The Effect of Partial Sleep Deprivation on Retinal Microvasculature in Myopia With Different Axial Lengths

Abstract

Purpose: To investigate the effects of partial sleep deprivation (PSD) on the retinal microvasculature in individuals with myopia, using optical coherence tomography angiography (OCTA).

Methods: In total, 87 right eyes were categorized by axial length (AL) into three groups: A (AL < 24 mm, n = 20), B (24 ≤ AL < 26 mm, n = 41), and C (AL ≥ 26 mm, n = 26). Participants underwent macular (6 mm × 6 mm) and optic disc (4.5 mm × 4.5 mm) OCTA scans. Vessel density (VD) parameters—including superficial capillary plexus VD (SCP-VD), deep capillary plexus VD (DCP-VD), radial peripapillary capillary VD (RPC-VD), inside disc RPC VD (iVD), and peripapillary RPC VD (ppVD)—were compared at baseline, before and after PSD, and following 3 days of regular sleep.

Results: Compared with group A, groups B and C had lower baseline DCP-VD in all subregions (P < 0.05), except for the perifovea–inferior area, where only groups C and A showed significant differences (P = 0.001). After PSD, group C showed a decrease in SCP-VD in certain subregions of the parafovea (P = 0.013 and P = 0.022 for parafovea–temporal and parafovea–inferior, respectively), along with an increase in ppVD (P = 0.012). All VD parameters recovered after 3 days of regular sleep (P < 0.05).

Conclusions: The retinal microvasculature of myopic eyes with an AL of ≥26 mm show greater susceptibility to PSD than those with an AL of <26 mm. However, short-term PSD effects can be restored by setting a regular sleep schedule.

Myopia is an ophthalmic disease in which the visual images come to a focus in front of the retina of the eye, resulting in defective vision of distant objects, and it has become a serious public health issue.¹ Its prevalence among adolescents in East Asia has reached 80% to 90%.^2,3 High myopia (HM) is often accompanied by an abnormally elongated ocular axis, which causes retinal vascular damage and reduces vessel density (VD)^4,5 and subsequently leads to a variety of serious complications.⁶ Therefore, the early monitoring of retinal microvasculature alterations in patients with myopia is of great significance. 

Research has revealed that partial sleep deprivation (PSD), defined as sleeping <6 hours per night,^7,8 is closely linked to various eye diseases. Studies have shown that the prevalence of myopia increases with a reduction in sleep duration, and PSD is also considered an influencing factor for high myopia.^9,10 Additionally, decreased VD in the macula and optic disc is correlated with axial length (AL) elongation in myopia.^4,5 Therefore, PSD may regulate the onset and progression of myopia and its complications by affecting the retinal microvasculature. Several neovascular retinal diseases have been linked to PSD.^11,12 A previous study reported that short sleep duration could be associated with age-related macular degeneration, potentially through mechanisms such as the induction of neuroinflammation, activation of apoptosis in cells, disruption of energy metabolism, and dysfunction of the immune system.¹¹ Furthermore, PSD may be involved in the pathogenesis of diabetic retinopathy, a microvascular disease.¹² Therefore, we designed an experiment to study the tolerance of myopic individuals with different ALs to acute PSD. 

Although there have been numerous investigations on the effects of sleep disorders on eye diseases, there has been a relative lack of research on their effects on retinal VD in HM patients over a short period. Therefore, in this study, we used optical coherence tomography angiography (OCTA) to investigate the effects of PSD on VD in the macular and optic disc regions in a myopic population with different ALs. 

Methods

This prospective observational study was approved by the Ethical Review Committee of Renmin Hospital of Wuhan University and was conducted in accordance with the tenets of the Declaration of Helsinki. Each participant provided informed consent prior to recruitment. The study included individuals with myopia and a spherical equivalent refraction (SER) of −0.50 diopters (D) or less. Participants were categorized into three groups based on the AL of the eye: group A (AL < 24 mm), group B (24 ≤ AL < 26 mm), and group C (AL ≥ 26 mm). 

Inclusion criteria were as follows: (1) 20 to 30 years of age; (2) best-corrected visual acuity (BCVA) ≥ 0.8; (3) SER of −0.50 D or less; and (4) no history of major ophthalmic diseases, surgery, or trauma. The exclusion criteria were as follows: (1) absence of HM-related retinal damage; (2) BCVA < 0.8; (3) history of eye surgery or trauma; (4) use of orthokeratology lenses, contact lenses, or scleral lenses within the past 1 month; (5) presence of other eye diseases affecting vision, such as corneal diseases, congenital cataracts, and retinal diseases; (6) presence of systemic diseases, such as diabetes mellitus, migraines, high blood pressure (BP), or other vascular diseases, or the use of vasoactive medications; and (7) inability to cooperate with the designated study examinations. 

Ocular Examination

All participants underwent a complete ophthalmologic assessment, including the measurement of intraocular pressure (IOP), corneal thickness, corneal curvature, and OCTA scans. A single evaluator (KX) consistently assessed all images and data. 

Optical Coherence Tomography Angiography

The participants underwent OCTA scans using AngioVue device prototype software (RTVue XR Avanti with AngioVue; Optovue, Fremont, CA, USA). This device has a scanning rate of 100,000 A-scans per second and a central wavelength of approximately 1050 nm. The OCTA tests covered a macular area of 6 mm × 6 mm and an optic disc area of 4.5 mm × 4.5 mm. Eye movement artifacts were reduced using an eye-tracking program built into the software. 

The VD was defined as the proportion of the total image area occupied by blood vessels to the total image area after binarization.¹³ In the 6 mm × 6 mm macular scan, the foveal region was defined as a round area around the fovea with a diameter of 1 mm, the parafoveal region as a ring between two circles with diameters of 1 mm and 3 mm, and the perifoveal region as a ring between two circles with diameters of 3 mm and 6 mm. Each parafoveal and perifoveal region had four fan-shaped subregions: temporal, superior, nasal, and inferior (Fig. 1a). The macular superficial capillary plexus (SCP) is the microvasculature extending from the internal limiting membrane (to the interface between the inner plexiform layer (IPL) and the inner nuclear layer (INL), whereas the deep capillary plexus (DCP) is the microvasculature extending from the IPL/INL junction to 25 µm below the outer plexiform layer. In the 4.5 mm × 4.5 mm optic disc scan, the inner disc area was a circular area centered on the optic nerve with a diameter of 2 mm, and the peripapillary area was an annular area between circles with diameters of 2 mm and 4 mm centered on the optic nerve area (Fig. 1b). 

Figure 1.

 

Division of macular and optic disc subregions.

Procedures

In our study, the participants underwent a five-night clinical trial. On the baseline night (Day 1), they slept from 23:00 to 07:00 (time in bed [TIB] = 8 hours). On Day 2, participants were subjected to partial sleep deprivation; they were not allowed to sleep from 23:00 to 02:00 (TIB = 5 hours) and were permitted to sleep from 02:00 to 07:00 (PSD). This was followed by three recovery nights with sleep from 23:00 to 07:00 (Days 3 to 5; TIB = 8 hours) (Fig. 2). The participants’ sleep behavior was monitored using a wrist-worn Actiwatch. The OCTA examination was conducted by three well-trained operators (KX, YRQ, YFZ), operating on the mornings after Day 1, Day 2, and Day 5. All participants were required to take a 5-minute break before the OCTA examination, which was performed from 7:00 to 8:00 AM. The VD values of the macular SCP, DCP, and optic disc were recorded in detail according to the subregions described above, and the results were analyzed using SPSS Statistics 26.0 for Windows (IBM, Chicago, IL, USA). 

Figure 2.

 

Experimental procedure.

Statistics

SPSS Statistics 26.0 was used to analyze the data. Three normally distributed OCTA data groups were compared using a one-way analysis of variance (ANOVA), and the least significant difference was used for pairwise comparisons. All data are expressed as mean ± SD. Statistical significance was set at P < 0.05. 

Results

A total of 87 right eyes were included in this study. The eyes were classified into three separate groups based on the AL: <24 mm, 24 ≤ AL < 26 mm, and ≥26 mm. Parameters such as age, IOP, and average sleep duration were comparable and may potentially influence the interpretation of the quantitative OCTA findings. The relevant demographic details are listed in Table 1 and the Supplementary Table. 

Table 1.

 

Demographic and Clinical Characteristics of Groups A, B, and C

 

We examined baseline VD differences in the SCP, DCP, and radial peripapillary capillary (RPC) among groups A, B, and C (Table 2, Fig. 3) and performed a correlation analysis (Table 3). The VD parameters in the fovea of both layers were significantly higher in group C than in group A (P = 0.015 and P = 0.024 for the superficial and deep layers, respectively) (Table 2). Superficial capillary plexus VD (SCP-VD) was statistically significant in only a few areas (P < 0.05) (Table 2), whereas deep capillary plexus VD (DCP-VD) changed significantly with an increase in AL (P < 0.05) (Table 2). Compared with group A, groups B and C had lower baseline DCP-VD in all subregions of the parafovea and perifovea (P < 0.05) (Table 2), except for the perifovea–inferior area, where only group C and group A showed significant differences (P = 0.001) (Table 2). For radial peripapillary capillary VD (RPC-VD), inside disc RPC VD (iVD) in group C increased (P = 0.027) (Table 2), whereas the peripapillary RPC VD (ppVD) decreased (P = 0.010) (Table 2). 

Table 2.

 

Comparison of VD at Baseline Nights in the Three Groups

 

Figure 3.

 

Comparison of VD at baseline nights in the three groups. *P < 0.05, **P < 0.01.

Table 3.

 

Correlation Between AL and VD at Baseline Nights

 

Correspondingly, in the correlation analysis, AL was positively correlated with VD in the fovea of both the SCP and DCP groups (r = 0.301, P = 0.005 and r = 0.243, P = 0.023, respectively). Furthermore, AL correlated with VD only in parafovea–temporal and perifovea–nasal regions of the SCP (r = −0.278, P = 0.009 and r = 0.297, P = 0.005, respectively) but in all eight subregions of the parafovea and perifovea of the DCP (r = −0.315, P = 0.003; r = −0.304, P = 0.004; r = −0.257, P = 0.016; r = −0.397, P < 0.001; r = −0.309, P = 0.004; r = −0.275, P = 0.010; r = −0.372, P < 0.001; and r = −0.333, P = 0.002, respectively). Similarly, a negative correlation was observed between AL and ppVD (r = −0.245, P = 0.022) (Table 3). 

After PSD, the SCP-VD, DCP-VD, and RPC-VD in groups A and B remained relatively stable (P > 0.05) (Table 4). Additionally, in group C, the SCP-VD in certain subregions of the parafovea decreased (P = 0.013 and P = 0.022 for parafovea–temporal and parafovea–inferior, respectively), whereas the ppVD increased after PSD (P = 0.012). All of the statistical measures recovered after 3 days of normal work and rest (P < 0.05). Additionally, for the SCP-VDs of the parafovea–nasal and perifovea–inferior groups, although there were only downward trends observed after Day 1 (P = 0.094 and P = 0.142, respectively), they significantly increased after recovery (P = 0.038 and P = 0.043, respectively). In group A, the SCP-VD tended to be higher than that at baseline after recovery; however, this increase was not statistically significant, except in the perifovea–superior region (P = 0.031). The results of the comparisons and statistical analyses are presented in Table 4. 

Table 4.

 

Comparison of Morning VD After Day 1, Day 2, and Day 5 for Participants in the Three Groups

 

Discussion

Circadian rhythms are temporal programs present in all living organisms and are regulated by the biological clock system.¹⁴ The mammalian retina has an independent internal clock,¹⁵ and disruptions to this clock can affect the physiological homeostasis of the eye.^16,17 The retinal vasculature is affected by increased AL in patients with HM, and additional PSD may further affect the stability of ocular blood flow, thereby contributing to disease onset and progression. 

Retinal microvascular changes are widely recognized risk factors for HM-related complications. This study found that, with an increase in AL, DCP-VD in the parafovea and perifovea decreased, which is consistent with the findings of most previous studies.^4,5,18–21 A possible reason is that axial elongation mechanically stretches the retina and choroid, causing blood vessels to straighten and narrow, directly leading to a decrease in VD.²² Another explanation is that axial elongation causes the thinning of both the choroid and retina; choroidal thinning reduces the oxygen supply to the retina, whereas retinal thinning decreases the demand for oxygen in the photoreceptor segments and deep capillary circulation, resulting in a decrease in DCP-VD.⁴ In our study, alterations in RPC-VD were observed—with ppVD decreasing and iVD increasing—as AL increased, consistent with previous study findings.^5,23 Jonas et al.^24,25 found that the distance between the optic disc center and fovea increased with axial elongation. This may be related to the mechanical stretching forces on the retinal and disc regions caused by the elongation of the posterior sclera, which may lead to ppVD reduction.²⁶ In our study, we found an increase in iVD that is consistent with the findings of Wang et al.,²³ where iVD increased because of the automatic adjustment mechanism of retinal vascular regulation. Their study indicated that decreased peripapillary vascular perfusion may affect regional oxygen demand and increase blood flow inside the disc to compensate for the decreased blood flow around the optic disc and ensure the normal function of the retinal tissue.²³ However, further research is required to verify these findings. Changes in central foveal VD in patients with HM have also been reported; however, the correlation between AL and central foveal VD of the macula remains controversial. Our study found that, with an increase in AL, the foveal VD increased, which is consistent with the findings of a study on adolescent myopia.²⁷ This may be related to the reduction in the FAZ area, which was speculated to be a compensatory mechanism in patients with nonpathological HM, where retinal thickening in the foveal center helps maintain visual function.²⁷ In contrast, Cheng et al.⁴ reported that decreased foveal VD was related to an increased FAZ area in HM. However, their study included participants 16 to 42 years old, which differed from ours (20–29 years old). A study of individuals under 20 years of age showed decreased FAZ, which may lead to increased foveal VD, indicating that age differences in the samples may affect the results.²⁷ Both our study sample and the previously mentioned study sample²⁷ were relatively young and may have had greater abilities for compensatory foveal center thickening. 

After PSD, group C showed a decrease in parafoveal SCP-VD; however, groups A and B showed no significant changes. This may be caused by the varying tolerances to PSD among different ALs in myopia. Several studies^27–29 have shown that, as the AL increases, the vascular resistance index increases, leading to a decrease in pulsatile ocular blood flow, which presents as a VD reduction on OCTA. In addition, blood pressure elevated by short sleep duration can also affect ocular blood flow.^30–32 Elevated BP may increase IOP by enhanced aqueous humor production through increased ciliary blood flow and capillary pressure, as well as reduced aqueous humor outflow caused by higher episcleral venous pressure,^33,34 which causes direct compression of the vasculature and a reduction in VD.³⁵ A study of healthy participants demonstrated that the Valsalva maneuver, which induces an increase in systemic BP, results in decreased VD in the parafovea, supporting our hypothesis.³² Therefore, we propose that a possible reason for our findings is that the vascular resistance index in the elongated AL is already high, resulting in a lower VD under the influence of a further increase in BP following PSD. We also observed a decreasing trend in DCP-VD, although the difference was not statistically significant. This phenomenon may be attributed to a greater autoregulatory capacity in the DCP than in the SCP when blood pressure is elevated.³⁶ 

Interestingly, we found that ppVD increased significantly after PSD in group C, which contrasted with the change in macular VD. This difference may be attributed to individual variations in blood structure and supply. The surface nerve fiber layer in the peripapillary region is supplied by both the retinal arterioles and cilioretinal artery branches from the posterior ciliary arteries (PCAs),^37,38 which ensures relatively ample blood flow. Conversely, the foveal region lacks PCA distribution and is solely supplied by the retinal artery system, making it most susceptible to ischemia in cases of vascular insufficiency.³⁹ In addition, Jeppesen et al.⁴⁰ observed a negative correlation between retinal arteriole dilation following an increase in BP and the starting vessel diameter. This may explain regional differences in VD changes, as the parafoveal region contains small vessels or capillaries, whereas the peripapillary region contains a rich capillary network along with four primary intraretinal arteries and secondary arterioles.³² Additionally, PSD has been associated with reduced angiotensin-I-converting enzyme activity and mRNA levels.⁴¹ Therefore, we speculate that this may contribute to the vasodilation of peripapillary PCAs and retinal arterioles. Furthermore, we observed that there was no significant alteration in iVD in group C after PSD, possibly because of the intricate structure of the large vessels inside the disc region, making intergroup differences less detectable.⁴² 

Our study has several limitations. First, group C had a small sample size, and future studies may require larger groups for more detailed assessments. Second, other factors affecting high myopia, such as refractive error and corneal curvature, were not analyzed further. Third, fixation changes in HM can cause transverse motion artifacts in OCTA images.⁴ The implementation of motion correction technology within the RTVue system notably enhanced image quality.⁴³ However, completely preventing artifacts is not feasible and may affect the assessment of retinal blood vessels in myopic eyes. Advancements in technology are expected to address projection artifacts and enable further research on long-term rhythm abnormalities. 

In conclusion, this study showed that myopia alters the retinal microvasculature. Individuals with an elongated AL are more susceptible to retinal microvasculature changes and fluctuations after PSD. This finding can serve as a warning, increase awareness of retinal damage caused by an unhealthy lifestyle, help avoid more serious complications caused by PSD, and establish a foundation for the systematic management and prevention of myopia. 

Acknowledgments

The authors gratefully acknowledge the participation of our study subjects, who generously contributed their time and effort to this research. 

Supported by the Undergraduate Training Program for Innovation and Entrepreneurship at Wuhan University (project number S202410486315), the Teaching Research Project of Wuhan University School of Medicine (project number 2024YB26), and the China Primary Health Care Foundation. 

Disclosure: K. Xu, None; Y. Xu, None; Y. Qin, None; Y. Zhang, None; H. Zheng, None; C. Chen, None; Y. Su, None