SEARCH

SEARCH BY CITATION

Keywords:

  • prostate cancer;
  • PSA;
  • polyunsaturated fatty acids;
  • Jamaica

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

OBJECTIVE

To investigate the relationship between erythrocyte membrane polyunsaturated fatty acid (PUFA) and serum prostate- specific antigen (PSA) levels in Jamaican men, as there may be an association between prostate cancer incidence and dietary fatty acids, and prostate cancer incidence in Jamaica is among the highest in the world.

PATIENTS AND METHODS

Blood from 107 Jamaican men was analysed for 32 individual fatty acids and PSA levels. Special attention was given to correlations between Ω3 and Ω6 PUFAs and PSA. Data were analysed using standard linear regression methods.

RESULTS

The mean PSA was 18.6 ng/mL (normal 0–4.0); for age groups of 51–60, 61–70 and 71–80 years the levels were 14, 26 and 23 ng/mL, respectively. Eicosapentaenoic acid (Ω3) levels decreased as PSA exceeded 10 ng/mL (P = 0.02). Arachidonic acid (Ω6) levels decreased as PSA was < 2 ng/mL (P = 0.02). Linoleic acid (Ω6) levels decreased in men with PSA levels of 2–10 ng/mL (P = 0.04). In men with a PSA of > 10 ng/mL there was a positive correlation between the ratio of Ω6 to Ω3 PUFAs and PSA (P = 0.036); there was also a negative correlation between the ratio of Ω3 to Ω6 PUFAs and PSA (P = 0.08). When the ratio of Ω3 PUFAs over the products of Ω6 PUFAs were used, this trend was significant (P= 0.03).

CONCLUSIONS

Increased levels of Ω6 PUFAs and the ratio of Ω6/Ω3 PUFAs in Jamaican men are associated with an increased mean PSA level and risk of prostate cancer. Additional studies are needed to establish a causal link between dietary fatty acid intake and the development of prostate cancer in Jamaican men.


Abbreviations
PUFA

polyunsaturated fatty acids

DGLA

dihomo-γ linolenic acid

AA

arachidonic acid

EPA

eicosapentaenoic acid.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

Jamaica has the highest reported incidence of prostate cancer in the world, with a rate of 304/100 000 across all age groups [1]. Although most of the Jamaican population is of African descent and shares a common genetic lineage with African-Americans [2] in the USA, the age-specific incidence of prostate cancer in Jamaican men aged 61–80 years is more than double that in African-American men [1,3]. Investigations have been undertaken to establish an epigenetic or environmental explanation for the discrepancy in prostate cancer incidence between such genetically related populations [4].

The dietary consumption of fatty acids has been proposed as a contributing factor to prostate cancer incidence [5]. The Ω6 polyunsaturated fatty acids (PUFAs) have been shown to promote prostate cancer cell growth in vitro[6–8], and increased levels of Ω6 PUFAs have been measured in men with prostate cancer [9–12]. Conversely, levels of Ω3 PUFAs are diminished in men with prostate cancer, perhaps because they have a protective effect on the prostate. In vitro data suggest that Ω3 PUFAs directly suppress prostate cancer cell growth [13,14]. The ratio of Ω3 to Ω6 PUFAs is inversely associated with prostate cancer risk [15]. Certain foods high in PUFAs, such as the ackee fruit, are staples of the Jamaican diet [16].

We hypothesized that, as a result, Jamaican men would have increased levels of Ω6 PUFAs which may be associated with prostate carcinogenesis. The current study was designed to determine whether there was a correlation between erythrocyte membrane Ω6 and Ω3 PUFAs, and serum PSA levels in Jamaican men.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

In all, 107 men were enrolled through the Public Hospital Urology outpatient clinic of Kingston, Jamaica, between June and July 2002. Located in the capital city of Kingston this hospital is recognized by the Jamaican Ministry of Health as the only multidisciplinary (Type A) public hospital in the city, serving as a final referral point for secondary and tertiary public healthcare services. It is the largest (422 beds) of the two Type A public healthcare institutions in the entire island and was selected as the study site to obtain a representative sample of patients.

The enrolment criteria included age > 35 years, no diagnosis of prostate cancer, no prostate biopsy or surgery during the previous month, no history of diabetes, and no history of using medications that alter cholesterol levels. Patients were being seen in the urology clinic for genitourinary problems not related to the prostate. Written informed consent was obtained from each participant; the informed consent document and study design were approved by both The University of Chicago Institutional Review Board and the Advisory Panel on Ethics and Medico-Legal Affairs, Ministry of Health, Kingston, Jamaica, West Indies.

The patients were interviewed either before or after their clinic appointment. After consent was obtained from a patient, information was collected by a trained medical student about race, socio-economic status and any family history of prostate cancer. Blood was collected from the patient after consent was obtained; samples were drawn only between 10.00 and 13.00 hours to ensure equivalent processing, and two separate samples were collected. The first was used for serum isolation for PSA determination. The first tube was allowed to clot for 45 min and then centrifuged at 600 g for 15–20 min, the serum layer isolated and then frozen at − 80 °C. The second sample was used for erythrocyte membrane fatty-acid profiles. The blood was collected in an EDTA-infused tube and refrigerated. The matched pair of samples were specially packed and then shipped overnight by courier service to Metametrix Clinical Laboratory (Norcross, Georgia), where serum PSA levels were measured by immunometric analysis. The normal serum PSA range was set at 0.0–4.0 ng/mL [17]. Erythrocyte membrane fatty-acids were analysed by gas chromatography after esterification by a 1:4 acetyl chloride iso-octane mixture (Metametrix).

Linear regression models were applied (by a faculty epidemiologist-statistician, K.E.P.) with serum PSA levels separated into three categories (<2, 2–10 and > 10 ng/mL) used to define subsets of men who could be considered as having a very low, low to medium and high risk of developing prostate cancer, respectively. Using PSA strata based on this risk and correlating these with fatty acid levels allowed a clinically meaningful relationship to be inferred.

The Ω3/Ω6 and Ω6/Ω3 PUFA ratios were measured and reported consistently with previously described methods [11]. The ratios reported are intended to assess eicosanoid precursor intake and balance. The precursor fatty acids are dihomo-γ linolenic (DGLA, series 1), arachidonic acid (AA, series 2), and eicosapentaenoic (EPA, series 3), and it is the relative ratios of these fatty acids that determine the magnitude of cell response. The AA/EPA (Ω6/Ω3) ratio in plasma estimates the series 2/3 relationship, and the EPA/DGLA (Ω3/Ω6) the series 3/1 relationship relative to the supply of PUFAs for constructing new membrane phospholipids.

Age-controlled correlation coefficients were determined for the individual Ω3 and Ω6 PUFAs, and Ω3/Ω6 and Ω6/Ω3 PUFA ratios (as described) and PSA. The ratio of Ω3 to the products of Ω6 was calculated using erythrocyte membrane levels of EPA over the product of erythrocyte membrane levels of linoleic acid and AA.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

In all, 114 men were asked to participate in the study but three declined (fear of the needle used for venepuncture); 111 matched blood samples were collected but four samples were damaged in shipping, leaving 107 available for analysis. All 107 men were of Afro-Caribbean descent and were being evaluated in the urology clinic for problems not related to prostate cancer. The mean age was 65.2 years; a minority (6.5%) had a family history of prostate cancer and most were from Kingston (population 700 000–800 000), with the remainder residing in parishes across the island.

The mean PSA of the 107 men was 18.6 ng/mL (normal range 0–4.0 ng/mL). Fig. 1A shows PSA levels grouped by age and compared between the present men and published information on other ethnic groups [18,19]. Fig. 1B shows the prostate cancer incidence rates for the same four ethnic populations [1,3,20]. For the two older groups the incidence in Jamaican men is more than double that for African-Americans.

image

Figure 1. (A) The mean serum PSA level in matched age groups among Jamaican (black), African-American (green), Caucasian-American (red) and Chinese (open) men. (B) The incidence of prostate cancer in matched age groups among Jamaican (black), African-American (green), Caucasian-American (red) and Chinese (open) men. Across the entire male population, the incidence rate of prostate cancer in Jamaican men is 304 (all per 100 000). For the older groups the incidence in Jamaican men is nearly double that for African-Americans.

Download figure to PowerPoint

The Appendix lists the fatty acids evaluated by gas chromatography in this study; Table 1 summarizes the linear regression analysis results of the PUFAs. There was a statistically significant relationship with serum PSA; as PSA values increased there was a corresponding decrease in erythrocyte membrane levels of the Ω3 PUFA EPA (P < 0.02 for PSA > 10 vs ≤ 10 ng/mL). Conversely, erythrocyte membrane levels of AA, an Ω6 PUFA, correlated negatively with PSA, being lower in men with a PSA of < 2 than with > 2 ng/mL (P = 0.02). Erythrocyte membrane levels of linoleic acid, another Ω6 PUFA, were negatively correlated with PSA in men with intermediate PSA levels (P = 0.04). Correlations between Ω6/Ω3 and Ω3/Ω6 PUFA ratios and the ratio Ω3/(Ω6) products with serum PSA are also shown in Table 1. There was a positive correlation between the Ω6/Ω3 ratio and men with a PSA of ≥ 10 ng/mL (P = 0.036), a negative correlation between the Ω3/Ω6 ratio and a PSA of ≥ 10 ng/mL (P = 0.09) and with Ω3/Ω6 products (P= 0.03). Likewise the ratio of Ω3/Ω6 products for men with a PSA of ≤ 2 was also nearly significantly correlated (P = 0.08).

Table 1.  Linear regression analysis between the relationships of erythrocyte membrane levels of EPA, AA and linoleic acids, and W3/W6 and Ω6/Ω3 PUFA ratios, with serum PSA
VariablePSA level, ng/mL
≤22–10 ≥10
N383237
t, P values
EPA (Ω3)1.45, NS −0.66, NS −2.47, 0.02
AA (Ω6) −2.45, 0.02 −0.38, NS 0.26, NS
Linoleic (Ω6)0.27, NS −2.13, 0.04 −0.54, NS
Ω3/Ω60.79, 0.437 −0.12, 0.904 −1.74, 0.091
Ω6/Ω3 −1.13, 0.2681.08, 0.2902.19, 0.036
Ω3/Ω6 product1.83, 0.080.79, NS −2.25, 0.03

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

Jamaica has the highest incidence of prostate cancer in the world [1]. Jamaican men have a significantly higher age-adjusted incidence rate of prostate cancer than African-American men, with whom they share a common genetic ancestry [2]. In Jamaica certain foods high in PUFAs comprise a substantial portion of the diet. One such example is the Ackee; this fruit, unique to Jamaica and a staple of its diet, is rich in Ω6 PUFAs [16].

We postulated that a diet rich in fatty foods might contribute to prostate cancer incidence in Jamaica, hypothesising that Jamaican men would have increased erythrocyte membrane levels of Ω6 PUFAs, decreased levels of Ω3 PUFAs, and increased Ω6/Ω3 and decreased Ω3/Ω6 PUFA ratios, and that all of these would be associated with increased serum PSA values. Kobayashi et al.[21] showed a three-fold decrease in Ω3 PUFA levels in Japanese men living in Brazil (3.9%), where prostate cancer mortality is 7.8/100 000, compared with Japanese men living in Akita, Japan (10.9%), where prostate cancer mortality is 2.7/100 000.

In the present study the mean PSA level among Jamaican men was 18.6 ng/mL. Once these men reached the age of 51 years, the mean PSA value was significantly higher than in age-matched men of African-American, Caucasian-American and Chinese ethnicities (Fig. 1A). Although serum PSA is an imperfect biomarker for prostate cancer, this trend parallels the incidence of prostate cancer in Jamaican men. However, the present mean serum PSA levels are substantially lower than those reported in Jamaican men diagnosed with prostate cancer (37 ng/mL) [22].

The present data show a statistically significant decrease in EPA (Ω3 PUFA) levels in men with a PSA of ≥ 10 ng/mL; as PSA values increased, EPA levels decreased. These results are consistent with other studies that suggest a protective association between Ω3 PUFAs and prostate cancer risk [7,8]. Similarly, there was a significant negative correlation between erythrocyte-membrane levels of AA (Ω6 PUFA) and men with a PSA level of ≤ 2 ng/mL. This is also consistent with other studies indicating a correlation between prostate cancer and increasing levels of Ω6 PUFAs [9–12]. There was a significant negative relationship between erythrocyte-membrane levels of linoleic acid (Ω6 PUFA) and men with a PSA of 2–10 ng/mL.

Freeman et al.[15] advocated the use of Ω3 and Ω6 PUFA ratios, as opposed to absolute values of the individual Ω-PUFAs, when attempting to assess associations with prostate cancer, because of competitive immunoregulatory effects. The present results show a significant positive relationship between Ω6/Ω3 PUFAs, and a negative association between Ω3/Ω6 PUFAs, and increasing PSA. This relationship became more significant when the products of linoleic acid and AA were substituted for their sums in the ratio. This supports other studies indicating that a low-fat, fish-oil supplemented diet rich in Ω3 PUFAs significantly increases the Ω3/Ω6 ratio in plasma and adipose tissue, which may have important implications for prostate cancer prevention [23,24].

Both Ω3 and Ω6 PUFAs are substrates for the cyclooxygenase-2 (COX-2) enzyme, responsible for prostaglandin production and elevated in men with clinical prostate cancer [25]. Prostaglandin and leukotriene metabolism can be modulated by diets high in Ω3 PUFA, as the displacement of Ω6-derived AA from cell membrane phospholipids by Ω3 PUFAs interferes with function of Δ6-desaturase, lipoxygenase and COX [13,25,26]. Hence, the present men with elevated Ω6/Ω3 PUFA and diminished Ω6/Ω3 PUFA ratios would be expected to be at a greater risk of prostate cancer. The finding of decreased erythrocyte membrane levels of EPA, a potent inhibitor of COX activity, in men with elevated PSA levels is consistent with this theory.

Ω6 PUFAs are rapidly metabolized by prostate cancer cells to produce prostaglandins and leukotrienes, thereby causing malignant prostate tissue to have a lower AA and a higher prostaglandin E2 (Ω6 PUFA-derived) content than normal tissue [11,27]. Studies also show that human prostate cancer cells lack feedback regulation for the low-density lipoprotein receptor, thereby increasing the uninhibited delivery of AA and linoleic acids into the cell [28]. Consequently, erythrocyte membrane levels of AA and linoleic acid would be expected to decrease in the early stages of prostate carcinogenesis (PSA 2–10 ng/mL) as they are rapidly being incorporated into the growing tumour cells, which is supported by the present data.

A reduction in erythrocyte membrane levels of linoleic acid and AA may also facilitate androgen-mediated development of prostate cancer and would be expected in men in the early stages of prostate disease (PSA < 10 ng/mL). The mechanism of action is based on an increase in the bioavailability of testosterone. Unbound linoleic acid can increase the fraction of testosterone made available to cells by inhibiting its binding to albumin and sex hormone-binding globulin [29]. The increased level of bioavailable testosterone can stimulate the growth of androgen-dependent prostate tumours and, subsequently, result in an elevation of PSA levels.

The present data suggest that there is a relationship between dietary PUFA intake and prostate cancer risk. Further studies are needed to identify putative dietary staples that may contribute to the high PSA and erythrocyte membrane Ω6 PUFA levels. Ackee fruit, the dry weight of which is mostly Ω6 PUFAs, may be one such factor. Dietary studies are currently underway using a dietary questionnaire designed for the Jamaican population [30].

Relationships between diet, prostate tissue PUFA levels and prostate cancer remain unclear. In the present study we used serum PSA levels to represent the risk of prostate cancer. We are currently expanding our studies by incorporating prostate needle biopsies from Jamaican men to study erythrocyte membrane fatty-acid levels in benign and malignant prostate tissue. Tissue levels of Ω3 and Ω6 PUFAs will also be evaluated. We are also broadening our studies to assess whether correlations exist between diet, erythrocyte membrane Ω6 and Ω3 PUFA and serum PSA levels, and prostate cancer in rural Chinese men, whose risk of prostate cancer is low.

In conclusion, serum PSA levels are dramatically increased in Jamaican men, paralleling their increased risk of developing prostate cancer. There is a direct correlation between increased PSA levels and increased Ω6/Ω3 PUFA ratios, and an indirect correlation between increased PSA levels and Ω3/Ω6 PUFA ratios. If dietary associations can be established these data may provide a possible environmental explanation for the difference in incidence rates in prostate cancer between African-Americans and Jamaican men of African descent.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

This study was supported by the New York Academy of Medicine and University of Chicago Section of Urology RESCUE fund: a philanthropic endowment established to promote research and educational endeavours. All authors contributed significantly to the preparation of this manuscript. There are no disclosures of conflict of interest.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

Appendix

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  10. Appendix

The 32 different fatty acids examined in the 107 Jamaican men.

Polyunsaturated Ω3

α-Linolenic (18:3n3)

Eicosapentaenoic(20:5n3)

Docosapentanaenoic (22:5n3)

Docosahexaenoic (22:6n3)

Polyunsaturated Ω-6

Linoleic (18:2n6)

γ-Linolenic (18:2n6)

Eicosadienic (20:2n6)

Dihomo-≈γ Linolenic (20:3n6)

Arachidonic (20:4n6)

Docosadienoic (22:2n6)

Docosatretraenic (22:4n6)

Monounsaturated

Vaccenic (18:1n7)

Myristoleic (14:1n5)

Palmitoleic (16:1n7)

Oleic (18:1n9)

11-Eisosenoic (20:1n9)

Erusicic (22:1n9)

Nervonic (24:1n9)

Saturated

Capric (10:0)

Lauric (12:0)

Myristic (14:0)

Palmitic (16:0)

Stearic (18:0)

Arachidic (20:0)

Behenic (22:0)

Lignoceric (24:0)

Hexacosanoic (26:0)

Odd chain

Pentadecanoic (15:0)

Heptadecanoic (17:0)

Nonadecanoic (19:0)

Heneicosanoic (21:0)

Tricosanoic (23:0)