In total, 1030 potentially relevant articles were recognized from the following databases: 131 from PubMed, 353 from Science Direct, 138 from Cochrane library and 408 from Embase. After the primary evaluation, articles with unsuitable title or abstract were excluded (N=851). After removing duplicate articles (N=147), 32 articles remained for further assessment and 4 other articles were excluded due to: book chapters (2 studies) [22, 23], short communication (1 study) [24] and letter (1 study) [25] . We scrutinized the full text of remaining 28 articles (Figure 1).
Study characteristics
The features of 28 animal studies evaluated in this meta-analysis are demonstrated in Table 1. These studies were carried out in Asia[26-33], South America [34, 35] , Australia [36] , the USA [37-46] , Canada [47-51]and Europe, [52, 53] and published between 2007 and 2021. One study was done on pigs [47] , one on hamsters[28] , 15 on mice [27, 29-33, 35, 36, 38, 40, 42, 43, 45, 46, 51, 54] and 11 on rats [26, 34, 37, 39, 41, 44, 48-50, 52, 53]. Among these 28 studies, 5 studies were about hyperstension, 8 were about obesity, 8 were about insulin resistance and 7 were about lipid profile. The studies utilized different kinds of blueberry that included blueberry powder [36, 41, 46] , freeze-dried blueberry [26, 37-40, 43-45, 48, 49] , freeze-dried fermented blueberry [52] , blueberry purified anthocyanin [31], fermented blueberry beverage[42] , juice [29, 32, 51, 53] , puree [50] , blueberry extract [27, 28, 30, 33-35].Furthermore, blueberry dosage varied among different studies (shown in table 1) with different units. Study duration was different and varied between 2 and 21 weeks, and animal age was between 3-13 weeks.
Findings
Effects of blueberry on blood pressure
In the studies about blood pressure which were included in this review, there was a significant effect of blueberry on blood pressure reduction. For instance, Ahren et al. reported a significant decrease in both systolic and diastolic blood pressure (p = 0.008 and 0.013 respectively) after 4 weeks of blueberry feeding[52]. Shaughnessy et al. reported a 30 percent reduction in systolic blood pressure following 6 weeks of blueberry feeding[48] .
Systolic blood pressure Meta-analysis: Five studies [26, 37, 48, 49, 52] were eligible for this outcome, according to assessment of study homogeneity results, I2=40.88 and p=0.14 in Cochrane test, the studies were homogenous. The standardized mean difference between control and intervention groups was -0.59 (p<0.001, CI95%: (-0.916, -0.283). It was confirmed that the intervention had a significant effect on decreasing systolic blood pressure (Figure2.a, b)
Diastolic blood pressure Meta-analysis: Three studies [48, 49, 52] were eligible for this outcome. According to assessment of study homogeneity results, I2=62.11and p=0.07 in Cochrane test, the studies were homogenous. The standardized mean difference between control and intervention groups was -0.61 (p=0.01, CI95%: (-1.10, -0.13). It was confirmed that the intervention had a significant effect on decreasing diastolic blood pressure.
Distribution bias was not significant according to Kendall tau test (Begg-Mazumdar: Kendall's tau = 0, P = 1) (Figure3.a, b).
Effects of blueberry on Insulin Resistance and glucose
Lee et al., demonstrated that providing 10g/100g diet BB for 8 weeks for rats did not affect fasting glucose and insulin (p>0.05), but stimulated insulin sensitivity and signaling. Blood insulin and glucose had lower concentrations in the blueberry fed group after oral-glucose-challenge test (p<0.05 and p=0.07, respectively) [39] . Seymour et al, reported that 90days supplementation with 2% blueberry powder in obesity‐prone Zucker Fatty rats remarkably reduced fasting insulin and glucose(p<0/05) [41] .
Blood insulin and glucose meta-analysis: Regarding these outcomes four studies [29, 30, 40, 46]were eligible for meta-analysis. According to assessment of study homogeneity results, I2=37.76and p=0.18 in Cochrane test, the studies were homogenous. The standardized mean difference between control and intervention groups was -1.20 (p<0.001, CI95%: (-1.711, -0.688) and p< 0.001, CI95 %:( -1.638, -0.623) for insulin and glucose respectively. It was confirmed that the intervention had a significant effect on decreasing insulin and glucose.
Distribution bias was not significant according to Kendall tau test (Begg-Mazumdar: Kendall’s tau = -.83, P = 0.08 and Begg-Mazumdar: Kendall’s tau = -.16, P = 0.73 for glucose and insulin respectively) (Figure 4.a,b,c,d).
Effects of blueberry on Lipid Profile
Prior et al, reported that supplementing high-fat fed male mice with 10 % blueberry freeze-dried powder along with drinking water for 85 days caused a significant decline in triglycerides level(p<0.012) .[45] In the study conducted by Si et al, it was observed that blueberry caused a significant decrease in triglycerides (23.48%) and total cholesterol (23.29%) and a significant increase in HDL (33.06%) amounts [30].
Tg meta-analysis: Five studies [28, 29, 34, 35, 38] were eligible for meta-analysis of this outcome. According to assessment of study homogeneity results, I2=23.046and p=0.26 in Cochrane test, the studies were homogenous. The standardized mean difference between control and intervention groups was -0.56 (p=0.01, CI95%: (-1.018, -0.109). It was confirmed that the intervention had a significant effect on decreasing triglycerides.
Distribution bias was not significant according to Kendall tau test (Begg-Mazumdar: Kendall's tau = -.7, P = .08) (Figure5.a,b).
HDL meta-analysis: five studies [28, 29, 34, 38, 47]were eligible for meta-analysis. According to assessment of study homogeneity results, I2=40.84 and p=0.14 in Cochrane test, the studies were homogenous. The standardized mean difference between control and intervention groups was 0.56 (p=0.02, CI95%: (0.077, 1.057). It was confirmed that the intervention had a significant effect on increasing HDL.
Distribution bias was not significant according to Kendall tau test (Begg-Mazumdar: Kendall's tau =.5, P = .22) (Figure 5.c,d)
LDL meta-analysis: five studies [28, 29, 34, 38, 47]were eligible for meta-analysis. According to assessment of study homogeneity results, I2=64.59 and p=0.02 in Cochrane test, the studies were heterogeneous. The standardized mean difference between control and intervention groups was -1.26 (p=0.008, CI95%: (-2.182, -0.334). It was confirmed that the intervention had a significant effect on decreasing LDL.
Distribution bias was not significant according to Kendall tau test (Begg-Mazumdar: Kendall's tau =-.9, P = .02) (Figure 5.e,f).
TC meta-analysis: Six studies [28, 29, 34, 35, 38, 47]were eligible for the meta-analysis.. According to the results of study homogeneity assessment, based on Cochrane test (p=0.62) and I2=0.00, studies were considered homogeneous. Standardized mean difference between control and intervention groups was -1.23 (p<0.001 CI95%: (-1.693, -0.782)). It was confirmed that the intervention had a significant effects on reduction of TC.
Distribution bias was significant according to Kendall tau test (Begg-Mazumdar: Kendall's tau = -.9, P = .01) (Figure 5.g,h)
Effects of blueberry on body weight
In a study carried out by Wu et al., it was demonstrated that supplementation with 200mg/kg blueberry purified anthocyanin for 12 weeks in 60male mice prevented body weight gain by 55/4% (p<0.05) [31].
Also another study conducted by Wu et al, showed that consumption of blueberry juice for 12 weeks in 48 male mice reduced body weight by 7.3 % [32]. In another study Johnson et al, reported that 72 diet-induced obese mice fed with 30% blueberry beverage for 12 weeks had the lowest body weight gain in comparison to other groups(p<0.05)[42]. Wu et al, reported that 200 mg/kg artificial planting blueberry supplementation for 8 weeks in male high-fat diet-induced obese mice caused body weight gain to be effectively reduced by 19.4% [33]. In another study conducted by Liu et al., it was reported that 54 high-fat diet mice fed with 1% and 2% blueberry extract had decreased body weight gain by 20–23% (p<0.05).[27] DeFuria et al., found that 8 week supplementation with 20g/d blueberry powder in male mice did not influence high fat diet induced body weight gain(p<0.05) [27].
Body weight meta-analysis: Four studies [36, 44, 51, 53] were eligible for meta-analysis. According to the results of study homogeneity assessment, based on Cochrane test (p<0.001) and I2=93.27, studies were considered heterogeneous. Standardized mean difference between control and intervention groups was -0.57 (p=0.65 CI95%: (-3.029, 1.892)). It was confirmed that the intervention had no significant effects on reduction of body weight.
Distribution bias was not significant according to Kendall tau test (Begg-Mazumdar: Kendall's tau = 0.0, P = 1)(Figure 6.a,b).