Design: double-blind randomised clinical trial at a single level III neonatal unit
Population: 130 very preterm infants born at 24-30 weeks; mean gestation 28.5 weeks, mean birth weight 1165 g
Primary outcome: weight and length z scores, fat-free mass, and quantitative brain MRI near term-equivalent age
Effect: no difference between groups on any primary outcome; intakes met or exceeded recommendations in both arms

Human milk varies in macronutrient content between mothers and across a feed, which is the argument for measuring it and fortifying each infant's milk to a target rather than adding a standard fortifier. A double-blind randomised trial at a single level III neonatal unit assigned 130 infants born at 24 to 30 completed weeks to standard fortification or to targeted fortification guided by point-of-care mid-infrared milk analysis, aiming for at least 1 g/dL true protein and 67 kcal/dL in the base milk. Mean gestation was 28.5 weeks and mean birth weight 1165 g; 87% completed the study diet through 36 weeks.

There was no difference in z scores for weight, length or head circumference, in fat-free mass by air displacement plethysmography, or in quantitative brain MRI indices near term-equivalent age. Results held in the per-protocol analysis.

The explanation is in the intake data and it is worth reading before dismissing the technique. Base milk protein and total protein intake were slightly higher in the control group, and nutrient intakes in both arms met or exceeded recommendations. The trial did not compare adequate nutrition with inadequate nutrition; it compared two ways of achieving adequate nutrition, and found the measurement added nothing. Where standard fortification already delivers target intakes, buying a milk analyser will not improve outcomes.

Do not purchase point-of-care milk analysis expecting better growth or neurodevelopmental imaging outcomes.
Confirm that your standard fortification protocol actually achieves recommended protein and energy intakes - that is the condition under which this null result applies.
Keep clinical growth faltering as the trigger for additional protein or fat, as both arms did.
Note the single-centre design: this is one unit's standard of care, and yours may differ.
In units where fortifier supply is intermittent, the relevant question is achieving target intake at all, which this trial does not address.

Why it matters

It removes the rationale for a technology already being purchased by neonatal units on the assumption that individualised nutrition must be better.

The statistics, in plain English

A null result is only as informative as the comparison behind it, and here the comparison was unusually clean: both groups met or exceeded recommended intakes, and the control group actually received marginally more protein. So this is not a trial that failed to detect a difference through poor delivery - it is a trial in which the intervention had little room to add anything. With 130 infants it could still miss a small effect, but not one large enough to justify the equipment.