Statins and Multiple Noncardiovascular Outcomes: Umbrella Review of Meta-analyses of Observational Studies and Randomized Controlled Trials.
Ann Intern Med. 2018 Oct 16;169(8):543-553
Authors: He Y, Li X, Gasevic D, Brunt E, McLachlan F, Millenson M, Timofeeva M, Ioannidis JPA, Campbell H, Theodoratou E
Background: Many effects of statins on non-cardiovascular disease (non-CVD) outcomes have been reported.
Purpose: To evaluate the quantity, validity, and credibility of evidence regarding associations between statins and non-CVD outcomes and the effects of statins on these outcomes.
Data Sources: MEDLINE and EMBASE (English terms only, inception to 28 May 2018).
Study Selection: Meta-analyses (published in English) of observational studies and of randomized controlled trials (RCTs) that examined non-CVD outcomes of statin intake.
Data Extraction: Two investigators extracted data from meta-analyses and individual studies. Credibility assessments based on summary effect sizes from a random-effects model, between-study heterogeneity, 95% prediction interval, small-study effect, excess significance, and credibility ceilings were devised to classify evidence.
Data Synthesis: This review explored 278 unique non-CVD outcomes from 112 meta-analyses of observational studies and 144 meta-analyses of RCTs. For observational studies, no convincing (class I) evidence, 2 highly suggestive (class II) associations (decreased cancer mortality in patients with cancer and decreased exacerbation in patients with chronic obstructive pulmonary disease), 21 suggestive (class III) associations, and 42 weak (class IV) associations were identified. One outcome from the RCTs (decreased all-cause mortality in patients with chronic kidney disease) attained a sufficient amount of evidence with no hints of bias. For adverse events, observational studies showed suggestive evidence that statins increase the risk for diabetes and myopathy. Among the RCTs, no statistically significant effects were found on myopathy, myalgia, or rhabdomyolysis.
Limitations: Studies with relevant data and outcomes not included in the meta-analyses may have been missed. Credibility assessments relied on several assumptions and arbitrary thresholds.
Conclusion: The absence of convincing evidence of an association between statins and non-CVD outcomes supports leaving the current recommendations unchanged.
Primary Funding Source: None.
PMID: 30304368 [PubMed – in process]