Resting heart rate and all-cause and cardiovascular mortality in the general population: a meta-analysis

Levine [1] found a semi-logarithmic relationship between the rest heart rate (RHR) and life expectancy in mammals and concluded that the total number of heart beats per lifetime is remarkably constant. Furthermore, he calculated that the inverse relationship of RHR and life expectancy in mammals could be expressed mathematically with on average 7.3 +/- 5.6 x 10 [8] heart beats per lifetime. Only humans fall out of this alignment and have reached a about fourfolded life expectancy of about 80 years. However, in 2016 Zhang et al. [2] in this journal presented the first meta-analysis indicating that also in humans the RHR is positively correlated to mortality (9% increase per 10 bpm). However, comparison of the data presented by Zhang et al. [2] with those published in the original publications incorporated in their meta-analysis revealed numerous differences. Here a revised version of the meta-analysis of Zhang et al. [2] adding two references missing by these authors and two more recently published references is presented. Additionally, the impact of an increase of the RHR by 10 bpm on life expectancy in the United States of America (U.S.A.) has been calculated.

Methods

Search strategy The literature searche has been conducted in the scientific database PubMED of the United States National Library of Medicine and EMBASE. Search term was "(resting heart frequency OR resting heart rate OR resting pulse rate) AND all-cause mortality". The search has been conducted by one author (CDR) with the last update on August, 31th, 2016. The literature search was conducted in three steps. First, all articles were screened based on titles. In a second step, abstracts of potentially relevant articles were reviewed. Third, if the abstract indicated that the study potentially fullfilled the eligibility criteria or it did not provide sufficient information for decision making, the full text of the study was screened for eligibility.

Eligibility criteria The following eligibility criteria were used: (i) the article provides information on mean values plus 95% confidence intervals (CI), (ii) the study excluded participants with pregnancy or comorbidities exclucing overweight, obesity, arterial hypertension, and dyslipidamia, (iii) the article has been published in English or German language.

Data analysis The meta-analysis was carried out using the packages meta and metafor in the freely available statistical software R, version 3.2.3 [3]. We report the classical meta-analysis confidence interval as well as the modified CI proposed by Hartung, Knapp [4]. Two-sided p-values < 0.05 are considered as significant. We judged the heterogeneity with Cochran's Q, the homogeneity test statistic, and Higgins-Thompson I2, a measure which judges the impact of heterogeneity.

Calculation of the life expectancy The effects of an increase of RHR by 10 bpm on the probability to die at the age of x years (dx) has been calculated using the formula: dx = d'x x M-OR10 (d'x = probability to die at the age of x years according to the most recently published mortality tables of the U.S.A.[5], M-OR10 = mortality due to increase of RHR by 10 bpm as presented in table 2). The probability to survive until the age of x years (Sx) was calculated employing the following algorithm: Sx = II_k (1 - dx). Applying these algorithms the median life expectancies, i. e. the age at which 50 percent of all individuals have died (Sx = 0.5), have been calculated.

Results The following original publications cited by Zhang et al. [2] have not been included in our meta-analysis due to missing data on the effects of increasing RHR (mean and 95%-CI) on all-cause mortality: Abadag et al. [2009], Benetos et al. [1999], Reunanen et al. [2002], and Saxena et al. [2013] (references: see Zhang et al. [2]). Also, the study of Mai et al. [2009], written in Chinese language, was not included in our meta- analysis. All data included in our meta-analysis and - for comparison - those of Zhang et al. [2] are presented in table 1. As indicated in table 2 an increase of the RHR by 10 bpm was associated with an excess of all-cause mortality of 10% in the studies of women, of 13% in those of men, and 20% of those both in women and men. These elevated mortality rates correspond to a loss of median life expectancy by 1.0, 1.5, and 2.1 years in the US American population.

Discussion In principle, the positive correlation between the resting heart rate as calculated by Zhang et al. [2] has been confirmed. However, the present meta-analysis revealed a considerable greater increase of mortality per 10 bpm. Employing the life tables of the U.S.A. the median life expectancy due to increase of the resting heart rate by 10 bpm corresponds to a median loss of one to two years of life expectancy. Thus, the RHR is a relevant, but less well known mortality factor. However it is unclear, whether the RHR by itself is the mortality risk factor.or only reflects another pathophysiological mechanism increasing mortality, e. g. the sympathetic tone, or both.

References

1. Levine HJ. Rest heart rate and life expectancy. J Am Coll Cardiol 1997;30:1104-6.

2. Zhang D, Shen X, Qi X. Resting heart rate and all-cause and cardiovascular mortality in the general population: a meta-analysis. CMAJ 2016;188:E53-E63.

3. Venables WN, Smith DM, R Core Team. An Introduction to R - Notes on R: A Programming Environment for Data Analysis and Graphics Version 3.2.3. 2015 doi:https://cran.r-project.org/doc/manuals/r-release/R-intro.pdf.

4. Hartung J, Knapp G. On tests of the overall treatment effect in meta- analysis with normally distributed responses. Stat Med 2001;20:1771-82.

5. Arias E. National Vital Statistics Reports: United States Life Tables 2010. 2014;63:7

6. Hozawa A, Inoue R, Ohkubo T, et al. Predictive value of ambulatory heart rate in the Japanese general population: the Ohasama study. J Hypertens 2008;26:1571-6.

7 . Korshoj M, Lidegaard M, Kittel F, et al. The relation of ambulatory heart rate with all-cause mortality among middle-aged men: a prospective cohort study. PLoS ONE 2015;10:e0121729.

8. Nilsson G, Rosenblad A, Leppert J. The resting heart rate strongly predicts all-cause death among middle aged women and men from a general population. Eur J Cardiovasc Prev 2010;Suppl 11:47

9. Parikh KS, Greiner MA, Suzuki T, et al. Resting heart rate and long- term outcomes among the African American population: Insights from the Jackson Heart Study. JAMA Cardiol 2016 Sep 28. doi: 10.1001/jamacardio.2016.3234. [Epub ahead of print]

Conflict of Interest:

None declared