Period 1: Uric Acid 538 → 183
The lab report on October 4th was rock bottom: uric acid at 538 µmol/L, lipids and liver panels flashing red across the board. That night I had a simple thought: when things are this bad, anything you do next is an improvement.
Fourteen monk-like days later, the needle went back into my vein. Here is the full readout.
Period 1 in numbers
This is the table I was actually waiting for—bloodwork from October 4th (Day 1) side-by-side with the draw on the morning of October 18th:
| Marker | Oct 4 | Oct 18 |
|---|---|---|
| Uric Acid | 538 µmol/L | 183 µmol/L |
| Total Cholesterol | 5.6 mmol/L | 3.0 mmol/L |
| LDL Cholesterol | 3.5 mmol/L | 1.5 mmol/L |
| Triglycerides | 4.1 mmol/L | 1.4 mmol/L |
| HDL Cholesterol | 0.9 mmol/L | 0.8 mmol/L |
| Fasting Glucose | — | 4.7 mmol/L |
Uric acid didn't just drop—it landed way below the 360 gout management threshold. Both LDL and triglycerides retreated from danger zones straight into normal ranges. HDL barely budged and stays low, so there is still plenty of hill left to climb.
What I actually did
- Rebuilt my plate from dinner one. Solid protein every meal (chicken breast, tofu, eggs, milk), swapped refined carbs for sweet potatoes, pumpkin, and corn, loaded up on vegetables twice a day, cut added sugar down to an average of 5g/day, and touched zero organ meats, shellfish, or alcohol.
- Kept a buffer below target. Averaged 1,505 kcal against a 1,750 target. Never overshot the daily ceiling by more than 10%.
- Walked. 5 km brisk walks on most days—11 active days out of 14, totaling 577 minutes on my feet.
- Took the pills, every single dose. Atorvastatin for lipids, Benzbromarone for uric acid, plus Fluconazole. I didn't miss one. (The app says 64%, and on Day 9 it said 72% and I called that a fail. Both numbers count the first three days, before Anlo could track doses at all.)
Which lever moved the bloodwork most? Honestly, I can't tell yet. Clean eating, walking, rapid weight loss, and targeted prescriptions all collided in the same two-week window. My logs prove that pulling all four levers at once worked dramatically, even if they can't isolate the MVP. (And dropping 5.3 kg in two weeks is steep—plenty of week one was water weight.)
Five days outside the ICU
A big chunk of this period didn't happen at home. My father underwent major surgery and spent time in the ICU for observation. I stayed at the hospital with him for five days straight.
The easiest move would have been grabbing whatever the hospital cafeteria served and hitting pause on the whole protocol. I didn't. I prepped my own food, packed it into the ward in containers, and logged every single meal right on schedule.
I also brought my MacBook. Half the features listed below were coded on a cramped hospital chair in the quiet gaps between doctors' rounds. My dad's surgery went well, and he is recovering steadily now. Looking back, those five grueling days in the hospital were the exact moment this system stopped being a "plan" and became second nature.
What got built in two weeks
On October 4th, Anlo didn't exist. By Day 14—partially written from that hospital ward—it was doing almost everything I wished the apps on the App Store could do:
- Zero-typing capture: Snap a plate, an ingredient label, or a menu, or hold the button and speak. Items eaten together merge into a single meal, and recurring meals duplicate to new days in one tap.
- The full clinical picture: Snap a photo of a hospital lab sheet to extract uric acid, lipid panels, glucose, and HbA1c automatically. Track blood pressure, manage pausable medications with hard alarm reminders, and sync weight, walks, sleep, and resting heart rate two-way with Apple Health.
- Grounded daily coaching: Real-time daily wrap-ups after every meal, pre-meal sanity checks ("Can I eat this?" via photo), and the automated 14-day review that crunched all the numbers above.
- Self-custody privacy: No email, no password. You hold the cryptographic key, and cloud sync encrypts everything locally before a single byte leaves your iPhone.
Dogfooding exposed two real bugs right away. First, the period summary initially claimed I only lost 0.1 kg instead of 5.3 kg because Apple Health logs multiple scale readings a day; it now anchors cleanly to one reading per day without dropping the baseline weight. Second, finishing Day 14 now prompts you to roll into the next period right on the Today screen instead of burying the button in settings.
The unexpected part
My wife posted my before-and-after lab numbers on her social feed.
The next day, a friend called asking how I pulled it off. Then my mom asked. Before long, my wife and kids said they wanted in too.
Period 1 started with just me, a scary blood report, and an empty code editor. Period 2 starts with the whole family. My wife is already logging her meals on the web build while I prep the App Store release, and I suddenly have a much better design benchmark: make it simple enough that my mom can use it without asking me for help.
A quick reflection
When you are starting from rock bottom, you don't need a master plan. You just need to start, and write down what happens. Every metric that ticks in the right direction is a foothold, and logging is how you see it moving.
I built Anlo to isolate variables one by one, yet in Period 1 I threw four things at the wall at once. Standing where I was on October 4th, throwing the kitchen sink at the problem was the right call. Now I know the numbers can move. Next, I want to know what keeps them there.
Period 2: Food only, no meds
So for Period 2, I'm changing one massive variable: I've paused the lipid and uric acid prescriptions. No statins, no uric acid meds. Just three meals a day—slightly more relaxed than Period 1—and another blood draw at the end of 14 days.
If uric acid and lipids hold steady, it means the plate is doing far more heavy lifting than I dared to hope two weeks ago. If they spike right back up, I'll know the pills were carrying the load—which is equally valuable to know. Either way, in two weeks I'll know something concrete about my body that I didn't know on October 4th.
(Disclaimer: This is my personal N=1 experiment, not medical advice. Uric acid and lipid-lowering medications are serious clinical tools. Any decision to pause or adjust prescriptions belongs between you and your doctor, not a developer's blog.)
The Period 2 review will sit right next to this one for a clean comparison. Anlo handles the side-by-side math automatically now.
— Anton