Background
Atrial fibrillation (AF) is a common cardiac rhythm disturbance which can lead to severe consequences such as stroke [
1]. AF can be triggered by various stressful conditions; for example, about 30% of patients undergoing cardiac operations, such as coronary artery bypass grafting (CABG), suffer from post-operative AF (POAF) [
2]. AF is associated with oxidative stress and it seems that the cause-effect relation may work in both directions. In animal studies, tachycardia increased the levels of oxidative stress markers [
3,
4], while oxidative stress increased the susceptibility of isolated hearts to tachycardia [
5]. Since several ion channels which are expressed in the atria are sensitive to the redox state, oxidative stress and antioxidants might influence the electrophysiology of the atria [
6].
Vitamin C is a water soluble antioxidant which may protect against oxidative stress. In some studies, cardiac surgery decreased vitamin C levels consistent with increased oxidative stress [
7‐
9]. There is also evidence that vitamin C may have a treatment effect on some cardiovascular disorders. In patients undergoing cardiac operations, vitamin C increased cardiac perfusion after surgery [
10], and decreased the level of creatine kinase MB [
11,
12]. A meta-analysis of patients with atherosclerosis or heart failure found that vitamin C improved endothelial function [
13], and another found that vitamin C reduced blood pressure [
14]. In addition, a recent meta-analysis found that vitamin C decreased the risk of contrast-induced acute kidney injury in patients undergoing coronary angiography [
15].
In 2001, Carnes et al. reported that vitamin C administration seemed to prevent POAF, but they used historical controls instead of concurrent randomized controls [
3]. Nevertheless, that study led to a series of randomized trials on vitamin C against POAF.
AF is a common cardiac arrhythmia and vitamin C is a safe and inexpensive essential nutrient. Approximately 0.01 g/day of vitamin C prevents scurvy but the safe dose range extends to grams per day [
16,
17]. The possibility that vitamin C might have preventive effects against AF, even in restricted population groups, is worth examination. The goal of this systematic review was to analyse the preventive effect of vitamin C against AF in patients with a high risk of AF.
Discussion
Our systematic review was formulated as an examination of the effects of vitamin C on the occurrence of AF in people at a high risk for AF. Since patients undergoing cardiac surgery and cardioversion may suffer from acute oxidative stress, vitamin C administration might have an influence in such special conditions. We found 15 randomized trials on vitamin C for preventing AF in high risk patients. On average, vitamin C decreased the occurrence of AF by 27%, but there was significant heterogeneity in the results. We also found a 10% decrease in hospital stay in cardiac surgery patients. Heterogeneity indicates that a single average estimate of effect cannot be valid for all of the 15 trials and justified exploration to identify the causes of this heterogeneity.
Five of the trials included in Fig.
2 were carried out in Iran and the pooled estimate indicates a 51% reduction in POAF incidence with vitamin C administration. In contrast, 5 trials in the USA found that vitamin C produced no benefit. These 2 sets of trials are significantly incompatible. We do not consider that methodological differences between the US and Iran trials are the most likely explanations for the divergence between these 2 sets of trials. Previous research has pointed out that treatment effects can differ between less and more developed countries. Panagiotou et al. [
36] found several cases in which trials in less developed countries showed more favourable treatment effects than trials in more developed countries. Although methodological variations may explain some of the differences, it is also likely that there are genuine differences between many treatment effects between substantially different cultures. Wealth is strongly correlated with life-style factors including nutrition, and with differences in hospital treatments. Such differences might explain the divergence between the results in the 5 US and the 5 Iran POAF trials.
Among the non-US trials, oral vitamin C had a greater effect on POAF occurrence than intravenous vitamin C (Fig.
3). There is a substantial difference in the pharmacokinetics between oral and intravenous vitamin C, so that the same intravenous dose leads to much higher vitamin levels in plasma than the oral dose [
16,
37]. Furthermore, intravenous vitamin C may be more reliable for postoperative patients since delayed gastric emptying is a frequent concern. Therefore the greater effect of oral vitamin C on POAF occurrence was against our expectation. On the other hand, the greater effect of intravenous vitamin C on the hospital stay is consistent with greater effects of higher plasma levels.
POAF has previously been correlated with a longer stay in hospital, but it is not known whether the longer hospital stay is caused by the episode of POAF or whether both of them are caused by other factors [
38]. Among the non-US trials, oral vitamin C had a greater effect on POAF, but a smaller effect on hospital stay compared with intravenous vitamin C (Fig.
6). This negative correlation between the effects on POAF and on hospital stay conflicts with the notion that POAF is the cause of a longer hospital stay. The divergence we found in the effects of oral and intravenous vitamin C administration indicates that the 2 methods of vitamin C administration should be further studied by head-to-head comparisons in 3-arm randomized trials instead of just comparing independent trials that have various other differences simultaneously.
As the methodological inclusion criterion, we required that the trials be randomized. Six trials with cardiac surgery patients did not use an explicit placebo [
19,
20,
23,
24,
29,
32]. However, other medications serve as a functional placebo to vitamin C in cardiac surgery trials. It seems highly unlikely that such patients might notice whether vitamin C is administered along with the other medications, so that the lack of an explicit placebo would substantially bias observations in the 5 trials. Three POAF trials [
23,
24,
29] did not report that physicians in charge of treatments and outcome assessment were blinded for the trial groups, but in other trials they were blinded. In the Korantzopoulos et al. trial on the recurrence of AF after cardioversion, patients bought vitamin C tablets themselves and knew their treatment [
32]. However, a large meta-analysis showed that placebo has minimal or no effects on binary outcomes [
39], such as the occurrence of AF.
In sensitivity analyses, we excluded trials that had some concerns about possible bias in the comparison. However, the estimates of vitamin C effects were not changed. In addition, the heterogeneity between the US and the non-US trials was not influenced by the exclusion of trials that had some concerns about possible bias.
Two relatively large US trials on POAF were not published because their results were negative [
27,
30]. Nevertheless, we do not consider that publication bias is a likely explanation for our main findings. The evidence of the benefit of vitamin C in the non-US trials is very strong and there should be a particularly large number of unpublished non-US trials to explain the findings purely as a result of researchers publishing just the positive findings. In particular, publication bias is not a reasonable explanation for the significant difference between oral and intravenous administration.
In the first study on vitamin C for preventing POAF, Carnes et al. described their protocol as follows: “patients scheduled for primary CABG surgery were given 2 g ascorbic acid (extended release) the night before surgery, followed by 500-mg doses twice daily for the 5 days after surgery” [
3]. All subsequent POAF trials used dosage protocols that are only minor modifications of the Carnes method. Although the first question after the publication of a positive result should be whether the result can be repeated, subsequent trials should also investigate protocol variations to determine the optimal protocol.
Vitamin C is water soluble, and its concentration in plasma increases within 1–2 h of oral administration and decreases thereafter [
16,
37]. On the basis of such pharmacokinetics, it seems unlikely that several days administration before the cardiac operation might lead to further benefits. In 2 POAF trials, vitamin C was administered intravenously 3 h before the operation [
21] or immediately before [
31] and both produced a significant benefit. Longer administration before the operation might not be needed.
Another time-dependent question is about the length of vitamin C administration after the operation. In the POAF trials, vitamin C administration was continued for 5 days after the operation and the cardioversion trial continued administration for 7 days. Assuming that the greatest peak of oxidative stress occurs during and soon after the operation or cardioversion, a shorter administration period might be sufficient. In a trial with severe burn patients, vitamin C was administered intravenously for only 24 h after hospitalization but the dose was particularly high with 66 mg/kg/h (i.e., 110 g per 70 kg per 24 h) [
40]. Compared with the control group, the level of vitamin C remained much higher in the vitamin groups for 3 days. In the vitamin C group, the length of mechanical ventilation was 43% shorter (
P = 0.03) and there was a significant decrease in the requirement for infusion fluids [
40], indicating that such a 1-day administration at a particularly high dose level might also be effective. Accordingly, the dose and the length of vitamin C administration should also be investigated in further trials.