Associations of dietary intakes of anthocyanins and berry fruits with risk of type 2 diabetes mellitus: a systematic review and meta-analysis of prospective cohort studies

Abstract

To investigate the associations of dietary intakes of anthocyanins and berry fruits with type 2 diabetes mellitus (T2DM) risk and to evaluate the potential dose–response relationships based on prospective cohort studies. Cochrane library, Embase and PubMed databases were systematically searched up to Jan 2016 for relevant original studies. Summary relative risks (RRs) were calculated with a random effects model comparing the highest with lowest category. Dose–response was estimated using restricted cubic spline regression models. Three cohort studies reporting dietary anthocyanin intake with 200 894 participants and 12 611 T2DM incident cases, and five cohort studies reporting berry intake with 194 019 participants and 13 013 T2DM incident cases were investigated. Dietary anthocyanin consumption was associated with a 15% reduction of T2DM risk (summary RR=0.85; 95% confidence interval (CI): 0.80–0.91; I2=14.5%). Consumption of berries was associated with an 18% reduction of T2DM risk (summary RR=0.82, 95% CI: 0.76–0.89; I2=48.6%). Significant curvilinear associations were found between dietary intake of anthocyanins (P for nonlinearity=0.006) and berries (P for nonlinearity=0.028) and T2DM risk, respectively. The risk of T2DM was decreased by 5%, with a 7.5 mg/day increment of dietary anthocyanin intake (RR=0.95; 95% CI: 0.93-0.98; I2=0.00%) or with a 17 g/day increment of berry intake (RR=0.95, 95% CI: 0.91–0.99; I2=0.00%), respectively. Higher dietary intakes of anthocyanins and berry fruits are associated with a lower T2DM risk.

Introduction

In recent decades, non-communicable diseases have increased unprecedentedly in both developed and developing nations, and become one of the leading fiscal expenditures for medical care.1 As one of the most common non-communicable diseases, the prevalence of type 2 diabetes mellitus (T2DM) is expected increase to 592 million in 2035 all over the world.2 An optimal dietary approach would be a wise strategy for the prevention and management of insulin resistance and T2DM. The Mediterranean dietary pattern, the low-glycemic index diet, as well as whole grains and fiber have shown inverse associations with risk of T2DM.3, 4 In addition, fruits and vegetables, which are rich in fiber and phytochemicals, are universally believed to have beneficial health effects for preventing and managing T2DM.5, 6

Anthocyanins are phytochemicals found in fruits, vegetables, flowers and grains. As secondary metabolites of plant origin, anthocyanins naturally exist in the vacuolar sap of the epidermal tissues in plant kingdom, imparting pink, red, purple and blue colors.7, 8 Mounting findings have revealed that anthocyanins may have a crucial role in the prevention of obesity,9 progression of cancers10 and cardiovascular diseases (CVDs).11 As a major source of dietary anthocyanins, berries are widely consumed as fresh fruit, jams and juices in daily life and most commonly comprise blackberries, blueberries, raspberries, strawberries as well as other berries.12 Owing to their ability to ameliorate non-communicable diseases, the dietary intake of berry fruits has received increasing attention and research.13, 14 Prospective cohort studies including Nurses’ Health Study (NHS; 1984–2008) and Nurses’ Health Study II (NHS II; 1991–2007) revealed that dietary consumption of anthocyanins and berry fruits was inversely associated with the risk of T2DM, but not the Health Professionals Follow-Up Study (HPFS; 1986–2006).15, 16 Another prospective cohort study conducted in the US, however, showed that anthocyanin intake had no significant association with T2DM risk.17

So far, the associations of dietary intakes of anthocyanins and berry fruits with T2DM risk are inconsistent. Therefore, we conducted a systematic review and meta-analysis to clarify the associations of dietary anthocyanin and berry consumption with T2DM risk. Available data from included studies were pooled with the highest versus the lowest category to calculate overall risk and assess the dose–response associations.

Methods

Search strategy

A systematic literature search was performed in Cochrane library, Embase and PubMed databases up to Jan 2016 with anthocyanidin(s), anthocyanin(s), berries or individual names of both berries and anthocyanins paired with type 2 diabetes as the search terms. Moreover, references of the retrieved studies and recent reviews were scrutinized.

Eligibility criteria

The studies included in the present study met the following criteria: (1) prospective study including prospective cohort, nested case–control and case–cohort studies; (2) anthocyanins or berry fruits as exposure; (3) T2DM outcome of interest; (4) the hazard risk (HR), relative risk (RR) or odds ratio (OR) with 95% confidence interavl (CI) was provided. If data were duplicated in more than one study, the most complete and detailed study was included in the present study.

Data extraction and quality evaluation

The data of original studies were independently extracted by Xiao-fei Guo and Bo Yang. The following information was extracted from each study: the author’s family name, published year, region, gender, length of follow-up, mean age at baseline, T2DM incident cases and participants, RRs and 95% CIs. Any discrepancies were resolved with the third investigator (Duo Li). The dietary intake of anthocyanins was calculated in NHS, and NHS II and HPFS, respectively, using the United States of Agriculture Department (USDA) flavonoid content of the foods database.16 As the studies used different measurements for dietary intake of berries such as serving/week, serving/day and g/day, using a standard serving of 106 g,18 we standardized all data into g/day. Quality evaluation was performed based on the Newcastle–Ottawa Scale with a 9-star system.19 The scoring system summarized nine aspects of each study. The full score was defined as 9 stars, and a study with stars of 0–3, 4–6 and 7–9 was regarded as low, moderate and high quality, respectively.

Statistical analysis

As the Cox’s model was used for the included studies, RR was identical to HR in the studies. Therefore, multivariate-adjusted RRs (HRs) with 95% CIs (comparing the highest with the lowest category) were pooled to estimate summary RRs for the associations of dietary intakes of anthocyanins and berries with T2DM risk. The RRs and 95% CIs were logarithm transformed, and the summary RR with 95% CI was calculated using a random effects model.20

The heterogeneity among studies was evaluated by I2 statistics. The degree of heterogeneity was defined by I2 values of 25, 50 and 75% as cut-off points regarded as low, moderate and high, respectively.21 The I2 statistic represents the proportion of total variation, which is attributed to between-study heterogeneity.22 The I2 value >50% was considered as indicative of heterogeneity based on the Cochrane handbook.23 Subgroup analysis was carried out to explore the potential sources of heterogeneity by mean age at baseline (50 and <50 years), gender (female and male) and geographical region (US and Europe). Meanwhile, meta-regression with restricted maximum likelihood estimation was performed to assess the potential impact of covariates on the heterogeneity between groups. Sensitivity analysis was performed by omitting one study at a time and evaluating the influence of each individual study for potential bias on the overall risk estimate.

Furthermore, dose–response meta-analysis was performed to explore the potential trend estimation. Independent studies with three or more categories were eligible for the dose–response analysis. Midpoint of the upper and lower boundaries in respective quantiles was assigned as the dose if the study only reported the range. The dose was defined as 1.2-fold highest boundary if the highest quantile was open-ended;24 the midpoint of lowest boundary and zero was calculated as the dose of the lowest quantile when the lowest quantile or reference category was open-ended.25 A curvilinear association was conducted with a 2-stage random dose–response meta-analysis to evaluate the association.22, 26 The curvilinear trend was examined by modeling dietary consumption of anthocyanins and berries using restricted cubic splines with three knots (two spline transformations) at fixed percentiles (25, 50 and 75%) of the distribution, respectively.27 A P value for possible curvilinear association was calculated by examining the null hypothesis that the coefficient of the second spline is equal to 0.28 Linear trend was tested by using the method described by Greenland and Longnecker and Orsini et al.26, 29 In the presence of substantial linear trends (P value for nonlinear trend >0.05), the linear dose–response analysis was performed to evaluate the associations between increment of anthocyanins (7.5 mg/day) and berries (17 g/day) and decline in RRs of T2DM, respectively.26 Stata 11.0 was used for statistical analysis and P value was two-tailed with a statistically significant level of 0.05 (Stata Corporation, College Station, TX, USA).

Results

Literature search

Figure 1 shows the process of literature search. A total of 814 articles were identified from Cochrane library, Embase and PubMed after deleting duplicated articles. After ruling out irrelevant research, cell and animal models and review articles, 11 studies were remained. After review of full-text, four studies were eligible for the present meta-analysis.14, 15, 16, 30 Jacques et al.17 have reported an association between each 2.5-fold increase in dietary anthocyanin intake and T2DM risk, but did not report the association (that is, HR and CI) between dietary anthocyanin intake and T2DM risk. Therefore, the study was not eligible for this present study.

Figure 1
figure1

The PRISMA flow diagram for detailed steps of literature search.

Study characteristics

A total of five prospective cohort studies were eligible for the present study. NHS, NHS II and HPFS were performed in the US;15, 16 Finnish Mobile Clinic Health Examination Survey (FMCHES) and Kuopio Ischaemic Heart Disease Risk Factor Study (KIHDRFS) were conducted in Finland.14, 30 One study reported the association of dietary anthocyanin intake with T2DM risk, including NHS, NHS II and HPFS with a total of 200 894 participants, of which 12 611 participants were diagnosed as T2DM incident cases.16 The association of berry intake with T2DM contained five separate prospective cohort studies involved with 194 019 participants, of which 13 013 subjects were ascertained as T2DM incident cases during follow-up periods ranging from 16 to 24 years.14, 15, 30 The detailed information of the studies is listed in Table 1. NHS, NHS II and HPFS did not reported the association of total berry intake with T2DM risk, but reported the associations with T2DM risk for blueberries, grapes (and raisins) and strawberries individually. Therefore, the summary RR was pooled with a fixed effect model in the separate study, respectively.15 The total of included studies showed high quality according to the Newcastle–Ottawa Scale (Table 2).

Table 1 Characteristics of included cohorts for dietary anthocyanin intake in relation to T2DM risk
Table 2 Quality assessment of each cohort with the Newcastle–Ottawa Scale

Highest versus lowest category

As shown in Figure 2, the pooled effect indicated that anthocyanin intake significantly decreased the risk of T2DM (summary RR=0.85; 95% CI: 0.80–0.91; P<0.001) with low between-study heterogeneity (I2=14.5%, P=0.310). The pooled effect of berry intake also significantly decreased the risk of T2DM (summary RR=0.82, 95% CI: 0.76–0.89; P<0.001) with moderate heterogeneity (I2=48.6%, P=0.100). Therefore, stratified analysis was conducted to confirm the potential source of heterogeneity. In subgroup analysis (Table 3), stratified by region indicated that berry intake was significantly associated with decreased risk of T2DM in both the US (summary RR=0.85; 95% CI: 0.81–0.89; P<0.001) and Europe (summary RR=0.64; 95% CI: 0.52–0.79; P<0.001). Although berry intake was significantly associated with decreased risk of T2DM in female participants (summary RR=0.84; 95% CI: 0.81–0.88; P<0.001), no significant association was observed in male participants. There was no evidence that the pooled RRs differed significantly between subgroups according to gender and mean age at baseline (P for meta-regression >0.05). Incidentally, there was a borderline significant heterogeneity between subgroups stratified by location (US and Europe) with meta-regression analysis (P for meta-regression=0.09). For sensitivity analysis, each study was sequentially omitted at a time and the remaining data were re-analyzed. The pooled effects of dietary intakes of anthocyanins and berries with T2DM risk did not substantially change the overall associations (Supplementary Figure 1).

Figure 2
figure2

The associations of dietary intakes of anthocyanins (a) and berries (b) with risk of T2DM comparing the highest with lowest category. The size of the gray box representing each risk estimate was proportional to the weight. The diamonds represent summary relative risk.

Table 3 Subgroup analysis for the association between berry intake and T2DM risk according to study characteristics

Dose–response analysis

Three separate cohort studies were eligible for the dose–response analysis between anthocyanin intake and T2DM risk.16 The association of dietary anthocyanin intake with T2DM risk demonstrated a pronounced nonlinear trend (P for nonlinearity=0.006, P for trend<0.001) (Figure 3a), with an increase in 7.5 mg/day anthocyanin intake associated with a 5% reduction in the risk of T2DM (RR=0.95; 95% CI: 0.93–0.98; I2=0.00%). There were four independent cohorts that were eligible for the dose–response analysis between berry intake and T2DM risk.14, 15 The association between berry consumption and T2DM risk also demonstrated a pronounced nonlinearity trend (P for nonlinearity=0.028, P for trend<0.001) (Figure 3b), with an increase in 17 g/day berry consumption associated with a 5% reduction in the risk of T2DM (RR=0.95, 95% CI: 0.91–0.99; I2=0.00%).

Figure 3
figure3

Dose–response analysis between dietary intakes of anthocyanins (a) and berries (b) and risk of T2DM. Nonlinear dose–response trend analysis was assessed by restricted cubic spline model with three knots. Adjusted RRs from all categories in each study were separately represented by the small gray circle, and corresponding nonlinear relationships were represented by the black solid line using restricted cubic splines functional model with three knots at fixed percentiles 25, 50 and 75% of the distribution.

Discussion

As far as we know, the present study is the first to evaluate the association of dietary consumption of anthocyanins and berries with T2DM risk. Collectively, the findings indicate that dietary intakes of anthocyanins and berries significantly decreased the risk of T2DM (P<0.001). Dose–response meta-analysis demonstrated nonlinear relationships of dietary intakes of anthocyanins (P for nonlinearity=0.006, P for trend <0.001) and berries (P for nonlinearity=0.028, P for trend <0.001) with T2DM risk, respectively. The risk of T2DM decreased by 5%, with a 7.5 mg/day increment of dietary anthocyanin (RR=0.95; 95% CI: 0.93-0.98; I2=0.00%) or a 17 g/day berry intake (RR=0.95, 95% CI: 0.91–0.99; I2=0.00%), respectively.

Owing to the lack of anthocyanin food composition data and the difficulties in separation and quantification for these compounds, the number of randomized controlled trails (RCTs) was limited. A double-blinded, randomized and placebo-controlled 6-week intervention clinical study was performed with 32 obese participants who had insulin resistance. The results revealed that supplementation with anthocyanin extracts from blueberries significantly improved insulin sensitivity compared with the control group, but not adiposity, energy intake and inflammatory biomarkers.31 In addition, a 12 week RCT intervention in Guangdong Province, China, was carried out with 58 T2DM participants. Supplementation with purified anthocyanins significantly decreased fasting blood glucose and improved serum adiponectin levels, compared with the control group.32 Furthermore, a 24 week RCT was conducted with 150 hypercholesterolemic participants. Supplementation with purified anthocyanins significantly decreased the serum levels of C-reactive protein, soluble cell adhesion molecule-1, plasma interleukin-1β and low-density lipoprotein-cholesterol and significantly increased high-density lipoprotein-cholesterol.33

Consistent with RCTs, in a cross-sectional study of 1997 female participants from Britain, dietary anthocyanin consumption, which was calculated by food-frequency questionnaires using the USDA database, was associated with decreased insulin resistance and inflammatory biomarkers in blood.34 On the contrary, in a Framingham Offspring cohort study of 2915 participants, no association was found between dietary anthocyanin intake and T2DM risk after adjusting for gender, age, CVDs, smoke, BMI and cumulative mean energy intake.17 In line with Framingham Offspring cohort, HPFS also indicated no association of dietary intakes of anthocyanins and berries with T2DM risk in male participants.15, 16 Although the associations of dietary anthocyanin and berry consumption with T2DM risk were inconsistent, the pooled estimates of RRs revealed that dietary intakes of anthocyanins and berries significantly decreased the risk of T2DM, respectively. The underlying mechanisms of anthocyanins that reduce the risk of T2DM involve the following aspects:

(1) Antioxidant capacity of anthocyanins is a key reason for the prevention of T2DM. Supplementation with antioxidants has been shown to reduce T2DM risk via regulation of inflammatory response, reduction of blood glucose and insulin resistance.8 The antioxidant properties of anthocyanins were assessed systematically in vitro, including 1,1-diphenyl-2-picrylhydrazyl, radical-scavenging activity, oxygen radical absorbance capacity and ferric reducing antioxidant potential.35, 36, 37 Moreover, anthocyanins decreased free radical generation and increased hepatic superoxide dismutase and catalase (H2O2) capacities in a rodent model.38 Pelargonidin, an anthocyanin without a sugar, increased levels of superoxide dismutase and glutathione (GSH) and decreased levels of malondialdehyde and fructosamine in diabetic rats.39 In another study, hydrogen bonding and hydrophobic forces had crucial roles in forming DNA–anthocyanin complexes that prevented hydroxyl radical DNA damage.40 Finally, anthocyanins improve antioxidant defense by the reduction of reactive oxygen species and oxidative stress.38, 39

(2) The anti-inflammatory capability of anthocyanins is of remarkable importance for the progression of T2DM. Chronic low-grade inflammation is implicated in the initiation, propagation and progression of metabolic syndrome, CVDs and T2DM. There is convincing evidence that the transcription factor NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) signaling pathway has a crucial role in coordinating inflammatory responses.41 Anthocyanin extracts from berries or black soybean showed anti-inflammatory responses by blocking nuclear translocation of NF-κB in cell models.42 Cell model findings appear to be consistent with rodent studies, indicating that intervention of anthocyanins was associated with down regulation of inflammatory cytokine genes (tumor necrosis factor-α, interleukin-6, monocyte chemoattractant protein-1) in adipose tissue.8 It has been reported that lipopolysaccharide was shown to induce cyclooxygenase-2 expression in murine macrophages, while supplementation with anthocyanins inhibited cyclooxygenase-2 expression in lipopolysaccharide-evoked macrophages.43 In line with the cell mode, a rodent model also showed that anthocyanin intervention suppressed biological activity of cyclooxygenase-2.44 Therefore, anthocyanins inhibit the turnover of omega-6 eicosanoids (for example, prostaglandin E2 (PGE2) and leukotriene B4 (LTB4)).

(3) The third aspect of anthocyanin intake in preventing and ameliorating T2DM pertains to the metabolism of glucose and lipids. Anthocyanin administration has beneficial effects in maintaining glucose and lipid homeostasis. Anthocyanin extracts from berries revealed effective inhibition of α-glucosidase and maltase activity. Thus, postprandial glucose levels decreased, by delaying gastric emptying and blocking transport of glucose through intestinal brush border membrane.42 Glucose transporter 4 (GLUT4) is primarily present in skeletal muscle, myocardium, and adipose tissue. Several compelling findings revealed that administration of anthocyanins significantly enhanced GLUT4 membrane localization in diabetic rodent models, thereby reducing hyperglycemia and insulin resistance.45, 46 Peroxisome proliferator-activated receptor-γ is associated with promoting insulin sensitivity, anti-inflammatory actions and controlling glucose and lipid metabolism.47 Consistent findings indicated that peroxisome proliferator-activated receptor-γ could be activated by anthocyanins in cell models, thereby increasing glucose and lipid uptake and ameliorating insulin resistance.46, 48 In addition, adenosine monophosphate-activated protein kinase (AMPK) has crucial roles in regulating cellular energy homeostasis and in stimulating insulin secretion by pancreatic β-cell.49 Several studies found that AMPK can be triggered by administration of anthocyanins in murine models.50, 51 The activation of AMPK can improve glucose utilization in muscle and suppress gluconeogenesis. The triggering of AMPK was also accompanied by the suppression of cholesterol and triacylglycerol synthesis.49, 50, 51 As mentioned above, the underlying mechanisms of anthocyanins that may prevent T2DM depend mainly on reduction of oxidative stress and inflammatory signaling and the modulation of glucose and lipid homeostasis.

Strengths and limitations

The strengths of the present study should be highlighted. First, the large sample size provided strong statistical power to quantitatively evaluate the associations. In addition, the studies included in the meta-analysis were prospective cohort studies with high-quality scores, thereby reducing the possibility of recall errors and selection bias.

Simultaneously, there are also several potential limitations. First, the meta-analysis was performed on observational studies, which are inevitably susceptible to inherent biases. Although the extracted RRs were adjusted for covariates, report biases and residual confounding factors might influence the pooled effects. Second, due to the lack of blood biomarkers, dietary anthocyanin intake, which was calculated using food-frequency questionnaires, might have recording errors. Third, a limited number of cohorts recorded detailed information on the associations of anthocyanin and berry consumption with T2DM risk, thus only five cohort studies were eligible for the present study. In addition, all of the studies were conducted in western countries, thereby lacking information on other regions and ethnic groups.

Conclusion

T2DM and its associated complications cause serious medical and socioeconomic burdens. The findings from the present meta-analysis provide sufficient evidence that dietary intakes of anthocyanins and berries are associated with a lower risk of T2DM, respectively. More prospective studies in other regions and ethnic groups are warranted to further explore the associations of dietary anthocyanins and berries with T2DM risk.

References

  1. 1

    Daar AS, Singer PA, Persad DL, Pramming SK, Matthews DR, Beaglehole R et al. Grand challenges in chronic non-communicable diseases. Nature 2007; 450: 494–496.

    CAS  Article  Google Scholar 

  2. 2

    Guariguata L, Whiting D, Hambleton I, Beagley J, Linnenkamp U, Shaw J . Global estimates of diabetes prevalence for 2013 and projections for 2035. Diabetes Res Clin Pract 2014; 103: 137–149.

    CAS  Article  Google Scholar 

  3. 3

    Ajala O, English P, Pinkney J . Systematic review and meta-analysis of different dietary approaches to the management of type 2 diabetes. Am J Clin Nutr 2013; 97: 505–516.

    CAS  Article  Google Scholar 

  4. 4

    Montonen J, Knekt P, Järvinen R, Aromaa A, Reunanen A . Whole-grain and fiber intake and the incidence of type 2 diabetes. Am J Clin Nutr 2003; 77: 622–629.

    CAS  Article  Google Scholar 

  5. 5

    Cooper AJ, Forouhi NG, Ye Z, Buijsse B, Arriola L, Balkau B et al. Fruit and vegetable intake and type 2 diabetes: EPIC-InterAct prospective study and meta-analysis. Eur J Clin Nutr 2012; 66: 1082–1092.

    CAS  Article  Google Scholar 

  6. 6

    Ley SH, Hamdy O, Mohan V, Hu FB . Prevention and management of type 2 diabetes: dietary components and nutritional strategies. Lancet 2014; 383: 1999–2007.

    CAS  Article  Google Scholar 

  7. 7

    Castaneda-Ovando A, de Lourdes Pacheco-Hernández M, Páez-Hernández ME, Rodríguez JA, Galán-Vidal CA . Chemical studies of anthocyanins: a review. Food Chem 2009; 113: 859–871.

    CAS  Article  Google Scholar 

  8. 8

    Guo H, Ling W . The update of anthocyanins on obesity and type 2 diabetes: experimental evidence and clinical perspectives. Rev Endocr Metab Dis 2015; 16: 1–13.

    Article  Google Scholar 

  9. 9

    Kwon S-H, Ahn I-S, Kim S-O, Kong C-S, Chung H-Y, Do M-S et al. Anti-obesity and hypolipidemic effects of black soybean anthocyanins. J Med Food 2007; 10: 552–556.

    CAS  Article  Google Scholar 

  10. 10

    Wang L-S, Stoner GD . Anthocyanins and their role in cancer prevention. Cancer Lett 2008; 269: 281–290.

    CAS  Article  Google Scholar 

  11. 11

    Wallace TC . Anthocyanins in cardiovascular disease. Adv Nutr 2011; 2: 1–7.

    CAS  Article  Google Scholar 

  12. 12

    Wu X, Beecher GR, Holden JM, Haytowitz DB, Gebhardt SE, Prior RL . Concentrations of anthocyanins in common foods in the United States and estimation of normal consumption. J Agric Food Chem 2006; 54: 4069–4075.

    CAS  Article  Google Scholar 

  13. 13

    Seeram NP . Berry fruits: compositional elements, biochemical activities, and the impact of their intake on human health, performance, and disease. J Agric Food Chem 2008; 56: 627–629.

    CAS  Article  Google Scholar 

  14. 14

    Mursu J, Virtanen JK, Tuomainen T-P, Nurmi T, Voutilainen S . Intake of fruit, berries, and vegetables and risk of type 2 diabetes in Finnish men: the Kuopio Ischaemic Heart Disease Risk Factor Study. Am J Clin Nutr 2014; 99: 328–333.

    CAS  Article  Google Scholar 

  15. 15

    Muraki I, Imamura F, Manson JE, Hu FB, Willett WC, van Dam RM et al. Fruit consumption and risk of type 2 diabetes: results from three prospective longitudinal cohort studies. Brit Med J 2013; 347: f5001.

    Article  Google Scholar 

  16. 16

    Wedick NM, Pan A, Cassidy A, Rimm EB, Sampson L, Rosner B et al. Dietary flavonoid intakes and risk of type 2 diabetes in US men and women. Am J Clin Nutr 2012; 95: 925–933.

    CAS  Article  Google Scholar 

  17. 17

    Jacques PF, Cassidy A, Rogers G, Peterson JJ, Meigs JB, Dwyer JT . Higher dietary flavonol intake is associated with lower incidence of type 2 diabetes. J Nutr 2013; 143: 1474–1480.

    CAS  Article  Google Scholar 

  18. 18

    Dauchet L, Amouyel P, Hercberg S, Dallongeville J . Fruit and vegetable consumption and risk of coronary heart disease: a meta-analysis of cohort studies. J Nutr 2006; 136: 2588–2593.

    CAS  Article  Google Scholar 

  19. 19

    Stang A . Critical evaluation of the Newcastle–Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010; 25: 603–605.

    Article  Google Scholar 

  20. 20

    DerSimonian R, Laird N . Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–188.

    CAS  Article  Google Scholar 

  21. 21

    Higgins JPT, Thompson SG, Deeks JJ, Altman DG . Measuring inconsistency in meta-analyses. Brit Med J 2003; 327: 557–560.

    Article  Google Scholar 

  22. 22

    Jackson D, White IR, Thompson SG . Extending DerSimonian and Laird's methodology to perform multivariate random effects meta-analyses. Stat Med 2010; 29: 1282–1297.

    Article  Google Scholar 

  23. 23

    Higgins JPT, Green S . Cochrane Handbook for Systematic Reviews of Interventions 4.2.5: The Cochrane Library. Chichester: John Wiley, 2005.

    Google Scholar 

  24. 24

    Liu Q, Cook NR, Bergström A, Hsieh C-C . A two-stage hierarchical regression model for meta-analysis of epidemiologic nonlinear dose-response data. Comput Stat Data Anal 2009; 53: 4157–4167.

    Article  Google Scholar 

  25. 25

    Yang B, Wang F-L, Ren X-L, Li D . Biospecimen long-chain n-3 PUFA and risk of colorectal cancer: a meta-analysis of data from 60,627 individuals. PLoS ONE 2014; 9: e110574.

    Article  Google Scholar 

  26. 26

    Orsini N, Bellocco R, Greenland S . Generalized least squares for trend estimation of summarized dose-response data. Stata J 2006; 6: 40–57.

    Article  Google Scholar 

  27. 27

    Harre FE, Lee KL, Pollock BG . Regression models in clinical studies: determining relationships between predictors and response. J Natl Cancer 1988; 80: 1198–1202.

    Article  Google Scholar 

  28. 28

    Orsini N, Li R, Wolk A, Khudyakov P, Spiegelman D . Meta-analysis for linear and nonlinear dose-response relations: examples, an evaluation of approximations, and software. Am J Epidemiol 2012; 175: 66–73.

    Article  Google Scholar 

  29. 29

    Greenland S, Longnecker MP . Methods for trend estimation from summarized dose-response data, with applications to meta-analysis. Am J Epidemiol 1992; 135: 1301–1309.

    CAS  Article  Google Scholar 

  30. 30

    Montonen J, Järvinen R, Heliövaara M, Reunanen A, Aromaa A, Knekt P . Food consumption and the incidence of type II diabetes mellitus. Eur J Clin Nutr 2005; 59: 441–448.

    CAS  Article  Google Scholar 

  31. 31

    Stull AJ, Cash KC, Johnson WD, Champagne CM, Cefalu WT . Bioactives in blueberries improve insulin sensitivity in obese, insulin-resistant men and women. J Nutr 2010; 140: 1764–1768.

    CAS  Article  Google Scholar 

  32. 32

    Liu Y, Li D, Zhang Y, Sun R, Xia M . Anthocyanin increases adiponectin secretion and protects against diabetes-related endothelial dysfunction. Am J Physiol Endocrinol Metab 2014; 306: E975–E988.

    CAS  Article  Google Scholar 

  33. 33

    Zhu Y, Ling W, Guo H, Song F, Ye Q, Zou T et al. Anti-inflammatory effect of purified dietary anthocyanin in adults with hypercholesterolemia: a randomized controlled trial. Nutr Metab Cardiovasc Dis 2013; 23: 843–849.

    CAS  Article  Google Scholar 

  34. 34

    Jennings A, Welch AA, Spector T, Macgregor A, Cassidy A . Intakes of anthocyanins and flavones are associated with biomarkers of insulin resistance and inflammation in women. J Nutr 2014; 144: 202–208.

    CAS  Article  Google Scholar 

  35. 35

    Kano M, Takayanagi T, Harada K, Makino K, Ishikawa F . Antioxidative activity of anthocyanins from purple sweet potato, Ipomoea batatas cultivar Ayamurasaki. Biosci Biotechnol Biochem 2005; 69: 979–988.

    CAS  Article  Google Scholar 

  36. 36

    Wang H, Cao G, Prior RL . Oxygen radical absorbing capacity of anthocyanins. J Agric Food Chem 1997; 45: 304–309.

    CAS  Article  Google Scholar 

  37. 37

    Li H, Deng Z, Zhu H, Hu C, Liu R, Young JC et al. Highly pigmented vegetables: anthocyanin compositions and their role in antioxidant activities. Food Res Int 2012; 46: 250–259.

    CAS  Article  Google Scholar 

  38. 38

    Chiang A-N, Wu H-L, Yeh H-I, Chu C-S, Lin H-C, Lee W-C . Antioxidant effects of black rice extract through the induction of superoxide dismutase and catalase activities. Lipids 2006; 41: 797–803.

    CAS  Article  Google Scholar 

  39. 39

    Roy M, Sen S, Chakraborti AS . Action of pelargonidin on hyperglycemia and oxidative damage in diabetic rats: implication for glycation-induced hemoglobin modification. Life Sci 2008; 82: 1102–1110.

    CAS  Article  Google Scholar 

  40. 40

    Zhang C, Guo X, Cai W, Ma Y, Zhao X . Binding characteristics and protective capacity of cyanidin-3-glucoside and its aglycon to calf thymus DNA. J Food Sci 2015; 80: H889–H893.

    CAS  Article  Google Scholar 

  41. 41

    Baker RG, Hayden MS, Ghosh S . NF-κB, inflammation, and metabolic disease. Cell Metab 2011; 13: 11–22.

    CAS  Article  Google Scholar 

  42. 42

    Dembinska-Kiec A, Mykkänen O, Kiec-Wilk B, Mykkänen H . Antioxidant phytochemicals against type 2 diabetes. Brit J Nutr 2008; 99: ES109–ES117.

    Article  Google Scholar 

  43. 43

    Hou D-X, Yanagita T, Uto T, Masuzaki S, Fujii M . Anthocyanidins inhibit cyclooxygenase-2 expression in LPS-evoked macrophages: structure–activity relationship and molecular mechanisms involved. Biochem Pharmacol 2005; 70: 417–425.

    CAS  Article  Google Scholar 

  44. 44

    Hassimotto NMA, Moreira V, Nascimento N.G.D, Souto P.C.M.D.C, Teixeira C, Lajolo FM . Inhibition of carrageenan-induced acute inflammation in mice by oral administration of anthocyanin mixture from wild mulberry and cyanidin-3-glucoside. Biomed Res Int 2013; 2013: 146716.

    Article  Google Scholar 

  45. 45

    Nizamutdinova IT, Jin YC, Chung JI, Shin SC, Lee SJ, Seo HG et al. The anti-diabetic effect of anthocyanins in streptozotocin-induced diabetic rats through glucose transporter 4 regulation and prevention of insulin resistance and pancreatic apoptosis. Mol Nutr Food Res 2009; 53: 1419–1429.

    CAS  Article  Google Scholar 

  46. 46

    Sasaki R, Nishimura N, Hoshino H, Isa Y, Kadowaki M, Ichi T et al. Cyanidin 3-glucoside ameliorates hyperglycemia and insulin sensitivity due to downregulation of retinol binding protein 4 expression in diabetic mice. Biochem Pharmacol 2007; 74: 1619–1627.

    CAS  Article  Google Scholar 

  47. 47

    Calder PC . Mechanisms of action of (n-3) fatty acids. J Nutr 2012; 142: 592S–599S.

    CAS  Article  Google Scholar 

  48. 48

    Jia Y, Kim J-Y, Jun H-j, Kim S-J, Lee J-H, Hoang MH et al. Cyanidin is an agonistic ligand for peroxisome proliferator-activated receptor-alpha reducing hepatic lipid. Biochim Biophys Acta 2013; 1831: 698–708.

    CAS  Article  Google Scholar 

  49. 49

    Winder W, Hardie D . AMP-activated protein kinase, a metabolic master switch: possible roles in type 2 diabetes. Am J Physiol Endocrinol Metab 1999; 277: E1–E10.

    CAS  Article  Google Scholar 

  50. 50

    Wei X, Wang D, Yang Y, Xia M, Li D, Li G et al. Cyanidin-3-O-β-glucoside improves obesity and triglyceride metabolism in KK-Ay mice by regulating lipoprotein lipase activity. J Sci Food Agric 2011; 91: 1006–1013.

    CAS  Article  Google Scholar 

  51. 51

    Kurimoto Y, Shibayama Y, Inoue S, Soga M, Takikawa M, Ito C et al. Black soybean seed coat extract ameliorates hyperglycemia and insulin sensitivity via the activation of AMP-activated protein kinase in diabetic mice. J Agric Food Chem 2013; 61: 5558–5564.

    CAS  Article  Google Scholar 

Download references

Acknowledgements

We thank Dr Chao Zhang (Beijing Academy of Agriculture and Forestry Sciences) and Dr Jusheng Zheng (MRC Epidemiology Unit, University of Cambridge, Cambridge, United Kingdom) for the help of data analysis and discussion. This study was funded by the National Basic Research Program of China (973 Program: 2015CB553604); by National Natural Science Foundation of China (NSFC: 81273054); and by the PhD Programs Foundation of Ministry of Education of China (20120101110107). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Author information

Affiliations

Authors

Corresponding author

Correspondence to D Li.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on European Journal of Clinical Nutrition website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Guo, X., Yang, B., Tan, J. et al. Associations of dietary intakes of anthocyanins and berry fruits with risk of type 2 diabetes mellitus: a systematic review and meta-analysis of prospective cohort studies. Eur J Clin Nutr 70, 1360–1367 (2016). https://doi.org/10.1038/ejcn.2016.142

Download citation

Further reading

Search

Quick links