European Journal of Cardiovascular Medicine

Neonatal hypoglycemia remains one of the most frequent metabolic problems encountered in the immediate postnatal period and is a leading cause of neonatal monitoring, interventions, and neonatal intensive care unit (NICU) admissions. Infants of diabetic mothers (IDM) constitute a well-recognized high-risk group, with consistently reported higher rates of hypoglycemia compared with the general newborn population. In a systematic review focused on neonates born to diabetic mothers, neonatal hypoglycemia was identified as a common and clinically relevant complication, underscoring the need for structured screening and early management in this population [1].

The increased vulnerability of IDM to hypoglycemia is largely explained by altered fetal metabolic adaptation to maternal hyperglycemia. Chronic maternal hyperglycemia promotes fetal hyperinsulinemia, and after delivery the abrupt interruption of transplacental glucose supply in the presence of persistent hyperinsulinism predisposes the neonate to early and sometimes recurrent hypoglycemia. This pathophysiologic sequence is central to the clinical risk profile of IDM and forms the basis for early glucose surveillance protocols [2].

While hypoglycemia screening is routinely practiced in many neonatal units, considerable heterogeneity persists regarding eligibility criteria and screening intensity. A recent systematic review on infants eligible for hypoglycemia screening highlights ongoing variation in screening thresholds and target populations, reflecting differences in guideline interpretation and local practices [3]. In this context, defining the local incidence and clinical predictors of hypoglycemia among IDM is important for optimizing monitoring protocols, prioritizing high-risk infants, and minimizing avoidable separation of mother and newborn.

Several maternal and neonatal factors have been associated with neonatal hypoglycemia, including type of maternal diabetes, glycemic control, insulin therapy, prematurity, and abnormal fetal growth. Recent work has emphasized the value of multivariable risk prediction approaches to identify infants at greatest risk; however, models are often developed in retrospective cohorts and may not generalize across settings [4,5]. Prospective, setting-specific data are therefore needed to quantify incidence, characterize timing and severity, and determine predictors relevant to clinical care in Indian tertiary care settings.

Accordingly, this prospective observational study aims to determine the incidence and predictors of hypoglycemia in neonates born to diabetic mothers at SSPM Medical College and Lifetime Hospital, Padve, Sindhudurg, Maharashtra, over an 18-month period.

Objectives

1. To determine the incidence of hypoglycemia among neonates born to diabetic mothers during the early neonatal period (first 48–72 hours of life).
2. To identify maternal and neonatal predictors associated with hypoglycemia in neonates born to diabetic mothers.
3. To describe the timing, clinical profile (symptomatic versus asymptomatic; severe and/or recurrent episodes), and short-term outcomes of hypoglycemia including treatment required, NICU admission, and length of hospital stay..

Study design and setting This prospective observational study was conducted in the Department of Pediatrics/Neonatology at SSPM Medical College and Lifetime Hospital, Padve, Sindhudurg, Maharashtra. Study duration Eighteen months. Study population and sample size Consecutive live-born neonates delivered to mothers with diabetes (gestational diabetes mellitus or pregestational diabetes mellitus) were enrolled until a sample size of 200–250 infants of diabetic mothers (IDM) was achieved. Eligibility criteria All live-born IDM were eligible after written informed consent from parent/guardian. Neonates with major congenital malformations, suspected or confirmed inborn errors of metabolism identified early, those transferred out before completing minimum glucose monitoring, and those whose parents/guardians declined consent were excluded. Operational definitions Hypoglycemia was defined using IAP operational thresholds for at-risk neonates. Symptomatic hypoglycemia was defined as blood glucose <40 mg/dL in the presence of compatible clinical features (e.g., jitteriness, poor feeding, lethargy, apnea/cyanosis, hypotonia, seizures). For asymptomatic at-risk neonates, hypoglycemia thresholds were defined as blood glucose <25 mg/dL (<4 hours of age), <35 mg/dL (4–24 hours), <45 mg/dL (24–48 hours), and <60 mg/dL (>48 hours). Point-of-care (glucometer) testing was used for screening; low values were sent for laboratory confirmation (venous/plasma glucose) wherever feasible, without delaying treatment. Recurrent hypoglycemia was defined as two or more documented episodes after initial correction. Persistent hypoglycemia was considered when hypoglycemia continued beyond 72 hours or when higher glucose infusion rates were required, as per guideline-based triggers for further evaluation. Data collection Maternal and neonatal information was recorded in a structured case record form. Maternal variables included type of diabetes, treatment during pregnancy (medical nutrition therapy/oral hypoglycemics/insulin), available indicators of glycemic control (antenatal glucose records and/or HbA1c where available), obstetric complications, and mode of delivery. Neonatal variables included sex, gestational age, birth weight and growth category (SGA/AGA/LGA as per the unit chart), APGAR scores, need for resuscitation, time to initiation of feeding, associated clinical conditions (e.g., respiratory distress, hypothermia, suspected sepsis), and NICU admission details. Glucose monitoring protocol Glucose monitoring protocol: As infants of diabetic mothers are high-risk, blood glucose monitoring was performed using a pre-feed schedule consistent with Indian standard workflows, with early screening soon after birth. Blood glucose was measured at 2, 6, 12, 24, and 72 hours of life, and additionally whenever the neonate developed symptoms suggestive of hypoglycemia or if previous values were low. A scheduled 48-hour measurement was not included in the routine monitoring protocol; hypoglycemia beyond 24 hours was captured at the 72-hour assessment or during symptom-driven testing. [7] Clinical management All enrolled neonates received standard care as per institutional protocol. Management of hypoglycemia (early feeding support and intravenous dextrose when indicated) was provided by the treating team in accordance with guideline-based principles; the study did not alter clinical decisions. [6] Outcome measures The primary outcome was the incidence of hypoglycemia among IDM during the monitoring period. Secondary outcomes included time to first episode, lowest glucose value, recurrence, symptomatic status, requirement of treatment (feeding measures and/or IV dextrose), NICU admission, seizures (if any), and length of hospital stay. Statistical analysis Data were entered in Microsoft Excel and analyzed using appropriate statistical software (SPSS/R). Categorical variables were summarized as frequency and percentage, and continuous variables as mean (SD) or median (IQR) depending on distribution. Incidence was reported with 95% confidence interval. Comparisons between neonates with and without hypoglycemia were performed using Chi-square/Fisher’s exact test for categorical variables and Student’s t-test/Mann–Whitney U test for continuous variables. Multivariable logistic regression was used to identify independent predictors of hypoglycemia; results were expressed as adjusted odds ratios with 95% confidence intervals. A p-value <0.05 was considered statistically significant. Ethical considerations Institutional Ethics Committee approval was obtained prior to initiation. Written informed consent was taken from parents/guardians and confidentiality was maintained using anonymized identifiers.

1. Study recruitment and participant profile

A total of 220 consecutive neonates born to diabetic mothers were included in the analysis. Most mothers had gestational diabetes mellitus (80.0%), and 38.2% required insulin during pregnancy; the mean maternal HbA1c was 6.58 ± 0.86% (Table 1A). The neonates had a mean gestational age of 38.04 ± 1.58 weeks and mean birth weight of 2.93 ± 0.44 kg; 70.0% were appropriate for gestational age, and feeding was initiated at a mean of 85.52 ± 27.85 minutes after birth (Table 1B).

Table 1A. Maternal characteristics stratified by neonatal hypoglycemia status (n=220)

Data are presented as n (%) unless otherwise specified.

Table 1B. Neonatal characteristics stratified by hypoglycemia status (n=220)

Data are presented as n (%) unless otherwise specified.

2. Incidence of neonatal hypoglycemia

During the predefined monitoring period as per Indian operational thresholds for at-risk neonates, 69 of 220 infants of diabetic mothers developed at least one episode of hypoglycemia, giving an incidence of 31.4% (95% CI 25.6–37.8). The incidence estimate is presented along with subsequent clinical pattern and outcomes in Table 2. The majority of first hypoglycemic episodes were detected within the first 12 hours of life, with a peak at 6 hours (Figure 1).

Table 2. Incidence, timing, clinical pattern, management, and short-term outcomes among neonates with hypoglycemia (n=69)

Data are presented as n (%) unless otherwise specified. Incidence denominator is the total cohort (n=220); all other parameters are among neonates with hypoglycemia (n=69).

Fgure1. Timing of first detected hypoglycemic episode among neonates born to diabetic mothers

The line graph shows the distribution of the first detected hypoglycemic episode, with a clear peak at 6 hours of life, followed by a decline thereafter.

3. Predictors of neonatal hypoglycemia

On bivariate analysis, several maternal and neonatal variables showed a significant association with neonatal hypoglycemia. Maternal factors significantly associated with hypoglycemia included type of diabetes, antenatal treatment modality, and higher mean HbA1c levels. Among neonatal factors, large-for-gestational-age (LGA) status and longer time to initiation of first feed were significantly associated with hypoglycemia. Variables such as mode of delivery, sex of the neonate, gestational age, prematurity, respiratory distress, and hypothermia were not significantly associated with hypoglycemia on unadjusted analysis (Tables 1A and 1B).

To identify independent predictors, variables with clinical relevance and/or p values <0.20 on bivariate analysis were entered into a multivariable logistic regression model. After adjustment, higher maternal HbA1c levels and LGA status demonstrated a borderline independent association with neonatal hypoglycemia. Other variables, including type of maternal diabetes, insulin use during pregnancy, prematurity, and time to first feed, were not independently associated with hypoglycemia in the adjusted model. The results of the multivariable analysis are presented in Table 3.

Table 3. Multivariable logistic regression analysis identifying predictors of neonatal hypoglycemia (n=220)

OR: odds ratio; CI: confidence interval; GDM: gestational diabetes mellitus.

4. Short-term outcomes by hypoglycemia status

Neonates who developed hypoglycemia had a higher requirement for escalation of care, reflected by greater need for intravenous dextrose therapy and NICU admission. Among hypoglycemic neonates, 26/69 (37.7%) required intravenous dextrose, while the remaining 43/69 (62.3%) were managed with feeding measures alone. NICU admission was required in 28/69 (40.6%) hypoglycemic neonates compared with 35/151 (23.2%) neonates without hypoglycemia, and this difference was statistically significant (p = 0.009). No deaths were recorded in the cohort during the hospital stay. Short-term outcomes stratified by hypoglycemia status are summarized in Table 4.

Table 4. Short-term outcomes stratified by hypoglycemia status (n=220)

Data are presented as n (%). Intravenous dextrose is a hypoglycemia-directed intervention and was not applicable in the non-hypoglycemia group.

In this prospective observational study of 220 infants of diabetic mothers (IDM), the incidence of neonatal hypoglycemia during the early neonatal monitoring period was 31.4%. This incidence is consistent with prior reports indicating that hypoglycemia remains a frequent metabolic complication in IDM, though the reported rates vary widely across studies due to differences in operational thresholds, timing/frequency of glucose monitoring, feeding practices, and case-mix (pre-gestational diabetes vs GDM, glycemic control, prematurity, and fetal growth patterns) [1,10,14]. The burden observed in our cohort reinforces the need for structured surveillance among IDM as recommended in standard care pathways and Indian guidance for at-risk neonates [6,7]. A key clinical finding in our cohort was the early timing of hypoglycemia, with the peak detection at 6 hours and the majority of first episodes occurring within the first 12 hours of life. This pattern aligns with the established pathophysiology in IDM, where fetal hyperinsulinism persists after delivery while maternal glucose supply is abruptly interrupted, resulting in early postnatal glucose instability [2]. Similar early clustering of events in IDM has been described in prior prospective and observational cohorts, supporting the practice of early and pre-feed glucose checks in this high-risk group [10,14]. From a service perspective, this timing emphasizes the importance of immediate postnatal feeding support and close monitoring during the first day of life—an area where adherence to protocolized screening is likely to yield the greatest benefit [6,7]. With regard to predictors, bivariate analysis showed significant associations between neonatal hypoglycemia and maternal diabetes type, antenatal treatment modality, higher maternal HbA1c, LGA status, and longer time to initiation of feeds. In the multivariable model, higher HbA1c and LGA status demonstrated a borderline independent association with hypoglycemia. The directionality of these associations is biologically plausible. Maternal hyperglycemia increases fetal insulin secretion and fat deposition, predisposing to both LGA status and early neonatal hypoglycemia [2]. Prior literature supports an association between poor maternal glycemic control and neonatal metabolic complications; HbA1c has been evaluated as a surrogate of glycemic exposure, though predictive performance varies by population and timing of measurement [12]. Our finding of a borderline adjusted association may reflect limited power for some strata (e.g., type 1 diabetes subgroup) and collinearity between glycemic control, insulin use, and fetal overgrowth—factors that are mechanistically linked and often move together clinically. The association between LGA status and neonatal hypoglycemia observed in our cohort is consistent with the broader evidence that fetal overgrowth is a marker of intrauterine hyperinsulinism and postnatal glucose instability. While much of the literature on LGA and hypoglycemia involves IDM, studies in LGA infants of non-diabetic mothers also demonstrate a higher risk of hypoglycemia, underscoring that fetal size and metabolic programming can independently signal risk even when maternal diabetes is absent or unrecognized [11]. Therefore, LGA status remains a clinically useful bedside risk marker, especially in settings where detailed maternal glycemic metrics may be incomplete. In our cohort, longer time to initiation of feeding was associated with hypoglycemia on bivariate analysis, although it did not remain independently significant after adjustment. Early feeding is a modifiable factor in preventing or attenuating transitional hypoglycemia in at-risk neonates and is embedded in standard workflows [6,7]. The attenuation of effect in the multivariable model may reflect that delayed feeding co-occurs with other drivers of NICU transition or observation (e.g., early respiratory issues, post-caesarean logistics, maternal recovery), and thus may act as a mediator rather than a standalone predictor. Nonetheless, from a practical standpoint, the observed association supports reinforcing early feed initiation, lactation support, and structured monitoring during the first few hours of life. Clinically, hypoglycemia in our cohort was not only frequent but also consequential in terms of escalation of care. Approximately two-fifths of hypoglycemic neonates required NICU admission, and over one-third required intravenous dextrose. This is broadly in keeping with the recognition that IDM frequently require enhanced surveillance and intervention compared with non-IDM neonates, even when many episodes are detected before severe symptoms occur [1,10,14]. Importantly, more than one-third of hypoglycemic neonates in our cohort were symptomatic, highlighting that reliance on symptoms alone would miss a substantial proportion of cases, while also risking delayed treatment in those who do become symptomatic. This aligns with screening-focused approaches that target at-risk groups rather than symptom-triggered testing alone [3,6,7,15]. The question of symptomatic versus asymptomatic hypoglycemia has ongoing relevance because neurodevelopmental risk appears to relate to severity, recurrence, and duration of low glucose rather than symptoms alone. Recent narrative reviews and evolving evidence highlight that neonatal hypoglycemia remains an important and complex risk factor for neurodevelopmental outcomes, and that improving early recognition and management is central to reducing avoidable harm [13]. While our study focused on short-term outcomes, the proportion of severe episodes and the early clustering of events underscore the need for robust early surveillance and timely correction as recommended in standard protocols [6,7]. In other clinical contexts, symptomatic hypoglycemia has been associated with worse outcomes than asymptomatic events, which conceptually supports the emphasis on preventing progression to symptomatic or severe episodes, though neonatal-specific inference must be drawn from neonatal literature [16]. Our results also support the value of risk stratification beyond a “one-size-fits-all” approach. Prediction modeling work in neonatal hypoglycemia has demonstrated that combining maternal and neonatal variables can improve identification of high-risk infants, though models are often retrospective and may not generalize across practice settings [4]. Biomarkers such as cord C-peptide have also been explored as early indicators of fetal hyperinsulinism and subsequent hypoglycemia risk in IDM, with promising findings that may supplement clinical risk factors where feasible [9]. In resource-constrained settings, however, pragmatic predictors such as LGA status, maternal glycemic control indicators, and early feeding patterns may be more immediately actionable than biomarker-based strategies. Strengths and limitations The strengths of this study include its prospective design, inclusion of a clinically relevant IDM cohort, and systematic early monitoring consistent with standard workflows for at-risk neonates. Limitations include a single-center design, potential variability in confirmatory laboratory glucose testing for low point-of-care readings, and the possibility of residual confounding due to interrelated predictors (HbA1c, insulin use, and fetal size). In addition, the study evaluated short-term hospital outcomes; longer-term neurodevelopmental follow-up would be required to relate hypoglycemia phenotype (severity/recurrent episodes) to developmental endpoints [13]. Implications for practice The observed incidence and early timing of hypoglycemia support maintaining structured early screening in IDM, with particular vigilance during the first 12 hours of life. LGA status and poorer maternal glycemic control should prompt heightened surveillance and early feeding reinforcement. These findings align with standard guideline-based approaches for at-risk neonates and support strengthening protocol adherence to reduce NICU admissions and the need for intravenous therapy where preventable [6,7,15].

Our Neonatal hypoglycemia occurred in nearly one-third of infants of diabetic mothers, with most first episodes detected within the first 12 hours of life. Large-for-gestational-age status and higher maternal HbA1c showed an association with hypoglycemia, and affected neonates more frequently required intravenous dextrose and NICU admission. These findings support strict early feeding and protocol-based glucose monitoring in this high-risk group

[1] Alemu, B. T., Olayinka, O., Baydoun, H. A., Hoch, M., & Akpinar-Elci, M. (2017). Neonatal hypoglycemia in diabetic mothers: A systematic review. Current Pediatric Research, 21(1).

[2] Eidelman, A. I., & Samueloff, A. (2002). The pathophysiology of the fetus of the diabetic mother. Seminars in Perinatology, 26(3), 232–236.

[3] O’Brien, M., Gilchrist, C., Sadler, L., Hegarty, J. E., & Alsweiler, J. M. (2023). Infants eligible for neonatal hypoglycemia screening: A systematic review. JAMA Pediatrics, 177(11), 1187–1196.

[4] Wu, T., Huang, Y. Y., Song, W., Redding, S. R., Huang, W. P., & Ouyang, Y. Q. (2023). Development of a prediction model for neonatal hypoglycemia risk factors: A retrospective study. Frontiers in Endocrinology, 14, 1199628.

5]. Diggikar, S., Trif, P., Mudura, D., Prasath, A., Mazela, J., Ognean, M. L., ... & Galis, R. (2024). Neonatal Hypoglycemia and Neurodevelopmental Outcomes—An Updated Systematic Review and Meta-Analysis. Life, 14(12), 1618.

[6] Manoj, V. C., Jajoo, M., & Rao, N. (2022). Neonatal hypoglycemia. Indian Academy of Pediatrics. (Standard Treatment Guidelines 2022).

[7] Department of Health Research, Ministry of Health and Family Welfare, Government of India. (2020, July). Standard Treatment Workflow (STW): Neonatal hypoglycemia (ICD-10-P70.4). Indian Council of Medical Research.

[8]. Pildes, R. S., Cornblath, M., Warren, I., Page-El, E., di Menza, S., Merritt, D. M., & Peeva, A. (1974). A prospective controlled study of neonatal hypoglycemia. Pediatrics, 54(1), 5-14.

9. Saber, A. M., Mohamed, M. A., Sadek, A. A., & Mahmoud, R. A. (2021). Role of umbilical cord C-peptide levels in early prediction of hypoglycemia in infants of diabetic mothers. BMC pediatrics, 21(1), 85.
10. Maayan-Metzger, A., Lubin, D., & Kuint, J. (2009). Hypoglycemia rates in the first days of life among term infants born to diabetic mothers. Neonatology, 96(2), 80-85.
11. Schaefer-Graf, U. M., Rossi, R., Bührer, C., Siebert, G., Kjos, S. L., Dudenhausen, J. W., & Vetter, K. (2002). Rate and risk factors of hypoglycemia in large-for-gestational-age newborn infants of nondiabetic mothers. American journal of obstetrics and gynecology, 187(4), 913-917.
12. Kumar, N., Kumar, P., Harris, N., Monga, R., & Sampath, V. (2020). Impact of maternal HbA1c levels≤ 6% and race in nondiabetic pregnancies on birthweight and early neonatal hypoglycemia. The Journal of pediatrics, 227, 121-127.
13. De Rose, D. U., Perri, A., Maggio, L., Salvatori, G., Dotta, A., Vento, G., & Gallini, F. (2024). Neonatal hypoglycemia and neurodevelopmental outcomes: Yesterday, today, tomorrow. European Journal of Pediatrics, 183(3), 1113-1119.
14. Agrawal, R. K., Lui, K., & Gupta, J. M. (2000). Neonatal hypoglycaemia in infants of diabetic mothers. Journal of paediatrics and child health, 36(4), 354-356.
15. Enni, J. B. A., Narasimhan, S. R., Huang, A., & Jegatheesan, P. (2023). Screening and diagnosis of neonatal hypoglycaemia in at-risk late preterm and term infants following AAP recommendations: a single centre retrospective study. BMJ Paediatrics Open, 7(1), e001766.
16. Cardona, S., Gomez, P. C., Vellanki, P., Anzola, I., Ramos, C., Urrutia, M. A., ... & Umpierrez, G. E. (2018). Clinical characteristics and outcomes of symptomatic and asymptomatic hypoglycemia in hospitalized patients with diabetes. BMJ open diabetes research & care, 6(1).