Cancer Causes Control (2012) 23:23–33 DOI 10.1007/s10552-011-9850-4
ORIGINAL PAPER
Associations of whole-blood fatty acids and dietary intakes with prostate cancer in Jamaica Maria D. Jackson • Susan P. Walker • Candace M. Simpson-Smith Carole M. Lindsay • Garret Smith • Norma McFarlane-Anderson • Franklyn I. Bennett • Kathleen C. M. Coard • William D. Aiken • Trevor Tulloch • Tomlin J. Paul • Robert L. Wan
•
Received: 2 July 2010 / Accepted: 27 September 2011 / Published online: 9 October 2011 Ó Springer Science+Business Media B.V. 2011
Abstract Objective To investigate the association of whole-blood fatty acids and reported intakes of fats with risk of prostate cancer (PCa). Design Case–control study of 209 men 40–80 years old with newly diagnosed, histologically confirmed prostate cancer and 226 cancer-free men attending the same urology clinics. Whole-blood fatty acid composition (mol%) was measured by gas chromatography and diet assessed by food frequency questionnaire. Results High whole-blood oleic acid composition (tertile 3 vs. tertile 1: OR, 0.37; CI, 0.14–0.0.98) and moderate palmitic acid proportions (tertile 2: OR, 0.29; CI, 0.12–0.70) (tertile 3: OR, 0.53; CI, 0.19–1.54) were
inversely related to risk of PCa, whereas men with high linolenic acid proportions were at increased likelihood of PCa (tertile 3 vs. tertile 1: OR, 2.06; 1.29–3.27). Blood myristic, stearic and palmitoleic acids were not associated with PCa. Higher intakes of dietary MUFA were inversely related to prostate cancer (tertile 3 vs. tertile 1: OR, 0.39; CI 0.16–0.92). The principal source of dietary MUFA was avocado intake. Dietary intakes of other fats were not associated with PCa. Conclusions Whole-blood and dietary MUFA reduced the risk of prostate cancer. The association may be related to avocado intakes. High blood linolenic acid was directly related to prostate cancer. These associations warrant further investigation.
M. D. Jackson (&) T. J. Paul Department of Community Health and Psychiatry, University of the West Indies, Mona, Kingston, Jamaica e-mail:
[email protected]
F. I. Bennett K. C. M. Coard Department of Pathology, University of the West Indies, Mona, Kingston, Jamaica e-mail:
[email protected]
T. J. Paul e-mail:
[email protected]
K. C. M. Coard e-mail:
[email protected]
S. P. Walker C. M. Simpson-Smith Tropical Medicine Research Institute, University of the West Indies, Mona, Kingston, Jamaica e-mail:
[email protected]
W. D. Aiken T. Tulloch Department of Surgery, Radiology, Anaesthesia, and Intensive Care, University of the West Indies, Mona, Kingston, Jamaica e-mail:
[email protected]
C. M. Simpson-Smith e-mail:
[email protected]
T. Tulloch e-mail:
[email protected]
C. M. Lindsay G. Smith N. McFarlane-Anderson Department of Basic Medical Sciences, University of the West Indies, Mona, Kingston, Jamaica e-mail:
[email protected]
R. L. Wan Kingston Public Hospital, Kingston, Jamaica e-mail:
[email protected]
G. Smith e-mail:
[email protected] N. McFarlane-Anderson e-mail:
[email protected]
123
24
Keywords Prostate cancer Whole-blood fatty acids Dietary fat intake Avocado intake
Introduction Globally, the incidence of prostate cancer ranges widely [1]. Evidence of a higher incidence and mortality rates in Western societies, particularly among men of African descent, than in Asian countries suggests that genetic, environmental and behavioral factors (such as diet) play an important role in the development of prostate cancer [2]. Studies suggest that diet is a part of prostate cancer risk that is modifiable [3]. In Jamaica, morbidity and mortality data over the last three decades highlight prostate cancer as the most commonly diagnosed solid malignancy/neoplasm in Jamaican men. Recent information on cancer-related deaths showed prostate cancer to be the leading cancer site among men (36.6%) and that prostate cancer increased from 65.5 to 78.1 per 100,000 per year, 2003–2007 [4]. Greenlee et al. [5] report that men of African descent, when compared to Caucasians and Asian races, show highest incidence and mortality from the disease. Our data rank prostate cancer incidence in Jamaica lower than men in Europe (86.7/ 100,000) [6], blacks in the USA (101/100,000) but substantially higher than in men in Asia (2–10/100,000) [7]. Although dietary fat has important structural and physiological functions [8], its association as a risk factor for prostate cancer has been suggested [9, 10]. The World Cancer Research Fund and American Institute for Cancer Research Report (WCRF/AICR) (2007) indicated that the results of investigations of associations between fat and prostate cancer were inconclusive [11]. However, the metaanalyses were conducted on predominantly white men as few studies have been conducted in blacks. Biomarkers represent proxies for diet and have provided an alternative as indicators of dietary exposure [12, 13]. Biological measurements in blood and urine have been used as objective indices to reflect dietary intake without the limitations associated with reported intake [14]. In this report, we investigated the relationship of whole-blood fatty acids and dietary fats for their association with prostate cancer.
Materials and methods Cases and controls were recruited from men attending urology clinics at the 2 main tertiary hospitals and from private practitioners in the Kingston Metropolitan area in
123
Cancer Causes Control (2012) 23:23–33
Jamaica. The hospitals receive referrals from primary care clinics, hospitals and private practitioners, island-wide. Regardless of presentation, most patients over 40 years presenting to the urology clinic were asked to do a prostatespecific antigen (PSA) test. Exceptions were those patients referred for non-prostate-related conditions who in addition had comorbid illnesses which would have compromised life expectancy. Study population Subjects were recruited from men attending for the first time at urology clinics, March 2005–July 2007, and comprised 243 men with prostate cancer and 273 controls. To detect a twofold increase in risk (OR = 2) with 90% power and at the 5% significance level, the sample size required is 200 cases and 200 controls. Calculations were based on total prostate cancer. The study had sufficient power to detect a twofold increase in risk (tertile 3 vs. tertile 1: OR = 2) for total prostate cancer. The study was approved by the Ethics Committee of the University of the West Indies, and subjects gave written informed consent to participate in the study. Selection of cases and controls Cases Cases were men 40–80 years old with histological confirmation of prostate cancer. They were newly diagnosed (incident cases) and were enrolled consecutively. Cases were classified by tumor stage according to the TNM classification, and histological grade was reported using the Gleason system [15]. Controls Controls were subjects attending the same urology outpatient clinics. PSA testing was done in the same laboratory for cases and controls. Similar to cases, controls were enrolled consecutively in the study. Control subjects presented with diagnostic conditions that were primarily related to lower urinary tract symptoms secondary to BPH and urinary stones. To a lesser extent, patients were referred for the investigation of and treatment for other suspected genitourinary malignancies or disorders (e.g., male-factor infertility); there were few reports of erectile dysfunction or decreased libido as the primary complaint. Men were assigned as controls based on a normal digital rectal examination (DRE) and PSA concentration as follows:
Cancer Causes Control (2012) 23:23–33
criteria:
(a) total PSA \ 2.0 l/L or free: total PSA [ 0.25 (n = 174) (b) total PSA 2.0–4.0 l/L or free : total PSA 0.15–0.24 (n = 99)
Exclusion criteria The following persons were excluded from the study: men (1) with previous prostate surgery, (2) on hormonal treatment or (3) taking finasteride. Data collection After obtaining informed consent, all first-time clinic attendees were interviewed before pathological diagnosis was obtained. A standardized questionnaire was administered, and participants were asked to donate an additional 10 mL blood sample for biological measurements. Medical charts and pathology reports were examined to ensure that controls did not have a prior history of cancer. Measurements Dietary assessment Dietary intakes were assessed using a previously validated food frequency questionnaire (FFQ) [16] that was expanded to assess diet and cancer in the Jamaican population [17]. In addition to frequency of consumption for each food item, subjects were asked to estimate the portion size usually consumed by using food models, commonly used household utensils, measuring cups and a measuring tape. The FFQ was interviewer-administered.
25
acid methyl esters. Fatty acid analysis was conducted at the Department of Basic Medical Sciences of the University of the West Indies. Fatty acid concentrations in each sample were expressed as the percentage of total fatty acids. Quality control samples were embedded randomly in each box of study samples. Coefficients of variation for saturated fatty acids [myristic (C14:0), palmitic (C16:0) and stearic (C18:0)] were 3.2, 1.1 and 4.0%, respectively. For monounsaturated fatty acids, the CVs were 5.3 and 3.5% for palmitoleic (C16:1) and oleic (C18:1), respectively. The CVs for polyunsaturated fatty acids were 5.3 and 1.5% for linolenic (18:n-3) and linoleic (18:2n-6). Laboratory personnel were blinded to the status of the samples, and samples from cases and controls were analyzed together to reduce the effect of inter-assay variability. Anthropometry Body weight (without shoes) was measured in clothing to the nearest 0.1 kg on a bathroom-type digital scale. Height was measured without shoes on a floor-standing stadiometer fitted with a spirit level, to the nearest 0.1 centimeters. Body mass index (BMI) was calculated, and the World Health Organization’s classification of BMI was used to determine overweight (BMI [ 25.00 kg/m2) and obesity (BMI C 30.00 kg/m2). Socio-demographic and health questionnaire Information on demographic and socioeconomic factors, medical history, lifestyle (cigarette smoking and alcohol use) and sexual behavior was obtained by questionnaire. Statistical analysis
Whole-blood fatty acids Venous blood samples were collected in the mornings and were placed on ice packs in a cooler before being taken to the laboratory in the afternoons. Then, 5.0-mL aliquots of whole blood were taken and the remaining blood samples were centrifuged at 8009g for 10 min and separated into aliquots of plasma, and buffy coat fractions and aliquots were stored. Total lipids were extracted from whole-blood samples (200 ll) with chloroform/methanol (2:1 v/v) containing 0.005% antioxidant, butylated hydroxytoluene (Sigma) and tricosonoic acid as internal standard. Fatty acids were transmethylated with methanol and sulfuric acid as described by Zock et al. [18]. After esterification, samples were evaporated and the fatty acid methyl esters were re-dissolved in hexane and analysis carried out on a 7890 Gas Chromatograph (Agilent, DE, USA). An HP-88 (100 m 9 0.25 mm 9 0.2 lm) capillary column was used for the separation of the fatty
Chi-square statistics or Fisher’s exact test was used to examine differences between cases and controls for categorical variables, and the Mann–Whitney U, t test and analysis of variance were used for continuous data. Fatty acids were presented as the percentage of total fatty acid in whole blood. Dietary intake variables were expressed as grams per day. Total fat was also expressed as percentage of total energy. Subjects were classified into tertiles according to blood composition of fatty acids and dietary intakes based on the distribution of the controls. For myristic acid and linolenic acid, the number of controls with no detectable concentrations exceeded the number in each tertile; hence, tertiles were grouped as follows: not detected (reference group), and subjects with detectable concentrations were divided in halves. Adjusted binary logistic regression analyses were performed to test whether dietary or whole-blood fatty acids examined
123
26
Cancer Causes Control (2012) 23:23–33
Table 1 Characteristics of cases and controls
Controls (n = 226)
Cases (n = 209)
p
62.6 ± 10.5
67.5 ± 7.9
0.0001
Primary or less
80.1 (181)
89.2 (183)
0.011
Secondary
12.8 (29)
5.9 (14)
Tertiary
7.1 (16)
4.9 (12)
Waist circumference: mean ± SD
84.4 ± 12.7
88.0 ± 12.2
0.590
Body mass index (BMI) kg/m2
25.1 ± 4.3
25.1 ± 4.6
0.852
B24.99
52.2 (118)
54.7 (106)
0.576
25.00–29.99
36.3 (82)
32.4 (62)
C30.00
11.5 (26)
12.9 (25)
–
43.1 (90)
–
54.5 (114)
–
2.4 (5)
Socio-demographic Age: mean ± SD (years) Education: (n)
Anthropometry
BMI categories
Medical/behavioral Cases only: % (n) Low-grade prostate cancera High-grade prostate cancer
b
Unknown PSA prostate-specific antigen, DRE digital rectal examination a
Low-grade cancer indicates Gleason score of \7
b
High-grade cancer indicates Gleason score of C7
PSA: lg/L median (25th, 75th percentile)
1.6 (0.8, 3.5)
25.8 (12.0, 100.1)
0.0001
Previous PSA screens : % (n)
59.0 (133)
88.6 (182)
0.0001
Previous DRE: % (n)
75.0 (170)
85.9 (180)
0.005
Family history of prostate cancer: % (n)
11.1 (25)
16.3 (34)
0.113
Current smoker: % (n)
16.8 (38)
13.2 (28)
0.508
Supplement use: % (n)
29.5 (67)
23.5 (49)
0.185
were associated with prostate cancer. In all multivariate models, adjustments were made for potential confounders and included age, education (as a measure of socioeconomic status), family history of prostate cancer in first-degree relatives, smoking and BMI; for dietary intake analyses, alcohol intake and total energy intake were added to the model for consideration of potential confounding effects. To test for linear trends across tertiles of blood composition of fatty acid and dietary intake variables, the median of the variable within each category was determined and used as a continuous variable. Multiple linear regression analyses were used to ascertain the extent to which dietary intake of avocado and MUFA was associated with whole-blood oleic acid. All analyses were performed using the Statistical Package for Social Sciences (SPSS) version 17. Statistical significance was achieved when p \ 0.05.
Results Characteristics of cases and controls The characteristics of the 209 men with prostate cancer and 226 controls are displayed in Table 1. Controls were on
123
average 5 years younger than cases and were more likely to have secondary or higher education. Men diagnosed with prostate cancer had similar mean waist circumference and BMI to the controls. Just over one-half of prostate cancer cases were of high grade. Cases and controls differed in median PSA levels (cases, 25.8 lg/L; controls, 1.6 lg/L). Significantly more men with prostate cancer than controls reported that they were previously screened for the disease by the PSA or the DRE. Reports of a family history of the disease were low in both groups. Smoking and supplement use were similar in cases and controls. Associations of whole-blood fatty acids The mean percentages of fatty acids in whole blood for cases and controls are given in Table 2. Palmitic fatty acid was significantly higher among controls compared with cases. On average, the percentages of all other fatty acids were not significantly different between cases and controls. Table 2 also shows the multivariable-adjusted associations of percent whole-blood fatty acids with total prostate cancer. Moderate palmitic fatty acid compositions were negatively associated with prostate cancer (tertile 2 vs. tertile 1: OR, 0.29; CI, 0.12–0.70) but showed no
Cancer Causes Control (2012) 23:23–33
27
Table 2 Odds ratios (OR) and 95% CI for association of whole-blood fatty acid with prostate cancer Controls (mean ± SD) mol%
Cases (mean ± SD) mol%
Tertile of fatty acids
ptrend
1 (reference)
2 OR (CI)
3 OR (CI)
0.0 133/124
0.01–1.00 36/51
1.01–5.93 40/51
1.0
0.99 (0.23, 4.29)
2.58 (0.65, 6.26)
Saturated fatty acids Myristic acida
1.24 ± 0.98
1.19 ± 0.59
Range (mol%) Cases/controls (n) OR (CI)c Palmitic acid
35.72 ± 4.78
33.89 ± 4.73
Range (mol%)
5.31–34.94
34.95–39.58
39.59–58.75
Cases/controls (n)
64/75
61/76
84/75
1.0
0.29 (0.12, 0.70)
0.53 (0.19, 1.54)
Range (mol%)
10.54–16.25
16.26–19.30
19.31–76.47
Cases/controls (n)
64/75
62/76
84/75
OR (CI)c
1.0
0.94 (0.56, 1.58)
1.48 (0.89, 2.44)
Range (mol%)
0.37 -18.42
18.43–24.24
24.25–40.53
Cases/controls (n)
69/75
79/76
61/76
1.0
0.52 (0.20, 1.34)
0.37 (0.14, 0.98)
OR (CI)c Stearic acid
17.21 ± 3.41
0.177
b
0.061
17.81 ± 3.10
0.115
Monounsaturated fatty acids Oleic acid
22.12 ± 5.45
OR (CI)c Palmitoleic acid
1.51 ± 1.00
21.56 ± 5.43
0.049
1.60 ± 1.32
Range (mol%)
0.06–0.94
0.95–1.57
1.58–5.77
Cases/controls (n)
60/75
56/76
93/75
OR (CI)c
1.0
0.86 (0.33, 2.27)
0.81 (0.32, 2.04)
0.266
Polyunsaturated fatty acids Linoleic acid
14.08 ± 4.92
13.93 ± 5.29
Range (mol%)
4.25–11.55
11.56–17.05
17.06–38.43
Cases/controls (n)
59/75
55/76
85/75
OR (CI)c
1.0
1.07 (0.64, 1.80)
1.37 (0.83, 2.25)
a-Linolenic acid
a
0.12 ± 0.35
0.213
0.04 ± 0.03
Range (mol%)
0.0
0.014–0.029
0.030 -1.92
Cases/controls (n)
108/120
11/53
90/53
OR (CI)c
1.0
0.23 (0.11–0.49)
2.06 (1.29–3.27)
0.043
a
Analyte was not detected (i.e., zero value or below detection limit of the assay); myristic acid, 133 cases and 124 controls; linolenic acid, 108 cases and 120 controls
b
p \ 0.05; cases significantly different from controls
c
Adjusted for age, family history of prostate cancer, education, smoking and body mass index
relationship with disease at the highest tertile (OR, 0.53; CI, 0.19–1.54). There was an inverse association between the percentage of oleic acid and prostate cancer; the adjusted OR for men in the highest tertile of oleic acid was 0.37 (CI, 0.14, 0.98; ptrend = 0.049) compared with men in the lowest tertile of oleic acid. Examinations of linolenic acid in which men with undetected levels of the metabolite served as the reference group showed that men in the second tertile had decreased odds for total prostate cancer (tertile 2 vs. tertile 1: OR, 0.14; CI, 0.04–0.47) and suggested increased risk at higher levels (tertile 3 vs. tertile 1:
OR, 2.06; CI, 1.29–3.27) (ptrend, 0.010). The remaining fatty acids were not associated with prostate cancer risk. Dietary intakes In this report, we excluded from analyses subjects who provided incomplete information on the FFQ (n = 45) or whose reported intakes were outside the range of 800–5,000 kcal (n = 36). Comparison of the demographic characteristics of men with missing data with the sample providing dietary information showed that they were
123
28
Cancer Causes Control (2012) 23:23–33
Table 3 Consumption (g/day) of selected foods and food groups and prostate cancer risk according to tertiles of intakes Controls (g/day) Mean ± SD (median) Unprocessed meata
Processed meata
Seafood
Dairy
Fruits
Dark green leafy and yellow vegetables
Other vegetables*
Ackee (Blighia sapida)
Avocado**
Beans and legumes
92.0 ± 64.9 (79.6)
25.9 ± 23.7 (19.0)
58.5 ± 45.9 (46.5)
63.6 ± 109.5 (17.1)
252 ± 171 (121)
98.6 ± 40.5 (62.9)
79.1 ± 38.4 (56.0)
18.2 ± 22.5 (10.9)
56.2 ± 70.8 (30.2)
99.0 ± 127.7 (59.7)
Cases (g/day) Mean ± SD (median)
Odds ratio (95% CI)
93.1 ± 74.7 (83.5)
29.2 ± 27.7 (21.0)
67.1 ± 53.9 (56.0)
54.1 ± 81.3 (22.0)
246 ± 193 (180)
93.2 ± 41.0 (58.9)
59.3 ± 41.9 (43.0)
19.5 ± 25.3 (10.9)
36.9 ± 40.5 (25.6)
101.6 ± 98.9 (71.3)
Tertile 1 (low)
Tertile 2
Tertile 3 (high)
Cases/ controls
60/80
65/82
68/80
Adjustedb
1.00 (ref)
1.12 (0.65, 1.94)
1.24 (0.71, 2.19)
Cases/ controls
57/81
60/81
76/82
Adjustedb
1.00 (ref)
1.01 (0.59, 1.74)
1.21 (0.70, 2.10)
Cases/ controls
60/80
57/84
76/80
Adjustedb
1.00 (ref)
0.81 (0.47, 1.40)
1.39 (0.79, 2.41)
Cases/ controls
51/81
72/83
66/80
Adjustedb
1.00 (ref)
1.29 (0.76, 2.21)
1.06 (0.60, 1.89)
Cases/ controls
65/78
67/80
61/81
Adjustedb
1.00 (ref)
0.89 (0.51, 1.57)
0.86 (0.49, 1.53)
Cases/ controls
69/82
76/82
48/80
Adjustedb
1.00 (ref)
0.83 (0.48, 1.44)
0.65 (0.37, 1.15)
Cases/ controls
58/83
83/81
52/80
Adjustedb
1.00 (ref)
1.50 (0.86, 2.60)
0.87 (0.50, 1.53)
Cases/ controls
58/83
64/79
71/82
Adjustedb
1.00 (ref)
1.02 (0.58, 1.79)
1.31 (0.76, 2.27)
Cases/ controls
79/84
60/70
53/89
Adjustedb
1.00 (ref)
0.90 (0.52, 1.55)
0.41 (CI, 0.21–0.78)
Cases/ controls
62/81
60/81
71/82
Adjustedb
1.00 (ref)
0.99 (0.57, 1.73)
1.30 (0.75, 2.24)
* p \ 0.05; ** p \ 0.0001: cases significantly different from controls a
Unprocessed meat: poultry, beef, mutton, pork, liver. Processed meat: corned beef, ham/, bacon, sausage, pickled mackerel, salted cod fish, pig’s tail
b
Adjusted for age, family history of prostate cancer, education, smoking, body mass index and total energy
similar. A total of 193 cases and 244 controls were used for the analyses of diet. Specific foods or food groups hypothesized to be associated with prostate cancer were examined in Table 3. With the exception of intakes of other vegetables and avocado, cases reported consumption of
123
similar amounts of foods or food groups as controls. Men with prostate cancer reported significantly lower consumption of other vegetables that included tomato, cabbage, string beans and cauliflower when compared to controls. Cases also had lower mean intakes of avocados than controls.
Cancer Causes Control (2012) 23:23–33
29
Table 3 also shows odds ratios (CI) for association with prostate cancer according to tertiles of the food and food groups adjusted for potential confounders. With the exception of the consumption of avocado, which was inversely related, no significant association was found between any individual food or food groups and risk of prostate cancer. Significant decrease in risk was observed for unadjusted avocado intake at the highest tertile (tertile 3 vs. 1: OR, 0.61; CI, 0.38–0.97) only (data not shown), and this remained significant after controlling for potential confounders (adjusted tertile 3 vs. 1: OR, 0.41; CI, 0.21–0.78). Men who ate C60.00 g/day of avocado (highest tertile), when compared with men with intakes of 12.8 g/day or less (lowest tertile), had reduced risk of prostate cancer. Table 4 displays intakes and the associations of energy, macronutrients, percentage of energy from fat and fatty acid intakes with prostate cancer risk. Mean intakes of energy, total fat, fat percent energy and the fatty acids were similar among cases and non-cases, except for monounsaturated fat that was significantly higher among controls than cases. Prostate cancer was not related to total energy intake, with or without adjustment for potential confounders. Risk was unrelated to intakes of fat as total fat or
as a percentage of energy. Odds ratios for associations of fatty acids show that higher intakes of MUFA were protective of prostate cancer. Comparing the highest to the lowest tertile of monounsaturated fat, the adjusted odds ratio was 0.39 (95% CI, 0.16–0.92), suggesting a 61% reduction in cancer risk with intakes of 25.9 g/day or higher. There were no associations with dietary intakes of saturated and polyunsaturated fats or with linoleic and linolenic acids. The inverse relationship between monounsaturated fat intakes and prostate cancer was further examined by investigating associations of foods contributing to monounsaturated fat intakes. The major sources of monounsaturated fat included avocado (18%), peanuts (8%) and baked/roasted chicken (8%). In view of our findings of associations of whole-blood oleic acid and avocado intake with prostate cancer, we further evaluated their independent contribution to prostate cancer. With additional adjustments for whole-blood oleic acid and avocado intake, at the highest tertile, both oleic acid (tertile 3 vs. 1: OR, 0.77; CI, 0.45–0.98) (ptrend = 0.049) and avocado intake (tertile 3 vs. 1: OR, 0.50; CI, 0.28–0.89) (ptrend = 0.020) were inversely related to prostate cancer (data not shown).
Table 4 Odds ratios for associations of dietary energy, macronutrients and fat intake with risk of prostate cancer
Total energy (kcal) Protein (g)
Controls (g/day) Mean ± SD
Cases (g/day) Mean ± SD
3,076 ± 1,003
2,944 ± 980
116.6 ± 49.5
113.9 ± 46.4
Odds ratio (95% CI) Tertile 1 (low)
Tertile 2
Tertile 3 (high) 54/84
Cases/controls
71/80
68/80
Adjusteda
1.00 (ref)
1.13 (0.62, 2.06)
0.67 (0.36, 1.29)
Cases/controls
71/80
57/82
67/83
Adjusteda
1.00 (ref)
1.01 (0.54, 1.89)
1.26 (0.58, 2.73)
Carbohydrate
430.2 ± 190.9
410.5 ± 195.0
Cases/controls Adjusteda
1.00 (ref)
0.79 (0.38, 1.58)
0.51 (0.17, 1.50)
Total fat (g)
108.0 ± 31.5
103.4 ± 28.6
Cases/controls
74/80
66/82
53/82
Fat (% energy) Saturated fat (g) Monounsaturated fat (g)* Poly-unsaturated fat Linoleic acid (n-6) Linolenic acid (n-3)
33.0 ± 5.6 40.7 ± 14.4 32.6 ± 11.2 26.3 ± 6.7 7.69 ± 9.83 0.97 ± 0.38
32.5 ± 5.1 39.8 ± 13.9 30.0 ± 9.2 25.9 ± 7.4 8.84 ± 11.50 0.99 ± 0.40
Adjusteda
1.00 (ref)
0.96 (0.47, 1.96)
0.48 (0.18, 1.24)
Cases/controls
72/82
65/80
56/81
Adjusteda
1.00 (ref)
0.85 (0.46, 1.55)
0.61 (0.33, 1.15)
Cases/controls
66/75
56/85
70/83
Adjusteda
1.00 (ref)
0.85 (0.42, 1.73)
1.11 (0.48, 2.53)
Cases/controls
72/80
76/82
45/82
Adjusteda
1.00 (ref)
0.88 (0.45, 1.74)
0.39 (0.16, 0.92) 62/91
Cases/controls
60/79
70/73
Adjusteda
1.00 (ref)
1.58 (0.77, 3.22)
1.15 (0.49, 2.67)
Cases/controls
61/80
55/81
77/83
Adjusteda
1.00 (ref)
0.55 (0.28, 1.12)
1.33 (0.68, 2.58)
Cases/controls
60/78
67/76
63/82
Adjusteda
1.00 (ref)
0.97 (0.46, 2.07)
0.93 (0.35, 2.44)
* p \ 0.05; cases significantly different from controls a
Adjusted for age, family history of prostate cancer, education, smoking, body mass index, alcohol intake and total energy
123
30
Multiple linear regression analyses for PSA concentrations and avocado intake showed a non-significant inverse relationship (B ± SE, -1.16 ± 0.64; p = 0.070). Neither whole-blood oleic proportion (B ± SE, -1.91 ± 6.16; p = 0.756) or dietary MUFA (B ± SE, -3.81 ± 4.78; p = 0.426) were associated with PSA concentrations.
Discussion In the present study, our results suggested that higher proportions of whole-blood oleic acid and higher dietary intakes of monounsaturated fat were associated with decreased risk of prostate cancer. Moderate proportions of blood palmitic acid were inversely related to prostate cancer, whereas higher proportions of linolenic acid increased the likelihood of the disease. No other associations were found for other whole-blood or dietary fatty acids with prostate cancer risk. Studies of diet have been regarded to be less informative than those based on biological specimens because of their reliance on self-reports. Measurements of fatty acid composition in blood can provide an alternative to dietary assessment in the measurement of exposure to diet–disease studies. The associations between prostate cancer and fatty acid composition were measured in whole blood. Erythrocytes are deemed to be more appropriate than plasma or serum as a biological specimen to reflect long-term intake as the half-life of erythrocytes is 120 days; on the other hand, plasma is a reflection of short-term or recent fat intake [13]. We observed an inverse association between circulating oleic acid but not palmitoleic acid (monounsaturated fats) and the risk of prostate cancer. Our results are in contrast to a nested case–control study of pre-diagnostic fatty acids in serum, and subsequent risk of prostate cancer in Norwegian men showed an increased risk of prostate cancer with increasing quartiles of palmitoleic acid [19]. The AlphaTocopherol, Beta-Carotene (ATBC) Cancer Prevention Study of serum showed no association between circulating MUFA and the risk of prostate cancer [20]. Neither the European Prospective Investigation of Cancer’s (EPIC) investigation of plasma fatty acids [21] nor Park et al.’s [22] investigation of erythrocyte membrane MUFA and risk of prostate cancer in the multiethnic cohort reported a relationship with total prostate cancer or grade of disease. The comparison of our results with these studies is difficult because circulating MUFA was measured in whole blood and not in different fractions such as serum or plasma. It is possible therefore that whole blood MUFA used in this study to examine its relationship with prostate cancer provides a more comprehensive picture in relation to MUFA intakes when compared to other blood specimens.
123
Cancer Causes Control (2012) 23:23–33
It is also possible that circulating MUFA may be a marker for avocado intakes that are associated with several bioactive substances such as carotenoids, which in combination with other diet-derived phytochemicals may contribute to cancer risk reduction [23]. Avocado intake, the largest single source of dietary MUFA in this population, was also associated with reduced risk of prostate cancer, but the relationship between MUFA/avocado and prostate cancer may be more complex. Oleic acid accounts for approximately 50% of the monounsaturated fatty acid content of avocado [24]. However, avocado also contains phytochemicals that act as antioxidants exhibiting chemopreventive properties [25]. In this study, the association between avocado intake and prostate cancer risk remained significant with adjustment for oleic acid, thereby suggesting an independent association with the disease. To our knowledge, our study is the first to report an association between avocado consumption and prostate cancer. Studies of dietary MUFA and prostate cancer are inconsistent with some investigations showing a relationship [26–29] while others report no association [30–33]. Whereas reports of MUFA intakes from animal food sources and prostate cancer risk have shown no association [34, 35] or were positively related [27, 29], investigations of vegetable oils rich in MUFA have pointed to progressive reduction in prostate cancer [36]. For example, the Mediterranean diet, which features high consumption of olive oil, is rich in MUFA and has been recognized to be protective of cardiovascular disease and cancer. These effects have been ascribed to oleic acid, the main MUFA in olive oil, along with minor compounds such as squalene and phenolic compounds [37]. We found no previous studies that have specifically investigated the association between prostate cancer risk and avocado consumption. Oleic acid in avocado is less susceptible to oxidation than polyunsaturated fatty acids and thus contributes to its antioxidant properties [38]. Bioactive substances such as lutein, related carotenoids and vitamin E in avocado are recognized in modifying cancer risks [25]. Extracts of avocado containing carotenoids and tocopherols have been shown in vitro to induce cell cycle arrest, inhibit growth and induce apoptosis in precancerous and cancer cells [23]. More recent investigation of avocado showed that phytochemical extracts target multiple signaling pathways and increase intracellular reactive oxygen leading to apoptosis [25]. It is suggested that the monounsaturated fat found in avocado may facilitate the absorption of bioactive carotenoids into the bloodstream where, along with other diet-derived phytochemicals, they may contribute to cancer risk reduction [23]. The finding of an association between avocado and reduced risk of prostate cancer needs further investigation.
Cancer Causes Control (2012) 23:23–33
Whole-blood palmitic acid was the only saturated fatty acid proportion associated with prostate cancer; reported intakes of the saturated fat had no significant association with the disease. Palmitic acid was inversely related to prostate cancer; however, our study did not demonstrate a clear linear gradient of risk across tertiles of palmitic acid proportion as men in the second tertile, relative to the lowest tertile, were at reduced risk of prostate cancer. Three prospective studies on saturated fatty acid proportion and prostate cancer reported a positive association between palmitic acid and risk of prostate cancer [19, 21, 22]. Harvei et al.’s [19] study of blood donors in Norway showed that serum palmitic acid was positively associated with prostate cancer (OR, 2.3; CI, 1.1–4.7). The European Prospective Investigation into Cancer and Nutrition (EPIC) nested case–control study reported that a higher proportion of palmitic acid was associated with a greater risk of overall prostate cancer as well as localized and low-grade disease [21]. In the multiethnic cohort, Japanese Americans were associated with a 2.6-fold increased risk of total prostate cancer [22]. Other studies have found no association of palmitic acid with prostate cancer [20, 39]. Dietary sources of palmitic acid include meat, fish, milk and cheese; we found no association between the consumption of these foods and prostate cancer risk. Palmitic acid may also be synthesized from other fatty acids; it is the major fatty acid by de novo lipogenesis from acetyl CoA and malonyl CoA by the enzyme fatty acid synthase [40]. Our data do not provide strong evidence for a protective role of palmitic acid in prostate cancer, and the inverse relationship observed with the disease may be due to chance. Myristic and stearic fatty acids were not associated with prostate cancer. The absence of a relationship with myristic acid and prostate cancer in our population was not surprising. Myristic acid is largely derived from dairy products, and relatively small amounts are produced by de novo lipogenesis by fatty acid synthase [41]. In Jamaica, the consumption of dairy is low and dietary intakes of dairy were not related to prostate cancer. In the EPIC study, there was a suggestion of increased risk of advanced and highgrade prostate cancer [21]. A positive association was found in male smokers in the ATBC study [20]; however, no association was found in another prospective study [19]. Of note, our estimates of risk of disease may be unstable in view of the small sample. In contrast to our study and others [19, 20, 39] that showed no association between stearic acid and prostate cancer risk, the EPIC study [21] reported that stearic acid decreased the risk of overall, localized and low-grade prostate cancer. Although not a consistent finding, the association of polyunsaturated fatty acids with prostate caner risk has been reported in several studies [19, 21, 22, 39, 42, 43]. Lower linolenic fatty acid composition in our study
31
suggested decreased risk while men with higher molecular percentage of the fatty acid were at increased risk of the disease; dietary intakes of the nutrient were not associated with prostate cancer. We found one study that showed a positive association of biological linolenic fatty acid with prostate cancer. Harvei et al. [19] reported that high linolenic acid fatty acid composition was associated with a twofold increase in risk of prostate cancer. These findings are contrary to the suggestions that linolenic acid suppresses tumor growth through apoptosis or programmed cell death [44]. Other studies found no significant relationship of the fatty acid with prostate cancer [19–22, 43, 45, 46]. In this study, whole-blood linolenic acid composition was on average lower than levels reported elsewhere [19–22, 39, 42, 45]. Low linolenic acid proportion suggested reduced risk of prostate cancer; however, the finding must be viewed with caution as few subjects were in the second tertile (cases/controls, 11/53); this negative association could have been due to chance. Simon et al.’s systematic review and meta-analysis revealed that high dietary intakes or tissue compositions of linolenic acid increased the risk of prostate cancer by 20% (RR: 1.20; CI, 1.01–1.43); however, after adjustment for publication bias, similar to our findings, there was no relationship between linolenic acid and prostate cancer (RR: 0.96; CI, 0.79–1.17) [47]. The absence of a relationship between linoleic fatty acid proportion and prostate cancer is in contrast to animal and in vitro studies which show that fatty acid promotes prostate cancer growth [48–50]. Similar to us, other investigations report null findings [19, 21], while others suggest increased [51, 52] or inverse associations [43] between linoleic fatty acid and prostate cancer. Similar to our study, results from prospective [30–32] and case–control studies [26, 53] showed no association between dietary intake of linoleic acid and prostate cancer risk. The study has potential limitations. In addition to the DRE, we used the PSA to screen for prostate cancer. It has been reported that the PSA has low specificity and low predictive value [54]; to improve the diagnostic specificity we used free: total PSA to identify controls. However, it is possible that men with prostate cancer may have been included among the controls. Ideally, controls would be biopsy-negative men, but this was not done in all participants; we used the PSA to classify cancer-free men. The Prostate Cancer Prevention Trial showed that among men with PSA B 2.0 ng/mL (aged 62–91 years), 9.5% of men were diagnosed with prostate cancer [55]. Thus, it is possible that differences between groups may not be easily identified. Men without prostate cancer and a low PSA may be different from men without the disease. For example, a larger proportion of obese men may be in the control group. However, as found in this and other studies [56–59],
123
32
Cancer Causes Control (2012) 23:23–33
a higher BMI is unlikely to influence PSA levels. Dietary intake information was obtained with a FFQ, which was validated against 24-h recalls collected over 1 year and showed good agreement. Nonetheless, this methodology is subject to random and systematic errors [14], which may attenuate the relationship between dietary fat intake and the risk of prostate cancer. We adjusted for energy intake and to BMI to partially control for measurement error related to dietary intake [14]. Our analyses included many comparisons and cannot exclude chance as the cause of significant findings. Our study is case–control in design; hence measurements were conducted at the time of diagnosis. and temporal ordering among the relationships observed cannot be established. Residual confounding is also possible due to unexamined confounders.
Conclusion Our results suggest that high whole-blood oleic acid proportion and high dietary monounsaturated fat intake were inversely related to prostate cancer. The likelihood of prostate cancer was decreased among men reporting higher intakes of avocado, the largest single source of MUFA in this population. These findings warrant further investigation. Acknowledgments This study was supported by the National Health Fund and the Planning Institute of Jamaica and University of the West Indies. The authors wish to thank the research nurses— Barbara Panton, Elsa Brown, Nicola Meeks-Aitken, Donnahae Rhoden-Salmon—and study participants for their support in the investigation. Conflict of interest disclose.
None of the authors had a conflict of interest to
References 1. Parkin DM, Bray F, Ferlay J, Pisani P (2005) Global cancer statistics, 2002. CA Cancer J Clin 55:74–108 2. Stamatiou K, Delakas D, Sofras F (2007) Mediterranean diet, monounsaturated: saturated fat ratio and low prostate cancer risk. A myth or a reality? Minerva Urol Nefrol 59:59–66 3. Chan JM, Gann PH, Giovannucci EL (2005) Role of diet in prostate cancer development and progression. J Clin Oncol 23:8152–8160 4. Gibson TN, Hanchard B, Waugh N, McNaughton D (2010) Agespecific incidence of cancer in Kingston and St. Andrew, Jamaica, 2003–2007. West Indian Med J 59:456–464 5. Greenlee RT, Murray T, Bolden S, Wingo PA (2000) Cancer statistics, 2000. CA Cancer J Clin 50:7–33 6. Ferlay J, Autier P, Boniol M, Heanue M, Colombet M, Boyle P (2007) Estimates of the cancer incidence and mortality in Europe in 2006. Ann Oncol 18:581–592 7. Anonymous (1997) Cancer incidence in five continents, vol VII. IARC Sci Publ, pp i–1240
123
8. Institute of Medicine (2005) Dietary reference intakes for energy, carbohydrate, fibre, fat, fatty acids, cholesterol, protein and amino acids. National Academies Press, Washington, DC 9. Kolonel LN, Hankin JH, Lee J, Chu SY, Nomura AM, Hinds MW (1981) Nutrient intakes in relation to cancer incidence in Hawaii. Br J Cancer 44:332–339 10. Hursting SD, Thornquist M, Henderson MM (1990) Types of dietary fat and the incidence of cancer at five sites. Prev Med 19:242–253 11. World Cancer Research Fund/American Institute for Cancer Research (2007) Food, nutrition, physical activity and the prevention of cancer: a global perspective. Second expert report: a global perspective. AICR, Washington, DC (Ref type: report) 12. Bingham SA (2002) Biomarkers in nutritional epidemiology. Public Health Nutr 5:821–827 13. Arab L, Akbar J (2002) Biomarkers and the measurement of fatty acids. Public Health Nutr 5:865–871 14. Willett WC (1990) Nutritional epidemiology. Oxford University Press, New York 15. Gleason DF, Mellinger GT (1974) Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J Urol 111:58–64 16. Jackson M, Walker S, Cade J, Forrester T, Cruickshank JK, Wilks R (2001) Reproducibility and validity of a quantitative foodfrequency questionnaire among Jamaicans of African origin. Public Health Nutr 4:971–980 17. Jackson MD, Walker SP, Younger NM, Bennett FI (2011) Use of a food frequency questionnaire to assess diets of Jamaican adults: validation and correlation with biomarkers. Nutr J 10:28 18. Zock PL, Mensink RP, Harryvan J, de Vries JH, Katan MB (1997) Fatty acids in serum cholesteryl esters as quantitative biomarkers of dietary intake in humans. Am J Epidemiol 145:1114–1122 19. Harvei S, Bjerve KS, Tretli S, Jellum E, Robsahm TE, Vatten L (1997) Prediagnostic level of fatty acids in serum phospholipids: omega-3 and omega-6 fatty acids and the risk of prostate cancer. Int J Cancer 71:545–551 20. Mannisto S, Pietinen P, Virtanen MJ et al (2003) Fatty acids and risk of prostate cancer in a nested case–control study in male smokers. Cancer Epidemiol Biomarkers Prev 12:1422–1428 21. Crowe FL, Allen NE, Appleby PN et al (2008) Fatty acid composition of plasma phospholipids and risk of prostate cancer in a case–control analysis nested within the European Prospective Investigation into Cancer and Nutrition. Am J Clin Nutr 88:1353–1363 22. Park SY, Wilkens LR, Henning SM et al (2009) Circulating fatty acids and prostate cancer risk in a nested case–control study: the Multiethnic Cohort. Cancer Causes Control 20:211–223 23. Lu QY, Arteaga JR, Zhang Q, Huerta S, Go VL, Heber D (2005) Inhibition of prostate cancer cell growth by an avocado extract: role of lipid-soluble bioactive substances. J Nutr Biochem 16:23–30 24. Takenaga F, Matsuyama K, Abe S, Torii Y, Itoh S (2008) Lipid and fatty acid composition of mesocarp and seed of avocado fruits harvested at northern range in Japan. J Oleo Sci 57:591–597 25. Ding H, Chin YW, Kinghorn AD, D’Ambrosio SM (2007) Chemopreventive characteristics of avocado fruit. Semin Cancer Biol 17:386–394 26. Bidoli E, Talamini R, Bosetti C et al (2005) Macronutrients, fatty acids, cholesterol and prostate cancer risk. Ann Oncol 16:152–157 27. Giovannucci E, Rimm EB, Colditz GA et al (1993) A prospective study of dietary fat and risk of prostate cancer. J Natl Cancer Inst 85:1571–1579 28. Kristal AR, Cohen JH, Qu P, Stanford JL (2002) Associations of energy, fat, calcium, and vitamin D with prostate cancer risk. Cancer Epidemiol Biomarkers Prev 11:719–725
Cancer Causes Control (2012) 23:23–33 29. West DW, Slattery ML, Robison LM, French TK, Mahoney AW (1991) Adult dietary intake and prostate cancer risk in Utah: a case–control study with special emphasis on aggressive tumors. Cancer Causes Control 2:85–94 30. Neuhouser ML, Barnett MJ, Kristal AR et al (2007) (n-6) PUFA increase and dairy foods decrease prostate cancer risk in heavy smokers. J Nutr 137:1821–1827 31. Park SY, Murphy SP, Wilkens LR, Henderson BE, Kolonel LN (2007) Fat and meat intake and prostate cancer risk: the multiethnic cohort study. Int J Cancer 121:1339–1345 32. Wallstrom P, Bjartell A, Gullberg B, Olsson H, Wirfalt E (2007) A prospective study on dietary fat and incidence of prostate cancer (Malmo, Sweden). Cancer Causes Control 18:1107–1121 33. Schuurman AG, van den Brandt PA, Dorant E, Brants HA, Goldbohm RA (1999) Association of energy and fat intake with prostate carcinoma risk: results from The Netherlands Cohort Study. Cancer 86:1019–1027 34. Whittemore AS, Kolonel LN, Wu AH et al (1995) 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 87:652–661 35. Crowe FL, Key TJ, Appleby PN et al (2008) Dietary fat intake and risk of prostate cancer in the European Prospective Investigation into Cancer and Nutrition. Am J Clin Nutr 87:1405–1413 36. Norrish AE, Jackson RT, Sharpe SJ, Skeaff CM (2000) Men who consume vegetable oils rich in monounsaturated fat: their dietary patterns and risk of prostate cancer (New Zealand). Cancer Causes Control 11:609–615 37. Escrich E, Moral R, Grau L, Costa I, Solanas M (2007) Molecular mechanisms of the effects of olive oil and other dietary lipids on cancer. Mol Nutr Food Res 51:1279–1292 38. Waterman E, Lockwood B (2007) Active components and clinical applications of olive oil. Altern Med Rev 12:331–342 39. Gann PH, Hennekens CH, Sacks FM, Grodstein F, Giovannucci EL, Stampfer MJ (1994) Prospective study of plasma fatty acids and risk of prostate cancer. J Natl Cancer Inst 86:281–286 40. Wakil SJ, Stoops JK, Joshi VC (1983) Fatty acid synthesis and its regulation. Annu Rev Biochem 52:537–579 41. Wakil SJ (1989) Fatty acid synthase, a proficient multifunctional enzyme. Biochemistry 28:4523–4530 42. Chavarro JE, Stampfer MJ, Li H, Campos H, Kurth T, Ma J (2007) A prospective study of polyunsaturated fatty acid levels in blood and prostate cancer risk. Cancer Epidemiol Biomarkers Prev 16:1364–1370 43. Laaksonen DE, Laukkanen JA, Niskanen L et al (2004) Serum linoleic and total polyunsaturated fatty acids in relation to prostate and other cancers: a population-based cohort study. Int J Cancer 111:444–450 44. Williams GT (1991) Programmed cell death: apoptosis and oncogenesis. Cell 65:1097–1098
33 45. Brasky TM, Till C, White E et al (2011) Serum phospholipid fatty acids and prostate cancer risk: results from the prostate cancer prevention trial. Am J Epidemiol 173:1429–1439 46. Chavarro JE, Stampfer MJ, Campos H, Kurth T, Willett WC, Ma J (2008) A prospective study of trans-fatty acid levels in blood and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev 17:95–101 47. Simon JA, Chen YH, Bent S (2009) The relation of alpha-linolenic acid to the risk of prostate cancer: a systematic review and meta-analysis. Am J Clin Nutr 89:1558S–1564S 48. Connolly JM, Coleman M, Rose DP (1997) Effects of dietary fatty acids on DU145 human prostate cancer cell growth in athymic nude mice. Nutr Cancer 29:114–119 49. Rose DP (1997) Effects of dietary fatty acids on breast and prostate cancers: evidence from in vitro experiments and animal studies. Am J Clin Nutr 66:1513S–1522S 50. Wang Y, Corr JG, Thaler HT, Tao Y, Fair WR, Heston WD (1995) Decreased growth of established human prostate LNCaP tumors in nude mice fed a low-fat diet. J Natl Cancer Inst 87:1456–1462 51. Godley PA, Campbell MK, Gallagher P, Martinson FE, Mohler JL, Sandler RS (1996) Biomarkers of essential fatty acid consumption and risk of prostatic carcinoma. Cancer Epidemiol Biomarkers Prev 5:889–895 52. Newcomer LM, King IB, Wicklund KG, Stanford JL (2001) The association of fatty acids with prostate cancer risk. Prostate 47:262–268 53. Hedelin M, Chang ET, Wiklund F et al (2007) Association of frequent consumption of fatty fish with prostate cancer risk is modified by COX-2 polymorphism. Int J Cancer 120:398–405 54. Polascik TJ, Oesterling JE, Partin AW (1999) Prostate specific antigen: a decade of discovery—what we have learned and where we are going. J Urol 162:293–306 55. Thompson IM, Pauler DK, Goodman PJ et al (2004) Prevalence of prostate cancer among men with a prostate-specific antigen level \ or =4.0 ng per milliliter. N Engl J Med 350:2239–2246 56. Ando R, Nagaya T, Hashimoto Y et al (2008) Inverse relationship between obesity and serum prostate-specific antigen level in healthy Japanese men: a hospital-based cross-sectional survey, 2004–2006. Urology 72:561–565 57. Baillargeon J, Pollock BH, Kristal AR et al (2005) The association of body mass index and prostate-specific antigen in a population-based study. Cancer 103:1092–1095 58. Beebe-Dimmer JL, Faerber GJ, Morgenstern H et al (2008) Body composition and serum prostate-specific antigen: review and findings from Flint Men’s Health Study. Urology 71:554–560 59. Werny DM, Thompson T, Saraiya M et al (2007) Obesity is negatively associated with prostate-specific antigen in U.S. men, 2001–2004. Cancer Epidemiol Biomarkers Prev 16:70–76
123
