Multiple linear regression analysis for bone mineral content and bone mineral density of the lumbar spine and hip

Junhao GUO

ABSTRACT

BACKGROUND:

Bone densitometry using dual energy X-ray absorptiometry(DXA)is the "gold standard" for osteoporosis diagnosis.However,it has limited in clinical examination for it only supply area bone mineral density(BMD).

OBJECTIVE:

To form BMD coefficient by correcting area BMD based on body mass index(BMI),in addition,to predict whole body BMD via multivariate linear regression equation.DESIGN,TIME AND

SETTING:

A prospective study was performed at the Nanjing General Hospital of Nanjing Military Area Command of Chinese PLA between October 2004 and June 2008.

PARTICIPANTS:

Twenty subjects,who received DXA examination at the Nanjing General Hospital,Nanjing Military Area Command of Chinese PLA were selected.No relative operation which would affect BMD performed before examination.

METHODS:

The BMD of waist,hip as bone mineral content(BMC)of whole body was measured.Height,body weight of subjects was measured and calculates BMI,which called BMD coefficient after correction.It was obtained by BMD/BMI?10.BMC coefficient was obtained by BMC/body weight ?10.MAIN OUTCOME

MEASURES:

The BMD of waist,hip,as well as BMC of whole body.Independent variable BMC and dependent variable BMD were made respectively,and to do linear correlation analysis.

RESULTS:

There was a correlation between BMC and BMD(P=0.000),which was more significant after correction(P=0.000).The multivariate linear regression equation between BMC and BMD wasBMC=(0.065+0.415?G1+0.270?G2)?T/10.In this formula,G1 was corrected value of right hip BMD,G2 was the corrected value of waist BMD,and T was body weight.The predictive BMC was basically tallies with actual value.

CONCLUSION:

BMD of whole body can be predicted based on BMD of waist and hip,and the BMD coefficient may be the potential index for BMD evaluation.