Approval was obtained from the University and Medical Center Institutional Review Board (UMCIRB). The sample consisted of young adults with the inclusion criteria: (1) between 18 and 35 years of age, (2) family history of CVD, and (3) English speaking. Exclusion criteria consisted of (1) documented diagnosis of HTN, (2) currently prescribed antihypertensive medication, and (3) cognitive impairments that inhibit understanding of instructions. Recruitment flyers were posted in public locations. Those interested contacted the research team and were scheduled for an appointment to determine eligibility. Appointments were scheduled in a BF clinic setting where the study was explained in detail and informed consent was obtained on eligible subjects. A convenience sample of adults (N = 34) was recruited.
Investigator tools were developed to obtain participants’ demographics, health assessment data, and CV family history. Demographic intake data included: age, sex, ethnicity, marital status, educational level, occupation, and current employment status. Health assessment data included: current medical conditions, current medication, co-existing medical conditions, prescription medications, over-the-counter medications, exercise frequency, intensity, duration, dietary preferences, smoking history, and a self-rating of overall health.
Family history intake included identification of family members with CVD, the specific disease(s) including myocardial infarction, hypertension, stroke, hyperlipidemia, diabetes mellitus, CV related surgeries, overall general health of immediate family members, if family members were alive or deceased and if deceased at what age. All tools were administered via the Research Electronic Data Capture (REDCap) system.
Measurements of HRV were obtained using an ear-clip photoplethysmography and software (Heart Tracker, Biocom Technologies, USA) for analyses. The standard of HRV historically relied on electrocardiogram readings. However, with advanced technology, photoplethysmography has shown equivalence in measurement for short-term HRV readings.  For this short-term HRV study of young adults who were normatively healthy, photoplethysmography was appropriate for the measurement of HRV parameters.
Low frequency (LF), high frequency (HF), and very low frequency (VLF) are different frequencies with varying power hertz ranges.  These ranges equate to measurements that are expressed in milliseconds squared (ms2). LF can reflect sympathetic, parasympathetic activity, and blood pressure regulation by baroreceptor activity at rest. With a paced slow breathing exercise, the baroreceptor response may increase LF activity. HF is synchronous with respiratory efforts and reflects parasympathetic action. Causes of low HF may include stress and anxiety. VLF is associated with health. Abnormalities in this frequency may be associated with mortality. The ratio of LF to HF is expressed as LF/HF and is used to quantify the degree of sympathetic and parasympathetic balance in the body. LF/HF is also expressed in ms2. If the LF/HF ratio is low then parasympathetic action is dominant, while if high, it reflects sympathetic action. Total power (TP) is a cumulative reflection of the HF, LF, VLF and ultra-low frequency spectral bands and reflects how much power the body has in its ability to adapt. A decrease in this value will typically be seen when the body is under stress. .
The time domains measured in this study included root mean squared (RMS) and the standard deviation of N to N (SDNN). The RMS time domain related to HRV is the mathematical calculation resulting in the beat-to-beat variance in heart rate and reported as milliseconds (ms). HF domain correlates with RMS. SDNN is the standard deviation of normal sinus beat-to-beat variances measured in ms. Both sympathetic and parasympathetic activity contribute to SDNN measurement with it being highly correlated with the LF and VLF frequency bands.
Other biometrics included noninvasive continuous BP measurements using the Continuous Non-Invasive Arterial Pressure (CNAP) system with a slip-on design finger sensor (CNSystem®, Austria). The CNAP system is one of a few devices currently available for noninvasive continuous BP measurement. The CNAP is most commonly used during intraoperative procedures as it allows for uninterrupted recordings over long durations. CNAP devices, using an inflatable finger cuff or FINger Arterial PRESsure (Finapres) system, measure arterial pressure based on the principle of dynamic vascular unloading in the arterial walls within the finger. Finapres systems were originally developed in the 1980 s to provide reliable continuous blood pressure monitoring and have proven a reliable alternative for invasive measurements for mean and diastolic pressures . Imholz and colleagues  reviewed Finapres technology and concluded the accuracy and precision are sufficient for tracking BP changes. Though recent research has contended that CNAP systems are inaccurate for patient care decision making, a CNAP system was used in this study for its ability to noninvasively and continuously monitor patient blood pressure in the research setting.  Additionally, with each participant, an initial BP measurement was obtained using a traditional BP cuff and these measures were correlated with the initial CNAP measurement. A clear advantage of using continuous BP measurement is the change is noted instantly with intervention, such as paced breathing. However, it is not advantageous to use at home due to equipment complexity since was developed to be used in healthcare settings.
Home training tools consisted of a digital application (app) for paced breathing; Breath+ (iPhone) and Breathe2Relax (Androids). Although no research has been done crossing two applications, both are designed to allow users to set breathing pace rates depending on which type of mobile phone they possess. For our study, these applications were used primarily to reinforcement the paced breathing taught in the BF clinic during the initial session. The two free applications were provided to participants for their at-home training because both are designed so that users can set a paced breathing rate. These specific mobile applications have a similar user interface to the Heart Tacker software used during the initial face to face training session at the BF clinic. The applications allowed participants with either operating system (apple or android) to have access to a mobile paced breathing application to continue training at home. There are minimal differences between these applications.
In this pilot study, participants were seen in our BF clinic, a private, temperature-controlled room with no ambient noise. The room’s physical design remained consistent throughout the study with the same BF chair, BF equipment, software for collecting heart rate variability data, paced breathing training software, and a continuous blood pressure monitoring system. As we assessed the practicality of our study and wanted to assure participant retention, a 3-session design, with each session lasting 1 to 1 ½ hours, was agreed upon. A short-term HRV protocol was utilized. According to Voss and colleagues , short-term HRV is acceptable and has the benefits of providing immediate results in settings or with participants with time restraints. An individual’s heart rate and breathing synchronize at a specific breathing rate; this synchronization is called their resonance frequency. Each person has a unique resonance frequency related to their breathing rate, and breathing at a rate outside of this frequency can cause stress and impact HRV measures.  Many studies have found maximum effects while participants are breathe at approximately six breaths per minute or 0.1 Hz though there is evidence that resonance can occur at lower frequencies. .
The first session provided an intake of baseline data and pre-testing. The second session involved a training session with detailed home instruction for continuing training. The third session consisted of post-testing. All sessions were documented and completed by trained BF providers.
Before each session, participants were instructed not to consume caffeinated beverages within 3 h, not to eat a heavy meal within two hours, not to engage in aerobic exercise for at least one hour, and not to smoke within 30 min before the session. Participants were also asked to bring their mobile devices to each appointment.
At the baseline session, participants were placed in the quiet clinic room to complete all intake questionnaires. A brief description followed, describing the equipment, purpose, and procedures. Equipment was connected to the participant, who was asked to sit quietly for five minutes with legs uncrossed, and feet flat and comfortably on the floor. Baseline BP measurements were obtained, and a comparison was made between arm cuff pressures and finger cuff pressures for validation. Afterwards, skin preparation for BP monitoring and HRV sensor application was completed. BF sensors were connected, and with the participant breathing normally, baseline measurements for BP and HRV were obtained for five minutes. For consistency, in our protocol, all participants completed paced breathing at a rate of 5.5 to 6 breaths per minute, using guidance by an established HRV protocol. .
The second session included giving the participant thorough instructions on breathing using a paced breathing computer software application. The reinforcement of the training session was enhanced with a visual screen depicting each participant’s progress and compliance. Prior to the paced breathing training, a baseline BP reading was obtained. During the five-minute paced breathing cycle, BPs were obtained every minute. Participants were observed for any respiratory compromise. Signs of hyperventilation, including lightheadedness, dizziness, or an increase in heart rate were monitored. All data were recorded from the baseline and training sessions.
After the in-clinic session was completed, participants were assisted in loading the breathing app on their mobile devices. Two breathing apps were used with the breath settings standard across both devices to accommodate both iPhone and Android users. Theses apps modeled the same paced breathing as the software program used in the BF clinic setting. Participants were instructed to breathe at the resonant frequency of 5.5 to 6 breaths per minute, twice a day for 10 min each at-home session until returning for their next visit to the BF clinic.
The final session followed the same procedure as the second session for consistency in collecting and comparing data. In addition, participant intake also included the frequency and length of their mobile app training sessions. Again, participants were observed for any signs of hyperventilation. Participants were given visual copies of their progress from baseline to session 3, along with a detailed explanation, and then thanked for their participation. Participants were also encouraged to continue using the app for self-management purposes as needed.
Descriptive statistics were used for sample description. Independent t-tests were used for the continuous variables of heart rate (HR), SBP, DBP and, HRV measures at baseline and after training. Correlation using Pearson-r was used to determine the strength and direction of variable relationships. Multiple regression was used to determine the variance of HRV on both SBP and DBP. All data were analyzed using Statistical Package for the Social Sciences (SPSS), version 26.0.