Table 3.

Frequency, duration, equipment, HRV measures, and results

Study	Session frequency and duration1	BF intervention equipment	HRV parameters	HRV outcomes	Reported effect sizes
Bradley et al. (2010)	2 per week for 5 months	Freeze-Framer/emWave	RR SDRR Ln LF Ln HF LnTP LnCoherence ratio	Pre–post Exp vs. Ctrl: $\uparrow$ RR $\uparrow$ SDRR $\uparrow$ LnLF $\uparrow$ LnHF $\uparrow$ LnTP   $\uparrow$ Lncoherence ratio	ES: 0.46 0.64 0.55 0.58 0.64 0.52
de Bruin et al. (2016)	5 weeks	StressEraser	N/A
Ginsberg et al. (2010)	1 per week for 4 weeks	emWave	LF HF VLF TP Coherence ratio	Pre–post: $\uparrow$ LF ns ns $\uparrow$ TP $\uparrow$ coherence ratio	Missing
Groeneveld et al. (2019)2	30 sessions BF + NF (30–40-min sessions)	ProComp Infiniti + BioGraph	VLF LF HF	Pre–post: $\downarrow$ %VLF $\uparrow$ %LF $\downarrow$ %HF	Cohen’s d =  − 0.57 Cohen’s d = 1.32 Cohen’s d =  − 1.27
Groeneveld et al. (2019)			VLF LF HF	Pre–post: $\downarrow$ %VLF $\uparrow$ %LF $\downarrow$ %HF	Cohen’s d =  − 0.46 Cohen’s d = 0.89 Cohen’s d =  − 0.72
Jester et al. (2019)	2 per week, for 3 weeks (6 30-min sessions)	emWave	N/A
Kenien (2015)	12 weeks (12 20-min sessions)	emWave	N/A
Kim et al. (2013)	10 weeks (10 60-min sessions)	emWave	LF/HF Coherence ratio	Pre–post: $\uparrow$ LF/HF $\uparrow$ coherence ratio	${\eta }_{\mathrm{p}}^2=0.452$ ${\eta }_{\mathrm{p}}^2=0.390$
Lee and Finkelstein (2015)	2 visits (2 10-min sessions)	StressEraser + Zephyr	RR SDNN RMSSD LF HF LF/HF	Pre–post: ns ns ns ns ns ns	Missing
May et al. (2019)	3 weekly 20-min sessions 4 weeks	emWave	HFnu3 LFnu	Pre–post Exp: $\downarrow \mathrm{L}\mathrm{F}\mathrm{n}\mathrm{u}$ Pre–post HIIT: $\downarrow \mathrm{L}\mathrm{F}\mathrm{n}\mathrm{u}$ Pre–post Ctrl: $\downarrow \mathrm{L}\mathrm{F}\mathrm{n}\mathrm{u}$	${\eta }_{\mathrm{p}}^2=0.700$ ${\eta }_{\mathrm{p}}^2=0.700$ ${\eta }_{\mathrm{p}}^2=0.700$
Pop-Jordanova and Chakalaroska (2008)	5 sessions	Freeze-Framer/emWave	Coherence ratio4	Pre–post: Missing
Prinsloo et al. (2011)	1 session (1 10-min session)	BioPac/StressEraser	TP LF	Pre–post Exp vs. Ctrl: $\uparrow \mathrm{T}\mathrm{P}$ $\uparrow \mathrm{L}\mathrm{F}$	Missing
Raaijmakers et al. (2013)	7 sessions within 16 days	Active Two	RMSSD	Pre–post Exp vs. Ctrl: ns	Missing
Rusciano et al. (2017)	15 BF sessions 2 per week (30 min)	NeXus-10 Mark II	LF	Resting pre–post Exp: $\uparrow \mathrm{L}\mathrm{F}$	Cohen’s d = 0.89
Schumann et al. (2019)	5 sessions per week (4 home session and 1 lab session) 8 weeks	BioPac/Elite HRV	RMSSD SDNN TP RSA BRS	Pre–post Exp vs. Ctrl: $\uparrow$ RMSSD ns ns ns $\uparrow$ BRS	Missing
Sherlin et al. (2010)	1-day session (15 min)	StressEraser + NeXus-10	N/A
Sutarto et al. (2013)	5 weeks (30–50 min)	I-330 C2	TP LF	Pre–post Exp vs. Ctrl: TPbase ns TPstr ns TPrec ns $\uparrow$ LFbase $\uparrow$ LFstr $\uparrow$ LFrec	Missing

Arrows show the direction of the outcome measure. Upward arrows show an increase of the HRV parameter measure, and downward arrows show a decrease of the HRV parameter measure. “Pre–post Exp vs. Ctrl” refers to an interaction effect where the experimental group saw greater significant change following the intervention than the control group. “Pre–post” refers to a significant change following the intervention where the experimental group was not opposed to a control group (single group design or different comparison groups). “Pre–post Exp” and “Pre–post Ctrl” refers to a main effect of time for the experimental group following the intervention and for the control group not assigned to the intervention. “Post Exp vs. Ctrl” refers to a significant main effect of the experimental group compared to the control group after the intervention. HFnu and LFnu are expressed in normalized units. Nu is calculated by dividing the power of a given frequency component by the total power from which the power of very low frequencies has been subtracted (i.e., LFnu = LF / (LF + HF)); TP: sum of the energy in all the frequency bands; LF/HF: ratio LF [ms²] / HF [ms²]; TP/LF base, str, rec: measures referring to baseline, stressor, and recovery phase; Coherence ratio: Peak power / (Total power − Peak power)

SDNN standard deviation of NN intervals, RMSSD square root of the mean square differences between adjacent NN intervals, LF low frequency (0.04–0.15 Hz), HF high frequency (0.15–0.4 Hz)

¹Where available, duration (minute per session in parenthesis) is specified

²This study appears in two entries to separate the HRV outcomes of the two subgroups (adults and children)

³HF and LF are expressed in normalized units. LFnu = 1 − HFnu, LFnu = LF / (LF + HF). LFnu is considered as an index of cardiac sympathovagal tone, so that a reduction of LFnu is associated to an increase of autonomous nervous system modulation (Sgoifo et al., 2015)

⁴Only measured in the HRV-BF group