Falls, chronic diseases, and drug use in elderly women

Editor—We agree with Epstein that our study was unable to determine whether the participants had been appropriately prescribed cardiovascular drugs. This was not the aim of the study. He says that we may have underestimated the effect of cardiovascular drugs on the risk of falling because doctors may not prescribe these drugs to patients at greatest risk of falling, implying that our estimate of no effect on falls of cardiovascular drugs represents reduced risk of falling in those prescribed these drugs, appropriately matched by an increased fall risk among women prescribed these drugs inappropriately.

This issue of confounding by indication for treatment is problematic in observational studies of this nature. We could find no systematic review or randomised controlled trial examining the risk of falls associated with angiotensin converting enzyme (ACE) inhibitors or other cardiovascular drugs. However, risks of feeling dizzy or fainting were high in both those treated with enalapril and placebo (57% v 50%, relative risk increase 14%, 95% confidence interval 6% to 21%),¹ implying, firstly, that these drugs are still capable of causing falls in people with correct indications for treatment, and, secondly, feeling dizzy and faint is a common experience among patients with heart failure. Therefore, Epstein's suggested balance between falls caused and falls avoided due to confounding by indication seems an unlikely explanation for our findings.

We agree with Epstein that patients with heart failure and coronary heart disease should be prescribed appropriate cardiovascular treatment that may improve their prognosis and that their propensity to fall be assessed clinically. Our study strongly supports such practice as we showed no association between most classes of drugs and falling. Interestingly, we also showed that some factors commonly supposed to be associated with falls (and that Epstein says are used clinically by doctors to identify groups at high risk) such as postural hypotension, alcohol intake, and reduced physiological reserve (as evidenced by low forced expiratory volume in one second) were not associated with falls. It would be useful to validate prospectively such clinical markers of risk of falling as it may be that clinicians are unwittingly denying their patients life-saving, and life-enhancing, drugs such as ACE inhibitors, on spurious grounds.