Thursday, July 02, 2009

New Study Finds Additional Reason for Recent Declines in Heart Attacks; Casts Further Doubt on Conclusions of Helena Study and Others of Its Ilk

A new study published last month in the journal Medical Care found that reduced use of hormone replacement therapy in women since 2002 has led to a significant decline in heart attacks among women (see: Shetty KD, Vogt WB, and Bhattacharya J. Hormone replacement therapy and cardiovascular health in the United States. Med Care 2009; 47:600-606).

According to a Heartwire article: "A new study in the general US population has found that the sharp drop in use of hormone-replacement therapy (HRT) in the three years from 2002 onward was associated with significantly fewer MIs among women, but that the stroke rate was unaffected. Dr Kanaka D Shetty (RAND Corp, Santa Monica, CA) and colleagues report their findings in the May 2009 issue of Medical Care. They explain that the use of HRT was curtailed dramatically in 2002 following the results of the WHI trial, which found that HRT increased the risk of coronary heart disease in healthy postmenopausal women. This was in contrast to the prior perceived benefit of HRT on cardiovascular risk—Shetty et al thus exploited this "natural experiment" to see whether the cardiovascular harms (or benefits) of HRT seen in randomized controlled trials such as WHI extended to the general population of women. "We found that the decline in US HRT use after 2002 appeared to cause a decline in the AMI rate (while controlling for other factors); we did not observe statistically significant changes in the stroke rate," Shetty told heartwire."

The Rest of the Story

This finding has important implications for the interpretation of results from the Helena studies and others of its ilk. The present study suggests an alternative explanation for the observation of reduced heart attack rates following the recent implementation of smoking bans in places like Helena, Pueblo, Bowling Green, and elsewhere. A decline in the use of hormone replacement therapy may partially explain such findings. Since no previous studies have controlled for the decreased use of HRT, this remains a plausible alternative explanation for the Helena et al. conclusions and casts further doubt on those conclusions.

Note that I am not questioning the validity of the finding that heart attack rates have declined in localities that have enacted smoking bans. The issue is whether the observed declines are attributable to the smoking bans themselves, rather than to other secular changes that have occurred during the same time period. Most of the existing studies are before-after studies and the conclusions have been largely based on the finding of a reduction in heart attacks from before to after the smoking ban. However, secular changes concurrent with the implementation of these bans could also explain such findings. The demonstration that decreased use of HRT which occurred during exactly the same period when most of the smoking bans studies took place casts doubt on the conclusion of those studies, as it introduces an additional alternative explanation for the study results.

I have already explained that factors such as improved treatment for cardiovascular disease, including advances in angioplasty and thrombolytic techniques and increased use of statin drugs to control cholesterol have contributed to the recent observed declines in heart attack rates. This study simply adds an additional factor that would also explain the finding of reduced heart attacks in these studies.

The problem with the conclusions from the Helena et al. studies are two-fold. First, these studies have not been able to rule out the possibility that the observed declines in heart attacks are more than just random variation in the data. Second, these studies have not been able to rule out the possibility that other factors were responsible for the decline in heart attacks. The research I am reporting on today adds an additional factor that needs to be considered, but that has not been addressed in any of the existing research.

No comments: