I gained 13.6 pounds of muscle in four months. Probably not.

I got my fifth DEXA scan yesterday. The number was on the iPad before I’d swung my legs off the table: lean mass 252.6 lbs, up from 239.0 in January. Plus 13.6 pounds of lean tissue in four months.

I lift five days a week and ate in a surplus the whole time, so some of that is real. But I’m 35 and a half, several years into serious training, and not on anything. Thirteen-plus pounds of muscle in a quarter is not how the body works at this stage. So before I believed the number, I went looking for where it came from.

Start with what the report actually measures. It says “Lean Mass,” and right below, in grey, it tells you what’s in that bucket: Muscle Mass, Organs, Blood and Stomach Contents. It never says how those split. The thing I care about — pounds of contractile skeletal muscle — isn’t a number the machine reports.

Glycogen is the big confound. Trained muscle stores roughly 15–18 g of glycogen per kg, and every gram drags about three grams of water with it. Topped off, at my size, that’s plausibly 4–6 lbs of glycogen and bound water sitting inside the muscle, and DEXA reads all of it as lean mass. The two scans weren’t matched, either: January I went in fasted, after a low-carb day; May I went in fed, mid-training-week, pre-workout carbs on board. If my muscle was more depleted in January than in May — it almost certainly was — then four or five pounds of the “gain” is plumbing, not tissue.

Then there’s the noise. DEXA is good, not perfect. The same machine scanning the same person on back-to-back days can disagree by a pound or two before anything real has changed, and the spread widens across months as positioning, hydration, and machine drift stack up.

So here’s the 13.6, itemized:

• ~4–5 lbs — glycogen + bound water (optimistic)
• ~1–2 lbs — measurement noise
• ~1 lb — organ/blood/gut shifts at this size
• ~5–7 lbs — new muscle, best case

Five to seven pounds of real muscle in four months at 35 is still aggressive — that’s first-year-lifter territory, and my first year was a long time ago. My real guess is 2–4. And 2–4 lbs of actual muscle in four months is a good bulk. It’s just not the number on the report — the one I almost posted before I did the math.

The trend tells it cleaner than any single scan:

Two ~14-lb jumps, two boring scans in between. The boring ones — +5.2 over nine months, +3.2 over three — match both the literature and my actual lifting numbers. The jumps line up exactly with the scans I walked into hydrated, fed, and post-training-week. The real pattern isn’t “ten pounds of muscle a quarter, then plateau.” It’s “a couple pounds of muscle a quarter, and the rest is signal noise the DEXA faithfully logs as muscle.”

I start a twelve-week cut tomorrow, and the next scan settles it — fasted, low-carb day, matched to January. If lean drops 6–8 lbs while fat drops 15, that’s the glycogen-and-water tax coming due, and what’s left is what I built. If it drops 12+, the +13.6 was almost all smoke.

I don’t know which it’ll be yet.

Update, August 2026. The scan happened July 24, nine weeks into the cut — fasted, low-carb day before, the protocol I promised. One correction to the record before the numbers: I described January’s scan up top as fasted after a low-carb day. Going back through it honestly, I was closer to fed that morning — not May-level carbed up, but not empty either. So even the baseline I called clean was a little padded. Keep that in mind as you read the rest.

The numbers: lean mass 236.4 lbs, down 16.2 from May. Fat mass 41.1, down 8.5. Total mass down 24.5. Body fat 15.8% to 14.2%.

I gave two outcomes. Lean drops 6–8 while fat drops 15, and the gain was real minus the water tax. Lean drops 12+, and the +13.6 was almost all smoke. It dropped 16.2. Smoke — and I missed on the other side too. Both scenarios added up to roughly 21–23 pounds off; 24.5 came off. I got the total nearly right and the split wrong: I figured two-thirds of the loss would be fat. Two-thirds of it was “lean.”

So, the May ledger, regraded. Glycogen and bound water wasn’t the optimistic 4–5 lbs — closer to 8–10 once you put the fed-versus-drained gut in the same bucket. Noise is still noise. And the “5–7 lbs new muscle, best case” line item: lean is now 2.6 lbs below January. Before anyone panics on my behalf — my lifts held flat through nine weeks of deficit, which is the best evidence I have that nothing structural left. And July-me was drained in a way semi-fed January-me wasn’t, so the true January-to-July change is probably close to a wash. Call it one to three pounds of actual muscle built across the whole bulk, most of it kept. Against 13.6 on the report.

The cut itself did what it was supposed to: fat mass down 8.5 lbs, visceral fat from 84 cubic inches to 58 — out of the machine’s “increased risk” band and nearly into the healthy one — with lifts that didn’t move. Nothing about that required the May number to be true.

Next bulk, both ends get the same protocol: fasted, low-carb day before, morning scan. Then the comparison means something.