Anthocyanins

ACN comprise a subgroup of flavonoids abundant in many fruit and vegetables, in particular, berries and grapes and their products such as juice and wine. Particularly rich in ACN are berries such as blackberries, black currants, black elderberries and blueberries with some varieties producing about 400–500 mg ACN/100 g⁽Reference Khoo, Azlan and Tang^7, Reference Vendrame, Del Bo and Ciappellano⁸⁾. ACN are water-soluble glycosylated pigments produced through plant secondary metabolism and responsible for the red, purple or blue colours. Most predominant ACN compounds are derived from pelargonidin, cyanidin, delphinidin, petunidin, peonidin and malvidin base structures, differing with regard to position and number of hydroxyl groups, degree of methylation, type and number of sugar moieties, ultimately leading to a large diversity of ACN and their composition in different plants. The major ACN found in most plants is cyanidin-3-glucoside.

Reduction of weight gain following ACN supplementation in rodents has been associated with modulation of hepatic lipid metabolism, such as reduction of sterol regulatory element-binding protein-1 mRNA levels, inhibition of enzymes involved in fatty acid and TAG synthesis and up-regulation of lipolytic enzymes⁽Reference Belwal, Nabavi and Nabavi⁹⁾. Furthermore, energy expenditure has been found accelerated in high-fat diets-induced obese mice following blackberry and blueberry ACN supplementation⁽Reference Wu, Gao and Guo¹⁰⁾. Similarly, Solverson et al.⁽Reference Solverson, Rumpler and Leger¹¹⁾ reported an increase in fat oxidation in a recent randomised controlled trial (RCT) in twenty-seven overweight or obese males given blackberries (1500 mg/d) with high-fat diets for 7 d.

Daneshzad et al.⁽Reference Daneshzad, Shab-Bidar and Mohammadpour¹²⁾ conducted a systematic review and meta-analysis of nineteen RCT evaluating effects of ACN supplementation on cardio-metabolic biomarkers including BW, BMI, waist circumference (WC), blood pressure (BP), lipid profile and glycaemic status. Duration of supplementation ranged from 1 to 96 weeks with ACN doses ranging from 31·5 to 1050 mg daily. While there was no significant effect of ACN supplementation on BW, WC, BMI, BP (systolic and diastolic), a sub-group analysis revealed that ACN intake for more than 12 weeks led to a 2·42 kg reduction in BW (mean difference (MD) −2·42 kg; 95 % CI −4·46, −0·38; P = 0·020) and a 0·75 kg/m² decrease in BMI (MD −0·75 kg/m²; 95 % CI −1·38, −0·23; P = 0·005). Given the overall lack of effect on anthropometric markers and BP, duration as well as ACN dose may be the most likely sources for heterogeneity observed among different trials. This is in line with Amiot et al.⁽Reference Amiot, Riva and Vinet¹³⁾ who included six ACN supplementation studies in their systematic review on the effects of dietary polyphenols on MetS markers and reported highly variable results on BMI, WC, BP, lipid profile and glucose metabolism which are likely to relate to the different amounts of ACN provided through different berry food products (berry type, juice or powder product, extract) given over a supplementation periods of 6–8 weeks. Most effective was a mixture of berries (bilberry, blueberry, sea buckthorn) taken daily over 8 weeks to reduce BMI and WC⁽Reference Broncel, Kozirog and Duchnowicz¹⁴⁾; aronia extract (300 mg daily over 2 months) was able to significantly reduce BMI⁽Reference Lehtonen, Suomela and Tahvonen¹⁵⁾. Conversely, a 6-week daily supplementation with freeze-dried strawberry powder (equivalent to 500 g fresh strawberries) caused no changes in anthropometric indices and serum glucose⁽Reference Moazen, Amani and Rad¹⁶⁾.

ACN may exert hypoglycaemic effects through a combination of mechanisms including inhibition of carbohydrate digestion through inhibition of salivary and pancreatic α-amylase and α-glucosidase, inhibition of intestinal glucose absorption⁽Reference Castro-Acosta, Lenihan-Geels and Corpe¹⁷⁾, stimulation of insulin secretion⁽Reference Johnson and de Mejia¹⁸⁾ and increased glucose uptake in peripheral tissues through up-regulated GLUT4 and its utilisation⁽Reference Luna-Vital and Gonzalez de Mejia^19, Reference Rozanska and Regulska-Ilow²⁰⁾. Furthermore, cyanidin-3-glucoside has been shown to lead to increased differentiation of pre-adipocytes into smaller and insulin-sensitive adipocytes⁽Reference Matsukawa, Inaguma and Han²¹⁾ and exerts insulin-like effects in human adipocytes by up-regulating PPARγ activity⁽Reference Scazzocchio, Vari and Filesi²²⁾. Other mechanisms related to decreased insulin resistance (IR) involve activation of AMPK and insulin receptor substrate 1 and reduced inflammation⁽Reference Belwal, Nabavi and Nabavi⁹⁾. In addition, ACN may act in the gut to modulate postprandial blood glucose, insulin and incretin response⁽Reference Castro-Acosta, Smith and Miller²³⁾.

High intake of ACN has been associated with significantly lower peripheral IR and high-sensitivity C-reactive protein levels⁽Reference Jennings, Welch and Spector²⁴⁾. Soltani et al.⁽Reference Soltani, Gorji and Asgary²⁵⁾ have shown that the daily consumption of ACN-rich cornelian berry (Cornus mas L.) improved glycaemic control significantly by increasing insulin and reducing HbA1 levels in patients with type 2 diatetes (T2D), and a 12-week RCT in 138 Chinese adults with prediabetes or early untreated diabetes revealed that purified ACN favourably affects glycaemic control and lipid profile, in particular in patients with elevated metabolic markers⁽Reference Yang, Ling and Yang²⁶⁾. As well, a recent systematic review and meta-analysis involving thirty-two RCT with a minimum duration of 2 weeks demonstrated a consistently improved glycaemic control (reduced fasting glucose, 2 h postprandial glucose and glycated Hb) in both healthy and metabolically diseased populations, although in particular in subjects with existing hyperglycaemia⁽Reference Yang, Ling and Du²⁷⁾. This review also indicated significant reductions in total cholesterol and LDL-cholesterol levels across the thirty-two RCT.

Daneshzad et al.⁽Reference Daneshzad, Shab-Bidar and Mohammadpour¹²⁾ could not confirm effects on HbA1c, serum insulin and blood lipid profile in their systematic review/meta-analysis when all nineteen studies were included. Sub-grouping for interventions over 300 mg ACN/d and duration over 12 weeks significantly lowered Homeostasis model assessment (HOMA)-IR (−21 %). ACN supplementation periods over 12 weeks significantly increased HDL-cholesterol and reduced LDL-cholesterol levels, and ACN supplementation >300 mg significantly reduced total cholesterol by 6·69 mg/dl and LDL-cholesterol levels by 8·60 mg/dl. Hassellund et al.⁽Reference Hassellund, Flaa and Kjeldsen²⁸⁾, investigating the impact of ACN on cardiovascular risk factors and inflammation in pre-hypertensive men, emphasise the importance of ACN supplementation period over the dose in intervention studies, which is confirmed by Zhu et al.⁽Reference Zhu, Huang and Zhang²⁹⁾ demonstrating significantly reduced LDL-cholesterol and increased HDL-cholesterol levels after 24 weeks of ACN supplementation. Further, Alvarado et al.⁽Reference Alvarado, Schoenlau and Leschot³⁰⁾ confirmed that LDL-cholesterol only decreased significantly after 12 weeks and not after 4 and 8 weeks of ACN supplementation.

Also, in the systematic review/meta-analysis by Daneshzad et al.⁽Reference Daneshzad, Shab-Bidar and Mohammadpour¹²⁾ significant reductions for total cholesterol, TAG and LDL-cholesterol, and significant increase for HDL-cholesterol were observed among patients with hypercholesterolaemia, indicating that ACN supplementation may provide a higher benefit to these patients in comparison to healthy individuals. Similar conclusions were drawn from a previous systematic review of Wallace et al.⁽Reference Wallace, Slavin and Frankenfeld³¹⁾ evaluating effects of purified ACN and ACN-rich extracts on markers of CVD (total cholesterol, TAG, LDL-cholesterol, HDL-cholesterol, BP) in healthy and diseased subjects in supplementation trials ranging from 3 to 24 weeks and ACN doses from 7·4 to 640 mg/d stating that largest reductions (particularly LDL-cholesterol) could be achieved in subjects with elevated levels.

To summarise, ACN and ACN-rich foods are generally accepted to benefit (maintaining) healthy BW, improvement of glucose and lipid metabolism which has been demonstrated at least partially in a number of intervention studies. Variations seen in outcomes of individual studies may be due to varying ACN dose and duration of intervention trials with a duration of 12 weeks and amounts about 300 mg ACN being considered beneficial. However, the source of ACN per se might have a strong impact on its effectivity. Highly methylated ACN such as malvidin and petunidin have demonstrated to be more effective at reducing negative metabolic consequences (body composition, energy expenditure, mitochondrial dysfunction) in high-fat diet fed C57BL/6 mice⁽Reference Skates, Overall and DeZego³²⁾. At present, to classify ACN and/or ACN sources based on their effectiveness is not possible. Future studies need to consider the ACN concentration and profile, the possible synergism between different ACN and other bioactives within the same source, as well as factors such as processing and intake patterns.

Catechins

Catechins are a group of polyphenols, flavan-3-ols, belonging to one of most common group of polyphenolics in the human diet, the flavonoids. The name catechin is derived from Cutch tree (Acacia catechu L.f.). Catechins are present in abundant concentrations in a variety of fruit, vegetables and plant-based beverages such as apple, berries, cacao beans, black soya bean, hops, tea, beer, wine and fruit juice⁽Reference Braicu, Ladomery and Chedea³³⁾. The consumption of food rich in catechins is associated with potential health benefits partly based on the antioxidant properties of polyphenols⁽Reference Mielgo-Ayuso, Barrenechea and Alcorta³⁴⁾. The chemical structure of catechin consists of two benzene rings (A- and B-rings) and a dihydropyran heterocycle (the C-ring) with a hydroxyl group on C₃. There are two chiral centres on the molecule on C₂ and C₃. Catechin stereoisomers in cis ((−)-epicatechin) or trans ((+)-catechin) configuration, with respect to C₂ and C₃, are flavan-3-ol compounds. Through esterification with gallate groups, flavanols can form gallic acid conjugates epicatechin (ECG), epigallocatechin and epigallocatechin gallate (EGCG). Condensed catechins are obtained via polymerisation. The most common oligomers derived from ECG are A-type and B-type procyanidins⁽Reference Bernatoniene and Kopustinskiene³⁵⁾.

This review of the clinical trials performed to evaluate the potential health effects of catechins on reducing the risk factors of MetS is focused on the results of human trials performed with food or food supplement or extracts rich in catechins. Studies have mainly been performed with cocoa and green tea, which are considered the richest dietary sources of catechins. Particularly, cocoa contains catechin, ECG and oligomers, and green tea is rich in EGCG, which is considered to be the most potent catechin and responsible for its health properties⁽Reference Suzuki, Pervin and Goto^36, Reference Tokede, Gaziano and Djousse³⁷⁾.

Hibi et al.⁽Reference Hibi, Takase and Iwasaki³⁸⁾ studied the effects of continual intake of green tea catechins (GTC) in MetS. In particular, the authors led a post hoc pooled analysis of data obtained from published reports (six human trials) to assess the effects of continual intake of GTC-containing beverages (540–588 mg/d) on abdominal fat area reduction and improvements in MetS (total 921 subjects). The studies were run in healthy Japanese adults (BW: 71·8 (sd10) kg; WC: 88·9 (sd7·3) cm; BMI: 26·8 (sd2·3) kg/m²) that consumed GTC for 12 weeks. Volunteers were categorised as pre-MetS and MetS at the initiation of the trial. Results show that BW and BMI were significantly lower in the group receiving the high GTC dose, mean 564 (sd19) mg GTC/d (BW: −1·69 kg, 95 % CI −1·84, −1·53; BMI: −0·65 kg/m², 95 % CI −0·70, −0·59 from baseline). WC and abdominal fat area (total fat area, visceral fat area and subcutaneous fat area) decreased significantly from baseline in the high GTC group, and the decrease was significantly greater than that in the low GTC group (35 (sd50) mg/d; P < 0·001). Moreover, the analysis of the subclass exposed that in both groups, low and high catechins, an improvement was observed in the proportion of subjects who improved from pre-MetS to healthy, and from MetS to healthy or pre-MetS, in 30·2 % of subjects in the low-catechin group and 41·5 % of subjects in the high-catechin group. However, the rate was significantly higher in the high-catechin group than in the low-catechin group (P = 0·024, χ² test).

In contrast, a randomised, doubled-blind, placebo-controlled study by Mielgo-Ayuso et al.⁽Reference Mielgo-Ayuso, Barrenechea and Alcorta³⁴⁾ reported no effect after the consumption of EGCG (300 mg/d) for 12 weeks in eighty-three premenopausal women (BMI 30·0–39·9 kg/m²). It did neither improve BW nor metabolic risk factors such as blood lipids.

A review carried out by Keske et al.⁽Reference Keske, Ng and Premilovac³⁹⁾ showed the heterogenicity of the results in trials aimed to link consumption of EGCG/green tea with glucose tolerance and insulin sensitivity. In patients with T2D, green tea extract (EGCG 860 mg/d) for 16 weeks significantly reduced HOMA-IR, glycosylated haemoglobin (HbA1c) and fasting insulin levels⁽Reference Hsu, Liao and Lin⁴⁰⁾, and consumption of more than three cups of tea daily was associated with a 17–35 % lower risk of T2D⁽Reference Jing, Han and Hu⁴¹⁾. Shimada et al.⁽Reference Shimada, Kawarabayashi and Tanaka⁴²⁾ revealed that oolong tea consumption for 4 weeks (EGCG 45 mg/d) significantly increases plasma adiponectin levels by 9·9 % and lowers HbA1c levels by 3·3 % in patients with various coronary risk factors. Additionally, there was a slight, but NS, decrease in the fasting plasma glucose levels. Hosoda et al.⁽Reference Hosoda, Wang and Liao⁴³⁾ used a higher dose of oolong tea treatment (EGCG 390 mg/d) for 4 weeks and reported lower fasting plasma glucose levels in people with T2D. In contrast, green tea consumption (540 mg/d polyphenols, EGCG content unknown) for 2 months had no apparent effect on metabolic markers such as fasting serum glucose and insulin, HbA1c and HOMA-IR⁽Reference Fukino, Shimbo and Aoki⁴⁴⁾. The proportion of flavanols (ratio of catechins is different in oolong tea than in green tea) and the study duration are critical aspects to modulate glucose metabolism positively.

The results of intervention studies indicate that consumption of flavan-3-ols is associated with an improvement of lipid homeostasis parameters such as HDL-cholesterol and LDL-cholesterol. Tokede et al.⁽Reference Tokede, Gaziano and Djousse³⁷⁾ analysed ten RCT of interventions (total 320 participants) administering dark chocolate/cocoa products for 2–12 weeks. Eight of the studies were comparing flavanol-rich cocoa or dark chocolate with either flavanol-poor white chocolate or a matching placebo. One study compared milk chocolate with cocoa butter and one compared a supplemented diet with dark chocolate and cocoa podwer with an unsupplemented diet. Therefore, the intake of catechins was heterogenous, from 963 to 88 mg/d compared with control intake from 0 to 75 mg catechins. The differences in catechin intake between cocoa/chocolate group and control ranged from 8·74 % to more than 100 %. The authors reported a significant reduction in serum LDL-cholesterol and total cholesterol levels (−5·90 and −6·23 mg/dl, respectively; data as MD of the results of the ten studies). No statistically significant effects were observed for HDL-cholesterol and TAG. Hooper et al.⁽Reference Hooper, Kay and Abdelhamid⁴⁵⁾ described the marginally significant effects of cocoa products on LDL-cholesterol (−0·07 mm; 95 % CI −0·13, 0·00 mm), HDL-cholesterol (0·03 mm; 95 % CI 0·00, 0·06 mm) and cholesterol (data referred as MD of the differences in each study between cocoa group and control).

Hartley et al.⁽Reference Hartley, Flowers and Holmes⁴⁶⁾ carried out an analysis of RCT lasting al least 3 months which investigated the effects of black or green tea or tea extracts involving healthy adults or those at high risk of CVD. The global analysis of the consumption of black tea (1 g extract/d, 1·29 g black tea polyphenols/d; three serving of black tea (200 ml/serving) and 318 mg black tea catechin/d) was found to produce statistically significant reductions in LDL-cholesterol (MD −0·43 mm, 95 % CI −0·56, −0·31). Green tea (58·91 mg catechin in green tea, 500 mg green tea polyphenols/d, 375 mg green tea extract/d, 200 mg theanine and 400 mg decaffeinated catechin green tea extract/d) was also found to produce statistically significant reductions in total cholesterol (MD −0·62 mm, 95 % CI −0·77, −0·46) and LDL-cholesterol (MD −0·64 mm, 95 % CI −0·77, −0·52). When both tea types were analysed together they showed favourable effects on LDL-cholesterol (MD –0·48 mm, 95 % CI −0·61, −0·35).

The meta-analysis of Desch et al.⁽Reference Desch, Schmidt and Kobler⁴⁷⁾ and Hooper et al.⁽Reference Hooper, Kay and Abdelhamid⁴⁵⁾ confirmed the BP-lowering capacity of flavanol-rich cocoa products. Desch et al.⁽Reference Desch, Schmidt and Kobler⁴⁷⁾ analysed 297 participants including six cross-over and four parallel-group designs. Although the studies displayed a diverse spectrum of treatment regimens (duration 2–8 weeks and intake 6·8–902 mg/d flavanol), results revealed that the mean BP reduction was −4·5 mmHg (95 % CI −5·9, −3·2; I ² = 89 %) for systolic BP and −2·5 mmHg (95 % CI −3·9, −1·2; I ² = 90 %) for diastolic BP. Hooper et al.⁽Reference Hooper, Kay and Abdelhamid⁴⁵⁾ reviewed the effects of chocolate, cocoa and flavan-3-ols including forty-two acute or short-term chronic (≤18 weeks) RCT that comprised 1297 participants. They observed reductions in diastolic BP (−1·60 mm Hg; 95 % CI −2·77, −0·43 mm Hg) and mean arterial pressure (−1·64 mm Hg; 95 % CI −3·27, 0·01 mm Hg). Although some studies did not identify dose-dependent effects of ECG, subgrouping by ECG dose suggested greater effects for systolic and diastolic BP at doses >50 mg/d. In the earlier reported meta-analyses by Hibi et al.⁽Reference Hibi, Takase and Iwasaki³⁸⁾ a significant decrease of systolic BP compared with baseline was observed only in the high catechin group (−11 mmHg, 95 % CI −2·1, −0·1; P < 0·01). The eleven RCT analysed by Hartley et al.⁽Reference Hartley, Flowers and Holmes⁴⁶⁾ evidenced that black tea consumption significantly reduced systolic BP (MD −1·85 mmHg, 95 % CI −3·21, −0·48), and green tea consumption significantly decreased both systolic and diastolic BP (MD −3·18 mmHg, 95 % CI −5·25, −1·11 and MD −3·42, 95 % CI −4·54, −2·30, respectively).

Most of the studies in the literature have been performed using catechins from tea or cocoa, and have evidenced that the dose exerting positive effects strongly depends on the physiological parameters that are being studied. Overall, catechins from tea seem to be effective in most of the MetS risk factors at a daily intake above 390 mg. The effect of cocoa's C is evident on BP with an intake from 6·5 mg/d.

Overall, several in vitro and in vivo animal studies are elucidating the potential mechanisms of action of catechins and they are on the way to demonstrate that flavan-3-ols can modulate metabolic pathways of the glucose and lipid metabolism and BP. It has been reported that EGCG up-regulates LDL receptor mRNA, reduces ApoB levels and inhibits pancreatic lipase, thereby reducing the absorption of dietary lipids⁽Reference Legeay, Rodier and Fillon⁴⁸⁾. Therefore, the modulation of molecules in lipid and glucose metabolism and the reduction on the delivery of proinflammatory cytokines as IL-6⁽Reference Nieto, Jaime and Arranz⁴⁹⁾ by catechins could contribute to reducing cholesterolaemia (LDL-cholesterol and total cholesterol) and BW. In vitro studies in several cell types (myocytes, adipocytes and hepatocytes) have reported that green tea or EGCG have insulin-mimetic metabolic actions. EGCG stimulates the uptake of glucose by stimulation of GLUT4 translocation⁽Reference Keske, Ng and Premilovac³⁹⁾. Analysis in endothelial cells show the enhancement of nitric oxide production by EGCG⁽Reference Legeay, Rodier and Fillon^48, Reference Osakabe⁵⁰⁾. Apart from the metabolic regulation, recent studies are focusing on assessing the epigenetic modulation of candidate genes of MetS by flavan-3-ols⁽Reference Legeay, Rodier and Fillon⁴⁸⁾.

Although these mechanisms could justify positive effects of catechins in human subjects, results of clinical intervention studies are still controversial. This is probably due to discrepancies among studies, including varying experimental designs, type and doses of catechin. Further research is needed to draw robust conclusions.

β-Glucan

BG is a NSP found in the cell walls of endosperm and aleurone cells of grains. BG consists of short β-(1,4)-d-glycans (cellotriosyl and cellotetraosyl units) linked to each other by β-(1,3) linkages leading to polymers of high molecular weight ranging from 8 to 200 kDa⁽Reference Roubroeks, Mastromauro and Andersson⁵¹⁾. This specific chemical structure is responsible for its physical properties, such as high solubility and viscosity which may contribute to the health benefits attributed to BG⁽Reference Wang and Ellis⁵²⁾, in particular those attributed to improvements of cardiometabolic health. Oat and barley are rich in BG, and most of the studies have been performed using BG from oat or barley.

Elevated WC is one of the criteria for MetS. However, clinical studies on BG have not focused on this anthropological parameter. A 4 % decrease of the WC was observed following adoption of a healthy diet that included ‘viscous fibres’ amongst other dietary improvements⁽Reference Tovar, Johansson and Bjorck⁵³⁾, which also saw improvement in a number of metabolic markers including fasting glucose, total and HDL-cholestrol. Beck et al.⁽Reference Beck, Tapsell and Batterham⁵⁴⁾ observed a significant effect of oat BG consumption (5–9 g/d) at breakfast on BW and WC, together with improvements in metabolic markers and alterations in levels of satiety hormones including leptin, and peptide YY. The study, however, showed that an energy-restricted diet had similar effects compared with oat BG consumption, which did not enhance the effectiveness of energy restriction. It is worth noting that the EFSA panel did not find sufficient evidence to substantiate a link between BG consumption and a reduction in appetite or BW (maintenance or achievement of normal BW)⁽⁵⁵⁾, although the panel did not consider evidence related to WC.

Conversely, EFSA supported a health claim stating that regular consumption of BG contributes to the maintenance of normal blood cholesterol concentrations for foods that provide ‘at least 3 g/d BG from oats, oat bran, barley, barley bran, or from mixtures of non-processed or minimally processed BG in one or more servings’⁽⁵⁵⁾. The US Food and Drugs Administration provided a similar recommendation⁽⁵⁶⁾.

A meta-analysis of epidemiological studies reported beneficial effects on blood lipids associated with consumption of soluble fibre from both oats and barley, but reported high levels of heterogeneity and called for well-controlled intervention studies⁽Reference Tiwari and Cummins⁵⁷⁾. A meta-analysis of RCT showed that oat BG at doses higher than 3 g/d reduced LDL-cholesterol and total cholesterol significantly compared with control, with little or no effect on HDL-cholesterol and TAG irrespective of dose or study duration⁽Reference Whitehead, Beck and Tosh⁵⁸⁾. The authors specified that the effectiveness of oat BG is linked to its high molecular weight and associated physicochemical properties, however called for more dose–response and longer studies to evaluate impacts of chronic consumption of oat BG in healthy and MetS populations.

Ibrugger et al.⁽Reference Ibrugger, Kristensen and Poulsen⁵⁹⁾ compared the effects of BG from oats and barley and showed that neither affected blood lipids significantly compared with the control. However, the consumption of 3·3 g/d oat BG led to the largest observed decrease in total cholesterol and LDL-cholesterol, as well as significantly reducing TAG. The authors identified a lack of systematic studies, with great differences amongst studies in terms of study foods, dose and study duration. Few of the intervention studies investigated the dose–effect relationship between oat BG and blood cholesterol. Biorklund et al.⁽Reference Biorklund, van Rees and Mensink⁶⁰⁾ reported that consuming a drink containing 5 g/d oat BG resulted in a 6·7 % decrease in LDL-cholesterol, while consumption of the drink containing 10 g/d oat BG reduced LDL-cholesterol by only 3·7 %, compared with control drink. Kerckhoffs et al.⁽Reference Kerckhoffs, Hornstra and Mensink⁶¹⁾ highlighted that processing of oats could have an adverse effect on the cholesterol lowering effect. Charlton et al.⁽Reference Charlton, Tapsell and Batterham⁶²⁾ showed that 1·5 g/d provided as cereal flakes was just as effective as 3 g/d provided as porridge in lowering blood cholesterol. Wolever et al.⁽Reference Wolever, Tosh and Gibbs⁶³⁾ showed the importance of molecular weight for the effectiveness of oat BG towards cholesterol markers⁽Reference Wolever, Tosh and Gibbs⁶³⁾. The impact of processing on oat BG properties has been recently reviewed by Grundy et al.⁽Reference Grundy and Fardet⁶⁴⁾. In healthy people, BG consumption does not appear to affect lipid homeostasis⁽Reference Lovegrove, Clohessy and Milon⁶⁵⁾.

Epidemiological studies have supported the association between whole-grain intake and improved metabolic risk factors for T2D and MetS⁽Reference McKeown, Meigs and Liu^66, Reference Sahyoun, Jacques and Zhang⁶⁷⁾. The fasting glucose concentrations decreased across increasing quartile categories of whole-grain intake. However, few clinical trials have focused on the impact of the consumption of BG on glucose metabolism. Many studies investigating BG and lipid homeostasis have also investigated impacts on glucose homeostasis. The EFSA panel supported a claim that consuming 4 g BG from oats or barley for each 30 g available carbohydrate decreased post-prandial glycaemic response without disproportionally increasing insulin response. The effect was observed when BG was incorporated into carbohydrate-rich food (e.g. bread or pasta) and when combined into a meal⁽⁵⁵⁾. Consuming at least 4 g BG per meal, from either oats or barley, and where the BG is soluble and has a molecular weight >250 000g/mol is sufficient to significantly reduce post-prandial area under curve by 27 (sd3) mmol min/l for meals with about 30–80 g available carbohydrates⁽Reference Tosh⁶⁸⁾. He et al.⁽Reference He, Zhao and Huang⁶⁹⁾ carried out a meta-analysis of controlled intervention trials, and showed that consumption of either whole-grain oats or BG extracted from oats was associated with strong significant reducing effects on fasting glucose and fasting insulin in T2D patients, but no effect on hyperlipidaemic subjects. A moderate effect was observed for obese subjects without hyperlipidaemia. A long-term (6 months) substitution of regular white bread with a functional bread enriched with fibre (7·62 g/100 g of bread, mostly BG) in the everyday diet of subjects with T2D induced no statistical difference on the fasting glucose level, but a significant decrease was observed for the post-prandial plasma glucose (P = 0·001) and mean plasma glucose (P = 0·02) with the ‘functional bread’ compared with the control bread⁽Reference Tessari and Lante⁷⁰⁾. In this study, other metabolic markers such as blood lipids and BP were not affected.

Few clinical trials have specifically studied the effects of the consumption of BG on BP. The results of the different studies show discrepancies. Past results obtained with healthy volunteers generally did not demonstrate an effect of the consumption of fibres on BP compared with low-fibre grain supplementation⁽Reference Margetts, Beilin and Vandongen⁷¹⁾. However, a recent meta-analysis concluded that systolic and diastolic BP could be reduced by 2·9 mmHg (95 % CI 0·9, 4·9) and 1·5 mmHg (95 % CI 0·2, 2·7), respectively, by diets rich in BG, for a median difference in BG of 4 g in healthy volunteers⁽Reference Evans, Greenwood and Threapleton⁷²⁾. The consumption of BG should thus help to manage BP of non-healthy people, especially people at risk of MetS. In 2006, Behall et al.⁽Reference Behall, Scholfield and Hallfrisch⁷³⁾ demonstrated the effects of consuming controlled portions of whole-grain rice and barley BG on BP in twenty-five overweight/obese mildly hypercholesterolaemic women. Both whole-grain rice and barley BG interventions led to significant decreases in diastolic BP and the mean arterial pressure, especially in post-menopausal women. In a randomised cross-over design, the consumption of a diet enriched in legumes and barley by overweight women for 4 weeks induced a significant reduction (−3 %, P < 0·05) of the diastolic BP but no effect was observed on systolic BP compared with the equivalent diet without legumes and barley⁽Reference Tovar, Nilsson and Johansson⁷⁴⁾. A similar observation was made in healthy and obese men and women consuming multifunctional diets that included BG amongst other health-enhancing constituents⁽Reference Tovar, Johansson and Bjorck⁵³⁾. However, it is difficult to dissociate the effect of BG from other constituents in the diet.

Summarising, there is strong and consistent evidence that consumption of BG impacts on lipid metabolism, with strong caveats relating to the dose and molecular size required for effects. There are multiple mechanisms associated with the effects of BG on lipid metabolism which may be acting in concert to excerpt positive effects. Proposed mechanisms include increased gut permeability⁽Reference Mackie, Rigby and Harvey⁷⁵⁾, reduced lipid digestion and absorption⁽Reference Wang and Ellis⁵²⁾, decreased bile reabsorption through physical barrier and bile colonic metabolism⁽Reference Gunness, Michiels and Vanhaecke⁷⁶⁾, increased bile acid production and SCFA metabolism⁽Reference Thandapilly, Ndou and Wang⁷⁷⁾ which impact on cholesterol homeostasis. There is also strong evidence supporting a role for BG in control of post-prandial glucose, but its effect may be attributed to fibre in general, rather than specifically to BG.

The evidence for other markers of MetS including BW, fasting glucose and BP are less well established. It is clear that further research is needed, also focusing on BG-matrix interactions and implications of food processing.

n-3 Long chain PUFA

n-3 LCPUFA, namely EPA and DHA have been suggested as potential anti-obesity bioactives⁽Reference Torres-Fuentes, Schellekens and Dinan⁵⁾, and growing evidence is emerging about the role of white adipose tissue in mediating the beneficial effects of marine n-3 PUFA in obesity-associated metabolic disorders. EPA and DHA have been shown to reduce BW and fat deposition in human clinical studies⁽Reference Parra, Ramel and Bandarra⁷⁸⁾. Their mechanism of action is supposed to be multiple. After consumption, these fatty acids are incorporated into cell membranes where they modulate membrane protein function, cellular signalling and gene expression⁽Reference Bordoni, Di Nunzio and Danesi⁷⁹⁾. Incorporation of EPA and DHA into tissues may modify inflammatory and immune reactions, mainly by inhibiting pro-inflammatory interleukins, therefore counteracting low-grade chronic inflammation caused by obesity. It has been suggested that MetS is the consequence of adipose tissue abnormalities. Therefore, n-3 LCPUFA could target adipose tissue inflammation and improve systemic metabolism⁽Reference Kalupahana, Claycombe and Moustaid-Moussa⁸⁰⁾. In addition, several trials indicate that n-3 LCPUFA reduce hypertension, total cholesterol and TAG levels in the body, being a perfect candidate to develop nutritional strategies to counteract MetS.

In 2012 and 2013, the effects of EPA and DHA have been in the focus of two reviews emphasising several limitations, including varying experimental designs, type and doses of n-3 PUFA, making it impossible to draw robust conclusions⁽Reference Lorente-Cebrian, Costa and Navas-Carretero^81, Reference Martinez-Victoria and Yago⁸²⁾. Other trials have been performed in the following years using supplements, fish oil (FO) or enriched foods.

In the single-blind, parallel trial described by Oh et al.⁽Reference Oh, Koh and Sakuma⁸³⁾ a placebo or n-3 PUFA as supplement (1, 2 or 4 Omacor^® capsules each containing 460 mg EPA ethyl ester and 380 mg DHA ethyl ester) were randomly administered to 176 patients with primary hypertriglyceridaemia (>150 mg/dl) once daily for 2 months. n-3 PUFA treatment dose-dependently and significantly decreased TAG and TAG/HDL-cholesterol and improved flow-mediated dilation but caused no significant modification in BMI compared with placebo.

Likewise, no modification in fat-free mass, upper-body subcutaneous fat mass and visceral fat mass across the intervention or between groups was observed in the prospective, randomised, placebo-controlled, double-blind study by Hames et al.⁽Reference Hames, Morgan-Bathke and Harteneck⁸⁴⁾ involving insulin-resistant, overweight or obese adults aged 18–65 years. Participants were randomly assigned to placebo (4·2 g oleic acid/d) or received a supplement containing 3·9 g EPA + DHA/d. Although EPA and DHA concentration in plasma and adipose tissue significantly increased in the n-3 group, there was no improvement in adipose tissue markers of inflammation. BMI (+0·7; P = 0·03), percentage of body fat (0·9 %; P = 0·009) and leg fat mass (0·5 kg; P = 0·02) increased for participants in both groups at the end of the intervention, and the changes were not different between groups.

Supplementation with n-3 LCPUFA did not improve the effect of a hypoenergetic diet in the RCT by Tardivo et al.⁽Reference Tardivo, Nahas-Neto and Orsatti⁸⁵⁾. The trial included eighty-seven postmenopausal Brazilian women with MetS, who were randomised to diet alone or diet plus n-3 supplementation, 900 mg/d. After 6 months, despite significant reductions in BMI and WC observed in both groups, there were no changes in body fat or muscle mass. Intervention with n-3 LCPUFA was associated with significant reduction in systolic (<12·2 %) and diastolic (<8·2 %) BP, serum TAG concentration (<21·4 %) and IR (<13·1 %; P < 0·05), as well as a reduction in serum IL-6 concentration (<28·5 %; P = 0·034).

In contrast, a significant effect on body fat upon n-3 LCPUFA supplementation was observed by Barbosa et al.⁽Reference Barbosa, Melo and Damasceno⁸⁶⁾. In this double-blind, placebo-controlled, randomised clinical trial, a supplement containing n-3 LCPUFA (3 g/d; 37 % EPA and 23 % DHA) or placebo (3 g/d sunflower oil) were administered for 2 months. Study participants were eighty men and women, aged 30–74 years, with some classic CVD risk factors (overweight, hypertension, dyslipidaemia, diabetes, smoking) with or without treatment and without previous cardiovascular event. The n-3 group showed a significant reduction of body fat compared with the placebo group, without any significant modification in BW, BMI and WC. In the treated group, an increase in serum adiponectin was detected. Adiponectin synthesis is inversely proportional to the amount of adipose tissue⁽Reference Barbosa, Melo and Damasceno⁸⁶⁾; in animals, increased n-3 LCPUFA consumption is associated with increased adiponectin levels, however the results are controversial in human subjects⁽Reference Lopez-Huertas⁸⁷⁾. Results of this trial confirm that n-3 LCPUFA consumption reduces body fat, leading to increased concentration of adiponectin and this, in turn, could further influence the reduction of fat mass.

Overall, although n-3 LCPUFA as supplements modify some MetS and CVD-related parameters they seem to have no effect on BW and BMI. On the contrary, a significant reduction of body fat could be related to the administration of supplements containing 3 g/d EPA + DHA.

The effect of an increased dietary intake of n-3 LCPUFA could be different. The RCT of the LIPGENE study⁽Reference Yubero-Serrano, Delgado-Lista and Tierney⁸⁸⁾ involved volunteers aged 35–70 years with BMI 20–40 kg/m², characterised by at least three of the following five criteria: high WC, high fasting glycaemia, high TAG, high BP, low HDL-cholesterol. Each subject was randomly stratified to one of four dietary interventions for 12 weeks: high SFA; high MUFA; n-3 diet including 1·24 g/d n-3 LCPUFA with a ratio of 1·4 EPA:DHA; control diet, including control high-oleic acid sunflower seed oil capsules. Volunteers were stratified according to their IR. MetS subjects without IR (lower HOMA-IR) showed improvement in metabolic risk factors related to MetS, such as obesity, BP and lipid markers, after consumption of the n-3 LCPUFA diet. In addition, in subjects without IR, WC was reduced after consumption of the control and n-3 LCPUFA diets compared with the high SFA and high MUFA diets (all P < 0·05).

Based on the evidence of the health benefits related to the consumption of oily fish⁽Reference Lavie, Milani and Mehra⁸⁹⁾, some trials administered n-3 LCPUFA as FO, enriched oils or enriched food. The intervention study by Venturini et al.⁽Reference Venturini, Simao and Urbano⁹⁰⁾ included 102 patients (eighty-one women and twenty-one men) with MetS (mean age 51·45 (sd8·27) years) aimed to compare extra virgin olive oil (OO) and FO effects, also investigating their possible synergism. Patients in the control group were instructed to maintain their usual diet; FO group received 3 g/d FO (ten capsules, each one containing 180 mg EPA and 120 mg DHA); OO group received 10 ml/d OO; and the fourth group (FOO) received 3 g/d FO and 10 ml/d OO. After 90 d intervention, no intragroup changes in anthropometric parameters were observed compared with baseline. In the FOO group, after treatment a significant decrease in LDL-cholesterol, and total cholesterol/HDL-cholesterol and LDL-cholesterol/HDL-cholesterol indexes was observed compared with baseline.

Fifty-nine subjects with early-stage T2D or MetS participated in an 8-week, randomised, single-blind, parallel intervention study⁽Reference Lee, Ivester and Hester⁹¹⁾. Individuals received either corn oil, a botanical oil combination (borage (Borago officinalis L.)/echium oil (Echium plantagineum L.)) or FO (EPA 3·58 g/d and DHA 2·44 g/d). FO supplementation induced a marked increase in serum levels of n-3 LCPUFA, HDL-cholesterol and insulin, and a decrease in serum TAG. No indication of the effect on anthropometric data was reported by the researchers.

A randomised, cross-over, five diet period, controlled feeding study was conducted by Liu et al.⁽Reference Liu, Kris-Etherton and West⁹²⁾ on 130 participants with BMI 22–40 kg/m² with central obesity plus at least one other MetS criteria. Five treatment oils: rapeseed oil, CanolaOleic (high-oleic acid rapeseed oil), CanolaDHA (high-oleic acid rapeseed oil with DHA), maize/saff (maize/safflower oil) and flax/saff (flax/safflower oil) were incorporated into smoothies that participants consumed twice daily. The quantity of oil was calculated based on participant energy needs, and it provided 18 % of total energy. The impact of each test diet on BW and body composition was low, and mainly on android fat mass that significantly decreased from baseline on the rapeseed and CanolaOleic oil diets only. The reduction in android fat mass was positively correlated with decreases in cardiometabolic risk factors including TAG, systolic and diastolic BP after all diets except the maize/saff lower oil group.

In a double-blind randomised trial, thirty-six patients with MetS received 500 ml/d semi-skimmed milk (placebo) or 500 ml/d skimmed milk enriched with 275 mg EPA + DHA and 7·5 g oleate and underwent 24 weeks of high-intensity interval training⁽Reference Ortega, Morales-Palomo and Fernandez-Elias⁹³⁾. Treatment did not increase n-3 LCPUFA plasma concentration, and a similar decrease in BW, WC, body fat mass, trunk fat mass and BP were observed in placebo and treated group. However, insulin sensitivity, serum concentration of C-reactive protein and HDL-cholesterol improved only in the treated group.

As for supplements, the increase of n-3 LCPUFA intake by FO or enriched-food significantly improves different physiological parameters without clear effect on BW, BMI and other anthropometric parameters.

The effect of n-3 LCPUFA was also investigated in combination with other bioactives. Eventy-eight individuals (thirty-three men and forty-five women), aged 35–70 years, with a large BMI (27–35 kg/m²) and WC (men >102 cm, women >88 cm) and at least one more component of the MetS were recruited in the trial reported by Bondia-Pons et al.⁽Reference Bondia-Pons, Poho and Bozzetto⁹⁴⁾. Participants were randomly assigned to one of four different nutritional interventions for the duration of 8 weeks. Diets only differed for the content of n-3 LCPUFA and polyphenols. Dependency network analysis showed a different pattern of associations between lipidomics, dietary and clinical variables after the dietary interventions, but no modification in BMI or WC were observed in any group.

Foods with a combination of high-oleic acid rapeseed oil-DHA and barley BG have been used in the CONFIDENCE trial, a randomised, single-blind cross-over trial with four treatment phases of 28 d each⁽Reference Ramprasath, Thandapilly and Yang⁹⁵⁾. The possible synergism between DHA and other bioactives was also in the focus of the EU project PATHWAY-27 (Pivotal assessment of the effects of bioactives on health and wellbeing. From human genoma to food industry)⁽⁹⁶⁾ that investigated the role and mechanisms of action of DHA, oat BG and ACN, alone and in combination, in the counteraction of MetS considering them not as stand-alone molecules but as ingredients of food. In PATHWAY-27, three monocentric, parallel-arm, randomised, double-blind pilot trials and a multicentre, randomised, placebo-controlled, parallel-arm dietary intervention study were performed on subjects at risk of MetS. At present, neither CONFIDENCE nor PATHWAY-27 results are available in the literature to report on the outcomes of potential combined effects in interventions involving DHA, BG (and ACN).

Based on available results, the increased intake of n-3 LCPUFA seem to have an effect on BW and BMI only if it is associated to modification of the whole diet so we can argue that it is not simply due to LCPUFA themselves. The effectiveness of n-3 LCPUFA on other parameters has been evidenced in trials using both supplements and enriched food with differences related to the daily dose, the duration of the intervention and the EPA:DHA ratio.