Back

Dietary Factors and Predominant Eye Diseases in Sub-Saharan African Populations: A Comprehensive Systematic Review Protocol

Osei Duah Junior, I.; Akuffo, K. O.; Danso-Appiah, A.; Ampong, J.; Owiredu, D.

2025-02-21 ophthalmology
10.1101/2025.02.17.25322272 medRxiv
Show abstract

BackgroundEvidence linking diet and ocular diseases is growing, yet variations persist, with a paucity of data in sub-Saharan Africa. The proposed review will systematically synthesize evidence on dietary factors associated with predominant eye disorders (cataracts, refractive error, glaucoma, diabetic retinopathy, age-related macular degeneration, and dry eye disease) in the sub-Saharan African population. MethodsThe systematic review and meta-analysis will follow the gold standard for reporting systematic reviews and meta-analysis, thus Preferred Reporting Items for Reporting Systematic Reviews and Meta-analysis (PRISMA), and Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines. The study will identify all relevant studies published, unpublished, and pre-prints from PubMed, SciVerse Scopus, Web of Science, Google Scholar, Embase, Health Inter-Network Access to Research Initiative (HINARI), Cumulative Index to Nursing and Allied Health Literature (CINAHL), African Journal of Science, Biomed Central, and Cochrane Library and relevant institutional databases without date restrictions. We will check the reference lists of all studies identified by the methods and contact experts for additional relevant studies. Data will be analyzed with the Review Manager (RevMan) software. The random effects model will be used to pool study effect estimates and robustness assessed through sensitivity analyses. Heterogeneity across studies will be assessed using the I2-statistic. Publication bias will be investigated graphically with a funnel plot and statistically with Eggers regression intercept method or Beggs rank correlation tests. All estimates will be reported with their confidence intervals, and statistical analysis will be set at a significance level of p<0.05. The results of the review will be presented based on the PRISMA guidelines.

Matching journals

The top 1 journal accounts for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.