Obesity is considered to be an epidemic of the 21st century. Its prevalence is increasing and its importance to public health is considerable because it is associated with a higher risk of type 2 diabetes, coronary heart disease and certain cancers, and a shorter life expectancy. Although the mechanism underlying the development of obesity and its comorbidities are not well established, it has been recognised that these clusters of disorders are associated to chronic mild inflammation in which the metabolism of fat tissue plays an important roleReference Yudkin, Stehouwer, Emeis and Coppack1, Reference Bullo, Garcia-Lorda, Megias and Salas-Salvado2.
In recent decades, fat was considered to have a passive role in the development of obesity and was simply stored in the adipocytes. More recently, the adipocyte metabolism has been shown to change during the fat enlargement process associated to obesity. Within these mechanisms, a greater release of free fatty acids (FFA) and glycerol from adipocytes has been described in obese people than in lean individuals, probably promoting insulin resistance and type 2 diabetes through its specific actions that block insulin signal transductionReference Horowitz, Coppack, Paramore, Cryer, Zhao and Klein3. Several procoagulant proteins such as plasminogen activator inhibitor type 1, tissue factor or factor VII and also inducible nitric oxide synthase (iNOS) show higher expression in adipose tissue of obese people in comparison to lean peopleReference Samad, Yamamoto and Loskutoff4–Reference Perreault and Marette7. This overexpression could explain at least a part of the atherogenic and cardiovascular risk associated obesity. White adipose tissue is also specially characterised by the production and secretion of more than 50 adipokines, such as leptin, tumour necrosis factor (TNF) or interleukin-6 (IL6) with characterised local and systemic pro-inflammatory effectsReference Bullo, Garcia-Lorda, Megias and Salas-Salvado2, Reference Fried, Bunkin and Greenberg8, Reference Bullo, Garcia-Lorda, Peinado-Onsurbe, Hernandez, Del Castillo and Argiles9. Several of these molecules are over released into the circulation in obese subjects leading to a low grade of chronic systemic inflammation. Recent data indicate that white adipose tissue in obesity is infiltrated by macrophages, where they apparently function to scavenge moribund adipocytesReference Weisberg, McCann, Desai, Rosenbaum, Leibel and Ferrante10, Reference Cinti, Mitchell, Barbatelli, Murano, Ceresi and Faloia11. Nevertheless, infiltrated macrophages could be the major source of pro-inflammatory cytokine production in obese adipose tissue. Interestingly, weight loss is associated with an improvement in the circulating pro-inflammatory profile and a reduction in the macrophage infiltration of white adipose tissueReference Cancello, Henegar, Viguerie, Taleb, Poitou and Rouault12.
Although knowledge about the interplay of inflammation and obesity is still being gained, the available evidence indicates that dietary intervention will become an integral part of future approaches to prevent and treat obesity, the metabolic derangements associated with obesity, the metabolic syndrome and ultimately the cardiovascular disease associated with the atherosclerotic processes. For this reason, in this review we analyse the importance of a healthy diet if these metabolic alterations are to be prevented.
Adipose tissue and inflammation
Since the discovery of leptin in 1994, the perspective on the physiological role of white adipose tissue has dramatically changed and it is now considered to be a major endocrine organ with an important role in the regulation of energy intake and metabolism. Indeed, as explained above, adipose tissue secretes a multiplicity of factors commonly termed adipokines with different protein structures and functions: for example, cytokines or related-proteins (leptin, TNF, IL-6, neurotrophins), chemoattractant proteins (monocyte chemotactic protein-1 (MCP-1)), proteins of the complement system (adipsin), proteins involved in the regulation of blood pressure, vascular haemostasis or angiogenesis (angiotensinogen, plasminogen activator inhibitor-1 (PAI-1), vascular endothelial growth factor (VEGF)), and molecules involved in the glucose and lipid metabolism (adiponectin, resistin, visfatin, cholesteryl ester transfer protein (CETP)). The functional homology between adipocytes and cells from the immune system reinforces the previously observed relationship between adipose tissue and immunityReference Stallone13. In addition to cytokine secretion, many other features have been observed to be common to adipocyte and monocyte/macrophage lines: for example, phagocytic and microbicidal activitiesReference Cousin, Munoz, Andre, Fontanilles, Dani and Cousin14, and possibly a cellular plasticity of adipose precursorsReference Charriere, Cousin, Arnaud, Andre, Bacou and Penicaud15.
The theoretical role of adipose tissue in the development of local or systemic inflammatory status, especially in obesity, could be complicated by its heterogeneity at the cellular level. Mature adipocytes are not more than half of the total cell content of white fat. The remaining cell components contain pre-adipocytes, fibroblast, endothelial cells and macrophagesReference Weisberg, McCann, Desai, Rosenbaum, Leibel and Ferrante10. It is well-documented that during adipose tissue growth, for example in obesity, there is an increase in the size and number of mature adipocytes differentiated from progenitor cells like preadipocytes present in the stroma-vascular fraction. However, this phenomenon seems to be reversible during adult life and the relation between the number of mature adipocytes and preadipocytes can changeReference Zhang, Berger, Hu, Szalkowski, White-Carrington and Spiegelman16. Moreover, infiltration of macrophages from peripheral blood has been reported during fat mass expansion in obesityReference Weisberg, McCann, Desai, Rosenbaum, Leibel and Ferrante10, Reference Xu, Barnes, Yang, Tan, Yang and Chou17. Although the molecular mechanisms responsible for these microphagic infiltrations have not yet been elucidated, some adipokynes such as MCP-1 and leptin can favour the diapedesis of macrophages from circulation to adipose tissueReference Curat, Miranville, Sengenes, Diehl, Tonus and Busse18. Macrophages that infiltrate adipose tissue seem to be responsible for almost the total amount of TNFα and a significant part of IL-6 produced by this tissueReference Weisberg, McCann, Desai, Rosenbaum, Leibel and Ferrante10. Factors secreted by infiltrated macrophages inhibit the differentiation of preadipocytes into mature adipocytes probably because the extracellular matrix is remodelledReference Lacasa, Taleb, Keophiphath, Miranville and Clement19. So, the real capacity of adipose tissue to produce and release inflammatory proteins may be related to its cellular composition, and depends on the degree of obesity.
Obesity, metabolic syndrome and inflammation
Overweight and obesity are associated with an increased risk of such metabolic abnormalities as dyslipidemia, hypertension or type 2 diabetes mellitus and cardiovascular diseases – common features of the metabolic syndrome. Initially, insulin resistance or hyperinsulinemia was suggested as the origin of these abnormalitiesReference Reaven20. Nowadays, more recent studies indicate that adipokynes have an important role in obesity-associated metabolic complications, and suggest that chronically elevated local or systemic concentrations of adipokynes contribute to the development of complications associated with obesity and metabolic syndrome.
TNFα is a potent cytokine that induces the production of IL-6Reference Bullo, Garcia-Lorda, Peinado-Onsurbe, Hernandez, Del Castillo and Argiles9, which is the major determinant of the acute-phase responseReference Zhang, Berger, Hu, Szalkowski, White-Carrington and Spiegelman16, Reference Xu, Barnes, Yang, Tan, Yang and Chou17 and is also produced by adipose tissue. Plasma levels of TNF are persistently elevated in obese or type 2 diabetic patients and among patients at increased risk of recurrent coronary eventsReference Bullo, Garcia-Lorda, Peinado-Onsurbe, Hernandez, Del Castillo and Argiles9, Reference Ridker, Rifai, Pfeffer, Sacks, Lepage and Braunwald21. Its role favoring insulin resistance is mediated by the increase in adipocyte lipolysisReference Souza, Palmer, Kang, Yamamoto, Muliro and Paulson22 and through its role on insulin receptor substrate-1 phosphorylation and modulating glucose transporter type 4 (GLUT4) activityReference Bullo-Bonet, Garcia-Lorda, Lopez-Soriano, Argiles and Salas-Salvado23. Because TNF is the first step in the inflammatory cascade, it is not surprising that plasma levels of IL-6 are high in obese or diabetic subjectsReference Fried, Bunkin and Greenberg8. Like TNF, IL-6 concentrations correlate with insulin resistanceReference Kern, Ranganathan, Li, Wood and Ranganathan24, Reference van Hall, Steensberg, Sacchetti, Fischer, Keller and Schjerling25 and increase lipolysis and fat oxidation in humansReference van Hall, Steensberg, Sacchetti, Fischer, Keller and Schjerling25. Therefore, it is not surprising that elevated levels of IL-6 have been associated with an increased risk of type 2 diabetesReference Pradhan, Manson, Rifai, Buring and Ridker26 and future myocardial infarction in healthy menReference Ridker, Rifai, Pfeffer, Sacks, Lepage and Braunwald27.
One of the most important effects of IL-6 is that it controls hepatic C-reative protein (CRP) production, one of the most important inflammatory markers in humans, also positively associated with degree of obesity, fasting glucose and insulin, blood pressure and lipid profileReference Pradhan, Manson, Rifai, Buring and Ridker26. CRP induces the synthesis of other cytokines, cell adhesion molecules such as vascular cell adhesion molecule (VCAM) or intercellular adhesion molecule (ICAM), and tissue factor, and provides a link between inflammation and atherosclerosisReference Gabay and Kushner28. Elevated CRP levels have been described as an independent predictor of risk of type 2 diabetes, myocardial infarction, stroke and peripheral arterial disease even in healthy subjectsReference Pradhan, Manson, Rifai, Buring and Ridker26, Reference Torres and Ridker29. Other inflammatory markers, such as blood cell count, fibrinogen, ESR, inflammation-associated haemostatic variables and low serum albumin, have also been associated with obesity and its metabolic derangementsReference Dixon and O’Brien30.
In obesity, the decrease in production and expression of adiponectin, a molecule with an anti-inflammatory action, exacerbates the important role of inflammation in the obese stateReference Arita, Kihara, Ouchi, Takahashi, Maeda and Miyagawa31, Reference Bulló, Salas-Salvado and Garcia-Lorda32. Plasma adiponectin concentrations are lower in patients who have type 2 diabetes or coronary heart disease than in controls, even when body mass index or age are matchedReference Hotta, Funahashi, Arita, Takahashi, Matsuda and Okamoto33, Reference Ouchi, Kihara, Funahashi, Matsuzawa and Walsh34. Adiponectin may affect insulin sensitivity because it acts on muscle fatty acid oxidation and hormone-sensitivity lipaseReference Bullo, Garcia-Lorda, Megias and Salas-Salvado2, Reference Tomas, Tsao, Saha, Murrey, Zhang and Itani35. The stimulation of nitric oxide production and the reduction in the expression of adhesion molecules in endothelial cells explains the anti-hypertensive and anti-atherogenic properties of adiponectin.
Several clinical studies have demonstrated that it is abdominal fat distribution, not the amount of fat deposition, which is better associated with multiple metabolic risk factors. In fact, epidemiological studies have suggested that abdominal fat distribution is a significant predictor for coronary artery disease independent of body mass indexReference Larsson, Svardsudd, Welin, Wilhelmsen, Bjorntorp and Tibblin36. Quantitative and qualitative changes in serum lipids and lipoproteins such as small dense low-density lipoproteins (LDL), hypertriglyceridemia or the hyperinsulinemic state are more closely related to visceral than subcutaneous fat. Furthermore, differences in adipokyne expression and production between visceral and subcutaneous fat depots have also been reported. For example, TNF is significantly overproduced in visceral adipose tissue. Because this cytokine is a strong inhibitor of adiponectin promoter activityReference Maeda, Takahashi, Funahashi, Kihara, Nishizawa and Kishida37, it may explain the negative correlation observed between visceral adiposity and adiponectin levels. It has been speculated that approximately 20% of all genes in subcutaneous adipose tissue encode secretory proteins, and increase visceral fat by about 30%Reference Matsuzawa38. Recently, a visceral-fat-specific adipocytokine, visfatin, was describedReference Fukuhara, Matsuda, Nishizawa, Segawa, Tanaka and Kishimoto39. Plasma visfatin levels strongly correlated with the amount of visceral adipose tissue estimated by computed tomographyReference Fukuhara, Matsuda, Nishizawa, Segawa, Tanaka and Kishimoto39. Higher visfatin plasma levels have also been observed in patients with type 2 diabetes mellitusReference Hammarstedt, Pihlajamaki, Rotter Sopasakis, Gogg, Jansson and Laakso40 or in those with rheumatoid arthritisReference Chen, Chung, Chang, Tsai, Huang and Shin41. Although its physiological significance is not fully understood, visfatin mimics the effect of insulin and has a potent activity on adipogenesis.
Although knowledge about the relationship between inflammation and obesity is increasing, whether the inflammatory state is the cause or the consequence of obesity remains to be elucidated. It has been recently proposed that local hypoxia during white adipose tissue mass expansion is the first sign that the production of inflammatory adipokines and related substances is being stimulated. Hypertrophic adipocyte expansion during the development of obesity is associated with low oxygen tension and metabolic adaptation to hypoxiaReference Fleischmann, Kurz, Niedermayr, Schebesta, Kimberger and Sessler42, Reference Kabon, Nagele, Reddy, Eagon, Fleshman and Sessler43. Within the adaptative mechanism, the overexpression of hypoxia-inducible genes such as VEGF, erythropoietin (EPO) or hypoxia-inducible factor (HIF-1) has been described in the adipose tissue of human obese subjects. This overexpression decreases after weight lossReference Cancello, Henegar, Viguerie, Taleb, Poitou and Rouault12. It has been suggested that these molecules may stimulate the release of inflammatory cytokines and other inflammatory substancesReference Lolmede, Durand de Saint Front, Galitzky, Lafontan and Bouloumie44, Reference Segawa, Fukuhara, Hosogai, Morita, Okuno and Tanaka45.
Inflammation and weight loss
Evidence for a connection between obesity and inflammation has also been found in clinical weight loss studies. Since adipose tissue from obese patients overexpresses adipokynes, it seems reasonable that a reduction in adipose mass after weight loss will restore inflammatory levels because adipose tissue expression and secretion will decrease, and insulin resistance and other cardiovascular risk factors will improve. Several studies of weight loss due to dietary intervention or exercise, report a decrease in circulating levels of several peripheral inflammatory markersReference Esposito, Pontillo, Di Palo, Giugliano, Masella and Marfella46, Reference Troseid, Lappegard, Claudi, Damas, Morkrid and Brendberg47. Weight loss due to bariatric surgery or fat removal by liposuction is also associated with an improvement in the inflammatory stateReference Haider, Schindler, Schaller, Prager, Wolzt and Ludvik48. However, there are some discrepancies in terms of adipose tissue adipokyne expression after weight loss. While some studies describe a decrease in TNF or IL-6 adipose tissue expression, others have found an increase or no changesReference Bastard, Hainque, Dusserre, Bruckert, Robin and Vallier49, Reference Viguerie, Vidal, Arner, Holst, Verdich and Avizou50. As has been previously suggested by our group, these discrepancies may be due to the sign of the energy balance at the moment of the study51. While most of the studies describing a decrease in adipokyne expression were made a long time after weight loss, and the weight loss had stabilised, in our study we evaluated subcutaneous adipose tissue adipokyne expression in negative energy balance conditions when lipid metabolism is stimulated. Because several pro-inflammatory cytokines have lipolytic activity, our results were not surprising.
The restoration of inflammatory levels could also be explained by the decrease in the infiltration of macrophage adipose tissue after weight lossReference Weisberg, McCann, Desai, Rosenbaum, Leibel and Ferrante10 and changes in its distributionReference Cancello, Henegar, Viguerie, Taleb, Poitou and Rouault12. Although the mechanisms responsible for macrophage recruitment in white adipose tissue are unknown, such chemotactic factors as MCP-1 showed decreased expression after weight loss. Tissue and cellular hypoxia is a well-known cause of macrophage attraction and retention, particularly in tumours and atherosclerotic plaqueReference Murdoch, Giannoudis and Lewis52. The overexpression of HIF-1α in morbid obesity and its lower expression after weight loss also indicate that local white adipose tissue hypoxia contributes to macrophage movement, as suggestedReference Trayhurn and Wood53.
The role of diet in inflammatory modulation
Inflammation is a normal response to tissue injury or infection. However, several studies suggest that when a source of chronic inflammation exposure persists, a low-grade inflammation response develops, thus increasing the risk of obesity, insulin resistance, diabetes, metabolic syndrome, cardiovascular disease or cancerReference Pradhan, Manson, Rifai, Buring and Ridker26, Reference Ridker54–Reference Danesh, Wheeler, Hirschfield, Eda, Eiriksdottir and Rumley56. Dietary habits are probably one of the key determinants of the balance that influences the overall inflammatory process in chronic conditions. Genetic polymorphisms at multiple sites may alter the ability of bioactive food components to influence the overall process by modulating pro- and/or anti-inflammatory mediators. The modulation of this inflammatory response by diet would be fundamental if such metabolic derangements associated with obesity such as insulin resistance, type 2 diabetes, hypertension, atherogenic dyslipidaemia and metabolic syndrome are to be prevented and treated.
The inflammatory process is initiated by the synthesis and secretion of pro-inflammatory cytokines in response to an inflammatory insult (for example, a non-healthy diet). The increased production of cytokines and the subsequent increase in reactive oxygen and nitrogen species are recognised hallmarks of inflammation. This process is regulated by a negative feedback mechanism and is closely followed by the secretion of anti-inflammatory cytokines to reduce the accumulation of reactive species. The cellular anti-oxidant defence system is also activated to limit the development of chronic inflammation in which the risk of metabolic syndrome and subsequently atherosclerosis is much higher than normal.
The binding of pro-inflammatory cytokines to their receptors triggers the mitogen-activated protein kinase (MAPK) pathway that ultimately results in the activation of two redox-sensitive transcription factors: nuclear factor kappa B (NFκB) and the c-Jun part of activating protein-1 (AP-1). These transcription factors activate the expression of a wide variety of genes including cytokines, chemokines, adhesion molecules, and inducible effector enzymes such as iNOS and cycloxygenase-2 (COX-2).
There is evidence to suggest that such dietary components as eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), butyrate, curcumin, resveratrol, lutein, quercetin and oleuropein can influence the inflammatory process at various sites and thus modulate the balance within the process. For example, the production of IL-6 or IL-1β in THP-1 cellules was reduced after incubation with DHA or ALAReference Zhao, Etherton, Martin, Vanden Heuvel, Gillies and West57. Several phenolic compounds have also been identified as anti-inflammatory compounds; however, phenolic compounds from extra virgin olive oil have been shown to have different anti-inflammatory effects in human whole blood culturesReference Miles, Zoubouli and Calder58. Likewise, a number of dietary components, including resveratrol in red grapes, butyrate generated through the microbial metabolism of dietary fibre in colon, and curcumin in curry spice, have been reported to suppress the TNF-α-induced activation of NF-κB and COX-2 expression in vitro and in vivo. The mechanisms by which these bioactive components influence inflammation are not known with any certainty, but may involve transport, activation or inactivation processes. Resveratrol has been shown to inhibit the translocation of NF-κB from the cytoplasm to the nucleus by modulating the IκB kinase (IKK) activity in various cell types (for example, myeloid, lymphoid and epithelial cells)Reference Manna, Mukhopadhyay and Aggarwal59.
Cereals, fibre and glycaemic response
Fasting and postprandial hyperglycaemia are very frequently observed in obesity. They can reduce the availability of nitric oxideReference Giugliano, Marfella, Coppola, Verrazzo, Acampora and Giunta60, and increase the free radical production, which also activates inflammation by modulating protein kinase C (PKC) and NF-κB functionReference Giugliano, Ceriello and Esposito61. Refined starches and sugars cause a rapid increase in blood glucose and insulin levels, and a subsequent decrease in glycaemia, which leads to hunger and a decrease in fat oxidation. For this mechanism, the intake of foods with a high glycaemic index is associated with hyperglycaemia being a major stimulus for inflammation.
A cross-sectional study performed in 720 diabetic patients between 40 and 75 years old from the Health Profesionals’ Follow-up study supports this hypothesis. In this study, subjects with diets low in glycemic load and high in dietary fibre have higher plasma levels of adiponectin, a molecule with anti-inflammatory properties that protects against diabetes and atherosclerosisReference Qi, Meigs, Liu, Manson, Mantzoros and Hu62. Also recently, in two more cross-sectional studies, plasma levels of CRP have been observed to be negatively related to the total fibre intake in 524 healthy adult subjectsReference Ma, Griffith, Chasan-Taber, Olendzki, Jackson and Stanek63 and in patients with diabetes, hypertension or obesity from the NHANES study64. These associations were maintained when variables were adjusted for other confounding dietary and non-dietary factors. Some interventional studies support the hypothesis that fibre can improve inflammationReference Esposito, Pontillo, Di Palo, Giugliano, Masella and Marfella46. In fact, recent epidemiologic studies suggest that a high consumption of dietary fibre or cereals decreases the risk of diabetes or atherosclerosis and is associated with a decrease in weight gainReference Liu, Buring, Sesso, Rimm, Willett and Manson65–Reference Pereira, O’Reilly, Augustsson, Fraser, Goldbourt and Heitmann67.
Fish and omega-3 fatty acids
One of the most commonly studied dietary compounds is omega-3 (ω3) fatty acids. ω3 fatty acids (especially EPA and DHA) decrease the arachidonic acid content of cell membranes, which results in the synthesis of eicosanoids that have fewer inflammatory properties than those synthesised from omega-6 (ω6) fatty acids. Several studies support the idea that ω3 may play a major role in modulating inflammation associated with diseases and situations.
Several cross-sectional studies support the idea that ω3 fatty acids have anti-inflammatory properties. In a sample of individuals from the Nurses’ Health Study II and the Health Professionals Follow-up Study, Pischon et al.Reference Pischon, Hankinson, Hotamisligil, Rifai, Willett and Rimm68 observed that a high intake of ω3 (EPA and DHA) and a low intake of ω6 (linoleic acid) was associated with lower levels of soluble TNF receptors 1 and 2, indicating a low level of inflammation. This anti-inflammatory effect of EPA and DHA has also been observed in other cross-sectional studies performed in healthy individuals from the Nurses’ Health StudyReference Lopez-Garcia, Schulze, Manson, Meigs, Albert and Rifai69 and the Attica StudyReference Zampelas, Panagiotakos, Pitsavos, Das, Chrysohoou and Skoumas70 or in patients with established coronary artery diseaseReference Madsen, Skou, Hansen, Fog, Christensen and Toft71. Some interventional studies also support the anti-inflammatory effect of the ω3 in humans. For example, dietary fish oil has been observed to decrease CRP and IL-6 in postmenopausal womenReference Ciubotaru, Lee and Wander72. The anti-inflammatory effects of the α-linolenic acid have also been observed in several interventional studiesReference Rallidis, Paschos, Liakos, Velissaridou, Anastasiadis and Zampelas73–Reference Zhao, Etherton, Martin, West, Gillies and Kris-Etherton75.
Trans- and saturated fatty acids
Some observational and interventional studies in humans suggest that diets rich in trans- or saturated fatty acids are more closely associated with inflammation than vegetarian diets, which usually contain fewer of these compounds. For example, subjects consuming a diet rich in trans-fatty acids showed higher levels of CRP, IL-6, sTNFR2, E-selectin, sICAM-1 and sVCAM-1 than those in the low quintile of trans-fatty acid consumptionReference Lopez-Garcia, Schulze, Meigs, Manson, Rifai and Stampfer76. In a randomised cross-over study, CRP and IL-6 increased when 9% of fatty acids were replaced with trans-fatty acid or estearic acidReference Baer, Judd, Clevidence and Tracy77. However, some studies show that intake of trans-fatty acid has no effect on inflammatory parameters. CRP levels increase with trans-fatty acid substitution in a high fat diet in healthy subjects, whereas 6% substitution of trans-fatty acids in a standard fat diet (30% fat) showed no effects on CRP in moderately hypercholesterolemic subjects although TNF and IL-6 levels increasedReference Lichtenstein, Erkkila, Lamarche, Schwab, Jalbert and Ausman78.
Virgin olive oil and nuts
Other important foods with anti-inflammatory properties are virgin olive oil and nuts. Virgin olive oil is a rich source of MUFA, and retains all the lipophilic components of the olive fruit, especially the phenolic compounds with strong anti-oxidant and anti-inflammatory propertiesReference Perez-Jimenez, Alvarez de Cienfuegos, Badimon, Barja, Battino and Blanco79, Reference Fito, Cladellas, de la Torre, Marti, Alcantara and Pujadas-Bastardes80. The administration of olive oil with a high phenolic content has been shown to protect against inflammationReference Martinez-Dominguez, de la Puerta and Ruiz-Gutierrez81. In addition, phenolic compounds derived from extra virgin oil were recently shown to decrease the production of inflammatory mediators in human whole-blood culturesReference Miles, Zoubouli and Calder58 and to inhibit endothelial adhesion molecule expression in vitro Reference Carluccio, Siculella, Ancora, Massaro, Scoditti and Storelli82. These processes may at least partly explain why the Mediterranean-type pattern of food intake is thought to protect against atherosclerosis.
Nuts are rich in unsaturated fatty acidsReference Kris-Etherton, Yu-Poth, Sabate, Ratcliffe, Zhao and Etherton83 and, besides having favourable fatty acid profiles, they are a good source of bioactive compounds with potential benefits against metabolic syndrome and cardiovascular disease. Nuts are rich in fibreReference Salas-Salvado, Bullo, Perez-Heras and Ros84, phenolic compounds and other anti-inflammatory and anti-oxidant moleculesReference Halvorsen, Holte, Myhrstad, Barikmo, Hvattum and Remberg85. Also, nuts contain sizeable amounts of l-arginine, a precursor of the endogenous vasodilator nitric oxideReference Cooke, Tsao, Singer, Wang, Kosek and Drexler86, Reference Fitzpatrick, Hirschfield, Ricci, Jantzen and Coffey87. Walnuts differ from all other nuts in that they have a high α-linolenic acid contentReference Hepburn, Exler and Weihrauch88, Reference Zwarts, Savage and McNeil89 which also has anti-inflammatory properties. A high consumption of nuts and seeds has recently been shown to be associated with lower concentrations of inflammatory markers in the MESA atherosclerosis studyReference Jiang, Jacobs, Mayer-Davis, Szklo, Herrington and Jenny90. Also, clinical trials of nut consumption have reported decreases in inflammatory markersReference Zhao, Etherton, Martin, West, Gillies and Kris-Etherton75 and improvements in the endothelial functionReference Ros, Nunez, Perez-Heras, Serra, Gilabert and Casals91. This mechanism has been proposed as a potential explanation for the established cardio-protective effect of frequent nut intakeReference Nash and Westpfal92.
Fruit and vegetables
An increase in the consumption of fruit and vegetables has also been advocated for the prevention of strokes and other cardiovascular eventsReference He, Nowson and MacGregor93. Fruit and vegetables are very rich in anti-oxidants and other anti-inflammatory phytochemicals. Plasma CRP and homocysteine concentrations have been related to frequent fruit and vegetable intake in the Massachusetts Hispanic Elders Cross-sectional Study after adjusting for confounding variables. Moreover, a significant relationship between vitamin C or β-carotene and CRP levels has been observed in healthy subjectsReference van Herpen-Broekmans, Klopping-Ketelaars, Bots, Kluft, Princen and Hendriks94. Several interventional studies have demonstrated an anti-inflammatory effect of fruit and vegetable or some of their individual compoundsReference Watzl, Kulling, Moseneder, Barth and Bub95–Reference Sanchez-Moreno, Cano, de Ancos, Plaza, Olmedilla and Granado97.
Wine and alcohol
Both epidemiological and experimental studies have proposed that moderate consumption of red wine can contribute to promote health and prevent disease probably because of their content of resveratrol or other polyphenolic compounds and flavonoids, and their effect on the modulation of inflammation. A 4-week consumption of 30 g day−1 of red wine led to a significant decrease in the serum concentration of CRP, VCAM and ICAM in healthy adult menReference Estruch, Sacanella, Badia, Antunez, Nicolas and Fernandez-Sola98. Lower levels of circulating CRP have also been observed associated to a moderate consumption of red wine in a cross-sectional study performed on more than 2800 men and women of the PRINCE studyReference Albert, Glynn and Ridker99 or in subjects from the NHANES III studyReference Stewart, Mainous and Gilbert100. Some of the mechanisms that explain the biological activities of resveratrol involving downregulation of the inflammatory response are the inhibition of synthesis and release of pro-inflammatory cytokines, modification of eicosanoid synthesis, inhibition of activated immune cells, or inhibition of iNOS and COX-2 via the inhibitory effects on NFκB or the activator protein-1 (AP-1)Reference Rahman, Biswas and Kirkham101. The effect on NFκB may explain the suppression of the ex vivo production of TNF, IL-1 and IL-6 by mononuclear blood cells observed after the incubation with trans-resveratrolReference Marier, Chen, Prince, Scott, del Castillo and Vachon102. Flavonoids are another of the bioactive compounds of red wine with anti-oxidant and anti-inflammatory properties. Studies performed in cell cultures show that the expression of endothelial adhesion molecules is inhibited, which leads to a reduction in atherosclerotic lesion formationReference Lotito and Frei103.
Moderate alcohol intake has also been associated with beneficial effects on markers of inflammation, independently of the bioactive components of grapesReference Rimm and Stampfer104. This effect, however, is controversialReference Estruch, Coca and Rodicio105.
Healthy dietary pattern
Considering all the evidence relating to diet and inflammation, the best diet for protecting against the metabolic derangements associated with obesity and metabolic syndrome would be high in fibre-rich cereals, fruit, vegetables, fish, virgin olive oil and nuts; moderate in wine; and low in meat, processed meat foods and trans-fatty acids. In fact, in a cross-sectional study from the Nurse’s Health Study I Cohort, a prudent pattern was associated to low levels of some inflammatory and endothelial function markers in comparison to the results from a Western dietary patternReference Lopez-Garcia, Schulze, Manson, Meigs, Albert and Rifai69. Furthermore, the traditional Mediterranean-type diet has been associated to a low risk of cardiovascular diseaseReference Trichopoulou, Costacou, Bamia and Trichopoulos106 probably mediated by inflammation related factorsReference Esposito, Marfella, Ciotola, Di Palo, Giugliano and Giugliano107, Reference Estruch, Martinez-Gonzalez, Corella, Salas-Salvado, Ruiz-Gutierrez and Covas108 although some of the results are controversialReference Michalsen, Lehmann, Pithan, Knoblauch, Moebus and Kannenberg109. Some unpublished results from our group show lower serum concentrations of inflammatory markers, especially those related to endothelial function, in those subjects who consume more fruit, cereals, olive oil and nuts. However, controlled trials analysing the effects of Mediterranean-type diets on inflammation parameters are required if the cross-talk between diet compounds and the modulation of inflammatory state associated with obesity and metabolic syndrome is to be elucidated.
Acknowledgements
Sources of funding:Funded in part by the Ministerio de Educación y Ciencia CICYT–AGL2005-0365 and two grants from the Instituto de Salud Carlos III (PI041828 and PI051839), Madrid, Spain.
Conflict of interest declaration:No authors have any conflict of interest.
Authorship responsibilities:All the authors have participated in the redaction of the manuscript.
