“…For example, in terms of subject presentation, only a third of studies required subjects to be in a fasted state, a further third allowed subjects to eat lightly or drink fluids before the scan, and the remaining studies did not provide details of this aspect of their DXA assessment protocol, despite literature documenting acute food and fluid intake impacting reliability of measurement (Going et al, 1993;Horber et al, 1992;Thomsen et al, 1998;Vilaca et al, 2009). About 70% of reviewed studies failed to report any detail on subject positioning on the scanning bed and while most studies provided details of the type and model of DXA scanner used, ~30% failed to report the software version used during analysis, despite the fact that these factors are known to influence reliability of measurement (Kistorp & Svendsen, 1998;Lambrinoudaki et al, 1998;M. Lohman et al, 2009;Van Loan et al, 1995).…”