- Email: [email protected]
Neurodevelopmental Outcome of Asymptomatic Hypoglycemia Compared With Symptomatic Hypoglycemia and Euglycemia in High-Risk Neonates
Accepted Manuscript Neurodevelopmental outcome of asymptomatic hypoglycemia in comparison to symptomatic hypoglycemia and euglycemia in high risk neonates Gagan Mahajan, MD, DM, Kanya Mukhopadhyay, MD, DM, Savita Attri, PhD, Praveen Kumar, MD, DM PII:
S0887-8994(17)30282-5
DOI:
10.1016/j.pediatrneurol.2017.05.028
Reference:
PNU 9169
To appear in:
Pediatric Neurology
Received Date: 16 March 2017 Revised Date:
26 May 2017
Accepted Date: 29 May 2017
Please cite this article as: Mahajan G, Mukhopadhyay K, Attri S, Kumar P, Neurodevelopmental outcome of asymptomatic hypoglycemia in comparison to symptomatic hypoglycemia and euglycemia in high risk neonates, Pediatric Neurology (2017), doi: 10.1016/j.pediatrneurol.2017.05.028. 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
Title: Neurodevelopmental outcome of asymptomatic hypoglycemia in comparison to symptomatic hypoglycemia and euglycemia in high risk neonates
1
Kanya Mukhopadhyay (MD, DM)
2
Savita Attri (PhD)
1
Praveen Kumar (MD, DM)
SC
Gagan Mahajan (MD, DM)
M AN U
1
RI PT
Authors:
Affiliations: 1Neonatal unit and 2Pediatric biochemistry unit, Dept. of Pediatrics, PGIMER,
TE D
Chandigarh
Corresponding author: Kanya Mukhopadhyay, Professor, Neonatal unit, Department of Pediatrics, PGIMER, Chandigarh, India 160012. Fax: 0172-2744401, Phone: 0172-2755316,18.
EP
Email id: [email protected]
AC C
E-mail: [email protected],[email protected],[email protected] Running title: Neurodevelopmental outcome of hypoglycemic neonates Funding: nil
Conflict of interests: none
ACCEPTED MANUSCRIPT
Contribution of authors:
RI PT
GM: Prepared the protocol, collected and analyzed the data and drafted the manuscript KM: Conceptualized and designed the study, supervised data collection and analysis, and
SC
critically revised the manuscript SA: Conducted Biochemical analysis and reviewed the manuscript
AC C
EP
TE D
M AN U
PK: Reviewed data analysis and the manuscript
ACCEPTED MANUSCRIPT
ABSTRACT Aims: To assess the neurodevelopmental outcome of asymptomatic neonatal hypoglycemia at 1
RI PT
year and to compare it with symptomatic hypoglycemic and euglycemic neonates . Method: Seventy two hypoglycemic (plasma glucose <50 mg/dl) neonates, both symptomatic (n=27) and asymptomatic (n=45) and 70 weight and gestation matched euglycemic neonates who
SC
were > 32 weeks gestation, enrolled during 1st week of life and assessed for neurodevelopmental outcome at corrected age (CA) 6 and 12 months ( n=67 and 62 in hypoglycemia group and 63
M AN U
and 54 in euglycemia group, rest lost to follow up and death=1) .
Results: At 1 year, 8% (5 of 62, 4 in symptomatic and 1 in asymptomatic group) hypoglycemic neonates developed Cerebral Palsy. Mean Motor and Mental DQ were significantly lower at CA 6 and 12 months in any hypoglycemia (p <0.001) and if blood glucose was < 40 mg/dl (p <
TE D
0.001) as compared to euglycemia. Symptomatic infants had lower MoDQ (p= 0.004 and 0.003) and MeDQ (p= 0.001 and 0.001) at CA 6 and 12 months than asymptomatic infants and asymptomatic infants had lower MoDQ (p= < 0.001and 0.004) and MeDQ (p= 0.001 and 0.004)
EP
than euglycemic group at CA 6 and 12 months respectively. Blood glucose of < 40 mg/dl had
AC C
high sensitivity (83% for MoDQ and 81% for MeDQ) for DQ scores of < 85. Conclusion:
Hypoglycemia,
both
symptomatic
and
asymptomatic
lead
to
adverse
neurodevelopmental outcome as compared to euglycemia though it was worse in symptomatic group and at blood glucose < 40 mg/dl.
Key words: Neurodevelopment; MoDQ; MeDQ; DASII score; hypoglycemia; blood glucose
ACCEPTED MANUSCRIPT
INTRODUCTION Hypoglycemia in neonates has been recognized as a cause of serious long-term neuromorbidity
RI PT
for over 50 years which causes damage to both neuronal and glial cells of the brain resulting in death or handicap (1, 2) . Due to poor correlation between blood glucose levels and clinical manifestations and controversial treatment thresholds, it is difficult to define a safe blood using
SC
glucose. American Academy of Pediatrics (AAP) in 2011 proposed a guideline of
intravenous (IV) fluids in late preterm, term small for gestational age (SGA) and infant of
M AN U
diabetic mother (IDM)/large for gestational age (LGA) infants if blood glucose level is < 40 mg/dl and in symptomatic cases. In asymptomatic infants, a trial of enteral feed is given if blood glucose is < 25 mg/dl within 1st 4 hours and at < 35 mg/dl between 4-24 hours of age. If no improvement happens 1 hour after feed then IV glucose is recommended. In case of partial improvement, refeed/IV glucose as needed is recommended (3). Cornblath et al,proposed
TE D
operational thresholds of intervention at 36 mg/dl and to maintain a therapeutic level of more than 45 mg/dl in neonates (4). Several studies have looked into the effects of various ranges of hypoglycemia on their neurodevelopmental outcome (1-2,5-7) however the effect of
EP
asymptomatic hypoglycemia has been reported with variable results on the neurodevelopmental
AC C
outcome (1,5-6) without any clear conclusion. We planned this study to look at long term neurodevelopmental outcome in symptomatic and asymptomatic hypoglycemic neonates as compared to euglycemic neonates.
ACCEPTED MANUSCRIPT
METHODS This prospective study was conducted in a level III neonatal unit and subsequently in neonatal
RI PT
follow up clinic after discharge, from July 2010 to June 2012 in > 32 weeks gestation infants who developed hypoglycemia during first 7 days of life. As per our neonatal unit protocol, all neonates at risk of hypoglycemia (all preterm infants, term infants who are SGA or LGA and IDM) are routinely screened for hypoglycemia during first 72 hours of life with 6 hourly blood
SC
glucose monitoring. Neonates who developed hypoglycemia, as measured by glucose test strips
M AN U
(defined as blood glucose levels < 50 mg/dl, confirmed by lab glucose) were consecutively enrolled as cases after written informed consent from parents. Those who remained euglycemic were matched for weight and gestation and served as controls. In euglycemic group one lab plasma glucose level was estimated within 1st 6 hours of life. If hypoglycemia was associated with lethargy, poor feeding, seizures, jitteriness and apnea, they were considered as symptomatic
TE D
hypoglycemia. Those with major congenital malformations, severe birth asphyxia, Rh isoimmunization and grade III or IV intraventricular haemorrhage (IVH) were excluded. After 72 hours if any baby became symptomatic due to any illness or developed any symptoms
EP
suggestive of hypoglycemia , blood glucose was repeated and if found hypoglycemic were also
AC C
planned to be enrolled as cases till day 7 of life. The initial blood glucose was done with Optium Xceed Glucometer, Abbott Diabetes Care Inc. Alameda, CA 94502 USA using Optium blood glucose test strips and if it was < 50 mg/dl, plasma glucose level was done using hexokinase method on Clinical Chemistry Analyser. We used oxalate and sodium fluoride containing vials for estimation of plasma glucose. Infants with blood glucose < 40 mg/dl tested by glucose strips were started on IV fluids with glucose infusion rate (GIR) of 6 mg/kg/min. Blood glucose was repeated every 15 minutes
ACCEPTED MANUSCRIPT
initially till stabilization, subsequently ½ hourly for 2 hours and then 6 hourly for next 72 hours. If blood glucose level remained below 40 mg/dl, than GIR was increased by 2 mg/kg/min till blood glucose of 50 mg/dl or more was achieved. Those neonates having value > 40 mg/dl were
RI PT
given oral feeds and monitored every ½ hr till blood glucose increased to 50 mg/dl. They were monitored for next 72 hours with 6 hourly blood glucose. In infants who were on IV fluids, oral feeds were started as soon as possible, initially at the rate of 20-30 ml/kg/day and advancement
SC
of feed was done depending on tolerance and need for glucose infusion. After observing for 6 more hours, if blood glucose remained in euglycemic range ( > 50 mg/dl) and the baby tolerated
M AN U
feeds, then feeds were further increased and IV fluids tapered off. All blood glucose were done by optium glucose test strips. If the baby had intermittent hypoglycemia on subsequent days, everyday one sample for lab plasma glucose was sent during the episode of hypoglycemia. One documented episode of hypoglycemia on each subsequent day was counted as one day and was
TE D
added up for the total duration of hypoglycemia.
At birth all infants were assessed by New Ballard Score (9) for accurate gestational assessment. Growth was categorized as AGA, SGA and LGA as per Lubchenco’s intrauterine growth chart
EP
(10). After discharge infants were followed up at monthly intervals for 3 months for growth and feeding pattern and then at corrected ages (CA) of 6 and 12 months for neurodevelopmental
AC C
assessment. Neurological assessment was done by using Amiel Tison scale (11) and developmental assessment was done by using Developmental assessment score for Indian infants (DASII) scoring system which is an Indian adaptation of Bayley Scale of Infant Development (12). DASII scale was administered by the principal investigator (unblinded) who was trained by a certified and experienced trainer. Mental and Motor development quotients (MeDQ and MoDQ) were calculated as per the DASII instruction manual and a score of < 70 was considered
ACCEPTED MANUSCRIPT
as delay, between 70-85 borderline and > 85 average. Respiratory distress was defined as tachypnea, retractions or grunt and need for oxygen or any form of respiratory support. We included all cases of respiratory distress irrespective of causes. Sepsis included both culture
RI PT
positive and culture negative with septic screen positivity alongwith symptoms. Polycythemia was defined as packed cell volume of > 65% sampled from a major vein. The study protocol was
SC
approved by Institute research ethics committee. SAMPLE SIZE
M AN U
According to a previous study by Yamaguchi et al (13), DQ at 2 years was 103.8 ± 24.2 for cases of hypoglycemia and 116.8 ± 20.7 for controls. Based on this, to find out a difference of DQ of 10, with a SD of 20 and alpha error of 5% and a power of 80%, 128 samples were required. Anticipating 10% loss in follow up, we enrolled a total 142 (72 hypoglycemia and 70
TE D
euglycemia) patients. We finally assessed 130 (67 in hypoglycemia and 63 in euglycemia group ) at 6 months and 116 ( 62 in hypoglycemia and 54 in euglycemia group) at 1 year CA. We compared the outcome (DQ) in symptomatic and asymptomatic hypoglycemic infants as
EP
well as compared them with euglycemic infants at CA 6 months and 1 year.
AC C
STATISTICAL ANALYSIS
Statistical analysis was done using SPSS version19. Quantitative variables are reported as mean (SD), median (IQR) and qualitative variables are reported as proportions. Comparison was made by using student t test or chi square test, as appropriate. A cut off of blood glucose level was established by constructing a receiver operating curve (ROC) to define a normal DQ of > 85. A two tailed significance level of 0.05 was applied for all
ACCEPTED MANUSCRIPT
analysis. Multivariate regression analysis was performed to see the independent risk factors of abnormal neurodevelopment other than low blood glucose levels.
RI PT
RESULTS A total of 647 infants were screened who fulfilled eligibility criteria (Fig 1) and a final sample of 142 were enrolled (n=72, hypoglycemia and n=70, euglycemia). In the hypoglycemia group, 27
SC
(37.5%) were symptomatic (lethargy - 22, poor feeding – 15, seizures- 7, jitteriness- 5) and some infants had more than one symptoms .In the hypoglycemic group, 67 were followed up at 6
M AN U
months and 62 at 12 months CA and in euglycemic group , 63 were followed up at 6 months and 54 at 12 months CA. Baseline demographic profiles and morbidities of the 2 groups are given in Table1. Mean gestational age and proportion of SGA infants and morbidities (respiratory distress, polycythemia, any sepsis) were significantly higher in hypoglycemic group than
TE D
euglycemic group (Table1).
Forty five (62.5%) neonates had asymptomatic and 27 (37.5%) had symptomatic hypoglycemia and the demographic characteristics were similar in these 2 groups (Table 2) except
EP
polycythemia and sepsis which were significantly higher in symptomatic group. The mean lowest blood glucose level was significantly lower (p<0.001) in symptomatic hypoglycemia than
AC C
asymptomatic hypoglycemia group (20.4 + 8.2 vs. 30 + 8.4 mg/dl, p <0.001) respectively and these were significantly lower as compared to euglycemia group (59 + 4 mg/dl, p<0.001). In hypoglycemia group, there were 3 babies whose blood glucose was between 40-49 mg/dl and they were all asymptomatic but rest 69 babies had blood glucose < 40 mg/dl of which 27 were symptomatic and 42 were asymptomatic.
ACCEPTED MANUSCRIPT
Table 3 depicts the significant differences in MoDQ and MeDQ at CA 6 and 12 months among symptomatic , asymptomatic hypoglycemic and euglycemic infants.
RI PT
The mean MoDQ and MeDQ at 6 months ( 82 + 7 vs. 89 + 4, p<0.001 and 81 + 9 vs. 89 + 3, p <0.001) and at 12 months CA (82 + 9 vs. 89 + 3, p < 0.001 and 82 +10 vs. 90 + 2.5 , p <0.001) were significantly lower in overall hypoglycemia group as compared to euglycemia group
SC
respectively.
At 1 year, 8% (5 of 62) developed cerebral palsy in hypoglycemia group (4 in symptomatic and 1
M AN U
in asymptomatic group). In the infants who had seizures 33.4% (2 of the 6) were abnormal as compared to 66.6% (4 out of 6, p=0.069) normal at CA 1 year. There was no difference in DQ scores at CA 6 and 12 months whether hypoglycemia persisted for < 48 hours vs. > 48 hours. At 6 months CA, 28% (19 of 67) infants had MoDQ > 85 as compared to 75% (47 of 63 ) and
TE D
MeDQ > 85 in 25% (17 of 67) as compared to 71% (45 of 63) infants ) in hypoglycemia and euglycemia group respectively (p<0.001). At 1 year CA, 53% (29 of 62) infants had MoDQ > 85 as compared to 89% (48 of 54) and MeDQ > 85 in 45% (28 of 62) as compared to 93% (50
EP
of 54) in hypoglycemia and euglycemia group respectively (p<0.001). Multivariate linear regression analysis by forward step wise method was carried out by using
AC C
MoDQ and MeDQ as dependent variable and factors found significant on univariate analysis like gestational age, birth weight, growth status, lowest blood glucose, respiratory distress, polycythemia, sepsis and IVH as independent variables. It was found that sepsis, birth weight and lowest blood glucose were independent predictors of low MoDQ score and IVH and lowest blood glucose for low MeDQ (Table 4).
ACCEPTED MANUSCRIPT
We prepared a ROC curve taking a cut off of MeDQ and MoDQ of < 85 at 6 months and lowest blood glucose levels detected in hypoglycemic neonates. We found that a blood glucose cut-off level of 40 mg/dl had sensitivity and specificity of 83% and 67% respectively for MoDQ and
RI PT
81% and 70% for MeDQ. Area under curve was 0.806 (95% CI 0.72 to 0.88) for both MoDQ and MeDQ. Based on the ROC cut off, we divided hypoglycemic infants in 2 sub groups who had any episode of blood glucose < 40 mg/dl and > 40 mg/dl . DASII scores at CA 6 and 12
SC
months were significantly lower in < 40 mg/dl group as compared to > 40 mg/dl group (p <
AC C
EP
TE D
M AN U
0.001) (Fig2).
ACCEPTED MANUSCRIPT
DISCUSSION This study shows that both symptomatic and asymptomatic hypoglycemia have worse outcomes
RI PT
than euglycemic infants and a cut off of 40 mg/dl was associated with lower developmental score. We enrolled only > 32 weeks gestation neonates as co-morbidities would be less in this gestation and hence less confounding factors for adverse neurodevelopmental outcome.
SC
Various studies have used different cut off values (2-3 mmol/L) to define hypoglycemia (4, 6, 14, 15 ). Generally a cut off of 47 mg/dl is considered to define hypoglycemia however scientific
M AN U
evidence to adopt this policy is not robust and this level should be viewed with caution (14). Blood glucose levels which require active intervention is also controversial and American Academy of Pediatrics (AAP) in 2011 has suggested intravenous fluids only in symptomatic infants with blood glucose < 40 mg/dl and in asymptomatic infants at blood glucose < 25 mg/dl
TE D
within 1st 4 hours of life and < 35 mg/dl between 4-24 hours of life though the effects of asymptomatic hypoglycemia on neurodevelopmental outcome was not reviewed (3). Kalhan et al proposed a cut off value of 45 mg/dl in symptomatic infants and 36 mg/dl in asymptomatic
EP
infants though the targeted therapeutic values are maintained in the range of 72-90 mg/dl (16). Wayenberg et al suggested active intervention at < 45 mg/dl as repeated mild hypoglycemia
AC C
could be detrimental to developing preterm brain (17). We used 50 mg/dl of plasma glucose as cut off for defining hypoglycemia and 40 mg/dl for active intervention in the form of intravenous fluids in our study population. Various authors which conducted similar studies which included gestational age ranged between 28-43 weeks and birth weight between 1040-4850 grams ( 1,8,18,19,20,21,22) and in SGA infants (20). Mean gestation and proportion of SGA infants in our hypoglycemia group was significantly higher than euglycemia group though we wanted to match for this. Duvanel et al
ACCEPTED MANUSCRIPT
reported an incidence of 72.9% hypoglycemia in SGA infants ( hypoglycemia defined as <47 mg/dl) and found significantly lower head circumference and lower scores in psychometric tests at 3.5 and 5 years of age (20) . Though we did not analyze DQ in SGA subcategories but overall
RI PT
we had significantly high proportion of SGA infants in hypoglycemic group.
Literature on neurodevelopmental outcome in asymptomatic hypoglycemia is very controversial. Koivisto et al found normal neurodevelopment at 1 – 4 yrs of age in all asymptomatic
SC
hypoglycemic infants (n=66) except 4 having only abnormal ophthalmological findings (1),
M AN U
however we found neonates with even asymptomatic hypoglycemia had significantly lower MoDQ and MeDQ at 6 and 12 months CA when compared to euglycemic infants. Abnormal evoked potentials were recorded by Koh et al in 10 out of 11 children with blood glucose < 47 mg/dl and 5 of them were asymptomatic (5). Bland et al did not find any significant difference in developmental outcome in term euglycemic and mildly hypoglycemic infants at 4 years (18).
TE D
Griffiths et al (23) found normal neurodevelopment in all 12 asymptomatic hypoglycemic but 6 of 29 symptomatic hypoglycemic were abnormal at 51 months follow up which is similar to our observation that symptomatic group had lower DQ than asymptomatic. Fluge et al reported 5
EP
out 7 asymptomatic children to be normal at a mean of 3 and ½ years though only 1out of 9 symptomatic infants was normal (24). Singh M et al reported mental and psychomotor
AC C
developmental index were significantly lower in infants with hypoglycemia and seizures as compared to hypoglycemia with other symptoms and in euglycemic infants (7). Koivisto et al observed that outcome was worse in presence of seizures with hypoglycemia (1). In our study, 1/3rd were abnormal if they had seizure during hypoglycemia . Though we did not find any significant difference in the DQ in the infants with hypoglycemia persisting for < 48 hours vs ≥ 48 hrs, Singh et al found that the duration of hypoglycemia was
ACCEPTED MANUSCRIPT
directly related to mental developmental index ( r=-0.74, y=102.5-o.69x) and psychomotor developmental index (r= -0.81, y = 105.6-0.86x) (7). The reason for our no difference could be very small size in hypoglycemia lasting < 48 hours and our arbitrary cut off of 48 hours. It may
RI PT
be repeated episodes of hypoglycemia within a particular time period which causes adverse outcome rather than duration alone. Fluge at al concluded that duration and severity influences the adverse neurological outcome (24). Lucas et al in 1988 reported a strong correlation with
SC
duration of hypoglycemia and lower mental and motor developmental indices even after adjusting for other risk factors for hypoglycemia in a cohort of preterm infants (< 1850 grams).
M AN U
At 18 months corrected age MeDQ and MoDQ were reduced by 14 and 13 points respectively and incidence of neurodevelopmental impairment was increased by a factor of 3.5 ( 95% CI 1.39.4) if hypoglycemia was recorded on 5 or more separate days (8).
We found that MoDQ and MeDQ were significantly lower in the group with blood glucose < 40
TE D
mg/dl as compared to > 40 mg/dl group. Pildes et al (19) showed that infants with asymptomatic hypoglycemia and blood glucose between 20 – 30 mg/dl did not have poor neurological abnormality while Singh et al showed that infants with blood glucose of 17.6 ± 4.4 mg/dl had
EP
low mental and psychomotor developmental indices as compared to those with higher blood glucose (7). In a recent study reported by McKinlay et al concluded that maintaining blood
(25).
AC C
glucose above 47 mg/dl was not associated with adverse neurodevelopmental outcome at 2 years
In our study in addition to lowest blood glucose, sepsis and birth weight were found to be independent predictors of low MoDQ and IVH for low MeDQ however no such analysis were done in previous studies except duration or number of episodes of hypoglycemia or lowest level
ACCEPTED MANUSCRIPT
of blood glucose ( 7,8,20). We also made a ROC curve to find the cut off level of lowest blood glucose to predict low DQ scores (DQ <85).
RI PT
The limitations of our study is absence of blinding due to limited man power. Primary investigator who enrolled cases conducted the development assessments. We monitored blood glucose routinely till 72 hours of age in all infants hence some asymptomatic infants beyond 72
SC
hours may have been missed. This is also true that outcome at 6 months and 12 months may not correlate well with later cognitive outcomes, and may be more so if done in an unblinded set up.
M AN U
These cases also require longer follow up beyond 1 year for better prediction of outcome. The strengths of our study are that we assessed neurodevelopmental outcome of asymptomatic group which is reported to have normal outcome in previous studies (1, 6) though a recent study reported less proficient on fourth grade achievement in those who had transient early
TE D
hypoglycemia as compared to normoglycemia (26). We also had adequate sample size, matched euglycemic controls and a good follow up rate till CA 1 year (86%). We used DASII scale which is an Indian adaptation of Bayley scale. We also developed a regression equation to calculate DQ
EP
based on risk factors.
AC C
In conclusion, symptomatic hypoglycemia had worst neurodevelopmental outcome however even asymptomatic hypoglycemia had a worse outcome as compared to euglycemic group and a cut off level of 40 mg/dl was associated with lower developmental scores. All high risk neonates should be monitored adequately to prevent any kind of hypoglycemia.
ACCEPTED MANUSCRIPT
References: 1.
Koivisto M, Blanco-Sequeiros M, Krause U: Neonatal symptomatic and asymptomatic
2.
RI PT
hypoglycaemia: a follow-up study of 151 children. Dev Med Child Neurol. 1972; 14:603-14. Fluge G: Clinical aspects of neonatal hypoglycaemia. Acta Paediatr Scand. 1974; 63:826-
32.
SC
3. Adamkin DH. Postnatal glucose homeostasis in Late -Preterm and Term infants. Committee on fetus and newborn. Americal Academy of Pediatrics. Pediatrics 2011; 127: 575-9.
M AN U
4. Cornblath M, Hawdon JM, Williams AF, Aynsley-Green A, Ward-Platt MP, Schwartz R, et al. Controversies regarding definition of neonatal hypoglycemia: suggested operational thresholds. Pediatrics. 2000 ; 105:1141-5.
5. Koh TH, Aynsley-Green A, Tarbit M, Eyre JA. Neural dysfunction during hypoglycaemia. Arch Dis Child. 1988; 63:1353-8.
TE D
6.Griffiths AD BG: Assessment of effects of neonatal hypoglycaemia. A study of 41 cases with matched controls. Arch Dis Child. 1971; 46:819-27. 7. Singh M, Singhal PK, Paul VK, Deorari AK, Sundaram KR, Ghorpade MD, Agadi A:
EP
Neurodevelopmental outcome of asymptomatic & symptomatic infants with neonatal
AC C
hypoglycaemia. Indian J Med Res. 1991; 94:6-10. 8. Lucas A, Morley R, Cole TJ: Adverse neurodevelopmental outcome of moderate neonatal hypoglycaemia. BMJ. 1988; 297:1304-8. 9.
Ballard JL, Khoury JC, Wedig K, Wang L, Eilers-Walsman BL, Lipp R: New Ballard
Score, expanded to include extremely premature infants. J Pediatr. 1991; 119:417-23. 10. Lubchenco LO, Hansman C, Dressler M, Boyd E. Intrauterine growth as estimated from live born birth weight data at 24 to 42 weeks of gestation. Pediatrics. 1963; 32:793-800
ACCEPTED MANUSCRIPT
11.
Amiel-Tison C: Update of the Amiel-Tison neurologic assessment for the term neonate or
at 40 weeks corrected age. Pediatr Neurol. 2002; 27:196-212. 12.
Pathak P: Developmental Assessment Scales for Indian Infants (DASII), Revised Baroda
13.
RI PT
Norms 1997. Anand Agencies, Pune.
Yamaguchi K, Mishina J, Mitsuishi C, Takamura T, Nishida H: Follow-up study of
14.
Simmons R, Stanley C. Neonatal hypoglycemia studies — is there a sweet story of
M AN U
success yet? N Engl J Med 2015 ; 373 (16) : 1567-9.
15.
SC
neonatal hypoglycemia. Acta Paediatr Jpn. 1997; 39 Suppl 1:S51-53
Srinivasan G, Pildes RS, Cattamanchi G, Voora S, Lilien LD: Plasma glucose values in
normal neonates: a new look. J Pediatr. 1986;109:114-7.
16. Kalhan S, Peter Wohl S. Hypoglycemia : what is it for neonate? Am J Perinatology.
TE D
2000;17(1):11-8.
17. Wayenberg JL, Pardou A. Moderate hypoglycemia in the preterm infant: is it relevant? Arch Pediatr. 2008 ; 15(2) : 153-6.
EP
18. Brand PLP, Molenaar NLD, Kaaijk C, Wierenga WS: Neurodevelopmental outcome of
81. 19.
AC C
hypoglycaemia in healthy large for gestational age term newborns. Arch Dis Child. 2005; 90:78-
Pildes RS, Cornblath M, Warren I, Page-El E, Di Menza S, Merritt DM et al: A
prospective controlled study of neonatal hypoglycemia. Pediatrics. 1974; 54:5-14 20.
Duvanel CB, Fawer CL, Cotting J, Hohlfeld P, Matthieu JM: Long-term effects of
neonatal hypoglycemia on brain growth and psychomotor development in small-for-gestationalage preterm infants. J Pediatr. 1999; 134:492-8.
ACCEPTED MANUSCRIPT
21.
Montassir H, Maegaki Y, Ogura K, Kurozawa Y, Nagata I, Kanzaki S et al. Associated
factors in neonatal hypoglycemic brain injury. Brain Dev. 2009; 31:649-56. Cornblath M, Schwartz R: Outcome of neonatal hypoglycaemia. Complete data are
needed. BMJ. 1999; 318:194-5. 23.
Griffiths AD, Bryant GM: Assessment of effects of neonatal hypoglycaemia. A study of
41 cases with matched controls. Arch Dis Child. 1971; 46:819-827
Fluge G: Neurological findings at follow-up in neonatal hypoglycaemia. Acta Paediatr
SC
24.
RI PT
22.
M AN U
Scand. 1975; 64:629-34.
25. McKinlay CJD, Alsweiler JM, Ansell JM, Nicola S, Anstice NS, Chase JG et al for the CHYLD Study Group. Neonatal glycemia and neurodevelopmental outcomes at 2 years. N Engl J Med. 2015; 373:1507-18.
TE D
26. Kaiser JR, Bai S,Gibson N,Holland G, Lin TM, Swearingen CJ et al. Association between transient neonatal hypoglycemia and fourth grade achievementtest proficiency - A population
AC C
EP
based study.JAMA Pediatr 2015 ; 169(10) : 913-921.
ACCEPTED MANUSCRIPT
Table 1. Baseline characteristics and morbidities of Study subjects
Euglycemia
(n= 72)
(n = 70)
^Gestation (weeks)
35.6 ± 2.2
34.8 ± 1.9
^Birth weight (grams)
1711 ± 433
1819 ± 383
^OFC (cm)
30.5 ± 2.2
31.2 ± 1.6
0.022*
^Length (cm)
43.5 ± 3.6
44.4 ± 3.1
0.115
#
39 (54.2%)
Male #AGA
14 (19.4%)
SGA
58 (80.6%)
#
Maternal illness
$
Apgar Score 1 min
$
Apgar Score 5 min
#
Respiratory distress
#
Polycythemia
#
Any sepsis
SC
38 (54.3%)
0.029*
0.119
0.989
29 (41.4%) 41 (58.6%)
0.004*
28 (40%)
<0.001*
8 (7 - 8)
8 (8 – 8)
0.120
9 (9-10)
9 (9-10)
0.055
16(22%)
7(10%)
0.048*
26(36%)
1(1.4%)
<0.001*
21(29%)
2(3%)
<0.001*
EP
TE D
52 (72%)
AC C
p value
RI PT
Hypoglycemia
M AN U
Variables
Values are expressed in #n (%), ^Mean ± SD, $ Median (IQR) *p < 0.05 considered significant AGA- Appropriate for gestational age, SGA – Small for gestational age
ACCEPTED MANUSCRIPT
Table2. Comparison of demographic characteristics of asymptomatic and symptomatic hypoglycemic infants Symptomatic hypoglycemia
(n=45)
(n=27)
^Gestation (weeks)
35.3 ± 2.3
36.0 ± 1.9
^Birth weight (grams)
1680 ± 441
1761 ± 424
^OFC (cm)
30.2 ± 2.2
30.8 ± 2.2
^Length (cm)
42.9 ± 3.4
^Males
p value
M AN U
SC
RI PT
Asymptomatic hypoglycemia
Variables
0.226
0.766
0.296 0.046*
22 (48.9%)
17 (63%)
0.246
#AGA
10 (22.2%)
4 (14.8%)
SGA
35 (77.8%)
23 (85.2%)
0.442
26 (54%)
12 (44%)
0.691
8 ( 7 - 8)
8 (7 - 8)
0.130
TE D
44.6 ± 3.3
Maternal illness
$
Apgar Score 1 min
$
Apgar Score 5 min
9 (9 -10)
9 ( 9 -10)
0.140
#
Respiratory distress
10(22%)
6(22%)
0.142
#
Polycythemia
15(33%)
11(41%)
<0.001
#
Any sepsis
7(16%)
14(52%)
<.001
$
Duration of hospital stay 12.5 (5 – 60)
13.0 (5 – 56)
0.437
AC C
EP
#
(days) Values are expressed in #n (%), ^Mean ±SD, $Median (IQR),*p <0.05 considered significant
ACCEPTED MANUSCRIPT
Table 3. Comparison of MoDQ and MeDQ at 12 months in various subgroups Euglycemia
Asymptomatic
Symptomatic
hypoglycemia
hypoglycemia
n=63
n=41
n=26
MoDQ
89 + 4
84 + 6.5
79 + 7
MeDQ
89 + 3
84 +7
77 + 10
12 months
n=54
n=36
MoDQ
89 + 3
85 + 7
MeDQ
90 + 2.5
85 + 9
p*
RI PT
6 months
SC
DQ
<0.001 <0.001
n=26
M AN U
79 + 10 78 + 11
<0.001 <0.001
*Difference between groups by Bonferroni posthoc tests,
TE D
At 6 months , MoDQ, between euglycemia and asymptomatic hypoglycemia, p= < 0.001 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and asymptomatic hypoglycemia, p=0.004.
EP
At 6 months, MeDQ, between euglycemia and asymptomatic hypoglycemia, p=0.001 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and
AC C
asymptomatic hypoglycemia, p= <0.001. At 12 months , MoDQ between euglycemia and asymptomatic hypoglycemia, p=0.004 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and asymptomatic hypoglycemia, p=0.003. At 12 months, MeDQ, between euglycemia and asymptomatic hypoglycemia, p=0.004 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and asymptomatic hypoglycemia, p=0.001.
ACCEPTED MANUSCRIPT
MeDQ
Beta
95% CI
Constant
70.4
61.5 – 79.2
<0.001*
Sepsis
– 3.8
– 7.9 to – 0.38
0.03*
Birth weight
0.005
0.001 – 0.008
0.02*
Lowest blood glucose
0.19
0.02 – 0.36
0.03*
Constant
75.8
69.6 – 81.9
<0.001*
– 7.4
– 14.3 to – 0.57
0.03*
0.23
0.01 – 0.45
0.03*
IVH
TE D
Lowest blood glucose
p value
RI PT
MoDQ
Parameter
M AN U
Variable
SC
Table 4. Regression equation for MoDQ and MeDQ based on risk factors
(MoDQ and MeDQ as dependent variable). CI - Confidence Interval, IVH- Intraventricular
EP
haemorrhage *significant p value< 0.05
AC C
MoDQ = 70.4 + (0.005 x birth weight) + (0.19 x lowest blood sugar) -3.8, if sepsis is present and 70.4 + (0.005 x birth weight) + (0.19 x lowest blood sugar), if sepsis is not present
MeDQ = 75.8 + (0.23 x lowest blood sugar) – 7.4 if IVH is present and 75.8 + (0.23 x lowest blood sugar) if IVH is not present
ACCEPTED MANUSCRIPT
n = 647 Total infants assessed> 32 wks gestation, till 7 days age and all SGA, LGA, IDM or preterm n= 538 Euglycemia
n= 109 Hypoglycemia
n= 70 (Euglycemia) selected consecutively by weight and gestation matching
n = 72 Enrolled
TE D
n= 27 Symptomatic hypoglycemia
n= 63 Follow up till 6 months CA
AC C
n= 67 Follow up till 6 months CA
Lost to follow up = 5
n= 62 Follow up till 12 months CA
n = 70 Enrolled
Lost to follow up= 7
EP
Lost to follow up= 4 Death = 1
M AN U
SC
Excluded= 37 Lab blood glucose not available – 14, Investigator absent – 12, Deaths – 5, declined follow up - 6
n= 45 Asymptomatic hypoglycemia
RI PT
Fig. 1 - Flow of patients
Lost to follow up= 9
n= 54 Follow up till 12 months CA
ACCEPTED MANUSCRIPT
SGA – Small for gestational age, LGA – Large for gestational age, IDM – Infant of diabetic
AC C
EP
TE D
M AN U
SC
RI PT
mother
ACCEPTED MANUSCRIPT
Fig 2. Comparison of mean values of MoDQ and MeDQ in infants with blood glucose < 40
RI PT
mg/dl with > = 40 mg/dl group at CA 6 months and 1 year
92 90 88
SC
86 84
<40 mg/dl
82 80 78 76 MoDQ6 Mo
MeDQ 6 Mo
M AN U
>=40 mg/dl
MoDQ 1 yr
AC C
EP
TE D
All p values <0.001
MeDQ 1 yr
S0887-8994(17)30282-5
DOI:
10.1016/j.pediatrneurol.2017.05.028
Reference:
PNU 9169
To appear in:
Pediatric Neurology
Received Date: 16 March 2017 Revised Date:
26 May 2017
Accepted Date: 29 May 2017
Please cite this article as: Mahajan G, Mukhopadhyay K, Attri S, Kumar P, Neurodevelopmental outcome of asymptomatic hypoglycemia in comparison to symptomatic hypoglycemia and euglycemia in high risk neonates, Pediatric Neurology (2017), doi: 10.1016/j.pediatrneurol.2017.05.028. 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
Title: Neurodevelopmental outcome of asymptomatic hypoglycemia in comparison to symptomatic hypoglycemia and euglycemia in high risk neonates
1
Kanya Mukhopadhyay (MD, DM)
2
Savita Attri (PhD)
1
Praveen Kumar (MD, DM)
SC
Gagan Mahajan (MD, DM)
M AN U
1
RI PT
Authors:
Affiliations: 1Neonatal unit and 2Pediatric biochemistry unit, Dept. of Pediatrics, PGIMER,
TE D
Chandigarh
Corresponding author: Kanya Mukhopadhyay, Professor, Neonatal unit, Department of Pediatrics, PGIMER, Chandigarh, India 160012. Fax: 0172-2744401, Phone: 0172-2755316,18.
EP
Email id: [email protected]
AC C
E-mail: [email protected],[email protected],[email protected] Running title: Neurodevelopmental outcome of hypoglycemic neonates Funding: nil
Conflict of interests: none
ACCEPTED MANUSCRIPT
Contribution of authors:
RI PT
GM: Prepared the protocol, collected and analyzed the data and drafted the manuscript KM: Conceptualized and designed the study, supervised data collection and analysis, and
SC
critically revised the manuscript SA: Conducted Biochemical analysis and reviewed the manuscript
AC C
EP
TE D
M AN U
PK: Reviewed data analysis and the manuscript
ACCEPTED MANUSCRIPT
ABSTRACT Aims: To assess the neurodevelopmental outcome of asymptomatic neonatal hypoglycemia at 1
RI PT
year and to compare it with symptomatic hypoglycemic and euglycemic neonates . Method: Seventy two hypoglycemic (plasma glucose <50 mg/dl) neonates, both symptomatic (n=27) and asymptomatic (n=45) and 70 weight and gestation matched euglycemic neonates who
SC
were > 32 weeks gestation, enrolled during 1st week of life and assessed for neurodevelopmental outcome at corrected age (CA) 6 and 12 months ( n=67 and 62 in hypoglycemia group and 63
M AN U
and 54 in euglycemia group, rest lost to follow up and death=1) .
Results: At 1 year, 8% (5 of 62, 4 in symptomatic and 1 in asymptomatic group) hypoglycemic neonates developed Cerebral Palsy. Mean Motor and Mental DQ were significantly lower at CA 6 and 12 months in any hypoglycemia (p <0.001) and if blood glucose was < 40 mg/dl (p <
TE D
0.001) as compared to euglycemia. Symptomatic infants had lower MoDQ (p= 0.004 and 0.003) and MeDQ (p= 0.001 and 0.001) at CA 6 and 12 months than asymptomatic infants and asymptomatic infants had lower MoDQ (p= < 0.001and 0.004) and MeDQ (p= 0.001 and 0.004)
EP
than euglycemic group at CA 6 and 12 months respectively. Blood glucose of < 40 mg/dl had
AC C
high sensitivity (83% for MoDQ and 81% for MeDQ) for DQ scores of < 85. Conclusion:
Hypoglycemia,
both
symptomatic
and
asymptomatic
lead
to
adverse
neurodevelopmental outcome as compared to euglycemia though it was worse in symptomatic group and at blood glucose < 40 mg/dl.
Key words: Neurodevelopment; MoDQ; MeDQ; DASII score; hypoglycemia; blood glucose
ACCEPTED MANUSCRIPT
INTRODUCTION Hypoglycemia in neonates has been recognized as a cause of serious long-term neuromorbidity
RI PT
for over 50 years which causes damage to both neuronal and glial cells of the brain resulting in death or handicap (1, 2) . Due to poor correlation between blood glucose levels and clinical manifestations and controversial treatment thresholds, it is difficult to define a safe blood using
SC
glucose. American Academy of Pediatrics (AAP) in 2011 proposed a guideline of
intravenous (IV) fluids in late preterm, term small for gestational age (SGA) and infant of
M AN U
diabetic mother (IDM)/large for gestational age (LGA) infants if blood glucose level is < 40 mg/dl and in symptomatic cases. In asymptomatic infants, a trial of enteral feed is given if blood glucose is < 25 mg/dl within 1st 4 hours and at < 35 mg/dl between 4-24 hours of age. If no improvement happens 1 hour after feed then IV glucose is recommended. In case of partial improvement, refeed/IV glucose as needed is recommended (3). Cornblath et al,proposed
TE D
operational thresholds of intervention at 36 mg/dl and to maintain a therapeutic level of more than 45 mg/dl in neonates (4). Several studies have looked into the effects of various ranges of hypoglycemia on their neurodevelopmental outcome (1-2,5-7) however the effect of
EP
asymptomatic hypoglycemia has been reported with variable results on the neurodevelopmental
AC C
outcome (1,5-6) without any clear conclusion. We planned this study to look at long term neurodevelopmental outcome in symptomatic and asymptomatic hypoglycemic neonates as compared to euglycemic neonates.
ACCEPTED MANUSCRIPT
METHODS This prospective study was conducted in a level III neonatal unit and subsequently in neonatal
RI PT
follow up clinic after discharge, from July 2010 to June 2012 in > 32 weeks gestation infants who developed hypoglycemia during first 7 days of life. As per our neonatal unit protocol, all neonates at risk of hypoglycemia (all preterm infants, term infants who are SGA or LGA and IDM) are routinely screened for hypoglycemia during first 72 hours of life with 6 hourly blood
SC
glucose monitoring. Neonates who developed hypoglycemia, as measured by glucose test strips
M AN U
(defined as blood glucose levels < 50 mg/dl, confirmed by lab glucose) were consecutively enrolled as cases after written informed consent from parents. Those who remained euglycemic were matched for weight and gestation and served as controls. In euglycemic group one lab plasma glucose level was estimated within 1st 6 hours of life. If hypoglycemia was associated with lethargy, poor feeding, seizures, jitteriness and apnea, they were considered as symptomatic
TE D
hypoglycemia. Those with major congenital malformations, severe birth asphyxia, Rh isoimmunization and grade III or IV intraventricular haemorrhage (IVH) were excluded. After 72 hours if any baby became symptomatic due to any illness or developed any symptoms
EP
suggestive of hypoglycemia , blood glucose was repeated and if found hypoglycemic were also
AC C
planned to be enrolled as cases till day 7 of life. The initial blood glucose was done with Optium Xceed Glucometer, Abbott Diabetes Care Inc. Alameda, CA 94502 USA using Optium blood glucose test strips and if it was < 50 mg/dl, plasma glucose level was done using hexokinase method on Clinical Chemistry Analyser. We used oxalate and sodium fluoride containing vials for estimation of plasma glucose. Infants with blood glucose < 40 mg/dl tested by glucose strips were started on IV fluids with glucose infusion rate (GIR) of 6 mg/kg/min. Blood glucose was repeated every 15 minutes
ACCEPTED MANUSCRIPT
initially till stabilization, subsequently ½ hourly for 2 hours and then 6 hourly for next 72 hours. If blood glucose level remained below 40 mg/dl, than GIR was increased by 2 mg/kg/min till blood glucose of 50 mg/dl or more was achieved. Those neonates having value > 40 mg/dl were
RI PT
given oral feeds and monitored every ½ hr till blood glucose increased to 50 mg/dl. They were monitored for next 72 hours with 6 hourly blood glucose. In infants who were on IV fluids, oral feeds were started as soon as possible, initially at the rate of 20-30 ml/kg/day and advancement
SC
of feed was done depending on tolerance and need for glucose infusion. After observing for 6 more hours, if blood glucose remained in euglycemic range ( > 50 mg/dl) and the baby tolerated
M AN U
feeds, then feeds were further increased and IV fluids tapered off. All blood glucose were done by optium glucose test strips. If the baby had intermittent hypoglycemia on subsequent days, everyday one sample for lab plasma glucose was sent during the episode of hypoglycemia. One documented episode of hypoglycemia on each subsequent day was counted as one day and was
TE D
added up for the total duration of hypoglycemia.
At birth all infants were assessed by New Ballard Score (9) for accurate gestational assessment. Growth was categorized as AGA, SGA and LGA as per Lubchenco’s intrauterine growth chart
EP
(10). After discharge infants were followed up at monthly intervals for 3 months for growth and feeding pattern and then at corrected ages (CA) of 6 and 12 months for neurodevelopmental
AC C
assessment. Neurological assessment was done by using Amiel Tison scale (11) and developmental assessment was done by using Developmental assessment score for Indian infants (DASII) scoring system which is an Indian adaptation of Bayley Scale of Infant Development (12). DASII scale was administered by the principal investigator (unblinded) who was trained by a certified and experienced trainer. Mental and Motor development quotients (MeDQ and MoDQ) were calculated as per the DASII instruction manual and a score of < 70 was considered
ACCEPTED MANUSCRIPT
as delay, between 70-85 borderline and > 85 average. Respiratory distress was defined as tachypnea, retractions or grunt and need for oxygen or any form of respiratory support. We included all cases of respiratory distress irrespective of causes. Sepsis included both culture
RI PT
positive and culture negative with septic screen positivity alongwith symptoms. Polycythemia was defined as packed cell volume of > 65% sampled from a major vein. The study protocol was
SC
approved by Institute research ethics committee. SAMPLE SIZE
M AN U
According to a previous study by Yamaguchi et al (13), DQ at 2 years was 103.8 ± 24.2 for cases of hypoglycemia and 116.8 ± 20.7 for controls. Based on this, to find out a difference of DQ of 10, with a SD of 20 and alpha error of 5% and a power of 80%, 128 samples were required. Anticipating 10% loss in follow up, we enrolled a total 142 (72 hypoglycemia and 70
TE D
euglycemia) patients. We finally assessed 130 (67 in hypoglycemia and 63 in euglycemia group ) at 6 months and 116 ( 62 in hypoglycemia and 54 in euglycemia group) at 1 year CA. We compared the outcome (DQ) in symptomatic and asymptomatic hypoglycemic infants as
EP
well as compared them with euglycemic infants at CA 6 months and 1 year.
AC C
STATISTICAL ANALYSIS
Statistical analysis was done using SPSS version19. Quantitative variables are reported as mean (SD), median (IQR) and qualitative variables are reported as proportions. Comparison was made by using student t test or chi square test, as appropriate. A cut off of blood glucose level was established by constructing a receiver operating curve (ROC) to define a normal DQ of > 85. A two tailed significance level of 0.05 was applied for all
ACCEPTED MANUSCRIPT
analysis. Multivariate regression analysis was performed to see the independent risk factors of abnormal neurodevelopment other than low blood glucose levels.
RI PT
RESULTS A total of 647 infants were screened who fulfilled eligibility criteria (Fig 1) and a final sample of 142 were enrolled (n=72, hypoglycemia and n=70, euglycemia). In the hypoglycemia group, 27
SC
(37.5%) were symptomatic (lethargy - 22, poor feeding – 15, seizures- 7, jitteriness- 5) and some infants had more than one symptoms .In the hypoglycemic group, 67 were followed up at 6
M AN U
months and 62 at 12 months CA and in euglycemic group , 63 were followed up at 6 months and 54 at 12 months CA. Baseline demographic profiles and morbidities of the 2 groups are given in Table1. Mean gestational age and proportion of SGA infants and morbidities (respiratory distress, polycythemia, any sepsis) were significantly higher in hypoglycemic group than
TE D
euglycemic group (Table1).
Forty five (62.5%) neonates had asymptomatic and 27 (37.5%) had symptomatic hypoglycemia and the demographic characteristics were similar in these 2 groups (Table 2) except
EP
polycythemia and sepsis which were significantly higher in symptomatic group. The mean lowest blood glucose level was significantly lower (p<0.001) in symptomatic hypoglycemia than
AC C
asymptomatic hypoglycemia group (20.4 + 8.2 vs. 30 + 8.4 mg/dl, p <0.001) respectively and these were significantly lower as compared to euglycemia group (59 + 4 mg/dl, p<0.001). In hypoglycemia group, there were 3 babies whose blood glucose was between 40-49 mg/dl and they were all asymptomatic but rest 69 babies had blood glucose < 40 mg/dl of which 27 were symptomatic and 42 were asymptomatic.
ACCEPTED MANUSCRIPT
Table 3 depicts the significant differences in MoDQ and MeDQ at CA 6 and 12 months among symptomatic , asymptomatic hypoglycemic and euglycemic infants.
RI PT
The mean MoDQ and MeDQ at 6 months ( 82 + 7 vs. 89 + 4, p<0.001 and 81 + 9 vs. 89 + 3, p <0.001) and at 12 months CA (82 + 9 vs. 89 + 3, p < 0.001 and 82 +10 vs. 90 + 2.5 , p <0.001) were significantly lower in overall hypoglycemia group as compared to euglycemia group
SC
respectively.
At 1 year, 8% (5 of 62) developed cerebral palsy in hypoglycemia group (4 in symptomatic and 1
M AN U
in asymptomatic group). In the infants who had seizures 33.4% (2 of the 6) were abnormal as compared to 66.6% (4 out of 6, p=0.069) normal at CA 1 year. There was no difference in DQ scores at CA 6 and 12 months whether hypoglycemia persisted for < 48 hours vs. > 48 hours. At 6 months CA, 28% (19 of 67) infants had MoDQ > 85 as compared to 75% (47 of 63 ) and
TE D
MeDQ > 85 in 25% (17 of 67) as compared to 71% (45 of 63) infants ) in hypoglycemia and euglycemia group respectively (p<0.001). At 1 year CA, 53% (29 of 62) infants had MoDQ > 85 as compared to 89% (48 of 54) and MeDQ > 85 in 45% (28 of 62) as compared to 93% (50
EP
of 54) in hypoglycemia and euglycemia group respectively (p<0.001). Multivariate linear regression analysis by forward step wise method was carried out by using
AC C
MoDQ and MeDQ as dependent variable and factors found significant on univariate analysis like gestational age, birth weight, growth status, lowest blood glucose, respiratory distress, polycythemia, sepsis and IVH as independent variables. It was found that sepsis, birth weight and lowest blood glucose were independent predictors of low MoDQ score and IVH and lowest blood glucose for low MeDQ (Table 4).
ACCEPTED MANUSCRIPT
We prepared a ROC curve taking a cut off of MeDQ and MoDQ of < 85 at 6 months and lowest blood glucose levels detected in hypoglycemic neonates. We found that a blood glucose cut-off level of 40 mg/dl had sensitivity and specificity of 83% and 67% respectively for MoDQ and
RI PT
81% and 70% for MeDQ. Area under curve was 0.806 (95% CI 0.72 to 0.88) for both MoDQ and MeDQ. Based on the ROC cut off, we divided hypoglycemic infants in 2 sub groups who had any episode of blood glucose < 40 mg/dl and > 40 mg/dl . DASII scores at CA 6 and 12
SC
months were significantly lower in < 40 mg/dl group as compared to > 40 mg/dl group (p <
AC C
EP
TE D
M AN U
0.001) (Fig2).
ACCEPTED MANUSCRIPT
DISCUSSION This study shows that both symptomatic and asymptomatic hypoglycemia have worse outcomes
RI PT
than euglycemic infants and a cut off of 40 mg/dl was associated with lower developmental score. We enrolled only > 32 weeks gestation neonates as co-morbidities would be less in this gestation and hence less confounding factors for adverse neurodevelopmental outcome.
SC
Various studies have used different cut off values (2-3 mmol/L) to define hypoglycemia (4, 6, 14, 15 ). Generally a cut off of 47 mg/dl is considered to define hypoglycemia however scientific
M AN U
evidence to adopt this policy is not robust and this level should be viewed with caution (14). Blood glucose levels which require active intervention is also controversial and American Academy of Pediatrics (AAP) in 2011 has suggested intravenous fluids only in symptomatic infants with blood glucose < 40 mg/dl and in asymptomatic infants at blood glucose < 25 mg/dl
TE D
within 1st 4 hours of life and < 35 mg/dl between 4-24 hours of life though the effects of asymptomatic hypoglycemia on neurodevelopmental outcome was not reviewed (3). Kalhan et al proposed a cut off value of 45 mg/dl in symptomatic infants and 36 mg/dl in asymptomatic
EP
infants though the targeted therapeutic values are maintained in the range of 72-90 mg/dl (16). Wayenberg et al suggested active intervention at < 45 mg/dl as repeated mild hypoglycemia
AC C
could be detrimental to developing preterm brain (17). We used 50 mg/dl of plasma glucose as cut off for defining hypoglycemia and 40 mg/dl for active intervention in the form of intravenous fluids in our study population. Various authors which conducted similar studies which included gestational age ranged between 28-43 weeks and birth weight between 1040-4850 grams ( 1,8,18,19,20,21,22) and in SGA infants (20). Mean gestation and proportion of SGA infants in our hypoglycemia group was significantly higher than euglycemia group though we wanted to match for this. Duvanel et al
ACCEPTED MANUSCRIPT
reported an incidence of 72.9% hypoglycemia in SGA infants ( hypoglycemia defined as <47 mg/dl) and found significantly lower head circumference and lower scores in psychometric tests at 3.5 and 5 years of age (20) . Though we did not analyze DQ in SGA subcategories but overall
RI PT
we had significantly high proportion of SGA infants in hypoglycemic group.
Literature on neurodevelopmental outcome in asymptomatic hypoglycemia is very controversial. Koivisto et al found normal neurodevelopment at 1 – 4 yrs of age in all asymptomatic
SC
hypoglycemic infants (n=66) except 4 having only abnormal ophthalmological findings (1),
M AN U
however we found neonates with even asymptomatic hypoglycemia had significantly lower MoDQ and MeDQ at 6 and 12 months CA when compared to euglycemic infants. Abnormal evoked potentials were recorded by Koh et al in 10 out of 11 children with blood glucose < 47 mg/dl and 5 of them were asymptomatic (5). Bland et al did not find any significant difference in developmental outcome in term euglycemic and mildly hypoglycemic infants at 4 years (18).
TE D
Griffiths et al (23) found normal neurodevelopment in all 12 asymptomatic hypoglycemic but 6 of 29 symptomatic hypoglycemic were abnormal at 51 months follow up which is similar to our observation that symptomatic group had lower DQ than asymptomatic. Fluge et al reported 5
EP
out 7 asymptomatic children to be normal at a mean of 3 and ½ years though only 1out of 9 symptomatic infants was normal (24). Singh M et al reported mental and psychomotor
AC C
developmental index were significantly lower in infants with hypoglycemia and seizures as compared to hypoglycemia with other symptoms and in euglycemic infants (7). Koivisto et al observed that outcome was worse in presence of seizures with hypoglycemia (1). In our study, 1/3rd were abnormal if they had seizure during hypoglycemia . Though we did not find any significant difference in the DQ in the infants with hypoglycemia persisting for < 48 hours vs ≥ 48 hrs, Singh et al found that the duration of hypoglycemia was
ACCEPTED MANUSCRIPT
directly related to mental developmental index ( r=-0.74, y=102.5-o.69x) and psychomotor developmental index (r= -0.81, y = 105.6-0.86x) (7). The reason for our no difference could be very small size in hypoglycemia lasting < 48 hours and our arbitrary cut off of 48 hours. It may
RI PT
be repeated episodes of hypoglycemia within a particular time period which causes adverse outcome rather than duration alone. Fluge at al concluded that duration and severity influences the adverse neurological outcome (24). Lucas et al in 1988 reported a strong correlation with
SC
duration of hypoglycemia and lower mental and motor developmental indices even after adjusting for other risk factors for hypoglycemia in a cohort of preterm infants (< 1850 grams).
M AN U
At 18 months corrected age MeDQ and MoDQ were reduced by 14 and 13 points respectively and incidence of neurodevelopmental impairment was increased by a factor of 3.5 ( 95% CI 1.39.4) if hypoglycemia was recorded on 5 or more separate days (8).
We found that MoDQ and MeDQ were significantly lower in the group with blood glucose < 40
TE D
mg/dl as compared to > 40 mg/dl group. Pildes et al (19) showed that infants with asymptomatic hypoglycemia and blood glucose between 20 – 30 mg/dl did not have poor neurological abnormality while Singh et al showed that infants with blood glucose of 17.6 ± 4.4 mg/dl had
EP
low mental and psychomotor developmental indices as compared to those with higher blood glucose (7). In a recent study reported by McKinlay et al concluded that maintaining blood
(25).
AC C
glucose above 47 mg/dl was not associated with adverse neurodevelopmental outcome at 2 years
In our study in addition to lowest blood glucose, sepsis and birth weight were found to be independent predictors of low MoDQ and IVH for low MeDQ however no such analysis were done in previous studies except duration or number of episodes of hypoglycemia or lowest level
ACCEPTED MANUSCRIPT
of blood glucose ( 7,8,20). We also made a ROC curve to find the cut off level of lowest blood glucose to predict low DQ scores (DQ <85).
RI PT
The limitations of our study is absence of blinding due to limited man power. Primary investigator who enrolled cases conducted the development assessments. We monitored blood glucose routinely till 72 hours of age in all infants hence some asymptomatic infants beyond 72
SC
hours may have been missed. This is also true that outcome at 6 months and 12 months may not correlate well with later cognitive outcomes, and may be more so if done in an unblinded set up.
M AN U
These cases also require longer follow up beyond 1 year for better prediction of outcome. The strengths of our study are that we assessed neurodevelopmental outcome of asymptomatic group which is reported to have normal outcome in previous studies (1, 6) though a recent study reported less proficient on fourth grade achievement in those who had transient early
TE D
hypoglycemia as compared to normoglycemia (26). We also had adequate sample size, matched euglycemic controls and a good follow up rate till CA 1 year (86%). We used DASII scale which is an Indian adaptation of Bayley scale. We also developed a regression equation to calculate DQ
EP
based on risk factors.
AC C
In conclusion, symptomatic hypoglycemia had worst neurodevelopmental outcome however even asymptomatic hypoglycemia had a worse outcome as compared to euglycemic group and a cut off level of 40 mg/dl was associated with lower developmental scores. All high risk neonates should be monitored adequately to prevent any kind of hypoglycemia.
ACCEPTED MANUSCRIPT
References: 1.
Koivisto M, Blanco-Sequeiros M, Krause U: Neonatal symptomatic and asymptomatic
2.
RI PT
hypoglycaemia: a follow-up study of 151 children. Dev Med Child Neurol. 1972; 14:603-14. Fluge G: Clinical aspects of neonatal hypoglycaemia. Acta Paediatr Scand. 1974; 63:826-
32.
SC
3. Adamkin DH. Postnatal glucose homeostasis in Late -Preterm and Term infants. Committee on fetus and newborn. Americal Academy of Pediatrics. Pediatrics 2011; 127: 575-9.
M AN U
4. Cornblath M, Hawdon JM, Williams AF, Aynsley-Green A, Ward-Platt MP, Schwartz R, et al. Controversies regarding definition of neonatal hypoglycemia: suggested operational thresholds. Pediatrics. 2000 ; 105:1141-5.
5. Koh TH, Aynsley-Green A, Tarbit M, Eyre JA. Neural dysfunction during hypoglycaemia. Arch Dis Child. 1988; 63:1353-8.
TE D
6.Griffiths AD BG: Assessment of effects of neonatal hypoglycaemia. A study of 41 cases with matched controls. Arch Dis Child. 1971; 46:819-27. 7. Singh M, Singhal PK, Paul VK, Deorari AK, Sundaram KR, Ghorpade MD, Agadi A:
EP
Neurodevelopmental outcome of asymptomatic & symptomatic infants with neonatal
AC C
hypoglycaemia. Indian J Med Res. 1991; 94:6-10. 8. Lucas A, Morley R, Cole TJ: Adverse neurodevelopmental outcome of moderate neonatal hypoglycaemia. BMJ. 1988; 297:1304-8. 9.
Ballard JL, Khoury JC, Wedig K, Wang L, Eilers-Walsman BL, Lipp R: New Ballard
Score, expanded to include extremely premature infants. J Pediatr. 1991; 119:417-23. 10. Lubchenco LO, Hansman C, Dressler M, Boyd E. Intrauterine growth as estimated from live born birth weight data at 24 to 42 weeks of gestation. Pediatrics. 1963; 32:793-800
ACCEPTED MANUSCRIPT
11.
Amiel-Tison C: Update of the Amiel-Tison neurologic assessment for the term neonate or
at 40 weeks corrected age. Pediatr Neurol. 2002; 27:196-212. 12.
Pathak P: Developmental Assessment Scales for Indian Infants (DASII), Revised Baroda
13.
RI PT
Norms 1997. Anand Agencies, Pune.
Yamaguchi K, Mishina J, Mitsuishi C, Takamura T, Nishida H: Follow-up study of
14.
Simmons R, Stanley C. Neonatal hypoglycemia studies — is there a sweet story of
M AN U
success yet? N Engl J Med 2015 ; 373 (16) : 1567-9.
15.
SC
neonatal hypoglycemia. Acta Paediatr Jpn. 1997; 39 Suppl 1:S51-53
Srinivasan G, Pildes RS, Cattamanchi G, Voora S, Lilien LD: Plasma glucose values in
normal neonates: a new look. J Pediatr. 1986;109:114-7.
16. Kalhan S, Peter Wohl S. Hypoglycemia : what is it for neonate? Am J Perinatology.
TE D
2000;17(1):11-8.
17. Wayenberg JL, Pardou A. Moderate hypoglycemia in the preterm infant: is it relevant? Arch Pediatr. 2008 ; 15(2) : 153-6.
EP
18. Brand PLP, Molenaar NLD, Kaaijk C, Wierenga WS: Neurodevelopmental outcome of
81. 19.
AC C
hypoglycaemia in healthy large for gestational age term newborns. Arch Dis Child. 2005; 90:78-
Pildes RS, Cornblath M, Warren I, Page-El E, Di Menza S, Merritt DM et al: A
prospective controlled study of neonatal hypoglycemia. Pediatrics. 1974; 54:5-14 20.
Duvanel CB, Fawer CL, Cotting J, Hohlfeld P, Matthieu JM: Long-term effects of
neonatal hypoglycemia on brain growth and psychomotor development in small-for-gestationalage preterm infants. J Pediatr. 1999; 134:492-8.
ACCEPTED MANUSCRIPT
21.
Montassir H, Maegaki Y, Ogura K, Kurozawa Y, Nagata I, Kanzaki S et al. Associated
factors in neonatal hypoglycemic brain injury. Brain Dev. 2009; 31:649-56. Cornblath M, Schwartz R: Outcome of neonatal hypoglycaemia. Complete data are
needed. BMJ. 1999; 318:194-5. 23.
Griffiths AD, Bryant GM: Assessment of effects of neonatal hypoglycaemia. A study of
41 cases with matched controls. Arch Dis Child. 1971; 46:819-827
Fluge G: Neurological findings at follow-up in neonatal hypoglycaemia. Acta Paediatr
SC
24.
RI PT
22.
M AN U
Scand. 1975; 64:629-34.
25. McKinlay CJD, Alsweiler JM, Ansell JM, Nicola S, Anstice NS, Chase JG et al for the CHYLD Study Group. Neonatal glycemia and neurodevelopmental outcomes at 2 years. N Engl J Med. 2015; 373:1507-18.
TE D
26. Kaiser JR, Bai S,Gibson N,Holland G, Lin TM, Swearingen CJ et al. Association between transient neonatal hypoglycemia and fourth grade achievementtest proficiency - A population
AC C
EP
based study.JAMA Pediatr 2015 ; 169(10) : 913-921.
ACCEPTED MANUSCRIPT
Table 1. Baseline characteristics and morbidities of Study subjects
Euglycemia
(n= 72)
(n = 70)
^Gestation (weeks)
35.6 ± 2.2
34.8 ± 1.9
^Birth weight (grams)
1711 ± 433
1819 ± 383
^OFC (cm)
30.5 ± 2.2
31.2 ± 1.6
0.022*
^Length (cm)
43.5 ± 3.6
44.4 ± 3.1
0.115
#
39 (54.2%)
Male #AGA
14 (19.4%)
SGA
58 (80.6%)
#
Maternal illness
$
Apgar Score 1 min
$
Apgar Score 5 min
#
Respiratory distress
#
Polycythemia
#
Any sepsis
SC
38 (54.3%)
0.029*
0.119
0.989
29 (41.4%) 41 (58.6%)
0.004*
28 (40%)
<0.001*
8 (7 - 8)
8 (8 – 8)
0.120
9 (9-10)
9 (9-10)
0.055
16(22%)
7(10%)
0.048*
26(36%)
1(1.4%)
<0.001*
21(29%)
2(3%)
<0.001*
EP
TE D
52 (72%)
AC C
p value
RI PT
Hypoglycemia
M AN U
Variables
Values are expressed in #n (%), ^Mean ± SD, $ Median (IQR) *p < 0.05 considered significant AGA- Appropriate for gestational age, SGA – Small for gestational age
ACCEPTED MANUSCRIPT
Table2. Comparison of demographic characteristics of asymptomatic and symptomatic hypoglycemic infants Symptomatic hypoglycemia
(n=45)
(n=27)
^Gestation (weeks)
35.3 ± 2.3
36.0 ± 1.9
^Birth weight (grams)
1680 ± 441
1761 ± 424
^OFC (cm)
30.2 ± 2.2
30.8 ± 2.2
^Length (cm)
42.9 ± 3.4
^Males
p value
M AN U
SC
RI PT
Asymptomatic hypoglycemia
Variables
0.226
0.766
0.296 0.046*
22 (48.9%)
17 (63%)
0.246
#AGA
10 (22.2%)
4 (14.8%)
SGA
35 (77.8%)
23 (85.2%)
0.442
26 (54%)
12 (44%)
0.691
8 ( 7 - 8)
8 (7 - 8)
0.130
TE D
44.6 ± 3.3
Maternal illness
$
Apgar Score 1 min
$
Apgar Score 5 min
9 (9 -10)
9 ( 9 -10)
0.140
#
Respiratory distress
10(22%)
6(22%)
0.142
#
Polycythemia
15(33%)
11(41%)
<0.001
#
Any sepsis
7(16%)
14(52%)
<.001
$
Duration of hospital stay 12.5 (5 – 60)
13.0 (5 – 56)
0.437
AC C
EP
#
(days) Values are expressed in #n (%), ^Mean ±SD, $Median (IQR),*p <0.05 considered significant
ACCEPTED MANUSCRIPT
Table 3. Comparison of MoDQ and MeDQ at 12 months in various subgroups Euglycemia
Asymptomatic
Symptomatic
hypoglycemia
hypoglycemia
n=63
n=41
n=26
MoDQ
89 + 4
84 + 6.5
79 + 7
MeDQ
89 + 3
84 +7
77 + 10
12 months
n=54
n=36
MoDQ
89 + 3
85 + 7
MeDQ
90 + 2.5
85 + 9
p*
RI PT
6 months
SC
DQ
<0.001 <0.001
n=26
M AN U
79 + 10 78 + 11
<0.001 <0.001
*Difference between groups by Bonferroni posthoc tests,
TE D
At 6 months , MoDQ, between euglycemia and asymptomatic hypoglycemia, p= < 0.001 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and asymptomatic hypoglycemia, p=0.004.
EP
At 6 months, MeDQ, between euglycemia and asymptomatic hypoglycemia, p=0.001 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and
AC C
asymptomatic hypoglycemia, p= <0.001. At 12 months , MoDQ between euglycemia and asymptomatic hypoglycemia, p=0.004 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and asymptomatic hypoglycemia, p=0.003. At 12 months, MeDQ, between euglycemia and asymptomatic hypoglycemia, p=0.004 and between euglycemia and symptomatic hypoglycemia, p=<0.001 and between symptomatic and asymptomatic hypoglycemia, p=0.001.
ACCEPTED MANUSCRIPT
MeDQ
Beta
95% CI
Constant
70.4
61.5 – 79.2
<0.001*
Sepsis
– 3.8
– 7.9 to – 0.38
0.03*
Birth weight
0.005
0.001 – 0.008
0.02*
Lowest blood glucose
0.19
0.02 – 0.36
0.03*
Constant
75.8
69.6 – 81.9
<0.001*
– 7.4
– 14.3 to – 0.57
0.03*
0.23
0.01 – 0.45
0.03*
IVH
TE D
Lowest blood glucose
p value
RI PT
MoDQ
Parameter
M AN U
Variable
SC
Table 4. Regression equation for MoDQ and MeDQ based on risk factors
(MoDQ and MeDQ as dependent variable). CI - Confidence Interval, IVH- Intraventricular
EP
haemorrhage *significant p value< 0.05
AC C
MoDQ = 70.4 + (0.005 x birth weight) + (0.19 x lowest blood sugar) -3.8, if sepsis is present and 70.4 + (0.005 x birth weight) + (0.19 x lowest blood sugar), if sepsis is not present
MeDQ = 75.8 + (0.23 x lowest blood sugar) – 7.4 if IVH is present and 75.8 + (0.23 x lowest blood sugar) if IVH is not present
ACCEPTED MANUSCRIPT
n = 647 Total infants assessed> 32 wks gestation, till 7 days age and all SGA, LGA, IDM or preterm n= 538 Euglycemia
n= 109 Hypoglycemia
n= 70 (Euglycemia) selected consecutively by weight and gestation matching
n = 72 Enrolled
TE D
n= 27 Symptomatic hypoglycemia
n= 63 Follow up till 6 months CA
AC C
n= 67 Follow up till 6 months CA
Lost to follow up = 5
n= 62 Follow up till 12 months CA
n = 70 Enrolled
Lost to follow up= 7
EP
Lost to follow up= 4 Death = 1
M AN U
SC
Excluded= 37 Lab blood glucose not available – 14, Investigator absent – 12, Deaths – 5, declined follow up - 6
n= 45 Asymptomatic hypoglycemia
RI PT
Fig. 1 - Flow of patients
Lost to follow up= 9
n= 54 Follow up till 12 months CA
ACCEPTED MANUSCRIPT
SGA – Small for gestational age, LGA – Large for gestational age, IDM – Infant of diabetic
AC C
EP
TE D
M AN U
SC
RI PT
mother
ACCEPTED MANUSCRIPT
Fig 2. Comparison of mean values of MoDQ and MeDQ in infants with blood glucose < 40
RI PT
mg/dl with > = 40 mg/dl group at CA 6 months and 1 year
92 90 88
SC
86 84
<40 mg/dl
82 80 78 76 MoDQ6 Mo
MeDQ 6 Mo
M AN U
>=40 mg/dl
MoDQ 1 yr
AC C
EP
TE D
All p values <0.001
MeDQ 1 yr