Abstract
Molecular mechanisms linking obesity to prostate cancer involve steroid hormone and insulin/insulin-like growth factor-1 (IGF-1) pathways. We investigated the association of circulating serum markers (e.g., androgens & IGFs/IGFBPs) with BMI and in modifying the association of obesity with prostate cancer risk. Data and specimens for this nested case-control study are from the Prostate Cancer Prevention Trial, a randomized, placebo-controlled trial of finasteride for prostate cancer prevention. Presence or absence of cancer was determined by prostate biopsy. Serum samples were assayed for sex steroid hormone concentrations and IGF-1 axis analytes. Logistic regression estimated odds ratio and 95% confidence intervals (CIs) for risk of overall, low-grade (Gleason 2–6), and high-grade (Gleason 7–10) cancers. We found significant associations between BMI with serum steroids and IGFs/IGFBPs; the IGF-1 axis significantly associated with several serum steroids. Serum steroid levels did not affect the association of BMI with prostate cancer risk; however, IGFBP2 and IGFs modified the association of obesity with low- and high-grade disease. While serum steroids and IGFs/IGFBPs are associated with BMI, only the IGF-1 axis contributed to obesity-related prostate cancer risk. Understanding the biological mechanisms linking obesity to prostate cancer risk as it relates to circulating serum markers will aid in developing effective prostate cancer prevention strategies and treatments.
Keywords: obesity, BMI, sex steroids, insulin/insulin-like growth factor-1, prostate cancer risk
Introduction
Prostate cancer is the most common non-cutaneous malignancy and the second leading cause of cancer deaths among men in the United States (U.S.), with approximately 248,530 new cases and about 34,130 men expected to die of the disease in 2021 (Siegel et al. 2021). The prevalence of obesity (body mass index [BMI] ≥30 kg/m2) in the U.S. adult male population is 36.6% (Hales et al. 2018). Obesity and related metabolic alterations have been implicated in the risk of cancer (Nimptsch and Pischon 2016). While the link of obesity to overall prostate cancer risk is controversial, studies have consistently shown a strong association of obesity to increased risk of aggressive disease (Barrington et al. 2015, Gong et al. 2006, Guerrios-Rivera et al. 2017, Perez-Cornago et al. 2017, Vidal et al. 2014).
Molecular mechanisms that explain the link between obesity and prostate cancer have not been elucidated. Biological mechanisms hypothesized to underlie the relationship between the two may involve crosstalk among hormonal pathways that include sex-steroid hormones and the insulin/insulin-like growth factor-1 (IGF-1) axis, specifically adiposity-related changes in metabolism and endogenous hormone levels (De Pergola and Silvestris 2013, Hsing et al. 2007, Kaaks and Stattin 2010, Boibessot and Toren 2018). Notably, the androgen and IGF-1 pathways are known to play critical roles in prostate cancer development (Nimptsch and Pischon 2016, Holly et al. 2020).
Using data from the Prostate Cancer Prevention Trial (PCPT), we previously found obesity is associated with an increased risk of high-grade prostate cancer and a decreased risk of low-grade disease (Gong et al. 2006). In the current follow-up study, we investigated the association of serum sex steroid hormone and IGFs and their binding proteins (IGFPBs) with BMI and hypothesized that these circulating serum markers modify the association of obesity with prostate cancer risk. We first examined the relationship between obesity and serum markers, and then determined their effects on modulating obesity-related prostate cancer risk. Our findings suggest that serum steroids and IGFs/IGFBPs are associated with BMI and found that only the IGF-1 axis (IGFBP2 and IGFs) modified the association of BMI with low- and high-grade disease.
Methods
Participant and Study Description
All data for this study originated from the PCPT (SWOG-S9217); details of the trial and participant characteristics have been described previously (Feigl et al. 1995, Thompson et al. 2003). Briefly, 18,880 eligible men age 55 years and older with a normal digital rectal exam (DRE), prostate specific antigen (PSA) level of 3 ng/mL or below, and no history of prostate cancer or other clinically significant co-morbid conditions that would have precluded successful completion of the study protocol, were randomized to receive either finasteride (5 mg/day) or placebo daily for seven years. During the trial, men underwent annual DRE and PSA measures and a prostate biopsy was recommended for all men with an abnormal DRE or a finasteride-adjusted PSA of > 4.0 ng/mL. At the conclusion of the trial, those men not previously diagnosed with prostate cancer were offered an end-of-study biopsy. Details regarding age, race/ethnicity, family history, smoking, alcohol and physical activity habits (type, frequency, duration, pace, and intensity), and dietary consumptions were collected at baseline using self-administered questionnaires (Neuhouser et al. 1999, Neuhouser et al. 2001). Clinic staff measured height and weight at randomization, and body mass index (BMI) was calculated as weight (kg) divided by height2 (m).
The study was performed in accordance with the Declaration of Helsinki and conducted in accordance with the FDA Guidelines on Good Clinical Practice. All men signed informed consent and study procedures were approved by the Institutional Review Boards of the participating 221 study sites (Feigl et al. 1995, Thompson et al. 2003). The study reported here is part of a large nested case-control study designed to examine multiple hypotheses about prostate cancer risk. Cases were men with biopsy-determined prostate cancer identified either by a for-cause or end-of-study biopsy and controls were selected from men who completed the end-of-study biopsy procedure and had no evidence of prostate cancer. Controls were frequency matched to cases based on age, treatment arm, and family history of prostate cancer, and were oversampled to include all eligible non-whites.
Sample Collection and Measurement of Serum Androgens, SHBG Concentrations, and IGFs
Non-fasting blood specimens were collected at baseline and annually thereafter until diagnosis or the end of study. Venous blood was drawn into tubes without anticoagulant, refrigerated, and shipped to a central repository where it was centrifuged, aliquoted, and stored at −70°C. Concentrations of testosterone (T), sex hormone binding globulin (SHBG), and 5α-androstane-3α,17β-diol glucuronide (or 3α-androstanediol glucuronide, 3α-dG) were measured from serum for this study in 2008, and a subset of men (313 cases and 346 controls) had serum androstenedione measured in 2009 as part of a different study. For the placebo arm, serum samples were collected at baseline and year 3 and pooled before analysis to better characterize androgen levels and reduce intraindividual variability; alternate years were selected if men were missing a year 3 sample or were diagnosed before year 3, and a single, baseline sample was used if a post-baseline, pre-diagnostic sample was unavailable. For the finasteride arm, all the androgen and SHBG measures were from baseline. Total testosterone, 3α-dG, androstenedione and SHBG were quantified in serum by highly specific immunoassays as described previously (Hoque et al. 2015, Kristal et al. 2012, Price et al. 2016). Serum concentrations of estrone (E1) and estradiol (E2) were determined by radioimmunoassay (Yao et al. 2011). Concentrations of IGF1, IGF2, IGFBP2 and IGFBP3 were assayed in the baseline serum samples with a standard ELISA as described previously (Neuhouser et al. 2013). Serum steroid hormone and IGF concentrations were available from 1787 cases and 1782 controls; 313 cases and 346 controls with androstenedione.
Statistical Methods
We compared baseline demographic and lifestyle characteristics of prostate cancer cases and controls by Student’s t test for continuous variables and chi-square test for categorical variables. For ordered categorical variables, a trend p-value was calculated based on an ordinal variable corresponding to rank (lowest to highest). Logistic regression was used to calculate odds ratio (ORs) and 95% confidence intervals (CIs) for risk of total prostate cancer, and polytomous logistic regression was used to calculate ORs and 95% CIs of both low-grade (Gleason 2–6) and high-grade (Gleason 7–10) prostate cancers. The polytomous regression with a generalized logit link permits a model including both low-grade and high-grade cancers as outcomes in the same model, contrasted with no cancer, with no assumption of ordinality in the outcome. Logistic models were stratified by treatment arm (finasteride vs placebo), and also combined arms; combined models were adjusted for treatment arm. All models were also adjusted for age (continuous), race (white vs nonwhite), and family history of prostate cancer (yes vs no), and adjustments for serum sex steroids and IGFs were added. Trend p-values were calculated for main effects of BMI, as well as interaction between BMI and treatment arm, and interactions between BMI and serum sex steroids and IGFs, using an ordinal variable corresponding to rank from lowest to highest category of BMI. Mean concentrations of androgens and IGFs were estimated for each BMI category (normal, <25 kg/m2, overweight, 25–30 kg/m2, and obese, ≥30 kg/m2) and p-values were calculated using linear regression, usuing an ordinal variable corresponding to rank from the lowest category to the highest of BMI. Among controls, mean concentrations of serum steroids and BMI were estimated for each IGF and IGFBP tertile, and trend p-values were calculated using linear regression, using an ordinal variable corresponding to rank of IGF tertile, from the lowest category to the highest category. IGF tertiles were calculated based on the observed distribution in the data. All statistical tests were two-sided with statistical significance set at p<0.05 and carried out using SAS statistical software (version 9.4, SAS Corporation, Cary, NC).
Results
Association of serum sex steroids and IGFs/IGFBPs with BMI
Table 1 depicts the demographic and health-related variables of the current PCPT case-control population. Cases were less likely to be diabetic at baseline than controls, but cases and controls were similar with respect to BMI, smoking status, and physical activity, as well as other dietary and lifestyle factors (e.g., consumption of proteins, fat, carbohydrates, fruit, vegetables, calories and alcohol; data not shown).
Table 1 –
Baseline demographics of trial participants
| (n=1787) | (n=1782) | (n=3569) | ||
|---|---|---|---|---|
| Age, mean (SD)2 | 63.7 (5.5) | 63.6 (5.5) | 63.6 (5.5) | NA2 |
| Race2 | NA2 | |||
| Non-hispanic white | 1658 (92.8%) | 1415 (79.4%) | 3073 (86.1%) | |
| Minority | 129 (7.2%) | 367 (20.6%) | 496 (13.9%) | |
| BMI, N (%) | 0.07 | |||
| Normal | 498 (27.9%) | 444 (24.9%) | 942 (26.4%) | |
| Overweight | 915 (51.2%) | 944 (53.0%) | 1859 (52.1%) | |
| Obese | 374 (20.9%) | 394 (22.1%) | 768 (21.5%) | |
| Smoking Status, N (%) | 0.55 | |||
| Never | 636 (35.6%) | 610 (34.2%) | 1246 (34.9%) | |
| Former | 124 (6.9%) | 137 (7.7%) | 261 (7.3%) | |
| Current | 1027 (57.5%) | 1035 (58.1%) | 2062 (57.8%) | |
| Physical Activity, N (%) | 0.56 | |||
| Sedentary | 306 (17.2%) | 311 (17.5%) | 617 (17.4%) | |
| Light | 740 (41.6%) | 734 (41.4%) | 1474 (41.5%) | |
| Moderate | 586 (32.9%) | 543 (30.6%) | 1129 (31.8%) | |
| Active | 148 (8.3%) | 186 (10.5%) | 334 (9.4%) | |
| Treatment Arm2 | NA2 | |||
| Finasteride | 755 (42.2%) | 757 (42.5%) | 1512 (42.4%) | |
| Placebo | 1032 (57.8%) | 1025 (57.5%) | 2057 (57.6%) | |
| Diabetes, N (%) | 84 (4.7%) | 131 (7.4%) | 215 (6.0%) | <0.001 |
| Family History of prostate cancer, N (%)2 | 380 (21.3%) | 377 (21.2%) | 757 (21.2%) | NA2 |
P-values comparing cases to controls were calculated using t-tests for continuous variables and chi-square tests for categorical variables. For ordered categorical variables, linear regression was used to calculate a trend p-value, using an ordinal variable corresponding to rank from the lowest category to the highest.
P-values not included due to controls being frequency matched to cases based on age, treatment arm, and family history, and were oversampled to include all eligible non-whites.
We investigated the association of serum steroids and IGFs/IGFBPs with BMI. Table 2 shows the association between BMI and serum sex steroids (testosterone, free testosterone, 3α-dG, estrone, estradiol, SHBG, androstenedione), separately by treatment arm and combined treatment arms. Serum testosterone, free testosterone, and SHBG have inverse relationships with BMI (all p-values <0.0001). Serum 3α-dG, estrone, and estradiol have a positive, linear association with BMI (all p-values <0.0001). Androstenedione did not appear to have any association with BMI. Table 2 also shows significant associations between BMI and serum IGF2 and IGFBP2. In particular, serum concentrations of IGF2 increased with increasing BMI (p-value <0.01) while IGFBP2 significantly decreased with increasing BMI (p-value <0.0001). IGF1 and IGFBP3 did not have any association with BMI. When stratified by treatment arm, the associations are similar. Our findings demonstrate significant associations between BMI and serum steroids or IGFs/IGFBPs.
Table 2 –
Association between mean concentrations* of serum steroids and IGFs with BMI in cases and controls
| Normal | Overweight | Obese | p-value1 | |
|---|---|---|---|---|
| (<25 kg/m2) | (25–30 kg/m2) | (≥30 kg/m2) | ||
| Combined Treatment Arms | ||||
| Count | 942 | 1859 | 768 | |
| Testosterone, ng/dL | 428.6 (146.0) | 380.0 (130.4) | 330.1 (112.5) | <0.0001 |
| Free Testosterone, pg/mL | 9.1 (2.7) | 8.6 (2.6) | 7.9 (2.4) | <0.0001 |
| 3a-DG, ng/mL | 6.0 (4.2) | 6.8 (4.8) | 7.0 (4.5) | <0.0001 |
| Estrone, pg/mL | 44.3 (14.4) | 45.1 (14.6) | 49.7 (16.5) | <0.0001 |
| Estradiol, pg/mL | 33.4 (10.3) | 33.9 (10.8) | 35.5 (10.6) | <0.0001 |
| SHBG, nmol/L | 45.1 (18.1) | 38.3 (14.7) | 33.1 (12.9) | <0.0001 |
| Androstenedione2, ng/mL | 0.6 (0.2) | 0.6 (0.2) | 0.6 (0.2) | 0.09 |
| IGF1, ng/mL | 211.6 (60.6) | 213.4 (64.0) | 207.0 (72.3) | 0.19 |
| IGF2, ng/mL | 1707.9 (400.1) | 1758.5 (423.1) | 1770.2 (502.9) | 0.002 |
| IGFBP2, ng/mL | 723.7 (358.4) | 504.6 (270.7) | 370.0 (201.7) | <0.0001 |
| IGFBP3, ng/mL | 4004.8 (907.8) | 4083.8 (946.1) | 4067.1 (1115.2) | 0.16 |
| Finasteride | ||||
| Count | 399 | 766 | 347 | |
| Testosterone, ng/dL | 430.1 (146.0) | 379.1 (135.6) | 326.9 (113.9) | <0.0001 |
| Free Testosterone, pg/mL | 9.2 (2.8) | 8.7 (2.7) | 8.0 (2.6) | <0.0001 |
| 3a-DG, ng/mL | 6.3 (4.7) | 6.8 (5.1) | 6.9 (4.4) | 0.08 |
| Estrone, pg/mL | 45.3 (15.1) | 45.8 (15.6) | 50.1 (16.2) | <0.0001 |
| Estradiol, pg/mL | 33.9 (10.4) | 34.4 (12.4) | 35.5 (10.8) | 0.07 |
| SHBG, nmol/L | 45.2 (19.2) | 37.9 (15.3) | 31.9 (11.7) | <0.0001 |
| Androstenedione2, ng/mL | 0.7 (0.3) | 0.6 (0.2) | 0.6 (0.2) | 0.15 |
| IGF1, ng/mL | 210.9 (62.9) | 214.0 (63.5) | 211.7 (73.3) | 0.84 |
| IGF2, ng/mL | 1705.0 (412.3) | 1745.3 (429.4) | 1785.2 (475.1) | 0.01 |
| IGFBP2, ng/mL | 718.9 (331.0) | 510.7 (273.1) | 367.1 (211.5) | <0.0001 |
| IGFBP3, ng/mL | 4001.2 (948.4) | 4059.4 (948.6) | 4107.1 (1071.4) | 0.14 |
| Placebo | ||||
| Count | 543 | 1093 | 421 | |
| Testosterone, ng/dL | 427.6 (146.1) | 380.7 (126.7) | 332.8 (111.4) | <0.0001 |
| Free Testosterone, pg/mL | 9.0 (2.6) | 8.6 (2.5) | 7.9 (2.2) | <0.0001 |
| 3a-DG, ng/mL | 5.9 (3.8) | 6.8 (4.6) | 7.2 (4.6) | <0.0001 |
| Estrone, pg/mL | 43.5 (13.8) | 44.6 (13.8) | 49.3 (16.8) | <0.0001 |
| Estradiol, pg/mL | 33.0 (10.3) | 33.6 (9.6) | 35.6 (10.5) | <0.01 |
| SHBG, nmol/L | 44.9 (17.4) | 38.6 (14.3) | 34.1 (13.8) | <0.0001 |
| Androstenedione2, ng/mL | 0.6 (0.2) | 0.6 (0.2) | 0.6 (0.2) | 0.29 |
| IGF1, ng/mL | 212.1 (58.9) | 212.9 (64.4) | 203.1 (71.3) | 0.05 |
| IGF2, ng/mL | 1710.0 (391.2) | 1767.8 (418.6) | 1757.8 (525.0) | 0.06 |
| IGFBP2, ng/mL | 727.3 (377.6) | 500.4 (269.0) | 372.5 (193.5) | <0.0001 |
| IGFBP3, ng/mL | 4007.5 (877.7) | 4100.8 (944.5) | 4033.8 (1150.6) | 0.55 |
Serum steroids and IGFs concentrations reported as mean (SD).
Linear regression was used to calculate a trend p-value, using an ordinal variable corresponding to rank from the lowest category to the highest of BMI.
Counts for androstenedione concentrations are lower due to the subset of men with available serum data
[Normal – Overweight – Obese: Combined treatment arm (184 – 338 – 137); Finasteride (69 – 142 – 69); Placebo (115 – 196 – 68)]
No adjustments for multiple testing have been done.
Association between serum sex steroids and IGFs/IGFBPs
We next determined the association between IGFs/IGFBPs and serum steroids. Table 3 gives mean values of serum steroids by tertiles of serum IGF1, IGF2, IGFBP2 and IGFBP3 in the controls. Serum concentrations of T, free T, and SHBG increase significantly, and levels of 3α-dG and estrone decrease, with increasing concentrations of IGFBP2. There was evidence of significant associations of IGF1, IGF2, and IGFBP3 with several serum steroids. SHBG has associations with all IGFs while androstenedione did not appear to have any association with any of the IGFs. Testosterone and 3α-dG were significantly associated with IGF2, IGFBP2, and IGFBP3. Free testosterone was significantly associated with increasing concentrations of IGF1 and IGFBP2. In the finasteride treatment arm, we observed a similar association of serum steroids (testosterone and SHBG) with serum IGFs (Supplemental Table 1). Thus, we found significant associations between IGFs/IGFBPs with several serum steroids.
Table 3 –
Association of serum steroids with IGFs among controls
| Count | Testosterone (ng/dL) | Free Testosterone (pg/mL) | 3α-DG (ng/mL) | Estrone (pg/mL) | Estradiol (pg/mL) | SHBG (nmol/L) | Androstenedione (ng/mL) | |
|---|---|---|---|---|---|---|---|---|
| IGF1, ng/mL | ||||||||
| <181 | 590 | 388.2 (143.6) | 8.5 (2.7) | 6.9 (5.0) | 45.7 (17.0) | 34.5 (11.8) | 42.0 (16.9) | 0.6 (0.2) |
| 181 to <231 | 591 | 383.4 (144.6) | 8.6 (2.7) | 6.5 (3.7) | 45.6 (14.3) | 34.2 (10.4) | 39.1 (17.0) | 0.6 (0.2) |
| ≥231 | 593 | 377.5 (123.0) | 8.9 (2.6) | 6.5 (3.8) | 45.2 (14.2) | 33.7 (12.0) | 35.7 (12.7) | 0.6 (0.2) |
| p-value1 | 0.18 | 0.006 | 0.15 | 0.57 | 0.19 | <0.0001 | 0.23 | |
| IGF2, ng/mL | ||||||||
| <1543 | 590 | 407.1 (155.1) | 8.7 (2.8) | 6.4 (4.2) | 44.8 (15.6) | 35.1 (11.9) | 43.9 (18.7) | 0.6 (0.2) |
| 1543 to <1903 | 592 | 387.3 (129.3) | 8.7 (2.6) | 6.4 (3.7) | 45.5 (14.7) | 34.3 (12.2) | 39.4 (14.3) | 0.6 (0.2) |
| ≥1903 | 592 | 354.9 (120.7) | 8.5 (2.6) | 7.1 (4.6) | 46.1 (15.4) | 33.0 (10.0) | 33.5 (12.2) | 0.6 (0.2) |
| p-value1 | <0.0001 | 0.36 | 0.001 | 0.14 | 0.001 | <0.0001 | 0.09 | |
| IGFBP2, ng/mL | ||||||||
| <343 | 590 | 332.4 (121.2) | 8.2 (2.6) | 6.9 (4.2) | 46.9 (16.2) | 33.1 (9.9) | 31.3 (12.2) | 0.6 (0.2) |
| 343 to <563 | 591 | 380.1 (123.9) | 8.7 (2.6) | 6.9 (4.2) | 45.8 (15.4) | 35.0 (13.2) | 38.3 (14.7) | 0.6 (0.2) |
| ≥563 | 593 | 436.3 (145.6) | 9.1 (2.7) | 6.2 (4.1) | 43.7 (13.8) | 34.3 (10.9) | 47.1 (16.3) | 0.6 (0.2) |
| p-value1 | <0.0001 | <0.0001 | 0.004 | 0.0003 | 0.07 | <0.0001 | 0.20 | |
| IGFBP3, ng/mL | ||||||||
| <3601 | 590 | 404.7 (156.5) | 8.7 (2.8) | 6.5 (4.5) | 45.0 (15.8) | 34.8 (11.8) | 43.5 (18.7) | 0.6 (0.2) |
| 3601 to <4424 | 591 | 381.1 (124.0) | 8.6 (2.5) | 6.2 (3.5) | 45.6 (15.0) | 34.4 (12.2) | 39.2 (13.7) | 0.6 (0.2) |
| ≥4424 | 593 | 363.5 (126.5) | 8.7 (2.7) | 7.2 (4.5) | 45.8 (14.9) | 33.2 (10.1) | 34.1 (13.3) | 0.6 (0.2) |
| p-value1 | <0.0001 | 0.87 | 0.005 | 0.36 | 0.02 | <0.0001 | 0.14 |
Linear regression was used to calculate a trend p-value, using an ordinal variable corresponding to rank from the lowest category to the highest of IGF. This table includes controls only.
No adjustments for multiple testing have been made.
Association between obesity and prostate cancer risk modified by sex steroid hormones and IGFs/IGFPBs
Table 4 examines the association between obesity and total, low-grade and high-grade prostate cancer risk stratified by PCPT treatment arm (placebo or finasteride) and factoring in several covariates (e.g., age, race, family history, treatment arm, serum steroids, and IGFs/IGFBP2) to determine their effect on modifying risk. BMI did not have any significant association with total prostate cancer risk. However, when cancer was stratified by grade, risk of low-grade (Gleason 2–6) cancer decreased with increasing BMI (OR (95% CI) for obese participants 0.66 (0.50–0.88), p-trend=0.003) in the placebo arm. Risk of high-grade (Gleason 7–10) cancer was associated with increasing BMI when treatment arms were combined (OR (95% CI) for obese participants 1.34 (1.00–1.79), p-trend=0.05). Results when high-grade cancer was restricted to Gleason sum 8 to 10 demonstrated further increased risk with increasing BMI (OR (95% CI) for obese participants 1.69 (1.00–2.85), p-trend=0.05, data not shown).
Table 4 –
Association of obesity with prostate cancer risk (BMI vs total, low, high-grade) adjusted for covariates
| All Prostate Cancer | Low-Grade (Gleason <7) | High-Grade (Gleason ≥7) | |||||
|---|---|---|---|---|---|---|---|
| N ctl | N case | OR (95% CI) | N case | OR (95% CI) | N case | OR (95% CI) | |
| Adjusted for age, race, family history of prostate cancer, and treatment arm 1 | |||||||
| Normal (<25 kg/m2) | 444 | 498 | ref | 363 | ref | 115 | ref |
| Overweight (25–30 kg/m2) | 944 | 915 | 0.86 (0.74–1.01) | 628 | 0.81 (0.68–0.97) | 247 | 1.03 (0.80–1.32) |
| Obese (≥30 kg/m2) | 394 | 374 | 0.89 (0.73–1.08) | 231 | 0.75 (0.60–0.93) | 126 | 1.34 (1.00–1.79) |
| p-trend | 0.21 | 0.01 | 0.05 | ||||
| p-interaction5 | 0.10 | 0.21 | 0.85 | ||||
| Adjusted for age, race, family history, treatment arm, and serum steroids 2 | |||||||
| Normal (<25 kg/m2) | 442 | 492 | ref | 359 | ref | 113 | ref |
| Overweight (25–30 kg/m2) | 940 | 909 | 0.86 (0.73–1.02) | 623 | 0.82 (0.68–0.98) | 246 | 1.01 (0.78–1.30) |
| Obese (≥30 kg/m2) | 393 | 371 | 0.88 (0.72–1.08) | 230 | 0.76 (0.61–0.96) | 125 | 1.27 (0.93–1.73) |
| p-trend | 0.20 | 0.02 | 0.14 | ||||
| p-interaction5 | 0.08 | 0.18 | 0.92 | ||||
| Adjusted for age, race, family history, treatment arm, and IGFBP2 3 | |||||||
| Normal (<25 kg/m2) | 442 | 497 | ref | 362 | ref | 115 | ref |
| Overweight (25–30 kg/m2) | 941 | 910 | 0.92 (0.78–1.09) | 625 | 0.88 (0.73–1.05) | 245 | 1.09 (0.83–1.42) |
| Obese (≥30 kg/m2) | 391 | 374 | 1.00 (0.81–1.24) | 231 | 0.85 (0.67–1.08) | 126 | 1.48 (1.08–2.04) |
| p-trend | 0.99 | 0.16 | 0.01 | ||||
| p-interaction5 | 0.12 | 0.24 | 0.81 | ||||
| Adjusted for age, race, family history, treatment arm, and IGFs 4 | |||||||
| Normal (<25 kg/m2) | 442 | 497 | ref | 362 | ref | 115 | ref |
| Overweight (25–30 kg/m2) | 941 | 910 | 0.93 (0.79–1.10) | 625 | 0.88 (0.73–1.06) | 245 | 1.08 (0.83–1.41) |
| Obese (≥30 kg/m2) | 391 | 374 | 1.01 (0.82–1.26) | 231 | 0.87 (0.68–1.10) | 126 | 1.47 (1.07–2.02) |
| p-trend | 0.93 | 0.21 | 0.02 | ||||
| p-interaction5 | 0.12 | 0.25 | 0.82 | ||||
| Adjusted for age, race, and family history of prostate cancer 1 | |||||||
| Normal (<25 kg/m2) | 202 | 197 | ref | 125 | ref | 66 | ref |
| Overweight (25–30 kg/m2) | 383 | 383 | 0.97 (0.75–1.24) | 227 | 0.89 (0.67–1.19) | 141 | 1.08 (0.76–1.52) |
| Obese (≥30 kg/m2) | 172 | 175 | 1.08 (0.80–1.46) | 94 | 0.88 (0.62–1.25) | 69 | 1.33 (0.89–2.00) |
| p-trend | 0.65 | 0.46 | 0.17 | ||||
| Adjusted for age, race, family history, and serum steroids 2 | |||||||
| Normal (<25 kg/m2) | 201 | 191 | ref | 121 | ref | 64 | ref |
| Overweight (25–30 kg/m2) | 379 | 378 | 0.98 (0.76–1.27) | 222 | 0.91 (0.68–1.23) | 141 | 1.09 (0.76–1.55) |
| Obese (≥30 kg/m2) | 171 | 172 | 1.08 (0.79–1.49) | 93 | 0.92 (0.63–1.33) | 68 | 1.29 (0.84–1.98) |
| p-trend | 0.64 | 0.62 | 0.25 | ||||
| Adjusted for age, race, family history, and IGFBP2 | |||||||
| Normal (<25 kg/m2) | 201 | 197 | ref | 125 | ref | 66 | ref |
| Overweight (25–30 kg/m2) | 382 | 379 | 1.04 (0.80–1.36) | 225 | 0.97 (0.72–1.31) | 139 | 1.17 (0.81–1.68) |
| Obese (≥30 kg/m2) | 172 | 175 | 1.24 (0.89–1.73) | 94 | 1.02 (0.70–1.50) | 69 | 1.57 (1.01–2.44) |
| p-trend | 0.20 | 0.92 | 0.05 | ||||
| Adjusted for age, race, family history, and IGFs | |||||||
| Normal (<25 kg/m2) | 201 | 197 | ref | 125 | ref | 66 | ref |
| Overweight (25–30 kg/m2) | 382 | 379 | 1.05 (0.80–1.37) | 225 | 0.98 (0.72–1.33) | 139 | 1.16 (0.81–1.67) |
| Obese (≥30 kg/m2) | 172 | 175 | 1.24 (0.89–1.73) | 94 | 1.04 (0.71–1.52) | 69 | 1.54 (0.99–2.40) |
| p-trend | 0.20 | 0.86 | 0.06 | ||||
| Adjusted for age, race, and family history of prostate cancer 1 | |||||||
| Normal (<25 kg/m2) | 242 | 301 | ref | 238 | ref | 49 | ref |
| Overweight (25–30 kg/m2) | 561 | 532 | 0.78 (0.64–0.97) | 401 | 0.75 (0.60–0.94) | 106 | 0.95 (0.66–1.38) |
| Obese (≥30 kg/m2) | 222 | 199 | 0.76 (0.59–0.99) | 137 | 0.66 (0.50–0.88) | 57 | 1.35 (0.88–2.06) |
| p-trend | 0.03 | 0.003 | 0.17 | ||||
| Adjusted for age, race, family history, and serum steroids 2 | |||||||
| Normal (<25 kg/m2) | 241 | 301 | ref | 238 | ref | 49 | ref |
| Overweight (25–30 kg/m2) | 561 | 531 | 0.77 (0.62–0.95) | 401 | 0.75 (0.60–0.94) | 105 | 0.89 (0.61–1.31) |
| Obese (≥30 kg/m2) | 222 | 199 | 0.75 (0.57–0.98) | 137 | 0.67 (0.50–0.90) | 57 | 1.24 (0.79–1.94) |
| p-trend | 0.03 | 0.01 | 0.33 | ||||
| Adjusted for age, race, family history, and IGFBP2 | |||||||
| Normal (<25 kg/m2) | 241 | 300 | ref | 237 | ref | 49 | ref |
| Overweight (25–30 kg/m2) | 559 | 531 | 0.83 (0.67–1.04) | 400 | 0.81 (0.64–1.02) | 106 | 0.97 (0.66–1.44) |
| Obese (≥30 kg/m2) | 219 | 199 | 0.85 (0.64–1.12) | 137 | 0.75 (0.55–1.01) | 57 | 1.40 (0.88–2.23) |
| p-trend | 0.22 | 0.06 | 0.14 | ||||
| Adjusted for age, race, family history, and IGFs | |||||||
| Normal (<25 kg/m2) | 241 | 300 | ref | 237 | ref | 49 | ref |
| Overweight (25–30 kg/m2) | 559 | 531 | 0.84 (0.67–1.05) | 400 | 0.82 (0.64–1.03) | 106 | 0.98 (0.66–1.45) |
| Obese (≥30 kg/m2) | 219 | 199 | 0.86 (0.65–1.15) | 137 | 0.76 (0.56–1.04) | 57 | 1.41 (0.88–2.25) |
| p-trend | 0.28 | 0.08 | 0.14 | ||||
Adjusted for age (continuous) and race (white vs nonwhite). Combined models also adjusted for treatment arm.
Adjusted for the above factors, as well as serum testosterone, SHBG, and 3α-DG. P-value testing interaction between BMI and serum steroids is 0.29.
P-value testing interaction between BMI and IGFBP2 is 0.71.
IGFs include IGF1, IGF2, IGFBP2, and IGFBP3. P-value testing interactions between BMI and IGFs is 0.97.
P-value tests the interaction between treatment arm and BMI.
No adjustments for multiple testing have been made.
Odds ratios for low- and high-grade cancer were calculated using polytomous logistic regression, with no assumption of ordinality in the outcome.
When serum steroids were included as an adjustment factor in the risk model, the associations remained the same, suggesting that any association BMI had with prostate cancer was likely not mediated by serum steroid levels. However, when adjusting for confounders IGFBP2 and IGFs, the risk of high-grade (Gleason 7–10) cancer increasing with increasing BMI in the combined treatment arms became significant (IGFBP2: OR (95% CI) for obese participants 1.48 (1.08–2.04), p-trend=0.01; IGFs: OR (95% CI) 1.47 (1.07–2.02), p-trend=0.02). Alternatively, the association of risk of low-grade cancer decreasing with increasing BMI was no longer significant, suggesting IGFBP2 and IGFs may modify the association of BMI with low- and high-grade prostate cancer. However, when evaluated by separate treatment arms (placebo vs finasteride), IGFBP2 and IGFs had no effects on risk, perhaps due to a smaller sample size vs. the increase in number of cases by combining the treatment arms. Therefore, while serum steroid levels did not affect the association of BMI with prostate cancer risk, IGFBP2 and IGFs modified the association of BMI with low- and high-grade disease.
Discussion
Obesity is a pandemic of increasing proportion and a risk factor for prostate cancer. The link between obesity and prostate cancer is complex, with inconsistent results observed across studies that depend on the anthropometric indicators used to characterize obesity. Many studies including several meta-analyses evaluating the association of BMI with prostate cancer have reported null (Harding et al. 2015) or weak results (Renehan et al. 2008, MacInnis and English 2006) as well as positive associations between BMI and aggressive disease (Discacciati et al. 2012, Gong et al. 2006, Fang et al. 2018). Other studies have examined whether obesity plays a role in the racial disparity of prostate cancer incidence. While blacks had a significant increased risk of prostate cancer incidence, BMI, however, was not found to be a significant mediator (Akinyemiju et al. 2018). In a separate study, obesity was strongly associated with an increased risk of prostate cancer among black compared to white men in the SELECT trial (Barrington et al. 2015), which might reflect a difference in the biological effects of obesity on inflammation or insulin secretion between racial groups. However, a recent re-analysis of the trial data after taking into account differential biopsy assessment, which may affect prostate cancer diagnosis and subsequent cancer risk factors, changed the previously reported positive association between BMI and prostate cancer in black men to null (Tangen et al. 2019). Furthermore, the association of BMI with prostate cancer was generally stronger in studies that reported mortality rather than incidence (Fang et al. 2018, Zhong et al. 2016, Jochems et al. 2020). Future studies are warranted to investigate the role of high BMI in increased mortality among patients with prostate cancer including our PCPT population.
Biological mechanisms that explain the apparent link between obesity and prostate cancer have not been fully elucidated. Several mechanisms have been proposed to explain the association between obesity and prostate cancer, which include inter-related pathways that involve altered sex steroid hormones, insulin/IGF-1 axis, and adipokine signaling caused by inflammation. We undertook the current study to understand the association between circulating serum markers (e.g., sex serum steroid, serum IGFs/IGFBPs) and obesity with prostate cancer risk. The impact of these serum markers on obesity-related prostate cancer risk may play a critical role in prostate carcinogenesis and disease progression.
We identified significant associations between BMI with serum steroids and IGFs/IGFBPs. Our results showed obese men having lower concentrations of serum testosterone, free testosterone and SHBG and higher levels of serum 3α-dG, estrone and estradiol. Serum SHBG, a carrier protein that binds circulating sex steroid homones and reduces their availability to tissues, correlates inversely with BMI due mostly to increases in serum insulin, which inhibits hepatic SHBG synthesis (Kaaks et al. 2000, Kaaks and Stattin 2010). Obesity is associated with a lower concentration of free testosterone, resulting in the growth of aggressive prostate tumors (Allott et al. 2013) and men with prostate cancer who have low testosterone tend towards a more aggressive phenotype. Our data are consistent with a recent international collaboration that combined data from over 12,300 men from 25 studies that found BMI was strongly associated with concentrations of all the sex hormones and SHBG (Watts et al. 2017). We also found significant associations between BMI and serum IGF2 and IGFBP2. Specifically, obese men tended to have higher serum concentrations of IGF2 and lower concentrations of IGFBP2. Moreover, we found significant associations between IGFs/IGFBPs with several serum steroids (Table 3).
We next investigated whether sex steroid hormones and IGFs/IGFPBs are involved in modifying the association between obesity and prostate cancer risk since we previously found that BMI increased the risk of high-grade but decreased the risk of low-grade disease in PCPT (Gong et al. 2006). We showed that serum steroid levels did not modify the association of BMI with prostate cancer risk. This result is not surprising since circulating sex steroid hormones were not associated with the risk of prostate cancer in our PCPT studies (Kristal et al. 2012, Schenk et al. 2016, Yao et al. 2011) as well as others (Endogenous et al. 2008).
Our results revealed that the association between BMI and low- and high-grade prostate cancer may be modified by the IGF axis, specifically IGFBP2 and IGFs. We previously showed that only high serum IGFBP2 was a risk factor for low-grade disease, which was attenuated for men on finasteride (Neuhouser et al. 2013). In the current study, after adjusting for IGFs and IGFBP2, the association of obesity with risk reduction of low-grade disease diminished, while the risk of high-grade (Gleason 7–10) prostate cancer was increased. IGFs have mitogenic and anti-apoptotic effects and the IGF-1 axis (comprised of IGF-1, IGF-2, insulin, and their respective receptors, as well as binding proteins) regulate many physiological processes. This axis has been implicated in playing an important role in tumor development and progression. IGF-1 and −2 are responsible for cellular growth, and IGFBPs regulate the bioactivity of IGFs for binding of their receptors (Clemmons 1997). Obese men have higher levels of insulin and IGFs (Giovannucci and Michaud 2007). Obesity and hyperinsulinemia are associated with not only alterations in sex steroids, adipocytokines, but also with decreased concentrations of IGFBPs resulting in increased circulating levels of bioactive IGF-1, which in turn elevate circulating growth factors (Renehan et al. 2006). Studies have found that circulating IGFs and binding proteins may be associated with prostate cancer development or progression as well as increased risk of advanced prostate cancer (Roddam et al. 2008). In epidemiological studies, plasma IGF-1 levels were positively associated with risk of total prostate cancer and plasma fasting IGFBP-1 was strongly inversely associated with low- and intermediate-grade but not high-grade prostate cancer in the Health Professionals Follow-up Study (Cao et al. 2015). BMI was associated with IGFBP-1 but not IGF-1 (Cao et al. 2015). Further research would be needed to better understand the relationship between these proteins and prostate cancer.
There are strengths and limitations to our study. The PCPT was a large placebo-controlled randomized trial that used prostate biopsies to verify absence or presence of cancer; thus, the control group all had confirmed negative prostate biopsies, largely eliminating the possibility that controls may have had undiagnosed or undetected disease. Additional strengths included the use of a central pathology laboratory for uniform adjudication of all cases (including adjudication of Gleason grade), highly sensitive and specific assays for quantitating circulating serum steroids and IGFs, and the large sample size of our patient population. Our observational study was limited in that single time point measurement of biomarkers (e.g., serum steroids, SHBG, IGFs, IGFBPs) may not be reflect long-term circulating levels and laboratory measurements were performed on non-fasting blood samples which may affect IGF levels. Levels of circulating biomarkers may be affected by genetic and lifestyle factors, which may modify the serum concentrations of these biomarkers. Men on this trial underwent 6-core biopsy so there was a possibility that prostate cancer detection may be missed. Limitations of sample size by disease grade may have precluded detection of stronger interactions between obesity, circulating biomarkers, and treatment arm, and increased the possibility of spurious associations as a result of multiple comparisons. This analysis included largely white participants; however, this minimized concerns of selection bias or population stratification.
In conclusion, our previous study found an association between BMI and an increased the risk of high-grade and a decreased the risk of low-grade disease (Gong et al. 2006). The current follow-up study reveals that circulating serum biomarkers along the IGF-1 axis, specifically IGFBP2 and IGFs, may modify the association of obesity with prostate cancer risk. Future studies are warranted to delineate the relationship between obesity and circulating serum biomarkers and to elucidate the biological mechanism involved in driving prostate cancer progression. This is important as a thorough understanding of these mechanisms may be valuable in the development of effective prostate cancer prevention strategies and treatments. More research is necessary to ascertain the roles of these potential modifying factors in obesity-related prostate cancer risk.
Supplementary Material
ACKNOWLEDGMENTS
The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organization imply endorsement by the U.S. Government.
Funding/Support:
This work was supported by the Intramural Research Program of the Center for Cancer Research, National Cancer Institute, National Institutes of Health (ZIA BC 010453). This work was also supported by grants from the National Cancer Institute of the National Institutes of Health (P01CA108964, U10CA37429, 5UM1CA182883).
Footnotes
Conflict of Interest Disclosures: None reported.
References
- Akinyemiju T, Moore JX & Pisu M 2018. Mediating effects of cancer risk factors on the association between race and cancer incidence: analysis of the NIH-AARP Diet and Health Study. Ann Epidemiol, 28, 33–40 e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Allott EH, Masko EM & Freedland SJ 2013. Obesity and prostate cancer: weighing the evidence. Eur Urol, 63, 800–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barrington WE, Schenk JM, Etzioni R, Arnold KB, Neuhouser ML, Thompson IM Jr., Lucia MS & Kristal AR 2015. Difference in Association of Obesity With Prostate Cancer Risk Between US African American and Non-Hispanic White Men in the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA Oncol, 1, 342–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Boibessot C & Toren P 2018. Sex steroids in the tumor microenvironment and prostate cancer progression. Endocr Relat Cancer, 25, R179–R196. [DOI] [PubMed] [Google Scholar]
- Cao Y, Nimptsch K, Shui IM, Platz EA, Wu K, Pollak MN, Kenfield SA, Stampfer MJ & Giovannucci EL 2015. Prediagnostic plasma IGFBP-1, IGF-1 and risk of prostate cancer. Int J Cancer, 136, 2418–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Clemmons DR 1997. Insulin-like growth factor binding proteins and their role in controlling IGF actions. Cytokine Growth Factor Rev, 8, 45–62. [DOI] [PubMed] [Google Scholar]
- De Pergola G & Silvestris F 2013. Obesity as a major risk factor for cancer. J Obes, 2013, 291546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Discacciati A, Orsini N & Wolk A 2012. Body mass index and incidence of localized and advanced prostate cancer--a dose-response meta-analysis of prospective studies. Ann Oncol, 23, 1665–71. [DOI] [PubMed] [Google Scholar]
- Endogenous H, Prostate Cancer Collaborative G, Roddam AW, Allen NE, Appleby P & Key TJ 2008. Endogenous sex hormones and prostate cancer: a collaborative analysis of 18 prospective studies. J Natl Cancer Inst, 100, 170–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fang X, Wei J, He X, Lian J, Han D, An P, Zhou T, Liu S, Wang F & Min J 2018. Quantitative association between body mass index and the risk of cancer: A global Meta-analysis of prospective cohort studies. Int J Cancer, 143, 1595–1603. [DOI] [PubMed] [Google Scholar]
- Feigl P, Blumenstein B, Thompson I, Crowley J, Wolf M, Kramer BS, Coltman CA Jr., Brawley OW & Ford LG 1995. Design of the Prostate Cancer Prevention Trial (PCPT). Control Clin Trials, 16, 150–63. [DOI] [PubMed] [Google Scholar]
- Giovannucci E & Michaud D 2007. The role of obesity and related metabolic disturbances in cancers of the colon, prostate, and pancreas. Gastroenterology, 132, 2208–25. [DOI] [PubMed] [Google Scholar]
- Gong Z, Neuhouser ML, Goodman PJ, Albanes D, Chi C, Hsing AW, Lippman SM, Platz EA, Pollak MN, Thompson IM, et al. 2006. Obesity, diabetes, and risk of prostate cancer: results from the prostate cancer prevention trial. Cancer Epidemiol Biomarkers Prev, 15, 1977–83. [DOI] [PubMed] [Google Scholar]
- Guerrios-Rivera L, Howard L, Frank J, De Hoedt A, Beverly D, Grant DJ, Hoyo C & Freedland SJ 2017. Is Body Mass Index the Best Adiposity Measure for Prostate Cancer Risk? Results From a Veterans Affairs Biopsy Cohort. Urology, 105, 129–135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hales CM, Fryar CD, Carroll MD, Freedman DS, Aoki Y & Ogden CL 2018. Differences in Obesity Prevalence by Demographic Characteristics and Urbanization Level Among Adults in the United States, 2013–2016. JAMA, 319, 2419–2429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Harding JL, Shaw JE, Anstey KJ, Adams R, Balkau B, Brennan-Olsen SL, Briffa T, Davis TM, Davis WA, Dobson A, et al. 2015. Comparison of anthropometric measures as predictors of cancer incidence: A pooled collaborative analysis of 11 Australian cohorts. Int J Cancer, 137, 1699–708. [DOI] [PubMed] [Google Scholar]
- Holly JMP, Biernacka K & Perks CM 2020. The role of insulin-like growth factors in the development of prostate cancer. Expert Rev Endocrinol Metab, 15, 237–250. [DOI] [PubMed] [Google Scholar]
- Hoque A, Yao S, Till C, Kristal AR, Goodman PJ, Hsing AW, Tangen CM, Platz EA, Stanczyk FZ, Reichardt JK, et al. 2015. Effect of finasteride on serum androstenedione and risk of prostate cancer within the prostate cancer prevention trial: differential effect on high- and low-grade disease. Urology, 85, 616–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hsing AW, Sakoda LC & Chua S Jr., 2007. Obesity, metabolic syndrome, and prostate cancer. Am J Clin Nutr, 86, s843–57. [DOI] [PubMed] [Google Scholar]
- Jochems SHJ, Stattin P, Haggstrom C, Jarvholm B, Orho-Melander M, Wood AM & Stocks T 2020. Height, body mass index and prostate cancer risk and mortality by way of detection and cancer risk category. Int J Cancer, 147, 3328–3338. [DOI] [PubMed] [Google Scholar]
- Kaaks R, Lukanova A & Sommersberg B 2000. Plasma androgens, IGF-1, body size, and prostate cancer risk: a synthetic review. Prostate Cancer Prostatic Dis, 3, 157–172. [DOI] [PubMed] [Google Scholar]
- Kaaks R & Stattin P 2010. Obesity, endogenous hormone metabolism, and prostate cancer risk: a conundrum of “highs” and “lows”. Cancer Prev Res (Phila), 3, 259–62. [DOI] [PubMed] [Google Scholar]
- Kristal AR, Till C, Tangen CM, Goodman PJ, Neuhouser ML, Stanczyk FZ, Chu LW, Patel SK, Thompson IM, Reichardt JK, et al. 2012. Associations of serum sex steroid hormone and 5alpha-androstane-3alpha,17beta-diol glucuronide concentrations with prostate cancer risk among men treated with finasteride. Cancer Epidemiol Biomarkers Prev, 21, 1823–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Macinnis RJ & English DR 2006. Body size and composition and prostate cancer risk: systematic review and meta-regression analysis. Cancer Causes Control, 17, 989–1003. [DOI] [PubMed] [Google Scholar]
- Neuhouser ML, Kristal AR, Mclerran D, Patterson RE & Atkinson J 1999. Validity of short food frequency questionnaires used in cancer chemoprevention trials: results from the Prostate Cancer Prevention Trial. Cancer Epidemiol Biomarkers Prev, 8, 721–5. [PubMed] [Google Scholar]
- Neuhouser ML, Kristal AR, Patterson RE, Goodman PJ & Thompson IM 2001. Dietary supplement use in the Prostate Cancer Prevention Trial: implications for prevention trials. Nutr Cancer, 39, 12–8. [DOI] [PubMed] [Google Scholar]
- Neuhouser ML, Platz EA, Till C, Tangen CM, Goodman PJ, Kristal A, Parnes HL, Tao Y, Figg WD, Lucia MS, et al. 2013. Insulin-like growth factors and insulin-like growth factor-binding proteins and prostate cancer risk: results from the prostate cancer prevention trial. Cancer Prev Res (Phila), 6, 91–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nimptsch K & Pischon T 2016. Obesity Biomarkers, Metabolism and Risk of Cancer: An Epidemiological Perspective. Recent Results Cancer Res, 208, 199–217. [DOI] [PubMed] [Google Scholar]
- Perez-Cornago A, Appleby PN, Pischon T, Tsilidis KK, Tjonneland A, Olsen A, Overvad K, Kaaks R, Kuhn T, Boeing H, et al. 2017. Tall height and obesity are associated with an increased risk of aggressive prostate cancer: results from the EPIC cohort study. BMC Med, 15, 115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Price DK, Chau CH, Till C, Goodman PJ, Leach RJ, Johnson-Pais TL, Hsing AW, Hoque A, Parnes HL, Schenk JM, et al. 2016. Association of androgen metabolism gene polymorphisms with prostate cancer risk and androgen concentrations: Results from the Prostate Cancer Prevention Trial. Cancer, 122, 2332–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Renehan AG, Frystyk J & Flyvbjerg A 2006. Obesity and cancer risk: the role of the insulin-IGF axis. Trends Endocrinol Metab, 17, 328–36. [DOI] [PubMed] [Google Scholar]
- Renehan AG, Tyson M, Egger M, Heller RF & Zwahlen M 2008. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet, 371, 569–78. [DOI] [PubMed] [Google Scholar]
- Roddam AW, Allen NE, Appleby P, Key TJ, Ferrucci L, Carter HB, Metter EJ, Chen C, Weiss NS, Fitzpatrick A, et al. 2008. Insulin-like growth factors, their binding proteins, and prostate cancer risk: analysis of individual patient data from 12 prospective studies. Ann Intern Med, 149, 461–71, W83–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schenk JM, Till C, Hsing AW, Stanczyk FZ, Gong Z, Neuhouser ML, Reichardt JK, Hoque AM, Figg WD, Goodman PJ, et al. 2016. Serum androgens and prostate cancer risk: results from the placebo arm of the Prostate Cancer Prevention Trial. Cancer Causes Control, 27, 175–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Siegel RL, Miller KD, Fuchs HE & Jemal A 2021. Cancer Statistics, 2021. CA Cancer J Clin, 71, 7–33. [DOI] [PubMed] [Google Scholar]
- Tangen CM, Schenk J, Till C, Goodman PJ, Barrington W, Lucia MS & Thompson IM 2019. Variations in prostate biopsy recommendation and acceptance confound evaluation of risk factors for prostate cancer: Examining race and BMI. Cancer Epidemiol, 63, 101619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG, Lieber MM, Cespedes RD, Atkins JN, Lippman SM, et al. 2003. The influence of finasteride on the development of prostate cancer. N Engl J Med, 349, 215–24. [DOI] [PubMed] [Google Scholar]
- Vidal AC, Howard LE, Moreira DM, Castro-Santamaria R, Andriole GL Jr., & Freedland SJ 2014. Obesity increases the risk for high-grade prostate cancer: results from the REDUCE study. Cancer Epidemiol Biomarkers Prev, 23, 2936–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Watts EL, Appleby PN, Albanes D, Black A, Chan JM, Chen C, Cirillo PM, Cohn BA, Cook MB, Donovan JL, et al. 2017. Circulating sex hormones in relation to anthropometric, sociodemographic and behavioural factors in an international dataset of 12,300 men. PLoS One, 12, e0187741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yao S, Till C, Kristal AR, Goodman PJ, Hsing AW, Tangen CM, Platz EA, Stanczyk FZ, Reichardt JK, Tang L, et al. 2011. Serum estrogen levels and prostate cancer risk in the prostate cancer prevention trial: a nested case-control study. Cancer Causes Control, 22, 1121–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhong S, Yan X, Wu Y, Zhang X, Chen L, Tang J & Zhao J 2016. Body mass index and mortality in prostate cancer patients: a dose-response meta-analysis. Prostate Cancer Prostatic Dis, 19, 122–31. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.