Contents lists available at ScienceDirect Free Radical Biology and Medicine journal homepage: www.elsevier.com/locate/freeradbiomed Original article Erythrocyte SOD1 activity, but not SOD1 polymorphisms, is associated with ICU mortality in patients with septic shock Nara Aline Costaa, Natália Baraldi Cunhaa, Ana Lucia Guta, Paula Schmidt Azevedoa, Bertha Furlan Polegatoa, Leonardo Antonio Mamede Zornoffa, Sergio Alberto Rupp de Paivaa, Bruna Zavarize Reisb, Ana Angelica Henrique Fernandesc, Marcelo Macedo Rogerod, Marina Maintinguer Norded, Marcos Ferreira Minicuccia,⁎ a Department of Internal Medicine, Botucatu Medical School, UNESP – Univ Estadual Paulista, Botucatu, Brazil bDepartment of Food and Experimental Nutrition, Faculty of Pharmaceutical Science, USP - University of São Paulo, São Paulo, Brazil c Chemistry and Biochemistry Department, Institute of Biosciences, Univ Estadual Paulista (UNESP), Botucatu, Brazil d Department of Nutrition, School of Public Health, USP – University of São Paulo, São Paulo, Brazil A R T I C L E I N F O Keywords: SOD1 activity SOD1 polymorphisms Septic shock Oxidative stress A B S T R A C T The objective of our study was to evaluate the influence of the superoxide dismutase 1 (SOD1) polymorphisms on erythrocyte SOD1 activity and the mortality of patients with septic shock. We prospectively evaluated 175 patients aged over 18 years with septic shock upon ICU admission. However, 38 patients were excluded. Thus, 137 patients were enrolled in the study. Blood samples were taken within the first 24 h of the patient's admission to determine erythrocyte SOD1 activity and nine SOD1 gene polymorphisms. The mean patient age was 63 ± 16 years, 58% were men, and ICU mortality rate was 66%. The patients who died were older and more severely ill, with higher Acute Physiology and Chronic Health Evaluation (APACHE II) and Sequential Organ Failure Assessment (SOFA) scores, as well as higher lactate, urea, and protein carbonyl levels. In the logistic regression model, erythrocyte SOD1 activity was associated with ICU mortality. This relationship was also maintained in the highest tertile of SOD1 activity (odds ratio [OR]: 0.02; 95% confidence interval [CI]: 0.00–0.78; p=0.037). Only SNP rs2070424 of the SOD1 gene influenced erythrocyte SOD1 activity. For patients with the AA allele, the activity of SOD1 was lower in relation to G-carriers (A/G+G/G genotype) (p=0.019). None of the nine SOD1 SNPs were associated with ICU mortality. In conclusion, the SNP rs2070424 of the SOD1 gene interferes with erythrocyte SOD1 activity, and higher activity of SOD1 was associated with decreased mortality in patients with septic shock. 1. Introduction Septic shock is one of the most important causes of intensive care unit (ICU) admissions, and the leading cause of death in critically ill patients worldwide [1,2]. The incidence of septic shock is not well known; however, it is increasing, likely as a result of the progressive aging of the population, the large number of individuals with co- morbidities, and the increasing recognition of sepsis [2]. Sepsis is characterized by life-threatening organ dysfunction caused by a dys- regulated host response to infection [1,2]. In this scenario, oxidative stress plays a key role, as it is a major promoter and mediator of sys- temic inflammation and organ failure [3–5]. Oxidative stress is defined as an imbalance between the production of reactive oxygen species (ROS) and reactive nitrogen species and the body's antioxidant defenses [5–7]. Oxidative stress can be measured by different products derived from lipids, proteins and DNA. Protein car- bonyl groups are markers of protein damage that are formed early during septic shock and are more stable than lipid peroxidation pro- ducts [8,9]. Regarding the endogenous antioxidant defense system, it can be divided into an enzymatic and a non-enzymatic group [5–7]. Superoxide dismutase (SOD) is the most abundant enzyme of the an- tioxidant system, which catalyzes the conversion of superoxide into hydrogen peroxide, and it is considered the first line of defense against ROS [5–7]. SOD1 is the cytoplasmic isoform of SOD, and has copper https://doi.org/10.1016/j.freeradbiomed.2018.06.013 Received 17 April 2018; Received in revised form 6 June 2018; Accepted 11 June 2018 ⁎ Correspondence to: Departamento de Clínica Médica, Faculdade de Medicina de Botucatu, Rubião Júnior s/n, 18618-970 Botucatu, SP, Brazil. E-mail address: minicucci@fmb.unesp.br (M.F. Minicucci). Abbreviations: APACHE II score, Acute Physiology and Chronic Health Evaluation II score; CI95%, confidence interval 95%; CRP, C-reactive protein; ICU, intensive care unit; MDA, malondialdehyde; OR, odds ratio; ROS, reactive oxygen species; SNPs, single nucleotide polymorphisms; SOD, superoxide dismutase; SOFA score, Sequential Organ Failure Assessment score Free Radical Biology and Medicine 124 (2018) 199–204 Available online 12 June 2018 0891-5849/ © 2018 Elsevier Inc. All rights reserved. T http://www.sciencedirect.com/science/journal/08915849 https://www.elsevier.com/locate/freeradbiomed https://doi.org/10.1016/j.freeradbiomed.2018.06.013 https://doi.org/10.1016/j.freeradbiomed.2018.06.013 mailto:minicucci@fmb.unesp.br https://doi.org/10.1016/j.freeradbiomed.2018.06.013 http://crossmark.crossref.org/dialog/?doi=10.1016/j.freeradbiomed.2018.06.013&domain=pdf and zinc as cofactors. Recently, our group published a sub-analysis showing that lower erythrocyte activity of SOD1 was an early predictor of the development of acute kidney injury in patients with septic shock [10]. However, it is important to note that cellular responses to ROS reflect a complex in- tegration of ROS type, location, levels, and polymorphisms of anti- oxidant enzymes. Single nucleotide polymorphisms (SNPs) are the most common type of variation in the human genome (about 90% of all variations), and refer to the replacement of only one nucleotide in a certain DNA po- sition. The presence of these polymorphisms can affect protein structure and function, leading to altered enzymatic responses. The decreased activity of antioxidant enzymes could lead to oxidative stress and affect the risk and prognosis of several diseases [11]. Variation in the SOD1 gene has been found to be associated with cardiovascular deaths in the general population, and the SNP rs1041740 might be associated with the development of ascites in patients with cirrhosis [12,13]. Data in the literature are scarce and there have been no studies evaluating the association between erythrocyte SOD1 activity and SOD1 polymorphisms and mortality in patients with septic shock. Therefore, the aim of the present study was to evaluate the influence of SOD1 polymorphisms on erythrocyte SOD1 activity as well as on the mortality of patients with septic shock. 2. Materials and methods 2.1. Study design This was a prospective observational clinical study, conducted from April 2014 to May 2015. The protocol was approved by the Ethics Committee of Botucatu Medical School (30457414.7.0000.5411). Written informed consent was obtained from all patients or relatives prior to their inclusion in the study. The sample size was calculated using the following variables: mortality rate of septic shock 40–60%, 95% confidence interval (CI), and 10% sample error. The result was a minimum sample size of 96 patients. The inclusion criteria were all individuals older than 18 years of age, of both sexes, with a diagnosis of septic shock at ICU admission. Septic shock was defined according to the Surviving Sepsis Guidelines [14]. Exclusion criteria were delayed diagnosis of septic shock (longer than 24 h), pregnancy, noradrenaline dose> 2.0 μg/kg/min, con- firmed brain death, palliative care, technical problems in the determi- nation of SOD1 SNPs and association of other types of shock. At the time of patient enrollment, demographic information, the Acute Physiology and Chronic Health Evaluation (APACHE II) score, and the Sequential Organ Failure Assessment (SOFA) score were re- corded. Blood samples were taken within the first 24 h of the patient's admission to determine erythrocyte SOD1 activity, SOD1 gene poly- morphisms, serum malondialdehyde (MDA), protein carbonyl, and zinc and copper levels. The ICU mortality rate was also recorded. 2.2. Laboratory analysis The hemogram was performed with a Coulter STKS hematological auto analyzer (Beckman Coulter, Inc., Brea, CA, USA). Total serum le- vels of C-reactive protein (CRP), albumin, creatinine, and urea were measured using the dry chemistry method (Ortho-Clinical Diagnostics VITROS 950®, Johnson & Johnson, New Brunswick, NJ, USA). Lactate was measured using the Roche OMNI S™ Blood Gas Analyzer (Roche Diagnostics, Basel, Switzerland). Serum MDA levels were analyzed based on the reaction with thio- barbituric acid by high-performance liquid chromatography, as pre- viously specified [15]. Protein carbonyl concentration was determined according to the method described by Reznick and Packer [16]. Zinc levels in plasma and erythrocytes was determined by flame atomic absorption spectrophotometry [17]. The same methodology was used for the analysis of plasma copper concentration. The SOD1 activity in erythrocytes was determined in a Lyasis biochemical analyzer ac- cording to the methodology recommended by the manufacturer (Ransod kit; Randox Laboratories Ltd., Crumlin, Antrim, UK) [18]. The reference range of normality was 70–110 μg/dL for plasma zinc concentration; 40–44 μg/g Hb for erythrocyte zinc concentration; 63.7–140.12 μg/dL for plasma copper concentration; and 1102–1601 U/g Hb for erythrocyte SOD1 activity according to the Ransod/Randox kit [19,20]. For comparison, the respective levels in 17 normal volunteers were 78.8 ± 18.2 μg/dL for plasma zinc con- centration; 68.3 ± 18.1 μg/g Hb for erythrocyte zinc concentration; 53.2 ± 10.0 μg/dL for plasma copper concentration; and 4340.4 ± 1430.2 U/g Hb for erythrocyte SOD1 activity. 2.3. SOD1 gene polymorphisms DNA was isolated from frozen blood samples using a method pre- viously described [21]. DNA integrity was verified using a 1% agarose gel, while DNA concentration was measured using a Nanodrop 8000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). The fol- lowing SNPs of the SOD1 gene were analyzed: rs4998557; rs2070424; rs10432782; rs1041740; rs11910115; rs202446; rs2173962; rs202449; and rs17880135. The genotyping assay was performed using the Taqman Open Array® system (Life Technologies Corporation, Carlsbad, CA, USA), with replicates of the SNP analyses of 20% of the samples and following the manufacturer's instructions. 2.4. Statistical analysis Data are expressed as mean± standard deviation (SD), in cases of a normal distribution, and median (including the lower and upper quartiles) in cases of a non-normal distribution or percentage. Comparisons between two groups for continuous variables were per- formed using Student's t-test or the Mann-Whitney U test. Comparisons between two groups for categorical variables were performed using the chi-square test or Fisher's exact test. The chi-square test with continuity correction was used to de- termine if genotype frequencies followed the Hardy-Weinberg equili- brium. Linkage disequilibrium between SNPs and haplotype block for- mation were tested using Haploview software version 4.2 (Broad Institute, Cambridge, MA, USA). Only haplotype blocks with fre- quencies above 1% were included in the subsequent analysis. Simple linear regression models were used to evaluate the associa- tion between erythrocyte SOD1 activity, demographics, and laboratory data. Simple and multiple linear regression models were designed to test differences in enzyme activity among the SOD1 SNP genotypes and were adjusted for variables that exhibited a significant difference in the univariate analyses (serum hemoglobin and MDA concentrations, plasma copper levels, and erythrocyte zinc concentration). To evaluate the influence of erythrocyte SOD1 activity on mortality, two logistic regression models were created. In the first model, the activity of the enzyme was adjusted for sex, age, and APACHE II score, and in the second model, the adjustment was made for age and protein carbonyl concentration (best fit model). Logistic regression models were also performed to evaluate the as- sociation between mortality and each SOD1 SNP. These models were adjusted by age and serum protein carbonyl concentration (best fit model). Data analysis was performed using Stata version 13.1 SE (StataCorp, College Station, TX, USA). The significance level was 5%. 3. Results During the study, 175 patients diagnosed with septic shock at the time of admission to the ICU were prospectively evaluated. However, 38 patients were subsequently excluded, 26 owing to technical pro- blems in the determination of SOD1 SNPs and 12 owing to a delayed N.A. Costa et al. Free Radical Biology and Medicine 124 (2018) 199–204 200 diagnosis of septic shock. Thus, 137 patients with septic shock were evaluated and included in the study (Fig. 1). The mean age of study patients was 63 ± 16 years and 58% were men. The median length of hospital stay and the ICU mortality rate were 8 (4–14) days and 66%, respectively. Concerning the cause of sepsis, 58% of the patients had pulmonary infection, 21% abdominal, 8% urinary, and 13% other types of focal infection. (Table 1) Regarding laboratory data, the median MDA concentration was 1.5 (0.8–2.2) μmol/L, protein carbonyl concentration was 24.0 (12.5–32.7) nmol/mL, plasma zinc level was 47.2 (35.4–70.3) μg/dL, erythrocyte zinc concentration was 61.0 (50.0–77.2) μg/g Hb, plasma copper con- centration was 66.6 ± 18.0 μg/dL, and erythrocyte SOD1 activity was 3191 (2230–4013) U/g Hb. The individuals who experienced mortality were older and more severely ill, with higher APACHE II and SOFA scores and higher levels of lactate, urea, and protein carbonyl concentrations. There were no differences in sex or in other biochemical variables. (Table 1) Several demographic and clinical variables were evaluated in rela- tion to SOD1 activity. Erythrocyte SOD1 activity had a positive asso- ciation with hemoglobin concentration (p=0.022), MDA concentra- tion (p=0.016), erythrocyte zinc level (p < 0.001), and plasma copper concentration (p=0.029) (Table 2). Higher values of erythrocyte SOD1 activity were associated with decreased ICU mortality when adjusted for age and protein carbonyl concentration in the logistic regression model (p=0.025). This re- lationship was also maintained in the highest tertile of erythrocyte SOD1 activity (odds ratio [OR]: 0.02; 95% CI: 0.00–0.78; p= 0.037). The other regression models found no association of SOD1 with ICU mortality (Table 3). SOD1 activity was also assessed according to the SNPs of the SOD1 gene. The nine SNPs were genotyped and all showed Hardy-Weinberg equilibrium. The genotypes of each SNP were separated into two groups: individuals with the dominant homozygous allele or individuals with the presence of the recessive allele (heterozygotes and homo- zygotes). Only the SNP SOD1 rs2070424 influenced erythrocyte SOD1 activity. For patients with the AA allele, the activity of SOD1 was lower in comparison to G-carriers (A/G+G/G genotype) (p= 0.019) (Table 4). On the other hand, there was no difference in SOD1 activity according to haplotype frequency (Table 5). None of the nine SOD1 SNPs were associated with ICU mortality, even when adjusted for age and protein carbonyl concentrations. The frequency of haplotypes was also not related to ICU mortality (Tables 6 and 7). 4. Discussion The objective of the present study was to evaluate the influence of SOD1 gene polymorphisms on the erythrocyte activity of the SOD1 enzyme and on the mortality of patients with septic shock. The results showed that increased activity of SOD1 was associated with a decreased mortality rate in patients with septic shock, and only the rs2070424 polymorphism interfered with SOD1 activity. Despite the publication of international guidelines for the early di- agnosis and appropriate treatment of septic shock, mortality rates re- main high [1,2]. In the present study, 66% of patients with septic shock progressed to death. Although this is a very high mortality rate, it is in Fig. 1. Flow diagram of study patients with septic shock. N.A. Costa et al. Free Radical Biology and Medicine 124 (2018) 199–204 201 accordance with the septic shock mortality observed in other studies in Latin America [22,23]. The determining factors for the progressive deterioration of patients with septic shock are not completely clear. However, it is believed that in addition to the profile of the infectious agent, the modulation of the pro and anti-inflammatory response resulting from the individual characteristics of the host is also important [1,2]. In critical patients, especially those with septic shock, there is a drastic increase in ROS production. Therefore, adequate antioxidant defense is essential to combat oxidative stress and improve survival [3–5]. In addition to increased ROS production, decreased antioxidant status has been reported in sepsis. Several causes such as dilution sec- ondary to fluid resuscitation, inadequate nutritional status, redistribu- tion, and losses through body fluids could explain these results [3,24]. When total redox capacity is measured is sepsis, a clear decrease is seen in septic patients compared to healthy controls [25]. In addition, this redox capacity usually remains low in patients who die, whereas it Table 1 Demographic and laboratory parameters in patients with septic shock according to ICU mortality. Variable ICU mortality Yes (n= 91) No (n=46) p value Age (years) 64.7 (14.6) 58.9 (18.0) 0.044 Male, n (%) 51 (56) 29 (63) 0.432 APACHE II score 20.1 (6.6) 14.8 (6.5) < 0.001 SOFA score 10.6 (2.8) 8.2 (2.3) < 0.001 Sepsis focus, n (%) 0.529 Respiratory 55 (61.1) 24 (52.2) Abdominal 18 (20.0) 11 (23.9) Urinary 5 (5.6) 6 (13.0) Other 12 (13.3) 5 (10.9) Hemoglobin (g/dL) 11.1 (2.2) 11.4 (2.0) 0.429 Hematocrit (%) 33.3 (6.7) 34.1 (5.7) 0.495 Leukocytes (10 ³/mm³)a 15.7 (12.1–22.2) 16 (12.2–22.9) 0.545 Albumin (g/dL) 2.27 (0.57) 2.43 (0.65) 0.162 Urea (mg/dL) 114.4 (69.7) 88.9 (62.6) 0.039 Creatinine (md/dL)a 2.0 (0.9–3.6) 1.4 (0.7–2.2) 0.169 Lactate (mmol/L)a 2.8 (1.5–3.8) 1.5 (1.1–2.5) 0.005 CRP (mg/dL) 32.2 (16.9) 30.4 (16) 0.539 Erythrocyte SOD1 activity (U/g Hb) 2957 (2205–4157) 3293 (2302–3936) 0.757 Protein carbonyl (nmol/ mL) 28.9 (7.1) 11.1 (4.3) < 0.001 MDA (nmol/L)a 1.61 (0.92–2.41) 1.53 (0.70–2.11) 0.122 APACHE - Acute Physiology and Chronic Health Evaluation; SOFA - Sequential Organ Failure Assessment; CRP – C reactive protein; MDA – malondialdehyde; Values are expressed as mean (SD) or median (25–75%). Student´s t – test was applied to determine test mean differences in continuous variables and Chi- squared test was applied for association between categorical variables. *p < 0.05 was considered statistically significant. a Log transformed variables. Table 2 Association between demographic or laboratory parameters and erythrocyte SOD1 activity. Parameters β ρ p value Age (years) 0.003 0.11 0.198 APACHE II score − 0.001 − 0.03 0.753 SOFA score − 0.007 − 0.05 0.606 Hemoglobin (g/dL) 0.045 0.20 0.022 Hematocrit (%) 0.007 0.09 0.305 Leukocytes (10 ³/mm³)a − 0.033 − 0.06 0.526 Albumin (g/dL) 0.050 0.06 0.468 Urea (mg/dL) − 0.001 − 0.08 0.334 Creatinine (md/dL)a − 0.031 − 0.06 0.479 Lactate (mmol/L)a − 0.007 − 0.01 0.911 CRP (mg/dL) − 0.002 − 0.07 0.403 MDA (nmol/L)a 0.135 0.21 0.016 Plasma zinc (μg/mL) − 0.001 − 0.01 0.895 Erythrocyte zinc (μg/g Hb) 0.550 0.170 < 0.001 Plasma copper (μg/dL) 0.005 0.19 0.029 Protein carbonyl (nmol/mL) 0.003 0.06 0.475 APACHE - Acute Physiology and Chronic Health Evaluation; SOFA - Sequential Organ Failure Assessment; CRP – C reactive protein; MDA – malondialdehyde. Simple linear regression models with erythrocyte SOD1 activity as dependent variable. *p < 0.05 was considered statistically significant. a Log transformed variables. Table 3 Logistic regression models for association between ICU mortality and tertiles of erythrocyte SOD1 activity in patients with septic shock. Variable OR (95% CI) p value p trend SOD1 erythrocyte activitya 0.881 1st tertile (957.3–2515.9 U/g Hb)a 1 2nd tertile (2515.9–3807.0 U/g Hb)a 1.11 (0.46–2.68) 0.822 3rd tertile (3807.0–11283.7 U/g Hb)a 0.94 (0.39–2.23) 0.884 SOD1 erythrocyte activityb 0.754 1st tertile (957.3–2515.9 U/g Hb)b 1 2nd tertile (2515.9–3807.0 U/g Hb)b 1.14 (0.44–2.98) 0.792 3rd tertile (3807.0–11283.7 U/g Hb)b 0.86 (0.33–2.22) 0.758 SOD1 erythrocyte activityc 0.025 1st tertile (957.3–2515.9 U/g Hb)c 1 2nd tertile (2515.9–3807.0 U/g Hb)c 0.15 (0.01–1.53) 0.109 3rd tertile (3807.0–11283.7 U/g Hb)c 0.02 (0.00–0.78) 0.037 *p < 0.05 was considered statistically significant. a Unadjusted. b adjusted by age, sex, and APACHE II score. c adjusted by protein carbonyl concentration and age. Table 4 Differences in erythrocyte SOD1 activity among SOD-1 SNP genotypes in pa- tients with septic shock. SOD-1 SNP SOD1 erythrocyte activity (U/g Hb) Median (25–75%) p valuea p valueb rs4998557 0.618 0.117 GG (n= 94) 3064 (2205–3925) GA+AA (n=43) 3459 (2438–4312) rs2070424 0.203 0.019 AA (n= 111) 2924 (2188–3925) AG+GG (n=26) 3798 (2516–4312) rs10432782 0.529 0.060 TT (n=100) 3064 (2205–3925) TG+GG (n=37) 3547 (2483–4312) rs1041740 0.749 0.406 GG (n= 61) 3321 (2380–4027) GT+TT (n=76) 3064 (2205–3942) rs11910115 0.285 0.321 AA (n= 129) 3288 (2344–4074) AC (n= 8) 2663 (1988–2890) rs202446 0.122 0.207 GG (n= 100) 2999 (2032–3942) GT+TT (n=37) 3459 (2490–4157) rs2173962 0.503 0.745 TT (n=125) 3288 (2344–3999) TC+CC (n= 12) 2663 (1988–3659) rs202449 0.102 0.255 AA (n= 100) 3350 (2456–4139) AT+TT (n= 37) 2669 (1912–3870) rs17880135 0.405 0.538 TT (n=128) 3101 (2188–3972) TG+GG (n=9) 3459 (2870–4157) Values are expressed as median (25–75%). Multiple linear regressions were applied to test differences in erythrocyte SOD 1 activity among genotypes in a dominant model. *p < 0.05 was considered statistically significant. a Unadjusted model. b Adjusted by hemoglobin, MDA, plasma copper, and erythrocyte zinc. N.A. Costa et al. Free Radical Biology and Medicine 124 (2018) 199–204 202 returns to normal in patients who survive sepsis [25–27]. It is also important to observe that mitochondria are a major source of ROS, and at the same time, a target for oxidative damage [3–6]. In addition, damage to mitochondria from oxidative stress appears to be funda- mental to the pathophysiology of organ failure and death in sepsis [3–6]. Experimental and clinical studies evaluated the association of SOD and sepsis [27–31]. Macarthur et al. showed in Sprague-Dawley rats that, postinfection treatment with the superoxide dismutase mimetic, protects against hypotension, vascular hyporeactivity to catechola- mines, and mortality associated with septic shock [28]. It is also in- teresting to observe that heterozygous nor homozygous SOD-1 over- expression did not improve the sepsis-related impairment of carbohydrate metabolism nor sustained improvement of the sepsis-re- lated impairment of myocardial-norepinephrine responsiveness, pos- sibly because of the lacking increase of the tissue catalase and the mi- tochondrial SOD activity [29,30]. Warner et al. showed that the plasma SOD concentration was ele- vated in septic patients, and levels were higher in non-survivors of sepsis compared to survivors [27]. Recently, we showed that lower erythrocyte SOD1 activity was an early predictor of the development of acute renal injury in patients with septic shock [8]. In the present study, we showed that increased erythrocyte SOD1 activity was associated with decreased ICU mortality in patients with septic shock. Although these results initially appear contradictory, Warner et al. evaluated the plasma levels of SOD1, but not its activity [27]. In addition, they in- cluded only 32 patients, with sepsis and septic shock, and they did not evaluated SOD1 polymorphisms or nutrition status of them [27]. It is also important to note that in the present study, the higher activity of SOD1 was related to erythrocyte zinc and plasma copper concentrations. Thus, adequate nutritional status of these minerals, known as cofactors of SOD1, contributes to improved action of the enzyme [7]. However, despite their relevant role, the concentrations of these minerals were not predictors of mortality in patients with septic shock. These results suggest that other factors, such as gene poly- morphisms, contribute to erythrocyte SOD1 activity. Polymorphisms of SOD1 gene may affect the enzymatic activities, leading to altered antioxidant defense. Several studies have evaluated the influence of SOD1 polymorphisms in patients with cirrhosis, car- diovascular disease, and cancer; however, there have been no previous studies in patients with septic shock [11–13,32]. The Yamagata study genotyped 639 SNPs of 2799 healthy in- dividuals and followed them for 10 years. The authors found that the SNPs rs1041740 and rs17880487 of the SOD1 gene were associated with cardiovascular mortality [12]. In the present study, we found no association of SNP rs1041740 with erythrocyte SOD1 activity or with mortality in patients with septic shock. This study was the first that evaluated nine polymorphisms of the SOD1 gene. Among these SNPs, only one showed influence on SOD1 activity. The presence of the dominant homozygous (AA) allele of the SNP rs2070424 was re- sponsible for the lower activity of SOD1. Thus, in addition to the nutritional status of copper and zinc, the presence of SOD1 gene polymorphisms may have directly interfered Table 5 Association between SOD-1 SNP and ICU mortality in patients with septic shock. SOD-1 SNP ICU mortality OR (95% CI) p value Adjusted OR (95% CI) p value rs4998557 GG (n=94) 1 1 GA+AA (n=43) 0.79 (0.37–1.68) 0.543 0.96 (0.18–5.30) 0.963 rs2070424 AA (n=111) 1 1 AG+GG (n=26) 0.94 (0.38–2.32) 0.901 0.80 (0.12–5.45) 0.817 rs10432782 TT (n= 100) 1 1 TG+GG (n=37) 0.66 (0.30–1.44) 0.295 0.95 (0.17–5.26) 0.950 rs1041740 GG (n=61) 1 1 GT+TT (n=76) 1.22 (0.60–2.49) 0.581 2.05 (0.43–9.73) 0.367 rs11910115 AA (n=129) 1 1 AC (n= 8) 0.83 (0.19–3.65) 0.809 2.21 (0.17–28.82) 0.545 rs202446 GG (n=100) 1 1 GT+TT (n=37) 0.56 (0.26–1.23) 0.147 1.00 (0.21–4.84) 0.995 rs2173962 TT (n= 125) 1 1 TC+CC (n=12) 0.47 (0.14–1.55) 0.215 1.16 (0.11–12.40) 0.904 rs202449 AA (n=100) 1 1 AT+TT (n=37) 0.77 (0.35–1.70) 0.521 0.23 (0.03–1.52) 0.126 rs17880135 TT (n= 128) 1 1 TG+GG (n=9) 1.83 (0.37–9.20) 0.461 0.58 (0.03–13.3) 0.734 Simple and multiple logistic regression, adjusted by age and carbonyl, were applied to test the association of SNPs in SOD-1 gene and ICU mortality. *p < 0.05 was considered statistically significant. Table 6 Association between SOD-1 haplotypes (rs4998557, rs10432782, rs2070424) and tertiles of erythrocyte SOD1 activity. Erythrocyte SOD1 activity Haplotypes 1st tertile 2nd tertile 3rd tertile Haplotype OR (95% CI) OR (95% CI) OR (95% CI) frequency (%)a AGG 1 1.21 (0.37–3.93) 2.70 (0.92–7.87) 9.5 AGA 1 0.32 (0.06–1.66) 0.51 (0.12–2.18) 4.0 ATA 1 0.98 (0.19–5.14) 0.68 (0.11–4.28) 2.9 Haplotype allele sequences correspond to SNP rs4998557, rs10432782, rs2070424, respectively. Multiple logistic regression, adjusted by hemoglobin, MDA, plasma copper, and erythrocyte zinc, was applied to test the association of haplotypes in SOD-1 gene and erythrocyte SOD1 activity tertiles. Haplotype “000” (prevalence of 83.6%a) was the reference haplotype. *p < 0.05 was considered statistically significant. a Generated with Haploview software. Table 7 Association between SOD-1 haplotypes (rs4998557, rs10432782, rs2070424) and ICU mortality in patients with septic shock. Haplotypes ICU mortality OR (95% CI) p value Haplotype frequency (%)a AGG 1.34 (0.53–3.37) 0.538 9.5 AGA 0.33 (0.09–1.25) 0.104 4.0 ATA 3.24 (0.38–27.7) 0.284 2.9 Haplotype allele sequences correspond to SNP rs4998557, rs10432782, rs2070424, respectively. Multiple logistic regression, adjusted by age and protein carbonyl concentration, was applied to test the association of haplo- types in SOD-1 gene and ICU mortality. Haplotype “000” (prevalence of 83.6%a) was the reference haplotype. *p < 0.05 was considered statistically significant. a Generated with Haploview software. N.A. Costa et al. Free Radical Biology and Medicine 124 (2018) 199–204 203 with the enzyme activity in our patients. Although one SNP was related to differences in SOD1 behavior, none were associated with ICU mor- tality. These results reinforce the complexity of the oxidative stress response and the participation of other antioxidants in the mortality of patients with septic shock. Some limitations of this study should be considered. Only patients from a single medical center were included. The timing of antibiotic dosing was not recorded, and blood samples were taken within the first 24 h, which is a relatively large window. In addition, we could not measure erythrocyte copper concentration owing to technical pro- blems. Despite these limitations, we strongly believe that our data contribute relevant knowledge regarding the role of the polymorphisms of the SOD1 gene in the activity of the SOD enzyme as well as its re- lationship with mortality in patients with septic shock. In conclusion, the SNP rs2070424 of the SOD1 gene interferes with erythrocyte SOD1 activity, and higher activity of SOD1 is associated with decreased mortality in patients with septic shock. Acknowledgements This study was funded by the State of São Paulo Research Foundation (FAPESP−2014/17262-0; 2014/07988-4) and CAPES (“Coordenação de Aperfeiçoamento de Pessoal de Nível Superior”). Declarations of interest None. Role of funding source The funding source had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication References [1] M. Singer, C.S. Deutschman, C.W. Seymour, M. Shankar-Hari, D. Annane, M. Bauer, R. Bellomo, G.R. Bernard, J.D. Chiche, C.M. Coopersmith, R.S. Hotchkiss, M.M. Levy, J.C. Marshall, G.S. Martin, S.M. Opal, G.D. Rubenfeld, T. van der Poll, J.L. Vincent, D.C. Angus, The third international consensus definition for sepsis and septic shock (Sepsis-3), J. Am. Med. Assoc. 315 (2016) 801–810. [2] A. Rhodes, L.E. Evans, W. Alhazzani, M.M. Levy, M. Antonelli, R. Ferrer, A. Kumar, J.E. Sevransky, C.L. Sprung, M.E. Nunnally, B. Rochwerg, G.D. Rubenfeld, D.C. Angus, D. Annane, R.J. Beale, G.J. Bellinghan, G.R. Bernard, J.D. Chiche, C. Coopersmith, D.P. De Backer, C.J. French, S. Fujishima, H. Gerlach, J.L. Hidalgo, S.M. Hollenberg, A.E. Jones, D.R. Karnad, R.M. Kleinpell, Y. Koh, T.C. Lisboa, F.R. Machado, J.J. Marini, J.C. Marshall, J.E. Mazuski, L.A. McIntyre, A.S. McLean, S. Mehta, R.P. Moreno, J. Myburgh, P. Navalesi, O. Nishida, T.M. Osborn, A. Perner, C.M. Plunkett, M. Ranieri, C.A. Schorr, M.A. Seckel, C.W. Seymour, L. Shieh, K.A. Shukri, S.Q. Simpson, M. Singer, B.T. Thompson, S.R. Townsend, T. Van der Poll, J.L. Vincent, W.J. Wiersinga, J.L. Zimmerman, R.P. Dellinger, Surviving sepsis campaign: international guidelines for management of sepsis and septic shock: 2016, Intensive Care Med. 43 (2017) 304–377. [3] W.A. Koekkoek, A.R. van Zanten, Antioxidant vitamins and trace elements in cri- tical illness, Nutr. Clin. Pract. 31 (2016) 457–474. [4] C.A. Prauchner, Oxidative stress in sepsis: pathophysiological implications justi- fying antioxidant co-therapy, Burns 43 (2017) 471–485. [5] M.E. Bernal, J. Varon, P. Acosta, L. Montagnier, Oxidative stress in critical care medicine, Int. J. Clin. Pract. 64 (2010) 1480–1488. [6] H.F. Galley, Oxidative stress and mitochondrial dysfunction in sepsis, Br. J. Anaesth. 107 (2011) 57–64. [7] W. Manzanares, R. Dhaliwal, X. Jiang, L. Murch, D.K. Heyland, Antioxidant mi- cronutrients in the critically ill a systematic review and meta-analysis, Crit. Care 16 (2012) R66. [8] I. Dalle-Donne, R. Rossi, D. Giustarini, A. Milzani, R. Colombo, Protein carbonyl groups as biomarkers of oxidative stress, Clin. Chim. Acta 329 (2003) 23–38. [9] M. Andresen, T. Regueira, A. Bruhn, D. Perez, P. Strobel, A. Dougnac, G. Marshall, F. Leighton, Lipoperoxidation and protein oxidative damage exhibit different ki- netics during septic shock, Mediat. Inflamm. 2008 (2008) 168652. [10] N.A. Costa, A.L. Gut, P.S. Azevedo, S.E. Tanni, N.B. Cunha, E.S. Magalhães, G.B. Silva, B.F. Polegato, L.A.M. Zornoff, S.A.R. de Paiva, A.L. Balbi, D. Ponce, M.F. Minicucci, Erythrocyte superoxide dismutase as a biomarker of septic acute kidney injury, Ann. Intensive Care 6 (2016) 95. [11] T.I. Kopp, U. Vogel, L.O. Dragsted, A. Tjonneland, G. Ravn-Haren, Association be- tween single nucleotide polymorphisms in the antioxidant genes CAT, GR and SOD1, erythrocyte enzyme activities, dietary and life style factors and breast cancer risk in a Danish, prospective cohort study, Oncotarget 8 (2017) 62984–62997. [12] Y. Otaki, T. Watanabe, S. Nishiyama, H. Takahashi, T. Arimoto, T. Shishido, T. Miyamoto, T. Konta, Y. Shibata, H. Sato, R. Kawasaki, M. Daimon, Y. Ueno, T. Kato, T. Kayama, I. Kubota, The impact of superoxide dismutase-1 genetic var- iation on cardiovascular and all-cause mortality in a prospective cohort study: the Yamagata (Takahata) study, PLoS One. 11 (2016) e0164732. [13] S. Schwab, J. Lehmann, P. Lutz, C. Jansen, B. Appenrodt, F. Lammert, C.P. Strassburg, U. Spengler, H.D. Nischalke, J. Trebicka, Influence of genetic variations in the SOD1 gene on the development of ascites and spontaneous bac- terial peritonitis in decompensated liver cirrhosis, Eur. J. Gastroenterol. Hepatol. 29 (2017) 800–804. [14] R.P. Dellinger, M.M. Levy, A. Rhodes, D. Annane, H. Gerlach, S.M. Opal, J.E. Sevransky, C.L. Sprung, I.S. Douglas, R. Jaeschke, T.M. Osborn, M.E. Nunnally, S.R. Townsend, K. Reinhart, R.M. Kleinpell, D.C. Angus, C.S. Deutschman, F.R. Machado, G.D. Rubenfeld, S. Webb, R.J. Beale, J.L. Vincent, R. Moreno, Surviving Sepsis Campaign Guidelines Committee including The Pediatric Subgroup, Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2012, Crit. Care Med. 41 (2013) 580–637. [15] F. Nielsen, B.B. Mikkelsen, J.B. Nielsen, H.R. Andersen, P. Grandjean, Plasma malondialdehyde as biomarker for oxidative stress: reference interval and effects of life-style factors, Clin. Chem. 43 (1997) 1209–1214. [16] A.Z. Reznick, L. Packer, Oxidative damage to proteins: spectrophotometric method for carbonyl assay, Methods Enzymol. 233 (1994) 357–363. [17] M.P. Rodriguez, A. Narizano, V. Demczylo, A. Cid, A simple method for the de- termination of zinc human plasma levels by flame atomic absorption spectro- fotometry, At. Spectrosc. 10 (1989) 68–70. [18] F.E. Ferro, V.B. de Sousa Lima, N.R. Soares, K.G. de Sousa Almondes, L.V. Pires, S.M. Cozzolino, do Nascimento, D. Marreiro, Parameters of metabolic syndrome and its relationship with zincemia and activities of superoxide dismutase and glu- tathione peroxidase in obese women, Biol. Trace Elem. Res. 143 (2011) 787–793. [19] R.S. Gibson, Principles of Nutritional Assessment, 2 ed, Oxford University Press, Nova York, 2005. [20] R.K. Murray, M. Jacob, J. Varghese, Plasma Proteins & Immunoglobulins, in: D.A. Bender, K.M. Botham, P.A. Weil, P.J. Kennelly, R.K. Murray, V.W. Rodwell (Eds.), Harper's Illustrated Biochemistry, 29th., McGraw-Hill, New York, 2011. [21] S.A. Miller, D.D. Dykes, H.F. Polesky, A simple salting out procedure for extracting DNA from human nucleated cells, Nucleic Acids Res. 16 (1988) 1215. [22] N.A. Costa, A.L. Gut, J.A. Pimentel, S.M. Cozzolino, P.S. Azevedo, A.A. Fernandes, B.F. Polegato, S.E. Tanni, R.D. Gaiolla, L.A. Zornoff, S.A. Paiva, M.F. Minicucci, Erythrocyte selenium concentration predicts intensive care unit and hospital mor- tality in patients with septic shock: a prospective observational study, Crit. Care 18 (2014) R92. [23] F.R. Machado, A.B. Cavalcanti, F.A. Bozza, E.M. Ferreira, F.S. Angotti Carrara, J.L. Sousa, N. Caixeta, R. Salomao, D.C. Angus, L.C. Pontes Azevedo, SPREAD Investigators, Latin American Sepsis Institute Network, The epidemiology of sepsis in Brazilian intensive care units (the Sepsis PREvalence Assessment Database, SPREAD): an observational study, Lancet Infect. Dis. 17 (2017) 1180–1189. [24] J. Abilés, A.P. de la Cruz, J. Castaño, M. Rodríguez-Elvira, E. Aguayo, R. Moreno- Torres, J. Llopis, P. Aranda, S. Argüelles, A. Ayala, A.M. de la Quintana, E.M. Planells, Oxidative stress is increased in critically ill patients according to antioxidant vitamins intake, independent of severity: a cohort study, Crit. Care 10 (2006) R146. [25] H.C. Cowley, P.J. Bacon, H.F. Goode, N.R. Webster, J.G. Jones, D.K. Menon, Plasma antioxidant potential in severe sepsis: a comparison of survivors and nonsurvivors, Crit. Care Med. 24 (1996) 1179–1183. [26] D. Bar-Or, R. Bar-Or, L.T. Rael, E.N. Brody, Oxidative stress in severe acute illness, Redox Biol. 4 (2015) 340–345. [27] A. Warner, A. Bencosme, D. Healy, C. Verme, Prognostic role of antioxidant en- zymes in sepsis: preliminary assessment, Clin. Chem. 41 (1995) 867–871. [28] H. Macarthur, D.M. Couri, G.H. Wilken, T.C. Westfall, A.J. Lechner, G.M. Matuschak, Z. Chen, Salvemini, Modulation of serum cytokine levels by a novel superoxide dismutase mimetic, M40401, in an Escherichia coli model of septic shock: Correlation with preserved circulating catecholamines, Crit Care Med. 31 (2003) 237–245. [29] V. Simkova, K. Baumgart, J. Vogt, U. Wachter, S. Weber, M. Gröger, G. Speit, P. Radermacher, G. Albuszies, E. Barth, The effects of superoxide dismutase over- expression on hepatic gluconeogenesis and whole-body glucose oxidation during resuscitated normotensive murine septic shock, Shock 30 (2008) 578–584. [30] K. Baumgart, V. Simkova, F. Wagner, S. Weber, M. Georgieff, P. Radermacher, G. Albuszies, E. Barth, Effect of SOD-1 over-expression on myocardial function during resuscitated murine septic shock, Intensive Care Med. 35 (2009) 344–349. [31] T. Brenner, S. Hofer, C. Rosenhagen, J. Steppan, C. Lichtenstern, J. Weitz, T. Bruckner, I.K. Lukic, E. Martin, A. Bierhaus, U. Hoffmann, M.A. Weigand, Macrophage migration inhibitory factor (MIF) and manganese superoxide dis- mutase (MnSOD) as early predictors for survival in patients with severe sepsis or septic shock, J. Surg. Res. 164 (2010) e163–e171. [32] D.A. Jones, S.L. Prior, T.S. Tang, S.C. Bain, S.J. Hurel, S.E. Humphries, J.W. Stephens, Association between the rs4880 superoxide dismutase 2 (C>T) gene variant and coronary heart disease in diabetes mellitus, Diabetes Res. Clin. Pract. 90 (2010) 196–201. N.A. Costa et al. Free Radical Biology and Medicine 124 (2018) 199–204 204 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref1 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref1 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref1 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref1 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref1 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref2 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref3 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref3 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref4 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref4 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref5 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref5 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref6 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref6 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref7 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref7 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref7 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref8 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref8 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref9 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref9 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref9 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref10 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref10 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref10 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref10 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref11 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref11 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref11 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref11 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref12 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref12 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref12 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref12 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref12 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref13 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref13 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref13 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref13 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref13 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref14 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref15 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref15 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref15 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref16 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref16 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref17 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref17 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref17 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref18 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref18 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref18 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref18 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref19 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref19 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref20 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref20 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref20 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref21 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref21 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref22 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref22 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref22 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref22 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref22 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref23 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref23 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref23 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref23 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref23 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref24 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref24 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref24 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref24 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref24 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref25 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref25 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref25 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref26 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref26 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref27 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref27 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref28 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref28 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref28 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref28 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref28 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref29 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref29 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref29 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref29 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref30 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref30 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref30 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref31 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref31 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref31 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref31 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref31 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref32 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref32 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref32 http://refhub.elsevier.com/S0891-5849(18)31040-2/sbref32 Erythrocyte SOD1 activity, but not SOD1 polymorphisms, is associated with ICU mortality in patients with septic shock Introduction Materials and methods Study design Laboratory analysis SOD1 gene polymorphisms Statistical analysis Results Discussion Acknowledgements Declarations of interest Role of funding source References