Meta-analysis of age-based maximum heart rate prediction equations: validating existing models across diverse populations

Purpose: Maximal heart rate (MHR) plays a crucial role in guiding exercise recommendations and monitoring in both clinical and sports contexts. Nevertheless, prediction equations designed for adults may not accurately predict MHR in youth. This study aims to systematically review and analyze the existing evidence on the validity of commonly used age-based MHR prediction models across participants of different ages. Methods: The inclusion criteria encompassed peer-reviewed articles published in English that compared measured and predicted MHR values in male and female participants. To gauge the accuracy of age-predicted MHR values, the standardized mean difference effect size (ES) was employed. Furthermore, predefined moderators were examined to identify potential sources of variability. Results: The cumulative findings from 29 effects obtained from nine articles demonstrated that prediction equations did not statistically significantly differ from zero MHR (ES= 0.24, p = 0.48), while individual effects (z = 1.99, p < .0.05) varied across the studies. Subgroup analyses indicated that the Fox, Nes and Londeree equations tended to overestimate MHR, while the Tanaka, Gelish and Arena equations have better accuracy with less mean bias. Conclusion:  Age-based MHR equations vary across the different age groups. However, if the use of age-based equations is unavoidable, our recommendation is to employ the Tanaka equation, taking into account the reported range of error in this study.