The adjusted cubic spline model demonstrates the flexible association between changes in HR from the preceding visit and the hazard of all-cause mortality (A) and hazard of incident HF (B), when no change in HR is taken as the reference (ie, 0 bpm). Both curves display a near linear relationship between change in HR and the end points of all-cause mortality and incident heart failure (HF), such that any rise in HR (>1 bpm) from the preceding visit appears to increase the risk significantly. However, any drop in HR from the preceding visit significantly reduced risk of all-cause mortality, and a drop in HR from the preceding visit of more than 12 bpm is associated with a significant reduction in risk of incident HF. The adjusted cubic spline model demonstrates the flexible relationship between time-updated HR (ie, most recent HR value before an event or end of the study) and the hazard of all-cause mortality (C) and hazard of incident HF (D) when a resting HR of 60 bpm is taken as the reference. Time-updated HR between 50 and 60 bpm during follow-up was associated with a lower risk of all-cause mortality and risk of incident HF compared with a HR of 60 bpm. A time-updated HR between 60 and 66 bpm was not associated with an increase in risk of all-cause mortality compared with a HR of 60 bpm; however, the risk of incident HF was greater above a HR of 60 bpm compared with a HR of 60 bpm. The most recent resting HR above 66 bpm during follow-up was associated with a higher risk of all-cause mortality. The dashed black curves represent the upper and lower 95% confidence limits. The horizontal blue line represents the hazard ratio of 1. The domain was defined by excluding the smallest 1% and the largest 1% of values of change in HR values.