select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

Benefits of statins in elderly patients with no pre-existing heart disease

Scientists publishing in JAMA Internal Medicine examine statin treatment among adults aged 65 to 74 years and 75 years and older when used for primary prevention in the Lipid-Lowering Trial (LLT) component of the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT-LLT). Roundup comments accompanied this analysis.


Title, Date of Publication & Journal

Effect of Statin Treatment vs Usual Care on Primary Cardiovascular Prevention Among Older Adults, The ALLHAT-LLT Randomized Clinical Trial

Published: May 22nd 2017

JAMA Internal Medicine


Study’s main claims – and are they supported by the data

This paper is a subgroup analysis of an earlier trial of statin, focussing on the effect in older patients without history of CVD.

The key finding claimed in the abstract and press release is that statins provided no benefit to the elderly, but the data does not support this claim.  The paper is inconclusive as the sample-size is too small (post-hoc subset analysis of the ALLHAT-LLT trial) with no documented statistical justification for this unplanned analysis in either this paper or the original paper of the full dataset.

The primary reported results were a higher mortality rate in the statins group, although it was not significant (p=0.09), and a lower risk of CVD in the statins group, which was also non-significant (p=0.12).

This study provides interesting data on a relevant question, but does not provide any definitive answers. No findings were statistically significant, and this subgroup analysis is at high risk of spurious findings.




The key strength of this study is that it comes from a randomised trial. Data on effect of statin for primary prevention in older adults are scarce and conflicting.


–  The key limitation was that the sample size was too small, which makes it very hard to confidently test for significant differences between groups. Regarding the lack of significant effect on CVD reduction, it is important to understand that in such a small sample it not appropriate to interpret the lack of statistically significant differences as demonstrating the absence of benefit of a statin.

– This study reports a subgroup analysis, conducted after the main trial data had already been analysed (post-hoc). Results focusing on a subgroup of a larger trial are at a high risk of spurious findings, and the risks are particularly high in this study as the subgroup investigated was defined after the data had already been analysed. There was also no indication from the original trial that age was an issue regarding the benefits of statins, so it is hard to justify a further analysis of this sub-group and also hard to be confident of the findings.

–  There was a high-level of cross over between the groups receiving a statin and those with usual care. By year 2, only 86% of the statin group were still taking the statin (14% discontinuation) increasing to 22% at year 6, while by the same point in the trial 29% of patients in the UC group had started taking a statin (there is no information about which sort of statin). Statin use in both groups (rather than just one) makes it hard to assign observed group differences to treatment with a statin and no further sub- analysis between the groups to clarify this finding is reported.

– This is not a new study and does not use new data; it reports results of a trial that was published 15 years ago, and the effects of statins on mortality and CHD on the over-65s with no history of CHD were already reported in the original study.

–  There was no attempt to adjust for multiple significance testing despite a high number of tests. The more tests conducted the more likely it is at least one will appear significant due to chance alone.

– One other limitation is that this study did not use a placebo in the control arm of the study, and so participants in that group receiving usual care could have made changes to their lifestyle that would affect comparisons between the groups.



(IQR) Inter-quartile-range: range of the middle 50% of the data

Confounding: Characteristics/traits that explain an apparent relationship

Fecundity: Biologic capacity for reproduction

(Measures of fecundity are distinct from fertility)

Fertility: Demonstrated fecundity, usually measured by live births


Before The Headlines is a service provided to the SMC by volunteer statisticians: members of the Royal Statistical Society (RSS), Statisticians in the Pharmaceutical Industry (PSI) and experienced statisticians in academia and research. A list of contributors, including affiliations, is available here.

in this section

filter Headlines by year

search by tag