Findings from a recent study published in Asia-Pacific Journal of Ophthalmology found that side sleeping is associated with reduced optic nerve blood flow in the eye closest to the pillow, a finding that may help explain asymmetric optic nerve damage in patients with normal-tension glaucoma (NTG) and obstructive sleep apnea (OSA)..
Give me some background first.
With an estimated 1 billion people worldwide impacted by OSA, this condition is a well-known systemic risk factor for glaucoma—though the mechanism behind the link is still unclear.
- New stats: A recent meta-analysis found that people with OSA are 6.72 times more likely to develop NTG than those without.
And why does that matter?
NTG often presents asymmetrically even though intraocular pressure (IOP) measurements in the sitting position are usually similar between eyes.
- That mismatch has pushed researchers to look at what happens during sleep, when IOP and ocular blood flow shift with body position.
Also worth noting: Between 30% and 60% of NTG patients prefer to sleep in the lateral decubitus (LD) position (side sleeping). OSA patients tend to favour it too, since lying on the side helps relieve airway obstruction.
Now, talk about the study.
Researchers at the Singapore National Eye Centre ran a cross-sectional study to characterize ocular features and ONH blood flow velocity across supine and right and left LD positions in patients with asymmetrical NTG and OSA who habitually sleep on their side.
The tools:
- Laser speckle flowgraphy (LSFG) was used to derive mean blur rates (MBR) for the vascular area (MV), tissue area (MT), and entire ONH area (MA) after 5 minutes of stabilization in each position.
- IOP was measured with a Tonopen at sitting, supine, and both LD positions.
Who was included in the study?
Investigators enrolled 40 NTG-OSA habitual side sleepers (21 with unilateral glaucoma, 19 with bilateral but asymmetrical disease) and 29 healthy controls.
To make the cut, NTG-OSA participants needed:
- An apnea-hypopnea index (AHI) ≥ 5 events/hour with symptoms or ≥ 15 events/hour regardless of symptoms
- A mean deviation (MD) difference of at least 2 dB between eyes on Humphrey visual field testing
- Self-reported habitual side sleeping (mostly right or mostly left)
- Pre-treatment IOP < 21 mmHg
Demographics: Mean age was 62.9 years, 70% male, with a mean body mass index of 26.4. Mean AHI sat at 33.9 events/hour.
Of the 40 NTG-OSA participants, 33 identified a predominant sleeping side and were used in subgroup analyses.
Findings?
In the worse glaucoma eyes, MT dropped to 9.00 ± 2.07 au in the lower LD position compared with 9.27 ± 1.97 au in the upper LD (P = 0.037) and 9.61 ± 1.95 au supine (P = 0.011).
- Worse eyes also had thinner central corneal thickness (CCT) than fellow eyes:
- 543.95 vs 548.10 µm, P = 0.019
The kicker: IOP showed no significant inter-eye difference in any position, supine or LD. The MBR drop in the dependent eye (eye against the pillow) was happening independently of measurable pressure changes.
Tell me more.
Investigators split the side sleepers into two groups based on whether the worse eye was habitually positioned lower (n = 20) or upper (n = 13) during LD sleep.
- Group 1 (worse eye lower): MV, MT, and MA were all significantly less in the worse eyes than in fellow eyes in both LD positions (all P < 0.05). No inter-eye difference appeared in supine.
- Group 2 (worse eye upper): Inter-eye differences only emerged when the worse eye was forced into the lower LD position.
- MV came in at 30.74 ± 10.05 au in the worse eye vs 36.31 ± 6.43 au in the fellow eye (P = 0.008)
- MT at 8.08 ± 1.48 au vs 10.39 ± 1.49 au (P = 0.008).
- No difference when the worse eye was upper, and none supine.
And in practice?
After adjusting for CCT, vertical cup-disc ratio (VCDR), MD, IOP-lowering medications, and inter-eye correlation, the change in MT from supine to the upper LD position was significantly greater in study eyes than fellow eyes.
The estimate: −1.15; 95% CI: −2.01, −0.30; P = 0.008
Any limitations?
This was a single-center, cross-sectional study, so the usual cautions about causality apply.
The authors flagged the findings as exploratory and hypothesis-generating, not confirmatory.
Other considerations:
- Sample size was modest, especially for subgroup analyses, raising the risk of both type I and type II errors
- Sleep position was self-reported, with no polysomnographic verification. A prior study found objective and self-reported sleep positions matched only 77% of the time
- Measurements followed a fixed sequence with no randomization or counterbalancing, which could introduce order effects
- All glaucoma eyes were on IOP-lowering medications that may improve ocular blood flow, which the authors statistically adjusted for but could not fully eliminate
- The 2 dB MD threshold for asymmetry may fall within test variability
Expert opinion?
The authors argued that habitual side sleeping combined with posture-related reductions in ONH blood flow—particularly when the worse eye is dependent—may contribute to asymmetric glaucomatous damage in susceptible eyes.
The supine position consistently showed no inter-eye MBR differences, suggesting the dependent position is doing something specific.
Anything else?
The lack of IOP differences across positions in this cohort contrasts with earlier studies that found higher IOP in the lower-positioned eye during LD sleep, especially in NTG.
- The authors attributed the discrepancy partly to the small sample size and uniform use of medical therapy, and called for future studies with larger samples and a medication washout period.
Worth noting: Repeated apnea-hypopnea episodes during sleep can reduce arterial oxygen levels and expose the ONH to hypoxia.
The combination of hypoxic stress and persistent perfusion reduction in one eye over years of side sleeping may be what tips the balance toward asymmetric disease.
Take home.
For clinicians managing NTG patients with OSA, sleep position appears to matter. The eye that spends years in the dependent LD position may be sustaining repeated drops in ONH perfusion that do not show up on standard IOP measurements.
As such: Asking patients about habitual sleeping side, and counseling those with asymmetric disease about positional sleep modifications, may be worth folding into routine glaucoma management while larger prospective studies sort out causation.