In this article· 13 sectionsClick to collapse

11. Common clinic scenarios
22. Common patient Q&As
33. 30-second takeaway: 8 things to remember
44. Two diseases — a quick primer
55. What is this 2026 systematic review and meta-analysis?
66. Result 1: Overall AMD risk (three sets of analyses)
77. Result 2: Late-AMD risk signals
88. Why might OSA damage the macula? (mechanism chain)
99. Limitations (why we can't jump to «OSA causes AMD»)
1010. Clinical implications and Taiwan context
1111. Red flags: when to seek care immediately
1212. Conclusion: shifting AMD from «unchangeable» towards «one modifiable factor»
13References

Enlarge diagram

Scroll horizontally and vertically to read the diagram. Close Enlarge diagram to restore its original size.

OSA-AMD link at a glance. ① The upper airway repeatedly collapses during sleep → recurrent nocturnal SpO₂ drops (intermittent hypoxia) → oxidative stress and complement activation → RPE injury and choroidal hypoperfusion at the macula. ② Five pooled estimates: adjusted aOR 1.44 (moderate certainty, I²=0%); time-to-event aHR 1.66; crude nAMD OR 1.76; crude non-neovascular OR 1.95; anti-VEGF requirement aHR 2.85 (single study, very low certainty). ③ Three clinical directions: (green) OSA is one of few treatable systemic AMD risk factors — CPAP, weight loss, smoking cessation; but whether OSA treatment reduces AMD risk is still unproven; (yellow) ophthalmology clinic can ask one more question about snoring and daytime sleepiness; refer to pulmonary or sleep medicine if positive — but routine screening is not yet guideline-recommended; (red) this is observational only — no causal inference, CPAP treatment status unrecorded in most studies, and confounders (DM, obesity) not uniformly adjusted. Source: Yaldo et al. 2026 AJO (8 studies, 3,536,314 people incl. Taiwan NHIRD).

1. Common clinic scenarios

'Doctor, my wife says I snore loudly and stop breathing for seconds at night. Recently I notice my central vision is a bit blurry and straight lines look bent. Online they say this might be macular degeneration. Is sleep apnea really related to macular degeneration?'

'Doctor, I've already been diagnosed with obstructive sleep apnea by pulmonology and I'm on CPAP. Both my parents have dry AMD. How should I monitor my eyes? Any special tests? How often should I come?'

'Doctor, I was diagnosed with wet AMD last year and have had several anti-VEGF injections. My wife says I snore badly and I'm sleepy during the day — should I also get a sleep study? Will treating sleep apnea slow down my AMD?'

All three scenarios revolve around the same core question: between obstructive sleep apnea (OSA) and age-related macular degeneration (AMD), is the link a coincidence or a real risk? A May 2026 systematic review and meta-analysis by Yaldo et al. in American Journal of Ophthalmology—one of the few AMD-focused meta-analyses, integrating multiple study designs and analysing both odds-ratio and time-to-event data—offers the most complete answer to date. This article translates it into clinic-friendly language.

2. Common patient Q&As

Q1: Does sleep apnea «really» increase the risk of AMD?

A: Yes, but only mildly to moderately, with certainty of evidence ranging from very low to moderate. Yaldo et al. 2026 pooled 8 observational studies (3,536,314 participants). The most credible analysis (4 studies, with adjustment): OSA patients have aOR 1.44 (95% CI 1.11-1.77) for AMD, moderate certainty, no heterogeneity (I²=0%). Another time-to-event analysis (4 studies): aHR 1.66 (95% CI 1.13-2.19, low certainty). In plain language: OSA patients are about 1.4 to 1.7 times more likely to develop AMD than non-OSA peers. In the crude analysis, the subgroup of studies without confounder adjustment showed no significant difference, while the adjusted estimates still showed an association — suggesting the OSA-AMD link persists after controlling for other factors and is not simply explained by OSA patients being older/heavier/more cardiovascular-comorbid; however, diabetes, obesity and coronary artery disease were not adjusted consistently across studies, so residual confounding cannot be fully ruled out.

Q2: Which type of AMD goes up more? Dry or wet?

A: In crude analyses not adjusted for confounders, both types increase, with similar magnitudes. Neovascular AMD (nAMD, «wet»): OR 1.76 (95% CI 1.06-2.93, low certainty, 2 studies). Non-neovascular AMD (including early/intermediate and geographic atrophy): OR 1.95 (95% CI 1.04-3.66, low certainty, 2 studies). From a single study: anti-VEGF requirement aHR 2.85; late GA aHR 6.13. But because these two come from a single study with «very low» GRADE certainty, they remain hypothesis-generating rather than conclusive. In time-to-event analyses by stage, neither the nAMD nor the non-neovascular hazard ratio was statistically significant (very low certainty). The overall message: OSA isn't selectively linked to one AMD subtype — both neovascular and non-neovascular AMD appear to be at risk.

Q3: Why does sleep apnea damage the macula?

A: The key term is «intermittent hypoxia». In OSA, the upper airway repeatedly collapses during sleep, dropping SpO₂ from normal to around 80% (in one US veterans' sleep-lab series, the mean overnight minimum in severe OSA was about 78%), then recovering — dozens to hundreds of cycles per night. This drives four mechanisms: (1) oxidative stress injures the retinal pigment epithelium (RPE); (2) abnormal complement activation — one of the strongest validated molecular pathways in AMD (CFH, C3, CFB variants are well-known AMD risk loci); (3) sympathetic surges and blood-pressure swings compromise choroidal perfusion; (4) endothelial dysfunction may promote choroidal neovascularization. Plain translation: the cost of nightly desaturation isn't just daytime fatigue — it's chronic macular inflammation.

Q4: If I treat sleep apnea (CPAP, weight loss), will my AMD risk drop?

A: A reasonable hypothesis — but currently no study has shown that treating OSA reduces AMD risk. This is the field's biggest unknown. In Yaldo et al. 2026, almost none of the 8 studies recorded whether participants used CPAP, lost weight, or had other OSA treatment. The comparison was «OSA» vs «no OSA», not «OSA on treatment» vs «OSA untreated». If a future RCT shows OSA treatment lowers AMD risk, OSA would become one of very few modifiable systemic AMD risk factors (age and genes are not modifiable). Until then, the established reasons to treat OSA remain cardiovascular event reduction, daytime alertness, and crash prevention — not «for the eyes».

Q5: I have a family history of AMD — should I proactively get a sleep study?

A: The US Preventive Services Task Force (USPSTF, 2022) concluded that the evidence is insufficient to assess screening asymptomatic adults for OSA. But you can self-check with the validated STOP-Bang questionnaire (8 items, 1 point per «yes»): (1) Snoring loudly (heard through closed doors); (2) Tired, fatigued or sleepy during the daytime; (3) Observed to stop breathing during sleep; (4) high blood Pressure, or being treated for it; (5) BMI >35 kg/m²; (6) Age >50 years; (7) Neck circumference >40 cm; (8) male Gender. The original study defined ≥3 points as high risk of OSA (the current official website further splits 3-4 as intermediate and 5-8 as high risk). With a score of ≥3, see a pulmonology or ENT sleep clinic and discuss polysomnography (PSG). The justification is overall health — not AMD prevention.

Q6: I already have AMD and I snore — what should ophthalmology do?

A: There is no OSA-specific change to AMD care. But three reasonable steps: (1) Ask one more question — snoring, daytime sleepiness, witnessed apneas? (2) If positive, refer to pulmonary or sleep medicine — not «to protect the eyes», but because OSA treatment helps cardiovascular health, cognition, and quality of life. (3) Follow-up cadence stays guided by AMD stage (per the 2024 AAO Preferred Practice Pattern: every 6-24 months for early AMD if asymptomatic, 6-18 months for intermediate AMD or advanced AMD in one eye, with prompt examination for new symptoms; individualised for late disease) — current evidence does not support intensifying follow-up specifically because of OSA.

Q7: Beyond AMD, what other eye diseases is OSA linked to?

A: OSA is a multi-system disease and ophthalmology has linked it to several eye conditions. Commonly cited: (1) diabetic retinopathy — associated in unadjusted analyses but not clearly after adjustment, so the link remains unsettled; (2) non-arteritic anterior ischemic optic neuropathy (NAION) — vision loss commonly noticed on waking; (3) central serous chorioretinopathy — the pooled confidence interval still crosses 1, so the evidence is less certain; (4) glaucoma; (5) corneal structural disorders such as keratoconus; (6) floppy eyelid syndrome. So if any of these is identified in clinic, asking «do you snore? daytime sleepy?» is a meaningful clue.

3. 30-second takeaway: 8 things to remember

📋 OSA × AMD: 8 things to remember

Associated, not yet causal: 8 studies, 3.5M people; adjusted aOR 1.44 (moderate certainty, I²=0%); aHR 1.66 (low)
Both wet and dry AMD may be affected: crude nAMD OR 1.76, non-neovascular 1.95, similar magnitude; stage-specific time-to-event analyses were not significant
Single-study late AMD signal: anti-VEGF aHR 2.85, GA aHR 6.13; very low certainty — hypothesis only
Mechanism plausible: intermittent hypoxia → oxidative stress + complement + choroidal hypoperfusion + endothelial dysfunction
One of few treatable systemic AMD risk factors: CPAP, weight loss, smoking cessation — but no evidence yet that treatment reduces AMD
Routine screening NOT recommended: authors state current evidence does not support routine cross-specialty screening; but ophthalmologists may consider a few targeted questions (sleep quality, snoring, witnessed apneas, daytime sleepiness, morning headaches)
OSA is multi-system: ophthalmic studies also report links with NAION, floppy eyelid, keratoconus and glaucoma; links with DR and CSR are less certain
Major limitations: observational only, variable diagnostic criteria, CPAP treatment status unrecorded in most studies, confounders (DM, obesity, CAD) inconsistently adjusted

4. Two diseases — a quick primer

Obstructive sleep apnea (OSA) is a disorder of recurrent upper-airway collapse during sleep, causing repeated drops in blood oxygen. The diagnostic gold standard is polysomnography (PSG) — overnight recording of EEG, respiration, oxygen saturation, and cardiac rhythm in a sleep lab. Severity is quantified by the apnea-hypopnea index (AHI): ≥5/hr mild, ≥15/hr moderate, ≥30/hr severe. The most-used treatment is continuous positive airway pressure (CPAP); mild-to-moderate cases may use a mandibular advancement device (MAD). Weight loss, smoking cessation, avoiding supine sleep, and limiting alcohol all help. A minority benefit from surgery (e.g., uvulopalatopharyngoplasty).

Age-related macular degeneration (AMD) is a leading cause of irreversible vision loss worldwide, mostly affecting people over 50. Clinically, the Beckman Initiative for Macular Research classification groups it into early, intermediate, and late stages by drusen size and pigmentary changes. Late AMD has two phenotypes: (1) neovascular AMD (nAMD, «wet») — choroidal neovessels grow under the retina causing haemorrhage or exudation, with rapid vision loss; mainstay treatment is anti-VEGF injection; (2) geographic atrophy (GA, dry late) — geographic loss of RPE and photoreceptors with gradual vision decline; in 2023 the US FDA approved two intravitreal complement inhibitors (pegcetacoplan, avacincaptad pegol) that slow lesion growth but do not restore lost vision; neither had been approved in Taiwan as of September 2026. Early/intermediate AMD is managed by regular surveillance; for intermediate AMD (or late AMD in one eye), AREDS/AREDS2 supplementation (lutein, zeaxanthin, vitamin C/E, zinc, copper) is proven to cut the risk of progression to late AMD by about 25%, with no proven benefit in early AMD.

5. What is this 2026 systematic review and meta-analysis?

Yaldo et al. 2026 (lead institution: Bascom Palmer Eye Institute) followed the PRISMA 2020 reporting guideline and pre-registered the protocol with PROSPERO (CRD420251119881). The search covered PubMed, Web of Science, and Scopus through 27 June 2025, supplemented by Google Scholar and citation tracking. Eligibility: (1) observational studies (cohort / case-control / cross-sectional); (2) reported effect estimates for OSA and AMD, or raw data sufficient to calculate them; (3) included both OSA and non-OSA groups; (4) reported AMD outcomes. Eight studies were included — 4 retrospective cohorts, 2 case-control, 1 prospective cohort, 1 cross-sectional. Geographies: US 2, Australia 2, UK 2, Canada 1, Italy 1, Taiwan 1 (Blazes et al. 2025, NHIRD cohort); Han 2021 analysed both UK and Canadian cohorts, so the country counts add up to 9. Study quality was rated with the NIH tool for observational studies — 3 «good», 5 «fair».

Certainty of evidence was assessed using the GRADE framework across five domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Observational studies start at «low» certainty and can be upgraded or downgraded. Publication bias was not formally tested because each analysis had <10 studies (under-powered by methodological consensus). Pooling used random-effects (DerSimonian-Laird) to handle between-study heterogeneity.

6. Result 1: Overall AMD risk (three sets of analyses)

The meta-analysis estimated the OSA-AMD association from three angles — crude (unadjusted) odds ratio, adjusted odds ratio, and adjusted hazard ratio — analysed independently to demonstrate robustness.

Scroll horizontally to read the table. With a keyboard, focus the table area and use the arrow keys.

Analysis type	Studies	Effect estimate (95% CI)	Heterogeneity I²	GRADE certainty
Crude OR	8	1.45（1.13-1.84）	95.10%	Low
Adjusted aOR	4	1.44（1.11-1.77）	0%	Moderate
Adjusted aHR	4	1.66（1.13-2.19）	97.87%	Low
Propensity-score matched OR	2	1.92（1.05-3.52）	99.19%	Low
No-adjustment subgroup	2	1.32 (0.90-1.93) NS	33.11%	Low
Regression-adjusted subgroup	4	1.31 (0.96-1.78) NS	74.01%	Low

Three observations: (1) The adjusted aOR 1.44 is the most credible single number in the entire meta-analysis — 4 studies, I²=0% (highly consistent), narrow CI, GRADE moderate. (2) The aHR 1.66 point estimate is larger but heterogeneity is extreme (I²=97.87%) — meaning studies differ substantially, no precise number is possible, only the direction (OSA → AMD). (3) Propensity-score matched (PSM) analyses give the largest effect, OR 1.92 — PSM can adjust for many potential confounders, supporting a genuine association. But simple regression-adjusted studies show no significant difference — suggesting residual confounding may not have been fully handled in less rigorous adjustments. That said, the difference between the three adjustment subgroups was not statistically significant (p = 0.52).

7. Result 2: Late-AMD risk signals

Subgroup analyses by AMD stage are limited in study numbers (1-2 each) and certainty is lower, but still provide directional signals:

Scroll horizontally to read the table. With a keyboard, focus the table area and use the arrow keys.

AMD subtype / outcome	Studies	Effect (95% CI)	Certainty
Neovascular AMD (wet)	2	OR 1.76（1.06-2.93）	Low
Non-neovascular AMD	2	OR 1.95（1.04-3.66）	Low
Time-to-event: nAMD	2	aHR 1.69（0.19-3.19）NS	Very low
Time-to-event: non-neovascular	2	aHR 1.79（0.13-3.46）NS	Very low
Anti-VEGF requirement	1	aHR 2.85（2.18-3.51）	Very low
Late AMD with GA	1	aHR 6.13（2.94-9.32）	Very low

The «anti-VEGF» aHR 2.85 and «late GA» aHR 6.13 are striking but come from a single study (Alshaikhsalama et al. 2025 Ophthalmology Retina), GRADE-rated «very low» — meaning «hypothesis-generating, not conclusive». The main text does not itemise the reasons for the «very low» rating beyond noting that both come from a single study and are not conclusive; in general, a single observational study cannot show consistency across studies and cannot rule out residual confounding. These signals are worth following up in prospective studies and larger datasets — if replicated, and especially if they survive adjustment for CPAP treatment status, they could inform clinical prioritisation of OSA treatment.

8. Why might OSA damage the macula? (mechanism chain)

OSA is a «systemic intermittent-hypoxia disease» and AMD is a «local chronic degenerative disease of the macula» — on the surface, no obvious link. But at the molecular level, they share several pathogenic pathways — which is why the observed association in the meta-analysis is biologically plausible.

🔬 Four OSA → AMD pathways

1. Intermittent hypoxia → oxidative stress: dozens to hundreds of «desaturation-reperfusion» cycles per night generate excess reactive oxygen species (ROS), damaging RPE cell membranes and mitochondria
2. Intermittent hypoxia → complement activation: C3 and C5 expression rises in RPE and choroidal endothelium, driving local inflammation and RPE apoptosis. CFH, C3, CFB variants are among the strongest AMD genetic risk loci — OSA's intermittent hypoxia may «add on» to genetic susceptibility
3. Sympathetic surges + BP swings → choroidal hypoperfusion: OSA causes repeated sympathetic storms during sleep, raising rather than lowering nocturnal BP, and reducing choroidal perfusion under the macula. Imaging studies confirm thinner choroid and reduced choroidal vessel density
4. Endothelial dysfunction + VEGF up → choroidal neovascularization: intermittent hypoxia upregulates VEGF via the HIF-1α pathway — the key driver of choroidal neovessels in wet AMD

This mechanistic chain may also help explain why studies report links between OSA and eye diseases other than AMD (such as NAION and glaucoma); but these links come mostly from observational studies, and whether they truly share a pathogenic basis remains to be confirmed.

9. Limitations (why we can't jump to «OSA causes AMD»)

⚠️ 5 caveats to remember

All observational, no causal inference: all 8 studies are observational; can only show association, not causation
Variable OSA diagnostic criteria: 4 studies used ICD-9/10 codes (missing undiagnosed OSA), 1 oxygen desaturation index (ODI, home oximetry), 1 Epworth, 1 NHANES self-report; only 2 used polysomnography gold-standard (one of them also used ICD codes, so the counts add up to more than 8). A sensitivity analysis limited to these 2 studies gave OR 1.63 (95% CI 0.56-4.72), not statistically significant
Variable AMD diagnostic criteria: 2 used Beckman classification (most rigorous), 4 used ICD-9/10 codes, others used imaging-based or thickness measurements
CPAP treatment status not tracked: almost none of the 8 studies reported OSA treatment; only 1 did subgroup by device use, insufficient for meta-analysis. «Well-treated OSA» vs «untreated OSA» may have very different AMD impact
Results differ across adjustment methods: residual confounding and diagnostic variability limit inference. Diabetes, coronary artery disease and obesity were not adjusted consistently across studies; observational association is not causation. Stratified by study quality, the 3 «good»-quality studies showed no association in the crude analysis (OR 1.00, 95% CI 0.68-1.45, I²=0%), whereas the 5 «fair»-quality studies were highly heterogeneous (I² ≈ 96%); the between-group difference was significant (p = 0.03).

The authors themselves state clearly in the conclusion: findings should be «considered hypothesis-generating»; current evidence does not support routine cross-specialty screening or diagnostic testing; but because AMD patients often fall within an age group at increased risk of OSA, greater awareness of OSA and its symptoms among ophthalmologists may be appropriate. This is a responsible scientific stance — acknowledging the association while respecting its limits and not overstating clinical implications.

10. Clinical implications and Taiwan context

This meta-analysis has two extra implications for Taiwan readers: (1) one of the 8 studies is a Taiwan NHIRD cohort (Blazes et al. 2025 Retina) — meaning the finding is not entirely «foreign» for Taiwan's population; (2) OSA is not rare in Asia — in a Korean community study of adults aged 40-69 using overnight polysomnography, sleep-disordered breathing (AHI ≥5) was found in 27% of men and 16% of women (Kim 2004); in Taiwan, the Shihpai Eye Study found early AMD in 9.2% and late AMD in 1.9% of adults aged 65+ (about 11% in total; Chen 2008), and prevalence is expected to keep rising as the population ages.

⚠️ Costs & coverage — always confirm with your hospital

Costs for polysomnography (PSG), home sleep apnea testing (HSAT), NHI coverage vs out-of-pocket — confirm directly with your pulmonary, ENT, or sleep medicine physician and your hospital's pre-service notice. No claims here.
CPAP purchase or rental, subsidies: rules and support vary by hospital and time; check with your prescriber, the device vendor, and your insurance (NHI / commercial) for current info.
NHI coverage of anti-VEGF injections (ranibizumab, aflibercept, faricimab): Under Taiwan NHI drug reimbursement rules (updated 2026-10-22), section 14.9.2, wet AMD qualifies at age 50 or older with best-corrected visual acuity (BCVA) of 0.05–0.5 in the affected eye, supported by colour fundus photos, fluorescein angiography (FAG) and OCT; pre-approval is required and an ophthalmologist must give the injections. Eyes with a subfoveal scar, subretinal fibrosis or advanced geographic atrophy are not covered. The first approval allows up to 8 injections and the second and third 3 each — at most 14 per eye, used within 5 years of the first approval; continuing requires evidence such as at least one line of BCVA gain, improvement on imaging, and remaining disease activity. Only one of the three drugs may be chosen. For costs outside these conditions or beyond the injection limit, check with your hospital; discuss details with your doctor.
AREDS2 supplements (lutein, zeaxanthin, vitamin C/E, zinc, copper) are not NHI-covered. Quality and formulation vary; choose products with the full AREDS2 ratio and confirm appropriateness with your ophthalmologist.
Smoking cessation, weight loss, limiting alcohol: not NHI-covered, but these three lifestyle interventions benefit OSA, AMD, and cardiovascular health simultaneously — cumulative daily gains, long-term payoff.

11. Red flags: when to seek care immediately

OSA and AMD each have «do-not-delay» red flags. Memorise them separately — they aren't necessarily linked to the OSA-AMD association directly, but they are signals every patient and family should recognise.

🚨 AMD red flags (ophthalmology ER)

Sudden central distortion, bending of lines, or scotoma — possible new haemorrhage or exudation
Unilateral significant vision drop within days to a week
«Fish-eye lens» effect — central magnification or minification, asymmetric image
Known AMD + new Amsler grid changes (new bending, breaks, or scotomas)

If any of these, return same-day to your ophthalmology or retina specialist — for wet AMD, earlier anti-VEGF means better preserved vision.

🚨 OSA red flags (pulmonary / sleep clinic)

Severe daytime sleepiness that causes you to nod off while driving — the most dangerous OSA complication, directly linked to fatal crashes
Witnessed apneas >10 seconds, multiple times per night
Morning headaches with refractory hypertension — possible severe OSA with nocturnal hypoxia + CO₂ retention
Known arrhythmia, heart failure, or stroke without prior sleep workup — OSA is a treatable comorbid driver, deserves evaluation

If any of these, see pulmonary, ENT, or sleep medicine clinic for evaluation.

12. Conclusion: shifting AMD from «unchangeable» towards «one modifiable factor»

For three decades, the AMD risk-factor list has essentially read: «age, family history (CFH/C3/ARMS2 variants), smoking» — the first two are unchangeable; the third is the only well-established modifiable factor and is limited to smoking cessation. Yaldo et al. 2026's message: «OSA» may be the next «treatable systemic risk factor» worth pursuing seriously on this list.

But «possible» ≠ «proven». Current evidence: aOR 1.44 (moderate), aHR 1.66 (low), mechanism plausible — enough to say «association exists», not enough to say «OSA causes AMD», and not enough to say «treating OSA reduces AMD risk». Next steps needed:

Large prospective cohorts — PSG-confirmed OSA, Beckman-classified AMD, recorded CPAP status, ≥10 years follow-up
Randomised controlled trials (RCT) — comparing «intensive CPAP» vs «standard care» on AMD progression rate; the key design to confirm causality
Mechanistic studies — direct measurement of submacular choroidal perfusion, complement biomarkers (C3a, C5a), and retinal layer thickness in OSA patients to validate the pathway

Until those studies are out, the message for the public is simple: (1) Known OSA — treat it seriously regardless of eyes: CPAP, weight loss, smoking cessation, avoid supine sleep, limit alcohol. (2) Known AMD — regular follow-up remains paramount; if you snore loudly or feel daytime sleepy, mention it to your ophthalmologist for referral. (3) Neither — no need to panic from this study or get routine screening, but treat sleep quality as part of overall health, alongside diet, exercise, and smoking cessation.

The macula is the key input to your brain's visual cortex — protecting it starts with a quiet night's sleep.

📚 HsiaoEye Related Articles — Retina / Systemic

DR severity and dementia risk — Retinal microvasculature as a window on cerebral microvascular disease
6 Floater Red Flags — Retinal emergencies you should not delay
Recurrent SK After HZO — Another eye disease needing long-term topical steroid management
Pterygium Fixation NMA — Outdoor workers and their ocular risks

References

Yaldo L, Ngo A, Yaldo M, Abdelaal A, Ong J, Kiryakoza L, Sengillo J. Obstructive sleep apnea and risk of age-related macular degeneration: A systematic review and meta-analysis. Am J Ophthalmol. 2026;285:276–287. doi:10.1016/j.ajo.2026.02.016
Alshaikhsalama AM, Alsoudi AF, Wai KM, Koo E, Mruthyunjaya P, Rahimy E. Association between obstructive sleep apnea and age-related macular degeneration development and progression. Ophthalmol Retina. 2025;9(6):537–545. doi:10.1016/j.oret.2024.12.004
Blazes M, Ngadisastra C, Li PR, et al. Incidence and progression of age-related macular degeneration among patients with and without obstructive sleep apnea: a national cohort study. Retina. 2025;45(2):198–206. doi:10.1097/IAE.0000000000004293
Chaudhary A, Abbott CJ, Wu Z, et al. Nocturnal hypoxia and age-related macular degeneration. Clin Exp Ophthalmol. 2024;52(9):973–980. doi:10.1111/ceo.14428
Fang WY, Rama Raj P, Wu Z, et al. Role of sleep-disordered breathing in age-related macular degeneration. BMJ Open Ophthalmol. 2023;8(1). doi:10.1136/bmjophth-2022-001203
Han X, Lee SS, Ingold N, et al. Associations of sleep apnoea with glaucoma and age-related macular degeneration: an analysis in the United Kingdom Biobank and the Canadian Longitudinal Study on aging. BMC Med. 2021;19(1):104. doi:10.1186/s12916-021-01973-y
Keenan TD, Goldacre R, Goldacre MJ. Associations between obstructive sleep apnoea, primary open angle glaucoma and age-related macular degeneration: record linkage study. Br J Ophthalmol. 2017;101(2):155–159. doi:10.1136/bjophthalmol-2015-308278
Pedrotti E, Demasi CL, Bruni E, et al. Prevalence and risk factors of eye diseases in adult patients with obstructive sleep apnoea: results from the SLE.E.P.Y Cohort Study. BMJ Open. 2017;7(10):e016142. doi:10.1136/bmjopen-2017-016142
Nau JA. Association between age-related macular degeneration and sleep-disordered breathing [doctoral dissertation]. Walden University; 2017. https://scholarworks.waldenu.edu/dissertations/3463
Ferris FL III, Wilkinson CP, Bird A, et al. Clinical classification of age-related macular degeneration. Ophthalmology. 2013;120(4):844–851.
Abdelaal A, Mahmoud YA, Alsouri W, et al. Bidirectional association between obstructive sleep apnea and diabetic retinopathy: a severity-stratified systematic review and meta-analysis with GRADE assessment. Acta Diabetol. 2026. doi:10.1007/s00592-026-02732-2
Bulloch G, Seth I, Zhu Z, Sukumar S, McNab A. Ocular manifestations of obstructive sleep apnea: a systematic review and meta-analysis. Graefes Arch Clin Exp Ophthalmol. 2024;262(1):19–32.
Lyons MM, Bhatt NY, Pack AI, Magalang UJ. Global burden of sleep-disordered breathing and its implications. Respirology. 2020;25(7):690–702.
Patil SP, Ayappa IA, Caples SM, Kimoff RJ, Patel SR, Harrod CG. Treatment of adult obstructive sleep apnea with positive airway pressure: an American Academy of Sleep Medicine Clinical Practice Guideline. J Clin Sleep Med. 2019;15(2):335–343.
US Preventive Services Task Force; Mangione CM, Barry MJ, Nicholson WK, et al. Screening for obstructive sleep apnea in adults: US Preventive Services Task Force recommendation statement. JAMA. 2022;328(19):1945–1950. doi:10.1001/jama.2022.20304
Vemulakonda GA, Bailey ST, Kim SJ, et al. Age-Related Macular Degeneration Preferred Practice Pattern®. Ophthalmology. 2025;132(4):P1–P74. doi:10.1016/j.ophtha.2024.12.018
Chung F, Yegneswaran B, Liao P, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108(5):812–821. doi:10.1097/ALN.0b013e31816d83e4
Chung F, Abdullah HR, Liao P. STOP-Bang questionnaire: a practical approach to screen for obstructive sleep apnea. Chest. 2016;149(3):631–638. doi:10.1378/chest.15-0903
Kim J, In K, Kim J, et al. Prevalence of sleep-disordered breathing in middle-aged Korean men and women. Am J Respir Crit Care Med. 2004;170(10):1108–1113. doi:10.1164/rccm.200404-519OC
Chen SJ, Cheng CY, Peng KL, et al. Prevalence and associated risk factors of age-related macular degeneration in an elderly Chinese population in Taiwan: the Shihpai Eye Study. Invest Ophthalmol Vis Sci. 2008;49(7):3126–3133. doi:10.1167/iovs.08-1803
Samson P, Casey KR, Knepler J, Panos RJ. Clinical characteristics, comorbidities, and response to treatment of veterans with obstructive sleep apnea, Cincinnati Veterans Affairs Medical Center, 2005-2007. Prev Chronic Dis. 2012;9:E46. doi:10.5888/pcd9.110117

Frequently asked questions

Does sleep apnea increase the risk of macular degeneration?

Yes, but the association is mild to moderate. After pooling 8 observational studies (~3.5 million people), Yaldo et al. 2026 found an adjusted odds ratio of 1.44 (moderate certainty, I²=0%) and a hazard ratio of 1.66 (low certainty); in other words, people with OSA are roughly 1.4 to 1.7 times more likely to develop AMD than those without OSA.

Why does sleep apnea damage the macula?

The key is intermittent hypoxia. In OSA the upper airway repeatedly collapses during sleep, so blood oxygen drops and recovers dozens to hundreds of times a night. This cycle drives four mechanisms: oxidative stress that injures the retinal pigment epithelium, abnormal complement activation, sympathetic surges that compromise choroidal perfusion, and endothelial dysfunction. The cost of nightly desaturation is not just daytime fatigue but chronic macular inflammation.

Will treating sleep apnea (e.g. with CPAP) lower my AMD risk?

Currently no study has shown that treating OSA reduces AMD risk — this is the field's biggest unknown. Almost none of the 8 pooled studies recorded whether participants used CPAP, lost weight, or had other treatment, so treated versus untreated OSA cannot be compared. Until more evidence emerges, the established reasons to treat OSA remain cardiovascular protection, daytime alertness, and crash prevention, not the eyes.

I have a family history of AMD — should I get a sleep study?

The US Preventive Services Task Force (2022) concluded that the evidence is insufficient to assess routine OSA screening in asymptomatic adults. However, you can self-check with the validated STOP-Bang questionnaire (8 items): loud snoring, daytime tiredness or sleepiness, observed stopped breathing during sleep, high blood pressure (or being treated for it), BMI over 35, age over 50, neck circumference over 40 cm, and male sex. Three or more indicates a higher risk of OSA; consider a pulmonary or ENT sleep clinic for evaluation — for overall health, not specifically AMD prevention.

I have AMD and I snore — should my eye follow-up be more frequent?

Current evidence does not support intensifying AMD follow-up simply because of OSA. Follow-up cadence is still guided by AMD stage: per the 2024 AAO Preferred Practice Pattern, about every 6 to 24 months for early disease without symptoms and every 6 to 18 months for intermediate disease (or advanced AMD in one eye), with prompt examination for new symptoms, and individualised for late disease. A reasonable step is for ophthalmology to ask about snoring and daytime sleepiness, and to refer to pulmonary or sleep medicine if symptoms are suspicious.

Does this study mean sleep apnea «causes» macular degeneration?

No. All studies in this meta-analysis are observational, so they can show association but cannot prove causation. Most did not track CPAP treatment status, and confounders such as diabetes and obesity were not uniformly adjusted; the authors themselves frame the findings as hypothesis-generating. Confirming causality will require large prospective cohorts, randomised controlled trials, and more direct mechanistic studies.

Does Sleep Apnea Raise the Risk of Age-Related Macular Degeneration?