Calcium and Vitamin D3 Supplementation Prevents Bone Loss in
the Spine Secondary to Low-Dose Corticosteroids in Patients with
Rheumatoid Arthritis
A Randomized, Double-Blind, Placebo-Controlled Trial
Lenore M. Buckley, MD, MPH; Edward S. Leib, MD; Kathryn S. Cartularo, RN; Pamela M. Vacek, PhD;
and Sheldon M. Cooper, MD
Background: Therapy with low-dose corticosteroids is
commonly used to treat allergic and autoimmune diseases.
Long-term use of corticosteroids can lead to loss of bone
mineral density and higher risk for vertebral fractures.
Calcium and vitamin D3 supplementation is rational ther-
apy for minimizing bone loss, but little evidence for its
effectiveness exists.
Objective: To assess 1) the effects of supplemental cal-
cium and vitamin D3 on bone mineral density of patients
with rheumatoid arthritis and 2) the relation between the
effects of this supplementation and corticosteroid use.
Design: 2 year randomized, double-blind, placebo-
controlled trial.
Setting: University outpatient-care facility.
Patients: 96 patients with rheumatoid arthritis, 65 of
whom were receiving treatment with corticosteroids
(mean dosage, 5.6 mg/d).
Intervention: Calcium carbonate (1000 mg/d) and vita-
min D3 (500 lU/d) or placebo.
Measurements: Bone mineral densities of the lumbar
spine and femur were determined annually.
Results: Patients receiving prednisone therapy who were
given placebo lost bone mineral density in the lumbar
spine and trochanter at a rate of 2.0% and 0.9% per year,
respectively. Patients receiving prednisone therapy who
were given calcium and vitamin D3 gained bone mineral
density in the lumbar spine and trochanter at a rate of
0.72% (P = 0.005) and 0.85% {P = 0,024) per year, respec-
tively. In patients receiving prednisone therapy, bone min-
eral densities of the femoral neck and the Ward triangle
did not increase significantly with calcium and vitamin D3.
Calcium and vitamin D3 did not improve bone mineral
density at any site in patients who were not receiving
corticosteroids.
Conclusion: Calcium and vitamin D3 prevented loss of
bonemineraidensity in the lumbar spine and trochanter in
patients with rheumatoid arthritis who were treated with
low-dose corticosteroids.
Ann Intern Med. 1996;125;96]-968.
From Medical College of Virginia, Richmond, Virginia; and the
University of Vermont, CollegL- of Medicine, Burlington, Ver-
mont. For current author addresses, see end of text.
L ong-term administration of low doses of corti-
costeroids is common in the treatment of auto-
immune diseases, chronic obstructive lung disease,
asthma, and allergic conditions. Although treatment
with high doses of corticosteroids causes osteoporo-
sis (especially in trabecular bone, such as that found
in the lumbar spine [1]), use of corticosteroids in
low doses (<10 mg/d) was thought to be associated
with few substantial side efFects (2). However, re-
cent research suggests that use of low doses of
corticosteroids is also associated with loss of bone
mineral density in the lumbar spine (3-6) and that
patients receiving continuous therapy with low-dose
corticosteroids have a higher rate of vertebral frac-
ture (7-9).
Corticosteroids cause osteoporosis by several
mechanisms, such as decreasing levels of sex ste-
roids (10, 11) and direct effects on osteoblast (12)
and osteoclast function (13). Use of corticosteroids
also decreases absorption of intestinal calcium (14-
17), thereby causing secondary hyperparathyroidism
(18). Some studies (19, 20) have shown that vitamin
D , and its more potent analogues can improve cal-
cium absorption in patients receiving corticoste-
roids. Treatment with 1,25-dihydroxyvitamin D3 (cal-
citriol) and calcium stabilized bone mineral density
in the lumbar spine in patients receiving moderate
to high doses of eorticosteroids (21). However, cal-
citriol is a potent vitamin D analogue that can cause
hypercalcemia and hypercalciuria (22). The associ-
ated toxicity and expense make it impractical for
widespread use for prevention of bone loss in all
patients receiving long-term treatment with low-
dose corticosteroids.
Vitamin D3, the parent compound, is generally
well tolerated and rarely causes side effeets when
used in the recommended dosage (400 to 800 IU/d).
It has been suggested that vitamin D-, and calcium
should be given to all patients receiving long-term
eorticosteroid therapy (23). However, few data sup-
port this approach. Early studies (24) suggested that
calcium and pharmacologic dosages of vitamin D3
(50 000 IU three times a week) increased bone min-
eral density in patients taking long-term corticoste-
roids, but bone mineral density was measured only
American CoJlege of Physicians 961
in the radius. Adachi and colleagues (25) recently
reported that treatment with calcium (1000 mg/d)
and vitamin D3 (50 000 IU/wk) did not prevent
bone loss associated with moderate- to high-dose
treatment with corticosteroids, but Healy and col-
leagues (26) found that calcium and vitamin D3
prevented substantial loss of bone mineral density in
a similar group of patients starting corticosteroid
treatment. No randomized, controlled, clinical trials
have shown the effectiveness of calcium and vitamin
D3 supplementation for prevention of bone loss sec-
ondary to long-term low-dose corticosteroid treat-
ment. Therefore, we studied the eifects of calcium
(1000 mg/d) and vitamin D3 (500 TU/d) supplemen-
tation on the bone mineral density of patients with
rheumatoid arthritis and contrasted the efficacy of
this supplementation in patients who were and
those who were not receiving corticosteroids.
Methods
Patients
Patients with rheumatoid arthritis who were fol-
lowed at the University of Vermont, College of Med-
icine (n = 354) were identified through a computer
listing by diagnostic code. All identified patients were
contacted by mail and asked whether they were
interested in participating in a study of the effects of
rheumatoid arthritis and medication on bone min-
eral density. Patients were eligible if they were be-
tween 18 and 65 years of age and had a diagnosis of
rheumatoid arthritis as defined by the revised
American College of Rheumatology criteria (27),
serum creatinine level less than 176.8 jixmoI/L, and
normal liver function. Patients were excluded if they
were receiving an anticonvulsant medication, hydro-
chlorothiazide, bisphosphonates, fluoride, calcitonin,
or calcitriol or if they had a history of malabsorp-
tion, hyperparathyroidism, immobilization, metabolic
bone disease, or thyroid disease with an abnormal
thyroid-stimulating hormone. One hundred thirty pa-
tients met these criteria, agreed to participate, and
entered the study between January 1991 and Janu-
ary 1992.
Study Design
The study was approved by the Committee on
Human Research of the University of Vermont, and
informed consent was obtained as patients entered
the study. Patients were stratified by sex, meno-
pausal status (premenopausal, postmenopausal with-
out estrogen replacement, or postmenopausal with
estrogen replacement), and current dosage of pred-
nisone (<10 mg/d o r > 1 0 mg/d). Patients were ran-
domly assigned in blocks of four to a treatment
962 15 December 1996 • Annals of Internal Medicine * Volume 125 • Number 12
group using a random-number table. Group assign-
ment was kept in sealed envelopes to ensure blind-
ing. Patients received either calcium carbonate
(1000 mg/d) and vitamin D3 (500 IU/d) (two tablets
twice a day of Os-Cal 250 + D, SmithKline Beecham
Healthcare, Pittsburgh, Pennsylvania) or an identi-
cal placebo. Study medication was taken with break-
fast and dinner.
Measurement
Demographic information and information about
risk factors for osteoporosis, such as smoking his-
tory, use of alcohol, menstrual history, parity, frac-
ture history, and use of medications (including thy-
roid medication and estrogen replacement therapy)
were obtained from questionnaires given at the be-
ginning of the study and annually thereafter. Cal-
cium intake was assessed using the Food Frequency
Questionnaire (28) and 3-day dietary history. Dis-
ease-related information included activity level, mea-
sured by the Framingham Activity Index (29); dis-
ease severity, measured by the Health Assessment
Questionnaire (30); and a severity score, based on
radiologic findings (31). Information about disease
duration and use of prednisone or disease-modiiying
antirheumatic drugs was ascertained by chart review.
Bone mineral densities of the femur (femoral
neck. Ward triangle, and trochanter) and lumbar
spine were measured in all patients by dual-energy
x-ray absorptiometry (Lunar Corp., Madison, Wis-
consin) at the beginning of the study and at yearly
follow-up visits. All studies were done using the
same machine. Stability of measurement was as-
sured with daily quahty-assurance calibration by the
manufacturer, weekly Lunar aluminum phantom in
water bath, and Hologic spine phantom three times
weekly. The coefficient of variation (calculated using
11 patients and three measurements) was 1.2% for
the anteroposterior spine and 2.2% for the femoral
neck. At the start of the study, the lateral scan of
the spine was thought to be the most sensitive mea-
surement for detecting changes induced by cortico-
steroids (32). However, we and others (33) have
found the reproducibility of this view with the avail-
able technology to be suboptimal, especially in pa-
tients with arthritis, because correct positioning is
difficult. We subsequently obtained anteroposterior
scans of the spine. For patients from whom only the
lateral view of the spine was available at baseline
{n = 48 [20 patients in the treatment group and 28
in the placebo group]), we determined the antero-
posterior value by adding the difference between the
anteroposterior and lateral scans obtained after the
first year of the study to the lateral scan obtained at
baseline. This provided adequate measurement be-
cause the differences between anteroposterior and
 lateral views for each patient remained relatively con-
stant (within 0.10 ± 0.07 g/em^ for 87% of patients).
Patients were seen by the study coordinator at
6-month intervals to assess compliance. At tbese
visits, pills were counted and patients were ques-
tioned about side effects of medication.
Statistical Analysis
Because loss of bone mineral density over time is
presumed to be proportional to the amount of bone
mineral density remaining, the yearly rate of change
(k) was estimated using data from the three time
points (baseline, year 1, and year 2) in the following
equation:
bone mineral density at year t =
(bone mineral density at baseline) • exp (k • t)
Differences in rates for the two treatment groups
were assessed by using the Mest. The effect of treat-
ment after adjustment for group differences in dis-
ease severity and duration was examined by linear
regression analysis. Patients who received pred-
nisone during the study and those who did not were
analyzed separately and in eombination. Clinical
eharacteristies of the two treatment groups were
compared using /-tests and chi-square tests. Poten-
tial interactions between treatment effects and pa-
tient eharacteristies were examined by analysis of
variance. All analyses were done using SPSS statis-
tical software (SPSS, Inc., Chicago, Illinois).
Results
Baseline eharacteristies of the patients eomplet-
ing both years of the study are listed in Table 1.
The patients studied were primarily middle-aged
white women. Although study groups were similar
for most eharacteristies, the patients receiving eal-
eium and vitamin Dj, tended to have disease of
greater severity (by radiologie stage) and longer du-
ration than did those receiving placebo. Among pa-
tients who were receiving cortieosteroids, the mean
dosage of prednisone at the initial visit was 5.9 mg/d
for the calcium and vitamin D3 group and 5.0 mg/d
for the plaeebo group. During the first year of
study, 68% of patients were taking prednisone
(mean dosage, 5.5 mg/d; range, 1 to 20 mg/d); dur-
ing the second year, 65% were taking prednisone
(mean dosage, 5.6 mg/d; range, 0.5 to 20 mg/d).
During the 2-year study period, 29 of the original
130 patients (22%) dropped out of the study. Fif-
teen of the patients who dropped out had been in
the ealeium and vitamin D3 group and 14 had been
in the placebo group. Fourteen patients withdrew
for personal reasons, 1 patient died, and 3 patients
developed serious illness (inflammatory bowel dis-
ease, caneer, and amyotrophic lateral sclerosis).
Eleven patients (7 in the calcium and vitamin D3
group and 4 in the placebo group) dropped out of
the study because of gastrointestinal toxicity. The
most common gastrointestinal symptoms in treated

Table 1. Baseline Characteristics of Patients Who Had Bone Mineral Density Measurements Available for All Time Points"

Characteristic Patients Receiving Prednisone Patients Not Receiving Prednisone

Calcium and Piacebo Group Calcium and Placebo Group
Vitamin D^ Group (n = 35) Vitamin D3 Group (n = 14)
(n = 31) (n = 16)

Age, y 51.9 ± 12-6 54.2 ±11.5 56.9 ± 8.6 53.3 ± 12,4

Femaie, n (%) 20 (65) 27 (77) 13(81) 13(93)
Postmenopausal women, n (%) 11 (55) 20 (74) 10(77) 10(77)
Postmenopajsai women using
estrogen, n (%) 2(18) 4(20) 1(10) 3(30)
Cumulative previous
corticosteroid use, mgf 10 950 (3550 to 20 075) 5657 (1525 to 21 900) 0 (0 to 585) 1011 (0 to 3832)
Duration of disease, y 10.5 ± 5,5 8.3 ± 7,6 10.6 ± 8 , 5 7,3 ± 4.4
Disease progression (on a scale of
1 to 4)* 2.8 ± 0 58 2,6 ± 0.66 2.9 ± 0.50 2.0 ± 0.88
Health Assessment Questionnaire
score (on a scale of 0 to 3) 0.97 + 0,74 0.93 ± 0,67 0.68 ± 0.82 0.89 ± 0,52
Functional ciass (on a scale of 1
to 4) 2.4 + 0,55 2,45 ± 0.56 2.4 ± 0.72 2,2 ±0.70
Activity level (on a scaie of 24 to
80)+ 30.7 ± 3.4 31.4 ± 3.4 29.6 ± 2,06 32 0 ± 2.80
Bone density, g/crrr^
Lumbar spine 1.10 + 0.16 1.13 ± 0.19 1.06 ± 0.16 '1.08 ± 0.15
Trochanter 0.75 ±0.18 0.74 ±0.16 0.70 ± 0,16 0.72 ±0,12
Femoral neck 0,85 ±0.17 0.83 ±0,13 0.79 ± 0.14 0.80 ± 0,07
Ward triangie 0.72 ± 0,20 0.73 ±0.15 0,66 ± 0.14 0.69 ± 0.12
Calcium intake by the Food
Frequency Questionnaire, mg/d 1091 ± 451 846 + 477 910 ± 569 904 ±410

- Time points for measurements were baseline, year 1, and year 2. Values are Ihe mean ± SD unless otherwise notec
t Values are median (interquartile range),
t P ^ 0,05 tor comparison between treatrnent groups.

15 December 1996 • Annals of internal Medicine • Volume 125 • Number 12 963

patients were indigestion (H = 6) and constipation
(« = 1). No cases of nephrolithiasis or hyperealee-
mia were seen in either group. Data are presented
for the 96 patients who remained in the study and
who had evaluahle data at all three time points. Of
these patients, only 9 (4 in the caleium and D^
group and 5 in the placebo group) took less than
80% of their medication, as assessed by pill count.
Bone Densitometry of the Lumbar Spine
The annual rate of change in bone mineral den-
sity is shown in the Figure. Patients in the caleium
and vitamin D3 group had an average increase in
bone mineral density in the lumbar spine of 0.63%
per year, and those in the placebo group had an
average decrease of 1.31% per year (F = 0.015).
The results of analyses done separately for patients
who were receiving prednisone and those who were
not indicate that this differenee was attributable to
the effects of caleium and vitamin D, supplementa-
tion in patients receiving prednisone. Patients who
reeeived prednisone during the study showed a
larger treatment effect attributable to calcium and
vitamin D3 than did those who did not receive
prednisone. The difference in the annual rate of
change in bone mineral density between patients in

Spine Femoral Neek Ward Triangle Trochanter

105 -1
104 -
103 -
102 -
101 -
100
99
98 -
-
- r
97-
96 -
9S-
94 -
93

105 -,
104 -
103 -
102 -
101 -
100
99
98
97
-
-
-
I
96 -
95 -
94 -
93

105
1 1

104
103
1

102
101 ___J
1

100
99
98 -
97 -
96
95
94
93 1

Years
Figure. Change in bone mineral density during the 2-year study period in patients with rheumatoid arthritis. Top. All patients Middle. Patients
receiving prednisone. Bottom. Patients not receiving prednisone. Soiid iines indicate values for patients in the calcium and vitamin D3 group; dashed lines
indicate values for patients in the placebo group. Bars represent 95% CIs. A significant difference was seen between tfie rate of cfiange in bone mineral
densities of the spine and trochanter over 2 years in all patients and in patients receiving prednisone.

964 15 December 1996 • Annals of Internal Medicine • Volume 125 • Number 12

the calcium and vitamin D3 group and those in the
placebo group is shown in Table 2. A positive value
indicates an increase in the rate of change in bone
mineral density in patients in the calcium and vita-
min D3 group compared with that in patients in the
placebo group. The difference in the annual rate of
change in bone mineral density of the lumbar spine
between patients in the calcium and vitamin D3
group and those in the placebo group was 2.65%
per year (adjusted for differences in disease severity
and duration). During the 2 years of the study, bone
mineral density changed only slightly in the lumbar
spine of patients who did not receive prednisone
(Figure). No statistically significant difference in the
annual rate of change in bone mineral density in the
lumbar spine was seen between patients in both
groups who did not receive prednisone (Table 2).
Bone Densitometry of the Femur
Treatment also had a statistically significant effect
at the trochanter in patients treated with corlicoste-
riods. Patients who received placebo lost 0.90% of
their bone mineral density per year, whereas pa-
tients who received calcium and vitamin D3 had a
gain of 0.85% per year (P = 0.024) (Figure). The
difference in the adjusted annual rate of change
between patients in the calcium and vitamin D, and
placebo groups was 2.08% per year (Table 2). Bone
mineral densities of the femoral neck and Ward
triangle were relatively stable during the 2 years of
the study in patients treated with corticosteroids,
and no difference was seen between treatment
groups in rates of change in bone mineral density in
these areas. No statistically significant benefit of
calcium and vitamin D3 supplementation was seen
at any femoral site in patients who were not treated
with corticosteroids.
Analysis of variance did not show statistically
significant interactions for change in bone mineral
density of the lumbar spine between the effect of
treatment with calcium and vitamin D3 and sex
(P > 0.2), menopausal status {P > 0.2), baseline
calcium intake (P > 0.2), or estrogen status (pre-
menopausal and postmenopausal women receiving
estrogen replacement therapy compared with post-
menopausal women not receiving estrogen replace-
ment therapy) {P = 0.144).
Discussion
Corticosteroids lead to a decrease in bone min-
eral density, particularly in areas rich in trabecular
bone, such as the lumbar spine and trochanter (1).
Dietary supplementation with calcium (calcium car-
bonate, 1000 mg/d) and vitamin D3 (500 lU/d) sta-
bilized bone mineral density in the lumbar spine
15 December 1996 • Annals of Internal Medicine • Volume 125 • Number 12
and trochanter during 2 years of treatment in a
group of patients with rheumatoid arthritis who
were receiving long-term low-dose prednisone ther-
apy (mean dosage, 5.6 mg/d). In patients treated
with corticosteroids who did not receive calcium and
vitamin D3, bone mineral densities of the lumbar
spine and the trochanter decreased by 2.0% and
0.9% per year, respectively. Bone mineral densities
of the femoral neck and Ward triangle were rela-
tively stable in patients treated with corticosteroids
and were unaffected by treatment with calcium and
vitamin D3. Patients who were not receiving corti-
costeroids had stable bone mineral density in the
lumbar spine and femur during the 2 years of the
study, and treatment with calcium and vitamin D3 did
not substantially improve bone mineral density in
these patients.
Tlie treatment effect seen in this group of patients
was probably not caused by correction of vitamin D
deficiency. Although levels of 25-hydroxyvitamin D
and 1,25-dihydroxyvitamin D were not measured at
baseline in our study, vitamin D deficiency was
rarely found in patients with osteoporosis in this age
group at our center (Leib E. Unpublished data). In
addition, intake of calcium and vitamin D3 in this
group of patients was greater than the national av-
erage, and conversion of vitamin D-^ to its active
metabolites is not affected by corticosteroid use
(34-37).
The design of this study limits its generalizability.
First, it is unclear whether the benefits of calcium
and vitamin D3 supplementation continue beyond 2
years. Second, bone mineral density is a surrogate
measure for the clinical outcomes of major interest:
pain, disability, and fracture related to osteoporosis.
Third, because we studied only patients who were
receiving low-dose corticosteroids, it is still unclear
whether calcium and vitamin D3 supplementation
prevents or minimizes loss of bone mineral density
associated with moderate- to high-dose corticoste-
roid therapy (25, 26).
Bone loss during treatment with corticosteroids
varies because of differences in corticosteroid use
(dose, timing, and duration) and such patient-specific
risk factors as age, sex, hormonal status, baseline
value of bone mineral density, and genetic and
other factors that have not yet been determined.
We found no statistically significant relation be-
tween sex, calcium intake, estrogen use, or meno-
pausal status and treatment effect; however, because
the numbers of men, women taking estrogen re-
placement therapy, and patients with low intake of
calcium were small, our study could not answer
these questions. The effects of calcium supplemen-
tation may differ between men and women. Calcium
supplementation may be less important in women
who are estrogen replete (premenopausal women or
965
Table 2. Difference in Annual Rate of Change in Bone Mineral Density between the Calcium and Vitamin D3 Group and
the Placebo Group*

Site and Treatmenl Patients, Unadjusted Difference betvween Groups Adjusted Difference betvi/een Groupst
n
Rate (95% Ci) P Value Rate (95% CI) P Value

-umbar spine
All patients 96 1.93 (0.38 to 3.48) 0.015 1.78(0.12 to 3.44) 0.036
Patients receiving prednisone 66 2.69 (0.86 to 4.52) 0.005 2.65 (0.73 to 4.57) 0.008
Patients not receiving prednisone 30 0,07 (-2.88 to 3 02) >0.2 -1.00(-4,40to2.40) >0.2
"emcral neck 0.69 (-0.82 to 2.20) >0.2
Ali patients
84- 0-30 (-1.07 to 1.67) >0.2
Patients receiving prednisone 5B 0.57 (-0.88 to 2.02) >0.2 0.75 (-0.86 to 2.36) >0.2
Patients not receiving prednisone 26 -0.32 (-3.62 to 3.00) >0.2 0.54 (-3.20 to 4.28) >0.2
Trochanter
All patients 84 1.66 (0.19tO 3.13) 0.028 2.02 (0.45 to 3.59) 0.013
Patients receiving prednisone 5& 1.75 (0.23 to 3.27) 0.025 2.08(0.43 to 3.73) 0.015
Patients not receiving prednisone 26 1.41 (-2.15to4.97) >0.2 1.26(-2.72to5.24) >0.2
Ward triangle
Ali patients 84 0.41 (-1.56 to 2.38) >0.2 0.32 (-1.85 to 2.49) >0.2
Patients receiving prednisone 58 0.49(-1.81 to2.79) >0.2 0.26 (-2.25 to 2.77) >0.2
Patients not receiving prednisone 26 0.17(-3.90to4.07) >0.2 -0.05 (-4.87 to 4.77) >0.2

* Po5i1ive values indicate an increased annuai rate of change of bone minerai density in patients in irie calcium and vitamin Dj group compared with those in the placebo group. Negative
vaiues indicate a decreased annual rate of charge of bone mineral density in patients in the calcium and vitamin D3 group compared wiih those in the placebo group.
t Adjusted for disease severity and duration using multiple regression analysis.

postmenopausal women receiving estrogen replace-
ment therapy) than in those who are estrogen defi-
cient. Lukert (38) and Hall (39) and their col-
leagues have shown the positive effects of estrogen
replacement on bone mineral density in women re-
ceiving corticosteroids.
Another limitation of our study is that half of the
basehne values of bone mineral density of the an-
teroposterior spine were imputed from the baseline
values of bone mineral density of the lateral spine.
The rate of change of bone mineral density of the
spine was determined by regressing data from all
three time points so that determination of the base-
line value of bone mineral density of the spine in
some patients had much less effect on the overall
rate. This type of conversion should introduce ran-
dom error. Despite the conversion, we found a sta-
tistically significant difference in the change in bone
mineral density at the lumbar spine in the patients
treated with calcium and vitamin D3 compared with
patients who received placebo. Bone mineral density
measurements at the trochanter were not imputed,
and a significant difference in the change in bone
mineral density in the patients treated with calcium
and vitamin D3 was also seen in this area. Both
areas are rich in trabecular bone, in which the det-
rimental effects of corticosteroids should be greatest
and treatment effects would be expected to be most
significant.
The deterioration of bone mineral density in pa-
tients not treated with corticosteroids does not
mean that calcium and vitamin D supplementation
is not beneficial in these patients. Reid (40) and
Dawson-Hughes (41) and their colleagues have
shown the benefits of calcium supplementation in
postmenopausal women, particularly those with low
966 15 December 1996 • Annals of Internal Medicine • Volume 125 • Number 12
calcium intake. We studied a more diverse group of
patients (men and pre- and postmenopausal wom-
en), and dietary intake of calcium among our par-
ticipants was average.
This study was not strictly an intention-to-treat
study: We did not have complete data on bone
mineral density for patients who dropped out of the
study. Therefore, data from these patients could not
be included in the analysis. Our results show the
effects of calcium and vitamin D3 supplementation
in a group of patients who were relatively motivated
and compliant.
Long-term, low-dose corticosteroid therapy is fre-
quently used to treat rheumatoid and other types of
inflammatory arthritis, polymyalgia rheumatica, and
lupus erythematosus. Corticosteroids are prescribed
by general ists who treat asthma and chronic ob-
structive lung disease as well as those who treat
such autoimmune and allergic diseases as multiple
sclerosis., inflammatory bowel disease, vasculitis, and
dermatologic conditions. An increasing number of
persons are taking corticosteroids for an indefinite
period after bone marrow or organ transplantation
(42-44). Corticosteroid use is associated with lower
bone mineral density in the spine and can result in
painful vertebral fractures that substantially de-
crease quality of life (45, 46). These fractures are a
delayed complication, often occurring years after
treatment with corticosteroids, when bone mineral
density has decreased substantially below fracture
threshold. Years of treatment are required to sig-
nificantly increase bone mass after a fracture occurs.
Thus, stabilization of bone mineral density during
treatment with corticosteroids appears to be the
optimal strategy for prevention of fractures induced
by corticosteroid use.
 Because cortieosteroids decrease absorption of
calcium from the gastrointestinal tract, supplemen-
tation with calcium and vitamin D3 is one strategy
to prevent bone loss induced by corticosteroids. Cal-
cium supplementation has been reported to sup-
press biochemieal markers of bone resorption in
patients treated with cortieosteroids (25, 47). Re-
eently, Sambrook and colleagues (21) found that
treatment with ealcium and 1,25-dihydroxyvitamin D
stabilized bone mineral density of the lumbar spine
in patients treated with moderate to high doses of
cortieosteroids. Although the more potent vitamin
D analogues may preserve bone mineral density in
the lumbar spine during moderate- to high-dose cor-
ticosteroid treatment (48), the expense and toxicity
associated with the use of these analogues have
preeluded widespread patient and provider accep-
tance. Caleitonin, estrogen, and the bisphosphonates
appear to be effeetive in the prevention and treat-
ment of osteoporosis induced by the use of eortico-
steroids (38, 39, 49-52), but eost and patient pref-
erence tend to limit widespread use of these agents
in preventing development of osteoporosis second-
ary to long-term, low-dose corticosteroid treatment.
The ideal preventive strategy should be effeetive,
inexpensive, uneomplicated, and safe. Calcium car-
bonate and vitamin D3 supplements can be pur-
chased without preseription and have low toxieity.
Our results suggest that 1000 mg of calcium carbon-
ate and 500 IU of vitamin D3 per day will stabilize
bone mineral density in the lumbar spine and tro-
chanter during low-dose prednisone treatment and
may therefore lead to a deerease in the vertebral
fraeture rate in patients with diseases neeessitating
long-term cortieosteroid use.
Acknowledgments: The authors thank John Orav, PhD, and Bruce
Hiliner, MD, for their advice and comments; Maiy Merchant,
RN, for research assistance; and Maia Eckler for secretarial
assistance.
Requests for Reprints: Lenore M. Buckley, MD, MPH, Medical
College of Virginia. 1200 East Broad Street, PO Box 980102,
Richmond, VA 23298-0102.
Current Author Addresses: Dr. Buckley; Medical College of Vir-
ginia, t200 East Broad Street, PO Box 9«OIO2, Richmond, VA
23298-0102.
Drs. Leib, Vacek. and Cooper and Ms. Cartularo: University of
Vermont, College of Medicine, Given D301, Burlington, VT
05405.
References
1. Rickers H, Deding A, Christiansen C, Rodbro P. Mineral loss in cortical
and trabecular bone during high-dose prednisone Ireatment. Caicif Tissue Int
1984;36:269-73.
2. Sambrook PN, Eisman JA, Yeates MG, Pocock NA, Eberl S, Champion
GD. Osteoporosis in rheumatoid arthritis: safety of low dose corticosteroids,
Ann Rheum Dis. 1986;4B:950-3.
3. Laan RF, van Riel PL van de Putte LB, van Erning U . van't Hof MA,
Lemmens JA, Low-dose prednisone induces rapid reversible axial bone loss in
patients with rheumatoid arthritis. A randomized, controlled study. Ann Intern
Med. 1993; 119:963-8.
15 December 1996 • Annals of Intemal Medicine • Volume 125 • Number 12
4. Buckley LM, Leib ES, Cartularo K5, Vacek PM, Cooper SM. The effeas of
low dose corttcosteroid5 on the bone mineral density of patients with rheu-
matoid arthritis. J Rheumatol. 1996;22:1055-9.
5. Hall G M , Spector TD, Griffin AJ, Jawad AS, Hall ML, Doyle DV. The
effects of rheumatoid arthritis and steroid therapy on bone density in post-
menopausal women. Arthritis Rheum 1993;35:1510-6.
6. Garton MJ, Reid DM. Bone mineral density of the hip and of the antero-
postenor and lateral dimensions of the spine in men with rheumatoid arthritis.
Effects of low-dose corticosteroids. Arthritis Rheum. 1993;36:222-8.
7. Clegg DO, Egger MJ, Ward JR. Osteoporotic vertebral compression frac-
tures in rheumatologic patients treated with prednisone. Arthritis Rheum
1983;26(Suppl}:S47,
8. Butler RC, Davie MW, Worsfotd M, Sharp CA. Bone mineral content in
patients with rheumatoid arthritis: relationship to low-dose steroid therapy.
Br J Rheumatol. 1991;30:86-90.
9. Verstraeten A, Dequeker J. Vertebral and peripheral bone mineral content
and fraaure incidence in post-menopausal patients with rheumatoid arthritis'
effect of low dose corticosteroids. Ann Rheum Dis. 1986;45:852-7.
10. MacAdams MR, White RH, Chipps BE. Reduction of serum testosterone levels
during chronic glucocorticoid therapy. Ann Intern Med. 1986;104:648-51,
11. Sambrook PN, Eisman JA, Champion GD, Pocock NA. Sex hormone
status and osteoporosis in postmenopausal women with rheumatoid arthritis.
Arthritis Rheum. 1988;31:973-8.
12. Chyun YS, Kream BE, Raisz LG. Cortisol decreases bone formation by
inhibiting periosteal cell proliferation. Endocrinology, 1984:114:447-80.
13. Canalis E. Effea of glucocorticoids on type I collagen synthesis, alkaline
phosphatase activity, and deosyribonudeic acid content in cultured rat cal-
variae Endocrinology. 1983;! 12:931-9.
14. Aloia JF, Semla HM. Yeh JK. Discordant effects of glucocorticoids on active
and passive transport of calcium in the rat duodenum. Calcif Tissue Int
1984:36:327-31.
15. Kimberg DV, Baerg RD, Gershon E, Graudusius RT. Effect of cortisone
treatment on active transport of catcium by the small intestine J Clin Invest
1971:50:1309-21.
16. Klein RG, Arnaud SB, Gallagher JC, DeLuca HF, Riggs BL Intestinal
calcium absorption in exogenous hypercortisonism. Role of 25-hydroxyvitamin
D and corticosteroid dose. J Clin Invest. 1977;60:253-9,
17. Lukert BP, Adams JS. Calcium and phosphorus homeostasis in man. Effect
of corticosteroids. Arch Intern Med. 1976,1361249-53.
18. Kukreja SC, Bowser EN, Hargis GK, Henderson WJ, Williams GA.
Mechanism of glucocorticoid-induced osteopenia: role of parathyroid glands.
Proc Soc Exp Biol Med, 1975:152.358-51.
19. Hahn TJ, Halstead LR, Teitelbaum SL, Hahn BH. Altered mineral metab-
olism in glucocorticoid-induced osfeopenia. effect of 25-hydro)(yvit3min D
administration, J Clin Invest. 1979:64:655-65.
20. Dykman TR, Haralson KM, Gluch OS, Murphy WA, Teitelbaum SL,
Hahn TJ, et al. Effect of oral 1,25-dihydroxyvitamin D and calcium on gluco-
corticoid-induced osteopenia in patients with rheumatic diseases. Arthritis
Rheum. 1984,27:1336-43.
21. Sambrook P, Birmingham J, Kelly P, Kempler S, Nguyen T, Pocock N,
et al. Prevention of corticosteroid osteoporosis, A comparison of calcium,
calcitriol, and caleitonin. N EngI J Med. 1993:328:1747-52
22. Chestnut CH 3d. Osteoporosis and its treatment [Editorial], N EngI J Med
1992:325:406-8.
23. Meunier PJ. Is steroid-induced osteoporosis preventable? [Editorial], N EngI
J Med. 1993:328:1781-2.
24. Hahn TJ, Hahn BH. Osteopenia in subjects with rheumatic diseases. Princi-
ples of diagnosis and therapy. Semin Arthritis Rheum. 1975:5:165-88.
25. Adachi JD, Bensen WG, Bianchi F, Cividino A, Pillersdorf S, Sebaldt RJ,
et al. Vitamin D and calcium in the prevention of corticosteroid induced
osteoporosis' a three year follow-up. J Rheumatol. 1996:23:995-1000.
26. Healy JH, Paget SA, Williams-Russo P, Szatrowski TP, Spiera H, Mit-
nick H, et al. A randomized controlled trial of salmon caleitonin to prevent
bone loss in corticosteroid-treated temporal arteritis and polymyalgia rheu-
matica. Caicif Tissue Int. 1995:58:73-8.
27. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS,
et al. The American Rheumatism Association 1987 revised criteria for the
classification of rheumatoid arthritis. Arthritis Rheum 1988:31.315-24,
28. Willet w e , Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, et al.
Reproducibility and validity of a semi-quantitative food frequency question-
naire. Am J Epidemiol. 1985:122:51-65,
29. Kannel WB, Sorlie P. Some health benefits of physical activity. The Eram-
ingham Study Arch Intern Med 1979:139:857-61,
30. Ramey DR, Reynould JP, Fries JF. The Health Assessment Questionnaire
1992: status and review. Arthritis Care Res. 1992:5:119-29.
31. Steinbrocker O, Traeger CH, Batterman RC. Therapeutic criteria in rheu-
matoid arthritis. JAMA. 1949:140:559-52.
32. Reid IR, Evans MC, Stapleton J, Lateral spine densitometry is a more
sensitive indicator of glucocorticoid-induced bone loss. J Bone Miner Res
1992,7:1221-5.
33. Larnach TA, Boyd SJ, Smart RC, Butler SP, Rohl PG, Diamond TH.
Reproducibility of lateral spine scans using dual energy X-ray absorptiometiy.
Caicif Tissue int. 1992:51.225-8.
34. Zerwekh JE, Emkey RD, Harris ED Jr. Low-dose predntsone therapy in
rheumatoid arthritis: effect on vitamin D metabolism. Arthritis Rheum 1984"
27:1050-2.
35. Hahn TJ, Halstead LR, Haddad JG Jr. Serum 25-hydroxyvitamin D concen-
trations in patients receiving chronic corticosteroid therapy. J Lab Clin Med
1977:90:399-404.
967
36. Seeman E, Kumar R, Hunder GG, Scott M. Heath H 3d, Rjggs BL.
Production, degradation, and circulating ieveis of 1,25-rJihydroxyvitamin D in
health and chronic glucocorticoid excess. J Clin In^^est. 1980;66:664-9.
37. Hahn TJ, Halstead LR, Baran DT. Effects of short terrr glucocortjcoid
administration on intestinal calcium absorption and circulating vitamin D me-
tabolite concentrations in man. J Clin Endccrinol Metab. 1981:52:111-5.
38. LukBrt BP, Johnson BE. Robinson RG. Estrogen and progesterone replace-
ment therapy reduces glucocorticoid-induced bone ioss. J Bone Miner Res.
1992:7:1063-9.
39. Hall GM, Daniels M, Doyle DV, Spector TD. Effect of hormone replace-
ment therapy on bone mass in rheumatoid arthritis patients treated with and
without steroids Arthritis fiheum 19g4;37'l499-505
40. Reid IR, Ames RW, Evans M C Gamble GD, Sharpe 5J. Effect of calcium
supplementation on bone loss in postmenopausal women. N Engl J Med,
1993:328:460-4
4 1 . Daw5on-Hughes B, Dallal CE, Krall EA, Sadowski L, 5ahyoun N, Tan-
nenbaum S. A controlled trial of the effect of calcium supplementation
on bone density m postmenopausal women N Engl J Med. 1990:323:878-
83
42. Julian BA, Laskow DA, Dubovsky J, Dubovsky EV, Curtis JJ, Quarles
LD, Rapid loss of vertebral mineral density after renal transplantation, N Engl
J Med. ! 991:325:544-50,
43. Lee AH, Mull RL, Keenan GF, Callegari PE, Dalinka MK, Eisen HJ, et al.
Osteoporosis and bone morbidity in cardiac transplant recipients. Am J Med,
1994:96:35-41.
44. Meys E, Fontanges E, Fourcade N, Thomasson A, Pouyet M, Delmas
PD. Bone loss after orthotopic liver transplantation. Am J Med, 1994:97:
445-50,
45. Silverman SL, Mason J, Greenwald M. Quality of life after osteoporotic
vertebral fraaure Arthritis Rheum. 1993,36:5122
46. Silverman SL The clinical consequences of vertebral compression frarture.
Bone, 1992;13(Suppl 2):S27-31.
47. Reid IR, Ibbertson HK. Calcium supplements in the prevention of steroid-
induced osteoporosis Am J Clin Nutr, 1986:44:287-90,
48. Lindhotm TS, Sevastikoglou JA, Lindgren U. Treatment of patients with
senile post-menopausal and corticosteroid-induced osteoporosis with 1 «-hydroxy-
vitamin D, and calcium: short- and long-term effects. Clin Endocrinol (O)cf),
1977:7(Suppl):183s-9s.
49. Ringe JD, Welzel D. Salmon calcitonin in the therapy of corticoid-induced
osteoporosis Eur J Clin Pharmacol, 1987;33:35-9.
50. Montemurro L, Schiraldi G, Fraioli P, Tosi G, Riboldi A, Rizzato G.
Prevention of corticosteroid-induced osteoporosis with salmon calcitonin in
wrcoid patients, Calcif Tissue Int. 1991:49 71-6,
51. Struys A, Snelder AA, Mulder H. Cyclic etidronate reverses bone loss of
the spine and proximal femur in patients with established corticosteroid-
induced osteoporosis Am J Med. 1995:99:235-42.
52. Adachi i, Cranney A, Goldsmith CH, Bensen WG, Bianchi F, Cividino A,
et al. Intermittent cyclic therapy with etidronate in the prevention of corti-
costeroid induced bone loss J Rheumatol. 1994:21:1922-6.

There are all kinds of silences and each of them means a different thing. There is the
silence that comes with morning in a forest, and this is different from the silence of
a sleeping city. There is silence after a rainstorm, and before a rainstorm, and these
are not the same. There is the silence of emptiness, the silence of fear, the silence of
doubt. There is a certain silence that can emanate from a lifeless object as from a
chair lately used, or from a piano with old dust upon its keys, or from anything that
has answered to the need of a man, for pleasure or for work. This kind of silence can
speak. Its voice may be melancholy, but it is not always so; for the chair may have
been left by a laughing child or the last notes of a piano may have been raucous and
gay. Whatever the mood or the cireumstance, the essence of its quality may linger in
the silence that follows. It is a soundless echo.

Beryl Markham
West with the Night
San Francisco: North Point Pr; 1983

Submitted by:
James A. Long, MD
Waynesboro, VA 22980

Submissions from readers are welcomed. If the quotation is published, the sender's name will be acknowl-
edged. Please include a complete citation, as done for any reference.—Tlie Editor

968 15 December 1996 • Annals of Internal Medicine • Volume 125 • Number 12