To investigate if an nonphysiological prolongation of the AV interval is common during activity sensor modulated atrial rate adaptive (AAIR) pacing, 21 patients with sinus node disease treated with fixed rate atrial (AAI) or AAIR pacemakers were examined. Spike-Q intervals were compared at different heart rates obtained by overdrive pacing at rest and during exercise (Study I), measured during exercise at unresponsive (AAI), optimal (AAIR) and over responsive programming (AAIR+) of the activity sensor (Study II), and finally examined by 24-hour Holter recording in AAI and AAIR pacing modes (Study III). Study I: The spike-Q interval increased significantly with increasing heart rate at rest, but not during exercise. At rest the spike-Q interval was significantly higher at all heart rates compared to exercise. There was a significant positive correlation between the maximal spike-Q interval at rest and the maximal spike-Q interval during exercise (r = 0.63). Study II: The spike-Q interval was shortest in the AAI and longest in the AAIR+ mode in all patients. Study III: During AAI or AAIR pacing the spike-Q interval was longest at night and shortest in the morning. The mean spike-Q interval was longer in AAIR than in AAI pacing. No statistical difference between the maximal spike-Q intervals observed during the two modes was, however, found. Variations in spike-Q interval are generally caused by changes in autonomic tone or medication with drugs with antiarrhythmic effect.(ABSTRACT TRUNCATED AT 250 WORDS)