Clinical Science (1982) 63,461472
46 1
Vitamin D from skin: contribution to vitamin D status
compared with oral vitamin D in normal and anticonvulsanttreated subjects
M. W . J . D A V I E * , D . E . M . L A W S O N * , C . E M B E R S O N ? ,
J. L . C . B A R N E S ? , G. E . R O B E R T S ? A N D N . D . B A R N E S $
'Dunn Nutritional Laboratory, Milton Road, Cambridge, t Ida Darwin Hospital, Fulbourne, Cambridge, and
wddenbrooke's Hospital, Hills Road, Cambridge, U.K.
(Received 24 August 1981123 March 1982; accepted I0 June 1982)
Summary
1. The plasma 25-hydroxycholecalciferol[25(OH)D,l response to measured U.V. irradiation
applied thrice weekly for 10 weeks was investigated in normal and in anticonvulsant-treated
subjects.
2. Levels of plasma 25-(OH)D, achieved after
U.V. irradiation were similar in both normal and
anticonvulsant-treated subjects, suggesting that
hepatic microsomal enzyme induction does not
lead to low plasma 25-(OH)D, concentrations.
3. Cholecalciferol was present in plasma of
normal subjects in a very low concentrations
(< 5-0 nmol/l) and did not increase until plasma
25-(OH)D, levels exceeded 62.5 nmolll.
4. Cholecalciferol occurred in significant
concentrations in plasma during whole body U.V.
irradiation or during oral dosage of 62.5 nmol
(1000 i.u) or more daily.
5 . Plasma 25-(OH)D, concentrations reached
a steady state after 5-6 weeks of U.V. irradiation
or of oral intake within the usual intake range.
6. Cholecalciferol synthesis in skin calculated
from the steady-state equation was 0.0015 k
04008 nmol1mJ.
7. Cholecalciferol synthesis in skin was also
calculated from the oral dosage required to yield
the same plasma 25-(OH)D, concentration as
U.V. irradiation and was 0.0024 & 0.0018
nmol/mJ.
8. Rates of cholecalciferol synthesis calculated
from these data suggest that many of the
Correspondence: Dr M. W. J. Davie, Metabolic
Unit, St Mary's Hospital, Praed Street, London W.2.
0143-5221/82/110461-12501.50
population of England receive insufficient U.V.
irradiation to maintain vitamin D status
throughout the year.
Key words: anticonvulsant drugs, cholecalciferol,
skin, ultraviolet irradiation, vitamin D.
Abbreviations: 25-(OH)D,, 25-hydroxycholecalciferol; 25-(OH)D,, 25-hydroxyergocalciferol; D,, cholecalciferol.
Introduction
Cutaneous synthesis of cholecalciferol in response to U.V. irradiation is considered to be
important in maintaining vitamin D status in man
[l, 21. Little is known about the influence of
environmental U.V. energy, or of the effect of age,
anatomical site or drugs associated with derangement of vitamin D metabolism on the
capacity of human skin to synthesize cholecalciferol in vivo. It is important to consider the
skin response in vivo since isolated skin may
respond differently [31. Acquisition of a skin
biopsy after U.V. irradiation may be unacceptable
to the patient, especially if it is taken from an area
such as the face, and the information must
therefore be obtained indirectly by measuring the
main circulating metabolite of cholecalciferol,
25-(OH)D3 [4,51.
Although the importance of sunlight in maintaining vitamin D status is recognized, the
relative contribution arising from diet or U.V.
energy is controversial [6,71, since many subjects
have little exposure to U.V. irradiation [81 or do
0 1982 The Biochemical Society and the Medical Research Society
462
M . W.J. Davie et al.
not go out of doors [91. It would thus be of value
to establish the amount of cholecalciferol that
may be synthesized in skin by a known quantity
of U.V. energy.
We have compared the pattern and magnitude
of the plasma 25-(OH)D, response both to
standardized U.V. irradiation and to oral dosage
with cholecalciferol. Cutaneous cholecalciferol
synthesis has been calculated as mass of cholecalciferol per millijoule of U.V. energy. The
influence of anticonvulsant drugs on the plasma
25-(OH)D, response to U.V. irradiation has also
been examined.
poses and from two male subjects taking ergocalciferol (1 -25 mglday) for iatrogenic hypoparathyroidism. Supplementary vitamin D or U.V.
irradiation had been administered for at least 9
weeks in all subjects receiving these treatments.
Details of the subjects receiving U.V. energy
and cholecalciferol and of the groups into which
they were divided appear in Table 1. Subjects in
groups 2-5 had been established on anticonvulsant drugs before the experimental period
and continued to receive these drugs during the
study.
U.V.irradiation
Subjects and methods
Subjects
Subjects investigated were white skinned and
had no clinical or biochemical features of
osteomalacia (Table 1). Subjects in groups 1-5,
who were selected from patients in the Ida
Darwin Hospital, Cambridge, rarely went outside. Since there was a considerable variation in
weight amongst the subjects, care was taken to
match the two groups receiving U.V. irradiation
(groups 1 and 2) for weight. The subjects in
group 6, who were older than the subjects in
groups 1-5, were selected from patients in the
Hinchingbrooke Hospital, Huntingdon, and over
the period of observation they did not go outside.
Subjects in whom the relationship between
plasma concentrations of 25-(OH)D, and cholecalciferol were investigated came from the above
groups, from volunteer subjects receiving no
therapy, from subjects receiving whole body U.V.
irradiation for experimental or therapeutic pur-
U.V. irradiation was administered to 600 cmz
dorsal skin for 3 midday for 3 dayslweek over a
10 week period in the subjects comprising groups
1 and 2 beginning 30 January 1978. Subjects in
group 6 received U.V. irradiation to 900 cmz
dorsal skin for 3 midday on 3 dayslweek over 5
weeks in FebruaryIMarch 1979. In all groups the
same area of skin was exposed to U.V. irradiation
on each occasion.
The U.V. energy source was a Hanovia I A
prescription lamp situated 50 cm from the skin
surface in groups 1 and 2, and at various
distances in group 6. The energy received by each
subject in group 6 was known accurately since
the power emitted was found to behave in a
manner predicted by the inverse square law of
irradiation (Dr J. Haybittle, Medical Physics
Department, Addenbrooke's Hospital, Hills Road,
Cambridge). The power spectrum of this lamp
has previously been described [lo]. Since the
energy required either to heal rickets or to
convert 7-dehydrocholesterol into cholecalciferol
TABLE1. Details of subjects receiving U.V.irradiation or oral cholecalciferol
Group
Age
(years)
No.
-
Weight
(kg)
Treatment
Vitamin
D, dosage
M F
U.V.energy
(mJ day-' cm-')
(nrnol/day)
~~
I
2
21.8 f 4.6 6
20.6 f 2.8 6
2 33.5 f 9.4
2 34.1 f 13.6
Skin area
Pheno- Phenytoin Initial plasma
irradiated barbitone
25-(OH)D,
(cm')
(mg day-' kg-')
(nmol/l)
~
U.V.
U.V.
-
18.9
18.9
~~~
600
600
3
21.2 f 7.6 7
2 50.2 f 17.6
Oral
25
-
4
21.4 f 8.9 7
2 45.9 f 14.9
Oral
62.5
-
5
22.5 f 9.1 6
2 45.6 f 18.4
Oral
6
77.2
7 55.5 f 12.1
U.V.
9.4 2
625
-
-
16.7 (n = 3)
8.4 (n = 3)
3.3 (n = 3)
900
Nil
Nil
2.9 f 1.5 3.3 k 0.6
(n = 5)
( n = 3)
( + 2 on primidone)
1.4 k 1.5 8.3 f 4.4
(n = 5)
(n = 4 )
(+ 1 on primidone)
2.7 f 1.1 3.6 f 2.5
(n = 3)
(n = 9)
1.4 f 0.7 6.5 f 5.8
( n = 7)
(n = 3)
(+ 1 on primidone)
Nil
Nil
~~
14.7 f 9.0
5 . 5 k 6.5
( n = 7)
16.5 f 4.8
13.3 f 7.5
13.2 f 3.5
10.6 f 6 . 8
Cutaneous vitamin D synthesis
is different for each wavelength in the spectrum of
this lamp, a factor was calculated from previously published data [l l-131 to express the
power of each wavelength as a proportion of the
most efficient wavelength. The total power
emitted from the Hanovia lamp, expressed in
terms of the most efficient wavelength, is 140.2
pW/cmZ calculated from the data of Knudsen &
Benford [121, 245.3 pW/cmZ calculated from the
figures of Kobayashi & Yasumura [ l l l and
168.7 pW/cm2 from activity spectrum of Bunker
& Harris [ 131. The total effective energy over 60
s was found by multiplying the power by 60 and
dividing by 1000 to obtain the result in mJ/cm2:
in turn this yielded 8.4, 14.7 and 10.1 mJ/cm2
for the power emissions stated above. The
procedure has been described in detail previously
[lo]. The energy (14.7 mJ/cm2) calculated from
the conversion factor (in vitro) of Kobayashi &
Yasumura [ 111 has been used in the calculation
of vitamin D synthesis. Thus each subject in
groups 1 and 2 irradiated for 9 midweek
received an average of 1.29 midday or 18.9
mJ/cm2 over 1 day. The energy received each
day appears in Table 1, corrected by use of the
conversion factor (in vitro).
Oral cholecalciferol intake
The supplementary cholecalciferol used in
groups 3-5 was made up in arachis oil and
administered in a once-daily dosage.
Vitamin D intake from the diet was assessed
by reference to food tables I 5 1.
Cholecalciferol and 25-(0H)D, measurements
Plasma cholecalciferol and 25-(OH)D, were
measured
by
high
pressure
liquid
chromatography [ 141.
25-(OH)D,kinetic studies
Tracer studies of 25-(OH)D, kinetics were
performed by using [26,27-,H125-(OH)D3 (5-15
Ci/mmol)
(The
Radiochemical
Centre,
Amersham, Bucks, U.K.), made up in ethanol.
Immediately before intravenous injections, 5 pCi
was added to 0.9% sodium chloride solution to
make a 10% solution. The radioactive 25(OH)D, in ethanol was rendered bacteria free by
filtration through a Millipore filter. Plasma
samples were taken over the 3 week period after
the injection, and ,H-labelled 25-(OH)D, was
separated by thin-layer chromatography [ 161.
The decline of radioactivity in plasma over 2-21
days was linear when plotted on a logarithmic
463
scale. The volume of distribution ( V d ) and half
life (T,,,) were calculated by standard techniques
[17, 181. These studies were undertaken in 11
subjects (eight females, three males) who had a
mean weight of 54.3 f 0.85 kg and a mean age
of 28.0 f 8.4 years. Six subjects were receiving
phenobarbitone (1.5 f 0.85 mg/kg) and seven
were receiving phenytoin (3.6 f 1.8 mg/kg).
Three subjects were receiving both phenobarbitone and phenytoin and one subject was
taking primidone.
Calculation of cholecalciferol synthesis
Effective cholecalciferol synthesis in skin was
calculated from the plasma 25-(OH)D, response
to U.V. irradiation. Use of this measurement may
result in a slight underestimate of cholecalciferol
synthesis in skin if some cholecalciferol is not
converted into 25-(OH)D,. However, at the
concentrations of 25-(OH)D, achieved during
U.V. irradiation, there is rapid conversion of tracer
dosages of cholecalciferol into 25-(OH)D, [4,
261, and there is no accumulation of cholecalciferol in plasma (see the Results section). Two
methods were used in the calculation of cholecalciferol synthesis. One made use of the steadystate equation [ 181 and in the second method the
oral dosage of cholecalciferol required to achieve
the same plasma 25-(OH)D, concentration as
occurred during U.V. irradiation was calculated.
The steady-state equation [17, 181 used in the
first method describes the relationship between
dosage of a drug and the plasma concentration
achieved at plateau level thus:
C, = (1.44 x F x T,,, x D)/Vd
in which C, = plasma concentration (nmol/l), F
= fractional absorption (assumed to be unity),
T,,, = half life determined by PHI25-(OH)D3
(days), Vd = volume of distribution (litres) and D
= dosage per unit time (nmol/day). The value for
the radioactive half life was used because it is less
influenced by exogenous sources of vitamin D
than is the actual half life, and it is unrelated to
the plasma 25-(OH)D, concentration [ 191. The
half life determined by the tracer method is also
closer to the half life of plasma 25-(OH)D, after
very high dosage of vitamin D, when any
extraneous vitamin D will make less impact on
the rate of decline of plasma 25-(OH)D, from
high levels [20]. Moreover the time required for a
drug to achieve a concentration of 90% of the
plateau level is equivalent to three to four half
lives [ 17, 181. The radioactive tracer 25-(OH)D,
half life meets this criterion more closely than
does the actual half life, which is longer [ 191. The
M.W.J. Davie et al.
464
volume of distribution was calculated from
a single dose of radiolabelled 25-(OH)D,.
Although the volume of distribution calculated
from the steady-state method is not identical with
that obtained by the single dose injection, the
difference between the two is very small [171. The
-
Cholecalciferol synthesis in skin may be
expressed as cholecalciferol synthesis (nmol/mJ)
x energy (mJ/cmz) x area irradiated (cmz). The
cholecalciferol intake being the same these two
terms will be equal, and can be rearranged such
that daily cholecalciferol synthesis (nmol/mJ)
cholecalciferol intake (nmol/kg) x body weight (kg)
energy (mJ/cmz) x area irradiated (cmz)
concentration in the compartment of volume V,
was assumed to equate to that of the plasma
compartment since the steady-state period (30
days) was long compared with the time required
for equilibration of the L3H125-(OH)D, (2 days).
For the purpose of the steady-state equation V,
was calculated from the term body weight (kg) x
21.6/100. This term represents the fraction of
body weight represented by Vd as calculated from
a single injection of [,H125-(OH)D3 (see the
Results section). Thus with values for C,, V,, F
and TI,, known, the steady-state equation may be
expressed in terms of D, this being the daily input
of cholecalciferol from skin with units of nmol/
day. Since the total cholecalciferol synthesis per
day is derived from a known area of skin, each
square centimetre contributes an amount of
cholecalciferol equal to total cholecalciferol per
day/area of exposure. The U.V. energy applied to
each square centimetre of skin during the plateau
period of plasma 25-(OH)D, being known (see
above), cholecalciferol synthesis per mJ of U.V.
energy is equal to
Total cholecalciferol per day/area of exposure
daily U.V. energy/cm2
(1)
Cholecalciferol synthesis in skin was also
estimated from the amount of oral cholecalciferol required to sustain a plasma 25(OH)D, level equivalent to that obtained during
U.V. irradiation treatment. Since plasma 25(OH)D, correlates with cholecalciferol intake per
kg body weight (see the Results section), a value
of cholecalciferol intake/kg could be found for
any plasma 25-(OH)D, level. Assuming that the
metabolism of cholecalciferol does not differ
according to the origin of cholecalciferol in the
body, and ignoring absorptive losses, any concentration of plasma 25-(OH)D, will reflect the same
intake of cholecalciferol, be it from the diet or
from skin. Thus the relationship of cholecalciferol dosage to plasma 25-(OH)D, concentration
allows the total amount of cholecalciferol required to sustain a given 25-(OH)D, level to be
calculated from cholecalciferol intake (nmol/kg)
x body weight (kg).
(2)
Statistics
Regression analysis was performed on a Texas
TI-5 1-1 1 1 calculator. Other statistical analyses
were performed as described in standard texts,
and expressed as mean f S.D.
Ethical approval was obtained for all these
experiments from the Ethical committee, Addenbrooke's Hospital, Cambridge. Permission was
obtained from all subjects or their guardians.
Results
Details of the subjects receiving U.V. irradiation or oral cholecalciferol appear in Table 1.
The initial plasma 25-(OH)D, concentration was
5.5- 16-5 nmol/l, the level being significantly
lower in group 2 subjects compared with those in
group 1 ( P < 0.05;Mann-Whitney U-test). The
intake of cholecalciferol in subjects in groups
1-6 was 2.6 f 0.5 nmol/day with no significant
difference between the groups.
Response to U.U. irradiation
Plasma 25-(OH)D, rose in all subjects receiving U.V. irradiation. In groups 1 and 2, plasma
25-(OH)D, levels attained a maximum level after
30-40 days: thereafter the levels remained steady
until irradiation was discontinued on day 70 (Fig.
la). In the subjects comprising group 6, plasma
25-(OH)D, levels at the end of irradiation were
28-0 f 14.4 nmol/l in those receiving 16.7 mJ
day-' cm-2, 17.9 f 6.3 nmol/l in those exposed
to 8.4 mJ day-' cm-z and 12.9 f 1.0 in those
receiving the least U.V. energy. In all the subjects
receiving U.V. irradiation the plasma 25-(OH)D,
concentration after 5 weeks was related to the
daily U.V. energy applied by a power function
(Table 2; Fig. 2). The incremental plasma
25-(OH)D, was linearly related to the daily U.V.
energy applied (Table 2; Fig. 3), the relationship
being significant for all subjects. One subject,
however, had a disproportionate effect on the
slope of the regression and thus the slope relating
subjects receiving daily U.V. energy up to 500 mJ
day-' kg-I is also shown (Fig. 3).
465
Cutaneous vitamin D synthesis
150
n
z
h
100
0
N
VI
2
-na.
50
0
35
70
0
Time (days)
35
70
Time (days)
FIG. 1. Plasma 25-(OH)D, response (mean k SD)by subjects in groups 1 and 2 to U.V. irradiation
(n), and by subjects in groups 3-5 to oral cholecalciferol (b). (a) M, Normal subjects; A---A,
anticonvulsant-treated subjects. (b) 0 , 6 2 5 nmol of cholecalciferol/day; D--U, 62.5 nmol/day ;
.A,25 nmol/day.
+
100
-
A
A
=
:
-B
d
A
h
X
0
v
A
N
a
-e:
0
0
10 J
I
I
I
I
M . W.J. Davie et al.
466
0
600
300
900
1200
U.V.energy (mJ day-I kg-l)
FIG. 3. Relationship between the increase of plasma 25-(OH)D, after 32-35 days of U.V.
irradiation and daily U.V. energy for subjects in group 1 (0,
group 2 (W) and group 6 (A). The
dotted line refers to all subjects, and the continuous line to subjects receiving less than 500 mJ
day-' kg-' (see the text). The equation of the continuous line is (vi) in Table 2). U.V. energy is
derived as in Fig. 2.
TABLE2. Relationships between oral vitamin D intake or U . V . irradiation and plasma concentrations of 25-(0H)D, or
25-(OH)D3+ D ,
Oral cholecalciferol
(i) log plasma 25-(OH)D, (nmol/l) = log 56.7 + 0.31 log D, intake.
(ii) log plasma (25-(OH)D, + D,) (nmol/l) = log 65.3 + 0.5 log D, intake'
(iii) log D, intake' =log 0.0003 + 2.1 log plasma 25-(OH)D, (nmol/l)
(iv) log D intake' = log 0.0012 + 1.62 log plasma (25-(OH)D, + D,) (nmol/l)
U.V.irradiation
( v ) log plasma 25-(OH)D, (nmol/l) = log 2.0 + 0.49 log daily U.V. irradiation?
(vi) Increase plasma 25-(OH)D, (nmol/l) = 1.21 + 0.08 daily U.V. irradiation
n = 26; r = 0.79; P
n = 25; r = 0.90; P
< 0.001
n = 25; r = 0.67; P
< 0.001
< 0.001
n = 2 6 ; r = 0 . 7 9 ; P <0.001$
n = 25; r = 0.90; P < 0.001
n = 23; r ='0.71; P
< O.OOl$
Oral intake of cholecalciferol = nmol of cholecalciferol day-' kg-'.
(of ml day-' kg-I).
$ Relationship used to determine oral equivalent of cutaneously derived cholecalciferol.
$ Regression equation omits two subjects; one initial value was not available and one subject had an exceptionally high degree of
exposure (see the text). This equation describes the relationshipup to 500 mJ day-' kg-l.
t U.V.irradiation = mJ/cm'. cm' day-I kg-'
Plasma 25-(OH)D, levels were followed for 6
weeks after cessation of U.V. irradiation in groups
1 and 2. At the end of the period of irradiation
plasma 25-(OH)D, levels were 44.0 f 15.5
nmol/l and 44.3 k 12.3 nmol/l in groups 1 and 2
respectively. Six weeks later the levels were 33.8
f 16.0 nmol/l and 33.5 k 15.3 nmol/l in the two
groups.
Oral cholecalciferol dosage
Plasma 25-(OH)D, levels also reached a
plateau during oral administration of cholecalci-
ferol (Fig. lb). Subjects receiving 62.5 nmol/day
had similar plateau levels of 25-(OH)D, as
subjects receiving 25 nmol/day. Oral cholecalciferol dosage (as nmol of cholecalciferol/kg
body weight) correlated with plasma 25-(OH)D,
concentration (Table 2). Cholecalciferol dosage
was also related to total plasma [25-(OH)D, +
D,] if this term were used for groups 4 and 5, and
25-(OH)D, only in group 3, since there was very
little cholecalciferol present in the plasma of
group 3 subjects. A power function also best
described this relationship (Table 2; Fig. 4), the
slope of which was similar to that relating U.V.
energy and plasma 25-(OH)D, concentration.
Cutaneous vitamin D synthesis
20
1
467
0
I
0 3
I
1
1
1
I
J
3
10
30
100
Cholecalciferol intake (nmol day-' kg-')
FIG. 4. Relationship between plateau plasma 25-(OH)D, + D, concentrations and cholecalciferol intake for subjects in group 3 (Q, group 4 (.) and group 5 (A). The equation of the line
is (ii) in Table 2.
Plasma cholecalciferol
Cholecalciferol was present in plasma at a
concentration below 5 nmol/l in subjects whose
plasma 25-(OH)D, levels were less than 62.5
nmol/l (Fig. 5). The plasma concentration of
cholecalciferol exceeded that of 25-(OH)D, in six
subjects (five females, one male), three of whom
were receiving U.V. irradiation and three oral
cholecalciferol. Cholecalciferol was present at a
concentration below 5 nmol/l in subjects from
groups 1-3, and was not measured in the subjects
in group 6. Plasma cholecalciferol reached a
plateau during oral administration of 62.5 nmol of
cholecalciferol/day, the mean level being 14.5 f
7.5 nmol/l. In subjects receiving 625 nmol daily
(group 5), high levels of cholecalciferol were
found in plasma (Table 3), but, unlike the levels in
group 4, plasma cholecalciferol rose rapidly in
group 5 subjects and in four (nos. 2, 4, 5, 6;
Table 3 ) levels gradually fell even though dosage
was continuing. Plasma cholecalciferol fell
rapidly when
discontinued.
oral
supplementation
was
Radioactive 2.5-(OH)D3studies
The disappearance of L3H125-(OH)D, fitted a
two-compartment system. The slow half life,
measured from 48 to 504 h (2-21 days) was 12.9
f 3.6 days, and the total volume of distribution
was 11.45 & 3.70 litres or 21.6 & 5.5% of body'
weight. The term (21.6/100) x body weight (kg)
was used to calculate the value for Vd in the
steady-state equation.
Cholecalciferol synthesis in skin
Effective cholecalciferol synthesis in skin was
calculated for subjects in group 2 (Table 4). The
energy available from the Hanovia lamp was
calculated as described and the value of 14.7
mJ/cm* derived from the activity spectrum in
vitro (see the Methods section) was used to
M . W.J. Davie et al.
468
I
A
'
A
A
.
A
A
A
..
62 5
-
a
. .
.
.
,.d.
25-Hydroxycalciferol (nmol/l)
FIG. 5. Relationship between plasma cholecalciferol and 25-(OH)D, levels, in subjects not
receiving U.V. irradiation or cholecalciferol supplements (A), subjects receiving U.V. irradiation
from an artificial U.V. source (A), subjects receiving supplementary cholecalciferol (.) and
subjects receiving supplementary ergocalciferol (0).The inset is an enlargement of the area
between 0 and 62.5 nmol/l on the ordinate and 0 and 125 nmol/l on the abscissa.
TABLE3. Plasma cholecalciferol levels in group 5 subjects during and a f e r dosage with 625 nmol
of cholecalciferol daily
-, Measurement was not done.
Subject
no.
Plasma cholecalciferol (nmol/l)
0
4.5
0.0
0.0
4.3
-
0.0
0.0
0.0
Day 1
Day 14
Day 21
Day 49
Day 70
-
145.5
93.8
63.8
107.3
107.3
109.8
135.0
121.5
-
-
100.5
28.3
0.0
47.5
40.5
0.0
35.8
49.8
293.3
46.5
232.5
-
11.5
65.5
106.0
188.0
-
250.3
-
calculate the results depicted in Table 4. Use of
the activity spectrum of Bunker & Harris [131
gives a result 45% higher and the spectrum of
Knudsen & Benford [12l gives a value 75%
higher.
Discussion
The pattern of plasma 25-(OH)D, response is
similar irrespective of the source of cholecalci-
14 days after dosage ceased
-
-
92.0
66.8
53.0
151.0
58.8
-
68.0
178.8
288.5
31.8
13.5
51.3
-
ferol, and plateau levels are achieved after
equivalent periods with low dosage of oral
cholecalciferol or of U.V. energy. Thus the initial
exposures of U.V. irradiation lead to a rapid
increase of plasma 25-(OH)D, from a low
starting level (101. By following the plasma
25-(OH)D, response to U.V. irradiation over a 10
week period, it was evident that a plateau level
was eventually reached.
Unmetabolized cholecalciferol appeared in
Cutaneous vitamin D synthesis
469
TABLE4. Cholecalciferol synthesis in skin
Subjects were from group 2 and received 18.9 mJ/cm2 over 600 cm2 daily. Calculations do not
allow for dietary vitamin D as this contributed less than 10%to the total plateau level (see the text).
Subject
no.
1
2
3
4
5
6
I
8
Plateau plasma
Cholecalciferolsynthesis (nmol/mJ)
25-(OH)D,
Calculated from plateau level
(nmol/l)
Equivalent to oral intake
(4
(b)
51.1
39.0
45.6
43.6
32.8
31.8
36.5
53.3
0.0048
0.0021
0.0020
0.0018
0,0024
0,0016
0.00 I3
0.0012
04015
0~0010
0.0005
0.0055
04014
(b)as % of ((I)
50.0
-
-
16.2
65.0
66.1
93.3
90.0
82.0
56.4
-
0.0024k 0.0018
0.00154 ? 0.0008
72.5 ? 15.6
plasma during oral dosage of 62-5 nmol of
cholecalciferol or more daily or during whole
body U.V. irradiation. Although subjects in group
4 had a plasma 25-(OH)D, concentration similar
to those in group 3, the total response of
25-(OH)D, + D, exceeded that of group 3
subjects. Plasma levels of cholecalciferol were
higher than 25-(OH)D, on several occasions in
subjects receiving 625 nmol of cholecalciferol
daily, such high dosages probably leading to
inhibition of cholecalciferol conversion into 25(OH)D, [211. In the course of dosing with large
quantities of vitamin D, changes in the metabolism of vitamin D may take place, leading to
lower ievels of plasma vitamin D activity than
might otherwise be expected. Not only did
plasma cholecalciferol levels start to fall even
during the course of cholecalciferol administration (Table 3), but levels fell very rapidly after
supplementation was stopped. Moreover in subjects receiving 25 nmol of cholecalciferol daily
(group 3), the concentration of 25-(OH)D,
predicted by the steady state equation assuming
75% absorption [27] was 36 nmol/l compared
with the value of 47.0 nmol/l calculated from the
relationship of plasma 25-(OH)D, + D, and oral
cholecalciferol dosage (ii; Table 2). In contrast
the predicted levels of vitamin D activity in
groups 4 and 5, derived from the steady-state
equation, are 90.3 and 909.0 nmol/l respectively;
these values compare with the levels of 70.0 and
215.0 nmol/l calculated from the actual regression (ii; Table 2). At high dosages of cholecalciferol, vitamin D activity may accumulate in
tissues 122, 231, resulting in a different value for
Vd. Moreover cholecalciferol excretion may be
increased.
U.V. irradiation was applied both to normal
0.0009
0.0004
0.003 I
subjects and to subjects receiving hepatic microsoma1 enzyme-inducing drugs. Initial plasma
25-(OH)D, levels were lower in subjects receiving
these drugs, but the resulting plateau levels and
the rate of decline on cessation of U.V. irradiation
were unaffected. The phenomenon of hepatic
microsomal enzyme induction has been proposed to explain the low levels of plasma
25-(OH)D, frequently found in association with
phenytoin and phenobarbitone administration
[241, together with the impaired response to oral
vitamin D [251. The present data show that it is
important to continue dosing until a plateau level
of 25-(OH)D, is reached: otherwise the plasma
25-(OH)D, response to vitamin D may appear to
be impaired. If the plasma level of 25-(OH)D,
had been compared after 2 weeks (Fig. la), the
level of plasma 25-(OH)D, would have been
lower in anticonvulsant-treated subjects than in
the normal subjects. This may explain the
differences of plasma 25-(OH)D, concentration
found in normal and anticonvulsant-treated subjects after oral dosage with vitamin D for 3 weeks
[251.
The existence of hepatic microsomal enzyme
induction might also be expected to reduce the
value of D in the steady-state equation by
restricting the amount of cholecalciferol available
for conversion into 25-(OH)D,, thereby leading
to lower plasma 25-(OH)D, concentrations if Vd
and T,,, remain unchanged. Tracer studies with
25-(OH)D, suggest that
is not affected by
anticonvulsant drugs, although V , may be
slightly lower [261. Diminished vitamin D storage
is also unlikely since the rate of decline of plasma
25-(OH)D, after cessation of U.V. irradiation was
similar in groups 1 and 2. The initial low plasma
25-(OH)D, levels in group 2 do, however,
470
M.W.J. Davie et al.
indicate that the dosage of anticonvulsant was
sufficient to exert an effect on vitamin D
metabolism.
In calculating cutaneous vitamin D synthesis
from the plasma 25-(OH)D, levels, it is assumed
that all cholecalciferol is converted into 25(OH)D,. Cholecalciferol undergoes rapid 25hydroxylation at low plasma 25-(OH)D, concentrations [4, 161, and there is no evidence that
cholecalciferol accumulates whilst plasma 25(OH)D, is below 62.5 nmol/l (Fig. 5): such a
level was not achieved by any subject receiving
limited area U.V. irradiation. Accumulation in
tissues other than plasma is unlikely, since, even
during high dosage of cholecalciferol, concentrations in plasma are equivalent to those in other
tissues. When dosing is discontinued, some
tissues, notably fat, release cholecalciferol more
slowly than plasma and may exhibit higher levels
1231. Cholecalciferol could also suffer biliary
excretion, rendering itself unavailable for 25hydroxylation; indeed excretion of radioactivity
after intravenous 3H or 14Cas cholecalciferol was
slightly greater in the faeces of a normal subject
[281, or in the urine of subjects with biliary
obstruction 1291 compared with the excretion after
administration of [,HI25-(OH)D3. If significant
losses occurred by means of this pathway,
plasma 25-(OH)D, concentrations during oral
administration would be lower than those predicted by the steady-state equation. Although this
was true for high intakes of cholecalciferol (see
above), it did not occur with oral doses of vitamin
D that were associated with plasma 25-(OH)D,
levels corresponding to those found during
irradiation.
Since 25-(OH)D, is the main circulating
metabolite in plasma, values of V, and T,,,
relating to this metabolite were used in the
steady-state equation. T1,, calculated by this
method is shorter than the TlI2 of plasma
25-(OH)D,, possibly as a result of daily vitamin
D intake 1201. The value for the radioactive
25-(OH)D, half life is at the lower end of the
range 13.3-23.0 days previously reported [30,
311, but compares well with the value of 13.5
days found in normal subjects ([26]; P. Siklos &
M. Davie, unpublished work). Vitamin D synthesis in human skin (Table 4) is similar to that
found in rat skin, 0.001 1-0-0027 nmol/mJ being
synthesized according to the wavelength used
I121. A higher value for synthesis calculated from
the oral equivalent (Table 4) is expected since not
all the cholecalciferol ingested will be absorbed
[271. At these rates of synthesis considerable
exposure would be required to make any impact
on the 2-4 nmol/cm2 of 7-dehydrocholesterol
present in skin [31. For the purpose of the
calculation, no account was taken of the initial
plasma 25-(OH)D, level or of the contribution of
dietary vitamin D, since both were low in subjects
making up group 2. In subjects from group 1, use
of the incremental plasma 25-(OH)D, relationship with U.V. energy (Fig. 3) suggested that
cholecalciferol synthesis was 0.001 f 0.0003
nmol/mJ. Dietary vitamin D intake was very low
in all subjects, and from the steady-state
equation, assuming 75% absorption to calculate
the expected steady-state level, it was apparent
that no more than 10% of the plateau level could
be accounted for from this source.
The potential contribution of U.V. energy to
vitamin D status may be assessed by using the
data in Table 4 relating to cholecalciferol synthesis and published incident U.V. energy data 18,
321. From the data of Bunker & Harris [13] and
Johnson et al. [321 up to 315 nm, the total
antirachitic activity was calculated in the same
manner as the activity of the Hanovia lamp [ 101,
for latitude 45"N. Calculation of the area under
the curve relating the wavelength with (energy at
1 x antirachitic activity at d) showed that 17.5
mJ day-' cmP2are incident in January compared
with 712.5 mJ day-' cmP2in July. However 60%
of daily U.V. energy falls in the 4 h around midday
G.M.T. 133, 341: thus about 15% of total energy
will be available in each hour around midday.
Partial cloud cover may reduce energy by 25%
and this hour might contain 15 x 75% = 11-25%
of daily U.V. energy. The incident U.V. energy over
this 1 h period would thus be 1.97 (17.5 x
0.1125) and 80.2 mJ/cm2 in January and July
respectively. However, direct measurement of
U.V. energy in June/July in southern England has
shown that gardeners were exposed to 21.8
mJ/cm2 per day and indoor workers only to 3.7
mJ/cm2 per day. Expected plateau plasma 25(OH)D, levels associated with these incident
energy values may be calculated either from the
relationship in Fig. 2 or by the steady-state
equation, with the values for cholecalciferol
synthesis depicted in Table 4. Assuming that the
dorsum of the hands and the face are exposed
(about 600 cm2) in a 70 kg man, the relationship
described in Fig. 2 predicts that daily exposure to
80.2 mJ/cm2 would lead to a plateau level of
plasma 25-(OH)D, of 82.7 nmol/l, and that the
gardeners and indoor workers would have 26.0
and 10.9 nmol/l respectively. To obtain values
from the steady-state equation, a value for D is
calculated from the product of area irradiated x
energy applied x cholecalciferol synthesis, using
the figure of 0.00154 nmol/mJ for cholecalciferol synthesis (Table 4), and values for T,,, and
Cutaneous vitamin D synthesis
V, as described above. For the same 70 kg
subject with exposure of 600 cm2, exposure to
80.2 mJ/cm2 is associated with a plasma 25(OH)D, level of 91.0 nmol/l, compared with
24.7 nmol/l for the gardeners and 4.2 nmol/l for
indoor workers. Since plasma 25-(OH)D, levels
can rise to 72 nmol/l in England [351, it is likely
that exposure must be longer or the area
irradiated greater than has been accounted for in
the above calculation. However, the low plasma
25-(OH)D, levels in groups 1-6 (Table 1) suggest
that institutionalized subjects lack both U.V.
exposure and vitamin D in the diet, since both are
effective at increasing plasma 25-(OH)D, levels
(Fig. 1). The summer months give an opportunity
for additional U.V. exposure, especially at weekends, and since the half life of 25-(OH)D, extends
over 2 weeks, additional exposure each weekend
will have a cumulative effect on plasma 25(OH)D, concentrations. This will have an effect
throughout the year since winter levels of plasma
25-(OH)D, are related to the summer levels
[2,361.
There is also an advantage in using U.V.
irradiation in anticonvulsant-treated subjects.
Although doubt exists about the effect of oral
vitamin D on the plasma 25-(OH)D, level [25,
371, it is evident from the present study that U.V.
exposure yields identical responses in both
normal and anticonvulsant-treated subjects.
Note added in proof
Since this paper was submitted, our attention has
been drawn to similar conclusions arising out of
work in animals [Takada, K. et al. (1981)
Journal of Steroid Biochemistry, 14,136 1-1367.
Acknowledgments
We are grateful to Dr S. G. P. Webster,
Consultant Physician, United Cambridge Hospitals, for allowing us to study elderly patients
under his care. We also acknowledge the help of
Mr J. S. Ashford in performing assays for
cholecalciferol and 25-hydroxycholecalciferol.
M.W.J.D. gratefully acknowledges the financial
assistance provided by The Royal Society and by
J. Sainsbury Ltd.
References
I11 HADDAD,J.G. & HAHN,T J . (1973) Natural and synthetic
sources of circulating 25-hydroxyvitamin D in man. Nature
(London), 244,515-517.
I21 LAWSON,D.E.M., PAUL,
A.A., BLACK,A.E., COLE, T.J.,
MANDAL, A.R. & DAVIE,M. (1979) Relative contributions of
diet and sunlight to vitamin D status in the elderly. British
Medical Journal, il, 303-305.
I31 BEKEMEIER,
H. (1966) Vitamin D der haut. In: International
47 1
Zeitsschrifl Vitamitfersch 10, chapter 6. Hans Huber,
Stuttgart.
I41 MAWER,E.B., LUMB,G.A., SCHAEFFER,
K. & STANBURY,
S.W.(1971) The metabolism of isotopically labelled vitamin
D, in man: the influence of the state of vitamin D nutrition.
Clinical Science, 40.39-53.
151 HADDAD,J.G. & CHYU, K.J. (1971) Competitive protein
binding radioassay for 25-hydroxycholecalciferol.Journal of
Clinical Endocrinology and Metabolism, 33,992-995.
C.A., PIETREK, J.,
I61 PREECE,M.A., TOMLINSON,S., RIEZOT,
KORN. H.T., DAVIES,D.M., FORD,
J.A., DUNNIGAN,
M.G. &
O'RIORDAN.J.L.H. (1975) Studies of vitamin D deficiencv in
man. Quarterly Journal' of Medicine (New Series), '44,
575-589.
AS., MACLENNAN,
W.J. HAMILTON,
J.C., ROSE,P. &
171 NAYAL,
KONG,M. (1978) 25-Hydroxy-vitaminD, diet and sunlight in
patients admitted to a geriatric unit. Gerontology, 24, 117-122.
I81 CHALLONER,A.V.J., CORLESS,D., DAVIS, A., DEANE,
G.H.W., DIFFEY, B.L., GUPTA,S.P. & MAGNUS,LA. (1976)
Personnel monitoring of exposure to ultraviolet radiation.
Clinical and Experimental Dermatology, 1,175-179.
D.E.M. (1979) Role of
I91 PIITET,P.G., DAVIE,M. & LAWSON,
nutrition in the development of osteomalacia in the elderly.
Nutrition and Metabolism, 23, 109-1 16.
D.E.M. (1980). Assessment of plasma
[ 101 DAVIE,M.& LAWSON,
25-hydroxyvitaminD response to ultraviolet irradiation over a
controlled area in young and elderly subjects. Clinical Science,
58,235-242.
T. & YASUMURA, M. (1973) Studies on the
I111 KOBAYASHI,
ultraviolet irradiation of provitamin D and its related compounds. 111. Effect of wavelength on the formation of potential
vitamin D, in the irradiation of ergosterol by monochromatic
ultraviolet rays. Journal of Nutritional Science and
Vitaminology, 19, 123-128.
F. (1938) Quantitative studies of
A. & BENPORD,
I121 KNUDSEN,
the effectiveness of ultraviolet radiation of various wavelengths in rickets. Journal of Biological Chemistry, 124,
287-289.
J.W.M. & HARRIS,
R.S. (1937) Precise evaluation of
I131 BUNKER,
ultraviolet therapy in experimental rickets. New England
Journal ofMedicine, 216,165-169.
I141 JONES, G. (1978) Assay of vitamins D, and D,, and
25-hydroxyvitamins D, and D, in human plasma by high
performance liquid chromatography. Clinical Chemistry, 24,
287-298.
E.M. (1960) The compoI151 MCCANCE,R.A. & WIDDOWSON,
sition of foods. Medical Research Council Special Report
Series, no. 297. HMSO, London.
1161 JUNG,R.T.,DAVIE,M.,HUNTER, J.O., CHALMERS,T.M. &
LAWSON,
D.E.M. (1978) Abnormal vitamin D metabolism in
cirrhosis. Gut, 19,290-293.
I171 GREENBLATT,D.J. & KOCH-WESER,J. (1975) Clinical
pharmacokinetics. New England Journal of Medicine, 293,
702-705 964-9713
I181 GOODMAN,L.S. & GILMAN,A. (1975) In: The Pharmacological Basis of Therapeutics, 5th edn, pp. 18-24. MacMilIan, New York.
1191 HUNTER,J.O.& DAVIE,M. (1980) Vitamin D metabolism in
patients treated with phenytoin and phenobarbitone. In:
Environmental Chemicals. Enzyme Function and Human
Disease, pp. 315-330. Ciba Symposium 76. Excerpta Medica.
Amsterdam.
[201 STANBURY, S.W., MAWER,E.B., TAYLOR,C.M. & DE SILVA,
P. (1980) The skin, vitamin D and the control of its
25-hydroxylation. An attempted integration. Mineral and
Electrolyte Metabolism, 3.5 1-60.
I211 MAWER,E.B. & REEVE,A. (1977) The use of an isolated
perfused liver to study the control of cholecalciferol 25hydroxylase activity in the rat. Calcified 7Yssue Research, 22
(Suppl.), 24-28.
I221 MAWER,E.B., BACKHOUSE,
J., HOLMAN,C.A., LUMB,G.A. &
STANBURY,
S.W. (1972) The distribution and storage of
vitamin D and its metabolites in human tissu&s. Clinical
Science, 43.4 13-43 1.
I231 ROSENSTREICH,S.J., RICH, C. & VOLWILER, W. (1971)
Deposition in and release of vitamin D, from body fat:
evidence for a storage site in the rat. Journal of Clinical
Investigation, 50,679-687.
I241 STAMP,T.C.B., ROUND,J.M., R o w , D.J.F. & HADDAD,
J.G.
412
M.W.J. Davie et al.
(1972) Plasma levels and therapeutic effect of 25-hydroxycholecalciferol in epileptic patients taking anticonvulsant
drugs. British Medical Journal, iv, 9-12.
1251 BOUILLON,
R., REYNAERT,
J., CLAES, J.H., LISSENS,W. & DE
MOOR, P. (1975) The effect of anticonvulsant therapy on
serum levels of 25-hydroxy-vitamin D, calcium, and parathyroid hormone. Journal of Clinical Endocrinology and
Metabolism, 41, 1130-1 135.
M.W.J. (1981) M.D. Thesis, University of Cambridge.
I261 DAVIE,
I271 THOMPSON,
G.R., LEWIS,B. & BOOTH,C.C. (1966) Vitamin D
absorption aRer partial gastrectomy. Lancet, i, 457-458.
1281 ARNAUD,S.B., GOLDSMITH,R.S., LAMBERT,P.W. & Go,
V.L.W. (1975) 25-Hydroxyvitamin D,: evidence of an enterohepatic circulation in man. Proceedings of the Society for
Experimental Biology and Medicine, 149,570-572.
I291 JUNG, R.T., DAVIE, M., SIKLOS, P., CHALMERS,T.M.,
HUNTER,J.O.& LAWSON,D.E.M. (1979) Vitamin D metabolism in acute and chronic cholestasis. Gut. 20,840-847.
1301 BEC, P., BAYARD,F. & LOUVET,J.P. (1972) 25-Hydroxycholecalciferol dynamics in human plasma. Revue Europeanne Etudes Cliniques el Biologique, 17, 793-796.
I311 GRAY,R.W., WEBER,H.P., DOMINGUEZ,
J.H. & LEMANN,J.
(1974) The metabolism of vitamin D, and 25-hydroxyvitamin
D, in normal and anephric humans. Journal of Clinical
Endocrinology and Metabolism, 39, 1045- 1056.
1321 JOHNSON,F.S., MO, T. & GREEN,
A.E.S. (1976) Average
latitudinal variation in ultraviolet radiation at the earth’s
surface. Photochemistry and Photobiology, 23,179- 188.
[331 SCHULZE,R. & GRAPE, K. (1969) Consideration of sky
ultraviolet radiation in the measurement of solar ultraviolet
radiation. In: The Biologic Effects of Ultraviolet Radiation, pp.
359-373. Ed. Urbach, F. Pergamon Press, Oxford.
1341 ADAMS, J.T. (1952) Livestock and their environments:
sunlight and vitamin D. Veterinary Record, 64, 15 1-157.
1351 STANBURY,
S.W. & MAWER,E.B. (1978) Physiological aspects
of vitamin D metabolism in man. In: Vitamin D, p. 315. Ed.
Lawson, D.E.M. Academic Press, London.
1361 LAWSON,D.E.M. & DAVIE, M. (1979) Aspects of the
metabolism and function of vitamin D. Vitamins and
Hormones, 37.1-67.
I371 HAHN,T.J., HENDIN,B.A., SCHAIW,C.A. & HADDAD,J.G.
(1972) Effect of anticonvulsant therapy on serum 25-hydroxycholecalciferol levels in adults. New England Journal of
Medicine, 287,90&904.