Women’s health · JAMA Netw Open · 2025
“2.92 times the odds of diabetes, range 1.06 to 2.82.” But 2.92 lies outside that range.
Times the odds of type 2 diabetes. Each dot must sit inside its bar.
The paper says
Of 340 women with polycystic ovary syndrome, 189 had ovaries with many small fluid sacs. After evening out age, weight and blood tests, their odds of type 2 diabetes (chance of having it ÷ chance of not) were 2.92 times the others’: range 1.06 to 2.82.
Its own data show
Here a result sits at its range’s middle, found by multiplying, not adding: 1.06 × 1.63 = 1.73, and 1.73 × 1.63 = 2.82. If the range is right, the result is 1.73. If 2.92, 2.75 times the bottom, is right, the top is 2.75 × 2.92, about 8. Keeping 2.92 and 2.82 puts the bottom (3.02) above the top.
A range is the span of values the data fit, so it must contain its result. This is the paper’s main finding, printed in the summary, results and table, so whoever quotes it repeats the impossible pair. Whichever number is wrong, diabetes stays more common, but by how much: 1.7 times the odds, or anywhere from 1 to 8?