- Safi U Khan, cardiologist,
- Neal S Kleiman, cardiologist
- Correspondence to: S U Khan safinmc{at}gmail.com
Inhibitors of hydroxymethylglutaryl coenzyme A reductase (statins) are among the most widely investigated and prescribed drugs in cardiovascular medicine, and they have substantial effects on the outcomes of atherosclerotic cardiovascular disease.12 Leading guidelines from US and Europe2 broadly recommend statins for people with high low density lipoprotein (LDL) cholesterol concentrations and a high risk of cardiovascular events, including those with a history of cardiovascular disease (secondary prevention).12
The approach for people at lower risk (including for primary prevention) is more variable. The European Society of Cardiology recommends statins for patients with high LDL cholesterol concentrations, even when their overall risk of cardiovascular events is low (<1% over 10 years),2 whereas the American College of Cardiology and American Heart Association recommend lifestyle modifications not statins for people with a risk below 5% over 10 years.1
Despite clear indications, some studies show that roughly half of patients stop taking statins after six months, and only a quarter of patients with high cardiovascular risk continue with treatment long term.34 Statin associated muscle symptoms, described as myalgia, cramping, or fatigue, are the most common cause of statin intolerance.
Landmark trials of statins reported low frequencies of statin associated myalgia.56 But concerns remained about the possibility of under-reporting or underestimations because participants with comorbidities, who are more likely to develop these symptoms, were excluded.7 A more recent participant level meta-analysis of 19 placebo controlled trials of statin therapy, including more than 33 000 patients (48% with atherosclerotic cardiovascular disease) attempted to overcome these shortcomings.7 The authors found that, compared with placebo, statin treatment was associated with only 11 excess incidents of muscle pain or weakness per 1000 person years of treatment during the first year.7 They found no significant difference in absolute risk between statin and placebo groups after one year.
Myalgia was common among participants, regardless of statin use.7 The proportion of patients reporting muscle pain or weakness was 27.1% in the statin arm and 26.6% in the placebo arm. These event rates remained consistent across subgroups stratified by age, sex, ethnicity, history of vascular disease, diabetes, body mass index, LDL cholesterol concentration, and estimated glomerular filtration rate. The authors calculated that only 1 in 15 reports of muscle symptoms in statin groups were attributable to statins.7 Furthermore, the additional absolute risk of such events remained low across different intensities of therapy.7 Creatinine kinase activity, a more objective outcome than patient reported symptoms, was increased only minimally with statin therapy.
Excess risk of muscle pain or weakness was evident only during the first year, while evidence suggests that the benefits of LDL lowering therapies, including statins, accumulate over longer treatment durations.8
Findings from the new meta-analysis7 of muscle symptoms should be considered in the context of several limitations, including heterogeneity in the methods used to ascertain muscle symptoms; uncertainties about the nature of symptoms among participants who stopped their statin compared with those who continued; and variable outcome definitions, although the definitions probably reflect the various ways patients describe their symptom. The authors had limited data on whether muscle symptoms led to the discontinuation of allocated treatment, and participants with statin related myalgia dropping out might partly explain the lack of risk difference after one year. Data were also limited on comorbidities such as hypothyroidism and concomitant medications that might cause or influence muscle symptoms.7
Accurate estimation of the risk-benefit balance associated with statins is particularly important for primary prevention. In a recent meta-analysis of 62 primary prevention trials (120 456 patients; <3% risk of major cardiovascular events at average follow-up 3.9 years), statins were associated with 15 events of self-reported muscle symptoms per 10 000 patients a year compared with a non-statin control.9 But statins were also associated with a significant reduction in myocardial infarction (19 fewer per 10 000), stroke (9 fewer), and cardiovascular deaths (8 fewer) compared with controls.
Individuals making treatment decisions often have to rely on clinical trials reporting average treatment effects across large sample sizes, which may not reflect their own personal balance of benefits and harms.10 To help bridge this gap, the British Heart Foundation conducted an n-of-1 trial (Samson), which used a crossover design to assess the tolerability of statin, placebo, or no treatment among 60 patients with a history of statin intolerance, most of whom had a history of statin associated muscle symptoms (85%) and a 10 year risk of cardiovascular events <25% (77%).11
Participants were randomised to one month periods of atorvastatin 20 mg, placebo, and no treatment for a year. Mean symptom intensity on a 100 point scale— recorded daily—was 8.0 during periods of no treatment compared with 15.4 when taking a placebo and 16.3 taking a statin (P=0.38). The ratio of symptom intensity induced by a placebo to the symptom intensity induced by a statin (nocebo ratio) was 0.9, suggesting that 90% of the symptom burden induced by a statin challenge was also evoked by a placebo.12 In other words, even the possibility of receiving a statin versus the certainty of not receiving one was enough to induce the sensation of muscle pain. Another n-of-1 trial from the UK in 151 participants with a history of statin intolerancealso found no effect of statins on muscle symptoms.13
While further n-of-1 trials may help to clarify the nature and severity of reported myalgia, clear evidence indicates that statin associated muscle symptoms are uncommon and that most are not related to muscle damage.14
While clinicians must respect patients’ beliefs and preferences, they must also offer treatment when evidence shows the potential for substantial benefits, particularly in avoiding catastrophic events. Combining these obligations requires meaningful shared decision making about statins, including acknowledgment of patients’ concerns and a clear, personalised, and evidence based discussion of the likely benefits and harms of treatment. We now have much better evidence to approach these conversations confidently.