Outcome . | Anticipated absolute effects* (95% CI) . | Participants (RCTs), n . | Certainty of the evidence (GRADE) . | Minimal clinically important difference . | Comments . |
---|---|---|---|---|---|
HbA1c (%) | MD 0.54 lower (0.90 lower to 0.17 lower) | 1,133 (16) | ⊕○○○ Very low†‡§‖ | ≥0.5 | Vitamin C supplementation may improve HbA1c to a clinically meaningful extent. Evidence rated down for inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, and 3) larger study sample sizes. |
Fasting glucose (mmol/L) | MD 0.74 lower (1.17 lower to 0.31 lower) | 1,305 (20) | ⊕○○○ Very low‡§‖ | ≥1 | Evidence shows a statistically significant but clinically insignificant reduction in fasting glucose. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, 3) larger study sample sizes, and 4) lower vitamin C supplementation doses. |
SBP (mmHg) | MD 6.27 lower (9.6 lower to 2.95 lower) | 466 (8) | ⊕⊕⊕○ Moderate§# | >2 | Evidence is suggestive of a hypotensive effect of vitamin C, with significant reductions observed that are consistent with a clinical improvement. Evidence rated down for indirectness (1 level). |
DBP (mmHg) | MD 3.77 lower (6.13 lower to 1.42 lower) | 466 (8) | ⊕○○○ Very low‡§‖ | >2 | Evidence is suggestive of a hypotensive effect of vitamin C, with significant reductions observed that are consistent with a clinical improvement. Evidence rated down for inconsistency (1 level), imprecision (1 level), and indirectness (1 level). |
Triglycerides (mmol/L) | MD 0.2 lower (0.36 lower to 0.04 lower) | 1,065 (17) | ⊕○○○ Very low‡§** | ≥1 | Evidence shows a small statistically significant but clinically insignificant reduction in triglycerides with vitamin C. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, and 3) larger study sample sizes. |
Total cholesterol (mmol/L) | MD 0.27 lower (0.43 lower to 0.1 lower) | 1,125 (19) | ⊕○○○ Very low‡§** | ≥1 | Evidence shows a small significant but clinically insignificant reduction in total cholesterol with vitamin C. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, and 3) larger study sample sizes. |
LDL cholesterol (mmol/L) | MD 0.23 lower (0.48 lower to 0.03 higher) | 988 (16) | ⊕○○○ Very low‡§** | ≥0.3 | Overall, evidence is not supportive of a statistically or clinically significant reduction in LDL cholesterol with vitamin C. Evidence rated down for inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, 3) larger study sample sizes, and 4) lower vitamin C supplementation doses. |
HDL cholesterol (mmol/L) | MD 0.06 higher (0.00–0.13 higher) | 1,022 (17) | ⊕○○○ Very low‡§| | ≥0.03 | Overall, evidence is not supportive of a statistically significant change in HDL cholesterol with vitamin C, although mean changes found are potentially clinically meaningful. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) a lower baseline BMI, 2) a lower adult age, 3) a longer study duration, and 4) larger sample sizes. |
Outcome . | Anticipated absolute effects* (95% CI) . | Participants (RCTs), n . | Certainty of the evidence (GRADE) . | Minimal clinically important difference . | Comments . |
---|---|---|---|---|---|
HbA1c (%) | MD 0.54 lower (0.90 lower to 0.17 lower) | 1,133 (16) | ⊕○○○ Very low†‡§‖ | ≥0.5 | Vitamin C supplementation may improve HbA1c to a clinically meaningful extent. Evidence rated down for inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, and 3) larger study sample sizes. |
Fasting glucose (mmol/L) | MD 0.74 lower (1.17 lower to 0.31 lower) | 1,305 (20) | ⊕○○○ Very low‡§‖ | ≥1 | Evidence shows a statistically significant but clinically insignificant reduction in fasting glucose. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, 3) larger study sample sizes, and 4) lower vitamin C supplementation doses. |
SBP (mmHg) | MD 6.27 lower (9.6 lower to 2.95 lower) | 466 (8) | ⊕⊕⊕○ Moderate§# | >2 | Evidence is suggestive of a hypotensive effect of vitamin C, with significant reductions observed that are consistent with a clinical improvement. Evidence rated down for indirectness (1 level). |
DBP (mmHg) | MD 3.77 lower (6.13 lower to 1.42 lower) | 466 (8) | ⊕○○○ Very low‡§‖ | >2 | Evidence is suggestive of a hypotensive effect of vitamin C, with significant reductions observed that are consistent with a clinical improvement. Evidence rated down for inconsistency (1 level), imprecision (1 level), and indirectness (1 level). |
Triglycerides (mmol/L) | MD 0.2 lower (0.36 lower to 0.04 lower) | 1,065 (17) | ⊕○○○ Very low‡§** | ≥1 | Evidence shows a small statistically significant but clinically insignificant reduction in triglycerides with vitamin C. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, and 3) larger study sample sizes. |
Total cholesterol (mmol/L) | MD 0.27 lower (0.43 lower to 0.1 lower) | 1,125 (19) | ⊕○○○ Very low‡§** | ≥1 | Evidence shows a small significant but clinically insignificant reduction in total cholesterol with vitamin C. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, and 3) larger study sample sizes. |
LDL cholesterol (mmol/L) | MD 0.23 lower (0.48 lower to 0.03 higher) | 988 (16) | ⊕○○○ Very low‡§** | ≥0.3 | Overall, evidence is not supportive of a statistically or clinically significant reduction in LDL cholesterol with vitamin C. Evidence rated down for inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) higher baseline HbA1c, 2) a longer supplementation duration, 3) larger study sample sizes, and 4) lower vitamin C supplementation doses. |
HDL cholesterol (mmol/L) | MD 0.06 higher (0.00–0.13 higher) | 1,022 (17) | ⊕○○○ Very low‡§| | ≥0.03 | Overall, evidence is not supportive of a statistically significant change in HDL cholesterol with vitamin C, although mean changes found are potentially clinically meaningful. Evidence rated down for risk of bias (1 level), inconsistency (1 level), imprecision (1 level), and indirectness (1 level). Subgroup and meta-regression analyses suggest the largest improvements with 1) a lower baseline BMI, 2) a lower adult age, 3) a longer study duration, and 4) larger sample sizes. |
The risk in the intervention group and its 95% CI are based on the assumed risk in the comparison group and the relative effect of the intervention and its 95% CI).
Overall, findings alternated from significantly favoring vitamin C (5 domains) to borderline (P = 0.05) significant effects (2 domains) when undertaking sensitivity analyses on the basis of different individual Cochrane risk-of-bias domains when using only low-risk studies; a decision was made to not rate down for risk of bias because of this relative consistency.
Significant heterogeneity in meta-analysis (I2 > 50%).
Surrogate outcome measure, not a patient-important end point.
Upper bound 95% CI of estimate outside of clinical meaningfulness. Overall findings alternated from significantly favoring vitamin C to null effects when undertaking sensitivity analyses on the basis of different individual Cochrane risk-of-bias domains when using only low-risk studies.
Overall findings alternated from significantly favoring vitamin C to null effects when undertaking sensitivity analyses on the basis of different individual Cochrane risk-of-bias domains (P < 0.05 for all domains except for allocation concealment [P = 0.06] and blinding of outcome assessment [P = 0.06]) when using only low-risk-of-bias studies; a decision was made to not rate down for risk of bias because of this relative consistency.
Upper and/or lower bounds of 95% CI not clinically meaningful.
Lower bound 95% CI of estimate outside of clinical meaningfulness.