Last winter I went with my mother to collect her health check report. It listed height 158 cm and weight 72 kg, and the conclusion column read BMI 28.8, obese. She stopped eating dinner that same day. I recalculated the numbers and found two things. For the same height and weight, using the cut-offs common internationally, where 25 and above counts as overweight and only 30 and above counts as obese, she came out merely overweight. And the waist column said 92 cm, which deserved more attention than the BMI. Add my own case, training four times a week at 77.5 kg and 1.75 m, BMI 25.3, flagged as overweight by the company check-up, and I spent the time working out what these two numbers can actually tell you.
1. It is one line of division, but first say which cut-offs you mean
The formula is BMI = weight (kg) ÷ height (m)². 72 ÷ (1.58×1.58) = 28.84, about 28.8.
The grades on Chinese health check reports come from the health industry standard WS/T 428-2013, Criteria of weight for adults, issued in 2013:
| Grade | Range in the standard | Notes |
|---|---|---|
| Normal weight | 18.5 ≤ BMI < 24.0 | Includes 18.5, excludes 24.0 |
| Overweight | 24.0 ≤ BMI < 28.0 | |
| Obese | BMI ≥ 28.0 | |
| Central obesity | Waist ≥ 90 cm in men, ≥ 85 cm in women | Judged by waist, independent of BMI |
One more note: the WHO figures of 25 for overweight and 30 for obese are an international classification, a different system from the table above, and the two should not be forced into one table for comparison. The Chinese cut-offs are lower, usually explained by Asian populations carrying a higher body fat percentage and more abdominal fat at the same BMI, so risk appears earlier. Reports in China are read against WS/T 428-2013, and "by WHO standards I am only overweight" means nothing in a mainland setting.
There is a grey zone between the two sets. Height 1.68 m, weight 69 kg: 69 ÷ (1.68×1.68) = 24.45, about 24.4. By the Chinese standard that is already overweight; by the international classification it is still normal (24.4 < 25). At the same height and 66 kg it is 66 ÷ 2.8224 = 23.38, about 23.4, normal either way. Between 24 and 25 the conclusion depends on which column you read. I do not argue about grade names at the boundary; I go straight to waist circumference and body fat percentage.
You can work the normal range backwards: multiply both ends of the range by height squared, calculate exactly first and then round to 0.1. At 1.58 m: 18.5×2.4964 = 46.18 and 23.9×2.4964 = 59.66, so 46.2-59.7 kg; at 1.75 m: 18.5×3.0625 = 56.66 and 23.9×3.0625 = 73.19, so 56.7-73.2 kg. The standard says BMI < 24.0, and with one decimal place you can only use 23.9, a gap of 0.1×height², which is 0.31 kg at 1.75 m. Working back from 23.9, someone at 1.72 m and 80 kg (80 ÷ 2.9584 = 27.04, about 27.0) has to drop to 70.7 kg to return to the normal range, so 80 − 70.7 = 9.3 kg is the real target, not "lose another five jin and see".
2. It cannot tell muscle from fat, and that is where every misreading starts
BMI takes only height and weight as inputs. It does not know whether what you lost was fat or muscle, or whether fat sits at the waist or on the limbs, so three kinds of error are unavoidable:
- False positives: strength trainees run a high BMI with low body fat. My 25.3 is one of these, at 17% body fat and a 79 cm waist, which by the numbers should not be called overweight.
- Missed cases: hidden obesity with a BMI around 22 but an 88 cm waist looks fine on paper while the metabolic risk is far from low. The central obesity line is 90 cm for men and 85 cm for women, and my mother's 92 cm deserved attention before the 28.8 did.
- Age is a dimension BMI cannot cover: the older you get, the less muscle sits behind the same BMI, and the number distorts. Whether to lose fat or build muscle has to be judged together with waist circumference, body fat percentage and existing conditions, and that part is for a doctor or a clinical dietitian to decide.
So my use of it now is this: BMI only acts as a switch for whether to look further, while waist circumference and body fat percentage decide whether to act.
3. How I actually calculate and track it
- Do it on paper first, then check it with a tool. The easiest mistake in hand calculation is how you handle height: forget to square it and divide once, 72 ÷ 1.58 = 45.6, which still falls in the usual BMI magnitude and is easy to accept as a real result; square the centimetres instead, 72 ÷ (158×158) = 0.0029, and you know at a glance it is wrong. The first one is the truly dangerous case. I use the BMI Calculator only to verify, and I trust the number only when both sides agree.
- Weigh yourself under the same conditions: after emptying in the morning, in the same clothing, once a week. A daily swing of 1-2 kg comes mostly from water and food, so watch the trend, not any single day.
- Measure the waist 0.5-1 cm above the navel, at the end of a breath, once every two weeks. It reflects abdominal change earlier than weight, and it is the one measure that genuinely improved for my mother this time.
- When you want body composition, estimate it once with the Body Fat Calculator from waist, hip and height, then work out how much to eat each day; treat the estimated body fat percentage as a trend only.
4. Details that are easy to overlook
- Children and teenagers do not fit the adult table: they are read against BMI percentile curves for their age and sex, and adult cut-offs mean nothing here.
- Groups it does not apply to: pregnancy, edema, amputation, athletes with very large muscle mass. Do not make decisions on this number for them.
- Height shrinks: leave it unmeasured for a few years and measuring 1-2 cm less than the record is common, and BMI is sensitive to the denominator, so a fresh measurement beats the old value.
- There are two yardsticks for the speed of weight loss: 0.3-0.5 kg a week is the common gym-world recommendation and what I do myself, with 1 kg of fat at roughly 7700 kcal, which spreads to a 330-550 kcal daily deficit, or about 1.2-2.0 kg a month. The National Health Commission's Dietary Guidelines for Adult Obesity (2024 edition) sets another level: 2-4 kg a month, or 5%-10% of current body weight within six months, which is a target for people who are overweight or obese and working under professional guidance. My number is even gentler than the lower bound of the guideline, so do not mix the two.
- A fixed rhythm: weight once a week, waist once every two weeks, BMI recalculated once a month, all three recorded in one table to read the trend. No single column is enough for me to draw a conclusion.
5. Common questions
Q: If my BMI is normal, can I stop worrying about weight?
Not necessarily. Someone with a normal BMI but an oversized waist and high body fat can carry more risk than someone with an elevated BMI and a normal waist. It is a screening measure, not a verdict on health.
Q: Are the grades different for men and women?
The BMI grades in WS/T 428-2013 use the same cut-offs for both sexes. What differs by sex is the waist line, 90 cm for men and 85 cm for women, along with body fat percentage, and that distinction only matters when you look at body composition.
Q: At what BMI should I see a doctor?
The grades themselves are not a diagnosis. A BMI of 28 or above, a waist past the 90/85 line, or a weight drop of more than 5% within six months without trying (5% of my mother's 72 kg is 3.6 kg) all justify checking blood sugar, blood lipids and thyroid function. This means unintended change, which is a different matter from the "lose 5%" in a weight-loss target; what to test and how is up to your doctor.
Q: I train with weights year round and my BMI stays above the line. Do I really need to lose weight?
Look at waist circumference and body fat percentage before answering. When both are normal, the high BMI is mostly muscle pushing weight up; once the waist has crossed into central obesity, what needs losing is fat, and years of training do not change that.
Q: Can I trust the numbers on a home body fat scale?
Bioelectrical impedance is affected by drinking, eating and skin moisture, so a single reading swings widely. The long-term trend means more than any one day. Like BMI, it is an estimate, not a measurement.
The conclusion from that report is this: 28.8 was not miscalculated, but it is a prompt about whether to look further, not an instruction to skip dinner. My mother ended up walking 3,000 more steps a day and halving the starch at dinner while keeping protein and vegetables. In three months she lost 4 kg and 5 cm off her waist, which worked better than slashing calories to match a grade and did not starve her muscle away with it.