The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
in the estrogen plus progestin arm they saw a non-statistically significant increased risk of breast cancer and it was this the relative risk relative now you know what this is but your Mone your listeners may not was 25% and and I hope I get the numbers right it was four out of a thousand women per year to five out of a thousand women per year
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.
when you look at the women who had a uterus these are women who got estrogen and progesterone synthetic progesterone progestin versus placebo there was a trend towards a higher breast cancer but when you look at the opposite the women who had no uterus and therefore only needed estrogen the trend was in the exact opposite direction to the exact same magnitude
the absolute increase in risk if you believe the 26 percent relative increase it turned it calculated at about 0.9 cases per thousand or 9 cases per 10,000 so that link was less than one percent absolute risk increase even if it turned out to be true
estrogen only had a 21% risk but the 95% confidence interval crossed unity which meant it was not statistically significant the p-value is point zero seven if my memory serves me correctly which meant there is a seven percent chance that that effect was random meaning it didn't hit the five percent threshold right but I remember that mostly pretty pretty close the the fact is there was no real increase among the women who got estrogen alone that we would hang our head on
in 2003 they came out with a second article showing borderline statistical significance for the increased risk of breast cancer and this is the point you alluded to earlier and this is actually the paper that I am looking at more often because this is the one that went from point one point two six as the hazard ratio to one point two four and this is the one where the p-value actually went to 0.5 from I believe point I was a show that it was that's right borderline statistically significant
in the estrogen plus mpa group there was a barely statistical significant increase in the risk of breast cancer i think the hazard ratio was i want to say it was about 1.24 1.25 um and the p-value was exactly 0.05 or 0.049 or something like that so at the surface you'd say gosh this is increasing the risk of breast cancer and what was talked about was it was it was a 25 percent increase in the risk of breast cancer but the you know to talk about the relative risk increase without talking about the absolute risk is obviously irresponsible if you look at the absolute risk change it was a 0.1 it was one in a thousand
in the estrogen only versus the placebo there was a non-statistical significant reduction in the risk of breast cancer so there was a hazard ratio of about 0.8 or 0.79 or 0.81 something like that but it didn't quite reach statistical significance
the relative risk increase of breast cancer in the estrogen plus MPA group versus the placebo was 25 27 percent and that became the only headline HRT increases risk of breast cancer by 27 percent now I don't think that's true at all today but let's even look at the data what was the ARR what was the absolute risk increase it was a difference between five cases per thousand and four cases per thousand so the ARR was 0.1 percent one case in a thousand
The absolute risk increase for these women was 0.1%.