Russiantortoise.net

Vitamin D Supplementation and Total Mortality
A Meta-analysis of Randomized Controlled Trials
Philippe Autier, MD; Sara Gandini, PhD Background: Ecological and observational studies sug-
varied from 300 to 2000 IU. The trial size–adjusted mean gest that low vitamin D status could be associated with higher daily vitamin D dose was 528 IU. In 9 trials, there was a 1.4- mortality from life-threatening conditions including cancer, to 5.2-fold difference in serum 25-hydroxyvitamin D be- cardiovascular disease, and diabetes mellitus that account tween the intervention and control groups. The summary for 60% to 70% of total mortality in high-income countries.
relative risk for mortality from any cause was 0.93 (95% con- We examined the risk of dying from any cause in subjects fidence interval, 0.87-0.99). There was neither indication who participated in randomized trials testing the impact of for heterogeneity nor indication for publication biases. The vitamin D supplementation (ergocalciferol [vitamin D2] or summary relative risk did not change according to the ad- cholecalciferol [vitamin D3]) on any health condition.
dition of calcium supplements in the intervention.
Conclusions: Intake of ordinary doses of vitamin D supple-
Methods: The literature up to November 2006 was
ments seems to be associated with decreases in total mor- searched without language restriction using the following tality rates. The relationship between baseline vitamin D databases: PubMed, ISI Web of Science (Science Citation status, dose of vitamin D supplements, and total mortal- Index Expanded), EMBASE, and the Cochrane Library.
ity rates remains to be investigated. Population-based,placebo-controlled randomized trials with total mortal- Results: We identified 18 independent randomized con-
ity as the main end point should be organized for confirm- trolled trials, including 57 311 participants. A total of 4777 deaths from any cause occurred during a trial size–adjustedmean of 5.7 years. Daily doses of vitamin D supplements Arch Intern Med. 2007;167(16):1730-1737 ECOLOGICALSTUDIESINNORTH productsmayalsorepresentasourceofvi-
tamin D, although of highly variable content (eg, fortified foods, oily fish, eggs, and but- ter). Hence, low vitamin D status could pro- ceed from the conjunction of insufficient in- cer, cardiovascular diseases, and diabetes takes (exogenous source) and of insufficient mellitus would increase with increasing lati- skin synthesis (endogenous source) of vita- tude, that is, with residence increasingly dis- min D. Biological findings have reinforced tant from the equator.1,2 Other studies have the likelihood of the vitamin D hypothesis.
shown that the survival of patients with car- First, vitamin D receptors have been found diovascular disease or with some cancer (eg, in various organs, and activation of these re- greaterifthediagnosiswasmadeduringsum-mer as compared with the winter.3,4 Increas- For editorial comment
ing distance from the equator and winter pe- see page 1709
riod were equated to decreasing exposure tosunlight, especially to UV-B radiation (280- CME available online at
315 nm) because with increasing latitude, www.archinternmed.com
amountsofUV-BradiationreachingtheearthsurfacedecreasefasterthanamountsofUV-A (calcitriol), the physiologically active form radiation(315-400nm).5Also,seasonalvaria- of vitamin D, induces cell differentiation and tions are more pronounced for UV-B radia- inhibitsproliferation,invasiveness,angiogen- tion than for the UV-A radiation.5 Because esis, and metastatic potential.11,12 These bio- UV-B radiation is necessary for the synthe- logical phenomena are typical of cancer gen- Author Affiliations:
sis of vitamin D in the skin, it has been hy- pothesized that associations found between proliferation) are also involved in cardiovas- latitudeorseasonalityandmortalityfromsev- cular ischemic diseases. Second, many tis- France (Dr Autier); and theEuropean Institute of Oncology, eral chronic conditions could be owing to sues express the 1␣-hydroxylase enzyme.11 variations in vitamin D status.6-10 Some food So, after 25-hydroxylation of vitamin D in ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
tervals (CIs) (ie, crude data or adjusted were used: vitamin D, cholecalciferol, and RRs and standard errors, 95% CIs, or P ergocalciferol. For methods, the follow- vitamin D intake vs placebo or control.
MeSH terms used were randomized con- trolled trial and placebo. A first general DESCRIPTION OF STUDIES
RETRIEVED
A total of 992 articles or abstracts were re- trieved and checked for relevance in terms words with the following outcome key-words (and their corresponding MeSH of intervention, design, and reporting of terms): congestive heart failure, coro- mortality data. This process resulted in re- nary heart disease, cardiovascular dis- trieving a total of 27 articles or abstracts ease, fracture, bone mineral density, and bone turnover. Mortality was not a help- ized clinical trials evaluating effects of vi- point and reporting data on deaths. Ofthese 27 articles, 9 were not included in 1 trial in the United Kingdom,16 had mor- sons: (1) Two articles referred to the same trial.17,18 (2) Three did not report deaths erences cited in the selected articles and trieved.19-21 (3) In 2 trials, the interven-tion consisted of a set of drugs including stract or article whose title or summarycontained at least 1 intervention key- vitamin D.22,23 (4) Two trials were basedon cluster randomization,24,25 and 1 of synthesis of randomized controlled trials cluding 3717 participants with a mean age SELECTION OF ARTICLES
of 85 years. The intervention was equiva-lent to a daily dose of 1100 IU of ergocal- INTERVENTION
ciferol. (5) A placebo-controlled random- AND OUTCOME
label trial with a subgroup of the placebo- ation of Calcium Or vitamin D) trial.28 We used the data from the open-label trial and trial to have independent studies. For the estimates derived from the same trial.
Table 1 summarizes the 18 studies that
calcidol), the physiologically active form D3 [calcitriol]), or other vitamin D ana- STATISTICAL ANALYSIS
end-stage renal disease or in patients un- rates in each randomization group were all (intent-to-treat analysis). Some trials, such LITERATURE SEARCH
as the RECORD trial,28 had a factorial de- ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
Table 1. Vitamin D Supplements and All-Cause Mortality: Overview of Trials Selected for Meta-analysis
Baseline,
in Control
Follow-up,
Point(s)
Population
Intervention
(300 IUϩcalcium [0.5 g]during 3 first years and 100IUϩcalcium [0.5 g] in thelast year) 1 group with single-injectionergocalciferol(300 000 IU)ϩoral calcium(1 g), 1 group with daily oralcholecalciferol (800IUϩcalcium [1 g]) (10 000 IU), followed by dailyoral ergocalciferol (1000 IU) a Women randomized to multivitamin supplement containing vitamin D were not included in the meta-analysis.
b Women randomized to hormone therapy or to hormone therapy and vitamin D groups were not included in the meta-analysis.
c Cod liver oil without cholecalciferol.
d The same article reported 2 randomized controlled trials. We took into account only the open label trial because the placebo-controlled trial was a part of the RECORD trial.28 Mortality data of the open label trial we used were those reported by Avenell et al15 in 2005.
e Intervention assumed to be the same as in the RECORD Trial.28 ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
Table 2. Serum Levels of 25-Hydroxyvitamin D in Randomized Trials With Vitamin D Supplementsa
Mean Serum 25-Hydroxyvitamin D3 (ng/mL)b
Daily Dose
of Vitamin D
Ratio for In-Study Level,
Intervention Group
Control Group
Follow-up,
in Intervention
Intervention vs
Group, IU
Baseline
Baseline
Control Group
a Trials in Table 1 not included in Table 2 did not report serum 25-hydroxyvitamin D levels.
b Measurements of serum levels were always performed in subsamples of subjects in intervention and control groups.
c In-study serum 25-hydroxyvitamin D levels were derived from Figure 1 in the original publication.33 d Estimated from oral cholecalciferol, 100 000 IU every 4 mo.
studies that is attributable to heterogeneity coming from the “intervention group” and data related to groups not receiving vita- ments in the largest trials (see “Trials tality was a relatively rare event, and we lication bias: the sensitivity analysis of Figure) was 48% in the RECORD
tween the various measures of relative risk Copas and Shi47 and the funnel plot regres- (ie, odds ratio, rate ratio, and risk ratio).
Initiative trial,42,43 63% in the trial by the corresponding variance using the for- mula proposed by Greenland44 in 1987.
culated them from tabular data, and weused the Woolf formula to evaluate the trial by Lips et al,30 and 95% in the trial standard error of the log odds ratio.44 Logit 0.5 in every cell of those tables that con- tality across selected trials was computed results of the 18 trials indicated a sig- as a summary RR (SRR) with 95% CIs. TheSRR was considered statistically signifi- cant if the 95% CI did not include 1.0.
effect across studies using the large sample test based on the ␹2 statistic. Since the ␹2 for heterogeneity (P=.52) or of pub- test has limited power, we considered that lication bias (P=.37 with the method statistically significant heterogeneity ex- was 528 IU. Table 2 indicates a sub-
of Copas and Shi47 and P=.77 with the isted when the P value was Յ.10.45 Sub- A subgroup analysis (Table 3)
heterogeneity focusing on type of study,type of control, length of follow-up, vita- groups, translating to a 1.4- to 5.2-fold using the I2 parameter, which represents ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
Subtotal SRR (95% CI)
0.92 (0.86-0.99)
Subtotal SRR (95% CI)
1.15 (0.79-2.73)
All trials
2330/28 500
2447/28 811
0.93 (0.87-0.99)
χ2 Test for heterogeneity: P = .52 Figure. Meta-analysis of data on all-cause mortality in 18 randomized controlled trials with vitamin D. SRR indicates summary relative risk.
Table 3. Vitamin D Supplements and All-Cause Mortality:
Subgroup and Sensitivity Analysis
on all-cause mortality is not likely be-cause total mortality did not consti- No. of Trials
Parameter, Heterogeneity,
Variable
Meta-Analysis
P Value
18 trials included in the meta-analysis except 1.16 Timing of deaths calcium supplements. No relation-ship was found with dose of vita- Abbreviation: CI, confidence interval; SRR, summary relative risk.
The I2 parameter represents the percentage of total variation across studies that is attributable to b No placebo for vitamin D in the control group.
tively narrow (ie, 400-830 IU), andlarge variations in size of trials andin compliance to interventions pre- neity. In this respect, exclusion of this ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
have less of an effect (or not at all) on probability to die because of falls.
P o p u l a t i o n - b a s e d , p l a c e b o - Accepted for Publication: April 18,
Correspondence: Philippe Autier,
Author Contributions: Drs Autier
sibility for the integrity of the data and Study concept and design: Autier. Ac- quisition of data: Autier and Gandini.
Analysis and interpretation of data: Autier and Gandini. Drafting of the manuscript: Autier. Statistical analy- sis: Gandini. Administrative, techni- cal, and material support: Autier. Study Financial Disclosure: None re-
en’s Health Initiative trial, rate ratios Additional Contributions: The li-
ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
20. Graafmans WC, Ooms ME, Hofstee HM, Beze- ized, controlled trial of quadriceps resistance ex- mer PD, Bouter LM, Lips P. Falls in the elderly: a ercise and vitamin D in frail older people: the Frailty prospective study of risk factors and risk profiles.
Interventions Trial in Elderly Subjects. J Am Geri- 1. Grant WB. Ecologic studies of solar UV-B radia- Am J Epidemiol. 1996;143(11):1129-1136.
tion and cancer mortality rates. Recent Results 21. Keane EM, Healy M, O’Moore R, Coakley D, Walsh 36. HarwoodRH,SahotaO,GaynorK,etal.Arandomised, Cancer Res. 2003;164:371-377.
JB. Vitamin D–fortified liquid milk: benefits for the controlled comparison of different calcium and vi- 2. Zittermann A, Schleithoff SS, Koerfer R. Putting elderly community-based population. Calcif Tis- tamin D supplementation regimens in elderly wom- cardiovascular disease and vitamin D into en after hip fracture: the Nottingham Neck of Femur perspective. Br J Nutr. 2005;94(4):483-492.
22. Hedstro¨m M, Sjoberg K, Brosjo E, Astrom K, (NoNOF) Study. Age Ageing. 2004;33(1):45-51.
3. Scragg R. Seasonality of cardiovascular disease Sjoberg H, Dalen N. Positive effects of anabolic 37. Meier C, Woitge HW, Witte K, Lemmer B, Seibel mortality and the possible protective effect of ultra- steroids, vitamin D and calcium on muscle mass, MJ. Supplementation with oral vitamin D3 and cal- violet radiation. Int J Epidemiol. 1981;10(4): bone mineral density and clinical function after a cium during winter prevents seasonal bone loss: hip fracture: a randomised study of 63 women.
a randomized controlled open-label prospective 4. Lim HS, Roychoudhuri R, Peto J, Schwartz G, J Bone Joint Surg Br. 2002;84(4):497-503.
trial. J Bone Miner Res. 2004;19(8):1221-1230.
Baade P, Moller H. Cancer survival is dependent 23. Sato Y, Kanoko T, Satoh K, Iwamoto J. The preven- 38. Brazier M, Grados F, Kamel S, et al. Clinical and labo- on season of diagnosis and sunlight exposure. Int tion of hip fracture with risedronate and ergocal- ratory safety of one year’s use of a combination J Cancer. 2006;119(7):1530-1536.
ciferol plus calcium supplementation in elderly wom- calcium ϩ vitamin D tablets in ambulatory elderly 5. IARC. Solar and Ultraviolet Radiation. Vol 55. Lyon, en with Alzheimer disease: a randomized controlled women with vitamin D insufficiency: results of a France: International Agency for Research on Can- trial. Arch Intern Med. 2005;165(15):1737-1742.
multicenter, randomized, double-blind, placebo- cer; 1992. IARC Monographs on the Evaluation 24. Law M, Withers H, Morris J, Anderson F. Vita- controlled study. Clin Ther. 2005;27(12):1885- min D supplementation and the prevention of frac- 6. Giovannucci E. The epidemiology of vitamin D and tures and falls: results of a randomised trial in el- 39. Porthouse J, Cockayne S, King C, et al. Ran- cancer incidence and mortality: a review (Unites derly people in residential accommodation. Age domised controlled trial of calcium and supple- States). Cancer Causes Control. 2005;16(2): mentation with cholecalciferol (vitamin D3) for pre- 25. Larsen ER, Mosekilde L, Foldspang A. Vitamin D vention of fractures in primary care. BMJ. 2005; 7. Kricker A, Armstrong B. Does sunlight have a ben- and calcium supplementation prevents osteopo- eficial influence on certain cancers? Prog Bio- rotic fractures in elderly community dwelling resi- 40. Flicker L, MacInnis R, Stein M, et al. Should all older phys Mol Biol. 2006;92(1):132-139.
dents: a pragmatic population-based 3-year in- people in residential care receive vitamin D to pre- 8. Garland CF, Garland FC, Gorham ED, et al. The role tervention study. J Bone Miner Res. 2004;19 vent falls? results of a randomised trial [abstract].
of vitamin D in cancer prevention. Am J Public J Bone Miner Res. 2004;19(suppl 1):S99.
26. Corless D, Dawson E, Fraser F, et al. Do vitamin 41. Schleithoff SS, Zittermann A, Tenderich G, Ber- 9. Poole KE, Loveridge N, Barker PJ, et al. Reduced D supplements improve the physical capabilities thold HK, Stehle P, Koerfer R. Vitamin D supple- vitamin D in acute stroke. Stroke. 2006;37(1): of elderly hospital patients. Age Ageing. 1985; mentation improves cytokine profiles in patients with congestive heart failure: a double-blind, ran- 10. Grimes DS. Are statins analogues of vitamin D? 27. Avenell A, Grant AM, McGee M, McPherson G, domized, placebo-controlled trial. Am J Clin Nutr.
Lancet. 2006;368(9529):83-86.
11. Ordonez-Moran P, Larriba MJ, Pendas-Franco N, agement Group. The effects of an open design on 42. Jackson RD, LaCroix AZ, Gass M, et al; Women’s et al. Vitamin D and cancer: an update of in vitro and trial participant recruitment, compliance and re- Health Initiative Investigators. Calcium plus vita- in vivo data. Front Biosci. 2005;10:2723-2749.
tention—a randomized controlled trial compari- min D supplementation and the risk of fractures. N 12. Banerjee P, Chatterjee M. Antiproliferative role of son with a blinded, placebo-controlled design. Clin Engl J Med. 2006;354(7):669-683.
vitamin D and its analogs—a brief overview. Mol 43. Wactawski-Wende J, Kotchen JM, Anderson GL, Cell Biochem. 2003;253(1-2):247-254.
28. Grant AM, Avenell A, Campbell MK, et al; RECORD et al; Women’s Health Initiative Investigators.
13. National Center for Health Statistics. Health, United Trial Group. Oral vitamin D3 and calcium for sec- Calcium plus vitamin D supplementation and the States, 2006, With Chart Book on Trends in the ondary prevention of low-trauma fractures in el- risk of colorectal cancer. N Engl J Med. 2006; Health of Americans. Hyattsville, MD; National Cen- derly people: a randomised placebo-controlled trial.
Lancet. 2005;365(9471):1621-1628.
44. Greenland S. Quantitative methods in the review 14. Niederlaender E. Causes of Death in the EU. Lux- 29. Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 of epidemiologic literature. Epidemiol Rev. 1987; embourg City, Luxembourg: Eurostat; October and calcium to prevent hip fractures in elderly 2006. Population and Social Conditions— women. N Engl J Med. 1992;327(23):1637-1642.
45. Sterne JA, Juni P, Schulz KF, Altman DG, Bartlett 30. Lips P, Graafmans WC, Ooms ME, Bezemer PD, C, Egger M. Statistical methods for assessing the 15. Avenell A, Gillespie WJ, Gillespie LD, O’Connell Bouter LM. Vitamin D supplementation and frac- influence of study characteristics on treatment ef- DL. Vitamin D and vitamin D analogues for pre- ture incidence in elderly persons. Ann Intern Med.
fects in “meta-epidemiological” research. Stat Med.
venting fractures associated with involutional and post-menopausal osteoporosis. Cochrane Data- 31. Baeksgaard L, Andersen KP, Hylstrup L. Calcium 46. Higgins JPT, Thompson SG. Quantifying hetero- base Syst Rev. 2005;(3):CD000227.
and vitamin supplementation increases spinal BMD geneity in a meta-analysis. Stat Med. 2002;21 16. Trivedi DP, Doll R, Khaw KT. Effect of four monthly in healthy, postmenopausal women. Osteoporos oral vitamin D3 (cholecalciferol) supplementa- 47. Copas JB, Shi JQ. A sensitivity analysis for pub- tion on fractures and mortality in men and women 32. Krieg MA, Jacquet AF, Bremgartner M, Cuttelod lication bias in systematic reviews. Stat Methods living in the community: randomized double blind S, Thiebaud D, Burckhardt P. Effect of supple- controlled trial. BMJ. 2003;326(7387): mentation with vitamin D3 and calcium on quan- 48. Macaskill P, Walter SD, Irwig L. A comparison of titative ultrasound of bone in elderly institution- m e t h o d s t o d e t e c t p u b l i c a t i o n b i a s i n 17. Komulainen MH, Kroger H, Tuppurainen MT, et al.
alized women: a longitudinal study. Osteoporos meta-analysis. Stat Med. 2001;20(4):641-654.
HRT and Vit D in prevention of non-vertebral frac- 49. Bischoff-Ferrari HA, Dawson-Hughes B, Willett WC, tures in postmenopausal women: a 5 year ran- 33. Chapuy MC, Pamphile R, Paris E, et al. Com- et al. Effect of vitamin D on falls: a meta-analysis.
domized trial. Maturitas. 1998;31(1):45-54.
bined calcium and vitamin D3 supplementation in JAMA. 2004;291(16):1999-2006.
18. Komulainen M, Kroger H, Tuppurainen MT, et al.
elderly women: confirmation of reversal of sec- 50. Bouillon R, Moody T, Sporn M, Barrett JC, Nor- Prevention of femoral and lumbar bone loss with ondary hyperparathyroidism and hip fracture risk: man AW. NIH deltanoids meeting on vitamin D and hormone replacement therapy and vitamin D3 in the Decalyos II Study. Osteoporos Int. 2002; cancer: conclusion and strategic options. J Ste- early postmenopausal women: a population based roid Biochem Mol Biol. 2005;97(1-2):3-5.
5-year randomized trial. J Clin Endocrinol Metab.
34. Meyer HE, Smedshaug GB, Kvaavik E, Falch JA, 51. Stolzenberg-Solomon RZ, Vieth R, Azad A, et al.
Tverdal A, Pedersen JI. Can vitamin D supple- A prospective nested case-control study of vita- 19. Bischoff HA, Stahelin HB, Dick W, et al. Effects of mentation reduce the risk of fracture in the el- min D status and pancreatic cancer risk in male vitamin d and calcium supplementation on falls: derly? a randomized controlled trial. J Bone Miner s m o k e r s . C a n c e r R e s . 2 0 0 6 ; 6 6 ( 2 0 ) : a randomized controlled trial. J Bone Miner Res.
35. Latham NK, Anderson CS, Lee A, et al. A random- 52. Palmieri C, MacGregor T, Girgis S, Vigushin D.
ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.
Serum 25 hydroxyvitamin D levels in early and ad- 55. Lee IM, Cook NR, Gaziano JM, et al. Vitamin E in 58. Pe´rez-Castrillo´n JL, Vega G, Abad L, et al. Effects vanced breast cancer. J Clin Pathol. 2006;59 the primary prevention of cardiovascular disease of atorvastatin on vitamin D levels in patients with and cancer: the Women’s Health Study: a ran- acute ischemic heart disease. Am J Cardiol. 2007; 53. Giovannucci E, Liu Y, Rimm EB, et al. A prospec- domized controlled trial. JAMA. 2005;294(1): t i v e S t u d y o f p r e d i c t o r s o f v i t a m i n D 59. Pittas AG, Dawson-Hughes B, Li T, et al. Vitamin D status and cancer incidence and mortality 56. Bouillon R, Eelen G, Verlinden L, Mathieu C, Car- in men. J Natl Cancer Inst. 2006;98(7): meliet G, Verstuyf A. Vitamin D and cancer. J Ste- diabetes in women. Diabetes Care. 2006;29(3): roid Biochem Mol Biol. 2006;102(1-5): 54. Bjelakovic G, Nikolova D, Simonetti RG, Gluud C.
60. Barbeau WE, Bassaganya-Riera J, Hontecillas R.
Antioxidant supplements for prevention of gas- 57. Penda´s-Franco N, Gonzalez-Sancho JM, Suarez Putting the pieces of the puzzle together—a se- trointestinal cancers: a systematic review and Y, et al. Vitamin D regulates the phenotype of hu- ries of hypotheses on the etiology and pathogen- meta-analysis. Lancet. 2004;364(9441):1219- man breast cancer cells. Differentiation. 2007; esis of type 1 diabetes. Med Hypotheses. 2007; Error in Figure. In the Original Investigation by Fito´ et al titled
“Effect of a Traditional Mediterranean Diet on Lipoprotein Oxi-
dation: A Randomized Controlled Trial” published in the June
11, 2007, issue of the ARCHIVES (2007;167[11]:1195-1203), an
error occurred in Figure 2 wherein the y-axis labels in parts
A and C were mistakenly transposed. A corrected figure and
legend appears below.
Figure 2. Mean±SD changes in plasma ␣-linolenic acid (A), urinary tyrosol (B),
and hydroxytyrosol (C) after 3-month interventions. *PϽ.05 vs the
corresponding baseline. †PϽ.05 vs low-fat diet group. ‡PϽ.05 vs TMDϩnuts
group. TMD indicates traditional Mediterranean diet; VOO, virgin olive oil.
ARCH INTERN MED/ VOL 167 (NO. 16), SEP 10, 2007 2007 American Medical Association. All rights reserved.

Source: http://www.russiantortoise.net/vitamin_d/Vit%20D%20and%20total%20mortality%207.09.pdf