CoChat
Sign Up
Cited by 0Body Composition Measurement TechniquesGestational Diabetes Research and ManagementPregnancy and preeclampsia studies

Trimester-Specific Bioelectrical Impedance Reference Curves in Physiological Pregnancy: Mapping Maternal Body Composition and Hydration Trajectories

Open access
Di Martino, Daniela Denis, Sabattini, Elisa, Viscioni, Lucrezia, Cerri, Serena, Radaelli, Tatjana, Ornaghi, Sara, Belfiore, Jennifer, Stampalija, Tamara +5 more·2026·Nutrients
DOI 10.3390/nu18203341
Open this paper in CoChat — free account

Abstract

Background/Objectives: Clinical risk measures such as pre-pregnancy body mass index (BMI) and gestational weight gain do not distinguish the expansion of fluid, lean, and adipose compartments. Standardized obstetric reference curves for bioelectrical impedance analysis (BIA) are lacking. This study aimed to establish longitudinal, trimester-specific BIA reference curves in uncomplicated pregnancies. Methods: This multicenter, prospective, longitudinal cohort study was conducted at three Italian tertiary research hospitals from September 2020 to September 2025. BIA assessments (InBody 270®) were performed at 10–15, 23–25, and 34–37 gestational weeks. Generalized additive models for location, scale, and shape were used to derive reference centiles and BMI-conditioned contour plots for total body water, lean mass, fat mass, and visceral fat level. Pregnancies with complications or incomplete longitudinal data were excluded from the primary reference cohort. Results: Among 442 healthy pregnant women, all parameters increased progressively across gestation. Each 1-unit increase in pre-pregnancy BMI was associated with a larger shift in fat mass (1.88 kg; 95% CI, 1.77–1.99) than in total body water (0.47 L; 95% CI, 0.38–0.57). Conclusions: These multicenter, BMI-conditioned, trimester-specific BIA reference curves provide normative benchmarks for maternal hydration and body composition and may support the identification of early deviations from physiological gestational trajectories.

AI snapshot

Written from the paper’s full text.

TL;DR

This multicenter prospective study derived trimester-specific bioelectrical impedance reference centiles for total body water, lean mass, fat mass, and visceral fat in 442 uncomplicated pregnancies. All parameters increased across gestation, with pre-pregnancy BMI shifting fat mass and visceral fat far more than total body water or lean mass.

Key findings

  • In the physiologic reference population, median values increased from first to third trimester: total body water 31.8 to 35.5 L, lean mass 43.4 to 48.3 kg, fat mass 17.7 to 22.7 kg, and visceral fat level 7.0 to 10.0 a.u.
  • For a representative nonsmoking patient with pre-pregnancy BMI 21.6, estimated medians rose from 12 to 34 weeks: total body water 31.2 to 35.0 L, lean mass 42.5 to 47.8 kg, fat mass 16.9 to 21.9 kg, and visceral fat level 6.9 to 9.3 a.u.
  • Each 1-unit increase in pre-pregnancy BMI was associated with a larger shift in fat mass (1.88 kg; 95% CI, 1.77–1.99) than in total body water (0.47 L; 95% CI, 0.38–0.57); lean mass changed by 0.63 kg (95% CI, 0.50–0.76) and visceral fat by 0.92 a.u. (95% CI, 0.85–0.98).
  • Model calibration was close to nominal: observed proportions below the 5th centile ranged from 4.7% to 5.2%, above the 95th centile from 4.4% to 5.6%, and empirical coverage of the central 90% interval from 89.4% to 90.8%.

Methods

  • Multicenter, prospective, longitudinal cohort study at three Italian tertiary hospitals from September 2020 to September 2025; 938 women were initially enrolled, and after post hoc exclusion of major complications and incomplete longitudinal data, 442 pregnancies formed the primary normative cohort.
  • BIA assessments (InBody 270) were performed at 10–15, 23–25, and 34–37 gestational weeks, with total body water, lean mass, fat mass, and visceral fat level as primary endpoints.
  • Generalized additive models for location, scale, and shape (GAMLSS) were used to derive gestational age-specific centiles, with gestational age modeled by restricted cubic spline with 3 internal knots and pre-pregnancy BMI and smoking status as covariates; participant-level bootstrap resampling accounted for within-participant correlation, with longitudinal sensitivity models using a Gaussian copula.

Limitations

  • The paper states the population was primarily Italian women of Caucasian ancestry, limiting generalizability to other ethnic groups with distinct baseline body composition.
  • Body composition parameters were estimated using the InBody 270’s proprietary multi-frequency algorithms rather than direct reference methods such as isotopic dilution or DXA, and alignment with supine tetrapolar BIA systems requires confirmation.
  • BIA was obtained at three discrete gestational windows, so continuous centile lines are model-based smooth interpolations and cannot capture short-term or acute fluid fluctuations; the post hoc filtering from 938 to 442 could introduce minor selection attrition bias, and calibration metrics reflect internal fit only, with external validation still required.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/nu18203341

Cites

22 of 22 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

9 of 37 cited sources checked · 0 supported · 0 flagged · accuracy not yet measured

  • Premisecitation
    § 4.1. Principal Findings and Originality · p. 9

    Previous obstetric studies utilizing bioelectrical impedance analysis routinely relied on reference equations or vectors calibrated on non-pregnant populations, which inherently fail to account for the unique, fluid-driven, and metabolic dynamics of gestation.

    “While previous obstetric studies had utilized bioelectrical impedance analysis [18,19], they routinely relied on reference equations or vectors calibrated on non-pregnant populations, which inherently fail to account for the unique, fluid-driven, and metabolic dynamics of gestation.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [18] Widen, 2014, Body composition changes in pregnancy: Measurement, predictors and outcomes, Eur. J. Clin. Nutr. source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [19] Sarno, 2021, Maternal body composition changes in pregnancy tracked by bioelectrical impedance analysis: A systematic review, J. Matern Fetal Neonatal Med.

      Not checked in the free run.

  • Premisecitation
    § 1. Introduction · p. 2

    Deviance from a physiological trajectory might be intertwined with gestational complications such as hypertensive disorders of pregnancy, fetal growth restriction, and others.

    “While these modifications are pivotal to support fetal development, their deviance from a physiological trajectory might be intertwined with some gestational complications such as hypertensive disorders of pregnancy, fetal growth restriction and others [2-5].”
    Quote in the paperSubstantial claim2 of 4 sources checked
    • cites [2] Valensise, 2000, Multifrequency bioelectrical impedance analysis in women with a normal and hypertensive pregnancy, Am. J. Clin. Nutr. source
    • cites [3] Ståhlberg, A., Gustafsson, C.K., Engtröm, K., Thomsen, C., Dolatabadi, S., Jonasson, E., Li, C.-Y., Ruff, D., Chen, S.-M., and Åman, P. (2014). Hypertensive disorders of pregnancy: A systematic review of fluid and hemodynamic changes. PLoS ONE, 9. source

      The Introduction and Results and Discussion address TP53 function and genetic stability in myxoid/round cell liposarcoma, not pregnancy. Searches for gestational complications, hypertensive disorders, fetal growth restriction, and physiological pregnancy changes found no passage addressing the claim. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
    • cites [4] Magee, 2022, The 2021 International Society for the Study of Hypertension in Pregnancy classification, diagnosis & management recommendations for international practice, Pregnancy Hypertens. source

      The available Abstract discusses management of hypertensive pregnancy and says the causes of pre-eclampsia remain uncertain. Searches for physiological adaptations and fetal growth restriction found no passage linking deviations from a normal pregnancy trajectory to those complications; the available copy contains no substantive body text to check further. (LLM, conf=0.70)

      Experimental — accuracy not yet measured
    • cites [5] Fatima, 2021, Fetal programming and the impact of maternal body composition on international benchmarks of infant growth, Am. J. Clin. Nutr.

      Not checked in the free run.

  • Premisecitation
    § 4.2. Maternal Hydration Stability vs. Adipose-Driven Meta-Inflammation · p. 10

    Excess visceral adiposity does not merely represent a passive energy reserve; rather, it acts as an active immunometabolic organ.

    “Excess visceral adiposity does not merely represent a passive energy reserve; rather, it acts as an active immunometabolic organ [8-10].”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites [8] Catalano, 2017, Obesity and pregnancy: Mechanisms of short term and long term adverse consequences for mother and child, BMJ source
    • cites [9] Hotamisligil, 2017, Foundations of immunometabolism and metainflammation, Nature source
    • cites [10] Kirwan, 2002, TNF-alpha is a predictor of insulin resistance in human pregnancy, Diabetes source

      The Discussion (page unavailable) says fat cells secrete TNF-α and leptin, supporting the point that adipose tissue is metabolically active rather than merely an energy reserve. It does not examine visceral adiposity specifically or establish that excess visceral fat acts as an immunometabolic organ; for the pregnancy-related TNF-α increase studied here, it identifies the placenta as the apparent primary source. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Findingown data
    § 3.1. Study Population · p. 5

    Among 774 eligible pregnancies, 582 (75.2%) underwent complete BIA interrogations across all three trimesters.

    “Among these pregnancies, 582 patients (75.2%) underwent complete bioimpedance interrogations across all three trimesters.”
    Quote in the paperSubstantial claim
  • Findingown data
    § 3.2. Gestational Trajectories and Reference Curves · p. 8

    For a representative non-smoking patient with pre-pregnancy BMI 21.6, estimated median visceral fat level increased from 6.9 a.u. (T1) to 8.2 a.u. (T2) and 9.3 a.u. (T3).

    “- Visceral Fat Level: The estimated median increased from 6.9 a.u. (T1) to 8.2 a.u. (T2) and 9.3 a.u. (T3).”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.
  • May combine two claims. 24
  • Classified as probably not a substantial claim. 21

  • May combine two claims.

    Premisecitation
    § 1. Introduction · p. 2

    BIA offers a non-invasive, safe, and highly reproducible method to estimate body fluid compartments and active tissue components.

    “Bioelectrical Impedance Analysis (BIA) offers a non-invasive, safe, and highly reproducible method to estimate body fluid compartments and active tissue components [17-19].”
    Quote in the paperSubstantial claim2 of 3 sources checked
    • cites [17] Moroni, 2022, Bioelectrical Impedance Vector Analysis (BIVA) for the monitoring of body composition in pregnancy, Eur. J. Clin. Nutr. source

      The Introduction (p. 4) calls BIA fast and reliable and says it is used to estimate body liquids. The Methods (p. 7) say BIVA vectors can indicate changes in cellular mass, and the Conclusions (p. 10) describe the approach as non-invasive. The paper does not establish that it is safe or highly reproducible, and it cautions in the Introduction (p. 4) that conventional BIA cannot quantify fat-free mass during pregnancy using standard hydration assumptions. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites [18] Widen, 2014, Body composition changes in pregnancy: Measurement, predictors and outcomes, Eur. J. Clin. Nutr. source

      The BIA section (page not provided) describes BIA as non-invasive and says it estimates total body water, from which fat and fat-free mass can be derived. It describes a very-low-amperage current but does not explicitly establish safety or high reproducibility. The same section cautions that BIA estimates are difficult to validate in pregnancy. (LLM, conf=0.87)

      Experimental — accuracy not yet measured
    • cites [19] Sarno, 2021, Maternal body composition changes in pregnancy tracked by bioelectrical impedance analysis: A systematic review, J. Matern Fetal Neonatal Med.

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § 4.2. Maternal Hydration Stability vs. Adipose-Driven Meta-Inflammation · p. 10

    This persistent inflammatory cascade promotes systemic insulin resistance and endothelial dysfunction, directly predisposing patients to gestational diabetes and hypertensive disorders.

    “This persistent inflammatory cascade promotes systemic insulin resistance and endothelial dysfunction, directly predisposing patients to gestational diabetes and hypertensive disorders [13,15,16].”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites [13] Jansson, 2007, Placental nutrient transport and fetal growth for gestational age, Int. J. Obes.

      Not checked in the free run.

    • cites [15] Maynard, 2003, Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction in preeclampsia, J. Clin. Investig. source

      The Introduction links excess placental sFlt1 to endothelial dysfunction and hypertension in preeclampsia, and the summary (p. 654) says it contributes to preeclamptic hypertension. These passages do not support the attributed inflammatory cascade, systemic insulin resistance, or predisposition to gestational diabetes; the mechanism described is antagonism of VEGF and PlGF. (LLM, conf=0.88)

      Experimental — accuracy not yet measured
    • cites [16] McIntyre, 2019, Gestational diabetes mellitus, Nat. Rev. Dis. Primers source

  • May combine two claims.

    Study design and conductreasoning
    § 4.1. Principal Findings and Originality · p. 10

    Compared to prior longitudinal BIA studies in healthy pregnant cohorts, which were often limited by small single-center sample sizes or lacked rigorous post hoc complication screening, the multicenter design and strict exclusion criteria ensured that these curves reflected a highly reliable physiological baseline.

    “Compared to prior longitudinal BIA studies in healthy pregnant cohorts, which were often limited by small single-center sample sizes or lacked rigorous post hoc complication screening [25-27], our multicenter design and strict exclusion criteria ensured that these curves reflected a highly reliable physiological baseline.”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites [25] Piuri, 2017, Longitudinal changes and correlations of bioimpedance and anthropometric measurements in pregnancy: Simple possible bed-side tools to assess pregnancy evolution, J. Matern Fetal Neonatal Med. source
    • cites [26] Larciprete, 2003, Body composition during normal pregnancy: Reference ranges, Acta Diabetol. source
    • cites [27] Bosaeus, 2020, Body Composition During Pregnancy: Longitudinal Changes and Method Comparisons, Reprod. Sci. source

      Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationcitation
    § 4.2. Maternal Hydration Stability vs. Adipose-Driven Meta-Inflammation · p. 10

    The remarkable stability of the TBW trajectory across different pre-pregnancy BMI strata suggests that gestational fluid expansion is a highly preserved, 'hardwired' evolutionary adaptation designed to support placental perfusion.

    “The remarkable stability of the TBW trajectory across different pre-pregnancy BMI strata suggested that gestational fluid expansion is a highly preserved, 'hardwired' evolutionary adaptation designed to support placental perfusion [1-3].”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites [1] Hytten, 1985, Blood volume changes in normal pregnancy, Clin. Haematol. source
    • cites [2] Valensise, 2000, Multifrequency bioelectrical impedance analysis in women with a normal and hypertensive pregnancy, Am. J. Clin. Nutr. source
    • cites [3] Ståhlberg, A., Gustafsson, C.K., Engtröm, K., Thomsen, C., Dolatabadi, S., Jonasson, E., Li, C.-Y., Ruff, D., Chen, S.-M., and Åman, P. (2014). Hypertensive disorders of pregnancy: A systematic review of fluid and hemodynamic changes. PLoS ONE, 9. source

      The Results and Discussion passages concern TP53 expression and genetic stability in myxoid/round cell liposarcoma. Searches for TBW trajectories, pre-pregnancy BMI, gestational fluid expansion, and placental perfusion found no passage addressing the claim. The cited paper therefore cannot verify this part of the sentence. (LLM, conf=0.79)

      Experimental — accuracy not yet measured
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given
    § 3.1. Study Population · p. 5

    These 140 pregnancies were excluded from the primary normative pool and retained exclusively for sensitivity analyses.

    “These 140 pregnancies were excluded from the primary normative pool and retained exclusively for sensitivity analyses (Supplementary Table S1).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Abstract · p. 1

    In the 442 healthy pregnant women (physiologic reference population), all BIA parameters (total body water, lean mass, fat mass, and visceral fat level) increased progressively from the first to the third trimester across all three centers.

    “Among 442 healthy pregnant women, all parameters increased progressively across gestation.”
    • Also stated in: § 3.2. Gestational Trajectories and Reference Curves · p. 6 · plain
      “Across all three clinical centers, a progressive, consistent increase from the first to the third trimester was observed for Total Body Water, Lean Mass, Fat Mass, and Visceral Fat level.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The device-derived estimates of total body water, lean mass, fat mass, and visceral fat level from the InBody 270's proprietary DSM-BIA algorithms are valid measures of maternal body composition and hydration in pregnancy.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § 3.1. Study Population · p. 5

    Within the longitudinal subgroup, 83 (14.3%) developed GD, 9 (1.5%) had preterm birth, 4 (0.7%) developed intrahepatic cholestasis of pregnancy, and 46 (7.9%) had other recorded minor obstetric complications.

    “Within this longitudinal subgroup, 83 (14.3%) developed GD, 9 (1.5%) had preterm birth, 4 (0.7%) developed intrahepatic cholestasis of pregnancy, and 46 (7.9%) had other recorded minor obstetric complications.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 3.2. Gestational Trajectories and Reference Curves · p. 8

    For a representative non-smoking patient with pre-pregnancy BMI 21.6 (median of the healthy reference population), estimated median total body water increased from 31.2 L at 12 weeks to 33.3 L at 24 weeks and 35.0 L at 34 weeks, with 90% reference intervals (5th-95th centiles) spanning 25.9-37.1 L, 27.6-39.6 L, and 29.0-41.7 L respectively.

    “- Total Body Water: Increased from 31.2 L at 12 weeks to 33.3 L at 24 weeks and 35.0 L at 34 weeks, with 90% reference intervals (5th-95th centiles) spanning 25.9 to 37.1 L, 27.6 to 39.6 L, and 29.0 to 41.7 L, respectively;”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § 3.1. Study Population · p. 5

    Maternal age and parity differed significantly between centers, whereas no significant between-center differences were observed in pre-pregnancy maternal height or BMI.

    “Maternal age and parity differed significantly between centers, whereas no significant between-center differences were observed in pre-pregnancy maternal height or BMI.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 8

    Similar associations were observed in the longitudinal sensitivity analysis.

    “Similar associations were observed in the longitudinal sensitivity analysis (Supplementary Table S6).”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • May combine two claims.

    Findingown data
    § 3.1. Study Population · p. 5

    Baseline characteristics between women with complete longitudinal visits (n=582) and those with missing visits (n=192) showed no significant differences in maternal age (p=0.42), baseline BMI (p=0.68), or parity (p=0.31).

    “Evaluation of baseline characteristics between women with complete longitudinal visits ( n = 582) and those with missing visits (n = 192) demonstrated no significant differences in maternal age ( p = 0.42), baseline BMI ( p = 0.68), or parity ( p = 0.31).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data
    § 3.2. Gestational Trajectories and Reference Curves · p. 7

    Recruitment targeted three primary gestational windows (10-15, 23-25, and 34-37 weeks), and the continuous trajectories across gestational age represent model-based smooth interpolations linking these empirical assessment windows.

    “Although recruitment was targeted around three primary gestational windows (10-15, 23-25, and 34-37 weeks), the continuous trajectories presented across gestational age represent model-based smooth interpolations linking these empirical assessment windows (Figure 2).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The continuous centile lines from model-based smooth interpolation accurately reflect gestational trajectories between the three discrete assessment windows.

  • May combine two claims.

    Study design and conductown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 8

    Normalized quantile residuals exhibited mean values close to 0 and standard deviations of 1.00 across all endpoints, with Q-Q plots demonstrating no relevant tail mis-calibration.

    “Normalized quantile residuals exhibited mean values close to 0 and standard deviations of 1.00 across all endpoints, with Q-Q plots demonstrating no relevant tail mis-calibration.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 8

    Within the healthy reference population, the observed proportion of patients falling below the 5th centile ranged from 4.7% to 5.2% across all endpoints, while the observed proportion above the 95th centile ranged from 4.4% to 5.6%.

    “Within the healthy reference population, the observed proportion of patients falling below the 5th centile ranged from 4.7% to 5.2% across all endpoints, while the observed proportion above the 95th centile ranged from 4.4% to 5.6%.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § 1. Introduction · p. 2

    Standard BIA equations and reference vectors are traditionally calibrated on non-pregnant populations, failing to account for the unique, fluid-driven impedance shifts that characterize normal gestation.

    “Standard BIA equations and reference vectors are traditionally calibrated on non-pregnant populations, failing to account for the unique, fluid-driven impedance shifts that characterize normal gestation [19-23].”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [19] Sarno, 2021, Maternal body composition changes in pregnancy tracked by bioelectrical impedance analysis: A systematic review, J. Matern Fetal Neonatal Med.

      Not checked in the free run.

    • cites [20] Staelens, 2016, Bioelectrical impedance analysis in pregnant women: A comparison of dynamic changes in body composition, Physiol. Meas.

      Not checked in the free run.

    • cites [21] Lukaski, 1994, Total body water in pregnancy: Assessment by using bioelectrical impedance, Am. J. Clin. Nutr. source
    • cites [22] Ward, 2019, Bioelectrical impedance analysis for body composition assessment: Reflections on accuracy, clinical utility, and standardisation, Eur. J. Clin. Nutr. source
    • cites [23] Lukaski, H.C., Piccoli, A., and Kyle, U.G. (2019). Refinement of bioelectrical impedance vector analysis (BIVA) to appreciate tissue hydration and cellularity in human health and disease. Nutrients, 11.

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § 2.5. Statistical Analysis · p. 4

    Model goodness of fit and calibration were assessed using normalized quantile residuals, observed-to-expected centile agreement, and empirical coverage of the central reference intervals.

    “Model goodness of fit and calibration were assessed using normalized quantile residuals, observed-to-expected centile agreement, and empirical coverage of the central reference intervals.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § 2.5. Statistical Analysis · p. 4

    Longitudinal sensitivity models were fitted using a Gaussian copula with an unstructured within-participant correlation matrix to assess robustness of estimated centiles.

    “To assess the robustness of the estimated centiles, longitudinal sensitivity models were fitted using a Gaussian copula with an unstructured within-participant correlation matrix, retaining the same marginal distributions and covariate specifications as the primary models.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 8

    These calibration metrics reflect internal model fit within the development cohort and should not be interpreted as external validation.

    “These metrics reflect internal model fit within the development cohort and should not be interpreted as external validation.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § 2.4. Variables and Measurements · p. 3

    A final postpartum follow-up visit was systematically conducted to confirm the physiological nature of the pregnancy before including the patient's data in the normative reference pool.

    “A final postpartum follow-up visit was systematically conducted to collect mode of delivery, neonatal birth weight and outcome, and final maternal-fetal data, thereby definitively confirming the physiological nature of the pregnancy before including the patient's data in the normative reference pool.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § 2.5. Statistical Analysis · p. 4

    An additional sensitivity analysis included GD and other minor obstetric complications in the model-derivation population to assess whether reference trajectories were influenced by them.

    “To assess whether the reference trajectories were influenced by GD or other minor obstetric complications, an additional sensitivity analysis was performed including these pregnancies in the model-derivation population.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data
    § 2.5. Statistical Analysis · p. 4

    Calibration was evaluated across gestational age, pre-pregnancy BMI strata, and recruiting centers.

    “Calibration was further evaluated across gestational age and pre-pregnancy BMI strata and recruiting centers.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § 4.1. Principal Findings and Originality · p. 9

    Pre-pregnancy BMI has a minimal, tightly calibrated effect on total body water and lean mass, but a profound, disproportionate impact on maternal fat mass and visceral fat trajectories.

    “While a 1-unit increase in pre-pregnancy BMI exerted a minimal, tightly calibrated effect on TBW and Lean Mass, it demonstrated a profound, disproportionate impact on maternal Fat Mass and Visceral Fat trajectories.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § 2.3. Post Hoc Exclusion Criteria for the Normative Cohort · p. 3

    Patients who failed to complete the bioelectrical impedance assessments across all three main gestational time points were excluded due to incomplete longitudinal data.

    “Patients who failed to complete the bioelectrical impedance assessments across all three main gestational time points were excluded due to incomplete longitudinal data.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation
    § 2.4. Variables and Measurements · p. 3

    BIA measurements were performed in accordance with the manufacturer's clinical guidelines and standardized research protocols.

    “The measurements were performed in accordance with the manufacturer's clinical guidelines and standardized research protocols [24].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [24] (2026, October 01). Available online: https://uk.inbody.com/wp-content/uploads/2019/01/270_manual_ENG_C_170714.pdf.

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 9

    Center-adjusted sensitivity models did not materially alter primary model selection or calibration parameters.

    “Finally, center-adjusted sensitivity models did not materially alter primary model selection or calibration parameters (Table S5).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 8

    The pre-pregnancy BMI gradient was modest for total body water and lean mass, whereas it was markedly more pronounced for fat mass and visceral fat level.

    “The pre-pregnancy BMI gradient was modest for Total Body Water and Lean Mass, whereas it was markedly more pronounced for Fat Mass and Visceral Fat level.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § 2.4. Variables and Measurements · p. 3

    Patients were verified to have emptied their bladders before BIA assessment to avoid confounding fluid-volume artifacts.

    “Patients were verified to have emptied their bladders before the assessment to avoid confounding fluid-volume artifacts.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § 2.5. Statistical Analysis · p. 4

    The conditional outcome distribution was selected by comparing converged candidate distributional models using the Bayesian information criterion (BIC).

    “The conditional outcome distribution was selected by comparing converged candidate distributional models using the Bayesian information criterion (BIC).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data
    § 3.3. BMI-Conditioned Contour Plots and Model Calibration · p. 9

    All endpoints retained a Box-Cox t-distribution with a constant scale, tail calibration remained close to nominal values, and estimated center-related shifts did not alter the overall gestational and BMI-dependent architecture of the normative reference curves.

    “All endpoints retained a Box-Cox t-distribution with a constant scale, tail calibration remained close to nominal values, and estimated center-related shifts did not alter the overall gestational and BMI-dependent architecture of the normative reference curves.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § 2.5. Statistical Analysis · p. 4

    Within-participant correlation was accounted for in the estimation of confidence intervals and p-values using participant-level bootstrap resampling.

    “Within-participant correlation was accounted for in the estimation of confidence intervals and p -values using participant-level bootstrap resampling.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning
    § 4.4. Strengths and Limitations · p. 11

    BIA assessments were obtained at three discrete gestational windows (10-15, 23-25, and 34-37 weeks); consequently, the continuous centile lines represent model-based smooth interpolations and cannot capture short-term or acute temporal fluid fluctuations.

    “Third, BIA assessments were obtained at three discrete gestational windows (10-15, 23-25, and 34-37 weeks); consequently, the continuous centile lines represent model-based smooth interpolations and cannot capture short-term or acute temporal fluid fluctuations.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The continuous centile lines from model-based smooth interpolation accurately reflect gestational trajectories between the three discrete assessment windows.

  • May combine two claims.

    Premisecitation
    § 4.2. Maternal Hydration Stability vs. Adipose-Driven Meta-Inflammation · p. 10

    Hypertrophic visceral adipose tissue is a primary driver of chronic, low-grade meta-inflammation, characterized by an accelerated secretion of pro-inflammatory cytokines (such as TNF-alpha and IL-6) and a concomitant reduction in protective adiponectin.

    “Hypertrophic visceral adipose tissue is a primary driver of chronic, low-grade meta-inflammation, characterized by an accelerated secretion of pro-inflammatory cytokines (such as TNF-alpha and IL-6) and a concomitant reduction in protective adiponectin [9,11,12].”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [9] Hotamisligil, 2017, Foundations of immunometabolism and metainflammation, Nature source
    • cites [11] Gibney, 2008, Obesity in pregnancy stimulates macrophage activation and inflammation in the placenta, J. Clin. Endocrinol. Metab.

      Not checked in the free run.

    • cites [12] Jaworowski, 2017, Low-Grade Metabolically-Induced Inflammation Mediators Interleukin-6, Adiponectin, and TNF-α Serum Levels in Obese Pregnant Patients in the Perinatal Period, Med. Sci. Monit. Basic Res.

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § 4.4. Strengths and Limitations · p. 11

    Despite limitations, the multicenter design and broadly consistent findings from GAMLSS and longitudinal sensitivity analyses support the derived curves as a reference framework for maternal bioimpedance in uncomplicated pregnancy, although external validation remains necessary before clinical application.

    “Despite these limitations, the multicenter design and broadly consistent findings from GAMLSS and longitudinal sensitivity analyses support the derived curves as a reference framework for maternal bioimpedance in uncomplicated pregnancy, although external validation remains necessary before clinical application.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.
    • Rests on an assumption: The continuous centile lines from model-based smooth interpolation accurately reflect gestational trajectories between the three discrete assessment windows.
    • Rests on an assumption: The post-hoc filtering did not introduce substantial selection attrition bias.
    • Rests on an assumption: The final cohort of 442 women with uncomplicated pregnancies is free of undiagnosed conditions affecting body composition.

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationcitation
    § 4.2. Maternal Hydration Stability vs. Adipose-Driven Meta-Inflammation · p. 10

    Deviations above the 95th centile describe statistical non-conformity relative to a physiological cohort, presenting a hypothesis-generating marker that requires prospective testing in cohorts with fluid overload or hypertensive complications.

    “Deviations above the 95th centile describe statistical non-conformity relative to a physiological cohort, presenting a hypothesis-generating marker that requires prospective testing in cohorts with fluid overload or hypertensive complications [2,28].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [2] Valensise, 2000, Multifrequency bioelectrical impedance analysis in women with a normal and hypertensive pregnancy, Am. J. Clin. Nutr. source
    • cites [28] Staelens, 2016, Maternal body fluid composition in uncomplicated pregnancies and preeclampsia: A bioelectrical impedance analysis, Eur. J. Obstet. Gynecol. Reprod. Biol. source
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationreasoning
    § 5. Conclusions · p. 11

    These normative benchmarks could support the identification of deviations in maternal hydration and body composition and facilitate future research on personalized nutritional and obstetric monitoring.

    “These normative benchmarks could support the identification of deviations in maternal hydration and body composition and facilitate future research on personalized nutritional and obstetric monitoring.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.
    • Rests on an assumption: The continuous centile lines from model-based smooth interpolation accurately reflect gestational trajectories between the three discrete assessment windows.
    • Rests on an assumption: The post-hoc filtering did not introduce substantial selection attrition bias.
    • Rests on an assumption: The final cohort of 442 women with uncomplicated pregnancies is free of undiagnosed conditions affecting body composition.

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § 4.4. Strengths and Limitations · p. 11

    Body composition parameters were estimated using the built-in multi-frequency algorithms of a commercial standing BIA device (InBody 270), rather than direct reference methods (e.g., isotopic dilution techniques or DXA).

    “Therefore, body composition parameters were estimated using the built-in multi-frequency algorithms of a commercial standing BIA device (InBody 270), rather than direct reference methods (e.g., isotopic dilution techniques or DXA); further studies should confirm alignment with supine tetrapolar BIA systems.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § 4.2. Maternal Hydration Stability vs. Adipose-Driven Meta-Inflammation · p. 10

    By demonstrating that fat mass and visceral fat trajectories were highly dependent on baseline BMI, the study provided clinicians with a sensitive tool to monitor these inflammation-prone compartments.

    “By demonstrating that Fat Mass and Visceral Fat trajectories were highly dependent on baseline BMI, our study provided clinicians with a sensitive tool to monitor these inflammation-prone compartments.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given
    § 4.4. Strengths and Limitations · p. 11

    Model calibration metrics reflect internal goodness-of-fit within the development cohort; external validation in independent international cohorts is still required.

    “Finally, model calibration metrics reflect internal goodness-of-fit within the development cohort; external validation in independent international cohorts is still required.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning
    § 4.4. Strengths and Limitations · p. 11

    The multi-step post-hoc filtering (which reduced the cohort from 938 enrolled to 442 analyzed) was necessary to establish a strictly uncomplicated reference pool, but could introduce minor selection attrition bias.

    “Fourth, the multi-step post-hoc filtering (which reduced the cohort from 938 (enrolled) to 442 (analyzed)) was necessary to establish a strictly uncomplicated reference pool, but could introduce minor selection attrition bias.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The post-hoc filtering did not introduce substantial selection attrition bias.
39 of 39

  • The paper relies on this without support.

    The device-derived estimates of total body water, lean mass, fat mass, and visceral fat level from the InBody 270's proprietary DSM-BIA algorithms are valid measures of maternal body composition and hydration in pregnancy.

    “It must be highlighted that BIA does not measure tissue mass or volume components directly. Parameters such as Total Body Water (TBW), Lean Mass (LM), Fat Mass (FM), and Visceral Fat Level (VF) represent device-derived estimates calculated using the manufacturer's proprietary Direct Segmental Multi-frequency BIA (DSM-BIA) algorithms.”

  • The paper relies on this without support.

    The 442 healthy pregnant women in this multicenter cohort are representative of the general population of uncomplicated pregnancies, so the derived reference curves are generalizable.

    “Consequently, the final cohort on which physiologic model derivation was performed included 442 pregnancies (Figure 1).”

  • The paper relies on this without support.

    No prior study has established longitudinal, trimester-specific bioelectrical impedance reference curves tailored exclusively to a large, strictly selected cohort of physiologically uncomplicated pregnancies.

    “To the best of our knowledge, this was the first multicenter, prospective study to establish longitudinal, trimester-specific bioelectrical impedance reference curves tailored exclusively to a large, strictly selected cohort of physiologically uncomplicated pregnancies.”

  • The paper relies on this without support.

    The InBody 270 multi-frequency algorithms provide valid estimates of maternal body composition in pregnancy.

    “body composition parameters were estimated using the built-in multi-frequency algorithms of a commercial standing BIA device (InBody 270), rather than direct reference methods (e.g., isotopic dilution techniques or DXA); further studies should confirm alignment with supine tetrapolar BIA systems.”

  • The paper relies on this without support.

    The continuous centile lines from model-based smooth interpolation accurately reflect gestational trajectories between the three discrete assessment windows.

    “the continuous centile lines represent model-based smooth interpolations and cannot capture short-term or acute temporal fluid fluctuations.”

  • The paper relies on this without support.

    The post-hoc filtering did not introduce substantial selection attrition bias.

    “the multi-step post-hoc filtering (which reduced the cohort from 938 (enrolled) to 442 (analyzed)) was necessary to establish a strictly uncomplicated reference pool, but could introduce minor selection attrition bias.”

  • The paper relies on this without support.

    The final cohort of 442 women with uncomplicated pregnancies is free of undiagnosed conditions affecting body composition.

    “our final modelderivation population represented a highly filtered reference cohort of true 'normality.'”
7 of 7

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Gestational diabetes appears in the cohort after its stated exclusion

    Major

    The paper says women diagnosed with major complications including gestational diabetes were excluded before defining the 774-person baseline population. It then reports 83 gestational diabetes cases among the 582 women drawn from that population, leaving the cohort-selection sequence inconsistent.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 7

  1. Fatima, 2021, Fetal programming and the impact of maternal body composition on international benchmarks of infant growth, Am. J. Clin. Nutr., 114, 1432
  2. Gibney, 2008, Obesity in pregnancy stimulates macrophage activation and inflammation in the placenta, J. Clin. Endocrinol. Metab., 93, 3999
  3. Jaworowski, 2017, Low-Grade Metabolically-Induced Inflammation Mediators Interleukin-6, Adiponectin, and TNF-α Serum Levels in Obese Pregnant Patients in the Perinatal Period, Med. Sci. Monit. Basic Res., 23, 1
  4. Jansson, 2007, Placental nutrient transport and fetal growth for gestational age, Int. J. Obes., 31, 389
  5. Pantham, 2015, Review: The role of mitochondrial dysfunction and elevated cytokines in the pathogenesis of preeclampsia and gestational diabetes, Placenta, 36, 1475
  6. Sarno, 2021, Maternal body composition changes in pregnancy tracked by bioelectrical impedance analysis: A systematic review, J. Matern Fetal Neonatal Med., 34, 2715
  7. Staelens, 2016, Bioelectrical impedance analysis in pregnant women: A comparison of dynamic changes in body composition, Physiol. Meas., 37, 115

Authors
Di Martino, D. D., Sabattini, E., Viscioni, L., Cerri, S., Radaelli, T., Ornaghi, S., Belfiore, J., Stampalija, T., Nuredini, R., Privitera, G., Tinè, G., Ferrazzi, E., & 3P Study Group Collaborators
Affiliations
  • Department of Mother and Child, Mangiagalli Center, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, 20122 Milan, Italy
  • University of Milan (IT)
  • Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico (IT)
  • Azienda Ospedaliera San Gerardo (IT)
  • University of Milano-Bicocca (IT)
  • Imperial College London (GB)
  • IRCCS Materno Infantile Burlo Garofolo (IT)
Corresponding author
Belfiore, Jennifer
ORCID
Venue
Nutrients
Publisher
MDPI AG
Published
2026-10-12
Volume
18
Issue
20
Pages
3341
ISSN
2072-6643 (ISSN-L)
DOI
10.3390/nu18203341
OpenAlex
W7221543556
Semantic Scholar
65ffde8f5b3df0ca3d6c8ed898f3b0806715716a
Cited by
0
References
20
FWCI
0.00
Type
journal article
Language
en
Open access
Licence
Locations
  • Nutrients (publishedVersion, cc-by)
  • BOA (University of Milano-Bicocca) (submittedVersion, cc-by)
Funders
  • Italian Ministry of Health—Current Research IRCCS
  • Roche Italia S.p.A.
Topics
  • Health Sciences › Medicine › Physiology › Body Composition Measurement Techniques
  • Health Sciences › Medicine › Obstetrics and Gynecology › Gestational Diabetes Research and Management
  • Health Sciences › Medicine › Obstetrics and Gynecology › Pregnancy and preeclampsia studies
Keywords
bioelectrical impedance analysis, pregnancy, body composition, maternal hydration, total body water, lean body mass, fat mass, visceral fat, reference curves, pre-pregnancy body mass index, gestational weight gain

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-23.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-23.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review