Nutritional and Exercise-Focused Lifestyle Interventions and Glycemic Control in Women with Diabetes in Pregnancy: A Systematic Review and Meta-Analysis of Randomized Clinical Trials

Diabetes in pregnancy (DIP)—including gestational diabetes mellitus (GDM) and pre-existing type 1 (T1D) or type 2 diabetes (T2D)—affects an estimated 16.7% of live births and is associated with a threefold increase in adverse maternal and neonatal outcomes such as macrosomia, stillbirth, neonatal metabolic disturbances, preeclampsia, and cesarean delivery. Women with DIP face elevated long-term risks of T2D, and offspring have higher risks of early-life glucose intolerance and later-life obesity. Clinical guidelines (e.g., NICE, ADA) recommend glycemic targets during pregnancy and emphasize lifestyle strategies—healthy diet and regular physical activity—to manage dysglycemia, yet effectiveness varies and many studies do not account for interactions between diet and physical activity. The study’s research question was: Do diet and/or exercise interventions improve maternal glycemia (fasting and postprandial glucose, HbA1c, and insulin resistance) in women diagnosed with DIP compared with control? The purpose was to synthesize and quantify evidence on nutritional supplements, dietary, and exercise interventions to inform antenatal care and identify gaps, particularly for women with pre-existing diabetes.

Prior work and guidelines support balanced dietary patterns (e.g., lower glycemic index carbohydrate quality) and physical activity for glycemic management in pregnancy, but effects on hypoglycemia/hyperglycemia episodes and maternal/offspring outcomes have been heterogeneous. Earlier systematic reviews reported moderate benefits of dietary modifications on maternal glycemia, with high heterogeneity and limited attention to exercise or combined interventions. Reviews of exercise in GDM have shown reductions in fasting and postprandial glucose, while meta-analyses of specific supplements (e.g., probiotics, vitamin D, magnesium) suggest improvements in fasting glucose and insulin resistance. However, evidence has been inconsistent, often imprecise, and few trials have included women with pre-existing T1D or T2D. This review aimed to update and expand on earlier SRMAs by including newer RCTs and conducting subgroup analyses (age, BMI, gestational age, geography, diagnostic criteria, study duration) to address heterogeneity.

The review followed PRISMA guidelines and was registered with PROSPERO (CRD42021268977). Databases searched included Cochrane, AMED, EMBASE, MEDLINE (via OVID), PubMed, and Scopus, supplemented with manual reference checks. Eligible studies were randomized controlled trials or randomized crossover studies (acute <2 weeks or long-term >2 weeks) published in English since 2000, involving pregnant women with DIP (GDM, T1D, or T2D) and assessing diet and/or exercise interventions against control, with outcomes measured via capillary or venous blood. Exclusions included non-diet/exercise interventions, populations <18 years or >45 years with comorbidities, or lack of glycemic outcomes. Data extracted included study characteristics, diagnostic criteria, intervention details (type, dose, format, duration), participant characteristics (age, BMI, gestational age), and glycemic outcomes (FPG, PPG, HbA1c, HOMA-IR). Units were standardized to mmol/L for glucose (mg/dL divided by 18.0182). Where needed, figures were digitized using WebPlotDigitizer. Risk of bias was assessed using RoB2 across domains (randomization, deviations from interventions, missing data, outcome measurement, reporting selection). Publication bias was explored via funnel plots. Meta-analyses were conducted in RevMan v5.4.1 using primarily random-effects models due to expected heterogeneity, reporting mean differences (MD) with 95% CIs for change from pre- to post-intervention. Heterogeneity was evaluated via Tau² and I², with prediction intervals where relevant. Prespecified subgroup analyses (where ≥2 RCTs) examined maternal age, gestational age, BMI, country/region, diagnostic criteria (e.g., ADA/IADPSG vs other), and study duration. Evidence certainty was graded with GRADE considering risk of bias, inconsistency, indirectness, imprecision, and other factors.

- Study selection: 24 RCTs and 3 randomized crossover trials (23 RCTs and 3 crossover in meta-analysis), totaling 1653 participants, all with GDM; no RCTs in pregnant women with pre-existing T1D or T2D were found. Interventions were categorized as nutritional supplements (n=8 RCTs), diet (n=10 RCTs + 2 crossover), and exercise (n=5 RCTs + 1 crossover). - Nutritional supplements (alpha-lipoic acid, probiotics, ginger, fish oil, zinc+vitamin): • Fasting plasma glucose (FPG): significantly reduced (8 RCTs), MD −0.30 mmol/L (95% CI −0.55, −0.06; p=0.02; I²=95%). • HOMA-IR: significantly reduced (6 RCTs), MD −0.40 (95% CI −0.58, −0.22; p<0.0001; I²=14%). • PPG and HbA1c: only 1 RCT each; no pooled analysis. Funnel plots for FPG and HOMA-IR did not indicate asymmetry. Subgroup analyses suggested earlier initiation, younger age, and normal weight may be associated with greater HOMA-IR improvements; later pregnancy initiation and non-Western settings may show less FPG effect. - Diet interventions (higher complex carbohydrates/lower GI, energy restriction, DASH vs standard care): • HOMA-IR: significantly reduced (5 RCTs), MD −1.15 (95% CI −2.36, −1.44; p=0.02; I²=94%). • FPG: trend toward reduction, not statistically significant (10 RCTs), MD −0.17 mmol/L (95% CI −0.35, 0.01; p=0.06; I²=89%). • PPG: not significant (5 RCTs), MD −0.23 mmol/L (95% CI −0.69, 0.32; p=0.34; I²=95%). • HbA1c: not significant (4 RCTs), MD −0.08% (95% CI −0.23, 0.08; p=0.34; I²=70%). Subgroup analyses indicated greater HbA1c benefit in overweight participants and in studies not using ADA criteria; HOMA-IR benefits appeared greater in longer studies with younger participants at earlier gestation and in non-Western countries not using ADA criteria. - Exercise interventions (brisk walking, resistance training, supervised mixed exercise, home-based programs): • FPG: significantly reduced (5 RCTs), MD −0.10 mmol/L (95% CI −0.20, −0.01; p=0.04; I²=0%). • PPG: not significant (4 RCTs), MD −0.24 mmol/L (95% CI −0.59, 0.12; p=0.17; I²=82%). • HbA1c: not significant (3 RCTs), MD 0.04% (95% CI −0.19, 0.27; p=0.73; I²=56%). • HOMA-IR: insufficient data for meta-analysis (reported in 1 RCT only). Subgroup analyses suggested possible modification by maternal age, gestational age, and pre-pregnancy weight, but effects were not consistently significant. Funnel plot for FPG did not indicate asymmetry. - Risk of bias: Most studies had low risk or some concerns, commonly due to limited details on allocation concealment and blinding of outcome assessors; one high-risk study (Valentini et al., 2012) was excluded from meta-analysis. - GRADE: • Nutritional supplements: moderate certainty for HbA1c and HOMA-IR; low/very low for FPG and PPG. • Diet: moderate for HOMA-IR; low/very low for FPG, PPG, HbA1c. • Exercise: moderate for FPG; low/very low for PPG, HbA1c, HOMA-IR. - Overall: Specific nutritional supplements and exercise modestly improved FPG; supplements and diet improved insulin resistance (HOMA-IR). Effects on PPG and HbA1c were generally not significant. Heterogeneity was frequently high, except for several HOMA-IR and exercise FPG analyses.

The review addressed whether diet and/or exercise interventions improve maternal glycemia in DIP by synthesizing RCT evidence. Findings indicate that, among women with GDM, certain nutritional supplements and exercise reduce fasting glucose, while supplements and dietary interventions reduce insulin resistance (HOMA-IR). These results support current clinical recommendations emphasizing lifestyle management during pregnancy and suggest that supplement-based approaches may offer standardized, adherent strategies with measurable benefits. High heterogeneity across trials underscores variability in populations, diagnostic criteria, timing and duration of interventions, and adherence. Subgroup trends suggest potential moderators—age, gestational age at intervention start, BMI, region, and diagnostic criteria—that may influence effectiveness and point to a need for more personalized diet and exercise prescriptions. Mechanistic plausibility exists: exercise enhances skeletal muscle glucose uptake and insulin sensitivity; selected supplements (e.g., probiotics, vitamin D, magnesium) may influence insulin action, inflammation, oxidative stress, and beta-cell function. Importantly, a major evidence gap persists for pre-existing T1D and T2D in pregnancy, limiting generalizability to the full DIP population.

This systematic review and meta-analysis demonstrates that nutritional supplements, dietary modifications, and exercise can favorably influence glycemic control in women with GDM, with the strongest and most consistent effects observed for insulin resistance (HOMA-IR) and modest improvements in fasting glucose. Effects on postprandial glucose and HbA1c were generally not significant. Future research should prioritize large, well-designed RCTs that include women with pre-existing T1D and T2D, assess combined lifestyle strategies, start interventions earlier in gestation, incorporate standardized adherence measures, and evaluate longer-term maternal and offspring outcomes. HOMA-IR should be incorporated as a key outcome given its sensitivity and favorable evidence profile.

- Evidence limited to GDM; no RCTs in pregnant women with pre-existing T1D or T2D were identified, restricting generalizability across DIP. - Several included studies were pilot or underpowered, limiting precision. - High between-study heterogeneity for many outcomes (especially FPG, PPG, HbA1c) due to variability in interventions, populations, diagnostic criteria, and timing/duration. - Short intervention durations and initiation later in gestation may have attenuated effects. - Inability to perform network meta-analysis due to diverse intervention strategies. - Methodological limitations lowered GRADE for many outcomes (e.g., incomplete allocation concealment, limited blinding, inadequate adherence reporting). - Subgroup analyses were constrained for some outcomes by small numbers of trials; some moderators could not be fully evaluated.