Association of pre-pregnancy body mass index and rate of weight gain during pregnancy with maternal indicators of cardiometabolic risk

The study investigates how pre-pregnancy BMI (pBMI) and monthly gestational weight gain (MGWG) relate to longitudinal trajectories of maternal cardiometabolic risk indicators—total cholesterol (TC), triglycerides (TG), glucose, and blood pressure (BP)—through pregnancy. Physiological adaptations in lipid and glucose metabolism and cardiovascular function occur during normal pregnancy, and obesity may overload these adaptations, increasing the risk of complications such as preeclampsia, gestational diabetes, and adverse birth outcomes. Understanding how pBMI and MGWG influence these biomarkers can improve risk stratification in populations with high obesity prevalence, such as in Mexico.

Prior research shows that lipid trajectories during pregnancy differ by pBMI, with obesity linked to altered lipid patterns (studies from Brazil, the U.S., Austria). Evidence on the effects of gestational weight gain (GWG), particularly MGWG, on maternal lipids is limited and inconsistent. No prior studies were identified analyzing the association of pBMI and MGWG with glucose trajectories across pregnancy, though maternal obesity increases gestational diabetes risk. pBMI is positively associated with BP (SBP, DBP, MAP) in several cohorts, and one study found GWG positively associated with SBP and DBP. Most trajectory studies have been in high-income countries, with limited data from Latin America despite high obesity prevalence and potential unique genomic and environmental factors.

Design: Prospective cohort (PRINCESA) conducted at Hospital Materno Infantil Inguarán, Mexico City. Recruitment 2010–2015; entry ≤18 gestational weeks; monthly follow-up. Participants: n=794 enrolled; for this analysis, n=720 women aged 18–49 with at least two anthropometric and biochemical measurements. Exclusions: development of pregnancy complications (e.g., gestational diabetes, preeclampsia) and active smoking. Ethics approvals obtained; informed consent signed. Exposure assessment: pBMI calculated from measured height and weight at first visit (<18 weeks), chosen due to identified systematic error in self-reported pBMI and timing before the linear weight gain phase. Height measured with Seca stadiometer (0.1 cm); weight with Tanita scale (0.01 kg; rounded to nearest 100 g). MGWG defined as month-to-month weight change; total GWG computed from first to last weight at ~36 weeks and categorized by IOM recommendations (not used as main exposure due to monthly design). Outcomes: Fasting serum TC, TG, glucose; BP (SBP, DBP, MAP). Blood sampling after 8 h fast using fluoride/oxalate tubes; serum processed within 3 h on Adaltis automated system with SpinReact reagents. BP measured with mercury sphygmomanometer per AHA guidelines. Outlier handling: TC and glucose values beyond ±3 SD (TC <67 or >397 mg/dL; glucose <50 or >122 mg/dL) removed; TG >500 mg/dL (≈4 SD) removed. Covariates: age, education, marital status, parity, gestational month; dietary energy and micronutrients initially considered but excluded from final models due to non-significance and loss of observations; physical activity not included due to no between-participant differences. Time scaling: weeks grouped into months (month 2: weeks 5–8.6; 3: 9–13.6; 4: 14–17.6; 5: 18–22.6; 6: 23–27.6; 7: 28–31.6; 8: 32–35.6; 9: 36–40.6). Statistical analysis: Linear mixed-effects models with random intercepts and slopes to model biomarker trajectories and associations with pBMI and MGWG. Functional form of time selected by model fit (log-likelihood) for each outcome: TC quadratic; TG linear; glucose cubic; SBP/DBP/MAP cubic (small differences across forms). Models adjusted for pBMI categories (underweight, normal, overweight, obesity), MGWG (where specified), age, height, education, marital status, parity, and time terms. Interaction terms between pBMI and time included to distinguish cross-sectional and longitudinal effects. Significance thresholds: p<0.05 for main effects; p<0.10 for interactions. Analyses performed in STATA/SE 15.0.

Sample: 720 women; median gestational age at enrollment 12.6 weeks (8.1–18); average 4 visits (range 2–8). pBMI distribution: underweight 4.4%, normal 45.8%, overweight 33.2%, obesity 16.6%. Total GWG mean 8.8±5.6 kg; 52.3% below IOM recommendations (final weights missing in ~50% due to referral to multiple hospitals). Lipids: • TC: Overweight and obese women had higher TC at pregnancy start vs normal pBMI (overweight β=26.62, p=0.008; obesity β=33.08, p=0.010). However, their TC rose less over time (interaction with month: overweight β=−10.35, p=0.002; obesity β=−14.18, p=0.001; with month²: overweight β=0.69, p=0.014; obesity β=0.93, p=0.009). By late pregnancy, obese women had lower TC than normal pBMI. Inclusion of MGWG attenuated pBMI associations with TC (lost significance), suggesting potential confounding/mediation. • TG: Higher at start for overweight/obesity vs normal (overweight β=17.00, p=0.024; obesity β=31.29, p=0.001). Obesity showed smaller increases over time (month interaction β=−5.42, p<0.001), ending pregnancy with lower TG vs normal. MGWG not associated with TG (β=0.06, p=0.931). Glucose: Obesity associated with higher glucose (β=45.73, p=0.011). Obese women’s glucose decreased over months relative to normals (month interaction β=−19.92, p=0.033; month² β=3.03, p=0.056; month³ β=−0.15, p=0.079), yet remained slightly higher by the end. MGWG not associated with glucose (β=0.16, p=0.352). Blood pressure: • SBP: No significant effect of pBMI over pregnancy (non-significant main and interaction terms); MGWG not associated. • DBP and MAP: Overweight and obesity showed higher baseline levels vs normal (DBP: overweight β=22.60, p=0.003; obesity β=31.61, p=0.003. MAP: overweight β=18.19, p=0.014; obesity β=27.12, p=0.008). Trajectories showed initial decreases followed by increases (significant interactions with month and month²), with overweight/obesity remaining higher throughout pregnancy. MGWG not associated with DBP or MAP. Overall: pBMI is a strong determinant of trajectories for lipids, glucose, and BP (DBP, MAP); MGWG showed no significant associations with these cardiometabolic indicators in this cohort.

The findings indicate that maternal adiposity status before pregnancy (pBMI) influences the trajectory and levels of cardiometabolic biomarkers during gestation, while short-term month-to-month weight changes (MGWG) do not materially affect these markers. Obese women started pregnancy with higher TC and TG but exhibited attenuated increases, resulting in lower late-pregnancy lipid levels compared with normal pBMI women; this aligns with prior studies. In contrast, obese women had persistently higher glucose and higher DBP/MAP throughout pregnancy. The paradox of reduced circulating lipids but higher glucose among women with pre-gestational obesity may reflect increased placental transfer and utilization of fatty acids and cholesterol, or differential insulin action in adipose tissue, leading to altered maternal-fetal lipid partitioning. The consistent association between higher pBMI and elevated BP (particularly DBP and MAP) suggests that excess maternal adiposity amplifies cardiovascular load even in normotensive pregnancies, potentially contributing to long-term cardiovascular risk. Collectively, results support prioritizing preconception weight optimization and risk stratification based on pBMI over MGWG when predicting adverse cardiometabolic trajectories in pregnancy.

Pre-pregnancy BMI is a key predictor of maternal cardiometabolic trajectories during pregnancy. Women with higher pBMI showed altered lipid, glucose, and blood pressure patterns, implying increased cardiometabolic risk during pregnancy and potentially later in life. MGWG was not significantly associated with these indicators, suggesting pBMI is more relevant for risk prediction. Continued follow-up is recommended to assess long-term cardiometabolic outcomes and to relate maternal risk trajectories to fetal growth and development. Clinical approaches should emphasize achieving a healthy pBMI before conception and consider stratifying care based on cardiometabolic risk profiles.

- Not all participants had the full set of eight monthly measurements; although mixed-effects models utilized available data, variable follow-up could affect precision. - The cohort was not a probabilistic sample and was drawn from Mexico City’s Ministry of Health system serving largely low-income populations, limiting generalizability to other settings. - Final term weights were missing for about 50% due to deliveries occurring across multiple hospitals, limiting total GWG assessment. - Dietary covariates and physical activity were not included in final models (dietary variables were non-significant and reduced sample size; physical activity showed no between-participant differences), which could leave residual confounding by lifestyle factors.