Personal Activity Intelligence and mortality in patients with cardiovascular disease: the HUNT study

Abstract Objective To test whether Personal Activity Intelligence (PAI), a personalized metric of physical activity (PA) tracking, is associated with all-cause and cardiovascular disease (CVD) mortality in patients with self-reported CVD and to determine whether these associations change depending on whether contemporary PA recommendations are met. Patients and Methods A total of 3133 patients with CVD (mean [SD] age, 67.6 [10.3] years; 64% men) were followed from the date of participation in the Nord-Trondelag Health Study (between January 1, 1984, and February 28, 1986) until the date of death or the end of follow-up (December 31, 2015). The participants' weekly PAI score was calculated and divided into 4 groups (PAI scores of 0, ≤50, 51-99, and ≥100). We used Cox proportional hazards regression models to estimate hazard ratios for CVD and all-cause mortality rates. Results After mean follow-up of 12.5 years (39,157 person-years), there were 2936 deaths (94%), including 1936 CVD deaths. Participants with weekly PAI scores of 100 or greater had 36% (95% CI, 21%-48%) and 24% (95% CI, 10%-35%) lower risk of mortality from CVD and all causes, respectively, compared with the inactive group. Participants had similar risk reductions associated with their weekly PAI scores regardless of following contemporary PA recommendations or not. Conclusion Obtaining a weekly PAI score of at least 100 was associated with lower mortality risk from CVD and all causes in individuals with CVD regardless of whether the current PA recommendations were met.