Skip to main content
. 2023 Aug 15;10:1207751. doi: 10.3389/fnut.2023.1207751

Table 2.

Subgroup analysis for the effect of anthocyanin on high-density lipoprotein-cholesterol.

Subgroup Effect size I2 2 (%) P-value3
K 1 (95% CI)
Overall 49 0.42 (0.20, 0.65) 81
Study design
Parallel study 41 0.52 (0.25, 0.79) 83 <0.001
Crossover study 8 −0.04 (−0.20, 0.12) 0
Study duration
<8 weeks 24 0.59 (0.18, 0.99) 85 0.220
≥8 weeks 25 0.30 (0.08, 0.51) 75
Cholesterol level 4
Normal (<200 mg/dL) 18 1.00 (0.42, 1.57) 91 0.020
Higher (≥ 200 mg/dL) 30 0.19 (0.06, 0.31) 51
Study area
East 20 0.80 (0.41, 1.19) 85 0.007
West 29 0.17 (−0.06, 0.40) 72
Participant's age
<40 years old 11 1.30 (0.54, 2.05) 91 0.007
≥40 years old 38 0.21 (0.07, 0.35) 68
No. of participants
<50 19 0.77 (0.27, 1.27) 86 0.052
≥50 30 0.24 (0.06, 0.42) 75
Anthocyanin dosage
<50 mg/day 10 0.54 (−0.06, 1.14) 83 0.099
≥50 and <350 mg/day 30 0.50 (0.18, 0.81) 85
≥350 mg/day 9 0.14 (−0.05, 0.32) 0
Formula
Low processing 20 0.15 (0.02, 0.27) 79 0.194
High processing 29 0.54 (0.26, 0.83) 81
Risk of bias
Low risk 42 0.51 (0.25, 0.77) 83 0.002
High risk 7 0.01(−0.17, 0.20) 0

1Number of studies combined. 2Overall test for heterogeneity within subgroups by the random effects model. 3Test for subgroup difference by random effect model. 4Excluding one study in which total cholesterol level was not presented at the baseline.