Bone density test results during osteoporosis treatment can be misleading, a UCSF study has found

By Leslie Harris on March 07, 2000

Many women are being treated for osteoporosis with medications and their
progress is checked periodically with measurements of bone density. About one
out of five women taking these agents appear to lose bone density during the
first year of treatment, causing doctors to change the treatment.  But this
loss may be misleading, according to a UC San Francisco study that showed women
with the greatest amount of bone density loss during the first year of
treatment for osteoporosis were the most likely to gain bone density when the
same treatment is continued.

The study results are published in the March 8 issue of the Journal of the
American Medical Association.

When an osteoporosis patient loses bone density during the first year of
treatment with alendronate or raloxifene, physicians often change the treatment
because they believe the medicine is not working for the patient, said lead
author Steve Cummings, MD, UCSF professor of internal medicine and
epidemiology. But this may not be the best course of action for physicians to
take, he said.

“It’s important if women are being monitored by periodic tests of bone density
and they lose bone in the first year or two to not change treatment,” Cummings
said. “Most doctors would have changed or stopped the treatment. They say the
woman hasn’t responded so they often switch her to another drug or add a second
drug. Adding a second drug adds costs, risks and side effects. This study shows
that patients should make sure they are taking the drug regularly and properly
but should not change the treatment or add another one. “

The study examined the results of studies of 2,634 women who were taking five
milligrams of alendronate daily for treatment of osteoporosis and 3,954 women
taking 60 or 120 milligrams of raloxifene daily. Of the women who lost bone
density during the first year of treatment with either agent, more than 80
percent gained bone density the second year.

“Even those who seemed to lose more than four percent of their bone, the most
extreme amount, were the ones most likely to gain in the next year and they
were the ones to gain the most,” Cummings said. On average, the women who lost
bone density the first year saw an increase in bone density of 4.7 percent the
following year.

Of the 2,634 women taking alendronate, 484, or 18 percent, lost bone density
the first year. And of the 3,954 women taking raloxifene, 1,392, or 35 percent,
saw a loss of bone density during the first year of treatment.

Conversely, when the researchers looked at women who gained a large amount of
bone density that first year-a gain of more than eight percent—the majority
lost an average of one percent bone density the second year.

Usually, women taking raloxifene or alendronate for osteoporosis gain one to
three percent in bone density during the first year of treatment, Cummings
said   “When your changes are different than the average-if you gain more or
lose more than is expected—in first year you are probably gaining or losing
because of errors in the measurement,” he said  

These variable results are due to subtle errors in the test and not problems
with patients or the treatment, Cummings said. How a patient is positioned
while undergoing the test or how the machine is working on a particular day can
cause small errors. This variability leads to a statistical principle called
“regression to the mean,” which predicts that patients with unusual responses to
treatment are likely to have more typical responses if treatment is continued
and the measurement is repeated.

The bone density test works by passing low doses of x-ray beams through a
patient’s spine, hip or in some cases, the forearms and entire body as she lays
on a table. The machine measures the amount of x-ray absorbed by calcium in the
body. This shows how much calcium is in the bone, an indicator of how much bone
density has been lost or gained.

While the test is an excellent way to determine a woman’s risk of getting a
fracture, whether she has osteoporosis or what kind of treatment she needs, it
may not be the best method of monitoring treatment results for this disease
that affects 10 to 15 percent of postmenopausal Caucasian and Asian women. The
rate is lower in Hispanic and African American women because they have higher
bone density, Cummings said.

“This study does raise doubts about whether it’s really worth while to get
periodic measurements of bone density to monitor treatment,” Cummings said.
“Monitoring treatments with bone density is new, it has only been used for a few
years. We aren’t certain how to interpret the results of the test very well.”

“The practical meaning of this is don’t be stressed out by losing bone density
during treatment,” he said. “If the tests show that you have lost bone, make
sure you are taking the drug regularly and correctly. But don’t change
treatment. If you gain a lot, don’t get too excited because your measurements
may slip back the next year.”

Data for the study was collected through the Fracture Intervention Trial funded
by Merck Research Laboratories and the Multiple Outcomes of Raloxifene
Evaluation trial, funded by Eli Lilly and Company. UCSF investigators conducted
these analyses without support from either company.

Other study authors include: Lisa Palermo, MA, associate specialist in the UCSF
department of epidemiology and biostatistics; Dennis Black, PhD, UCSF assistant
professor of epidemiology and biostatistics; Jim Pearson, UCSF project
assistant in the department of epidemiology and biostatistics; Warren Browner,
MD, UCSF associate professor in residence of general internal medicine; Terri
Blackwell, MA, UCSF statistician, department of epidemiology and biostatistics;
Robert Marcus, MD, Veterans Affairs Medical Center, Palo Alto, California;
Robert Wallace, MD, professor, department of preventive medicine, University of
Iowa, Iowa City, Iowa; Stephen Eckert, PhD, Eli Lilly and Company,
Indianapolis, Indiana.