JOURNAL OF BONE AND MINERAL RESEARCH Volume 5, Number It, 1990 Mary Ann Liebert, Inc., Publishers
Letter to the Editor Effect of Season on Activity and Bone Mineral Density To the Editor: Bergstralh et al. recently reported (JBkIR 5371-377, 1990) that bone mineral density (BMD) in the lumbar spine of postmenopausal women was 1.4% (0.015 g/cm2) higher in August-November than in February-May when measured at the Mayo Clinic by dual-photon absorptiometry using gadolinium 153. It is not clear whether this reflects changes in the women measured or in the equipment used to measure them. Because the study extended over 36 months it was doubtless necessary to replace the gadolinium 153 source several times. Gadolinium replacements can increase or decrease the results of bone mineral density measurements in patients and quality control phantoms by 0.015 g/cm', or even more, using dual-photon absorptiometry. ('-') Sources are often replaced annually, which means seasonally. Before it can be accepted that bone mineral density fluctuated seasonally in their patients, the authors need to exclude seasonal variations in their equipment by displaying the results of weekly measurements of BMD on an anthropomorphic phantom. Stability measurements on bone powders or aluminum cylinders are not sufficient to document instrumental stability because the artifacts introduced by gadolinium source changes can be undetectable when measurements are carried out on a thin object, such as a bone powder or small cylinder, yet grossly apparent
To the Editor: Doctor Neer raises an important point pertinent to all longitudinal measurements of bone mineral made with the newly replaced dual-energy (isotope) absorptiometry method. We believe that this aspect of our study was well controlled and that instrument-related factors can be excluded as the explanation for the seasonal variation described. Our study was performed in 1987 and 1988 using a Mayo-built dual-photon absorptiometry (DPA) instrument with a commercial lsaGd source. On this instrument sources were changed every 12-15 months, and two source changes occurred during the course of this 2-year study. The quality control program in use at this time was de-
when carried out simultaneously on patients or a thick phantom.".') If the authors d o not have appropriate quality control data they may at least be able to demonstrate whether the observed seasonal variations in bone mineral density are still statistically significant after changes in gadolinium 153 source, and source age, are taken into account. Their results are so important for the design of future clinical trials, and for the evaluation of bone mineral density in individual subjects, that it is important to show that the small fluctuations observed are not merely artifacts. Robert M. Neer Chief, Mineral Metabolism Unit Director, Mallinckrodt General Clinical Research Center Massachusetts General Hospital Boston. MA 02114
REFERENCES 1. Kelly TL, Slovik DM, Neer RM 1989 J Bone Min Res 4:663669. 2. Dawson-Hughes 9, et al. 1989 Calcif Tissue Int 44:251-257. 3. Lindsay R , et al. 1987 Calcif Tissue In1 41:293-294. 4. Ross PD, et al. 1988 Radiology 166:523-527. 5 . Nilas L, et al. 1988 Bone Min 3:305-315.
scribed in detail in a previous publication.'') We recognized the need for independent calibration to correct errors introduced by the changing strength over the useful life of a source and by potential differences in the energy spectrum of different sources due to impurities, and we published our observations with different instruments at that time.") Thus the instrument was calibrated every 3 weeks using five pieces of human bone of known weight (ashed at 600°C; BMD 1-1.9 g/cm*), which were scanned in triplicate in 20 cm water. From these samples a standard curve was constructed for conversion of computer units into grams of bone ash. The standard curves were always linear and had the same slope but occasionally differed slightly in actual bone mineral content (BMC) value. These samples
1271
1272 had bone edges similar to those encountered in the spine, being of varying thickness and shape. When these samples were scanned on the Lunar DP3 instrument we were able to recognize the source strength dependency of bone mineral density (BMD) measurements and the changes related to a new source, which are described by Doctor Neer. Because source strength and different sources also affected our instrument as described previously,(ll the actual and baseline BMD of one of the ashed bones was used as a BMD correction factor for the measured patient BMD. As a consequence these effects were neutralized in the data that we reported for BMD seasonality. In addition we calculated the T I , ,of the isotope from the count rate in air to check instrument hardware and isotope and scanned daily bone samples in water for short-term precision estimations. This has also been described in more detail elsewhere."' In addition, since patients were entered into the study at
LETTER TO THE EDITOR different times during the year, the instrument-related problems outlined here would have been obscured.
E.J. Bergstralh, M. Sinaki, M.D., K.P. Offord, H.W. Wahner, M.D., and L. Joseph Melton 111, M.D. Mayo Clinic Rochester, MN
REFERENCE 1. Dunn WL, Kan SH, Wahner HWW 1987 Errors in longitudi-
nal measurements of bone mineral: Effect of source strength in single and dual photon absorptiometry. J Nucl Med 28(1 I): I751- 1757.
Letter to the Editor Effect of Season on Activity and Bone Mineral Density To the Editor: Bergstralh et al. recently reported (JBkIR 5371-377, 1990) that bone mineral density (BMD) in the lumbar spine of postmenopausal women was 1.4% (0.015 g/cm2) higher in August-November than in February-May when measured at the Mayo Clinic by dual-photon absorptiometry using gadolinium 153. It is not clear whether this reflects changes in the women measured or in the equipment used to measure them. Because the study extended over 36 months it was doubtless necessary to replace the gadolinium 153 source several times. Gadolinium replacements can increase or decrease the results of bone mineral density measurements in patients and quality control phantoms by 0.015 g/cm', or even more, using dual-photon absorptiometry. ('-') Sources are often replaced annually, which means seasonally. Before it can be accepted that bone mineral density fluctuated seasonally in their patients, the authors need to exclude seasonal variations in their equipment by displaying the results of weekly measurements of BMD on an anthropomorphic phantom. Stability measurements on bone powders or aluminum cylinders are not sufficient to document instrumental stability because the artifacts introduced by gadolinium source changes can be undetectable when measurements are carried out on a thin object, such as a bone powder or small cylinder, yet grossly apparent
To the Editor: Doctor Neer raises an important point pertinent to all longitudinal measurements of bone mineral made with the newly replaced dual-energy (isotope) absorptiometry method. We believe that this aspect of our study was well controlled and that instrument-related factors can be excluded as the explanation for the seasonal variation described. Our study was performed in 1987 and 1988 using a Mayo-built dual-photon absorptiometry (DPA) instrument with a commercial lsaGd source. On this instrument sources were changed every 12-15 months, and two source changes occurred during the course of this 2-year study. The quality control program in use at this time was de-
when carried out simultaneously on patients or a thick phantom.".') If the authors d o not have appropriate quality control data they may at least be able to demonstrate whether the observed seasonal variations in bone mineral density are still statistically significant after changes in gadolinium 153 source, and source age, are taken into account. Their results are so important for the design of future clinical trials, and for the evaluation of bone mineral density in individual subjects, that it is important to show that the small fluctuations observed are not merely artifacts. Robert M. Neer Chief, Mineral Metabolism Unit Director, Mallinckrodt General Clinical Research Center Massachusetts General Hospital Boston. MA 02114
REFERENCES 1. Kelly TL, Slovik DM, Neer RM 1989 J Bone Min Res 4:663669. 2. Dawson-Hughes 9, et al. 1989 Calcif Tissue Int 44:251-257. 3. Lindsay R , et al. 1987 Calcif Tissue In1 41:293-294. 4. Ross PD, et al. 1988 Radiology 166:523-527. 5 . Nilas L, et al. 1988 Bone Min 3:305-315.
scribed in detail in a previous publication.'') We recognized the need for independent calibration to correct errors introduced by the changing strength over the useful life of a source and by potential differences in the energy spectrum of different sources due to impurities, and we published our observations with different instruments at that time.") Thus the instrument was calibrated every 3 weeks using five pieces of human bone of known weight (ashed at 600°C; BMD 1-1.9 g/cm*), which were scanned in triplicate in 20 cm water. From these samples a standard curve was constructed for conversion of computer units into grams of bone ash. The standard curves were always linear and had the same slope but occasionally differed slightly in actual bone mineral content (BMC) value. These samples
1271
1272 had bone edges similar to those encountered in the spine, being of varying thickness and shape. When these samples were scanned on the Lunar DP3 instrument we were able to recognize the source strength dependency of bone mineral density (BMD) measurements and the changes related to a new source, which are described by Doctor Neer. Because source strength and different sources also affected our instrument as described previously,(ll the actual and baseline BMD of one of the ashed bones was used as a BMD correction factor for the measured patient BMD. As a consequence these effects were neutralized in the data that we reported for BMD seasonality. In addition we calculated the T I , ,of the isotope from the count rate in air to check instrument hardware and isotope and scanned daily bone samples in water for short-term precision estimations. This has also been described in more detail elsewhere."' In addition, since patients were entered into the study at
LETTER TO THE EDITOR different times during the year, the instrument-related problems outlined here would have been obscured.
E.J. Bergstralh, M. Sinaki, M.D., K.P. Offord, H.W. Wahner, M.D., and L. Joseph Melton 111, M.D. Mayo Clinic Rochester, MN
REFERENCE 1. Dunn WL, Kan SH, Wahner HWW 1987 Errors in longitudi-
nal measurements of bone mineral: Effect of source strength in single and dual photon absorptiometry. J Nucl Med 28(1 I): I751- 1757.
