This retrospective study of patients hospitalized for sepsis revealed an association between IV vitamin C usage and subsequent AKI and in-hospital mortality. When examining the association of IV vitamin C therapy and AKI, the odds of experiencing AKI for individuals treated with IV vitamin C were 107% higher than the odds for individuals not treated with IV vitamin C, after controlling for a number of demographic and clinical variables. Findings also revealed that risk of AKI was increased for those who were male and older, which is supported in previous research (26, 27). Findings are consistent with results from a similar study of patients admitted via the ED for sepsis, which found that individuals over 65 years were at significantly higher risk of AKI and subsequent in-hospital death (28). Other research has also highlighted the relationships between increased LOS and older age, and the occurrence of AKI and mortality (29, 30). Other notable factors in the current study associated with increased risk of AKI were longer ICU LOS and identifying as black as compared to white, both of which are risk factors supported by previous research (34–36). The association between IV vitamin C and AKI is not unique to this study; a study by Litwak et al. observed a 12% increase in AKI in patients that received IV vitamin C but this did not reach statistical significance, likely due to small sample size (22).
When examining in-hospital mortality, results revealed that being older, having AKI, longer ICU LOS, and receiving IV vitamin C were associated with increased risk of mortality. Specifically, the odds of mortality were 67% higher for the IV vitamin C group compared to those in the comparison group. The association between high dose vitamin C and death was unexpected and not easily explained. Another variable found to be associated with an increased risk of death, which may lend insight into the relationship between vitamin C and mortality, was the diagnosis of AKI. Previous research indicates kidney failure increases the risk of death six to eight times in septic patients (37). However, when examining the data in the current study, an increased risk of death was evident irrespective of AKI. Among patients who did not have AKI and did not receive IV vitamin C, 3.03% died compared to 7.46% of those who did not have AKI but received IV vitamin C. This suggests there may be a direct effect of high dose vitamin C that may exacerbate the risk of mortality. The current study lacked the data that would allow for a more comprehensive examination between high doses of IV vitamin C and potential increased toxicity outside of the kidney, however, future research should explore the relationship between vitamin C and mortality, irrespective of AKI.
The current finding that IV vitamin C is associated with increased risk of death is inconsistent with another large retrospective study which found IV vitamin C to be associated with reduced mortality among patients with sepsis (21). One reason for the disparate results may be differences in treatment dose. Specifically, vitamin C in the Byerly et al. study was not limited to high dose and was not defined by a threshold defining the treatment group, whereas the current study examines what is considered high dose IV vitamin C (≥ 1.5g), which may account for outcome differences. One study that examined higher doses (200mg/kg/day) of vitamin C showed evidence of increasing rates of mortality, with higher mortality (50.6%) among a high dose vitamin C group compared to patients who received lower doses of vitamin C (50mg/kg/day) (38%), though both treatment groups had lower mortality than the placebo group (63%) (14). Notably, this study was limited by a small sample size and did not reach statistical significance. Additional research examining other clinical markers including SOFA scores in patients receiving high dose IV vitamin C over long periods of time is needed.
Sub-analyses of the ICU sample revealed that age, longer ICU LOS, and receiving IV vitamin C remained predictors of AKI, while identifying as black and being male were no longer significant predictors. Older age, and ICU LOS are both known risk factors supported by previous research (34–36). A second sub-analysis revealed that age and AKI remained significant predictors of in-hospital mortality, however high dose IV vitamin C use was no longer a statistically significant predictor of mortality once the sample was limited to ICU patients only.
The current study observed a moderate relationship between IV vitamin C and mortality in septic patients, which might suggest potential nephrotoxicity of vitamin C therapy. The association between high dose vitamin C and AKI due to oxalate deposition in renal tubules is well established (38). Since early reports in 1985, there are multiple cases of AKI resulting from high dose vitamin C, administered parenterally or orally (39–43). Ascorbic acid is metabolized intracellularly and converted to oxalic acid (44); the oxalic acid is then filtered in the kidney where it precipitates into crystals of calcium oxalate, potentially causing obstruction and tubular injury. Previous research has shown that vitamin C doses as low as 480–960 mg/D, taken orally over several months, have resulted in oxalate deposition kidney failure that requires temporary dialysis (45). Dietary supplementation of 2g per day can increase oxalate excretion by 21.8% (46). Oxalic acid and oxalate toxicity have been shown to occur in myocardial tissue of patients with hereditary hyperoxaluria (47). Others have suggested that similar myocardial tissue damage can occur in secondary forms of hyperoxalosis (48, 49). Acutely elevated oxalate levels in the serum could cause myocardial dysfunction, leading to increased risk of mortality regardless of AKI.
Limitations
This study was limited in the retrospective nature of the design, which precludes causal-effect conclusions and prevents patient follow-up. Relatedly, the EHR data utilized did not include time stamps for diagnoses and treatment, thus temporal order is unknown. However, strict inclusion and exclusion criteria were applied in an effort to reduce selection bias and focus on community-acquired sepsis. Further, as vasoactive drug use and ICU admission were the only available surrogates for illness severity, these factors were considered in the PSM and sample stratification, respectively. Although the current study was limited to one hospital, the large sample size provided adequate power to detect statistically significant associations between patient-level demographic and clinical variables, and specifically, vitamin C usage, AKI, and mortality. Additionally, existing literature has commonly reported SOFA scores, which were unable to be utilized in the current study due to inconsistencies in collection. Many studies have used SOFA scores as a representation of mortality risk, whereas in the current study, we were able to directly examine in-hospital mortality as a primary outcome. Additionally, the usage of serum lactic acid and procalcitonin levels were reviewed as a potential measure of disease severity but could not be included in analyses due to data missing not at random. In other words, these laboratory tests were not performed equally across all groups of patients. Finally, the current study did not assess for other nephrotoxic drugs that could be related to AKI, such as antimicrobials and radiologic contrast dye. Future research should include these as well as other factors to assess the unique relationship between IV vitamin C and AKI.