Understanding BMI and Its Limits: What Body Mass Index Can and Cannot Tell You

Body mass index, or BMI, is one of the simplest measurements used in health care: body weight in kilograms divided by height in metres squared. Its simplicity is precisely why it became so widely used. With two measurements that can be collected almost anywhere, clinicians and public-health systems can classify weight relative to height, monitor population trends and identify people who may benefit from further assessment.

That simplicity is also the source of much of the criticism surrounding BMI. The calculation cannot directly tell how much of a person's body is fat, muscle, bone or water. It cannot show where fat is stored, assess physical fitness or reveal whether blood pressure, blood glucose and cholesterol are healthy. Two people with identical BMI values can therefore have substantially different body composition and health profiles.

The most accurate way to understand BMI lies between two extremes. It is neither a useless number that should be abandoned nor a complete diagnosis of someone's health. The CDC describes BMI as a quick, inexpensive and reliable screening measure, while explicitly recommending that an individual's BMI be interpreted alongside factors such as blood pressure, cholesterol, medical history and physical examination. The World Health Organization similarly describes BMI as a surrogate marker of fatness and notes that additional measures such as waist circumference can improve assessment.

How BMI Is Calculated and What the Categories Mean

For adults, the calculation is straightforward:

BMI = weight in kilograms ÷ height in metres²

For example, a person weighing 80 kilograms who is 1.75 metres tall has a BMI of approximately:

80 ÷ (1.75 × 1.75) = 26.1 kg/m²

The widely used WHO adult categories are:

WHO's December 2025 obesity guidance continues to define overweight in adults as a BMI of 25 or above and obesity as 30 or above. These categories are useful for population monitoring and risk stratification, but they should not be imagined as biological switches. A person does not undergo an abrupt physiological transformation because BMI moves from 24.9 to 25.0. Health risk usually changes along a continuum rather than appearing suddenly at one decimal point.

Children and adolescents require a different interpretation because body composition changes substantially during growth and puberty. Their BMI is generally considered relative to age and sex using growth references or percentiles rather than applying adult thresholds directly. WHO uses BMI-for-age standards for people aged 5–19 years, while the CDC likewise uses age- and sex-specific BMI growth charts for children and adolescents.

Pregnancy is another situation in which ordinary adult BMI categories should not be interpreted as though pregnancy-related weight gain represents ordinary changes in body fat. Very frail older adults, people with substantial muscle mass and people whose height cannot be measured conventionally may also require more individualised assessment.

The value of BMI comes from standardisation. A clinician in one city, a researcher in another country and a national health survey can all calculate the same index from the same two measurements. That makes BMI extremely useful for comparing large groups and studying how weight-for-height relates to disease.

The mistake is assuming that because BMI is easy to calculate, it must also explain everything about the body.

What BMI Measures—and What It Cannot See

BMI directly measures only weight relative to height. Everything else associated with BMI is an inference.

This distinction matters because body weight contains several different components. Skeletal muscle contributes to weight. So do fat, bone, organs, body water and other tissues. BMI cannot separate them.

Imagine two adults who are the same height and both weigh 90 kilograms. One has spent years doing resistance training and carries a large amount of skeletal muscle. The other has substantially less muscle and more body fat. Their BMI can be identical even though their body composition is very different.

The CDC explicitly lists this as one of BMI's principal limitations: it does not distinguish between fat, muscle and bone mass and does not identify the location or type of body fat. At the same time, BMI is moderately to strongly associated with more direct measures of adiposity across populations, which is why the measurement remains useful despite its limitations.

This also explains the familiar athlete example. A muscular athlete can fall into the overweight or even obesity BMI range without carrying the amount of body fat that the category might suggest in an average person. That criticism of BMI is valid.

But the athlete example can itself be misused. The existence of muscular people with high BMI does not mean that every person with a high BMI has been incorrectly classified because of exceptional muscle mass. In large populations, higher BMI remains strongly associated with greater adiposity and increased risk of several chronic conditions.

The correct conclusion is therefore not “BMI fails because athletes exist.” It is that BMI requires context when body composition may differ substantially from what the population average would predict.

The reverse problem also occurs. Someone can have a BMI inside the conventional reference range while carrying relatively little skeletal muscle and more body fat than expected. This may become particularly relevant with ageing, when muscle can decline while fat mass rises without producing a dramatic change in total body weight.

BMI may therefore underestimate some forms of risk just as it can overestimate adiposity in highly muscular people.

That is one reason the phrase “normal BMI” should be used carefully. A BMI within the conventional reference range means weight-for-height falls within that category. It does not certify that blood pressure, cholesterol, glucose metabolism, physical fitness, diet, smoking status or every other dimension of health is normal.

BMI Is a Screening Tool, Not a Complete Diagnosis

A good screening tool is not expected to diagnose every aspect of a condition. Its purpose is to identify patterns that may justify closer examination.

BMI performs that function unusually well because it is quick, inexpensive, non-invasive and reproducible. The CDC continues to describe it as a valuable population health measure and an individual screening tool when considered together with other clinical information.

A screening result becomes more meaningful when combined with other evidence.

A clinician interpreting a person's BMI may also consider:

• waist circumference or another measure of central adiposity;

• blood pressure;

• blood glucose or HbA1c where appropriate;

• blood lipid levels;

• medical and family history;

• smoking and alcohol use;

• physical activity;

• medications;

• recent weight change;

• physical function and fitness;

• and signs or symptoms of weight-related disease.

This distinction has become even more important in recent obesity research. A major Lancet Diabetes & Endocrinology Commission published in 2025 argued that BMI alone can both underestimate and overestimate adiposity at the individual level. The commission recommended using BMI primarily as a screening or epidemiological tool and confirming excess adiposity with direct body-fat measurement, where available, or additional anthropometric measures such as waist circumference, waist-to-height ratio or waist-to-hip ratio.

The commission went further by distinguishing clinical obesity, in which excess adiposity is associated with objective impairment in organ or tissue function or limitations in daily activities, from preclinical obesity, in which excess adiposity is present without current organ dysfunction but future risk is increased. This framework is influential but comparatively new, and clinical guidelines around the world have not all adopted identical terminology. It nevertheless illustrates the direction of modern thinking: health assessment increasingly looks beyond BMI alone.

Research published in 2026 reinforces the value of combining measurements. An analysis comparing BMI and waist-based measures with DEXA-defined excess adiposity found that BMI offered greater specificity while waist-based approaches were more sensitive; incorporating waist measurements could therefore identify excess adiposity that BMI alone might miss.

The takeaway is not that every person needs an expensive body scan. It is that BMI becomes most useful when treated as one piece of evidence rather than the final diagnosis.

Why Waist Size and Fat Distribution Matter

BMI cannot determine where body fat is stored, yet fat distribution is relevant to cardiometabolic risk.

Fat concentrated around the abdomen—often described as central or visceral adiposity—has different metabolic associations from fat stored predominantly in some other regions of the body. This is why waist circumference and waist-to-height ratio are increasingly used alongside BMI in some clinical guidelines.

WHO's current obesity guidance explicitly notes that waist circumference can provide additional information beyond BMI. NICE guidance in the United Kingdom goes further, recommending the assessment of central adiposity alongside BMI in many adults.

The broader principle is more important than memorising one measurement. Two people with identical BMI values can have different fat distribution and therefore different cardiometabolic profiles.

Direct body-composition techniques can provide more detail. Dual-energy X-ray absorptiometry, or DEXA, can estimate fat mass, lean soft tissue and bone mineral. The CDC notes that DEXA measures body composition more precisely than BMI but is more expensive and less widely available.

Bioelectrical impedance devices—including many consumer “smart scales”—estimate body composition using electrical properties of the body. They are convenient but sensitive to assumptions, hydration and device algorithms. Their outputs can sometimes be useful for following trends under consistent conditions, but a body-fat estimate displayed to one decimal place should not be confused with an exact laboratory measurement.

Skinfold measurements and other anthropometric techniques offer additional options but depend substantially on measurement technique and prediction equations.

For most routine health decisions, the objective is not to obtain the most technologically sophisticated body-composition measurement possible. It is to gather enough information to understand risk more accurately than BMI alone allows.

The Same BMI Does Not Always Mean the Same Risk

BMI categories are useful precisely because they simplify a continuous biological relationship. But simplification can hide differences among populations.

Age, sex, ancestry, fat distribution, muscle mass, cardiorespiratory fitness, smoking, socioeconomic conditions and existing disease can all influence the relationship between BMI and health risk.

One particularly important issue concerns ethnic and ancestral differences in cardiometabolic risk. NICE currently recommends lower BMI thresholds for people from South Asian, Chinese, other Asian, Middle Eastern, Black African and African-Caribbean backgrounds because cardiometabolic risk and central adiposity can occur at lower BMI values in these groups. Under this framework, overweight begins at BMI 23 rather than 25 and obesity at 27.5 rather than 30.

This is especially relevant in South Asia, where type 2 diabetes and metabolic risk can occur at BMI levels that might appear relatively modest under conventional global categories.

The American Diabetes Association's 2026 Standards of Care provide another example. For diabetes screening, the ADA recommends considering a BMI threshold of 23 kg/m² for people of Asian ancestry, compared with 25 kg/m² for the broader adult population when other risk factors are present.

These differences should not encourage people to invent their own personalised BMI categories from online charts. They demonstrate something more fundamental: a universal BMI value does not translate into identical metabolic risk for every population.

Age creates another complication. NICE advises particular caution when interpreting BMI in adults aged 65 and over because illness, functional status, muscle loss and the possible protective effects associated with somewhat higher BMI in some older populations can alter interpretation.

An older adult who unintentionally loses substantial weight and muscle may remain within a conventional BMI category while experiencing an important decline in health. The direction and speed of weight change can therefore be as clinically important as the category itself.

BMI Works Better for Populations Than for Personal Judgements

One of the most useful ways to understand BMI is to distinguish epidemiology from individual diagnosis.

When researchers study hundreds of thousands or millions of people, BMI allows them to group participants using a common measure. Large patterns then become visible. Higher BMI categories are associated at the population level with increased risk of conditions including type 2 diabetes, hypertension and cardiovascular disease, while very low body weight can also be associated with poor health.

WHO continues to use BMI extensively for global surveillance because it allows countries to estimate and compare the prevalence of underweight, overweight and obesity using a standardised method.

Population associations, however, describe probabilities rather than destinies.

If a population group with BMI above 30 has a higher average risk of diabetes than a group with BMI between 18.5 and 24.9, this does not mean every individual in the first group has diabetes or that every individual in the second group is metabolically healthy.

The same statistical principle applies throughout medicine. Blood pressure, cholesterol and glucose thresholds are useful because they identify meaningful risk categories, even though biology does not change instantaneously at one precise value.

Problems arise when a population-level relationship becomes a personal label detached from other evidence.

BMI can tell a clinician that a patient deserves further consideration. It cannot, by itself, tell the clinician everything that consideration will reveal.

When BMI Can Be Particularly Misleading

Some situations deserve more caution than others.

Highly muscular people are the obvious example because substantial muscle can raise body weight relative to height without equivalent excess adiposity. CDC guidance explicitly recommends interpreting BMI alongside other clinical factors in these individuals.

Older adults can experience the opposite problem: low muscle mass and increased body fat without a dramatic change in BMI. Loss of muscle may affect mobility and resilience even when body weight appears stable.

Children and adolescents require age- and sex-specific interpretation because their bodies are still developing. Adult thresholds should not simply be applied to them.

Pregnancy changes body weight through the growth of the fetus, placenta, fluid and other physiological adaptations, making ordinary adult BMI categories inappropriate as a direct description of current body composition.

People with oedema or substantial fluid retention can also gain body weight for reasons unrelated to an increase in body fat.

People experiencing unintentional weight loss deserve assessment even if the final BMI remains within a conventional reference category. A fall from someone's previous stable weight can signal illness, malnutrition or muscle loss.

Finally, people from populations with higher cardiometabolic risk at lower BMI values may require assessment earlier than standard international BMI thresholds alone would suggest.

These examples all point toward the same conclusion: context determines whether BMI is answering the question you actually need answered.

A “Normal” BMI Does Not Guarantee Good Health

One of the most persistent misunderstandings is that a BMI between 18.5 and 24.9 functions as a certificate of health.

It does not.

Someone can fall within that range while smoking, being physically inactive, having uncontrolled blood pressure, carrying substantial abdominal fat or living with high blood glucose or abnormal cholesterol.

Similarly, improvements in health can occur without a dramatic movement in BMI.

A person who begins exercising regularly may become stronger, improve cardiorespiratory fitness, reduce waist circumference and lower blood pressure while losing relatively little total body weight. Resistance training can also preserve or increase muscle while fat mass decreases, making the change on the scale smaller than the change in body composition.

This is why health goals should not be defined entirely through one target BMI.

The article on muscle and body weight provides a useful companion principle: the scale measures total mass, not the quality or function of the tissues making up that mass.

BMI inherits the same limitation because weight is one half of the BMI equation.

At the same time, it would be equally misleading to use these examples to argue that body weight never matters. Excess adiposity can increase the risk of substantial disease, and obesity is now recognised by WHO as a chronic, relapsing disease arising from complex biological, behavioural and environmental influences.

The balanced position is therefore not “ignore BMI.” It is “do not ask BMI to diagnose what it cannot directly measure.”

A Better Way to Use BMI

BMI is most useful when it begins a conversation rather than ends one.

If BMI is high, useful follow-up questions include whether excess adiposity is actually present, where fat is distributed, whether blood pressure or metabolic markers are abnormal, how physically active the person is and whether weight-related symptoms or conditions are present.

If BMI is low, clinicians may consider nutrition, unintentional weight loss, chronic illness, muscle loss and whether the person's weight has changed significantly from their usual level.

If BMI falls inside the conventional reference range, assessment should not automatically stop. Other risk factors can exist independently of BMI.

A practical framework is:

Not everyone needs every measurement. Someone undergoing routine preventive care may need only BMI, waist measurement and standard clinical risk assessment. Another person with substantial muscle mass, unexplained weight change or complex metabolic disease may benefit from more detailed evaluation.

The purpose is not to collect as many numbers as possible.

It is to use the right measurement for the right question.

Frequently Asked Questions

What does BMI actually measure?

BMI measures body weight relative to height. It does not directly measure body fat, muscle mass, fitness or metabolic health.

Is BMI accurate?

BMI is accurate as a calculation of weight relative to height. The limitation is interpretation. It is useful for population research and screening but can misrepresent body fatness in some individuals, particularly those with unusually high or low muscle mass.

Can BMI tell how much body fat I have?

No. BMI is associated with body fat at the population level but does not directly measure it. DEXA and some other techniques estimate body composition more directly.

Can someone have obesity-related health risks with a normal BMI?

Yes. Central adiposity, low muscle mass, metabolic abnormalities and other health risks can occur in people whose BMI is below standard obesity thresholds. This is one reason waist measurements and clinical risk factors can add useful information.

Why are BMI thresholds sometimes lower for Asian populations?

Some Asian populations develop type 2 diabetes and other cardiometabolic risks at lower BMI levels. NICE uses lower practical thresholds for several ethnic groups, while the ADA uses BMI 23 kg/m² rather than 25 kg/m² as part of diabetes-screening criteria for people of Asian ancestry.

Is BMI useless for athletes?

No, but it may be less informative about body fat in highly muscular individuals. A muscular athlete can have a high BMI because of muscle rather than excess fat, so body composition and other clinical measures become more important.

Is waist circumference better than BMI?

Neither measurement answers every question. Waist measurements provide information about central adiposity that BMI cannot capture, while BMI remains convenient and useful for overall screening and population monitoring. Recent research suggests the two measurements can provide complementary information.

Does a normal BMI mean I am healthy?

No. BMI does not measure blood pressure, cholesterol, glucose, fitness, diet, smoking or many other dimensions of health. A conventional BMI range should be interpreted as one part of a broader assessment.

BMI Is Useful Information—Not a Personal Verdict

BMI has survived for so long because it solves a genuine public-health problem remarkably efficiently. Weight and height can be measured almost anywhere, the calculation costs virtually nothing and the resulting index is sufficiently associated with health risk to be useful for research, surveillance and initial clinical screening.

Its limitations are just as real.

BMI cannot distinguish muscle from fat. It cannot show where fat is stored. It cannot measure fitness, strength or metabolic health. It does not account perfectly for differences related to age, ancestry or unusual body composition, and the same BMI can correspond to different levels of risk in different people.

Modern guidance is increasingly explicit about these limitations. The CDC recommends interpreting individual BMI alongside other health information. WHO describes BMI as a surrogate marker and recognises the usefulness of additional measures such as waist circumference. The 2025 Lancet Commission went further, arguing that BMI should primarily serve as a screening and epidemiological measure rather than functioning alone as an individual diagnosis of obesity.

The right response is therefore neither to worship the number nor to throw it away.

Use BMI for the question it answers well: How does body weight relate to height, and does that result suggest that a closer health assessment may be useful?

Then ask the questions BMI cannot answer.

What is the person's body composition? Where is fat stored? Is weight changing? What are the blood pressure, glucose and lipid levels? How strong and physically active is the person? Are there symptoms or diseases that change the meaning of the number?

That is the difference between using BMI intelligently and allowing a screening category to become a health verdict.

Medical note: This article provides general health information and is not a substitute for individual medical assessment. Unexplained weight loss or gain, significant weakness, pregnancy-related weight concerns, eating disorders or symptoms of metabolic or cardiovascular disease should be discussed with an appropriate health professional.