What is the most powerful respiratory stimulant in a healthy person?

Medications that increase ventilation

Acetazolamide: Carbonic anhydrases (CA) catalyze the interconversion between carbon dioxide and bicarbonate. In its active form, it has a hydroxide radical bound to its zinc ion. When it attacks a carbon dioxide molecule, it forms bicarbonate, and exchanges it with a water molecule. This enzyme plays an important role in alveolar capillary transport of carbon dioxide, acid-base equilibrium, and control of ventilation.47

The development of the first CA inhibitors and their use as respiratory stimulants arose in the late 1930 when it was noticed that sulfonamide antibiotics (weak inhibitors of CA) caused a mild metabolic acidosis and compensatory hyperventilation. Only few years earlier, CA had been discovered in red blood cells and kidneys, and it was established that inhibiting it resulted in metabolic acidosis. These observations led to the development of CA inhibitors such as acetazolamide and other less known ones such as methazolamide, diclorphenamide, and benzolamide.48

After oral intake, acetazolamide is absorbed quickly and reaches peak concentration in 2 h. Its half-life is 4–8 h. It is eliminated by the kidneys and liver. Its renal clearance is affected by changes in plasma protein binding. This is particularly important to understand in the elderly. They have a reduced capacity to clear unbound acetazolamide from the plasma because of age-related changes in renal function and have reduced plasma protein binding of acetazolamide. Therefore, the resulting excretory rates is like that of young subjects. The reduced protein binding, however, predisposes the elderly to the accumulation of acetazolamide in red blood cells, which increases the risk of experiencing concentration dependent side effects.49

At a low dose of 250–500 mg, acetazolamide reaches high concentrations in the kidneys that inhibits all their CA with little or no effect in other organs.48 When acetazolamide binds to CA, it decreases the tubular reabsorption of bicarbonate and inhibits the distal secretion of hydrogen. The resulting alkaline diuresis is maximal at 24 h.47 In healthy individuals, the resultant metabolic acidosis increases ventilation by 10%–20% leading to a 5–6 mmHg decrease in PaCO2.48

High doses of acetazolamide can induce a near-complete inhibition of CA in red blood cells and other tissues. This can lead to profound retention of carbon dioxide in all tissues. In the brain, this increase in carbon dioxide near the central chemoreceptors stimulates ventilation.48 When only the CA in the central chemoreceptors is inhibited, the rate of hydrogen ion production decreases resulting in a 50% decrease in the rate of increase in compensatory ventilation to hypercapnia. This however does not occur in whole brains because the large increase in carbon dioxide induces ventilation. Inhibition of tissue CA also creates a unique situation where the blood acid-base status does not reflect tissue acidification.48

Furosemide, one of the most commonly used diuretics, reduces the effect of acetazolamide by many potential mechanisms. It increases distal tubular excretion of hydrogen, directly counteracting the effect of acetazolamide. It also competes with acetazolamide for the same transport mechanisms from the plasma to its site of action in the renal tubules and for the same binding sight on CA.47 Hemming and colleagues developed a pharmacodynamic model of bicarbonate response to acetazolamide in mechanically ventilated patients with chronic obstructive pulmonary disease and found that higher doses of acetazolamide might be needed in patients receiving furosemide or corticosteroids and in patients with elevated chloride levels.50

Acetazolamide is generally safe. The main side effects are blood dyscrasias, dysthyroidism, and gout attacks. It can aggravate hepatic encephalopathy and can potentially increase carbon dioxide levels from suppressing CA of red blood cells. The main side effect to watch for in acutely ill patients is hypokalemia.47

Whyte and colleague studied the effect of acetazolamide in 10 patients with OSA whose BMI was 48 kg/m2.51 Their AHI was 50 ± 26 events/h and their lowest oxygen saturation during sleep was 70% ± 24%. These characteristics suggest that some of them were hypercapnic, but the authors did not report the arterial blood gases despite measuring them. Acetazolamide was administered at 250 mg twice daily for 1 week and then increased to 250 mg four times daily for another week. It reduced the AHI to 26 ± 20 events/h but did not increase the lowest oxygen saturation, change the duration of apneas, or improve sleep quality and symptoms.

Eskandari and colleagues compared the effects of acetazolamide, acetazolamide with CPAP, and CPAP alone on OSA and blood pressure.52 The patients were not morbidly obese (mean BMI <  30 kg/m2) and had moderate-to-severe OSA (mean AHI <  40 events/h). They found that given alone, acetazolamide reduced the AHI by 42%. Sleepiness and sleep quality did not improve with acetazolamide, CPAP, or their combination.52

Sharp and colleagues studied the effect of acidosis induced by acetazolamide on sleep apnea.53 They found that it had no effect in patients with pure obstructive or central apnea. In two patients categorized as having mixed sleep apnea, the acidosis made more apneas obstructive, and the apneas became longer and caused more hypoxemia.53 These findings were contradicted by Tojima and colleagues who studied the effect of 250 mg taken once daily for a week on respiratory events and the control of ventilation in eucapnic patients with moderate sleep apnea.54 They found that acetazolamide reduced sleepiness and the frequency of respiratory events and desaturations but did not change the types of respiratory events. It also increased the ventilatory response to hypercapnia from 1.22 ± 0.17 to 1.89 ± 0.33 L/min/mmHg. It had no effect on the ventilatory response to hypoxemia.

More recently, Edwards and colleagues identified mechanisms by which acetazolamide decreased the AHI. They assessed the four traits associated with OSA before and after treatment with acetazolamide (500 mg twice daily for 1 week) in 13 patients with moderate-to-severe sleep apnea (BMI: 34 kg/m2, NREM AHI: 50 events/h, REM AHI: 43 events/h, lowest oxygen saturation during sleep: 81%). They demonstrated that acetazolamide reduced loop gain and had no effect on upper airway collapsibility, upper airway gain, and arousal threshold. The AHI decreased by 47%, more in NREM than in REM. The percentage decrease in NREM AHI was modestly correlated with the percentage decrease in loop gain (r = 0.66, P = 0.05).55 In a later study, the same group of investigators identified another mechanism by which acetazolamide decreased the AHI.56 They demonstrated that before treatment with acetazolamide the increase in ventilation in response to an arousal correlated directly with AHI (r2 = 0.44, P = 0.01). Acetazolamide did not reduce the frequency of arousals but reduced the increase in ventilation after an arousal. The percentage decrease in the increase in ventilation after an arousal correlated modestly with the percentage decrease in the AHI (r2 = 0.66, P = 0.01) among patients whose AHI decreased with acetazolamide.56

Faisy and colleagues compared the effect of acetazolamide to placebo in patients receiving invasive mechanical ventilation for COPD.57 Acetazolamide improved oxygenation more than placebo but had no effect on ventilation and PaCO2 and did not reduce the duration of mechanical ventilation. Patients were enrolled in this study independent of their acid-base status.57 Most were not alkalotic (pH 7.32 ± 0.11), the PaCO2 was not severely high at enrollment (52 ± 16 mmHg), and the serum bicarbonate was normal in most patients (26 ± 7 mmoL/L).

Metabolic alkalosis frequently complicates the treatment of patients with acute on chronic hypercapnic respiratory failure. It occurs when PaCO2 decreases faster that the kidneys excrete the excess bicarbonates. It can also result from the administration of diuretics and corticosteroids. This alkalosis can be corrected with the administration of acetazolamide for no more than 2 days.58

This indication was evaluated by Fontana and colleagues in patients that developed alkalosis after treating acute hypercapnic failure from acute exacerbation of COPD with NIV.59 The patients were treated with acetazolamide (500 mg daily for 2 days) while receiving NIV. They were compared to a matched group of historical controls that were treated with NIV only. At the time of admission, the pH was 7.32 ± 0.05 and the PaCO2 was 74 ± 11 mmHg. After NIV alone, pH increased to 7.46 ± 0.06 and the PaCO2 decrease to 64 ± 10 mmHg, and the bicarbonate level was 44 ± 6 mmoL/L. Adding acetazolamide reduced the pH to 7.41 ± 0.06, the PaCO2 to 55 ± 8 mmHg, and the bicarbonate level to 36 ± 5 mmoL/L. These findings, however, were not confirmed in a randomized trial by Gulsvik and colleagues in patients hospitalized with respiratory failure (PaO2 <  60 mmHg or PaCO2 >  52 mmHg) and alkalosis (base excess ≥  8 mmoL/L). Most of the patients had COPD, none had OHS, and only 18% were treated with NIV. Acetazolamide was administered at 250 mg three times daily for 5 days. The pH was 7.42 ± 0.05, and PaCO2 was 59 mmHg in the acetazolamide arm and 57 mmHg in the placebo arm. Although acetazolamide increased the PaO2 (12 mmHg) more than the placebo (6 mmHg) and decreased the pH, it did not reduce the PaCO2 more than the placebo.60

In contrast, the study by Rialp Cervera and colleagues supported the use of acetazolamide in patients with hypercapnic respiratory failure and alkalosis.61 They enrolled patients mechanically ventilated for less than 3 days that had a pH >  7.35 and a bicarbonate >  28 mmoL/L. The patients enrolled to the acetazolamide arm had a pH of 7.43 ± 0.06, a PaCO2 of 52 ± 7 mmHg, and a bicarbonate level of 34 ± 4 mmoL/L. Only 24% of the patients had OHS. In comparison to placebo, acetazolamide resulted in a lower PaCO2 (48 versus 55 mmHg) and bicarbonate levels (29 versus 34 mmoL/L). Acetazolamide, as in the Faisy study, did not reduce the duration of mechanical ventilation.

Raurich and colleagues studied the effect of acetazolamide on the hypercapnic response of intubated patients with OHS.62 They enrolled 25 patients with OHS, eight of them also had mild-to-moderate COPD. The patients were tested once they were ready to commence weaning trials. The response to hypercapnia was assessed by two indexes: the hypercapnic drive response (ratio of the change in the occlusion pressure to the change in PaCO2 (normal: 0.6 ± 0.5 cmH2O/mmHg)) and the hypercapnic ventilatory response (ratio of the change in ventilation to change in PaCO2, normal 2.6 ± 1.2 L/min/mmHg). Both indexes were reduced (0.21 ± 26 cm H2O/mmHg and 0.39 ± 0.28 L/min/mmHg) and similar between patients with COPD and patients without COPD. They correlated with the bicarbonate level and not with the BMI.62 These indexes were assessed again after treatment with acetazolamide in patients with baseline bicarbonate >  34 mmoL/L.62 These patients had higher PaCO2 (almost 70 mmHg), higher bicarbonate level 39 ± 4 mmoL/L, and a pH of 7.38 ± 06. After treatment with acetazolamide, the pH did not change, but the bicarbonate level decreased to 31 ± 3 mmoL/L and the PaCO2 decreased to around 55 mmHg. The hypercapnic drive response increased from 0.10 ± 0.05 to 0.23 ± 0.14 cmH2O/mmHg. The hypercapnic ventilatory response increased from 0.21 ± 0.17 to 0.32 ± 0.19 L/min/mmHg.62

The aforementioned studies do not offer direct guidance on the use of acetazolamide in patients with acute hypercapnic respiratory failure from OHS. They do however suggest that clinicians can consider administering a short course of acetazolamide to mechanically ventilated patients with hypercapnic respiratory failure when they become alkalotic from ventilation and medications (pH >  7.35 and bicarbonate >  34 mmoL/L) and especially if they have OHS.

This also means that clinicians must determine the etiology of hypercapnia and closely monitor the changes in blood gases in response to treatment. Administering acetazolamide when ventilation is ineffective (COPD) or maximal (controlled mechanical ventilation) might be harmful. Carter and colleagues studied the effect of acetazolamide on acid-base status in rats breathing 10% CO2.63 Breathing CO2 increased the PaCO2 to 102 mmHg (34 mmHg breathing room air) and a decrease in the pH to 7.28 (on room air 7.43). When acetazolamide was added, the pH decreased to 6.98 and the PaCO2 paradoxically increased to 175 mmHg because, the authors speculated, of the effect of acetazolamide on the carbonic anhydrase of red blood cells.63

Rapoport and colleagues studied the effect of adjunct treatments in three patients with OHS who remained hypercapnic after tracheostomy. All patients reduced their PaCO2 with voluntary hyperventilation. Eucapnia was induced in two patients after 2 days of continuous mechanical ventilation and in one patient with acetazolamide for 1 week. After discontinuing these treatment, PaCO2 levels returned to their original levels.64

Progesterone: The effects of sex hormones on respiration are complex and vary with age.65 This section will focus on progesterone because it is the only one that was tested in patients with OHS. It has been known for more than a 100 years now that women hyperventilate during pregnancy and during the luteal phase of the menstrual cycle and that these cyclical changes in ventilation cease with menopause. The correlation between changes in ventilation and the concentration of progesterone suggested that it played a role in the stimulation of ventilation and led to testing its effects in men in the 1940s and 1950s.66 Later, it became also apparent that upper airway resistance was lower during the luteal phase than during the follicular phase and that it increased with menopause and decreased with the replacement of sex hormones.67, 68

Bayliss and Millhorn summarized the neural mechanisms of the effect of progesterone on respiration.66 These effects arise from its action on the central nervous system. Progesterone binds to its receptors, the production of which is increased by estrogen. The area of the brain associated with the effects of progesterone on respiration was the diencephalone and is limited to the hypothalamus and the preoptic area, particularly in the infundibular and rostroventral periventricular nuclei. The experiments by Bayliss and Millhorn suggested that the line of connection between the hypothalamus and the respiratory centers is probably indirect. Other mechanisms of progesterone include altering the release of serotonin in the respiratory nuclei in the brain stem and binding directly to the GABAA receptor and modulating its function.65

Skatrud and colleagues demonstrated that medroxyprogesterone acetate (MPA) increased ventilation in normal subjects and caused alkalosis in the blood and the cerebrospinal fluid. This effect was apparent after 2 days of administration and was maximal at 7 days. It was associated with a 5-mmHg decrease in PaCO2. It did not however change the ventilatory response to carbon dioxide but increased it in response to exercise.69 These findings were supported by Zwillich and colleagues.70 They found that MPA increase minute ventilation by 0.5 ± 0.2 L/min. But in contrast to the finding by Skatrud and colleagues, they found the MPA increased the ventilatory response to hypercapnia from 2.9 ± 0.3 to 4.0 ± 0.3 L/min/mmHg.70

Skatrud and colleagues later studied the effect of MPA in 17 patients with chronic hypercapnic respiratory failure: 14 with COPD, 2 were morbidly obese, and 1 had hypercapnia from unknown reasons. MPA was administered orally, 20 mg three times every day for 4 weeks.71 They defined correctors as those who lowered their PaCO2 by 5 mmHg. Ten patients met the corrector definition. Three of the correctors were obese (>  25% ideal body weight), and two noncorrectors were massively obese (150 and 200 kg). All correctors were able to increase their ventilation and decrease their PaCO2 by 5 mmHg during voluntary hyperventilation. Three of the noncorrectors were able to decrease their PaCO2 by 5 mmHg, two of them were obese.71

The correctors lowered their PaCO2 by 8 ± 1 mmHg. After stopping MPA, PaCO2 returned to within 2 mmHg of their baseline PaCO2 in 8 of the 10 correctors. Correctors increased their minutes ventilation to PaCO2 production ratio by 15%. This increase was mostly due to an 11% increase in tidal volume that was achieved by an increase in effort. On average, their respiratory rate and inspiratory time did not change.71 On the other hand, minute ventilation increased in noncorrectors but mostly from an increase in respiratory rate from 17 to 19 breaths/min. Despite an increase in the inspiratory effort in noncorrectors, the increase in tidal volume was minimal.71

The effect of MPA on sleep apnea was tested in patients with BMI of 40 kg/m2 and an apnea index of 48 events/h. The apnea rate did not change. But there was a small improvement in gas exchange during wake and sleep. The sample included seven patients only, and gas exchange data before and after MPA was available on four of them only.72 In contrast, Strohl and colleagues found that MPA significantly reduced the apnea index and improved the symptoms in a study of nine patients with an apnea index of 69 events/h and a lowest oxygen saturation during sleep of 73%, four of whom were hypercapnic. MPA also improved gas exchange.73

The earliest detailed report on the effects of treatment with progesterone on patients with OHS was by Lyons and colleagues and was published in 1968.74 They extensively examined the ventilation of eight patients with a mean weight of 148 kg and a pH of 7.35, PaCO2 of 62 mmHg, and PaO2 of 54 mmHg. After treatment with diuretics, digitalis, and diet, their weight decreased to 137 kg and the blood gases improved to a pH of 7.39, a PaCO2 of 51 mmHg, and a PaO2 of 64 mmHg. They then administered intramuscularly 100 mg of progesterone daily for 18–40 days. With no further decrease in weight, blood gases improved to a pH of 7.39, a PaCO2 of 38 mmHg, and a PaO2 of 78 mmHg. These changes in blood gases were associated with improvement in minute ventilation and ventilatory response. Ventilation increased from 7.7 L/min at baseline to 9.7 L/min after medical therapy and then to 11.7 L/min after treatment with progesterone. The ventilatory response to hypercapnia increased from 0.99 L/min/mmHg (range: 0.59–1.52 L/min/mmHg) after medical therapy to 1.7 L/min/mmHg (range: 1.18–2.05 L/min/mmHg) after treatment with progesterone. After discontinuing progesterone in five patients, PaCO2 increased in two and ventilatory response decreased in all.

These results were confirmed by Sutton and Colleagues.75 They prescribed sublingual MPA to 10 patients with OHS for 1 month and reduced the PaCO2 from 51 ± 2 to 38 ± 1 mmHg and increased the PaO2 from 49 ± 3 to 62 ± 2 mmHg with no change in weight. PaCO2 and PaO2 deteriorated to pretreatment level after a month of withdrawing medroxyprogesterone acetate and improved once again after reinstating it.

Rapoport and colleagues studied the effect of tracheostomy or CPAP on ventilation in patients with OHS. All of them were able to reduce their PaCO2 with voluntary hyperventilation. Seven were treated with tracheostomy and one with CPAP. Four patients became eucapnic and four remained hypercapnic. Three of the four patients with residual hypercapnia were treated with MPA: 20–40 mg three times daily for 1–2 months. It led to no change in PaCO2, minute ventilation, or response to CO2 rebreathing.64 It is important to note that the PaCO2 levels were not very high and the patients that had residual hypercapnia had higher pH at the start of the study (calculated from provided data).

OSA was not effectively treated in the two studies that showed an effect of progesterone in patients with OHS because they were published in 1968 and 1975 when CPAP was not a standard therapy. In contrast, OSA was effectively treated with tracheostomy or CPAP in the study that showed no improvement in ventilation with MPA. Does this imply that progesterone increases ventilation in patients with untreated OSA but fails to do so in patients with remain hypercapnic after treating OSA?

Leptin: Leptin is produced by adipose tissue. It crosses the blood brain barrier to the hypothalamus and medulla, where it suppresses appetite and increases energy expenditure. This feedback loop maintains weight. Leptin also influences various neural mechanisms that maintain normal ventilation. However, obese patients have elevated levels of leptin, signifying that they are resistant to its effects. Because of leptin’s effects on ventilation, resistance to leptin has been proposed as a mechanism for OHS.

Leptin resistance is thought to result from reduced transfer through the blood brain barrier. The blood brain barrier can be bypassed by administering leptin nasally. Berger and colleagues compared the effects of intranasal leptin to intraperitoneal leptin on sleep-disordered breathing in mice with diet-induced obesity. They found that only intranasal leptin improved ventilation and reduced upper airway obstruction during sleep. Leptin’s effects on breathing were independent from its effects on metabolism.76 These findings make leptin an attractive molecule to treat OHS.

Medications that reduce ventilation

Furosemide: Furosemide is frequently used in patients with OHS to reduce edema. The alkalosis induced by furosemide can compound the compensatory metabolic alkalosis leading to respiratory depression and worsening of hypercapnia. Unfortunately, most studies on the effects of furosemide in humans are in subjects with normal lungs or in patients with COPD.

Hazinski and colleagues evaluated the effect of furosemide on ventilation in rabbits.81 They demonstrated that it induced metabolic alkalosis, suppressed ventilation, and increased the PaCO2, but it did not change the ventilatory response to hypercapnia. In other words, the ventilatory response curve shifted to the right without changing its slope.

Ventilatory failure from metabolic alkalosis was described in depth by Jarboe and colleagues in a set of experiments in humans and dogs. The human data showed that in patients with normal lungs hypercapnia was a common accompaniment of metabolic alkalosis and lead to hypoxemia. The dog experiments confirmed these findings and demonstrated that alkalosis changed the ventilatory response to hypercapnia.82 Javaheri later summarized the effect of metabolic alkalosis on ventilation in humans and noted that metabolic alkalosis in humans results in hypoventilation from a decrease in tidal volume and that this hypoventilation ultimately leads to hypoxemia. The slope in the regression equations to predict PCO2 from bicarbonate level was 0.70–0.9 mmHg/mEq/L with the lower values observed in normal subjects, and the intercept was 9–21 mmHg.83

More recently, Feldman and colleagues studied in 52 patients the effect of diuretic and vomiting-induced alkalosis on PaCO2.84 These patients were not obese. The mean bicarbonate level was 32 mmoL/L and the mean pH was 7.48. The PaCO2 correlated strongly with the bicarbonate level (r = 0.97) and the regression equation that described the relation was PaCO2 = [bicarbonate × 1.2] + 6. Meaning that hypercapnia should be expected in patients with metabolic alkalosis when the bicarbonate level is above 33 mmoL/L. This was consistent with the findings of Jarboe and colleagues. But the rate of increase in PaCO2 in response to the increase in bicarbonates was higher than that found by Javaheri.

Urano and colleagues studied the effect of the dose of furosemide on the regression line between PaCO2 and hydrogen [hydrogen = intercept + PaCO2 × slope].85 They found that as the dose of furosemide increased, the intercepts decreased [intercept = −  6.9 × furosemide dose + 30.9] and the slope of the regression line increased [slope = 0.094 × furosemide dose + 0.22]. The regression lines of all the doses of furosemide crossed the PaCO2 value of 75 mmHg suggesting that the effect of furosemide is negligible at severe hypercapnia.

Discontinuing furosemide can improve ventilation. This was demonstrated in patients with COPD where discontinuing furosemide increased ventilation from 10.4 to 11.6 L/min and reduced the PaCO2 from 45 (range: 35–64 mmHg) to 41 mmHg (range: 32–61 mmHg).86

There is also a reciprocal effect of hypercapnia and hypoxia on the pharmacodynamics of furosemide. Babini and colleagues report that in rabbits, hypercapnia with hypoxemia reduced renal blood flow, increased the reabsorption of sodium in the tubules and therefore reduced its excretion, and reduced the diuresis induced by furosemide.87 Hypercapnia with hypoxia also increased reabsorption of furosemide and reduced its renal clearance of furosemide because of decreased renal blood flow.

Oxygen: Oxygen therapy can also depress ventilation in patients with OHS. Wijesinghe and colleagues studied the effect of breathing 100% oxygen for 20 min on ventilation of patients newly diagnosed with OHS that had a BMI 52 ± 11 kg/m2. On average, the PaCO2 increased by 5.0 mmHg (95% CI: 3.1–0.8 mmHg).88 However, it increased in three subjects by more than 10 mmHg in <  15 min. The increase in PaCO2 was larger, the higher the baseline PaCO2 and the lower the baseline ventilation. This hypoventilation resulted from a decrease in tidal volume from 0.69 to 0.56 L with no change in respiratory rate.

These findings were extended by Hollier and colleagues who studied the effect of lower concentrations of oxygen therapy (28% and 50%) on the ventilation of 14 patients with OHS with a BMI of 53 ± 7 kg/m2. Arterialized venous blood was used to monitor changes on PCO2.89 The PCO2 increase by 2.3 ± 1.5 mmHg on 25% oxygen and by 3.8 ± 3.0 mmHg on 50% oxygen. Minute ventilation did not change with 28% oxygen and decreased by 1.2 ± 2.1 L. By 20 min, ventilation returned to baseline in some patients, increased above baseline in some, and remained low in others. The decrease in ventilation was also caused by a decrease in tidal volume. After breathing 50% oxygen, the tidal volume decreased from 0.58 ± 0.11 to 0.49 ± 0.11 L.

Opiates and benzodiazepines: Opiates reduce upper airway tone and respiratory rate and induce irregular ventilation.90 These effects can induce hypoventilation especially in older and in morbidly obese patients. The elderly are particularly susceptible because of decreased elimination of opiates and increased sensitivity to them. Benzodiazepines have a variety of effects on sleep and breathing.91 Although they can reduce the severity of certain types of respiratory events, benzodiazepines can also worsen sleep hypoventilation and the severity of sleep-disordered breathing in patients with COPD and can increase the frequency and duration of apneas and worsen the desaturations in patients with sleep-disordered breathing. These medications are better avoided in patients with untreated OHS.

Key points

Weight loss of more than 25% is needed to improve ventilation of patients with OHS.

Weight loss of this extent can be achieved with certain types of bariatric surgery.

Acetazolamide can be used at low doses (250–500 mg daily for 3–7 days) to treat the metabolic alkalosis (pH >  7.35 and bicarbonate level >  34 mmoL/L) that emerges after treating patients with OHS with mechanical ventilation or furosemide, which prevents spontaneous ventilation from increasing. Hypokalemia is a common side effect.

Medroxyprogesterone acetate (20 mg three time daily for less than a month) improves ventilation in patients with OHS that have not been treated with PAP.

Furosemide and oxygen can rapidly worsen hypercapnia in patients with OHS.

Tracheostomy improved ventilation in patients with OHS but is rarely performed.

Opiates and benzodiazepines should be avoided in patients with untreated OHS.