Exercise Frequency Determines Heart Rate Variability Gains in Older People: A Meta-Analysis and Meta-Regression

• Published in: Sports medicine (Auckland) 2019-05, Vol.49 (5), p.719-729
• Main Authors: Raffin, Jérémy, Barthélémy, Jean-Claude, Dupré, Caroline, Pichot, Vincent, Berger, Mathieu, Féasson, Léonard, Busso, Thierry, Da Costa, Antoine, Colvez, Alain, Montuy-Coquard, Claude, Bouvier, Rémi, Bongue, Bienvenu, Roche, Frédéric, Hupin, David
• Format: Article
• Language: English
• Subjects: Aerobics; Aged; Aging; Bias; Chaos theory; Endurance; Endurance Training; Exercise; Fatigue tests; Fitness training programs; Frailty; Frequency response; Geriatrics; Heart Rate; Heterogeneity; Human health and pathology; Humans; Life Sciences; Medicin och hälsovetenskap; Medicine; Medicine & Public Health; Meta-analysis; Middle age; Older people; Parasympathetic nervous system; Physical fitness; Physical training; Randomized Controlled Trials as Topic; Regression Analysis; Sports Medicine; Standard deviation; Statistical analysis; Systematic Review; Tissues and Organs; Variability
• ISSN: 0112-1642; 1179-2035; 1179-2035
• DOI: 10.1007/s40279-019-01097-7

Background Previous studies have suggested that exercise training improves cardiac autonomic drive in young and middle-aged adults. In this study, we discuss the benefits for the elderly. Objectives We aimed to establish whether exercise still increases heart rate variability (HRV) beyond the age of 60 years, and to identify which training factors influence HRV gains in this population. Methods Interventional controlled and non-controlled studies were selected from the PubMed, Ovid, Cochrane and Google Scholar databases. Only interventional endurance training protocols involving healthy subjects aged 60 years and over, and measuring at least one heart rate global or parasympathetic index, such as the standard deviation of the normal-to-normal intervals (SDNN), total frequency power (Ptot), root mean square of successive differences between adjacent NN intervals (RMSSD), or high frequency power (HF) before and after the training intervention, were included. HRV parameters were pooled separately from short-term and 24 h recordings for analysis. Risks of bias were assessed using the Methodological Index for Non-Randomized Studies and the Cochrane risk of bias tool. A random-effects model was used to determine effect sizes (Hedges’ g ) for changes, and heterogeneity was assessed using Q and I statistics. Results Twelve studies, seven of which included a control group, including 218 and 111 subjects, respectively (mean age 69.0 ± 3.2 and 68.6 ± 2.5), were selected for meta-analysis. Including the 12 studies demonstrated homogeneous significant effect sizes for short-term (ST)-SDNN and 24 h-SDNN, with effect sizes of 0.366 (95% CI 0.185–547) and 0.442 (95% CI 0.144–0.740), respectively. Controlled study analysis demonstrated homogeneous significant effect sizes for 24 h-SDNN with g  = 0.721 (95% CI 0.184–1.257), and 24 h-Ptot with g  = 0.731 (95% CI 0.195–1.267). Meta-regression analyses revealed positive relationships between ST-SDNN effect sizes and training frequency ( Tau res 2  = 0.000; I res 2  = 0.000; p  = 0.0462). Conclusion This meta-analysis demonstrates a positive effect of endurance-type exercise on autonomic regulation in older adults. However, the selected studies expressed some risks of bias. We conclude that chronic endurance exercise leads to HRV improvements in a linear frequency–response relationship, encouraging the promotion of high-frequency training programmes in older adults.