But senior author of the paper, Dr Christoph R Meier (University Hospital, Basel, Switzerland), told that the "practical implications" of these findings are few, because "we studied a sample, a subset of patients on monotherapy, and therefore one cannot generalize to a more complicated hypertensive population."

Nevertheless, he says, "If a patient is believed to be at risk for AF, this may be good to know for the doctor, that certain drugs might have a beneficial effect [on AF].

It could be yet another criterion to look at when selecting an antihypertensive drug, just one more piece of the puzzle."

Largest study of its kind to date

Schaer et al explain that all antihypertensive drugs lower the risk for AF simply by lowering blood pressure, but some agents might also reduce the risk of AF through other mechanisms.

Post hoc analyses of large placebo-controlled trials in patients with heart failure indicate that ACE inhibitors or ARBs may reduce the risk for new-onset AF, but current opinions on the effect of various antihypertensives on risk for AF are "conflicting."

In their nested case-control study using the UK-based General Practice Research Database—which is the largest epidemiological study of its kind so far to examine this issue, according to Meier—the researchers looked at first-time occurrence of AF in hypertensive patients already receiving one blood-pressure-lowering agent who were free of any cardiovascular disease.

This relatively homogeneous sample of hypertensive patients who were unlikely to have structural heart disease or other risk factors for AF allowed a focus on the possible effects of the drug of interest, the researchers note.

From a total study population of more than 650 000 hypertensive patients who had at least one prescription for an antihypertensive drug in their computer record, they identified 4661 hypertensive patients with a recorded diagnosis of new AF between 1998 and 2008. For each case, up to four controls without AF were identified (18 642 controls total).

They compared the risk of AF among those who received only ACE inhibitors, ARBs, or beta blockers relative to the reference group of patients who used only CCBs.

Use of diuretics was allowed sequentially or concomitantly with other antihypertensive drugs, however, they note, because so many antihypertensive drugs are now administered as combinations that contain small amounts of diuretics.

Compared with the reference groups using CCBs, current exclusive long-term recipients of ACE inhibitors, ARBs, and beta blockers had lower risks for AF, with adjusted odd ratios of 0.75, 0.71, and 0.78 respectively.

"Our findings provide some evidence that long-term receipt (approximately 12 months or more) of ARBs, ACE inhibitors, or beta blockers may decrease the risk for AF compared with receipt of CCBs," say Schaer et al.

The fact that beta blockers also seemed to reduce the risk of AF in the study could be due to a positive effect on premature atrial contractions, the researchers hypothesize.

But they stress that the results "are not generalizable to patients with severe hypertension that requires therapy with multiple drugs."

Another protective effect to throw into the risks/benefits equation

Meier says these findings could be put into the "protective-effects" basket of certain drug groups and taken into consideration in, for example, pharmacoeconomic analyses.

"When you perform such cost-effectiveness studies, we know, for instance, that ARBs cost more than other antihypertensive drugs, but the more we know about their beneficial effects, the more information can go into the equation.

If we can prevent cases of AF, which are very problematic and cost so much money, we have to put this in relation to the additional costs for the more modern drug treatment. This is just one aspect, but it's a nice add-on."

A logical extension of this work, he said, would be to look at patients who already have atrial fibrillation and perform a study to see whether any particular antihypertensive agent is better than any other in this population, "but this needs to be prospectively tested," he concluded.

• Schaer BA, Schneider C, Jick SS, et al. Risk for incident AF in patients who receive antihypertensive drugs. Ann Intern Med 2010; 152:78-84.