Equine Veterinary Journal ISSN 0425-1644

DOI: 10.1111/evj.12553

Factors associated with outcome in 94 hospitalised foals

diagnosed with neonatal encephalopathy
 †, M. F. MALLICOTE, R. J. MACKAY and L. C. SANCHEZ*
J. LYLE-DUGAS, S. GIGUERE
Hofmann Neonatal Intensive Care Unit, College of Veterinary Medicine, University of Florida, Gainesville, USA
†
Veterinary Medical Center, University of Georgia, Athens, USA.

*Correspondence email: sanchezl@ufl.edu; Received: 13.08.15; Accepted: 11.12.15

Summary
Reasons for performing study: Neonatal encephalopathy is the most common neurological abnormality identified in neonatal foals, but its clinical
course has been rarely characterised.
Objectives: To describe factors associated with nonsurvival in a population of foals diagnosed with neonatal encephalopathy.
Study design: Retrospective cross-sectional clinical study.
Methods: Cases were selected from equine neonatal (≤14 days of age) admissions between 1996 and 2007. Multivariable logistic regression was used
to identify clinical parameters, laboratory variables and therapeutic interventions associated with nonsurvival.
Results: A total of 94 foals were included in the study. Median age at admission was 12 h (range 0–96 h). The most frequently identified clinical signs
included abnormal udder seeking (59%), abnormal suckle (55%), inability to stand (42%), abnormal gastrointestinal motility (37%), abnormal consciousness
(34%) and seizure activity (22%). Overall, 75 (79.8%) foals survived to be discharged from the hospital and 19 foals died or were subjected to euthanasia.
Variables significantly associated with nonsurvival in the multivariable model were serum total calcium concentration, serum activity of alkaline
phosphatase, recumbency, number of concurrent diseases, and use of vasopressors/inotropes. The model correctly classified 92.0% of cases.
Conclusions: Overall survival was good and similar to previous reports. Vasopressors/inotropes were the only therapeutic intervention associated with
nonsurvival, suggesting that persistent hypotension is associated with nonsurvival in the current population. Foals with concurrent disease, high total
calcium and low alkaline phosphatase at admission, and that were recumbent or required treatment with vasopressors/inotropes during hospitalisation,
were significantly less likely to survive.

Keywords: horse; neonatology; foal; encephalopathy; neonatal maladjustment

Introduction Factors included in analysis

Neonatal encephalopathy (NE) is typically recognised as the most common
neurological disorder affecting neonatal foals [1]. The syndrome has many
synonyms, including hypoxic ischaemic encephalopathy, perinatal asphyxia
syndrome, neonatal maladjustment syndrome and dummy foal syndrome.
The nomenclature describes the clinical appearance and proposed
aetiologies, which include brain hypoxia and ischaemia, and persistence of
the fetal hypothalamic–pituitary–adrenocortical axis [2,3]. Affected animals
are either born showing abnormalities or develop clinical signs of disease
within a few days after birth. The most commonly reported clinical signs
include alterations of consciousness, inability to suckle, lack of affinity for
the mare, tongue protrusion, star-gazing, and localised and generalised
seizures, among many others. Diagnosis is typically made by excluding
infectious or congenital disorders, and prognosis is generally good in
uncomplicated cases [1].
Despite several anecdotal reports about prognostic factors and effective
treatments for NE, scientific evaluation of this information is lacking. Our
primary study goal was to evaluate factors associated with nonsurvival in
foals diagnosed with NE. A secondary study goal was to characterise the
most common clinical and clinicopathological abnormalities in affected
foals.
Materials and methods
Inclusion criteria
Cases were selected retrospectively from equine neonatal (≤14 days of
age) admissions to the Large Animal Hospital at the University of Florida
between January 1996 and June 2007 for which a diagnosis of neonatal
encephalopathy or any of its synonyms was made. Foals were excluded if
meningitis or congenital neurological abnormalities were suspected or
confirmed.
Information recovered from the medical records included the following
dichotomous (yes/no) variables: male sex, Thoroughbred breed, survival to
hospital discharge, presence of concurrent problems, IgG <4 g/l, IgG <8 g/l,
inability to stand, abnormal gastrointestinal motility, abnormal
consciousness, cranial nerve dysfunction, oedema, abnormal posture/
muscle tone, abnormal pupils, abnormal respiratory pattern, seizure
activity, sporadic myoclonus, abnormal suckle, abnormal udder seeking,
antibiotic administration prior to admission, positive blood culture, delivery
via caesarean section, corneal ulceration, diarrhoea, delivery deemed a
dystocia, onset of clinical signs prior to admission, mare’s placenta
deemed abnormal, recumbent (unable to stand with assistance at any
point), and use of the following: mechanical ventilation, respiratory
stimulants (doxapram and caffeine), vasopressors/inotropes, dimethyl
sulfoxide (DMSO), mannitol, magnesium sulfate, vitamin E.
Vital signs and laboratory variables were taken at or near the time of
admission. Clinical signs and therapeutic interventions covered the
duration of hospitalisation. The following were recorded as continuous or
discrete variables: age at admission (hours), gestation length (days),
temperature (F), duration of hospital stay (days), initial weight (kg), final
weight (kg), weight gain (kg), respiratory rate (breaths/min), heart rate
(beats/min), sepsis score [4], number of concurrent problems and litres of
plasma administered. In addition the durations (days) of the following were
recorded: inability to stand, abnormal gastrointestinal motility, abnormal
consciousness, cranial nerve dysfunction, oedema, abnormal posture/
muscle tone, abnormal pupils, abnormal respiratory pattern, seizure
activity, sporadic myoclonus, abnormal suckle, abnormal udder seeking,
enteral feeding, intranasal oxygen insufflation, parenteral nutrition
administration and intravenous fluid administration. Continuous laboratory
parameters recorded included the following: total white blood cell count,
total red blood cell count, packed red cell volume, mean corpuscular
volume, mean corpuscular haemoglobin, mean corpuscular haemoglobin

Equine Veterinary Journal 0 (2015) 1–4 © 2016 EVJ Ltd 1

Outcome in equine neonatal encephalopathy
concentration, peripheral myelocyte count, peripheral metamyelocyte
count, peripheral band count, peripheral neutrophil count, peripheral
lymphocyte count, peripheral monocyte count, plasma protein, fibrinogen,
platelet count, sodium, potassium, chloride, magnesium, creatinine, blood
urea nitrogen, calcium, phosphorus, bilirubin, alkaline phosphatase,
aspartate aminotransferase, c-glutamyl transpeptidase, creatine kinase,
arterial pH, arterial partial pressure of carbon dioxide, arterial partial
pressure of oxygen, arterial bicarbonate, standard base excess, arterial
oxygen saturation and blood lactate.
Data analysis
Data were summarised as median (10th to 90th percentiles) for continuous
data and percentages for categorical data. Normality of the data and
equality of variances were assessed using Shapiro–Wilk’s and Levene’s
tests, respectively. Comparisons between proportions were done using
Fisher’s exact test. Because most data were not normally distributed, the
Mann–Whitney U test was used to compare continuous variables between
survivors and nonsurvivors.
Initial screening of variables potentially associated with nonsurvival was
done by univariate logistic regression. Variables with screening P values
<0.10 and for which missing data represented <20% of the cases were
considered for inclusion in the multivariable model. When correlated
variables had variance inflation factors >5.0, only the variable most
significantly associated with nonsurvival was included in the model to
avoid multicollinearity. When a continuous variable had a nonlinear
association with nonsurvival, the variable was dichotomised based on the
best cut point as assessed by receiver operating characteristic (ROC)
curve analysis. The multivariable model was a backward stepwise model,
whereby variables were removed sequentially starting with that having
the largest P value until only those variables with P<0.05 remained.
Likelihood ratio tests were used to calculate P values for comparison of
nested models. Goodness of fit of the final model was evaluated using the
Hosmer and Lemeshow test and -2 log likelihood fit statistic. Odds ratios
(ORs) and 95% confidence intervals were calculated. An OR greater than
one corresponds to a positive association with nonsurvival whereas an OR
less than one corresponds to a negative association. The overall ability of
the multivariable regression model to predict nonsurvival was assessed
by use of ROC curve analysis and the optimal cut point to maximise
sensitivity and specificity was reported. For all analyses, P≤0.05 was
considered statistically significant.
Results
Ninety-four foals were included in the study. Overall, 75 (79.8%) foals
survived to be discharged from the hospital and 19 foals died or were
subjected to euthanasia. Of the nonsurvivors, 4 died, 15 were subjected to
euthanasia and financial considerations did not appear to play a role in any
case.
Comparison of recorded demographic data, clinical signs, therapeutic
interventions and laboratory data between survivors and nonsurvivors are
presented in Supplementary Items 1–4, respectively. The most frequently
identified clinical signs included abnormal udder seeking (59%), abnormal
suckle (55%), inability to stand (42%), abnormal gastrointestinal motility
(37%), abnormal consciousness (34%) and seizure activity (22%).
Nonsurviving foals were more likely to receive directed therapy with
mechanical ventilation, respiratory stimulants and vasopressors/inotropes,
but therapy with mannitol, magnesium sulfate, vitamin E, or DMSO did not
differ significantly between survivors and nonsurvivors (Supplementary
Item 3).
Of the 19 nonsurvivors, necropsy reports were available for 17, and
microscopic descriptions of brain sections were available for 11. One of the
nonsurvivors died after discharge from the hospital and did not return for
necropsy. Eleven nonsurviving foals had severe pneumonia or other forms
of disseminated sepsis or sepsis-associated complications such as vascular
thrombosis. Ten of 11 foals had histological evidence of neuronal necrosis
and/or degeneration within the central nervous system (CNS) consistent
with ischaemia; 6 were apparently septic, 4 were not. One foal had a spinal
haematoma without accompanying microscopic description of the brain.
One foal died of a gastric rupture associated with mesenteric root volvulus.
2 Equine Veterinary Journal 0 (2015) 1–4 © 2016 EVJ Ltd
J. Lyle-Dugas et al.
Of the concurrent diseases other than sepsis, the following were
reported most frequently: pneumonia (17 foals), prematurity/dysmaturity
(12), patent urachus/omphalitis (11), limb deformity (9), colic (4) and
uroperitoneum (3). Fifteen other disorders were reported in one foal each;
14 foals had a sole diagnosis of neonatal encephalopathy.
Variables potentially associated with nonsurvival by univariate
logistic regression are presented in Supplementary Item 5. Variables
significantly associated with nonsurvival in the multivariable model
were total calcium, alkaline phosphatase, recumbency, number of
concurrent diseases, and the need for use of vasopressors/inotropes
(Table 1). Complete data on all 5 variables included in the final model
were available for 87 foals, of which 69 survived and 18 died. The
model correctly classified 92.0% of the cases. The area under the
ROC curve for the ability of the variables in the final model to predict
nonsurvival was 0.948 (95% confidence interval, 0.878–0.984;
P<0.0001). By use of a cut point of >0.451, the sensitivity of the
model was 77.8% and specificity was 98.6%.
Discussion
Although neonatal encephalopathy is frequently encountered by equine
clinicians, this is the first detailed report of clinical and clinicopathological
findings in such foals along with an analysis of their impact on short-term
survival. Overall survival was high (79.8%), similar to other retrospective
reports [5,6] and reviews [1,7,8]. Of the 19 nonsurvivors, 15 were subjected
to euthanasia, and financial consideration did not appear to be the primary
reason for euthanasia in any of the foals. Pneumonia, disseminated sepsis,
or any of several sepsis-associated complications was the primary reason
for death or euthanasia in 11 of 17 foals for which necropsy results were
available. One foal suffered from an intestinal accident. In 5 foals, death
appeared related to primary neurological disease or multi-organ failure
secondary to ischaemia.
A major limitation of this study is the lack of definitive antemortem tests
for the diagnosis of neonatal encephalopathy (NE). As a result, the
diagnosis of NE relied strictly on clinical diagnosis. Post mortem
examination confirmed the diagnosis in 10 cases in which histopathology
of the brain was performed. However, it is possible that altered mentation
in some foals classified as having NE in the present study was the result of
a systemic illness rather than the result of brain injury. Moving forward,
evaluation of neuroactive progestagen concentrations may be helpful in
some cases [2,9]. Another limitation is the retrospective collection of data
from the medical records. Thus, it is possible that some subjective criteria,
such as diarrhoea or pneumonia, were either under- or over-reported.
None of the demographic or historical variables were retained in the final
model with the exception of multiple diagnoses. Foals with multiple
comorbidities were less likely to survive. While this is not a surprising
finding, it is infrequently evaluated in reports of factors associated with
TABLE 1: Multivariable logistic regression model to predict
nonsurvival (death or euthanasia) in 87 foals with a diagnosis of
neonatal encephalopathy
Variable Coefficient s.e. P Odds ratio (95% CI)
Constant -12.36
Calcium 0.911 0.369 0.01 2.49 (1.21–5.31)
Number of 1.66 0.602 0.006 5.28 (1.62–17.2)
concurrent
diseases
Recumbent 2.22 0.960 0.02 9.18 (1.40–60.2)
Vasopressor/ 2.94 1.14 0.01 19.0 (2.03–177)
inotrope use
Alkaline -0.00206 0.000784 0.009 0.998 (0.996–0.999)
phosphatase
CI, confidence interval. Hosmer and Lemeshow test: Chi-squared 8.25
with 8 degrees of freedom, P = 0.4; null model -2 log likelihood = 88.7,
full model -2 log likelihood = 37.6, P<0.0001.
J. Lyle-Dugas et al.
survival in neonatal foals. Interestingly, maternal factors were not retained
in the final model. Thus, neither commonly encountered factors (dystocia,
25.6%; abnormal placentation, 35.5%) nor uncommonly encountered ones
(caesarean section, 2%) appeared strongly associated with nonsurvival in
the current population. Median gestational length was shorter in
nonsurvivors, and the 10th percentile extended well below the typically
accepted upper limit of prematurity (320 days) for nonsurvivors (306 days)
but not survivors (322 days). Of the 12 foals diagnosed with prematurity/
dysmaturity, 6 survived and 11 had at least one other comorbidity. Thus,
while gestational length was not specifically retained in the final model, the
effect of prematurity on survival was probably multifactorial and evident in
the negative impact of multiple diagnoses upon survival in the current
population.
Clinical signs were similar to those previously described, with abnormal
udder seeking and abnormal suckling most commonly reported [1,7]. Of
the clinical signs, only recumbency was significantly associated with
nonsurvival in the final multivariable analysis.
A previous retrospective study of 78 foals with NE described high serum
creatinine concentration (32%) and high serum creatine kinase activity (61%)
as the most common clinicopathological abnormalities [5]. Although both
abnormalities were common in the current population, neither had a
significant impact on nonsurvival. The only clinicopathological values
retained in the final model were total calcium concentration and serum
alkaline phosphatase activity. The positive association between calcium
concentration and nonsurvival is interesting, given that many (if not most)
of the nonsurvivors had multiple comorbidities, including sepsis.
Hypocalcaemia has been reported in septic foals [10,11]. Septic foals also
reportedly have higher serum parathyroid hormone (PTH) and phosphorus
concentrations [11] and lower serum 25-hydroxyvitamin D3 (25(OH)D3) and
1,25-dihydroxyvitamin D3 (1,25(OH)2D3) relative to healthy and sick,
nonseptic foals [10]. In one of those reports, high PTH concentrations were
associated with nonsurvival whereas neither total nor ionised calcium
concentration had a significant association [11]. In the other, low serum 25
(OH)D3 was associated with nonsurvival in hospitalised foals whereas PTH,
total calcium and phosphorus did not have a significant association [10].
Because ionised calcium, 25(OH)D3, 1,25(OH)2D3 and vitamin D binding
protein were not recorded and/or measured, a complete picture of calcium
homeostasis in the current population is not available.
Many therapies have been recommended anecdotally for foals
diagnosed with neonatal encephalopathy [1]. In this study
population, the only therapy retained in the final model was the use
of vasopressors/inotropes. The odds of nonsurvival in foals
administered vasopressors was 19 times higher than that of foals
that did not receive vasopressors. It is most likely that vasopressor
or inotrope therapy is a proxy for hypotension associated with shock
or multi-organ system dysfunction. Because the study population
covered a wide range of years and supervising clinicians, a wide
variety of therapeutic interventions were evaluated, none of which
retained a significant impact on nonsurvival in the final model. This
finding supports the clinical impression that supportive nursing care
is the therapy most essential to a positive outcome in affected foals
[1,8].
The study population consisted solely of foals referred for intensive care.
Thus, selection bias probably favoured severely affected foals and the
prognosis for survival in an ambulatory population may be higher than that
reported. Thus, some factors associated with nonsurvival, such as the use
of vasopressors/inotropes, may not be directly applicable to foals treated
in an ambulatory setting. Nevertheless, an overall good prognosis can be
relayed when discussing the pros of referral for supportive care vs. the
obvious con of relatively high cost with owners of foals with neonatal
encephalopathy in a field setting. Similarly, the potential impact of
recumbency, shock and multiple comorbidities can be factored into
treatment and referral decisions, along with discussions of their potential
impact upon prognosis.
In conclusion, the overall prognosis in this study population of foals with
neonatal encephalopathy was good. Foals with high total calcium or low
alkaline phosphatase at admission, or those that became recumbent,
required treatment with vasopressors/inotropes or had multiple
comorbidities at any point during hospitalisation were significantly less
likely to survive.
Equine Veterinary Journal 0 (2015) 1–4 © 2016 EVJ Ltd
Outcome in equine neonatal encephalopathy
Author’s declarations of interest
No competing interests have been declared.
Ethical animal research
Research ethics committee oversight was not required by this journal:
retrospective study of clinical records. Explicit owner informed consent for
inclusion of animals in this study was not stated.
Source of funding
None.
Acknowledgements
We thank Dr Elizabeth Weber for assistance with data entry and all
University of Florida clinicians, technicians, and students for their care of
the foals.
Authorship
re and L.C. Sanchez contributed to study design.
J. Lyle-Dugas, S. Gigue
re
J. Lyle-Dugas and L.C. Sanchez contributed to data collection. S. Gigue
re, M.F. Mallicote,
performed statistical analysis. J. Lyle-Dugas, S. Gigue
R.J. MacKay and L.C. Sanchez contributed to manuscript preparation and
reviewed and approved the final manuscript.
References
1. MacKay, R.J. (2005) Neurologic disorders of neonatal foals. Vet. Clin. N.
Am.: Equine Pract. 21, 387-406.
2. Aleman, M., Pickles, K.J., Conley, A.J., Stanley, S., Haggett, E., Toth, B. and
Madigan, J.E. (2013) Abnormal plasma neuroactive progestagen
derivatives in ill, neonatal foals presented to the neonatal intensive care
unit. Equine Vet. J. 45, 661-665.
3. Drummond, W.H. (1988) Neonatal maladjustment syndrome: its
relationship to perinatal hypoxic-ischaemic insults. Equine Vet. J. 20,
Suppl. 5, 41-43.
4. Brewer, B.D. and Koterba, A.M. (1988) Development of a scoring system
for the early diagnosis of equine neonatal sepsis. Equine Vet. J. 20, 18-
22.
5. Bernard, W.V., Reimer, J.M., Cudd, T.A. and Hewlett, L. (1995) Historical
factors, clinicopathologic findings, clinical features, and outcome of
equine neonates presenting with or developing signs of central nervous
system disease. Proc. Am. Ass. Equine Practnrs. 41, 222-224.
6. Bryant, J.E., Bernard, W., Wilson, W.D. and Gardner, I. (1994) Race
earnings as an indicator of future performance in neonatal foals treated
for neurological disorders. Proc. Am. Ass. Equine Practnrs. 40, 198.
7. Wong, D., Wilkins, P.A., Bain, F.T. and Brockus, C. (2011) Neonatal
encephalopathy in foals. Comp. Cont. Educ. Vet. 33, E5.
8. Furr, M. (1996) Perinatal asphyxia in foals. Comp. Cont. Educ. Pract. Vet.
18, 1342-1351.
9. Madigan, J.E., Haggett, E.F., Pickles, K.J., Conley, A., Stanley, S., Moeller,
B., Toth, B. and Aleman, M. (2012) Allopregnanolone infusion induced
neurobehavioural alterations in a neonatal foal: is this a clue to the
pathogenesis of neonatal maladjustment syndrome?. Equine Vet. J. 44,
109-112.
10. Kamr, A.M., Dembek, K.A., Reed, S.M., Slovis, N.M., Zaghawa, A.A., Rosol,
T.J. and Toribio, R.E. (2015) Vitamin D metabolites and their association
with calcium, phosphorus, and PTH concentrations, severity of illness, and
mortality in hospitalized equine neonates. PLoS ONE 10, e0127684.
11. Hurcombe, S.D.A., Toribio, R.E., Slovis, N.M., Saville, W.J., Mudge, M.C.,
MacGillivray, K. and Frazer, M.L. (2009) Calcium regulating hormones
and serum calcium and magnesium concentrations in septic and
critically ill foals and their association with survival. J. Vet. Intern. Med.
23, 335-343.
3
Outcome in equine neonatal encephalopathy J. Lyle-Dugas et al.

Supplementary Item 3: Therapeutic interventions used in 94 foals

Supporting Information diagnosed with neonatalencephalopathy.
Additional Supporting Information may be found in the online version Supplementary Item 4: Laboratory parameters in 94 foals diagnosed
of this article at the publisher’s website: with neonatal encephalopathy.
Supplementary Item 5: Factors potentially associated with non-survival
Supplementary Item 1: Demographic and clinical data from 94 foals
by univariate logistic regression in 94 foals (P<0.10).
diagnosed with neonatalencephalopathy.
Supplementary Item 2: Clinical signs in 94 foals diagnosed with
neonatal encephalopathy.

4 Equine Veterinary Journal 0 (2015) 1–4 © 2016 EVJ Ltd