Benefits of Nut Consumption on Insulin Resistance and Cardiovascular Risk Factors: Multiple Potential Mechanisms of Actions

Epidemiological and clinical studies have indicated that nut consumption could be a healthy dietary strategy to prevent and treat type 2 diabetes (T2DM) and related cardiovascular disease (CVD). The objective of this review is to examine the potential mechanisms of action of nuts addressing effects on glycemic control, weight management, energy balance, appetite, gut microbiota modification, lipid metabolism, oxidative stress, inflammation, endothelial function and blood pressure with a focus on data from both animal and human studies. The favourable effects of nuts could be explained by the unique nutrient composition and bioactive compounds in nuts. Unsaturated fatty acids (monounsaturated fatty acids and polyunsaturated fatty acids) present in nuts may play a role in glucose control and appetite suppression. Fiber and polyphenols in nuts may also have an anti-diabetic effect by altering gut microbiota. Nuts lower serum cholesterol by reduced cholesterol absorption, inhibition of HMG-CoA reductase and increased bile acid production by stimulation of 7-α hydroxylase. Arginine and magnesium improve inflammation, oxidative stress, endothelial function and blood pressure. In conclusion, nuts contain compounds that favourably influence glucose homeostasis, weight control and vascular health. Further investigations are required to identify the most important mechanisms by which nuts decrease the risk of T2DM and CVD.


Introduction
The prevalence of diabetes mellitus will increase worldwide from 382 million individuals in 2013 to 592 million individuals in 2035. Type 2 diabetes mellitus (T2DM) accounts for 90-95% of individuals with diabetes mellitus [1]. T2DM is attributable to poor diet, obesity, age, genetics, smoking, sedentary lifestyle and hypertension [2,3]. T2DM is associated with an increased risk of adverse cardiovascular events [3]. A 1 mmol/L increase in fasting plasma glucose is related to a 17% increase in the risk of developing of cardiovascular disease (CVD) and death [4]. Individuals with diabetes have a 2-to 4-fold higher risk of CVD than individuals without diabetes [5]. As a consequence, T2DM dramatically increases health care costs and disease burden [6]. Lifestyle modification including diet can decrease the risk of developing T2DM [7]. A large number of epidemiological and clinical studies have shown that consumption of different types of nuts (almonds, walnuts, hazelnuts, pecans, pistachios, macadamia nuts, cashews and Brazil nuts) is associated with a reduced risk of T2DM [8][9][10] and CVD [8][9][10][11][12].
This review aims to address the potential mechanisms of nut consumption on the prevention and treatment of T2DM and CVD, focusing on body weight control, glucose control, gut microbiota composition, inflammation, oxidative stress, lipid profiles, endothelial function and hypertension. All human interventions available on PubMed that covered these areas were included. This is the only systematic review that covers all these areas.      The negative study [37], which was included in the meta-analysis of T2DM subjects [20], consisted of two studies to assess insulin sensitivity. Both studies showed no effect on insulin sensitivity in healthy subjects or on glycemic control in subjects with T2DM, but beneficial effects were seen on lipid profiles. In study 1, 20 healthy adults were instructed to adhere to their habitual diet supplemented with 100 g almonds/day for 4 weeks. No changes in fasting glucose and fasting insulin, and insulin sensitivity index (SI) and glucose effectiveness (SG) as calculated using the Minimal Model method were observed before and after the diet. Body weight significantly increased (p = 0.006) and a significant time-by-sex interaction for SI and SG was observed due to an increase in women with a decrease in men. Changes in body weight had no effect on insulin sensitivity. In study 2 using a randomised, double-blind, crossover design, volunteers were randomly assigned to begin one of four diets for 4 weeks: high-fat, high-almond (37% total fat, 10% from almonds); low-fat, high-almond (25% total fat, 10% from almonds); high-fat control (37% total fat, 10% from olive oil or canola oil); low-fat control (25% total fat, 10% from olive or canola oil). The almond diets contained 57-113 g almonds/day based on the total energy requirements. No changes in fasting glucose, fasting insulin, and 2-h glucose and insulin as assessed by oral glucose tolerance test were observed overall with almonds, while a significant interaction between fat source and fat level for both fasting and 2-h glucose levels was observed due to the lower glucose levels on the low fat, high almond diet than a high fat control diet [37].

Possible Nutrients Involved in Glucose-Lowering Effects in Nuts
In summary, unsaturated fatty acids in nuts in place of both carbohydrate and saturated fat appear to improve glucose levels by a small amount but more studies at the molecular level should be undertaken along with a detailed investigation of different nut types and amounts, study duration, background diet, subject ethnicity in subjects with T2DM and at risk of T2DM.

Modification in Micro RNAs Related to Insulin Sensitivity
Only two nut intervention studies have explored microRNA modification in relation to T2DM risk [46,47]. Ortegaa et al. [46] found significant modification of several common miRNAs in plasma from 10 healthy women who consumed a PUFA-enriched normocaloric diet containing 15 g/day of almonds and 15 g/day of walnuts (55-60% carbohydrates 15% proteins, 30% fat, 10% SFA, 10-15% MUFAs and 10% PUFAs) for 8 weeks. The changes of plasma miR-106a were correlated with changes in circulating PUFAs. The changes in plasma miR-130b and miR-221 (r = 0.46, p = 0.03) were correlated with changes in plasma C-reactive protein. The changes in plasma miR-125a-5p were correlated with changes in plasma fasting triglycerides and adiponectin [46]. Very recently, a 4-month diet supplemented with 57 g/day of pistachio (50% carbohydrates, 33% fat) modified circulating microRNAs (miR-192 and miR-375) and these changes were positively correlated with plasma glucose, insulin and HOMA-IR in 49 subjects with prediabetes [47]. Thus, microRNAs can be modulated by nut consumption and may be involved in glucose metabolism, lipid metabolism and inflammation. Circulating miRNAs have stable and reproducible levels in serum and may be useful biomarkers examining molecular mechanisms in T2DM and CVD [48].
The PREDIMED study showed that adherence to the Mediterranean diets enriched with either 50 mL/day (350 mL/week) of extra virgin olive oil or 30 g/day of nuts for 3 years resulted in a reduction in body weight. The increased energy density was not associated with body weight gain [78]. Other short-term interventions also showed a greater weight loss effect of a diet enriched in nuts in comparison with a low-fat diet [79,80], or a calorie-matched complex carbohydrate diet [26]. In a randomized controlled trial comparing the effect of a moderate fat low energy diet (fat: 35% of energy; predominantly PUFAs), with a low-fat, low-energy diet (fat: 20% of energy) on weight loss in 101 overweight subjects for 18 months, the nut-enriched diet resulted in greater sustained weight reduction at 18 months compared with a low-fat diet [79]. Supplementation with either a high dose (70 g/day) or a recommended dose of pistachio nuts (42 g/day) for 12 weeks in a parallel design did not alter BMI or waist-to-hip ratio compared with a control diet in Chinese subjects with metabolic syndrome [73].
In summary, it would appear that nuts can be added to a diet without significant body weight gain under most circumstances.

Appetite
Two studies showed a favourable effect on appetite of almond snacks in subjects at increased risk of T2DM [81] and in healthy women [82]. In a 4-week randomized, parallel-arm study, 43 g almonds/day with breakfast or lunch, or alone as a morning or afternoon snack for 4 weeks, attenuated postprandial glucose responses and decreased hunger and desire to eat in subjects with increased risk for T2DM [81].
Cholecystokinin (CCK), GLP-1 and peptide YY (PYY) are gut satiety hormones. Cholecystokinin (CCK-8) is released from duodenal enteroendocrine cells in response to fatty acid and protein. GLP-1 and PYY are secreted from the ileum in response to carbohydrates, proteins and fatty acids. Ghrelin is also a gut hunger hormone released from the stomach which is suppressed by food intake. It increases appetite [83,84].
A favourable acute and second meal effect of added nuts was seen with almonds ingested by subjects with impaired glucose tolerance [31] and peanuts consumed by obese women with a high T2DM risk [29] with breakfast on postprandial glucose responses and appetite.
Leptin is a hormone secreted from adipocytes which regulates energy balance by suppressing appetite along with other metabolic effects. Leptin plays an opposite role to that of ghrelin [86]. A recent meta-analysis of 20 RCTs showed that nut consumption (tree nut, peanut, and soy nut) significantly reduced leptin but did not significantly alter high sensitivity C-reactive protein (hs-CRP), IL-6, adiponectin, IL-10, and tumor necrosis factor-α (TNF-α) [87].
The physical texture of nuts may also affect satiety through the enhanced mastication required [88,89].

Bioaccessibility of Nutrients
Bioaccessibility refers to the proportion of nutrients released from a complex food matrix, which become potentially available for absorption in the gastrointestinal tract [92]. One proposed mechanism of low or no weight gain with nuts includes a reduced level of lipid bioaccessibility from nuts. The properties of cell walls and fibre-rich skins in nuts may play a role in the rate and extent of lipid release [93].
Several studies [93,94] have investigated this question. In the first study [93], only the first layer of cells at the fracture surface was ruptured by chewing and released lipids. In faecal samples from healthy subjects who consumed the almond rich diet, intact almond tissues, intact cotyledon cells containing intracellular lipid, increased lipid concentrations, and large numbers of bacteria adherent to the cotyledon cells were observed [93]. Cell walls of almonds consist of non-starch polysaccharides (especially, arabinose-rich polysaccharides) and phenolic compounds (predominantly, protocatechuic acid, p-hydroxybenzoic acid, and vanillic acid).
The poor lipid bioaccessibility of almonds reduces postprandial lipidemia which may be associated with a lower risk of coronary heart disease [99][100][101]. A randomised crossover study [102] in healthy human subjects compared 54 g fat in four different meals (whole almond seed, almond oil and defatted almond flour, or a sunflower oil blend), and showed a favourable effect of whole almond seed on postprandial lipidemia compared with other forms of lipid. However, no significant differences in insulin, 8-isoprostane F2 α (a marker of oxidative stress) and peripheral augmentation index (a measure of vascular tone) were seen between meals. The iAUC for plasma glucose, however, was higher on the whole almond seed meal than on the almond oil and defatted almond flour meal [102].

Gut Microbiota Modification
Fiber and polyphenols (flavonoids and nonflavonoids) abundant in nuts may exert a prebiotic effect and influence glucose metabolism [103]. Nuts with the highest contents of polymerized polyphenols are hazelnuts and pecans followed by pistachios, almonds and walnuts [104]. Walnuts, pecans and chestnuts are high in ellagitannins (hydrolysable tannins), whereas proanthocyanidins (condensed tannins) are abundant in the kernel of nuts such as hazelnuts, pecans, pistachios, almonds and cashew [19].
Proanthocyanidins are oligomers or polymers of polyhydroxy flavan-3-ol units, such as (−)-epicatechin, and (+)-catechin [105]. Epidemiological studies have shown an inverse association between proanthocyanidin intake and incident T2DM [106,107]. No dietary intervention examining the prebiotic effects of nut-derived proanthocyanidin has been performed while two dietary interventions have shown a favourable prebiotic effect of proanthocyanidin derived from grape seeds [108] and cocoa [109] modifying the proportion of Bifidobacterium and Enterobacteriaceae [108,109] which was not dose related. The doses of proanthocyanidins administered were 190 mg/day for healthy adults aged 40-59 years for 2 weeks [108], 380 mg/day for elderly adults aged 67-98 years for 2 weeks [108], and 494 mg/day for young healthy adults with a mean age of 30 years for 4 weeks [109].

Urolithin (Microflora Metabolite)
Walnuts and pecans which have the highest total phenol contents are the richest sources of ellagic acid that possesses 2,2-diphenyl-1-picrylhydrazyl (DPPH•) scavenging capacities [19]. Brazil nuts contain no polyphenols nor have an antioxidant capacity [19]. Ellagitannins and ellagic acid are not absorbed in the gastrointestinal tract but are metabolized by the human gut microbiota to urolithins (dibenzo[b,d]pyran-6-one derivatives with different hydroxyl substitutions) which are much better absorbed than the parent ellagitannins and ellagic acid [110] while the type of urolithin formed depends on particular species in the gut microflora [111]. There is a wide range of urolithins based on a decreasing number of phenolic hydroxyl groups; urolithin D→C→A→B [112]. Aromatic and phenolic structures including phenylvalerolactones and phenylvaleric, phenylpropionic, phenylacetic, hippuric, and benzoic acids, with different hydroxylation patterns which are microbial metabolites of flavan-3-ol dimers, oligomers, flavan-3-ol polymers [113][114][115][116]. Hydroxyphenylvalerolactones in plasma or urine are proposed to be the main metabolites of flavan-3-ol compounds as they come from reactions of the flavonoid C-ring-opening followed by lactonization [113,114]. In vitro and in vivo studies have shown beneficial effects of urolithins on oxidation, inflammation and glycation [110,117].
In a randomized parallel intervention in subjects with the metabolic syndrome comparing a 12-week healthy diet supplemented with 30 g/day mixed nuts (15 g walnuts, 7.5 g almonds and 7.5 g hazelnuts) with a healthy diet without nuts [118][119][120][121], a healthy diet with added nuts significantly increased walnut ellagitannin-derived urolithins A and B in urine [118,119], significantly improved fasting insulin, insulin sensitivity assessed by HOMA-IR, and IL-6 and significantly decreased DNA damage assessed by urinary 8-oxo-7,8-dihydro-2 -deoxyguanosine (a measure of oxidative stress) [121]. Adjustment for weight loss attenuated the statistical significance of IL-6 (p = 0.08) [120]. No effects were seen on cholesterol levels [120], biomarkers of oxidative stress (plasma oxidized LDL, plasma conjugated diene and urinary 8-isoprostanes), antioxidant capacity or endothelial function determined by non-invasive peripheral artery tonometry [121].
In a study with a metabolomic approach, nut consumption resulted in increased metabolism and excretion of unsaturated fatty acids with increased detection of fatty acid conjugated metabolites and increased excretion of serotonin metabolites in 24-h urine samples [118].
In summary, urolithins are phenolic metabolites which are increased following nut consumption, but to our knowledge, no human intervention study has investigated the prebiotic effect of nut derived urolithins. Therefore, it is unclear what role urolithins play in decreasing the risk of T2DM.

Butyrate (Microflora Metabolite)
A few studies have examined the effect of nut consumption on gut microbiota and the gut metabolome [122][123][124][125]. In vitro, chestnut extract enhanced the viability of Lactobacillus strains in a simulated GI tract [122]. In vitro using a model stomach, small intestine and colon finely ground almond seeds significantly increased the growth of bifidobacteria and Eubacterium rectale with a butyrogenic prebiotic effect [123].
Butyrate is a short chain fatty acid (SCFAs). SCFAs stimulate the expression of peptide YY via the G-protein-coupled receptors (Gpr41 and Gpr43) leading to the inhibition of gut motility and suppression of appetite [126,127]. Activation of G-protein-coupled receptors (Gpr43; and Gpr41) on L-cells by SCFAs triggers the secretion of GLP-1 which improve glucose homeostasis by increasing insulin and decreasing glucagon secretion [128,129]. SCFAs play a role in controlling body energy utilization by suppressing insulin signalling in adipocytes and fat accumulation in adipose tissues through Gpr43 activation [130]. Intestinal gluconeogenesis (IGN) may have a favourable impact on the risk of T2DM by decreasing food intake, body weight, hepatic glucose production and plasma glucose [131][132][133]. Butyrate promotes IGN via cyclic adenosine monophosphate (cAMP) in a Gpr41-independent manner, while propionate (a longer SCFA) promotes IGN via a gut-brain neuronal circuit by binding to Gpr41 [134].
Ukhanova et al. [124] conducted two separate randomised, controlled, crossover nut feeding studies in healthy subjects (n = 18 for an almond group; n = 16 for a pistachio group) who consumed 0, 43 g and 86 g per day for 18 days. Analyses of faecal samples using a 16S rRNA-based approach showed that nuts altered the composition of the faecal bacterial and fungal microbiota with an increase in butyrate producing bacteria. This effect was much stronger in the pistachio group. Neither nut altered Lactobacillus or Bifidobacteria levels [124]. An intervention with a much longer period (6 weeks) in healthy subjects indicated that whole almond or almond skin could beneficially modify gut microbiota and bacterial activities [125]. In this study [125], daily consumption of 56 g roasted, unsaltednwhole almonds or 10 g almond skin significantly increased populations of Bifidobacterium spp. and Lactobacillus spp., and suppressed the growth of Clostridium perfringens, leading to changes in bacterial enzyme activities (increased faecal β-galactosidase and decreased faecal β-glucuronidase, nitroreductase and azoreductase) [125].
In summary, butyrate is a major product of a nut-related change in the microbiota. Butyrate appears to be a strong candidate molecule which can be mechanistically linked to a reduced T2DM risk.

Changes in Blood Lipids and Lipoproteins
Nut consumption can favourably alter blood lipids and lipoproteins. Mediterranean diets enriched with 30 g/day of nuts after 1 year of follow up in subjects with a high CVD risk showed an increase in cholesterol efflux capacity and an increase in the percentage of large HDL-C particles [153].
Meta-analyses of nut intervention studies showed a favourable alteration in lipid profiles and apolipoproteins [9,[154][155][156]. A meta-analysis of 61 interventions with the study period ranging from 3 to 26 weeks [9] showed that nut consumption reduced TC (−4.7 mg/dL), LDL-C (−4.8 mg/dL), apolipoprotein B (apo B) (−3.7 mg/dL), and triglyceride (TG; −2.2 mg/dL) regardless of type of nut or background diets. A favourable effect on Apo B was stronger in individuals with T2DM. An association between tree nut consumption and a reduction in TC and LDL-C was seen at ≥60 g/day of nuts. The LDL-C lowering effect of 100 g/day of nuts was up to 35 mg/dL (0.9 mmol/L) [9]. Coronary deaths may be reduced by 20% (95% CI: 0.76, 0.85) per 1.0 mmol/L reduction in LDL-C [157]. Moreover, a separate meta-analysis of only almonds [154] or only walnut [156] intervention studies showed a significant reduction in TC, LDL-C and TG and no alteration in HDL-C [154] and reduction in TC, LDL-C and no alteration in HDL-C and TG compared with the diet [156]. Incorporation of 50-100 g/day nuts providing 35% of energy to a low-fat diet five or more times per week can lower TC and LDL-C for the prevention of CVD [158][159][160].
In a randomised, cross-over, controlled study, which was not included in the meta-analyses [9,[154][155][156], comparing three 4-week iso-energetic dietary regimes in subjects with an increased risk of CVD, a diet providing 20% of energy from pistachios (63-126 g/day of pistachios; 34% total fat and 8% SFA) significantly reduced small, dense LDL-C levels (sdLDL, associated with more CVD events than total LDL-C) and the triacylglycerol:HDL-C ratio compared with the low-fat control diet (control; 25% total fat and 8% SFA), as well as improved ATP-binding cassette transporter A1 (ABCA1)-mediated serum cholesterol efflux capacity but only in participants with low hs-CRP status at baseline (<10 mg/L), compared with a diet providing 10% of energy from pistachios (32-63 g/day of pistachio; 30% total fat and 8% SFA) [74].
Favourable effects of nut consumption on lipid profiles could in part result from unsaturated fat or bioactive compounds including plant sterols and dietary fibre [158,161]. Meta-analyses of prospective cohort studies showed that PUFAs in nuts (especially in walnuts) and ALA (n-3 PUFA; 18:3n-3) were inversely associated with the risk of CVD [162,163].

Unsaturated Fat
Unsaturated fat in nuts may alter the composition of very low density lipoprotein (VLDL) [164] and the activities of lipoprotein lipase and hepatic lipase [165,166] and decrease apo C-III in HDL.
Park et al. demonstrated that PUFAs decreased the expression of cholesterol transporter NPC1Ll in Caco-2 cells (human colon adenocarcinoma cell line) and HepG2 cells (human liver carcinoma cell line) compared with SFAs and phytosterols [167]. Harini et al. [168] proposed several potential mechanisms for PUFA to lower lipids. The expression of several genes involved in lipid metabolism can be mediated by PUFAs via nuclear receptors including the nuclear receptors (PPAR), liver X receptor (LXR), and hepatocyte nuclear factor-(HNF)-4α], via NFκB and the transcription factors sterol-regulatory element binding protein (SREBP) [168].

The Lysine to Arginine Ratio
One potential mechanism of the hypolipidemic effects of nuts may be the low content of lysine and the relatively high content of arginine. In animals, a diet with a low Lys: Arg ratio promoted 7-α hydroxylase activity resulting in elevated bile acid production and excretion of neutral and acidic steroids decreasing the hepatic cholesterol pool size. This can result in a decrease in LDL-C [169,170].

Phytosterols
A meta-analysis of RCTs showed that a daily intake of approximately 3 g phytosterols lowered LDL-C concentrations by 12% [171]. Possible molecular mechanisms have been proposed but they are still controversial [172]. Phytosterols are plant-specific phytochemicals and their chemical structure is similar to that of cholesterol. Phytosterols have been known to compete against dietary cholesterol and biliary cholesterol incorporation into mixed micelles in the intestinal lumen [172][173][174].
Intestinal transport proteins and receptors have also been proposed as a mechanism underlying the cholesterol-lowering effect of phytosterols [172]. The NPC1L1 (a target of ezetimibe which lowers cholesterol uptake) is a protein expressed in jejunal enterocytes and localized to the brush border membrane and which is required for intestinal cholesterol absorption and phytosterol absorption [175,176].
On the other hand, the heterodimer of ATP-binding cassette (ABC) transporters, G5 (ABCG5) and G8 (ABCG8) inhibit the absorption of cholesterol and phytosterols from the diet by driving the efflux of cholesterol and phytosterols from enterocytes back into the intestinal lumen, and by facilitating the release of cholesterol and phytosterols from hepatocytes into bile [177].
Intestinal acyl CoA: cholesterol acyltransferase (ACAT2) converts cholesterol to cholesteryl ester which is packed into chylomicrons by microsomal triglyceride protein (MTP-a lipid transfer protein) for secretion into chyle [178]. In hamsters fed a high cholesterol diet, β-sitosterol and stigmasterol down-regulated mRNA levels of intestinal ACAT2 and MTP [179]. Brauner et al. [180] demonstrated that phytosterols have lower uptake and lower esterification by ACAT2 in the intestine than cholesterol. In Caco-2 enterocytes, phytosterols decreased cholesterol absorption by suppressing 27-hydroxycholesterol generation by CYP27, liver X receptor α (LXRα) and the basolateral sterol exporter ABCA1 expression, and delivering sterol back into the gut lumen by ABCG5/G8 [180].
3-hydroxy-3-methylglutary-CoA (HMG-CoA) reductase is the rate-limiting enzyme in cholesterol synthesis. HMG-CoA reductase gene expression was reduced in CaCo-2 cells incubated with beta-sitosterol [181]. An in vitro study showed that stigmasterol, campesterol and β-sitosterol decreased secretion of apo B48 in Caco2 human intestinal cells by 15%, 16% and 19% respectively. Moreover, those three phytosterols led to a 30% reduction in VLDL levels as measured by secretion of apo B100 in HepG2 human liver cells [182].
In a placebo-controlled, crossover feeding trial, phytosterol intakes at moderate (458 mg/day) and high (2059 mg/day) doses significantly increased total faecal cholesterol excretion and biliary cholesterol excretion and decreased intestinal cholesterol absorption [183].

Fiber
Meta-analyses of prospective cohort studies showed that total dietary fibre consumption (average of 7 g/day) was inversely associated with the risk of CVD [184,185]. Insoluble fiber in nuts increases satiety and faecal bulk. Soluble fiber (viscous fiber) decreases gastric emptying and impairs diffusion across the unstirred water layer in the small intestine and increases bile acid excretion [186]. These favourable effects of fiber may partly contribute to the cholesterol-lowering effects of nuts.

Antioxidant Activity and Decreased Oxidative Stress
Nuts have very different antioxidant capacities varying from 1.2 to 120 mg of Trolox equivalents per 100 g fresh weight [19]. In a systematic review of in vitro, in vivo and human intervention studies, most nuts showed favourable effects on oxidation while walnuts showed inconsistent antioxidant effects [187]. The PREDIMED study showed that adherence to the Mediterranean diets enriched with either an average of 50 mL/day of virgin olive oil or 30 g/day of nuts for 5 years in subjects with metabolic syndrome decreased xanthine oxidase (an enzymatic source of reactive oxygen species (ROS)) activity and increased plasma SOD and catalase activities compared with a low-fat diet [188].
Walnuts have the highest content of polyphenols in both raw and roasted nuts. Walnut polyphenols have the highest antioxidant efficacy compared with other nuts. Walnuts, almonds, Brazil, cashews, hazelnuts, macadamias, peanuts, pecans, pistachios and peanut butter have greater antioxidant power compared with α-tocopherol ranging from 15 times for raw walnuts to 1.9 times for roasted cashews [14].
Walnut extracts consisting of ellagic acid monomers, polymeric ellagitannins and other phenolics, principally nonflavonoid compounds inhibited Cu 2+ -induced LDL oxidation by 84% while ellagic acid alone at the same level inhibited Cu 2+ -induced LDL oxidation by 14% [196]. Approximately 120 phenolic compounds have been identified in walnuts [214]. Almond polyphenols and α-tocopherol synergistically inhibited Cu 2+ -induced LDL oxidation [198,204]. A decrease in Cu 2+ -induced LDL oxidation was also observed in clinical studies following 4 weeks of almond consumption (60 g/day) in men with mild hyperlipidemia [215] and in Chinese patients with T2DM [216] but not in normocholesterolemic healthy subjects who consumed 66 g/day of almonds for 6 weeks [65].
Walnuts and cashews (63-108 g/day for each nut) significantly improved antioxidant status with a reduction in oxidized glutathione (GSSG) and an increase in oxygen radical absorbance capacity (ORAC) levels but only in comparison with baseline and not the control diet [206]. An acute study examined postprandial oxidative stress compared a walnut meal containing 90 g walnuts and a isocaloric control meal (50% carbohydrate; 20% protein; 30% fat) in healthy young subjects. The walnut meal reduced oxidative stress by significantly increasing iAUC for hydrophilic and lipophilic ORAC and decreasing iAUC for MDA over the course of a 5-h test. There was decreased oxidized LDL at 2 h after a walnut meat. The walnut meal also significantly elevated plasma levels of epicatechin gallate (ECG), epigallocatechin gallate (EGCG) and gallocatechin gallate (GCG) 1 h after a walnut meal. A significant amount of urolithin-A in urine samples was detected after the walnut meal. A diet effect for plasma γ-tocopherol but not for α-tocopherol was observed in the mixed linear models [211].    AOP, antioxidant potential index; 8-OH-dG, urinary 8-hydroxy-2 -deoxyguanosine; FRAP, ferric reducing antioxidant power; GSH-Px, glutathione peroxidase; GSSG, oxidized glutathione; MDA, malondialdehyde; ORAC, oxygen radical absorbance capacity; PGI2, prostacyclin; SOD, superoxide dismutase; TAP, total antioxidant performance; TXB2, plasma thromboxane; ↑, increase; ↓, decrease; ↔, no effect.
An acute study showed that the glucose and insulin-lowering effects of almonds were also related to a lower risk of oxidative damage to proteins [32]. An almond meal improved postprandial protein damage by increasing serum protein thiol (-SH) levels compared with control meals (bread, potato and rice meals containing similar amounts of carbohydrate, protein and fat) when data from three control meals were combined. There were negative correlations between postprandial changes in protein thiols and 2-h glucose iAUC or 2-h insulin iAUC or glucose peak height or insulin peak height, indicating that blood glucose levels increase as damage to protein thiols increases [32].
Four-week pecan [212] and 3-week pistachios [71] consumption in healthy individuals also showed decreased MDA concentrations. Pistachio diets (32-63 or 63-126 g/day) consumed by hypercholesterolemic subjects for 4 weeks significantly reduced serum oxidized-LDL levels compared with a low-fat control diet and elevated plasma lutein, α-carotene, and β-carotene levels compared with baseline [210]. Daily supplementation of two Brazil nuts added to a normal diet for 12 weeks increased plasma selenium concentrations and GPx activities compared with the normal diet [203]. In addition, the consumption of pistachios (65-75 g/day) for 3 weeks and hazelnuts (1 g/kg body weight/day) for 30 days in healthy subjects [71,77] and macadamias (40-90 g/day) for 4 weeks in hypercholesterolemic subjects [207] enhanced oxidative status but these studies had the limitation that the amount of SFA in the nut diet was lower than that in a control diet. Nuts also reduce oxidative DNA damage [121,205,208].
However, whether the antioxidant effects of nuts play a role in their apparent effects in CVD prevention is not clear and more clinical trials of nuts with hard endpoints are required.

Anti-Inflammatory Actions
Elevated levels of pro-inflammatory markers (e.g., hs-CRP, IL-6 and TNF-α) are strong predictors of the development of T2DM [217,218] and CVD [219]. In the Nurses' Health Study (NHS) and the Health Professionals Follow-Up Study (HPFS) of US men and women with no diabetes, nut consumption (≥5 times per week, 28 g per serving) was associated with lower inflammatory markers of hs-CRP (relative concentrations 0.80 (95% CI: 0.69, 0.90); p = 0.0003) and IL-6 (0.86 (95% CI: 0.77, 0.97); p = 0.006) after adjusting for demographic, medical, dietary, and lifestyle variables, compared with no or almost no nut consumption [220]. A sub-study of the PREDIMED study (two Mediterranean diets enriched with either 50 mL/day of virgin olive oil or 30 g/day of nuts and a control lower fat Mediterranean diet) investigated methylation changes in inflammation-related genes in peripheral blood cells between baseline and 5 years in 36 subjects at high cardiovascular risks. This study showed a correlation between methylation changes of eight genes (EEF2, COL18A1, IL4I1, LEPR, PLAGL1, IFRD1, MAPKAPK2, and PPARGC1B) related to inflammation and adherence to a Mediterranean diet. A positive correlation between levels of EEF2 methylation and TNF-α and hs-CRP was observed, suggesting that Mediterranean diets might have anti-inflammatory actions possibly via epigenetic mechanisms [221]. Human interventions evaluating the effect of nut consumption on inflammation are shown in Table 5. Most clinical studies in Table 5 showed beneficial effects on inflammatory markers. Only one pistachio study showed no change in inflammatory markers [22]. In a pre-post intervention study [226], a diet supplemented with raw almond (20% of total energy substituted for fat and carbohydrate) and physical activity (45 min of walking at least 5 days a week) for 24 weeks in subjects with T2DM showed a significant decrease in hs-CRP (p < 0.01), along with a reduction in waist circumference, waist-to-height ratio, serum TC, TG, LDL-C, and HbA1c compared with a baseline 3-week almond-free diet and same level of physical activity [226].
Zhao et al. [222,223] conducted a randomized, controlled cross-over study in 23 hypercholesterolemic subjects who were randomly assigned to one of three diets for 6 weeks each: an average American diet ((AAD)-13% energy from SFA, 13% energy from MUFAs and 8.7% energy from PUFAs (7.7% LA; 0.8% ALA)); two diets high in PUFAs-a LA diet and ALA diet. The LA diet consisted of 16.4% energy from PUFAs (12.6% LA; 3.6% ALA). The ALA diet consisted of 17% energy from PUFAs (10.5% LA; 6.5% ALA). Walnuts and walnut oil were used for half of total fat in the 2 high-PUFA diets. The ALA diet reduced hs-CRP (p < 0.01) while the LA diet had a trend to decrease hs-CRP (p = 0.08) compared with the AAD [222]. The levels of IL-6, IL-1β and TNF-α cultured in peripheral blood mononuclear cells (PBMCs) was significantly lower on the ALA diet than on the LA diet (p < 0.05) [223]. In addition, the ALA diet decreased VCAM-1 by 16% while the LA diet decreased it by only 3% (p < 0.01). This study indicates that ALA and EPA derived from ALA exert a key role in the anti-inflammatory responses after walnut consumption [222] but a large proportion of dietary ALA is oxidized after digestion and only a small proportion is converted to EPA/DHA [227].
Other bioactive compounds in nuts may influence anti-inflammatory activity. Ellagic acid, a major polyphenolic compound in especially walnuts, has shown anti-inflammatory activity [228,229]. Nuts are a good source of magnesium which is inversely associated with the risk of CVD [230,231] and hs-CRP [232,233], and E-selectin levels [233].
Hernandez-Alonso et al. [225] showed an anti-inflammatory effect of pistachio consumption (pistachios are higher in β-carotene, γ-tocopherol and lutein than other nuts) compared with a control diet with olive oil to compensate for the energy from pistachios) [225]. Another pistachio diet (57 g/day) for 4 months significantly decreased IL-6 mRNA (p = 0.004) and resistin gene expression (p = 0.04) in lymphocytes compared with the control diet [225]. Elevated serum resistin levels are associated with insulin resistance, T2DM, and CVD [234][235][236].
Luo et al. [237] demonstrated the mechanism of resistin in increased glucose production and decreased insulin action in human liver HepG2 cells. Resistin led to insulin resistance through both 5 adenosine monophosphate-activated protein kinase (AMPK)-dependent and AMPK-independent pathways in hepatocytes: resistin upregulated cytokine signalling 3 (SOCS-3) expression, and downregulated insulin receptor substrate 2 (IRS-2) and Akt phosphorylation via an AMPK-independent mechanism. In addition, resistin downregulated AMPK leading to upregulated mRNA expression of gluconeogenic enzymes, G6Pase and PEPCK, and downregulated glucose transporter 2 (GLUT2) mRNA and glycogen synthesis [237].
Endothelial dysfunction which is characterized by decreased nitric oxide (NO) bioavailability (either decreased NO synthesis or destruction of NO by reactive oxygen species such as superoxide) [240], is associated with obesity and insulin resistance and is involved in initiation and promotion of atherosclerosis [241] and is related to the risk of CVD [242,243].
Insulin can induce both vasodilation and vasoconstriction in the endothelium. It has been postulated that insulin binding to the insulin receptor (IR) activates IR substrate (IRS) proteins leading to activation of phosphoinositide 3-kinase (PI3K)/protein kinase B (PKB)/endothelial NO synthase (eNOS) at the Ser1177-P pathway [244][245][246][247][248]. Activated eNOS synthesizes NO and L-citrulline from L-arginine and molecular oxygen. NO then stimulates cyclic GMP to induce smooth muscle relaxation.
Unsaturated fatty acids, especially oleic acid in nuts, appear to prevent atherosclerosis by modulating gene expression of endothelial leukocyte adhesion molecules. In HUVECs, oleic acid suppressed the expression of VCAM-1 stimulated with cytokines (e.g., TNF-α, IL-1α, IL-1β and IL-4) or lipopolysaccharide (LPS) or protein-kinase C activator (PMA) and oleic acid also inhibited LPS-stimulated NF-kB activation. Moreover, oleic acid decreased SFAs (palmitic and stearic acids) and increased oleic acid with no change PUFAs in the cell membrane phospholipids which influences prostaglandin production [254].
Adiponectin activates AMPK and PKB signalling resulting in NO generation in endothelial cells [255,256], and suppresses the expression of VCAM-1, ICAM-1 and E-selectin [257]. Walnut consumption (48 g/day) for 4 days increased circulating apo AI and total adiponectin levels in obese subjects with the metabolic syndrome, indicating that even short-term walnut consumption benefits the lipid profile and adiponectin levels which may insulin resistance and CVD [258].
Nut consumption may favourably affect endothelial function. Improved endothelial function was observed after consumption of walnuts in overweight subjects with visceral obesity [259], patients with hypercholesterolemia [192,260], and patients with T2DM [261], as well as after consumption of pistachios in subjects with mild dyslipidemia [262].
In a randomized crossover study of hypercholesterolemic subjects comparing a cholesterollowering Mediterranean diet and a walnut diet containing similar energy and fat substituting walnuts for 32% of the energy from MUFA in the Mediterranean diet, walnuts significantly improved endothelium-dependent vasodilation and decreased VCAM-1, TC and LDL-C. Decreased cholesterol correlated with increases in dietary α-linolenic acid and LDL γ-tocopherol levels. Changes in endothelium-dependent vasodilation was inversely correlated with those in the TC/HDL-C ratios suggesting the favourable effect of walnut on endothelial function could be explained in part by improved lipid profiles [192].
Other nut bioactive such as L-arginine [264] and magnesium [265,266] are possible candidates to account for improved endothelial function [254].
In a cross-sectional study, subjects who consumed more than 7.5 g/day of arginine had a 30% lower likelihood of having an hs-CRP above 3.0 mg/L compared with those who consumed 2.5-5.0 g/day of arginine [267]. Intervention studies [268,269] also showed that dietary L-arginine improved endothelial function and decreased monocyte adhesion to endothelial cells in young men with coronary artery disease [269] and decreased P-selectin, ICAM-1, E-selectin, IL-1 and IL-6 levels in subjects with intractable angina pectoris [268]. Similarly, daily supplementation of 10 g of arginine for 4 weeks improved endothelial function and decreased LDL-C oxidation in subjects with stable coronary artery disease without significant changes in inflammatory markers [270].

Blood Pressure
A meta-analysis of RCTs [271] showed that nut consumption reduced systolic blood pressure (SBP) by 1.3 mm Hg (p = 0.02) and diastolic blood pressure (DBP) in subjects without T2DM. Only pistachios and mixed nuts showed a significant effect on BP [271].
Mediterranean diets enriched with either 350 mL/week of extra virgin olive oil or 30 g/day of nuts lowered BP after a 4-year follow-up period [272] and 24-h ambulatory BP after a 1-year follow-up period [273]. Polyphenol consumption may have been responsible for the hypotensive effects [274]. The reduction in BP in 200 participants was positively associated with an increase in total polyphenol excretion and plasma NO markers and a positive correlation between urinary total polyphenol excretion and plasma NO was observed. These findings suggest that polyphenols could also exert a protective effect on endothelial function [274].

Conclusions
In conclusion, frequent nut consumption could play a role in reducing the risk of T2DM and CVD through improvement in glucose and lipid metabolism, weight maintenance and improved endothelial function. The protective effect of nuts could be explained by their distinctive nutrient profile and non-nutrient bioactive compounds. However, the specific mechanisms underlying these effects are not fully understood. Further targeted research should be undertaken to clarify the biological mechanisms.