In this study, we have demonstrated a reduction in retinal thickness within the central and parafoveal ETDRS subfields from 9 a.m. to 1 p.m., when the participants were in upright position. When the participants were positioned recumbent from 1 p.m. to 3 p.m., an increased retinal thickness in the central and parafoveal subfields was observed. We also found a positive correlation between the reduction in MOPP and parafoveal retinal thickness in upright position from 9 a.m. to 1 p.m. This implies that macular oedema associated with MH behaves similarly to oedema related to diabetes mellitus, central retinal vein occlusion, and uveitis [8,9,10,11]. In these conditions, the BRB is damaged, resulting in leakage of fluid from the capillary bed. Such leakage will be affected by the MOPP, which is reported to decrease from early morning until after noon [16]. However, in conditions with NVCM like MH, where the capillaries are not known to leak, the role of MOPP in the observed alterations in macular thickness remains unknown. The intraretinal fluid in MH-related oedema probably originates from the vitreous and the retinal metabolism, independent of MOPP [17, 18]. On fluorescein angiography, leakage from retinal capillaries is normally not detected, but subtle leakage may evade detection due to active fluid absorption by the Müller cells and the RPE [5, 19]. In a previous study of 43 MH patients, a hyperfluorescent fluid cuff was seen in 47%, while there was absence of hyperfluorescence in 37% of the patients. A late fluorescein staining pattern, typical for CMO, was not found in any of the cases, but microvascular changes resembling microaneurysms or telangiectatic capillaries in the cuff area were noted in 12% of the patients [20]. In addition, a tractional trauma on the Müller cells could stimulate pro inflammatory factors and lead to gliosis and vascular leakage [12, 21, 22]. We propose that our finding of MOPP-related changes in MH-related oedema implies some degree of capillary leakage. The same pattern has also been observed in one other NVCM condition, namely X-linked retinoschisis [23, 24]. These studies also documented a daytime reduction in CRT.

The accumulation of fluid in a tissue depends not only on leakage or influx but also on the capacity of absorption [18, 25]. In MH cases, the absorption can be impaired in at least two ways. First, the retina surrounding the hole is detached, which means that the underlying RPE pump is overwhelmed by vitreous fluid. This induces a reduced capability of the RPE to clear fluid from the foveal tissue [26]. Second, the foveal Müller cells which dehydrate the inner retina by transporting water into the capillaries are traumatised. Impairment of the Müller cell ion pumps will lead to decreased fluid absorption, resulting in inner retinal cyst formation [27]. The increased hydrostatic pressure resulting from elevated MOPP may also impair the ability of the RPE and the Müller cells to transport fluid into the choroidal and retinal capillaries and lead to an increase in macular oedema [28, 29]. In the healthy fellow eyes, an opposite posture effect on retinal thickness could be observed. The thickness changes were minimal, but in line with earlier reports on circadian CRT patterns [30].

Compared with real-life MH surgery settings, our study has certain restrictions, as more detailed analyses involving various body positions, particularly face-down positioning, would be relevant. However, our primary focus was on comparing the major positions, upright versus recumbent, while also minimising the number of variables. Additionally, the potential impact of an intraocular gas bubble on our measurements remains unknown. Gravitational fluid shift has been proposed to have an impact on the CMO in diabetes and in MH eyes with gas tamponade [31, 32]. One would therefore expect a relatively higher increase in retinal thickness in the temporal ETDRS subfields because the participants were positioned recumbent on the side with their MH eye down. As there was no systematic variation in retinal thickness across the four ETDRS quadrants, we assume that gravitational forces have little or no effect on MH-related oedema under the studied conditions. However, since rapid displacements of oedema may occur, and the OCT was performed 10–20 s after transitioning from a recumbent to a sitting position, a gravitational impact on the MH-related oedema could not be completely ruled out.

Compared to the changes in retinal thickness, the minimum and basal MH diameters were less directly influenced by the body position. Nonetheless, the basal MH diameter decreased when upright whereas the minimum MH diameter increased when recumbent, indicating that upright positioning promotes MH closure.

It is debated whether the macular cystic cavities represent pseudocysts from tractional forces or oedematous cysts due to fluid influx. A combined picture, at least in cases with VMT is likely [26]. In established MH cases without VMT, fluctuations of cyst size have been observed, which is a behaviour typical for oedematous cysts. Vitreoretinal interface disorders, such as ERM and VMT, have the potential to influence how macular oedema responds to different body positions. We found that the presence of VMT significantly influenced this response. The reduction in macular oedema when upright was less prominent in eyes with VMT compared to those without VMT. It seems logical that anteroposterior traction exerted by the posterior hyaloid counteracts the reduction of macular oedema, but one would also expect that when recumbent, VMT should facilitate its aggravation. However, if the VMT also contributes to tangential traction and imparts some rigidity to the tissue, it may explain our findings. We found no effect of the presence or absence of ERM on changes in retinal thickness.

To the best of our knowledge, this is the first study assessing the impact of positioning on MH-related oedema. The low number of participants is a limitation of the study. In addition, the OCT scans at 3 p.m. were obtained from participants who transitioned to a sitting position seconds after being recumbent. This could potentially result in an undervaluation of the true increase in retinal thickness due to the rapid fluctuations in macular oedema [8]. Another possible bias is the shorter duration of recumbency compared to upright positioning, which could lead to an underestimation of the effect of recumbency. However, the assumed rapid changes in the amount of oedema makes this unlikely. During periods of upright positioning, the participants were allowed to sit, stand, or walk around. Consequently, their activity level was somewhat higher than during recumbency, which might have affected the results.

In summary, we have demonstrated that macular oedema related to MH decreases during the day when upright and increases, in parallel with an increase in mean minimum MH diameter, when recumbent. The reduction of macular oedema is correlated with a reduction in MOPP. This suggests that MH-related oedema is caused not only by influx from the vitreous and the retinal metabolism but also, to some extent, by vascular leakage and the ability of Müller cells and RPE to transport fluid back into the bloodstream. The results imply that maintaining a standing or sitting position, is beneficial in the early postoperative period of MH surgery. Further research is required to confirm these observations and to explore their implications for optimal positioning after MH surgery.