1000/1000
Hot
Most Recent
Adiponectin is one of the most important adipokines. It is a bioactive peptide composed of 244 amino acids constituting approximately 0.01% of plasma proteins.
The adipose tissue, being the reservoir of energy, is also an active endocrine organ which synthesizes and secretes a variety of adipokines influencing the regulation of human metabolism. Adiponectin is one of the most important adipokines. It is a bioactive peptide composed of 244 amino acids constituting approximately 0.01% of plasma proteins. Plasma adipokine occurs in three types of complexes: (i) low-, (ii) medium-, and (iii) high-molecular-weight. High-molecular-weight (HMW) adiponectin is considered to be the most common and active form of adiponectin [1]. Furthermore, the activity of this adipokine also depends on the appropriate ratio between low- and high-molecular-weight adiponectin [2]. Currently, two isoforms of the adiponectin receptor are known: AdipoR1 and AdipoR2, which are located mainly in the skeletal muscles and the liver [1].
Adiponectin presents antineoplastic, cardioprotective and anti-inflammatory properties [3]. Additionally, it sensitizes tissues to insulin activity which contributes to its hypoglycemic properties [1]. Its hypolipidemic properties involve increasing the oxidation of fatty acids, reducing the storage of triglycerides in the skeletal muscles and increasing high density lipoprotein in the plasma via the activation of PPARα (Peroxisome Proliferator-Activated Receptor α). Its hypoglycemic effect is mostly due to the activation of PPARα, AMPK (AMP-Activated Protein Kinase), glucose transporters in the cell membrane such as GLUT4 (Glucose Transporter Type 4) and the reduction of gluconeogenesis in the liver [4][5].
Adiponectin expression is influenced by numerous factors including age, physical activity and ethnicity. Factors related to sex are also important determinants of its concentration. Women are characterized by a higher concentration of adiponectin compared to men, which is mainly due to the presence of different sex hormones [6]. Genetic factors also seem to be particularly important, because the concentration of adiponectin can be inherited by up to 55% [7]. Moreover, body weight and BMI (Body Mass Index) are also strongly correlated with adiponectin concentrations. Cruz-Mejía et al. [8] observed that adiponectin concentrations were markedly lower in obese individuals compared to participants with normal body weight (16.03 ± 2.53 μg/mL vs. 28.18 ± 1.97 μg/mL; p = 0.01). In addition, adiponectin concentrations were negatively correlated with the degree of obesity in obese patients (r = −0.477; p = 0.001). Interesting results were obtained in the POUNDS Lost Trial [9] which revealed statistically significant correlations between adiponectin concentrations, body composition and adipose tissue distribution. Increased adiponectin concentrations were significantly correlated with the reduction in the total adipose tissue (β = −0.68; p = 0.005), adipose tissue located within the trunk (β = −0.57; p = 0.005), subcutaneous adipose tissue (β = −0.42; p = 0.002) and visceral adipose tissue (β = −0.22; p = 0.02). Similar results were obtained by Gariballa et al. [10]. They demonstrated that the increased amount of visceral adipose tissue was related to the reduction in total adiponectin concentration. Furthermore, Meshkini et al. [11] noted that the adipose tissue content within the trunk was negatively correlated with adiponectin values (r = −0.44; p < 0.001). A high amount of adipose tissue in the area was also a strong prognostic factor of adiponectin concentration (β = −0.487; p < 0.001).
Apart from the above mentioned factors, adiponectin concentration is also influenced by dietary patterns and the share of individual products and nutrients in the diet. An appropriate dietary structure seems to be one of the most important factors increasing adiponectin concentrations. Beneficial dietary habits are the Mediterranean diet (MD), DASH diet (Dietary Approach to Stop Hypertension), diet based on plant products and diet with a reduced energy value. The products and nutrients which increase adiponectin concentrations include monounsaturated fatty acids, polyunsaturated omega-3 fatty acids, dietary fiber, polyphenols, alcohol and milk products. Conversely, dietary ingredients which have a negative effect on the concentration of adiponectin are saturated fatty acids, trans fatty acids, monosaccharides and disaccharides, and red meat, which are typical components of the Western diet. Furthermore, a diet characterized by a high glycemic index and a high-carbohydrate low-fat diet also seem to be unfavorable [1][2][12].
As regards factors which influence adiponectin expression, a key role is attributed to those which are associated with dietary habits and adhering to a healthy dietary pattern. The prospective Nurses’ Health Study [13] including 1922 women revealed that total adiponectin concentrations were 24% higher and HMW adiponectin concentrations were 32% higher in women from the highest quartile of adherence to the Alternate Healthy Eating Index (AHEI) compared to the women from the lowest quartile. Similar results were obtained by Volp et al. [14] who demonstrated a direct significant correlation between Healthy Eating Index and adiponectin concentrations.
The Mediterranean diet is one of the healthiest dietary patterns. Numerous observational and interventional studies have shown its correlation with increased adiponectin concentrations [5][15][16][17][18][19]. A traditional Mediterranean diet is characterized by a plentiful supply of vegetables, fruit, olive oil, fatty saltwater fish, whole-grain cereal products, moderate alcohol consumption and low red meat consumption [16]. It is suggested that the favorable activity of the MD on adiponectin expression may be due to the structure of the diet and the synergistic activity of its bioactive components, such as omega-3 fatty acids, fiber, vitamins and polyphenols which have a positive influence on adiponectin concentrations [20].
The ATTICA epidemiological study [4] demonstrated that adiponectin concentrations were 41% higher in persons from the highest tertile of adherence to the MD diet compared to those from the lowest tertile. Moreover, the score obtained for the diet was significantly correlated with adiponectin concentrations, both in women and in men. The described correlation was confirmed with a systematic review and a meta-analysis of 20 interventional studies in which adherence to the MD was associated with a significantly higher increase in adiponectin concentration compared to the control diet [21]. Comparable outcomes were also obtained by Sureda et al. [16] in a group of 598 inhabitants of the Balearic Islands. Significantly higher adiponectin concentrations were observed in adult men who adhered to the MD compared to non-adherent men. However, such a correlation was not observed in women and adolescents of both sexes.
Mantzoros et al. [17] conducted a study in a group of women with diabetes. They demonstrated that serum adiponectin concentrations significantly improved as a result of the MD in these women. Interesting results were also obtained by Spadafranca et al. [20] who studied the changes in serum adiponectin concentrations in 99 pregnant women in terms of the degree of adherence of the dietary pattern to the MD. Women from the highest tertile of adhering to MD were characterized by a significantly lower decrease in the percentage of adiponectin concentrations during pregnancy compared to women from the lowest tertile.
The Mediterranean diet is associated with numerous benefits in both sexes, including those related to adiponectin concentrations, despite substantial differences between sex and the response to the MD [22]. This was corroborated by a prospective cohort study conducted by Kouvari et al. [18] in a group of 1514 men and 1528 women. Serum adiponectin was markedly improved in both sexes after the introduction of the MD. Additionally, participants with a higher degree of adherence to the MD were at a lower risk of developing liver steatosis, which was strongly correlated with adiponectin levels. The MÉDITA randomized trial [15] was conducted in 215 T2DM (Type 2 Diabetes Mellitus) patients whose adiponectin concentrations increased by 43% after a year of following the MD. A similar correlation was observed for HMW adiponectin. Furthermore, a study by Luisi et al. [19] confirmed that the beneficial effect of the MD, additionally enhanced with 40 g of extra virgin olive oil daily, on adiponectin concentrations was independent of body weight, because a significant increase in adiponectin concentrations was observed both in participants with normal body weight and with excessive body weight. The authors suggested that olive oil contributed to the strong anti-inflammatory effect of DM.
The DASH diet is another healthy dietary pattern which, if adhered to, is associated with less severe systemic inflammation [23][24]. The DASH diet is based on vegetables, fruit, nuts, seeds of pulses, whole-grain cereal products and low-fat milk products. It is also characterized by a low content of red processed meat, sweetened beverages and products with high sodium content [25]. The mechanism through which the diet may be associated with an increased adiponectin concentration may, similarly to the Mediterranean diet, result from the presence of bioactive components with strong anti-inflammatory properties, such as polyphenols and omega-3 fatty acids. Nilsson et al. [25] conducted a study in 122 elderly women. Serum adiponectin concentrations were 20% higher in women from the highest tertile of the adherence to the DASH diet compared to women from the lowest tertile. AlEssa et al. [26] demonstrated an increasing tendency of adiponectin concentrations together with increased adherence to the DASH diet. However, the correlations were on the border of statistical significance. A beneficial effect of the diet on adiponectin concentrations may also be related to low sodium supply. According to Prates et al. [27] the dietary content of sodium was negatively correlated with adiponectin concentrations (r = −0.19, p = 0.03). Despite the paucity of studies linking this dietary pattern with adiponectin concentrations, it may be speculated that the diet may have a positive influence on the concentrations of this adipokine.
A plant-based diet is another dietary pattern which may exert a beneficial effect on serum adiponectin concentrations. It may be presumed that plant-based diets have a positive effect on adiponectin concentrations, but this is not as explicit as in the case of the MD. Adiponectin may also be influenced by bioactive components. Low animal protein and total protein content seem to be beneficial in terms of adipocyte function [28][29]. Kahleova et al. [30] noted that the concentrations of total adiponectin and HMW adiponectin increased by 19% and 15%, respectively, compared to baseline in the study group which followed a vegetarian diet for 24 weeks. Furthermore, a case-control study conducted in healthy non-obese adults revealed adiponectin concentrations to be significantly higher in women following a vegetarian diet than in those following a traditional diet. However, such a correlation was not observed in men [31]. The influence of reproductive hormones on the regulation of adiponectin concentrations may be a possible mechanism explaining why the described correlation was observed only in women [11]. A cross-sectional study conducted in a group of women showed that adiponectin concentrations were strongly correlated with FSH and SHGB concentrations [32]. Besides, sex-related differences in the expression of adiponectin may depend on differences in the distribution of the adipose tissue in men and women [11].
However, according to some authors, plant-based diets did not influence adiponectin concentrations [29][33][34] or might be associated with its lower expression [35]. Conversely, the authors explained the described correlation by the fact that the study was conducted in India in a group of 464 women (261 vegetarians and 203 non-vegetarians), where traditional vegetarian diet was characterized by a very high consumption of carbohydrates and a low consumption of fats (including omega-3 fatty acids). It was stated that various proportions between those macronutrients contributed to the difference between the studied groups of women. In addition, systematic reviews and meta-analyses of cross-sectional [36], interventional [37] and observational studies [38] demonstrated that vegetarian and vegan diets were associated with total lower inflammation compared to the traditional diet. However, no statistically significant relationship was found between this dietary pattern and adiponectin concentrations. Ambroszkiewicz et al. [29] revealed that children following a vegetarian diet were characterized by a significantly higher ratio of anti-inflammatory adiponectin and proinflammatory leptin compared to children consuming products of animal origin, which also indicated the anti-inflammatory properties of the diet.
It was also demonstrated that low-energy diets had a beneficial effect on adiponectin concentrations. It seems particularly beneficial to follow the negative energy balance pattern for a prolonged time, which resulted in effective reduction in body weight [2]. Monda et al. [39] conducted a study in 20 obese men and women. They observed that an eight-week balanced low-calorie ketogenic diet contributed to a significant increase in adiponectin concentrations both in women and in men. Furthermore, an increase was observed for all types of adiponectin of various molecular weights.
A randomized case-control study including 107 obese adults also showed that the reduction in calorie content by 500–700 kcal contributed to a significant increase in adiponectin concentrations. A similar correlation was observed in a group of individuals using diet combined with physical activity [40]. Similar results were also obtained by Christiansen et al. [41] and Abbenhardt et al. [42]. The observed correlations were confirmed by a systematic review and meta-analysis of 13 interventional studies which demonstrated that a low-calorie diet might considerably increase adiponectin concentrations. Particularly beneficial effects were observed if the diet was followed for at least 16 weeks. The authors suggested that the beneficial effect of the reduced-calorie diet on adiponectin concentrations depended predominantly on its duration and the degree of body weight reduction [43].
Song et al. [44] demonstrated that adiponectin concentrations significantly increased with the degree of body weight loss. Moreover, the activation of the PPARαreceptor and the reduction of inflammation resulting from the low-calorie diet seem to underlie this correlation. Alternatively, it is believed that body weight reduction may strengthen the expression of adiponectin receptors in the liver and skeletal muscles [43].
Apart from healthy dietary patterns the regulation of serum adiponectin concentrations also depends on individual nutrients, i.e., monounsaturated fatty acids, polyunsaturated omega-3 fatty acids, fiber, polyphenols and products included in the diet, i.e., dairy or alcohol. Seemingly, both physical activity and the use of low-energy diets influence adiponectin concentrations, mostly via the influence on body weight. However, the influence on adiponectin concentrations seems to be direct in case of some the nutrients such as monounsaturated fatty acids or polyunsaturated omega-3 fatty acids [11].
Omega-3 acids, including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), seem to be of particular importance in the context of adiponectin concentrations. Their main dietary sources are fatty saltwater fish and seafood [45]. The mechanism through which omega-3 acids induce adiponectin expression is mostly associated with PPARγ activation, the increased expression of adiponectin genes and the inhibition of the receptors of calcium ion channels [46][47]. Moreover, omega-3 acids reduce the concentrations of TNFα (Tumor Necrosis Factor α) and IL-6 (Interleukin 6), which inhibit the activity of the gene of this adipokine [48].
A randomized case-control study by Mazaherioun et al. [49] revealed that adiponectin concentrations significantly increased compared to baseline values in a group of individuals supplementing omega-3 fatty acids at a dose of 2.7 g/day. Furthermore, the study showed that PUFA (Polyunsaturated Fatty Acids) supplementation increased the expression of AdipoR1 and AdipoR2, adiponectin receptor genes, in persons with T2DM. Similar results were obtained by Barbosa et al. [48] who demonstrated that omega-3 supplementation at a dose of 3 g/day was also associated with significantly increased adiponectin concentrations. A study by Khorrami et al. [50] conducted in patients with atrial fibrillation showed that adiponectin concentrations significantly increased as a result of eight weeks of daily supplementation with 2 g of fish oil compared to the placebo arm. Similarly, Balfegó et al. [51] conducted a study in individuals who enhanced their diet with 100 g of sardines for 6 months. Adiponectin concentrations significantly increased from 2.1 ± 0.3 μg/mL at baseline to 3.0 ± 0.3 μg/mL after six months (p = 0.04). Interesting results were obtained by Song et al. [52] who demonstrated that adiponectin concentrations increased in each of the three groups over the period of 12 weeks. The groups differed in terms of the dose of EPA and DHA (gr. 1–3.1 g/d; gr. 2–6.2 g/d; gr. 3–12.4 g/d). However, the highest increase in adiponectin concentrations was observed in the group in which the dose of PUFA was the highest. This suggests that the amount of consumed omega-3 acids might also be of importance as regards the influence on adiponectin concentrations.
Similar correlations were also observed in women with PCOS [53][54]. Furthermore, the beneficial effect of omega-3 acids on adiponectin was observed in women with insulin resistance and with normal sensitivity to insulin [53]. The described correlations between omega-3 acids and adiponectin concentrations were confirmed by systematic reviews and meta-analyses of randomized case-control studies [45][46].
Linseed is another source of omega-3 acids. Linseed exerts a positive effect on adiponectin expression because it contains alpha-linolenic acid (ALA), which may also act as a ligand for the PPARγ receptor. Haidari et al. [55] conducted a randomized study in a group of women with PCOS. They demonstrated that linseed might significantly increase adiponectin concentrations. A significant increase in adiponectin concentrations was observed in comparison with baseline values in a study group which enhancing their diet with 30 g of ground linseed for 12 weeks. Gomes et al. [56] also demonstrated that the supplementation with 3 g of ALA significantly increased serum adiponectin concentrations after 60 days. However, a systematic review and a meta-analysis of seven randomized case-control studies revealed no statistically significant correlation between adiponectin concentrations and linseed consumption [57].
Notably, other studies revealed an important role of polyunsaturated omega-6 fatty acids and monounsaturated fatty acids on the regulation of adiponectin concentrations. A study by Kalgaonkar et al. [58] conducted in a group of women with PCOS showed that both walnuts and almonds significantly increased adiponectin concentrations. The results suggested that both linoleic acid found in walnuts and oleic acid found in almonds had a positive effect on adipokine concentrations. The described correlation was confirmed by a systematic review and meta-analysis of three randomized case-control studies which showed that walnuts were a dietary component with the potential of increasing adiponectin concentrations [59]. It is worth noting the results of a study by Kabiri et al. [60] conducted in obese women, which revealed that a diet rich in olive oil had a tendency to increase adiponectin concentrations (correlation on the border of statistical significance (p = 0.06)).
Furthermore, suitable ratios of dietary omega-3 and omega-6 acids [61] and polyunsaturated and saturated fatty acids [13] seem to be important factors. Fargnoli et al. [13] demonstrated significantly higher total adiponectin and HMW concentrations in women with the lowest ratio of polyunsaturated to saturated fatty acids compared to women with the highest ratio.
Dietary fiber is another component of food which has a positive effect on adiponectin concentrations. A review of 52 studies conducted by Silva et al. [2] showed that the presence of fiber in the diet contributed to an increase in adiponectin concentrations, even up to 60–115%. The prospective Nurses’ Health Study [13] demonstrated that women from the highest quartile of cereal fiber consumption had significantly higher total adiponectin and HMW adiponectin concentrations compared to women from the lowest quartile. A cross-sectional observational study by Pereira et al. [62] showed that a higher consumption of fiber included in vegetables and fruit was associated with higher adiponectin concentrations. The concentrations of adiponectin were 4.7 μg/mL (p = 0.03) higher in individuals from the highest quartile of cereal fiber consumption compared to participants from the lowest quartile.
The cross-sectional Health Professionals Follow-up Study [63], which included 780 men with T2DM, revealed that adiponectin concentrations were significantly higher in men from the highest quartile of cereal fiber consumption compared to men from the lowest quartile. Notably, the correlations between total fiber and vegetable fiber consumption and adiponectin concentrations were not statistically significant. Similar results were obtained by AlEssa et al. [64], who demonstrated that the consumption of both total fiber and fiber from cereals, vegetables and fruit was positively associated with adiponectin concentrations. Similar results were also obtained by Mantzoros et al. [17]. They observed that the consumption of whole-grain cereal products which were the source of dietary fiber was associated with significantly higher adiponectin concentration. Such an explicit correlation was also observed in the case of the consumption of fruit and nuts which also constitute an important source of dietary fiber. The seeds of pulses, including lentils, chickpeas, beans, broad beans and soy, are also excellent sources of both soluble and insoluble dietary fiber [65]. A study by Mirmiran et al. [65] conducted in a group of T2DM patients demonstrated that a diet in which two servings of red meat were replaced with pulse seeds effectively increased adiponectin concentrations.
Adiponectin concentrations and the expression of its receptors seem to be influenced by polyphenols which are secondary plant metabolites characterized by strong anti-inflammatory properties [66]. Coffee and green tea are rich sources of polyphenols. Caffeine and the catechin it includes have a beneficial effect on adiponectin mainly through the stimulation of the PPARγ receptor expression [67]. Some studies have revealed a correlation between the consumption of coffee and adiponectin concentrations [68][69][70]. A cross-sectional study conducted in Japan included 665 men. It revealed that higher coffee consumption was associated with higher adiponectin concentrations [71]. Similar results were also observed in relation to green tea [72][73][74]. Moreover, a study by Fragopoulou et al. [5] demonstrated that adiponectin concentrations were positively correlated with green tea consumption. However, a systematic review and a meta-analysis of 14 randomized case-control studies revealed no statistically significant correlation between serum adiponectin concentrations and green tea consumption. Nevertheless, the authors suggested that this might be due to the high heterogeneity of the analyzed studies (I2 D 91.7%; p < 0.0001) [75].
Curcumin is another polyphenol which has a beneficial effect on the expression of adiponectin [76][77]. Mirhafez et al. [78] conducted a study in a group of patients with nonalcoholic fatty liver disease and found that supplementation with curcumin for eight weeks contributed to a significant increase in adiponectin concentrations. Comparable results were also obtained by Adibian et al. [79] whose 10-week study included 44 patients with T2DM. The described correlations were confirmed by systematic reviews and meta-analyses of randomized case-control studies which showed a significant increase in adiponectin concentrations resulting from curcumin supplementation [80][81]. Furthermore, the highest increase was observed in case of interventions of at least 10-week duration [81].
It may be presumed that other polyphenols also increase the expression of adiponectin, but the data are too scarce to confirm correlations between such substances as anthocyanins, lignans, resveratrol or quercetin and the concentration of adiponectin [82][83][84][85][86][87][88]. A study by Tucakovic et al. [82] revealed that a four-week diet enhanced with the Queen Garnet plum, which is rich in anthocyanins, increased adiponectin concentrations by an average of 3.83 μg/mL (p = 0.048). Yang et al. [83] also demonstrated a correlation between anthocyanin supplementation for 12 weeks and adiponectin concentrations compared to placebo. Comparable results were also obtained by Jeong et al. [84], who demonstrated that the daily consumption of black raspberry for 12 weeks was associated with a considerable increase in adiponectin concentrations.
Resveratrol is another polyphenol influencing adiponectin. A randomized case-control study by Tomé-Carneiro et al. [85] revealed that a six-month dietary supplementation with grape extract increased adiponectin concentrations by 9.6% (p = 0.01) compared to placebo. The correlation was confirmed by a systematic review and a meta-analysis of 10 randomized case-control studies which showed that resveratrol supplementation contributed to a marked increase in adiponectin concentrations [89].
A study by Shahi et al. [86] revealed that lignans present in sesame seeds also had strong anti-inflammatory properties, and a diet enriched with sesame seeds significantly increased adiponectin concentrations in T2DM patients.
In addition, quercetin, mainly found in onion skin, also presented a beneficial effect in terms of adiponectin concentrations. Kim et al. [87] conducted a study in a group of women with excessive body weight. Quercetin supplementation for 12 weeks resulted in a significant increase in adiponectin concentration by 3.3 μg/mL compared to baseline. Additionally, a study by Rezvan et al. [88] including women with PCOS showed that oral quercetin supplementation increased the expression of the AdipoR1 and AdipoR2 adiponectin receptors.
Dairy products also seem to exert positive effects on adiponectin concentrations. However, there is a paucity of studies to confirm such a relationship. Nevertheless, due to the anti-inflammatory properties of such products, especially natural yoghurt, we may presume their positive effect on adiponectin concentrations [90]. A positive correlation between adiponectin concentrations and dairy products may be related to the content of milk fat, whey protein, vitamin D, calcium, potassium, magnesium and the reciprocal relations between those components [91]. A cross-sectional study including 612 Japanese individuals revealed that a diet characterized by the consumption of milk products was associated with higher adiponectin concentrations [91]. Yannakoulia et al. [92] also demonstrated a correlation between adiponectin concentrations and a dietary pattern characterized by a high intake of whole-grain cereal products and low-fat milk products. Similar results were obtained by Niu et al. [93], but a correlation was only observed for low-fat milk products. The correlation with high-fat milk products remained statistically insignificant. Fragopoulou et al. [5] also confirmed that the intake of low-fat milk products was positively correlated with adiponectin concentrations.
Moderate alcohol consumption also proved to be beneficial in relation to adiponectin concentrations. The prospective Nurses’ Health Study [13] revealed that the respective total adiponectin and HMW adiponectin levels were 28% and 45% higher in women from the highest quintile of alcohol intake (0.62–7.19 servings/d) compared to women who consumed no alcohol. Pischon et al. [12] and Bell et al. [94] also showed a significant positive correlation between moderate alcohol consumption and serum adiponectin concentrations. A similar correlation was reported in a study by Beulens et al. [95]. They noted that moderate alcohol intake for four weeks contributed to an increase in total adiponectin concentrations of 12.5% (p < 0.001). Furthermore, a cross-sectional study by Nova et al. [96] revealed that the relationship between adiponectin concentrations and moderate alcohol consumption was particularly visible in the case of wine (p = 0.017). The possible beneficial effect of this type of alcohol on adiponectin concentrations may be due to the content of polyphenols which are characterized by strong anti-inflammatory properties. However, some studies revealed no effect of alcohol consumption on the concentration of this adipokine [14][97][98]. Additionally, completely different results were observed as regards excessive alcohol consumption which was distinctly associated with low adiponectin concentrations [97][99]. The mechanism through which chronic heavy alcohol consumption is associated with reduced adiponectin concentrations is related to increased oxidative stress, the intensified expression of CYP2E1 (Cytochrome P450 2E1) and the reduced expression of PPARγ [100]. Detailed results of studies on the positive effects of dietary patterns on the concentration of adiponectin are described in Table 1.
Table 1. Dietary patterns and the concentration of adiponectin (AD)—A positive effect.
Author/Reference | Year | Study Design | Sample | Results |
---|---|---|---|---|
HEALTHY DIET | ||||
Fargnoli et al. [13] | 2008 | Prospective cohort study | 1922 women free of CVD, diabetes and cancer, aged 30–55 y | Total AD concentration was 24% higher (15.68 ± 1.03 μg/mL vs. 12.61 ± 1.03 μg/mL; p < 0.0001) and HMW AD was 32% higher (5.71 ± 1.04 μg/mL vs. 4.34 ± 1.04 μg/mL; p < 0.0001) in women from the highest quartile of adherence to AHEI compared to women from the lowest quartile. |
Volp et al. [14] | 2016 | Cross-sectional study | 157 apparently healthy men and women, aged 18–35 y | A correlation between the Healthy Eating Index and AD concentrations (r = 0.20074; p = 0.02). |
THE MEDITERRANEAN DIET | ||||
Mantzoros et al. [17] | 2006 | Cross-sectional study | 987 diabetic women, aged 30–55 y | Higher adherence to the MD was associated with markedly higher AD concentrations compared to the lowest adherence (6.91 ± 1.06 μg/mL vs. 5.49 ± 1.04 μg/mL; p < 0.01). |
Fragopoulou et al. [5] | 2010 | Cross-sectional study | 532 men and women free of CVD, aged > 18 y | Higher adherence to the MD was associated with markedly higher AD concentrations compared to the lowest adherence (4.8 ± 2.0 μg/mL vs. 3.4 ± 1.9 μg/mL; p < 0.001). A correlation between scores obtained for the MD and AD concentrations (women: rho = 0.156; p = 0.02), (men: rho = 0.130; p = 0.02). |
Schwingshackl et al. [21] | 2014 | A systematic review and meta-analysis of 17 interventional studies | 2300 men and women, aged 25–77 y | Adherence to the rules of the MD was related to significantly higher AD concentrations compared to the control diet (WMD: 1.69 mg/mL, 95% CI 0.27, 3.11; p = 0.02). |
Maiorino et al. [15] | 2016 | Randomized control study | 215 men and women with newly diagnosed T2DM, aged > 18 y | Following the MD for a year was associated with an increase in total AD concentrations by 43% (6.12 vs. 8.80 μg/mL; p < 0.001) and HMV AD by 54% (2.41 vs. 3.72 μg/mL; p < 0.01). |
Sureda et al. [16] | 2018 | Cross-sectional study | 598 men and women, aged 12–65 y | Adherence to the rules of the MD was related to significantly higher AD concentrations compared to non-adherence 13.1 ± 6.7 μg/mL vs. 9.5 ± 2.4 μg/mL; p < 0.05). No correlation found in women and adolescents of both sexes. |
Spadafranca et al. [20] | 2018 | Cohort study | 99 normal weight, pregnant women, aged 25–43 y | Women from the highest tertile of adhering to the MD were characterized by a lower decrease in the percentage of AD concentrations compared to women from the lowest tertile (10% ± 11% vs. −34% ± 3%; p = 0.01). |
Luisi et al. [19] | 2019 | Interventional study | 36 men and women, aged > 18 y | Following the MD enhanced with 40 g/d of extra virgin olive oil was associated with increased AD concentrations (increase by 0.6 ± 0.26 μg/mL; p < 0.01 in a group with normal body weight and an increase by 1.6 ± 0.2 μg/mL; p < 0.01 in a group with excessive body weight). |
Kouvari et al. [18] | 2020 | Prospective cohort study | 3042 apparently healthy men and women, aged > 18 y | Higher adherence to the MD was associated with markedly higher AD concentrations compared to the lowest adherence 4.8 ± 2.0 μg/mL vs. 3.4 ± 1.9 μg/mL; p < 0.001) |
THE DASH DIET | ||||
Nilsson et al. [25] | 2019 | Cross-sectional study | 112 women, aged 65–70 y | The highest tertile of adherence to the DASH diet was associated with markedly higher AD concentrations compared to the lowest tertile (12.9 ± 3.3 μg/mL vs. 11.5 ± 3.4 μg/mL; p = 0.008). |
PLANT-BASED DIET | ||||
Kahleova et al. [30] | 2011 | Randomized control study | 74 men and women with T2DM, aged 30–70 y | An increase in total AD by 19% (95% CI 7.5–25.4; p < 0.05) and HMV AD by 15% (95% CI 3.6–23.6; p < 0.05) after 24 weeks of following a vegetarian diet. |
Ambroszkiewicz et al. [29] | 2018 | Cross-sectional study | 117 prepubertal children, aged 5–10 y | Following a vegetarian diet was associated with a significantly higher adiponectin to leptin ratio (0.70 (0.37–0.93) vs. 0.39 (0.28–0.74); p = 0.005) compared to the traditional diet. |
Mirmiran et al. [65] | 2019 | Randomized cross-over clinical trial | 31 men and women with T2DM, aged 50–75 y | The consumption of two servings of pulses instead of red meat for eight weeks was associated with an increase in AD concentrations (10.5 ± 3.0 μg/mL vs. 13.1 ± 3.0 μg/mL; p < 0.05). |
Lovrenčić et al. [31] | 2020 | Case-control study | 76 non-obese men and women, aged 19–59 | Following a vegetarian diet was associated with significantly higher AD concentrations compared to the traditional diet (p = 0.03). No correlation in men. |
LOW-CALORIE DIET | ||||
Christiansen et al. [41] | 2010 | Randomized controlled trial | 79 obese men and women, aged 18–45 y | VLCD diet (800 kcal/d) was associated with a 19% increase in AD concentrations after 12 weeks (p < 0.01). |
Abbenhardt et al. [42] | 2013 | Randomized controlled trial | 439 overweight or obese postmenopausal women, aged 50–75 y | AD concentrations increased by 9.5% after 12 months of following LCD (12.4 μg/mL (11.3–13.5) vs. 13.5 μg/mL (12.5–14.6); p < 0.0001) and by 6.6% (12.8 μg/mL (11.7–13.9) vs. 13.6 μg/mL (12.5–14.8); p = 0.0001) as a result of combining LCD with physical activity. |
Bouchonville et al. [40] | 2014 | Randomized controlled trial | 107 obese men and women, aged ≥65 y | Calorie reduction of the diet by 500–700 kcal contributed to an increase in AD concentration by 8.9 μg/mL (3.5–14.8); p < 0.01), while the combination of reduction diet and physical activity contributed to an AD increase by 6.5 μg/mL (0.8–12.3; p = 0.02). |
Salehi-Abargouei et al. [43] | 2015 | Systematic review and meta-analysis of interventional trials (13 interventional studies) | 937 men and women, aged 20–75 y | The use of LCD was associated with an increase in AD concentration (Hedges’ g = 0.34, 95 % CI 0.17–0.50; p < 0.001), especially if the diet was followed for at least 16 weeks (Hedges’ g for ≤ 16 weeks = 0.48, 95% CI: 0.12–0.83; p = 0.01, (Hedges’ g for > 6 weeks = 0.30, 95 % CI: 0.11–0.48; p = 0.002). |
Monda et al. [39] | 2020 | Interventional study | 20 obese men and women, aged 20–60 y | The use of ketogenic VLCD for 8 weeks was associated with a significant increase in AD concentrations both in women (12.44 ± 1.07 μg/mL vs. 27.3 ± 1.33 μg/mL; p < 0.05), and in men (9.23 ± 0.7 μg/mL vs. 32.67 ± 1.6 μg/mL; p < 0.05). |
POLYUNSATURATED FATTY ACIDS | ||||
Fargnoli et al. [13] | 2008 | Prospective cohort study | 1922 women, free of CVD, diabetes and cancer, aged 30–55 y | Women from the group characterized by the lowest ratio of PUFA to SFA consumption had significantly higher total AD (12.66 ± 1.03 μg/mL vs. 11.47 ± 1.03 μg/mL; p = 0.01) and HMW (4.19 ± 1.04 μg/mL vs. 3.60 ± 1.03 μg/mL; p = 0.005) compared to women with the highest ratio. |
Kalgaonkar et al. [58] | 2011 | Randomized, prospective study | 36 women with PCOS, aged 20–45 y | The consumption of walnuts and almonds significantly increased AD concentrations (walnuts: 9.5 ± 1.6 μg/mL vs. 11.3 ± 1.8 μg/mL; p = 0.0241; almonds: 10.1 ± 1.5 μg/mL vs. 12.2 ± 1.4 μg/mL; p = 0.0262). |
Nadjarzadeh et al. [53] | 2015 | Randomized double-blind placebo-controlled clinical trial. | 84 women with polycystic ovary syndrome, aged > 18 y | Omega-3 supplementation (180 mg EPA and 120 mg DHA) for eight weeks significantly increased AD concentrations (4.44 ± 1.92 μg/mL vs. 5.62 ± 2.68 μg/mL; p < 0.005). |
Gomes et al. [56] | 2015 | Randomized double-blind, placebo-controlled trail | 20 men and women with T2DM, aged 30–65 y | Supplementation with 3 g of ALA increased AD concentrations after 60 days (10.61 ± 6.53 μg/mL vs. 15.01 ± 11.68 μg/mL; p = 0.01). |
Balfegó et al. [51] | 2016 | Pilot randomized trial | 35 men and women with T2DM, aged 40–70 y | Introducing 10 g of sardines into the diet (five times a week for six months) was associated with a significant increase in AD concentrations (2.1 ± 0.3 μg/mL vs. 3.0 ± 0.3 μg/mL; p = 0.04) |
Barbosa et al. [48] | 2017 | Randomized, double-blind placebo-controlled clinical trial | 80 men and women with at least one cardiovascular risk factor, aged 30–74 y | Omega-3 supplementation (3 g/d) for two months significantly increased AD concentrations (14.8 ± 10.0 μg/mL vs. 18.2 ± 12.1 μg/mL; p = 0.021). |
Mazaherioun et al. [49] | 2017 | Randomized, placebo-controlled, double-blind clinical trial | 88 men and women with T2DM, aged 30–65 y | Omega-3 supplementation (2.7 g/d) significantly increased AD concentrations (5.09 ± 2.79 μg/mL vs. 5.58 ± 3.13 μg/mL; p < 0.001). |
Mejia-Montilla et al. [54] | 2018 | Prospective study | 195 women with PCOS, aged > 18 y | N-3 supplementation (180 mg EPA and 120 mg DHA) significantly increased AD concentrations (3.9 ± 1.1 μg/mL vs. 5.3 ± 1.4 μg/mL; p = 0.001), both in women with HOMA-IR <3.5 (3.9 ± 1.1 μg/mL vs. 5.3 ± 1.4 μg/mL; p < 0.0001), and in those with HOMA-IR >3.5 (4.1 ± 1.1 μg/mL vs. 5.6 ± 1.3 μg/mL; p = 0.005). |
Song et al. [52] | 2018 | Double-blind randomized controlled trial | 201 healthy men and women, aged > 40 y | An increase in AD concentrations over 12 weeks as a result of omega-3 supplementation at a dose of: 3.1 g/d (5.79 ± 2.68 μg/mL vs. 6.36 ± 2.64 μg/mL; p < 0.05), 6.2 g/d (5.72 ± 2.07 μg/mL vs. 6.87 ± 2.58 μg/mL; p < 0.01) and 12.4 g/d (5.81 ± 2.13 μg/mL vs. 7.43 ± 2.63 μg/mL; p < 0.01). |
Bahreini et al. [47] | 2018 | A systematic review and meta-analysis of interventional trials (10 randomized controlled trails) | 177 men and women with T2DM, aged > 18 y | An increase in AD concentrations by 0.57 μg/mL as a result of omega-3 supplementation (95% CI 0.15–1.31; p = 0.01). |
Becic et al. [45] | 2018 | A systematic review and meta-analysis of interventional trials (10 randomized controlled trails) | 460 men and women with prediabetes and T2DM, aged > 18 y | An increase in AD concentrations by 0.48 μg/mL as a result of omega-3 supplementation (95% Cl 0.27–0.68; p < 0.00001). |
Haidari et al. [55] | 2020 | Randomized open-labeled controlled clinical trial | 41 women with PCOS, aged 18–45 y | An increase in AD concentrations over 12 weeks as a result of supplementation with 30 g of ground linseed (13.04 ± 3.36 μg/mL vs. 17.36 ± 4.1 μg/mL; p = 0.002). |
Khorrami et al. [50] | 2020 | Randomized double-blind, placebo-controlled study | 80 overweight or obese men and women with atrial fibrillation, aged > 50 y | An increase in AD concentrations over eight weeks as a result of supplementation with 2 g/d of fish oil (11.88 ± 6.94 μg/mL vs. 13.15 ± 7.33 μg/mL; p = 0.026). |
Yang et al. [59] | 2020 | A systematic review and meta-analysis of randomized clinical trials (3 randomized controlled trails) | 823 men and women, aged > 18 y | The consumption of walnuts significantly increased AD concentrations (WMD: 0.440 μg/mL; 95% CI: 0.323 to 0.557; p < 0.001). |
DIETARY FIBER | ||||
Qi et al. [63] | 2005 | Cross-sectional study | 780 men with T2DM, aged 40–75 y | Men from the highest quartile of dietary fiber consumption had significantly higher AD concentrations compared to men from the lowest quartile (17.3 μg/mL vs. 14.2 μg/mL; p = 0.006). |
Mantzoros et al. [17] | 2006 | Cross-sectional study | 987 diabetic women, aged 30–55 y | The consumption of whole-grain cereal products was associated with significantly higher AD concentrations (6.11 ± 1.06 μg/mL vs. 4.92 ± 1.05 μg/mL; p < 0.01). |
Fargnoli et al. [13] | 2008 | Prospective cohort study | 1922 women free of CVD, diabetes and cancer, aged 30–55 y | Women from the highest quartile of cereal fiber consumption were characterized by significantly higher total AD concentrations (14.73 ± 1.03 μg/mL vs. 13.36 ± 1.04 μg/mL; p < 0.04) and AD HMW (5.32 ± 1.04 μg/mL vs. 4.56 ± 1.04 μg/mL; p < 0.02) compared to women from the lowest quartile. |
Pereira et al. [62] | 2016 | Observational, cross-sectional study | 43 men and women, 18–60 y | A higher consumption of fiber included in vegetables and fruit was associated with higher AD concentrations (r = 0.50; p = 0.0007). The concentrations of adiponectin were 4.7 μg/mL (p = 0.03) higher in individuals from the highest quartile of cereal fiber consumption compared to participants from the lowest quartile. |
AlEssa et al. [64] | 2016 | Cross-sectional study | 2458 women, free of diabetes, aged 43–70 y | Women from the highest quintile of total fiber (p < 0.001), cereal fiber (p < 0.001), fruit fiber (p = 0.014) and vegetable fiber (p = 0.011) consumption had significantly higher AD concentrations compared to women from the lowest quintile. |
CURCUMIN | ||||
Campos-Cervantes et al. [77] | 2011 | Randomized, single blind, placebo-controlled trial | 50 obese men, aged 25–30 y | An increase in AD concentrations after six and 12 weeks of supplementation with 500 mg of curcumin (after six weeks: 16.0 μg/mL vs. 18.5 μg/mL; p < 0.01 and after 12 weeks: 16.0 μg/mL vs. 18. μg/mL; p < 0.02). |
Panahi et al. [76] | 2016 | Randomized controlled trial | 117 men and women, aged > 18 y | An increase in AD concentrations after eight weeks of supplementation with 1000 mg of curcumin (12.67 ± 2.13 μg/mL vs. 21.28 ± 4.40 μg/mL; p < 0.001). |
Mirhafez et al. [78] | 2019 | Randomized, double blind, placebo-controlled, cross-over trial | 65 men and women with nonalcoholic fatty liver disease, aged > 18 y | Supplementation with 250 mg/d of curcumin for wight weeks caused a significant increase in AD concentrations (14.35 ± 7.72 μg/mL vs. 18.23 ± 9.75 μg/mL; p < 0.001). |
Adibian et al. [79] | 2019 | Randomized, double blind, placebo-controlled trial | 44 men and women with T2DM, aged 40–70 y | Supplementation with 1500 mg/d of curcumin for 10 weeks caused a significant increase in AD concentrations (52.0 ± 8.0 μg/mL vs. 64.0 ± 3.0 μg/mL; p < 0.0001). |
Clark et al. [81] | 2019 | A systematic review and meta-analysis of interventional trials (10 randomized controlled trails) | 652 men and women with type 2 diabetes, prediabetes subjects, obese men or with metabolic syndrome, aged 18–84 y | Supplementation with curcumin caused a significant increase in AD concentrations compared to placebo (WMD: 0.82 Hedges’ g; 95% CI 0.33–1.30; p˂0.001). A particularly beneficial effect of at least 10 weeks of supplementation (WMD: 1.05 Hedges’ g; 95% CI: 0.64 to 1.45; p ˂ 0.001). |
Akbari et al. [80] | 2019 | Systematic review and meta-analysis of randomized controlled trials (21 randomized controlled trails) | 1646 men and women with metabolic syndrome | An increase in AD concentrations after supplementation with curcumin (SMD 1.05; 95% CI 0.23–1.87; p = 0.01). |
ANTHOCYANINS | ||||
Jeong et al. [84] | 2014 | Prospective randomized double-blind study | 77 men and women with metabolic syndrome, aged 18–75 y | Daily black raspberry consumption for 12 weeks was associated with an increase in AD concentrations (5.7 ± 5.1 μg/mL vs. 7.7 ± 5.0 μg/mL; p < 0.05). |
Tucakovic et al. [82] | 2018 | Randomized, double-blind, placebo-controlled, cross-over trial | 20 apparently healthy men and women, aged 18–65 y | Supplementation with the Queen Garnet plum for four weeks increased AD concentrations by the average of 3.83 μg/mL (p = 0.048). |
Yang et al. [83] | 2020 | Randomized controlled trial | 160 men and women with T2DM or prediabetes | Anthocyanin supplementation for 12 weeks was associated with an increase in AD concentrations compared to placebo (increase by 0.46 μg/mL; p = 0.038). |
RESVERATROL | ||||
Tomé-Carneiro et al. [85] | 2013 | Triple-blind, placebo-controlled clinical trial | 75 men and women, aged > 18 y | Supplementation with grape extract for six months increased AD concentrations by 9.6% (p = 0.01). |
Mohammadi-Sartang et al. [89] | 2017 | Systematic review and meta-analysis of randomized controlled trials (9 randomized controlled trails) | 590 men and women, aged > 18 y | Resveratrol supplementation significantly increased AD concentrations (WMD: 1.10 μg/mL, 95% CI 0.88, 1.33; p < 0.001) |
QUERCETIN | ||||
Kim et al. [87] | 2016 | Randomized double-blind, placebo-controlled study | 37 healthy overweight and obese women | AD increase after 12 weeks of quercetin supplementation (3.6 ± 2.0 μg/mL vs. 6.9 ± 2.3 μg/mL; p < 0.05). |
Rezvan et al. [88] | 2018 | Randomized double-blind, placebo-controlled study | 81 women with PCOS, aged 20–40 y | An increased expression of the AD receptors (AdipoR1 and AdipoR2) after 12 weeks of supplementation with 1 g/d of quercetin (p < 0.01). |
LIGNANS | ||||
Shahi et al. [86] | 2017 | Randomized double-blind, placebo-controlled study | 48 men and women with T2DM, aged 30–60 y | AD increase after eight weeks of supplementation with 200 mg/d of sesamin (6.21 ± 1.33 μg/mL vs. 7.34 ± 2.88 μg/mL; p = 0.024). |
COFFEE | ||||
Williams et al. [70] | 2008 | Prospective cohort study | 982 women with T2DM and 1058 nondiabetic women | The consumption of ≥4 cups of coffee daily was associated with significantly higher AD compared to the consumption of <1 cup a week (women with T2DM: 7.7 vs. 6.1 μg/mL; p = 0.002, nondiabetic women: 15.0 vs. 13.2 μg/mL; p = 0.04). |
Kempf et al. [68] | 2010 | Single-blind clinical trial | 47 men and women, free of T2DM, aged 18–65 y | The consumption of eight cups of coffee daily was associated with significantly higher AD concentrations compared to consuming no coffee (8421 (6634–11256) ng/mL vs. 7957 (6317, 10901) ng/mL; p < 0.05). |
Imatoh et al. [71] | 2011 | Cross-sectional study | 665 men, aged > 18 y | The consumption of ≥3 cups of coffee daily was associated with significantly higher AD compared to consuming no coffee (6.9 ± 3.3 μg/mL vs. 6.0 ± 2.6 μg/mL; p < 0.01). |
Yamashita et al. [69] | 2012 | Cross-sectional study | 3317 men and women, aged 35–69 y | The consumption of ≥4 cups of coffee daily was associated with significantly higher AD compared to the consumption of <1 cup a week (7.23 (6.84–7.65) μg/mL vs. 6.58 (6.40–6.76) μg/mL; p = 0.005). |
GREEN TEA | ||||
Hsu et al. [73] | 2008 | Randomized, double-blind, placebo-controlled clinical trial | 78 obese women, aged 16–60 y | An increase in AD concentrations after 12 weeks of supplementation with 400 mg of green tea extract (18.9 ± 6.7 μg/mL vs. 21.4 ± 8.7 μg/mL; p < 0.01). |
Fragopoulou et al. [5] | 2010 | Cross-sectional study | 532 men and women free of CVD, aged > 18 y | A correlation was found between green tea consumption and AD concentrations (rho = 0.108; p = 0.04). |
Liu et al. [74] | 2014 | Randomized, double-blind, and placebo-controlled trial | 102 men and women with T2DM, aged 20–65 y | An increase in AD concentrations after 16 weeks of supplementation with 500 mg of green tea extract (20.2 ± 5.1 μg/mL vs. 21.7 ± 5.1 μg/mL; p < 0.046). |
Chen et al. [72] | 2016 | Randomized, double-blind trial | 92 obese women, aged 20–60 y | An increase in AD concentrations after 12 weeks of supplementation with 856.8 mg of green tea extract (20.9 ± 11.0 μg/mL vs. 24.0 ± 10.7 μg/mL; p = 0.009). |
DAIRY PRODUCTS | ||||
Yannakoulia et al. [92] | 2008 | Cross-sectional study | 196 apparently healthy women, aged 18–84 y | A correlation occurred between AD and a dietary pattern rich in low-fat dairy and whole-grain cereal products (r = 0.15; p = 0.04). |
Niu et al. [93] | 2013 | Cross-sectional one-year longitudinal study | 938 apparently healthy men and women, aged > 18 y | The consumption of low-fat milk products (58.9–375 g/d) was associated with significantly higher AD concentrations compared to no consumption of such products (8.3 (7.8, 8.9) μg/mL vs. 7.3 (6.9, 7.6) μg/mL; p < 0.01). |
Fragopoulou et al. [5] | 2010 | Cross-sectional study | 532 man and women free of CVD, aged > 18 y | A correlation occurred between the consumption of low-fat milk products and AD concentrations (rho = 0.119, p = 0.04). |
Bahari et al. [91] | 2018 | Cross-sectional study | 612 men and women, 35–69 y | A diet characterized by the higher consumption of milk products was associated with higher AD concentrations (4.78 (3.24, 7.38) μg/mL vs. 3.68 (2.42, 6.12) μg/mL; p = 0.004). |
ALCOHOL | ||||
Pischon et al. [12] | 2005 | Prospective cohort study | 532 men, aged 40–75 y | Men from the highest quintile of AD concentrations (>24.9 μg/mL) consumed significantly more alcohol (16.2 ± 1.06 g/d vs. 13.05 ± 0.7 g/d) compared to men from the lowest quintile of AD concentrations (<10.6 μg/mL); p = 0.006). A correlation occurred between AD concentrations and alcohol consumption (r = 0.14; p = 0.002). |
Fargnoli et al. [13] | 2008 | Prospective cohort study | 1922 women free of CVD, diabetes and cancer, aged 30–55 y | Total AD concentrations were 28% higher (16.01 ± 1.03 vs. 12.50 ± 1.03; p < 0.0001) and HMW AD concentrations were 45% higher (6.10 ± 1.04 vs. 4.21 ± 1.03; p < 0.0001) in women from the highest quintile of alcohol consumption compared to those who consumed no alcohol. |
Beulens et al. [95] | 2007 | Randomized, controlled, cross-over trial | 17 apparently healthy men, aged 18–40 y | Moderate alcohol consumption (32 g/d) for four weeks caused an increase in total AD concentrations by 12.5% (p < 0.001). |
Bell et al. [94] | 2015 | Prospective cohort study | 2855 men and women, aged 40–63 y | Alcohol consumption was cross-sectionally associated with AD concentrations (β = 0.003; p < 0.001). |
Nova et al. [96] | 2019 | Observational cross-sectional study | 240 men and women, aged 55–85 y | Wine consumption was associated with higher AD (β = 204, 95% CI: 37–370; p = 0.017). |
Abbreviations: CVD, Cardiovascular disease; y, years; AHEI, Alternate Healthy Eating Index; AD, adiponectin; AD HMW, high-molecular-weight adiponectin; MD, Mediterranean diet; WMD, Weighted Mean Difference, T2DM, Type 2 Diabetes Mellitus; DASH, Dietary Approach to Stop Hypertension; VLCD, very low calorie diet; LCD, low calorie diet; PUFA, Polyunsaturated Fatty Acids; SFA, Saturated Fatty Acids; PCOS, polycystic ovary syndrome; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; ALA, α-linolenic acid; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance.
To conclude, adherence to the Mediterranean diet is related to particularly significant beneficial effect on serum adiponectin concentrations. The reciprocal relations between the components of this diet (i.e., monounsaturated fatty acids, polyunsaturated omega-3 fatty acids, fiber and polyphenols) and their individual properties contribute to such a positive effect. Moreover, the advantages of the Mediterranean diet are visible regardless of body weight, health status and sex. The properties of the DASH diet also seem promising. However, more research is necessary to provide an explicit confirmation of its positive effect on adiponectin concentrations. Furthermore, the plant-based diet and low-calorie diet seem to be beneficial in the context of adiponectin concentrations. Additionally, moderate alcohol consumption and dairy product intake seem to be of importance in terms of the regulation of adiponectin concentrations. However, more research is needed to determine the influence of such products on the expression of adiponectin.