Predicting total body water in infants aged 0.5–24 months and the implications for measuring breast milk intake: a secondary analysis of isotope dilution data

  • Marieke AJ de Sévaux
  • , Pernille Kæstel
  • , Hinke Haisma
  • , Teresa HM da Costa
  • , Adriana Vázquez-Vázquez
  • , Tinku Thomas
  • , Regien Biesma
  • , Rebecca Kuriyan
  • , Grace Munthali
  • , Aly Diana
  • , Ina S. Santos
  • , Andrew P. Hills
  • , V. Pujitha Wickramasinghe
  • , Shabina Ariff
  • , M. Nishani Lucas
  • , Caroline S. Costa
  • , Shane A. Norris
  • , Vandana Jain
  • , Helen Mulol
  • , Ute Feucht
  • Birna Thorisdottir, Inga Thorsdottir, Peter SW Davies, Cornelia U. Loechl, Jonathan CK Wells

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

Background: Total body water (TBW) is commonly used to derive estimates of body composition. The deuterium oxide dose-to-mother (DTM) technique for measuring breast milk intake requires an estimate of infant TBW. The DTM calculation employs a prediction equation for estimating infant TBW from body weight (TBWpred), but the general validity of this equation is unknown. Objectives: The objectives of this study were to assess the bias of TBW predicted by the equation used in DTM calculations (TBWpred); derive a new equation based on a multicountry dataset of TBW measured by isotope dilution (TBWid) in breastfed infants, 0.5 (2 wk) to 24 mo old; and quantify the impact of the new equation on measurements of breast milk and nonmilk water intake using the DTM technique. Methods: We conducted a secondary analysis of 3756 TBWid measurements from 1457 infants, which were compared to TBWpred using the Bland-Altman method and used to generate a new prediction equation. Values of breast milk intake and nonmilk water intake from 130 DTM calculation spreadsheets were compared with recalculated values using the Bland-Altman method. Results: TBWpred underestimates TBW, with absolute and relative bias increasing with age (3 mo: –1.4% ± 9.2%; 24 mo: –11.3% ± 8.3%) and differential by sex (males: –6.4% ± 9.6%; females: –2.4% ± 9.9%). The best prediction equation incorporates log weight, age, and sex, explaining 93.7% of the variability in log TBWid. Applying this new equation in DTM calculations results in higher estimates of breast milk and nonmilk water intake; for example, in infants aged 6–12 mo, differences averaged 26 g/d (limits of agreement: –17, 69) and 13 g/d (limits of agreement: –10, 36), respectively. Conclusions: The equation used in DTM calculations provides biased TBW estimates. This has implications for measuring breast milk intake and nonmilk water intake and our understanding of nutritional needs during the first 2 y of life.

Original languageEnglish (US)
Pages (from-to)820-829
Number of pages10
JournalAmerican Journal of Clinical Nutrition
Volume122
Issue number3
DOIs
Publication statusPublished - Sept 2025

Keywords

  • anthropometry
  • body composition
  • breast milk intake
  • deuterium
  • doubly labeled water
  • infant
  • isotope dilution
  • total body water

Fingerprint

Dive into the research topics of 'Predicting total body water in infants aged 0.5–24 months and the implications for measuring breast milk intake: a secondary analysis of isotope dilution data'. Together they form a unique fingerprint.

Cite this