Cat Health Rhythm: Reading the 30-Day Drift Most Owners Miss

Cats decline slowly. Almost every chronic feline condition — kidney disease, hyperthyroidism, dental disease, arthritis, diabetes, cardiomyopathy — develops over months to years before the cat is acutely unwell. The sum of small daily changes becomes a trajectory. Spotting that trajectory early changes outcomes: a vet workup at week 6 of slow weight loss is a different conversation than week 26.

The problem is humans are bad at noticing slow change in animals we live with. Daily comparison to yesterday hides drift; yesterday looks like the day before. The fix is structured multi-week tracking with explicit pattern-matching rules — what modern cat-care apps call health rhythm or drift detection. This piece walks through how it works, the six rules vets pattern-match on, and where it fits into the broader is-my-cat-sick decision flow.

Why drift, not just "is it bad today"

Cats evolved in conditions where visible illness attracted predators. The species selected for behavioural masking that persists in domestic cats. Practical implication: when something is wrong with a cat, you usually can't see it day-to-day. The cat compensates, hides, eats less but not zero, sleeps more but not obviously, loses weight but you don't weigh them.

What you CAN see, given enough data points across enough time, is drift. The cat that has been logged as "off mood" 4 days in the last 10 isn't having a bad week — that's a 40% rate where baseline was probably 5%. The cat that lost 200g across 30 days is showing a sustained pattern, not a normal weight fluctuation.

This is what vets pattern-match on. The "she's been off for a few weeks" patient gets a more focused workup than the "she threw up yesterday" patient. Drift is the language vets work in, and structured tracking gives them that language directly.

The six drift rules — what to actually flag

The simplest deterministic drift detection — no machine learning required, just rules — uses these six. Each is a pattern any owner can implement themselves with a notebook, or that a structured app applies automatically.

Rule 1 — 3+ "off" mood days in past 7 → CONCERN

Mood is the cheapest data point to log (one tap a day) and one of the most diagnostic. A cat that's "off" for nearly half the week is showing more than noise. Possible causes range from minor (stress, weather change, household disruption) to serious (early illness, pain, neurological change). Flag is "concern" tier — book a vet within 48 hours if no obvious environmental cause.

Rule 2 — Any "none" appetite day in past 7 → CONCERN

Anorexia for a full day in an adult cat is the most reliable single warning sign. Within 24-48 hours of stopping eating, cats begin breaking down fat for energy, which the liver processes inefficiently — the resulting fat infiltration of liver tissue is hepatic lipidosis, a serious and sometimes fatal condition. The prevention is cheap (eat normally); the treatment is expensive and prolonged. Any "none" day = vet call within 24 hours.

Rule 3 — 3+ "half" appetite days in past 7 → WATCH

Less acute than full anorexia, but the pattern of consistently reduced appetite is a leading indicator. Schedule a vet conversation within 1-2 weeks if the pattern persists.

Rule 4 — Weight ±5% in 30-day window → WATCH; ±10% → CONCERN

Cats are small. A 4 kg cat losing 200g (5%) is the equivalent of a 70 kg human losing 3.5 kg unintentionally — significant. 10% loss over 30 days is dramatic and almost always reflects underlying disease. The most common differentials in older cats are kidney disease, hyperthyroidism, and diabetes — see the unexplained weight loss guide for the full workup.

Rule 5 — 7-day or 30-day check-in streak → GOOD

The positive flag. Streaks confirm baseline data quality. A 30-day streak means the system has 30 reliable data points to pattern-match against; a 60-day streak means trends are confidently established. The streak rule isn't medical — it's data quality. Without it, the other rules are unreliable.

Rule 6 — <40% logging coverage in 14 days → WATCH

Not a health flag — a data flag. If you've logged fewer than 6 days in the last 14, the system can't reliably pattern-match. Coverage warning surfaces here so the owner knows the silent assumption (that no log = nothing changed) is wrong; no log = no data.

Where drift detection fits in the bigger triage flow

Drift is one of three signals in a complete cat-health monitoring stack:

• Acute symptoms — single-point events (vomiting, lethargy, hiding) that may or may not warrant immediate action. Covered in the is-my-cat-sick checklist.
• Drift — multi-day pattern deviation in mood, appetite, weight, or activity. Covered here.
• Vital-sign monitoring — sleeping respiratory rate (HCM screening — see SRR guide), Feline Grimace Scale (pain — see FGS guide), gum colour, weight trend.

An acute symptom (rule 1: vomiting × 3 today) → immediate triage. A drift pattern (rule 4: 5% weight loss in 30 days) → vet workup conversation. A vital-sign deviation (SRR 38 in a Maine Coon) → cardiology referral. Different signals, different actions, all important.

The data-quality lesson

Drift detection is only as good as the data you feed it. Two practical implications.

First, log the 10-second daily check-in even when nothing is wrong. The 30-day streak is what makes the drift cards reliable when something IS wrong. The instinct to skip logging on "boring" days is the same instinct that misses drift later.

Second, weigh monthly. Weight is the single most pattern-rich signal in the dataset; without it, half the rules are unusable. A cheap kitchen scale, the cat in a carrier, weighed once at the same time of day each month, is enough.

What this changes day-to-day

You stop relying on your unaided observation of a species evolved to mask illness from observers. The drift cards do the pattern-matching for you, the way a vet would if a vet sat in your house every day. Most months, no drift fires — that's the system working. The months it does fire, you're catching things weeks earlier than you would have, which is exactly the lead time that turns a "manageable chronic condition" diagnosis into the calm clinic visit instead of the panicked ER one.

Frequently asked questions

What is "health drift" and why does it matter for cats?

Drift is the slow, day-by-day deviation of a cat's normal patterns — mood, appetite, weight, activity, behaviour — that adds up to a meaningful change over weeks. Cats hide acute illness reflexively (prey-animal evolution), so most chronic conditions surface as drift rather than dramatic episodes. A cat that has gradually become 200g lighter, 10% less active, and 30% less interested in food over six weeks is showing classic drift. Day-to-day, the change is invisible. Pattern-matched over 30 days, it's the early signal of kidney disease, hyperthyroidism, dental pain, arthritis, or a dozen other slowly-progressing conditions.

How is drift different from a "bad day"?

A bad day is one off-mood entry, one skipped meal, or one sleep-heavy afternoon — surrounded by normal data. Drift is a SUSTAINED deviation: 3+ off-mood days in 7, or 5% weight loss over 4 weeks, or appetite consistently below baseline for 10+ days. A bad day resolves; drift accumulates. The whole point of multi-week tracking is to distinguish single-data-point noise from sustained pattern.

Why can't I just notice my cat is off?

Two reasons. (1) Cognitive — humans are bad at noticing slow change. The "boiled frog" effect is real for cat owners; what looked normal yesterday compared to today doesn't register, but yesterday compared to 60 days ago is dramatic. (2) Cats hide illness behaviourally — the prey-animal masking is automatic. By the time changes are visible enough to articulate without notes ("she's been off for a few weeks"), the underlying problem has typically been progressing for longer than that. Structured logging removes both biases.

What's the most actionable single drift signal?

Anorexia for 24+ hours in an adult cat. Cats that don't eat for 48 hours start developing hepatic lipidosis (fatty liver), which is much harder to treat than to prevent. Any drift card that flags "any 'none' appetite in past 7 days" warrants an immediate vet call regardless of whether the cat has other symptoms. The runner-up signals: weight ±5% in a 30-day window (warrants check-up), 3+ "off" mood days in past 7 (warrants observation + likely vet call within 48h).

Will drift detection replace my vet's exam?

No, and shouldn't. Drift cards flag patterns worth bringing to a vet conversation; they don't diagnose. Their value is two-fold: (1) catching drift earlier than you would unaided, and (2) giving the vet a structured 30-day timeline to look at rather than your verbal "she's been off for a while." Vets pattern-match on exactly this kind of timeline — drift detection is essentially packaging the data the way the vet would want to see it.