Visolie-info.nl

Effect of Different Antilipidemic Agents
and Diets on Mortality
A Systematic Review
Marco Studer, MD; Matthias Briel, MD; Bernd Leimenstoll, MD; Tracy R. Glass, MSc; Heiner C. Bucher, MD, MPH Background: Guidelines for the prevention and treat-
CI, 0.91-1.11), 0.84 for resins (95% CI, 0.66-1.08), 0.96 ment of hyperlipidemia are often based on trials using for niacin (95% CI, 0.86-1.08), 0.77 for n-3 fatty acids combined clinical end points. Mortality data are the most (95% CI, 0.63-0.94), and 0.97 for diet (95% CI, 0.91- reliable data to assess efficacy of interventions. We aimed 1.04). Compared with control groups, risk ratios for car- to assess efficacy and safety of different lipid-lowering diac mortality indicated benefit from statins (0.78; 95% interventions based on mortality data.
CI, 0.72-0.84), resins (0.70; 95% CI, 0.50-0.99) and n-3fatty acids (0.68; 95% CI, 0.52-0.90). Risk ratios for non- Methods: We conducted a systematic search of ran-
cardiovascular mortality of any intervention indicated no domized controlled trials published up to June 2003, com- association when compared with control groups, with the paring any lipid-lowering intervention with placebo or exception of fibrates (risk ratio, 1.13; 95% CI, 1.01- usual diet with respect to mortality. Outcome measures were mortality from all, cardiac, and noncardiovascularcauses.
Conclusions: Statins and n-3 fatty acids are the most fa-
vorable lipid-lowering interventions with reduced risks
Results: A total of 97 studies met eligibility criteria, with
of overall and cardiac mortality. Any potential reduc- 137 140 individuals in intervention and 138 976 indi- tion in cardiac mortality from fibrates is offset by an in- viduals in control groups. Compared with control groups, creased risk of death from noncardiovascular causes.
risk ratios for overall mortality were 0.87 for statins (95%confidence interval [CI], 0.81-0.94), 1.00 for fibrates (95% LIPID-LOWERINGAGENTSARE theefficacyofthesedrugsinvariousrisk
groups and settings as well as in generally analyses of randomized controlled trials are drugs with proven efficacy to lower both car- benefit of interventions and to explore effect diovascular morbidity and overall mortal- ity in a large-scale clinical trial were 3-hy- present meta-analysis is to investigate the reductase inhibitors (statins).1 In previous meta-analyses, only statins showed statis- lowering interventions in the primary and tically significant and clinically relevant re- Author Affiliations: Basel
and overall mortality.2,3 In addition to the potent lipid-lowering capacity of statins, more recent findings indicate that the posi- tive effects of statins could also be the re- SEARCH FOR RELEVANT STUDIES
sult of reductions in platelet aggregability We included references from previous meta- Over the past 5 years, large trials of sev- analyses2,6 and 2 of us (M.S. and M.B.) searched eral statins and other lipid-lowering inter- Financial Disclosure: None.
ventions provided important information on Cochrane Controlled Trials Register together (REPRINTED) ARCH INTERN MED/ VOL 165, APR 11, 2005 2005 American Medical Association. All rights reserved.
with a professional librarian to identify s e e h t t p : / / w w w . b i c e . c h / e n g l all benefit of lipid-lowering interven-tions were overall mortality and cardiac STUDY SELECTION
AND DATA ABSTRACTION
farction, sudden death, or heart failure) meta-analysis if they compared any lipid- STATISTICAL ANALYSIS
a follow-up of at least 6 months, and re- ported mortality data. We excluded trials recipients; trials in coronary artery by- groups by using a random effects model.
vention2,7); trials using any combination of lipid-lowering intervention (not allow- Cochran Q test and measured inconsis- tency (I2; the percentage of total varia- 1%-24%) (Table 1).
drug); and trials with outdated interven- tion across studies that is due to hetero- OVERALL MORTALITY
tails of included and excluded trials are effects across different lipid-lowering in- calculating a z score, the difference in quality blinded to one another’s rating the subgroup logarithmic relative risk di- of heterogeneity, P = .05; I2= 30% ference.13 For sensitivity analysis we ex- stracted in duplicate, and authors of the CI, 0.63-0.94; P = .01; I2= 53% [95% UI, 14%-75%]) (Figure). For stat-
mary and secondary prevention of CHD.
quality of included trials with respect to meta–regression analysis to investigate blinding of patients, caregivers, or asses- ness of follow-up.8 When the article failed tion, items about trial quality, percent- p r i m a r y p r e v e n t i o n o f C H D trials, and the type and duration of lipid- 95% CI, 0.91-1.11; P = .01; I2= 33% istics, we classified trials according to the following groups9: statins (35 trials [A1- e r a t e h e t e r o g e n e i t y ( P < . 1 0 ; A35]), fibrates (17 trials [A36-A52]), res- ins (8 trials [A53-A60]), niacin (2 trials [A39 and A61]), n-3 fatty acids (14 trials category.15 All statistical analyses were ited the analysis to interventions with at cally significant (Table 2). Hetero-
/publications_reports.htm. Trials in pri- (REPRINTED) ARCH INTERN MED/ VOL 165, APR 11, 2005 2005 American Medical Association. All rights reserved.
Table 1. Effects of Different Lipid-Lowering Interventions on Overall Mortality
Follow-up,
Cholesterol
Individuals,
Mean ± SD,
Reduction, Mean
Mortality,
Heterogeneity,
Inconsistency,
Type of Intervention*
(Range), %
RR (95% CI)
P Value
I 2 (95% UI), %
Abbreviations: CHD, coronary heart disease; CI, confidence interval, NA, not applicable; RR, risk ratio; T/C, number of individuals in treatment/control groups; *Trials in primary prevention of CHD were defined as trials with less than 10% of participants with CHD, while secondary prevention trials comprised 100% participants with CHD. There are some trials with mixed study populations (eg, 55% of participants with CHD) that could not be confined to either category; theseare therefore not included in this subgroup analysis.
†Sensitivity analysis without the trial of Burr et al (A79; see http://www.bice.ch/engl/publications_reports.htm): RR for overall mortality, 0.80 (95% CI, 0.69-0.92; P=.40; I 2=6% [95% UI, 0%-69%]).
the remaining n-3 fatty acid trials.
CI, 0.61-0.80; P = .47; I2= 0% [95% META-REGRESSION ANALYSIS
for overall mortality was 0.75 (95%CI, 0.65-0.87), and heterogeneity MORTALITY FROM
was substantially reduced (P = .36; CAUSES OTHER THAN
CARDIOVASCULAR DISEASE
with established CHD (coefficient,–0.001; 95% CI, –0.003 to –0.0003) CARDIAC MORTALITY
0.84; P = .42; I2= 3% [95% UI, 0%- CI, 1.01-1.27; P = .80, I2= 0% [95% 0.99; P = .83; I2= 0% [95% UI, 0%- CI, 0.52-0.90; P=.001; I2=66% [95% decrease in trials of longer duration.
(REPRINTED) ARCH INTERN MED/ VOL 165, APR 11, 2005 2005 American Medical Association. All rights reserved.
Risk Ratio (95% CI); Heterogeniety (P ); diac mortality in patients with CHD.
0.87 (0.81-0.94); P = .05; I 2 = 30 [0-54] 1.00 (0.91-1.11); P = .01; I 2 = 33 [0-63] 0.84 (0.66-1.08); P = .86; I 2 = 0 [0-68] 0.96 (0.86-1.08); P = .81; I 2 = 0 0.77 (0.63-0.94); P = .01; I 2 = 53 [14-75] 0.97 (0.91-1.04); P = .19; I 2 = 23 [0-56] CI, 585-1852) patients in a primaryprevention situation with a mortal- to be treated with a statin for 1 yearto prevent 1 death. For n-3 fatty ac- ids, 140 (95% CI, 87-538) patients ina secondary prevention situation have Risk Ratio (95% CI); Heterogeniety (P ); 0.78 (0.72-0.84); P = .42; I 2 = 3 [0-30] 0.93 (0.81-1.08); P = .13; I 2 = 28 [0-60] 0.70 (0.50-0.99); P = .83; I 2 = 0 [0-68] 0.95 (0.82-1.10); P = .75; I 2 = 0 0.68 (0.52-0.90); P <.001; I2 = 66 [37-81] 0.91 (0.82-1.02); P = .14; I 2 = 27 [0-59] viduals in placebo or control groups.
We found little evidence of hetero-geneity in the summary estimates fornoncardiovascular mortality in fi- Mortality From Causes Other Than Cardiovascular Diseases Risk Ratio (95% CI); Heterogeniety (P ); P=.80; I2=0%), suggesting a consis- 0.97 (0.91-1.04); P = .78; I 2 = 0 [0-43] 1.13 (1.01-1.27); P = .80; I 2 = 0 [0-54] 1.08 (0.74-1.58); P = .82; I 2 = 0 [0-85] 1.13 (0.77-1.67); P = .81; I 2 = 0 0.97 (0.84-1.13); P = .95; I 2 = 0 [0.65] 1.01 (0.96-1.07); P = .71; I 2 = 0 [0-54] rent guidelines recommend the use ofeither drug in patients with hypertri- density lipoproteins, and meta-bolic syndrome.17,18 If used in ap- Figure. Summary estimates for overall mortality (A), cardiac mortality (B), and mortality from causes
other than cardiovascular diseases (C) for different types of lipid-lowering interventions. The CochraneQ test for heterogeneity. I 2 as measure of inconsistency (in percent). CI indicates confidence interval; UI, uncertainty interval; n, number of trials available for analysis; n-3 FA, n-3 fatty acid.
glyceride levels19 but are associatedwith a reduction in overall mortal-ity. However, n-3 fatty acids lower vs trials of interventions other than fi- (REPRINTED) ARCH INTERN MED/ VOL 165, APR 11, 2005 2005 American Medical Association. All rights reserved.
Table 2. Sensitivity Analysis of Quality Components for Statins, Fibrates, and n-3 Fatty Acids
Fibrates
n-3 Fatty Acids
Mortality,
Difference Trials,
Mortality,
Difference Trials,
Mortality,
Difference
Quality Component
RR (95% CI)
P Value
RR (95% CI)
P Value
RR (95% CI)
P Value
Abbreviations: CI, confidence interval; RR, risk ratio.
*Sensitivity analysis without the trial of Burr et al (A79; see http://www.bice.ch/engl/publications_reports.htm): RR for overall mortality, 0.75 (95% CI, 0.60-0.92); †Sensitivity analysis without the trial of Burr et al (A79): RR for overall mortality, 1.25 (95% CI, 0.31-5.07); difference P value, .78.
1. 4S-Group. Randomised trial of cholesterol low- to retrieve all relevant eligible trials.
ering in 4444 patients with coronary heart dis- ease: the Scandinavian Simvastatin Survival Study (4S). Lancet. 1994;344:1383-1389.
for such bias, it cannot be ruled out.
2. Bucher HC, Griffith LE, Guyatt GH. Systematic re- view on the risk and benefit of different cholesterol- lowering interventions. Arterioscler Thromb Vasc be less prone to ascertainment bias.
3. Ross SD, Allen IE, Connelly JE, et al. Clinical out- comes in statin treatment trials: a meta-analysis.
cluded trials (PϽ.001; I2=37% [95% Arch Intern Med. 1999;159:1793-1802.
4. Rosenson RS, Tangney CC. Antiatherothrom- botic properties of statins: implications for car- diovascular event reduction. JAMA. 1998;279: 5. Pogue J, Yusuf S. Overcoming the limitations of current meta-analysis of randomised controlledtrials. Lancet. 1998;351:47-52.
Accepted for Publication: Novem-
6. Smith GD, Song F, Sheldon TA. Cholesterol low- ering and mortality: the importance of consider- Correspondence: Heiner C. Bucher,
ing initial level of risk. BMJ. 1993;306:1367- 7. Rossouw JE, Anderson GL, Prentice RL, et al.
Risks and benefits of estrogen plus progestin in results from the Women’s Health Initiative ran-domized controlled trial. JAMA. 2002;288:321- Funding/Support: Drs Bucher and
8. Juni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA. 1999;282:1054-1060.
9. McAlister FA, Laupacis A, Wells GA, Sackett DL. Users’ Guides to the Medical Literature, Acknowledgment: We are grateful
XIX: applying clinical trial results B: guidelines for determining whether a drug is exerting(more than) a class effect. JAMA. 1999;282: 10. Egger M, Davey SG, Schneider M, Minder C.
Bias in meta-analysis detected by a simple, graphi- action.9 For example, trials of n-3 fatty cal test. BMJ. 1997;315:629-634.
(REPRINTED) ARCH INTERN MED/ VOL 165, APR 11, 2005 2005 American Medical Association. All rights reserved.
11. Higgins JP, Thompson SG, Deeks JJ, Altman DG.
rived from meta-analysis: a word of caution. ACP 18. Grundy SM, Cleeman JI, Merz CN, et al. Implica- Measuring inconsistency in meta-analyses. BMJ.
tions of recent clinical trials for the National Cho- 16. Din JN, Newby DE, Flapan AD. Omega 3 fatty ac- lesterol Education Program Adult Treatment Panel 12. Higgins JP, Thompson SG. Quantifying hetero- ids and cardiovascular disease—fishing for a natu- III guidelines. Arterioscler Thromb Vasc Biol. 2004; geneity in a meta-analysis. Stat Med. 2002;21: ral treatment. BMJ. 2004;328:30-35.
17. Expert Panel on Detection EaToHBCiA. Execu- 19. Bucher HC, Hengstler P, Schindler C, Meier G.
13. Fleiss JL. The statistical basis of meta-analysis.
N-3 polyunsaturated fatty acids in coronary heart Stat Methods Med Res. 1993;2:121-145.
disease: a meta-analysis of randomized con- 14. Thompson SG, Sharp SJ. Explaining heterogene- (NCEP) Expert Panel on Detection, Evaluation, trolled trials. Am J Med. 2002;112:298-304.
ity in meta-analysis: a comparison of methods. Stat and Treatment of High Blood Cholesterol in 20. Oxman AD, Guyatt GH. A consumer’s guide to sub- Adults (Adult Treatment Panel III). JAMA. 2001; group analyses. Ann Intern Med. 1992;116: 15. Marx A, Bucher HC. Numbers needed to treat de- Omissions in Byline. In the Original Investigation by
Cohen et al titled “Emerging Credentialing Practices, Mal-
practice Liability Policies, and Guidelines Governing
Complementary and Alternative Medical Practices and
Dietary Supplement Recommendations: A Descriptive
Study of 19 Integrative Health Care Centers in the United
States,” published in the February 14 issue of the
ARCHIVES (2005;165:289-295), 2 authors were inadvert-
ently omitted from the byline on page 289. The byline
should have appeared as follows: “Michael H. Cohen, JD;
Andrea Hrbek; Roger B. Davis ScD; Steven C. Schachter,
MD; Kathi J. Kemper, MD, MPH; Edward W. Boyer, MD,
PhD; David M. Eisenberg, MD.”
(REPRINTED) ARCH INTERN MED/ VOL 165, APR 11, 2005 2005 American Medical Association. All rights reserved.

Source: http://www.visolie-info.nl/pdf/A&A%20omega-3%20pufa%20and%20effects%20on%20mortality%202005.pdf