Paperonium

Multiple sclerosis · Proc Natl Acad Sci U S A · 2022

More MS patients have milk-reactive immune cells, the paper says. Its counts don’t rule out luck.

1 in 100every time1 in 20the usual barabout1 in 26The paper’s text: passesabout1 in 5Its own counts, 80 people: fails
1 in 100every time1 in 20the usual barabout1 in 26The paper’s text:passesabout1 in 5Its own counts,80 people: fails

How often chance alone would produce this gap if the groups didn’t differ

The paper says

Immune cells reacted to casein, a cow’s-milk protein, in 19 of 45 people with multiple sclerosis (42%) versus 10 of 35 with other nerve diseases (28%). The paper says chance alone would produce a gap this big only 1 time in 26 (p = 0.0379).

Its own data show

On those counts, the test the paper names gives about 1 time in 5 (p = 0.21); eight other standard tests give 1 in 3 to 1 in 10. The paper’s 0.0379 comes out exactly only if its percentages are used as head counts: 42 of 100 and 28 of 100, 200 people, not 80.

A p-value is how often chance alone would produce a gap this big if the groups didn’t differ; the paper counts a gap as real below 1 in 20. The same percentages pass that bar with 200 people, not 80. This was the paper’s only evidence that these cells are more common in MS; its milk-antibody blood test is unaffected.

“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 →

Antibody cross-reactivity between casein and myelin-associated glycoprotein results in central nervous system demyelination · Proc Natl Acad Sci U S A (2022) · doi:10.1073/pnas.2117034119