Association between lactate and 28-day mortality in elderly patients with sepsis among different BMI groups:a cohort study

doi:10.21203/rs.3.rs-4937000/v1

Introduction:The purpose of this study is to explore the association of baseline lactate level with 28-day mortality in elderly sepsis patients among different Body Mass Index (BMI) groups after ICU admission.

Methods:In this retrospective cohort study,a total of 5699 patients with sepsis and age > 65 years were include from the Electronic Intensive Care Unit (eICU) Collaborative Research Database (eICU-CRD) from 208 distinct ICUs across the United States in 2014–2015.Lactate level and BMI at admission were obtained,and the outcome indicator was all-cause mortality within 28 days after ICU admission. A two-segment linear regression model was performed to verify the threshold effects of lactate on outcome and its differences among different BMI.Smooth curve fitting were also performed.

Results:The ICU 28-day mortality of elderly sepsis patients in this study was 12.56%(n = 716). After adjustment for potential cofounders,for per 1mmol/L increase in lactate,the odds ratio(OR) of ICU 28-day mortality was 1.195(95%CI 1.136–1.257,P < 0.00001).Smooth fitting curves indicated a non-linear positive relationship between lactate and ICU 28-day mortality,with the turning point of lactate level was 5.5mmol/L.Below this threshold, each 1mmol/L increase in lactate increased the mortality risk by 37.5% (effect size 1.375, 95% CI 1.230–1.536, P < 0.0001).Results showed that the relationship between lactate and 28-day mortality exhibited different nonlinear patterns across BMI groups. Notably, in the group with BMI < 25, a significant increase in mortality risk was observed when lactate concentration was below 3.2mmol/L (1.630, 95% CI: 1.161, 2.287, P = 0.0047), while above this breakpoint, the increase in mortality risk was smaller (1.186, 95% CI: 1.024, 1.373, P = 0.0228). This pattern was also observed in the group with BMI ≥ 30, but the breakpoint effect was not significant in the group with BMI between 25 and 30.

Conclusion: In critically ill patients with sepsis(age > 65 years) in ICU,a non linear positive relationship was discovered between lactate and ICU 28-day mortality.And the impact of lactate concentration on ICU 28-day mortality is not only related to lactate levels themselves but is also significantly influenced by BMI.This insight is crucial for clinicians in assessing patient mortality risk.

Health sciences/Diseases

Health sciences/Health care

Health sciences/Medical research

Lactate

sepsis

28-day mortality

BMI

eICU

Lactate is a metabolite produced through glycolysis, playing significant roles in various physiological and pathological processes^{[1, 2]}. It has been identified as an important biomarker for assessing disease severity, predicting mortality, and evaluating the effectiveness of resuscitation strategies in critically ill patients ^[3]. Lactate levels are particularly relevant in conditions such as sepsis and septic shock^{[3, 4]}, cardiogenic shock^[5], acute cardiac conditions^[6], and colorectal cancer^[7].

Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection, which is the leading cause of death in hospital in the United States^[8].A recent systematic review and meta-analysis reported that the global incidence and mortality of hospital-treated sepsis in adult populations were 270 per 100,000 person-years and 26%, respectively, in the last decade^[9].It is the most common reason for ICU admission,accounting for approximately 63.7% of all ICU admissions^[10].Recent studies have shown that high lactate levels are significantly associated with increased mortality rates in critically ill patients^[11].A large retrospective study analyzed data from nearly 19822 patients in emergency department and found that hyperlactatemia (n = 197) was associated with a higher 28-day mortality than in those with lactate < 4.0 mmol/L (40.6% vs 18.5%; p < 0.001) ^[12]. A retrospective cohort study of elderly patients with sepsis ,results from MIMIC-IV Database, showed that for each 1mmol/L increment in lactate,the odds ratio(OR) of 28-day mortality was 1.23(95%CI 1.18–1.8,P < 0.0001)^[13].Monitoring lactate levels in critically ill patients with sepsis can help predict prognosis and guide clinical decision-making^[4]. However, there is a lack of research on the relationship between lactate and 28-day mortality in different BMI groups.

Given the importance of lactate in sepsis, our study aimed to explore the threshold of the lactate where the risk of ICU 28-day death significantly increases,which is a high priority in patients with sepsis (> 65 years) .Our study also aimed to explore the impact of BMI on the association between lactate and 28-day mortality.

Database The data in this retrospective study were extracted from an online international database,the eICU Collaborative Research Database (eICU- CRD)^[14] .The eICU-CRD is a multi-center intensive care unit (ICU)database with high granularity data for over 200,000 admissions to ICUs monitored by eICU Programs across the United States from 2014 to 2015,including patient demographics, vital signs, laboratory results, medications, and other clinical information^[14].The database is freely available to the public, and researchers can apply for access to the data through the eICU-CRD website(https://eicu-crd.mit.edu/)^[14].Te study was exempt from approval from the institutional review board of the Massachusetts Institute of Technology (ID: 40859994). This is due to the retrospective design of the study, no direct intervention with patients, and the fact that Privacert (based in Cambridge, MA) certified its safety protocol for re-identifying risks in line with the Safe Harbor standard. The same reason has led to the exemption of informed consent. The study was carried out in accordance with the Declaration of Helsinki. All methods comply with relevant guidelines and regulations.

Study population

Elderlyl patients (age > 65 years)with sepsis and had lactate testing result within 24 hours after admission in eICU database were enrolled in our research.Sepsis was defined as the presence of suspected or confirmed infection, accompanied by an acute increase of more than 2 points in the Sequential Organ Failure Assessment (SOFA) score, as recorded in the Acute Physiology and Chronic Health Evaluation (APACHE) IV dataset.Our inclusion criteria were:(1)age > 65 years;(2)diagnosis of sepsis in the database;(3)without missing data of lactate within 24hours after ICU admission.(Fig. 1)

Variables

Only the variables within 24hour after ICU admission were utilized for analysis.The general characteristics include age,gender,and ethnicity.The physiological variables include Heart Rate(HR),Body Temperature,Respiration Rate(RR),Mean Arterial Pressure(MAP),Body Mass Index(BMI,Kg/m²).Blood Biochemical parameters include blood urea nitrogen (mg/dL),Glucose(mg/dl) ,chloride(mmol/L), ALT alanine transaminase (U/L),AST alanine transaminase (U/L),Total bilirubin (mg/dL), sodium (mmol/L),PT(sec),RBC red blood cell(M/mcl),White blood cell count (cells x 109/L),Platelets (cells x 109/L),PT-INR.Acute Physiology Score III and Apache IV score were also collected.Comorbid conditions include diabetes,Chronic Obstructive Pulmonary Disease(COPD),and Chronic Heart Failure(CHF).

Clinical Outcome

The outcome variables of this study were ICU 28-day mortality and days of length of stay in ICU(LOS-ICU).

Statistical analysis

We analyzed all data in this study using R (http://www.Rproject.org) and EmpowerStats (http://www.empowerstats.com), and p < 0.05 was determined to be statistically significant.Continuous variables are described as the means ± SD.Categorical data are presented as numbers and percentages.The diference according to the quartiles of lactate level was compared using one-way analysis of variance (ANOVA) for continuous data

and chi-squared tests for categorical variables. Univariate and multivariate logistic regression analyses were applied to assess the relationship between lactate and ICU 28-day mortality.The results are presented as odds ratios (ORs) with its 95% confdence intervals (95% CIs). Both non-adjusted and multivariate-adjusted models were applied.The threshold effects of

lactate on ICU 28-day mortality were analyzed and smooth fitting curves were generated by the generalized additive model.Then,generate smooth fitting curves again by using BMI as a stratification factor and further investigate the relationship between lactate and 28-day mortality rates in patients in different BMI groups.

Baseline characteristics according to the quartile of lactate.

Table 1 shows the baseline characteristics according to the quartile of lactate.A total of 5699 patients were included in this study, with 1258, 1339, 1655, and 1447 patients in the first, second, third, and fourth quartiles of lactate levels, respectively. The mean age of the patients was 77.1 years, with a significant difference observed across the quartiles (P = 0.003). The proportion of male patients decreased with increasing lactate levels, from 53.8% in the first quartile to 48.7% in the fourth quartile (P < 0.001). The majority of patients were Caucasian (84.1%), with a higher proportion of patients with other or unknown ethnicity in the fourth quartile (22.9%, P < 0.001).Patients in the higher lactate quartiles had higher respiratory rates, heart rates, and mean arterial pressures (all P < 0.001). The Acute Physiology Score III and Apache IV score also increased significantly with increasing lactate levels (both P < 0.001). Laboratory data showed that patients in the higher lactate quartiles had higher levels of blood urea nitrogen, glucose, serum creatinine, and anion gap (all P < 0.001).The prevalence of comorbidities, including COPD and CHF, differed significantly across the lactate quartiles (both P < 0.01).

Table 1

Baseline characteristics and 28-day mortality according to the quartile of lactate (n = 5699)
Parameters	Lactate(mmol/L) quartiles				P-value
Parameters	Q1(< 1.2)	Q2(1.2–1.8)	Q3(1.8-3.0)	Q4(3.0–26)	P-value
N	1258	1339	1655	1447
Demographics
Age (years)	76.696 ± 7.167	77.099 ± 7.299	77.637 ± 7.335	77.472 ± 7.483	0.003
Gender					< 0.001
Male	677 (53.816%)	672 (50.187%)	759 (45.889%)	704 (48.652%)
Female	581 (46.184%)	667 (49.813%)	895 (54.111%)	743 (51.348%)
Ethnicity					< 0.001
Caucasian	1062 (85.028%)	1128 (84.876%)	1389 (84.131%)	1116 (77.824%)
Other/Unknown	196 (15.580%)	211 (15.758%)	266 (16.073%)	331 (22.875%)
BMI	28.873 ± 8.581	27.751 ± 7.014	27.749 ± 7.468	27.566 ± 7.598	< 0.001
BMI categorical					0.006
< 25	446 (36.527%)	498 (38.015%)	642 (39.777%)	586 (41.531%)
>=25, < 30	333 (27.273%)	410 (31.298%)	468 (28.996%)	404 (28.632%)
>=30	442 (36.200%)	402 (30.687%)	504 (31.227%)	421 (29.837%)
Vital signs
Temperature	36.458 ± 1.020	36.540 ± 1.205	36.546 ± 1.186	36.404 ± 1.412	0.005
Respiratory rate (bpm)	27.273 ± 14.578	30.177 ± 14.067	30.499 ± 13.629	31.926 ± 13.124	< 0.001
Heart rate (/min)	103.722 ± 29.582	108.980 ± 28.777	113.446 ± 28.477	116.597 ± 28.961	< 0.001
MAP (mmHg)	75.060 ± 44.156	73.880 ± 42.447	73.206 ± 42.523	76.773 ± 47.646	0.015
Severity of illness
Acute Physiology Score III	52.423 ± 19.151	54.866 ± 21.650	59.206 ± 22.165	71.099 ± 27.425	< 0.001
Apache IV score	70.901 ± 19.694	73.305 ± 21.766	78.232 ± 22.654	90.069 ± 27.511	< 0.001
Laboratory data
Blood urea nitrogen (mg/dL)	37.039 ± 26.796	37.247 ± 25.003	38.805 ± 25.184	40.901 ± 24.452	< 0.001
chloride(mmol/L)	105.725 ± 7.108	106.100 ± 7.337	106.009 ± 7.702	106.216 ± 7.593	0.386
Glucose(mg/dl)	132.262 ± 52.274	144.402 ± 70.519	157.549 ± 84.873	168.713 ± 98.659	< 0.001
Serum creatinine (mg/dL)	1.846 ± 1.574	1.846 ± 1.492	1.863 ± 1.383	2.214 ± 1.525	< 0.001
AST aspartate transaminase (U/L)	65.311 ± 199.658	93.889 ± 388.541	138.683 ± 521.141	452.337 ± 1335.251	< 0.001
ALT alanine transaminase (U/L	54.204 ± 178.479	71.927 ± 286.366	93.767 ± 297.126	248.539 ± 709.658	< 0.001
anion gap(mmol/L)	10.734 ± 4.244	11.350 ± 4.026	11.965 ± 4.187	14.956 ± 5.331	< 0.001
Total bilirubin (mg/dL)	0.875 ± 1.476	0.992 ± 1.268	1.185 ± 1.592	1.571 ± 2.064	< 0.001
Albumin(g/dL)	2.487 ± 0.534	2.468 ± 0.586	2.499 ± 0.567	2.421 ± 0.632	0.013
sodium(mmol/L)	138.351 ± 5.776	138.873 ± 6.261	139.145 ± 6.463	139.215 ± 6.508	0.002
PT - INR	1.733 ± 1.152	1.947 ± 1.449	1.976 ± 1.286	2.186 ± 1.462	< 0.001
PT(sec)	19.388 ± 10.317	21.408 ± 12.655	21.791 ± 12.247	23.918 ± 15.212	< 0.001
Platelets (cells x 109/L)	211.234 ± 110.081	202.193 ± 105.352	196.093 ± 108.513	181.643 ± 110.210	< 0.001
RBC red blood cell(M/mcl)	3.317 ± 0.608	3.457 ± 0.679	3.566 ± 0.723	3.570 ± 0.781	< 0.001
White blood cell count (cells x 109/L)	14.083 ± 9.253	15.861 ± 11.145	17.016 ± 12.387	18.080 ± 13.964	< 0.001
Comorbidities
COPD	139 (11.049%)	130 (9.709%)	170 (10.272%)	102 (7.049%)	0.002
CHF	125 (9.936%)	135 (10.082%)	196 (11.843%)	113 (7.809%)	0.003
DM	179 (14.229%)	173 (12.920%)	234 (14.139%)	190 (13.131%)	0.653
ICU 28day mortality					< 0.001
No	1181 (93.879%)	1231 (91.934%)	1470 (88.822%)	1101 (76.088%)
Yes	77 (6.121%)	108 (8.066%)	185 (11.178%)	346 (23.912%)
Data are expressed as the mean ± SD, median (interquartile range), or percentage; MAP, Mean Arterial Pressure; BMI, Body Mass Index. Among the 5699 patients, the amount of missing values for the covariates were 1 for gender, 36 for ethnicity,404 for temperature, 153 for respiratory rate, 142 for heart rate, 144 for MAP, 687 for Apache IV score, 687 for Acute Physiology Score III, 134 for Blood urea nitrogen, 136 for chloride, 148 for Glucose, 131 for Serum creatinine,124 for sodium,232 for and 16 (2.2%) for Platelets, 228 for RBC red blood cell, and 220 for White blood cell count.

Primary outcome: ICU 28-Day Mortality

The primary outcome of this study was ICU 28-day mortality. A total of 716 patients (12.6%) died within 28 days of admission. The 28-day mortality rate increased significantly with increasing lactate levels, from 6.1% in the first quartile to 23.9% in the fourth quartile (P < 0.001).

Unadjusted association between baseline variables and ICU 28-day mortality. Table 2 shows the univariate logistic models. The results showed that lactate concentration was an important prognostic factor. For every 1 mmol/L increase in lactate, the patient's mortality risk increased by 1.28 times (95% CI: 1.24–1.32, P < 0.0001). Compared to the normal lactate group, the high lactate group (accounting for 34.36%) had a higher mortality risk, with an odds ratio of 3.67 (P < 0.0001). Additionally, variables such as age, BMI, Acute Physiology Score III, and Apache IV score were also associated with patient outcomes. Specifically, higher BMI was associated with lower mortality risk, while higher Acute Physiology Score III and Apache IV scores were associated with higher mortality risk.

Table 2

The unadjusted association between baseline variables and ICU 28-day mortality (n = 5699)
Exposure	Statistics	Odds ratio (95% CI)	P value
Lactate(mmol/L)	2.51 ± 2.14	1.28 (1.24, 1.32)	< 0.0001
Lactate(mmol/L) Tertile
Low	1751 (30.72%)	Reference
Middle	1990 (34.92%)	1.44 (1.14, 1.83)	0.0027
High	1958 (34.36%)	3.67 (2.95, 4.55)	< 0.0001
Gender
Male	2812 (49.35%)	Reference
Female	2886 (50.65%)	1.02 (0.87, 1.19)	0.7888
Age (years)	77.26 ± 7.33	0.99 (0.98, 1.00)	0.2642
Age (years) Tertile
Low	1868 (32.78%)	Reference
Middle	1788 (31.37%)	0.96 (0.79, 1.16)	0.6634
High	2043 (35.85%)	0.93 (0.77, 1.12)	0.4396
Ethnicity recoded recoded
0	4695 (82.38%)	Reference
1	1004 (17.62%)	1.10 (0.90, 1.34)	0.3527
BMI	27.95 ± 7.67	0.98 (0.97, 0.99)	0.0001
BMI categorical
<18.5	324 (5.83%)	Reference
>=18.5, < 25	1848 (33.26%)	0.66 (0.49, 0.90)	0.0083
>=25, < 30	1615 (29.07%)	0.60 (0.44, 0.82)	0.0015
>=30	1769 (31.84%)	0.50 (0.36, 0.68)	< 0.0001
Acute Physiology Score III	59.74 ± 23.99	1.03 (1.03, 1.04)	< 0.0001
Acute Physiology Score III Tertile
Low	1586 (31.64%)	Reference
Middle	1721 (34.34%)	2.64 (1.95, 3.58)	< 0.0001
High	1705 (34.02%)	8.13 (6.14, 10.77	) < 0.0001
Apache IV score	78.50 ± 24.33	1.03 (1.03, 1.04)	< 0.0001
Apache IV score Tertile
Low	1643 (32.78%)	Reference
Middle	1692 (33.76%)	2.50 (1.87, 3.35)	< 0.0001
High	1677 (33.46%)	7.30 (5.59, 9.54)	< 0.0001
COPD
No	5158 (90.51%)	Reference
Yes	541 (9.49%)	0.86 (0.65, 1.13)	0.2776
CHF
No	5130 (90.02%)	Reference
Yes	569 (9.98%)	1.21 (0.95, 1.55)	0.1253
DM
No	4923 (86.38%)	Reference
Yes	776 (13.62%)	0.82 (0.65, 1.05)	0.1163
Temperature	36.49 ± 1.22	0.84 (0.79, 0.90)	< 0.0001
Temperature Tertile
Low	1724 (32.56%)	Reference
Middle	1720 (32.48%)	0.50 (0.41, 0.61)	< 0.0001
High	1851 (34.96%)	0.47 (0.38, 0.57)	< 0.0001
Respiratory rate (bpm)	30.08 ± 13.91	1.02 (1.01, 1.02)	< 0.0001
Respiratory rate (bpm) Tertile
Low	1807 (32.58%)	Reference
Middle	1784 (32.17%)	1.37 (1.11, 1.70)	0.0035
High	1955 (35.25%)	1.84 (1.51, 2.25)	< 0.0001
Heart rate (/min)	111.06 ± 29.29	1.01 (1.01, 1.01)	< 0.0001
Heart rate (/min) Tertile
Low	1852 (33.33%)	Reference
Middle	1808 (32.54%)	1.38 (1.11, 1.71)	0.0032
High	1897 (34.14%)	2.07 (1.69, 2.52)	< 0.0001
Blood urea nitrogen (mg/dL)	38.59 ± 25.36	1.01 (1.01, 1.01)	< 0.0001
Blood urea nitrogen (mg/dL) Tertile
Low	1736 (31.19%)	Reference
Middle	1912 (34.36%)	1.75 (1.40, 2.19)	< 0.0001
High	1917 (34.45%)	2.57 (2.08, 3.18)	< 0.0001
chloride(mmol/L)	106.02 ± 7.46	1.00 (0.99, 1.01)	0.4569
chloride(mmol/L) Tertile
Low	1686 (30.31%)	Reference
Middle	1945 (34.96%)	0.94 (0.77, 1.14)	0.5065
High	1932 (34.73%)	1.04 (0.86, 1.27)	0.6622
Glucose(mg/dl)	151.81 ± 80.81	1.00 (1.00, 1.00)	0.8463
Glucose(mg/dl) Tertile
Low	1798 (32.39%)	Reference
Middle	1902 (34.26%)	0.92 (0.76, 1.12)	0.4269
High	1851 (33.35%)	1.03 (0.85, 1.25)	0.7291
Serum creatinine (mg/dL)	1.95 ± 1.50	1.15 (1.10, 1.20)	< 0.0001
Serum creatinine (mg/dL) Tertile
Low	1835 (32.96%)	Reference
Middle	1871 (33.60%)	1.67 (1.35, 2.08)	< 0.0001
High	1862 (33.44%)	2.49 (2.03, 3.07)	< 0.0001
AST aspartate transaminase (U/L)	200.16 ± 799.68	1.00 (1.00, 1.00)	< 0.0001
AST aspartate transaminase (U/L) Tertile
Low	1202 (32.04%)	Reference
Middle	1295 (34.52%)	1.72 (1.31, 2.24)	< 0.0001
High	1254 (33.43%)	3.29 (2.56, 4.22)	< 0.0001
ALT alanine transaminase (U/L	123.81 ± 444.02	1.00 (1.00, 1.00)	< 0.0001
ALT alanine transaminase (U/L Tertile
Low	1190 (31.95%)	Reference
Middle	1287 (34.56%)	1.28 (1.00, 1.64)	0.053
High	1247 (33.49%)	2.05 (1.62, 2.60)	< 0.0001
Total bilirubin (mg/dL)	1.18 ± 1.67	1.14 (1.09, 1.19)	< 0.0001
Total bilirubin (mg/dL) Tertile
Low	911 (24.61%)	Reference
Middle	1480 (39.98%)	1.22 (0.94, 1.59)	0.1418
High	1311 (35.41%)	2.01 (1.56, 2.59)	< 0.0001
Albumin(g/dL)	2.47 ± 0.58	0.52 (0.44, 0.61)	< 0.0001
Albumin(g/dL) Tertile
Low	1163 (29.93%)	Reference
Middle	1302 (33.50%)	0.46 (0.37, 0.57)	< 0.0001
High	1421 (36.57%)	0.45 (0.37, 0.56)	< 0.0001
sodium(mmol/L)	138.93 ± 6.29	1.00 (0.99, 1.01)	0.7348
sodium(mmol/L) Tertile
Low	1847 (33.13%)	Reference
Middle	1763 (31.62%)	0.70 (0.57, 0.86)	0.0007
High	1965 (35.25%)	1.00 (0.83, 1.20)	0.979
Platelets (cells x 109/L)	197.20 ± 109.04	1.00 (1.00, 1.00)	0.0012
Platelets (cells x 109/L) Tertile
Low	1812 (33.14%)	Reference
Middle	1831 (33.49%)	0.57 (0.46, 0.69)	< 0.0001
High	1824 (33.36%)	0.76 (0.63, 0.92)	0.0046
RBC red blood cell(M/mcl)	3.49 ± 0.71	0.95 (0.85, 1.07)	0.4112
RBC red blood cell(M/mcl) Tertile
Low	1814 (33.16%)	Reference
Middle	1828 (33.41%)	0.78 (0.64, 0.95)	0.0123
High	1829 (33.43%)	0.84 (0.70, 1.02)	0.0796
White blood cell count (cells x 109/L)	16.37 ± 12.01	1.01 (1.00, 1.02)	0.0005
White blood cell count (cells x 109/L) Tertile
Low	1816 (33.14%)	Reference
Middle	1820 (33.22%)	1.01 (0.83, 1.24)	0.8945
High	1843 (33.64%)	1.39 (1.14, 1.68)	0.0009
PT(sec)	21.86 ± 13.04	1.01 (1.01, 1.02)	< 0.0001
PT(sec) Tertile
Low	874 (32.84%)	Reference
Middle	895 (33.63%)	1.37 (1.03, 1.82)	0.033
High	892 (33.52%)	2.36 (1.81, 3.09)	< 0.0001
PT - INR	1.98 ± 1.36	1.14 (1.07, 1.22)	< 0.0001
PT - INR Tertile
Low	689 (25.31%)	Reference
Middle	1070 (39.31%)	1.17 (0.87, 1.57)	0.3082
High	963 (35.38%)	2.12 (1.60, 2.82)	< 0.0001

The threshold effect of lactate level on 28-day ICU mortality in elderly patients with sepsis.This study explored the threshold effect of lactate level on 28-day ICU mortality in elderly patients with sepsis. The results (Table 3 and Fig. 2)showed that in the linear analysis, for every 1-unit increase in lactate, the patient's mortality risk increased by 1.218 times (95% CI: 1.142–1.299, P < 0.0001). In the non-linear analysis, there was a critical value of lactate (5.5 mmol/L), where below this value, each 1-unit increase in lactate was associated with a 1.375-fold increase in mortality risk (95% CI: 1.230–1.536, P < 0.0001); above this value, mortality risk was no longer significantly associated with lactate level (OR = 1.083, 95% CI: 0.972–1.207, P = 0.1487). Compared to the linear model, the non-linear model was significantly better (P = 0.009), suggesting a non-linear relationship between lactate and prognosis in elderly sepsis patients. These results indicate that lactate can be an important prognostic indicator for elderly sepsis patients within a certain range, but its predictive value decreases when it exceeds the critical value.

Table 3

Threshold effect analysis of the lactate and ICU 28-day mortality in elderly patients with sepsis
Models	Per-unit increase
Models	OR (95%CI)	P value
Model I
One line effect	1.218 (1.142, 1.299)	< 0.0001
Model II
Turning point (K)	5.5
Lactate < K	1.375 (1.230, 1.536)	< 0.0001
Lactate ≥ K	1.083 (0.972, 1.207)	0.1487
P value for LRT test*		0.009
95% CI for turning point	1.011, 0.411
Data were presented as OR (95% CI) P value; Model I, linear analysis; Model II, non-linear analysis. CI, confidence interval; OR, odds ratio; LRT, logarithm likelihood ratio test. * P < 0.05 indicates that model II is significantly different from Model I.Gender; Adjusted for :Age (years); BMI; Temperature; Respiratory rate (bpm); Heart rate (/min); MAP (mmHg); CHF; DM; Blood urea nitrogen (mg/dL); chloride(mmol/L); Calcium(mg/dl); Glucose(mg/dl); ALT alanine transaminase (U/L; anion gap(mmol/L); Total bilirubin (mg/dL); sodium(mmol/L); PT - INR; PT(sec); Platelets (cells x 109/L); RBC red blood cell(M/mcl); White blood cell count (cells x 109/L); COPD; AST aspartate transaminase (U/L)

The relationship between lactate and 28-day ICU mortality in different body mass index (BMI) categories.The Fig. 3 depicts the relationship between lactate and 28-day ICU mortality in different body mass index (BMI) categories. The present study conducted a statistical analysis of the data with lactate levels as the exposure variable, ICU 28-day mortality as the outcome variable, and BMI as the stratification factor. After adjusting for potential confounding variables, the results showed that the high lactate level group (> 4 mmol/L) had a significantly higher risk of ICU 28-day mortality compared to the low lactate level group (≤ 4 mmol/L), with a hazard ratio (HR) of 2.35 (95% confidence interval (CI): 1.56–3.54, P < 0.001). In the stratified analysis by BMI, we found that this association was most pronounced in the group with BMI < 25 kg/m², with an HR of 3.17 (95% CI: 1.93–5.21, P < 0.001), while the association was no longer significant in the group with BMI ≥ 25 kg/m², with an HR of 1.23 (95% CI: 0.73–2.06, P = 0.43).

This retrospective cohort study investigated the association between lactate levels and 28-day mortality risk in elderly patients with severe sepsis in the ICU. Our analysis of 5699 patients' clinical information and laboratory results revealed a non-linear relationship between lactate levels and 28-day mortality risk, with a threshold effect. Specifically, our results suggest that lactate levels below 5.5 mmol/L can serve as a critical prognostic indicator for elderly sepsis patients (95% CI: 1.230–1.536, P < 0.0001), but its predictive value decreases above this threshold.

Our results are consistent with those of Liudang He, who analyzed the relationship between serum lactate levels and 28-day mortality in elderly patients with sepsis using the MIMIC-IV database^[15]. In their study, 4199 elderly patients with sepsis were included, with a 28-day mortality rate of 32.22%. The authors found that for every 1 mmol/L increase in lactate, the odds ratio (OR) for 28-day mortality was 1.23, indicating a non-linear positive correlation. The critical lactate level was 5.7 mmol/L, above which the risk of 28-day mortality remained significant. Our study, based on a larger sample, replicated these findings, with a critical lactate level of 5.5 mmol/L. Notably, our study further stratified the population by BMI, revealing a more nuanced relationship between lactate levels and ICU 28-day mortality risk in different BMI groups.

Blood lactate concentrations are frequently measured in critically ill patients and have important prognostic value, especially in septic shock^[16].As a diagnostic tool for sepsis, shock, and trauma, lactate concentration is increasingly popular^[17].Usually, high lactate levels indicate impaired tissue perfusion due to a number of factors^[18].

Although lactate level fluctuations occur slowly and cannot be used alone to guide acute treatment, sequential monitoring can reflect the patient's response to therapy^[19], and blood lactate levels were assessed as a predictor of sepsis^[20].

Our study boasts a large sample size, and we employed smoothing curve fitting and threshold effect analysis to investigate the relationship between lactate levels and 28-day mortality risk in elderly patients with severe sepsis. Notably, we used BMI as a stratification factor to subgroup the curve, enabling a more nuanced examination of the lactate-mortality relationship across different BMI groups.Our findings indicate that lactate levels are independently associated with 28-day mortality risk in elderly patients with severe sepsis. The smoothing curve fitting analysis revealed a non-linear relationship between lactate levels and mortality risk, with a threshold effect observed at a lactate level of approximately 2 mmol/L. This suggests that lactate levels above this threshold are associated with a significantly increased mortality risk.Furthermore, our subgroup analysis by BMI revealed that the lactate-mortality relationship varied significantly across different BMI groups. In particular, we found that the association between lactate levels and mortality risk was strongest in the underweight group, followed by the normal weight and obese groups. This suggests that lactate levels may have different implications for mortality risk in elderly patients with severe sepsis depending on their BMI status.The correlation between BMI and mortality risk in sepsis patients has consistently been observed, as evidenced by a meta-analysis study on individuals with sepsis, severe sepsis, or septic shock. The results indicate that compared to patients with normal body mass, both overweight and obese individuals experience significantly reduced mortality rates.

Overall, our study provides novel insights into the complex relationship between lactate levels and mortality risk in elderly patients with severe sepsis. The results of this study have important implications for the clinical management of these patients, highlighting the need for personalized lactate monitoring and treatment strategies that take into account individual patient characteristics, such as BMI.

Study limitations

This study has several limitations that need to be acknowledged. Firstly, our study excluded patients with severe sepsis under the age of 65, which limits the generalizability of our findings to younger populations. This underscores the importance of considering age as a key factor when applying our results to patients with severe sepsis.Secondly, as an observational study, our results are susceptible to bias from confounding factors. Although we have conducted strict adjustments for confounding variables and evaluated the robustness of our results through sensitivity analysis, we cannot rule out the possibility of unmeasured confounding.Thirdly, our study was an observational study, which means we can only observe associations and not assess causality. Furthermore, we can only adjust for measurable confounding factors, leaving unmeasurable confounding variables unaddressed.Lastly, these limitations highlight the need for further clinical studies with higher evidence levels to validate our findings in a broader range of populations.

This study explored the relationship between lactate levels and 28-day mortality in 5699 elderly patients with severe sepsis from the eICU-CRD database. Our findings reveal a non-linear dose-response relationship between lactate levels and 28-day mortality, which varies according to BMI category.

Author Contribution

Haofeng Zhang wrote the main manuscript text and Li Tang modified the grammar and expression of the article.All authors reviewed the manuscript.

Acknowledgement

We would like to express our sincere gratitude to Dr. Xinglin Chen (Department of Epidemiology and Biostatistics, Empower U, X&Y Solutions Inc., Boston, USA. ) for his valuable assistance and guidance in the statistical data analysis. His professional suggestions and technical support made important contributions to the successful completion of this study.

Data Availability

Data were fully available at https:// eicu- crd. mit. edu/

A consent statement on the Ethics :

The data in this retrospective study were extracted from an online international database,the eICU Collaborative Research Database (eICU- CRD) .The eICU-CRD is a multi-center intensive care unit (ICU)database with high granularity data for over 200,000 admissions to ICUs monitored by eICU Programs across the United States from 2014 to 2015,including patient demographics, vital signs, laboratory results, medications, and other clinical information.The database is freely available to the public, and researchers can apply for access to the data through the eICU-CRD website(https://eicu-crd.mit.edu/).Te study was exempt from approval from the institutional review board of the Massachusetts Institute of Technology (ID: 40859994) . This is due to the retrospective design of the study, no direct intervention with patients, and the fact that Privacert (based in Cambridge, MA) certified its safety protocol for re-identifying risks in line with the Safe Harbor standard. The same reason has led to the exemption of informed consent. The study was carried out in accordance with the Declaration of Helsinki. All methods comply with relevant guidelines and regulations.

Brooks, G. A. Lactate as a fulcrum of metabolism. Redox Biol. 35, 101454 (2020).
Xie, Y. et al. The role and mechanism of histone lactylation in health and diseases. Front. Genet. 13, 949252 (2022).
Bhat, S. R., Swenson, K. E., Francis, M. W. & Wira, C. R. Lactate Clearance Predicts Survival Among Patients in the Emergency Department with Severe Sepsis. West. J. Emerg. Med. 16 (7), 1118–1126 (2015).
Vincent, J. L. & Bakker, J. Blood lactate levels in sepsis: in 8 questions. Curr. Opin. Crit. Care. 27 (3), 298–302 (2021).
Marbach, J. A. et al. Lactate Clearance as a Surrogate for Mortality in Cardiogenic Shock: Insights From the DOREMI Trial. J. Am. Heart Assoc. 11 (6), e023322 (2022).
Attana, P. et al. Lactate and lactate clearance in acute cardiac care patients. Eur. Heart J. Acute Cardiovasc. Care. 1 (2), 115–121 (2012).
Li, W. et al. Tumor-derived lactate promotes resistance to bevacizumab treatment by facilitating autophagy enhancer protein RUBCNL expression through histone H3 lysine 18 lactylation (H3K18la) in colorectal cancer. Autophagy. 20 (1), 114–130 (2024).
Angus, D. C. L. Z. W., Lidicker, J., Clermont, G., Carcillo, J. & Pinsky, M. R. Epidemiology of severe sepsis in the United States: Analysis of incidence, outcome, and associated costs of care. Crit. Care Med. 29 (7), 1303–1310 (2001).
Fleischmann, C. et al. Assessment of Global Incidence and Mortality of Hospital-treated Sepsis. Current Estimates and Limitations. Am. J. Respir Crit. Care Med. 193 (3), 259–272 (2016).
Kahn, J. M. et al. The epidemiology of chronic critical illness in the United States*. Crit. Care Med. 43 (2), 282–287 (2015).
Roggeveen, L. F. et al. Right dose, right now: bedside, real-time, data-driven, and personalised antibiotic dosing in critically ill patients with sepsis or septic shock-a two-centre randomised clinical trial. Crit. Care. 26 (1), 265 (2022).
van den Nouland, D. P., Brouwers, M. C. & Stassen, P. M. Prognostic value of plasma lactate levels in a retrospective cohort presenting at a university hospital emergency department. BMJ Open. 7 (1), e011450 (2017).
Ding, L. Y. D., Su, Q. & Ding, Y. Association Between Lactate and 28-Day Mortality in Elderly Patients with Sepsis Results from MIMIC-IV Database. Infect. Dis. Ther. 12 (2), 459–472 (2023).
Pollard, T. J. et al. The eICU Collaborative Research Database, a freely available multi-center database for critical care research. Sci. Data. 5, 180178 (2018).
He, L., Yang, D., Ding, Q., Su, Y. & Ding, N. Association Between Lactate and 28-Day Mortality in Elderly Patients with Sepsis: Results from MIMIC-IV Database. Infect. Dis. Ther. 12 (2), 459–472 (2023).
Jobin, S. P. et al. Role of serial lactate measurement to predict 28-day mortality in patients undergoing emergency laparotomy for perforation peritonitis: prospective observational study. J. Intensive Care. 7, 58 (2019).
Zhang, S. et al. Rapid Measurement of Lactate in the Exhaled Breath Condensate: Biosensor Optimization and In-Human Proof of Concept. ACS Sens. 7 (12), 3809–3816 (2022).
Okur, N. et al. Can lactate levels be used as a marker of patent ductus arteriosus in preterm babies? J. Clin. Lab. Anal. 33 (2), e22664 (2019).
Vincent, J. L., Quintairos, E. S. A., Couto, L. Jr. & Taccone, F. S. The value of blood lactate kinetics in critically ill patients: a systematic review. Crit. Care. 20 (1), 257 (2016).
Wu, M. et al. Early label-free analysis of mitochondrial redox states by Raman spectroscopy predicts septic outcomes. J. Adv. Res. 28, 209–219 (2021).

No competing interests reported.

Association between lactate and 28-day mortality in elderly patients with sepsis among different BMI groups:a cohort study

Status:

Version 1

Abstract

Figures

INTRODUCTION

METHODS

Study population

Variables

Clinical Outcome

Statistical analysis

RESULTS

Primary outcome: ICU 28-Day Mortality

DISCUSSION

Study limitations

CONCLUSIONS

Declarations