Table 1

Studies investigating MBD and ALP

Lucas et al 15Preterm infants; birth weight <1850 g857Randomised to either: breast milk alone or preterm formula alone; or breast milk (banked) plus maternal milk or preterm formula plus maternal milk. ALP measured with other serum variables. Growth followed until 18 monthsInfants stratified as to peak ALP levels; infants ≥1200 IU/l had lower phosphate and reduced stature at 9 and 18 monthsObservational study from larger nutritional intervention trial: no formal/direct measurement of bone density
Backström et al 19Preterm infants <33 weeks, appropriate weight for gestation (median 1490 g)43ALP and bone-ALP measured at 3weeks, 12 weeks chronological age; DEXA at 3 and 6 months corrected gestational ageNo clear correlation between ALP and bone-ALP with bone density. Peak ALP ‘cut-off’ of >900 IU/l was only 88% sensitive and 71% specific, for subsequent low bone densityLow bone density as proxy for MBD defined as <95 mg/cm3
Faerk et al 16Preterm infants <32 weeks, mean birth weight 1129 g108Part of larger nutritional intervention RCT (randomised to either: mother's milk or breast milk+phosphate or breast milk+fortifier or preterm formula). ALP measured from 1 week of age until 37 weeks corrected gestation, DEXA scans at term correctedNo association between ALP levels and measured bone mineralisation. No difference in bone mineralisation between those with peak ALP<600 and ALP>1200
Hung et al 20Preterm infants ≤34 weeks, Mean GA 29+2 weeks; mean birthweight 1224 g46Samples collected fortnightly from 3 weeks old until term for ALP, BoneALP, Ca, P; Radiographic assessment at term equivalent to look for osteopenia. Infants given preterm formula or breast milk (fortifier incl. Ca/P added once intake 80 ml/kg/day)ALP and Bone ALP levels closely correlated. ALP> 700 IU/l at 3 weeks after delivery was 73% sensitive and 74% specific for osteopenia assessed radiographically at term equivalent. Bone-specific ALP 73% sensitive, 80% specific at same intervalInfants without osteopenia were significantly older; only 75% of infants completed study. Infants without osteopenia had significantly higher Pi at 3 weeks. Radiographic scoring used not DEXA
Walters et al 21Infants <38 weeks843 or more measurements of ALP and contemporary radiographs. Cut-off of 10× local upper limit of adult ‘normal’ as proposed predictor of bone diseaseVariable times to peak ALP level noticed. Only one infant with ALP>10×ULN had radiological evidence of bone diseaseObservational; inclusion of quite mature infants
Glass et al 14Infants 26–40 weeks gestation349ALP and bone-ALP measured at 5-10d after birth (to establish ‘normal range’ for these infants). Subset analysis of those who had radiographs done for clinical reasonsALP reached higher peak in more preterm infants, peak predated radiograph change by 2–4 weeks. Radiographic change only seen if ALP>500 IU/lCohort study, observational and not powered for radiographic comparison
James et al 22Preterm infants (28–32 weeks, 650–2000 g); Comparison group of term infants17; 15ALP measurements plus radiographs and photon absorptiometry of radius/ulna compared between two groups at term corrected agePreterm infants have lower bone mineralisation at term. Weak correlation between radiograph appearances of demineralisation and directly measured. No relationship found between mineralisation and ALP levelsComparative cohort study. ALP measured at term CGA therefore association with mineralisation at term rather than an examination of predictive value
  • ALP, alkaline phosphatase; CGA, corrected gestational age; DEXA, dual-energy absorptiometry; MBD, metabolic bone disease.