Hyperglycemia and glycemic variability are associated with the severity of sepsis in nondiabetic subjects

    Hyperglycemia and glycemic variability are associated with the severity of sepsis in non-diabetic subjects Lukana Preechasuk MD, Nattakarn Suwansaksri MD, Nantawan Ipichart MD, Sathit Vannasaeng MD, Chairat Permpikul MD, Apiradee Sriwijitkamol MD PII: DOI: Reference:

S0883-9441(16)30345-8 doi: 10.1016/j.jcrc.2016.12.005 YJCRC 52363

To appear in:

Journal of Critical Care

Please cite this article as: Preechasuk Lukana, Suwansaksri Nattakarn, Ipichart Nantawan, Vannasaeng Sathit, Permpikul Chairat, Sriwijitkamol Apiradee, Hyperglycemia and glycemic variability are associated with the severity of sepsis in non-diabetic subjects, Journal of Critical Care (2016), doi: 10.1016/j.jcrc.2016.12.005

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Hyperglycemia and glycemic variability are associated with the severity of sepsis in nondiabetic subjects

RI P

T

Lukana Preechasuk MDa, Nattakarn Suwansaksri MDa, Nantawan Ipichart MDa, Sathit

a

SC

Vannasaeng MDa, Chairat Permpikul MDb, Apiradee Sriwijitkamol MDa,*

Division of Endocrinology and Metabolism, Department of Medicine, Faculty of Medicine

Division of Critical Care, Department of Medicine, Faculty of Medicine Siriraj Hospital,

ED

Mahidol University, Thailand, 10700

MA

b

NU

Siriraj Hospital, Mahidol University, Thailand, 10700

*

Corresponding author: Tel: + 66 2 419 7799; fax: +66 2 419 7792.

AC

CE

PT

E-mail address: [email protected]

ACCEPTED MANUSCRIPT Abstract Purpose: To compare glucose variability (GV) obtained via continuous glucose monitoring

RI P

between GV and sepsis severity in non-diabetic sepsis patients.

T

(CGM) between non-diabetic sepsis patients and healthy subjects and to seek associations

Methods: Non-diabetic sepsis inpatients and healthy controls received a 72-hour CGM (iPro2,

SC

Medtronic) post-admission and post-OGTT, respectively. The mean glucose level (MGL) along with GV represented by standard deviation (SD) and the mean amplitude of glycemic excursion

NU

(MAGE) were calculated at 24 and 72 hours. Sepsis severity was evaluated with the Sepsis-

MA

related Organ Failure Assessment Score (SOFA). MGL and GV in patients with SOFA ≥9 and <9 were compared.

ED

Results: Thirty non-diabetic sepsis and 10 healthy subjects were recruited. No differences were found between groups except for higher patient age in sepsis patients. The MGL and MAGE72h

PT

of sepsis patients were significantly higher than those of healthy subjects. MGL and GV24h were

CE

higher in patients with SOFA ≥9 than in patients with SOFA <9 (MGL24h 13927 vs. 19517;

p=0.01).

AC

p<0.001, SD24h 32 [28, 36] vs.19 [5, 58]; p=0.02 and MAGE24h 94 [58, 153] vs. 54 [16, 179];

Conclusion: Non-diabetic sepsis patients had higher MGL and GV values than healthy subjects. MGL and GV24h were associated with sepsis severity.

Keyword glycemic variability; continuous glucose monitoring; mean amplitude of glycemic excursion; sepsis1

1

Abbreviations glucose variability (GV), continuous glucose monitoring (CGM), mean glucose level (MGL), standard deviation (SD), mean amplitude of glycemic excursion (MAGE), Sepsis-related Organ Failure Assessment Score (SOFA)

ACCEPTED MANUSCRIPT Introduction Hyperglycemia is associated with increased mortality in critically ill patients [1, 2]. In 2001, a study demonstrated that intensive glucose control decreased mortality in surgical ICU

RI P

T

patients compared with conventional glucose control [3]. However, several subsequent studies have not confirmed this survival benefit [4-9]. Glucose variability (GV), a fluctuation of glucose

SC

levels, might explain these discordant results. Interventional studies both in animals and humans have shown that intermittent high glucose levels stimulate reactive oxygen species

NU

overproduction and increase cellular apoptosis, leading to further impaired endothelial function

MA

compared to persistent high glucose [10-12]. Evidence showing the relationship between GV and mortality in sepsis has been increasing in both diabetic and non-diabetic patients [13, 14]. Most

ED

studies have calculated a standard deviation (SD) to represent GV from capillary, venous, or arterial glucose values four to seven times per day. However, GV from such data has not

PT

included the highest or lowest glucose levels or the glucose pattern. In contrast, continuous

CE

glucose monitoring (CGM) can demonstrate all glucose fluctuations over a 24-hour period because it measures interstitial glucose every five minutes. To date, there have been sparse

AC

prospective GV data calculated from CGM in non-diabetic patients with sepsis. Furthermore, no prospective study has demonstrated whether GV in non-diabetic patients with sepsis is different from GV in healthy subjects or whether GV is associated with sepsis severity in such patients. Therefore, the present study aimed to compare glucose profiles, including mean glucose level (MGL) and GV, between non-diabetic sepsis patients and healthy subjects and to determine the association of GV and sepsis severity. In this study, we also recruited normal healthy subjects to study GV because there is still no normal reference range of GV for non-diabetic persons [15, 16].

ACCEPTED MANUSCRIPT Methods

Patient selection

RI P

T

This study was a prospective study conducted from May 2013 to April 2015. All subjects were aged >18 years and without diabetes. Patients in the sepsis/septic shock group were

SC

identified by the Surviving Sepsis Campaign Guidelines 2012 [17], had no history of diabetes, had an HbA1c of less than 6.5% (48 mmol/mol), and had been admitted to a medical ward or a

NU

medical intensive care unit (ICU). We excluded patients with very severe disease, including

MA

patients with a Sequential Organ Failure Assessment (SOFA) score >15, a platelet count <20,000 /mm3, and uncorrected coagulopathy. Patients who refused to participate were also excluded.

ED

The inclusion criteria for the healthy control group were (1) normal glucose metabolism defined as a fasting plasma glucose <100 mg/dL, (2) a 2-hour plasma glucose <140 mg/dL after a 75-

PT

gram oral glucose tolerance test, and (3) hemoglobin A1C (HbA1c) <6.5% (48 mmol/mol). All

CE

participants provided written informed consent. The Institutional Review Board of the Faculty of

Procedure

AC

Medicine Siriraj Hospital approved the protocol.

Each sepsis/septic shock subject underwent a 72-hour CGM (iPro2, Medtronic, Minnesota, USA) within 12 hours of admission. The glucose sensor (Sof-Sensor, MMT-7002C, Medtronic, Minnesota, USA) was attached to subcutaneous tissue at the lower abdominal wall using an insertion device and measured interstitial fluid glucose every five minutes via the glucose oxidase method. After 72 hours, the CGM device was removed, and the digital recorder was connected to the docking station of the iPro2 to upload the data to the CareLink®iPro software program. CGM calibration was performed by point-of-care glucose testing four to six times per day using the SureStep Flexx Meter (Johnson & Johnson, New Jersey, USA). Blood

ACCEPTED MANUSCRIPT used for point-of-care glucose testing was drawn either from an arterial line or a venous line if present. Otherwise, capillary blood was used. All glucose tests were performed by trained health care personnel, and all glucose meters were calibrated daily. Glycemic control and insulin

RI P

T

treatment for sepsis patients were managed by physicians to control the glucose level between 140-180 mg/dL. This is the range recommended by our hospital’s clinical policy.

SC

The healthy control group underwent a 2-hour oral glucose tolerance test. Next, each healthy subject was trained to use the SureStep Meter (Johnson & Johnson, New Jersey, USA)

NU

and was instructed to measure capillary plasma glucose (CPG) four times per day (before each

MA

meal and at bedtime). Two hours after the oral glucose tolerance test was finished, each healthy subject underwent the 72-hour CGM using iPro2 (Medtronic, Minnesota, USA). All healthy

ED

subjects were allowed normal activity and ate ad libitum.

PT

Instrument

CE

The iPro2 digital recorder is a CGM device that collects and stores data from a glucose sensor. A previous study [18] demonstrated the high accuracy of the iPro2 (Medtronic,

AC

Minnesota, USA) and the Sof-Sensor glucose sensor (MMT-7002C, Medtronic, Minnesota, USA). The mean absolute relative difference, which indicates the average difference between the blood glucose measurement and the interstitial fluid glucose values from the sensor, was 9.9% in adults, while the Clarke error grid analysis showed that 99.0% of adult paired values were within zones A and B [18]. The SureStep Flexx Meter (Johnson&Johnson, New Jersey, USA) meets the ISO15197: 2013 standard. The coefficient of variation of SureStep Flexx Meter ranged from 4.2% to 5.9%. The HbA1c testing method was turbidimetric inhibition immunoassay (Roche Cobas Integra 800, Rotkreuz, Switzerland). The National Glycohemoglobin Standardization Program certifies our laboratory. The coefficient of variation of HbA1c testing ranged from 1.5% to 1.62%.

ACCEPTED MANUSCRIPT Data collection Baseline characteristics, including age, gender, body mass index (BMI), HbA1c level, and underlying conditions, were recorded. In the sepsis subjects, the SOFA scores, the diagnostic

RI P

T

criteria of septic shock, the number of organ failure, and the clinical outcomes were also recorded. Sepsis patients were divided into groups based on a sepsis SOFA score <9 and a sepsis

SC

SOFA score 9 following a previous study [19] showing that an initial SOFA score >8 discriminated between survival and non- survival. We used glucose data obtained from CGM to

NU

calculate the glucose profiles, including the mean glucose level (MGL) and GV as reflected by

MA

SD values as well as the mean amplitude of glycemic excursion (MAGE), which is the mean absolute value of any delta blood glucose levels (consecutive values) that are >1 SD of the entire

ED

set of blood glucose level data [20]. The MAGE was calculated with a computerized calculator (Glyculator program) provided by Czerwoniuk D [21]. Due to variations in glucose levels over

PT

24 hours, we used glucose data obtained over 24 hours and 72 hours to calculate 24-hour and 72-

AC

Statistical analyses

CE

hour glucose profiles, respectively.

SPSS version 18 (Chicago, IL, USA) was used for all analyses. The data with normal distribution patterns were presented as the means±SD, while the data with non-normal distribution patterns were presented as the medians (min, max). Categorical variables were shown as frequency and percentage. The differences between sepsis and healthy subjects were determined using the t-test for normally distributed data or the chi-square test for non-normally distributed data. Relationships between the glucose profiles obtained from CGM and glucose meter were evaluated by Pearson or Spearman correlations. For all analyses, a p-value <0.05 was considered statistically significant.

ACCEPTED MANUSCRIPT Results

Baseline characteristics

RI P

T

Thirty sepsis patients and 10 healthy control subjects were recruited. The patients in the sepsis group were significantly older (67±20 vs.44±5 years old, respectively; p < 0.001) and

SC

included more male subjects than the control group (n=18 vs. n=3, respectively; p= 0.15). There

NU

were no differences in either HbA1c (5.6±0.6 vs.5.3±0.3%, respectively; p = 0.12) or BMI (20.7±3.8vs.22.9±3.6 kg/m2, respectively; p= 0.12) between the groups (Table 1). The median

MA

SOFA score in the sepsis group was 4.5 (1, 15). Only 29 patients were included in the 72-hour data analysis because one patient in the sepsis group died within 24 hours of admission.

ED

No CGM sensors were removed because of localized pain, irritation, or bleeding, and no infectious complications occurred. No serious adverse events were reported during CGM usage

CE

PT

in this study.

Mean glucose levels and glucose variability between the sepsis and control groups

AC

Table 2 shows the glucose profiles according to CGM in the sepsis and control groups. The MGL at 24 hours (MGL24h) and at 72 hours after starting CGM (MGL72h) were significantly higher in the sepsis group. Moreover, the GV as assessed by MAGE at 72 hours (MAGE72h) was significantly greater in the sepsis group than in the other group. There was a trend of higher GV at 24 hours in the sepsis group than in the control group; however, these differences were not statistically significant. One subject in the sepsis group had episodes of hypoglycemia (a glucose level of 40 mg/dL) that were detected by CGM and not by glucose meter. However, the events spontaneously recovered without treatment.

ACCEPTED MANUSCRIPT MGL and GV in sepsis patients with differing severities The observed 28-day mortality rate of sepsis patients in this study was 13%. We used SOFA scores to assess the severities of sepsis. Sepsis patients with SOFA scores >9 were older

RI P

T

than those with SOFA scores <9. There were no differences in BMI or HbA1c levels between groups. Patients with sepsis SOFA scores ≥9 had significantly higher MGL24h and MGL72h

SC

values as well as a higher GV at 24 hours than the other group. There was a trend towards a higher GV at 72 hours in more severe sepsis patients; however, this was not statistically

NU

significant (Table 3). Patients with sepsis SOFA scores ≥9 received insulin treatment,

MA

intravenous glucose infusion, and inotropic drug treatment more often than the less severe patients 24 and 72 hours after starting CGM.

ED

There was a positive correlation between MGL and SOFA score at 24 hours (r = 0.72, p<0.01) (figure 1). As shown in figure 1, the correlation between SD as well as MAGE and

PT

SOFA score at 24 hours were 0.31, but this correlation did not reach statistically significance

CE

(p=0.09). At 72 hours, the correlation between MGL and SOFA score still had a positive correlation with the Pearson’s correlation of 0.63, and p<0.01 whereas there was no significant

AC

correlation between GV and SOFA (r= 0.2, p=0.3). There were very good correlations between SD and MAGE obtained from CGM at 24 hours and 72 hours in both sepsis patients and healthy subjects. Spearman’s correlation was 0.97 at 24 hours and 72 hours in the sepsis group and 0.94 at 24 hours and 72 hours in the healthy group.

Correlations between glucose profiles obtained from CGM and a glucose meter MGL and SD obtained from CGM and a glucose meter were strongly correlated in the sepsis group, whereas there was no correlation in the control group (Table 4). The overall

ACCEPTED MANUSCRIPT Spearman’s correlation between MAGE obtained from CGM and SD obtained from a glucose meter was 0.89 at the 24-hour and 72-hour periods.

RI P

T

Discussion

This study demonstrates that non-diabetic patients with sepsis had a higher mean daily

SC

glucose level than healthy subjects. It also demonstrates that sepsis patients with SOFA scores ≥9 had a higher MGL at 24 hours and 72 hours after starting CGM than patients with SOFA

NU

scores <9. This finding is consistent with those of other studies that have shown that critically ill

MA

patients and severe sepsis patients have stress-induced hyperglycemia due to increased cortisol and inflammatory cytokines levels, which results in increased gluconeogenesis and insulin

ED

resistance [22-24]. However, patients in the sepsis group and the high SOFA score (≥9) group in the present study were older than patients in the healthy subject group and the low SOFA score

PT

(<9) group, respectively. Several studies have shown that the elderly have greater insulin

CE

resistance than younger subjects [25-27]. Therefore, the difference in age might explain the higher MGL in sepsis patients, especially in those with severe sepsis. Despite this association

AC

between sepsis severity and hyperglycemia, studies [3-9] that have aimed to normalize glucose levels have shown inconclusive findings on sepsis outcomes. Subsequent studies have thus explored other aspects of glucose, such as GV and sepsis outcomes, and have demonstrated a relationship between GV and mortality in sepsis patients [13, 14, 28]. Interestingly, retrospective studies by Krinsley et al. showed that there was a relationship between GV and mortality only in critically ill patients without diabetes [29, 30]. This study demonstrates that sepsis patients without diabetes had higher GV as assessed by MAGE72h and had a trend toward a higher of SD24h compared to patients in the healthy control group. The reasons for higher MAGE at 72 hours but not during the first 24 hours after starting CGM in sepsis patients are unclear. Initially, we considered the possibility that

ACCEPTED MANUSCRIPT differences in dietary intake, insulin, and steroid administration between the 24- and 72-hour periods may have explained the discrepancy. We found that there were similar numbers of patients who received insulin, steroids, and intravenous glucose during the 24-hour and 72-hour

RI P

T

periods. However, there were differences in the type, dose, and duration of medications during these periods. These differences might explain the observed differences in MAGE during the 24-

SC

hour and 72-hour periods. Nevertheless, the present study shows that the median SD and MAGE values in sepsis patients without diabetes were higher than the proposed normal references for

NU

healthy Asian subjects [16].

MA

We have demonstrated that sepsis patients with SOFA scores ≥9 had higher GV during the first 24 hours after starting CGM than patients with SOFA scores <9. The percentages of

ED

patients who received insulin, intravenous glucose, and inotropic drugs were significantly higher in the more severe sepsis group than in the less severe sepsis group. These proportions might

PT

explain the high GV in severe sepsis patients. Although the patients with SOFA scores ≥9 had a

CE

trend of higher GV in the 72-hour period compared to the patients with SOFA scores <9, this difference was not statistically significant. These non-significant results might have been caused

AC

by the very small number of patients in the SOFA score ≥9 group (n=5). Therefore, further study is needed to explore the association between GV and the severity of sepsis and the benefit of controlling GV in sepsis patients. Few studies have determined the advantage of controlling GV using real time CGM compared to intermittent glucose monitoring in critically ill patients [31, 32]. However, those studies could not demonstrate a difference in GV between continuous and intermittent glucose monitoring due to the use of a well-established insulin protocol in the control groups. A previous study showed a good correlation between SD and MAGE in diabetic nonhospitalized patients [33]. Our data confirm that SD and MAGE values obtained from CGM also had a strong correlation in non-diabetic sepsis patients. In addition, we demonstrated a strong

ACCEPTED MANUSCRIPT correlation between glucose profiles obtained by CGM and the glucose meter in the sepsis group. Therefore, SD calculated from the glucose meter may be used to determine the GV in sepsis patients. We also demonstrated that the use of CGM can detect unrecognized hypoglycemia in

RI P

T

sepsis patients.

The present study has several strengths and limitations. One strength is its prospective

SC

design that aimed to assess the association between GV and the severity of sepsis via CGM. This mode of monitoring can record glucose levels every five minutes and can record every glycemic

NU

excursion during the day. Another strength is that we included only non-diabetic patients.

MA

Therefore, the effect on the glucose profiles of sepsis patients in the present study must have been caused by the direct effect of sepsis. Limitations included the small number of patients,

ED

especially of patients with severe sepsis; we excluded patients with very severe sepsis as defined as a SOFA score >15.

PT

In conclusion, non-diabetic sepsis patients had abnormal glucose profiles, as evidenced

CE

by higher MGL and GV values than those of healthy subjects. We found that a greater severity of sepsis was associated with a greater glycemic level and GV, especially in the first 24 hours after

AC

starting CGM. Therefore, further studies aimed at the benefit of controlling GV may provide important clinical data for glucose management in non-diabetic sepsis patients.

Acknowledgements This work was supported by grants from the Siriraj Grant for Research Development, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand. The authors give special thanks to Prof. Wojciech Mlynarski, M.D., Ph.D. for providing the Glyculator program and Ms. Kemjira Karngateklang for her assistance with statistics. Conflicts of interest All authors have no potential conflicts of interest to disclose.

ACCEPTED MANUSCRIPT References 1

Falciglia M, Freyberg RW, Almenoff PL, D'Alessio DA, Render ML. Hyperglycemiarelated mortality in critically ill patients varies with admission diagnosis. Crit Care Med

RI P

2

T

2009;37:3001-9.

Krinsley JS. Association between hyperglycemia and increased hospital mortality in a

3

SC

heterogeneous population of critically ill patients. Mayo Clin Proc 2003;78:1471-8. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, et al.

Nice-Sugar Study Investigators, Finfer S, Chittock DR, Su SY, Blair D, Foster D, et al.

MA

4

NU

Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-67.

Intensive versus conventional glucose control in critically ill patients. N Engl J Med

5

ED

2009;360:1283-97.

Arabi YM, Dabbagh OC, Tamim HM, Al-Shimemeri AA, Memish ZA, Haddad SH, et al.

PT

Intensive versus conventional insulin therapy: a randomized controlled trial in medical

6

CE

and surgical critically ill patients. Crit Care Med 2008;36:3190-7. De La Rosa GDC, Donado JH, Restrepo AH, Quintero AM, Gonzalez LG, Saldarriaga

AC

NE, et al. Strict glycaemic control in patients hospitalised in a mixed medical and surgical intensive care unit: a randomised clinical trial. Crit Care 2008;12:R120. 7

Preiser JC, Devos P, Ruiz-Santana S, Melot C, Annane D, Groeneveld J, et al. A prospective randomised multi-centre controlled trial on tight glucose control by intensive insulin therapy in adult intensive care units: the Glucontrol study. Intensive Care Med 2009;35:1738-48.

8

Coiitss Study Investigators, Annane D, Cariou A, Maxime V, Azoulay E, D'Honneur G, et al. Corticosteroid treatment and intensive insulin therapy for septic shock in adults: a randomized controlled trial. JAMA 2010;303:341-8.

ACCEPTED MANUSCRIPT 9

Song F, Zhong LJ, Han L, Xie GH, Xiao C, Zhao B, et al. Intensive insulin therapy for septic patients: a meta-analysis of randomized controlled trials. Biomed Res Int 2014;2014:698265.

T

Piconi L, Quagliaro L, Assaloni R, Da Ros R, Maier A, Zuodar G, et al. Constant and

RI P

10

intermittent high glucose enhances endothelial cell apoptosis through mitochondrial

11

SC

superoxide overproduction. Diabetes Metab Res Rev 2006;22:198-203. Turina M, Fry DE, Polk HC, Jr. Acute hyperglycemia and the innate immune system:

Ceriello A, Esposito K, Piconi L, Ihnat MA, Thorpe JE, Testa R, et al. Oscillating

MA

12

NU

clinical, cellular, and molecular aspects. Crit Care Med 2005;33:1624-33.

glucose is more deleterious to endothelial function and oxidative stress than mean

13

ED

glucose in normal and type 2 diabetic patients. Diabetes 2008;57:1349-54. Ali NA, O'Brien JM, Jr., Dungan K, Phillips G, Marsh CB, Lemeshow S, et al. Glucose

Krinsley JS. Glycemic variability: a strong independent predictor of mortality in critically

CE

14

PT

variability and mortality in patients with sepsis. Crit Care Med 2008;36:2316-21.

ill patients. Crit Care Med 2008;36:3008-13. Hill NR, Oliver NS, Choudhary P, Levy JC, Hindmarsh P, Matthews DR. Normal

AC

15

reference range for mean tissue glucose and glycemic variability derived from continuous glucose monitoring for subjects without diabetes in different ethnic groups. Diabetes Technol Ther 2011;13:921-8. 16

Zhou J, Li H, Ran X, Yang W, Li Q, Peng Y, et al. Establishment of normal reference ranges for glycemic variability in Chinese subjects using continuous glucose monitoring. Med Sci Monit 2011;17:CR9-13.

17

Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013;41:580-637.

ACCEPTED MANUSCRIPT 18

Welsh JB, Kaufman FR, Lee SW. Accuracy of the Sof-sensor glucose sensor with the iPro calibration algorithm. J Diabetes Sci Technol 2012;6:475-6.

19

Ferreira FL, Bota DP, Bross A, Melot C, Vincent JL. Serial evaluation of the SOFA score

20

RI P

T

to predict outcome in critically ill patients. JAMA 2001;286:1754-8. Siegelaar SE, Holleman F, Hoekstra JB, DeVries JH. Glucose variability; does it matter?

21

SC

Endocr Rev 2010;31:171-82.

Czerwoniuk D, Fendler W, Walenciak L, Mlynarski W. GlyCulator: a glycemic

NU

variability calculation tool for continuous glucose monitoring data. J Diabetes Sci

22

MA

Technol 2011;5:447-51.

Marik PE, Raghavan M. Stress-hyperglycemia, insulin and immunomodulation in sepsis.

23

ED

Intensive Care Med 2004;30:748-56.

Leonidou L, Michalaki M, Leonardou A, Polyzogopoulou E, Fouka K, Gerolymos M, et

PT

al. Stress-induced hyperglycemia in patients with severe sepsis: a compromising factor

Dungan KM, Braithwaite SS, Preiser JC. Stress hyperglycaemia. Lancet 2009;373:1798807.

25

AC

24

CE

for survival. Am J Med Sci 2008;336:467-71.

Facchini FS, Hua N, Abbasi F, Reaven GM. Insulin resistance as a predictor of agerelated diseases. J Clin Endocrinol Metab 2001;86:3574-8.

26

Lechleitner M. Obesity and the metabolic syndrome in the elderly--a mini-review. Gerontology 2008;54:253-9.

27

Ghosh S, Lertwattanarak R, Garduno Jde J, Galeana JJ, Li J, Zamarripa F, et al. Elevated muscle TLR4 expression and metabolic endotoxemia in human aging. J Gerontol A Biol Sci Med Sci 2015;70:232-46.

28

Eslami S, Taherzadeh Z, Schultz MJ, Abu-Hanna A. Glucose variability measures and their effect on mortality: a systematic review. Intensive Care Med 2011;37:583-93.

ACCEPTED MANUSCRIPT 29

Krinsley JS. Glycemic variability and mortality in critically ill patients: the impact of diabetes. J Diabetes Sci Technol 2009;3:1292-301.

30

Krinsley JS, Egi M, Kiss A, Devendra AN, Schuetz P, Maurer PM, et al. Diabetic status

RI P

T

and the relation of the three domains of glycemic control to mortality in critically ill patients: an international multicenter cohort study. Crit Care 2013;17:R37. Brunner R, Adelsmayr G, Herkner H, Madl C, Holzinger U. Glycemic variability and

SC

31

glucose complexity in critically ill patients: a retrospective analysis of continuous glucose

De Block CE, Gios J, Verheyen N, Manuel-y-Keenoy B, Rogiers P, Jorens PG, et al.

MA

32

NU

monitoring data. Crit Care 2012;16:R175.

Randomized evaluation of glycemic control in the medical intensive care unit using real-

ED

time continuous glucose monitoring (REGIMEN Trial). Diabetes Technol Ther 2015;17:889-98.

PT

Rodbard D. Interpretation of continuous glucose monitoring data: glycemic variability

CE

and quality of glycemic control. Diabetes Technol Ther 2009;11 Suppl 1:S55-67.

AC

33

ACCEPTED MANUSCRIPT Table 1. Characteristics of patients in the sepsis and control groups Sepsis/septic shock

Healthy n=10

Male gender (%)

18 (60)

3 (30)

0.148

Age (years)

67±20

44±5

<0.001

22.9±3.6

0.123

SC

2

T

n=30

RI P

p-value

20.7±3.8

HbA1C (%)

5.6±0.6

5.3±0.3

0.115

HbA1C (mmol/mol)

37.3±6

34.8±3

0.115

MA

NU

BMI (kg/m )

4.5 (1,15)

Patients with insulin treatment (%)

PT

Patients with intravenous glucose

ED

SOFA score

CE

treatment (%)

8 (27)

-

17 (57)

-

Patients with ventilator support (%)

12 (40)

Patients with inotropic drug use (%)

16 (53)

AC

-

Source of infection Pneumonia (%)

16 (53)

Urinary tract infection (%)

7 (23)

Intra-abdominal infection (%)

3 (10)

Other (%)

4 (14)

The data are presented as the meansSD, median (min, max) or percentage as appropriate. BMI body mass index, HbA1C hemoglobin A1C, SOFA Sequential Organ Failure Assessment

RI P

T

ACCEPTED MANUSCRIPT

Glucose profiles

SC

Table 2. MGL and GV from CGM in the sepsis and control groups Sepsis

97.49.0

<0.001

14.2 (9.7-21.0)

0.08

57.2 (15.9-179.1)

44.2 (33.7-60.7)

0.118

n=29

n=10

150.233.7 19 .5 (4.7-58.2)

MAGE24h (mg/dL)

ED

72-hour glucose profiles

MA

SD24h (mg/dL)

MGL72h (mg/dL)

143.531.42

99.36.8

<0.001

22.1 (8.8-61.1)

14.3 (7.9-20.7)

0.142

46.6 (25.5-62.6)

0.005

PT

SD72h (mg/dL) MAGE72h (mg/dL)

p-value

n=10

NU

n=30

24-hour glucose profiles MGL24h (mg/dL)

Control

69.5 (31.0-176.6)

CE

The data are presented as the meansSD, median (min, max), or percentage as appropriate.

AC

MGL mean glucose level, SD standard deviation, MAGE mean amplitude of glycemic excursion

ACCEPTED MANUSCRIPT Table 3. Glucose profiles in patients with sepsis SOFA scores <9 and sepsis SOFA scores ≥9 SOFA score ≥9

p-value

64±20

81±8

0.048

BMI (kg/m2)

20.6±4.0

21.0±3.0

HbA1c (%)

5.5±0.6 37±6

24-hour glucose profiles

n=24

MGL (mg/dL)

RI P

HbA1c (mmol/mol)

SC

Age (year)

T

SOFA score <9

Characteristics

SD (mg/dL)

19 (5,58)

54 (16,179)

MA

MAGE (mg/dL)

NU

13927

0.82

5.6±0.5

0.60

38.4±5

0.60

n=6 19517

<0.001

32 (16,55)

0.02

94 (58,153)

0.01

3 (12.5)

5 (83.3)

<0.001

Steroid treatment (n, %)

11 (45.8)

4 (66.7)

0.65

8 (33.3)

5 (83.3)

0.03

10 (41.7)

6 (100)

0.01

n=24

n=5

13528

1848

0.001

21 (9, 61)

33 (17,42)

0.11

65 (26,177)

105 (53,128)

0.11

Insulin treatment (n, %)

3 (12.5)

5 (83.3)

<0.001

Steroid treatment (n, %)

12 (50)

4 (66.7)

0.66

Intravenous glucose (n, %)

11 (45.8)

6 (100)

0.02

Inotropic drug regimen (n, %)

10 (41.7)

6 (100)

0.01

Intravenous glucose (n, %)

72-hour glucose profiles

MAGE

AC

SD

CE

MGL

PT

Inotropic drug regimen (n, %)

ED

Insulin treatment (n, %)

The data are presented as the meanSD, median (min, max), or percentage as appropriate. MGL mean glucose level, SD standard deviation, MAGE mean amplitude of glycemic excursion

ACCEPTED MANUSCRIPT Table 4. Correlations between MGL and SD at 24 and 72 hours from CGM and glucose meter Sepsis

Pearson’s

p-value

24 hours

0.974

<0.001

0.98

<0.001

0.657

0.87

<0.001

0.87

<0.001

SD

Spearman’s correlation

AC

CE

PT

ED

MA

NU

MGL mean glucose level, SD standard deviation

0.630

SC

correlation

p-value

72 hours

p-value

0.063

0.814

0.010

0.077

0.661

0.062

T

MGL

p-value 72 hours

RI P

24 hours

Control

AC

CE

PT

ED

MA

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Hilight

SC

NU MA ED PT CE

-

To evaluate glucose variability from continuous glucose monitoring in sepsis without DM. Non- diabetic sepsis patients had higher glucose variability than control group. Glucose variability during first 24 hours associated with severity of sepsis.

AC

-

RI P

T

“Hyperglycemia and glycemic variability associated with severity of sepsis in non-diabetic subjects”