Inter- and intraindividual variability in daily resting heart rate and its associations with age, sex, sleep, BMI, and time of year: Retrospective, longitudinal cohort study of 92,457 adults

Abstract

Background: Heart rate is routinely measured as part of the clinical examination but is rarely acted upon unless it is well outside a population-based normal range. With wearable sensor technologies, heart rate can now be continuously measured, making it possible to accurately identify an individual's "normal" heart rate and potentially important variations in it over time. Our objective is to describe inter- and intra-individual variability in resting heart rate (RHR) collected over the course of two years using a wearable device, studying the variations of resting heart rate as a function of time of year, as well as individuals characteristics like age, sex, average sleep duration, and body mass index (BMI).

Methods and findings: Our retrospective, longitudinal cohort study includes 92,457 de-identified individuals from the United States (all 50 states), who consistently-over at least 35 weeks in the period from March 2016 to February 2018, for at least 2 days per week, and at least 20 hours per day-wore a heart rate wrist-worn tracker. In this study, we report daily RHR and its association with age, BMI, sex, and sleep duration, and its variation over time. Individual daily RHR was available for a median of 320 days, providing nearly 33 million daily RHR values. We also explored the range in daily RHR variability between individuals, and the long- and short-term changes in the trajectory of an individual's daily RHR. Mean daily RHR was 65 beats per minute (bpm), with a range of 40 to 109 bpm among all individuals. The mean RHR differed significantly by age, sex, BMI, and average sleep duration. Time of year variations were also noted, with a minimum in July and maximum in January. For most subjects, RHR remained relatively stable over the short term, but 20% experienced at least 1 week in which their RHR fluctuated by 10 bpm or more.

Conclusions: Individuals have a daily RHR that is normal for them but can differ from another individual's normal by as much as 70 bpm. Within individuals, RHR was much more consistent over time, with a small but significant seasonal trend, and detectable discrete and infrequent episodes outside their norms.

Fig 1. Subject selection for the study.

We first removed users of resting heart rate (RHR) tracking devices based on demographic characteristics: age under 18 years, no sex specified, or body-mass index (BMI) outside the range of 15–50 kg/m². Then we removed users who did not have at least 2 days per week of use, with the device worn 20 hours or more per day, for a minimum of 35 weeks. The demographic and clinical data for the resulting cohort divided by gender are shown.

Fig 2. Distribution of average daily resting heart rates.

The average daily RHR for 57,836 individual women (green line) and 34,621 men (blue line). The overall mean for each group is indicated by the dashed lines.

Fig 3

Influence of age (A), body-mass index (BMI) (B), and hours of sleep (C) on average daily RHR, influence of age (D) and body-mass index (BMI) (E) on the standard deviation (STD), and influence of the mean RHR on the coefficient of variation (Coeff. Var.) of the RHR (F). The average, the standard deviation, and the coefficient of variation of the average daily RHR are shown for individual women (green) and men (blue line) in the cohort. Vertical lines indicate 95% confidence intervals in the estimation of the average for each subgroup.

Fig 4. Weekly variation in overall average RHR for women (green line) and men (blue line) over 1 year.

For each individual, and for each of the 52 weeks in a year, the average RHR was calculated considering all available data and interpolating over weeks with no data.

Fig 5

Distribution of the median (A) and maximum (Max) (B) difference in the resting heart rate (RHR) within 1 week and examples of longitudinal RHR variations in three women (C, D, E) over 1 year. For each individual, the median (A) and the maximum difference (B) in RHR over a week has been calculated considering all weeks with at least four RHR measurements. In the longitudinal examples, (C) shows a stable RHR for a woman over age 60 years; (D) a single episode of increased RHR in a woman between 50 and 60 years old; and (E) recurring monthly variability in a woman of childbearing age.