Vitamin D status in postmenopausal women living at higher latitudes
in the UK in relation to bone health, overweight,
sunlight exposure and dietary vitamin D
Helen M. Macdonalda,⁎, Alexandra Mavroeidia, Rebecca J. Barra, Alison J. Blacka,
William D. Fraserb, David M. Reida
aDepartment of Medicine and Therapeutics, University of Aberdeen, Medical School Buildings, Foresterhill, Aberdeen, AB25 2ZD, UK
bDepartment of Clinical Chemistry, Royal Liverpool University Hospital, Liverpool L69 3GA, UK
Received 30 July 2007; revised 14 January 2008; accepted 20 January 2008
Available online 9 February 2008
For 5 months a year the UK has insufficient sunlight for cutaneous synthesis of vitamin D and winter requirements are met from stores made
the previous summer. Although there are few natural dietary sources, dietary intake may help maintain vitamin D status.
We investigated the relationship between 25-hydroxyvitamin D (25(OH)D), bone health, overweight, sunlight exposure and dietary vitamin D
in 3113 women (age 54.8 [SD 2.3] years) living at latitude 57°N between 1998–2000. Serum 25(OH)D was measured by high performance liquid
chromatography (HPLC), dietary intakes (food frequency questionnaire, n=2598), sunlight exposure (questionnaire, n=2402) and bone markers
were assessed. Bone mineral density (BMD) was measured by dual x-ray absorptiometry in all women at the sampling visit and 6 years before.
Seasonal variation in 25(OH)D was not substantial with a peak in the autumn (23.7 [9.9] ng/ml) and a nadir in spring (19.7 [7.6] ng/ml). Daily
intake of vitamin D was 4.2 [2.5] μg from food only and 5.8 [4.0] μg including vitamin D from cod liver oil and multivitamins. The latter was
associated with 25(OH)D at each season whereas vitamin D simply from food was associated with 25(OH)D in winter and spring only. Sunlight
exposure was associated with 25(OH)D in summer and autumn. 25(OH)D was negatively associated with increased bone resorption and bone loss
(Pb0.05) remaining significant after adjustment for confounders (age, weight, height, menopausal status/HRT use, physical activity and socio-
economic status). Using an insufficiency cut-off of b28 ng/ml 25(OH)D, showed lower concentrations of bone resorption markers in the upper
category (fDPD/Cr 5.1 [1.7] nmol/mmol compared to 5.3 [2.1] nmol/mmol, P=0.03) and no difference in BMD or bone loss. 25(OH)D was lower
(Pb0.01) and parathyroid hormone higher (Pb0.01) in the top quintile of body mass index. In conclusion, low vitamin D status is associated with
greater bone turnover, bone loss and obesity. Diet appears to attenuate the seasonal variation of vitamin D status in early postmenopausal women
at northerly latitude where quality of sunlight for production of vitamin D is diminished.
© 2008 Elsevier Inc. All rights reserved.
Keywords: 25-hydroxy vitamin D; Postmenopausal women; Bone loss; Dietary vitamin D; Sunlight exposure
Vitamin D, the ‘sunshine vitamin’, is currently the subject of
controversy. Its role in preventing rickets is undisputed but how
much vitamin D is required for maintaining bone strength is
unclear. One argument is that current vitamin D intakes fall far
short of what is required, a view promoted particularly by
workers in the US . In the UK, until recently, calcium and
vitamin D supplementation was the treatment of choice for
preventing fractures in the elderly . Since the publication of
the RECORD trial, the importance of vitamin D has been called
into question and it appears that this regime may not be enough
to prevent further fractures in the ‘healthy’ elderly . Other
recent studies also failed to show a benefit in fracture reduction
with vitamin D supplementation [4,5], with the exception of one
Bone 42 (2008) 996–1003
⁎Corresponding author. Osteoporosis Research Unit, Health Sciences
Building, University of Aberdeen, Foresterhill, Aberdeen, AB25 2ZD, UK.
Fax: +44 1224 559348.
E-mail address: firstname.lastname@example.org (H.M. Macdonald).
8756-3282/$ - see front matter © 2008 Elsevier Inc. All rights reserved.
5-year study in retired doctors and general practice . A meta-
analysis suggests that vitamin D intakes of 800 IU (20 μg) a day
are required to show any benefit . UK government guidelines
are focused on the elderly and those at risk, with no dietary
vitamin D recommendations for healthy individuals b65 years,
who, it is believed obtain sufficient vitamin D from casual
sunlight exposure .Vitamin D cannot be synthesised between
October and April in the UK and it is assumed that winter
requirements are met from the store accumulated the previous
summer. There are limited data that can be directly compared,
on the variation in vitamin D status between different countries.
The contribution of sunlight and diet on vitamin D status has
been reported for 861 women in Australia (latitude 38°)  and
in small studies involving the elderly in Boston (42°) .
Dietary vitamin D has been shown to be important for vitamin D
status in 443 middle-aged women in Northern Norway (65–
71°) . A barrier preventing comparison of data on vitamin D
status from different studies is that of measurement standardiza-
tion . It is generally accepted that high performance liquid
chromatography (HPLC) is the gold standard method for the
measurement of 25(OH)D but this is not always available.
There is also a paucity of data from different regions within
countries, but it would be expected that those further from the
equator may not be able to synthesise as much vitamin D as
those nearer the equator. This has been borne out in France 
and the UK although for Scotland the data was not given
according to region so that comparisons between the North and
South Scotland are not possible .
The results presented in this paper are from a cohort of 3113
well-characterized early postmenopausal women from Aberd-
een, North East Scotland (57° latitude) in whom vitamin D
status was measured using HPLC and whose diet, dietary sup-
plement use and sunlight exposure had been assessed. The aims
were (i) to study the association between dietary vitamin D and
sunlight exposure on vitamin D status according to season and
(ii) to test the relationship between vitamin D status, markers
of bone health and overweight.
Subjects were Caucasian women recruited from the Aberdeen Prospective
Osteoporosis Screening Study (APOSS) . 5119 women attended a baseline
visit between 1990 and 1993, of whom 3883 early postmenopausal women
attended a follow-up visit between December 1997 and August 2000. Blood
samples were collected from 3113 women. A total of 2933 women, who
provided a blood sample, attended between December 1997 and July 1999 (the
majority attending between March 1998 and July 1999 [n=2827] and an
additional 7 women attending in August 1999 [total n=2834]); and 99 women
attending during the period December 1997 to March 1998. Later on, women
who had been unable to attend their original appointment were re-invited; 180
attended and gave a blood sample from May 2000 to August 2000. No women
were seen from September 1999 to May 2000. Written informed consent was
obtained for all the women and the study was approved by Grampian Research
Nutrient intakes were assessed by validated food frequency questionnaire
(FFQ) . Vitamin D was estimated using McCance and Widdowsons’
Composition of Foods version 5, which does not include the estimates of the
vitamin D contribution from metabolites in meat . It has been suggested that
the vitamin D metabolites in meat could have a 5 times potency factor, which
Subject characteristics of women who provided a blood sample at the follow-up visit
n BL/FU Mean±SDMin–maxMean±SDMin–maxMean±SDMin–max
40.0–160.0 0.001 +0.64±1.30 −6.06–+8.44
Height (cm) 161.2±5.9 135.5–
0.001 −0.56±0.59 −5.92–+3.07
FN BMD (g/cm2) 0.881±0.1250.837±0.121 0.001 −0.74±1.07 −8.69–+5.01
LS BMD (g/cm2) 1.051±0.161 1.009±0.1700.001 −0.61±1.22 −4.61–+6.20
Physical activity level
Dietary calcium intake (mg/d)
Diet and supplement calcium (mg/d)
Dietary vitamin D (μg/d)
Dietary vitamin D including supplements (μg/d)
Energy intake (MJ/d)
25 hydroxy vitamin D
Parathyroid hormone (pmol/L)
Menopausal status (pre-, peri-, postmenopausal or
HRT user) (%)
HRT use (none, past, present) (%)
Socio-economic status (categories 1 to 6)
5.8 ± 4.0
3.1, 6.0, 90.941.7, 12.7, 45.6
76.7, 6.6, 16.740.2, 22.0, 37.8
24, 42, 9, 15, 7, 3
FN BMD Femoral neck bone mineral density, LS BMD lumbar spine bone mineral density, P1NP N-terminal propeptide of type 1 collagen, fPYD/Cr free pyridinoline
crosslinks expressed relative to creatinine, fDPD/Cr free deoxypyridinoline crosslinks expressed relative to creatinine.
aSignificant difference between baseline (BL) and follow-up visits (FU) by paired t-test.
997 H.M. Macdonald et al. / Bone 42 (2008) 996–1003
would increase the contribution from meat  although this has not been
validated. Vitamin D intake from supplements and cod liver oil was calculated
from the frequency of use and amount of vitamin D present in the brand used as
indicated by the manufacturer on the product label. Preformed retinol intake
from cod liver oil capsules and vitamin supplements was estimated assuming a
dose 800 μg per tablet or capsule. 2614 women completed the FFQ and women
were excluded if their daily energy intakes were unreasonable (b3 MJ or
N18 MJ) leaving 2598 women for dietary analysis. Physical activity was
assessed by the same questions that had been used in the Scottish Heart Health
Sunlight was estimated by questionnaire previously designed for a vitamin
D/vitamin K intervention study in Dundee (60 miles south of Aberdeen). A
sunshine exposure score was obtained based on (a) how often the subjects were
outside (‘seldom’ 10 min, ‘occasionally’ 20 min and ‘often’ 80 min), (b) what
parts of the body were exposed as a percentage of total body surface area (face
4%, plus hands 6%, plus arms 20%, plus legs (30%), plus some or all trunk
(50%), and (c) weighted according to the intensity of sunlight based on the
latitudinal position for Aberdeen (63.3 for spring, 106.7 for summer, 24.6 for
autumn and 8.27 for winter, weighting figures provided by the meteorological
office courtesy of Colin Driscoll). The sunshine score from that season and the
previous 3 months season were combined. We also obtained details regarding
holidays from the women who had taken holidays abroad or in the South of
England during that period. A total of 2402 women completed the sunlight
questionnaire, the majority (n=2212) hadattendedin 1998–99. Socio-economic
status was assessed by national deprivation scoring, based on residential postal
codes, from 1 (least deprived) to 7 (most deprived) . The more deprived
categories (6 and 7) were combined.
Bone mineral density
was measured by dual energy x-ray absorptiometry using Norland scanners
(Cooper Surgical Inc, Trumbull, CT) . At the previous (baseline) visit, all the
scanned using an XR36. There was a small difference (1.258%) in mean BMD
when comparing 50 phantom measurements on both machines and a correction
Fig. 1. 25-hydroxyvitamin D (25(OH)D) (Mean±SEM) according to season of
visit. ● All women attending in 1998–1999 (n 2391), □ taken holiday abroad in
the same season of sampling or the previous season (n 635), Δ not taken holiday
abroad recently (n 1756).
Relationship between 25(OH)D, sunshine exposure, dietary vitamin D intake and total dietary vitamin D intake according to season
Season (months)Vitamin D status (all women) Correlations between 25(OH)D, diet and sunlight expsoure
% in cut-off groups:
25(OH)D with total
intake of vitamin D
Pearson rankPearson rank Spearman
1016 19.7 (7.6)38.3/84.4882 19.7 (7.6)0.075⁎
971 22.7 (8.5)24.8/72.7790 22.5 (8.4)0.069
569 23.1 (9.5) 25.7/72.6458 23.7 (9.9)0.147⁎⁎
557 21.5 (9.4) 34.1/75.4474 21.6 (9.4)0.133⁎⁎
3113 21.6 (8.7)31.0/77.0 2598 21.6 (8.7)0.081⁎⁎
25(OH)D 25-hydroxyvitamin D.
Relationship between 25(OH) D and markers of bone health
Marker of bone health Correlation with 25(OH)D
Unadjusted Adjusted for
n Pearson rankn Pearson rank
FN BMD (g/cm2)
LS BMD (g/cm2)
FN BMD change (%/y)
LS BMD change (%/y)
Parathyroid hormone (pmol/L)
⁎Pb0.05⁎⁎Pb0.01 with natural logarithm transformation if required.
†Adjusted for age, height, weight, smoking, menopausal status/HRT use, socio-
economic status and physical activity level.
25(OH)D 25-hydroxyvitamin D, FN BMD Femoral neck bone mineral density,
LS BMD lumbar spine bone mineral density, P1NP N-terminal propeptide of
type 1 collagen, fPYD/Cr free pyridinoline crosslinks expressed relative to
creatinine, fDPD/Cr free deoxypyridinoline crosslinks expressed relative to
998 H.M. Macdonald et al. / Bone 42 (2008) 996–1003
precision (coefficient of variation, root mean squared %) at our unit was 1.2% for
the lumbar spine (LS), and 2.3% for the femoral neck (FN).
A fasted second morning void urine sample was used for analysis of bone
crosslinks (fDPD) by HPLC . Creatinine (Cr) was measured in urine by
standard automated techniques (Roche, Lewes, UK) and results were expressed
as fPYD/Cr and fDPD/Cr (nmol/mmol). The bone formation marker, serum N-
terminal propeptide of type 1 collagen (P1NP) was measured using an Enzyme
Chemiluminescent Immuno-Assay (ECLIA) supplied by Roche products Ltd
(Penzberg, GmbH). Non fasted serum samples were analysed for parathyroid
hormone (PTH) by a 2-site immunoradiometric assay (IRMA) for the intact 84
amino acid chain of PTH (Nichols Institute Diagnostics, San Juan, Capistrano,
USA). Vitamin D status was determined by measuring 25-hydroxycholecalciferol
[25(OH)D3] and 25-hydroxyergocalciferol (25(OH)D2) by HPLC. Serum
samples were de-proteinised with acetonitrile and then centrifuged. The su-
pernatant was introduced onto a Varian Bond Elut C18 cartridge. The internal
standard was provided by Roche in the form of ampoules containing 1.0 mg 25
were used as internal controls. The assay was external quality assured by
participation in the DEQAS scheme. The DEQAS data confirmed that the assay
performed well and the results always aligned with the method mean, across the
range of concentrations tested, throughout the period when the analysis for this
study were performed.
Relatively small numbers of results were available for analysis each month
and so the data were combined into 3 monthly seasons. For the investigation of
seasonal variation the data from each year were combined and the difference in
mean 25(OH)D between spring (March to May), summer (June to August),
autumn (September to November) and winter (December to February) was
tested by one way ANOVA. The data for 1998 and 1999 (the time period when
the majority of women were assessed) were also analysed separately. Statistical
analysis was performed using SPSS version 14. Pearson correlations were used
to test the relationship between 25(OH)D and markers of bone health (BMD and
bone turnover markers) with adjustment for cofounders (age, weight, height,
menopausal status/HRT use, physical activity and socio-economic status) and
for testing the association between dietary intake of vitamin D and 25(OH)D
(both variables log transformed for normality) for each season. For the skewed
sunshine score, Spearman rank correlations were used. Linear regression
analysis with adjustment for confounders (age, weight, height, menopausal
status/HRT use and socio-economic status) was used for testing the percentage
variance of 25(OH)D explained by diet and sunshine at each season.
The characteristics of the women are given in Table 1. With
regard to the contributions from different foods: 37% dietary
vitamin D originated from oily fish, 23% from spreads, 13%
from eggs and 10% from breakfast cereals. The remainder came
from cakes (5%), puddings (3%), milk (3%), meat (2%), cheese
(1%) and other foods (3%). Of the 3113 women who provided
blood samples for analysis of 25-hydroxycholecalciferol (25
(OH)D3), only 3 women had detectable 25-hydroxyergocalci-
ferol (the equivalent measure for vitamin D2 status). Two of
these women reported taking dietary supplements, but not
specifically ergocalciferol. A total of 711 (27.4%) of women
who provided dietary information obtained vitamin D from
dietary supplements. Over half (58.3%) of the supplement users
consumed fish liver oils every day (16% of the total population),
which was cod liver oil in all but 2 cases where it was halibut
liver oil. Most of the brands of cod liver oil that the women
purchased were from shops (health food stores or super-
markets), which were claimed to contain 5 μg vitamin D as
cholecalciferol per capsule, whereas internet sources of cod
liver oil contained 2.5 μg or less per dose. Consumption of fish
oils (omega 3), applicable to 56 women in this study, was not
included in this analysis because fish oils contain negligible
amounts of vitamin D. For the 381 women who took a
multivitamin tablet daily, vitamin D intake increased by an
additional 5 μg or more a day in the majority of cases (83%).
2595 women gave information on their holidays including
the destination and sunshine exposure during their holiday. In
the season of the assessment visit or the season prior to the
assessment visit, 666 women had been on holiday abroad or in
the South of England. Of the 2834 women who attended
between spring 1998 and summer 1999, a total of 2391 gave
holiday information. The mean 25(OH)D was identical whether
they had given holiday information or not (data not shown).
Relationship between dietary preformed retinol and markers of bone health
n Retinol from fish
Retinol from vitamin
Retinol from fish oil and vitamin
Retinol from food and all
FN BMD (g/cm2)
LS BMD (g/cm2)
Change FN BMD
Change LS BMD
Parathyroid hormone (pmol/L)
25910.000 +0.022 +0.024
999 H.M. Macdonald et al. / Bone 42 (2008) 996–1003
Women who had taken holidays in the South of England or
abroad in the period prior to the blood sample being taken (n
635) had higher mean 25(OH)D (24.0±8.7 ng/ml) compared to
those who had not (n 1756) (20.7±8.6 ng/ml, Pb0.001),
although the underlying seasonal trends were similar (Fig. 1).
When the results for each year were combined, mean 25(OH)
D was lowest in spring and highest in autumn, which was
statistically different by one way ANOVA (Pb0.001) (Table 2).
The association between sunshine score and 25(OH)D was
only significant in the summer and autumn, whereas vitamin D
from food sources was associated with 25(OH)D in winter and
spring. The association between 25(OH)D and total intake of
vitamin D (including cod liver oil and multivitamins) was
significant for all seasons (Table 2). Adjusting for confounders,
linear regression analysis showed that the variation in 25(OH)D
was predicted by total dietary vitamin D intake (including
supplements) which explained 4.7% of the variation in spring
(with 0.4% additional contribution from sunshine exposure)
and 3.8% of the variation in winter (with no contribution from
sunshine exposure). In summer, the percentage variation of 25
(OH)D explained by diet was 1.8% with 3.4% explained by
sunshine exposure. In autumn, diet explained 4.0% and sun-
shine exposure 3.0% of the total variation in 25(OH)D.
Vitamin D status was positively associated with bone change
(ie reduced bone loss) and reduced bone turnover and the rela-
tionships remained significant after adjustment for confounders
(Table 3). The relationship between markers of bone health and
preformed retinol showed that retinol from fish liver oil was not
supplements was negatively associated with BMD at the sam-
pling visit but with bone change, bone resorption markers and
PTH the trend appeared to show a protective role. Retinol from
food was associated with increased bone resorption.
Comparing subjects who had 25(OH)D≥28 ng/ml with those
b28 ng/ml, showed significant differences in total dietary intake,
weight and bone resorption markers but no difference in BMD,
bone loss or bone formation markers (Table 5). Those with 25
(OH)DN28 ng/ml also had increased home sunshine exposure,
holidaying abroad, sun-bed use, higher socio-economic status,
more dietary supplement use and fewer smokers. The highest
quintile of BMI had the lowest mean circulating 25(OH)D and
highest PTH concentrations (Fig. 2). For 25(OH)D, one way
ANOVA with Dunnett's post hoc analysis showed that Q5 was
(P=0.001) and for PTH Dunnett's post hoc analysis showed Q5
was significantly different from Q1 (P=0.001), Q2 (Pb0.001),
Q3 (P=0.001) and Q4 (P=0.010)). The relationship was sig-
nificant after adjustment for confounders (age, physical activity
level, current smoking, menopausal status/HRT use and socio-
The seasonal variation in 25(OH)D and PTH was similar within
each quintile of BMI although for the lowest BMI quintile the
trend in seasonal variation was not significant (data not shown).
This cross-sectional study shows that although vitamin D
status varies with season at northerly latitude, there is not a large
difference between the peak in the autumn (23.7 ng/ml) and the
nadir in spring (19.7 ng/ml). In contrast, other studies that have
examined seasonal variation of 25(OH)D, measured using
HPLC, have reported substantial differences between summer/
autumn and winter/spring. At 42°N in the US, the summer mean
of 23 ng/ml had decreased to 12 ng/ml in winter . In
Finland, a mean of 25.6 ng/ml 25(OH)D was reported in the
period May to October and 12 ng/ml in November to May .
A study of 179 Korean women (aged 20 to 75 years) (mean 25
(OH)D of 23.3±12.3 ng/ml in September) found that dietary
vitamin D and time spent outside between midday and 2 pm
accounted for 5.5% and 5.0% respectively of the total variation
Characteristics of women with serum 25(OH)D above and below a cut-off of
28 ng ml−1
28 ng ml−1
Percentage of subjects
Home sunshine score
Recent holiday abroad
Once or twice a year
Once a month
Once a week
n=2006 (77.2%)Dietary vitamin D intake
Total intake including
Physical activity level
BMI (kg m−2)
FN BMD (g cm−2)⁎
LS BMD (g cm−2)⁎
FN BMD change (%/y) −0.76±1.08
FN BMD change (%/y) −0.61±1.23
P chi square or independent t-test with natural logarithm transformation if
1000 H.M. Macdonald et al. / Bone 42 (2008) 996–1003
in 25(OH)D , which is similar to the variation of 25(OH)D
accounted for by dietary vitamin D and sunshine in our study
(4% and 3% respectively). A recent UK study of men and
women aged 45 years showed high prevalence of hypovitami-
nosis D (b16 ng/ml 25(OH)D) in Scotland: 60.0% in winter/
spring (n=154/256) and 27.5% in summer/autumn (n=125/
456) but it is not known how many subjects resided in the north
of Scotland . Although the summer/autumn figures are
similar to our findings (24.8% summer and 25.7% in autumn
having b16 ng/ml 25(OH)D) our data reveal less women with
25(OH)D b16 ng/ml for the winter/spring period (34.1% winter
and 38.8% spring), which could be explained by differences in
sunlight exposure the previous summer, differences in dietary
intake of vitamin D (perhaps reflecting an effect of initial BMD
assessment on dietary behaviour), or the use of IDS ELISA
instead of HPLC for measuring 25(OH)D.
The association between total dietary vitamin D intake and
vitamin D status throughout the year suggests that, in this po-
pulation, diet (including dietary supplements) may be impor-
tant in reducing seasonal differences. The dietary supplements
(mainly cod liver oil and multivitamins) provide small amounts
not advocate increasing cod liver oil usage to increase dietary
vitaminDas these supplements contain preformed retinol which
may adversely affect bone health  although our data do not
support this. Either supplementation with vitamin D or food
fortification would be alternative approaches to increasing die-
tary vitamin D. Vitamin D2 (ergocalciferol) may be 70% as
effective as cholecalciferol . Since only 3 women in our
study had detectable 25(OH)D2, the effectiveness of this form
of vitamin D not would not change the outcome of our findings.
The relationship with retinol was not clear-cut. Retinol derived
from cod liver oil showed no detrimental associations with
markers of bone health although retinol from other vitamin
supplements was associated with lower BMD (but a trend for
lower bone resorption at the follow-up visit and less bone loss
over 6 years), and retinol from food was associated with
increased bone resorption. It would appear that retinol from
supplements and food have different effects, which may in part
be due to whether the source of retinol also provides vitamin D.
We found a significant association between vitamin D status,
higher bone turnover and greater BMD change. Since vitamin D
status is associated with being outdoors, a potential confounder
is physical activity level, with women who spend time walking
and gardening having more sunlight exposure. We adjusted
for physical activity level in our analysis and still found a
significant association between 25(OH)D, reduced bone loss
and bone turnover.
Although there are issues regarding methods employed for
25(OH)D analysis, and the suggested cut-offs to denote vitamin
D deficiency, we used the gold standard of HPLC and found
7% of our population had 25(OH)D values below 10 ng/ml.
This is the concentration considered at risk of bone disease in
the National Diet and Nutrition Survey . Our data are
insufficient to conclude that vitamin D status in the rest of our
population is adequate. It is possible that much higher intakes of
vitamin D are required to prevent fracture risk and it is unclear
about how much is required to help prevent other diseases. It
has been suggested that the minimum desirable serum 25(OH)D
should be 70–80 nmol/l (28–32 ng/ml) . Using the 28 ng/ml
cut-off, 77% of our women would be vitamin D insufficient.
However, as we did not observe a significant difference in BMD
(nor PTH) between women with 25(OH)D above or below
28 ng/ml we cannot be certain that increasing 25(OH)D above
28 ng/ml would have any long-term benefit for bone health.
Longer sun exposure times are required to synthesise vitamin
D at a distance from the equator . Recent calls have been
made to increase intakes of dietary vitamin D in order to reach
exposure at latitude 21° had 25(OH)D below this concentration
. Recent vitamin D supplementation studies have not shown
One way ANOVA for 25(OH)D, Pb0.001 with Dunnett's post hoc analysis showing that Q5 was significantly different from Q1 (P=0.036), Q2 (Pb0.001) and
Q3 (P=0.001). One way ANOVA for PTH Pb0.001 (Dunnett's post hoc analysis: Q5 was significantly different from Q1 (P=0.001), Q2 (Pb0.001), Q3 (P=0.001) and
Q4(P=0.010)).BMI Mean±SDQ121.3±1.3kg m−2(n624),Q223.9±0.6 kgm−2(n657),Q325.9±0.6kgm−2(n 627),Q4 28.5±0.9kg m−2(n622),Q534.2±3.9 kg
1001 H.M. Macdonald et al. / Bone 42 (2008) 996–1003
any benefit in reducing fractures. Compliance may have been an
issue in the RECORD trial since the supplements were sent by
post and vitamin D status was measured only in a small sample
of subjects . Two other UK trials also question the value of
vitamin D supplementation [4,32]. A meta-analysis showed a
benefit of vitamin D supplements on fracture  but this was
weighted by a study performed in French nursing homes . It
was suggested that vitamin D dosage is a critical issue since
successful vitamin D supplementation trials have used 700 to
800 IU (17.5–20 μg) a day whereas studies using 400 IU
(including those using 800 IU where an assumption for lower
compliance was made) were found to be ineffective . The
large US Women's Health Initiative study has also not provided
evidence that vitamin D supplementation is effective  but
again critics have suggested that compliance was an issue in that
Recent evidence suggests that obesity is linked to lower
serum 25(OH)D [33,34]. The lower serum concentrations could
be due to decreased sunshine exposure in the obese although
there are data to suggest that there is decreased bioavailability in
obese subjects . Our study found that 25(OH)D was lowest
(and PTH highest) in the highest quintile of BMI. This group
also had reduced physical activity but the association did not
change after adjustment for this confounder. It is recognised that
BMI is not a precise measure of adipose tissue since lean tissue
will also contribute to weight resulting in a high BMI for
muscular subjects. However, as the women in our population do
not participate in competitive sports or activities that would lead
to a substantial increase in lean tissue, we believe that BMI is a
good surrogate measure for adiposity.
The strength of our data is that vitamin D status has been
measured using the current gold standard of HPLC in a large
study in the North of the UK. We also included measures of
PTH, bone turnover markers; and bone mineral density mea-
surements performed on two occasions. A limitation is that
the 25(OH)D data were cross-sectional and we do not have
measures of central obesity.
In conclusion, based on this cross-sectional study at 57°N,
dietary intakes of vitamin D (including multivitamins and cod
liver oil) contribute to vitamin D status throughout the year in
early postmenopausal women. Seasonal differences were not
marked in spite of the northerly latitude and reduced quality of
sunlight of the appropriate wavelength for cutaneous synthesis
of vitamin D. Vitamin D status was associated with reduced
bone turnover and less bone loss. Obese subjects had lower
vitamin D status and raised PTH concentrations. More research
is required to establish what concentration of 25(OH)D is re-
quired for optimal health.
The authors wish to thank Colin Driscoll for providing data
on sunlight intensity weighting, Caroline Bolton Smith for the
sunlight exposure questionnaire and Brian Diffey for advice on
calculating the sunshine exposure score. We are indebted to the
staff at the Osteoporosis Research Unit and to all the women
who kindly took part. The study was part funded by Grampian
Osteoporosis Trust and the UK Foods Standards Agency. Any
views expressed are the authors' own. Thanks to Dr N Hoyle
(Roche) for supplying the P1NP kits.
 Dawson-Hughes B, Heaney RP, Holick MF, Lips P, Meunier PJ, Vieth R.
Estimates of optimal vitamin D status. Osteoporos Int 2005;16:713–6.
 Chapuy MC, Arlot ME, Delmas PD, Meunier PJ. Effect of calcium and
cholecalciferol treatment for three years on hip fractures in elderly women.
Brit Med J 1994;308:1081–2.
 Grant AM, Avenell A, Campbell MK, McDonald AM, MacLennan GS,
McPherson GC, et al. Oral vitamin D3 and calcium for secondary
prevention of low-trauma fractures in elderly people (Randomised
Evaluation of Calcium Or vitamin D, RECORD): a randomised placebo-
controlled trial. Lancet 2005;365:1621–8.
 Porthouse J, Cockayne S, King C, Saxon L, Steele E, Aspray T, et al.
Randomised controlled trial of calcium and supplementation with
cholecalciferol (vitamin D3) for prevention of fractures in primary care.
Calcium plus vitamin D supplementation and the risk of fractures. N Engl
J Med 2006;354:669–83.
 Trivedi DP, Doll R, Khaw KT. Effect of four monthly oral vitamin D3
(cholecalciferol) supplementation on fractures and mortality in men and
women living in the community: randomised double blind controlled trial.
 Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E, Dietrich T,
Dawson-HughesB. FracturepreventionwithvitaminDsupplementation: a
meta-analysis of randomized controlled trials. JAMA 2005;293:2257–64.
 Department of Health. Nutrition and bone health: with particular reference
to calcium and vitamin D. London: HMSO; 1998.
 Pasco JA, Henry MJ, Nicholson GC, Sanders KM, Kotowicz MA. Vitamin
D status of women in the Geelong Osteoporosis Study: association with
diet and casual exposure to sunlight. Med J Aust 2001;175:401–5.
 Webb AR, Pilbeam C, Hanafin N, Holick MF. An evaluation of the relative
contributions of exposure to sunlight and of diet to the circulating
concentrations of 25-hydroxyvitamin D in an elderly nursing home
population in Boston. Am J Clin Nutr 1990;51:1075–81.
 Brustad M, Alsaker E, Engelsen O, Aksnes L, Lund E. Vitamin D status of
middle-aged women at 65-71 degrees N in relation to dietary intake and
exposure to ultraviolet radiation. Public Health Nutr 2004;7:327–35.
 Binkley N, Krueger D, Cowgill CS, Plum L, Lake E, Hansen KE, et al.
Assay variation confounds the diagnosis of hypovitaminosis D: a call for
standardization. J Clin Endocrinol Metab 2004;89:3152–7.
 Chapuy MC, Preziosi P, Maamer M, Arnaud S, Galan P, Hercberg S, et al.
Prevalence of vitamin D insufficiency in an adult normal population.
Osteoporos Int 1997;7:439–43.
 Hypponen E, Power C. Hypovitaminosis D in British adults at age 45 y:
nationwide cohort study of dietary and lifestyle predictors. Am J Clin Nutr
et al. Large-scale population-based study shows no evidence of association
between common polymorphism of the VDR gene and BMD in British
women. J Bone Miner Res 2006;21:151–62.
 Macdonald HM, New SA, Reid DM. Longitudinal changes in dietary
intake in Scottish women around the menopause: changes in dietary
pattern result in minor changes in nutrient intake. Public Health Nutr
 Holland B, Welch AA, Unwin ID, Buss DH, Paul AA. Southgate DAT.
McCance and Widdowson's The composition of foods. Cambridge, UK:
Royal Society of Chemistry; 1992.
 Gibson SA, Ashwell M. New vitamin D values for meat and their
implication for vitamin D intake in adults. Proc Nutr Soc 1997;56:116A.
 Tunstall-Pedoe H, Smith WCS, Crombie IK, Tavendale R. Coronary risk
factor and lifestyle variation across Scotland: results from the Scottish
Heart Health Study. Scot Med J 1989;34:556–60.
1002 H.M. Macdonald et al. / Bone 42 (2008) 996–1003
 McLoone. Carstairs scores for Scottish postcode sectors from the 1991 Download full-text
census. Glasgow: University of Glasgow; 1994.
 MacDonald AG, Birkinshaw G, Durham B, Bucknall RC, Fraser WD.
Biochemical markers of bone turnover in seronegative spondylarthro-
pathy: relationship to disease activity. Br J Rheumatol 1997;36:50–3.
 Stryd RP, Gilbertson TJ, Brunden MN. A seasonal variation study of 25-
hydroxyvitamin D3 serum levels in normal humans. J Clin Endocrinol
 Dabek JT, Harkonen M, Wahlroos O, Adlercreutz H. Assay for plasma 25-
hydroxyvitamin D2 and 25-hydroxyvitamin D3 by “high-performance”
liquid chromatography. Clin Chem 1981;27:1346–51.
 Kim JH, Moon SJ. Time spent outdoors and seasonal variation in serum
concentrations of 25-hydroxyvitamin D in Korean women. Int J Food Sci
 Michaelsson K, Lithell H, Vessby B, Melhus H. Serum retinol levels and
the risk of fracture. N Engl J Med 2003;348:287–94.
 Trang HM, Cole DE, Rubin LA, Pierratos A, Siu S, Vieth R. Evidence that
vitamin D3 increases serum 25-hydroxyvitamin D more efficiently than
does vitamin D2. Am J Clin Nutr 1998;68:854–8.
 Finch S, Doyle W, Lowe C. National Diet and Nutrition Survey: people
aged 65 years or over. London: TSO; 1998.
 Holick MJ. Vitamin D. In: Holick MJ, Dawson-Hughes B, editors. Nutrition
and bone health. Totowa, NJ: Humana Press Inc; 2004. p. 403–40.
 Vieth R, Bischoff-Ferrari H, Boucher BJ, Dawson-Hughes B, Garland CF,
Heaney RP, et al. The urgent need to recommend an intake of vitamin D
that is effective. Am J Clin Nutr 2007;85:649–50.
 Vieth R. Vitamin D supplementation, 25-hydroxyvitamin D concentra-
tions, and safety. Am J Clin Nutr 1999;69:842–56.
 Binkley N, Novotny R, Krueger D, Kawahara T, Daida YG, Lensmeyer G,
et al. Low vitamin D status despite abundant sun exposure. J Clin
Endocrinol Metab 2007;92:2130–5.
 Anderson FH, Smith HE, Raphael HM, Crozier SR, Cooper C. Effect of
annual intramuscular vitamin D3 supplementation on fracture risk in 9440
community-living older people: the Wessex Fracture Prevention Trial.
J Bone Miner Res 2004;19:S57.
 Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF. De-
creased bioavailability of vitamin D in obesity. Am J Clin Nutr
 Snijder MB, van Dam RM, Visser M, Deeg DJ, Dekker JM, Bouter LM, et al.
Adiposity in relation to vitamin D status and parathyroid hormone levels: a
population-based study in older men and women. J Clin Endocrinol Metab
1003 H.M. Macdonald et al. / Bone 42 (2008) 996–1003