We employed EDIH, EDIP, GI, and GL as measures to examine the associations between dietary insulinemic, inflammatory, and glycemic factors and the likelihood of child neurodevelopmental delay. Elevated levels of EDIH, EDIP, and GI were associated, while GL was not found to be linked with an elevated relative risk of child neurodevelopmental delay. Notably, the robust positive correlations between EDIH, EDIP, and GI and child neurodevelopmental delay persisted even following comprehensive adjustments for the remaining three dietary indices. Sex-stratified analysis of our data revealed the associations of EDIH, EDIP, and GI with child neurodevelopmental delay were stronger among boys than girls. Thus, we conduct a combined analysis. The combined effect of EDIH and EDIP is similar to those of the combined effects of EDIH, EDIP, and GI, and both can effectively indicate delayed neurodevelopment in boys.
As far as we know, there is a lack of research on the relationship between diets characterized by elevated inflammatory and insulinemic properties and the occurrence of child neurodevelopmental delay. The methodology used to create EDIH and EDIP was based on circulating biomarkers and was not reliant on specific nutrient factors. The approach considered the dietary pattern without any preconceived notions about healthy or unhealthy foods. It encompassed complex interactions between foods, dietary nutrients, and food matrices. Prior studies have noted the importance of intrauterine inflammation on brain injury in children (30, 31). Exposure to maternal inflammation during pregnancy can result in irregular neural structure and a reduction in dendritic growth (32). Research in animal models has demonstrated that maternal immune activation can lead to ASD in offspring (33, 34). Our findings indicated a significant association between maternal diet and the risk of child neurodevelopmental delay in the overall study population.
In addition to the potential inflammatory effects, another pertinent consideration is the presence of glucose dysregulation and insulin resistance. We found that the EDIH and GI were associated with child neurodevelopmental delay risk in the overall study sample. The EDIH makes predictions about fasting C-peptide, which serves as an indicator of b-cell secretory activity (35, 36), and provides a direct measurement of the influence of diet on insulin levels. The GI represents the extent of food’s impact on blood glucose. Foods with a high glycemic index (GI) undergo quick digestion and absorption, resulting in a rapid elevation of blood glucose levels. In response to this rapid blood glucose elevation, the pancreas may need to release more insulin, which can potentially contribute to insulin resistance. The long-lasting influence of a high-glycemic load diet, as well as inflammation due to chronic excess insulin signaling, may be some of the mechanisms explaining these associations.
On the contrary, we did not observe any significant correlation between GL and the prospective risk of neurodevelopmental delay in both the overall sample and subgroup analyses. GL to some extent reflects the intake of carbohydrates, which may indicate that the effect estimate between maternal carbohydrate intake and child neurodevelopment is limited. Alternatively, GL possesses a limited capability to consider noncarbohydrate elements that could influence insulin response. Additionally, GL does not encompass the impact of diet on insulin resistance, a pivotal factor affecting insulin response. There are significant correlations between EDIH, EDIP, and GI with the risk of neurodevelopmental delay persisted even after accounting for glucose levels, underscoring those diets characterized by high inflammatory and insulinemic potential, or the capacity to induce rapid blood sugar spikes, may exert distinctive and substantial influences on the probability of neurodevelopmental delay.
For neurodevelopment to have sex-specific effects, we analyze our data stratified by sex. Findings from our longitudinal study indicate that elevated EDIH, EDIP, and GI scores might be associated with an elevated risk of neurodevelopmental delay specifically in boys. It is not clear how the effects vary by sex. Some possible mechanisms that have been recently reviewed and identified are the involvement of genetic sex(37) or steroid hormones(38). In addition, underlying neuroimmunologic sex differences could also contribute to explaining the different rates of neurodevelopmental delay (39). Research has found that maternal malnutrition and obesity are closely linked to a pro-inflammatory state, leading to offspring microglial activation. Interestingly, the brain undergoes more inflammatory processes in males, as shown by the higher levels of neuroinflammatory signals in males than in females during normal brain development (40).
To understand the prognosis of child neurodevelopment, we further analyzed the combined impact of these diets on neurodevelopmental delay. In our study, EDIH, EDIP, and GI were independent risk factors for child neurodevelopmental delay, especially in boys. We demonstrated that a value of the combined prediction of EDIH and EDIP had a better discriminative ability for the evaluation than the combined effect of EDIH and GI. The outcome is similar to the combined evaluation of EDIH, EDIP, and GI, and is more concise and convenient to perform. The dietary patterns assessed by EDIH and EDIP directly contribute to persistent hyperinsulinemia, insulin resistance, and chronic inflammation, which may help to explain why these scores are more predictive of risk for neurodevelopmental delays in children compared to GI and GL. We could find no similar prospective studies of predictions on child neurodevelopmental delay. Through the study, we may be able to make early predictions of the poor prognosis of child development, and timely intervention and treatment are crucial.
Our research has several strengths. A key advantage of our study was the use of a prospective cohort design, which can offer robust evidence concerning the link between maternal dietary patterns and the occurrence of child neurodevelopmental delay, while mitigating the potential for reverse causality biases. As far as we know, this is the inaugural study presenting the correlation between maternal dietary patterns and neurodevelopmental delay in children. Besides, the study’s large sample size allowed for conducting multiple subgroup analyses and combined analyses. Finally, we further investigated the combined effect of different maternal diet patterns on child neurodevelopmental delay. The study also has limitations. First, the evaluations of the child's diet were not accessible for incorporation into the path models. Disparities in children's diets have been demonstrated to exert enduring impacts on neurodevelopment (41). Although we have adjusted for the breastfeeding duration, a subsequent study is necessary to assess the impact of the influence of a child’s diet on offspring neurodevelopment. Second, self-reported dietary information and other covariates obtained from questionnaires could be subject to measurement errors. Nevertheless, prior validation studies have indicated a reasonably strong association between FFQs and dietary records. Third, while significant risk factors have been preselected from confounding variables typically considered in prior studies and subsequently adjusted for in the models, the presence of unknown covariates may still introduce potential residual confounding. Finally, the study population consisted mainly of Chinese women, so further research on different ethnicities and races is needed to validate the results.