Deeper Dive: Investigating the relationship between vitamin C, type 2 diabetes, and heart disease

Vitamin C supplementation seems to make a dent in blood pressure in people with diabetes. Its effect on glycemic control is less certain.

You are reading a free Examine Deep Dive.
Become an Examine+ member to get full access to the Examine Database, Study Summaries, and Supplement Guides.
Try for free for 7 days and stay on top of the latest research.
Introduction

Type 2 diabetes mellitus (T2DM) is characterized by elevated glucose (hyperglycemia), insulin resistance, and failure of insulin-secreting pancreatic beta cells[1] (seen as relative insulin deficiency and pancreatic inflammation and destruction[2]). T2DM is associated with obesity[2], oxidative stress[3], systemic inflammation[4], and heart disease. About 2 in 3 people with T2DM are insufficient or deficient in vitamin C, compared to about 1 in 4 people with normal glycemia, even if the people with T2DM are consuming enough vitamin C, according to one study[5] of 89 people. Across most studies, vitamin C levels are at least 30% lower among people with T2DM[6]. Mechanistically, vitamin C’s role in ameliorating aspects of T2DM also makes some sense. Some of the pathways by which it may help are laid out in Figure 1.

Figure 1: Some ways in which vitamin C may help dampen the impact of diabetes

image

References: Ellulu et al. Drug Des Devel Ther. 2015 Jul.[7]
Newsholme et al. Am J Physiol Cell Physiol. 2019 Sep.[8]
Will et al. Nutr Rev. 1996 Jul.[6]
Crabbe et al. Prog Retin Eye Res. 1998 Jul.[9]
Yan. Animal Model Exp Med. 2018 Mar.[10]
Haidara et al. Curr Vasc Pharmacol. 2006 Jul.[11]
Krone et al. Med Hypotheses. 2004.[12]
Taddei et al. Circulation. 1998 Jun.[13]
Mortensen et al. Nitric Oxide. 2014 Jan.[14]

Vitamin C supplementation has been associated with some improved glycemia outcomes (blood glucose, HbA1c, serum insulin, and insulin resistance) in previous meta-analyses, but results have been inconsistent. A 2011 systematic review and meta-analysis looking at the effect of supplementation with vitamin C and/or E on T2DM showed no effect on glucose or insulin and a small lowering effect on HbA1c. A 2014 meta-analysis of observational and randomized controlled trials showed a significant effect of vitamin C on decreasing fasting blood glucose, but no effect on HbA1c[15]. In this analysis, vitamin C was associated with lower levels of some oxidative stress markers, but was not associated with HbA1c or glucose outcomes. The most recent meta-analysis showed that, overall, vitamin C did not modify glucose, HbA1c or insulin concentrations[16]. Looking at subgroups in this study showed that vitamin C may have a more pronounced effect on fasting insulin than on post-meal insulin levels, on people diagnosed with T2DM than on people without a diagnosis, and when the duration of intervention was longer rather than shorter.

Other meta-analyses have focused on heart health measures related to T2DM. One 2012 meta-analysis concluded that vitamin C lowers blood pressure in all populations, and suggested that its effect on diastolic blood pressure may be greater for people with T2DM than for people without[17] T2DM. More recently, a meta-analysis looking at the effect of vitamin C on markers of endothelial function showed a strong beneficial relationship[18]. Finally, a 2016 meta-analysis looking exclusively at blood lipids found no significant effect of vitamin C[19].

None of these systematic reviews and meta-analyses assessed the certainty of evidence. With 10 new studies published since the previous 2017 systematic review on vitamin C and glycemic control, the present study was designed to clarify the risk of bias and update earlier estimates of vitamin C’s effect on T2D.

Some evidence suggests that vitamin C supplements may have positive effects on multiple aspects of diabetes, but the results of previous systematic reviews are mixed. Putative benefits include improvements of glycemia, blood lipids, and blood pressure. The most recent reviews of vitamin C’s effects on glycemic control and on blood lipids reported no significant associations.

What was studied?

This preregistered systematic review and meta analysis of randomized controlled trials looked at the effects of oral vitamin C supplementation on the primary outcomes of HbA1c, fasting blood glucose, blood lipids (triglycerides, cholesterol, LDL-C, HDL-C) and blood pressure (systolic and diastolic). Secondary outcomes included postprandial glucose, fasting insulin, insulin sensitivity measured by HOMA-IR, insulin sensitivity measured by clamp, and measures of oxidative stress. Studies with a duration of less than two weeks or using multiple antioxidants were excluded. Certainty of results was assessed using the GRADE system and Cochrane risk of bias tools. This is the first meta-analysis of these outcomes to use assessments of certainty. A random-effects model was used to pool the studies. The most prominent effects on heterogeneity and efficacy were due to study sample size, study duration, and baseline HbA1c.

The researchers included a total of 28 studies, though only 8 had a mostly low risk of bias, defined as 4 of 7 opportunities for bias being evaluated as “low risk.” Differences in included studies were large: average participant age ranged from 38 to 72 years, participants may or may not have been using medications, baseline HbA1c ranged from 6.5% (minimum for T2DM diagnosis[20]) to 9.1% (a very high level associated with adverse complications[20]) and baseline serum vitamin C levels ranged from deficient to mid-level sufficiency. Doses ranged from 200 to 3,000 mg per day (median: 500 mg) for 14 days to 1 year. Studies ranged from 8 to 456 participants.

This study was a systematic review and meta analysis of randomized controlled trials testing the effect of vitamin C on glycemia, blood lipids, and blood pressure. This is the first systematic review to use the GRADE system to assess certainty, which is important because previously reviewed studies were inconclusive. This systematic review and meta-analysis included 10 additional studies published after 2017, including more recent large and long-term (longer than 12 weeks) studies.

What were the findings?

Vitamin C supplementation resulted in statistically significant, clinically meaningful decreases in HbA1c and blood pressure. HbA1c was lowered by a little more than 0.5%. For reference, metformin, the first-line approach for pre-diabetes management, lowers HbA1c by a little more than 1%[21]. Fasting blood glucose decreased, but only to a very small extent. Statistically significant reductions in triglycerides and total cholesterol were not clinically meaningful, and no changes were seen for LDL-C and HDL-C. The evidence in support of an effect on systolic blood pressure was moderately certain, but all other results had a very low level of certainty. Very low quality evidence suggests that fasting blood glucose decreased by a statistically significant, but clinically irrelevant, level of 0.74 mmol/L. The statistically significant and clinically relevant changes are summarized in Figure 2.

Figure 2. Statistically Significant and Clinically Relevant Findings.

Outcome (units)Mean difference with vitamin C supplementationGRADE-measured certaintyHeterogeneity (I^2^; low, medium, high)Notes
HbA1c (%)-0.54Very Low88.70%, highEffect size seems to increase in larger and longer studies
Systolic Blood Pressure (mmHg)-6.27Moderate48.76%, moderate
Diastolic Blood Pressure (mmHg)-3.77Very Low57.91%, high

In terms of secondary endpoints, vitamin C reduced postprandial glucose and showed a trend (p=.06) for an association with lower fasting insulin levels. No effect was seen on insulin resistance. The only measure of lipid peroxidation that vitamin C had an positive effect on was malondialdehyde, which decreased by a standard mean difference of 1.25, 95% confidence interval: -1.88 to -0.62. For reference, other research shows that people with T2DM using insulin have about 1.5 times the malondialdehyde levels as people without diabetes serving as controls[22].

In terms of subgroups, the largest and longest studies were the ones with the most significant effect on HbA1c and fasting blood glucose. Studies of participants with the highest baseline HbA1c also had the largest reductions in HbA1c. This is potentially a spurious finding and an example of confounding due to regression to the mean. The authors also found a significant difference in the beneficial effect of supplementation on HbA1c for people with vitamin C insufficiency (less than 23 μmol/L), compared to people with sufficient levels. Baseline age, BMI, diabetes duration, and vitamin C dose had no significant effect on vitamin C efficacy.

Vitamin C supplementation was associated with statistically significant and clinically meaningful reductions in HbA1c, systolic blood pressure, and diastolic blood pressure. The secondary outcome malondialdehyde improved as well. Associations with fasting blood glucose, total cholesterol, HDL-C, LDL-C, and triglycerides were either not statistically significant or not clinically relevant. The only measure with a moderate degree of certainty was systolic blood pressure. All other measures had very low certainty. Secondary analysis suggested a relationship between study size and duration and vitamin C efficacy.

The bigger picture

This analysis is the first to include the largest and longest studies conducted so far on this topic. The largest, longest, highest quality and most recent studies increase confidence in what past studies have hinted at: that vitamin C may have a positive effect on T2DM. The effect just takes time, and is of moderate magnitude. The authors’ assessment of risk of bias is another major improvement over previous meta-analyses, as previous meta-analyses did not explicitly look at the certainty of the evidence that informed their results.

The inclusion of cutoffs for clinical relevance is a welcome addition in this study. A statistically significant finding is more likely to be a true effect, but whether the results have practical value is determined by the magnitude of the effect size. Clinical practice guidelines suggest that target blood pressure for people with hypertension and diabetes is 135/85 mmHg or less[23]. If vitamin C has the potential to lower blood pressure by 6.3/3.8, then it could help people meet that target. HbA1c targets for people with diabetes are in the 6.5–7.0% range, meaning the 0.54% absolute reduction seen in this analysis could represent a practical advantage[20], if it is representative of a true lowering of HbA1c. Even the fasting blood glucose-lowering effects of vitamin C could pose a benefit, however small they may be compared to post-meal glucose variations.

To explain vitamin C’s effect on blood pressure, the authors rationalize that vitamin C raises the vasodilator nitric oxide[13] by helping reset the cofactor involved in nitric oxide synthesis from l-arginine[14]. The issue is that even though T2DM is linked with heart conditions known for low nitric oxide[24], T2DM is, for some reason, also associated with high—not low—nitrite levels, on average[25]. (Nitrite is used here as a surrogate measure for nitric oxide[25], which might not be the most appropriate biomarker.) This discrepancy might be accounted for by an unmeasured variable like crosstalk or imbalance between different nitric oxide pathways[26][27], differential effects of this signaling molecule, different locations of nitric oxide[28], or some other variable. Alternatively, this study showed what had been found in a previous meta-analysis: that vitamin C can lower malondialdehyde[29], thus potentially lowering the burden of oxidative stress and downstream cardiovascular effects[30]. The benefits of vitamin C for people with diabetes and nitric oxide dysregulation still needs to be studied directly.

In spite of the inclusion of novel longer-term studies, most of the studies included in this meta-analysis were short in duration. Vitamin C depletion/repletion studies show that ascorbic acid levels reach saturation within 40 days of supplementation at doses of 500 mg per day[31]. This means that tissue concentrations of vitamin C would not have reached steady states in studies with a duration of less than 40 days, yet 7 of the 28 studies included in this systematic review were shorter than that. Additionally, recommendations indicate that HbA1c should be measured every 3 months[20], but only 8 of the included studies lasted 90 days or longer. Going further, the average lifespan of a red blood cell is 115 days[32]. To measure the effect of vitamin C on a long-term marker like glycated hemoglobin, investigators would need to supplement participants with vitamin C to sufficiency and then allow enough time for a complete renewal of red blood cells. The only two studies[33] to reach this 155 day duration were also the studies to show the most dramatic effect of vitamin C on HbA1c. One (n=135) showed that metformin plus 500 mg of vitamin C resulted in a 1.4 point decrease in HbA1c over 9 months, compared to metformin plus placebo. The largest (n=456) and longest study showed that metformin plus 500 mg of vitamin C resulted in a 2.99 point decrease in HbA1c over 12 months, compared to metformin plus placebo[33]. Neither of these studies were, however, free from bias: both were determined to have an unclear risk of bias due to poor randomization in the first study, and inadequate blinding procedures in both. The physiology and pharmacology of vitamin C indicate that positive effects would take time to see, yet the short duration of most of the studies analyzed in this meta-analysis means that few were even adequately equipped to assess the relationship between vitamin C supplementation and long-term measures of glycemia.

A caveat to this argument is that HbA1c reduction might not fully reflect the pathophysiology most relevant to T2DM. HbA1c is a long-term measure of blood glucose concentrations, but the reductions found in fasting and postprandial glucose were small and uncertain. Considering the structural resemblance of vitamin C and glucose, vitamin C could be interfering with the HbA1c assay[34], which would explain why it didn’t have a very dramatic effect on blood glucose measures. This could make the observed reduction in HbA1c not as relevant as the researchers hoped. This masking effect has been observed in vitro, but a small amount of clinical data suggests that vitamin C does not interfere with HbA1c assays[34].

The previously cited meta-analysis[16] looking at vitamin C and cardiometabolic measures in all populations showed no significant effect of vitamin C on HbA1c in a subgroup analysis including only people with T2DM, which is contrary to the findings of the present meta-analysis and attributable to the inclusion of recent, long-term studies. The present meta-analysis revealed a 1.7-fold greater reduction in fasting blood glucose than the previous review, potentially due to the additional 10 studies published between the two reviews. The previous meta-analysis suggested that studies longer than 30 days were more likely to report a beneficial effect and that vitamin C may have a more pronounced effect on fasting blood glucose than on post-meal blood glucose. The present meta-analysis also suggests both of these effects, with the cutoff for “longer studies” being 12 or more weeks, in subgroup analysis, showing significant consistency between these and previous results.

Ultimately, the results of this meta-analysis mean that vitamin C probably reduces blood pressure, and may improve glycemia in people with T2DM. The large proportion of people with T2DM and vitamin C insufficiency/deficiency are subpopulations most likely to experience a benefit.

Despite the high level of methodological rigor of the present meta-analysis, most of the included studies were not designed to fully capture the effects of vitamin C on glycated hemoglobin due to their relatively short duration. This may explain why previous meta-analyses have arrived at conflicting results regarding the effect of vitamin C on glycemia, whereas the present meta-analysis, which includes larger and longer studies, demonstrates a significant positive effect on HbA1c. Even though the certainty of its effect on systolic blood pressure is highest, the mechanism by which vitamin C may improve blood pressure has not yet been clearly elucidated.

Frequently asked questions

Q. Do common diabetes medications affect the function of nutritional antioxidants?

Some medications commonly used to treat diabetes and comorbidities have antioxidant properties and may affect vitamin status[35]. In randomized controlled trials, simvastatin was found to reduce circulating alpha-tocopherol (vitamin E) and beta-carotene (pro-vitamin A) in hypercholesterolemic men[36] and people[37] with T2DM and normal cholesterol. However, these results were not adjusted for the total lipid-lowering effect of a statin, which is relevant because both substances occur in the lipid fraction of blood. Other studies on simvastatin have shown no effect on vitamin E[38], and no effect on vitamin C[36] either.

Metformin, an antioxidant in its own right[35], may increase vitamin A and E levels and lower oxidative stress in people with T2DM[39]. Biguanides, thiazolidinediones, and statins consistently show antioxidant properties across studies[35], but it’s unclear whether these drugs pose an antioxidant benefit in addition to the antioxidant benefits of vitamin C, or if they pose an antioxidant benefit that subsumes the effects of vitamin C.

A newer class of diabetes drugs, sodium-glucose co-transporter 2 inhibitors, commonly known as SGLT inhibitors, has been reported in one case study to increase vitamin C levels due to their function in the kidneys. Note that this is a very low certainty of evidence because it comes from a case study. More research on the effect of SGLT inhibitors on vitamin C must validate this result before a hypothesis can be developed.

Q. Diabetes and vitamin C supplementation are both associated with kidney stones. Are there other ways to reduce risk?

Oxalate levels are an independent indicator of kidney stone risk, particularly in men[40]. T2DM is associated with increased risk of kidney stones[41]. In one meta-analysis, vitamin C supplements dosed at 250–499 mg per day and 1,000–1,499 mg per day were correlated with modestly increased risk of kidney stones in men, but the association between doses of vitamin C at 500–999 mg, or more than 1,500 mg, was suggestive, but not statistically significant[42]. Supplemental vitamin C is not linked with kidney stone risk in women, and vitamin C from diet is not linked either[42][43].

Based on the available literature, it’s probably best to get vitamin C from the diet, rather than supplements, in order to avoid increasing kidney stone risk while still getting the benefits of vitamin C. Pharmacokinetic studies show that absorption of vitamin C is limited when supplementing more than 400 mg[44] at a time, suggesting that there may be no need to take doses in excess of the median 500 mg dose studied in the meta-analysis under review.

Finally, one systematic review suggests adequate water intake and a healthy, plant-based diet high in calcium and low in sodium can help reduce the risk of kidney stones[45]. Since elevated urinary pH is a risk factor for oxalate kidney stones, alkalizing oral citrate supplements are suggested to have a positive effect, according to a Cochrane systematic review[46], although the study quality these results come from is generally poor.

What should I know?

This systematic review and meta-analysis is the first to assess risk of bias in studies measuring the effect of vitamin C on glycemia, blood lipids, and blood pressure. Moderately-certain evidence shows that vitamin C supplementation can reduce systolic blood pressure to a clinically relevant extent, though the mechanism behind this effect is unclear. Very-low certainty evidence suggests that vitamin C can lower diastolic blood pressure, HbA1c, and fasting blood glucose to a clinically meaningful extent. The analysis indicates that long-term (more than 12 weeks) supplementation is more likely to produce a greater effect on HbA1c, which makes good sense mechanistically. The results of this study depart from previous meta-analyses because of its inclusion of more recent, longer, larger, and higher-quality clinical trials—and with the GRADE measure of certainty analysis, the reliability of input data and results is easier to validate than in any previous systematic review and meta-analysis on vitamin C for T2DM.