Association of Visual Health With Depressive Symptoms and Brain Imaging Phenotypes Among Middle-Aged and Older Adults

Abstract

Importance: Vision loss and depression are common conditions with major health implications. However, mechanisms of the association of visual health (across the full acuity spectrum) with depression remain unclear.

Objective: To characterize the association between visual health and depression and investigate the association between depression and brain microstructure and macrostructure in subgroups divided by visual acuity.

Design, setting, and participants: In the UK Biobank Study cohort, 114 583 volunteers were included at baseline from March to June 2006 to July 2010. Habitual distance visual acuity was examined using the logarithm of the minimum angle of resolution (LogMAR) characters. Depression was identified based on Patient Health Questionnaire (PHQ) or through an interview-based psychiatric diagnosis. Subgroup participants completed multimodal magnetic resonance imaging (MRI) of the brain and PHQ evaluation during the imaging visit after 2014. Data were analyzed from May 5 to August 9, 2022.

Main outcomes and measures: Depression, depressive symptoms, and imaging-derived phenotypes from T1-weighted and diffusion MRI.

Results: Of the 114 583 participants from the UK Biobank Study, 62 401 (54.5%) were women, and the mean (SD) age was 56.8 (8.1) years (range, 39-72 years). A 1-line worse visual acuity (0.1 LogMAR increase) was associated with 5% higher odds of depression (odds ratio, 1.05 [95% CI, 1.04-1.07]) after adjustment for age, sex, race and ethnicity, Townsend index, educational qualifications, smoking, alcohol consumption, obesity, physical activity, history of hypertension, diabetes, hyperlipidemia, and family history of depression. Of the 7844 participants eligible for MRI analysis, there were linear associations between PHQ score and the left volume of gray matter in supracalcarine cortex (coefficient, 7.61 [95% CI, 3.90-11.31]) and mean isotropic volume fraction (ISOVF) in the right fornix (cres) and/or stria terminalis (coefficient, 0.003 [95% CI, 0.001-0.004]) after correction for multiple comparison. In addition, their association could be moderated by visual acuity, whereby increased PHQ score was associated with higher ISOVF levels only among those with poorer visual acuity (P = .02 for interaction).

Conclusions and relevance: This study suggests an association between visual health and depression and that the diffusion characteristic of ISOVF in the fornix (cres) and/or stria terminalis is associated with depressive symptoms in participants with poorer visual acuity.

Figure 1.. Selection of Eligible Participants From the UK Biobank Cohort

Abbreviations: MRI, magnetic resonance imaging; PHQ-2, 2-item Patient Health Questionnaire; PHQ-9, 9-item PHQ.

Figure 2.. Association Between 2-Item Patient Health Questionnaire (PHQ-2) Score and Magnetic Resonance Imaging (MRI)–Derived Indices of Interest

A and B, Restricted cubic spline curves show adjusted odds ratios and 95% CIs (gray shading) for greater gray matter volume (GMV) in the left supracalcarine cortex (A) and greater isotropic volume fraction (ISOVF) in the right fornix (cres) and/or stria terminalis (B) associated with PHQ-2 score. C, For GMV in the left supracalcarine cortex, among those in the highest and lowest quartile of visual acuity, a greater PHQ-2 score was associated with higher volume. D, For mean ISOVF in the right fornix (cres) and/or stria terminalis, greater PHQ-2 scores were associated with higher ISOVF levels in those with poorer visual acuity. LogMAR indicates logarithm of the minimum angle of resolution. ^aP for linear trend < .001; P for nonlinear trend = .20. ^bP for linear trend < .001; P for nonlinear trend = .23.