Obstructive vs central vs complex sleep apnea — pathophysiology and detection

The word “apnea” in sleep medicine describes a single observation — airflow stopped for at least 10 seconds — but hides three distinct physiological mechanisms. A patient with an AHI of 40 might have pure obstructive disease, pure central disease, or a mixed picture that shifts further when CPAP therapy is applied. The treatment for each is different. The diagnostic distinction hinges on one question: during the apnea, was there continuing respiratory effort, or not? That question is answerable only with effort signals recorded alongside flow, and it is answered wrongly surprisingly often in the Indian sleep-lab ecosystem, where effort belts are occasionally skipped in hurried studies.

This article covers the three mechanisms, how they look on the flow-and-effort traces during a PSG, the relevant AASM scoring rules, the incidence of complex sleep apnea (CompSAS) in heart-failure populations, why CPAP can worsen central apnea in some patients, and the indications for adaptive servo-ventilation (ASV) or BiPAP-ST.

The three mechanisms in one sentence each

Obstructive sleep apnea (OSA). The airway collapses; the respiratory drive from the brainstem and the effort from the diaphragm and intercostals continue, trying to push air through a closed upper airway. Thoracoabdominal effort is preserved or often increased. Flow at the mask/thermistor drops to near zero. The patient is “trying to breathe against a closed door.”

Central sleep apnea (CSA). The brainstem’s respiratory drive pauses. No neural command reaches the diaphragm. Effort ceases. Flow ceases. The patient is “not trying to breathe.”

Mixed apnea. An apnea begins as central (no effort) and transitions mid-event into obstructive (effort returns while the airway is still closed). Scored as mixed under AASM rules; clinically shares drivers with both obstructive and central disease.

Complex sleep apnea syndrome (CompSAS) / treatment-emergent CSA. A patient initially presents with predominantly obstructive events. CPAP therapy suppresses those obstructive events but reveals or induces central apneas. Residual CAI on CPAP is > 5 with CAI/total > 50%. This is physiologically a CSA pattern that was masked by the OSA; in some patients, the CSA component was present subclinically and only became visible once the obstruction was relieved. In others, CPAP itself may contribute by altering CO₂ set-points via washout.

How effort belts distinguish obstructive from central

A Type I polysomnography records respiratory effort via two respiratory inductance plethysmography (RIP) belts — one around the thorax, one around the abdomen. The belts detect rib-cage and abdominal circumference changes breath-to-breath. During normal breathing, both bands expand synchronously during inspiration. Scoring criteria (AASM Practice Guidelines):

• Obstructive apnea. Thoracic and abdominal bands continue to show effort (often increased effort, sometimes paradoxical movement with the thorax going in while the abdomen goes out). Flow channel shows ≥ 90% reduction for ≥ 10 seconds. The patient is trying hard.
• Central apnea. Both thoracic and abdominal bands go flat. No detectable effort. Flow is zero. The patient is not trying.
• Mixed apnea. First part of the event shows no effort (central); second part shows returning effort with flow still obstructed (obstructive). The event is scored as mixed.

Without effort belts — for example in a Type III HSAT that only records flow, SpO₂, and position — the distinction between obstructive and central apnea is not directly possible. Some Type III devices attempt to infer effort from flow-signal characteristics or from pulse-wave amplitude variations, but the inference is less reliable than direct RIP measurement. This matters because Type III HSAT is increasingly used in India for diagnostic workup, and a patient with significant central apnea may be diagnosed as “OSA” on a Type III and have the central component missed until CPAP is started and CompSAS becomes obvious.

Physiology — why centrals happen

Central apnea fundamentally reflects instability in the ventilatory control system. The respiratory drive is feedback-driven: brainstem CO₂ / H⁺ chemoreceptors detect changes in arterial CO₂ and adjust minute ventilation to keep PaCO₂ near a setpoint. In health, this system is stable. Several conditions make it unstable:

• Heart failure with reduced ejection fraction (HFrEF). Prolonged circulation time between lungs and chemoreceptors causes delayed CO₂ signal. The feedback loop overshoots and undershoots, producing the classic Cheyne-Stokes waxing-and-waning breathing pattern — crescendo breaths, then hypopnea/apnea, then crescendo again. Central events cluster at the troughs.
• High-altitude exposure. Hypoxic hyperventilation drives CO₂ down below the apneic threshold; the brainstem pauses drive until CO₂ re-accumulates. Periodic breathing of altitude is a predictable CSA variant. Indian patients in Leh, Manali, Srinagar, Gangtok develop this pattern seasonally.
• Opioid use. Chronic opioids (and some other CNS depressants) blunt chemoreceptor sensitivity and can cause irregular central events.
• Stroke involving brainstem / cerebellar respiratory centres. Direct central control disruption.
• Idiopathic CSA. No identifiable cause; rarer.
• Treatment-emergent CSA (CompSAS). CPAP reduces upper airway resistance and can wash out CO₂, briefly dropping below the apneic threshold and triggering central events. Usually emerges in the first weeks of CPAP therapy.

Cheyne-Stokes respiration (CSR) specifically refers to the waxing-waning pattern seen in heart failure. It’s a subset of central sleep apnea — all CSR is CSA, but not all CSA is CSR. The crescendo-decrescendo breathing morphology is distinctive on the flow trace.

CompSAS incidence — heart failure and beyond

Population estimates for treatment-emergent CSA in CPAP-initiated patients run roughly 5–15% of newly-treated OSA patients showing CompSAS criteria at 1–3 months post-initiation. The figure is higher in heart-failure cohorts and in opioid-using populations. Published heart-failure cohorts show CSA/CompSAS prevalence of 25–40% in HFrEF with NYHA II–IV disease, making this a common rather than rare presentation in that subgroup. (AASM Practice Guidelines)

In Indian practice, two groups are under-recognised:

1. Heart failure patients who present with fatigue and loud breathing. These are often worked up for OSA, given a PSG showing AHI 30–50, prescribed CPAP — and then the central component emerges. A HF patient with Cheyne-Stokes on the initial PSG should not be started on standard CPAP without a central-apnea evaluation.
2. Patients on chronic opioids for cancer pain or post-surgical pain. CSA can be prominent here and may not resolve with CPAP alone.

Why CPAP can worsen central apnea

The CO₂ washout story in one paragraph: CPAP delivers continuous positive pressure, reducing the work of breathing. In some patients, this results in slight hyperventilation and a drop in PaCO₂. If the drop pushes PaCO₂ below the patient’s apneic threshold (the CO₂ level below which the brainstem stops driving respiration during sleep), central apneas ensue. The CPAP has “improved” the obstructive component by doing its job, but exposed or caused a central component. This is the pathophysiological core of CompSAS.

Clinically, this appears as a patient who starts CPAP, reports subjective improvement in the first week, then plateaus or regresses. The download shows obstructive AHI trending down and central AHI trending up. Total AHI may be only modestly improved — or in some cases unchanged — from baseline.

When ASV or BiPAP-ST is indicated

Adaptive servo-ventilation (ASV) is a pressure-support device that dynamically adjusts inspiratory pressure to counteract irregular breathing patterns — including Cheyne-Stokes cycles and central apneas. The device targets a minute-ventilation or peak-flow setpoint and adds inspiratory pressure when the patient’s breathing falls below it. ASV is effective at suppressing CompSAS and idiopathic CSA and has historically been a preferred therapy for CompSAS after standard CPAP fails.

Important caveat: the SERVE-HF trial. In patients with HFrEF (EF ≤ 45%) and predominant CSA, ASV therapy was associated with increased cardiovascular mortality compared to control. As a result, ASV is contraindicated in HFrEF with LVEF ≤ 45% and predominant CSA. In other populations (CompSAS after OSA treatment, preserved EF, idiopathic CSA), ASV remains appropriate and effective.

BiPAP-ST (spontaneous-timed) provides two pressure levels (IPAP / EPAP) with a backup respiratory rate. If the patient fails to trigger a breath within the backup interval, the device delivers a timed breath at IPAP. BiPAP-ST is indicated for:

• CSA with hypercapnia (e.g., chronic hypoventilation syndromes).
• Neuromuscular disease with respiratory muscle weakness.
• Obesity hypoventilation syndrome (often with AVAPS for target-volume assurance).
• HFrEF with predominant CSA, where ASV is contraindicated by SERVE-HF.
• Post-surgical / post-stroke patients with unstable ventilation.

Standard CPAP remains first-line for straightforward OSA without CSA components. For mixed pictures, the decision tree is:

1. AHI primarily obstructive? CPAP or APAP.
2. Treatment-emergent CSA on CPAP, preserved EF? Consider ASV.
3. CSA with hypercapnia, or HFrEF with EF ≤ 45%? BiPAP-ST (backup rate).
4. CSR in heart failure, preserved EF? ASV is an option; HFrEF with EF ≤ 45% — BiPAP-ST or optimised medical therapy.

Diagnostic workflow for a suspected CSA / CompSAS patient

1. Diagnostic PSG with effort belts. Type III HSAT is inadequate; the distinction requires RIP.
2. Review the flow-and-effort trace morphology. Cheyne-Stokes crescendo-decrescendo, or flat effort with central events, or preserved effort during events (OSA)?
3. Cardiac workup. Echocardiogram, NT-proBNP, NYHA class assessment. Essential in any patient with centrals.
4. Medication review. Opioid use? CNS depressants?
5. CPAP trial (if predominantly obstructive). Download at 1 month; re-evaluate central component.
6. If CompSAS emerges: decide between ASV (preserved EF) and BiPAP-ST (HFrEF ≤ 45% or hypercapnia).
7. Titration for the chosen device. ASV and BiPAP-ST both benefit from attended titration given the complexity.

High-altitude Indian context — periodic breathing and Leh / Manali patients

Indian patients traveling to or residing at high altitudes deserve a specific note. Above approximately 2500 metres, hypoxia-driven hyperventilation lowers PaCO₂ toward the apneic threshold, and periodic breathing emerges in a substantial fraction of healthy travellers. In Leh (3500m), Manali (2050m, borderline), Gangtok (1600m, generally subclinical), Darjeeling (2000m), Shimla (2200m), Ooty (2200m), Munnar (1500m), Mussoorie (2000m), and Srinagar (1600m), the phenomenon ranges from mild to clinically significant.

A patient with known CSA or OSA who travels to altitude may see dramatic worsening. A patient without any prior sleep-disordered breathing may develop new-onset periodic breathing that resolves on descent. The distinction between “altitude-induced periodic breathing” and “unmasked CSA” requires a descent trial: if the pattern resolves at sea-level testing, the diagnosis is altitude-related; if it persists, there is underlying instability.

For Indian patients on BiPAP-ST or ASV who travel to altitude, the therapy usually handles the additional instability well, though the backup rate may need temporary adjustment. CPAP-only patients at altitude sometimes develop CompSAS that was not present at sea level. A clinician counselling a patient before a Leh or Manali trip should include this in the conversation.

Differential and confounders

Several conditions can mimic CSA on PSG and should be considered in the differential:

• Periodic limb movement disorder (PLMS) with arousal-driven breathing disturbances. Limb-EMG channels distinguish.
• Nocturnal seizures with respiratory disruption. EEG distinguishes.
• Anxiety-driven sighing dysregulation in young patients. Clinical history and psychiatric evaluation.
• Thyroid disease (severe hypothyroidism) with altered respiratory drive. Biochemistry.
• Chronic metabolic alkalosis from diuretic use in HF patients, altering CO₂ set-point. Biochemistry.

Each of these can produce event patterns that the naïve reader calls “CSA” but which respond to the underlying condition rather than to ventilatory support.

Clinical takeaway

Obstructive, central, and complex sleep apnea are three different diseases with three different treatments, all sharing the word “apnea.” The flow-and-effort trace on a Type I PSG is the diagnostic substrate; Type III HSAT is insufficient for this distinction. Heart failure patients deserve particular scrutiny — CSA prevalence is high and misclassification has therapeutic consequences. CompSAS is common enough in newly-initiated CPAP patients that every CPAP follow-up at 1–3 months should include a specific look at central-event index.

HHZ’s editorial view: Indian sleep practice should not initiate CPAP in a patient with known HFrEF or prominent CSR on the diagnostic PSG without a cardiology review and a considered therapy-mode decision. Default-to-CPAP workflow misses CompSAS and, in HFrEF, ASV misuse risks harm per SERVE-HF.

Consult your sleep physician and cardiologist for any patient with combined sleep-disordered breathing and cardiac comorbidity — the therapy decision in this group is not a routine CPAP prescription.

References: AASM Manual for the Scoring of Sleep and Associated Events v3 (AASM Practice Guidelines); Cowie MR et al, SERVE-HF N Engl J Med 2015 [CITATION]; Javaheri S et al, CSA in heart failure; Berry RB et al — AASM central sleep apnea scoring updates.