Positive airway pressure improves nocturnal beat-to-beat blood pressure surges in obesity hypoventilation syndrome with obstructive sleep apnea

Abstract

Positive airway pressure (PAP) treatment has been shown to have a modest effect on ambulatory blood pressure (BP) in patients with obstructive sleep apnea (OSA). However, there is a paucity of data on the effect of PAP therapy on rapid, yet significant, BP swings during sleep, particularly in obesity hypoventilation syndrome (OHS). The present study hypothesizes that PAP therapy will improve nocturnal BP on the first treatment night (titration PAP) in OHS patients with underlying OSA, and that these improvements will become more significant with 6 wk of PAP therapy. Seventeen adults (7 men, 10 women; age 50.4 ± 10.7 years, BMI 49.3 ± 2.4 kg/m²) with OHS and clinically diagnosed OSA participated in three overnight laboratory visits that included polysomnography and beat-to-beat BP monitoring via finger plethysmography. Six weeks of PAP therapy, but not titration PAP, lowered mean nocturnal BP. In contrast, when nocturnal beat-to-beat BPs were aggregated into bins consisting of at least three consecutive cardiac cycles with a >10 mmHg BP surge (i.e., Δ10–20, Δ20–30, Δ30–40, and Δ>40 mmHg), titration, and 6-wk PAP reduced the number of BP surges per hour (time × bin, P < 0.05). PAP adherence over the 6-wk period was significantly correlated to reductions in nocturnal systolic (r = 0.713, P = 0.001) and diastolic (r = 0.497, P = 0.043) BP surges. Despite these PAP-induced improvements in nocturnal beat-to-beat BP surges, 6 wk of PAP therapy did not alter daytime BP. In conclusion, PAP treatment reduces nocturnal beat-to-beat BP surges in OHS patients with underlying OSA, and this improvement in nocturnal BP regulation was greater in patients with higher PAP adherence.

over the last few decades, there has been a disproportionately higher increase in the rates of severe obesity (defined as body mass index ≥40 kg/m²) within the ongoing obesity epidemic (32). Indeed, a recent study from the U.S. Centers for Disease Control and Prevention estimates that 6.4% of adults in the United States have severe obesity, and the prevalence is substantially higher at 12.2% among non-Hispanic blacks (28). Obesity hypoventilation syndrome (OHS), a condition defined as daytime hypercapnia [partial pressure of arterial CO₂ (Pa_CO2) ≥45 mmHg] combined with obesity in the absence of other possible causes of hypoventilation, is more prevalent with severe obesity, and the vast majority of patients with OHS have concomitant obstructive sleep apnea (OSA) (21). Several studies from various geographic regions have examined the prevalence of OHS among patients with OSA who were referred to clinical sleep laboratories, and the aggregate prevalence of OHS among these studies was 17%, with increasing prevalence in the more severely obese patients (1).

Thus far, the preponderance of research on cardiovascular morbidity and mortality associated with sleep-disordered breathing has been performed in patients with OSA, typically with no inclusion/exclusion criteria regarding OHS. These studies have confirmed that untreated severe OSA is associated with increased cardiovascular morbidity and mortality. Positive airway pressure (PAP) therapy remains the gold-standard treatment for severe OSA, as well as OHS. In recent years, a number of meta-analyses have shown that PAP therapy elicits modest, yet significant, reductions in 24-h ambulatory blood pressure (BP) (10, 13, 25, 30). Moreover, all four of these meta-analyses report that PAP therapy elicits greater reductions of nocturnal BP compared with daytime BP. However, to date, there have been no comprehensive studies examining the impact of PAP therapy on nocturnal BP control in OHS patients with OSA.

In contrast to the large number of studies examining PAP therapy and 24-h ambulatory blood pressure monitoring (ABPM), there remains a paucity of studies examining PAP therapy on nocturnal beat-to-beat BP (7, 9). In contrast to 24-h ABPM, which typically measures BP in snapshots of 15- to 30-min intervals, beat-to-beat BP monitoring via finger plethysmography offers a reliable, noninvasive approach that can capture rapid oscillations and surges of BP (i.e., up to >40 mmHg lasting 20–30 s), which often occur immediately following obstructive respiratory events, such as apneas and hypopneas, likely via pathways initiated by sympathetic activation (31). Studies examining nocturnal beat-to-beat BP after PAP therapy have primarily focused on baroreflex function. Bongsignore et al. (5) reported that 3 to 14 mo of PAP therapy improved nocturnal baroreflex function in patients with severe OSA. In a subsequent study, it was demonstrated that as little as one night of PAP can improve nocturnal baroreflex function in severe OSA patients (4). It remains unclear whether chronic PAP offers additional benefit to immediate (i.e., one-night) PAP with respect to nocturnal beat-to-beat BP regulation. Moreover, compared with traditional OSA studies, which presumably include mostly eucapnic patients, there is no data on potential nocturnal beat-to-beat BP benefits of PAP therapy in OHS patients with OSA.

It is well recognized that arterial baroreflex dysfunction is associated with heightened cardiovascular risk (18, 26); thus, improved nocturnal baroreflex function associated with PAP therapy is clinically relevant. However, traditional baroreflex analyses do not adequately detail and/or quantify the rapid and extreme changes in BP that occur during disordered breathing while asleep. Therefore, the purpose of the present study was to determine the impact of both immediate (initial PAP titration night) and short-term (6-wk) PAP treatment on nocturnal beat-to-beat BP surges in patients with OHS. We hypothesized that initial exposure to PAP therapy (PAP titration night) would reduce the extreme surges of nocturnal beat-to-beat BP and that these improvements would be more pronounced after 6 wk of PAP treatment.

METHODS

Participants.

This was a substudy of a two-center randomized controlled trial testing the efficacy of three different modalities of PAP therapy in patients with OHS (clinical trial registration NCT01368614). These PAP modalities included average volume-assured pressure support with auto-expiratory positive airway pressure (AVAPS-AE), continuous positive airway pressure (CPAP), and bilevel PAP. The primary study was designed to test the hypothesis that in patients with OHS, the AVAPS-AE mode provides a benefit in daytime gas exchange at 6 wk that is equivalent or no worse than bilevel PAP and CPAP. Our inclusion criteria were age ≥18 and ≤75 yr, diagnosis of OHS in the past 3 mo but no initiation of PAP therapy, body mass index (BMI) of ≥30 kg/m², daytime partial pressure of arterial CO₂ (Pa_CO2) ≥45 mmHg and daytime pH >7.35 from an arterial blood gas, the presence of OSA with an apnea-hypopnea index (AHI) ≥5 events/h of sleep and absence of chronic obstructive pulmonary disease based on forced expiratory volume in 1 s (FEV₁)/forced vital capacity (FVC) > 70%. We excluded patients who were acutely ill and unstable; had been hospitalized for cardiac or respiratory exacerbation <6 wk prior to screening visit; had evidence of alkalosis (pH > 7.45) on the arterial blood gas measurement; had impaired upper airway function (e.g., obstruction due to infections, craniofacial malformations, tumors, uvulopalatopharyngoplasty, presence of tracheostomy, or bilateral vocal cord palsy); or had facial trauma, burns, surgery, or anatomical abnormalities interfering with mask fit.

Between November 2011 and February 2014, 51 patients evaluated in the Sleep Disorders Clinic were referred to the University of Chicago research site; 46 patients attended a screening visit. Two patients did not consent to the substudy, while the remaining 44 patients provided written informed consent. After initial evaluation, 19 were excluded for not having hypercapnia on the arterial blood gases, 2 participants were excluded because they were undergoing active PAP therapy, 1 was excluded due to obstructive airways disease on spirometry (FEV₁/FVC <70%), and 1 had a BMI less than 30 kg/m². An additional four participants did not show up for the first appointment after the initial screening. Thus, the final analytic cohort for the substudy consisted of 17 patients with OHS. The main protocol and the substudy were approved by the University of Chicago Institutional Review Board, and all participants provided written informed consent.

Our 17 participants (7 men, 10 women; age 50.4 ± 10.7 years; 16 black and 1 Hispanic) were severely obese at baseline, and there was no change in the BMI at the end of the study (49.3 ± 2.4 vs. 49.3 ± 2.6 kg/m²). Hypertension was present in 94%, Type 2 diabetes was present in 35%, and congestive heart failure was present in 35% of the participants. Fourteen participants were taking antihypertensive medications (five on two classes and three participants on three classes of antihypertensives). There were no changes in any of the participants' medication regimens during the study. Severe OSA was present in 15 out of 17 (88%) participants.

Polysomnography.

Each participant underwent three in-laboratory polysomnograms (PSG). Bedtimes were from 10:00 PM to 12:00 AM and wake times ranged from 7:00 AM to 9:00 AM. Each PSG included 8 h of recording. PSG (Nihon Kohden, Foothill Ranch, CA) included recordings of six electroencephalographic channels, bilateral electrooculograms, chin and tibialis electromyogram, electrocardiogram, airflow by nasal pressure transducer and oronasal thermocouples, chest and abdominal wall motion by respiratory inductance plethysmography belts, and oxygen saturation by finger pulse oximeter. Transcutaneous CO₂ monitoring was performed during baseline and 6-wk PSG. All PSGs were staged and scored according to the 2007 American Academy of Sleep Medicine's Manual for the Scoring of Sleep and Related Events (14). Apneas were defined as a reduction of airflow of at least 90% on the oronasal thermistor for at least 10 s (it was considered obstructive if respiratory effort was present and central if respiratory effort was absent). Hypopneas were scored when the magnitude of the signal decreased by at least 30% of the baseline amplitude of the nasal pressure transducer for at least 10 s and were associated with a 4% or greater drop in oxygen saturation, as measured by finger pulse oximetry. The total AHI was defined as the number of apneas and hypopneas per hour of sleep. OSA was defined as AHI ≥ 5. Severity of OSA was measured by the AHI. A subject was considered to have mild OSA when the AHI was 5–14, moderate OSA when the AHI was 15–29, and severe OSA when the AHI was ≥30. The oxygen desaturation index (ODI) was defined as the total number of oxygen desaturations of at least 4% per total sleep time in hours. Severity of hypoxemia was further quantified using percent of total recording time with oxygen saturation (SpO₂) below 90%, 80%, and 70%. The arousal index was calculated as the total number of microarousals per hour of sleep. Severity of nocturnal hypoventilation was assessed using the percentage of total recording time with transcutaneous CO₂ above 50 mmHg and 60 mmHg using SenTec Digital Monitor System (Therwil, Arlesheim, Switzerland). The transcutaneous CO₂ sensor is automatically calibrated using one-point dry gas calibration with 8% CO₂. To induce local vasodilation and to enhance skin permeability to carbon dioxide to improve gas diffusion at the measurement sites, the sensor is heated to 42°C.

Positive airway pressure titration and follow up.

Following the completion of the baseline diagnostic PSG, participants were randomly assigned to one of three PAP therapies: 1) average volume-assured pressure support-auto-expiratory positive airway pressure (AVAPS-AE mode), 2) continuous positive airway pressure (CPAP), or 3) bilevel PAP in spontaneous mode. Randomization was generated via a randomization website after the patient underwent the baseline PSG. The PAP titration PSG had to be completed within 7 days of the baseline diagnostic PSG. A single ventilator, OmniLab Advanced (Philips Respironics, Murrysville, PA), was used that had all three modes of therapy. Participants were blinded to the mode of PAP provided. CPAP and bilevel PAP titration were performed following the American Academy of Sleep Medicine's recommendations (3, 17). For AVAPS-AE mode, the target tidal volume was set at 8 ml/kg of ideal body weight with pressure support minimum of 4 and maximum of 26 cmH₂O, expiratory postive airway pressure minimum of 4 and maximum of 20 cmH₂O, and a backup rate in the auto mode. Seven subjects were randomized to AVAPS-AE mode (mean EPAP minimum and maximum 10 ± 1 and 18 ± 1 cm H₂O, mean pressure support minimum and maximum of 6 ± 1 and 20 ± 1 cm H₂O, and mean set tidal volume 546 ± 46 ml), six were randomized to CPAP mode (mean pressure 17 ± 1.7 cm H₂O), and four were randomized to bilevel PAP mode without a backup rate (mean IPAP of 24.5 ± 1.4 and mean EPAP of 13.2 ± 1.7 cm H₂O).

Following the titration PSG, participants were sent home with the OmniLab device set with their randomized therapy. During the initial week, the participants were contacted daily to determine whether there were any problems with the device and to encourage compliance. After the initial 2 wk of home PAP use, participants were asked to return to the research center for a follow-up visit. During this visit, the PAP device was downloaded for compliance information, and issues related to suboptimal compliance and mask fit were addressed. The PAP devices were downloaded once again at the end of the study after 6 wk of PAP therapy to ascertain objective adherence with therapy, which was quantified two ways: 1) PAP therapy use of ≥4 h per night during 6 wk (% of nights) and 2) mean PAP use during 6 wk (hours per night).

Blood pressure.

Clinically measured BP was assessed using automatic standardized sphygmomanometer. Baseline BP included the average of three measured BPs, all of which were obtained in the sitting position while awake and after resting for 20 min. These three BP measurements were obtained during the screening visit, the night, and the morning of the baseline PSG. The 6-wk follow-up BP included the average of three BPs, including the clinic visit at 6 wk, and the night and the morning of the 6-wk PSG.

Continuous beat-to-beat BP was monitored during all three in-laboratory PSGs using the Nexfin HD device (BMEYE B.V., Amsterdam, The Netherlands). Nexfin is an FDA-approved device and uses the latest development of the Finapres methodology, which allows for accurate continuous beat-to-beat noninvasive measurement of BP parameters with high within-subject precision (12, 15, 33). The measurement method is based on the volume-clamp method using a finger plethysmography, which has been shown to be in good concordance with invasive measures of blood pressure and highly reproducible (12, 15, 33). Measurements were regularly and automatically calibrated throughout all recording sessions. Briefly, an appropriately sized finger cuff was applied to the mid-phalanx of the middle finger. A “heart reference system” was used to correct and compensate for hydrostatic level effects due to movements of the measured finger with respect to the reference point at the level of the heart. For each night of recording, we quantified the average of the beat-to-beat BPs from the period of quiet wakefulness prior to initiation of sleep as identified by PSG, and this average obtained from each PSG was used as the baseline noninvasive wake BP for each night of recording. The duration of quiet wakefulness used to obtain the average baseline wake BP was similar across the baseline PSG (807 ± 163 s), titration PSG (669 ± 96 s), and 6-wk PSG (706 ± 117 s; P = 0.749).

To quantify the number of nocturnal BP surges, we used a “binning” approach in which we identified any interval when beat-to-beat BP increased at least three consecutive cardiac cycles (MATLAB, The Mathworks, Natick, MA). When ≥3 cardiac cycle “surge” of BP was identified, the change or delta in BP for that time interval was calculated. These acute surges were then binned into the following five BP bins: 1) number of nocturnal BP surges of 0–10 mmHg, 2) number of nocturnal BP surges of 10–20 mmHg, 3) number of nocturnal BP surges of 20–30 mmHg, 4) number of nocturnal BP surges of 30–40 mmHg, and 5) number of nocturnal BP surges of >40 mmHg. Each bin was normalized for the number of nocturnal hours of recording.

In addition to the primary “binning” analysis described above, we performed an additional analysis (MATLAB, The Mathworks) that binned every nocturnal BP data point into one of five BP bins: 1) less than +10 mmHg compared with quiet wakefulness, 2) 10–20 mmHg above quiet wakefulness, 3) 20–30 mmHg above quiet wakefulness, 4) 30–40 mmHg above quiet wakefulness, and 5) >40 mmHg above quiet wakefulness. From each of those bins, we calculated the percentage of the nocturnal recording spent within each BP bin. Binning analyses were performed for both systolic and diastolic arterial BP.

Gas exchange.

Oxygen saturation by pulse oximetry (SpO₂) and end-tidal CO₂ were measured during baseline visit and after 6 wk of therapy while awake, sitting and resting for 20 min, breathing room air. These measurements were performed between approximately 12:00 and 1:00 PM. Measurement of arterial blood gases were also performed at baseline and after 6 wk of therapy at ∼1:00 PM. After arriving to the pulmonary function laboratory, the participants rested for at least 20 min while breathing room air. Arterial blood gas sample was obtained from the radial artery (RapidPoint 405 series, Siemens Medical Solutions, Malvern, PA).

Questionnaires.

Subjective sleepiness and quality of life were assessed in each patient at baseline and after 6 wk of therapy using the Epworth Sleepiness Scale (ESS) and Severe Respiratory Insufficiency (SRI) questionnaire, respectively. The ESS is an eight-item questionnaire developed to assess subjective sleepiness. The score can range from 0 to 24. A score above 10 is considered to be hypersomnolent (16). The SRI questionnaire is a 49-item health-related quality of life measure developed for use by patients with chronic respiratory failure from a variety of underlying diseases. The score can range from 0 to 100, with higher scores consistent with better quality of life (34).

Statistical analysis.

Continuous variables are presented as means ± SE. For all variables except the BP “binning” comparisons, differences within the three conditions (i.e., baseline, titration, and 6-wk PSG) were tested using one-way ANOVA. Variables measured at only baseline and during visits at 6 wk were compared using paired t-test. For the nocturnal beat-to-beat BP “binning” surge comparisons, we used a one-between (bin: >10, 10–20, 20–30 mmHg), one-within (PSG: baseline, titration, 6 wk) repeated-measures ANOVA. When appropriate, post hoc analyses were conducted using least significant difference method. Pearson correlation tests were used to examine relations between PAP adherence and nocturnal surges in beat-to-beat BP. All tests of significance were two-sided, and a P value <0.05 was considered statistically significant. Data analyses were conducted using IBM SPSS statistics version 22 (SPSS Statistics v20 Armonk, NY: IBM).

RESULTS

PAP adherence.

The mean nightly usage of PAP therapy on all nights (i.e., nights used and not used) was 275 ± 32 min. On average the PAP device was used on 78 ± 7% of the nights. The PAP device was used more than 4 h per night on 58 ± 7% of the nights. There was no significant difference in adherence to PAP therapy among the three PAP modalities. The mean residual AHI, as estimated by the PAP device was 5 ± 3 events/h, suggesting effective resolution of OSA, while the PAP device was being used. The residual AHI was not different among the three different PAP modalities. Moreover, there was no significant difference in the degree of improvement in Pa_CO2 between the three different PAP modalities.

Polysomnography.

Table 1 depicts that compared with baseline PSG, both titration and 6-wk PAP significantly reduced AHI, arousal index, 4% ODI, and amount of recording time below 80% and 70% oxygen saturation (P < 0.05); these changes were not different between titration and 6-wk PAP. In contrast, the amount of time below 90% oxygen saturation decreased with titration PAP, and decreased further after 6-wk PAP. Compared with baseline PSG, total sleep time, REM sleep, and SWS increased during titration PSG, and stage 1 non-REM sleep decreased compared with baseline PSG (Table 1).

Table 1.

Polysomnography measurements

Variable	Baseline PSG	Titration PSG	Six-Week PSG	P Value
TST, min	320.5 ± 21.0	398.1 ± 17.5*	363.5 ± 18.2	0.021
REM, min	41.5 ± 7.4	124.4 ± 10.0*	81.3 ± 9.4*†	<0.001
Slow wave sleep, min	23.4 ± 8.4	70.7 ± 15.3*	52.3 ± 12.7*	0.034
Stage 1, min	73.1 ± 15.2	27.9 ± 5.4*	30.4 ± 4.7*	0.002
Stage 2, min	185.9 ± 22.9	175.0 ± 13.6	199.3 ± 12.7	0.604
AHI, events/h	85.9 ± 8.2	19.7 ± 3.4*	14.5 ± 3.2*	<0.001
Arousal index, events/h	73.3 ± 6.5	17.8 ± 2.8*	20.3 ± 2.2*	<0.001
4% ODI, events/h	94.8 ± 8.8	21.0 ± 3.3*	13.6 ± 3.4*	<0.001
T90, min	306.5 ± 27.4	209.3 ± 38.1*	77.9 ± 25.0*†	<0.001
T80, min	157.8 ± 27.9	61.1 ± 24.6*	16.2 ± 10.5*	<0.001
T70, min	63.8 ± 19.0	23.5 ± 10.7*	4.2 ± 3.7*	0.006

TST, total sleep time; REM, rapid eye movement; AHI, apnea-hypopnea index; ODI, oxygen desaturation index; PSG, polysomnography; T90, amount of recording time below 90% oxygen saturation; T80, amount of recording time below 80% oxygen saturation; T70, amount of recording time below 70% oxygen saturation.

^*P < 0.05 vs. baseline PSG.

^†P < 0.05 vs. titration PSG.

Gas exchange.

Six weeks of PAP therapy decreased Pa_CO2, serum venous bicarbonate, end-tidal CO₂ during wakefulness, peak TcCO₂ during sleep (P < 0.05). Six weeks of PAP therapy, however, did not change pH, Pa_O2, and SpO₂ (Table 2).

Table 2.

Gas exchange measurements

Variable	Baseline Visit	Six-Week Visit	P Value
Gas exchange
pH	7.4 ± 0.01	7.4 ± 0.01	0.890
PaO2, mmHg	58.8 ± 1.9	61.7 ± 2.2	0.115
PaCO2, mmHg*	50.6 ± 1.7	44.0 ± 1.8	<0.001
Serum venous bicarbonate, mEq/l	30.6 ± 1.1	28.5 ± 0.5	0.021
SpO2 during wakefulness, %	91.7 ± 1.02	92.8 ± 1.3	0.379
End-tidal CO2 during wakefulness, mmHg	51.6 ± 1.7	46.4 ± 2.1	0.017
Peak TcCO2 during sleep, mmHg	70.4 ± 3.3	62.8 ± 2.7	0.048
TcCO2 >50 mmHg during sleep, min	238.8 ± 32.4	88.6 ± 23.0	0.002
TcCO2 >60 mmHg during sleep, min	81.4 ± 35.5	21.0 ± 11.6	0.160

Pa_O2, partial pressure of arterial oxygen; Pa_CO2, partial pressure of arterial carbon dioxide; SpO₂, oxygen saturation by pulse oximetry; TcCO₂, transcutaneous carbon dioxide.

^*There was no significant difference in the degree of improvement in PaCO₂ between the three different PAP modalities.

Blood pressure.

Seated measurements of awake daytime systolic (136 ± 4 to 134 ± 4 mmHg; P = 0.538) and diastolic (86 ± 3 to 84 ± 3 mmHg; P = 0.432) arterial BP were not different between baseline and after 6 wk of PAP therapy.

Figure 1 demonstrates that compared with quiet wakefulness, 6 wk of PAP therapy decreased mean values of nocturnal systolic and diastolic BP compared with baseline PSG. Although BP during titration PAP tended to be lower than baseline PSG, this did not reach statistical significance.

Fig. 1.

Changes (Δ) in mean values of nocturnal blood pressure compared with quiet wakefulness. Six weeks of positive airway pressure (PAP) therapy significantly reduced Δsystolic (SAP) and Δdiastolic (DAP). *P < 0.05 vs. baseline polysomnogram (PSG). SAP, systolic arterial pressure; MAP, mean arterial pressure; DAP, diastolic arterial pressure; HR, heart rate.

Figure 2 demonstrates that both titration and 6-wk PAP significantly reduced the number of nocturnal BP surges per hour (i.e., 20–30, 30–40, and >40 mmHg) compared with baseline. These reductions in BP surges were not different between titration and 6-wk PAP. Figure 3 demonstrates that subjects spent a similar percentage of the night in the BP bin depicting 10–20 mmHg above quiet wakefulness. However, there was a significant reduction in the percentage of the night spent within the higher BP bins (i.e., 20–30, 30–40, and >40 mmHg) above quiet wakefulness when comparing baseline to both titration and 6-wk PAP. These reductions were not different between titration and 6-wk PAP.

Fig. 2.

Number (#) of systolic (left) and diastolic (right) blood pressure (BP) surges per hour of nocturnal recording. The higher BP bins (i.e., 20–30, 30–40, and >40 mmHg) revealed that both titration and 6 wk of PAP therapy significantly reduced the number of hourly nocturnal BP surges compared with baseline. *P < 0.05 vs. baseline PSG.

Fig. 3.

Percentage (%) of night when systolic (left) and diastolic (right) blood pressure (BP) bins were above quiet wakefulness. The higher BP bins (i.e., 20–30, 30–40, and >40 mmHg) revealed that titration and 6 wk of positive airway pressure (PAP) therapy significantly reduced the percentage of time spent above quiet wakefulness. *P < 0.05 vs. baseline PSG.

Figure 4 demonstrates that PAP adherence, defined as percentage of nights with ≥4 h of use over the 6 wk of therapy, was significantly correlated to reductions in the number of systolic (r = 0.713, P = 0.001) and diastolic (r = 0.497, P = 0.043) nocturnal BP surges >10 mmHg per hour. When PAP adherence was quantified as mean hours of PAP use per night over the 6 wk, it remained significantly correlated to reductions in the number of systolic (r = 0.612, P = 0.008), but not diastolic (r = 0.390, P = 0.122), nocturnal BP surges > 10 mmHg/h. Adherence to PAP therapy was not correlated to changes in awake daytime systolic (r = −0.154, P = 0.555) or diastolic (r = −0.245, P = 0.342) BP.

Fig. 4.

Associations between PAP adherence and changes in the number of nocturnal surges of BP after 6 wk of PAP therapy. PAP adherence, defined as % of nights with ≥4 h of use over the 6 wk of therapy, was significantly correlated to reductions in the number of SAP and DAP nocturnal BP surges of >10 mmHg per hour. Open circles (○) denote subjects treated with average volume assured pressure support (AVAPS), while closed circles (●) denote subjects treated with bilevel PAP, and triangles (▲) denote subjects treated with continuous positive airway pressure (CPAP).

The reductions in T90 (6-wk PAP-baseline) were highly correlated to reductions in systolic (r = 0.668, P = 0.003) and diastolic (r = 0.658, P = 0.004) nocturnal BP surges after 6 wk of PAP therapy. Moreover, there was a trend for a correlation between the improvement of T90 at 6 wk and PAP adherence (r = −0.471, P = 0.056). Consistent with the T90 data, the 4% ODI tended to correlate to changes in systolic (r = 0.361, P = 0.155) and diastolic (r = 0.470, P = 0.057) nocturnal BP surge.

Questionnaires.

After 6 wk of PAP, there was a significant improvement in SRI summary score (57 ± 4 to 66 ± 4 units; P = 0.014) and ESS (14 ± 1 to 8 ± 1 units; P = 0.001). Changes in SRI and ESS were not correlated to changes in either nocturnal BP surges > 10 mmHg per hour or seated awake daytime BP.

DISCUSSION

The present study aimed to determine the impact of both immediate (initial titration night) and short-term (6-wk) PAP treatment on nocturnal beat-to-beat BP reactivity (i.e., BP surges) in OHS patients with underlying OSA. We present several novel findings. First, when mean values of beat-to-beat BP for quiet nocturnal wakefulness and nocturnal sleep hours are examined, the therapeutic effects of PAP were observed only after 6 wk of treatment (i.e., titration PAP was not significantly different from baseline PSG). Second, in contrast to the mean BP comparisons, both titration and 6-wk PAP reduced the number of nocturnal BP surges per hour within the higher BP bins. The reductions in the number of nocturnal BP surges were not different between titration and 6-wk PAP. Third, PAP adherence over the 6-wk period was strongly correlated to reductions in the number of nocturnal BP surges. Finally, despite PAP-induced significant reductions in nocturnal beat-to-beat BP surges, 6 wk of PAP therapy did not alter daytime BP. Taken together, our findings demonstrate that PAP treatment elicits immediate reductions of nocturnal beat-to-beat BP surges in patients with OHS and that PAP adherence has a marked impact on the level of this reduction after 6 wk of therapy. These improvements in nocturnal BP control with PAP therapy occurred despite no observed change in daytime BP. Thus, even when PAP does not lower daytime BP, it can reduce extreme surges of nocturnal BP, which might have important long-term cardiovascular benefits.

Our study population offers some unique advantages relevant to current societal health issues. First, OHS is one of the most significant obesity-related conditions in the world. Compared with patients with eucapnic OSA and a similar degree of obesity, patients with OHS have significantly increased risk of cardiovascular mortality (6). Given the increasing rates of severe obesity (28), it is more critical than ever to examine this patient population. Second, all of our subjects were minority men and women, with the vast majority (16 of 17 subjects) being non-Hispanic black. Although an estimated 6.4% of adults in the United States have severe obesity, the prevalence is nearly double (∼12.2%) among non-Hispanic blacks (28). Given the recent emphasis by the National Institutes of Health and other leading health organizations on health disparities, our study population offers a unique insight into the racial group most severely impacted by OHS.

To our knowledge, this is the first study to systematically quantify the nocturnal beat-to-beat BP surges in patients with OHS, and determine the impact of both immediate and short-term PAP therapy on nocturnal BP regulation. Prior work examining the potential therapeutic influence of PAP on BP has primarily focused on 24-h ABPM in OSA patients without specific inclusion/exclusion criteria regarding OHS status. Several meta-analyses have converged to conclude that PAP therapy reduces 24-h ambulatory BP in OSA patients and that these reductions are greater at night compared with daytime (10, 13, 25, 30). Unfortunately, 24-h ABPM does not provide insight into the beat-to-beat BP oscillations and surges that occur throughout sleep. Moreover, the limited number of studies that have examined the influence of PAP on nocturnal beat-to-beat BP have focused on baroreflex function (4, 5). These studies have reported that PAP elicits both immediate (4) and short-term (5) improvements in baroreflex function, but what do these “improvements” in baroreflex function mean from a clinical perspective? Do they translate to better nocturnal BP control and fewer dangerous BP surges during sleep?

Consistent with the nocturnal baroreflex studies (4, 5), we found that both immediate (titration night) and short-term (6-wk) PAP therapy significantly improved nocturnal BP control. Specifically, our findings demonstrate that PAP therapy significantly reduces the number of hourly nocturnal beat-to-beat BP surges, particularly BP surges >20 mmHg. This quantification provides clinicians with meaningful numbers that can be understood, and potentially translated, to their patients. For example, this cohort of OHS patients demonstrated nearly 50 systolic BP surges of >40 mmHg per hour, but one night of PAP reduced the number of surges by ∼50%, and 6 wk of PAP reduced the number of surges by nearly 75%. It is certainly plausible that the initial reduction in BP surges observed on the night of PAP titration may have been related to the significant reduction of intermittent hypoxemia and microarousals associated with obstructive apneas and hypopneas, thereby reducing sympathetic surges associated with untreated obstructive events (29). Additional reduction in BP surges at 6 wk may have been due to an improvement in baroreflex function, further improvements in nocturnal hypoxemia (i.e., T90), or both. We speculate that the lack of improvement in daytime BP may be related to the fact that 14 out of 17 patients were being treated with antihypertensive medications (eight were on two or more antihypertensive agents), and baseline daytime BP was relatively well controlled (mean BP ∼136/86 mmHg). Moreover, several meta-analyses have reported that PAP therapy elicits greater reductions of nocturnal BP, when patients experience sympathetic surges due to arousals and intermittent hypoxemia, than daytime BP (10, 13, 25, 30).

As depicted clearly in Fig. 4, 6 wk of PAP adherence had a pronounced impact on nocturnal beat-to-beat BP control. Specifically, we observed a very robust association between 6-wk adherence and the reduction in the number of nocturnal systolic BP surges. We believe this to be one of the most clinically impactful findings of the present study, particularly when coupled with the finding that daytime BP did not change with 6 wk of PAP regardless of adherence. Moreover, these associations between PAP adherence and reductions in nocturnal BP surges do not appear to be related to PAP modality. Thus, our data suggest PAP therapy has very important beneficial effects on nocturnal BP that might not be captured with simple daytime BP measurements, or even 24-h APBM.

It is important to note that in contrast to simple eucapnic OSA, it is not uncommon for some degree of nocturnal hypoxemia to persist during PAP titration in patients with OHS. This may be due to restrictive thoracic limitations related to severe obesity or the incomplete resolution of apneas and hypopneas (2, 24, 29). As demonstrated in the present study, we observed that reductions in T90 (i.e., 6-wk PAP-baseline) were strongly correlated to reductions in nocturnal BP surges after 6 wk of PAP therapy. These improvements of T90 at 6 wk were moderately correlated to PAP adherence, a finding that is is consistent with a prior study (24), demonstrating that higher levels of adherence are significantly associated with improvements in gas exchange. This finding suggests that residual intermittent hypoxemia may play an important role in the degree of nocturnal BP surge reduction after 6 wk of PAP therapy.

We acknowledge several limitations. First, this study focused on OHS patients, while prior studies examining PAP therapy on 24-h ABPM or baroreflex function have focused on OSA patients. Thus, some might suggest our findings have limited generalizability. However, given prior studies that have demonstrated pronounced oscillations and surges of nocturnal beat-to-beat BP in severe OSA patients (8, 31), we are confident that the general conclusions of the present study (i.e., PAP therapy reduces severe nocturnal BP surges and PAP adherence matters) hold true for non-OHS patients with severe OSA. Second, although three different modalities of PAP therapy were used in this study, there was no significant difference in adherence or the residual AHI (as estimated by the PAP device) between the modalities. Moreover, several randomized controlled trials comparing various PAP modalities have not been able to demonstrate significant differences in resolution of hypercapnia after 2–3 mo of PAP therapy (19, 27, 29). Third, we do not report any nocturnal BP data specific to sleep stages (i.e., REM, slow wave sleep, etc.). While we had simultaneous PSGs with the beat-to-beat BP, the Nexfin device was not synced to the PSG. This also prevented traditional baroreflex analysis, as the ECG was recorded through PSG software. Given the findings of this study, future work might probe more specifically into nocturnal beat-to-beat BP regulation during particular sleep stages and should include traditional baroreflex analysis to provide further mechanistic insight. Of particular interest would be REM sleep, as recent studies have begun to establish that obstructive apneas and hypopneas during REM sleep can have a more deleterious influence on 24-h ABPM than events in non-REM sleep (22, 23). Lastly, daytime BP was calculated from three controlled measurements taken throughout the day before and after PAP therapy, but we acknowledge that more frequent and regular measurements of daytime BP (e.g., ABPM) would have been valuable.

Perspectives and Significance

Our findings demonstrate that PAP therapy significantly reduces surges in nocturnal BP in OHS patients with underlying OSA and that this improvement in nocturnal BP regulation was much greater in patients with higher PAP adherence. This is of particular relevance given that PAP therapy did not appear to alter daytime BP. These findings suggest that when PAP therapy is effectively administered (i.e., high adherence), it can have a marked impact on improving nocturnal BP control, even if daytime or ambulatory BP is unaltered. We suggest that future studies examining the role of PAP therapy on BP, either in OHS patients or eucapnic OSA patients, could substantially benefit from noninvasive beat-to-beat finger plethysmography methods and analyses. Moreover, we are hopeful that the novel quantitative approach of examining nocturnal BP surges as presented in this study (i.e., binning method) can prove to be complementary to traditional baroreflex analysis in future studies.

GRANTS

This study was supported in part by Philips/Respironics. B. Mokhlesi is supported by National Institutes of Health Grant R01HL-119161. J. R. Carter is supported by National Heart, Lung, and Blood Institute (HL-122919-01).

DISCLOSURES

B. Mokhlesi has served as a consultant to Philips/Respironics and has received research support from Philips/Respironics. He has also received honorarium from Zephyr Medical Technologies and has served on the advisory board of Itamar Medical. The other authors have no conflicts of interest to declare.

AUTHOR CONTRIBUTIONS

Author contributions: B.M. conceived and designed the study and performed the experiments; B.M. obtained study funding; J.R.C., I.T.F., D. Grimaldi, and B.M. analyzed data; J.R.C., I.T.F., D. Grimaldi, L.E., D. Gozal, C.E.S., and B.M. interpreted results of experiments; J.R.C., I.T.F., and D. Grimaldi prepared figures; J.R.C. drafted manuscript; J.R.C., I.T.F., D. Grimaldi, L.E., D. Gozal, C.E.S., and B.M. edited and revised manuscript; J.R.C., I.T.F., D. Grimaldi, L.E., D. Gozal, C.E.S., and B.M. approved final version of manuscript.