Not Everything Fits into a Label: What BMI Reveals—and Conceals—About the Risk of Chronic Diseases

Body mass index (BMI) remains one of the most widely used indicators for estimating health risks, although its explanatory capacity has limits that are worth understanding.

A couple of years ago I went for a company medical check-up. As usual, they weighed me and measured my height. The result: a body mass index (BMI) of 26 kg/m2 and the doctor’s recommendation—in line with the World Health Organization tables—to lose weight because I had a slight excess weight (defined as between 25 and 29.9 kg/m2)1.

I remember thinking: “It’s only 26 and I don’t have any serious health problem!” In the end, everything seemed to come down to a single number. Is there really an elevated risk of health problems in all people with slight excess weight?

At that moment I was working on a project on multimorbidity, so I did not hesitate to look into it in greater detail.
That said, we cannot deny that obesity has become one of the major public health concerns of recent decades. In the Spanish adult population, the prevalence of obesity (that is, the proportion of people who have it within a population) rose from 7.3% in 1987 to 15.7% in 2020. Although in recent years this increase seems to have stabilized, obesity continues to be associated with numerous chronic diseases and with a poorer quality of life.

However, the way we usually study this relationship has an important limitation. Most research classifies the population into standard BMI categories: underweight, normal weight, overweight, and obesity. These are useful categories as a general reference, but they simplify a more complex reality: not all people follow the same pattern, nor does risk begin to increase exactly at the conventional boundaries of the categories established to identify increased health risk. For that reason, instead of grouping people into large blocks, it is useful to observe what happens across the entire BMI continuum.

Beyond the labels

Using six national health surveys conducted in Spain between 2006 and 2023, we analyzed the prevalence of 23 chronic diseases in the population aged 20 to 84 and their relationship with BMI. These included musculoskeletal, respiratory, cardiovascular, metabolic, intestinal, and mental health conditions.

We first analyzed how these diseases were distributed according to the usual BMI categories. In general, prevalence was lower among people with normal weight and higher in the obesity categories, but that view concealed many nuances.

For that reason, in a second stage we treated the exact BMI value, with values within the range from 16 kg/m2 up to 40 or more kg/m2, using a technique that made it possible to identify more precisely at which levels the prevalence of each disease was lower or higher.

The results show that the relationship between BMI and chronic diseases and conditions is far from uniform. In some cases, such as diabetes or stroke, prevalence increases rather continuously as BMI rises. In others, such as hypertension, heart disease, or osteoarthritis, risk remains low at normal values or even somewhat lower ones and increases clearly from a certain threshold onward.

We also found other patterns. Some diseases, such as chronic skin problems, remain stable until overweight is reached and then increase rapidly. Others, such as high cholesterol, prostate problems, gastric ulcer, or menopausal problems, increase with BMI up to a certain point and then tend to stabilize.

Perhaps the most striking cases are those that, when plotted on a graph, draw a “U” shape in which prevalence is higher at both the lowest and the highest BMI values. This pattern appeared in respiratory diseases (COPD) and asthma, as well as in mental health problems, migraine, and allergies. In other words, both underweight and obesity seem to be associated with a higher frequency of these conditions.

We also observed important differences according to sex or age. Women show a higher prevalence than men in several chronic diseases across almost the entire BMI range, especially in osteoporosis, back pain, migraine, varicose veins, allergies, and mental health problems. Age also marks clear contrasts. As might be expected, older people concentrate a higher prevalence of most diseases, although there are exceptions: migraine is more frequent at Middle Ages and allergies stand out more among younger adults.

What this can contribute to public health

These results are not only of academic interest. If each disease is related to BMI in a different way, prevention policies should also be more precise. Broad categories are practical and easy to communicate, but from a clinical and preventive point of view they may fall short. Knowing that the risk of a disease begins to increase at a given BMI level—and not necessarily at the classic threshold of overweight or obesity—can help detect problems earlier, better guide interventions, and better plan health care resources.

Looking beyond the label

In public debate, BMI usually appears as a figure that classifies people into closed groups. But our results suggest that it is useful to look beyond that label.

The relationship between body weight and chronic diseases is neither uniform nor simple. Sometimes risk increases gradually. Sometimes it shoots up only from a certain point onward. And sometimes problems appear at both the lowest and the highest levels.

So, if we want to better understand how weight is related to health and design more useful policies, perhaps the time has come to stop thinking only in categories and begin to think also in trajectories and thresholds.
Because, in the end, between “normal weight” and “obesity” there is much more story than it seems.

And yes: after doing this work, I have to admit that that doctor was probably partly right when he noted in his report that I should lose some weight. Although with a BMI of 26 I had no comorbidities, our results suggest that, in many of the diseases analyzed, the minimum prevalence is located below that value and that, from there on, the association begins to increase.

*BMI is obtained by dividing weight (in kilograms) by height (in meters) squared. In general terms, it is considered underweight when it is below 18.5; normal weight, between 18.5 and 24.9; overweight, between 25 and 29.9; and obesity, from 30 onward.

References

Cámara AD, Spijker J (2010). Super-size Spain? A cross-sectional and quasi-cohort trend analysis of adult overweight and obesity in an accelerated transition country. Journal of Biosocial Science 42(3):377–93.

Spijker J, Castro-Prieto P. (2024). The prevalence of chronic diseases according to Body Mass Index categories and integer. Paper presented at the Health, Morbidity and Mortality Working Group (EAPS) meeting, Bilbao, September 25–27. https://hmmwg.vse.cz/wp-content/uploads/page/300/S5_Spijker.pdf