Objective Obstructive coronary artery disease (CAD) is evident in only half of patients referred for diagnostic angiography. Five-minute heart rate variability (HRV) is a non-invasive marker for autonomic control of the vasculature, which this study hypothesised could risk-stratify cardiac patients and reduce unnecessary angiograms.

Design A prospective observational study (the Alternative Risk Markers in Coronary Artery Disease (ARM–CAD) study).

Setting Three cardiac centres in Melbourne, Australia.

Patients 470 consecutive patients undergoing elective angiography (with predominantly normal cardiac rhythm), regardless of co-morbidity.

Main outcome measures The presence of obstructive CAD (≥50% stenosis) on angiography.

Results Patients with obstructive CAD had significantly reduced HRV, particularly in the low frequency (LF) range (median 180 vs 267 ms² without CAD; p<0.001). There was a linear trend with the severity of CAD; median LF power (IQR) in patients with normal coronaries was 275 (612), with minor coronary irregularities 255 (400), single-vessel CAD 212 (396) and more severe disease 170 (327) ms²; p value for trend 0.003. There was a similar reduction in LF power regardless of the anatomical location of coronary stenoses. Comparing patients with LF less than 250 and 250 ms² or greater, the adjusted OR for obstructive CAD using multivariate regression was 2.42, 95% CI 1.33 to 4.38 (p=0.004). No interactions were noted in subgroup analysis and HRV added to risk prediction irrespective of the baseline Framingham risk (p<0.0001).

Conclusion Low HRV is strongly predictive of angiographic coronary disease regardless of other co-morbidities and is clinically useful as a risk predictor in patients with sinus rhythm.

Clinical trial registration information http://clinicaltrials.gov/ct2/show/NCT00403351 www.armcad.com