Diabetes in pregnancy (DIP), encompassing gestational diabetes mellitus (GDM), type 1 diabetes (T1D), and type 2 diabetes (T2D), significantly increases the risk of adverse maternal and fetal outcomes. GDM accounts for 80% of DIP cases. Women with DIP face a threefold higher risk of complications such as fetal macrosomia, stillbirth, neonatal metabolic issues, preeclampsia, and cesarean delivery. Long-term risks include the development of T2D in the mother and increased risk of glucose intolerance and obesity in the offspring. Current guidelines recommend diet and exercise for glycemic control, aiming for fasting plasma glucose below 5.3 mmol/L and 1-hour post-meal glucose below 7.8 mmol/L or 2-hour post-meal glucose below 6.4 mmol/L. While these guidelines promote healthy lifestyle habits like a balanced diet (whole grains, fruits, vegetables) and regular physical activity, evidence on their effectiveness across different diabetes types and subpopulations remains uncertain. This systematic review and meta-analysis aimed to quantify the effects of nutritional supplements, diet, and/or exercise on glycemic control (fasting and postprandial glucose, glycated hemoglobin, and insulin resistance) in women with DIP, investigating generalizability and the role of key lifestyle moderators.

Existing research supports the use of balanced diets for managing mean glucose levels in pregnant women with diabetes. However, there's significant heterogeneity in the effectiveness of these diets in reducing hypo- and hyperglycemia episodes and mitigating maternal and offspring complications. Moreover, most studies have not adequately considered the interactive effects of diet and physical activity on glycemic control and overall health. Recent studies using continuous glucose monitoring (CGM) have provided more detailed insights into lifestyle-dysglycemia associations, but concerns remain regarding the long-term risks to both mother and offspring.

This systematic review and meta-analysis followed PRISMA guidelines and was registered with PROSPERO. Randomized controlled trials (RCTs) and crossover studies published after 2000, investigating the effects of diet and/or exercise interventions on glycemic control in women with DIP (T1D, T2D, or GDM), were included. Studies focusing on children/adolescents (<18 years), women >45 years with comorbidities, or lacking reported glycemic control measures were excluded. Six databases (Cochrane, AMED, EMBASE, MEDLINE, PubMed, Scopus) were searched, supplemented by manual searches. Two authors independently screened titles, abstracts, and full texts. Data extraction included publication details, sample size, GDM diagnostic criteria, intervention type, duration, participant characteristics (age, BMI, gestational age), and glycemic control measures (fasting plasma glucose (FPG), postprandial glucose (PPG), glycated hemoglobin (HbA1c), and HOMA-IR). Risk of bias was assessed using the Cochrane Collaboration tool (RoB 2), and publication bias was checked using funnel plots. Random effects models were used for meta-analysis, with subgroup analysis considering maternal age, gestational age, BMI, country, diagnostic criteria, and study duration. The GRADE tool assessed the certainty of evidence.

The search identified 5304 records, with 26 studies (8 nutritional supplement, 12 dietary, and 6 exercise interventions) meeting inclusion criteria. All included studies involved women with GDM; no studies on women with pre-existing T1D or T2D were found. **Nutritional Supplement Interventions:** Eight RCTs (541 participants) showed that supplement interventions significantly reduced FPG (−0.30 mmol/L; 95% CI −0.55, −0.06; p = 0.02) and HOMA-IR (−0.40; 95% CI −0.58, −0.22; p < 0.0001). Subgroup analyses suggested that supplements might be less effective in later pregnancy and non-Western countries for FPG, while for HOMA-IR, earlier initiation, younger age, and non-Western countries showed greater effectiveness. Limited data prevented analysis for PPG and HbA1c. **Diet Interventions:** Ten RCTs and two crossover trials (676 participants) indicated a significant reduction in HOMA-IR (−1.15; 95% CI −2.36, −1.44; p = 0.02) with a non-significant trend towards reduced FPG. PPG and HbA1c were not significantly affected. Subgroup analysis suggested diet effectiveness is driven by overweight individuals when using criteria other than ADA and in younger participants at earlier gestational ages in non-Western countries using non-ADA criteria. **Exercise Interventions:** Five RCTs and one crossover trial (416 participants) demonstrated a significant reduction in FPG (−0.10; 95% CI −0.20, −0.01; p = 0.04). PPG and HbA1c were not significantly affected. Subgroup analysis suggested that maternal age, gestational age, and pre-pregnancy weight may modify the effectiveness of exercise interventions but not significantly. Risk of bias assessment indicated low risk or some concerns for most studies, mainly due to insufficient information on randomization concealment and blinding of outcome assessors. One study was deemed high risk and excluded from meta-analysis.

This is the first systematic review and meta-analysis comprehensively analyzing the impact of nutritional supplements, diet, and exercise on glycemic control in GDM. The findings demonstrate that supplement-based interventions offer improvements in FPG and HOMA-IR, while diet and exercise interventions show benefits in HOMA-IR and FPG, respectively. The lack of studies on women with pre-existing T1D or T2D highlights a significant research gap. Subgroup analysis helped to account for heterogeneity, but differences in diagnostic criteria, intervention strategies, and study duration influenced the results. The large effect sizes and low heterogeneity observed for HOMA-IR suggest it is a valuable outcome measure for future studies. The variation in findings across different types of interventions may stem from differences in adherence and standardization of interventions. Supplements are easier to standardize than diet and exercise plans, leading to potentially higher adherence rates and more consistent results. Future research should consider the potential interaction effects of combined interventions.

This meta-analysis emphasizes the importance of nutritional supplements, diet, and exercise in managing GDM, showing promising effects on glycemic control. HOMA-IR showed the most significant effects, low heterogeneity, and highest GRADE. Future large, well-designed RCTs are needed to investigate the most effective interventions (alone or in combination) across various diabetes types during pregnancy and to incorporate long-term maternal and offspring outcomes, refining lifestyle recommendations for women with DIP.

Several included studies were pilot studies or underpowered, limiting the statistical power to detect significant differences for all primary outcomes. Subgroup analyses were limited by the number of studies available for certain subgroups. The study's inability to perform a network analysis due to heterogeneous intervention strategies is a limitation. The short duration of some interventions and their late initiation in pregnancy may have affected the impact on glycemic outcomes. Lastly, the low GRADE quality scores for many outcomes, owing to limitations in study design (allocation concealment, blinding), might influence interpretations of the results. Despite these limitations, the substantial number of participants and diverse backgrounds of the women in the included studies offer valuable insights into the effectiveness of these lifestyle interventions.