 31.9 98 29.1
$50,000 and higher 111 33.1 130 38.6
Not reported 19 5.7 12 3.6

TABLE 2--Adjusted odds ratios for selected risk factors for renal cell
carcinoma, by sex

 Men

 Cases (%) Controls (%) AOR (a) 95% CI
Medical history
Hypertension 104 (59%) 89 (51%) 1.1 0.7, 1.7
Kidney infections 16 (9%) 7 (4%) 2.1 0.8, 5.4
Kidney stones 44 (25%) 31 (18%) 1.5 0.9, 2.6
Diuretics (b) 51 (29%) 32 (19%) 1.4 0.8, 2.5

Tobacco use (x)
Never-smokers 57 (32%) 57 (32%) 1.0
Less than 5 PY 12 (7%) 31 (18%) 0.4 0.2, 0.9
5 to < 10 PY 11 (6%) 12 (7%) 0.9 0.3, 2.1
10 to < 20 PY 21 (12%) 15 (9%) 1.4 0.6, 3.0
20 or more PY 80 (44%) 61 (35%) 1.3 0.8, 2.1

Daily caloric intake (d)
1st quartile 46 (25%) 43 (25%) 1.0
2nd quartile 52 (29%) 36 (21%) 1.2 0.7, 2.3
3rd quartile 34 (19%) 54 (31%) 0.6 0.3, 1.1
4th quartile 49 (27%) 39 (23%) 1.3 0.7, 2.4

Daily fat intake (g) (e)
1st quartile 45 (25%) 43 (25%) 1.0
2nd quartile 46 (25%) 42 (24%) 0.9 0.5, 1.8
3rd quartile 41 (23%) 47 (27%) 0.8 0.4, 1.6
4th quartile 49 (27%) 40 (23%) 1.2 0.6, 2.3

 Women

 Cases (%) Controls (%) AOR (a) 95% CI
Medical history
Hypertension 80 (53%) 70 (44%) 1.3 0.8, 2.0
Kidney infections 40 (27%) 20 (13%) 2.5 1.4, 4.7
Kidney stones 18 (12%) 12 (8%) 1.7 0.8, 3.8
Diuretics (b) 49 (33%) 37 (24%) 1 0.6, 1.9

Tobacco use (c)
Never-smokers 73 (47%) 78 (48%) 1.0
Less than 5 PY 18 (12%) 18 (11%) 1.1 0.5, 2.3
5 to < 10 PY 3 (2%) 12 (8%) 0.3 0.1, 1.1
10 to < 20 PY 12 (8%) 18 (11%) 0.8 0.4, 1.8
20 or more PY 48 (31%) 35 (22%) 1.4 0.8, 2.4

Daily caloric intake (d)
1st quartile 33 (22%) 45 (28%) 1.0
2nd quartile 34 (23%) 44 (27%) 1.1 0.6, 2.0
3rd quartile 38 (25%) 40 (25%) 1.4 0.7, 2.6
4th quartile 46 (31%) 32 (20%) 2.3 1.2, 4.6

Daily fat intake (g) (e)
1st quartile 27 (18%) 51 (32%) 1.0
2nd quartile 40 (27%) 38 (24%) 2 1.0, 3.8
3rd quartile 38 (25%) 40 (25%) 1.8 0.9, 3.4
4th quartile 46 (31%) 32 (20%) 2.7 1.4, 5.2

(a) Adjusted for age, race, smoking (lifetime pack-years), and BMI

(b) Daily diuretic use for one year or longer; adjusted for age, race,
smoking (lifetime pack-years), BMI, and history of hypertension

(c) Wald x2 test for trend: Men, P = 0.008; Women, P = 0.639

(d) Wald x2 test for trend: Men, P = 0.697; Women, P = 0.027
 Quartile distributions: Men: (1) [less than or equal to] 1430;
 (2) 1431 - 1872;
 (3) 1873 - 2291;
 (4) [greater than or equal to] 2292
 Women: (1) [less than or equal to] 1199;
 (2) 1200 - 1528;
 (3) 1529 - 2031;
 (4) [greater than or equal to] 2032
(e) Wald x2 test for trend: Men, P = 0.957; Women, P = 0.006
 Quartile distributions Men: (1) [less than or equal to] 54.64;
 (2) 54.65 - 74.49;
 (3) 74.50 - 101.29;
 (4) [greater than or equal to] 101.30
 Women: (1) [less than or equal to] 43.14;
 (2) 43.15 - 61.14;
 (3) 61.15 - 84.44;
 (4) [greater than or equal to] 84.45

TABLE 3--Differences in means between renal cell carcinoma cases and
controls on continuous body composition measures

 Cases (n = 335)

 No. Mean

Height (cm) 335 167.48

Weight (kg)
Age 20 334 66.07
Age 40 331 76.58
Age 60 244 80.68
Maximum lifetime (a) 334 95.22
Time of interview (a) 335 87.74
Gain age 20 to maximum lifetime 333 29.18
Loss from maximum to time of interview 334 7.47

BMI
Age 20 (a) 334 23.44
Age 40 331 27.17
Age 60 244 28.82
Maximum lifetime (b) 334 33.74
Time of interview (b) 335 30.97
Gain age 20 to maximum lifetime 333 10.3
Loss from maximum to time of interview 334 2.78

Waist (cm) (b) 334 105.84
Hip (cm) 333 113.67
Waist/Hip (c) 333 0.93

Skinfold thickness (mm)
Abdomen (a) 330 30.2
Suprailiac (b) 333 23.38
Triceps 331 23.77
Thigh (b) 328 27.86
Core (Abdomen + Suprailiac) (b) 330 53.69
Total (all 4 sites) (b) 322 105.62
Core/Total 322 0.509

Bodyfat percentage (%)
Bioelectrical impedance method (c) 315 35.63
Skinfold thickness method (c) 322 29.61

 Controls (n = 337)

 No. Mean

Height (cm) 336 168.04

Weight (kg)
Age 20 335 64.57
Age 40 316 75.21
Age 60 211 79.77
Maximum lifetime (a) 337 91.59
Time of interview (a) 335 84.55
Gain age 20 to maximum lifetime 335 26.82
Loss from maximum to time of interview 335 6.93

BMI
Age 20 (a) 334 22.83
Age 40 315 26.52
Age 60 211 28.39
Maximum lifetime (b) 336 32.39
Time of interview (b) 335 29.86
Gain age 20 to maximum lifetime 334 9.48
Loss from maximum to time of interview 335 2.53

Waist (cm) (b) 336 101.57
Hip (cm) 336 111.43
Waist/Hip (c) 336 0.908

Skinfold thickness (mm)
Abdomen (a) 334 28.78
Suprailiac (b) 335 21.69
Triceps 336 22.99
Thigh (b) 332 26.08
Core (Abdomen + Suprailiac) (b) 334 50.46
Total (all 4 sites) (b) 330 99.39
Core/Total 330 0.509

Bodyfat percentage (%)
Bioelectrical impedance method (c) 324 33.55
Skinfold thickness method (c) 330 28.10

 T-test for difference

 t-statistic P value

Height (cm) 0.741 0.459

Weight (kg)
Age 20 -1.313 0.189
Age 40 -0.905 0.366
Age 60 -0.577 0.564
Maximum lifetime (a) -1.893 0.059
Time of interview (a) -1.925 0.055
Gain age 20 to maximum lifetime -1.548 0.122
Loss from maximum to time of interview -0.59 0.556

BMI
Age 20 (a) -1.748 0.081
Age 40 -1.403 0.161
Age 60 -0.894 0.372
Maximum lifetime (b) -2.229 0.026
Time of interview (b) -2.133 0.033
Gain age 20 to maximum lifetime -1.553 0.121
Loss from maximum to time of interview -0.736 0.462

Waist (cm) (b) -2.161 0.031
Hip (cm) -1.153 0.249
Waist/Hip (c) -3.055 0.002

Skinfold thickness (mm)
Abdomen (a) -1.921 0.055
Suprailiac (b) -2.323 0.02
Triceps -1.124 0.262
Thigh (b) -2.126 0.034
Core (Abdomen + Suprailiac) (b) -2.365 0.018
Total (all 4 sites) (b) -2.435 0.015
Core/Total 0.000 1.000

Bodyfat percentage (%)
Bioelectrical impedance method (c) -2.751 0.006
Skinfold thickness method (c) -2.960 0.003

(a) P < 0.10, (b) P < 0.05, (c) P < 0.01

TABLE 4--Adjusted odds ratios for renal cell carcinoma by sex: height,
weight, and body mass index (BMI)

 Men

 Cases Controls AOR (a) 95% CI
 (%) (%)

Height at interview (cm) (b)
1st quartile 48 (27%) 40 (23%) 1.0
2nd quartile 52 (29%) 37 (21%) 1.4 0.7, 2.6
3rd quartile 41 (23%) 49 (28%) 0.8 0.4, 1.5
4th quartile 40 (22%) 49 (28%) 0.7 0.4, 1.3

Weight at interview (kg) (c)
1st quartile 42 (23%) 46 (26%) 1.0
2nd quartile 46 (25%) 42 (24%) 1.1 0.6, 2.1
3rd quartile 45 (25%) 44 (25%) 1.2 0.6, 2.1
4th quartile 48 (27%) 42 (24%) 1.3 0.7, 2.5

BMI at interview (d)
18.5-24.9 20 (11%) 29 (17%) 1.0
25.0-29.9 69 (38%) 72 (42%) 1.2 0.6, 2.5
30.0-39.9 75 (41%) 62 (36%) 1.7 0.8, 3.4
40.0 or greater 17 (9%) 8 (5%) 3.5 1.2, 10.4

Maximum weight (kg) (e)
1st quartile 43 (24%) 40 (23%) 1.0
2nd quartile 44 (24%) 51 (29%) 0.8 0.5, 1.5
3rd quartile 41 (23%) 45 (26%) 0.8 0.4, 1.5
4th quartile 53 (29%) 40 (23% 1.3 0.7, 2.5

Maximum BMI (f)
18.5-24.9 3 (2%) 14 (8%) 1.0
25.0-29.9 53 (30%) 55 (33%) 4.3 1.1, 16.8
30.0-39.9 96 (54%) 82 (49%) 5.4 1.4, 20.7
40.0 or greater 27 (15%) 17 (10%) 7.3 1.7, 31.5

 Women

 Cases Controls AOR (a) 95% CI
 (%) (%)

Height at interview (cm) (b)
1st quartile 39 (25%) 39 (24%) 1.0
2nd quartile 35 (23%) 44 (27%) 0.9 0.4, 1.6
3rd quartile 42 (27%) 37 (23%) 1.3 0.7, 2.5
4th quartile 38 (25%) 41 (26%) 1.0 0.5, 2.0

Weight at interview (kg) (c)
1st quartile 31 (20%) 47 (29%) 1.0
2nd quartile 42 (27%) 37 (23%) 1.6 0.8, 3.1
3rd quartile 37 (24%) 42 (26%) 1.3 0.7, 2.6
4th quartile 44 (29%) 35 (22%) 2.0 1.0, 4.0

BMI at interview (d)
18.5-24.9 30 (20%) 41 (26%) 1.0
25.0-29.9 49 (32%) 51 (32%) 1.2 0.6, 2.3
30.0-39.9 60 (39%) 49 (31%) 1.6 0.8, 3.0
40.0 or greater 14 (9%) 17 (11%) 1.2 0.5, 2.8

Maximum weight (kg) (e)
1st quartile 31 (20%) 46 (29%) 1.0
2nd quartile 39 (26%) 38 (24%) 1.4 0.7, 2.6
3rd quartile 43 (28%) 36 (22%) 1.7 0.9, 3.2
4th quartile 40 (26%) 41 (26%) 1.6 0.8, 3.1

Maximum BMI (f)
18.5-24.9 16 (11%) 26 (16%) 1.0
25.0-29.9 35 (23%) 45 (28%) 1.1 0.5, 2.3
30.0-39.9 75 (50%) 61 (38%) 1.8 0.9, 3.8
40.0 or greater 25 (17%) 28 (18%) 1.4 0.6, 3.2

(a) Adjusted model controlled for age, race, smoking (lifetime
pack-years), daily caloric intake, daily fat intake, and education

(b) Wald x2 test for trend: Men, P = 0.092; Women, P = 0.653
 Quartile distributions: Men: (1)[less than or equal to] 170.2 cm;
 (2) 170.3 cm - 174.3 cm;
 (3) 174.4 cm - 178.4 cm;
 (4) [greater than or equal to] 178.5 cm
 Women: (1)[less than or equal to] 156.6 cm;
 (2) 156.6 cm - 160.6 cm;
 (3) 160.7 cm - 164.7 cm;
 (4) [greater than or equal to] 164.8 cm

(c) Wald x2 test for trend: Men, P = 0.118; Women, P = 0.146
 Quartile distributions: Men: (1)[less than or equal to] 78.39 kg;
 (2) 78.40 kg - 89.39 kg;
 (3) 89.40 kg - 103.19 kg;
 (4) [greater than or equal to] 103.20 kg
 Women: (1)[less than or equal to] 65.59 kg;
 (2) 65.60 kg - 76.39 kg;
 (3) 76.40 kg - 88.79 kg;
 (4) [greater than or equal to] 88.80 kg
(d) Wald x2 test for trend: Men, P = 0.028; Women, P = 0.295

(e) Wald x2 test for trend: Men, P = 0.043; Women, P = 0.425
 Quartile distributions Men: (1)[less than or equal to] 83.90 kg;
 (2) 83.91 kg - 97.06 kg;
 (3) 97.07 kg - 111.12 kg;
 (4) [greater than or equal to] 111.13 kg
 Women: (1)[less than or equal to] 70.75 kg;
 (2) 70.76 kg - 80.73 kg;
 (3) 80.74 kg - 94.79 kg;
 (4) [greater than or equal to] 94.80 kg
(f) Wald x2 test for trend: Men, P = 0.012; Women, P = 0.643

TABLE 5--Adjusted odds ratios for renal cell carcinoma by sex: body
fat composition at time of interview

 Men

Body fat method Cases (%) Controls (%) AOR (a) 95% CI

Bioimpedance (b)
1st quartile 37 (22%) 45 (27%) 1.0
2nd quartile 45 (27%) 40 (24%) 1.2 0.7, 2.4
3rd quartile 40 (24%) 42 (25%) 1.0 0.6, 2.0
4th quartile 46 (27%) 39 (24%) 1.3 0.7, 2.5

Skinfolds '(c)
1st quartile 36 (21%) 49 (29%) 1.0
2nd quartile 42 (25%) 44 (26%) 1.4 0.7, 2.5
3rd quartile 37 (22%) 49 (29%) 0.9 0.5, 1.8
4th quartile 56 (33%) 30 (17%) 2.5 1.3, 4.7

 Women

Body fat method Cases (%) Controls (%) AOR (a) 95% CI

Bioimpedance (b)
1st quartile 28 (19%) 48 (30%) 1.0
2nd quartile 38 (26%) 36 (23%) 1.8 0.9, 3.5
3rd quartile 39 (27%) 39 (25%) 1.6 0.8, 3.1
4th quartile 42 (29%) 35 (22%) 1.9 1.0, 3.7

Skinfolds (c)
1st quartile 33 (22%) 44 (28%) 1.0
2nd quartile 35 (23%) 42 (27%) 1.0 0.5, 1.9
3rd quartile 40 (26%) 38 (24%) 1.3 0.7, 2.5
4th quartile 43 (29%) 34 (22%) 1.5 0.8, 2.9

(a) Adjusted model controlled for age, race, smoking (lifetime
pack-years), daily caloric intake, daily fat intake, and education

(b) Wald x2 test for trend: Men, P = 0.042; Women, P = 0.032
 Quartile categories: Men: (1) [less than or equal to] 24.29%;
 (2) 24.30% - 29.04%;
 (3) 29.05% - 33.99%;
 (4) [greater than or equal to] 34.00%
 Women: (1) [less than or equal to] 35.69%;
 (2) 35.70% - 40.79%;
 (3) 40.80% - 45.99%;
 (4) [greater than or equal to] 46.00%
(c) Wald x2 test for trend: Men, P = 0.047; Women, P = 0.089
 Quartile categories: Men: (1) [less than or equal to] 22.64%;
 (2) 22.65% - 26.30%;
 (3) 26.31% - 30.07%;
 (4) [greater than or equal to] 30.08%
 Women: (1) [less than or equal to] 28.21%;
 (2) 28.22% - 32.30%;
 (3) 32.31% - 35.72%;
 (4) [greater than or equal to] 35.73%

TABLE 6--Adjusted odds ratios for renal cell carcinoma by sex: body
fat distribution

 Men

 Cases (%) Controls (%) AOR (a) 95% CI
Waist (cm) (b)
1st quartile 42 (23%) 46 (26%) 1.0
2nd quartile 46 (25%) 43 (25%) 1.1 0.6, 2.1
3rd quartile 38 (21%) 52 (30%) 0.8 0.4, 1.4
4th quartile 55 (30%) 34 (19%) 1.6 0.9, 3.1

Waist/Hip ratio (c)
1st quartile 34 (19%) 53 (30%) 1.0
2nd quartile 41 (23%) 49 (28%) 1.4 0.7, 2.6
3rd quartile 53 (29%) 36 (21%) 2.5 1.3, 4.8
4th quartile 52 (29%) 37 (21%) 2.0 1.0, 3.9

Suprailiac crest
skinfold (mm) (d)
1st quartile 42 (23%) 46 (26%) 1.0
2nd quartile 44 (24%) 44 (25%) 1.2 0.6, 2.1
3rd quartile 39 (22%) 51 (29%) 0.9 0.5, 1.7
4th quartile 55 (31%) 33 (19%) 2.1 1.1, 3.9

 Women

 Cases (%) Controls (%) AOR (a) 95% CI
Waist (cm) (b)
1st quartile 31 (20%) 47 (29%) 1.0
2nd quartile 40 (26%) 37 (23%) 1.6 0.8, 3.1
3rd quartile 41 (27%) 40 (25%) 1.5 0.7, 2.9
4th quartile 41 (27%) 37 (23%) 1.5 0.7, 2.8

Waist/Hip ratio (c)
1st quartile 34 (22%) 44 (27%) 1.0
2nd quartile 33 (22%) 45 (28%) 0.9 0.5, 1.8
3rd quartile 39 (26%) 41 (26%) 1.2 0.6, 2.3
4th quartile 47 (31%) 31 (19%) 1.7 0.8, 3.4

Suprailiac crest
skinfold (mm) (d)
1st quartile 36 (24%) 42 (26%) 1.0
2nd quartile 36 (25%) 43 (27%) 1.0 0.5, 1.9
3rd quartile 39 (26%) 40 (25%) 1.2 0.6, 2.3
4th quartile 42 (28%) 36 (22%) 1.3 0.7, 2.5

(a) Adjusted model controlled for age, race, smoking (lifetime
pack-years), daily caloric intake, daily fat intake, and education

(b) Wald x2 test for trend: Men, P = 0.211; Women, P = 0.305
 Quartile categories Men: (1) [less than or equal to] 96.99 cm;
 (2) 97.00 cm - 104.99 cm;
 (3) 105.00 cm - 115.07 cm;
 (4) [greater than or equal to] 115.08 cm
 Women: (1)[less than or equal to] 86.34 cm;
 (2) 86.35 cm - 96.99 cm;
 (3) 97.00 cm - 107.52 cm;
 (4) [greater than or equal to] 107.53 cm
(c) Wald x2 test for trend: Men, P = 0.052; Women, P = 0.147
 Quartile categories Men: (1) [less than or equal to] 0.921;
 (2) 0.922 - 0.962;
 (3) 0.963 - 1.007;
 (4) [greater than or equal to] 1.008
 Women: (1) [less than or equal to] 0.816;
 (2) 0.817 - 0.865;
 (3) 0.866 - 0.915;
 (4) [greater than or equal to] 0.916
(d) Wald x2 test for trend: Men, P = 0.007; Women, P = 0.426
COPYRIGHT 2009 Renaissance Medical Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:original article
Author:Asal, Nabih R.; Theis, Ryan P.; Grieb, Suzanne M. Dolwick; Burr, Deborah; Benardot, Dan; Siddiqui, T
Publication:International Journal of Health Science
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2009
Words:7929
Previous Article:The relationship between kidney function tests and systemic lupus erythematosus.
Next Article:Health seekers and healthy misery; the perfection of a healthy death.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters