Estimation of the dietary requirement for vitamin D in healthy
adults
Kevin D Cashman, Tom R Hill, Alice J Lucey, Nicola Taylor, Kelly M Seamans, Siobhan Muldowney,
Anthony P FitzGerald, Albert Flynn, Maria S Barnes, Geraldine Horigan, Maxine P Bonham, Emeir M Duffy, JJ Strain,
Julie MW Wallace, and Mairead Kiely
ABSTRACT
Background: Knowledge gaps have contributed to considerable
variation among international dietary recommendations for vitamin D.
Objective: We aimed to establish the distribution of dietary vitamin
D required to maintain serum 25-hydroxyvitamin D [25(OH)D]
concentrations above several proposed cutoffs (ie, 25, 37.5, 50, and
80 nmol/L) during wintertime after adjustment for the effect of
summer sunshine exposure and diet.
Design: A randomized, placebo-controlled, double-blind 22-wk intervention
study was conducted in men and women aged 20–40 y (n
238) by using different supplemental doses (0, 5, 10, and 15g/d)
of vitamin D3 throughout the winter. Serum 25(OH)D concentrations
were measured by using enzyme-linked immunoassay at baseline
(October 2006) and endpoint (March 2007).
Results: There were clear dose-related increments (P 0.0001) in
serum 25(OH)D with increasing supplemental vitaminD3. The slope
of the relation between vitamin D intake and serum 25(OH)D was
1.96 nmol L1 g1 intake. The vitamin D intake that maintained
serum 25(OH)D concentrations of 25 nmol/L in 97.5% of
the sample was 8.7 g/d. This intake ranged from 7.2 g/d in those
who enjoyed sunshine exposure, 8.8 g/d in those who sometimes
had sun exposure, and 12.3 g/d in those who avoided sunshine.
Vitamin D intakes required to maintain serum 25(OH)D concentrations
of 37.5, 50, and 80 nmol/L in 97.5% of the sample were
19.9, 28.0, and 41.1 g/d, respectively.
Conclusion: The range of vitamin D intakes required to ensure
maintenance of wintertime vitamin D status [as defined by incremental
cutoffs of serum 25(OH)D] in the vast majority (97.5%)
of 20 – 40-y-old adults, considering a variety of sun exposure
preferences, is between 7.2 and 41.1 g/d. Am J Clin Nutr
2008;88:1535– 42.
INTRODUCTION
It is well established that prolonged and severe clinical vitaminD
deficiency, represented as serum or plasma 25-hydroxyvitamin D
[25(OH)D] concentrations of 10–25 nmol/L, leads to rickets in
children and osteomalacia in adults (1). Less severe vitamin D deficiency
causes secondary hyperparathyroidism and increases bone
turnover and bone loss (2–4). Currently, in the United Kingdom, a
plasma concentration of 25 nmol 25(OH)D/L is used as the lower
threshold for vitamin D status (1). There is, however, a lack of
consensus on the cutoffs of plasma 25(OH)D that define the lower
limit of adequacy or sufficiency, and values between 30 and 80
nmol/L have been suggested (5–7). In addition, a growing body of
evidence suggests that serum 25(OH)D concentrations of 50
nmol/L may be associated with greater risk of a wide range of other
nonskeletal chronic diseases (8, 9). With this in mind, it is of concern
that a high prevalence of low vitamin D status has been reported in
adults from many countries, as reviewed in several reports (10–13).
In addition, the age profile of those with low vitamin D status is
contrary to previously accepted wisdom; for example, younger
adults in the United Kingdom are more likely to have serum
25(OH)D values of 25 nmol/L than are older adults (20.2% and
11.7% of adults aged 19–34 y and 35–64 y, respectively) (14).
In humans, vitaminDis obtained primarily through cutaneous
biosynthesis in the presence of ultraviolet B (UVB) sunlight and
also from the diet (1, 5). In the absence of sufficient sun exposure
for dermal synthesis, vitamin D becomes an essential nutrient.
Considerable variation exists between authoritative dietary recommendations
for vitamin D intakes (1, 5, 15, 16). The UK
Committee on Medical Aspects of Food and Nutrition Policy
(COMA) chose in 1991 not to set a reference nutrient intake
(RNI) for persons aged 4–64 y on the basis of the expectation that
skin synthesis of vitamin D would generally ensure adequacy
(15), a recommendation upheld in 1998 by the UK COMA subgroup
on bone health (1).
In contrast, the US Dietary Reference Intake (DRI) panel for
calcium and related nutrients set adequate intakes (AIs) for vitamin
D in 1997 (5). The US DRI panel concluded that there was
insufficient evidence to set estimated average requirements
[(EAR)], which are the foundation for setting recommended
dietary allowances (RDA), for vitamin D, and the panel emphasized
the fact that contributions from sunlight and food are difficult
to measure (5). An AI for vitamin D was set on the basis of
intakes necessary to achieve “normal” ranges of serum 25(OH)D
concentrations. However, in establishing the AI, the US DRI
1 From the Departments of Food and Nutritional Sciences (TRH, AJL, NT,
KS, SM, AF, MK, and KDC), Medicine (KDC), Epidemiology and Public
Health (APF), and Statistics (APF), University College, Cork, Ireland, and
the Northern Ireland Centre for Food and Health, University of Ulster, Coleraine,
United Kingdom (GH, MSB, MPB, EMD, JMWW, and JJS).
2 Supported by the UK Food Standards Agency.
3 Reprints not available. Address correspondence to KD Cashman, Department
of Food and Nutritional Sciences, and Department of Medicine,
University College, Cork, Ireland. E-mail:
k.cashman@ucc.ie.
Received June 25, 2008. Accepted for publication August 5, 2008.
doi: 10.3945/ajcn.2008.26594.
Am J Clin Nutr 2008;88:1535– 42. Printed in USA. © 2008 American Society for Nutrition 1535
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
panel assumed that there was no cutaneous synthesis of vitamin
D through sun exposure (5). The European Union (EU) dietary
recommendation [population reference intake (PRI)] for vitamin
D in adults ranges from 0 to 10 g/d to account for the widely
varying latitudes in which EU citizens live (35–70 oN) (16).
The aim of the present study was to perform a randomized
controlled intervention study in adults (aged 20–40 y) by using
supplemental intakes (0, 5, 10, and 15 g/d) of vitamin D3
throughout the winter to establish the distribution of dietary
requirements for the maintenance of nutritional adequacy of vitamin
D during late winter, as indicated by serum 25(OH)D
concentrations ranging from 25 nmol/L to 80 nmol/L. In
addition, the effect of summer sunshine exposure (and the resulting
tissue vitaminDstores) on these dietary requirements was
assessed.
SUBJECTS AND METHODS
Subjects
A total of 245 apparently healthy adults were recruited to this
2-center 22-wk vitamin D intervention trial. Subjects were recruited
in Cork, Ireland (n 123), and Coleraine, Northern
Ireland (United Kingdom) (n 122), with the use of advertisements
placed around the universities, at shopping centers, and at
various workplaces. We aimed to recruit equal numbers of men
and women and equal numbers of participants aged from 20 to
30 y and from 30 to 40 y. Inclusion criteria were consenting
white men and women aged 20–40 y. Volunteers were excluded
if they consumed vitamin D–containing supplements for 12 wk
before initiation of the study or if they planned to take a winter
vacation (during the course of the 22-wk intervention) to a location
at which either the altitude or the latitude would be predicted
to result in significant cutaneous vitamin D synthesis from solar
radiation (eg, a mountain ski resort or a sunny winter coastal
resort). Severe medical illness, hypercalcemia, known intestinal
malabsorption syndrome, excessive alcohol use, current medications
known to interfere with vitaminDmetabolism, and pregnancy
or plans to become pregnant during the 22-wk intervention
also were reasons for exclusion.
All participants gave written informed consent according to
the Helsinki Declaration. The study was approved by the Clinical
Research Ethics Committee of the Cork Teaching Hospitals,
University College Cork, and by the Research Ethics Committee
of the University of Ulster, Coleraine. The study was also registered
on the Current Controlled Trials Register (ISRCTN Reg.
no. ISRCTN20236112; Internet:
http://www.controlled-trials.
com/ISRCTN20236112).
Design and conduct of study
The present study was a double-blind, placebo-controlled trial
in which adult subjects at 2 centers were randomly assigned to
receive 0 (placebo), 5, 10, or 15g vitamin D3/d for 22 wk. This
range of supplemental vitamin D was estimated to provide a
range of intakes of vitamin D that fit closely within the 2.5th and
97.5th percentiles of intakes for UK adults (data from the National
Diet and Nutrition Survey [NDNS (14)]. The upper end of
the estimated range of daily total intake was well below the
tolerable upper intake level (UL) for vitamin D (50 g/d) established
by the EU Scientific Committee on Food (16) and the US
DRI panel (5). Randomization was centralized, computergenerated,
stratified by center, and adjusted for age (20–30 or
30–40 y) and sex. The vitamin D3 capsules and matching
placebo capsules were produced by Banner Pharmacaps (Tilburg,
Netherlands) and were identical in appearance and taste.
The vitamin D3 content of the capsules was independently confirmed
by laboratory analysis (Consultus Ltd, Glanmire, Ireland).
Compliance was assessed by capsule counting. An a priori
decision was made to include only those subjects whose compliance
exceeded 85%. The allocation remained concealed until
the final analyses, and all data were reported by persons who
were blinded to the allocation scheme.
The study was carried out in 2 locations: Cork, Ireland (latitude
51 °N), and Coleraine, Northern Ireland, United Kingdom (latitude
55 °N). A 2-center approach was chosen because of the
differences in summer weather patterns and cloud cover between
the 2 centers and because the 2 centers, which are separated by 4 °
of latitude, provide a geographic spread that covers a sizeable
area of Ireland and the United Kingdom. [Data from the NDNS
show that mean serum 25(OH)D concentrations in older adults
were10 nmol/L lower in the northern part of the United Kingdom
(55–57 °N) than in London and the Southeast (51 °N) (17)].
All subjects were recruited between March 2006 and June
2006, and they were asked to keep a sunshine-exposure diary and
answer a sunshine-exposure questionnaire during a defined period
in July 2006. Instructions on recording and completing the
sunshine diary were provided during a screening visit to the study
centers. The 7-d diary was developed as part of the EU Framework
V–funded OPTIFORD project (18). Variables recorded
included time spent outdoors, weather conditions, and manner of
dress.
All subjects commenced the intervention study between October
2 and November 2, 2006, and they finished the study 22 wk
later, between February 27 and April 7, 2007; this timespan
represents a period during which vitamin D status would be
expected to decline to a nadir (19). During the intervention phase,
each participant made 2 further visits to the study centers, at
baseline (week 0) and endpoint (week 22). At each visit, an
overnight fasting blood sample was taken from each participant
by a trained phlebotomist between 0830 and 1030. Blood was
collected by venipuncture into an evacuated tube without an
additive and processed to serum, which was immediately stored
at 80 °C until required for analysis. Anthropometric measurements
including height, weight, waist circumference, and biceps,
triceps, subscapular and suprailiac skinfold thicknesses were
taken as described previously (20). Habitual intakes of calcium
and vitamin D were estimated by using a validated foodfrequency
questionnaire (FFQ) (21, 22), which was administered
by a research nutritionist; a health and lifestyle questionnaire,
which assessed physical activity, general health, smoking status,
and alcohol consumption, also was completed. Participants were
contacted monthly by phone, E-mail, or both to promote compliance
and encourage completion of the study protocol.
Laboratory analysis
Serum 25-hydroxyvitamin D
25(OH)D concentrations were measured in serum samples by
using an enzyme-linked immunosorbent assay [(ELISA) OCTEIA
25-Hydroxy Vitamin D; Immuno Diagnostic Systems Ltd,
Boldon, United Kingdom]. The intraassay and interassay CV for
1536 CASHMAN ET AL
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
the ELISA method was 5.9% and 6.6%, respectively. This
ELISA is used for the quantitative measurement of 25(OH)D,
further details of which have been described previously (23). The
quality and accuracy of serum 25(OH)D analysis in our laboratory
are ensured on an ongoing basis by participation in the
Vitamin D External Quality Assessment Scheme [(DEQAS)
Charing Cross Hospital, London, United Kingdom].
Serum intact parathyroid hormone
Serum intact parathyroid hormone (iPTH) concentrations
were measured in serum with the use of an ELISA (MD Biosciences
Inc, St Paul, MN). The intraassay and interassayCVwas
3.4% and 3.8%, respectively.
Serum total calcium
Total calcium and albumin concentrations in serum were measured
by using an automated system (Instrumentation LaboratoriesUKLtd,
Warrington, United Kingdom). Serum calcium concentrations
were adjusted for albumin concentration.
Mathematical modeling of the relation of vitamin D
intake and status
The aim of the modeling was to describe the conditional distribution
of serum 25(OH)D at specific values of vitamin D
intake. Given the skewed distribution of serum 25(OH)D, the
mean value of log-transformed serum 25(OH)D was modeled as
a linear function of vitamin D intake. The linear model was
chosen after a series of models were assessed for best fit. A
regression model was used to estimate the variation in 25(OH)D
concentrations around the mean, and Q-Q plots were used to
examine the assumption that variation around the predicted value
was normally distributed. The distribution of log serum
25(OH)D was transformed to obtain the distribution for serum
25(OH)D as a function of total vitamin D intake. Finally, we
estimated the dietary requirements for vitamin D to maintain
selected percentages of the population above specific serum
25(OH)D concentrations. The 95% CIs of required vitamin D
intakes were calculated by using a bias-corrected bootstrap based
on 10 000 replications.Amore complex model that included sun
preference as a categorical variable allowed the mean concentrations
of log serum 25(OH)D to vary with sun preference. Sun
preference and total vitamin D intake were independent predictors
of serum 25(OH)D concentrations. There was no evidence
that the association between serum 25(OH)D and vitamin D
intake depended on sun preference. Results were verified by
using robust regression models that minimized the effect of outliers
and heteroscedasticity.
Statistical analysis
Because of the relative paucity of data on the relation between
habitual vitamin D intake and serum 25(OH)D concentrations,
power calculations were performed under relatively pessimistic
assumptions about the magnitude of any relation and the residual
variation in serum 25(OH)D concentration, after the effect of
background dietary intake has been removed. Specifically, a
value of 0.5 was assumed to represent the minimum clinically
important slope, and the residual variation of serum concentration
of 25(OH)D around the mean line was assumed to be normal.
On the basis of the distribution of data from older women from
our group’s previous study (22), it was assumed that the distribution
of dietary intakes in the current study would be similar.
With these assumptions, a study design recruiting 240 volunteers,
60 ofwhomwere assigned to 1 of 4 dose levels (0, 5, 10, and
15 g vitamin D/d), and including 20% to cover potential dropouts,
had 90% power to show a dose-response relation.
Statistical analysis of the data were conducted by using SPSS
for WINDOWS software (version 12.0; SPSS Inc, Chicago, IL)
and STATA software (version 10.0; StataCorp LP, College Station,
TX). The distributions of all variables were tested with the
use of Kolmogorov-Smirnov tests. Descriptive statistics (x SD
or median and interquartile range, where appropriate) were determined
for all variables. Serum concentrations of 25(OH)D and
PTH, as well as baseline dietary vitaminDand calcium, were not
normally distributed and thus were log transformed to achieve
near-normal distributions. Serum concentrations of albumincorrected
calcium, endpoint dietary calcium and total vitamin D
concentrations, and age, weight, height, and body mass index
(BMI; in kg/m2) were normally distributed. Baseline characteristics
of subjects in both study centers were compared by using
chi-square (for male-to-female ratio and sun preference) or unpaired
Student’s t tests. Baseline characteristics of subjects in the
different intervention groups were compared by using chi-square
tests (for male-to-female ratio and sun preference) and one-factor
analysis of variance (ANOVA). Changes in calcium and vitamin
Dintake from baseline to endpoint were tested by usingANOVA
and Tukey’s test. Linear models of the response in a repeatedmeasures
ANOVA for the differences in serum 25(OH)D and
PTH concentrations were also constructed. The main effects
included were dietary treatment and sex. The linear models also
included 2-way interactions between the main effects. P 0.05
was considered to be statistically significant.
RESULTS
Baseline characteristics of subjects
Of the 245 subjects recruited into the study, 238 returned for
the intervention phase, and 221 completed the intervention
phase. The progress of these subjects through the trial is shown
in Figure 1. Subjects in Cork were slightly but significantly (P
0.01) younger than those in Coleraine (Table 1), but there was
no significant (P 0.5) difference in mean age between males
and females (data not shown). There was no significant difference
in mean weight, height, orBMIat baseline between subjects
from the 2 centers (Table 1).
Two-factor ANOVA showed that, whereas baseline serum
25(OH)D concentrations did not differ by sex (P 0.5), they
differed significantly (P0.001) by center (Table 1). There was
no significant interaction (P0.2) between these 2 main factors.
Baseline serum PTH concentrations were similar in subjects
from both centers (P0.7; Table 1) but were significantly higher
in women than in men [median (interquartile range); 49.2 (35.3–
63.7) and 40.7 (30.1–54) ng/mL, respectively; P 0.05). Mean
SDbaseline serum albumin– corrected calcium concentrations
were significantly lower in subjects from Cork than in those from
Coleraine (P 0.001; Table 1) and significantly higher in men
than in women (8.80.3 and 8.70.2 nmol/L, respectively; P
0.01).
There was no significant between-center difference in habitual
vitamin D or calcium intake in subjects at baseline (Table 1);
DIETARY VITAMIN D REQUIREMENT IN HEALTHY ADULTS 1537
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
however, men had significantly (P 0.006) higher intakes of
vitamin D and calcium than did women [3.8 (2.4 –5.
and 3.3
(1.7–5.0) g/d, respectively, for vitamin D; 1128 (857–1485)
and 803 (587–1045) mg/d, respectively, for calcium), which was
expected, because men typically have higher food and nutrient
intakes than do women.
Baseline serum 25(OH)D concentrations in subjects who described
themselves as often having exposure to summer sunshine
(n 84) were significantly (P 0.01) higher than those in
subjectswhodescribed themselves as avoiding (n27) or sometimes
having exposure to summer sunshine (n 107) [82.4
(61.6 –105.9), 50.5 (43.7–78.2), and 65.2 (51.0–86.3) nmol/L,
respectively]. The difference between the latter 2 groups was not
significant (P 0.3).
Effects of vitamin D intervention
There difference in mean age, weight, height or BMI at baseline
among the 4 treatment groups was not significant (P 0.7;
data not shown). Similarly, there was no significant difference in
the proportion of men to women, in sun exposure preferences, in
mean habitual dietary vitamin D or calcium intake, or in mean
preintervention serum 25(OH)D, PTH, or albumin-corrected calcium
concentrations among the treatment groups (Table 2).
No adverse events were reported during the study. Of the 17
dropouts, 5, 6, 2, and 4 were from the placebo and 5, 10, and 15
g vitamin D/d groups, respectively. Subjects dropped out for a
variety of reasons (eg, pregnancy, loss of interest, illness unrelated
to the intervention, or desire to take sun holiday), and in no
instance was dropping out related to the intervention. Six subjects
failed to exceed the minimum 85% compliance, and they
were excluded from the main analysis. In the remaining subjects,
there was good supplement adherence based on pill count [100%
(97.4 –100%)], and compliance did not differ significantly
among the 4 treatment groups (P 0.7).
As expected, total vitamin D intake (diet plus supplemental
vitamin D) increased in a dose-related manner with supplementation
(4.4 3.6, 9.1 2.4, 13.9 2.0, and 19.2 3.1
g/d in the placebo and 5, 10, and 15 g vitamin D/d groups,
62 Placebo
Dropouts = 5
Noncompliers = 0
57 (5 μg/d) Vitamin D3
Dropouts = 6
Noncompliers = 3
58 (15 μg/d) Vitamin D3
Dropouts = 4
Noncompliers = 1
48 Endpoint
61 (10 μg/d) Vitamin D3
Dropouts = 2
Noncompliers = 2
57 Endpoint 57 Endpoint 53 Endpoint
FIGURE 1. Flow of subjects through the study.
TABLE 1
Baseline characteristics of the subjects who entered the intervention study1
All subjects
(n 221)
Cork
(n 108)
Coleraine
(n 113)
Male:female (n) 111:111 54:54 57:56
Age (y) 29.9 6.22 28.7 6.0 31.1 6.33
Weight (kg) 77.0 15.8 76.6 15.9 77.3 15.7
Height (m) 1.71 0.09 1.72 0.10 1.71 0.08
BMI (kg/m2) 26.1 4.3 25.8 4.0 26.3 4.5
Dietary calcium (mg/d) 976 (682–1301)4 955 (676–1301) 990 (718–1307)
Dietary vitamin D (g/d) 3.6 (2.1–5.4) 3.4 (2.1–5.1) 3.6 (2.3–5.7)
Serum 25(OH)D (nmol/L) 70.3 (53.4–90.3) 76.2 (57.4–104.1) 64.9 (48.5–84.9)4
Serum PTH (ng/mL) 43.8 (32.3–59.3) 43.6 (31.5–57.6) 44.1 (34.4–60.1)
Serum calcium (mmol/L)5 8.8 0.3 8.7 0.3 8.9 0.34
Summer sun exposure preferences (%)
Sun avoiders 12.7 13.0 12.4
Some exposure 48.8 54.0 44.2
Frequent exposure 38.5 33.0 43.4
1 PTH, parathyroid hormone; 25(OH)D, 25-hydroxyvitamin D.
2 x SD (all such values).
3 Significantly different from subjects in Cork, P 0.001 (unpaired Student’s t tests).
4 Median; interquartile range in parentheses (all such values); used in the case of nonnormally distributed variables.
5 Albumin corrected.
1538 CASHMAN ET AL
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
respectively; P 0.0001). In contrast, calcium intake at endpoint
did not differ significantly (P0.5) among the 4 groups
(data not shown).
There was a significant (P 0.0001) effect of treatment on
mean postintervention serum 25(OH)D concentrations, with
clear dose-related increments with increasing supplemental vitamin
D3 (Table 2). There was no significant difference in mean
postintervention serum albumin– corrected calcium concentrations
among the treatment groups (8.60.3, 8.70.3, 8.60.3,
and 8.60.3 mmol/L in the placebo and 5, 10, and 15g vitamin
D/d groups, respectively; P 0.526) and no significant change
over time (P for time treatment0.336). None of the subjects
had hypercalcemia. There was a trend (P 0.06) for postintervention
serum PTH concentration to be affected by treatment,
and post hoc analysis showed a significantly (P 0.009) lower
mean concentration in the group receiving 15 g/d than in the
group receiving placebo (Table 2). However, the treatment
time interaction in repeated-measures ANOVA was not significant
(P 0.274) for the effect of vitamin D supplementation on
serum PTH concentrations.
Relation between vitamin D intake and vitamin D status
The relation between serum 25(OH)D concentrations in late
winter 2007 and the total vitamin D intake (diet and supplemental)
in the 20–40-y-old subjects is shown in Figure 2. The slope
of the relation between total vitamin D intake and serum
25(OH)D concentrations in the entire group was 1.96 nmol/Lg
intake. There was no significant difference between the slope
estimates for men and women (1.82 and 2.15 nmolL1 g1
intake, respectively; P 0.26).
Using mathematical modeling of the vitamin D intake–status
data, we estimated that the vitamin D intakes that maintained
serum 25(OH)D concentrations 25 nmol/L in 90%, 95%, and
97.5% of the 20–40-y-old adults were 2.7, 5.9, and 8.7 g/d,
respectively. An EAR [the vitaminDintake required to maintain
serum 25(OH)D concentrations 25 nmol/L in 50% of the
adults] could not be estimated because, at the lowest vitamin D
intake (0.1 g), the serum 25(OH)D concentrations in the 50th
percentile were 34.5 nmol/L. Data on sun preference also were
incorporated into the model; the vitamin D intakes that maintained
serum 25(OH)D concentrations of 25 nmol/L in 97.5%
of the sample were 7.2, 8.8, and 12.3g/d in those who reported
often having sunshine exposure, those who sometimes had sunshine
exposure, and sunshine avoiders, respectively. The vitamin
D intakes that maintained serum 25(OH)D concentrations above
2 other commonly suggested cutoffs in 97.5% of the sample were
26.1, 27.7, and 31.0 g/d (for 50 nmol/L) and 38.9, 40.6, and
43.9 g/d (for 80 nmol/L) in those who reported often having
TABLE 2
Habitual dietary intake, summer sunshine exposure preference, and biochemical measures of vitamin D status among treatment groups before and after
intervention1
Treatment group
P2
Placebo
(n 57)
5 g vitamin D/d
(n 48)
10 g vitamin D/d
(n 57)
15 g vitamin D/d
(n 53)
Habitual dietary vitamin D (g/d) 3.4 (2.0–5.0)3 4.3 (2.2–5.7) 3.5 (2.3–4.7) 3.6 (1.8–5.
0.856
Habitual dietary calcium (mg/d) 924 (694–1197) 905 (655–1314) 976 (681–1286) 1014 (744–1387) 0.600
Summer sun exposure preferences (%)
Sun avoider 12.5 12.5 8.9 17.0
Some sun exposure 50.0 50.0 46.4 49.1
Frequent sun exposure 37.5 37.5 44.6 34.0 0.885
Serum 25(OH)D (nmol/L)
Before intervention4 65.7 (58.4–94.1) 60.0 (50.0–89.7) 72.2 (55.7–91.9) 75.9 (55.9–89.3) 0.623
After intervention5,6 37.4 (31.4–47.9)a 49.7 (44.6–60.0)b 60.0 (51.0–69.1)c 69.0 (59.1–84.2)d 0.0001
Serum PTH (ng/mL)
Before intervention4 49.7 (32.9–62.1) 46.9 (34.0–70.3) 43.1 (35.6–57.9) 38.4 (29.0–50.3) 0.145
After intervention5,6 56.2 (41.3–67.
a 52.0 (35.9–67.9)a 50.5 (41.1–69.4)a 43.0 (33.1–62.0)b 0.060
1 PTH, parathyroid hormone; 25(OH)D, 25-hydroxyvitamin D. Values in a row with different superscript letters are significantly different, P 0.05.
2 One-factor ANOVA followed by Tukey’s test.
3 Median; interquartile range in parentheses (all such values), used in the case of nonnormally distributed variables.
4 All baseline blood samples were taken between October 2 and November 7, 2006.
5 Repeated-measures ANOVA was used to test the treatment time interaction; and the same trend was observed for serum 25(OH)D (P0.0001), but
the treatment time interaction was not significant for serum PTH (P 0.274).
6 All endpoint blood samples were taken between February 27 and April 7, 2007.
FIGURE 2. The relation between serum 25-hydroxyvitamin D
[25(OH)D] concentrations (in late winter 2007) and total vitamin D intake
(diet and supplemental) in 20–40-y-old healthy persons (n 215) living at
northerly latitudes (51 and 55 oN). Mean response and 95% CIs in the shaded
area.
DIETARY VITAMIN D REQUIREMENT IN HEALTHY ADULTS 1539
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
sunshine exposure, thosewhosometimes had sunshine exposure,
and sunshine avoiders, respectively.
Whereas a serum 25(OH)D concentration of25 nmol/L has
been used by several authorities as the traditional indicator of
adequacy for vitamin D (1, 5, 15, 16), several other biochemical
cutoffs for serum 25(OH)D, ranging from 37.5 to 80 nmol/L,
have been suggested (6 –9, 24). The 50th, 90th, 95th, and 97.5th
percentile estimates for vitamin D intake, obtained by using a
range of alternative indicators of adequacy for vitamin D status,
are shown in Table 3.
DISCUSSION
The RDA for nutrients is generally established as the average
daily intake level that is sufficient to meet the nutrient requirement
for nearly all (97–98%) persons in a life-stage and sex group
(1, 5). Uncertainty and gaps in the available data about the relative
contribution of sunshine and diet to vitamin D status and
vitamin D requirements for health maintenance have presented
international authorities with considerable difficulty in setting
dietary requirements for vitamin D. An approach that prioritizes
the identification of the intake values that will maintain serum
25(OH)D concentrations above chosen cutoffs when dermal production
of vitaminDis absent or markedly diminished is urgently
required. We examined the relation between vitamin D intake
and serum 25(OH)D concentrations in late winter, after a 4-dose
vitamin D intervention study that lasted 22 wk (from October
2006 to March 2007) in 221 healthy 20–40-y-old whites living
at 51 °N and 55 oN. We found that a daily intake of 8.7 g
vitamin D/d would have maintained serum 25(OH)D concentrations
25 nmol/L in 97.5% of the sample. Because the thresholds
for vitaminDadequacy are widely disputed, we also reported the
50th, 90th, 95th, and 97.5th percentiles of vitamin D intakes
required to maintain serum 25(OH)D concentrations in excess of
37.5, 50, and 80 nmol/L (6, 7).
These data could provide a basis for reconsideration of the
establishment of anRNIfor vitaminDby the authoritative bodies
responsible for devising nutrition policy. In the United Kingdom,
COMA concluded in 1998 that there was no evidence on which
to base a recommendation to establish an RNI (1). When it was
establishing the AI for vitamin D for persons aged 19–50 y (5),
the US DRI panel for calcium and related nutrients relied heavily
on data from a study by Kinyamu et al (25), which showed that
an average intake of 3.3–3.4 g vitamin D/d was sufficient to
keep serum 25(OH)D concentrations above 30 nmol/L during
winter months in most (94%) women 25–35 y old (n 52) in
Nebraska (latitude: 41 oN). The panel rounded down this intake
value to 2.5g/d and then doubled it to achieve an AI of 5.0g/d
(5). Working from a lack of data in men, the panel also made an
assumption that the AI for men would be similar to that for
women (5). The findings of the present study suggest that a
vitamin D intake of approximately twice the AI is required by
healthy white men and women at latitudes of50 oNto maintain
25(OH)D at these concentrations (30 nmol/L).
In setting the AI, theUSDRI panel also assumed that there was
no cutaneous synthesis of vitamin D through sun exposure (5).
This is true in winter, and, whereas summertime dermal synthesis
can be viewed as a supplement (“top-up”) to help generate wintertime
tissue stores of vitamin D, individual variation is likely to
be high, which makes summertime dermal synthesis an unreliable
contributor to vitamin D status. In the current study, sun
exposure preference was assessed as a surrogate for tissue stores;
as expected, whereas most people liked some (50%) or a lot
(38%) of sun,12% of subjects were sun avoiders. One might
expect this minority of subjects to have the highest dietary requirement
for vitamin D in winter as a consequence of their low
tissue stores. In fact, the vitamin D intake required to maintain
serum 25(OH)D concentrations in late winter above 25 nmol/L
was 12.3, 8.8, and 7.2 g/d for sunshine avoiders, those who get
some sunshine, and those who enjoy the sun, respectively. This
analysis, although perhaps limited by the relatively small number
of sunshine avoiders, serves to illustrate not only the greater
dietary requirement for vitamin D in persons who steer clear of
the sun but also the potential importance of high vitaminDstores
from sun exposure during summer in offsetting potentially deleterious
effects of low dietary intakes of vitaminDduring winter.
It is interesting that the UK COMA subgroup on bone health,
in their re-evaluation of dietary vitamin D requirement (1), took
an approach completely opposite to that of the US authority (5)
in terms of contribution of sun exposure to vitamin D requirement.
The COMA subgroup suggested that most of the adult
population in the United Kingdom can achieve adequate vitamin
Dstatus if the skin of the face and arms is exposed for30 min/d
between April and October. Some have argued, however, that
this degree of surface exposure may not be sufficient (26). Moreover,
the COMA subgroup concluded there had been no new
evidence to suggest that persons aged 4–64 y rely on dietary
intake for adequate vitamin D status. The data from the present
study clearly show that vitaminDtissue stores, developed during
summer via exposure of skin to sunshine, were not sufficient to
maintain serum 25(OH)D concentrations of25 nmol/L in most
of the population, and that dietary vitamin D is an absolute
requirement to maintain status above this minimum threshold.
TABLE 3
Estimated dietary requirements for vitamin D at selected percentiles in 215 men and women aged 20–40 y to maintain serum 25-hydroxyvitamin D
25(OH)D concentrations above selected biochemical cutoffs during winter1
Cutoff 50th percentile2 90th percentile 95th percentile 97.5th percentile
g/d
Serum 25(OH)D 25 nmol/L — 2.7 (0.0, 4.7) 5.9 (3.6, 8.0) 8.7 (6.5, 11.1)
Serum 25(OH)D 37.5 nmol/L 2.3 (0.0, 4.2) 13.8 (12.1, 15.9) 17.0 (14.8, 19.9) 19.9 (17.2, 23.5)
Serum 25(OH)D 50 nmol/L 10.2 (8.9, 11.4) 21.7 (19.3, 25.0) 25.0 (21.9, 29.1) 28.0 (24.2, 32.
Serum 25(OH)D 80 nmol/L 23.1 (21.0, 26.0) 34.8 (30.4, 40.6) 38.3 (33.0, 44.
41.1 (35.4, 48.7)
1 All values are estimate; 95% CI in parentheses. Results were based on a log-linear model of serum 25(OH)D as a function of vitamin D intake; the 95%
CIs were calculated by using a bias-corrected bootstrap based on 10 000 replications.
2 The vitamin D intake value that will maintain serum 25(OH)D concentrations in 50% of 20–40-y-old adults above the indicated cutoff during winter.
1540 CASHMAN ET AL
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
Survey data from the UK NDNS showed that up to 20% of UK
adults 19–34 y old (whose median vitamin D intake was 2.5
g/d) have plasma 25(OH)D concentrations of 25 nmol/L
(14), which underscores our findings. We acknowledge that cutaneous
vitamin D synthesis during summer months probably
offsets the dietary requirement for vitamin D that would ensure
adequacy during wintertime. However, it is worth noting that the
percentage of the population with unprotected sun exposure may
be rapidly declining, as a consequence of public education campaigns
in relation to skin cancer (27).
In the current analysis, we placed strong emphasis on using a
cutoff of 25 nmol/L for serum 25(OH)D on the basis that concentrations
of20–27.5 nmol/L are considered to be consistent
with vitamin D deficiency and rickets or osteomalacia (1, 28),
and the 25 nmol/L threshold has been in use to date by various
important authorities (1, 5, 15, 16). However, we also reported
dietary requirements for vitaminDin the current sample of white
20–40-y-old persons by using several other serum 25(OH)D
cutoffs (37.5, 50, and 80 nmol/L) (6, 7). The rationale for these
alternative definitions of adequacy for vitamin D in relation to
skeletal and nonskeletal health benefits has been detailed elsewhere
(8, 9). In an extended vitamin D supplementation study
(supplementation range: 0–250 g/d) in adult males (x age:
38.7 y) in Omaha, NE (latitude: 41.2 oN), Heaney et al (29) used
pharmacokinetic modeling to estimate the vitamin D intake required
to maintain prewinter serum 25(OH)D concentrations, to
reach concentrations of 80 nmol/L during winter, or both. They
reported a slope estimate of 0.70 nmol L1 g1 intake (29),
a figure that has been used widely to predict dietary requirements
for the US adult population (27, 30). Although derived by a
different means, the slope estimate in our study was 1.96 nmol/
Lg intake. It is not clear why there is a large variation between
these estimates, because both studies were in young adults and
both were conducted throughout winter. Despite similar concentrations
of 25(OH)D (70 nmol/L) at baseline (October), the
placebo group in the study by Heaney et al (29) experienced a
mean decline in serum 25(OH)D of only 11.4 nmol/L between
October and March, whereas the concentration in our placebo
group decreased by 28.3 nmol/L over the same period. The men
in the study by Heaney et al (29) may have had higher tissue
stores after a summer at 41 oN in the United States, whereas our
subjects presumably had less sun at latitudes of 51–54 oN in
Ireland. It is interesting that our slope estimate agrees well with
the estimates ranging from 1.6 –2.2 nmol L1 g1 intake
derived in several studies in older adults (31–34). Heaney et al
(29) suggested that tissue stores in the subjects in those studies
may have made a lower contribution to serum 25(OH)D concentrations
than did the tissue stores in the younger men in their own
study. Our estimate of the dietary vitamin D requirement needed
to maintain serum 25(OH)D concentrations above 80 nmol/L in
97.5% of our sample of 20–40-y-olds was 41 g/d, which is
considerably less than the 114 g/d suggested by Heaney et al
(29). Our data also show that, even for the lower cutoff of 50
nmol/L serum 25(OH)D, which may be associated with a lower
risk of a wide range of nonskeletal chronic diseases (8, 9), the
dietary requirement (28.0 g/d) is still much higher than the
amount currently being consumed by adult populations (14, 22,
35).Apotential limitation of the present study was that relatively
few subjects (17%) achieved winter serum 25(OH)D concentrations
of80 nmol/L, because of our use of a maximum of 15g
supplemental vitamin D/d. This fact may have influenced the
accuracy with which we can estimate the dietary requirement to
achieve such high serum 25(OH)D concentrations. To absolutely
confirm that our recommended intakes can achieve 25(OH)D
concentrations in the range of 50 to 80 nmol/L, a wintertime
intervention study using higher doses of vitamin D (at least
20–40 g/d) would be required.
In conclusion, to ensure that the needs of 97.5% of 20–40-
y-old persons are met in relation to vitamin D status during
winter, 8.7 g vitamin D/d is required to maintain serum
25(OH)D concentrations above the most conservative threshold
of adequacy (ie, 25 nmol/L).
The authors’ responsibilities were as follows: MK, JMWW, AF, MPB,
EMD,JJS, andKDC:the conception of work and are grant holders; TRH, NT,
AJL, KMS, GH, MSB, JMWW, MK, and KDC: the execution of the study;
SM, NT, TRH, GH, AJL, and MB: sample analysis; and all authors: data
analysis and writing of the manuscript. None of the authors had a personal or
financial conflict of interest.
REFERENCES
1. UK Department of Health. Nutrition and bone health: with particular
reference to calcium and vitamin D. Report on Health and Social Subjects
(49). London, United Kingdom: The Stationary Office, 1998.
2. Lips P. Vitamin D deficiency and secondary hyperparathyroidism in the
elderly: consequences for bone loss and fractures and therapeutic implications.
Endocr Rev 2001;22:477–501.
3. Parfitt AM, Gallagher JC, Heaney RP, Johnston CC, Neer R, Whedon
GD. Vitamin D and bone health in the elderly. Am J Clin Nutr 1982;36:
1014–31.
4. Ooms ME, Lips P, Roos JC, et al. Vitamin D status and sex hormone
binding globulin: determinants of bone turnover and bone mineral density
in elderly women. J Bone Miner Res 1995;10:1177– 84.
5. Institute of Medicine Food and Nutrition Board. Dietary reference intakes:
calcium, magnesium, phosphorus, vitamin D, and fluoride. Washington,
DC: National Academy Press, 1997.
6. Lips P. Which circulating level of 25-hydroxyvitamin D is appropriate?
J Steroid Biochem Mol Biol 2004;89 –90:611– 4.
7. Chapuy MC, Preziosi P, Maamer M, Arnaud S, Galan P, Hercberg S,
Meunier PJ. Prevalence of vitamin D insufficiency in an adult normal
population. Osteoporos Int 1997;7:439–43.
8. Zittermann A. Vitamin D in preventive medicine: are we ignoring the
evidence? Br J Nutr 2003;89:552–72.
9. Holick MF. Sunlight and vitamin D for bone health and prevention of
autoimmune diseases, cancers and cardiovascular disease. Am J Clin
Nutr 2004;80(suppl):1678S– 88S.
10. McKenna MJ. Differences in vitamin D status between countries in
young adults and the elderly. Am J Med 1992;93:69 –77.
11. van der Wielen RP, Lowik MR, van den Berg H, et al. Serum vitamin D
concentrations among elderly people in Europe. Lancet 1995;
346(8969):207–10.
12. Lips P, Duong T, Oleksik A, et al.Aglobal study of vitaminDstatus and
parathyroid function in postmenopausalwomenwith osteoporosis: baseline
data from the multiple outcomes of raloxifene evaluation clinical
trial. J Clin Endocrinol Metab 2001;86:1212–21.
13. Lips P, Hosking D, Lippuner K, et al. The prevalence of vitamin D
inadequacy amongst women with osteoporosis: an international epidemiological
investigation. J Intern Med 2006;260:245–54.
14. Henderson L, Irving K, Gregory J, et al. The National Diet and Nutrition
Survey: adults aged 19 to 64 years—vitamin and mineral intake and
urinary analytes. London, United Kingdom: The Stationery Office,
2003.
15. UK Department of Health. Dietary reference values for food energy and
nutrients for the United Kingdom. Report on Health and Social Subjects
(41). London, United Kingdom: Her Majesty’s Stationery Office, 1991.
16. Commission of the European Communities. Vitamin D: In nutrient and
energy intakes of the European Community. Report of the Scientific
Committee for Food (31st series). Brussels, Luxembourg, 1993:132–9.
17. S. Finch, Doyle W, Lowe C, et al. National Diet and Nutrition Survey:
people aged 65 years and over. Volume 1: report of the diet and nutrition
survey. London, United Kingdom: The Stationery Office, 1998.
DIETARY VITAMIN D REQUIREMENT IN HEALTHY ADULTS 1541
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
18. Andersen R, Molgaard C, Skovgaard LT, et al. Teenage girls and elderly
women living in northern Europe have low winter vitamin D status. Eur
J Clin Nutr 2005;59:533– 41.
19. Webb AR, Kline L, Holick MF. Influence of season and latitude on the
cutaneous synthesis of vitamin D3: exposure to winter sunlight in Boston
and Edmonton will not promote vitamin D3 synthesis in human skin.
J Clin Endocrinol Metab 1988;67:373–78.
20. Lucey AJ, Paschos GK, Cashman KD, Martínéz JA, Thorsdottir I, Kiely
M. Influence of moderate energy restriction and seafood consumption on
bone turnover in overweight young adults. Am J Clin Nutr 2008;87:
1045–52.
21. Collins A. Development and validation of methods for the measurement
of micronutrient intakes relevant to bone health. Doctoral dissertation.
The National University of Ireland, Cork, 2006.
22. Hill T, Collins A, O’Brien M, Kiely M, Flynn A, Cashman KD. Vitamin
D intake and status in Irish postmenopausal women. Eur J Clin Nutr
2005;59:404 –10.
23. Cashman KD, Hill TR, Cotter AA, et al. Low vitaminDstatus adversely
affects bone health parameters in adolescents. Am J Clin Nutr 2008;87:
1039–44.
24. Bischoff-Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson-
Hughes B. Estimation of optimal serum concentrations of 25-
hydroxyvitamin D for multiple health outcomes. Am J Clin Nutr 2006;
84:18 –28.
25. Kinyamu HK, Gallagher JC, Balhorn KE, Petranick KM, Rafferty KA.
Serum vitamin D metabolites and calcium absorption in normal young
and elderly free-living women and in women living in nursing homes.
Am J Clin Nutr 1997;65:790 –7.
26. Matsuoka LY, Wortsman J, Hollis BW. Use of topical sunscreen for the
evaluation of regional synthesis of vitamin D3. J Am Acad Dermatol
1990;22:772–5.
27. Hollis BW. Circulating 25-hydroxyvitamin D levels indicative of vitamin
D sufficiency: implications for establishing a new effective dietary
intake recommendation for vitamin D. J Nutr 2005;135:317–22.
28. Specker BL, Ho ML, Oestreich A, et al. Prospective study of vitamin D
supplementation and rickets in China. J Pediatr 1992;120:733–9.
29. Heaney RP, Davies KM, Chen TC, Holick MF, Barger-Lux MJ. Human
serum 25-hydroxycholecalciferol response to extended oral dosing with
cholecalciferol. Am J Clin Nutr 2003;77:204 –10.
30. Weaver CM, Fleet JC. VitaminDrequirements: current and future.AmJ
Clin Nutr 2004;80(suppl):1735S–9S.
31. Byrne PM, Freaney R, McKenna MJ. Vitamin D supplementation in the
elderly: review of safety and effectiveness of different regimes. Calcif
Tissue Int 1995;56:518 –20.
32. Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 and calcium to
prevent hip fractures in elderly women. N Engl J Med 1992;327:1637–
42.
33. Krall EA, Sahyoun N, Tannenbaum S, Dallal GE, Dawson-Hughes B.
Effect of vitamin D input on seasonal variations in parathyroid hormone
secretion in postmenopausal women. N Engl J Med 1989;321:1777– 83.
34. Kyriakidou-Himonas M, Aloia JF, Yeh JK. Vitamin D supplementation
in postmenopausal black women. J Clin Endocrinol Metab 1999;84:
3988–90.
35. Hill TR, O’Brien MM, Cashman KD, Flynn A, Kiely M. Vitamin D
intakes in 18-64-y-old Irish adults. Eur J Clin Nutr 2004;58:1509 –17.
1542 CASHMAN ET AL
Downloaded from
www.ajcn.org at SCD Université Paris 5 on December 19, 2008
