Leptin Is Associated with Increased Prostate Cancer Risk: A Nested ...

4 downloads 55 Views 313KB Size Report
in normal, high-grade prostatic intraepithelial neoplasia lesions and malignant prostatic epithelium. Moderately elevated plasma leptin concentrations are ...
0021-972X/01/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 2001 by The Endocrine Society

Vol. 86, No. 3 Printed in U.S.A.

Leptin Is Associated with Increased Prostate Cancer Risk: A Nested Case-Referent Study* ¨ R STATTIN, STEFAN SO ¨ DERBERG, GO ¨ RAN HALLMANS, ANNIKA BYLUND, PA RUDOLF KAAKS, ULF-HÅKAN STENMAN, ANDERS BERGH, AND TOMMY OLSSON Departments of Urology and Andrology (P.S.), Public Health and Clinical Medicine (S.S., G.H., T.O.), Geriatric Medicine (A.By.), and Pathology (A.Be.), Umeå University Hospital, SE-901 85 Umeå, Sweden; International Agency for Research on Cancer (R.K.), F-69372 Lyon, France; and Department of Clinical Chemistry (U.-H.S.), Helsinki University Central Hospital, SF-00290 Helsinki, Finland ABSTRACT A Western lifestyle has been implicated in the pathogenesis of prostate cancer. However, no clear association between obesity and prostate cancer has been shown. Leptin may stimulate prostate growth and angiogenesis, and receptors for leptin are present in the prostate. Leptin may, thus, be associated with increased risk of prostate cancer. One hundred forty-nine men with prostate cancer were identified (together with 298 matched referents) who, before diagnosis, had participated in population-based health surveys in Northern Sweden. Blood pressure, body mass index, and use of tobacco were recorded. Leptin, insulin, insulin-like growth factor I (IGF-I), IGF-I-binding proteins 1–3, testosterone, and sex hormone-binding globulin were analyzed in stored samples. Their influences on prostate cancer were estimated by conditional logistic regression analysis. Prostate cancer specimens were investigated for immunoreactivity for the leptin receptor.

M

EN LIVING IN Western countries have a much higher incidence of prostate cancer than men in the Far East, and immigrating Asian men gradually adapt to the incidence in the West (1), strongly suggesting that lifestyle factors are important in the pathogenesis of prostate cancer. A Western lifestyle, encompassing a high caloric intake from an energydense diet and a sedentary lifestyle, resulting in a positive energy balance and often obesity, has consistently been implicated as a risk factor for prostate cancer in international comparisons and experimental studies (2– 4). However, individually based epidemiological studies on prostate cancer and diet have overall shown weaker associations (2), and epidemiological studies on prostate cancer in relation to obesity and physical activity have shown little or no association (2, 5). Besides insulin-like growth factor I (IGF-I), which has been shown to be a risk marker for prostate cancer in several epidemiological studies (6 – 8), no biomarker associated with a Western lifestyle has been linked to prostate cancer risk. Leptin, the adipocyte-derived hormone, regulates satiety and energy expenditure by conveying information to the brain about the size of energy stores (9), and leptin deficiency Received June 28, 2000. Revision received November 20, 2000. Accepted December 2, 2000. Address correspondence and requests for reprints to: Dr. Stefan So¨derberg, Department of Medicine, Umeå University Hospital, se-901 85 Umeå, Sweden. E-mail: [email protected]. * Supported by grants from The Swedish Cancer Society, The Lions Research Foundation (Umeå, Sweden), and The Medical Faculty (Umeå University, Umeå, Sweden).

Relative risk (95% confidence intervals) estimates of prostate cancer over the quintiles of leptin were 1.0, 2.1 (1.1– 4.1), 2.6 (1.4 – 4.8), 1.4 (0.7–2.7), and 1.6 (0.8 –3.2). Adjustments for metabolic variables, testosterone, and IGF-I and its binding proteins did not attenuate this increased risk. Immunoreactivity for the leptin receptor was detected in normal, high-grade prostatic intraepithelial neoplasia lesions and malignant prostatic epithelium. Moderately elevated plasma leptin concentrations are associated with later development of prostate cancer. This may be due to direct effects of leptin on prostatic intraepithelial neoplasia lesions, or to indirect actions through other mechanisms. A critical fat mass related to an interior milieu favorable for prostate cancer development seems to exist, because intermediate but not high leptin levels are related to prostate cancer risk. (J Clin Endocrinol Metab 86: 1341–1345, 2001)

is associated with gross overweight in humans (10). However, apart from rare cases of congenital leptin deficiency (10), circulating levels of leptin are elevated in obesity (11) and increases with high-energy food intake (12). Leptin has also been associated with features of the insulin resistance syndrome (13). High expression of the leptin receptor has been observed in the prostate (14), and circulating leptin levels increase in parallel with prostate growth at puberty in the rat (15). Interestingly, leptin stimulates angiogenesis (16). Our hypothesis was that increased plasma leptin levels are associated with the development of prostate cancer. Subjects and Methods Study populations The two northernmost counties in Sweden (Va¨sterbotten and Norrbotten) with a total population of 510,000 constitute 1 of 39 collaborating centers in the WHO MONICA (Monitoring of Trends and Determinants in Cardiovascular Disease) study (17). Population-based surveys were performed in 1986, 1990, and 1994, and men and women aged 25–74, selected by stratified randomization for age and sex, were recruited. A community intervention program concerning cardiovascular disease and diabetes prevention—The Va¨sterbotten Intervention Program (VIP)—was launched in 1985 in one of the counties (i.e. Va¨sterbotten with a total population of 260,000; Ref. 18). Men and women were invited to a health survey (the same design as the MONICA population surveys) at their primary health care center the year they became 30, 40, 50, and 60 yr of age. Participants in both the MONICA and the VIP surveys were requested to donate a fasting blood sample to be stored at the Medical Biobank at Umeå University Hospital for future research. Between January 1, 1985, and March 31, 1999, 26,856 men participated in the VIP

1341

1342

STATTIN ET AL.

health surveys and 2,704 men in three MONICA surveys. More than 90% of these participants donated blood samples to the Medical Biobank. These samples were collected, processed, and stored in an identical way in the MONICA and VIP surveys. This study was approved by the Research Ethics Committee of Umeå University, and informed consent was obtained from all subjects at the time of health survey.

Case-finding All cases of prostate cancer in the study region are registered in the Regional Cancer Registry. Using a nationwide individual identification number as the identity link, we identified 166 incident cases of prostate cancer that fulfilled two criteria: 1) the case was recorded in the Regional Cancer Registry from January 1, 1985 to March 31, 1999; 2) the case had participated in the MONICA or VIP health survey, and he had donated blood to the Research Blood Bank before the diagnosis of the prostate cancer. Inclusion in other projects precluded the use of 17 cases. Thus, we retained 149 cases for this study. The median time from the health survey to time of diagnosis of prostate cancer was 3.9 yr (interquartile range, 2.0 – 6.0 yr). Additional data on tumor classification was obtained from the Local Primary Prostate Cancer Registry, a regional registry established in 1992. Data on tumors diagnosed before 1992 and after 1997 were extracted from patient medical records, as the registry has a delay of 1 yr from the time of diagnosis to final registration. In the majority of cases, the tumor was localized and highly or intermediately differentiated. Tumors were locally advanced in 16 cases (11%), lymph node metastases were present in 6 cases (4%), metastases at bone scan were present in 14 cases (9%), and the tumors were poorly differentiated in 19 cases (13%). Referents were randomly selected from all cohort members alive and free of cancer at the time of diagnosis of the case, and they were matched for sex, age (⫾1 yr), type (MONICA or VIP), and date (⫾1 yr) of health survey, and geographical region. Finally, two referents for each case were selected.

Biomedical factors and biochemical analyses Smokers were defined as those who reported smoking cigarettes, cigarillos, cigars, or a pipe daily. Ex-smokers or “occasional smokers” were classified as nonsmokers. Snuffers were defined as those who reported daily use of snuff, and all others, including ex-snuffers, were classified as nonsnuffers. Blood pressure, weight, height, body mass index (BMI), glucose, and total cholesterol were measured as described recently (19). BMI was divided into three groups: less than 27 kg/m2, 27–30 kg/m2, and more than 30 kg/m2 according to the classification by Bray (20); and for total cholesterol, three groups were considered clinically relevant: less than 6.5 mmol/L, 6.5–7.8 mmol/L, and more than 7.8 mmol/L. Most subjects were screened in the morning after an overnight fast before blood sampling even if the requested minimum fasting period was only 4 h, for practical reasons. This period was extended to 8 h from 1992 and onward. Plasma samples were kept stored in the Medical Biobank at ⫺80 C before analysis. The leptin analysis was performed

JCE & M • 2001 Vol. 86 • No. 3

with a double antibody RIA with rabbit antihuman leptin antibodies I-labeled human leptin as tracer, and human leptin was used as standard (Linco Research, Inc., St. Charles, MO). Intrabatch coefficients of variation were between 2% and 5%, and the interassay coefficients of variation were 4% for low values (2.1–3.9 ng/mL) and 1% for high values (16.4 –24.6 ng/mL). Data from separate studies on prostate-specific antigen (PSA), testosterone, sex hormone-binding globulin (SHBG), IGF-I, and IGF-I-binding proteins (IGFBPs) 1–3, and insulin was used in multivariate models to address the specific importance of leptin. The PSA levels in plasma were determined by time-resolved immunofluorometric assays (Prostatus PSA; Wallac, Inc. Turku, Finland). Testosterone was determined by a direct RIA (Orion Diagnostica, Oulunsalo, Finland). SHBG was determined by a time-resolved immunofluorometric method (Delfia; Wallac, Inc.). Insulin, IGF-I, IGFBP-1, and IGFBP-3 were measured by double antibody, immunoradiometric assays, whereas IGFBP-2 was measured by a single antibody RIA. Reagents for the insulin assay were obtained from Sanofi Pharmaceuticals, Inc. (Marnes la Coquette, France); for the IGF-I and IGFBP-3 assays from Immunotech (Marseille, France); and for the IGFBP-1 and IGFBP-2 assays from Diagnostics Systems Laboratories, Inc. (Webster, TX). 125

Immunohistochemistry Paraffin-embedded formalin-fixed prostatic tissue from diagnostic needle biopsies from five patients and from five patients operated with radical prostatectomy was retrieved from the files of the Department of Pathology, Umeå University Hospital. Five-micrometer thick sections were cut, treated in a microwave oven at 600 watts for 2 ⫻ 10 min in 0.01 m citrate buffer (pH 6.0), and immunostained with an antibody reactive to the long and short form of human leptin receptor (Ob-R, M-18, sc1834; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), according to the manufacturer. The sections were incubated with antibody (1:500) overnight, and the immunoreaction was detected using a Vectastain Elite kit (Vector, CA), following the supplier’s instructions. Control sections were incubated with an antibody neutralized with excess of Ob-R peptide (sc 1834p; Santa Cruz Biotechnology, Inc.).

Statistical analysis Logarithmic transformation (ln) was used for main study variables, with reduced skewness and kurtosis. Geometric means and 95% confidence intervals (CIs) are presented. Possible interactions between study variables were explored with bivariate (Pearson) and partial (adjusted for age, BMI, and insulin) correlation analysis. Two-tailed tests were used and a P less than 0.05 was considered significant. The influence of studied variables on prostate cancer was analyzed by univariate and multivariate logistic regression analysis using the conditional maximum likelihood routine designed for matched analysis to estimate relative risk and 95% CI. To test the relation between increasing levels of risk factors and the risk of prostate cancer, we initially categorized the continuous variables, except BMI, blood pressure (BP), and cholesterol into quintiles by the distribution of the referent values. For assessment of the influence of leptin, we controlled for metabolic variables, andro-

TABLE 1. Baseline characteristics for cases and referents

Age at health survey (yr) (matched) Mean follow-up (yr) (matched) BMI (kg/m2) Diastolic BP (mm Hg) Systolic BP (mm Hg) Insulin (mIU/mL)a Fasting glucose (mmol/L) Postload glucose (mmol/L) Testosterone (nmol/L)a Leptin (ng/mL)a Smoking (%) Snuffing (%) Values are means and (95% CIs). a Geometric means.

Cases (n ⫽ 149)

Referents (n ⫽ 298)

58.5 (57.7–59.3) 6.5 (6.1– 6.9) 26.3 (25.8 –26.8) 86.0 (84.6 – 87.3) 135.7 (132.9 –138.6) 6.1 (5.4 – 6.8) 5.4 (5.2–5.6) 6.2 (5.9 – 6.5) 22.2 (20.2–24.4) 4.9 (4.4 –5.4) 20.6 14

58.6 (58.1–59.2) 6.5 (6.2– 6.8) 26.5 (26.0 –26.9) 86.6 (85.6 – 87.6) 137.6 (135.6 –139.6) 6.0 (5.6 – 6.5) 5.5 (5.4 –5.7) 6.5 (6.3– 6.7) 23.2 (21.9 –24.6) 4.7 (4.3–5.0) 17.9 19.2

LEPTIN AND PROSTATE CANCER genicity, and IGF-I and their binding proteins in separate and combined models. The number of categories was reduced in these models, a strategy adopted to improve precision (narrower CI), and guided by the results in the univariate analysis. The analysis of leptin consistently showed very similar risk estimates for quintiles 2 and 3 and quintiles 4 and 5, respectively, and these quintiles were, thus, analyzed in combination. Missing values for categorical variables were treated as a separate category in the analysis and omitted from the tables, whereas missing continuous values in the logistic regression analysis were replaced by the mean value representing the reference, thus ensuring a conservative result. All calculations were made with the statistical programs SPSS version 6.1 (SPSS, Inc., Chicago, IL) and STATA version 5.0 (STATA, College Station, TX).

Results

Baseline characteristics and plasma leptin levels are shown in Table 1. There were no significant differences in means between cases and referents for leptin or other variables. Because blood samples had been collected well before time of diagnosis (median, 3.9 yr), we combined data of cases and referents for analysis of interrelationships between leptin and other variables (Table 2). Plasma leptin was strongly correlated to aberrations associated with the insulin resistance syndrome, notably high BMI and hyperinsulinemia. After adjustments for age, BMI, and insulin, a significant correlation with glucose remained. Furthermore, a strong inverse correlation between leptin on one side and testosterone and SHBG on the other remained. The TABLE 2. Bivariate and partial correlation coefficients between leptin and selected variables

a b c

Adjusted for



Age BMI Diastolic BP Systolic BP Cholesterol Insulin Fasting glucose Postload glucose Testosterone SHBG Total PSA IGF-I IGFBP-1 IGFBP-2 IGFBP-3

0.04 0.62a 0.23a 0.16b ⫺0.05 0.53a 0.19a 0.30a ⫺0.29a ⫺0.29a 0.04 ⫺0.08 ⫺0.38a ⫺0.47a 0.04

Age/BMI/insulin

0.09 0.00 0.00 0.11c 0.18b ⫺0.20a ⫺0.13b 0.05 ⫺0.04 ⫺0.05 0.18c 0.05

P ⬍ 0.001. P ⬍ 0.01. P ⬍ 0.05

1343

associations with IGFBP-1 and IGFBP-2 were attenuated after these adjustments. In logistic regression analysis, the unadjusted relative risk of prostate cancer was associated with an intermediate range of plasma leptin. Relative risk estimates of prostate cancer and 95% CI over the quintiles of leptin were 1.0, 2.1 (1.1– 4.1), 2.6 (1.4 – 4.8), 1.4 (0.7–2.7), and 1.6 (0.8 –3.2). Increasing quintiles of leptin were associated with increasing levels of BMI but decreasing levels of testosterone (Table 3). No increase in risk was associated with increasing quintiles of insulin [1.0, 1.0 (0.5–1.8), 1.2 (0.6 –2.3), 1.0 (0.6 –1.9), and 0.9 (0.5–1.6)] nor for increasing BMI [1.0, 1.04 (0.65–1.65), and 0.86 (0.46 –1.62)]. Current use of tobacco was not associated with increased risk of prostate cancer: 1.2 (0.7–2.0) and 0.7 (0.4 –1.2) for smokers and snuffers, respectively. Adjustments for metabolic variables, androgenicity, and IGF-I and its binding proteins in separate and combined multivariate models did not attenuate the increased risk associated with moderately elevated leptin levels (Table 3). Neither was the risk affected by testing these potential confounders as categorical or continuous variables. Subgroup analyses were performed to explore the relative risk in relation to time interval between date of health survey and diagnosis and of age at health survey. In the group with a follow-up time more than median (⬎3.9 yr), the increased risk for intermediately elevated levels of leptin remained significant; the relative risk was 2.7 (95% CI, 1.2– 6.4) for Q2–3 and 1.4 (95% CI, 0.6 –3.4) for Q4 –5, whereas the risk was no longer increased in the group with a short follow-up time (⬍3.9 yr); relative risks of 1.0, 1.5 (95% CI, 0.7–3.4) and 1.4 (95% CI, 0.6 –3.1), respectively. We used 59 yr as the cut-point in the analysis of age effect on risk estimates to avoid dividing the cohort in the middle of a large cluster, because 74% of the entire cohort was surveyed at the age of 60 yr. No difference in risk was seen between age groups (data not shown). Immunoreactive leptin-receptors were observed in the epithelial cells in both benign and malignant glands (Fig. 1). No staining was observed in the prostatic stroma. In control sections, using an antibody preincubated with the Ob-R protein, no immunoreaction was detected (Fig. 1). Notably, a particularly strong immunoreaction was observed in the high-grade prostatic intraepithelial neoplasia (PIN) lesions (Fig. 1). There was no obvious association between staining intensity and histological grade in the manifest tumors.

TABLE 3. Relative risk for prostate cancer by leptin quintiles (Q)

a

Quintiles of leptin

Q1

Q2–3

Q4 –5

Cut-off values for leptin (ng/mL) No of cases/controls Mean BMI (95% CI) Mean testosterone (95% CI) (nmol/L)a Univariate leptin Multivariate leptinb Multivariate leptinc Multivariate leptind Multivariate leptine

ⱕ2.6 18/66 23.2 (22.7–23.6) 27.4 (25.0 –30.0) 1.0 1.0 1.0 1.0 1.0

2.6 –5.5 80/113 25.9 (25.4 –26.3) 24.3 (22.7–26.0) 2.4 (1.3– 4.2) 2.4 (1.3– 4.5) 2.6 (1.3–5.2) 2.4 (1.2– 4.6) 2.3 (1.2– 4.2)

⬎5.5 51/119 28.6 (28.1–29.1) 19.2 (17.4 –21.1) 1.5 (0.8 –2.7) 1.5 (0.7–3.2) 1.5 (0.7–3.4) 1.4 (0.6 –3.2) 1.5 (0.7–3.2)

Geometric means. Leptin adjusted for BMI and insulin. Leptin adjusted for BMI, insulin, cholesterol, fasting and postload glucose, diastolic and systolic BP, smoking, and snuffing. d Leptin adjusted for BMI, insulin, IGF-I, IGFBPs 1–3, testosterone, and SHBG. e Leptin adjusted for all variables included in models c and d. b c

1344

STATTIN ET AL.

JCE & M • 2001 Vol. 86 • No. 3

FIG. 1. Sections of prostatic tissue immunostained for leptin receptors (A and C; bar, 50 ␮m). Leptin receptors are observed in normal prostatic epithelial cells (A, filled arrow) and in prostatic cancer cells (A, unfilled arrows). In control sections incubated with antibody plus excess of leptin-receptor peptide, no immunoreaction was observed (B; bar, 50 ␮m). Particularly strong immunoreactivity was observed in high-grade PIN lesions (C, unfilled arrows) compared with that in cancer (C, filled arrows) or normal prostatic epithelial cells (A, filled arrow).

Discussion

This study, the first nested case referent study on leptin and cancer risk to the best of our knowledge, showed a significantly increased risk for prostate cancer in the intermediate range of circulating leptin. We also found immunoreactivity for leptin receptors in prostatic cancer specimens, with a strong expression in high-grade PIN lesions, a precursor lesion for manifest cancer (21). To date, only two small case referent studies have been published on a possible association between leptin and cancer [i.e. prostate cancer (22) and breast cancer (23), respectively]. None of these studies detected a significant association between leptin and cancer risk, consistent with our findings in the subgroup of men with a short follow-up time. The prospective design of our study ensures that the samples from cases and referents were handled in an identical fashion and also minimizes the risk that alterations in leptin are an effect and not a cause of disease, which may have confounded earlier studies. Acute and chronic disease may, thus, influence leptin levels (24). In addition, there seems to be a specific interval of leptin levels associated with development of prostate cancer. Leptin may promote cancer development through direct effects in the prostate. High leptin receptor messenger RNA levels have been demonstrated in the prostate (14). We now demonstrate the presence of immunoreactive leptin receptors in normal prostatic epithelial cells, in high-grade PIN, and in prostatic cancer cells. Leptin may be involved in the secondary sexual organ growth spurt during puberty (15), and it is, therefore, possible that leptin by direct cellular effects could influence the development and progression of prostate cancer. Other possible modes through which leptin could promote tumor growth is by stimulating angiogenesis, and in experimental studies leptin has been shown to induce angiogenesis (16). This suggests that leptin may be a tumor growth promoter, which is distinctly relevant for prostate cancer, given the fact that PIN lesions are known to exist in 30% of men already at the age of 30 yr (21). Leptin might, thus, stimulate tumor progression by inducing angiogenesis in high-grade PIN lesions and microfocal cancers. Functional leptin receptors have also been demonstrated in the endothelium of certain arteries and veins

(16), but we were not able to detect leptin receptors in vessels within the prostate. If paracrine/autocrine effects of local production of leptin exists remains to be studied. In preliminary studies, we have found leptin expression within the prostate by immunohistochemistry. There is, thus, a need for additional studies on this subject, including studies of possible messenger RNA expression in the prostate. Leptin could also influence prostate growth indirectly through increased sympathetic nervous system activity. In experimental studies, leptin increases sympathetic activity (25) whereas sympathetic denervation decreases prostate size in rats (26). Ecological studies have repeatedly shown a strong association between prostate cancer and high caloric intake (3). The interpretation of these studies is, however, hampered by the presence of confounders related to prostate cancer risk that differ between countries with high and low incidences of prostate cancer (1–3). In contrast, overt obesity is not a risk factor for prostate cancer in epidemiological studies, but mild overweight may be a risk factor (27). This conundrum might be explained by imprecise measurements of fat mass and body composition, and leptin may be a more precise measure of fat mass than BMI (28). An intermediate fat mass may, thus, be associated with an interior milieu favorable for early prostate cancer growth. Hormones known to stimulate prostate growth may be at concentrations inducing a consorted stimulation of prostate cancer development at an intermediate fat mass range. Such hormones include, besides leptin, testosterone, GH, and IGF-I. Experimental studies have consistently shown that testosterone induces prostate growth (29), and one prospective study has shown free circulating testosterone to be positively related to prostate cancer risk (30). GH also induces prostate growth (31) and stimulates hepatic production of IGF-I, which has been shown to be a risk marker for prostate cancer (6 – 8). Additive effects of leptin, testosterone, GH, and IGF-I could, thus, be present at an increased fat mass. Interestingly, caloric restriction has a protective effect against many tumors, including prostate cancer in animal models (4, 32). However, overall, a high serum testosterone level has not been shown to be a consistent risk factor for prostate cancer

LEPTIN AND PROSTATE CANCER

in men. Notably, the insulin resistance syndrome is characterized by low levels of testosterone (33). Marked overweight is also associated with decreasing levels of GH and IGF-I (34). In contrast, estrogens, which may act as protective hormones for prostate cancer (35), are known to be high in overt obesity due to increased aromatization of androgens. This shift in hormone balance with increasing body weight may counteract adverse effects of leptin and other hormones and explain the critical BMI level (i.e. 25.4 –26.3 in our study), associated with a more than 2-fold increased risk for prostate cancer. It is, thus, of major interest in future studies to control for potential confounding effects of estrogens on the development of prostate cancer. The validity of the MONICA and VIP studies has been carefully evaluated (18). The mean coverage for men was ⬃50% in the VIP study with only small differences between participants and nonparticipants with regard to socioeconomic and educational factors. Furthermore, the distribution by tumor stage and grade seen in this study closely resembles the distribution seen in other screening trials (36). Accordingly, we believe that present data are be representative for the male population in Northern Sweden within this age range. Notably, due to the age distribution of our cohort (mean age of diagnosis for the cases in this study was 64 yr vs. 74 yr for the mean prostate cancer case in the county of Va¨sterbotten; our unpublished data), the number of cases with prostate cancer is rather low compared with the total number of cases occurring in the area. Thus, of 261 new cases of prostate cancer in the county of Va¨sterbotten during 1998, only 36 cases occurred within our study cohort (our unpublished data). The results from the present study should, thus, primarily be compared with prospective screening studies with a similar age range at inclusion. In summary, moderately increased leptin levels are associated with the development of prostate cancer. Leptin may be a possible key link between Western lifestyle and the transition from premalignant lesions to overt prostate cancer. Acknowledgments We are grateful to the Northern Sweden MONICA project, the Va¨sterbotten Intervention Program (Åsa Ågren, Lars Weinehall, Bernt Lindahl, Go¨sta Dahle´n, and Birgitta Stegmayr), and the grants supporting them; to Karin Hjertkvist, Sabine Rinaldi, Birgitta Ekblom, Elisabeth Dahlgren, and Charlotte Ingri for expert technical assistance; and to Prof. Jonathan R. Seckl for constructive criticism of the manuscript.

References 1. Yu H, Harris RE, Gao YT, Gao R, Wynder EL. 1991 Comparative epidemiology of cancers of the colon, rectum, prostate and breast in Shanghai, China versus the United States. Int J Epidemiol. 20:76 – 81. 2. Bosland MC, Oakley-Girvan I, Whittemore AS. 1999 Dietary fat, calories, and prostate cancer risk. J Natl Cancer Inst. 91:489 – 491. 3. Hebert JR, Hurley TG, Olendzki BC, Teas J, Ma Y, Hampl JS. 1998 Nutritional and socioeconomic factors in relation to prostate cancer mortality: a crossnational study. J Natl Cancer Inst. 90:1637–1647. 4. Mukherjee P, Sotnikov AV, Mangian HJ, Zhou JR, Visek WJ, Clinton SK. 1999 Energy intake and prostate tumor growth, angiogenesis, and vascular endothelial growth factor expression. J Natl Cancer Inst. 91:512–523. 5. Giovannucci E, Leitzmann M, Spiegelman D, et al. 1998 A prospective study of physical activity and prostate cancer in male health professionals. Cancer Res. 58:5117–5122. 6. Chan JM, Stampfer MJ, Giovannucci E, et al. 1998 Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science. 279:563–566. 7. Mantzoros CS, Tzonou A, Signorello LB, Stampfer M, Trichopoulos D,

8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

35. 36.

1345

Adami HO. 1997 Insulin-like growth factor 1 in relation to prostate cancer and benign prostatic hyperplasia. Br J Cancer. 76:1115–1118. Wolk A, Mantzoros CS, Andersson SO, et al. 1998 Insulin-like growth factor I and prostate cancer risk: a population-based, case-control study. J Natl Cancer Inst. 90:911–915. Friedman JM, Halaas JL. 1998 Leptin and the regulation of body weight in mammals. Nature. 395:763–770. Montague CT, Farooqi IS, Whitehead JP, et al. 1997 Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature. 387:903–908. Considine RV, Sinha MK, Heiman ML, et al. 1996 Serum immunoreactiveleptin concentrations in normal-weight and obese humans. N Engl J Med. 334:292–295. Larsson H, Elmstahl S, Berglund G, Ahre´n B. 1998 Evidence for leptin regulation of food intake in humans. J Clin Endocrinol Metab. 83:4382– 4385. So¨derberg S, Olsson T, Eliasson M, Johnson O, Ahre´n B. 1999 Plasma leptin levels are associated with abnormal fibrinolysis in men and postmenopausal women. J Intern Med. 245:533–543. Cioffi JA, Shafer AW, Zupancic TJ, et al. 1996 Novel B219/OB receptor isoforms: possible role of leptin in hematopoiesis and reproduction. Nat Med. 2:585–589. Nazian SJ, Cameron DF. 1999 Temporal relation between leptin and various indices of sexual maturation in the male rat. J Androl. 20:487– 491. Sierra-Honigmann MR, Nath AK, Murakami C, et al. 1998 Biological action of leptin as an angiogenic factor. Science. 281:1683–1686. WHO MONICA Project Principal Investigators (prepared by H. Tunstall Pedoe). 1988 The World Health Organization MONICA Project (Monitoring Trends and Determinants in Cardiovascular Disease): a major international collaboration. J Clin Epidemiol. 41:105–114. Weinehall L, Hallgren CG, Westman G, Janlert U, Wall S. 1998 Reduction of selection bias in primary prevention of cardiovascular disease through involvement of primary health care. Scand J Prim Health Care. 16:171–176. So¨derberg S, Ahre´n B, Stegmayr B, et al. 1999 Leptin is a risk marker for first-ever hemorrhagic stroke in a population-based cohort. Stroke. 30:328 –337. Bray GA. 1992 An approach to the classification and evaluation of obesity. In: Bjo¨rntorp P, Brodoff BN, eds. Obesity. Philadelphia: J.B. Lippincott Co; 294 –308. Sakr WA, Grignon DJ, Haas GP, Heilbrun LK, Pontes JE, Crissman JD. 1996 Age and racial distribution of prostatic intraepithelial neoplasia. Eur Urol. 30:138 –144. Lagiou P, Signorello LB, Trichopoulos D, Tzonou A, Trichopoulou A, Mantzoros CS. 1998 Leptin in relation to prostate cancer and benign prostatic hyperplasia. Int J Cancer. 76:25–28. Mantzoros CS, Bolhke K, Moschos S, Cramer DW. 1999 Leptin in relation to carcinoma in situ of the breast: a study of premenopausal cases and controls. Int J Cancer. 80:523–526. Bornstein SR, Licinio J, Tauchnitz R, et al. 1998 Plasma leptin levels are increased in survivors of acute sepsis: associated loss of diurnal rhythm, in cortisol and leptin secretion. J Clin Endocrinol Metab. 83:280 –283. Haynes WG, Morgan DA, Walsh SA, Mark AL, Sivitz WI. 1997 Receptormediated regional sympathetic nerve activation by leptin. J Clin Invest. 100:270 –278. Wang JM, McKenna KE, McVary KT, Lee C. 1991 Requirement of innervation for maintenance of structural and functional integrity in the rat prostate. Biol Reprod. 44:1171–1176. Giovannucci E, Rimm EB, Stampfer MJ, Colditz GA, Willett WC. 1997 Height, body weight, and risk of prostate cancer. Cancer Epidemiol Biomark Prev. 6:557–563. Rosenbaum M, Nicolson M, Hirsch J, et al. 1996 Effects of gender, body composition, and menopause on plasma concentrations of leptin. J Clin Endocrinol Metab. 81:3424 –3427. English HF, Santen RJ, Isaacs JT. 1987 Response of glandular versus basal rat ventral prostatic epithelial cells to androgen withdrawal and replacement. Prostate. 11:229 –242. Gann PH, Hennekens CH, Ma J, Longcope C, Stampfer MJ. 1996 Prospective study of sex hormone levels and risk of prostate cancer. J Natl Cancer Inst. 88:1118 –1126. Colao A, Marzullo P, Ferone D, et al. 1998 Prostatic hyperplasia: an unknown feature of acromegaly. J Clin Endocrinol Metab. 83:775–779. Pollard M, Luckert PH, Snyder D. 1989 Prevention of prostate cancer and liver tumors in L-W rats by moderate dietary restriction. Cancer. 64:686 – 690. Bjo¨rntorp P. 1993 Androgens, the metabolic syndrome, and non-insulindependent diabetes mellitus. Ann NY Acad Sci. 676:242–252. Attia N, Caprio S, Jones TW, et al. 1999 Changes in free insulin-like growth factor-I and leptin concentrations during acute metabolic decompensation in insulin withdrawn patients with type 1 diabetes. J Clin Endocrinol Metab. 84:2324 –2328. Schatzl G, Reiter WJ, Thurridl T, et al. 2000 Endocrine patterns in patients with benign and malignant prostatic diseases. Prostate. 44:219 –224. Maattanen L, Auvinen A, Stenman UH, et al. 1999 European randomized study of prostate cancer screening: first-year results of the Finnish trial. Br J Cancer. 79:1210 –1214.