P. Mystkowski et al., Validation of whole-body magnetic resonance spectroscopy as a tool to assess murine body composition, INT J OBES, 24(6), 2000, pp. 719-724
OBJECTIVE: To evaluate proton magnetic resonance spectroscopy (MRS) as a to
ol for the non-invasive assessment of murine body composition.
DESIGN: Twenty C57/BL6 male mice with a wide range of body adiposities unde
rwent both pre- and post-mortem whole-body MRS to assess body composition.
MRS measures were compared to the results obtained by chemical carcass anal
ysis, the current 'gold standard' for determination of body composition.
MEASUREMENTS: Areas under the curve (AUC) for lipid and water peaks of whol
e body MRS spectra (AUC(lipid) and AUC(H2O), respectively) were used to det
ermine percentages of body fat (%FAT(MRS)) and fat free mass by MRS (%FFMMR
S) Total body fat, total body water, fat free mass, and total lean mass wer
e determined by chloroform/methanol extraction of lipid from dessicated who
le carcass and compared to MRS measures (%FAT(MRS), %FFMMRS, AUC(lipid), an
d AUC(H2O)) The variability of the MRS technique was assessed by determinin
g the coefficients of variation (COV) associated with %FAT(MRS), AUC(lipid)
, and AUC(H2O) for mice of three different adiposities.
RESULTS: %FAT(MRS) in live mice was highly correlated with body fat percent
age (r = 0.994, P < 0.001) and total body fat (r = 0.980, P < 0.001) derive
d from chemical carcass analysis over a broad range of adiposities (7-48% b
ody fat content by carcass analysis). There was no difference in %FAT(MRS)
measured pre- vs post-mortem (r = 1.00, P < 0.001). AUC(lipid) was highly c
orrelated with chemically derived total fat mass (r = 0.996, P < 0.001) and
body fat percentage (r = 0.981, P < 0.001), while %FFMMRS was strongly cor
related to chemical determinations of percentage body water (r = 0.994, P <
0.001), percentage fat free mass (r = 0.993, P < 0.001), and percentage le
an mass (r = 0.792, P < 0.001). AUC(H2O) was strongly associated with carca
ss analysis determinations of total body water (r = 0.964, P < 0.001), tota
l fat free mass (r = 0.953, P < 0.001), and total lean mass (r = 0.89, P <
0.001). In mice of 6%, 12%, and 43% body fat, COVs determined for %FAT(MRS)
and AUC(lipid) were less than 10%. The COVs for AUC(H2O) were less than 2%
.
CONCLUSIONS: MRS provides precise, accurate, rapid, and non-invasive measur
es of body fat, body water, fat free mass, and lean mass in living mice wit
h a broad range of adiposities.