Skip to main content
NCBI home page
Chest logoLink to Chest
. 2020 Apr 9;158(2):765–775. doi: 10.1016/j.chest.2020.02.074

Effect of Venlafaxine on Apnea-Hypopnea Index in Patients With Sleep Apnea

A Randomized, Double-Blind Crossover Study

Christopher N Schmickl a, Yanru Li a,b,∗, Jeremy E Orr a, Rachel Jen c, Scott A Sands d, Bradley A Edwards e,f, Pamela DeYoung a, Robert L Owens a, Atul Malhotra a
PMCID: PMC7417375  PMID: 32278781

Abstract

Background

One of the key mechanisms underlying OSA is reduced pharyngeal muscle tone during sleep. Data suggest that pharmacologic augmentation of central serotonergic/adrenergic tone increases pharyngeal muscle tone.

Research Question

We hypothesized that venlafaxine, a serotonin-norepinephrine reuptake inhibitor, would improve OSA severity.

Study Design and Methods

In this mechanistic, randomized, double-blind, placebo-controlled crossover trial, 20 patients with OSA underwent two overnight polysomnograms ≥ 4 days apart, receiving either 50 mg of immediate-release venlafaxine or placebo before bedtime. Primary outcomes were the apnea-hypopnea index (AHI) and peripheral oxygen saturation (Spo2) nadir, and secondary outcomes included sleep parameters and pathophysiologic traits with a view toward understanding the impact of venlafaxine on mechanisms underlying OSA.

Results

Overall, there was no significant difference between venlafaxine and placebo regarding AHI (mean reduction, –5.6 events/h [95% CI, –12.0 to 0.9]; P = .09) or Spo2 nadir (median increase, +1.0% [–0.5 to 5]; P = .11). Venlafaxine reduced total sleep time, sleep efficiency, and rapid eye movement (REM) sleep, while increasing non-REM stage 1 sleep (Pall < .05). On the basis of exploratory post hoc analyses venlafaxine decreased (“improved”) the ventilatory response to arousal (–30%; P = .049) and lowered (“worsened”) the predicted arousal threshold (–13%; [P = .02]; ie, more arousable), with no effects on other pathophysiologic traits (Pall ≥ .3). Post hoc analyses further suggested effect modification by arousal threshold (P = .002): AHI improved by 19% in patients with a high arousal threshold (–10.9 events/h [–3.9 to –17.9]) but tended to increase in patients with a low arousal threshold (+7 events/h [–2.0 to 16]). Other predictors of response were elevated AHI and less collapsible upper airway anatomy at baseline (|r| > 0.5, P ≤ .02).

Interpretation

In unselected patients, venlafaxine simultaneously worsened and improved various pathophysiologic traits, resulting in a zero net effect. Careful patient selection based on pathophysiologic traits, or combination therapy with drugs countering its alerting effects, may produce a more robust response.

Trial Registry

ClinicalTrials.gov; No.: NCT02714400; URL: www.clinicaltrials.gov.

Key Words: lung, OSA, pharmacotherapy, pharyngeal muscle tone, venlafaxine

Abbreviations: AHI, apnea hypopnea index; ArTH, arousal threshold; NREM, nonrapid eye movement; PSG, polysomnogram; REM, rapid eye movement

OSA is characterized by a repetitive collapse of the upper airway at night, which affects at least 10% of the US population and has been associated with various adverse health outcomes.1, 2, 3 The standard of care is CPAP,4 but many patients are unable to tolerate this therapy long-term.5, 6, 7, 8 Thus, over the past three decades various pharmacologic treatments have been evaluated as alternative options, but so far results have been mixed.9

One of the key goals of pharmacologic interventions for OSA has been to augment pharyngeal muscle tone during sleep: during wakefulness patients with OSA maintain airway patency via increased pharyngeal muscle tone compared with control subjects.10 However, central activation of the pharyngeal muscles is state-specific, and in patients with OSA muscle tone drops both rapidly and substantially during the wake-sleep transition; this can contribute to the development of apneas and hypopneas.11 Initially it was thought that this change in activity was due primarily to a drop in serotoninergic input to the hypoglossal motoneurons.12,13 Therefore, several studies tried to mitigate OSA by raising serotonin levels (eg, with selective serotonin reuptake inhibitors),14, 15, 16 but overall success has been limited. More recent data suggest that a drop in noradrenaline levels may play a key role as well.17, 18, 19 Therefore, we tested the hypothesis that venlafaxine (a serotonin-norepinephrine reuptake inhibitor) improves OSA severity.

Methods

Study Design

We conducted a mechanistic, randomized, double-blind, placebo-controlled crossover study.20 All participants underwent two in-laboratory overnight polysomnograms (PSGs) at least 4 days apart, based on the pharmacologic half-lives of venlafaxine and O-desmethylvenlafaxine (the only major active metabolite; t1/2 = 5 ± 2 and 11 ± 2 h, respectively).21,22 Given the lack of prior data on venlafaxine in patients with OSA and consistent with other studies18,23 we chose a priori a single-night design to increase feasibility and to minimize risks to participants. This single-night design precluded gradual up-titration of venlafaxine. Thus, during the two study visits subjects received either 50 mg of venlafaxine (based on the usual adult starting dose range of 37.5-75 mg21) or an indistinguishable placebo 1 h before sleep. The allocation sequence was generated by one investigator (R. J.), using an online random sequence generator (random.org). All subjects, investigators interacting with participants, and outcome assessors were blinded to subjects’ allocation sequence until the analysis stage.

Power Calculation

We estimated that a sample size of 20 subjects will provide a power > 0.8 to detect a change in apnea-hypopnea index (AHI) of 10 (± 15) events/h with an α level of 0.05.

Participants

Patients were eligible for the study if they were 18 to 70 years of age. All subjects had a history of untreated OSA with an AHI greater than 5 events/h and were recruited from the University of California San Diego sleep clinic. Exclusion criteria included the presence of pulmonary, cardiac, neurologic, or other active severe medical or psychiatric diseases; or current use of CPAP therapy. No drugs that might interact with the investigational medication or known to affect sleep were taken during the trial or 1 month before the study. We also excluded patients with known allergy to venlafaxine, currently smoking, or taking alcohol (> 3 oz/d). Written informed consent was provided by each subject before participation in the study, and the study protocol was approved by the Human Research Committee, University of California San Diego (IRB#141272).

Polysomnograms

EEGs, electro-oculograms, and surface electromyograms were applied to score arousals, leg movements, and sleep stage. Abdominal and chest movements (magnetometers), pulse oxygen saturation, and oral and nasal flow were recorded to detect respiratory events. Participants were instructed to sleep supine as much as possible throughout the duration of the night, and position was recorded on the basis of visual inspection via an infrared camera system. A registered polysomnographic technologist, who was blinded to the allocation sequence, scored the PSGs according to the Chicago criteria (ie, hypopneas were defined as a > 50% decrease in airflow, or a ≤ 50% reduction in airflow associated with an oxygen desaturation of ≥ 3% or arousal).24

Endotype Measurements

OSA is a multifactorial disease caused by the interplay of an anatomical predisposition and several nonanatomical physiologic traits or endotypes.25, 26, 27 A detailed discussion of these endotypes and their various implications can be found in the literature,28 and e-Appendix 1 and e-Figures 1 and 2 in the online article provide details of how endotypes can be quantified. In brief, based on a validated custom algorithm analyzing routine PSG data (MATLAB; MathWorks)27,29 we quantified the following traits for each subject off CPAP during the venlafaxine vs placebo night as a percentage of eupneic ventilation/drive (Veupnea):

  • •

    Anatomical upper airway collapsibility (Vpassive: Ventilation when upper airway dilator muscles are hypotonic/passive; higher %Veupnea values correspond with a stiffer airway).

  • •

    Pharyngeal muscle recruitment (Vactive: Ventilation when upper airway dilator muscles are maximally activated; higher %Veupnea values indicate more muscle recruitment in response to accumulating respiratory stimuli during partial airway collapse).

  • •

    Ventilatory instability (loop gain in response to 1 disturbance/min [LG1]: this metric is dimensionless; larger values reflect greater instability, and values > 1 correspond with periodic breathing/central sleep apnea).

  • •

    Respiratory arousal threshold (ArTH: respiratory drive that causes arousal from sleep; when the airway narrows then ventilation drops while ventilatory drive rises; thus, larger %Veupnea values correspond with greater drops in ventilation that can be sustained before arousal from sleep occurs).

  • •

    Ventilatory response to arousal (amount of ventilatory overshoot during an arousal that is not explained by the respiratory drive at the time of arousal; greater values reflect a greater ventilatory overshoot, which can exacerbate ventilatory instability; see Discussion).26

Furthermore, we calculated the upper airway gain, which is a function of both pharyngeal muscle recruitment and the ArTH (dimensionless; larger values indicate a greater ability to dilate pharyngeal muscles before respiratory arousals).

In addition to quantifying the ArTH on the basis of analysis of the raw PSG data (“measured” ArTH), we also calculated the “predicted” ArTH on the basis of a validated clinical score incorporating age, sex, BMI, AHI, peripheral oxygen saturation (Spo2) nadir, and the percentage of hypopneas.30,31 The predicted ArTH provides an estimate of the negative epiglottic pressure swing preceding arousal in –cm H2O; for simplicity we used the absolute value (cm H2O) for all analyses.

Statistical Analysis

Statistical analyses were performed with R (version 3.6.1).32 In preparation for each analysis we assessed normality, using a Shapiro-Wilk test in combination with histogram inspection to determine the need for nonparametric testing. Thus, as appropriate, we compared outcome data on venlafaxine vs placebo using paired t tests or Wilcoxon signed ranks tests, explored correlations between baseline characteristics and changes in AHI using Pearson or Spearman correlation coefficients, performed exploratory subgroup analyses (comparing AHI changes stratified by OSA severity and low vs high ArTH), and compared responder (arbitrarily defined as AHI-reduction by > 10 events/h) vs nonresponder characteristics using two-sample t tests. Data were summarized accordingly as mean (SD) or median (interquartile range). Statistical significance was set at P < .05. Comparisons of AHI, Spo2 nadir, loop gain, ArTH, and sleep parameters were prespecified; all other analyses were performed post hoc and considered exploratory (ie, no adjustment for multiple comparisons).

Results

Six of the 26 subjects screened did not meet inclusion criteria; 20 patients were randomized and completed the study (Fig 133). Two subjects reported mild nausea shortly after venlafaxine administration, which resolved within 1 h. In addition, one subject complained of nausea in the morning after placebo administration. No serious or unanticipated problems related to venlafaxine were observed during the study.

Figure 1.

Figure 1

Open in a new tab

CONSORT flow diagram (for crossover trials). CONSORT = Consolidating Standards of Reporting Trials.

(Reprinted with permission from Dwan et al.33)

As shown in Table 1, subjects were middle-aged, overweight to obese, with Epworth Sleepiness Scale scores ranging clinically from no to severe sleepiness. Thirty percent of the subjects were women and 25% were nonwhite. Of note, there was no difference in BMI, percentage of time slept supine, or percentage of “first night” in the laboratory between the venlafaxine vs placebo nights (Pall > .2; e-Table 1).

Table 1.

Baseline Demographics

Demographic Mean/Median (SD/IQR) or Percent (No.) Range
Age, y 53.8 (8.1) 35-64
Sex, female 30 (6) …
Race
 White 75 (15) …
 Black 10 (2) …
 Asian 15 (3) …
Epworth Sleepiness Scale 8 (6-15) 4-22
Pittsburgh Sleep Quality Index 8.6 (3.3) 4-15
BMI, kg/m2 30.9 (3.9) 24.8-39.5
Neck circumference, cm 42.3 (7.5) 34.5-63
Time between studies, d 7 (7-15) 4-22
Open in a new tab

IQR = interquartile range.

Effect of Venlafaxine on Sleep

Venlafaxine vs placebo significantly decreased total sleep time, sleep efficiency, and percentage of rapid eye movement (REM) sleep but increased the percentage of non-REM (NREM) stage 1 sleep (Table 2). REM sleep was absent in seven subjects (35%) receiving venlafaxine and two subjects receiving placebo (10%).

Table 2.

Sleep Parameters During Placebo vs Venlafaxine Night: n = 20

Sleep Parameter Placebo
 Venlafaxine
 Δ Venlafaxine-Placebo
 P Value
[Mean (SD) or Median (IQR)] [Mean (SD) or Median (IQR)] [MDa (95% CI)]
TST, min 334.6 (79.9) 284.4 (81.5) –50.0 (–91.4 to –8.6) .02b
Sleep efficiency, % 72.5 (15.5) 63.1 (16.7) –9.4 (–17.9 to –1) .03b
Sleep onset latency, min 4.2 (1.5-11.2) 5.0 (3.0-18.1) 1.5 (–0.7 to 21.5) .09
Total arousal index, h–1 39.3 (17.8) 44.2 (16.5) 4.9 (–3.8 to 13.6) .3
NREM stage 1 sleep, % TST 23.9 (16.6-29.4) 39.0 (29.6-45.2) 12.1 (2.4 to 20.6) .02b
NREM stage 2 sleep, % TST 54.4 (13.3) 53.2 (11.8) –1.2 (–8.3 to 5.9) .7
NREM stage 3 sleep, % TST 7.5 (7.2) 5.2 (7.0) –2.3 (–5.6 to 0.9) .2
REM sleep, % TST 10.8 (7.8) 4.2 (5.9) –6.6 (–11.6 to –1.6) .01b
Open in a new tab

Δ = change; MD = mean or median of differences for normally or nonnormally distributed differences, respectively; NREM = nonrapid eye movement; REM = rapid eye movement; TST = total sleep time. See Table 1 legend for expansion of other abbreviation.

a

The median of differences is mathematically not necessarily the same as the difference of medians; here the median of differences is the more informative measure as it takes into account the paired nature of the data shown.

b

P < .05.

Effect of Venlafaxine on OSA Severity

Overall, there was no significant difference between venlafaxine and placebo regarding AHI (mean reduction, –5.6 events/h [95% CI, –12.0 to 0.9]; P = .09) or Spo2 nadir (median increase, +1.0% [–0.5 to 5]; P = .11) (e-Fig 3). However, individual response to venlafaxine appeared to be heterogeneous: the AHI decreased by > 10 events/h in one-third of subjects (n = 7; 35%), whereas in one subject (5%) the AHI increased by > 10 events/h (Fig 2). Thus, we sought to explore potential subgroups of patients who may respond to venlafaxine.

Figure 2.

Figure 2

Open in a new tab

Changes in apnea-hypopnea index (AHI, events/h) with venlafaxine vs placebo. Diamonds and bars show mean and SD for each condition (red, placebo; blue, venlafaxine). Circles denote individuals’ AHI.

On the basis of post hoc analyses, there was no change in REM-sleep AHI (+1.9 events/h, n = 11; P = .8), and thus any improvement in total AHI appeared to be due primarily to changes in NREM-sleep AHI (–5.0 events/h; P = .14) (Table 3). Allocation order did not modify the effect of venlafaxine on the AHI (P = .3) (e-Table 2).

Table 3.

Sleep Apnea Severity on Venlafaxine vs Placebo

Sleep Apnea Measure Placebo
 Venlafaxine
 Δ Venlafaxine-Placebo
 No.a P Value
[Mean (SD) or Median (IQR)] [Mean (SD) or Median (IQR)] [MD (95% CI)]
AHI, events/h
Total 46.1 (21.9) 40.5 (16.5) –5.6 (–12.0 to 0.9) … .09
By sleep stage
 NREM 45.1 (22.7) 40.1 (16.3) –5.0 (–11.8 to 1.8) … .14
 REM 47.1 (17.2) 48.9 (22.2) 1.9 (–15.9 to 19.6) 11 .8
By time of night
 NREM AHI, first half of night 44.4 (26.2) 35.6 (21.1) –8.8 (–18.7 to 1.1) … .08
 NREM AHI, second half of night 45.7 (21.5) 44.6 (14.7) –1.1 (–8.7 to 6.5) … .8
By event type
 Obstructive apnea index 17.7 (12.7) 15.3 (13.1) –2.5 (–6.9 to 2.0) … .3
 Central apnea index 1.1 (0.3 to 2.4) 0.3 (0 to 2.8) –0.6 (–3.3 to –0.1) … .02b
 Mixed apnea index 0.2 (0 to 3.7) 0.2 (0 to 1.3) 0 (–3.8 to 0.6) … 1
 Hypopnea index 21.2 (9.7) 20.6 (11.2) –0.5 (–4.5 to 3.4) … .8
O2 Saturation, %
Nadir 81.5 (75.8 to 84.2) 82.0 (80.0 to 84.0) 1.0 (–0.5 to 5) … .11
Mean 92.7 (1.6) 92.5 (1.4) –0.2 (–0.9 to 0.5) … .6
Open in a new tab

AHI = apnea-hypopnea index. See Table 1 and 2 legends for expansion of other abbreviations.

a

n = 20 unless stated otherwise.

b

P < .05.

Effect of Venlafaxine on Pathophysiologic Traits

As shown in Table 4, venlafaxine vs placebo decreased the ventilatory response to arousal by 30% (–6.9% Veupnea; P = .049) and lowered the predicted ArTH by 13% (–2.4 cm H2O; P = .02). There was no significant change in any of the other traits including measures of pharyngeal muscle response, loop gain, or the ArTH quantified from raw PSG signals.

Table 4.

Changes in Endotypes With Venlafaxine vs Placebo

Endotype Placebo
 Venlafaxine
 Δ Venlafaxine-Placebo
 n P Value
[Mean (SD) or Median (IQR)] [Mean (SD) or Median (IQR)] [MD (95% CI)]
Vpassive, %Veupnea 95.0 (86.1 to 97.3) 97.2 (91.6 to 98.1) 0.5 (–3.5 to 6.3) 19 .4
Vactive, %Veupnea 96.1 (82.4 to 102.5) 101.7 (92.0 to 103.8) –0.9 (–16.6 to 4.9) 19 .7
UAG, dimensionless 0.02 (–0.12 to 0.69) 0.43 (0.06 to 0.6) 0.25 (–0.14 to 0.58) 19 .3
Loop gain 1, dimensionless 0.56 (0.14) 0.58 (0.15) 0.01 (–0.05 to 0.07) 19 .7
Loop gain n, dimensionless 0.39 (0.12) 0.37 (0.09) –0.02 (–0.07 to 0.03) 19 .4
Arousal threshold (ArTH), %Veupnea 118.6 (113.4 to 137.9) 115.6 (108.2 to 130.8) –3.1 (–10.7 to 6.5) 19 .6
Predicted ArTH, cm H2O 18.4 (6.5) 16.0 (5.8) –2.4 (–4.3 to –0.5) 20 .02a
VRA, %Veupnea 23.2 (14.0 to 32.9) 14.2 (3.6 to 19.6) –6.9 (–14.6 to –0.1) 19 .049a
Open in a new tab

ArTH = respiratory arousal threshold; UAG = upper airway gain; Vactive = ventilation when upper airway dilator muscles are maximally activated; Veupnea = eupneic ventilation; Vpassive = ventilation when upper airway dilator muscles are hypotonic/passive; VRA = ventilatory response to arousal. See Table 1 and 2 legends for expansion of other abbreviations.

a

P < .05.

Both reductions in ventilatory response to arousal (rS = 0.47, P = .045) (e-Table 3) and reductions in ArTH correlated with improvements in AHI (measured ArTH: rS = 0.68, P = .002; predicted ArTH: rP = 0.75, P < .001) (e-Table 3).

Baseline Characteristics Predicting Response

Elevated AHI, higher respiratory ArTH (ie, harder to arouse), and higher Vpassive (less collapsible upper airway) at baseline correlated with greater improvements in AHI on venlafaxine vs placebo (|r| > 0.48, P < .05) (Table 5). Similarly, responders (AHI improvement > 10 events/h) vs nonresponders had a higher baseline AHI (P = .04) and tended to have a higher ArTH (P = .07) (e-Table 4).

Table 5.

Predictors of Response: Correlations Between Baseline Characteristics and Change in Apnea-Hypopnea Indexa

Baseline Characteristic (ie, Measured on Placebo) rP/rS (95% CI) Typeb n P Value
Age, y –0.23 (–0.6 to 0.2) P 20 .3
Epworth Sleepiness Scale score 0.07 (–0.4 to 0.5) S 20 .8
AHI, events/h –0.66 (–0.9 to –0.3) P 20 .002c
Vpassive, %Veupnea 0.52 (0.1 to 0.8) S 19 .02c
Vactive, %Veupnea 0.16 (–0.3 to 0.6) S 19 .5
UAG, dimensionless 0.02 (–0.4 to 0.5) P 19 1
Loop gain 1, dimensionless –0.28 (–0.7 to 0.2) P 19 .2
ArTH, %Veupnea –0.48 (–0.8 to –0.03) S 19 .04c
Predicted ArTH, cm H2O –0.53 (–0.8 to –0.1) P 20 .02c
VRA, %Veupnea –0.14 (–0.6 to 0.3) S 19 .6
Open in a new tab

See Table 1, 2, and 4 legends for expansion of abbreviations.

a

More negative AHIvenlafaxine-placebo values reflect greater improvement in AHI. Higher baseline AHI and higher baseline ArTH (ie, being hard to arouse) were significantly correlated with more pronounced improvements in AHI on venlafaxine.

b

Type denotes Pearson (P) vs Spearman (S) for normally vs nonnormally distributed variables, respectively.

c

P < .05.

Baseline AHI and predicted ArTH were highly collinear (rS = 0.78, P = 10–4), precluding a meaningful comparison of their relative predictive value via multivariable regression. On the basis of exploratory post hoc subgroup analyses (using standard cutoffs), venlafaxine improved the AHI in subjects with a high ArTH (≥ 15 cm H2O, ie, hard to arouse) but tended to worsen the AHI in those with a low ArTH (Fig 3). Similarly, venlafaxine vs placebo improved the AHI in those with severe OSA (AHI ≥ 30 events/h) but tended to worsen the AHI in patients with mild/moderate OSA. This finding may reflect the low ArTH in most patients with mild/moderate OSA; however, the number of subjects with mild/moderate OSA was small, precluding firm conclusions. Results were similar when dichotomizing baseline ArTH and AHI using the mean of treatment and placebo values (e-Fig 4), and thus these findings were not explained by regression to the mean.

Figure 3.

Figure 3

Open in a new tab

Change in AHI with venlafaxine vs placebo stratified by OSA severity and the predicted arousal threshold, using standard cutoffs. Diamonds and bars show mean and 95% CI for each group; red circles denote individuals’ change in AHI. See Figure 2 legend for expansion of abbreviation.

Discussion

We note three key findings: (1) Overall, a single 50-mg dose of venlafaxine did not significantly improve the AHI, but individual responses appeared variable; (2) baseline ArTH and AHI predicted treatment response; and (3) counter to our expectation, we failed to detect an improvement in muscle recruitment. Instead, exploratory analyses suggest that venlafaxine lowered the ventilatory response to arousal and the ArTH. On the basis of the current pathophysiologic model of OSA these changes are expected to improve and worsen OSA severity, respectively, potentially explaining the zero net effect on the AHI overall and the differential response based on subjects’ baseline ArTH.

After a mechanical collapse of the upper airway, accumulation of respiratory stimuli (ie, rise in intrapharyngeal negative pressure plus CO2) leads to a gradual activation of pharyngeal muscles; if sleep is maintained until airway patency is restored (high ArTH) then the result is stable breathing, but if the person arouses before airway opening (low ArTH) then repetitive cycles of airway collapse and arousals (ie, OSA) are expected.34,35 On the basis of indirect evidence from other studies, we expected that serotonergic/noradrenergic augmentation by venlafaxine would increase pharyngeal muscle activity14,23,36 while lowering the ArTH.18,23 As expected, venlafaxine lowered the ArTH predicted on the basis of a clinical score (and shifted sleep toward lighter stages); however, for unclear reasons we failed to detect a change in the ArTH estimated from PSG data using a custom algorithm, and did not find a change in pharyngeal muscle recruitment using the same methodology. Interestingly, exploratory analyses suggested that venlafaxine blunts the ventilatory response to arousal. During partial airway obstruction accumulation of respiratory stimuli gradually increases respiratory drive. Thus, when pharyngeal dilators restore airway patency, ventilation rises temporarily to levels in excess of baseline ventilation. This excess ventilation is greatly increased when the airway reopening is associated with a cortical arousal (267% vs 180% prehypopnea baseline; P < 10–10), which may exacerbate the subsequent drop in respiratory drive and pharyngeal muscle activation, thus potentially triggering repetitive cycles of airway collapse and arousals (ie, OSA).34 Our findings are intriguing for two reasons: First, whereas results from previous studies have been mixed,26,37,38 our findings provide further support for ventilatory response to arousal being an independent and important pathophysiologic mechanism underlying OSA in some patients. Second, aside from acetazolamide,26 venlafaxine is the only drug that has been shown to lower ventilatory response to arousal, which may have important implications for a personalized, multimodal therapy approach in the future.39

To our knowledge this is the first study assessing the effect of venlafaxine on OSA. However, several studies evaluated pharmacodynamically similar medications: two small randomized controlled trials assessed protriptyline, a tricyclic antidepressant with noradrenergic, serotonergic, and some anticholinergic effects. One of these studies found no overall effect on the AHI but noted substantial interindividual variability,40 and the other reported a small reduction in OSA severity due to a reduction in REM sleep.41 Neither of these studies assessed effects on the pathophysiologic traits, making comparisons difficult. Similarly, a study of clonidine (α2-agonist with sympatholytic effects) reported minor improvements in OSA severity primarily due to a reduction in REM sleep.42 In addition, in a study of healthy subjects,43 clonidine reduced (central) hypocapnic ventilatory response (a component of loop gain), thus lowering the propensity to central apneas (lower apnea threshold, higher CO2 reserve). In contrast, in our study venlafaxine did not affect loop gain, suggesting that these two drugs have different effects on control of breathing. Further, in a small crossover trial the potent noradrenaline-reuptake inhibitor atomoxetine alone lowered the ArTH with a zero net effect on the AHI, but coadministration of the anticholinergic oxybutynin mitigated its effect on the ArTH44 (and possibly augmented pharyngeal muscle tone in REM),45, 46, 47, 48 resulting in a large improvement in OSA severity by ?70%.18 Moreover, a crossover trial testing the effect of single-dose desipramine (a tricyclic antidepressant with potent noradrenergic, some serotonergic and mild anticholinergic properties) in 14 patients with OSA showed many similar findings as in our study, including a shift toward lighter sleep stages, significant lowering of the ArTH, and a nonsignificant reduction of overall AHI by 18%, with about one-third of patients being classified as responders (drop in AHI by >20 events/h).23 Unfortunately, neither of these trials measured ventilatory response to arousal. However, these data support the notion that drugs increasing noradrenergic tone have opposing effects on the underlying OSA traits, resulting in a small to zero net effect (and high interindividual variability) when used alone, but combination with agents increasing the ArTH (eg, eszopiclone35) may produce synergistic effects that may result in more robust responses and substantially improved OSA severity.

Strengths and Limitations

Strengths of our study include the use of a randomized, double-blind, placebo-controlled trial design, which makes selection or measurement biases unlikely, and detailed physiologic assessments providing mechanistic insights. Limitations include the lack of “gold standard” assessment of pathophysiologic traits, which may have reduced our ability to detect changes in muscle tone. Also, although the predicted ArTH has been validated as a measure of interindividual baseline ArTHs, it has not been validated as a way to estimate changes in intraindividual ArTHs. Further, given the risk of chance findings, the results from our post hoc analyses do not allow firm conclusions but may be valuable for hypothesis generation guiding future research: for example, any potential beneficial effects of venlafaxine on the NREM-AHI tended to occur only during the first half of the night, which may be due to sleep stage effects (more deep sleep with high ArTH during the first vs second half of the night) or pharmacokinetic reasons (ie, drop in concentration below therapeutic levels over time). Thus, future research should consider use of extended-release formulations, longer durations of administration, and/or higher doses. Such research will also help address another important limitation of our study: the pharmacokinetic effects of a single 50-mg dose of immediate-release venlafaxine on serotonergic/noradrenergic tone are difficult to predict and were not measured: chronic administration of low-dose venlafaxine (50-75 mg) is sufficient to achieve in vivo the same level of serotonin transporter occupancy (?80%) as with high-dose venlafaxine,49,50 but blocks only about 10% to 30% of norepinephrine transporters vs 30% to 60% with doses > 150 mg.51 This dose-dependent effect has also been corroborated by studies using functional assessments.52,53 Moreover, many effects of antidepressants are mediated by adaptive changes related to chronic (> 2-6 weeks) administration.54 Thus, our results may reflect more changes in serotonergic than noradrenergic tone, and the effects of venlafaxine on OSA may be quite different when administered in higher doses and/or long-term. Another limitation is that that some of the changes in AHI as well as changes in endotypes were likely mediated by shifts in sleep stages. Moreover, in this study reductions in ArTH correlated with improvements in AHI (e-Table 3); this is in contrast to the current pathophysiologic model of OSA, which predicts that lowering the ArTH causes worsening OSA (e-Appendix 1, e-Figs 1, 2). This association may be explained by the observation that only subjects with a high baseline ArTH (hard to wake) were able to “tolerate” the alerting effects of venlafaxine and benefit from its effect on ventilatory response to arousal (ie, had improvements in AHI despite reductions in ArTH); conversely, in subjects with a low baseline ArTH (easy to wake) the potential magnitude of ArTH reductions is relatively limited, but the effects on the ArTH likely outweighed any beneficial effects on the ventilatory response to arousal, resulting in the observed zero effect on AHI in this subgroup. To test this hypothesis future studies combining venlafaxine with drugs that increase the ArTH are warranted.

Interpretation

Overall, a single dose of venlafaxine had no significant effect on the AHI, but worsened sleep architecture. There may be subgroups of patients defined by physiologic traits in whom venlafaxine may have future clinical value, and combination with drugs improving sleep continuity may result in synergistic effects. Pharmacotherapy for OSA may be achievable, but patient-centered outcomes (eg, sleepiness) will need to be assessed in rigorous multicenter trials.

Acknowledgments

Author contributions: C. N. S. and Y. L. serve as the guarantors of the articled, taking responsibility for the integrity of the work as a whole, from inception to publication of the article. Y. L., R. L. O., A. M., study design; Y. L., C. N. S., R. J., J. E. O., P. DeY., data acquisition; C. N. S., Y. L., S. A. S., R. L. O., A. M., data analysis; all authors contributed to interpretation of data and manuscript drafting or revisions.

Financial/nonfinancial disclosures: The authors have reported to CHEST the following: S. A. S. reports income from Nox Medical (consulting), Merck (consulting), Inspire (consulting), Apnimed (grant support), and Prosomnus (grant support). B. A. E. reports income from Apnimed (grant support). A. M. received income from Merck for medical education related to drug discovery. ResMed provided a philanthropic donation to UC San Diego in support of a sleep center. None declared (C. N. S., Y. L., J. E. O., R. J., P. DeY., R. L. O.).

Role of sponsors: The sponsors had no role in the design of the study, the collection and analysis of the data, or in the preparation of the manuscript.

Other contributions: The authors thank Erik Smales, BS, RPSGT, Dillon Gilbertson, MS, Naa-Oye Bosompra, BA, and especially Angelica Moore, BSRC-RRT, for help with completing this study.

Additional information: The e-Appendix, e-Figures, and e-Tables can be found in the Supplemental Materials section of the online article.

Footnotes

Drs Schmickl and Li contributed equally to this manuscript.

FUNDING/SUPPORT: Dr Schmickl is supported by the NIH [T32 grant HL134632]. Dr Edwards is supported by a Heart Foundation of Australia Future Leader Fellowship [grant 101167].

Supplementary Data

e-Online Data
mmc1.pdf (682.3KB, pdf)

References

  • 1.Jordan A.S., McSharry D.G., Malhotra A. Adult obstructive sleep apnoea. Lancet. 2014;383(9918):736–747. doi: 10.1016/S0140-6736(13)60734-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Peppard P.E., Young T., Barnet J.H., Palta M., Hagen E.W., Hla K.M. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol. 2013;177(9):1006–1014. doi: 10.1093/aje/kws342. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Benjafield A.V., Ayas N.T., Eastwood P.R. Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis. Lancet Respir Med. 2019;7(8):687–698. doi: 10.1016/S2213-2600(19)30198-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Epstein L.J., Kristo D., Strollo P.J., Jr., Adult Obstructive Sleep Apnea Task Force of the American Academy of Sleep Medicine Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med. 2009;5(3):263–276. [PMC free article] [PubMed] [Google Scholar]
  • 5.Zinchuk A., Edwards B.A., Jeon S. Prevalence, associated clinical features, and impact on continuous positive airway pressure use of a low respiratory arousal threshold among male United States veterans with obstructive sleep apnea. J Clin Sleep Med. 2018;14(5):809–817. doi: 10.5664/jcsm.7112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Alves C., Caminha J.M., da Silva A.M., Mendonça D. Compliance to continuous positive airway pressure therapy in a group of Portuguese patients with obstructive sleep apnea syndrome. Sleep Breath. 2012;16(2):555–562. doi: 10.1007/s11325-011-0542-9. [DOI] [PubMed] [Google Scholar]
  • 7.Campos-Rodriguez F., Martinez-Alonso M., Sanchez-de-la-Torre M., Barbe F., Spanish Sleep Network Long-term adherence to continuous positive airway pressure therapy in non-sleepy sleep apnea patients. Sleep Med. 2016;17:1–6. doi: 10.1016/j.sleep.2015.07.038. [DOI] [PubMed] [Google Scholar]
  • 8.Cistulli P.A., Armitstead J., Pepin J.L. Short-term CPAP adherence in obstructive sleep apnea: a big data analysis using real world data. Sleep Med. 2019;59:114–116. doi: 10.1016/j.sleep.2019.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mason M., Welsh E.J., Smith I. Drug therapy for obstructive sleep apnoea in adults. Cochrane Database Syst Rev. 2013;(5):CD003002. doi: 10.1002/14651858.CD003002.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Mezzanotte W.S., Tangel D.J., White D.P. Waking genioglossal electromyogram in sleep apnea patients versus normal controls (a neuromuscular compensatory mechanism) J Clin Invest. 1992;89(5):1571–1579. doi: 10.1172/JCI115751. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Mezzanotte W.S., Tangel D.J., White D.P. Influence of sleep onset on upper-airway muscle activity in apnea patients versus normal controls. Am J Respir Crit Care Med. 1996;153(6 Pt 1):1880–1887. doi: 10.1164/ajrccm.153.6.8665050. [DOI] [PubMed] [Google Scholar]
  • 12.Kubin L., Davies R.O., Pack A.I. Control of upper airway motoneurons during REM sleep. News Physiol Sci. 1998;13:91–97. doi: 10.1152/physiologyonline.1998.13.2.91. [DOI] [PubMed] [Google Scholar]
  • 13.Sood S., Liu X., Liu H., Nolan P., Horner R.L. 5-HT at hypoglossal motor nucleus and respiratory control of genioglossus muscle in anesthetized rats. Respir Physiol Neurobiol. 2003;138(2-3):205–221. doi: 10.1016/j.resp.2003.07.001. [DOI] [PubMed] [Google Scholar]
  • 14.Berry R.B., Yamaura E.M., Gill K., Reist C. Acute effects of paroxetine on genioglossus activity in obstructive sleep apnea. Sleep. 1999;22(8):1087–1092. doi: 10.1093/sleep/22.8.1087. [DOI] [PubMed] [Google Scholar]
  • 15.Kraiczi H., Hedner J., Dahlöf P., Ejnell H., Carlson J. Effect of serotonin uptake inhibition on breathing during sleep and daytime symptoms in obstructive sleep apnea. Sleep. 1999;22(1):61–67. [PubMed] [Google Scholar]
  • 16.Hanzel D.A., Proia N.G., Hudgel D.W. Response of obstructive sleep apnea to fluoxetine and protriptyline. Chest. 1991;100(2):416–421. doi: 10.1378/chest.100.2.416. [DOI] [PubMed] [Google Scholar]
  • 17.Fenik V.B., Davies R.O., Kubin L. REM sleep-like atonia of hypoglossal (XII) motoneurons is caused by loss of noradrenergic and serotonergic inputs. Am J Respir Crit Care Med. 2005;172(10):1322–1330. doi: 10.1164/rccm.200412-1750OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Taranto-Montemurro L., Messineo L., Sands S.A. The combination of atomoxetine and oxybutynin greatly reduces obstructive sleep apnea severity. a randomized, placebo-controlled, double-blind crossover trial. Am J Respir Crit Care Med. 2019;199(10):1267–1276. doi: 10.1164/rccm.201808-1493OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Rukhadze I., Carballo N.J., Bandaru S.S., Malhotra A., Fuller P.M., Fenik V.B. Catecholaminergic A1/C1 neurons contribute to the maintenance of upper airway muscle tone but may not participate in NREM sleep-related depression of these muscles. Respir Physiol Neurobiol. 2017;244:41–50. doi: 10.1016/j.resp.2017.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.National Institutes of Health Clinical Center The impact of venlafaxine on apnea hypopnea index in obstructive sleep apnea. NCT02714400. ClinicalTrials.gov. Bethesda, MD: National Institutes of Health. 2016. https://clinicaltrials.gov/ct2/show/NCT02714400 Last updated April 17, 2020.
  • 21.www.uptodate.com. Venlafaxine: drug information. https://www.uptodate.com/contents/table-of-contents/drug-information/general-drug-information [content available only to UpToDate subscribers]. Accessed April 30, 2020.
  • 22.www.uptodate.com. Desvenlafaxine: drug information. https://www.uptodate.com/contents/table-of-contents/drug-information/general-drug-information [content available only to UpToDate subscribers]. Accessed April 30, 2020.
  • 23.Taranto-Montemurro L., Sands S.A., Edwards B.A. Desipramine improves upper airway collapsibility and reduces OSA severity in patients with minimal muscle compensation. Eur Respir J. 2016;48(5):1340–1350. doi: 10.1183/13993003.00823-2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.American Academy of Sleep Medicine Task Force Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research: the report of an American Academy of Sleep Medicine Task Force. Sleep. 1999;22(5):667–689. [PubMed] [Google Scholar]
  • 25.Wellman A., Edwards B.A., Sands S.A. A simplified method for determining phenotypic traits in patients with obstructive sleep apnea. J Appl Physiol (1985) 2013;114(7):911–922. doi: 10.1152/japplphysiol.00747.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Edwards B.A., Connolly J.G., Campana L.M. Acetazolamide attenuates the ventilatory response to arousal in patients with obstructive sleep apnea. Sleep. 2013;36(2):281–285. doi: 10.5665/sleep.2390. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Terrill P.I., Edwards B.A., Nemati S. Quantifying the ventilatory control contribution to sleep apnoea using polysomnography. Eur Respir J. 2015;45(2):408–418. doi: 10.1183/09031936.00062914. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Schmickl C.N., Owens R., Edwards B.A., Malhotra A. OSA endotypes: what are they and what are their potential clinical implications? Curr Sleep Med Rep. 2018;4:231–242. [Google Scholar]
  • 29.Orr J.E., Sands S.A., Edwards B.A. Measuring loop gain via home sleep testing in patients with obstructive sleep apnea. Am J Respir Crit Care Med. 2018;197(10):1353–1355. doi: 10.1164/rccm.201707-1357LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Edwards B.A., Eckert D.J., McSharry D.G. Clinical predictors of the respiratory arousal threshold in patients with obstructive sleep apnea. Am J Respir Crit Care Med. 2014;190(11):1293–1300. doi: 10.1164/rccm.201404-0718OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lee R.W.W., Sutherland K., Sands S.A. Differences in respiratory arousal threshold in Caucasian and Chinese patients with obstructive sleep apnoea. Respirology. 2017;22(5):1015–1021. doi: 10.1111/resp.13022. [DOI] [PubMed] [Google Scholar]
  • 32.R Core Team . R Foundation for Statistical Computing; Vienna, Austria: 2019. R: A Language and Environment for Statistical Computing.https://www.R-project.org/ [Google Scholar]
  • 33.Dwan K., Li T., Altman D.G., Elbourne D. CONSORT 2010 statement: extension to randomised crossover trials. BMJ. 2019;366:I4378. doi: 10.1136/bmj.l4378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Younes M. Role of arousals in the pathogenesis of obstructive sleep apnea. Am J Respir Crit Care Med. 2004;169(5):623–633. doi: 10.1164/rccm.200307-1023OC. [DOI] [PubMed] [Google Scholar]
  • 35.Eckert D.J., Owens R.L., Kehlmann G.B. Eszopiclone increases the respiratory arousal threshold and lowers the apnoea/hypopnoea index in obstructive sleep apnoea patients with a low arousal threshold. Respirology. 2011;120(12):505–514. doi: 10.1042/CS20100588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Taranto-Montemurro L., Edwards B.A., Sands S.A. Desipramine increases genioglossus activity and reduces upper airway collapsibility during non-REM sleep in healthy subjects. Am J Respir Crit Care Med. 2016;194(7):878–885. doi: 10.1164/rccm.201511-2172OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Jordan A.S., Cori J.M., Dawson A. Arousal from sleep does not lead to reduced dilator muscle activity or elevated upper airway resistance on return to sleep in healthy individuals. Sleep. 2015;38(1):53–59. doi: 10.5665/sleep.4324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Amatoury J., Azarbarzin A., Younes M., Jordan A.S., Wellman A., Eckert D.J. Arousal intensity is a distinct pathophysiological trait in obstructive sleep apnea. Sleep. 2016;39(12):2091–2100. doi: 10.5665/sleep.6304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Light M., Owens R.L., Schmickl C.N., Malhotra A. Precision medicine for obstructive sleep apnea. Sleep Med Clin. 2019;14(3):391–398. doi: 10.1016/j.jsmc.2019.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Whyte K.F., Gould G.A., Airlie M.A., Shapiro C.M., Douglas N.J. Role of protriptyline and acetazolamide in the sleep apnea/hypopnea syndrome. Sleep. 1988;11(5):463–472. doi: 10.1093/sleep/11.5.463. [DOI] [PubMed] [Google Scholar]
  • 41.Brownell L.G., West P., Sweatman P., Acres J.C., Kryger M.H. Protriptyline in obstructive sleep apnea: a double-blind trial. N Engl J Med. 1982;307(17):1037–1042. doi: 10.1056/NEJM198210213071701. [DOI] [PubMed] [Google Scholar]
  • 42.Issa F.G. Effect of clonidine in obstructive sleep apnea. Am Rev Respir Dis. 1992;145(2 Pt 1):435–439. doi: 10.1164/ajrccm/145.2_Pt_1.435. [DOI] [PubMed] [Google Scholar]
  • 43.Sankri-Tarbichi A.G., Grullon K., Badr M.S. Effects of clonidine on breathing during sleep and susceptibility to central apnoea. Respir Physiol Neurobiol. 2013;185(2):356–361. doi: 10.1016/j.resp.2012.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.White R.P., Daigneault E.A. The antagonisms of atropine to the EEG effects of adrenergic drugs. J Pharmacol Exp Ther. 1959;125(4):339–346. [PubMed] [Google Scholar]
  • 45.Liu X., Sood S., Liu H., Horner R.L. Opposing muscarinic and nicotinic modulation of hypoglossal motor output to genioglossus muscle in rats in vivo. J Physiol. 2005;565(Pt 3):965–980. doi: 10.1113/jphysiol.2005.084657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Grace K.P., Hughes S.W., Horner R.L. Identification of the mechanism mediating genioglossus muscle suppression in REM sleep. Am J Respir Crit Care Med. 2013;187(3):311–319. doi: 10.1164/rccm.201209-1654OC. [DOI] [PubMed] [Google Scholar]
  • 47.Naji M., Komarov M., Krishnan G.P. Computational model of brain-stem circuit for state-dependent control of hypoglossal motoneurons. J Neurophysiol. 2018;120(1):296–305. doi: 10.1152/jn.00728.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Ohtake A., Saitoh C., Yuyama H. Pharmacological characterization of a new antimuscarinic agent, solifenacin succinate, in comparison with other antimuscarinic agents. Biol Pharm Bull. 2007;30(1):54–58. doi: 10.1248/bpb.30.54. [DOI] [PubMed] [Google Scholar]
  • 49.Lundberg J., Tiger M., Landén M., Halldin C., Farde L. Serotonin transporter occupancy with TCAs and SSRIs: a PET study in patients with major depressive disorder. Int J Neuropsychopharmacol. 2012;15(8):1167–1172. doi: 10.1017/S1461145711001945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Meyer J.H., Wilson A.A., Sagrati S. Serotonin transporter occupancy of five selective serotonin reuptake inhibitors at different doses: an [11C]DASB positron emission tomography study. Am J Psychiatry. 2004;161(5):826–835. doi: 10.1176/appi.ajp.161.5.826. [DOI] [PubMed] [Google Scholar]
  • 51.Arakawa R., Stenkrona P., Takano A. Venlafaxine ER blocks the norepinephrine transporter in the brain of patients with major depressive disorder: a PET study using [18F]FMeNER-D2. Int J Neuropsychopharmacol. 2019;22(4):278–285. doi: 10.1093/ijnp/pyz003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Harvey A.T., Rudolph R.L., Preskorn S.H. Evidence of the dual mechanisms of action of venlafaxine. Arch Gen Psychiatry. 2000;57(5):503–509. doi: 10.1001/archpsyc.57.5.503. [DOI] [PubMed] [Google Scholar]
  • 53.Debonnel G., Saint-André E., Hébert C., de Montigny C., Lavoie N., Blier P. Differential physiological effects of a low dose and high doses of venlafaxine in major depression. Int J Neuropsychopharmacol. 2007;10(1):51–61. doi: 10.1017/S1461145705006413. [DOI] [PubMed] [Google Scholar]
  • 54.Frazer A. Pharmacology of antidepressants. J Clin Psychopharmacol. 1997;17(suppl 1):2s–18s. doi: 10.1097/00004714-199704001-00002. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

e-Online Data
mmc1.pdf (682.3KB, pdf)

Articles from Chest are provided here courtesy of American College of Chest Physicians

RESOURCES