SOUL oral semaglutide MACE data — am I reading this right?

The 14% is relative risk. I saw an absolute difference around 1.5% to 2%, but I haven't seen the curves either.

Same pattern a sSELECT. Relative huge, absolute modest.

Modest absolute over three years in a high-risk group still matters. Did they publish NNT anywhere?

I think it was around 200 over the trial. Don't quote me, I only skimmed teh release.

The NNT guesses are fine, but I watched the conference session and the investigators were careful about the secondary endpoints. The reported 14% seemed driven mostly by nonfatal MI, and CV death looked basically flat.

The part that got lost in the headlines is that the trial wasn't designed to compare oral versus injectable. So all the 'will this replace shots' chatter is premature. I want the full paper and subgroup data before I get excited.

This is why I wait for the paper. The discussion section usually tells you what the press release avoided.

Fair, but the oral formulation angle is what keeps me clicking. I've been keeping a boring spreadsheet of labs, sleep and training for 19 weeks now, and my own numbers moved more than I expected: hsCRP 3.8 to 1.4, fasting glucose 108 to 94. That doesn't tell me anything about MACE.

Still, if an oral version has a similar CV signal with different GI tolerability, access and persistence could change fast. I just wish the coverage separated the mechanism news from the outcomes news.

I'm 31 weeks in and the numbers that actually stopped me doom-scrolling were resting HR 78→64, BP 132/86→118/76, and hsCRP 3.8→1.6, even though my A1c chilled at 5.5 for 10 weeks and the scale did that lovely bounce around the same 1.5 pounds before dropping again. The surprise was that my waist kept shrinking 2 inches during that plateau, and once I got my sleep from 5.5 to 7.5 hours, the labs finally caught up—so now I do Sunday-morning waist plus a two-week BP log before I let any relative-risk headline rent space in my head.

Are you tracking morning BP or resting HR at all, or just the every-8-week labs?

I'm only 18 weeks in, and the number that flipped my brain wasn't a MACE percentage—it was going from a tight 36 waist to a 34 with a 1.5-inch drop on the tape while my scale sat at the same 212 for nine straight days. The surprise was that the scale finally moved after I stopped killing myself with extra cardio and instead fixed sleep: 5.5 hours to 7.2, plus a 15-minute walk after dinner on weeknights.

My resting HR went 77 to 69, but Monday mornings still lie to me after a salty weekend—up 2.3 lbs, back down by Wednesday. I get wanting the absolute risk and KM curves, and I’d still read the actual paper if it drops, but for my own tracking the waist tape and sleep log have been way more honest than the daily scale.

i’m 40 weeks in and the thing I’m watching is whether the 14% holds in people without prior CVD. SOUL enrolled high-risk T2D, so baseline MACE was already elevated; a 14% relative drop might be a small absolute difference for someone whose main issue is weight/insulin resistance.

If the paper reports annualized event rates, I’d love to see the placebo arm number. Also curious if GI side effects caused more discontinuations—tolerability can shift the real-world calculus.

Same — I can only find the press-release version, not the full paper. I’m less interested in the 14% than in the absolute numbers: events per 100 patient-years, and whether the benefit was drivn by nonfatal MI/stroke or CV death. Also, did they pre-specify eGFR subgroups? My dad’s CKD is why I care.

Meanwhile I’m 6 weeks into a plateau and my home BP log is the only thing still moving, lol.

The thing I keep looking for is whether SOUL reported an absolute risk difference per 100 patient-years and an NNT, plus whether the MACE split was driven by nonfatal MI/stroke or CV death. A 14% relative drop can look very different depending on baseline risk. Did the conference materials mention baseline event rate?

For my own tracking, I’ve switched to 4-week rolling averages for BP and sleep, since single bad readings/nights were making me over-interpret.

Any word on the MACE component splits? I keep seeing the composite but not how much was nonfatal MI vs CV death vs stroke. If it’s mostly one component, that changes how I read it.

Also want the KM curves—early separation feels more pleiotropic, late separation feels more like risk-factor cleanup. I lift 4x and track protein/sleep, but CV events are the metric I actually care about beyond scale weight.

I’m 21 weeks in and stuck at the same weight for a month despite logging everything, so the headline number isn’t what I’m chewing on. I want to know if SOUL’s CV effect was independent of weight/A1c change, or if the benefit mostly tracks with how much people actually lost.

If it’s the latter, my plateau feels a lot less “just cosmetic.” Did they prespecify a weight-adjusted analysis, or is that just something people are speculating about until the full paper drops?

Tbh ugh, same — scale’s frozen but my jeans are looser, so I’m choosing to believe the jeans.

I’ve only seen the conference headline too, not the full paper. The detail I’m hunting for is the enrolled population: was SOUL mostly people with established CV disease/T2D, or did it include lower-risk weight-only folks?

A 14% relative reduction reads very differently depending on baseline risk. I’ve been tracking home BP and resting HR weekly alongside labs, which helps separate real trends from noisy one-offs.

If anyone spots registry results or a peer-reviewed link, please drop it.

Yeah, the headline is doing a lot of work. I’m hunting for the absolute risk difference/NNT, not just the 14% relative.

If the composite event rate was ~4% over the trial, a 14% relative cut is tiny in absolute terms; if it was 15%, different story. Did they report events per 100 patient-years or just the HR? My spreadsheet made me contrarian: I now log absolute deltas next to every relative trendline.

My dorm spreadsheet brain wants the event curves: did the lines split early or did the benefit show up after year 2? Also which component drove it—nonfatal MI, stroke, or CV death? And did the trial enroll people with established CV disease/diabetes, or lower-risk folks like a lot of us?

Because 14% relative on a small placebo event rate would hit different than on a high-risk group. If anyone finds a free PDF, I’ll trade you my legendary 89-cent tuna-chickpea situation.