Greater editorial staff representation correlated moderately with more published articles reporting research conducted in LMICs (Spearman ρ = 0.51; 95% CI, 0.25-0.70; P < .001).

It is funny how different people interpret correlations. I would not call a correlation of 0.5 for just one predictor “moderate”. That may be low for lab experiments or artificial objects, but it is pretty high for real life data.