- Email: [email protected]
Association between body mass index and prostate cancer detection rates in Japanese urologic patients
ADULT UROLOGY
ASSOCIATION BETWEEN BODY MASS INDEX AND PROSTATE CANCER DETECTION RATES IN JAPANESE UROLOGIC PATIENTS TAKASHI KOBAYASHI, KENJI MITSUMORI, KOJI NISHIZAWA, TAKASHI KAWAHARA, KEIJI OGURA, AND YOSHIHIRO IDE
ABSTRACT Objectives. To examine whether an association exists between body mass index (BMI) and prostate cancer detection rates or pathologic features of cancer in Japanese urologic patients. Methods. We studied the age, BMI, and biopsy results of 481 patients who underwent transrectal prostate biopsy. They were stratified by BMI into three groups according to the cutoffs recommended for Asian populations: normal, BMI less than 23.0 (n ⫽ 248); overweight, BMI 23.0 to 25.0 (n ⫽ 116); and obese, BMI greater than 25.0 (n ⫽ 117). We then compared the cancer detection rates and pathologic features among the three groups. Multivariate logistic regression analysis was also performed. Results. No significant differences in the cancer detection rate were found among the three groups (40.2% to 43.1%, P ⫽ 0.87) on univariate analysis. Multivariate logistic regression analyses revealed significant associations between the BMI and cancer detection (P ⫽ 0.029, 95% confidence interval [CI] 0.85 to 0.99), but no significant associations were observed between BMI and the presence of Gleason components 4 or 5 (P ⫽ 0.061, 95% CI 0.79 to 1.01), poor cell differentiation (P ⫽ 0.174, 95% CI 0.96 to 1.24), or clinically organ-confined disease (P ⫽ 0.45, 95% CI 0.84 to 1.08). Conclusions. BMI seems to have a significant impact on prostate cancer detection rates, although it seems difficult to apply the BMI directly to the management of patients at risk of prostate cancer in urologic clinics. UROLOGY 66: 130–134, 2005. © 2005 Elsevier Inc.
S
ome population-based studies have been done on the body mass index (BMI) as a risk factor for prostate cancer incidence, but these have provided inconclusive results. Some have shown a greater incidence for those with a greater BMI,1 some, a greater incidence for those with a lower BMI,2 and some, no significant association between the prostate cancer incidence and the BMI.3 In a recent study of a referral-based biopsy population in the United States, Presti et al.4 clearly showed a greater cancer detection rate in patients with a relatively lower BMI. It is well known that substantial racial differences exist between whites and Japanese in prostate gland properties such as prostate size, serum prostate-specific antigen (PSA) level, From the Departments of Urology and Surgical Pathology, Hamamatsu Rosai Hospital, Hamamatsu, Japan Reprint requests: Takashi Kobayashi, M.D., Department of Urology, Hamamatsu Rosai Hospital, Shogen-cho 25, Hamamatsu 4308525, Japan. E-mail: [email protected] Submitted: November 23, 2004, accepted (with revisions): January 21, 2005 © 2005 ELSEVIER INC. 130
ALL RIGHTS RESERVED
and the clinical incidence of prostate cancer.5 To our knowledge, no studies have been done on the association of BMI with the cancer detection rate in an Asian referral-based biopsy population, for whom the clinical BMI values differ from those in Western countries. We conducted a retrospective review of the association between BMI and biopsy results (presence of cancer, cancer Gleason grade and cell differentiation, and clinical stage) in Japanese patients who had undergone prostate biopsy in a urologic clinic. MATERIAL AND METHODS Between January 2000 and October 2004, 515 Japanese patients suspected of having prostate cancer because of an elevated serum PSA level or abnormal digital rectal examination findings underwent first-time prostate biopsy with 10 cores or more in our department. Although body weight and height were recorded using the measurements or interviews at biopsy in most patients, 34 patients were excluded because of a lack of records mainly due to their bedridden status. Consequently, 481 patients were included in this study (Table I). The BMI was cal0090-4295/05/$30.00 doi:10.1016/j.urology.2005.01.041
TABLE I. Clinical characteristics of 497 study patients and their biopsy results stratified by body mass index Patient Group Overall (n ⴝ 497)
BMI <23.0 (n ⴝ 260)
BMI 23.0–25.0 (n ⴝ 117)
BMI >25.0 (n ⴝ 120)
P Value
71 (41–92) 6.4 (1.2–8500) 39.7 (12.1–223) 20.7 (2.6–192) 41.9 7
73 (41–92) 6.6 (1.8–8500) 36.2 (12.1–230) 19.1 (2.6–192) 42.3 7
71 (49–89) 5.6 (1.2–240) 40.8 (14.9–168) 19.9 (4.2–121) 42.7 6
69 (53–85) 6.6 (2.0–1900) 47.2 (17–132) 23.2 (6.5–94.7) 40.0 7
0.0018† 0.075† ⬍0.0001† 0.0099† 0.89‡ 0.83†
Characteristic Age (yr)* PSA (ng/mL)* Prostate volume (cm3)* Transition zone volume (cm3)* Cancer detection rate (%) Median Gleason score§ WHO nuclear grade (%)§ Well Moderate Poor Clinical stage (%)§ Organ-confined Locally advanced Disseminated
48 (23) 118 (57) 42 (20)
23 (21) 66 (60) 21 (19)
16 (32) 23 (46) 11 (22)
9 (19) 29 (60) 10 (21)
0.43‡
129 (62) 46 (22) 33 (16)
69 (63) 20 (18) 21 (19)
34 (68) 11 (22) 5 (10)
26 (54) 15 (31) 7 (15)
0.27‡
KEY: BMI ⫽ body mass index; PSA ⫽ prostate-specific antigen; WHO ⫽ World Health Organization. * Data presented as the median, with the range in parentheses. † Kruskal-Wallis test. ‡ Chi-square test. § A total of 208 patients with positive biopsy results were analyzed.
culated as the body weight in kilograms divided by the square of the height in meters. The clinical records of the 481 study patients were retrospectively reviewed with regard to age, body weight and height, serum PSA level, total and transition zone prostate volumes, and prostate biopsy results. Serum PSA was quantified using a Tandem-R total PSA assay (Hybritech, San Diego, Calif) before manipulation of the prostate gland. The total and transition zone prostate volumes were measured using transrectal ultrasonography according to a previously described method.6 Prostate biopsy was performed transrectally with a minimum of 10 cores. Although some patients had additional cores to the 10-core systemic biopsy taken at the surgeon’s discretion, the findings from the additional cores were not included in the analysis of the current study. Pathologic examinations were performed by a single pathologist (Y.I.), and the specimens diagnosed as an atypical gland or prostatic intraepithelial neoplasia were scored as negative for cancer. In addition to the diagnosis of malignant or benign disease, the Gleason grade was scored in the patients diagnosed with cancer. The patients were stratified into three groups by the BMI cutoffs generally used in Japan: normal, less than 23.0; overweight, between 23.0 and 25.0; and obese, greater than 25.0. The cancer detection rates and Gleason scores were compared among the BMI groups. To control for the influence of age, PSA level, and prostate volume, logistic regression analysis was performed incorporating these variables in addition to the BMI. Statistical analyses were performed using commercially available software. All tests were two-sided, and statistical significance was applied to P values of less than 0.05.
RESULTS Table I shows the clinical characteristics of the 481 study patients stratified by BMI. Patient age (r ⫽ ⫺0.213, P ⬍0.0001) was inversely correlated UROLOGY 66 (1), 2005
with BMI, but total prostate volume (r ⫽ 0.169, P ⫽ 0.002) and transition zone volume (r ⫽ 0.113, P ⬍0.0001) showed a mild positive correlation with BMI. The PSA level did not show a significant association with the BMI (P ⫽ 0.075). A significant and strong correlation was found between the prostate volume and transition zone volume (r ⫽ 0.92, P ⬍0.0001). A total of 203 patients (42.2%) were diagnosed with prostate cancer. The cancer detection rate did not vary significantly among the BMI groups (40.2% to 43.1%; Table I). In the 228 patients who were 70 years old or younger, the BMI of the patients with prostate cancer was significantly lower than that of the patients without prostate cancer (median 22.7 versus 23.6, P ⫽ 0.025, Mann-Whitney U test). However, logistic regression analysis incorporating age, BMI, PSA level, and prostate volume revealed that a greater BMI was an independent factor for a positive biopsy result, and associations with a greater Gleason score, poor cell differentiation, and clinically organ-confined diseases were not significant (Table II). COMMENT Although some studies have reported on the association between BMI and prostate cancer incidence and mortality, they have yielded inconsistent results.1–3 Few studies have been done on BMI with regard to prostate cancer detection in a referral population.4 To the best of our knowledge, ours 131
TABLE II. Results of multivariate logistic regression analysis of biopsy results Variable Positive biopsy results (overall 497 patients) Age BMI PSA PV Positive biopsy results (464 patients after 33 cases with disseminated disease were excluded) Age BMI PSA PV Gleason 4 or 5 components* Age BMI PSA PV Poorly differentiated nuclear grade on WHO criteria* Age BMI PSA PV Organ-confined disease* Age BMI PSA PV
OR (95% CI)
P Value
0.971 (0.947–0.997) 0.917 (0.849–0.991) 0.942 (0.918–0.967) 1.071 (1.053–1.089)
0.029 0.029 ⬍0.0001 ⬍0.0001
0.974 (0.949–0.999) 0.907 (0.840–0.907) 0.946 (0.923–0.946) 1.069 (1.052–1.069)
0.047 0.014 ⬍0.0001 ⬍0.0001
0.919 (0.878–0.963) 0.872 (0.790–1.005) 0.933 (0.898–0.968) 1.015 (0.993–1.037)
0.0004 0.0613 0.0002 0.1740
1.080 (1.027–1.136) 1.092 (0.962–1.240) 1.001 (1.000–1.002) 0.989 (0.965–1.013)
0.0029 0.1740 0.0463 0.3512
0.932 (0.886–0.981) 0.951 (0.835–1.084) 0.921 (0.889–0.955) 1.001 (0.979–1.024)
0.0073 0.4537 ⬍0.0001 0.9217
KEY: OR ⫽ odds ratio; CI ⫽ confidence interval; BMI ⫽ body mass index; PSA ⫽ prostate-specific antigen; PV ⫽ prostate volume; WHO ⫽ World Health Organization. * A total of 208 patients with positive biopsy results were analyzed.
is the first study on the association between prostate cancer detection and BMI in an Asian population. Because the present study was not a population-based cohort study or a case-control study of a screening population without lower urinary tract symptoms, the aim of this study was not to clarify whether a lower BMI is a risk factor for carcinogenesis and more aggressive biologic potential in prostate cancer. In this regard, we agree entirely with the investigators of a previous study conducted in the United States.4 In the previous study, a lower BMI was significantly associated with greater overall prostate cancer detection rates and rates in patients younger than 70 years, although it was not true in patients 70 years old or older.4 Furthermore, they did not find a significant association between BMI and cancer extent, as measured according to the total core length and cancer grade according to the Gleason score. The same tendencies were found in the current study. The results of these two studies indicate that BMI has a similar effect between the two populations with different genetic and environmental backgrounds. It is a very interesting observation when considering the effect of BMI on prostate cancer detection in referral populations. The association between a lean BMI and a greater cancer 132
detection rate may be explained by the correlations of a lean BMI with older age and a smaller total and transition zone prostate volume, which are associated with greater cancer detection rates. Multivariate logistic regression analysis revealed that a greater BMI was an independent factor for a positive biopsy result. In each case, BMI seems to have some effect on prostate cancer detection among referral urologic patients beyond ethnicity. As the results of the previous study4 and current study have shown, the BMI seems to have some effect on the prostate cancer detection in men 70 years old or younger, although it is not clear why an effect of BMI is not observed in the older subjects. It may be because older patients have numerous other factors affecting their BMI, including other diseases or nutritional problems. The BMI may affect the development of prostate cancer in an earlier period of the male life. A previous report by Schuurman et al.1 showed that a significant association is present between the prostate cancer incidence and BMI at 20 years of age. Their findings suggested that body composition in young adulthood may already exert an effect on the later risk of prostate cancer. No conclusive evidence of the effect of BMI on cancer occurrence or the acquirement of clinically UROLOGY 66 (1), 2005
aggressive properties of prostate cancer has been demonstrated. One possible rationale is a hormonal effect. Excessive adipose tissue is associated with higher estrogen and lower testosterone levels and is compatible with the greater cancer detection rates in patients with a low BMI. Another possible explanation is the “lipid raft” theory. A number of studies have demonstrated that cholesterol accumulates in solid tumors and that cholesterol homeostasis breaks down in the prostate with aging and with the transition to a malignant state.7 However, the relationship between a lower BMI and greater cancer detection is not compatible with the hypothesis that fat consumption is related to prostate cancer risk. With regard to the relationship between fat accumulation and prostate cancer incidence, the two theories conflict. In addition, a number of growth factors are currently thought to be associated with prostate cancer risk and progression. In particular, recent reports have demonstrated that greater levels of insulin-like growth factor-I adjusted by insulinlike growth factor-I binding protein-1 are associated with a greater prostate cancer risk in relatively young (younger than 59 years) men8 and that insulin-like growth factor-I adjusted by insulin-like growth factor-I binding protein-1 is positively related to the BMI in middle-age men.9 However, these findings, which were derived from population-based studies, are incompatible with the results of studies of referral biopsy populations, including the current study, which have demonstrated that a lower BMI is associated with a greater cancer detection rate. This difference might be accounted for by the differences between screening and referral populations. The influence of fat accumulation on carcinogenesis in prostate cancer might alter according to age, prostatic profile, or endocrinologic status. In the present study, the patients were stratified by BMI using cutoffs for Japanese populations. It is well known that the distribution and clinical impact of BMI differs between whites and Japanese.10,11 In recent years, accumulating evidence has suggested that the relationship between BMI and body fat deposits differs between ethnic populations. Asian populations have greater body fat deposits at a lower BMI than do whites.11,12 Considering that the association of abundant adipose tissue with greater estrogen and lower testosterone levels is believed to be one of the possible explanations for the effect of BMI on prostate cancer incidence, it was reasonable to adopt ethnic-based BMI criteria in the present study. The present study did not show a significant association between a lower BMI and advanced cancer, which was observed in the previous study.4 This discrepancy might have resulted from the difUROLOGY 66 (1), 2005
ferent methods used to evaluate the cancer extent. We studied the clinical stage but the cancer extent on the needle cores was evaluated in the previous study. Because the BMI shows a positive correlation with prostate volume, a larger cancer extent on a biopsy core might be observed in patients with a lean BMI in the previous study. Our study had some limitations. In some cases, the BMI was calculated using the body weight and height according to the patient’s report, which seemed less reliable than measurements obtained by the medical staff. Because a cross-sectional value of the BMI was analyzed in this study, the effect of longitudinal BMI exposure on carcinogenesis and biologic features of prostate cancer could not be clarified. More detailed information on the lipid and hormonal profiles such as triglyceride, low and high-density lipoprotein cholesterol, total and free testosterone, estrogens, and insulin-like growth factor would have been useful in the analysis. CONCLUSIONS Although BMI seemed to have a significant effect on the prostate cancer detection rates in this Japanese referral population, the effect of BMI did not manifest directly, and we did not observe significant differences among the BMI groups in the prostate cancer detection rate. The results of this study confirmed that there is a similar tendency in the association between BMI and prostate cancer risk among referral urologic patients of a different ethnic population. REFERENCES 1. Schuurman AG, Goldbohm RA, Dorant E, et al: Anthropometry in relation to prostate cancer risk in the Netherlands Cohort Study. Am J Epidemiol 151: 541–549, 2000. 2. Giovannucci E, Rimm EB, Liu Y, et al: Body mass index and risk of prostate cancer in U.S. health professionals. J Natl Cancer Inst 95: 1240 –1244, 2003. 3. Andersson SO, Wolk A, Bergstrom R, et al: Body size and prostate cancer: a 20-year follow-up study among 135,006 Swedish construction workers. J Natl Cancer Inst 89: 385– 389, 1997. 4. Presti JC Jr, Lee U, Brooks JD, et al: Lower body mass index is associated with a higher prostate cancer detection rate and less favorable pathological features in a biopsy population. J Urol 171: 2199 –2202, 2004. 5. Oesterling JE, Kumamoto Y, Tsukamoto T, et al: Serum prostate-specific antigen in a community-based population of healthy Japanese men: lower values than for similarly aged white men. Br J Urol 75: 347–353, 1995. 6. Terris MK, and Stamey TA: Determination of prostate volume by transrectal ultrasound. J Urol 145: 984 –987, 1991. 7. Freeman MR, and Solomon KR: Cholesterol and prostate cancer. J Cell Biochem 91: 54 – 69, 2004. 133
8. Stattin P, Rinaldi S, Biessy C, et al: High levels of circulating insulin-like growth factor-I increase prostate cancer risk: a prospective study in a population-based nonscreened cohort. J Clin Oncol 22: 3104 –3112, 2004. 9. Sandhu MS, Gibson JM, Heald AH, et al: Association between insulin-like growth factor-I: insulin-like growth factor-binding protein-1 ratio and metabolic and anthropometric factors in men and women. Cancer Epidemiol Biomarkers Prev 13: 166 –170, 2004. 10. Examination Committee of Criteria for “Obesity
134
Disease” in Japan Society for the Study of Obesity: New criteria for “obesity disease” in Japan. Circ J 66: 987–992, 2002. 11. WHO/IASO/IOTF: The Asia-Pacific Perspective: Redefining Obesity and Its Treatment. Bentley WA, Australia, Health Communications Australia Pty Ltd, 2000. 12. Wang J, Thornton JC, Russell M, et al: Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am J Clin Nutr 60: 23–28, 1994.
UROLOGY 66 (1), 2005
ASSOCIATION BETWEEN BODY MASS INDEX AND PROSTATE CANCER DETECTION RATES IN JAPANESE UROLOGIC PATIENTS TAKASHI KOBAYASHI, KENJI MITSUMORI, KOJI NISHIZAWA, TAKASHI KAWAHARA, KEIJI OGURA, AND YOSHIHIRO IDE
ABSTRACT Objectives. To examine whether an association exists between body mass index (BMI) and prostate cancer detection rates or pathologic features of cancer in Japanese urologic patients. Methods. We studied the age, BMI, and biopsy results of 481 patients who underwent transrectal prostate biopsy. They were stratified by BMI into three groups according to the cutoffs recommended for Asian populations: normal, BMI less than 23.0 (n ⫽ 248); overweight, BMI 23.0 to 25.0 (n ⫽ 116); and obese, BMI greater than 25.0 (n ⫽ 117). We then compared the cancer detection rates and pathologic features among the three groups. Multivariate logistic regression analysis was also performed. Results. No significant differences in the cancer detection rate were found among the three groups (40.2% to 43.1%, P ⫽ 0.87) on univariate analysis. Multivariate logistic regression analyses revealed significant associations between the BMI and cancer detection (P ⫽ 0.029, 95% confidence interval [CI] 0.85 to 0.99), but no significant associations were observed between BMI and the presence of Gleason components 4 or 5 (P ⫽ 0.061, 95% CI 0.79 to 1.01), poor cell differentiation (P ⫽ 0.174, 95% CI 0.96 to 1.24), or clinically organ-confined disease (P ⫽ 0.45, 95% CI 0.84 to 1.08). Conclusions. BMI seems to have a significant impact on prostate cancer detection rates, although it seems difficult to apply the BMI directly to the management of patients at risk of prostate cancer in urologic clinics. UROLOGY 66: 130–134, 2005. © 2005 Elsevier Inc.
S
ome population-based studies have been done on the body mass index (BMI) as a risk factor for prostate cancer incidence, but these have provided inconclusive results. Some have shown a greater incidence for those with a greater BMI,1 some, a greater incidence for those with a lower BMI,2 and some, no significant association between the prostate cancer incidence and the BMI.3 In a recent study of a referral-based biopsy population in the United States, Presti et al.4 clearly showed a greater cancer detection rate in patients with a relatively lower BMI. It is well known that substantial racial differences exist between whites and Japanese in prostate gland properties such as prostate size, serum prostate-specific antigen (PSA) level, From the Departments of Urology and Surgical Pathology, Hamamatsu Rosai Hospital, Hamamatsu, Japan Reprint requests: Takashi Kobayashi, M.D., Department of Urology, Hamamatsu Rosai Hospital, Shogen-cho 25, Hamamatsu 4308525, Japan. E-mail: [email protected] Submitted: November 23, 2004, accepted (with revisions): January 21, 2005 © 2005 ELSEVIER INC. 130
ALL RIGHTS RESERVED
and the clinical incidence of prostate cancer.5 To our knowledge, no studies have been done on the association of BMI with the cancer detection rate in an Asian referral-based biopsy population, for whom the clinical BMI values differ from those in Western countries. We conducted a retrospective review of the association between BMI and biopsy results (presence of cancer, cancer Gleason grade and cell differentiation, and clinical stage) in Japanese patients who had undergone prostate biopsy in a urologic clinic. MATERIAL AND METHODS Between January 2000 and October 2004, 515 Japanese patients suspected of having prostate cancer because of an elevated serum PSA level or abnormal digital rectal examination findings underwent first-time prostate biopsy with 10 cores or more in our department. Although body weight and height were recorded using the measurements or interviews at biopsy in most patients, 34 patients were excluded because of a lack of records mainly due to their bedridden status. Consequently, 481 patients were included in this study (Table I). The BMI was cal0090-4295/05/$30.00 doi:10.1016/j.urology.2005.01.041
TABLE I. Clinical characteristics of 497 study patients and their biopsy results stratified by body mass index Patient Group Overall (n ⴝ 497)
BMI <23.0 (n ⴝ 260)
BMI 23.0–25.0 (n ⴝ 117)
BMI >25.0 (n ⴝ 120)
P Value
71 (41–92) 6.4 (1.2–8500) 39.7 (12.1–223) 20.7 (2.6–192) 41.9 7
73 (41–92) 6.6 (1.8–8500) 36.2 (12.1–230) 19.1 (2.6–192) 42.3 7
71 (49–89) 5.6 (1.2–240) 40.8 (14.9–168) 19.9 (4.2–121) 42.7 6
69 (53–85) 6.6 (2.0–1900) 47.2 (17–132) 23.2 (6.5–94.7) 40.0 7
0.0018† 0.075† ⬍0.0001† 0.0099† 0.89‡ 0.83†
Characteristic Age (yr)* PSA (ng/mL)* Prostate volume (cm3)* Transition zone volume (cm3)* Cancer detection rate (%) Median Gleason score§ WHO nuclear grade (%)§ Well Moderate Poor Clinical stage (%)§ Organ-confined Locally advanced Disseminated
48 (23) 118 (57) 42 (20)
23 (21) 66 (60) 21 (19)
16 (32) 23 (46) 11 (22)
9 (19) 29 (60) 10 (21)
0.43‡
129 (62) 46 (22) 33 (16)
69 (63) 20 (18) 21 (19)
34 (68) 11 (22) 5 (10)
26 (54) 15 (31) 7 (15)
0.27‡
KEY: BMI ⫽ body mass index; PSA ⫽ prostate-specific antigen; WHO ⫽ World Health Organization. * Data presented as the median, with the range in parentheses. † Kruskal-Wallis test. ‡ Chi-square test. § A total of 208 patients with positive biopsy results were analyzed.
culated as the body weight in kilograms divided by the square of the height in meters. The clinical records of the 481 study patients were retrospectively reviewed with regard to age, body weight and height, serum PSA level, total and transition zone prostate volumes, and prostate biopsy results. Serum PSA was quantified using a Tandem-R total PSA assay (Hybritech, San Diego, Calif) before manipulation of the prostate gland. The total and transition zone prostate volumes were measured using transrectal ultrasonography according to a previously described method.6 Prostate biopsy was performed transrectally with a minimum of 10 cores. Although some patients had additional cores to the 10-core systemic biopsy taken at the surgeon’s discretion, the findings from the additional cores were not included in the analysis of the current study. Pathologic examinations were performed by a single pathologist (Y.I.), and the specimens diagnosed as an atypical gland or prostatic intraepithelial neoplasia were scored as negative for cancer. In addition to the diagnosis of malignant or benign disease, the Gleason grade was scored in the patients diagnosed with cancer. The patients were stratified into three groups by the BMI cutoffs generally used in Japan: normal, less than 23.0; overweight, between 23.0 and 25.0; and obese, greater than 25.0. The cancer detection rates and Gleason scores were compared among the BMI groups. To control for the influence of age, PSA level, and prostate volume, logistic regression analysis was performed incorporating these variables in addition to the BMI. Statistical analyses were performed using commercially available software. All tests were two-sided, and statistical significance was applied to P values of less than 0.05.
RESULTS Table I shows the clinical characteristics of the 481 study patients stratified by BMI. Patient age (r ⫽ ⫺0.213, P ⬍0.0001) was inversely correlated UROLOGY 66 (1), 2005
with BMI, but total prostate volume (r ⫽ 0.169, P ⫽ 0.002) and transition zone volume (r ⫽ 0.113, P ⬍0.0001) showed a mild positive correlation with BMI. The PSA level did not show a significant association with the BMI (P ⫽ 0.075). A significant and strong correlation was found between the prostate volume and transition zone volume (r ⫽ 0.92, P ⬍0.0001). A total of 203 patients (42.2%) were diagnosed with prostate cancer. The cancer detection rate did not vary significantly among the BMI groups (40.2% to 43.1%; Table I). In the 228 patients who were 70 years old or younger, the BMI of the patients with prostate cancer was significantly lower than that of the patients without prostate cancer (median 22.7 versus 23.6, P ⫽ 0.025, Mann-Whitney U test). However, logistic regression analysis incorporating age, BMI, PSA level, and prostate volume revealed that a greater BMI was an independent factor for a positive biopsy result, and associations with a greater Gleason score, poor cell differentiation, and clinically organ-confined diseases were not significant (Table II). COMMENT Although some studies have reported on the association between BMI and prostate cancer incidence and mortality, they have yielded inconsistent results.1–3 Few studies have been done on BMI with regard to prostate cancer detection in a referral population.4 To the best of our knowledge, ours 131
TABLE II. Results of multivariate logistic regression analysis of biopsy results Variable Positive biopsy results (overall 497 patients) Age BMI PSA PV Positive biopsy results (464 patients after 33 cases with disseminated disease were excluded) Age BMI PSA PV Gleason 4 or 5 components* Age BMI PSA PV Poorly differentiated nuclear grade on WHO criteria* Age BMI PSA PV Organ-confined disease* Age BMI PSA PV
OR (95% CI)
P Value
0.971 (0.947–0.997) 0.917 (0.849–0.991) 0.942 (0.918–0.967) 1.071 (1.053–1.089)
0.029 0.029 ⬍0.0001 ⬍0.0001
0.974 (0.949–0.999) 0.907 (0.840–0.907) 0.946 (0.923–0.946) 1.069 (1.052–1.069)
0.047 0.014 ⬍0.0001 ⬍0.0001
0.919 (0.878–0.963) 0.872 (0.790–1.005) 0.933 (0.898–0.968) 1.015 (0.993–1.037)
0.0004 0.0613 0.0002 0.1740
1.080 (1.027–1.136) 1.092 (0.962–1.240) 1.001 (1.000–1.002) 0.989 (0.965–1.013)
0.0029 0.1740 0.0463 0.3512
0.932 (0.886–0.981) 0.951 (0.835–1.084) 0.921 (0.889–0.955) 1.001 (0.979–1.024)
0.0073 0.4537 ⬍0.0001 0.9217
KEY: OR ⫽ odds ratio; CI ⫽ confidence interval; BMI ⫽ body mass index; PSA ⫽ prostate-specific antigen; PV ⫽ prostate volume; WHO ⫽ World Health Organization. * A total of 208 patients with positive biopsy results were analyzed.
is the first study on the association between prostate cancer detection and BMI in an Asian population. Because the present study was not a population-based cohort study or a case-control study of a screening population without lower urinary tract symptoms, the aim of this study was not to clarify whether a lower BMI is a risk factor for carcinogenesis and more aggressive biologic potential in prostate cancer. In this regard, we agree entirely with the investigators of a previous study conducted in the United States.4 In the previous study, a lower BMI was significantly associated with greater overall prostate cancer detection rates and rates in patients younger than 70 years, although it was not true in patients 70 years old or older.4 Furthermore, they did not find a significant association between BMI and cancer extent, as measured according to the total core length and cancer grade according to the Gleason score. The same tendencies were found in the current study. The results of these two studies indicate that BMI has a similar effect between the two populations with different genetic and environmental backgrounds. It is a very interesting observation when considering the effect of BMI on prostate cancer detection in referral populations. The association between a lean BMI and a greater cancer 132
detection rate may be explained by the correlations of a lean BMI with older age and a smaller total and transition zone prostate volume, which are associated with greater cancer detection rates. Multivariate logistic regression analysis revealed that a greater BMI was an independent factor for a positive biopsy result. In each case, BMI seems to have some effect on prostate cancer detection among referral urologic patients beyond ethnicity. As the results of the previous study4 and current study have shown, the BMI seems to have some effect on the prostate cancer detection in men 70 years old or younger, although it is not clear why an effect of BMI is not observed in the older subjects. It may be because older patients have numerous other factors affecting their BMI, including other diseases or nutritional problems. The BMI may affect the development of prostate cancer in an earlier period of the male life. A previous report by Schuurman et al.1 showed that a significant association is present between the prostate cancer incidence and BMI at 20 years of age. Their findings suggested that body composition in young adulthood may already exert an effect on the later risk of prostate cancer. No conclusive evidence of the effect of BMI on cancer occurrence or the acquirement of clinically UROLOGY 66 (1), 2005
aggressive properties of prostate cancer has been demonstrated. One possible rationale is a hormonal effect. Excessive adipose tissue is associated with higher estrogen and lower testosterone levels and is compatible with the greater cancer detection rates in patients with a low BMI. Another possible explanation is the “lipid raft” theory. A number of studies have demonstrated that cholesterol accumulates in solid tumors and that cholesterol homeostasis breaks down in the prostate with aging and with the transition to a malignant state.7 However, the relationship between a lower BMI and greater cancer detection is not compatible with the hypothesis that fat consumption is related to prostate cancer risk. With regard to the relationship between fat accumulation and prostate cancer incidence, the two theories conflict. In addition, a number of growth factors are currently thought to be associated with prostate cancer risk and progression. In particular, recent reports have demonstrated that greater levels of insulin-like growth factor-I adjusted by insulinlike growth factor-I binding protein-1 are associated with a greater prostate cancer risk in relatively young (younger than 59 years) men8 and that insulin-like growth factor-I adjusted by insulin-like growth factor-I binding protein-1 is positively related to the BMI in middle-age men.9 However, these findings, which were derived from population-based studies, are incompatible with the results of studies of referral biopsy populations, including the current study, which have demonstrated that a lower BMI is associated with a greater cancer detection rate. This difference might be accounted for by the differences between screening and referral populations. The influence of fat accumulation on carcinogenesis in prostate cancer might alter according to age, prostatic profile, or endocrinologic status. In the present study, the patients were stratified by BMI using cutoffs for Japanese populations. It is well known that the distribution and clinical impact of BMI differs between whites and Japanese.10,11 In recent years, accumulating evidence has suggested that the relationship between BMI and body fat deposits differs between ethnic populations. Asian populations have greater body fat deposits at a lower BMI than do whites.11,12 Considering that the association of abundant adipose tissue with greater estrogen and lower testosterone levels is believed to be one of the possible explanations for the effect of BMI on prostate cancer incidence, it was reasonable to adopt ethnic-based BMI criteria in the present study. The present study did not show a significant association between a lower BMI and advanced cancer, which was observed in the previous study.4 This discrepancy might have resulted from the difUROLOGY 66 (1), 2005
ferent methods used to evaluate the cancer extent. We studied the clinical stage but the cancer extent on the needle cores was evaluated in the previous study. Because the BMI shows a positive correlation with prostate volume, a larger cancer extent on a biopsy core might be observed in patients with a lean BMI in the previous study. Our study had some limitations. In some cases, the BMI was calculated using the body weight and height according to the patient’s report, which seemed less reliable than measurements obtained by the medical staff. Because a cross-sectional value of the BMI was analyzed in this study, the effect of longitudinal BMI exposure on carcinogenesis and biologic features of prostate cancer could not be clarified. More detailed information on the lipid and hormonal profiles such as triglyceride, low and high-density lipoprotein cholesterol, total and free testosterone, estrogens, and insulin-like growth factor would have been useful in the analysis. CONCLUSIONS Although BMI seemed to have a significant effect on the prostate cancer detection rates in this Japanese referral population, the effect of BMI did not manifest directly, and we did not observe significant differences among the BMI groups in the prostate cancer detection rate. The results of this study confirmed that there is a similar tendency in the association between BMI and prostate cancer risk among referral urologic patients of a different ethnic population. REFERENCES 1. Schuurman AG, Goldbohm RA, Dorant E, et al: Anthropometry in relation to prostate cancer risk in the Netherlands Cohort Study. Am J Epidemiol 151: 541–549, 2000. 2. Giovannucci E, Rimm EB, Liu Y, et al: Body mass index and risk of prostate cancer in U.S. health professionals. J Natl Cancer Inst 95: 1240 –1244, 2003. 3. Andersson SO, Wolk A, Bergstrom R, et al: Body size and prostate cancer: a 20-year follow-up study among 135,006 Swedish construction workers. J Natl Cancer Inst 89: 385– 389, 1997. 4. Presti JC Jr, Lee U, Brooks JD, et al: Lower body mass index is associated with a higher prostate cancer detection rate and less favorable pathological features in a biopsy population. J Urol 171: 2199 –2202, 2004. 5. Oesterling JE, Kumamoto Y, Tsukamoto T, et al: Serum prostate-specific antigen in a community-based population of healthy Japanese men: lower values than for similarly aged white men. Br J Urol 75: 347–353, 1995. 6. Terris MK, and Stamey TA: Determination of prostate volume by transrectal ultrasound. J Urol 145: 984 –987, 1991. 7. Freeman MR, and Solomon KR: Cholesterol and prostate cancer. J Cell Biochem 91: 54 – 69, 2004. 133
8. Stattin P, Rinaldi S, Biessy C, et al: High levels of circulating insulin-like growth factor-I increase prostate cancer risk: a prospective study in a population-based nonscreened cohort. J Clin Oncol 22: 3104 –3112, 2004. 9. Sandhu MS, Gibson JM, Heald AH, et al: Association between insulin-like growth factor-I: insulin-like growth factor-binding protein-1 ratio and metabolic and anthropometric factors in men and women. Cancer Epidemiol Biomarkers Prev 13: 166 –170, 2004. 10. Examination Committee of Criteria for “Obesity
134
Disease” in Japan Society for the Study of Obesity: New criteria for “obesity disease” in Japan. Circ J 66: 987–992, 2002. 11. WHO/IASO/IOTF: The Asia-Pacific Perspective: Redefining Obesity and Its Treatment. Bentley WA, Australia, Health Communications Australia Pty Ltd, 2000. 12. Wang J, Thornton JC, Russell M, et al: Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am J Clin Nutr 60: 23–28, 1994.
UROLOGY 66 (1), 2005