Relationship between serum insulin and point-of-admission blood glucose of ill neonates at a tertiary health facility in Nigeria

Relationship between serum insulin and point-of-admission blood glucose of ill neonates at a tertiary health facility in Nigeria

Emmanuel Oluwatosin Adeniji^1,&, Bankole Peter Kuti², Jerome Boluwaji Elusiyan²

 

¹Department of Paediatrics and Child Health, College of Health Sciences, Ladoke Akintola University of Technology, Osogbo, Nigeria, ²Department of Paediatrics and Child Health, Obafemi Awolowo University, Ile-Ife, Nigeria

 

 

^&Corresponding author
Emmanuel Oluwatosin Adeniji, Department of Paediatrics and Child Health, College of Health Sciences, Ladoke Akintola University of Technology, Osogbo, Nigeria

 

 

 

 

Blood glucose homeostasis is an important aspect of newborn physiology as abnormalities including hypoglycaemia and hyperglycaemia constitute significant risks to the newborn health [1, 2]. Insulin is secreted by the beta cells of the islet of Langerhans of the pancreas [3]. It causes rapid uptake, storage, and use of glucose by the tissues of the body especially the muscles, adipose tissue, and liver [4]. The counter-regulatory hormones are hormones that oppose the action of insulin leading to increased blood glucose levels [4]. They include glucagon, cortisol, growth hormone, adrenaline and noradrenaline. The levels of the counter-regulatory hormones are increased during stressful clinical conditions [3]. Neonatal hyperglycaemia (NNH) and its relationship with serum insulin in ill neonates is largely understudied especially in the West African sub-region. Many studies have attributed increased level of stress hormones which are known to stimulate gluconeogenesis as being important in the pathogenesis of hyperglycaemia in ill neonates [1, 5, 6] but the importance of serum insulin in this respect needs to be further studied. There is paucity of studies relating blood glucose levels in ill newborns and the respective serum insulin levels at the point of admission especially in Nigeria. Knowing the serum insulin levels in ill babies with hyperglycaemia may justify insulin use or otherwise early in the management of NNH. The aim of the study was to determine the point-of-admission serum insulin and its relationship with the blood glucose levels in neonates admitted to a Nigerian tertiary health facility.

 

 

The study was a prospective cross-sectional study carried out at the Special Care Baby Unit (SCBU) of the wesley Guild Hospital (WGH), Ilesa, Osun State, Southwest Nigeria. Ethical approval for this study was obtained from the Ethics and Research Committee of the Obafemi Awolowo University Teaching Hospital (OAUTH) with approval number ERC/2016/03/17. Babies aged 0-28 days admitted to the SCBU of the WGH irrespective of gestational age, birth weight and initial diagnosis were recruited into the study.

 

Study location: the WGH is one of the units of the OAUTH. The SCBU consists of eight glass-walled cubicles that can admit up to 32 babies. The ward is manned by six clinicians including a consultant paediatrician, 3 resident doctors and 2 house officers. These are assisted by 15 nurses and other supportive members of staff. Three hundred babies aged ≤28 days admitted to the SCBU of the WGH were recruited into the study at the point of admission into the SCBU. Exclusion criteria included babies who had major congenital anomalies or parental refusal of consent. History obtained about the recruited babies included the age at presentation, parent's socioeconomic class (using Oyedeji's classification) [7], place of antenatal care, place of delivery, mode of delivery, gestational age (GA) and the mother's parity. These were documented in the proforma designed for the study. Physical examination of each baby was done and significant signs were documented in the proforma.

 

The weight of the babies were classified into normal (2.50 - 3.99kg), low birth weight (LBW) (<2.50kg) and macrosomia (>05; 4.00kg) [8]. Babies were also classified as preterm (GA <37 weeks), term (GA between 37-42weeks), and postterm (GA >42 weeks). The babies were also classified into small-for-gestational age (SGA) i.e. weight <10^th percentile of the expected for the gestational age, appropriate-for-gestational age (AGA) when the weight is between the 10^th and 90^th percentile, and large-for-gestational age (LGA) when weight is >90^th percentile using an intrauterine growth chart [8]. Babies with sepsis were further classified as presumed sepsis, probable sepsis, and confirmed sepsis. Babies with confirmed sepsis not only had risk factors or clinical features of sepsis, but they also had bacteria growth on blood culture [9]. Presumed sepsis were babies with one or more risk factors to suggest sepsis including prolonged rupture of membrane, peripartum fever, or chorioamnionitis but yet to manifest clinical features of sepsis [9]. Babies with probable sepsis were babies with clinical features to suggest sepsis such as fever, hypothermia, poor suck, lethargy, respiratory distress etc. [9].

 

Blood glucose and serum insulin determination: peripheral venous blood was collected using aseptic procedure from each baby following venepuncture at the dorsum of the hand. A drop of venous blood was used for cot-side measurement of blood glucose using the Accu-Chek Active^® glucometer (Roche Diagnostics GmbH, Germany) while two millilitres of blood sample was collected in a plain bottle. Blood glucose value >6.9mmol/L was taken as hyperglycaemia while blood glucose value <2.2mmol/L was taken as hypoglycaemia [2, 5, 10, 11]. The two ml of blood sample that was collected in the plain bottle was centrifuged at 3000rpm for five minutes shortly after collection. The supernatant serum was then separated into another plain bottle and immediately stored inside the freezer in the hospital laboratory at temperature of -20° centigrade. The supernatant serum was then transported in ice packs to the Chemical Pathology Laboratory of the Obafemi Awolowo University Teaching Hospital, Ile-Ife where the serum insulin was determined by a Chemical Pathologist using the Accu-Ɔ-Bind Enzyme-Linked Immunosorbent Assay (ELISA) Microwells which was manufactured by Monobind Inc. Lake Forest, CA 92630, USA.

 

Data analysis: the data for the study was analyzed with Statistical Package for Social Sciences (SPSS) for windows version 23.0 (IBM Corp. Armonk 2015, NY, USA). Descriptive statistics including measures of central tendency like mean with standard deviation (SD) and median with interquartile range (IQR) were used to summarize normally and non-normally distributed continous variables respectively, while proportions and percentages were used to summarize categorical variables such as sex, and age range. Differences in serum insulin level of the various categories of babies which were not normally distributed were assessed using the Mann Whitney U test (U) or Kruskal Wallis test (k) as appropriate. Spearman's correlation was used to determine the relationship between serum insulin level and the blood glucose of the babies. Statistical significance was set at p value <0.05 at 95% confidence interval.

 

 

Over a six-month study period (January to June 2017), 300 babies were recruited into the study.

 

Age of the babies at presentation: the median (IQR) age of the babies at presentation was 10.0 (0.5-70.0) hours. Majority (59.0%) of the babies presented within 24 hours of delivery, while 37 (12.3%) presented after 7 days of life.

 

Sex: there were 179 (59.7%) males and 121 (40.3%) females among the study participants. The male to female ratio was 1.5: 1.

 

Parental socio-economic class: one hundred and fifty-eight (52.7%) of the recruited babies were born to parents from the middle socio-economic class while 111 (37.0%) babies were born to parents from upper class. Only 31 (10.3%) were from the low socio-economic class.

 

Maternal age: the mean (SD) maternal age was 29.4 (6.2) years with a range of 17 to 51 years. The highest proportion (82.4%) of the mothers were in the age range 20-35 years.

 

Some obstetric characteristics and the babies' classification: the obstetric characteristics of the mothers and the babies' classification are shown in Table 1.

 

Gestational age (GA): two hundred and eighteen (72.6%) of the babies were term while 74 (24.7%) were preterms. The mean (SD) GA of the babies was 37.7 (2.9) weeks with range of 26 to 44 weeks.

 

Weight for gestational age: two hundred and fifty-four (84.6%) of the babies were AGA, 35 (11.7%) were SGA and 11 (3.7%) were LGA.

 

Maternal parity: this ranged from one to seven with a mean (SD) of 2.2 (1.3). One hundred and twenty (40.0%) of the mothers were primiparous, while 162 (54.0%) were multiparous.

 

Place of antenatal care: the mothers of 123 (41.0%) babies received antenatal care (ANC) at the WGH (booked) while the mothers of the remaining 177 (59.0%) babies were unbooked at the WGH.

 

Place of delivery: one hundred and forty-six (48.7%) of the babies were delivered at the WGH (inborn) while the remaining 154 (51.3%) were outborns. Among the outborns, 58 (37.7%) were delivered at maternity centres, 47 (30.5%) were delivered at private hospitals, 25 (16.2%) were delivered at general hospitals while 15 (9.7%) and 9 (5.8%) were delivered at mission houses or at the mother's homes respectively.

 

Mode of delivery: one hundred and eighty-three (61.0%) of the babies were delivered by vaginal delivery while the remaining 117 (39.0%) were delivered through caesarian section.

 

Gestational diabetes mellitus in mother: five (1.7%) of the mothers had gestational diabetes mellitus (GDM).

 

Serum insulin measurements: the median (IQR) serum insulin for all the babies studied was 9.79 (3.0 -35.3) µIU/ml with a range of 0.1 to 383.5µIU/ml.

 

Serum insulin levels in the different categories of babies

 

Sex: the median (IQR) serum insulin level for the male babies in this study was 10.7 (3.0 - 34.0) µIU/ml while that of female babies was 8.9 (3.2-38.0) µIU/ml. There was no statistically significant difference between the median (IQR) serum insulin levels in both groups of babies (Mann Whitney U = -0.069, p = 0.945).

 

Age: the babies that were older than or equal to 168 hours had the highest median (IQR) serum insulin level of 25.6 (7.3-43.3) µIU/ml. The median (IQR) serum insulin levels of the babies aged 72-167 hours, 24-71 hours, and babies less than 24 hours of age were 10.6 (3.3-47.2), 6.3 (2.0-34.0) and 6.8 (3.0-34.0) µIU/ml respectively. The age of the babies at presentation was significantly related with the median (IQR) serum insulin level (k = 8.602, p = 0.035).

 

Gestational age: the median (IQR) serum insulin of the post term babies was 34.0 (15.5-53.8) µIU/ml, while the median (IQR) serum insulin of term and preterm babies were 10.1 (3.1-35.2) and 6.4 (2.7-35.1) µIU/ml respectively. However, there was no significant difference in the median (IQR) insulin level in the various groups (k = 4.968, p = 0.083).

 

Classification based on weight: the median (IQR) serum insulin levels of the LBW babies was 5.0 (2.5-34.0) µIU/ml, that of normal birth weight was 10.9 (3.1-35.0) µIU/ml while that of the macrosomic babies was 5.2 (2.0-36.0) µIU/ml. There was no statistically significant difference in the median (IQR) serum insulin levels in the various birth weight categories (k = 3.991, p = 0.136).

 

Weight for gestational age: the median (IQR) serum insulin levels of the AGA babies was 10.0 (3.0-35.1) µIU/ml, median (IQR) serum insulin levels of SGA babies was 5.0 (3.0-35.0) µIU/ml, and median (IQR) serum insulin levels of LGA babies was 5.2 (2.0-36.0) µIU/ml. There was no statistically significant difference in the median (IQR) serum insulin levels in the different groups of babies (k = 0.426, p = 0.808).

 

Serum insulin levels in the different categories of clinical diagnosis of the babies: the median (IQR) serum insulin level in different clinical diagnosis was summarized in Table 2. The median (IQR) serum insulin level among babies with probable sepsis was 24.0 (7.3-43.2) µIU/ml while the median (IQR) serum insulin level among babies with confirmed sepsis was 9.9 (3.5-34.0) µIU/ml. The median (IQR) serum insulin level was significantly higher in babies with probable sepsis (k = 9.348, p = 0.025) compared with other groups of babies. The median (IQR) serum insulin level was significantly lower among babies with birth asphyxia compared with babies without birth asphyxia (U = -2.082, p = 0.037).

 

Serum insulin measurements and relationship with blood glucose levels: the median (IQR) serum insulin level for the hyperglycemic babies was the highest: 36.0 (7.7-82.2) µIU/ml. The least median (IQR) serum insulin level was found among the hypoglycaemic babies: 5.6(2.0-35.3) µIU/ml. This difference was statistically significant (k = 9.072, p = 0.011). The median serum insulin in the different blood glucose categories was summarized in Table 3. There was weak positive correlation between the random blood glucose and the serum insulin levels of the babies (r = 0.197, p = 0.001). Figure 1 shows a scatter plot of the relationship between the serum insulin levels and random blood glucose of the babies.

 

 

The age of the babies in this study was significantly related to serum insulin levels as the babies that were older than seven days tend to have higher serum insulin levels. This is in agreement with the finding of Abdel-Azeem et al. [12] who also noted that serum insulin level tend to increase with the age of the babies. This is possibly because by seven days of life most mothers would have established breastfeeding. It is expected that with established or adequate feeding, the insulin secretion is increased to ensure adequate glucose uptake, use, and storage [3, 4]. The present study showed that serum insulin level in the babies with probable sepsis was significantly higher than that of the other groups of babies. To the best of the investigator's knowledge, no study was found to relate neonatal sepsis with elevated serum insulin. However, it is possible that since this study showed that probable sepsis was a risk factor for neonatal hyperglycaemia, serum insulin was most likely elevated as a compensatory mechanism to correct hyperglycaemia in those babies. In addition, serum insulin could possibly be elevated in babies with sepsis in reaction to the effect of stress and other counter-regulatory hormones like glucagon and cortisol [3-5].

 

These counter-regulatory hormones mobilise stored glucose from the liver and muscles which insulin tends to catabolise [3-5]. Could insulin be a biomarker of sepsis in newborn like C reactive protein and procalcitonin which are known biomarkers of sepsis [13]. This will be a good future research. Contrary to the investigator's expectation [14-16], there was no significant difference in the serum insulin levels of babies who were born low birth weight or macrosomic and those who had normal birth weight. Also, the serum insulin level was not significantly related to whether baby was small, appropriate or large for gestational age. These may be due to the small proportion of macrosomic, large for gestational age and small for gestational age babies compared to other groups of babies. Furthermore, in this study, the babies were recruited at different presenting ages when other factors such as inadequate lactation by mothers leading to reduced breastmilk intake must have set in which could have brought down the insulin level to normal, unlike the cord blood estimation of insulin level done in them at birth [14-16].

 

It was also shown in this study that the serum insulin level was not significantly related to the gestational age of the babies. This was against the finding of Wang et al. [17] in the United States of America who found that the serum insulin level was higher in preterm babies. The reason for this difference was because the babies recruited into the present study were recruited at different ages unlike the study by Wang et al. [17] in which babies were recruited at birth. Also, the proportion of preterm babies in this study was fewer compared to those of term babies. This study equally showed that serum insulin level was significantly lower in babies who had birth asphyxia. This was contrary to previous studies, most of which implicated birth asphyxia as a cause of high serum insulin levels [18-20]. This was possibly because a major portion of the recruited babies for this study were not admitted immediately after birth even when they had birth asphyxia, hence, the serum insulin most likely had reduced after an initial rise. The explanation for this is the negative feedback mechanism which is normally stimulated in the body to stop insulin secretion when it is elevated [3, 21]. Furthermore, when there is high level of insulin, it predisposes to hypoglycaemia which also turns off the insulin secretion thus reducing its level [3, 18].

 

The current study showed that babies with hyperglycaemia have significantly higher level of serum insulin more than the babies with normoglycaemia (36.0 versus 8.6µIU/ml). This was also supported by Verhoeven et al. [22] who also found elevated serum insulin in hyperglycaemic babies compared to normoglycaemic babies (30.8 versus 8.3µIU/ml). This shows that insulin does not constitute an important factor in the pathogenesis of hyperglycaemia at point of admission in most neonates, but it may be elevated as a compensatory mechanism to reduce the blood glucose back to normal [3]. Also, because there is elevated counter-regulatory hormones which is thought to be responsible for stress hyperglycaemia, there is associated relative insulin resistance [23].

 

 

Serum insulin level increases with increasing blood glucose in ill Nigerian babies at presentation to the hospital. Babies with asphyxia and sepsis particularly tend to have abnormal serum insulin at admission. Hyperinsulinaemia in ill babies may connote a compensatory mechanism to normalise abnormal blood glucose rather than playing significant role in its aetio-pathogenesis.

 

 

The authors declare no competing interests.

 

 

EOA was involved in study conception, sample collection, data analysis and he also wrote the first draft of the manuscript. BPK was involved in sample collection, data analysis and interpretation as well as critical review of manuscript. JBE was involved in data interpretation and critical review of manuscript. All authors read and approved the final version of the manuscript.

 

 

The authors hereby acknowledge the contributions of the clinicians and other health professionals who were involved in the management of the babies admitted into the Special Care Baby Unit, WGH, Ilesa. We also acknowledge the babies and the parents/guardians who participated in the study.

 

 

Table 1: obstetric characteristics and the babies' classification

Table 2: serum insulin level of babies with some clinical diagnosis

Table 3: serum insulin measurements and relationship with blood glucose levels

Figure 1: scatter plot of the relationship between serum insulin level and blood glucose of the study participants

 

 

1. Anusha HH, Hemachandra MD, Richard MC. Neonatal Hyperglycemia. Peds in review. 1999;20:16-24. Google Scholar


2. Sunehag AL, Haymond MW. Glucose extremes in newborn infants. Clin perinatol. 2002;29(2):245-60. PubMed | Google Scholar


3. Guyton AC, Hall JE. Textbook of medical physiology-10^th Edition. 2001. Elsevier Inc.


4. Ganong WF. Review of Medical Physiology-22nd edition. 2003. United States of America. McGraw-Hill Companies Inc.


5. Lilien LD, Rosenfield RL, Baccaro MM, Pildes RS. Hyperglycemia in stressed small premature neonates. J Pediatr. 1979;94(3):454-9. PubMed | Google Scholar


6. Dungan MK, Braithwaite SS, Preiser JC. Stress Hyperglycemia. Lancet. 2009;373(9677):1798-1807. PubMed | Google Scholar


7. Oyedeji GA. Socioeconomic and cultural background of hospitalized children in Ilesa. Nig J Paediatr. 1985;13:111-18. Google Scholar


8. Battaglia FC, Lubchenco LO. A practical classification of newborn infants by weight and gestational age. J Pediatr. 1967;71(2):159-63. PubMed | Google Scholar


9. Anday EK, Stanley CA, Baker L. Delivoria-Papadopoulos M. Plasma ketones in newborn infants: absence of suckling ketosis. J Pediatr. 1981;98(4):628-30. PubMed | Google Scholar


10. Volpe JJ. Neurology of the Newborn-4th Edition. 2001. Philadelphia. WB Saunders Company.


11. Halamek LP, Benaron DA, Stevenson DK. Neonatal hypoglycaemia, Part I: Background and definition. Clin Pediatr. 1997;36(12):675-80. PubMed | Google Scholar


12. Abdel-Azeem ME, Khalid AN, Gehan LA, Tahra, Ebtesam F, Ebtesam ME et al. Adiponectin, leptin and insulin levels at birth and in early postnatal life in neonates with hypoxic ischemic encephalopathy. J Diabetes Metab Disord. 2015;14(1): 87. PubMed | Google Scholar


13. Dima M, Lacob D, Marginean O, Lacob ER. New emerging biological markers of neonatal sepsis. J Res Med Sci. 2017;22:65. PubMed | Google Scholar


14. Hoegsberg B, Gruppuso PA, Coustan DR. Hyperinsulinemia in macrosomic infants of nondiabetic mothers. Diabetes Care. 1993;16(1):32-36. PubMed | Google Scholar


15. Simental-Mendía LE, Castañeda-Chacón A, Rodríguez-Morán M, Guerrero-Romero F. Birth-weight, insulin levels, and HOMA-IR in newborns at term. BMC Pediatrics. 2012;12:94-98. PubMed | Google Scholar


16. Hou R-L, Jin W-Y, Chen X-Y, Jin Y, Wang X-M, Shao J et al. Cord Blood C-peptide, Insulin, HbA1c, and Lipids Levels in Small- and Large-for-Gestational-Age Newborns. Med Sci Mon Int Med J Exp Clinl Res. 2014;20:2097-105. PubMed | Google Scholar


17. Wang G, Divall S, Radovick S, Paige D, Ning Y, Chen Z et al. Preterm Birth and Random Plasma Insulin Levels at Birth and in Early Childhood. JAMA. 2014;311(6):587-96. PubMed | Google Scholar


18. Mizumoto H, Uchio H, Yamashita S, Hata D. Transient neonatal hyperinsulinism with adaptation disorders: a report of three cases. J Pediatr Endocrinol Metab. 2015;28(3-4):337-40. PubMed | Google Scholar


19. Schultz K, Soltész G. Transient hyperinsulinism in asphyxiated newborn infants. Paediatr Hung. 1991; 31(1):47-52. PubMed | Google Scholar


20. Clark W, O'Donovan D. Transient hyperinsulinism in an asphyxiated newborn infant with hypoglycemia. Am J Perinatol. 2001;18(4):175-8. PubMed | Google Scholar


21. Koeslag JH, Saunders PT, Terblanche E. A reappraisal of the blood glucose homeostat which comprehensively explains the type 2 diabetes mellitus-syndrome X complex. J Physiol. 2003 Jun 1;549(Pt 2):333-46. Epub 2003 Apr 25. PubMed | Google Scholar


22. Verhoeven JJ, Brinker MD, Hokken-Koelega ACS, Hazelzet JA, Joosten KFM. Pathophysiological aspects of hyperglycaemia in children with meningococcal sepsis and septic shock: a prospective, observational cohort study. Critical care. 2011;15(1):R44. PubMed | Google Scholar


23. Marik PE, Raghavan M. Stress hyperglycaemia, insulin and immunomodulation in sepsis. Intensive care med. 2004;30(5):748-56. PubMed | Google Scholar