Dpi.dipa.co.id

Copyright 2002 by The Endocrine Society VIII: Meta-Analysis of the Efficacy of Vitamin D Treatment in
Preventing Osteoporosis in Postmenopausal Women
EMMANUEL PAPADIMITROPOULOS, GEORGE WELLS, BEVERLEY SHEA, WILLIAM GILLESPIE,BRUCE WEAVER, NICOLE ZYTARUK, ANN CRANNEY, JONATHAN ADACHI, PETER TUGWELL,ROBERT JOSSE, CAROL GREENWOOD, GORDON GUYATT, THE OSTEOPOROSIS METHODOLOGYGROUP, AND THE OSTEOPOROSIS RESEARCH ADVISORY GROUP are based. The evidence supporting the current guidelines,particularly with respect to the administration of vitamin D, Objective: To review the effect of vitamin D on bone
is limited. Gillespie et al. (4) have conducted a meta-analysis density and fractures in postmenopausal women.
Data Source: We searched MEDLINE and EMBASE from
addressing the effect of vitamin D on vertebral and nonver- 1966 to 1999 and examined citations of relevant articles and tebral fractures. This meta-analysis met major methodolog- proceedings of international meetings. We contacted osteo- ical criteria: the question was clear and sensible, inclusion porosis investigators and primary authors to identify addi- and exclusion criteria were explicit, and the search for studies tional studies and to obtain unpublished data.
Study Selection: We included 25 trials that randomized
The Gillespie et al. meta-analysis is, however, limited in women to standard or hydroxylated vitamin D with or with- that it did not address the effect of vitamin D on bone density.
out calcium supplementation or a control and measured Furthermore, Gillespie et al. took a relatively conservative bone density or fracture incidence for at least 1 yr.
approach to pooling, and made little use of regression meth- Data Extraction: For each trial, three independent review-
ods to explore the appropriateness of combining data across ers assessed the methodological quality and abstracted data.
different forms of vitamin D and variations in study design.
Data Synthesis: Vitamin D reduced the incidence of ver-
As a result, the Gillespie study was largely descriptive and tebral fractures [relative risk (RR) 0.63, 95% confidence in- permitted few definitive conclusions (4). As part of our series terval (CI) 0.45– 0.88, P Ͻ 0.01) and showed a trend toward of systematic reviews of osteoporosis treatment, we therefore reduced incidence of nonvertebral fractures (RR 0.77, 95% CI conducted another systematic review to address these lim- 0.57–1.04, P ϭ 0.09). Most patients in the trials that evaluated itations using the Cochrane methodology. We describe the vertebral fractures received hydroxylated vitamin D, and methods of our review in detail in Section I.
most patients in the trials that evaluated nonvertebral frac-tures received standard vitamin D.
Hydroxylated vitamin D had a consistently larger impact on bone density than did standard vitamin D. For instance, 1. Inclusion criteria. Studies satisfied the following inclusion total body differences in percentage change between hy- criteria; 1) participants were women older than 45 yr with droxylated vitamin D and control were 2.06 (0.72, 3.40) and absence of menses for a minimum of 6 months; 2) the treat- 0.40 (Ϫ0.25, 1.06) for standard vitamin D. At the lumbar spineand forearm sites, hydroxylated vitamin D doses above 50 ␮g ment group received some form of vitamin D greater than yield larger effects than lower doses.
400 IU daily, or some form of dihydroxyvitamin D; 3) a Vitamin D resulted in an increased risk of discontinuing follow-up of at least 1 yr; 4) results reported on x-ray evi- medication in comparison to control as a result of either dence of fractures of hip, vertebrae, or wrist, or bone mineral symptomatic adverse effects or abnormal laboratory results density measured in grams per centimeter or grams per (RR 1.37, 95% CI 1.01–1.88), an effect that was similar in trials centimeter squared, by single-photon absorptiometry, dual- of standard and hydroxylated vitamin D.
photon absorptiometry, or dual x-ray absorptiometry in at Conclusions: Vitamin D decreases vertebral fractures and
least one of the following sites: femoral neck, total hip, tro- may decrease nonvertebral fractures. The available data are chanter, lumbar spine, total body, and the combined forearm, uninformative regarding the relative effects of standard and and reporting results on individual patients (as opposed to number of fractures); 5) the study was designed as a ran-domized control trial (RCT).
We included studies irrespective of whether calcium was added to vitamin D in the treatment or provided to the NUMBER OF groups have developed guidelines for the prevention and treatment of osteoporosis (1–3).
control group. We considered doses of vitamin D of no more Guidelines are only as strong as the evidence on which they than 100 IU daily to be negligible, and thus included studiesin which control patients received vitamin D in these lowdoses. We excluded studies that compared different types or Abbreviations: CI, Confidence interval; RCT, randomized control Guyatt et al. • Meta-Analyses of Osteoporosis Therapies Endocrine Reviews, August 2002, 23(4):560 –569 2. Study search and selection. The structured and tested Co- early postmenopausal women with bone density in the nor- chrane Collaborative approach for identifying RCTs, as de- mal or near normal range (prevention) vs. women with es- scribed by Dickersin et al. (5) and modified for the Cochrane Muscular Skeletal Group, guided our MEDLINE and EMBASE searches. We also conducted hand searches of bib- liographic references and the Cochrane Controlled Trials 7) level of calcium supplementation (Ͻ500 mg or Ͼ500 mg) Register and included all references in the Cochrane reviews 6. Statistical analysis. For fractures, we calculated a RR using update to September 2000 (5). We asked content experts to methods described by Fleiss (6). We constructed two-by-two identify published or unpublished relevant RCTs we had tables for both vertebral and nonvertebral fractures in each overlooked. Two reviewers (E.P., B.S.) examined each title study for which the data were available, and calculated the generated from the search and identified potentially eligible associated risk ratios. We tested for heterogeneity using a ␹2 articles for which we obtained the abstracts. For abstracts procedure (6). We tested whether our a priori hypotheses consistent with study eligibility, we obtained the full article could explain variability in the magnitude of treatment ef- fects across studies using a procedure described by Hedges 3. Methodological quality. We rated the methodological quality and Olkin (7). For study design, which had the four levels of each eligible study with respect to concealment of ran- described above (A, B, C, and D), we used the following domization; whether patients, caregivers, and those measur- planned orthogonal contrasts: A vs. [B, C, D]; [B, C] vs. D; ing outcome were blind to allocation; the extent of loss to follow-up; and whether the analysis was intention to treat.
We used analytic strategies similar to those for fracture We used more than one reviewer in the selection of studies, rates in examining the incidence of side effects and toxicity.
the assessment of methodological quality, and the extraction For each bone density site (lumbar spine, total body, com- of data. For all aspects of the review in which raters made bined hip, and combined forearm), we calculated the duplicate judgements, they resolved disagreements by weighted mean difference in bone density between treatment and control groups using the percentage change from base-line in the treatment and placebo groups and the associated 4. Data collection. Reviewers abstracted data regarding study sd values. We constructed regression models in which the design, patient characteristics, treatment duration, dosage, independent variables were year and dose and the depen- mean change, and sd values for bone density, and number dent variable the effect size, and we used this regression to of vertebral and nonvertebral fractures. For toxicity, we ex- determine the years across which pooling was appropriate.
amined the rate of withdrawal due to side effects and the rate To assess whether the magnitude of heterogeneity (differ- of withdrawal due to investigator-labeled adverse laboratory ences in apparent treatment effect across studies) was greater results. On most occasions, the adverse laboratory result than one might expect by chance, we conducted a test based was hypercalciuria. We sought key data that were missing on the ␹2 distribution with N-1 degrees of freedom, where N from the original reports through correspondence with the 5. A priori hypotheses regarding heterogeneity. To explore rea-sons for differences in results between studies (heterogene- 1. Search results. Electronic and hand searching resulted in the ity), we developed a priori hypotheses relating to the study retrieval of a total of 83 published papers that addressed the design, the methodological quality of the study, and the relationship between vitamin D and bone mineral density or study population. We describe these hypotheses below: fracture incidence (Fig. 1). Forty described RCTs (8 – 47). Rea- 1) We identified four study designs; given that calcium sons for excluding 15 of these trials (33– 47) were: 10 trials itself increases bone density relative to ordinary diet, we compared different types or doses of vitamin D, or studied anticipated that we would see the largest effects with trial combinations of vitamin D with other agents, without in- design A, intermediate effects with trial designs B and C, and cluding a control group that did not receive vitamin D (33–37, the smallest effect with trial design D.
39 – 43); 4 trials because trial duration was less than 1 yr (38, A) vitamin D and calcium supplementation vs. normal diet 44 – 46); and 1 trial because bone mineral density was mea- B) vitamin D alone vs. normal diet sured at the metacarpal site only (47). Thus, 25 RCTs fulfilled C) vitamin D combined with calcium supplementation vs. our eligibility criteria (Table 1) (8 –32).
Of the 25 trials included in this analysis, we had to contact D) vitamin D alone vs. calcium supplementation 10 authors for additional information (9 –11, 14, 16, 18, 19, 21, 2) whether the experimental intervention was standard 27, 31). Six investigators supplied the information we needed vitamin D or 25-OH vitamin D on the one hand, or hydroxy- lated vitamin D (1,25-OH vitamin D or calcitriol) on the other Table 1 describes these 25 studies in which a total of 4017 patients received some form of vitamin D and 4107 a con- 3) different methodological quality (randomization con- trolled intervention. Seventeen trials enrolled patients with cealed or unconcealed; blinded or unblinded; extent of loss decreased bone density; 10 used some form of standard vi- to follow-up; intention-to-treat analysis); tamin D, 14 hydroxylated vitamin D, and 1 trial had both a 4) primary prevention vs. secondary treatment, hypothe- standard and a hydroxylated vitamin D group in comparison sizing that the magnitude of the treatment effect may vary in to a control group (24). Follow-up ranged from 1 to 5 yr; loss Endocrine Reviews, August 2002, 23(4):560 –569 Guyatt et al. • Meta-Analyses of Osteoporosis Therapies FIG. 1. Search results for calcium/vitamin D review.
to follow-up was less than 10% in two studies, between 10 pooling of years and doses was determined by the regression and 20% in 8 studies, 20% or greater in 13, and unknown in analyses described in detail in Section I.
2 trials (10, 27). Eighteen trials were blinded (8, 9, 11–21, When sample size was adequate, the data showed large, 23–25, 30, 31), 5 trials were not (22, 26, 28, 29, 32), and the consistent, statistically significant effects of hydroxylated vi- blinding status was not clear in 2 of the trials (10, 27).
tamin D in all sites for all doses above 0.43 ␮g. The effect ofstandard vitamin D on bone density was consistently much 2. Fractures. Of the 25 eligible studies, 8 (total 1130 patients) smaller, and reached statistical significance only for lumbar measured the effect of vitamin D on morphometric vertebral spine at 1 yr and the femoral neck at final year. The difference fractures; all but 1 tested hydroxylated vitamin D. Rates of between standard and hydroxylated vitamin D was statis- vertebral fractures in the control groups varied from 1% to tically significant for total body (P ϭ 0.03) and for combined 58%. Figure 2 depicts the results of the individual studies and forearm (P ϭ 0.01) after the final year of treatment. Figure 4 the pooled estimates of the effect of vitamin D on vertebral depicts the results for combined forearm.
fractures, and Table 2 summarizes the pooled estimates. The For three of the analyses, there were large differences in pooled estimate indicates a 37% reduction in RR (95% CI results between trials reflected in small P values associated 0.45– 0.88) (Table 2). The point estimates from the individual with the formal test of heterogeneity. We found a number of trials are somewhat disparate, although the formal test of apparent contributing factors (Table 4). We have already heterogeneity did not reach conventional levels of statistical noted the differential impact by type of vitamin D. Contrary significance. None of the factors we identified in advance to our prediction of little difference between trial designs B explained the heterogeneity that does exist.
(vitamin D supplementation vs. normal diet) and C (vitamin Six studies (a total of 6187 patients) measured the effect of D and calcium supplementation vs. calcium supplementa- vitamin D on nonvertebral fractures. Fracture rates in the tion), the lumbar spine site did show significance. The result control group varied from 0% to 21%. Studies with standard was anomalous: standard vitamin D when compared with vitamin D enrolled far more patients then studies of hy- regular diet showed a substantial negative effect on bone droxylated vitamin D (Table 2). The pooled estimate suggests density, whereas when vitamin D and calcium were com- a RR reduction of 23%, but the CI includes a RR increase of pared with calcium alone there was no effect. However, this 4%. The studies show quite disparate results (Fig. 3), al- result was based upon the comparison of only two trials.
though the CIs are widely overlapping. The test of hetero- High vs. low levels of calcium supplementation yielded sig- geneity reaches our threshold for statistical significance for nificantly different effects for all three analyses presented in standard vitamin D and the combined results. None of the Table 4. For hydroxylated doses (0.50 –1.00 ␮g) at the com- factors we identified in advance, however, explained the bined forearm sites at yr 1–3, the effect of vitamin D was greater for patients receiving lower levels of calcium; but for For both fracture analyses, funnel plots showed no sug- lumbar spine and total body with standard doses at final year, the effect was greater for those receiving higher levels 3. Bone density. Table 3 summarizes the impact of vitamin D of calcium. Intention-to-treat analysis and loss-to-follow-up on bone mineral density at the four sites we examined. The results were also statistically significant for total body and Guyatt et al. • Meta-Analyses of Osteoporosis Therapies Endocrine Reviews, August 2002, 23(4):560 –569 300 IU Cholecalciferol vs. placebo 700 IU Cholecalciferol vs. placebo 0.75 ␮g ␣-Hydroxyvitamin D vs. 400 IU Cholecalciferol vs. placebo 400 IU Cholecalciferol vs. placebo 1 ␮g ␣-Calcidiol vs. control (B) 0.5 ␮g ␣-Cacidiol vs. placebo (C) 800 IU Cholecalciferol vs. placebo 0.5 ␮g Calcitriol vs. 1 g calcium a, Trial design: A ϭ Vitamin D (VD)ϩ Calcium (Ca) vs. Normal Diet (ND); B ϭ VD vs. ND; C ϭ VDϩ Ca vs. Ca; D ϭ VD vs. Ca.
b, Data not available/collected.
Endocrine Reviews, August 2002, 23(4):560 –569 Guyatt et al. • Meta-Analyses of Osteoporosis Therapies 0.62 ␮g Calcitriol vs. placebo (C) 1600 IU ␣-Calcidial vs. control (C) 0.43 ␮g Calcitriol vs. placebo (C) 1.0 ␮g ␣-Calcidiol vs. control (D) 0.5 ␮g or 1.0 ␮g ␣-Calcidiol vs. 0.5 ␮g ␣-Calcidiol vs. placebo (B) 0.5 ␮g Dihydroxycholecalciferol vs. 0.25 ␮g Dihydroxycholecalciferol vs. Guyatt et al. • Meta-Analyses of Osteoporosis Therapies Endocrine Reviews, August 2002, 23(4):560 –569 FIG. 2. RR with 95% CI for vertebral fractures after treatment with vitamin D.
TABLE 2. Weighted RR with 95% CI after treatment with vitamin D We interpreted P Յ 0.05 as indicating important between-study differences in results.
FIG. 3. RR with 95% CI for nonvertebral fractures after treatment with vitamin D.
combined forearm bone density. Larger effects were seen in hydroxylated vitamin D on forearm bone density. One trial, studies that did conduct an intention-to-treat analysis. The appreciably larger than the rest, showed a negligible effect of small number of studies makes inferences from these anal- hydroxylated vitamin D on forearm bone density. A number of small trials showed a substantial effect (Fig. 5). Although For all bone density analyses, we found only one instance by no means definitively demonstrating publication bias, suggesting publication bias, the investigation of the effect of these results do raise the possibility.
Endocrine Reviews, August 2002, 23(4):560 –569 Guyatt et al. • Meta-Analyses of Osteoporosis Therapies TABLE 3. Weighted mean difference of bone density after treatment with vitamin D FIG. 4. Weighted mean difference of distal forearm after treatment with vitamin D.
4. Side effects and toxicity. Our pooled estimate of the RR of Two issues of study design are particularly problematic.
discontinuing medication as a result of either symptomatic First, the methods of supplementation, and the use of cal- adverse effects or abnormal laboratory results from the 12 cium in addition to vitamin D, vary from study to study. In trials that reported such events was 1.37 (95% CI 1.01–1.88, the prior section in this series, we presented a meta-analysis P value 0.05, heterogeneity P value 0.99). The RR of with- suggesting that calcium alone increases bone density from drawal was similar in the trials of standard Vitamin D (RR 1.5% to 2%. We therefore anticipated that we might see 1.40, 95% CI, 0.94 to 2.06, p values 0.10) and hydroxylated largest effects when the intervention, but not the control vitamin D (RR 1.34, 95% CI 0.80 –2.24, P value 0.27), respec- group, received calcium in addition to vitamin D. We ex- tively (P value on the difference between the two estimates pected intermediate effects when calcium was withheld from, or offered to, both treatment and control groups, and smallest effects when calcium was given to only the controlarm.
Inferences from the results of these analyses are limited by The fracture data revealed quite a different pattern: the the variability in study designs, methodological weaknesses largest effect and most precise estimate of vitamin D effect in the primary studies (including lack of blinding in many on fracture came from a study in which only the control studies), the paucity of data, and the inconsistency of results.
group received calcium (21). Similarly, we failed to see the Guyatt et al. • Meta-Analyses of Osteoporosis Therapies Endocrine Reviews, August 2002, 23(4):560 –569 expected pattern in our examination of heterogeneity of find- ings in studies of bone density (Table 4). These puzzling results highlight the uncertainties regarding the effects of vitamin D that the studies to date have not resolved.
The second major issue in the possibility that vitamin D will have a different impact in different populations. We could not explore this issue adequately because studies typ- ically did not record baseline levels of vitamin D, the most likely explanation of heterogeneity of treatment effect across The variability in study results further limits any infer- ences one can make on the basis of the studies to date. We found statistically significant heterogeneity not only in a number of our bone density analyses, but also in the analysis of the effect of standard vitamin D, and the pooled analysis of all formulations, on nonvertebral fractures. Our a priori hypotheses failed to adequately explain this variability.
Nevertheless, combining across all trials, we found a sig- nificant effect of vitamin D in reducing vertebral fractures and a trend toward reduction in nonvertebral fractures (Ta- ble 2 and Figs. 2 and 3). The case for a biological mechanism for the vitamin D effect gains some strength from our anal- ysis of bone density, which suggested a positive impact on bone density at every site, particularly with hydroxylated The biology of standard and hydroxylated vitamin D is sufficiently different that one might be reluctant to pool in the first place (48). Both forms showed a similar effect on frac- tures. However, the confidence intervals around these effects are extremely wide, and it is quite possible that true effects differ greatly. Thus, the available data provide little guidance on the choice of vitamin D formulation.
In summary, secure inferences from the available random- ized trials of vitamin D are very limited. Vitamin D formu- lations probably reduce vertebral fractures. Their impact on nonvertebral fractures is uncertain. Moreover, the relative impact of different formulations on fracture rates, and the extent to which vitamin D effects vary in different popula- tions, is extremely uncertain. These issues offer potentially fruitful questions for subsequent investigation.
1. 1996 AACE Clinical practice guidelines for the prevention and treatment of postmenopausal osteoporosis. J Fla Med Assoc 83: 2. Kanis JA, Delmas P, Burckhardt P, Cooper C, Torgerson D 1997
Guidelines for diagnosis and management of osteoporosis. The Endocrine Reviews, August 2002, 23(4):560 –569 Guyatt et al. • Meta-Analyses of Osteoporosis Therapies European Foundation for Osteoporosis and Bone Disease. Osteo- controlled 2-year study in 315 normal females. Eur J Clin Invest 3. Scientific Advisory Board, Osteoporosis Society of Canada 1996
25. Geusens P, Dequeker J 1986 Long-term effect of nandrolone de-
Clinical practice guidelines for the diagnosis and management of canoate, 1␣ hydroxyvitamin D3, or intermittent calcium infusion osteoporosis. Scientific Advisory Board, Osteoporosis Society of therapy on bone mineral content, bone remodeling and fracture rate in symptomatic osteoporosis: a double-blind controlled study.
4. Gillespie WJ, Henry DA, O’Connell DL, Robertson J 1999 Vita-
min D and vitamin D analogues for preventing fractures associated 26. Shiraki M, Orimo H, Ito H, Akiguchi I, Nakao J, Takahashi R,
with involutional and post-menopausal osteoporosis. Cochrane Ishizuka S 1985 Long-term treatment of postmenopausal osteo-
porosis with active vitamin D3 1-␣-hydroxycholecalciferol and 1,24 5. Dickersin K, Scherer R, Lefebvre C 1994 Identifying relevant
dihydroxycholecalciferol. Endocrinol Jpn 32:305–315 studies for systematic reviews. BMJ 309:1286 –1291 27. Orimo H, Shiraki M, Hayashi T, Nakamura T 1987 Reduced
6. Fleiss JL 1993 The statistical basis of meta-analysis. Stat Methods
occurrence of vertebral crush fractures in senile osteoporosis treated with 1␣-vitamin D3. Bone Miner 3:47–52 7. Hedges LV, Olkin I 1985 Statistical methods for meta-analysis. San
28. Komulainen M, Kroger H, Tuppurainen MT, Heikkinen AM,
Alhava E, Honkanen R, Jurvelin J, Saarikoski S 1999 Prevention
8. Baeksgaard L, Anderson KP, Hyldstrup L 1998 Calcium and vi-
of femoral and lumber bone loss with hormone replacement ther- tamin D supplementation increase spinal BMD in healthy post- apy and vitamin D3 in early postmenopausal women: A popula- menopausal women. Osteoporos Int 8:255–260 tion-based 5-year randomized trial. J Clin Endocrinol Metab 84: 9. Chapuy MC, Arlot ME, Duboeuf F, Brun J, Crouzet B, Arnaud S,
Delmas PD, Meunier PJ 1992 Vitamin D3 and calcium to prevent
29. Shiraki M, Orimo H 1991 The effect of estrogen, and sex-steroids
hip fractures in elderly women. N Engl J Med 327:1637–1642 and thyroid hormone preparation on bone mineral density in senile 10. Christiansen C, Christiansen MS, Rodbro P, Hagen C, Transbol
osteoporosis: a comparative study of the effect of 1␣-hydroxy- I 1981 Effect of 1,25-dihydroxy-vitamin D3 in itself or combination
cholecalciferol on senile osteoporosis. Nippon Naibunpi Gakkai with hormone treatment in preventing postmenopausal osteopo- 30. Caniggia A, Delling G, Nuti R, Lore F, Vattimo A 1984 Clinical,
11. Dawson-Hughes B, Dallal GE, Krall EA, Harris S, Sokoll LJ,
biochemical and histological results of a double-blind trial with Falconer G 1991 Effect of vitamin D supplementation on winter-
1,25-dihydroxyvitamin D3, estradiol and placebo in post-meno- time and overall bone loss in healthy postmenopausal women. Ann pausal osteoporosis. Acta Vitaminol Enzymol 6:117–128 31. Thomsen K, Riis B, Christiansen C 1986 Effect of estrogen/ge-
12. Gallagher JC, Goldgar D 1990 Treatment of postmenopausal os-
stagen and 24R, 25-dihydroxyvitamin D3 therapy on bone forma- teoporosis and high doses of synthetic calcitriol. Ann Intern Med tion in postmenopausal women. J Bone Miner Res 1:503–507 32. Chen JT, Shiraki M, Hasumi K, Tanaka N, Katase K, Kato T, Hirai
13. Jensen GF, Christiansen C, Transbol I 1982 Treatment of post-
Y, Nakamura T, Ogata E 1997 1-␣ Hydroxyvitamin D3 treatment
menopausal osteoporosis. A controlled therapeutic trial comparing decreases bone turnover and modulates calcium-regulating hor- oestorogen/gestgen, 1,25-dihydroxy-vitamin D3 and calcium. Clin mones in early postmenopausal women. Bone 20:557–562 33. Falch JA, Odegaard OR, Finnanger AM, Matheson I 1987 Post-
14. Lips P, Wilco C, Graafmans MS, Ooms ME, Bezemer PD, Bouter
menopausal osteoporosis: no effect of three years treatment with LM 1996 Vitamin D supplementation and fracture incidence in
1,25-dihydroxycholecalciferol. Acta Medica Scand 221:199 –204 elderly persons. Ann Intern Med 124:400 – 406 34. Arthur RS, Piraino B, Candib D, Cooperstein L, Chen T, West C,
15. Menczel J, Foldes J, Steinberg R, Leichter I, Shalita B, Bdolah-
Puschett J 1990 Effect of low-dose calcitriol and calcium therapy on
Abram T, Kadosh S, Mazor Z, Ladkani D 1994 Alfacalcidol (␣D3)
bone histomorphometry and urinary calcium excretion in osteo- and calcium in osteoporosis. Clin Orthop 300:241–247 porotic women. Miner Electrolyte Metab 16:385–390 16. Ooms ME, Roos JC, Bezemer PD, van der Vijgh WJ, Bouter LM,
35. Lyritis GP, Androulakis C, Magiasis B, Charalambaki Z, Tsakala-
Lips P 1995 Prevention of bone loss by vitamin D supplementation
kos N 1994 Effect of nandrolone decontae and 1-␣-hydroxy-
in elderly women: a randomized double-blind trial. J Clin Endo- calciferol on patients with vertebral osteoporosis collapse: a dou- ble-blind clinical trial. Bone Miner 27:209 –217 17. Orimo H, Shiraki M, Hayashi Y, Hoshino T, Onaya T, Miyazaki
36. Eriksson SA, Lindgren JU 1993 Combined treatment with calci-
S, Kurosawa H, Nakamura T, Ogawa N 1994 Effects of 1␣-
tonin and 1,25-dihydroxyvitamin D3 for osteoporosis in women.
hydroxyvitamin D3 on lumbar bone mineral density and vertebral fractures in patients with postmenopausal osteoporosis. Calcif Tis- 37. Caniggia A, Lore F, Nuti R, Martini G, Frediani B, DiCairano G
1992 Role of the active vitamin D metabolite and 1␣-hydroxylated 18. Orwoll ES, McClung MR, Oviatt SK, Recker RR, Weigel RM 1989
analogs in the treatment of postmenopausal osteoporosis. J Nutr Sci Histomorphometric effects of calcium or calcium plus 25- hydroxyvitamin D3 therapy in senile osteoporosis. J Bone Miner 38. Gallagher JC, Jerpbak CM, Jee WSS, Johnson KA, DeLuca HF,
Riggs BL 1982 1,25-Dihydroxyvitamin D3: short- and long-term
19. Ott SM, Chesnut III CH 1989 Calcitriol treatment is not effective
effects on bone and calcium metabolism in patients with post- in postmenopausal osteoporosis. Ann Intern Med 110:267–274 menopausal osteoporosis. Proc Natl Acad Sci USA 79:3325–3329 20. Riis BJ, Thomsen K, Christiansen C 1986 Does 25R,25(OH)2-
39. Masud T, Mulcahy B, Thompson AV, Donnelly S, Keen RW,
vitamin D3 prevent postmenopausal bone loss? Calcif Tissue Int Doyle DV, Spector TD 1998 Effects of cyclical etidronate combined
with calcitriol versus cyclical etidronate alone on spine and femoral 21. Tilyard MW, Speanrs GF, Thomsen J, Dovey S 1992 Treatment of
neck bone mineral density in postmenopausal osteoporotic postmenopausal osteoporosis with calcitriol or calcium. N Engl 40. Aloia JF, Vaswani A, Yeh JK, Ellis K, Yasumura S, Cohn SH 1988
22. Ushiroyama T, Okamura S, Ikeda A, Ueki M 1995 Efficacy of
Calcitriol in the treatment of postmenopausal osteoporosis. Am J ipriflavone and 1␣ vitamin D therapy for the cessation of vertebral bone loss. Int J Gynecol Obstet 48:283–288 41. Aloia JF, Vaswani A, Yeh JK, Ross PL, Flaster E, Dilmanian A 1994
23. Dawson-Hughes B, Harris S, Krall E, Dallal GE 1997 Effect of
Calcium supplementation with and without hormone replacement calcium and vitamin D supplementation on bone density in men therapy to prevent postmenopausal bone loss. Ann Intern Med and women 65 years of age or older. N Engl J Med 337:670 – 676 24. Christiansen C, Christiansen M, McNair P, Hagen C, Stocklund
42. Recker RR, Davies KM, Dowd RM, Heaney RP 1999 The effect of
K, Trasnbol IB 1980 Prevention of early postmenopausal bone loss:
low-dose continuous estrogen and progesterone therapy with cal- Guyatt et al. • Meta-Analyses of Osteoporosis Therapies Endocrine Reviews, August 2002, 23(4):560 –569 cium and vitamin D on bone in elderly women: a randomized, 1983 Calcium, vitamin D and anabolic steroid in treatment of aged controlled trial. Ann Intern Med 130:897–904 bones: double-blind placebo-controlled long-term clinical trial. Age 43. Dawson-Hughes B, Harris SS, Krall EA, Dallal GE, Falconer G,
Green CL 1995 Rates on bone loss in postmenopausal women
46. Veijo Hoikka, EM, Alhava AA, Paavo K, Veikko R 1980 Treatment
randomly assigned to one of two dosages of vitamin D. Am J Clin of osteoporosis with 1-␣-hydroxycholecalciferol and calcium. Acta 44. Grady D, Halloran B, Cummings S, Leveille S, Wells L, Black D,
47. Nordin BE, Baker MR, Horsman A, Peacock M 1985 A prospective
Byl N 1991 1,25-Dihydroxyvitamin D3 and muscle strength in
trial of the effect of vitamin D supplementation on metacarpal bone elderly: a randomized controlled trial. J Clin Endocrinol Metab loss in elderly women. Am J Clin Nutr 42:470 – 474 48. DeLuca HF 1988 The vitamin D story: a collaborative effort of basic
45. Inkovaara J, Gothoni G, Halttula R, Heikinheimo R, Tokola O
science and clinical medicine. FASEB J 2:224 –236

Source: http://dpi.dipa.co.id/xplod/prodart/vit%20d%20meta%20analysis.pdf