The incidence and prevalence of type 1 diabetes mellitus (T1DM) are increasing worldwide (1, 2). There were 8.4 million people with T1DM in 2021, and the International Diabetes Federation (IDF) projects this to rise to 17 million in 2040 (2). With a peak incidence at 10–14 years, more women of reproductive age are expected to have T1DM. The prevalence of T1DM in pregnancy is, therefore, expected to increase. Indeed, the incidence of pre-pregnancy T1DM has doubled between 1990-2020 (3). In Qatar, 0.2% of all pregnancies are affected by T1DM (4). Pregnancies with T1DM are associated with an increased risk of maternal and fetal complications such as congenital malformations, pre-term labor, pre-eclampsia, Cesarean section ( C-sections), large for gestational age (LGA), small for gestational age (SGA), and intra-uterine fetal death (IUFD) (5–7). Apart from glycaemic control, pre-gravid obesity, excessive gestational weight gain, and smoking are factors associated with poor pregnancy outcomes (8).

While improvements in pre-pregnancy care have resulted in a lower incidence of congenital malformation in women with DM, there has been a paradoxical increase in the rate of LGA (9). Glycaemic control plays a central role in determining birth weight. Most studies have shown that elevated third-trimester glycated hemoglobin (HBA1c) was associated with increased fetal weight (5, 10–14). However, there are contradicting data on the association between first-trimester HBA1c and birth weight, with some studies showing increased birth weight with higher pre-pregnancy HBA1c and others showing a reduced fetal weight with higher pre-pregnancy HBA1c (10, 15, 16).

The placenta plays a central role in the rates of nutrient transfer to the fetus. Desoye et al. suggested that pre-pregnancy glycaemic control is central to placental development (9). They hypothesized that poor glycaemic control in early pregnancy is associated with poor placentation, while adequate control is associated with healthy placentation ( Figure 1 ). Hence, in those with adequate glycaemic control, mild hyperglycemia during pregnancy may be associated with increased birth weight, while in those with poor pre-pregnancy glycaemic control, adequate glycaemic control during pregnancy might be associated with reduced birth weight (9). Data from a randomized control trial (RCT) showed that birth weight is higher in those with good placental health and sub-optimal third-trimester glycaemic control (17).

This proof of concept study aims to investigate the intricate relationship between glycaemic control, fetal weight, and placental weight during pregnancy. The primary objective is to analyze the impact of HbA1c levels recorded during early pregnancy and the third trimester on fetal birth weight. By examining these critical time points, the study aims to determine how varying glycaemic control influences the growth and development of the fetus. The second objective is to assess the role of the placenta in mediating the effects of glycaemic control on fetal body weight and the correlation between placental weight, maternal glycaemic control and fetal birth weight. This involves evaluating the correlation between placental weight, maternal glycemic control, and fetal birth weight. Through this detailed examination, the research hopes to identify potential implications for pre-conception and prenatal care for managing women with T1DM.

We included 265 women with T1DM; placental weight data was available for 154 women. As shown in Table 1 , the mean age (SD) was 29.4 (4.6) years, the mean (SD) duration was 14.1 (7.1) years, the mean BMI (SD) was 27.2 (5.2) kg/m²), and 62.5% of women were either overweight or obese. The mean (SD) HBA1c dropped from 7.9 (1.4) % in the first trimester to 6.9 (1.1) % in the third trimester. Similarly, the proportion of women within glycaemic targets improved from 28.8% in the first trimester to 40.0% in the third trimester. As shown in Table 1 , 20% were at target in both the first and third trimesters (T-T), 8.7% were at target in the first trimester but not in the third trimester (T-H), 20% were above target in the first trimester but at target in the third trimester (H-T), and 51.3% were above target at first and third trimesters (H-H). Supplementary Table 1 shows the demographics and pregnancy outcomes of the four groups (23).

As shown in Figure 3 , the median (IQR) of fetal weight was higher in the T-H group [3.56 kg(3.24-3.83)], followed by the T-T group [ 3.41 kg(2.99-3.70)]; then H-H[ 3.31 kg(2.86-3.66)] and T-H had the lowest birth weight [ 3.20 kg( 2.79-3.50)]; p=0.009. Post-hoc analysis showed that the H-T group has significantly lower birth weight when compared with T-T, T-H, and H-H, p= 0.013, 0.032, and 0.040, respectively.

Table 2 shows the demographic of the study population with placental data. Table 3 presents the results of the regression analysis. Model 1 showed that firsta1c was negatively associated with neonatal weight (β-coefficient -150.9 [ 95% CI -244.6, -57.2], p<0.01) while thirda1c was positively associated with neonatal weight (β-coefficient 162.5 [ 95% CI 44.8,280.2], p<0.01). This relationship remained unchanged in Models 2 and 3. However, after correction for log (placenta) in Model 4, the relationship between the firsta1c and neonatal weight was lost, while that with thirda1c remained significantly positive. This suggests that the placenta is the main mediator of the negative effect of the firsta1c on the neonatal weight.

Mediation analysis ( Supplementary Tables 2 - 4 ) shows that 64% of the effect of the firsta1c on neonatal weight was mediated through the placenta (23).

Our study supports the complex interaction between pre-pregnancy glycaemic control and birth weight, as proposed by Desoye et al. We have shown that neonatal birth weight varies based on the level of glycaemic control in the first and third trimesters. Specifically, we observed the lowest fetal weight in individuals with poor early glycaemic control but achieved glycaemic targets in the third trimester (H-T group). Our findings suggest that poor early glycaemic control is linked to lower birth weight, primarily due to inadequate placentation. On the other hand, elevated third-trimester HbA1c levels are independently associated with increased birth weight.

Our findings support the results of other studies. An in-vitro study compared the placental tissues of 12 women with T1DM with 11 controls showed a decrease in the protein levels of the proliferation markers, cell cycle regulators, and cell invasion cycle in those with T1DM (24). Furthermore, under hyperglycemic conditions, the number of trophoblast cells was reduced (24). A longitudinal analysis of 157 women with T1DM enrolled in the CONCEPTT trial assessed the placental function at baseline and again at 24- and 34-weeks' gestation by measuring placental growth factor (PlGF) and soluble fms-like tyrosine kinase 1 (sFlt-1) (17). The study reported improved placental health with adequate pre-conception glycaemic control and reduced placental health with inadequate pre-conception glycaemic control (17). Infants of women with suboptimal glycemia were heavier when the placenta was healthy and lighter when the placenta was unhealthy (17).

Adequate placentation is critical for a healthy pregnancy. People with T1DM are known to be at higher risk of microangiopathy, resulting in retinopathy, nephropathy, and peripheral neuropathy. The cumulative exposure to hyperglycemia mainly determines the risk and extent of microangiopathy. In a large cohort study of people with T1DM, hyperglycemia, duration of diabetes, and hypertension were the main risk factors for microangiopathy (25). Poor placentation is likely another manifestation of microangiopathy in women with T1DM (9). Another possible factor of poor placentation is hypertriglyceridemia, an independent risk of microangiopathy in people with T1DM (26). Pre-conception hypertriglyceridemia secondary to poor glycaemic control and obesity is a known association with poor placentation and a higher risk of pre-eclampsia (27).

Birth weight is associated with short and long-term metabolic complications. A large cohort study of women with T1DM showed that the risk of IUFD is increased by 3–6 folds with either LGA or SGA infants (8, 28). Regardless of the birth weight, offspring of mothers with T1DM have lower muscle mass and higher fat mass, which increases the future risk of metabolic disorders (29). SGA infants have normal visceral fat but low total and subcutaneous fat (30). Accelerated weight gain in the first two years leads to more visceral fat deposition in SGA infants, which increases their future risk of hypertension and T2DM (30). On the other hand, LGA offspring are born with increased total body fat and are at higher risk of childhood obesity, T2DM, and cardiovascular disease (31). Furthermore, Infants in the H-H group are more likely to have intrauterine hypoxia due to overnutrition with impaired placental function. A small study showed that the offspring of women with T1DM who had intra-uterine hypoxia have higher measures of adiposity and glycemia in adulthood than those with no hypoxia (32). While those with H-T are at higher risk of SGA, it is not clear if they exhibit the same body composition of increased visceral and reduced subcutaneous fat. Hence, thriving to achieve normal placentation and maintain a normal fetal weight is critical to reducing the short- and long-term complications for offspring of mothers with T1DM.

Our findings have significant clinical implications. It is clear that adequate glycaemic management during pregnancy is insufficient to achieve favorable pregnancy outcomes. Hence, pre-conception care should be offered to all women with T1DM at reproductive age. Furthermore, adequate glycaemic control should be achieved and maintained throughout pregnancy. For those women who conceive with suboptimal glycaemic control, third-trimester glycaemic targets should be adjusted according to the fetal size to avoid restricted fetal growth.

The main limitation of the study is its retrospective nature. We lacked data on critical confounding factors such as parity, smoking, history of hypertension, history of microvascular complications, and albuminuria. Additionally, we did not have trimester-specific gestational weight gain data. However, the study's strength lies in the relatively large number of subjects and the careful analysis.

In conclusion, our findings showed that poor early glycaemic control is linked to lower neonatal birth weight, while poor third-trimester control is associated with increased fetal birth weight. The impact of early glycaemic control on neonatal weight is mainly mediated through poor placentation. These findings emphasize the crucial role of pre-conception and early glycaemic control in pregnancy outcomes.