Contrast Sensitivity, Visual Field, Color Vision, Motion Perception, and Cognitive Impairment: A Systematic Review
Ying Xu,Htein Linn Aung,Negin Hesam-Shariati,Lisa Keay,Xiaodong Sun,Jack Phu,Vanessa Honson,Phillip J Tully,Andrew Booth,Ebony Lewis,Craig S Anderson,Kaarin J Anstey,Ruth Peters
DOI: https://doi.org/10.1016/j.jamda.2024.105098
Abstract:Objectives: To examine relationships between visual function (ie, contrast sensitivity, visual field, color vision, and motion perception) and cognitive impairment, including any definition of "cognitive impairment," mild cognitive impairment, or dementia. Design: Systematic review and meta-analyses. Setting and participants: Any settings; participants with (cases) or without (controls) cognitive impairment. Methods: We searched 4 databases (to January 2024) and included published studies that compared visual function between cases and controls. Standardized mean differences (SMD) with 95% CIs were calculated where data were available. Data were sufficient for meta-analyses when cases were people with dementia. The Joanna Briggs Institute checklists were used for quality assessment. Results: Fifty-one studies/69 reports were included. Cross-sectional evidence shows that people with dementia had worse contrast sensitivity function and color vision than controls: measured by contrast sensitivity (log units) on letter charts, SMD -1.22 (95% CI -1.98, -0.47), or at varied spatial frequencies, -0.92 (-1.28, -0.57); and by pseudoisochromatic plates, -1.04 (-1.59, -0.49); color arrangement, -1.30 (-2.31, -0.29); or matching tests, -0.51 (-0.78, -0.24). They also performed more poorly on tests of motion perception, -1.20 (-1.73, -0.67), and visual field: mean deviation, -0.87 (-1.29, -0.46), and pattern standard deviation, -0.69 (-1.24, -0.15). Results were similar when cases were limited to participants with clinically diagnosed Alzheimer disease. Sources of bias included lack of clarity on study populations or settings and definitions of cognitive impairment. The 2 included longitudinal studies with follow-ups of approximately 10 years were of good quality but reported inconsistent results. Conclusions and implications: In the lack of longitudinal data, cross-sectional studies indicate that individuals with cognitive impairment have poorer visual function than those with normal cognition. Additional longitudinal data are needed to understand whether poor visual function precedes cognitive impairment and the most relevant aspects of visual function, dementia pathologies, and domains of cognition.