Sequential L-lactate concentration in hospitalised equine neonates: A prospective multicentre study


  • Portions of these data were presented in abstract at the 2010 American College of Veterinary Internal Medicine Forum and the 2010 International Veterinary Emergency and Critical Care Symposium.


Reasons for performing study

Evaluation of serial blood lactate concentrations [LAC] are of prognostic value for morbidity and mortality in critically ill human patients and neonatal foals, but have not been prospectively evaluated in a large multicentre study of critically ill neonatal foals.


To prospectively evaluate the prognostic value of sequential [LAC] analysis in critically ill neonatal foals with risk of mortality.

Study design

Prospective, observational study.


Thirteen university and private equine referral hospitals enrolled 643 foals over the 2008 foaling season and [LAC] was measured at admission ([LAC]ADMIT) and 24 ([LAC]24), 48 ([LAC]48), 72 ([LAC]72), 96 ([LAC]96) and 120 h ([LAC]120) after admission. [LAC] changes over time ([LAC]Δ) were calculated between sampling points.


Nonsurvivors had significantly greater [LAC]ADMIT, [LAC]24 and [LAC]48 compared with surviving foals (P<0.001). In nonsurviving foals [LAC]Δ did not decrease over time while survivors showed significant positive [LAC]Δ between [LAC]ADM–24 and all other time periods (P<0.001). Logistic regression analysis showed that the odds of survival decreased for each 1 mmol/l [LAC] increase at all time points for all critically ill foals, independent of major final diagnoses as potential confounders. Septic foals had significantly greater [LAC] at all time points compared with nonseptic foals (P<0.001) and [LAC]Δ in septic foals was significantly more positive (suggesting better clearance of lactate from the blood) only at [LAC]ADM–24 and [LAC]72–96 (P<0.01), while in nonseptic foals [LAC]Δ was significantly positive between [LAC]ADM–24 compared with all other time periods (P<0.001).


Blood lactate concentration is a strong, independent biomarker used to predict mortality in critically ill foals. Lactate metabolism is impaired in nonsurviving and septic foals and [LAC]Δ can be utilised to identify patients at high risk for mortality.


Lactate is an established diagnostic, therapeutic and prognostic marker of global tissue hypoxia, inadequate oxygen utilisation, sepsis, disease severity and mortality in critically ill human patients [1-9] as well as in adult and neonatal equine emergencies [10-15] and in critically ill dogs [16-19].

Single, one time measurements of lactate concentration ([LAC]) have been associated with disease severity and mortality [8, 10, 20, 21] but may be inaccurate if evaluated without considering the course of the primary disease [3, 6]. Blood lactate concentrations reflect both production of lactate and its elimination/metabolism, and increases in blood lactate concentrations are classified as type A (due to inadequate oxygen delivery or increased oxygen demand) or type B (due to inadequate oxygen utilisation, congenital errors of metabolism, drugs/toxin or other causes) [6, 22]. Serial [LAC] measurements may be more reliable for monitoring response to therapy and predicting outcome than single measurements [3, 5, 13, 14, 23, 24]. The rate of [LAC] decrease has been associated with improved outcome (survival) in septic human patients and in critically ill foals and horses, supporting the concept that serial lactate measurements are more reliable predictors of survival [3, 5, 13, 14, 23].

Numerous retrospective and prospective studies have evaluated the role of [LAC] in critically ill neonatal foals and have consistently shown increased [LAC] in addition to either apparently impaired L-lactate removal or increased production, or both, in nonsurvivors [10, 12, 13, 15]. Similar results were obtained in a prospective serial evaluation of [LAC] in adult equine emergencies [14].

Serial [LAC] concentrations have not been evaluated prospectively as a prognostic indicator in critically ill equine neonates. Thus, the objective of this study was to investigate serial [LAC] and [LAC]Δ over a 120 h time period of hospitalisation using a prospective multicentre approach. We hypothesise that persistence of hyperlactataemia, or a slow decrease in [LAC] during hospitalisation, is associated with morbidity and mortality.

Materials and methods

This is the second part of a previously published worldwide, prospective multicentre study for which 13 university and private equine referral hospitals were recruited [25]. Neonatal foals aged 0–35 days were included during the 2008–2009 foaling season. Sequential samples to determine [LAC] were collected at admission ([LAC]ADMIT), and 24 h ([LAC]24), 48 h ([LAC]48), 72 h ([LAC]72), 96 h ([LAC]96) and 120 h ([LAC]120) after admission. Only routinely collected [LAC] data were reported. Not all samples were available at all time periods because [LAC] were not available at all time points and because of foals leaving the study for reasons of health status (died, discharge due to improved status, etc).

Collection and analysis of demographic data were previously reported in the first part of this prospective, multicentre study [25]. In this second part of the study, the 18 major final diagnoses were stratified into 8 groups for logistic regression analysis: Group 1: Sepsis/infectious/inflammatory, Group 2: Perinatal asphyxia syndrome, premature/dysmature, birth-related disease, Group 3: Respiratory, Group 4: Urogenital, Group 5: musculoskeletal, Group 6: gastrointestinal, Group 7: immune-related, Group 8: other. Survival was defined as discharge from the hospital. The nonsurvival group included foals that died or were subjected to euthanasia due to clinician-determined poor prognosis or financial reasons. Foals were assigned to the ‘septic group if blood culture results were positive and/or a calculated sepsis score was ≥14.

Arterial or venous blood samples were collected on admission using sampling techniques specific to each [LAC] analyser and analysed immediately for [LAC] with a blood gas/lactate analyser (Nova Biomedicala, Accu-check Lactate Monitorb, Accutrend Lactate Monitorb, Accu-check Aviva 400c, Lactate Scoutd, ABL 700 radiometere, I-Statf). No attempt was made to ensure that all sites used similar sampling techniques or similar lactate analysers. Similarly, mean arterial pressure (MAP) was determined by direct or indirect pressure measurement at some sites with no effort made to control techniques used.

display math

Lactate change [LAC]Δ between sequential sampling periods was calculated as follows:

display math

Where [LAC]a represents plasma lactate at any sampling period and [LAC]b represents plasma lactate of the sampling period following [LAC]a, divided by [LAC]a. A positive value indicates a decrease in lactate, while a negative value indicates an increase in [LAC] between the sampling periods. The [LAC]Δ was calculated for the following sampling intervals: admission to 24 h, [LAC]ADM–24, 24–48 h [LAC]24–48, 48–72 h [LAC]48–72,72–96 h [LAC]72–96 and 96–120 h [LAC]96–120.

Plasma [LAC] collected at admission [LAC]ADMIT, 24 h [LAC]24 and 48 h [LAC]48 were also stratified into groups low (<2.5 mmol/l), intermediate (2.5–3.9 mmol/l) or high (≥ 4 mmol/l).

Data analysis

Descriptive statistical analyses, including mean, standard deviation (s.d.), median and range, were used to summarise the data. All analyses were performed using commercially available statistics software (Stata 12.1g, GraphPadh).

Data were tested for normality using the Kolmogorov and Smirnov method. A nonparametric approach using Mann–Whitney was used to compare variables between 2 groups and Kruskal–Wallis, a nonparametric ANOVA test, was used to examine differences among [LAC] at different time points (survivor/nonsurvivor and septic/nonseptic foals), followed by Dunn's multiple comparison tests. Logistic regression analysis was used to examine the odds ratios (OR) for outcome based on unit changes in [LAC] in 1.0 mmol/l taking into account major final diagnoses as confounder and also for [LAC]Δ for each calculated unit change between time points. Receiver operator curve (ROC) analysis and sensitivity and specificity analysis were performed to determine a cut-point above which nonsurvival could be most accurately predicted using approaches suggested by Hilbe [26]. To ease the assumption of independence of multiple observations from centres in our analysis, logistic regression employing clustering on centre in the fashion by Collett [27] was used. Significance was determined at P≤0.05.


There were 643 foals enrolled in the study. Demographic data including major final diagnoses has been previously published [25].

Overall survival rate was 79% (505/643). Nonsurvivors (n = 129) included foals subjected to euthanasia for clinician-determined poor prognosis (n = 81) or financial reasons (n = 6) and foals that died (n = 48). None were subjected to euthanasia at admission without attempting stabilisation and establishment of a diagnosis and severity of condition that carried a poor prognosis for life. These foals were not, however, included in ROC analysis. Foals without [LAC]ADMIT were not included in relative risk, logistic regression or ROC analyses.

There was no difference between sites regarding survival percentage. Mean age of the foals at admission was 53 ± 16 h (range 0–840 h). Median duration of hospitalisation for survivors and nonsurvivors were 6 days (0.25–60 days) and 2 days (0–23 days), respectively, P<0.001.

Three hundred and twenty-nine foals (51%) had a negative sepsis score (<14) and were classified as nonseptic, of which 291 (89%) foals survived and 37 (11%) did not survive. One hundred and ninety-eight foals (31%) had a positive sepsis score (≥14) and/or positive blood cultures and were classified as septic, of which 132 (67%) foals survived and 66 (33%) did not survive. Mean arterial pressures for nonseptic and septic foals on presentation were 71.5 mmHg (range 43–101 mmHg, n = 74) and 67 mmHg (range 35–97 mmHg, n = 85), respectively and were not significantly different (P = 0.1201).

Sequential [LAC] analysis and outcome

Within-group analysis indicated that [LAC]ADMIT in nonsurvivors was significantly increased compared with other time points [LAC]48, [LAC]96 and [LAC]120 (P<0.001) (Table 1). Nonsurvivor [LAC]24 was also significantly greater compared with [LAC]72 and [LAC]120 (P = 0.010) (Table 1). Surviving foals had significantly greater [LAC]ADMIT compared with [LAC]24 – [LAC]120 and [LAC]24 compared with [LAC]48 – [LAC]120.

Table 1. Median lactate concentration ([LAC]) (mmol/l) by sample period, outcome (survivor/nonsurvivor) and septic/nonseptic
 Admission24 h48 h72 h96 h120 h
  1. a–b, c–dValues with different letters in a row are significantly different (P<0.05). *Values are significantly different between survivor/nonsurvivor (P<0.05). +Values are significantly different between septic/nonseptic foals (P<0.05).
Alive3.6 (0.3–36.6)*,a2.1 (0.5–14)*,b,c1.6 (0.4–10.4)*,b,d1.3 (0.3–7.7)b,d1.4 (0.5–5.7)b,d1.5 (0.4–4.9)b,d
Dead5.5 (0.9–31.7)*,a4.8 (0.6–16.4)*,c3.7 (0.7–9.8)*,b2.1 (0.4–7.3)d2.5 (0.3–9.3)b1.8 (0.3–9)b,d
Septic4.8 (0.6–36.6)+,a2.9 (0.6–16.4)+,b1.9 (0.5–8.5)+,b1.5 (0.5–7.3)+,b2.2 (0.7–9.3)+,b1.6 (0.4–9)+,b
Nonseptic3.3 (0.3–21)+,a2.1 (0.5–14)+,b1.5 (0.4–11)+,b1.3 (0.3–7.7)+,b1.4 (0.5–6.1)+,b1.3 (0.6–5.1)+,b

Median [LAC] was significantly increased in nonsurvivors compared with survivors at admission, 24 and 48 h after presentation (P<0.001) (Table 1).

Within-group analysis, when specifically evaluating septic and nonseptic foals, indicated that both groups had significantly greater [LAC]ADMIT when compared with all other time points ([LAC]24-−[LAC]120) (P<0.001) (Table 1).

Median [LAC] of septic foals was also significantly greater than nonseptic foals at all time points (P<0.001) (Table 1).

The OR for survival decreased (making survival less likely) significantly for every 1 mmol/l increase in [LAC] at all time points [LAC]ADMIT – [LAC]120. When taking into account grouped major final diagnoses as potential confounders for [LAC] as determinant of outcome, the OR remained significant at [LAC]ADM, [LAC]24, [LAC]48, [LAC]96 and [LAC]120 (Table 2).

Table 2. Logistic regression analysis (survival) and lactate concentration ([LAC]) at each time point without and with considering major final diagnoses as confounders for lactate as determinant of outcome
 [LAC] Time OR95% CIP*
  1. *P<0.05. [LAC]ADM = [LAC] at admission; [LAC]24 = [LAC] at 24 h; [LAC]48 = [LAC] at 48 h; etc. [LAC] ADM–120/Diag: Group 1: Sepsis/infectious/inflammatory, Group 2: Perinatal asphyxia syndrome, premature/dysmature, birth-related disease, Group 3: Respiratory, Group 4: Urogenital, Group 5: musculoskeletal, Group 6: gastrointestinal, Group 7: immune-related, Group 8: other.

When evaluating [LAC] groups at admission, using [LAC]ADMIT concentrations <2.5 mmol/l as reference group, the odds of survival for the intermediate [LAC]ADMIT group (2.5–3.9 mmol/l) were 1.07 (95% CI 0.52–2.24, P = 0.85) and for the high [LAC]ADMIT group (≥4 mmol/l) the odds of survival were 0.38 (95% CI 0.21–0.68, P = 0.001). For consecutive time points the ORs for survival are as follows: [LAC]24 intermediate group 0.45 (95% CI 0.23–0.87, P = 0.019), [LAC]24 high group 0.11 (95% CI 0.07–0.18, P = 0.000), [LAC]48 intermediate group 0.19 (95 % CI 0.09–0.4, P = 0.000), [LAC]48 high group 0.09 (95% CI 0.05–0.19, P = 0.000).

Receiver operator curve and sensitivity and specificity analysis revealed that a [LAC]ADMIT cut-point of 4.4 mmol/l with a probability cut-point of 0.82 yielded a sensitivity of 63% and specificity of 63% for prediction of survival. Mortality for patients with [LAC] >4.4 mmol/l was 30 and 13% for patients with [LAC] <4.4. The results for consecutive time points are displayed in Table 3 and Fig 1.

Figure 1.

Receiver operating curves (ROC) and sensitivity/specificity curves at [LAC]ADMIT and [LAC]48.

a) and b) Sensitivity/specificity and ROC curves at [LAC]ADMIT showing that a lactate cut-point of 4.4 mmol/l with a probability cut-point of 0.82 yielded a sensitivity of 63% and specificity of 63% for prediction of survival to discharge. c) and b) Sensitivity/specificity and curves at [LAC]48 showing that a lactate cut-point of 2.3 mmol/l with a probability cut-point of 0.88 yielded a sensitivity of 73% and specificity of 73% for prediction of survival to discharge. [LAC] = lactate concentration; [LAC]ADMIT = [LAC] at admission; [LAC]48 = [LAC] at 48 h.

Table 3. Receiver operating curve (ROC) and sensitivity/specificity analysis with mortality for certain cut-points for lactate concentration ([LAC]) at different time points
[LAC] Time[LAC]ctpSensitivitySpecificityMortality >ctpMortality <ctp
  1. [LAC]ADM = [LAC] at admission; [LAC]24 = [LAC] at 24 h; etc.; ctp = consecutive time points; [LAC]ctp in mmol/l, Sensitivity/Specificity, Mortality >ctp, Mortality <ctp in %, Mortality >ctp or <ctp: Mortality for [LAC] above or below selected cut-point.

When evaluating participating sites for outcome, the odds of survival for each 1 mmol/l increase of [LAC] were similarly decreased for most sites, indicating that [LAC] is an independent, precise, crude predictor of mortality, regardless of study site (Fig 2).

Figure 2.

Forest plot showing the odds ratio and confidence intervals for survival at [LAC]ADMIT by study site. [LAC]ADMIT = lactate concentration at admission.

[LAC]Δ metabolism and outcome

In survivors, [LAC]Δ was significantly more positive, indicating a decrease at [LAC]ADM–24 compared to all other time intervals ([LAC]24–48, [LAC]48–72, [LAC]72–96, [LAC]96–120) (P<0.001), while [LAC]Δ in nonsurvivors did not change significantly at any time point (Table 4). Survivors also had significant more positive [LAC]Δ values and hence [LAC] decreases at [LAC]ADM-24, compared with nonsurvivors at [LAC]24–48, [LAC]48–72, [LAC]72–96 and [LAC]96–120 (P<0.001) (Table 4).

Table 4. Median change in lactate concentration ([LAC]Δ) by sample period, outcome (survivor/nonsurvivor) and septic/nonseptic
 Admission–24 h24–48 h48–72 h72–96 h96–120 h
  1. a–bValues with different letters in a row are significantly different (P<0.05). *Values are significantly different between survivor/nonsurvivor (P<0.05). +Values are significantly different between septic/nonseptic foals (P<0.05).
Alive0.44 (-6.67–0.95)a,*0.21 (-2.8–0.83)b0.18 (-4.32–0.79)b0.09 (-4.69–0.65)b0.08 (-1.75–0.85)b
Dead0.13 (-3.3–0.94)0.14 (-2.86–0.59)*0.28 (-0.67–0.73)*0.2 (-0.79–0.43)*-0.01 (-1.46–0.77)*
Septic0.31 (-6.67–0.95)a0.16 (-2.86–0.72)+0.20 (-1.41–0.79)+0.05 (-4.69–0.65)b+0.09 (-1.75–0.86)+
Nonseptic0.41 (-2.94–0.9)a,+0.21 (-2.8–0.83)b0.16 (-4.31–0.73)b0.16 (-2.27–0.65)b0.01 (-1.46–0.63)b

When evaluating [LAC]Δ in nonseptic foals there were significantly more positive values and, therefore, lactate decreases, at [LAC]ADM–24 compared with all other time periods at [LAC]24–48, [LAC]48–72, [LAC]72–96 and [LAC]96–120 (P<0.001) while in septic foals [LAC]Δ was only significantly different between time periods at [LAC]ADM-24 and [LAC]72–96 (P<0.01) (Table 2). Also, when comparing nonseptic with septic foals, nonseptic foals had significantly greater positive values, hence [LAC] decreases at [LAC]ADM-24 when compared with septic foals at [LAC]24–48, [LAC]48–72, [LAC]72–96, [LAC]96–120 (P<0.01) (Table 4).

When considering all foals, logistic regression analyses of outcome for [LAC]Δ over all time intervals indicated that the OR for survival increased significantly for each positive 1-unit [LAC]Δ change in [LAC] at [LAC]ADMIT-24, (OR 2.24, 95% CI 1.19–4.22, P = 0.0121). The OR for survival decreased significantly for each positive 1-unit [LAC]Δ change in [LAC] at [LAC]48–72 (OR 0.39, 95% CI 0.20–0.79, P = 0.0089) and was not significant at [LAC]24–48, [LAC]72–96, and [LAC]96–120.


In the second part of this large, multicentre prospective study evaluating serial blood [LAC] and [LAC]Δ over a 120 h time period, we demonstrate that both persistently increased [LAC] and decreased [LAC]Δ over time are common in critically ill and septic hospitalised foals and definitively associated with poor outcome independent of grouped major final diagnoses. The strong association of blood [LAC] and outcome has been reliably confirmed in human medicine [1-5, 7-9, 20, 23, 24, 28-31] as well as veterinary medicine [10-19] and further strong, high-level evidence is provided by this multicentre prospective study.

Hyperlactataemia is typically the result of production that exceeds utilisation and ‘clearance’. While lactic acidosis is generally classified as type A or B lactic acidosis, often both types exist in critically ill patients [6, 22]. Interestingly, increased mortality associated with hyperlactataemia is independent of diagnosis and has been reported in human patients with sepsis and septic shock [3, 7, 8, 21, 23, 32], cardiac surgical patients [33], trauma patients [2, 29] and post operatively [4]. In veterinary medicine poor outcomes associated with hyperlactataemia have been reported in critically ill foals [10, 12, 13, 15, 25], critically ill adult horses [11, 14], critically ill dogs, dogs with infectious diseases [16, 18] and dogs with gastric dilation volvulus [17, 19]. The first part of this prospective study showed that, in foals, [LAC]ADMIT was statistically associated with nonsurvival for certain major diagnoses including sepsis, enterocolitis, unspecified colic, trauma, immune-related and respiratory disease [25] The association was not apparent in the second part of this study, possibly due to either the grouping of major final diagnoses in the assessment or diagnoses become less important than [LAC]Δ when related to survival, or a possible overlap of pathophysiological processes in the different diagnostic groups. The [LAC] was also increased in foals with prematurity/dysmaturity, neonatal encephalopathy, enteritis [12] and SIRS [10]. Given the broad spectrum of diagnoses that encompass the majority of critically ill neonatal foals commonly seen in neonatal intensive care units, it is tempting to speculate that persistently increased [LAC] is truly independent of diagnosis in foals.

Lactate kinetics are complex but almost always are an indicator of tissue perfusion, systemic oxygen delivery (DO2), tissue oxygen extraction (OE) and global tissue hypoxia when limits of oxygen extraction are reached. Additionally, impaired hepatic and renal metabolism of lactate may play a significant role in lactate kinetics. This may be due to disease affecting the liver and/or kidneys or secondary to shock states with severe hypoperfusion and impaired oxygen delivery to these organs. Under normal physiological conditions, the liver reconverts approximately 60–70% of plasma lactate to pyruvate while the kidneys reconvert approximately 20–30%. The majority of the produced pyruvate will then enter the Krebs cycle for oxidation and energy production or gluconeogenesis via the Cori cycle [6, 20]. Septic patients often suffer from ‘oxygen debt’, a term which stands for systemic oxygen uptake (VO2) failure due to inadequate DO2 or failure of OE on a microvascular level due to mitochondrial defects in oxygen utilisation [21, 34, 35]. ‘Cryptic shock’ is a condition in septic patients that describes oxygen utilisation issues on a microvascular and/or mitochondrial level which will lead to a rise of [LAC] but vital signs, especially blood pressure, are often not obviously compromised [21, 35, 36]. Recent research in man showed that septic human patients with ‘cryptic shock’ had similar mortality rates to septic human patients with overt septic shock [36]. Septic foals in our study had significantly greater lactate concentrations at all time points compared with nonseptic foals, but MAP was not different between groups indicating that some degree of ‘cryptic shock’ may have existed in the septic foal group.

It is important to note that even mild [LAC] increases (2–4 mmol/l) in septic human patients carry a poorer prognosis [8, 21, 32], suggesting that any increase in [LAC], especially in septic patients, should be taken seriously and goal-directed therapy initiated. We investigated if mild (<2.5 mmol/l; reference value), intermediate (2.5–4 mmol/l) and high (≥4 mmol/l) [LAC] increases carried equally poor prognoses, as previously reported in septic human patients at admission [8, 21, 32] and determined that in critically ill foals at admission 62% had decreased odds of survival for each 1 mmol/l increase of [LAC] in the high [LAC] group. Unlike in man, we did not observe significant decreased odds of survival in the intermediate [LAC] group. However, at 24 and 48 h after admission, the odds of survival decreased significantly in both the intermediate and high [LAC] groups for each 1 mmol/l increase of [LAC]. While the authors of the studies in man [8, 21] speculated that a single [LAC] measurement is sufficient to provide prognostic information, it appears that in foals the odds of survival decrease, in later stages of disease, even in intermediate [LAC] groups, supporting frequent [LAC] monitoring in critically ill foals. This is further supported by the observation that the OR for survival, when taking into account all sick foals, decreased from 12.5% at [LAC]ADMIT to 35% at [LAC]120 for each 1 mmol/l [LAC] increase (Table 2). The ROC analysis correctly identified overall mortality rates of 30, 40 and 34% for [LAC] cut-off values of >4.4 mmol/l, >3.21 mmol/l and >2.29 mmol/l at admission, 24 and 48 h, respectively, indicating that foals can be readily identified as ‘at risk of dying’ at certain blood lactate concentrations. Interestingly, the sensitivity at which mortality for a certain blood [LAC] cut-point was predicted increased until 48 h after presentation, supporting the use of sequential [LAC] analyses, rather than a single, one-time analysis. The results of ROC analysis and the identified cut-off points are somewhat comparable with previous studies in man [21, 31] and human paediatric studies [33] but lower than most veterinary foal and adult studies [10-13, 15] and may reflect different referral populations or disease.

The [LAC]Δ has been repeatedly evaluated as an haemodynamic endpoint for goal-directed therapy and prognosis [8, 37]. Available data regarding [LAC] ‘clearance’ have been fairly consistent with patients who fail to ‘clear’ or normalise blood [LAC] having worse outcomes than those patients that do normalise [LAC] within treatment time [4, 5, 38]. Specifically, in human patients, a [LAC] ‘clearance’ cut-off of <10% within 6–12 h after admission was consistent with greater mortality rates compared with those patients stratified into a high [LAC] ‘clearance’ group (≥10%) [23, 39]. Recent studies in equine neonates and critically ill adult horses reported significantly decreased [LAC]Δ in nonsurviving foals and adult horses [13, 14], a finding consistent with those of our study. The [LAC] returned to normal, or close to normal values in surviving and nonseptic foals, indicating treatment success with resolution of tissue oxygen debt as well as intact [LAC] metabolism, whereas it remained increased in nonsurviving foals and septic foals. Recent experimental research evaluating true exogenous [LAC] clearance in healthy adult horses, found the exogenous clearance to be similar to normolactaemic, haemodynamically stable septic human patients, whereas it was greater than that of hyperlactaemic sick human patients. Additionally, [LAC] production appeared to be lower in healthy adult horses compared with sick and healthy human patients, indicating species-specific [LAC] dynamics that may influence further research, investigating [LAC] dynamics [40].

In conclusion, this study confirms that sequential [LAC] monitoring is useful in identifying critically ill and septic foals at risk of dying. Nonsurviving and septic foals had significantly greater lactate concentrations than surviving and nonseptic foals. Additionally, blood [LAC] failed to return to normal or close to normal values in nonsurviving and septic foals at 24 h after admission and [LAC]Δ was significantly less than for surviving and nonseptic foals. The odds for survival for each 1 mmol/l [LAC] increase decreased significantly over time indicating that sequential monitoring, especially later in the course of the disease, is a powerful biomarker to predict outcome.

Authors’ declaration of interests

No competing interests have been declared.

Ethical animal research

All samples were collected as part of routine clinical management.

Sources of funding



A. Borchers and P.A. Wilkins contributed to study design, study execution, data analysis and interpretation, and preparation of the manuscript. All of the other authors contributed to study execution and data analysis and interpretation. All authors gave their final approval of the manuscript.

Manufacturers’ addresses

  1. aNova Biomedical, Waltham, Massachusetts, USA.

  2. bRoche Diagnostics, Mannheim, Germany.

  3. cRoche Diagnostic Corps, Indianapolis, Indiana, USA.

  4. dSensLab GmbH, Leipzig, Germany.

  5. eRadiometer, Copenhagen, Denmark.

  6. fAbbott Laboratories, Inc. Princeton, New Jersey, USA.

  7. gStata Corp, College Station, Texas, USA.

  8. hGraphPad, instat 3.1a, La Jolla, California, USA.