“A drug for the bone-thinning disease osteoporosis… nearly doubles the risk of a common heart condition”, The Guardian reports today. BBC News also reports that women who take the drug Fosamax, the generic name of which is alendronate, increased the risk of a particular type of abnormal heartbeat (atrial fibrillation) by 86%.
The news reports mention previous studies that have had contradictory findings for whether or not the drug increases risk of atrial fibrillation (AF). They both quote a spokesperson from The National Osteoporosis Society as saying that the study “should be considered in the context with other recent research, which has not shown the same increase in atrial fibrillation”.
These results are based on a study that compared alendronate use in just over 700 women who had experienced atrial fibrillation and almost a thousand women who had not. The main limitation to this study is that these two groups were not randomly chosen, and so there may have been differences between the groups, other than alendronate use, that are responsible for the differences seen in AF.
The apparent contradiction in the results of this study and the previous studies as mentioned by the newspapers, may be resolved by a systematic review and pooling of data from randomised trials. Until then, doctors will need to consider the balance of benefits and risks of alendronate for each patient individually to see whether the known reduction in the risk of fractures outweighs the possible increase in risk of AF.
Where did the story come from?
Dr Susan Heckbert and colleagues from the University of Washington and other US healthcare and research bodies carried out the research. The study was funded by the US National Heart, Lung, and Blood Institute. The study was published in the peer-reviewed medical journal: Archives of Internal Medicine.
What kind of scientific study was this?
In this retrospective case-control study, the researchers looked at whether the use of a drug taken to treat and prevent osteoporosis – alendronate – was associated with an increased risk of AF.
The researchers used the records of a large healthcare delivery system in Washington to identify women aged 30 to 84-years who had been diagnosed with AF for the first time at any inpatient or outpatient visit between October 2001 and December 2004. The date of diagnosis was defined as their “index date” (a date randomly selected from within the range of dates when the cases were diagnosed). To be included in the study, women had to have visited the healthcare provider at least four times before they were diagnosed with AF.
The AF patients were put into three groups that described the persistence and duration of their AF in the six months following diagnosis. The groups were: transitory AF (a single episode lasting up to seven days), persistent/intermittent AF (episodes lasting longer than seven days, or when there was more than one episode with periods of normal heartbeat in between), or sustained AF (continuous for six months). The researchers also recorded where the patient had been diagnosed: the urgent care clinic, emergency department, or on admission to hospital.
A control group of women were randomly selected from the same healthcare system. None of these women had been diagnosed with AF or had a pacemaker prior to an ‘index date’. This group had initially been selected for a different study, and were to be compared to a group of people who had experienced a heart attack. They had been matched to this group by age, being treated for high blood pressure, and the year in which they were matched to the case.
The researchers looked at all the women’s medical records for the past 20 years (on average) and recorded which major conditions they had experienced, such as osteoporosis, diabetes, high blood pressure, heart attack, heart failure, angina, stroke, and peripheral vascular disease. They also recorded the patient’s blood pressure and weight at the visit nearest to being diagnosed with AF. Where possible, the women were contacted by telephone to obtain demographic and health information, such as their race, whether they smoked and how much they drank before their index date.
The healthcare system’s pharmacy database was used to identify what medications the women were taking, including bisphosphonates (such as alendronate), blood pressure medication, and hormone replacement therapy. Those who had used a bisphosphonate other than alendronate were excluded. The researchers defined those who had ever used alendronate and had received at least two prescriptions for it as “ever users”. Women who had received an alendronate prescription that would have lasted up to the date of AF diagnosis were described as current users. The total amounts of alendronate taken were calculated and the total time a person had taken alendronate was recorded.
In their main analyses, the researchers compared the proportion of alendronate “ever users” among cases and controls. In subsidiary analyses, they looked at the relationship between alendronate use and AF in different subgroups of people. Analyses were adjusted for age, treatment for high blood pressure, calendar year of index date, osteoporosis and any cardiovascular disease.
What were the results of the study?
The researchers included 719 cases and 966 controls. The average age of women who experienced AF was 75 years, and 71 years for controls. The women who experienced AF had higher rates of diabetes and heart problems such as heart attacks than controls, but the rates of osteoporosis were similar in the two groups (about 10% in each).
More women who had experienced AF had ever used (i.e. were “ever users”) alendronate (about 7%) than those who had not (about 4%). The proportion of women currently using alendronate was similar in cases and controls. Women who had used alendronate differed in many ways from those who hadn’t. For example, alendronate users were older, had higher levels of good cholesterol (HDL) and were less likely to have diabetes or cardiovascular disease.
The researchers calculated that alendronate users increased their odds of having AF by about 86%, and estimated that alendronate use was responsible for about three out of every 100 cases of AF in this population.
What interpretations did the researchers draw from these results?
The researchers concluded that using alendronate (as the “ever users” had) increased the risk of AF in normal clinical practice.
What does the NHS Knowledge Service make of this study?
This study aimed to look at whether alendronate use was associated with an increased risk of AF in normal clinical practice. There are some limitations to this study, some of which the authors acknowledge:
- This type of study does not randomly assign people to receive alendronate or not, this means that the groups of people that are compared may be imbalanced with regards to factors other than the one tested. The authors did try to adjust for known factors that might affect risk of AF in their analyses, but it is not possible to adjust for unknown or unmeasured factors, and adjustments rely on the accuracy of the measurements and recording of the data by the healthcare system. For example, an overactive thyroid gland (hyperthyroidism) is linked to both osteoporosis and AF. If the researchers had been able to adjust for rates of this, they may have found different results.
- The majority of women in the healthcare system that was studied were prescribed alendronate, rather than other bisphosphonates. These results may not apply to other bisphosphonates. The authors report that a pooled analysis of trials of another bisphosphonate – risedronate – found no increase in AF.
- Controls in this study seemed to have been matched to a set of cases with heart attack, rather than the AF cases in this study. Poor matching of cases and controls may lead to greater inaccuracies in the results of this type of study.
- Cases of AF may have been missed the women did not go to their doctor.
- All the cases and controls in this study were women, and the results may not be applicable to men.
- Data was only available on prescriptions given through the healthcare system in which patients were enrolled. Prescriptions from other sources would not have been identified.
A possible link between bisphosphonates and AF was highlighted in a randomised controlled trial of zolendronate (another bisphosphonate). Another randomised controlled trial of alendronate found a trend for an increased risk of AF, but this increase did not reach statistical significance.
The apparent contradiction in the results of randomised and observational evidence may be resolved by a systematic review of the pooled results from randomised trials. For bisphosphonates to be accepted as a cause of atrial fibrillation, researchers will also need to prove a biological mechanism by which the drug is acting on the heart.
Until then, doctors will need to consider the balance of benefits and risks of alendronate for each patient individually to see whether the known reduction in the risk of fractures outweighs the possible increase in risk of AF.
Sir Muir Gray adds…
Drugs that don’t have side effects are rare. This type of study, carried out after the widespread use of a drug, is needed to complement the study of effectiveness usually done by a randomised controlled trial.