This was a retrospective single tertiary hospital-based cohort study at Zhongnan Hospital of Wuhan University, from 2018 to 2022, in two departments: Pediatric and Obstetrics & Gynecology. All Infants exposed to gestational diabetes mellitus and their gestational diabetic mothers were recruited in the study under the following inclusion and exclusion criteria:

a) Infants, Inclusion criteria: Infants aged from 0 - 12 months; Infants born to mothers with a confirmed diagnosis of gestational diabetes mellitus; Infants with available medical records and neurodevelopmental assessment scores; Infants whose mothers did not use any medications or substances known to affect neurodevelopment during pregnancy; Infants whose mothers received consistent prenatal care throughout their pregnancy; Infants born to mothers who did not have any other significant medical conditions or complications during pregnancy.

Exclusion Criteria: Infants with a known genetic or congenital abnormality that could independently affect neurodevelopment; Infants with a history of significant neonatal complications that could impact neurodevelopment; Infants with incomplete or unavailable medical records and neurodevelopmental assessments; Infants with a history of significant medical conditions or complications during the prenatal or postnatal period may confound the relationship between maternal gestational diabetes and neurodevelopmental outcomes; infants born to mothers with pre-existing diabetes; Infants born to mothers who used medications or substances known to affect neurodevelopment during pregnancy.

b) Gestational Diabetic mothers, Inclusion Criteria: Mothers with a confirmed diagnosis of gestational diabetes during pregnancy; Mothers who received prenatal care and regular monitoring for gestational diabetes; Mothers who delivered a live-born infant; Mothers who did not have any significant medical conditions or complications during pregnancy, apart from gestational diabetes; Mothers who did not use medications or substances known to affect neurodevelopment during pregnancy. Exclusion Criteria: Mothers with pre-existing diabetes (type 1 or type 2) prior to pregnancy; Mothers with other significant medical conditions or comorbidities that could independently impact neurodevelopment; Mothers who used medications or substances known to affect neurodevelopment during pregnancy; Mothers with a history of significant mental health disorders; Mothers with incomplete or unavailable medical records. Figure 1 provides details regarding inclusions and exclusion criteria of the study participants.

Clinical data assessment: Neurodevelopmental outcome and mental evaluation results were obtained from the record of the assessment of infants using the Denver II Developmental Screening Test (DDST-II). The Denver II Developmental Screening Test (DDST-2) is used to screen children’s development from the age of 0 month to 6 years. It consists of 125 items, and the development of a child is measured based on these 125 items. The test usually takes 10-20 minutes to perform. Each test item is scored as pass, fail, refused or no opportunity. In the Denver II Developmental Screening Test for infants, the categories of normal, abnormal, suspicious, and unable to judge were used to assess a child’s development across four domains: personal-social, fine motor-adaptive, language, and gross motor skills.

Ethics consideration

Ethical approval was sought from Ethics Committee of Zhongnan Hospital of Wuhan University with a number (Kelun 2022042K) and exempted from signing the informed consent form.

SPSS version 27 was used for statistical analysis. The independent samples t-test was used to compare the means of normally distributed data between the two groups, while the Mann-Whitney test was employed for continuous data that were not normally distributed. The chi-square test was used to assess the association between gestational diabetes mellitus and neurodevelopmental outcomes in infants. Results were indicated statistically significant for the p-value less than 0.05.

Regarding developmental assessment, of the 1430 subjects, 1370 (95.80%) showed normal results, 8 (0.56%) had abnormal results, 44 (3.08%) were categorized as suspicious, and 8 (0.56%) had results that were unable to be judged based on DDST findings. Table 2 presents the DDST results across all four domains.

DDST: Denver Developmental Screening Test.

There was no statistically significant difference in the probability of abnormal, suspicious, or unable to judge findings between male and female infants (p = 0.163). Similarly, no significant association was found between abnormal findings and maternal age (p = 0.193) or gestational age (p = 0.566). Delivery methods (spontaneous vaginal delivery, cesarean delivery, and forceps delivery) also showed no significant association (p = 0.702). However, there was a significant association with parity, where primiparous mothers had a higher likelihood of neurodevelopmental impairments in their infants’ DDST results compared to multiparous mothers (p < 0.001). Infants with normal head size and microcephaly showed significantly more abnormal DDST results (p = 0.002). The probability of abnormal DDST findings was highest in infants younger than six months, especially around three months of age (p < 0.001). Regarding anterior fontanelle status, infants with normal, sunken, closed, or not closed fontanelles showed significantly more abnormal DDST results than those with enlarged anterior fontanelles (p < 0.001). Birth length (p = 0.492), birth weight (p = 0.081), and weight level (p = 0.394) showed no significant effect on abnormal DDST findings. Table 3 summarizes the statistical relationships between maternal and infant variables and the overall DDST results.

The overall mental responses of infants were categorized as follows: 124 (8.70%) showed normal responses, 208 (14.50%) had good responses, 841 (58.80%) had acceptable responses, and 64 (4.50%) had poor responses. On the other hand, a significant number of infants exhibited abnormal behaviors, including annoying noise (64, 4.50%), incessant annoying noise (30, 2.10%), crying and arguing (2, 0.10%), incessant crying and arguing (154, 10.80%), crying and making noise (5, 0.30%), and others (2, 0.10%). Table 4 presents a summary of these mental responses.

There was a significant association between poor mental responses and infants with birth weights ranging from 1 - 1.5 kg, 1.5 - 2.5 kg, and 2.5 - 4 kg (p < 0.001). Infants with normal or microcephalic head circumferences also showed a statistically significant association with poor mental responses (p < 0.001). Infants with low weight levels had poorer responses compared to those in the Above Middle, Average Positive, Average Negative, and Below Middle weight categories, with a significant association (p < 0.001). Regarding the anterior fontanelle, infants with a normal anterior fontanelle had poor mental responses with a statistically significant association (p < 0.001). However, no statistically significant associations were found between gender (p = 0.354), maternal age (p = 0.362), gestational age (p = 0.663), birth length (p = 0.978), or delivery methods (p = 0.616) and general mental responses in infants. Table 5 provides a summary of the statistical relationships between maternal and infant factors and general mental responses in infants.

The present study found that gestational diabetes mellitus (GDM) has a significant impact on both the neurodevelopmental and mental development of infants. In our sample, 68.65% of the participants were affected by GDM. These findings on the prevalence of GDM differ from those reported in previous studies. Binbin Yin et al. reported that the prevalence of gestational diabetes mellitus (GDM) after the COVID-19 pandemic changed to 21.46%, 19.81%, and 18.48% in the years 2019, 2020, and 2021, respectively [32] . Wei et al. in the retrospective study which included 3211 singletons of mothers with GDM at the Shanghai First Maternity and Infant Hospital between January 2017 and December 2019 found that the prevalence of GDM was 7% [33] . However, Chenghan Gao et al. found an incidence of GDM in mainland China at 14.8% [34] . The prevalence of gestational diabetes mellitus (GDM) has been consistently rising over the past few decades and is expected to keep increasing [35] . In China, the risk factors for gestational diabetes mellitus (GDM) include advanced maternal age, being overweight or obese before pregnancy, excessive weight gain during pregnancy, certain dietary habits, smoking, exposure to secondhand smoke, higher parity, a previous history of GDM, a history of fetal macrosomia, polycystic ovarian syndrome, a family history of diabetes, lower levels of education, and limited health knowledge [35] . Globally, Behboudi-Gandevani et al. found that the overall pooled prevalence of gestational diabetes mellitus (GDM), regardless of screening thresholds, was 4.4%. However, when the diagnostic thresholds based on IADPSG criteria were applied, the pooled prevalence rose to 10.6%. A similar pattern was observed, with significant regional variation in GDM prevalence. East Asia had the highest prevalence at 11.4%, while Australia had the lowest at 3.6%, regardless of the screening criteria used [36] . In Switzerland, a retrospective cohort study showed a notable increase in the prevalence of gestational diabetes mellitus (GDM) when the IADPSG criteria were applied. The prevalence rose from 3.3% using a 2-step screening approach to 11.8% with a 1-step screening method [37] . Early identification of high-risk individuals is essential, as it enables the implementation of preventive measures that can help reduce the incidence of gestational diabetes mellitus (GDM) and improve perinatal outcomes. In China, prioritizing prevention and intervention for GDM is critical. Effective strategies for managing and preventing GDM include lifestyle changes such as improving diet and increasing physical activity. The GDM One-day Care Clinic, established in 2011, has become a successful model for managing GDM. It offers patient education on the basics of GDM, dietary modifications, exercise plans, weight control, and blood glucose monitoring. This model has been widely adopted in hospitals and maternal and child health centers throughout China [35] .

In this study, the overall DDST II results showed 1370 (95.80%) normal, 8 (0.60%) abnormal, 44 (3.10%) suspicious, and 8 (0.60%) unable to judge. Our results differed from those of Seyhan et al., who, in their study of 1000 subjects, reported that 741 (74.1%) had normal findings, 140 (14%) were considered at risk, and 119 (11.9%) had abnormal findings based on the DDST results [38] . Likewise, in their study, Meltem et al. found that the DDST-II results for Turkish children indicated that 82.1% had normal findings, 10.7% had questionable results, and 7.1% were categorized as abnormal [39] . In contrast, the study by Ibrahim et al. found that 134 subjects (38.5%) had normal results, 136 subjects (39%) had abnormal results, and 78 subjects (22.4%) had questionable results on the DDST [40] .

Gender showed no statistically significant association with neurodevelopmental outcomes. Similarly, a study by Ozkan et al. reported no differences between genders concerning DDST results for children aged 3 to 60 months [41] . While it is widely recognized that developmental disorders are more common in males, our study differed from Seyhan et al.’s study, which found that males were more likely to have suspicious and abnormal DDST findings compared to females [38] . Similarly, Xiang et al. discovered that male infants born to mothers with gestational diabetes were at a higher risk for certain neurodevelopmental issues, such as cognitive delays and behavioral problems, compared to female infants. This study suggests that gender differences may influence the neurodevelopmental outcomes of infants born to mothers with gestational diabetes [9] .

Infants with normal anterior fontanelles showed suspicion, abnormal findings, and an inability to judge neurodevelopmental outcomes compared to those with abnormal anterior fontanelles. Our study aligns with previous research, which suggests that despite having normal anterior fontanelles, these infants may still raise concerns for potential neurodevelopmental issues due to underlying metabolic and environmental factors linked to gestational diabetes. Infants born to mothers with gestational diabetes are exposed to maternal hyperglycemia during crucial periods of fetal development, which can have lasting effects on neurodevelopment, regardless of the status of the anterior fontanelle. Additionally, subtle neurodevelopmental delays may not always be evident right away and may require long-term monitoring to detect cognitive, motor, or behavioral delays that might not be apparent during early assessments. Furthermore, the impact of gestational diabetes on neurodevelopment can vary among infants, and those with normal anterior fontanelles may still be at risk for developmental challenges due to individual genetic factors, environmental influences, and the severity of maternal diabetes [42] - [44] .

Infants born at a gestational age between 32 and 42 weeks showed a higher number of abnormal and suspicious neurodevelopmental outcomes. A similar study found that neurodevelopmental impairments were most common at 32 weeks (the earliest gestational age), gradually decreased until 41 weeks, and were also more prevalent at 37 - 38 weeks (early term) compared to 39 - 40 weeks [45] . However, in their study, Hadden DR et al. found that early-onset gestational diabetes may have a greater impact on fetal growth and neurodevelopment than late-onset gestational diabetes [46] . Similarly, Ju Hyun Jin et al. in their study found that, while the highest rates of morbidities were observed in children born at the earliest gestational ages, those born moderate-to-late preterm also faced significantly more adverse neurodevelopmental outcomes compared to full-term children [47] .

The mode of delivery (spontaneous vaginal, cesarean, or forceps delivery) showed no differences in neurodevelopmental outcomes in infants. Our results align with a previous study by Hanne Trap Wolf et al., which found that the mode of delivery was not linked to neurodevelopmental impairments at 2 years of age [48] . Similarly, Yi-Ya Huang et al. used propensity scores and unit policies but were unable to establish a statistically significant association between the mode of delivery and neurodevelopmental outcomes at 2 years of corrected age in perceivable singleton infants [49] . Moreover, Seyhan et al. in their study found no significant effect on DDST results regarding mode of delivery [38] . A possible explanation for this observation is that neurodevelopmental outcomes might not appear right away and could need long-term monitoring to properly evaluate.

Infants born from gestational Primiparous mothers showed greater neurodevelopmental impairments compared to multiparous. This could be because first-time mothers often experience longer labor and higher rates of instrumental deliveries (like forceps or vacuum), which may lead to birth trauma and subsequent neurodevelopmental challenges. Additionally, first-time parents tend to experience more stress and anxiety, which can adversely affect infant care and early development. However, Abubakar et al. suggested that with each additional pregnancy, the risk of negative developmental outcomes in later-born children increases [50] .

Infants with normal head size and microcephaly exhibited more significant neurodevelopmental impairments compared to those with macrocephaly. A similar study found that a smaller head circumference was linked to lower cognitive scores during follow-up [51] . Similarly, Guellec I et al. in their longitudinal study that followed very preterm infants and found that head circumference at birth was strongly correlated with neurodevelopmental outcomes. Infants with smaller head circumference had higher rates of cognitive and motor delays [52] . Our study reinforces the importance of early head circumference as a predictor of neurodevelopmental outcomes. Moreover, careful monitoring of fetal growth—including head circumference measurements—can help identify infants born from gestational diabetic mothers who are at risk for later cognitive and motor impairments, allowing for early interventions and neurodevelopmental support.

Neurodevelopmental impairments were found to be statistically significant at a very early age, particularly before six months. A similar study by Bersain et al. indicated that neurodevelopmental outcomes were impacted as early as 3½ months in infants born to mothers with GDM. This suggests that more stringent diabetes control measures during pregnancy may be necessary to ensure the best neurodevelopmental outcomes. However, previous studies have reported that as preschool children get older, there tends to be an increase in abnormal developmental performance on the DDST [38] [53] .

In terms of mental response, there were no differences observed based on the infants’ gender. However, previous studies have indicated that female infants generally exhibit more advanced emotional regulation and social skills than males. Girls tend to engage more in social interactions and demonstrate higher levels of empathy, but they are also more likely to develop anxiety and depression later in life [54] [55] . However, Maenner et al. found that male infants are approximately four times more likely to be diagnosed with ASD than females. Male infants are at a greater risk for neurodevelopmental disorders like autism spectrum disorder (ASD) and ADHD, which can influence mental evaluation outcomes [56] . On the other hand, Tamis-LeMonda et al. emphasized that parents tend to interact with boys and girls in distinct ways, which can influence their developmental paths and later evaluations [57] . This suggests that the relationship between gender and parenting styles can impact mental evaluation outcomes. Additionally, these findings highlight that gender differences in mental evaluations are shaped by a mix of biological, cognitive, social, and environmental factors.

Infants with microcephaly showed significantly better and worse mental outcomes in mental response compared to those with macrocephaly. Ornoy et al. reported that these children are more prone to exhibiting symptoms of ADHD and emotional challenges, regardless of their head circumference at birth [58] . But one study has linked larger head circumference with better emotional regulation and fewer behavioral issues, but these advantages don’t seem to last into later childhood, indicating a complex relationship between prenatal growth patterns and postnatal outcomes [59] . Therefore, these findings suggest that while a larger head circumference in infants born to gestational diabetic mothers may offer some early mental benefits, the overall mental risk remains high due to the intricate interaction between prenatal metabolic conditions and postnatal environmental factors.

The mode of delivery showed no differences in mental response. Similar studies have found no statistically significant differences in mental outcomes between infants born via cesarean section and those born vaginally to mothers with GDM [60] [61] . However, Huang et al. found that children born via cesarean section to mothers with GDM were more likely to show behavioral problems and emotional difficulties than those born vaginally [62] . This link may be influenced by factors like changes in gut microbiota due to cesarean delivery, which can affect brain development and behavior.

Infants with a normal anterior fontanelle were found to have mental impairments compared to those with enlarged, small, closed, or open anterior fontanelles. A similar study by Kumar, M., & Dewan, P., reported that the size of the anterior fontanelle can vary significantly in healthy infants, and deviations from the typical size are more likely to signal potential issues rather than the normal size itself causing impairments [63] . This suggests that neurodevelopmental anomalies in infants with normal fontanelles could be linked to other underlying conditions that may not be immediately visible through fontanelle size alone.

Infants with Average positive and Average negative weight levels showed a higher likelihood of neurodevelopmental concerns compared to those with Above Middle, Below Middle, and Low weight levels. Our findings align with Taveras et al., who found that infants with an average positive weight level experience rapid weight gain in the first six months, which is linked to a higher risk of obesity in later childhood and may affect cognitive and behavioral development [64] . Whilst, Belsky et al. found that higher birth weight was associated with better initial cognitive performance, but rapid postnatal weight gain was linked to higher risks of behavioral problems [65] . Infants with weight above the middle range typically do not show a high rate of suspicious neurodevelopmental outcomes unless there are underlying health issues. Overweight infants might face metabolic challenges that could indirectly affect development.

Gestational age showed no significant difference for mental response in infants. However previous studies [66] [67] , reported that, while full-term infants (37 - 42 weeks) of gestational diabetic mothers showed generally better outcomes than preterm infants, they still had a higher likelihood of poor mental outcomes compared to infants of non-diabetic mothers. Infants born late preterm and early term had higher rates of ADHD and other behavioral problems compared to those born at 39 - 42 weeks.

Infants with birthweights 1 - 1.5 kg, 1.5 - 2.5 kg and 2.5 - 4 kg showed significant poor mental outcomes compared to infants at birthweight <1 kg and >4 kg. Our study was consistent with previous studies, whereby, The EPIPAGE-2 study found that VLBW infants born to gestational diabetic mothers had higher rates of cognitive and behavioral problems at 5 years of age compared to infants with normal birth weights [68] . But, our study, differed from a study by Crume et al., who found that macrosomic infants had higher rates of metabolic syndrome, which was associated with lower cognitive function and increased behavioral problems in early childhood [69] .

Infants with an AGA birth length demonstrated poorer mental outcomes compared to those with SGA or LGA birth lengths. One possible explanation for this finding in our study is that AGA infants may still experience unrecognized perinatal complications or less-than-ideal postnatal environments, both of which could negatively impact their neurodevelopment. Previous research indicates that factors such as socio-economic status, parental education, and the quality of postnatal care play a significant role in child development outcomes. For example, AGA infants might be more vulnerable to the negative effects of inadequate postnatal care and lower socio-economic conditions, whereas SGA or LGA infants may benefit from more focused interventions [70] . Another possible explanation is that some AGA infants may be at the lower end of the AGA spectrum, closer to the 10th percentile. These infants might experience less-than-ideal intrauterine growth, which isn’t severe enough to be classified as SGA but could still influence their neurodevelopment. This mild form of growth restriction might be enough to affect cognitive outcomes, even if it doesn’t meet the criteria for an SGA classification. One study found that SGA infants, who generally have shorter birth lengths, were more likely to have lower IQ scores at the age of 5 compared to those with birth lengths appropriate for their gestational age (AGA). Specifically, SGA infants experienced a notable decline in IQ and had a higher risk of scoring below 85. On the other hand, LGA infants did not exhibit significant differences in neurodevelopmental outcomes when compared to AGA infants [71] .

Infants with low birth weights demonstrated notably poorer mental outcomes. A similar study by Chuhao Xi et al. found that low birth weight infants are generally at a higher risk for negative mental evaluation results [72] . This increased risk is often linked to factors such as undernutrition, health complications during the prenatal and early postnatal periods, and less favorable growth environments. As a result, these infants are more likely to experience developmental delays and cognitive impairments compared to those with higher birth weights.