Paperonium

Infectious disease genetics · PLoS Pathog · 2020

“DNA variant raises leprosy odds 1.47 times.” The two studies it combines found 1.85 and 2.06.

Paper’s combined figurebelow both studies; tiny rangeStudy 1: patients andtheir parentsStudy 2: patientsvs. non-patients2.06×; no upper limitBoth studies, combined by us1.97×, between them1.47×1.46 to 1.481.85×1.392.441.71.652.361×1.5×2×3×no difference in odds
Paper’scombined figurebelow bothstudies; tiny rangeStudy 1: patientsand their parentsStudy 2: patientsvs. non-patients2.06×; noupper limitBoth studies,combined by us1.97×,between them1.47×1.46 to 1.481.85×1.392.441.71.652.361×1.5×2×3×no difference in odds

Times the odds of leprosy, per copy of the variant, with each range

The paper says

The summary, main table and text say each copy of a DNA variant (people carry 0, 1 or 2) multiplies the odds of leprosy by 1.47, with a range (values the data can’t rule out) of just 1.46 to 1.48. That figure combines two studies.

Its own data show

The same table lists both studies. Patients and their parents: 1.85 times (range 1.39 to 2.44). Patients versus people without leprosy: 2.06 times, with only a floor of 1.7, because that study tested only for higher odds. Combining them ourselves, weighting each by how precise it is, gives 1.97 times, range 1.65 to 2.36.

Combining two studies is a kind of average, so the result must land between their central figures, 1.85 and 2.06. A range this narrow needs far more people: 1.46 to 1.48 would take hundreds of thousands of patients; there were 1,803. The variant really is tied to leprosy, but probably about doubles the odds, with far more uncertainty.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Family-based genome-wide association study of leprosy in Vietnam · PLoS Pathog (2020) · doi:10.1371/journal.ppat.1008565