Detecting transdiagnostic retinal deviations in mental disorders through normative modeling

ObjectiveSeveral studies have found thinner retinal tissue in mental disorders compared to healthy controls. Because the retina is part of the human brain, this suggests that informative brain structure readouts can be obtained efficiently through retinal imaging. Instead of focusing on group-level case-control differences, we used normative modeling to estimate age-related decline of the human retina (and its expected variation) and compared it to the decline seen in schizophrenia (SZ), bipolar disorder (BD), and major depression (MDD). We hypothesized accelerated retinal decline in mental disorders compared to controls, with SZ being most affected, followed by BD, then MDD. MethodsUsing UK Biobank data, we estimated age-related retinal decline in healthy controls (HC, N = 56,545) for total macular thickness (including coronal subfields) and two sublayers (retinal nerve fiber layer; RNFL; and ganglion cell-inner plexiform layer; GC-IPL). We then compared the decline in SZ (N = 171), BD (N = 256), and MDD (N = 102) to the normative decline in HC. ResultsFor HC, the pattern of age-related decline for total macular thickness, RNFL, and GC-IPL was curve-like rather than linear and more pronounced in males compared to females. For mental disorders, the decline-pattern was generally faster, driven by SZ and disorder-specific macular subfields. There was also an enrichment of individuals with extremely low (infranormal) values. These results were confirmed in robustness checks that ruled out unspecific confounders. ConclusionThese findings suggest that mental disorders, particularly SZ, involve accelerated neurodegenerative decline that can be detected in the human retina.