Methods: Cholesterol measurements were available on 17,696 men. During the 17-year median follow-up, 2404 PCa cases were diagnosed. Cox regression model was used to estimate hazard ratios (HR) and their 95% confidence intervals (95% CI) for overall PCa risk and stratified by Gleason grade and tumor stage. We compared normolipidemic and hyperlipidemic men on four cholesterol parameters total cholesterol (TC), HDL, LDL, and triglycerides (TG), analyzed as time-dependent variables.
Results: TC in the highest tertile (above 5.1 mmol/l) and LDL above 3 mmol/l were associated with increased risk of Gleason 8–10 cancer (HR 1.42, 95% CI 1.04–1.95 and HR 1.38, 95% CI 1.02–1.86, respectively). Further, overall PCa risk was elevated in the 3-year lag time analysis by TC in the highest two tertiles (HR 1.27, 95% CI 1.05–1.54 for TC above 4.4 mmol/l, and HR 1.26, 95% CI 1.05–1.51 for TC above 5.1 mmol/l) and HDL in the highest tertile (HR 1.33, 95% CI 1.08– 1.64) and above 1 mmol/l (HR 1.29, 95% CI 1.01–1.65). In contrast, TC in the highest tertile was associated with a decreased risk of PCa with 20-year lag time. The risk associations for overall PCa grew stronger with added lag time but were observed only in the FinRSPC control arm. Statin use did not modify the risk association.
Conclusions: Hypercholesterolemia may increase overall PCa risk in short-term, inverse risk association was observed with 20-years’ time lag. Similar risk increase of overall PCa was also observed for elevated HDL, conflicting with previous findings on the subject.
Introduction:
The production of cholesterol through the mevalonate pathway is increased in prostate cancer (PCa) in vitro1. The accumulation of long-chain fatty acids and especially cholesterol in the cell membrane increases the amount and size of lipid rafts or detergent-resistant membranes (DRMs) and affects the protein composition of these structures. DRMs play a major role in oncogenic signaling and PCa progression through multiple signaling proteins2–4.
Increased intake of dietary saturated fat has been associated with higher levels of plasma cholesterol and increased risk of advanced PCa5-7. Dietary choices also play a pivotal role in the development of obesity, which has been shown to be a risk factor for high-grade and aggressive PCa and especially PCa mortality8,9.
Epidemiological studies investigating the association between hypercholesterolemia and overall prostate cancer risk have reported inconsistent results10–17. However, several studies point to an association between high total cholesterol and high-grade cancer18–21. Similarly, triglycerides have been linked to increased risk of prostate cancer, especially high-grade PCa16, 22. However, conflicting results have also been reported10, 12, 15, 18.
Of specific lipoproteins, elevated high-density lipoprotein (HDL) has been in some studies associated with lower risk of non-aggressive and overall PCa18, 21, 23, 24 while other studies have reported no effect12, 25. Studies examining the association between LDL and PCa risk have
produced mixed results10, 14, 18, 24–26.
A previous analysis on statin use and prostate cancer risk in the Finnish Randomized Study of Screening for Prostate Cancer (FinRSPC) found that statin use was associated with decreased overall prostate cancer risk in a dose-dependent fashion and this was not due to altered serum PSA 27. In a later study examining a cohort of 6537 prostate cancer cases in the FinRSPC statins were associated with decreased risk of PCa death especially among men managed with androgen deprivation therapy 28.
Prostate cancer has a long natural history and takes years to progress to a clinically significant disease 29. To reduce the likelihood of findings being the result of reverse causality, i.e., tumor growth influencing cholesterol levels instead of cholesterol driving prostate cancer development, we have included exposure lagging in our analyses. Time lagging is the exclusion of exposures occurring a given time period before the outcome 30, thus allowing the analysis of exposures that may have an effect on cancer development before it being clinically significant or screen positive. We have included 1-year and 3-year lag time analyses, as well as 10-year and 20-year lag time analyses to look at the effect of lipid parameters on cholesterol risk both in the
short term and long term, respectively.
Our aim in this population-based cohort study was to examine the effects of total cholesterol, triglycerides, high-density lipoprotein, and low-density lipoprotein on prostate cancer risk overall and by tumor Gleason grade and stage.
Materials and Methods
Study Cohort
The study was nested within the Finnish Randomized Study of Screening for Prostate Cancer (FinRSPC) with 80458 men aged 55, 59, 63, and 69 years at baseline and randomized during 1996– 1999 to the screening or control arm. All men were followed through national registries and those with a pre-existing PCa diagnosis at baseline were identified from nationally comprehensive Finnish Cancer Registry (FCR) and excluded (n = 203). Men in the screening arm were invited for three rounds of PSA-based screening at four-year intervals, except those aged 69 years at entry were invited twice. Men with PSA 4 ng/ml or higher were considered screen-positive and were referred for prostate biopsy. Additionally, men with PSA at 3.0– 3.9 ng/ml and free/total PSA ratio below 16% were also considered nscreen-positive. The final screening round ended in 2007. PCa cases were identified from FCR. The study population was limited to men from Pirkanmaa region for whom measurements of lipid parameters were available in the Fimlab laboratory database. In total 17696 men were included. The formation of the study cohort is illustrated in Fig. 1.
Fig 1. Flowchart of the study population
The study population was linked to the national prescription database by the Social Insurance Institution of Finland using personal identification numbers to obtain information on the purchases of blood pressure-lowering drugs, antidiabetic drugs and non-steroidal anti-inflammatory drugs (NSAIDs) during 1995– 2015 and cholesterol-lowering drugs during 1995– 2015.
A survey including questions on height and weight for calculation of BMI and on over-the-counter use of NSAIDs and acetylsalicylic acid (ASA) was mailed along with the invitations to participate in the third screening round, with 93% response rate 31.
Information on serum cholesterol parameters
The FinRSPC study cohort was linked to Fimlab laboratory database using personal identification numbers to obtain results from all serum cholesterol, high-density lipoprotein, low-density lipoprotein and triglyceride measurements during 1978– 2015. Fimlab is the leading provider of laboratory services in the Pirkanmaa region of Finland. In total 17 696 men had at least one serum cholesterol measurement recorded in the Fimlab database and of them n2404 were subsequently diagnosed with prostate cancer.
Values for each lipid parameter were determined separately for each calendar year that measurements were available for. If there was more than one measured value for a given calendar year, the mean was used. To estimate the risk association by currently recommended cut-points for cardiovascular risk according to Finnish guidelines, the cutpoint for hypercholesterolemia was set at 5.0 mmol/l (193 mg/dl), for HDL at 1.0 mmol/l (39 mg/dl), for LDL at 3.0 mmol/l (116 mg/dl) and for triglycerides at 1.7 mmol/l (151 mg/dl) 32. Further, to estimate dose-dependence of the risk association by cholesterol level we stratified each lipid parameter into three groups according to tertiles of the mean yearly for all available lipid parameter measurements available from 1978 to 2015. Tertile boundaries were 4.4 mmol/l (170 mg/dl) and 5.1 mmol/l (197 mg/dl) for total cholesterol, 2.4 mmol/l (93 mg/dl), and 3.0 mmol/l (116 mg/ dl) for LDL, 1.2 mmol/l (46 mg/dl) and 1.5 mmol/l (58 mg/dl) for HDL, and 1.0 mmol/l (89 mg/dl) and 1.5 mmol/l(133 mg/dl) for triglycerides.
Statistical analysis
Cox regression was used to estimate hazard ratios (HRs) and their 95% confidence intervals (95% CIs) for overall prostate cancer risk, as well as by Gleason score (categorized as ≤6, 7, 8–10) and tumor stage (M0 vs. M1 tumors). The follow-up started at the FinRSPC baseline and continued until PCa diagnosis, death, emigration or the end of the year 2015. Time metric was years and months since FinRSPC baseline. Missing data was coded as a separate category.
The analysis was adjusted for age and in the multivariable-adjusted analysis also for the FinRSPC trial arm, use of NSAIDs and ASA, blood pressure-lowering drugs, 5-alpha reductase inhibitors, and antidiabetic medication. Medication use was analyzed as a dichotomous variable, ever-users, and never-users.
Level of total cholesterol and other lipid parameters were analyzed as time-dependent covariates, i.e., the level was updated prospectively for each follow-up year after FinRSPC baseline based on available cholesterol measurements. Men with no cholesterol measurements on a given year were analyzed in a separate category of missing information. In addition, an additional analysis where the previously available cholesterol measurement is used for years without a measurement until a more recent measurement is available is included as supplementary material
(Supplementary Table 1.)
The long-term effects of cholesterol levels were analyzed using lagging (i.e., allowing for a latency period), where the exposure was related to outcomes occurring one to three years after the measurement, e.g., the cholesterol levels of 2001 were used to predict the risk in 2002 in the one-year lag time analysis. In addition to the 1-year lag time and 3- year lag time analyses, we also analyzed the long-term effects of cholesterol on prostate cancer risk by looking at exposures one and two decades before the outcome.
Reduction in cholesterol level after the start of statin use was calculated as the difference between the mean cholesterol level between the year 1978 and first prescription of statin drugs and the mean cholesterol level after first statin prescription and end of 2015.
All statistical tests were performed using IBM SPSS Statistics version 23.0.
Results
Population characteristics
The age distribution at baseline was similar among normolipidemic and hyperlipidemic men (Table 1). Among the 2404 prostate cancer cases diagnosed during the follow-up, there was a higher proportion of Gleason 8–10 cancer among men with high TC and LDL, and a higher proportion of Gleason 6 cancer among men with high triglycerides. Gleason scores were otherwise similarly distributed between men with cholesterol, triglycerides or lipoproteins in the normal range and men with cholesterol, triglycerides or lipoproteins outside of the recommended target area. The proportion of metastatic cases at diagnosis was similar between normolipidemic and hyperlipidemic men. The median PSA values were similarly distributed in the first and second screening round but were higher in men with TC and LDL levels above the recommendation in the third screening round. Medication use (use of cholesterol-lowering, antidiabetic and antihypertensive drugs, aspirin and other NSAIDs) was more common among men with median total cholesterol and LDL in the recommended target area. In contrast, men with HDL-C below the recommendation and triglycerides above it were more commonly medication users.
Prostate cancer risk by serum cholesterol level
None of the lipid parameters examined were associated with overall PCa risk in the age-adjusted or multivariate models when stratified by the recommended CVD risk cut-points (Table 2).
Note: Study cohort of 17,696 men participating in the Finnish Randomized Study of Screening for Prostate Cancer
*multivariable-adjusted analysis adjusted for the FinRSPC trial arm, use of NSAIDs and ASA, blood pressure-lowering drugs, 5-alpha-reductase inhibitors, and antidiabetic medication as dichotomous variables
a2404 events
b1158 events
c619 events
d441 events
e1412 events
f145 events
Total person-time: 265335.67 years
However, when dividing each lipid parameter by population tertiles, being in the highest tertile for total cholesterol was associated with increased risk of overall PCa compared with men in the lowest tertile in the age-adjusted model (HR 1.17, 95% CI 1.01–1.35), but not in the multivariate analysis.
When stratified by Gleason grade, both being in the highest tertile for total cholesterol (HR 1.42, 95% CI 1.04–1.95) compared with men in the lowest tertile and LDL level above the CVD cut-point (HR 1.38, 95% CI 1.02– 1.86) compared to men with levels below cut-point were associated with increased risk of Gleason 8–10 cancer. No risk association was found when stratified by tumor stage.
Lag time analyses
In the 3-year lag time analysis, the overall PCa risk was higher among men in the highest two tertiles for total cholesterol compared with those in the lowest tertile; (HR 1.27, 95% CI 1.05–1.54) for total cholesterol above 4.4 mmol/l and (HR 1.26, 95% CI 1.05–1.51) for above 5.1 mmol/l (Table 3). Similarly, the overall PCa risk in the 3-year lag time analysis was elevated for both HDL values in the highest tertile compared with men in the lowest tertile (HR 1.33, 95% CI 1.08–1.64) and HDL values above the CVD cut-point compared with values below the CVD cutpoint HR 1.29, 95% CI 1.01–1.65). The risk increased consistently in the main and 1-year analyses but did not become significant until the 3-year lag time analysis.
Note: Study cohort of 17,696 men participating in the Finnish Randomized Study of Screening for Prostate Cancer
*multivariable-adjusted analysis adjusted for the FinRSPC trial arm, use of NSAIDs and ASA, blood pressure-lowering drugs, 5-alpha-reductase inhibitors, and antidiabetic medication as dichotomous variables
a2404 events
b441 events
c145 events
Total person-time: 265335.67 years in main analysis, 265080.00 years in 1 year lag time, and 261927.50 years in 3 year lag time analysis
The increased Gleason score 8–10 cancer risk associated with LDL above the cut-point and total cholesterol in the highest tertile in the main analysis was no longer statistically significant in the 1-year and 3-year lag time analyses.
No association was found between elevated lipid parameter values and PCa risk when stratified by tumor Gleason score and stage at diagnosis in the lag time analyses.
Long-term association between serum cholesterol and PCa risk
Being in the highest tertile for total cholesterol was associated with decreased risk of overall PCa compared with the lowest tertile in the 20-year lag time analysis (HR 0.38, 95% CI 0.15–0.97), otherwise, no long-term risk associations were observed (Table 4). No risk association was found for elevated triglyceride levels. Hypercholesterolemia and hypertriglyceridemia were not associated with PCa risk in analyses stratified by tumor Gleason score or stage at diagnosis.
Note: Study cohort of 17,696 men participating in the Finnish Randomized Study of Screening for Prostate Cancer
*multivariable-adjusted analysis adjusted for the FinRSPC trial arm, use of NSAIDs and ASA, blood pressure-lowering drugs, 5-alpha-reductase inhibitors, and antidiabetic medication as dichotomous variables
a1820 events
b394 events
c339 events
d78 events
e119 events
f32 events
Total person-time: 201268.42 years in the 10 year lag time analysis, and 47692.50 years in the 20 year lag time analysis
Impact of change in cholesterol level after initiation of statin use
Reduction in cholesterol level after statin use was not associated with overall PCa risk or risk by Gleason grade or tumor stage. The result was similar independent of the magnitude of change after initiation of statin use (Supplementary Table 2).
Sensitivity analyses
To distinguish the role of hypercholesterolemia from that of statin use, we performed analyses limited to statin users and on-users separately. No statistically significant difference in overall prostate cancer risk between the groups was found; p for interaction = 0.085 (HR 1.07, 95% CI 0.94– 1.22 among statin users and HR 1.08, 95% CI 0.89–1.30 among for men with TC above 5 mmol/l compared to men with TC below 5 mmol/l). Having TC above 5 mmol/l was associated with increased risk of Gleason 8–10 cancer (HR 1.54, 95 % CI 1.01–2.37) among non-users but not in statin users (Supplementary Table 3).
FinRSPC study arm modified the risk association; p for interaction <0.001. Risk elevation for overall prostate cancer risk, albeit non-significant, by total cholesterol was observed only among men in the control arm, not in the screening arm. The results stratified by trial arm are included as supplementary material (Supplementary Table 4). Interestingly, when stratified by Gleason grade and tumor stage, increased risk for metastatic cancer among men with higher TC and LDL levels was only seen in the screening arm (HR 2.98, 95% CI 1.21– 7.36 for TC above 5 mmol/l and HR 2.92, 95% CI 1.18– 7.21 for LDL above 3 mmol/l). However, this risk increase was not consistent or dose-dependent in the analysis by tertiles. Increased risk for Gleason grade 8– 10 cancer was found among men in the highest tertile for TC compared to the lowest tertile in the screening arm (HR 1.85, 95% CI 1.05– 3.27). The risk association between total cholesterol and PCa risk was significant (HR 1.36, 95% CI 1.05– 1.77) for cases diagnosed before 2004, i.e., before opportunistic PSA testing became widespread in Finland 33, but not in cases diagnosed after 2004.
No significant risk modification was observed by a number of available cholesterol measurements; p for interaction = 0.78.
To estimate the effect of the difference in all-cause mortality between hyperlipidemic and normolipidemic men on the results of the 20-year lag time analysis, we performed an additional analysis with yearly cholesterol measures lagged 20 years as predictive variable and all-cause mortality as the outcome measure. Men with total cholesterol above 5 mmol/l were not statistically significantly more likely to die due to any cause compared to men with TC below 5 mmol/l (HR 1.17, 95% CI 0.63– 2.18).
When using the previously available cholesterol measurement for years with missing values the results were similar to the main analysis with the exception of TC above 5 mmol/l being associated with increased risk of Gleason 8–10 cancer compared with TC below 5 mmol/l (HR 1.31, 95 % CI 1.03– 1.68) even without time lagging.
Discussion
Elevated total cholesterol and HDL were associated with increased PCa risk overall within 3 years after the measurement. Further, elevated total cholesterol and LDL were associated with increased risk of Gleason 8– 10 PCa, but the risk increase did not persist long-term. PCa risk did not vary in relation to the degree of reduction in cholesterol level after the start of statin use.
Epidemiological studies examining the association between total cholesterol and PCa risk have been inconsistent. High total cholesterol may not affect overall PCa risk 10–17, 34. It has, however, been associated with a higher risk of high-grade and metastatic disease 18–21. Our study supports these previous findings in that a short-term risk increase of specifically high-grade PCa was found, though it did not persist in the lag time-analyses. Overall PCa risk in hypercholesterolemic men increased consistently with increasing lag time suggesting a more long-term association between elevated cholesterol and cancer development. Still, there was no significant risk increase in the 10-year lag time analysis and a decrease in risk in the 20-year lag time analysis. Cholesterol measurements for 10- year and 20-year analyses were, however, available only for a limited number of men, thus should be interpreted with caution. The decrease in prostate cancer risk seen in the 20- year lag time analysis could also be explained by competing risks since men with higher total cholesterol levels are at increased risk of dying from cardiovascular causes. Van Hemelrijck et al. found that considering competing risks attenuated the protective effect of higher glucose levels on prostate cancer risk. However, competing for risk analysis did not show a significant effect on the association between cholesterol and PCa15 . We found no statistically significant difference in all-cause mortality between hypercholesterolemic and normocholesterolemic men with cholesterol measurements lagged 20 years in our sensitivity analysis.
An interesting observation in the sensitivity analyses was that the risk increase for overall prostate cancer was only found in the control arm of the FinRSPC. This could be due to men with increased cholesterol levels having more contacts with healthcare professionals and being screened more often, thus increasing the probability of detecting prostate cancer in this population. Platz et al. showed previously in a study protocol aimed at reducing possible detection bias that statin drugs were not associated with prostate cancer in a population screened for prostate cancer yearly. However, this finding differs from what has been reported in the FinRSPC population previously, a decrease in overall prostate cancer risk among statin users in a dose-dependent manner was observed even when comparing men in the screening arm 27. Also, increased opportunistic PSA screening among men with higher cholesterol would more likely result in increased detection of low-grade cancer compared to the increase in high-grade cancer that was observed in our study. Further, when stratified by tumor stage only men in the screening arm with TC and LDL above the recommended CVD cut-point had an increased risk of metastatic cancer compared to normolipidemic men (HR 2.98, 95% CI 1.21– 7.36 for TC above 5 mmol/l and HR 2.92, 95% CI 1.18– 7.21 for LDL above 3 mmol/l). The risk association reached statistical significance only when comparing lipid values by CVD cut-point, not by tertiles, and a decreased risk was seen when comparing men in the second tertile to men in the first tertile for LDL. This finding does not fit well with the hypothesis that men with higher cholesterol are screened for PCa more often and may be due to a multitude of confounding factors.
Administration of LDL cholesterol increases cell number in prostate cancer cell lines relative to controls 35. In addition, oxidized LDL has been shown to be associated with PCa aggressiveness along with having a role in the proliferation, migration, and invasion of prostate cancer cell lines 36. There is relatively little epidemiological evidence on the association between LDL and PCa risk. Some studies have shown high LDL to be associated with an increased risk of PCa18 , 26 but many have shown no association 10, 14, 24, 25. Of the studies that showed a positive association between high LDL and PCa risk (Kok et al., 2011) included 2161 men. Information on medication use and BMI was collected from questionnaires. High LDL was associated with an increased risk of both overall and aggressive PCa in men who were not statin users. The results were non-significant after the statin users were included. Farwell et al. (2011) found increased risk of overall PCa in the highest two quartiles of LDL and increased risk of high grade PCa in the highest quartile of LDL in a cohort of 55,875 men. Concordantly, we showed high LDL to be associated with an increased risk of high-grade PCa. However, the association is not likely causal as it disappeared in the lag-time analyses. Of note is that statin use did not affect the risk association with LDL in our study.
HDL-C has an important role in reverse cholesterol transport and maintaining lipid homeostasis on a cellular level. HDL receptor SR-B1 expression has been shown to be increased in high-grade and metastatic PCa 37, 38. In addition, HDL induces cell proliferation and migration in androgen-independent but not androgen-dependent PCa cells, but does not affect intracellular cholesterol levels possibly due to the ability of SR-B1 receptors to donate cholesterol to cells 39. Previous epidemiological studies have shown that low HDL levels are not clearly associated with PCa risk or might slightly increase it 25, 40. The largest previous cohort study including 69,735 men, the Swedish AMORIS study 24, found a positive association between low HDL and PCa risk. It had a similar length of follow-up compared to our study, comparable mean age, and similarly relied on national registry data. The AMORIS study had data on a greater number of HDL measurements whereas our data on HDL measurements spanned a longer time. The results were statistically significant after the exclusion of the first 3 years of follow-up. The results were not significant using the cut-point of 1.03 mmol/l for HDL. Our study did not show a risk increase for HDL in the main analysis. However, conversely to the AMORIS study, high HDL was associated with an increased risk of overall PCa in the 3-year lag time analysis. Further, the risk association between high HDL and overall PCa increased consistently with added lag time suggesting a long-term risk association. Similarly, a recent study by Heir et al. (2016) found low BMI and high physical fitness to be associated with increased prostate cancer risk. Both high physical fitness and especially low BMI have been previously found to be associated with higher HDL levels 41, 42.
The strengths of this study include a large population-based cohort of 17,696 men, accurate data on lipid parameters spanning over twenty years, and the possibility to analyze the effect of initiation of statin use on PCa risk. Our findings add to the existing literature, as we were able to analyze multiple lipid parameters and the values were not gathered at baseline only, but over a period of several decades. This enabled us to assess long-term associations between cholesterol level and PCa risk, and account for changes in cholesterol level during follow-up. We had detailed data on multiple cholesterol measurements over the course of the follow-up allowing time-dependent prospective analysis, which to our knowledge has not been done in previous studies on this topic. Further, reliable information on the time of death and PCa diagnosis was available from national registries. In addition, data on Gleason score was available for most PCa cases.
However, our study was not without its limitations. Cholesterol measurements were available only for 22% of the eligible study subjects. The subjects with cholesterol data had a varying number of measurements and they were made at irregular intervals. Data on BMI was available only for 18% of men and we had no data on physical exercise, diet or smoking status. These confounding factors may affect the risk associations, increasing the risk estimates observed among hypercholesterolemic men. In addition, we did not use the correction for multiple testing as most results are only borderline significant even at the conventional p-value threshold. Nevertheless, the increased risk estimates for high-risk PCA among hypercholesterolemic were fairly consistent, arguing against the chance finding.
Hypercholesterolemia may increase overall PCa risk in short-term. However, an inverse risk association was observed with 20-years’ time lag. Similar risk increase of overall PCa was also observed for elevated HDL, conflicting with previous findings on the subject.
Authors: Teemu J. Murtola1,2, Tatu V. J. Kasurinen1, Kirsi Talala3, Kimmo Taari4, Teuvo L. J. Tammela1,2, Anssi Auvinen5
These authors contributed equally: Teemu J. Murtola, Tatu V. J. Kasurinen.
Author Affiliation:
1. Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland
2. Department of Urology, Tampere University Hospital, Tampere, Finland
3. Finnish Cancer Registry, Helsinki, Finland
4. Department of Urology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland
5. Faculty of Social Sciences, University of Tampere, Tampere, Finland
Conflicts of Interest: The manuscript was supported by a competitive research grant from the Expert Responsibility Area of The Pirkanmaa Hospital District to TJM. Dr. TJM has worked as a paid consultant for Astellas and Janssen-Cilag and received lecture fees from Astellas, Janssen-Cilag, Abbvie and MSD. Dr. KT has received consulting fees from Abbvie, research funding from Medivation, and travel support from Astellas, and Orion. Professor TLJT has worked as a paid consultant for Astellas and Janssen-Cilag, and received lecture fees from Astellas, Janssen-Cilag, Abbvie and MSD. Professor AA has received a lecture fee from MSD, and worked as a paid consultant for Epid Research Inc. The remaining authors declare that they have no conflict of interest.
References:
1. Krycer JR, Brown AJ. Cholesterol accumulation in prostate cancer: a classic observation from a modern perspective. Biochim Biophys Acta. 2013;1835:219–29.
2. Di Vizio D, Solomon KR, Freeman MR. Cholesterol and cholesterol-rich membranes in prostate cancer: an update. Tumori. 2008;94:633–9.
3. Hager MH, Solomon KR, Freeman MR. The role of cholesterol in prostate cancer. Curr Opin Clin Nutr Metab Care. 2006;9:379–85.
4. Zadra G, Photopoulos C, Loda M. The fat side of prostate cancer. Biochim Biophys Acta. 2013;1831:1518–32.
5. Gathirua-Mwangi WG, Zhang J. Dietary factors and risk for advanced prostate cancer. Eur J Cancer Prev. 2014;23:96–109.
6. Clarke R, Frost C, Collins R, Appleby P, Peto R. Dietary lipids and blood cholesterol: quantitative meta-analysis of metabolic ward studies. BMJ. 1997;314:112.
7. Whittemore AS, Kolonel LN, Wu AH, John EM, Gallagher RP, Howe GR, et al. Prostate cancer in relation to diet, physical activity, and body size in blacks, whites, and Asians in the United States and Canada. J Natl Cancer Inst. 1995;87:652–61.
8. Allott EH, Masko EM, Freedland SJ. Obesity and prostate cancer: weighing the evidence. Eur Urol. 2013;63:800–9.
9. Gong Z, Neuhouser ML, Goodman PJ, Albanes D, Chi C, Hsing AW, et al. Obesity, diabetes, and risk of prostate cancer: results from the prostate cancer prevention trial. Cancer Epidemiol Biomark Prev. 2006;15:1977–83.
10. His M, Zelek L, Deschasaux M, Pouchieu C, Kesse-Guyot E, Hercberg S, et al. Prospective associations between serum biomarkers of lipid metabolism and overall, breast and prostate cancer risk. Eur J Epidemiol. 2014;29:119–32.
11. Kitahara CM, Berrington de Gonzalez A, Freedman ND, Huxley R, Mok Y, Jee SH, et al. Total cholesterol and cancer risk in a large prospective study in Korea. J Clin Oncol. 2011;29:1592–8.
12. Martin RM, Vatten L, Gunnell D, Romundstad P, Nilsen TI. Components of the metabolic syndrome and risk of prostate cancer: the HUNT 2 cohort, Norway. Cancer Causes Control. 2009;20:1181–92.
13. Mondul AM, Clipp SL, Helzlsouer KJ, Platz EA. Association between plasma total cholesterol concentration and incident prostate cancer in the CLUE II cohort. Cancer Causes Control. 2010;21:61–68.
14. Jacobs EJ, Stevens VL, Newton CC, Gapstur SM. Plasma total, LDL, and HDL cholesterol and risk of aggressive prostate cancer in the Cancer Prevention Study II Nutrition Cohort. Cancer Causes Control. 2012;23:1289–96.
15. Van Hemelrijck M, Garmo H, Holmberg L, Walldius G, Jungner I, Hammar N, et al. Prostate cancer risk in the Swedish AMORIS study: the interplay among triglycerides, total cholesterol, and glucose. Cancer . 2011;117:2086–95.
16. Arthur R, Moller H, Garmo H, Holmberg L, Stattin P, Malmstrom H, et al. Association between baseline serum glucose, triglycerides and total cholesterol, and prostate cancer risk categories. Cancer Med. 2016;5:1307–18.
17. Heir T, Falk RS, Robsahm TE, Sandvik L, Erikssen J, Tretli S. Cholesterol and prostate cancer risk: a long-term prospective cohort study. BMC Cancer. 2016;16:643.
18. Kok DE, van Roermund JG, Aben KK, den Heijer M, Swinkels DW, Kampman E, et al. Blood lipid levels and prostate cancer risk; a cohort study. Prostate Cancer Prostatic Dis. 2011;14:340–5.
19. Platz EA, Clinton SK, Giovannucci E. Association between plasma cholesterol and prostate cancer in the PSA era. Int J Cancer. 2008;123:1693–8.
20. Platz EA, Till C, Goodman PJ, Parnes HL, Figg WD, Albanes D, et al. Men with low serum cholesterol have a lower risk of high-grade prostate cancer in the placebo arm of the prostate cancer prevention trial. Cancer Epidemiol Biomark Prev. 2009;18:2807–13
21. Mondul AM, Weinstein SJ, Virtamo J, Albanes D. Serum total and HDL cholesterol and risk of prostate cancer. Cancer Causes Control. 2011;22:1545–52.
22. Salgado-Montilla J, Soto Salgado M, Surillo Trautmann B, Sanchez-Ortiz R, Irizarry-Ramirez M. Association of serum lipid levels and prostate cancer severity among Hispanic Puerto Rican men. Lipids Health Dis. 2015;14:111.
23. Ahn J, Lim U, Weinstein SJ, Schatzkin A, Hayes RB, Virtamo J, et al. Prediagnostic total and high-density lipoprotein cholesterol and risk of cancer. Cancer Epidemiol Biomark Prev. 2009;18:2814–21.
24. Van Hemelrijck M, Walldius G, Jungner I, Hammar N, Garmo H, Binda E, et al. Low levels of apolipoprotein A-I and HDL are associated with risk of prostate cancer in the Swedish AMORIS study. Cancer Causes Control. 2011;22:1011–9.
25. Borgquist S, Butt T, Almgren P, Shiffman D, Stocks T, Orho-Melander M, et al. Apolipoproteins, lipids and risk of cancer. Int J Cancer. 2016;138:2648–56.
26. Farwell WR, D’Avolio LW, Scranton RE, Lawler EV, Gaziano JM. Statins and prostate cancer diagnosis and grade in a veterans population. J Natl Cancer Inst. 2011;103:885–92.
27. Murtola TJ, Tammela TL, Maattanen L, Huhtala H, Platz EA, Ala-Opas M, et al. Prostate cancer and PSA among statin users in the Finnish prostate cancer screening trial. Int J Cancer. 2010;127:1650–9.
28. Murtola TJ, Peltomaa AI, Talala K, Maattanen L, Taari K, Tammela TLJ, et al. Statin use and prostate cancer survival in the Finnish Randomized Study of Screening for Prostate Cancer. Eur Urol Focus. 2017;3:212–20.
29. Crawford ED. Understanding the epidemiology, natural history, and key pathways involved in prostate cancer. Urology. 2009;73:S4–S10.
30. Checkoway H, Checkoway H, Pearce N, Pearce N, Hickey JLS, Hickey JLS, et al. Latency analysis in occupational epidemiology. Arch Environ Health. 1990;45:95–100.
31. Sarre S, Maattanen L, Tammela TL, Auvinen A, Murtola TJ. Postscreening follow-up of the Finnish Prostate Cancer Screening Trial on putative prostate cancer risk factors: vitamin and mineral use, male pattern baldness, pubertal development and non-steroidal anti-inflammatory drug use. Scand J Urol. 2016;50:1–7.
32. Dyslipidemia. Current Care Guidelines. Working group set up by the Finnish Medical Society Duodecim and Finnish Society of Internal Medicine. Helsinki: The Finnish Medical Society Duodecim, 2013. www.kaypahoito.fi Accessed 30 Mar 2016.
33. Kilpelainen TP, Pogodin-Hannolainen D, Kemppainen K, Talala K, Raitanen J, Taari K, et al. Estimate of opportunistic prostate-specific antigen testing in the Finnish Randomized Study of Screening for Prostate Cancer. J Urol. 2017;198:50–57.
34. YuPeng L, YuXue Z, PengFei L, Cheng C, YaShuang Z, DaPeng L, et al. Cholesterol levels in blood and the risk of prostate cancer: a meta-analysis of 14 prospective studies. Cancer Epidemiol Biomark Prev. 2015;24:1086–93.
35. Murtola TJ, Syvala H, Pennanen P, Blauer M, Solakivi T, Ylikomi T, et al. The importance of LDL and cholesterol metabolism for prostate epithelial cell growth. PLoS One. 2012;7:e39445.
36. Wan F, Qin X, Zhang G, Lu X, Zhu Y, Zhang H, et al. Oxidized low-density lipoprotein is associated with advanced-stage prostate cancer. Tumour Biol. 2015;36:3573–82.
37. Schorghofer D, Kinslechner K, Preitschopf A, Schutz B, Rohrl C, Hengstschlager M, et al. The HDL receptor SR-BI is associated with human prostate cancer progression and plays a possible role in establishing androgen independence. Reprod Biol Endocrinol. 2015;13:88.
38. Gutierrez-Pajares JL, Ben Hassen C, Chevalier S, Frank PG. SR-BI: linking cholesterol and lipoprotein metabolism with breast and prostate cancer. Front Pharmacol. 2016;7:338.
39. Sekine Y, Demosky SJ, Stonik JA, Furuya Y, Koike H, Suzuki K, et al. High-density lipoprotein induces proliferation and migration of human prostate androgen-independent cancer cells by an ABCA1-dependent mechanism. Mol Cancer Res. 2010;8:1284–94.
40. Kotani K, Sekine Y, Ishikawa S, Ikpot IZ, Suzuki K, Remaley AT. High-density lipoprotein and prostate cancer: an overview. J Epidemiol. 2013;23:313–9.
41. Tall AR. Exercise to reduce cardiovascular risk—how much is enough? N Engl J Med. 2002;347:1522–4.
42. Lamon-Fava S, Wilson PWF, Schaefer EJ. Impact of body mass index on coronary heart disease risk factors in men and women. Arterioscler Thromb Vasc Bio. 1996;16:1509–15.
Received: 31 January 2018 / Revised: 26 April 2018 / Accepted: 16 May 2018 © Springer Nature Limited 2018
Read More: A Commentary from the Editor of PCAN - Stephen J. Freedland, MD