featureHealth Patterns
Health Patterns: What Your Own Data Says Actually Works for You
One observed pattern a week, drawn from comparing your fasting days against your non-fasting days, your training days against your rest days, and your body scans against the scale.
General nutrition advice is written for a population. You are one person, and the only dataset that describes you is the one you have been building every time you logged a meal, closed a fasting window, drank a glass of water, or finished a workout.
Health Patterns compares those records against each other and reports what it finds — with the sample size it was drawn from, phrased as an observation, never as a cause. Nine rules run over the same data your other screens already loaded, so nothing new is fetched and nothing new is stored.
3-day free trial · Cancel anytime · No card charged during trial
Why Most Trackers Cannot Tell You This
In most apps, fasting lives in one screen, water in another, workouts in a third, food in a fourth, and body composition somewhere else entirely. Each feature writes its own record and each screen reads only its own. Nothing can say anything that spans two of them.
That is a structural limit, not a design choice — and it is why a fasting app can tell you that you completed 18 fasts and still cannot tell you whether they helped. The comparison requires the food log and the weight log in the same place as the fasting log.
SlimAI mirrors one number per feature onto the same daily record: water, fasted hours, steps, workout minutes, when your first and last meals landed, how processed your food was, how many meals came from your plan, how many came from a scan, and your weight. Once every feature writes to one day, cross-feature comparison becomes possible.
The Nine Patterns SlimAI Looks For
Each rule carries its own minimum sample. Below it, the rule does not fire at all — you never see a comparison built on two days.
Does fasting actually help you?
Compares your calorie intake and results on days you completed a fast against days you did not. Needs at least 3 of each before it will say anything.
Are you eating back your training?
Compares your intake on training days against rest days. A common and completely invisible pattern: the workout is real, and so is the extra 400 kcal that followed it.
Protein gap on training days
Checks whether your protein actually rises on the days you train. Needs at least 3 training days.
Low water, higher intake
Compares your calorie intake on low-hydration days against normal ones. Requires 3 of each.
Losing weight or losing muscle
The flagship pattern, and the one that needs body composition to exist at all: it compares the change in your lean mass against the change on the scale across two body scans. A scale cannot tell the difference between fat lost and muscle lost. This can.
Late eating
Compares days when your last meal landed late against the rest of the week. Needs 4 days with meal times recorded.
Does the diet plan work for you?
Compares days where you logged meals from your plan against days you improvised. Requires 3 of each.
Scanning and accuracy
Compares how much of your food was scanned on the days you went over target against the rest of the week — a proxy for where estimation crept in.
Time since your last body scan
Flags when 30 days have passed without a body composition scan, so the trend line does not quietly go stale.
Ready to try Health Patterns?
Snap a photo and SlimAI does the rest. 50k+ people already log this way.
3-day free trial · Cancel anytime · No card charged during trial
The Lean Mass Pattern Is Why Body Composition Matters
Losing weight is not the goal. Losing fat is. The scale cannot distinguish between the two, which is why a diet that "works" for six weeks can leave you lighter, weaker, and with a lower resting metabolism than you started with.
Because SlimAI includes AI Body Assessment, it can compare fat mass and lean mass across two scans and put that next to what the scale did. If lean mass is falling alongside weight, the pattern fires — and it carries two actions rather than one: raise protein, and add resistance work.
This is the single pattern a paid coach is most often hired to catch. It requires two body scans, and nothing else in your log can substitute for them.
One Pattern a Week, Rotated
Even when several rules qualify, you see one. A screen listing six simultaneous observations is a dashboard, not a coach — and nobody acts on six things.
Which one you see rotates deterministically by week, so the surface stays fresh across a month without repeating the same finding every Monday, and the same week always shows the same pattern if you open it twice.
Any individual rule can be switched off remotely without an app update, so a pattern that turns out to be unhelpful stops appearing for everyone immediately rather than waiting on a release.
Observations, Not Causes
Every pattern sentence is written as something observed in your data over a stated number of days. None of them claims that one thing caused another, because that is not what a comparison of your own logs can establish.
Your data is also not a controlled study. Sleep, stress, illness, medication, menstrual cycle, and dozens of other variables are not in the comparison, and any of them can be the real story behind a pattern.
Treat what you see as a hypothesis worth testing for a couple of weeks, not a verdict. If a pattern touches something medical, that is a conversation for a qualified health professional — SlimAI provides tracking tools, not medical advice.
What Health Patterns Needs From You
Patterns light up as data accrues. Nothing is backfilled, so the clock starts when you begin logging, and the rules that compare two groups of days need both groups to exist.
Log most days
Four or more logged days a week is the floor for most comparisons. Consistency is what makes the rest possible.
Use more than one feature
A fasting pattern needs fasting days. A hydration pattern needs water logs. A training pattern needs workouts or synced activity.
Scan your body twice
The lean mass pattern is the highest-value rule in the set and needs exactly two body scans to fire — one is a starting point, two is a trend.
Weigh in regularly
Weight readings anchor most of the result-side comparisons. Three in the window is the practical minimum.
Explore More SlimAI Features
FAQ
Still Wondering About SlimAI
Health Patterns compares different parts of your own log against each other and reports one observed pattern a week. Nine rules run across fasting days versus non-fasting days, training days versus rest days, low-water days versus normal days, late-eating days, plan-based meals versus improvised ones, scan coverage, and the change in lean mass between two body scans. Each rule states the sample it was drawn from and is phrased as an observation, never a cause.
It can tell you what your own data shows. Once you have at least 3 completed fasting days and 3 non-fasting days in the window, SlimAI compares your calorie intake and results across the two groups and reports the difference it observed. That is a pattern in your log, not proof of cause — sleep, stress, and many other factors are not in the comparison.
It compares fat mass and lean mass between two AI Body Assessment scans and puts that next to what the scale did over the same period. If lean mass is falling alongside your weight, SlimAI surfaces the pattern with two suggested actions: raise protein and add resistance training. It needs two body scans to fire, and the results are estimates for fitness awareness, not clinical body composition measurements.
Because a list of six simultaneous observations is a dashboard, and nobody acts on six things at once. Even when several rules qualify, SlimAI shows one and rotates which one across weeks, so the surface stays useful over a month without repeating the same finding every time.
Every rule declares its own minimum. Most comparisons need at least 3 days in each group being compared, meal-timing rules need 4 days with recorded meal times, and the lean mass rule needs 2 body scans. Below those thresholds a rule does not appear at all rather than showing a weaker version. Nothing is backfilled, so patterns light up as you keep logging.
Cross-feature patterns are part of SlimAI Pro. The free tier includes the coach card, your weight trajectory, macros against target, attribution, the weekly score, and the weekly micro-lesson.
Start logging smarter with AI Meal Logging.
Log any meal, any way. Get instant calories and macros. Build the habit that sticks.
Free 3-day trial, then $4.99/month or $34.99/year. Cancel anytime.
- 50K+ food items in the food database
- 50k+ users
- 10M+ foods logged
- 80+ countries
- 4.8★ App Store · 4.6★ Google Play (as of June 3, 2026)
