original article

Bioavailability of vitamin D3 in
non-oily capsules: the role
of formulated compounds
and implications for
intermittent replacement
Biodisponibilidade da vitamina D3 em cápsulas não
oleosas: o papel das formulações manipuladas e as
implicações para a substituição intermitente

Iara Maria Gomes Coelho1, Luena Dias de Andrade2,

Lunnara Saldanha2, Erik Trovão Diniz2, Luiz Griz2, Francisco Bandeira2

ABSTRACT
Objective: To evaluate the bioavailability of vitamin D in capsules as compared with oily drops 1
Ambulatório de Endocrinologia
Ginecológica, Hospital
in nuns living in a closed community with very low sun exposure. Methods: A randomized, 2
Agamenon Magalhães,
x 2 crossover, open clinical trial was conducted, with 18 nuns aged between 20 and 75 years. Sistema Único de Saúde (SUS),
Samples were collected in the fasting state and at 4, 8, 12 and 24 hours following the adminis- Universidade de Pernambuco
(UPE), Recife, PE, Brasil
tration of capsules and oily drops (both containing vitamin D3 66,000 UI plus vitamin A 13,200 2
Divisão de Endocrinologia
UI) to determine serum 25 hydroxivitamin D concentrations (25OHD), at baseline and 90 days e Diabetes, Hospital
after. The evaluation was based on the maximum concentration (Cmax) and area under the curve Agamenon Magalhães, SUS/
UPE, Recife, PE, Brasil
(AUC0-24). Results: The capsule formulation presented Cmax and AUC0-24, 5.78% and 0.76%, res-
pectively, greater than the oily drops formulation. Conclusion: Both formulations were within
the limits for a bioequivalence study, namely C-90% for Cmax and AUC0-24, and the drugs were
considered bioequivalent. Arq Bras Endocrinol Metab. 2010;54(2):239-43
Keywords
Bioavailability; vitamin D; pharmacokinetics

RESUMO
Objetivo: Avaliar a biodisponibilidade da vitamina D3 em cápsulas, comparando com gotas Correspondence to:
Iara Maria Gomes Coelho
oleosas, em religiosas que vivem em comunidade fechada com baixa exposição solar. Méto- Rua Setubal, 596, ap. 602
dos: Ensaio clínico aberto, aleatório e cruzado 2 x 2, com 18 religiosas (20-75 anos de idade). 51030-010 − Recife, PE, Brasil
[email protected]
As amostras foram coletadas em jejum e 4, 8, 12 e 24 horas após a administração de cápsulas
e de gotas oleosas (ambos contendo 66.000 UI de vitamina D3 e 13.200 UI de vitamina A) para
dosagem da 25 hidroxivitamina D (25OHD), em data base e 90 dias após. A avaliação baseou- Received on Nov/16/2009
Accepted on Feb/28/2010
se nos resultados da concentração máxima (Cmáx) da 25OHD e da área sob a curva (ASC0-24).
Resultados: A análise descritiva desses parâmetros demonstrou que a cápsula apresentou Cmáx
e ASC0-24 de 5,78% e 0,76% a mais que as gotas oleosas, respectivamente. Conclusão: Ambas
as formulações encontravam-se dentro dos limites de aceitação em estudo de bioequivalência
IC-90% para Cmáx e ASC (0-24), daí as drogas serem consideradas bioequivalentes. Arq Bras Endocrinol
Copyright© ABE&M todos os direitos reservados.

Metab. 2010;54(2):239-43
Descritores
Biodisponibilidade; vitamina D; farmacocinética

Arq Bras Endocrinol Metab. 2010;54/2 239

 Bioavailability of vitamin D in non-oily capsules

Introduction drugs, with the second series of samples being taken

The prevention of vitamin D deficiency and insufficien- after administering the drug not given in the first phase
cy remains a priority of both domestic and international of the study, in a cross-over design.
health services (1-3). Different administration protocols The intervention therapies were as follows: 1) Oily
have been used taking into account the long half-life of drops − 13,200 IU of vitamin A acetate and 66,000
vitamin D, since there is interest in administering vitamin IU of vitamin D3 (Aderogil – Sanofi Aventis) in excipi-
D3 intermittently. A systematic review evaluated several ent peanut oil − (reference). 2) Manipulated capsules
studies since 1980, with intermittent doses of vitamin D, – 13,200 IU of vitamin A acetate and 66,000 IU of
ranging from 100,000 IU per month to 4,000 IU daily, vitamin D3 in lactose excipient – (test). The interval
and concluded that this treatment option is safe, effec- between the two periods of administration (washout
tive, simple to use and has better compliance (4). time) was 90 days in order to eliminate the residual ef-
Vitamin D is known to be liposoluble, and its rela- fects of the formulations administered. This time was de-
tive bioavailability could result in unfavorable conditions termined based on the half-life of the drug administered
when administered in solid form (capsules), since the pro- and the liposoluble characteristics of vitamin D3 (5).
cess of its release is a factor limiting the rate of absorption, Serum 25OHD was measured by radioimmunoas-
bearing in mind that bioavailability is related not only say (DiaSorin, Inc, Stillwater, MN, USA). The lowest
to the pharmaceutically active molecules, but also, and detectable limit was 1.5 ng/mL (3.7 nmol/L), and the
most importantly, to the formulation and excipients used. intra-assay and inter-assay coefficients variation were 6%
The aim of this study was to evaluate whether vitamin and 8%, respectively. On both occasions fifteen days af-
D would achieve bioavailability within the limits of safety ter the administration of the drugs, a clinical assessment
and efficacy when administered in non-oily capsules. was made in all the subjects, and there were no com-
plaints or side-effects being associated with the therapy.
The clinical study protocol was approved by the HAM
Patients and methods Ethics in Research Committee.
An open, randomized, 2 x 2 crossover trial was con-
ducted in healthy nuns living in a closed community
Comparative bioavailability −
whose garments exposed only 10% of the body (face,
hands and feet), who shared the same eating habits and
Pharmacokinetics analysis
who were engaged, in a disciplined manner, in outdoor The pharmacokinetic measures evaluated to determine
work (gardening) for 1 hour every 20 days, thereby bioavailability derived directly from the curve of serum
having very low sun exposure. These nuns spontane- concentration of the drug versus time, by quantifying
ously sought treatment at the gynecology outpatient the biological samples in relation to the previously es-
clinic of Agamenon Magalhães Hospital (HAM). tablished collection times (0, 4, 8, 12 and 24 hours).
The inclusion criteria were as follows: nuns with a The design of this study provided the subjects with
low sun exposure; members of the same religious order their own control, reducing external factors in the anal-
(Carmelites); aged between 20 and 75 years with body ysis of the parameters that define bioavailability (speed
mass index between 21.8 and 29 kg/m2, who sponta- and amount of drug entering the bloodstream). The
neously agreed to participate in the trial after giving in- following pharmacokinetic measurements were evalu-
formed consent. The exclusion criteria were as follows: ated: plasma concentration of the curve drug versus
systemic inflammatory or malignant disease, hepatic or time (AUC0-24), the maximum plasma concentration
renal failure, uncontrolled hypo− or hyperthyroidism, achieved (Cmax), and the time to peak concentration
the use of drugs that are known to affect bone metabo- (Tmax). All pharmacokinetic parameters were calculated
lism such as bisphosphonates, glucocorticoids and anti- using the statistical software WinNolin 5.0.
Copyright© ABE&M todos os direitos reservados.

convulsants. Eighteen patients were randomly assigned

and two dates were set with 90 days of interval between
them in order to measure serum 25OHD concentra- Statistical analysis
tions in response to both intervention therapies. The pharmacokinetic parameters AUC0-24 and Cmax
Blood samples were taken in the fasting state and at were converted into a natural logarithm for the anal-
4, 8, 12 and 24 hours after the administration of both ysis of variance (ANOVA), and the confidence inter-

240 Arq Bras Endocrinol Metab. 2010;54/2

 Bioavailability of vitamin D in non-oily capsules

vals were also calculated (CI). The construction of the test in both cases proved to be high, thus demonstrat-
confidence interval of 90% for the difference of means ing that the number of volunteers (7) was sufficient to
was based on the least square means of the data con- obtain the desired minimum power of 80%.
verted into logarithms and the residual mean square On the basis of the results obtained, the two drugs
of this ANOVA test. The antilogarithms of the con- were considered bioequivalent, since the 90% CI for
fidence limits obtained constitute the 90% CI for the the ratio between the means of the Cmax and AUC phar-
ratio of geometric means between the test and refer- macokinetic parameters were between 80% and 125%
ence drugs. The conclusion of mean bioequivalence is and thus interchangeable (8).
achieved when this confidence interval is between 80%
and 125%. This method is equivalent to the procedure
of two one-tailed tests corresponding to the null hy- 28
pothesis of bioequivalence (6).
27

Mean (ng/mL)
26
Results 25 Refer
Figure 1 shows the mean serum 25OHD concentra- 24 Test
tions after the reference and test drugs. Seventy-eight 23
percent of the study patients had 25OHD levels below
22
30 ng/mL in the basal state. In the descriptive analysis 0 5 10 15 20 25
of the pharmacokinetic parameters presented in Tables Time (hr)
1 and 2, the test formulation presented a Cmax 5.78% Figure 1. Mean serum 25OHD after the test and reference compounds
higher than the reference drug, while the AUC0-24
showed an absorption 0.76% greater than the reference
Table 1. Mean pharmacokinetics parameters of 18 volunteers for the
formulation. In relation to the Tmax, very similar times reference drug (R)
were observed − 15.11 h and 14.00 h – respectively, for
AUC(0-24)
the test and reference formulations. Cmax (ng/mL) Tmax (h)
(h ng/mL)
The confidence interval for the mean Cmax was Mean 28.51 14 592.07
96.91% (lower limit) and 111.57% (upper limit), ac- Geometric mean 27.34 9.80 556.51
cording to the classic confidence interval. Values ac- DP 8.44 8.92 182.86
cording to the Westlake test were 90.22 (lower limit) CV 29.60 63.70 30.88
and 109.78 (upper limit). The mean Cmax of the refer- Minimum 17.40 0.00 359.80
ence drug (oily drops) calculated by the method of the Maximum 45.20 24.00 949.20
least squares was 3.30 and 3.34 for the test drug (cap-
sules), resulting in a ratio (Cmax Test/Cmax Reference) of
Table 2. Mean pharmacokinetic parameters of 18 volunteers for the test
103.98 with a Cmax test power of 99.91% and an AUC drug (T)
of 99.99%. The Tmax values were 11.78 and 12.67 for AUC(0-24)
Cmax (ng/mL) Tmax (h)
the test and reference drugs, respectively. We also cal- (h ng/mL)
culated the classic confidence interval for the AUC0-24 Mean 30.16 15.11 596.57
values, yielding of 93.44 and 104.08 for the lower and Geometric mean 28.43 13.00 558.68
upper limits, respectively. The values of the confidence DP 10.56 7.71 233.92
interval using Westlake were 94.25 (lower limit) and CV 35.01 51.00 39.21
105.75 (upper limit). The mean AUC0-24 calculated by Minimum 11.70 4.00 243.00
the least squares method of the reference and test drugs Maximum 56.50 24.00 1260.60
Copyright© ABE&M todos os direitos reservados.

was 6.33 and 6.32, respectively. The ratio between the

means (AUC0-24 Test/AUC0-24 Reference) was 98.62,
showing a test power of 99.99%. Discussion
The statistical calculations for the Cmax and AUC0-24 There are important concerns regarding bioavailability,
confidence intervals were discriminatory for the asser- bioequivalence and interchangeability of the various
tion of bioequivalence as the power of the statistical formulations used in therapeutic compounds, for which

Arq Bras Endocrinol Metab. 2010;54/2 241

 Bioavailability of vitamin D in non-oily capsules

there are many factors that can modify the release, dis- reported by Alagol (12), and Abdullah (13) in Turkish
solution and absorption of the drug in the body. There and Saudi Arabian women who, due to covering their
are several ways of evaluating the bioavailability of bodies with clothing, usually have low serum 25OHD
drugs, but the recommended method is the quantita- concentrations.
tive determination of the drug and/or its metabolite in The coefficient of intra-individual variation (CV), of
body fluids (blood, plasma) as function of time, due to both formulations was less than 35% for Cmax and 39%
its greater precision and accuracy. for AUC0-24, demonstrating that vitamin D is not a drug
Vitamin D deficiency is common in individuals of of high variability (14). As the appropriate size of the
various ethnic groups around the world, in both de- sample was dependent on the power of the test (in our
veloping and developed countries, even in those close study > 80%), based on an estimate of the coefficient
to the equator (9). New applications for vitamin D are of individual variation, the sample size was sufficient to
being sought in an attempt to present them for major accurately determine the equivalence and interchange-
public health problems such as osteoporosis, which in- ability of the two formulations in question.
creases with advancing age, autoimmune diseases, type 1 In conclusion the amount and speed (bioavailabil-
and type 2 diabetes, and cancers, among others. ity) of vitamin D3 entering the bloodstream is not af-
Muindi and cols. (10) conducted a comparative phar- fected by capsule formulation, which suggests that pro-
macokinetic study comparing an oily liquid calcitriol for- phylaxis and treatment of vitamin D deficiencies may
mulation administered to patients with advanced cancer be carried out with manipulated capsules. Considering
with the soft gelatin capsule formulation and concluded intermittent high-dose of oral vitamin D3 is as effective
that the liquid formulation does not offer any advan- as a daily dose (in the same amount) in maintaining
tages in terms of pharmacokinetics or bioavailability over adequate serum 25OHD levels, thus improving long-
the capsule formulation. The results indicated that cal- term compliance, our results should have practical im-
citriol in liquid form failed to attain Cmax and AUC val- plications for the clinical use of vitamin D (15,16).
ues associated with an anti-tumor action (10).
Disclosure: no potential conflict of interest relevant to this article
A study conducted in the city of Boston set out to
was reported.
evaluate whether vitamin D, would attain bioavailabil-
ity within the levels of safety and efficacy when con-
tained in non-fatty products, such as skimmed milk,
References
orange juice, and corn oil on toast, comparing them
1. Schuch NJ, Garcia VC, Martini LA. Vitamin D and endocrine disea-
with whole milk enriched with that vitamin. The au- ses. Arq Bras Endocrinol Metab. 2009;53(5):625-33.
thors concluded that the level of fat in the milk does 2. Bandeira F, Griz L, Dreyer P, Eufrazino C, Bandeira C, Freese E.
not affect the bioavailability of vitamin D (11). Vitamin D deficiency: a global perspective. Arq Bras Endocrinol
Metab. 2006;50(4):640-6.
Our data show that the test drug (capsules), when
3. Holick MF. Vitamin D: the underappreciated D-lightful homone
compared with the reference drug (oily drops), pre- that is important for skeletal and cellular health. Curr Opin Endo-
sented a Cmax and an AUC0-24, 5.78% and 0.76%, respec- crinol Diabetes. 2002;9:87-98.
tively, higher than the reference drug. It was also ob- 4. Hathcock JN, Shao A, Vieth R, Heaney R. Risk assessment for vi-
tamin D. Am J Clin Nutr. 2007;85:6-18.
served that the Tmax showed very similar times, namely 5. Van den Berg, H. Bioavailability of vitamin D. Eur J Clin Nutr.
15.11 h and 14 h, respectively, for the test and refer- 1997;51(1 suppl):S76-9.
ence drugs. The above data indicate that the fact that 6. Anvisa − Agência Nacional de Vigilância Sanitária. [acesso 22 ago
2006]. Disponível em: http://www.anvisa.gov.br/legis/resol/44_77.htm.
the vehicle used in the administration of the capsules is
7. Tothfalusi L, Endrenyi L, Midha KK, Rawson MJ, Hubbard JW.
not of lipophilic origin it does not affect the bioavail- Evaluation of the bioequivalence of highly-variable drugs and
ability of vitamin D. drug products. Pharm Res. 2001;18(6):728-33.
From the results of the basal concentrations of 8. Storpirtis S, Marcolongo R, Gasparotto FS, Vilanova CM. A equi-
Copyright© ABE&M todos os direitos reservados.

valência farmacêutica no contexto da intercambialidade entre

25OHD, we concluded that 77.75% of the nuns had
medicamentos genéricos e de referência: bases técnicas e cientí-
vitamin D deficiency. We can thus state that the brief ficas. Gerência de Medicamentos Genéricos-Brasil. Agência Na-
and casual exposure of the face, arms and hands to sun- cional de Vigilância Sanitária. São Paulo: (sn), 2004.

light may not provide certain individuals with their vi- 9. Pasco JA, Henry MJ, Nichollson GC, Sanders KM, Kotowick MA.
Vitamin D status of women in the Geelong Osteoporosis Study:
tamin D requirements, and also that the garments limit association with diet and casual exposure to sunlight. Med J
exposure to sunlight. This is in agreement with the data Aust. 2001;175:401-5.

242 Arq Bras Endocrinol Metab. 2010;54/2

 Bioavailability of vitamin D in non-oily capsules

10. Muindi JR, Potter DM, Peng Y, Johnson CS, Trump DL. Pharma- 13. Abdullah MA, Salhi HS, Bakry LA, Okamoto E, Abomelha AM,
cokinetics of liquid calcitriol formulation in advanced solid tumor Stevens B, et al. Adolescent rickets in Saudi Arabia: a rich and
patients: comparison with caplet formulation. Cancer Chemother sunny country. J Pediatr Endocrinol Metab. 2002;15:1017-25.
Pharmacol. 2005;56(5):493-6. 14. Demirturk E, Oner L. Highly variable drugs in bioavailability and
11. Tangpricha V, Koutkia P, Rieke SM, Chen TC, Perez AA, Holick bioequivalence studies. J Pharm Sci. 2002;27(3):165-72.
MF. Fortification of orange juice with vitamin D: a novel ap- 15. Ilahi M, Armas LA, Heaney RP. The effect of a single large oral
proach for enhancing vitamin D nutritional health. Am J Nutr. dose of cholecalciferol on serum 25-hydroxyvitamin D levels in
2003;77:1478-83. humans. J Bone Miner Res. 2007;22(Suppl1):560.
12. Alagol F, Shihadeh Y, Boztepe H, Tanakol R, Yarman S, Azizlerli H, 16. Bacon CJ, Gamble GD, Horne AM, Reid IR. A comparison of three
et al. Sunlight exposure and vitamin D deficiency in Turkish wo- high dose oral vitamin D3 supplementation regimens. J Bone Mi-
men. J Endocrinol Invest. 2000;23:173-7. ner Res. 2007;22(Suppl):S126-7.

Copyright© ABE&M todos os direitos reservados.

Arq Bras Endocrinol Metab. 2010;54/2 243