Vitamin C supplementation showed greater effects on systolic blood pressure in hypertensive and diabetic patients: an updated systematic review and meta-analysis of randomised clinical trials

Abstract Results from randomised controlled trials (RCTs) testing the effect of vitamin C supplementation on blood pressure (BP) have been inconsistent. This systematic review evaluated the effects of vitamin C supplementation on BP and included RCTs testing the effects of vitamin C supplementation alone, on systolic and diastolic BP in adult participants (≥18 years). Random-effect models were conducted to estimate the pooled effects of vitamin C supplementation on BP. A total of 20 studies with 890 participants were included. The median dose of vitamin C was 757.5 mg/d, the median duration was 6 weeks. Vitamin C supplementation was found to reduce systolic BP by −3.0 mmHg (95%CI: −4.7, −1.3 mmHg; p = 0.001). Subgroup analysis showed a more pronounced effect on systolic BP in patients with hypertension (−3.2 mmHg, 95%CI −5.2, −1.2 mmHg, p = 0.002) and diabetes (−4.6 mmHg, 95%CI −8.9, −0.3 mmHg, p = 0.03). Further research needs to evaluate the long-term effect of vitamin C on BP in populations with impaired cardio-metabolic health.


Introduction
Vitamin C, also known as L-ascorbic acid, has been the subject of interest in relation to its potential impact on blood pressure (BP) regulation.The multifaceted mechanisms through which vitamin C may influence BP include its antioxidant properties, improvement of endothelial function, and modulation of nitric oxide (NO) production (Vinishdharma and Don 2019).Additionally, vitamin C has been suggested to enhance NO bioavailability by converting nitrite to NO under gastric conditions (Izumi et al. 1989).While previous epidemiological studies have suggested a link between higher vitamin C intake and reduced blood pressure (BP) (Ness et al. 1997;Ran et al. 2020), the evidence from randomised controlled trials (RCTs) has been contrasting (Guan et al. 2020;Juraschek et al. 2012).
Juraschek et al. conducted a comprehensive meta-analysis in 2012 involving 29 trials, which had an average duration of 8 weeks, and trial sizes ranging from 10 to 120 participants.The combined results showed a significant reduction in systolic blood pressure (SBP) by approximately 3.84 mm Hg and diastolic blood pressure (DBP) by approximately 1.48 mm Hg.In trials specifically involving hypertensive participants, SBP was significantly reduced by approximately 4.85 mm Hg.The meta-analysis however included studies that assessed the anti-hypertensive effects of vitamin C alone and in combination with other interventions (i.e.drugs or other supplements), which may have confounded the identification of the effects on BP attributable to vitamin C supplementation alone (Juraschek et al. 2012).
Therefore, the aim of this systematic review and meta-analysis of RCTs is to update the previous review but focussing specifically on the evaluation of the effects of vitamin C supplementation alone to provide an unbiased assessment of its effects on SBP and DBP.The efficacy of oral vitamin C supplementation obtained from the selected RCTs will be evaluated using random-effects models and meta-regression analyses will be conducted to evaluate the influence of potential confounding factors (age, study duration, baseline BP and vitamin C concentrations and quality of the RCTs) on the effect size estimates.

Search strategy and eligibility criteria
The protocol of the systematic review has been registered on the on PROSPERO database (CRD42022367173).We utilised the OVID search tool for Medline, Embase, and Cochrane Central Register of Controlled Trials, covering the period from January 2010 to July 2023.Furthermore, studies published before 2010 and included in a previous meta-analysis (Juraschek et al. 2012), that met the criteria of this review were included in the selection.Our search terms comprised of "blood pressure", "hypertension", "hypotension", "hypertensive", "endothelial dysfunction", "endothelial function", "ascorbic acid", "antioxidant/antioxidants", "vitamin/vitamins", "randomised controlled trial", "random allocation", "double-blind", "single-blind", and "clinical trials".A full description of the search algorithm is provided in the online supplementary material.
The search was confined to human studies.The studies were limited to RCTs, including parallel, crossover, and factorial studies, and were conducted in adult subjects (18 years and over).Crossover studies were only included if they declared an appropriate washout period to allow normal ascorbic acid plasma concentration to re-establish.The intervention in the studies involved oral vitamin C supplementation against placebo control groups and trials were included if the effect of vitamin C supplementation alone could be ascertained.Trials reporting changes in blood pressure, including SBP, DBP, or both were included.Studies with an intervention duration of at least 2 weeks were considered.No language restrictions were applied to the search.
Trials that included adolescents (under 18 years of age) and pregnant women were excluded due to discrepancies with recommended vitamin C intake and inconsistency in antihypertensive medication use.Similarly, trials including patients with end-stage renal disease were excluded due to vitamin C renal excretion/reuptake complications and inconsistent antihypertensive medication use.Editorials, commentaries, letters, and descriptive reviews were also excluded due to a lack of objective evidence.Case-controlled studies, case series, case reports, and cohort studies were not relevant to the review question and therefore excluded.Systematic reviews were removed due to the absence of original data.

Study selection
EndNote 20 (Clarivate) was used to manage all search results.A three-step method was employed for this systematic review: title, abstract, and full-text screening.Three researchers (KK, EL, MS) evaluated the search results and disregarded studies that did not match the aim of the review.Abstracts were compared to pre-set eligibility criteria, and relevant full texts were acquired from Medline and the University of Nottingham online library source (NUSearch).Final exclusion was dependent on relevance to exclusion criteria and the primary focus of the systematic review.Disagreements between researchers were reconciled using consensus.

Quality assessment
The methodological quality of each clinical trial was assessed using the Eight-item Modified Jadad Score developed by Jadad et al. (Jadad et al. 1996).The modified scale provides scores of 0 to 8 to assess trial quality, with a maximum score of 8 signifying a rigorous trial design with a low risk of reporting bias.The scale was structured as 8 questions which addressed reporting of: randomisation (yes or no), appropriate randomisation technique (yes, no, not reported), blinding (double-blinded, single-blinded, unblinded/not reported), appropriate blinding technique (yes, no, not reported), withdrawals and dropouts (yes or no), inclusion and exclusion criteria (yes or no), adverse effects (yes or no), and statistical analysis technique (yes or no).The allocation of points against these assessment criteria is defined in the Online Supplementary Material.Full texts were extracted from trials and Jadad Scores were independently screened by two researchers (KK, EL).Totalled scores differing by 1 point were assigned the mean average and scores differing by >1 were resolved via consensus.

Meta-Analysis
The study extracted SBP and DBP measurements from both the placebo and vitamin C groups, and calculated the BP changes by subtracting baseline BP from final BP.Any missing data not included in tables or the main body of text were obtained from figures.In parallel trials where baseline BP varied, the review used between-arm differences or determined BP change based on data presented in the study.When baseline BP was not reported in one parallel trial, the review used the difference between intervention and control final BP readings.In crossover trials, the review used the stated BP change or calculated the differences between intervention and control final BP readings based on the mean baseline BP for all participants.In trials that reported more than one BP measure, for example night and daytime ambulatory BP (ABPM), the review used the mean BP resulting from the greatest number of measurements.Comprehensive Meta-Analysis software (version 2, Biostat, Englewood, New Jersey, USA) was used to perform meta-analysis, subgroup analysis, meta-regression and publication bias assessments.Heterogeneity between participant and trial characteristics were taken into consideration by the implementation of random effect models during analysis.Inverse variance weighting was used to determine effect sizes and to calculate 95% confidence intervals (CIs).Calculated outcomes were presented graphically in separate forest plots for SBP and DBP.Publication bias was initially evaluated using visual assessment of funnel plots and statistically measured using Egger's regression intercept.Relative weightings of trials were calculated; considerable variance in study relative weightings would suggest heterogeneity.An I 2 test was used to assess study consistency: ≥75% suggested high risk, 25-75% suggested moderate risk and <25% suggested low risk of between-trial variations (Basaqr et al. 2021).Subgroup analyses were conducted to identify consistencies in results or varied magnitudes of effect within specific trial characteristic groups.Factors which may have resulted in varied results, and hence were evaluated, included: type of trial (crossover or parallel), type of patient (Healthy, cardiovascular diseases (CVD), essential hypertension (EH), type 2 diabetic (T2D)), BP measurement (resting or ambulatory), clean (vitamin C alone; i.e. vitamin C was not co-administered with other nutritional interventions) intervention (yes or no).Meta-regression analysis was carried out to test the influence of the following factors on effect estimates: age, trial duration, intervention dosage, Jadad score, baseline SBP and DBP values and baseline vitamin C concentrations.

Results
The search strategy yielded 912 results; 197 studies were automatically excluded due to duplication of records across databases.Following title and abstract screening, 18 out of 715 studies were selected.Twenty-nine studies from a previous meta-analysis were also included in this list for evaluation.(Juraschek et al. 2012).This resulted in 47 studies moving onto final exclusion, which were retrieved for full-text screening and adjudication via consensus.Of the initial 18 studies selected, 15 were excluded during full-text screening.Seventeen studies from the previous meta-analysis also met the criteria for inclusion.Therefore, 20 studies in total were selected for inclusion in the present review.The study selection process was presented in a flow chart following PRISMA guidelines (Figure 1 of the Online Supplementary Material).
Trial characteristics are summarised in Table 1.A total of 890 participants (517 males, 373 females) across 20 studies, conducted from 1982 to 2019, were included in this systematic review.Trial size ranged from 10 to 120 participants.The mean trial age was 54.3 (range 25.0 to 73.8) years.One trial included independent subgroups: smokers and non-smokers (Keith and Driskell 1982).
Methodological quality assessments were derived using the Eight-item Modified Jadad Score.The results of the qualitative assessment are presented in Table S1 of the Online Supplementary Material.Full texts were extracted for the majority of the trials (19 of 20).One study was published as an abstract and excluded from the quality assessment as not reporting sufficient information (Block et al. 2002).All trials reported randomisation, with most providing details of sufficient randomisation techniques (18 of 20).Most trials included blinding of both investigators and participants (14 of 20), with 9 of these studies reporting appropriate blinding methods (Mason et al. 2019;Shateri 2016;Hutchins et al. 2005;Ward et al. 2005;Magen et al. 2004;Brody et al. 2002;Mullan et al. 2002;Ghosh et al. 1994;Keith and Driskell 1982).Nine studies included presentation of withdrawals and dropouts (Mason et al. 2019;Hutchins et al. 2005;Ward et al. 2005;Magen et al. 2004;Brody et al. 2002;Mullan et al. 2002;Duffy et al. 2001;Gokce et al. 1999;Ghosh et al. 1994).All trials included information about eligibility criteria for participants.Methodological approach to adverse effects was addressed in eight studies (Mason et al. 2019;Nightingale et al. 2007;Hutchins et al. 2005;Ward et al. 2005;Magen et al. 2004;Brody et al. 2002;Gokce et al. 1999;Ghosh et al. 1994).All studies included information on the approach to statistical analysis (Overall, Jadad scores ranged from 4 to 8 suggesting a moderate to low risk of reporting bias across all trials.

Meta-Analysis
All 20 studies (890 participants) were included in the meta-analysis.Pooled effects were visualised and presented graphically in forest plots of SBP and DBP (Figure 1).Pooled estimates and 95% CIs of effect sizes were calculated using a random effects analysis model with relative weighting accounted for and presented.The pooled effect of vitamin C supplementation on BP showed a significant reduction in SBP by −3.0 mmHg (95% CI: −4.7, −1.3 mmHg; p = 0.001) and an average reduction in DBP by −0.7 mmHg (95% CI: −2.1, 0.6 mmHg; p = 0.293).There was a moderate degree of heterogeneity between studies detected in both SBP (Q-value: 68; I 2 : 70; p < 0.001) and DBP (Q-value: 77; I 2 : 74; p < 0.001).

Meta-regression
The effect of vitamin C on both SBP and DBP was not significantly influenced by baseline BP, age, duration, vitamin C dosage, Jadad score or baseline vitamin C concentrations after meta-regression analysis (data not shown).

Publication bias
Funnel plots were generated by plotting standard error against the difference in means (Figure S2 of the Online Supplementary Material).Visual assessment of the funnel plots showed evidence of asymmetry in SBP and statistical analysis using Egger's test confirmed this was significant (p = 0.03).DBP was symmetrical (Egger's test, p = 0.37), suggesting no evidence of publication bias.

Discussion
This review provides an update on the effects of oral vitamin C supplementation on BP in adult participants that was first reviewed by Juraschek et al. in 2012(Juraschek et al. 2012).However, our protocol was modified to isolate the specific effects of vitamin C more clearly by not including studies that used co-administration of other nutrients alongside vitamin C and/or not including an appropriate control group to clearly isolate the effects of vitamin C on BP.
The results showed that supplemental vitamin C is associated with significant reductions in SBP; changes in DBP were only significant in patients with diabetes.The review identified a moderate inter-trial heterogeneity, which could have implications on the pooled effect estimates calculated.Subgroup analysis and meta-regression analysis were carried out to address the impact of different trial qualities.Some trial characteristics (patient population, BP measurement technique, and clean trials) were found to cause significant changes to pooled effect estimates.However, the review has a shortcoming in that it can only comment on the short-term effects of supplemental vitamin C, and few studies have focused on the long-term effects of vitamin C, which may have contributed to the generation of inconsistent results.Kim et al. reported an increase in SBP and no change to DBP in a patient group at risk of stomach cancer and stroke, suggesting that any perceived anti-hypertensive effects of vitamin C may not be reflected in long-term supplementary use (Kim et al. 2002).This study was not included in the systematic review as it lacked of an appropriate control group.Some trials reported BP reductions with prolonged vitamin C supplementation (Mark et al. 1996;Zureik et al. 2004), and a long-term study also found significant, protective effects on vascular health as it slowed atherosclerotic progression in patients with hypercholesteremia (Salonen et al. 2003).
Longer trials would help assess the effects of sustained vitamin C interventions and provide insights into the potential use of vitamin C for preventing metabolic and cardiovascular diseases.Our subgroup analysis showed that vitamin C supplementation had no effect on BP in healthy subjects and a stronger effect on patients with higher oxidative stress levels, such as those with CVD, EH, and T2D, with T2D patients experiencing significant reductions in both SBP and DBP.A recent meta-analysis (Mason et al. 2021) on oral vitamin C supplementation and BP in T2D patients supported our findings.While the EH and T2D papers included in the review did not explicitly discuss the underlying causes for the observed associations between vitamin C supplementation and larger changes in BP, it is worth considering potential mechanisms that may contribute to these findings.Both hypertension and diabetes are characterised by increased oxidative stress and endothelial dysfunction, which can damage blood vessels and contribute to elevated blood pressure (Petrie et al. 2018).Vitamin C's antioxidant properties can help counteract oxidative stress and enhance NO production, which may lead to more pronounced effect on blood pressure in individuals with compromised endothelial function (Lbban et al. 2023).In addition, a recent meta-analysis showed that vitamin C intake had positive effects on improving metabolic parameters and reducing total cholesterol levels, thereby lowering cardiovascular disease (CVD) risk in individuals with type 2 diabetes or related metabolic conditions.Additionally, vitamin C intake was linked to decreased inflammation and oxidative stress markers such as C-reactive protein, interleukin-6, and malondialdehyde (Dludla et al. 2022).
Vitamin C's antioxidant properties and its ability to enhance NO production and improve endothelial function are potential explanations for its effects on BP (Vinishdharma and Don 2019).By reducing oxidative stress, vitamin C may help maintain the health and elasticity of blood vessels, promoting healthy blood pressure levels (Sinbad et al. 2019).Vitamin C supports endothelial function by promoting the production of NO via limiting the formation of reactive species and reducing their reactivity with NO (Mortensen and Lykkesfeldt 2014).(Tamari et al. 2013) and by preserving concentrations of the endothelial NO synthase cofactor tetrahydrobiopterin (Huang et al. 2000).Additional physiological mechanisms by which vitamin C can influence the regulation of BP include regulation of the synthesis of collagen and maintenance of the structural integrity of the vessel walls (May and Harrison 2013), modulation by vitamin C of the Renin-Angiotensin System (RAS) by inhibiting the production of angiotensin II (Hwang et al. 2022), and links with the regulation of release and activity of stress hormones such as cortisol (Moritz et al. 2020).
In-person BP measurements can be affected by white coat syndrome (Pioli et al. 2018), resulting in inaccurate readings.Studies using ambulatory measurements showed smaller changes in SBP compared to at-rest readings, which could be due to the latter's lower accuracy resulting from fewer readings.A significant reduction in SBP was observed in trials testing effects of vitamin C supplementation alone, supporting its potential as an intervention for hypertension.Vitamin E is commonly used alongside vitamin C, and a recent meta-analysis showed significant SBP reduction with vitamin E supplementation (Emami et al. 2019).Future research could explore the synergistic effects of both vitamins on BP.
The strength of the conclusions in this systematic review were driven by the rigour of its approach.A review protocol was registered with PROSPERO, and following the PICO framework, a pre-specified eligibility criterion was used to identify relevant trials.Three researchers independently screened papers to avoid excluding any relevant studies.Additionally, adhering to PRISMA guidelines ensured that the results of the review were repeatable.The review was focused on interventions consisting only of vitamin C, minimising potential bias from interactions, and the results indicate that vitamin C has antihypertensive qualities.
The results of this review are also subject to limitations.Moderate heterogeneity detected between studies due to various factors, such as trial design, duration, participant characteristics, which could affect the effectiveness of vitamin C interventions.75% of trials were conducted on patients with pre-existing conditions and the mean trial age was 54.3 years, which may require higher intervention dosages to reach saturation requirements.Small sample sizes, averaging 44.5 individuals across 20 studies, present complications of the small study effect and may exaggerate pooled effect estimates.There were limited estimates on the effects of CVD and ambulatory BP measurement in subgroup analysis due to the small number of trials that included these patients or used ABPM as the method of BP measurement.Results could be at risk of false positives due to the wide range of subgroup analysis and meta-regression analysis.There was also a moderate to low risk of reporting bias based on Jadad scores; future studies would benefit from improved reporting of methods of blinding, patient dropouts and withdrawals, and assessments of adverse effects.Finally, there was evidence of publication bias for SBP due to funnel plot asymmetry and a statistically significant Egger's regression model, which may need to be considered for the interpretation of the results.In addition to these limitations, we acknowledge the importance of considering baseline vitamin C concentration on blood pressure outcomes.While we investigated this aspect in our study using meta-regression, we also recognise that this information was not reported in all the trials included in the meta-analysis (16 out of 20 trials), which may have potentially influenced the validity of the analyses.The majority of the studies used resting BP measurements; trials with more rigorous measurements of BP, such as ambulatory blood pressure monitoring (ABPM), are needed for a more accurate evaluation of the effects of vitamin C on BP

Conclusions
The findings of this updated meta-analysis suggest that vitamin C supplementation may have a significant impact on reducing SBP.Subgroup analyses further revealed greater reductions in both SBP and DBP among patients with EH and T2D, compared to healthy.Further research is needed to evaluate the long-term effects of vitamin C in preventing hypertension and to determine how it can be best used alongside existing pharmacological interventions.

Figure 1 .
Figure 1.Pooled effect of vitamin c supplementation on systolic blood pressure (a) and diastolic blood pressure (B) in 20 randomised controlled studies.Squares indicate mean values, with the area of each square being proportional to its relative weight in the analysis.Horizontal lines denote 95% confidence intervals; arrows indicate a lower/upper limit outside the range of -12.00mmHg to +12.00mmHg.diamonds represent pooled estimates.Keith et al. included two distinct populations (nS, non-smokers; S, smokers) hence the total number of studies included in this analysis is 21.

Table 1 .
characteristics of studies included in the systematic review a .

Table 2 .
Subgroup analysis of mean change in BP a .
a one trial reported effects in two distinct subgroups, hence the number of trials included in meta-analysis totalled to 21. BP: blood pressure; cI: confidence interval; co: crossover; cVd: cardiovascular disease; eH: essential hypertension; P: parallel; t2d: type 2 diabetes.