BMI is failing to detect dangerous obesity as people age

By Dr. Liji Thomas, MDReviewed by Lauren HardakerNov 19 2025

A landmark analysis of 120,000 people reveals that hidden central fat is rising across the UK, and BMI is missing it, making waist-to-height ratio the clearer warning signal for growing obesity risk.

Study: Long-term trends in central obesity in England: an age-period-cohort approach. Image credit: Studio Romantic/Shutterstock.com

Several measures are used to track obesity, both to understand its prevalence and to predict its health impact. A recent study published in the International Journal of Obesity attempted to track the long-term changes in central obesity in the UK, stratifying by birth cohort, age, and time.

Obesity measurement gaps

The body mass index (BMI) has long been the go-to measure for identifying obesity and overweight individuals. However, it probably underestimates the prevalence of high-risk obesity when compared to measures of central obesity. These include the waist circumference, waist-to-hip ratio, and waist-to-height ratio (WC, WHR, and WHtR, respectively), all of which provide early warnings of possible cardiovascular and cardiometabolic health risks.

The BMI captures the total weight of the body, irrespective of whether it is derived from muscle mass, other lean mass, or fat mass. This is especially relevant for children and older people who have a different body composition compared to mature adults.

Using the BMI alone could potentially mislabel one in ten of the UK population as obese. Conversely, one in four high-risk obese people has a “healthy” BMI and is misclassified as being low risk. Moreover, the BMI overdiagnoses obesity among children.  

Again, the BMI changes over time, making follow-up essential to determine the associated health risks, especially among older individuals. Finally, the variation in BMI by age and sex also makes it less easily interpretable at the expert and population level.

These limitations prompted the NICE to suggest the use of uniform WHtR values in adults and children instead, since this better predicts cardiovascular risk in adults. It also outperforms dual-energy x-ray absorptiometry (DEXA) in measuring truncal and total fat mass in children, and it better correlates with the prevalence of fatty liver and fibrosis in both children and adults.

NICE currently recommends using both BMI and WtHR in individuals with a non-obese BMI (<35 kg/m²). Newer definitions of obesity are being developed to more accurately reflect health risks and identify the need for intervention. The European Association for the Study of Obesity (EASO) has published a new protocol for diagnosing, staging, and managing obesity.

Despite the wealth of knowledge about changing obesity trajectories over time in the UK, there is a lack of understanding about the direction of central obesity measures over time. The current study aimed to examine this variation while analyzing the results for the effects of age, time, and generational cohort.

Measuring obesity over time

The researchers used data from the Health Survey for England (2005–2021), covering 120,024 individuals aged 11 to 89 years. Their birth years ranged between 1919 and 2008. The study period spanned 16 years and 17 birth cohorts, including participants born in successive five-year spans.

The study used central obesity measures, namely, the WC, WHR, and WHtR, as well as the BMI. High risk thresholds were set for each, based on the definitions provided by the World Health Organization and the UK National Institute for Health and Care Excellence.

The participants were analyzed for obesity effects by age, period, and birth cohort (an APC analysis).

Emerging obesity trends

The researchers found time-dependent increases in high-risk obesity and central obesity irrespective of the obesity measures used. High-risk BMI increased from ~23 % in 2005 to ~27 % in 2021. Similarly, the high-risk WHtR increased from ~24 % in 2005 to 33.4 % in 2021.

High-risk WC also increased at the same rate, from ~38.5 % in 2005 to ~49 % in 2021. The highest prevalence of high-risk obesity was obtained with the WHR, from ~46.3 % in 2005 to 61 % in 2021.

How obesity changes with age

The prevalence of all central obesity measures, except WHtR, increased linearly with age until 65-70 years, then slowed thereafter. The WHtR increased from 11 years onwards to the age of 85, fluctuating more markedly thereafter.  

In contrast, the BMI-age graph exhibited an inverted U shape, indicating an early rise in BMI with age. After steadying at around 50 years, it began to decline. However, the risk of obesity-linked adverse health impacts increases rather than decreases with age. Thus, the BMI fails to identify this group of individuals at risk accurately.

By the age of 85 to 89 years, women and men had nearly five-fold and six-fold the odds of high-risk obesity, respectively, when compared with the baseline values at 18 to 19 years.

While similar increases occurred for all measures in both sexes, for men, the increase in the odds of high-risk WHtR increased steadily until 80 to 84 years, decreasing over the next five years, unlike the linear increase until 89 years for females. Similar differences were seen for high-risk WC, with females showing increasing odds until 80 to 84 years, but men only until 55 to 59 years. High-risk BMI odds also increased in men until 50 to 54 years, but in women until 65 to 69 years.

Obesity risk changes over time

When compared across periods, the odds for all measures of high-risk central and general obesity were somewhat higher for both women and men in 2019 to 2021 compared to 2005 to 2006.

Comparing birth cohorts

Significant differences were largely absent between birth cohorts. Only females born between 2004 and 2008, and men born after 1974, had slightly lower odds of high-risk central obesity.

Rethinking obesity measures

The study is the first to perform an APC analysis of general and central obesity measures. The linear increase in high-risk obesity with age agrees well with the observation that age increases the likelihood of obesity-related disease.

Age plays a significant role in driving obesity trends meaning an aging population could leading to further increases in the prevalence of obesity.

The findings underline the superiority of central obesity measures, particularly WtHR, in evaluating time-dependent obesity risk. The results suggest that both types of measures identify different risk groups at the population level, but not for individuals.

Currently, the WHtR appears to excel at measuring obesity more accurately than the BMI, and it should be used as a standard in clinical situations. This aligns with updated NICE guidance, which recommends WHtR alongside BMI rather than as a complete replacement.

In view of these findings, early interventions targeting children and adolescents are indicated to support healthy aging. Further work should examine other evidence-based obesity thresholds with the potential to yield the actual obesity prevalence with comparable accuracy across different groups. 

Download your PDF copy now!