Reproducibility of DXA: Potential Impact on Serial Measurements and Misclassification of Osteoporosis

Phillipov, George; Seaborn, C. J.; Phillips, Patrick J.

doi:10.1007/s001980170157

Cited by 37 publications

(9 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For an increase in BMD to be clinically significant, there must be a change of 0.050 g/cm 2 (or a 4-7% increase) (Phillipov, Seaborn & Phillips, 2001;Phillips & Phillipov, 2006). Although the participants in the present study experienced a statistically significant increase (p<0.05) in both the lumbar region of the spine (L1-L4) and the femur neck (p=0.00), the increase in BMD at the femur neck was the only area to have a clinically significant increase (9.4%).…”

Section: Discussionmentioning

confidence: 99%

The effect of a 12-week exercise programme on bone mineral density in young South African females

Egbunike¹,

Wood²,

Grant³

2010

Af. J. Phys, Health Edu. Rec. & Dance

View full text Add to dashboard Cite

A longer lifespan has increased the emphasis on bone health, which is often compromised with age. Studies have shown that exercise yields a positive influence on bone mineral density (BMD), especially when done during the early years of life (second to third decades). A 12-week study was undertaken using dual-energy X-ray absorptiometry (DEXA) to assess changes in BMD (total, lumbar and neck of the femur) in 70 females (mean age 20.1 years) before and after a 12-week intense exercise programme. The exercise programme consisted of a daily 5 km run and two isolated resistance exercises for each body region. The Wilcoxon signed-rank test was used to determine if any changes took place after the intervention. Statistically significant increases occurred in only the BMD of the lumbar spine (L1-L4) (2.54%) and neck of the femur (NOF) (9.35%), with clinical significance shown only in the NOF, as the increase was greater than 7%. This study's results thus indicate that a relatively short intense intervention of 12 weeks can improve BMD in young women, thereby possibly assisting in preventing osteoporosis and fractures later on in life. Further studies are recommended to determine the minimum time needed to achieve a clinically significant change in the total BMD as well as the lumbar spine BMD.

show abstract

Section: Discussionmentioning

confidence: 99%

The effect of a 12-week exercise programme on bone mineral density in young South African females

Egbunike¹,

Wood²,

Grant³

2010

Af. J. Phys, Health Edu. Rec. & Dance

View full text Add to dashboard Cite

show abstract

“…In addition, whereas such models enable statistical inference for the effect of several covariates on the mean change in BMD, statistical inference cannot be easily performed on the residual SD, and hence judgment on factors affecting precision remains qualitative. A different approach has been taken to analyze precision by Phillipov et al (26) They performed a nonparametric linear regression on 12 serial BMD measurements in 24 subjects. They were able to perform statistical inference (e.g., confidence interval) on the components of the overall variance using bootstrap techniques and analysis of covariance (ANCOVA).…”

Section: Discussionmentioning

confidence: 99%

Heteroscedastic Regression Analysis of Factors Affecting BMD Monitoring

Sadatsafavi

Moayyeri

Wang

et al. 2008

Journal of Bone and Mineral Research

View full text Add to dashboard Cite

Identifying factors affecting BMD precision and interindividual heterogeneity in BMD change can help optimize BMD monitoring. BMD change for the lumbar spine and total hip for short-term reproducibility (n ‫ס‬ 328) and long-term clinical monitoring (n ‫ס‬ 2720) populations were analyzed with heteroscedastic regression using linear prediction for mean (monitoring population only) and log-linear prediction for SD (both populations). For clinical monitoring, male sex, baseline body mass index (BMI), and systemic corticosteroid use were associated with greater SD of BMD change. Weight gain was negatively associated with SD for the hip, whereas height change was positively associated with SD for the spine. Each additional year of monitoring increased the SD by 6.5-9.2%. Osteoporosis treatment affected mean change but did not increase dispersion. For short-term reproducibility, performing scans on a different day increased the SD of measurement error by 38-44%. Baseline BMD, difference in bone area, and a repeat scan performed by different technologists were associated with higher measurement error only for the hip. For both samples, heteroscedastic regression outperformed models that assumed homogeneous variance. Heteroscedastic regression techniques are powerful yet underused tools in analyzing longitudinal BMD data and can be used to generate individualized predictions of BMD change and measurement error.

show abstract

“…Because 26.6% of women had initial BMDs performed by a pencil-beamed device, we attempted to make corrections when necessary on subsequent BMDs performed by fan-beamed devices. 16 We classified women using the lowest available T score at either the femoral neck, total hip, or lumbar spine into the following categories: normal (T scores > −1.0), mild osteopenia (T scores −1.0 to −1.49), moderate osteopenia (T scores −1.5 to —1.99), and severe osteopenia (T scores −2.0 to −2.49). Because all three sites were not always reported, the most frequent sites available for review were the spine plus total hip (54.6%), spine plus total hip and femoral neck (37.8%), and spine plus femoral neck (4.74%).…”

Section: Methodsmentioning

confidence: 99%

Changes in Nonosteoporotic Bone Density and Subsequent Fractures in Women

Pfister¹,

Welch²,

John³

et al. 2016

Southern Medical Journal

View full text Add to dashboard Cite

Objectives Osteopenia is considerably more common than osteoporosis and accounts for most of the fracture burden in women older than 50 years. It is uncertain when to initiate treatment in osteopenia. We sought to determine in women with osteopenia what effect transitioning to lower categories had on subsequent fracturing. Methods We surveyed 1150 women from office-based practices who had initial normal or osteopenic bone mineral densities (BMDs) and who were retested after 5.75 years. We classified categories related to baseline T scores as follows: normal (> −1.0), mild osteopenia (−1.0 to −1.49), moderate osteopenia (−1.5 to −1.99), and severe osteopenia (−2.0 to −2.49). We determined during a 9.6-year follow-up period the fracture occurrence in those who maintained their initial category status or transitioned into lower categories. Results Transitioning to lower categories was not significantly different among baseline osteopenic categories but significantly more than normal baseline BMDs. Total fractures, individuals fracturing, and major fractures were significantly more, with baseline T scores of ≤ −1.5 (<0.001). Although only 10.2% transitioned to osteoporosis, 90.5% of these transitions occurred with baseline T scores ≤ −1.5 and accounted for significantly more fractures than baseline T scores of > −1.5. Conclusions Most subsequent fractures and transitions to osteoporosis occurred with baseline T scores ≤ −1.5. Clinical risk factors need to be used to determine at what T score threshold treatment would be cost effective.

show abstract

Reproducibility of DXA: Potential Impact on Serial Measurements and Misclassification of Osteoporosis

Cited by 37 publications

References 17 publications

The effect of a 12-week exercise programme on bone mineral density in young South African females

The effect of a 12-week exercise programme on bone mineral density in young South African females

Heteroscedastic Regression Analysis of Factors Affecting BMD Monitoring

Changes in Nonosteoporotic Bone Density and Subsequent Fractures in Women

Contact Info

Product

Resources

About