**Background:** Diabetic macular oedema (DMO) is a vision-threatening complication of diabetic retinopathy. While structural changes are routinely assessed, functional changes beyond foveal vision are not commonly examined, despite evidence that retinal neurodegeneration may precede vascular changes. This study aimed to compare longitudinal changes in macular structure (OCT) with objective (OFA) and subjective (Matrix) perimetry in Type 2 diabetes (T2D) patients with mild DMO and good vision, hypothesizing that peripheral functional changes would be associated with DMO.
**Methods:** Sixteen T2D patients (10 with baseline DMO, 6 without) were followed for up to 27 months, providing 94 data sets. Inclusion criteria: LogMAR visual acuity <0.3, mild DMO defined as retinal thickness ≥320 μm on spectral-domain OCT (Spectralis) within the macula but not at the centre. Exclusion criteria: prior ophthalmic treatment for DR, other vitreoretinal or optic nerve disease, cataract surgery. Examinations at each visit included: logMAR BCVA, IOP, pachymetry, Humphrey Matrix 10-2 perimetry, and OFA (FDA-cleared prototype) using two multifocal pupillographic objective perimetry (mfPOP) variants: P30° (central ±15°) and P60° (central ±30°), each presenting 44 stimuli/eye. OCT posterior-pole thickness was measured on an 8×8 grid. OCT, Matrix, and P30° OFA data were mapped to a common 44-region grid (10-2 format) for region-by-region comparison. Longitudinal changes were assessed using linear regression; significant change was defined as slope p<0.05.
**Key Results:** Baseline characteristics: DMO group (n=10) had mean diabetes duration 11.81±6.6 years, HbA1c 7.6±1.35%, systolic BP 135.3±13.5 mmHg; non-DMO group (n=6) had duration 6.75±4.9 years, HbA1c 6.75±0.8%, systolic BP 132.1±10.4 mmHg (differences not significant). Mean central retinal thickness at baseline: DMO eyes 237±25 μm, non-DMO eyes 250.7±24.4 μm. At study end, DMO eyes reduced to 234.2±26.7 μm, while non-DMO eyes increased to 255.7±20.6 μm (p=0.023). Visual acuity did not change significantly.
In eyes with DMO at baseline, decreasing retinal thickness was associated with increasing OFA sensitivity and decreasing response delays (all p<0.021). For example, subject mf87157 showed reduced retinal thickness and improved P30° and P60° sensitivities over 571 days. Matrix perimetry showed significant change in only 5 regions (mean p=0.017), mostly within central 8°, while OFA P30° sensitivities showed 58 significant regions (mean p=0.020), P30° delays 55 regions (mean p=0.019), and OCT thickness 162 regions (mean p=0.016). For the wide-field P60°, sensitivities showed 53 significant regions (mean p=0.021) and delays 39 regions (mean p=0.020). In one non-DMO subject (mf87171), increasing retinal thickness was associated with decreasing OFA sensitivity and shorter delays. The centroids of oedematous regions were often off-axis, partially surrounding the fovea without involving it.
**Clinical Implications:** This study demonstrates that objective perimetry (OFA) can detect regional functional changes (sensitivity and delays) that correlate with structural changes in mild DMO, even when visual acuity is good. OFA identified more significant functional changes than Matrix perimetry, particularly in peripheral regions beyond 4° eccentricity. The ability to monitor both progression and recovery of function suggests OFA could serve as a sensitive biomarker for early DMO, potentially guiding treatment decisions (e.g., anti-VEGF, fenofibrate) before irreversible structural damage occurs. The rapid, non-invasive nature of OFA (under 7 minutes) makes it clinically feasible. Limitations include small sample size (16 patients) and single-centre design; larger studies are needed to confirm these findings and establish normative thresholds for clinical use.