A modified low-protein infant formula supports adequate growth in healthy, term infants: A randomized, double-blind, equivalence trial

Abstract

Background: A high protein intake in early life is associated with a risk of obesity later in life. The essential amino acid requirements of formula-fed infants have been reassessed recently, enabling a reduction in total protein content and thus in protein intake. Objectives: We aimed to assess the safety of an infant formula with a modified amino acid profile and a modified low-protein (mLP) content in healthy term-born infants. Outcomes were compared with a specifically designed control (CTRL) infant formula. Methods: In this double-blind, randomized controlled equivalence trial, infants received either mLP (1.7 g protein/100 kcal; n = 90) or CTRL formula (2.1 g protein/100 kcal; n = 88) from enrollment (age ≤ 45 d) to 6 mo of age. A breastfed group served as a reference (n = 67). Anthropometry and body composition were determined at baseline, 17 wk (including safety blood parameters), and 6 mo of age. The primary outcome was daily weight gain from enrollment up until the age of 17 wk (at an equivalence margin of ±3.0 g/d). Results: Weight gain from baseline (mean ± SD age: 31 ± 9 d) up to the age of 17 wk was equivalent between the mLP and CTRL formula groups (27.9 and 28.8 g/d, respectively; difference:-0.86 g/d; 90% CI:-2.36, 0.63 g/d). No differences in other growth parameters, body composition, or in adverse events were observed. Urea was significantly lower in the mLP formula group than in the CTRL formula group (-0.74 mmol/L; 95% CI:-0.97,-0.51 mmol/L; P < 0.001). Growth rates, fat mass, fat-free mass, and several essential amino acids were significantly higher in both formula groups than in the breastfed reference group. Conclusions: Feeding an infant formula with a modified amino acid profile and a lower protein content from an average age of 1 mo until the age of 6 mo is safe and supports an adequate growth, similar to that of infants consuming CTRL formula. This trial was registered at www.trialregister.nl as Trial NL4677.