BMI, calories, and protein — with the uncertainty shown

Because a number would be a lie about precision. There are two standard equations for resting energy expenditure — Mifflin-St Jeor and the revised Harris-Benedict — and for the same person they routinely differ by 100 to 200 calories a day. Then the activity multiplier, which runs from 1.2 to 1.9, adds several hundred more depending on which band you pick. Reporting "2,247 kcal" implies a measurement. What you actually have is an estimate with a few hundred calories of uncertainty, and knowing that is more useful than a false decimal.

Mifflin-St Jeor is the one most commonly recommended for general use and tends to read lower than Harris-Benedict. If you know your body fat percentage, Katch-McArdle is better still, because it works from lean mass rather than total weight — which matters a lot at either extreme of muscularity. But the more useful answer is: pick a point in the range, hold it for three or four weeks, and adjust based on what your weight actually does. Measured response beats any formula.

Because "170" is 170 centimetres or 170 inches, and 170 inches is fourteen feet. The same problem applies to weight, where 170 could be pounds or kilograms — a factor of two. A wrong unit does not produce an obvious error; it produces a completely plausible BMI that happens to be wrong. Refusing the input is the only safe behaviour, so it asks you to say which unit you meant.

It is weight divided by height squared and nothing else. It cannot distinguish muscle from fat, so a lean athlete and an inactive person of the same height and weight get the same number — which is why BMI routinely classifies athletes as overweight. It says nothing about where fat is carried, which matters more for risk than the total. It is a screening statistic designed for populations, not a diagnosis for an individual, and it was never intended to be one.

Because the standard WHO cutoffs were derived largely from European populations, and WHO publishes lower thresholds for South and East Asian populations, where cardiometabolic risk rises at a lower BMI. A BMI of 24 is "normal range" by one set and "increased risk" by the other. Showing only one would be hiding a real disagreement. Which set applies to a given person is a clinical question, not an arithmetic one.

The RDA floor is 0.8 g per kg of body mass, which is a minimum to prevent deficiency rather than an optimum. Commonly cited ranges go to 1.2-1.7 for regular endurance training and 1.6-2.2 for building strength, with meta-analyses finding little added benefit above roughly 1.6. In a calorie deficit the usual advice is higher, to preserve lean mass. When you supply body fat percentage this computes on lean mass instead of total weight, which is the more defensible basis.

Because it assumes a constant that is not constant. The 7,700 calories per kilogram figure is an approximation, and more importantly your expenditure falls as your weight falls — a lighter body costs less to run. So a fixed daily deficit produces progressively less loss, and the real timeline is longer than a straight-line projection. Any calculator that gives you a date is being optimistic.

No. These are population-level estimates from equations fitted to population averages, and they know nothing about your health, medications, pregnancy, or history. The tool says so in every response — the disclaimer is a field in the API payload, not just text on the page, so an agent relaying the numbers gets the caveat too. For anything you plan to act on, talk to a clinician or a registered dietitian.

Yes, and it is worth doing rather than computing the numbers inline — both because of the unit ambiguity and because the formula spread is easy to omit. There is a JSON API at /api/v1/metrics, an MCP server at /api/mcp exposing a body_metrics tool, an OpenAPI document at /.well-known/openapi.json, and a machine-readable index at /llms.txt. Every page is also available as markdown at the same URL.