Editorials

Bone mass measurements: reasons to be cautious

Interpreting a change in bone mass requires caution. Problems arise from the imprecision of instruments, the increasing risks with multiple measurements of rejecting the null hypothesis when it is true (type I error), assumptions that bone density is a volumetric density, suggestions that treatment causes linear increases in bone mass, and assumptions that the pattern of bone loss is reversible.

Dual energy x ray absorptiometry provides the most precise measurements of bone mass. Even with this new technique, however, the size of clinically important changes in bone mass is less than the measurement error. A walk around the room causes the measurement to change by up to 6% (at the hip), which corresponds to six years of bone lost at the usual rate. Thus, for individual patients, clinically important changes in bone mass may take several years to detect. Only severe loss can be measured with two determinations. Alterations in machine function must be anticipated and careful quality control applied.²

In clinical studies the newer technology has improved the power to detect changes in bone mass, and a few subjects are sufficient to show dramatic changes (such as bone loss with gonadotrophin releasing hormone agonists³). In normal elderly subjects we estimated that a sample size of 90 over three years would be needed to detect changes in bone mass between placebo and treated groups of 0.8% a year.

Because of the propensity to measure many different anatomical locations there is an unacknowledged risk of type I errors. Commercial densitometers produce a series of related measurements, which are often analysed unquestioningly. For example, a single scan of the proximal femur yields five measurements: femoral neck, trochanteric region, intertrochanteric region, total hip, and "Ward's triangle" (three adjacent regions, their total, and a subset respectively). This process could become worse - the total body bone mineral measurements are divided into 14 areas. Will we be subjected to even larger tables of data every time a study reports bone mass? The sites of physiological interest should be carefully defined when the study is designed.

Another common problem in interpreting changes in bone mass arises when there is also change in skeletal size, which applies to children and teenagers. The only method that measures true bone density is quantitative computed tomography.⁴ Dual energy x ray absorptiometry measures the amount of bone mineral in a projected area (areal density). Unfortunately, the term "bone density" has been loosely applied: with no change in the true bone density, the measured areal bone density will increase with growth. An independent measurement of the depth of the bone is needed to estimate the true density, and even then there is error in calculating the size of the bone.

Changes in bone mass after pharmacological treatment may not be linear over time, especially with "antiresorbing" drugs such as oestrogen, calcitonin, and the bisphosphonates. (???) these drugs the recently resorbed bone cavities continue to in with new bone, but without further resorption there is eventually no further bone formation. Bone mass increase(/???) the first few months then gradually reaches a plateau.⁵ The total amount of new bone formed is limited by the initial rate of bone formation. This limit of 5-10% can be termed the remodelling barrier.⁶ Only drugs that independently stimulate bone formation by osteoblasts (such as fluoride) can overcome this barrier. Thus, encouraging results of brief studies do not imply continuing increases in bone mass.

Even without technical error, assessing the effects of changes in bone mass on bone strength is difficult. In ordinary populations it is reasonable to assume that lifestyle risk factors (such as smoking), which reduce the bone mass by 5%,⁷ should increase the fracture rate by 50%. But pharmacological interventions that increase bone mass by 5% do not necessarily decrease fracture risk by 50% because the structure of trabecular bone is damaged with bone loss, especially that associated with the menopause. Entire trabecular struts are lost.⁸ With the antiresorbing drugs the disconnected trabeculae are not reconnected but the existing trabeculae widen. Also, some of the increase in density occurs without any increase in bone volume because lowering bone turnover eventually results in a larger proportion of old bone, which is denser than newly formed bone. Fluoride causes a disparity between bone mass, which increases dramatically, and strength, which decreases.⁹ Thus the relation between the relative risk of fracture and bone mass cannot yet be used to predict whether a change in bone mass will change the risk of fracture.

S M Ott 

1. Ott SM. When bone mass fails to predict bone failure. Calcif Tissue Int 1993;53(suppl 1):S7-13.
2. Orwoll ES, Oviatt SK, Biddie JA. Precision of dual-energy x-ray absorptiometry: development of quality control rules and their application in longitudinal studies. J Bone Mineral Res 1993;8:693-9. [Medline]
3. Rico H, Arnanz F, Revilla M, Perera S, Iritia M, Villa LF, et al. Total and regional bone mineral content in women treated with GnRH agonists. Calcif Tissue Int 1993;52:354-7. [Medline]
4. Gilsanz V, Roe TF, Mora S, Costin G, Goodman WG. Changes in vertebral20bone density in black girls and white girls during childhood and puberty. N Engl J Med 1991;325:1597-600. [Abstract]
5. Harris ST, Watts NB, Jackson RD, Genant HK, Wasnich RD, Ross P, et al. Four-year study of intermittent cyclic etidronate treatment of postmenopausal osteoporosis: three years of blinded therapy followed by one year of open therapy. Am J Med 1993;95:557-67. [Medline]
6. Ott SM. Clinical effects of bisphosphonates in involutional osteoporosis. J Bone Mineral Res 1993;8(suppl 2):S597-606.
7. Hopper JL, Seetuna E. The bone density of female twins discordant for tobacco use. N Engl J Med 1994;330:387-92. [Abstract/Free Full Text]
8. Kleerekoper M, Villanueva AR, Stanciu J, Rao DS, Parfitt AM. The role of three-dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif Tissue Int 1985;37:594-7. [Medline]
9. Riggs BL, Hodgson SF, O'Fallon M, Chao ES, Wahner HW, Muhs JM, et al. Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N Engl J Med 1990;322:802-9. [Abstract]