Modulation of the association between plasma intercellular adhesion ...

1 downloads 0 Views 288KB Size Report
Feb 29, 2012 - growth (3) and cell migration (6). In epidemiologic cross- ..... Bouloumie A. From blood monocytes to adipose tissue-resident mac- rophages: ...
Modulation of the association between plasma intercellular adhesion molecule-1 and cancer risk by n23 PUFA intake: a nested case-control study1–3 Mathilde Touvier, Emmanuelle Kesse-Guyot, Valentina A Andreeva, Le´opold Fezeu, Nathalie Charnaux, Angela Sutton, Nathalie Druesne-Pecollo, Serge Hercberg, Pilar Galan, Laurent Zelek, Paule Latino-Martel, and Se´bastien Czernichow ABSTRACT Background: Mechanistic data suggest that n23 PUFAs and endothelial function may interact and play a role in carcinogenesis, but epidemiologic evidence is lacking. Objective: Our objective was to investigate whether the prospective association between soluble intercellular adhesion molecule-1 (sICAM1) and cancer risk is modulated by n23 PUFA intake. Design: A nested case-control study was designed to include all firstincident cancer cases diagnosed in the SUpple´mentation en VItamines et Mine´raux AntioXydants cohort between 1994 and 2007, with available dietary data from 24-h records (n = 408). Cases were matched with 1 or 2 randomly selected controls (n = 760). Conditional logistic regression was used to estimate ORs and 95% CIs for the association between prediagnostic plasma concentrations of sICAM-1 and cancer risk, stratified by n23 PUFA intake. The interactions between sICAM1 and n23 PUFA intake were tested. Results: An interaction was observed between sICAM-1 and n23 PUFA intake, which was consistent across the studied cancer locations (P-interaction = 0.036 for overall, 0.038 for breast, and 0.020 for prostate cancer risk). sICAM-1 concentrations were positively associated with cancer risk among subjects with n23 PUFA intakes below the median (multivariate ORTertile3vsTertile1: 2.8; 95% CI: 1.5, 5.2; P-trend = 0.001), whereas this association was not observed for subjects with n23 PUFA intakes above the median (ORTertile3vsTertile1: 1.3; 95% CI: 0.8, 2.3; P-trend = 0.3). Conclusion: These findings suggest that n23 PUFA intake may counteract the procarcinogenic actions of sICAM-1. This trial was registered at clinicaltrials.gov as NCT00272428. Am J Clin Nutr 2012;95:944–50.

INTRODUCTION

ICAM-14 is a member of the immunoglobulin superfamily of adhesion molecules constitutively expressed in many human tissues. ICAM-1 is involved in cell–cell and cell–basement membrane interactions and plays a key role in immunologic and inflammatory responses (1). Its synthesis is induced by various pro-inflammatory cytokines (2, 3). Available data have provided evidence for the involvement of ICAM-1 in carcinogenesis. Mechanistic studies have shown that ICAM-1 could stimulate angiogenesis and neovascularization (4, 5) and promote tumor growth (3) and cell migration (6). In epidemiologic crosssectional studies, concentrations of sICAM-1 were elevated in

944

patients with various cancers (1, 7, 8). However, prospective studies investigating the association between prediagnostic concentrations of sICAM-1 and cancer risk are lacking. Several environmental factors, such as cigarette smoke (9) and high ethanol consumption (10), have been shown to increase ICAM-1 expression. Conversely, n23 PUFAs could modulate ICAM-1– induced carcinogenesis. Indeed, previous experimental studies have shown that n23 PUFAs downregulate inflammatory processes involving leukocyte-endothelial interactions and decrease endothelial expression of adhesion molecules and endothelial activation (11, 12). In addition, experimental studies suggest that n23 PUFAs are involved in the same carcinogenic pathways as ICAM-1, but with antagonist actions (13, 14). The potential mechanisms involved in their anticarcinogenic action include suppression of the biosynthesis of proinflammatory molecules, modification of transcription factor activity and gene expression, influence on signal transduction, alteration of hormone-stimulated cell growth, suppression of the production of free radicals and reactive oxygen species, and influence on insulin sensitivity and membrane fluidity (13, 14). Because no epidemiologic study has previously investigated whether the prospective association between sICAM-1 and cancer risk is modulated by n23 PUFAs, we tested the hypothesis of an interaction between sICAM-1 and n23 PUFA intake on cancer risk to estimate whether these fatty acids might counteract the 1 From INSERM U557, National Institute of Health and Medical Research, Inra, Cnam, Paris 13 University (UREN), Bobigny, France (MT, EK-G, VAA, LF, ND-P, SH, PG, LZ, and PL-M); Jean Verdier Hospital, Biochemistry Department, Bondy, France (NC and AS); INSERM U698, Paris 13 University, Bobigny, France (NC, AS); the Public Health Department, Avicenne Hospital, Paris 13 University, Bobigny, France (SH); and the Department of Nutrition, Ambroise Pare´ Hospital, INSERM U1018–Centre for Research in Epidemiology and Population Health, University Versailles St-Quentin, Boulogne-Billancourt, France (SC). 2 Supported by a grant from the French National Cancer Institute (Institut National du Cancer, INCa, 2007-1-SPC-3). 3 Address correspondence to M Touvier, INSERM U557, UREN, SMBH Paris 13, 74 rue Marcel Cachin, F-93017, Bobigny, France. E-mail: m.touvier@ uren.smbh.univ-paris13.fr. 4 Abbreviations used: ALA, a-linolenic acid; DPA, docosapentaenoic acid; ICAM-1, intercellular adhesion molecule 1; sICAM-1, soluble intercellular adhesion molecule 1; SU.VI.MAX, SUpple´mentation en VItamines et Mine´raux AntioXydants; T, tertile. Received September 27, 2011. Accepted for publication January 20, 2011. First published online February 29, 2012; doi: 10.3945/ajcn.111.027805.

Am J Clin Nutr 2012;95:944–50. Printed in USA. Ó 2012 American Society for Nutrition

sICAM-1, OMEGA 3 INTAKE, AND CANCER RISK

procarcinogenic actions of sICAM-1. Because n23 PUFAs from different food sources may play contrasting roles (because of differential bioavailability, type of n23 PUFA, internutrient and nutrient-food matrix interactions, etc), we also tested whether the potential interaction with sICAM-1 was restricted to specific food sources of n23 PUFAs. SUBJECTS AND METHODS

Subjects The SU.VI.MAX study is a population-based, double-blind, placebo-controlled, randomized trial initially designed to assess the effect of a daily antioxidant supplementation on the incidence of cardiovascular disease and cancer (15). A total of 13,017 subjects were enrolled in 1994–1995. The intervention study lasted 8 y, and follow-up of health events was maintained until July 2007. Subjects provided written informed consent, and the study was approved by the Ethics Committee for Studies with Human Subjects at the ParisCochin Hospital (CCPPRB n°706/2364) and the “Commission Nationale de l’Informatique et des Liberte´s” (CNIL n°334641/ 907094). Data collection At enrollment, all participants underwent a clinical examination and anthropometric measurements by the study nurses and physicians. They completed questionnaires regarding sociodemographic data, smoking, physical activity, and medication use. Fasting venous blood samples were also obtained. Plasma aliquots were immediately prepared and stored frozen in liquid nitrogen. Dietary data were collected by using the Minitel Telematic Network, a small terminal that was widely used as an adjunct to the telephone in France at the beginning of the study. Participants were invited to provide a 24-h dietary record every 2 mo. These records were randomly distributed over 2 weekend days and 4 weekdays per year, so that each day of the week and all seasons were covered to account for individual variability in intake. Only dietary records for the first 2 y of follow-up were considered in the current study for purposes of consistency with a prospective design. Portion sizes were estimated by using a validated picture booklet distributed to the participants at enrollment (16). French recipes validated by food and nutrition professionals were used to assess the amounts consumed from composite dishes. The mean daily energy, alcohol, and nutrient intakes were estimated by using a published French foodcomposition table (17). Descriptions of dietary intakes and sources of n23 and n26 PUFAs in the SU.VI.MAX study have been published previously (18). Subjects were advised against taking any self-prescribed dietary supplements during their participation in the SU.VI.MAX study. Case ascertainment Confirmed or suspected events were self-reported by subjects during the follow-up process. Investigations were conducted in all such cases to obtain medical data from participants, physicians, and/or hospitals. All information was reviewed by an independent expert committee and cases were validated by pathological reports and classified using the International Chronic Diseases Classification, 10th Revision, Clinical Modification (ICD-10) (19).

945

Nested case-control study All first primary incident cancer cases diagnosed between inclusion in the SU.VI.MAX cohort in 1994 and July 2007 were included in the current study. Controls with complete follow-up data and without a cancer diagnosis by the end of follow-up were randomly selected among the participants and were matched by sex, age (by 2-y strata), BMI (in kg/m2; ,25 or 25), and intervention group. One or 2 controls per case were selected, depending on the availability of subjects with respect to the required matching criteria. Baseline plasma samples were used to determine the concentrations of sICAM-1 with an ELISA sandwich technique (R & D Laboratory Systems). Intra- and interassay CVs were ,10%. Statistical methods The participants’ baseline characteristics were compared between cases and controls by using Student’s t tests or chi-square tests. The association between sICAM-1 and incident cancer was examined with conditional logistic regression models and expressed as ORs for sex-specific tertiles of sICAM-1 concentrations and 95% CIs. The interactions between plasma sICAM-1 concentration and n23 PUFA intake (overall and from different food sources), n26 PUFA intake, n26/n23 PUFA intake ratio, fatty fish intake, and fruit and vegetable intake were successively tested in multivariate models. A 3-factor interaction between sICAM-1, n23 PUFAs, and BMI was also tested. The OR for the association between sICAM-1 and cancer risk were computed overall and stratified by n23 and n26 PUFA intakes and by fatty fish intake (below or above the sex-specific median). Multivariate models were adjusted for sex, age, BMI, height, intervention group, alcohol intake, physical activity, smoking status, educational level, and number of dietary records. With the use of the residual method, dietary intakes were treated as without-alcohol energy-adjusted variables. In site-specific analyses, multivariate models were further adjusted for family history of breast cancer, number of children, menopausal hormone therapy and menopausal status at baseline (in breast cancer analyses), baseline prostate specific antigen concentration, and family history of prostate cancer (in prostate cancer analyses). Further adjustments were also tested for n23 PUFA intake as a continuous variable (to account for a potential residual confounding effect within n23 PUFA intake classes), total lipid intake, and fruit and vegetable intake. All statistical tests were 2-sided, and P , 0.05 was considered statistically significant. All analyses were performed with SAS software (v9.2; SAS Institute Inc). RESULTS

A total of 512 incident cancer cases were diagnosed during follow-up, for which dietary data were available for 408 cases (at least one 24-h dietary record during the first 2 y of follow-up): 178 breast, 129 prostate, and 101 other cancer cases (colorectal, lung, thyroid, skin melanomas, esophagus, and stomach). A total of 760 controls were randomly selected and matched to the cases. Median follow-up time was 6.3 y for cases and 13 y for controls. The characteristics of the cancer cases and controls are described in Table 1. Contributors to n23 PUFA intake were fish (30%), vegetable oils (11%), nuts and oil seeds (5%), other animal products (meat: 13%; milk and dairy products: 12%; animal fat: 7%; eggs: 3%),

946

TOUVIER ET AL TABLE 1 Baseline characteristics of cancer cases and controls1

Sex [n (%)] Men Women Age (y) BMI [n (%)] ,25 kg/m2 25 to ,30 kg/m2 30 kg/m2 Height (cm) Intervention group [n (%)] Yes No (placebo) Smoking status [n (%)] Never smoker Former smoker Current smoker Physical activity [n (%)] Low Moderate High Education level [n (%)] ,12 y 12 y Alcohol intake (g/d) Plasma sICAM-1 (ng/mL) n23 PUFA intake (g/d) 18:3n23 PUFA (ALA) intake (g/d) 20:5n23 PUFA (EPA) intake (g/d) 22:5n23 PUFA (DPA) intake (g/d) 22:6n23 PUFA (DHA) intake (g/d) n23 PUFA intake from animal foods (% of n23 intake) n23 PUFA intake from plant foods (% of n23 intake) Fatty fish intake (g/d)4 n26 PUFA intake (g/d) Energy intake (kcal/d)

Cancer cases (n = 408)

Controls (n = 760)

181 (44.4) 227 (55.6) 51.4 6 6.13

343 (45.1) 417 (54.9) 51.3 6 6.0

P2 0.80

261 (64.0) 106 (26.0) 41 (10.1) 167.6 6 8.5

483 (63.6) 221 (29.1) 56 (7.4) 166.7 6 8.3

203 (49.8) 205 (50.3)

380 (50.0) 380 (50.0)

199 (48.8) 143 (35.1) 66 (16.2)

388 (51.1) 293 (38.6) 79 (10.4)

104 (25.5) 119 (29.2) 185 (45.3)

193 (25.4) 206 (27.1) 361 (47.5)

239 (58.6) 169 (41.4) 16.8 6 19.1 248.4 6 75.1 1.35 6 0.77 0.85 6 0.48 0.15 6 0.17 0.07 6 0.05 0.28 6 0.31 65.8 6 13.1 34.2 6 12.9 21.0 6 32.2 11.22 6 5.42 2096.3 6 652.0

439 (57.8) 321 (42.2) 15.0 6 16.6 239.7 6 64.9 1.29 6 0.56 0.85 6 0.36 0.13 6 0.13 0.06 6 0.04 0.25 6 0.22 64.6 6 13.1 35.4 6 12.9 18.6 6 25.2 11.41 6 4.63 2117.0 6 639.9

0.81 0.20

0.07 0.94

0.02

0.72

0.79

0.70 0.08 0.56 0.51 0.17 0.35 0.48 0.19 0.18 0.96 0.08 0.55

ALA, a-linolenic acid; DPA, docosapentaenoic acid; sICAM-1, soluble intercellular adhesion molecule-1. P value for the comparison of cancer cases and noncases by Student’s t tests for continuous variables and chi-square tests for categorical variables. Continuous variables were log-transformed to improve normality. 3 Mean 6 SD (all such values). 4 Salmon, sardines, tuna, anchovies, sea bass, carp, eel, halibut, mullet, dogfish, trout, fish roe, herring, and mackerel. 1 2

fruit and vegetables (7%), starchy foods (5%), cakes and pastries (5%), and other (2%). In subjects with n23 PUFA intakes below the median, the range of sICAM-1 concentrations was 113–684 ng/mL; the corresponding range was 83–618 ng/mL in subjects with n23 PUFA intakes above the median (data not tabulated). Overall associations between sICAM-1, n23 PUFAs, and cancer risk are presented in Table 2. The multivariate OR of cancer risk for the third compared with the first tertile of sICAM-1 was 1.3 (95% CI: 0.9, 1.7; P-trend = 0.1). However, the relation between sICAM-1 and cancer risk was modulated by n23 PUFA intake (P-interaction = 0.036; Figure 1). sICAM-1 was positively associated with an increased cancer risk among subjects with n23 PUFA intakes below the median (ORT3vsT1: 2.8; 95% CI: 1.5, 5.2; P-trend = 0.001), whereas this association was not observed in subjects with n23 PUFA intake above the median (ORT3vsT1: 1.3; 95% CI: 0.8, 2.3; P-trend = 0.3). The interaction between n23 PUFA intake and sICAM-1 regarding cancer risk was observed for n23 PUFAs from animal

sources (P = 0.049) but not from plant foods (P = 0.5) (Figure 2). Within the animal sources, this interaction was observed with n23 PUFAs from fatty fish (P = 0.043) but not with those from nonmarine sources (P = 0.2; Figure 3). We also observed an interaction between fatty fish intake and sICAM-1 on cancer risk (P = 0.03, data not tabulated). sICAM-1 was positively associated with cancer risk among subjects with fatty fish intake below the median (ORT3vsT1: 1.9; 95% CI: 1.02, 3.5; P-trend = 0.04), whereas this association was not observed for subjects whose fatty fish intake was above the median (ORT3vsT1: 1.2; 95% CI: 0.7, 2.0; P-trend = 0.5). An interaction between the n26/n23 PUFA intake ratio and sICAM-1 on cancer risk was also observed (P = 0.039). However, no interaction was detected with n26 PUFA intake alone (P = 0.4; data not tabulated). None of the interactions between sICAM-1 and each n23 PUFA considered separately or as a combination of long-chain n23 PUFAs were statistically significant (P = 0.1, 0.2, 0.1, 0.2, 0.4, and 0.4 for ALA, EPA, DPA, DHA,

sICAM-1, OMEGA 3 INTAKE, AND CANCER RISK TABLE 2 Multivariate ORs for the association between tertiles of sICAM-1 concentrations and n23 PUFA intake and cancer risk1 OR (95% CI)

No. of cases/controls

1 1.0 (0.8, 1.4) 1.3 (0.9, 1.7) 0.13

126/262 133/258 149/240

1 1.1 (0.8, 1.5) 1.2 (0.8, 1.7) 0.38

137/251 131/259 140/250

2

Plasma sICAM-1 T1 T2 T3 P-trend n23 PUFA intake3 T1 T2 T3 P-trend

1 Conditional logistic regression models adjusted for sex, age, BMI, height, intervention group, alcohol intake, physical activity, smoking status, educational level, number of dietary records, and energy intake. The P values were derived by using the Wald test. sICAM-1, soluble intercellular adhesion molecule-1; T, tertile. 2 Cutoffs for sex-specific tertiles of sICAM-1 (in ng/mL) were as follows: 217.0 and 266.2 in men and 205.0 and 253.0 in women. 3 Cutoffs for sex-specific tertiles of n23 fatty acid intakes (g/d) were as follows: 1.2 and 1.6 in men and 0.9 and 1.2 in women.

EPA+DHA, and EPA+DPA+DHA, respectively; data not tabulated). No interaction was observed between sICAM-1 and fruit and vegetable intake on cancer risk (P = 0.9; data not tabulated). In site-specific analyses, the interaction between n23 PUFA intake and sICAM-1 was observed for both main cancer locations, ie, for breast cancer (ORT3vsT1: 4.7; 95% CI: 1.6, 13.4; P-trend = 0.004 in subjects with n23 PUFA intakes below the median and ORT3vsT1: 0.9; 95% CI: 0.4, 2.2; P-trend = 0.8 in subjects with n23 PUFA intakes above the median; P-interaction = 0.038) and for prostate cancer (ORT3vsT1: 6.1; 95% CI: 1.1, 34.5; P-trend = 0.03 in subjects with n23 PUFA intakes below the median and ORT3vsT1: 0.3; 95% CI: 0.1, 1.6 in subjects with n23 PUFA intakes above the median; P-trend = 0.2; P-interaction = 0.02) (data not tabulated). Further adjustment for total lipid intake, fruit and vegetable intake, and n23 PUFA intakes as a continuous variable did not modify the findings, nor did the sensitivity analysis excluding cases (n = 10) diagnosed during the first year of follow-up (data not shown). The 3-factor interaction between sICAM-1, n23 PUFAs, and BMI was not statistically significant (P = 0.66).

947

DISCUSSION

To date, no epidemiologic study had investigated the potential modulation of the prospective association between sICAM-1 concentrations and cancer risk by n23 PUFA intake. For the first time, the results of this nested case-control study suggest that n23 PUFAs modulate the prospective association between plasma ICAM-1 and cancer risk. sICAM-1 was associated with an increased cancer risk among subjects with low n23 PUFA intakes, whereas no association was observed for subjects with higher n23 PUFA intakes. This interaction was consistent across the studied cancer sites (ie, overall, breast, and prostate cancers). Several prevalent case-control studies have observed higher circulating concentrations of sICAM-1 in cancer cases compared with controls for various cancer sites (1, 7, 8). Genetic studies have shown that some single nucleotide polymorphisms of the ICAM-1 gene were associated with an increased cancer risk (20–23), although this point is debated (24). Higher concentrations of sICAM-1 were also associated with a poorer prognosis of cancer (8, 25). However, prospective studies investigating the association between prediagnostic concentrations of sICAM-1 and cancer risk are scarce. One study found a prospective association between sICAM-1 and non-Hodgkin lymphomas (26). Another study found no association between prediagnostic concentrations of sICAM-1 and overall cancer risk, but it involved only 46 cases (27). In contrast, the role of n23 PUFAs in carcinogenesis has been extensively investigated in epidemiologic studies for many cancer sites. Despite promising evidence from ecologic and experimental studies, analytic epidemiologic studies have reported somewhat conflicting results regarding the effects of fish and n23 PUFAs on cancer risk (28–39). Indeed, although some studies showed an inverse association between the intake of n23 PUFAs or fish and cancer risk, most did not. Several methodologic drawbacks may account for these inconsistencies, among which are n23 PUFA intakes too low to produce an effect, low within-population variability, nondifferential misclassification of estimated n23 PUFA intakes, and n23 PUFA exposure measured too late in life (close to cancer diagnosis), whereas the critical period may be several years earlier (13). Substantial evidence from mechanistic studies based on experimental, animal, or in vitro models indicates that n23 PUFAs may counteract the procarcinogenic ICAM-1–induced mechanisms, in line with the findings of the current study. It has been suggested that dietary n23 PUFAs modulate the risk condition known as

FIGURE 1. Multivariate ORs for the association between tertiles of sICAM-1 concentrations and cancer risk, stratified by median n23 PUFA intake (P-interaction = 0.036). Cutoffs for sex-specific tertiles of sICAM-1 (in ng/mL) were as follows: 217.0 and 266.2 in men and 205.0 and 253.0 in women. Sexspecific median n23 PUFA intakes (in g/d) were as follows: 1.4 in men and 1.1 in women. Conditional logistic regression analyses adjusted for sex, age, BMI, height, intervention group, alcohol intake, physical activity, smoking status, educational level, number of dietary records, and energy intake. The P values were derived by using the Wald test. The number of cases/controls in each tertile is indicated in parentheses. sICAM-1, soluble intercellular adhesion molecule-1; T, tertile.

948

TOUVIER ET AL

FIGURE 2. Multivariate ORs for the association between tertiles of sICAM-1 concentrations and cancer risk, stratified by median n23 PUFA intake from animal sources (P-interaction = 0.049) and from plant sources (P-interaction = 0.5). Cutoffs for sex-specific tertiles of sICAM-1 (in ng/mL) were as follows: 217.0 and 266.2 in men and 205.0 and 253.0 in women. Sex-specific median n23 PUFA intakes (in g/d) were as follows: 0.9 in men and 0.7 in women (from animal sources) and 0.4 in men and 0.3 in women (from plant sources). Conditional logistic regression analyses adjusted for sex, age, BMI, height, intervention group, alcohol intake, physical activity, smoking status, educational level, number of dietary records, and energy intake. The P values were derived by using the Wald test. The number of cases/controls in each tertile is indicated in parentheses. sICAM-1, soluble intercellular adhesion molecule-1; T, tertile.

“endothelial dysfunction” by reverting the endothelial alterations associated with it (40). Two kinds of potential mechanisms merit attention. First, previous experimental studies have shown that n23 PUFAs decrease the endothelial expression of adhesion molecules (11, 12). Chen et al (41) showed that DHA inhibited the expression of cytokine-induced cell adhesion molecules through suppression of nuclear transcription factor jB nuclear translocation and upstream I-jB a-phosphorylation and degradation. In addition, in a subsample of the Nurses’ Health Study, a Western dietary pattern with low n23 PUFA intakes was positively associated with sICAM-1 concentrations (42). In line with these findings, the range of sICAM-1 concentrations in our study was slightly lower in subjects with higher n23 PUFA intakes. Second, it has been shown experimentally that ICAM-1 plays a procarcinogenic role at different stages of cancer development and through several mechanisms, whereas n23 PUFAs intervene in the same carcinogenic pathways, but with an anticarcinogenic action (13, 14). ICAM-1 stimulates angiogenesis and neovascularization (4, 5) and promotes tumor growth (3), cell migration (6), and escape of the tumor cells from immune surveillance (3). ICAM-1 is also a predictor of oxidative stress,

as measured by 8-iso-prostaglandin F2a (43). In contrast, n23 PUFAs are involved in several mechanisms that counteract these carcinogenic processes (13, 44–47). They alter the immune response to cancer cells and decrease inflammation, cell proliferation, metastasis, and angiogenesis and contribute to restore apoptosis. They also limit tumor cell growth and promote differentiation by influencing transcription factor activity, gene expression, and signal transduction. They reduce estrogenstimulated cell growth and play a role in insulin sensitivity and membrane fluidity. Finally, these nutrients may reduce oxidative stress by modulating the production of free radicals and reactive oxygen species. In our study, the interaction between sICAM-1 and n23 PUFAs was statistically significant for overall n23 PUFA intakes, but was not statistically significant for each n23 PUFA type taken separately or for the combination of long-chain n23 PUFAs. Although experimental research suggests that longchain n23 PUFAs may more efficiently modulate both plasma concentrations of sICAM-1 (11, 12) and carcinogenesis (13) than ALA, the role of the latter needs clarification. ALA is by far the main n23 PUFA type consumed in occidental diets. Part

FIGURE 3. Multivariate ORs for the association between tertiles of sICAM-1 concentrations and cancer risk, stratified by median n23 PUFA intake from fatty fish (P-interaction = 0.043) and from nonmarine sources (P-interaction = 0.2). Cutoffs for sex-specific tertiles of sICAM-1 (in ng/mL) were as follows: 217.0 and 266.2 in men and 205.0 and 253.0 in women. Sex-specific median n23 PUFA intakes (in g/d) were as follows: 0.2 in men and 0.1 in women (from fatty fish) and 0.5 in men and 0.4 in women (from nonmarine sources). Conditional logistic regression analyses adjusted for sex, age, BMI, height, intervention group, alcohol intake, physical activity, smoking status, educational level, number of dietary records, and energy intake. The P values were derived by using the Wald test. The number of cases/controls in each tertile is indicated in parentheses. sICAM-1, soluble intercellular adhesion molecule-1; T, tertile.

sICAM-1, OMEGA 3 INTAKE, AND CANCER RISK

of ALA intake may be converted into long-chain fatty acids, although this process is limited in humans (13). In addition, ALA may also exert independent effects (48). Although evidence is still inconsistent (13, 30, 31), a protective effect of ALA on cancer risk has been suggested in human studies (30). Taken together, these findings support our results, which suggest that dietary intake of a combination of all types of n23 PUFAs may be the variable of interest regarding the modulation of the association between sICAM-1 and cancer risk. We observed an interaction between sICAM-1 and n23 PUFAs from animal but not plant sources, more specifically from fatty fish but not from nonmarine animal sources. These results might be explained by mechanisms other than the ones linked to the different types of n23 PUFAs contained in such dietary sources, as noted above. Indeed, it cannot be excluded that the observed modulation of sICAM-1–related cancer risk by n23 PUFAs/fatty fish intake is in fact due to an overall healthy diet, for which fish intake would be a marker. However, a broad range of known confounding factors has been accounted for in our models, including fruit and vegetable intakes (often considered as a proxy for a “healthy” diet), which did not modify the findings. In addition, the interaction between sICAM-1 and fruit and vegetable intake was not statistically significant, which suggests that the observed interaction was specific to n23 PUFAs and could not be attributed to an overall healthy diet. Thus, a more causal interpretation could be that beyond the type of n23 PUFAs, other factors such as bioavailability, nutrientfood matrix, nutrient-nutrient interactions, and overall nutritional quality of the vector food may play a discriminatory role. For instance, the intake of vegetable oils that are particularly rich in n23 PUFAs (such as walnut, rapeseed, and soybean oils) is low in France (especially in 1994–1995 when dietary data for the current study were collected) (18). Consequently, other oils, such as olive oil—which are less rich in n23 PUFAs (;0.9 compared with 12 g/100 g for walnut oil) but are consumed in greater amounts—are stronger contributors to n23 PUFA intakes. A large quantity of oils (and thus lipids) is consumed to achieve substantial amounts of n23 PUFAs with this vector food. Some limitations should be acknowledged. First, exposure assessment may have been affected by classification errors. Indeed, although the median number of 24-h dietary records per subject was 10, a small part (6.8%) of the subjects had only one 24-h dietary record. No data were available regarding plasma n23 PUFA concentrations in our sample. In addition, a single measurement of sICAM-1 was available at baseline. Although the probability of a differential bias between cases and controls is low because of the prospective design, these limitations in exposure assessment might have attenuated the observed associations due to intraindividual variation. This may have limited our ability to detect some associations. Second, the possibility that n23 PUFAs do not directly modulate sICAM-1, but other mediators involved in the same steps of carcinogenesis (inflammation, angiogenesis) cannot be excluded. This should be elucidated in further epidemiologic and mechanistic studies. Next, the observed relations might have been partly affected by unmeasured or residual confounding. For instance, no information was available in our study regarding the plasma profiling of carotenoid/antioxidant vitamins of subjects with low or high dietary intakes of n23 PUFAs. Finally, no information was available on trans fatty acid intakes. Some studies suggest direct associations

949

between trans fatty acids and both cancer risk and plasma concentrations of sICAM-1 (51). Thus, the potential modulation of the relation between sICAM-1 and cancer risk by trans fatty acid intake should also be investigated in future studies. Our findings suggest that dietary intake of n23 PUFAs may counteract the procarcinogenic actions of sICAM-1, which has never been investigated before in any epidemiologic study. The identification of behavioral and environmental factors that could modulate ICAM-1–induced carcinogenesis may pave the way for new preventive strategies. Further epidemiologic and experimental studies are needed to identify and verify potential mechanisms in humans to gain more insight into the effects of n23 PUFA intakes and ICAM-1 on cancer risk. The authors’ responsibilities were as follows—MT and SC: designed the research; SH, PG, NC, AS, and SC: conducted the research; MT: analyzed the data and led the writing; EK-G, VAA, LF, ND-P, LZ, PL-M, SH, PG, NC, AS, and SC: contributed to the data interpretation and revised each draft for important intellectual content; and MT: had primary responsibility for the final content. All authors read and approved the final manuscript. No conflicts of interest were declared. The funders had no role in the design, implementation, analysis, or interpretation of the data.

REFERENCES 1. O’Hanlon DM, Fitzsimons H, Lynch J, Tormey S, Malone C, Given HF. Soluble adhesion molecules (E-selectin, ICAM-1 and VCAM-1) in breast carcinoma. Eur J Cancer 2002;38:2252–7. 2. Curat CA, Miranville A, Sengenes C, Diehl M, Tonus C, Busse R, Bouloumie A. From blood monocytes to adipose tissue-resident macrophages: induction of diapedesis by human mature adipocytes. Diabetes 2004;53:1285–92. 3. Gho YS, Kim PN, Li HC, Elkin M, Kleinman HK. Stimulation of tumor growth by human soluble intercellular adhesion molecule-1. Cancer Res 2001;61:4253–7. 4. Byrne GJ, Ghellal A, Iddon J, Blann AD, Venizelos V, Kumar S, Howell A, Bundred NJ. Serum soluble vascular cell adhesion molecule-1: role as a surrogate marker of angiogenesis. J Natl Cancer Inst 2000;92: 1329–36. 5. Gho YS, Kleinman HK, Sosne G. Angiogenic activity of human soluble intercellular adhesion molecule-1. Cancer Res 1999;59:5128–32. 6. Kobayashi H, Boelte KC, Lin PC. Endothelial cell adhesion molecules and cancer progression. Curr Med Chem 2007;14:377–86. 7. Basoglu M, Atamanalp SS, Yildirgan MI, Aydinli B, Ozturk G, Akcay F, Oren D. Correlation between the serum values of soluble intercellular adhesion molecule-1 and total sialic acid levels in patients with breast cancer. Eur Surg Res 2007;39:136–40. 8. Boyano MD, Garcia-Vazquez MD, Lopez-Michelena T, Gardeazabal J, Bilbao J, Canavate ML, Galdeano AG, Izu R, Diaz-Ramon L, Raton JA, et al. Soluble interleukin-2 receptor, intercellular adhesion molecule-1 and interleukin-10 serum levels in patients with melanoma. Br J Cancer 2000;83:847–52. 9. Chen HW, Chien ML, Chaung YH, Lii CK, Wang TS. Extracts from cigarette smoke induce DNA damage and cell adhesion molecule expression through different pathways. Chem Biol Interact 2004;150: 233–41. 10. Sacanella E, Estruch R. The effect of alcohol consumption on endothelial adhesion molecule expression. Addict Biol 2003;8:371–8. 11. Calder PC. n23 polyunsaturated fatty acids, inflammation, and inflammatory diseases. Am J Clin Nutr 2006;83:1505S–19S. 12. De Caterina R, Liao JK, Libby P. Fatty acid modulation of endothelial activation. Am J Clin Nutr 2000;71:213S–23S. 13. Larsson SC, Kumlin M, Ingelman-Sundberg M, Wolk A. Dietary long-chain n23 fatty acids for the prevention of cancer: a review of potential mechanisms. Am J Clin Nutr 2004;79:935–45. 14. Pauwels EK, Kairemo K. Fatty acid facts, part II: role in the prevention of carcinogenesis, or, more fish on the dish? Drug News Perspect 2008; 21:504–10. 15. Hercberg S, Galan P, Preziosi P, Bertrais S, Mennen L, Malvy D, Roussel AM, Favier A, Briancon S. The SU.VI.MAX study: a randomized,

950

16.

17. 18. 19. 20. 21.

22.

23.

24. 25.

26.

27. 28. 29. 30. 31. 32. 33.

TOUVIER ET AL placebo-controlled trial of the health effects of antioxidant vitamins and minerals. Arch Intern Med 2004;164:2335–42. Le Moullec N, Deheeger M, Preziosi P, Montero P, Valeix P, Rolland-Cachera MF, Potier de Courcy G, Christides JP, Galan P, Hercberg S. Validation du manuel photos utilise´ pour l’enqueˆte alimentaire de l’e´tude SU.VI.MAX. [Validation of the food portion size booklet used in the SU.VI.MAX study.] Cah Nutr Diet 1996;31: 158–64 (in French). Hercberg S. Table de composition SU.VI.MAX des aliments. Paris, France: Les e´ditions INSERM/Economica, 2005. Astorg P, Arnault N, Czernichow S, Noisette N, Galan P, Hercberg S. Dietary intakes and food sources of n26 and n23 PUFA in French adult men and women. Lipids 2004;39:527–35. WHO. ICD-10, International classification of diseases and related health problems. 10th revision. Geneva, Switzerland: World Health Organization, 1993. Arandi N, Talei A, Erfani N, Ghaderi A. Intercellular adhesion molecule-1 genetic markers (+241G/A and +469A/G) in Iranian women with breast cancer. Cancer Genet Cytogenet 2008;183:9–13. Han W, Kang SY, Kang D, Park SK, Lee JY, Kim H, Park AK, Noh DY. Multiplex genotyping of 1107 SNPs from 232 candidate genes identified an association between IL1A polymorphism and breast cancer risk. Oncol Rep 2010;23:763–9. Kammerer S, Roth RB, Reneland R, Marnellos G, Hoyal CR, Markward NJ, Ebner F, Kiechle M, Schwarz-Boeger U, Griffiths LR, et al. Large-scale association study identifies ICAM gene region as breast and prostate cancer susceptibility locus. Cancer Res 2004;64: 8906–10. Theodoropoulos G, Papaconstantinou I, Felekouras E, Nikiteas N, Karakitsos P, Panoussopoulos D, Lazaris AC, Patsouris E, Bramis J, Gazouli M. Relation between common polymorphisms in genes related to inflammatory response and colorectal cancer. World J Gastroenterol 2006;12:5037–43. Cox DG, Hankinson SE, Hunter DJ. Polymorphisms in the ICAM gene locus are not associated with breast cancer Risk. Cancer Epidemiol Biomarkers Prev 2006;15:178–9. Giannoulis K, Angouridaki C, Fountzilas G, Papapolychroniadis C, Giannoulis E, Gamvros O. Serum concentrations of soluble ICAM-1 and VCAM-1 in patients with colorectal cancer. Clinical implications. Tech Coloproctol 2004;8(suppl 1):s65–7. Saberi Hosnijeh F, Krop EJ, Scoccianti C, Krogh V, Palli D, Panico S, Tumino R, Sacredote C, Nawroly N, Portengen L, et al. Plasma cytokines and future risk of non-Hodgkin lymphoma (NHL): a casecontrol study nested in the Italian European Prospective Investigation into Cancer and Nutrition. Cancer Epidemiol Biomarkers Prev 2010; 19:1577–84. Kamezaki S, Kurozawa Y, Iwai N, Hosoda T, Okamoto M, Nose T. Serum levels of soluble ICAM-1 and VCAM-1 predict pre-clinical cancer. Eur J Cancer 2005;41:2355–9. AICR/WCRF. American Institute for Cancer Research/World Cancer Research Fund—food, nutrition, physical activity and the prevention of cancer: a global perspective. Washington, DC: AICR, 2007. Astorg P. Dietary N-6 and N-3 polyunsaturated fatty acids and prostate cancer risk: a review of epidemiological and experimental evidence. Cancer Causes Control 2004;15:367–86. Carayol M, Grosclaude P, Delpierre C. Prospective studies of dietary alpha-linolenic acid intake and prostate cancer risk: a meta-analysis. Cancer Causes Control 2010;21:347–55. Simon JA, Chen YH, Bent S. The relation of alpha-linolenic acid to the risk of prostate cancer: a systematic review and meta-analysis. Am J Clin Nutr 2009;89:1558S–64S. Szymanski KM, Wheeler DC, Mucci LA. Fish consumption and prostate cancer risk: a review and meta-analysis. Am J Clin Nutr 2010;92: 1223–33. Terry PD, Rohan TE, Wolk A. Intakes of fish and marine fatty acids and the risks of cancers of the breast and prostate and of other

34. 35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48. 49.

50. 51.

hormone-related cancers: a review of the epidemiologic evidence. Am J Clin Nutr 2003;77:532–43. Gerber M. Background review paper on total fat, fatty acid intake and cancers. Ann Nutr Metab 2009;55:140–61. Hooper L, Thompson RL, Harrison RA, Summerbell CD, Ness AR, Moore HJ, Worthington HV, Durrington PN, Higgins JP, Capps NE, et al. Risks and benefits of omega 3 fats for mortality, cardiovascular disease, and cancer: systematic review. BMJ 2006;332:752–60. MacLean CH, Newberry SJ, Mojica WA, Khanna P, Issa AM, Suttorp MJ, Lim YW, Traina SB, Hilton L, Garland R, et al. Effects of omega-3 fatty acids on cancer risk: a systematic review. JAMA 2006;295:403– 15. Terry PD, Terry JB, Rohan TE. Long-chain (n23) fatty acid intake and risk of cancers of the breast and the prostate: recent epidemiological studies, biological mechanisms, and directions for future research. J Nutr 2004;134:3412S–20S. Bougnoux P, Maillard V, Chajes V. Omega-6/omega-3 polyunsaturated fatty acids ratio and breast cancer. World Rev Nutr Diet 2005;94:158– 65. Thie´baut AC, Chajes V, Gerber M, Boutron-Ruault MC, Joulin V, Lenoir G, Berrino F, Riboli E, Benichou J, Clavel-Chapelon F. Dietary intakes of omega-6 and omega-3 polyunsaturated fatty acids and the risk of breast cancer. Int J Cancer 2009;124:924–31. Serini S, Piccioni E, Calviello G. Dietary n23 PUFA vascular targeting and the prevention of tumor growth and age-related macular degeneration. Curr Med Chem 2009;16:4511–26. Chen W, Esselman WJ, Jump DB, Busik JV. Anti-inflammatory effect of docosahexaenoic acid on cytokine-induced adhesion molecule expression in human retinal vascular endothelial cells. Invest Ophthalmol Vis Sci 2005;46:4342–7. Lopez-Garcia E, Schulze MB, Fung TT, Meigs JB, Rifai N, Manson JE, Hu FB. Major dietary patterns are related to plasma concentrations of markers of inflammation and endothelial dysfunction. Am J Clin Nutr 2004;80:1029–35. Cottone S, Mule G, Nardi E, Vadala A, Lorito MC, Guarneri M, Arsena R, Palermo A, Cerasola G. C-reactive protein and intercellular adhesion molecule-1 are stronger predictors of oxidant stress than blood pressure in established hypertension. J Hypertens 2007;25:423–8. Calviello G, Serini S, Piccioni E. n23 polyunsaturated fatty acids and the prevention of colorectal cancer: molecular mechanisms involved. Curr Med Chem 2007;14:3059–69. Chapkin RS, Seo J, McMurray DN, Lupton JR. Mechanisms by which docosahexaenoic acid and related fatty acids reduce colon cancer risk and inflammatory disorders of the intestine. Chem Phys Lipids 2008; 153:14–23. Gleissman H, Johnsen JI, Kogner P. Omega-3 fatty acids in cancer, the protectors of good and the killers of evil? Exp Cell Res 2010;316: 1365–73. Spencer L, Mann C, Metcalfe M, Webb M, Pollard C, Spencer D, Berry D, Steward W, Dennison A. The effect of omega-3 FAs on tumour angiogenesis and their therapeutic potential. Eur J Cancer 2009;45: 2077–86. Anderson BM, Ma DW. Are all n23 polyunsaturated fatty acids created equal? Lipids Health Dis 2009;8:33. Chaje`s V, Thiebaut AC, Rotival M, Gauthier E, Maillard V, Boutron-Ruault MC, Joulin V, Lenoir GM, Clavel-Chapelon F. Association between serum trans-monounsaturated fatty acids and breast cancer risk in the E3N-EPIC Study. Am J Epidemiol 2008; 167:1312–20. Smith BK, Robinson LE, Nam R, Ma DW. Trans-fatty acids and cancer: a mini-review. Br J Nutr 2009;102:1254–66. Lopez-Garcia E, Schulze MB, Meigs JB, Manson JE, Rifai N, Stampfer MJ, Willett WC, Hu FB. Consumption of trans fatty acids is related to plasma biomarkers of inflammation and endothelial dysfunction. J Nutr 2005;135:562–6.