Abstract
Purpose:
To describe visual acuity (VA) and depressive symptom (DS) trajectories over 8 years among older adults and to estimate relationships between these trajectories.
Methods:
A population-based sample of 2520 white and black individuals, aged 65-84 years in 1993-1995 were re-assessed 2, 6, and 8 years later. Presenting and best-corrected ETDRS VA were obtained with refraction performed in participants with presenting VA of 20/30 or worse. DS were assessed using the Severe Depression subscale of General Health Questionnaire (GHQ)-28. Latent Growth Curve models were used to estimate VA and DS trajectories and to obtain age-adjusted associations between trajectories.
Results:
Both slopes and intercepts of the best-corrected logMAR VA trajectory (slope=0.026, intercept=0.013, p<0.001) and presenting VA trajectory (slope=0.022 and intercept=0.040, p<0.001) showed starting values and increases that differ from zero. The intercept of the DS trajectory was significant (intercept=1.180, p< 0.001), but the slope was not (slope=-0.001, p=0.762), indicating no change in DS averaged over time. Best-corrected VA at baseline correlated with baseline DS (r=0.14, p<0.001) indicating a cross sectional relationship between VA and DS. DS at baseline were associated with change of best-corrected VA over time (r=0.17, p=0.01). Baseline best-corrected VA was not associated with DS change (r=0.017, p=0.8). VA change was not associated with DS change (r=-0.03, p=0.7). Similar relationships were found between presenting VA and DS. Post-hoc analyses were performed in the subset of participants whose vision was 20/40 or worse at baseline, and in a subgroup that had better than 20/40 vision at baseline. In both cases after adjustment for age there was no significant change in DS over time.
Conclusions:
Consistent with previous research there is a significant cross-sectional association between DS and VA. Those with higher depression scores at baseline were more likely to experience worsening VA over time. However, change in VA was not associated with change in DS that may be attributed to the limited interpersonal variation in DS change over time. Future studies should include more frequent assessment of DS (i.e., monthly) using measures which also assess more mild/moderate symptomatology which may not be captured by the (GHQ)-28 Depression subscale.
Keywords: 463 clinical (human) or epidemiologic studies: prevalence/incidence •
459 clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology •
754 visual acuity