JOURNAL OF BONE AND MINERAL RESEARCH

Volume 19, Number 8, 2004

Published online on May 24, 2004; doi: 10.1359/JBMR.040511
2004 American Society for Bone and Mineral Research

Supplementation With Oral Vitamin D3 and Calcium During Winter

Prevents Seasonal Bone Loss: A Randomized Controlled Open-Label
Prospective Trial
Christian Meier,1 Henning W Woitge,2 Klaus Witte,3 Bjorn Lemmer,3 and Markus J Seibel1,2

ABSTRACT: Bone metabolism follows a seasonal pattern with high bone turnover and bone loss during the
winter. In a randomized, open-label 2-year sequential follow-up study of 55 healthy adults, we found that
supplementation with oral vitamin D3 and calcium during winter abolished seasonal changes in calciotropic
hormones and markers of bone turnover and led to an increase in BMD. Supplementation with oral vitamin
D3 and calcium during the winter months seems to counteract the effects of seasonal changes in vitamin D and
thus may be beneficial as a primary prevention strategy for age-related bone loss.
Introduction: Bone metabolism follows a seasonal pattern characterized by high bone turnover and bone loss during
winter. We investigated whether wintertime supplementation with oral vitamin D3 and calcium had beneficial effects
on the circannual changes in bone turnover and bone mass.
Materials and Methods: This prospective study comprised an initial observation period of 12 months (year 1),
followed by an intervention during parts of year 2. Fifty-five healthy subjects living in southwestern Germany
(latitude, 49.5 N) were randomized into two groups: 30 subjects were assigned to the treatment group and received
oral cholecalciferol (500 IU/day) and calcium (500 mg/day) during the winter months of year 2 (OctoberApril),
while 25 subjects assigned to the control group obtained no supplements. Primary endpoints were changes in
calciotropic hormones [serum 25(OH)D, 1,25(OH)2D, and parathyroid hormone], markers of bone formation (serum
bone-specific alkaline phosphatase) and of bone resorption (urinary pyridinoline and deoxypyridinoline), and changes
in lumbar spine and femoral neck BMD.
Results: Forty-three subjects completed the study. During year 1, calciotropic hormones, markers of bone turnover,
and BMD varied by season in both groups. During the winter months of year 1, bone turnover was significantly
accelerated, and lumbar spine and femoral BMD declined by 0.3 0.9%. In year 2, seasonal changes in calciotropic
hormones and markers of bone turnover were either reversed or abolished in the intervention group while unchanged
in the control cohort. In the subjects receiving oral vitamin D3 and calcium, lumbar and femoral BMD increased
significantly (lumbar spine: 0.8%, p 0.04 versus year 1; femoral neck: 0.1%, p 0.05 versus year 1), whereas
controls continued to lose bone (intervention group versus control group: lumbar spine, p 0.03; femoral neck, p
0.05).
Conclusions: Supplementation with oral vitamin D3 and calcium during winter prevents seasonal changes in bone
turnover and bone loss in healthy adults. It seems conceivable that annually recurring cycles of low vitamin D and
mild secondary hyperparathyroidism during the winter months contributes, at least in part and over many years, to
age-related bone loss. Supplementation with low-dose oral vitamin D3 and calcium during winter may be an efficient
and inexpensive strategy for the primary prevention of bone loss in northern latitudes.
J Bone Miner Res 2004;19:12211230. Published online on May 24, 2004; doi: 10.1359/JBMR.040511
Key words:

vitamin D, seasonal variation, bone turnover markers, bone densitometry

INTRODUCTION

STEOPOROSIS IS A WORLDWIDE health issue. It is anticipated that the number of affected individuals, and
thereby costs to health care systems, will increase substantially with further population aging.(1,2) Most approaches to

The authors have no conflict of interest.

the secondary or even tertiary prevention of osteoporosis

and osteoporotic fractures have proven expensive and of
limited efficacy. In contrast, primary prevention strategies
based on biological and mechanistic considerations may
have considerable potential within the public health setting.
Circannual variations in 25-hydroxyvitamin D [25(OH)D]
levels have been well established, and there also seems to be an
effect of season on bone turnover and bone mass.(312) It has

1
Bone Research Program, ANZAC Research Institute, University Sydney, Concord, New South Wales, Australia; 2Department of
Medicine I, University of Heidelberg, Heidelberg, Germany; 3Institute of Pharmacology and Toxicology, Faculty of Clinical Medicine
Mannheim, University of Heidelberg, Mannheim, Germany.

1221

1222

MEIER ET AL.

been shown that bone turnover follows a circannual rhythm,

with highest levels of bone markers during the northern winter
months (OctoberMarch). These circannual rhythms seem to
be directly related to variations in the hormones regulating
calcium homeostasis (i.e., parathyroid hormone [PTH],
vitamin D).(13,14) Although these observations have been
made in a number of longitudinal(1517) and cross-sectional
studies,(18 21) they were not confirmed in all reports,(18,2224)
indicating that seasonal variability in bone turnover may be
modulated by other factors such as age, gender, degree of and
individual sensitivity to vitamin D deficiency, fluctuations in
serum PTH levels, and in women, estrogen status. Further
evidence for a seasonal pattern in bone metabolism comes
from the observation that BMD decreases during the winter
months.(15,17,19,24 27) However, the biological significance of
this finding has yet to be determined, although it is conceivable
that it may, over many cycles, contribute to the decline in bone
mass seen with aging. Of note, intervention studies in elderly
subjects have shown that either calcium supplementation
alone(28 31) or in combination with vitamin D(32,33) partially
prevents bone loss and fractures and reverses secondary hyperparathyroidism and accelerated bone turnover.
We hypothesize that vitamin D insufficiency during wintertime induces, through PTH, accelerated bone turnover
and bone loss. We further hypothesize that supplementation
with oral vitamin D3 and calcium during the winter months
can prevent this increase in bone turnover and the associated
bone loss. In this prospective randomized open-label controlled study of healthy volunteers, we aimed (1) to evaluate
the circannual changes in bone turnover and BMD and (2)
to determine the effect of oral calcium and vitamin D3
supplementation on rates of bone turnover and bone loss
during winter months.

MATERIALS AND METHODS

Study population and design
Fifty-five healthy volunteers, 3378 years of age (19 men
and 36 postmenopausal), were randomly recruited from a
local health club in southwestern Germany (Heidelberg;
latitude 49.5 N) to participate in a 2-year prospective study.
At study entry, none of the participants had a history or
clinical evidence of significant skeletal or nonskeletal disease nor had they taken any medication known to affect
bone metabolism, including vitamin D and mineral supplements. Fourteen men (mean age, 60.6 10.3 years; range,
34 75 years) and 29 postmenopausal women (mean age,
54.1 10.8 years; range, 38 75 years) completed the study
over a period of 24 months. Twelve participants had to be
excluded from the final analysis because of incomplete data
or voluntary dropout (n 9), previously unknown medical
conditions (n 2; cancer and osteoporosis), and death (n
1; Fig. 1).
At study entry, subjects were randomly assigned to either
the intervention or the control group. The first year of the
study was designed as an observation period only, during
which subjects followed their usual daily routine with no
intervention per protocol. During the winter of the second
year, participants assigned to the intervention group received a daily supplement of oral vitamin D3 (cholecalcif-

FIG. 1.

Study design.

erol, 500 IU) and calcium (500 mg) from October to March,
whereas subjects in the control group received no supplements and were asked to remain off such agents. The study
medication was open label. Adherence to the protocol was
checked in monthly intervals by personal interview.
At baseline, detailed information on lifestyle habits (diet,
dietary calcium intake, physical activity, smoking, alcohol,
sun exposure), medical history, and past or present medication was obtained.(34) Weight and height measures were
taken in light clothing without shoes. Obesity was estimated
using body mass index (BMI; kg/m2). Throughout the entire
study, subjects were reviewed in monthly intervals for
changes in blood pressure, heart rate, BMI, medication, and
lifestyle habits.
Blood and urine samples were obtained from all subjects
in the nonfasting state before the study and every 4 weeks.
During their regular follow-up visits, blood and urine samples were collected at the same time-points during the day.
Blood samples were processed within 3 h after phlebotomy
and centrifuged at 1500g for 10 minutes, and serum was
stored in aliquots at 80C until analysis. Urine specimens
were protected from light exposure and stored within 3 h of
collection at 30C until measurement. All laboratory analyses were performed within 3 months after completion of
patient recruitment and follow-up.
The study protocol was approved by the Human Ethics
Committee of the University of Heidelberg, Germany, and
written informed consent was obtained from all participants
before inclusion into the study.

Laboratory measurements
Measurements of hormones and biochemical markers of
bone turnover were assessed once at baseline and every 4
weeks throughout the study period. All analyses were performed in duplicate, using the same assay batches and
running samples from a single subject back-to-back in one
assay.
25(OH)D was measured by radioimmunoassay (reference
range, 8.9 46.7 ng/ml; Instar, Stillwater, OK, USA). Serum
intact PTH was determined by immunoluminometric assay
(reference range, 1154 pg/ml; Chiron Diagnostics Magic
Lite, Cergy Pontoise, France). 1,25-dihydroxyvitamin D

PREVENTION OF SEASONAL BONE LOSS

[1,25(OH)2D] was measured using an in-house radioimmunoassay (reference range, 32 80 ng/liter) as previously described.(35)
A solid-phase, two site immunoradiometric assay
(Tandem-R Ostase; Hybritech, San Diego, CA, USA) was
used to determine serum bone-specific alkaline phosphatase
(BSAP).(36) Intra- and interassay CVs ranged between 3.7%
to 6.7% and 7.0% and 8.1%, respectively. The normal
reference range for this assay is 11.6 4.1 g/liter for
premenopausal women and 12.4 4.4 g/liter for men.
Total urinary pyridinoline (PYD) and total urinary deoxypyridinoline (DPD) were determined by HPLC. After complete acid hydrolysis of urine samples at 107C for 16 h, the
peptide-free forms of PYD and DPD were separated by
ion-paired HPLC, and concentrations were quantitated by
fluorometry using a fully automated method as described by
Pratt et al.(37) Standards were derived from sheep bone. The
overall reproducibility of the assay was between 3% and
12% for intra- and interassay variability. Urinary concentrations were expressed relatively to urinary creatinine (Cr)
levels. The normal reference range for urinary PYD is
17.6 47.3 nM/mM Cr in premenopausal females and 10.8
49.1 nM/mM Cr in males. The reference range for urinary
DPD is 2.712.7 nM/mM Cr in premenopausal females and
2.6 9.3 nM/mM Cr in males.

Other laboratory measurements

All participants had routine laboratory tests including
serum and urinary Cr, as well as serum calcium and albumin
(Beckman II Analyser; Beckman Instruments, Palo Alto,
CA, USA). These measurements were performed immediately after sample collection. Total serum calcium was
corrected for total protein content.

BMD
BMD was measured at the lumbar spine (L2L4) and at
the femoral neck by DXA in 6-month intervals, starting in
October of study year 1 and ending in April of study year 2
(QDR 4500; Hologic). These four time-points were chosen
to capture the summer and winter seasons and the effect of
vitamin D and calcium supplementation during the winter of
the second year, respectively. The CV for BMD measurements was 1.2% at the lumbar spine and 1.9% at the femoral
neck. The same instrument and software were used throughout the study. All scans were analyzed in a blinded manner
at a central location (Institute for Diagnostic Radiology,
University of Kiel, Kiel, Germany).

Statistical analysis
The SAS software package (SAS Institute, Cary, NC,
USA) was used. Descriptive statistics are presented as
mean SD if not stated otherwise. The significance of
group differences was tested using Students t-test or
Wilcoxon-Mann-Whitney U test in nonparametric distributions. Fishers exact test was used to test for differences in
the distribution of participants within categories. Differences in BMD were evaluated using the Students paired
t-test for single repeated measurements or an unpaired t-test
to compare changes between groups.

1223

Rhythm analyses were performed using the Pharmfit

method as described previously.(38) The mesor (rhythmadjusted annual mean) and amplitude (one-half of peak-totrough of rhythmic change) were calculated for each individual, assuming a period length of 365 days. Group data,
expressed as percent of the individual mesor, were used for
calculation of the acrophase (time of seasonal peak) and for
rhythm statistics. The following equation was used for
rhythm analyses: y mesor amplitude cos([x
acrophase]/365 2), in which y is the measured marker
value and x is the respective day of the year. Significance of
circannual rhythmicity was tested using the zero-amplitude
hypothesis. All statistical tests were two-tailed, and a significance level of p 0.05 was considered statistically
significant.

RESULTS
Characteristics of the study population
Table 1 summarizes the anthropometric, biochemical, and
lifestyle characteristics of the 43 study participants who
completed the study per the protocol.
At baseline, subjects in the control group (CTR, n 16)
and subjects in the intervention group (INT, n 27) were
similar with respect to age, BMI, 25(OH)D and PTH levels,
biochemical markers of bone turnover, and BMD. Postmenopausal women (n 29; age, 54.7 10.7 years) had
higher levels of bone resorption markers (U-PYD, p 0.01;
U-DPD, p 0.02) and lower BMD (spine BMD, p 0.01;
femoral neck BMD, p 0.001) than male study participants
(n 14; age, 60.9 10.2 years), whereas baseline levels of
osteotropic hormones and of bone formation markers were
similar in both sexes. In the first observational year, mesors
and amplitudes of bone resorption markers were higher in
women than in men (U-PYD: p 0.002, p 0.01; U-DPD:
p 0.001, p 0.01), whereas the respective values for
calciotropic hormones and bone formation markers were
similar in both sexes.
One female and one male subject (one in each group)
were vitamin D deficient [serum 25(OH)D 30 nM]. Both
individuals had normal serum calcium and PTH levels,
although the normal PTH levels in these individuals do not
preclude vitamin D deficiency.
There was no difference in smoking habits or alcohol
intake, but subjects in the control group were physically
more active (p 0.046) and had a higher intake of dairy
products (p 0.02) than subjects in the intervention group.
During the follow-up period, no significant changes in
lifestyle and nutritional factors were observed in either
group.

Circannual rhythms of osteotropic hormones and bone

turnover markers
During year 1 (observation only), calciotropic hormones
and markers of bone turnover followed a significant seasonal rhythm, with similar amplitudes and mesors (rhythmadjusted annual mean) in both groups (Table 2). The amplitude of seasonal change in controls was greatest for
serum 25(OH)D, urinary DPD, and serum 1,25(OH)2D levels (22.2%, 21.8%, and 19.3% of the mesor, respectively).

1224

MEIER ET AL.
TABLE 1. POPULATION CHARACTERISTICS
All

AT

BASELINE

Women

Men

Characteristics

CTR (n 16)

INT (n 27)

p Value

CTR (n 12)

INT (n 17)

p Value

CTR (n 4)

INT (n 10)

p Values

Age (years)
Height (cm)
Weight (kg)
BMI (kg/m2)
Calcium (mmol/liter)
25(OH)D (g/liter)
1,25(OH)2D (ng/liter)
PTH (pg/ml)
Serum BSAP (g/liter)
Urinary-PYD (nM/mM
Cr)
Urinary-DPD (nM/mM
Cr)
BMD, spine (g/cm2)
BMD, femoral neck
(g/cm2)
Smoking habits (smokers,
%)
Alcohol intake*
Intake of dairy products/
day
Physical activity/day (h)

57.9 11.3
165.0 5.1
71.6 15.7
26.2 4.8
2.5 0.1
30.8 9.3
49.9 9.1
40.0 22.7
11.0 3.5
47.8 31.7

55.2 10.9
166.4 9.5
72.5 15.7
26.1 4.4
2.5 0.1
30.1 11.4
48.2 16.9
41.9 23.6
11.6 4.5
41.2 20.1

NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

55.3 11.3
165.2 5.5
71.2 18.1
25.9 5.6
2.5 0.1
29.9 9.2
50.2 8.5
39.9 23.2
11.5 3.8
53.0 35.2

53.5 10.7
165.7 10.1
73.1 14.5
26.5 3.9
2.5 0.1
29.8 10.1
49.2 21.2
45.2 27.0
11.9 4.9
46.5 23.6

NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

66.4 7.5
164.5 3.9
73.0 5.8
26.9 1.1
2.5 0.1
33.5 10.5
49.0 13.0
40.3 24.6
9.3 1.9
32.4 7.9

58.6 10.6
167.5 9.2
71.9 18.1
25.5 5.4
2.5 0.1
30.6 13.9
47.1 11.7
36.3 16.0
11.1 3.8
32.2 5.9

NS
NS
NS
NS
NS
NS
NS
NS
NS
NS

12.2 10.5

9.9 5.8

NS

13.9 11.7

11.4 6.7

NS

7.2 1.5

7.3 2.7

NS

0.961 0.251
0.876 0.168

1.032 0.150
0.997 0.224

NS
NS

0.902 0.201
0.800 0.083

0.994 0.179
0.896 0.139

NS
NS

1.205 0.244
1.010 0.207

1.096 0.169
1.179 0.237

NS
NS

12.5

14.8

NS

16.7

17.7

NS

10.0

NS

2.3 1.0
3.1 0.7

2.2 1.1
2.5 0.9

NS
0.018

2.1 1.0
3.1 0.6

1.9 0.9
2.9 0.9

NS
NS

3.0 0.8
3.3 1.0

2.7 1.4
1.9 0.6

NS
NS

5.4 4.8

2.9 1.7

0.046

6.0 5.3

2.4 1.2

0.031

3.8 2.1

3.8 2.2

NS

NS, not significant.

Values are mean SD.
* Alcohol intake is characterized as mean score of the following: I, no alcohol; II, 2 times/week; III, 3 4 times/week; IV, 5 times/week.

Intake of dairy products is characterized as mean score of the following: I, no dairy products; II, consumption at 3 days/week; III, consumption at
3 6 days/week; IV, daily consumption.

Physical activity: hours of strenuous activity per week.

In subjects allocated to the intervention group, the amplitude of seasonal change was greatest for serum
1,25(OH)2D, 25(OH)D, and PTH (35.3%, 32.6%, and
18.8% of the mesor, respectively). The corresponding acrophases (time of seasonal peak) were as follows: serum
25(OH)D and 1,25(OH)2D, September/August; serum PTH,
February; serum BSAP, December/October; urinary PYD
and DPD, January/November (Fig. 2). Gender-specific analyses showed significant seasonal rhythms of serum
25(OH)D and PTH in postmenopausal women and men
with acrophases in late summer and winter, respectively.
Circannual changes of bone turnover markers were significant in women only (data not shown).
In year 2 (intervention), seasonal changes in calciotropic
hormones and markers of bone turnover were reversed or
abolished in the group receiving cholecalciferol and calcium
supplements during winter. In contrast, the circannual pattern was essentially unchanged in the control group (Table
2; Figs. 2 and 3). Highest serum 25(OH)D concentrations
were seen at the end of the replacement period in March,
with a flattened mean amplitude of 16.3% compared with
32.6% in the first study year. Serum 1,25(OH)2D levels
resembled the pattern of 25(OH)D, but the seasonal rhythm
was not significant, and the mean amplitude was lower
(4.9%) compared with the amplitude in year 1 (35.5%).
Serum PTH concentrations in the intervention group again
mirrored the changes in serum 25(OH)D levels, but peak
values were now shifted to summer (August) with a lower
amplitude (p 0.07). While there was significant seasonality in serum BSAP, with highest levels in September,

urinary PYD and DPD concentrations showed no significant

seasonal variation in the intervention group. However, peak
values were delayed and observed in May and July, respectively. While on vitamin D3 and calcium supplementation,
subjects in the intervention group showed significant differences for mesor (increase, p 0.05) and amplitude (decrease, p 0.05) levels of 25(OH)D, 1,25(OH)2D (decrease
in amplitude, p 0.05), PTH (decrease in mesor, p 0.05),
and serum BSAP (decrease in mesor, p 0.05) compared
with their levels in the first study year without intervention.
Comparable with the gender-related differences in seasonality during the first study year, the main effects of vitamin
D3 and calcium supplementation on bone turnover markers
were observed in women (data not shown).
Table 3 allows for a comparison of the group means for
25(OH)D, PTH, and bone markers as observed at the time
of the seasonal nadir and peak of serum 25(OH)D during the
three follow-up seasons. In the control group, serum
25(OH)D levels were significantly higher in summer than in
both winters, and there was a trend toward lower serum
PTH levels during summer. Mean urinary DPD levels were
lower in summer, whereas mean values for serum BSAP
were similar at all three time-points.
In the intervention group, similar seasonal changes in
25(OH)D, PTH and bone turnover were observed during the
period of observation only (Table 3). However, supplementation with oral vitamin D3 and calcium during the second
winter was associated with significantly increased serum
25(OH)D and suppressed PTH levels (p 0.001 and p
0.01, respectively, compared with values observed during

11.0 4.6

41.6 12.5

10.8 3.6

41.9 21.9
2.44 0.10

37.8 6.9

27.5 9.0

Year 1

AND

0.05
NS
NS

41.3 14.8
10.1 3.9

1.1 0.7
5.3 4.0
1.8 1.5

1.4 0.6
7.0 7.3
2.4 2.9

Feb.
Dec.

44.9 28.3 NS
7.7 8.9
7.3 8.9
2.35 0.08 0.05 0.08 0.10 0.08 0.03
9.7 3.7

0.05 Sept.*

13.6 5.2

NS

NS

NS

NS
NS

Jan.

Jan.*

Dec.*

Sept.

7.3 4.0

NS

38.4 8.0

NS

7.0 3.5

Jan.*

Feb.*

Oct.

Feb.
Oct.

Sept.

Aug.

THE

2.1 3.5

5.3 3.5

Year 2

8.8 3.6

7.8 3.1

37.2 12.3 36.9 10.2

9.4 4.0

NS

NS

0.05

1.2 1.0

5.1 3.3

1.2 0.9

1.5 1.1

4.7 3.3

1.3 0.7

NS

NS

Nov.

Nov.*

Oct.*

Feb.
Nov.

0.001 Aug.

Year 1

Jul.

May

Sept.

Aug.
Oct.

Feb.

March

Year 2

Acrophase

0.05 Sept.

NS

14.0 5.1

8.4 4.7

Year 1

11.3 4.3

NS

0.05

Amplitude

2-YEAR FOLLOW-UP

0.07
NS

42.1 9.5

STUDY GROUP DURING

Intervention group (n 27)

BY

40.8 14.7 35.3 11.4 0.05 7.7 7.1

5.3 3.4
2.42 0.06 2.34 0.06 0.05 0.08 0.06 0.09 0.05

39.6 8.3

32.6 5.8

25.8 8.8

Mesor

ALL PARTICIPANTS

Year 2

IN

Year 1

BONE TURNOVER

Year 1 Year 2

6.1 3.4

NS

25.7 8.5

Year 2

Year 1

Amplitude

OF

Acrophase

BIOCHEMICAL MARKERS

Control group (n 16)

HORMONE LEVELS

Year 2

OF

Data are mean SD.

Levels of significance for differences in mesor (rhythm-adjusted annual mean) and amplitude (half of peak-to-trough of rhythmic change) between year 1 and year 2 were determined by using Wilcoxon-matched
pairs test. Significance of circannual pattern of acrophases (time of seasonal peak) was determined by using the zero-amplitude hypothesis: * p 0.05, p 0.01, p 0.001, p 0.0001.

25(OH)D (g/
liter)
1,25(OH)2D
(ng/liter)
PTH (g/liter)
Calcium
(mmol/liter)
Serum BSAP
(g/liter)
Urinary-PYD
(nM/mM Cr)
Urinary-DPD
(nm/mM Cr)

Markers

Mesor

TABLE 2. CIRCANNUAL RHYTHMS

PREVENTION OF SEASONAL BONE LOSS

1225

FIG. 2. Circannual variation of calciotropic hormones in controls (E,

broken line, n 16) and subjects supplemented with oral vitamin D
and calcium during the winter months of year 2 (F, solid line, n 27;
the intervention period is symbolized by an open bar). Data are presented as mean values calculated as the percentage of the respective
mesor (rhythm-adjusted annual mean). Assessment of circannual
rhythms in marker values was performed using the Pharmfit method.
Significance of seasonal patterns was tested using the zero-amplitude
hypothesis as shown in Table 2.

the winter of year 1). At the same time, a significant reduction was seen in serum BSAP and urinary DPD levels (p
0.001 and p 0.03, respectively, compared with values
observed during the winter of year 1).

Seasonal changes in BMD

Figure 4 summarizes the seasonal changes in BMD at the

lumbar spine and the femoral neck for both study groups.
During the first winter, BMD at the lumbar spine numerically declined in both groups (October versus April; con-

1226

MEIER ET AL.

creased at the lumbar spine (1.028 0.180 1.035 0.185

g/cm2, 0.8%, p 0.04) and remained unchanged at the
femoral neck (0.999 0.225 0.999 0.221 g/cm2,
0.1%, p NS). In contrast, lumbar spine and femoral neck
BMD decreased in controls.
In the control group, changes in BMD at the lumbar spine
and at the femoral neck were not significantly different from
each other during all time-points measured (Fig. 4, W1
versus S versus W2). In the intervention group, supplementation with oral calcium and vitamin D3 resulted in a significant gain in BMD, which was significantly different
from the loss in BMD observed during the first winter (W1,
p 0.04). When comparing the change in BMD during the
second winter (W2) between the control and the intervention groups, the difference was also significant (p 0.03).
Similar observations at nearly identical p values were made
at the femoral neck (Fig. 4). The effect of vitamin D3 and
calcium supplementation on BMD was similar in women
and men in all of the above comparisons.

DISCUSSION

FIG. 3. Circannual variation of markers of bone turnover in controls

(E, broken line, n 16) and subjects supplemented with oral vitamin
D and calcium during the winter months of year 2 (F, solid line, n
27; the intervention period is symbolized by an open bar). Data are
presented as mean values calculated as the percentage of the respective
mesor (rhythm-adjusted annual mean). Assessment of circannual
rhythms in marker values was performed using the Pharmfit method.
Significance of seasonal patterns was tested using the zero-amplitude
hypothesis as shown in Table 2.

trols: 0.961 0.251 versus 0.951 0.235 g/cm2,

0.8%, p not significant [NS]; intervention group:
1.032 0.171 versus 1.028 0.178 g/cm2, 0.6%,
p NS). Changes of similar magnitude were seen at the
femoral neck (controls: 0.876 0.168 versus 0.871
0.172 g/cm2, 0.6%, p NS; intervention group:
0.997 0.224 versus 0.987 0.220 g/cm2, 0.9%,
p 0.006). These changes were similar for women and
men.
During the following summer, bone mass increased at the
femoral neck only (April year 1 versus October year 2;
controls: 0.871 0.172 versus 0.876 0.168 g/cm2,
0.7%, p 0.006; intervention group: 0.987 0.220 versus
0.999 0.225 g/cm2, 1.2, p 0.014), whereas lumbar
spine BMD remained unchanged (controls: 0.951 0.235
versus 0.953 0.254 g/cm2, 0.1%, p NS; intervention group: 1.031 0.178 versus 1.028 0.180 g/cm2,
0.4%, p NS). The seasonal patterns of changes in
BMD were similar for women and men.
During the second winter (October year 2 versus April
year 2), BMD in the intervention group significantly in-

In this 2-year prospective study of healthy adults, several

observations relevant to the seasonality of bone metabolism
were made. First, this study confirms that bone metabolism
follows a seasonal pattern. Bone turnover was highest during the winter, coinciding with a nadir in serum 25(OH)D
levels and with a rise in serum intact PTH. Second, we
observed a trend toward lower BMD at the lumbar spine and
the femoral neck during the winter months. Third, supplementation with 500 IU vitamin D3 and 500 mg calcium
daily during the winter months maintained serum 25(OH)D
levels on summer levels, suppressed serum PTH concentrations, reduced bone turnover, and most importantly, prevented the bone loss seen in a nonsupplemented group
during wintertime. Our observations suggest that seasonality in calciotropic hormones significantly affects bone turnover and BMD. These results add further evidence to the
hypothesis that supplementation with oral vitamin D and
calcium may be a useful and inexpensive strategy to prevent
age-related bone loss.
Our study differs from previous investigations in that
cholecalciferol and calcium was supplemented during the
winter months only (as opposed to continuous supplementation throughout the year). In addition, the randomized,
2-year sequential follow-up allows for comparisons within
identical individuals and groups through two winter and two
summer seasons with and without intervention.
Previous interventional studies showed that continuous
supplementation with either calcium alone or in combination with vitamin D3 prevents bone loss(28,30 33,39) or osteoporotic fractures.(32,40,41) So far, however, only three studies
have investigated the effect of vitamin D and/or calcium on
seasonal bone loss. Storm et al.(15) reported that, in older
postmenopausal women, bone loss during winter was prevented after 2 years of continuous supplementation with
1000 mg of calcium alone. In another study, healthy postmenopausal women had significantly reduced bone loss in
late winter when continuously treated with 400 IU of vitamin D3 and 377 mg of calcium over the course of 1 year.(25)

PREVENTION OF SEASONAL BONE LOSS

TABLE 3. SERUM 25(OH)D, SERUM PTH,

Markers
25(OH)D (g/liter)
PTH (g/liter)
Serum BSAP (g/
liter)
Urinary DPD (nm/
mM Cr)

AND

1227

BIOCHEMICAL MARKERS

Group

Feb./March,
year 1

Aug./Sept.,
year 2

CTR
INT
CTR
INT
CTR
INT
CTR
INT

24.6 9.4
19.5 10.0
49.7 29.8
40.6 12.3
10.9 3.4
11.1 4.5
13.3 9.3
9.4 5.1

33.8 11.5
32.1 8.2
42.2 22.3
38.6 19.5
10.5 4.6
10.0 4.0
8.8 4.8
8.6 6.1

OF

BONE TURNOVER DURING THREE SEASONS

Aug./Sept. year 2
vs. Feb./March
year 1
p
p
p
p
p
p
p
p

0.001
0.001
0.09
0.24
0.47
0.03
0.03
0.04

Feb./March,
year 2
20.5 8.5
35.1 8.1
47.3 32.2
34.9 14.4
10.1 4.7
8.8 3.5
10.3 4.5
7.7 3.5

Feb./March year
2 vs. Feb./
March year 1
p
p
p
p
p
p
p
p

0.02
0.001
0.76
0.01
0.25
0.001
0.23
0.03

OF

FOLLOW-UP

Aug./Sept. year 2
vs. Feb./March
year 2
p
p
p
p
p
p
p
p

0.001
0.09
0.23
0.16
0.50
0.008
0.11
0.67

Values represent the group means SD for each parameter as measured at the time of the seasonal nadir and peak of serum 25(OH)D. The INT group
was supplemented with calcium and vitamin D only during the winter of the second year, so that the group means listed for INT under Feb/March, year
2 represent values during supplementation. CTR, control group (n 16); INT, intervention group (n 27).

FIG. 4. Mean percent change in lumbar spine and femoral neck BMD
during winter (OctoberApril) of year 1 (W1) and year 2 (W2) and
summer (S; AprilOctober). Open bars represent the control group;
black bars represent the intervention group. *Comparison of change
during the winter months of year 2 (W2) between the intervention and
control groups: p 0.03 for change in spine BMD and p 0.05 for
change in femoral BMD. Values are mean SE.

In contrast, a recent study in pre- and postmenopausal

women was unable to confirm these findings.(23) So far, no
data are available on the bone-sparing effects of a supplementation with calcium and vitamin D3 during wintertime
only.

Previous studies suggested supplementation with lowdose vitamin D has only limited effects on BMD in
postmenopausal vitamin Dreplete women.(42) In contrast, elderly subjects with vitamin D deficiency and
secondary hyperparathyroidism usually experience significant gains in bone mass and substantial reductions in
fracture risk during and after supplementation with vitamin D. In this study, the mean semiannual increase in
BMD after supplementation with calcium and vitamin D3
was 0.8% at the lumbar spine and 0.1% at the femoral
neck. These differences were significant compared with
the changes in BMD observed in the same group during
the previous winter, when no supplementation was provided (lumbar spine, 0.6%, p 0.04; femoral neck,
0.9%, p 0.05). Although the observed semiannual
changes were small, the resulting net differences during
winter with or without intervention were clinically important (1.4% at the lumbar spine and 1.0% at the femoral
neck). The differences in BMD associated with supplementation were also significant compared with those seen
in control subjects during the same winter period (lumbar
spine, 0.3%, p 0.03; femoral neck, 1.1%, p
0.05). Importantly, the effect on BMD was not restricted
to patients with vitamin D deficiency, because only two
subjects had abnormally low levels of vitamin D at
baseline, and the results were unchanged when these
subjects were excluded from the analysis (data not
shown).
Seasonality in 25(OH)D levels has been extensively investigated, with most studies reporting a nadir during winter
and a peak during summer and early autumn.(13,14,19,43) We
recently showed that biochemical markers of bone remodeling vary by season and that the rise in marker levels seen
in winter coincides with significant variations in serum
vitamin D and PTH levels.(13) Seasonal changes in calciotropic hormones and markers of bone turnover has been
observed in a number of other studies.(1517,19,21) In this
study, we again confirm these observations by showing
lower levels of serum 25(OH)D, higher levels of serum
intact PTH, and accelerated bone turnover in winter. In
contrast to previous studies, however, the change in mean
serum PTH levels at the time of the seasonal nadir of serum

1228

25(OH)D was less pronounced. This may be because of the

fact that PTH is secreted in a pulsatile fashion, leading to
significant variability of random PTH measurements (Tables 13).
A number of studies reported no significant changes in
bone turnover or density with season. This lack in seasonality may be because of methodological issues such as
differences in the age of study participants,(44) the use of
bone markers with recognized low sensitivity,(25,45,46) or
small sample size.(18) Because circannual changes in bone
turnover are of comparatively small magnitude, these studies may have underestimated the effects of season on bone
turnover simply by lack of methodological sensitivity.(47)
However, biological factors also may play a role. Thus,
circannual effects on bone turnover are predominantly observed in either pre- or postmenopausal women. In contrast,
and as also shown in this study, seasonality in bone turnover
seems to be almost absent in healthy men.(13,16) Gender
seems to be an independent determinant of serum 25(OH)D
levels.(12) Also, compared with men, women in Northern
Europe seem to require higher serum 25(OH)D levels to
maintain normal serum PTH concentrations.(48) These gender differences may explain, at least in part, why women are
at greater risk for (subclinical) vitamin D deficiency and
why studies including predominantly males(17,18,46) were
unable to show significant seasonal variations in bone turnover.
Supplementation with calcium and vitamin D3 during the
winter of study year 2 resulted in major changes and shifts
of hormonal and bone marker circannual patterns. As expected, the decrease in vitamin D levels seen during the
previous winter was completely prevented in supplemented
subjects. More importantly, mean PTH levels in the supplemented group were significantly lower than in the previous
winter (within-group comparison) or in the control group
(within-season comparison). Furthermore, the PTH acrophase in the intervention group was not only less pronounced, but also shifted from February to August. Finally,
the changes in serum vitamin D and PTH levels were
associated with significantly lower levels of bone turnover
markers and a loss in their circannual rhythmicity. Taken
together, these and the previous observations support the
hypothesis that transient secondary hyperparathyroidism
caused by (subclinical) vitamin D deficiency may be, at
least in part, the driving force behind the winter increase in
bone turnover, and ultimately, bone loss. We further conclude that supplementation with oral calcium and vitamin
D3 during the winter months may be able to prevent high
bone turnover and bone loss during winter by preventing
secondary hyperparathyroidism caused by low vitamin D
levels.
Our study must be interpreted in the context of its
limitations. First, and most importantly, our study was
not placebo-controlled but open labeled. Study subjects
were asked to not change their lifestyle habits, and this
was confirmed and documented by monthly interview.
Although these control measures revealed no significant
changes between groups or with time, subtle changes in
lifestyle, calcium intake, and other factors cannot be
ruled out.

MEIER ET AL.

Second, our study does not provide evidence on the

relative contributions of vitamin D3 and calcium in the
prevention of seasonal changes in bone turnover and
mass. During the second year of the study, annual mean
25(OH)D levels were significantly higher in the intervention group than in controls. One might therefore speculate that some, or most, of the changes observed would be
caused by vitamin D3 supplementation. So far, only three
studies have investigated the effect of calcium and/or
vitamin D3 supplementation on seasonal changes in bone
metabolism. Storm et al.(15) found that supplementation
with 1000 mg/day of calcium alone prevents bone loss in
elderly women by suppressing bone turnover during the
winter. Another study showed that bone loss during winter was prevented in younger postmenopausal women by
low-dose supplementation with vitamin D3 (400 IU) and
calcium (377 mg/day).(23) Finally, a crossover study in
younger women (mean age, 47 years) using 800 IU of
vitamin D3 revealed no effect on BMD compared with
placebo.(25) Thus, in summary, the effects of oral vitamin
D and calcium on bone turnover and bone mass seem to
depend on a number of factors, including age, gender,
menopausal status, lifestyle, dietary calcium intake, baseline vitamin D levels, and of course, supplementary dosing. From a physiological point of view, further trials
would be helpful to dissect the relative contributions of
vitamin D3 and calcium to seasonal changes in bone
turnover and mass. However, given the efficiency, low
cost, and safety of the combined regimen, the question
could be raised whether such studies would be relevant to
the clinical/public health issue.
Third, subjects in the control group reported higher
levels of physical activity and higher intake of dairy
products at baseline compared with study participants
allocated to the intervention group. However, this fact
would only attenuate the effects observed between the
control and supplemented groups and therefore is unlikely to strongly bias the results. Fourth, vitamin D
levels decrease in winter because of limited sunlight
exposure. However, other confounders such as a more
sedentary lifestyle during winter and different eating
behaviors must be taken into account when evaluating
seasonal loss of bone mass. We did not record total body
fat and lean masses or muscle strength in this study.
However, as noted above, lifestyle factors such as smoking habits or alcohol intake, as well as physical activity,
remained unchanged throughout the whole study period.
The beneficial effect of calcium and vitamin D3 supplementation in our study is observed in a younger population (mean age of the study participants at enrollment,
56.4 years). Because vitamin D deficiency is common in
frail elderly and circannual rhythms in bone turnover are
preserved in advanced age,(14,15,19) our findings should
also be relevant to an elderly population.

REFERENCES
1. Schneider EL, Guralnik JM 1990 The aging of America. Impact on
health care costs. JAMA 263:23352340.
2. Cooper C, Campion G, Melton LJ III 1992 Hip fractures in the
elderly: A world-wide projection. Osteoporos Int 2:285289.

PREVENTION OF SEASONAL BONE LOSS

3. Stamp TC, Round JM 1974 Seasonal changes in human plasma
levels of 25-hydroxyvitamin D. Nature 247:563565.
4. Lester E, Skinner RK, Wills MR 1977 Seasonal variation in
serum-25-hydroxyvitamin-D in the elderly in Britain. Lancet
1:979 980.
5. Lawson DE, Paul AA, Black AE, Cole TJ, Mandal AR, Davie M
1979 Relative contributions of diet and sunlight to vitamin D state
in the elderly. BMJ 2:303305.
6. Poskitt EM, Cole TJ, Lawson DE 1979 Diet, sunlight, and 25hydroxy vitamin D in healthy children and adults. BMJ 1:221223.
7. Stryd RP, Gilbertson TJ, Brunden MN 1979 A seasonal variation
study of 25-hydroxyvitamin D3 serum levels in normal humans.
J Clin Endocrinol Metab 48:771775.
8. Bouillon RA, Auwerx JH, Lissens WD, Pelemans WK 1987 Vitamin D status in the elderly: Seasonal substrate deficiency causes
1,25-dihydroxycholecalciferol deficiency. Am J Clin Nutr 45:755
763.
9. Krall EA, Sahyoun N, Tannenbaum S, Dallal GE, Dawson-Hughes
B 1989 Effect of vitamin D intake on seasonal variations in
parathyroid hormone secretion in postmenopausal women. N Engl
J Med 321:17771783.
10. Sherman SS, Hollis BW, Tobin JD 1990 Vitamin D status and
related parameters in a healthy population: The effects of age, sex,
and season. J Clin Endocrinol Metab 71:405 413.
11. Harris SS, Dawson-Hughes B 1998 Seasonal changes in plasma
25-hydroxyvitamin D concentrations of young American black and
white women. Am J Clin Nutr 67:12321236.
12. Perry HM III, Horowitz M, Morley JE, Patrick P, Vellas B,
Baumgartner R, Garry PJ 1999 Longitudinal changes in serum
25-hydroxyvitamin D in older people. Metabolism 48:1028
1032.
13. Woitge HW, Knothe A, Witte K, Schmidt-Gayk H, Ziegler R,
Lemmer B, Seibel MJ 2000 Circaannual rhythms and interactions
of vitamin D metabolites, parathyroid hormone, and biochemical
markers of skeletal homeostasis: A prospective study. J Bone
Miner Res 15:24432450.
14. Woitge HW, Scheidt-Nave C, Kissling C, Leidig-Bruckner G,
Meyer K, Grauer A, Scharla SH, Ziegler R, Seibel MJ 1998
Seasonal variation of biochemical indexes of bone turnover: Results of a population-based study. J Clin Endocrinol Metab 83:68
75.
15. Storm D, Eslin R, Porter ES, Musgrave K, Vereault D, Patton C,
Kessenich C, Mohan S, Chen T, Holick MF, Rosen CJ 1998
Calcium supplementation prevents seasonal bone loss and changes
in biochemical markers of bone turnover in elderly New England
women: A randomized placebo-controlled trial. J Clin Endocrinol
Metab 83:38173825.
16. Carnevale V, Modoni S, Pileri M, Di Giorgio A, Chiodini I,
Minisola S, Vieth R, Scillitani A 2001 Longitudinal evaluation of
vitamin D status in healthy subjects from southern Italy: Seasonal
and gender differences. Osteoporos Int 12:1026 1030.
17. Hyldstrup L, McNair P, Jensen GF, Transbol I 1986 Seasonal
variations in indices of bone formation precede appropriate bone
mineral changes in normal men. Bone 7:167170.
18. Blumsohn A, Naylor KE, Timm W, Eagleton AC, Hannon RA,
Eastell R 2003 Absence of marked seasonal change in bone turnover: A longitudinal and multicenter cross-sectional study. J Bone
Miner Res 18:1274 1281.
19. Rapuri PB, Kinyamu HK, Gallagher JC, Haynatzka V 2002 Seasonal changes in calciotropic hormones, bone markers, and bone
mineral density in elderly women. J Clin Endocrinol Metab 87:
2024 2032.
20. Thomsen K, Eriksen EF, Jorgensen JC, Charles P, Mosekilde L
1989 Seasonal variation of serum bone GLA protein. Scand J Clin
Lab Invest 49:605 611.
21. Douglas AS, Miller MH, Reid DM, Hutchison JD, Porter RW,
Robins SP 1996 Seasonal differences in biochemical parameters of
bone remodelling. J Clin Pathol 49:284 289.
22. Tsai KS, Hsu SH, Cheng JP, Yang RS 1997 Vitamin D stores of
urban women in Taipei: Effect on bone density and bone turnover,
and seasonal variation. Bone 20:371374.
23. Patel R, Collins D, Bullock S, Swaminathan R, Blake GM, Fogelman I 2001 The effect of season and vitamin D supplementation
on bone mineral density in healthy women: A double-masked
crossover study. Osteoporos Int 12:319 325.

1229
24. Rosen CJ, Morrison A, Zhou H, Storm D, Hunter SJ, Musgrave K,
Chen T, Wei W, Holick MF 1994 Elderly women in northern New
England exhibit seasonal changes in bone mineral density and
calciotropic hormones. Bone Miner 25:8392.
25. Dawson-Hughes B, Dallal GE, Krall EA, Harris S, Sokoll LJ,
Falconer G 1991 Effect of vitamin D supplementation on wintertime and overall bone loss in healthy postmenopausal women. Ann
Intern Med 115:505512.
26. Bergstralh EJ, Sinaki M, Offord KP, Wahner HW, Melton LJ III
1990 Effect of season on physical activity score, back extensor
muscle strength, and lumbar bone mineral density. J Bone Miner
Res 5:371377.
27. Jones G, Dwyer T 1998 Bone mass in prepubertal children: Gender
differences and the role of physical activity and sunlight exposure.
J Clin Endocrinol Metab 83:4274 4279.
28. Dawson-Hughes B, Dallal GE, Krall EA, Sadowski L, Sahyoun N,
Tannenbaum S 1990 A controlled trial of the effect of calcium
supplementation on bone density in postmenopausal women.
N Engl J Med 323:878 883.
29. Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ 1993
Effect of calcium supplementation on bone loss in postmenopausal
women. N Engl J Med 328:460 464.
30. Recker RR, Hinders S, Davies KM, Heaney RP, Stegman MR,
Lappe JM, Kimmel DB 1996 Correcting calcium nutritional deficiency prevents spine fractures in elderly women. J Bone Miner
Res 11:19611966.
31. Riggs BL, OFallon WM, Muhs J, OConnor MK, Kumar R,
Melton LJ III 1998 Long-term effects of calcium supplementation
on serum parathyroid hormone level, bone turnover, and bone loss
in elderly women. J Bone Miner Res 13:168 174.
32. Dawson-Hughes B, Harris SS, Krall EA, Dallal GE 1997 Effect
of calcium and vitamin D supplementation on bone density in
men and women 65 years of age or older. N Engl J Med
337:670 676.
33. Chapuy MC, Arlot ME, Delmas PD, Meunier PJ 1994 Effect of
calcium and cholecalciferol treatment for three years on hip fractures in elderly women. BMJ 308:10811082.
34. ONeill TW, Cooper C, Cannata JB, Diaz Lopez JB, Hoszowski K,
Johnell O, Lorenc RS, Nilsson B, Raspe H, Stewart O 1994
Reproducibility of a questionnaire on risk factors for osteoporosis
in a multicentre prevalence survey: The European Vertebral Osteoporosis Study. Int J Epidemiol 23:559 565.
35. Wildermuth S, Dittrich K, Schmidt-Gayk H, Zahn I, ORiordan
JL 1993 Scintillation proximity assay for calcitriol in serum
without high pressure liquid chromatography. Clin Chim Acta
220:6170.
36. Woitge HW, Seibel MJ, Ziegler R 1996 Comparison of total
and bone-specific alkaline phosphatase in patients with nonskeletal disorder or metabolic bone diseases. Clin Chem 42:
1796 1804.
37. Pratt DA, Daniloff Y, Duncan A, Robins SP 1992 Automated
analysis of the pyridinium crosslinks of collagen in tissue and urine
using solid-phase extraction and reversed-phase high-performance
liquid chromatography. Anal Biochem 207:168 175.
38. Mattes A, Witte K, Hohmann W, Lemmer B 1991 PHARMFITa
nonlinear fitting program for pharmacology. Chronobiol Int
8:460 476.
39. Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ 1995
Long-term effects of calcium supplementation on bone loss and
fractures in postmenopausal women: A randomized controlled
trial. Am J Med 98:331335.
40. Chapuy MC, Arlot ME, Duboeuf F, Brun J, Crouzet B, Arnaud S,
Delmas PD, Meunier PJ 1992 Vitamin D3 and calcium to
prevent hip fractures in the elderly women. N Engl J Med
327:16371642.
41. Larsen ER, Mosekilde L, Foldspang A 2004 Vitamin D and calcium supplementation prevents osteoporotic fractures in elderly
community dwelling residents: A pragmatic population-based
3-year intervention study. J Bone Miner Res 19:370 378.
42. Christiansen C, Christensen MS, McNair P, Hagen C, Stocklund
KE, Transbol I 1980 Prevention of early postmenopausal bone
loss: Controlled 2-year study in 315 normal females. Eur J Clin
Invest 10:273279.
43. Tangpricha V, Pearce EN, Chen TC, Holick MF 2002 Vitamin D
insufficiency among free-living healthy young adults. Am J Med
112:659 662.

1230
44. Zittermann A, Scheld K, Stehle P 1998 Seasonal variations in
vitamin D status and calcium absorption do not influence bone
turnover in young women. Eur J Clin Nutr 52:501506.
45. Overgaard K, Alexandersen P, Riis BJ, Christiansen C 1996 Evaluation of a new commercial IRMA for bone-specific alkaline
phosphatase during treatment with hormone replacement therapy
and calcitonin. Clin Chem 42:973974.
46. Vanderschueren D, Gevers G, Dequeker J, Geusens P, Nijs J,
Devos P, De Roo M, Bouillon R 1991 Seasonal variation in
bone metabolism in young healthy subjects. Calcif Tissue Int
49:84 89.
47. Seibel MJ, Meier C, Woitge HW, Witte K, Lemmer B 2004
Seasonal variation of bone turnover? J Bone Miner Res 19:168
169.
48. Lamberg-Allardt CJ, Outila TA, Karkkainen MU, Rita HJ, Valsta
LM 2001 Vitamin D deficiency and bone health in healthy adults
in Finland: Could this be a concern in other parts of Europe?.
J Bone Miner Res 16:2066 2073.

MEIER ET AL.

Address reprint requests to:

Markus J Seibel, MD, PhD, FRACP
Bone Research Program
ANZAC Research Institute
Department of Endocrinology and Metabolism
The University of Sydney
Sydney, New South Wales 2139, Australia
E-mail: mjs@anzac.edu.au

Received in original form January 12, 2004; in revised form April

14, 2004; accepted April 19, 2004.