"This study followed nearly 7,000 older Americans for 14 years and reports two things at once. Men who were overweight or had class I or II obesity were around half as likely to die as men in the normal BMI range. At the same time, a large waist raised the risk of dying by about a quarter, independent of BMI. The authors are careful to say this does not mean body fat protects you, and they are right to be careful.

There are good reasons for caution. Height and weight were self-reported, while waist was measured with a tape by trained staff. Older people lose height and tend to report the height they had at 40, which flattens their calculated BMI. So a noisy, biased measure of overall size is being compared against a carefully measured one of belly fat. The apparent benefit also disappears at the top of the range, with no survival advantage at all in the most severe obesity, and in women most of it was not statistically significant. A protective effect that vanishes when you increase the dose is not usually biological.

None of this is permission to put on weight. BMI cannot tell fat from muscle, and after 60 a low BMI often means lost muscle rather than leanness. Protect muscle with resistance training and enough protein. That matters more than ever now that so many people are losing weight fast on the new obesity drugs like Ozempic or Mounjaro, where a good share of what comes off is muscle."

"I think we are going to see more papers like this - observational follow-up studies of people over 65 years. It suggests that it’s better to have a decent amount of fat and muscle in the right places as we age. To me, it makes evolutionary sense. Men and women with high muscle mass and high bone density are at an advantage. What’s abundantly clear is that fat around the waist is not. Men with a beer belly will be in trouble; women with a spare tyre below the belly button should be celebrating, not moaning. I like that!"

"This analysis of almost 7,000 US adults aged 65 and older who were followed for 14 years explored how their body weight and body composition influenced their risk of mortality. It demonstrates the ‘obesity paradox’: the finding that in older adults and some populations with chronic health conditions, higher body mass indexes (BMIs), even in the overweight or obese range, can protect against mortality. However, higher waist circumference, indicating greater visceral fat, was associated with higher mortality risk regardless of BMI status. These findings support recent calls for health assessments to go beyond measuring only body weight or BMI and additionally measure body composition.

An important limitation of BMI in older adults is that scores increase as height decreases, so as we get older, our BMI may increase even without gaining weight or fat mass. Furthermore, because BMI does not distinguish between fat and lean mass, the ‘obesity paradox’ may be explained by the fact that some people classified as overweight or obese have high amounts of lean mass and relatively low amounts of fat mass, while others who appear to be a healthy weight may actually have a high body fat percentage.

This concept is demonstrated in the image below, where a body composition scan reveals that two people of similar body size have markedly different amounts of muscle and fat mass. In summary, these findings are an important reminder for older adults to engage in behaviours that enable them to maintain healthy body composition, including resistance exercise and adequate protein intake to increase muscle mass."

"Like many prior studies, this analysis has demonstrated that having a high BMI in older age is actually associated with less risk of dying, which may seem counterintuitive, especially when the opposite is often found in younger adults.  However, as the authors point out, this should not be interpreted as high levels of fatness being protective of your health but instead reflects the inadequacies of BMI as a measure of body composition- a problem which is exaggerated in older adults. Many other studies also support the interpretation from the authors of this paper that the amount of lean muscle mass and perhaps better nutrition in older people with high body weights may be the driver of their improved lifespan.

However, the standout finding from this analysis is that whilst the relationship between BMI (which reflects total fatness) and mortality is reversed in older age, the relationship between increased waist circumference (which reflects abdominal fatness) and increased mortality remains.  This finding is another reminder that clinical assessment of an individual’s levels of fatness and associated health risk should not rely on BMI alone but needs to include additional indicators of abdominal fatness and lean muscle mass where possible."

"This paper followed up 6900 people over 65 years for 13 years and showed that being overweight was protective and being underweight was harmful for both men and women. In addition, having a higher waist circumference was harmful, with a 5- to 6-fold increase in mortality at the highest waist levels (>60  to 65 inches). All of these findings have been noted before in other studies and meta-analyses. Being obese in males up to a BMI of 40 was associated with protection in men but not in women. This may be due to increased lean tissue in men but not in women at these BMI levels, but we have no data on this as no tissue measurements were available in this cohort, which provided all of their own measurements. Having a high waist circumference in men with a BMI of 30 to 40 was not harmful, but it was not statistically different from the smaller low circumference group.

Clearly, this group who have survived to 65 without dying due to excess fat are a special group who can handle fat and may well benefit from a bit more, but these populations need to be studied in detail with objective measures of weight, fat and lean tissue and biological variables to understand the mechanisms of resistance and protection."

"The headline findings are largely things we already knew. That higher waist circumference predicts mortality independently of BMI was established in 2012, and normal-weight central obesity has been recognised as a high-risk phenotype for at least as long.
Two main concerns:

• Height and weight were self-reported while waist was measured by trained staff, so the comparison between the two measures is not like-for-like. Self-reported height and weight are subject to systematic error that increases with adiposity, whereas waist circumference was measured to protocol. The resulting mismeasurement trends would likely attenuate BMI's associations and inflate waist circumference's apparent independent contribution.
• The study updated body measurements annually until death, and excluded only deaths occurring within a year of the study's start rather than within a year of each measurement. Weight loss from illness in someone's final years is therefore counted against the normal-weight group. Dementia, cancer, heart failure and frailty all cause weight loss well before death, and all reduce muscle while sparing abdominal fat. The pattern the authors identify as their novel high-risk finding is not a surprise to a practising physician.
• Practically, the useful message is unclear. We know weight loss in an older person is often a warning sign, and that BMI alone is a poor guide in this age group. Routinely measuring waists is unlikely to change what we do for patients, as they're already screened for diabetes, hypertension, dyslipidaemia, etc and everyone is encouraged to exercise and move already. For now, that is what the measurement would be prompting."

"The takeaway from this study is that abdominal obesity, which implies an excess of visceral fat that sits around the organs, is associated with a higher risk of mortality even in individuals whose weight is considered within a normal range. This has been demonstrated in multiple studies now. BMI, on the other hand, calculated using height and weight, is a poor indicator of health and mortality, unable to explain body composition, fat distribution, or organ function. Some athletes can fall into high BMI ‘obese’ categories purely due to their body composition.

This study also showed that overweight and obesity, but not severe obesity, were associated with a lower risk of death in older individuals. This phenomenon has also been shown elsewhere. However, it should be interpreted with caution as BMI on its own is a poor indicator of health. Weight loss can occur with some chronic diseases such as cancer, a leading cause of death in older individuals. This and other similar studies do not look at participants’ disease status. Consequently, an individual in the ‘healthy’ BMI category against which all others are compared may in fact be a former obese individual who has lost weight due to a chronic illness."

“This study adds to a long-standing puzzle in ageing research. Some studies suggest that carrying extra body weight appears protective in older age, while a larger waist does not. My own earlier research in older adults found a similar pattern, with little relationship between body weight and survival.

The most likely explanation is that people often lose weight when they become seriously unwell, which can make lower body weight look riskier and higher body weight look safer than they really are, not because the weight itself is protective.

Interestingly, when this study looked at body weight and waist size together, the apparent protective effect was seen only in men and did not follow a consistent pattern across weight categories. That inconsistency suggests these simple measures do not reliably capture what is happening inside an older body, where muscle is often being lost while body fat increases, even when overall weight changes very little.

These findings should not be read as evidence that weight loss is harmful. Strong evidence from bariatric surgery and newer weight-loss medications shows that safe, intentional weight loss can improve health and extend life in people living with obesity.

The real message is that body weight and body size measures alone may not be a reliable guide to health in older adults, and findings like these should be interpreted with caution rather than taken at face value.”

“This US study reminds us that BMI is a crude screening tool, especially in later life. The finding that higher BMI was associated with lower mortality should not be interpreted as evidence that excess body fat protects older people. This observational study cannot establish cause and effect. Lower weight may reflect frailty, sarcopenia, illness or unintentional weight loss, and the analysis did not account for chronic disease burden, physical activity, diet or muscle mass. The association was stronger in men, and several associations for women with obesity were not statistically significant. The percentages reported are relative differences in mortality hazard (the rate at which deaths occurred over time), not absolute chances of dying.

From a behavioural science perspective, the wrong interpretation would be to encourage older Australians to gain weight based on BMI alone. Research on weight-loss maintenance, including my own, shows that sustained behaviour change is shaped by cognitive, emotional and contextual factors. Health goals should be individualised and focus on strength, mobility, adequate nutrition, cardiometabolic health and sustainable physical activity. The key message is that BMI and waist circumference provide complementary information, but neither can determine a person’s overall health on its own.”

Dominika has also included some brief quotes:

“Lower weight in later life does not necessarily mean better health; it may reflect muscle loss, frailty, illness or unintentional weight loss.”

“For older adults, the goal should not be to chase a particular BMI category, but to protect strength, mobility, nutrition and cardiometabolic health.”

“This study identifies associations. It should not be used to tell older people whether they should lose or gain weight.”

“Waist circumference and BMI tell us different things, but neither tells us everything about a person’s health.”

“The Australian takeaway is not that extra weight is protective; it is that older people’s health needs to be assessed more comprehensively.”

"I think this is an interesting study, particularly the associations observed for BMI and waist circumference (WC). Higher BMI was associated with lower mortality, whereas higher WC was associated with an increase in mortality risk. These findings highlight limitations of BMI as a measure of health and adiposity, as it does not distinguish between fat mass and lean mass. This may be particularly important in older adults, where ageing is accompanied by muscle loss and increasing frailty. Hence, the inverse association between BMI and mortality should be interpreted cautiously, as lower BMI may partly reflect underlying illness, frailty and unintentional weight loss, rather than indicating that body size itself is protective.

These findings also raise questions about  BMI cut-points across diverse populations. Although global thresholds are useful for population comparisons, differences in body composition and fat distribution mean that the same BMI may represent different health risks across populations. This is relevant to Australia given its diversity, including Aboriginal and Torres Strait Islander peoples and an increasingly diverse migrant population. More importantly, an individual's health should not be simply reduced to their weight or BMI. A more nuanced approach incorporating WC, body composition, physical capacity and metabolic health may provide a more meaningful assessment of health risk."

"This is an interesting study because it gives a different message about what is means to be a healthy weight in the elderly.

They found that older people who were overweight or obese actually lived longer.

In this day and age, it has become normal for adults to be obese. But this may not apply to the elderly. 

As people age they lose weight and their muscles and bones become weaker. It appears that if a person is able to retain their middle-aged body weight into old age, this may be a sign of good health. If they remain active, their bones and muscles will still be strong. On the other hand, weight loss is associated with increasing frailty and a higher risk of death. 

So the message we should be giving is that weight loss among the elderly is bad news. We should encourage our elderly friends and relatives to continue to eat a healthy, hearty diet, to remain active and continue to think of themselves as middle-aged for as long as they can."

"I think this is an important study because it adds to growing evidence that BMI on its own may be an incomplete indicator of health risk in older adults. What is particularly valuable is that the researchers examined BMI and waist circumference together, using repeated measurements over long-term follow-up in a nationally representative sample of older US adults. BMI cannot distinguish fat from muscle or indicate where fat is stored. In later life, low body weight or unintentional weight loss may also reflect frailty, sarcopenia or underlying illness, which can complicate the relationship between BMI and mortality. This makes the waist circumference finding particularly relevant, as central adiposity may identify risk that BMI alone misses.

I would therefore be cautious about interpreting the findings as evidence that gaining weight is protective. This remains an observational study, so it cannot establish causation, and reverse causation and residual confounding remain possible. For Australians, the broader message is not to disregard BMI, but to interpret it alongside waist circumference, recent weight change, physical function and overall clinical context. Measures of muscle mass and strength can also provide additional information when assessing health and mortality risk in older adults."

"This study adds to the evidence behind the Lancet Commission on Clinical Obesity, which determined that BMI should be considered alongside waist measurement. This is especially important because this is an older group, where these issues often catch up with a person and can lead to health problems. It is also useful because waist circumference is the type of measure that can be obtained at a GP visit and used to inform health decisions.

This is a reminder that patients and clinicians should move away from only relying on BMI, or a number on a scale, as an indicator of health. The relationship between health and weight is much more complex. Issues such as exercise frequency, diet and mobility all factor into someone's waist circumference. These lifestyle issues around overall wellbeing should be the focus rather than a single number.

In the current climate of GLP-1 medications, we must avoid putting weight loss on a pedestal and remember that weight alone doesn't determine someone's health. Discussions with a GP or other health professional about these drugs should include psychological wellbeing, physical activity and nutrition.

Our environments aren't designed around the factors this study shows reduce mortality. It is hard for older adults to stay active when we prioritise cars over encouraging walking, and the food industry pushes quick, unhealthy options over balanced meals. Voters have the power to tell policymakers that their health should come first."

"The findings of this study highlight the limitations of using BMI to predict mortality risk in older adults. Although higher BMI appeared protective, these findings should be interpreted with caution. While higher BMI is associated with increased mortality at a population level, many previous studies included relatively few older adults. Waist circumference may be a better predictor of mortality than BMI, and has been incorporated into definitions of obesity, including the recent Lancet Commission definition.

The authors acknowledge limitations of this study, and the finding that older adults with class III obesity (BMI >40) do not have an increased mortality in this study highlights this. Lower mortality was linked to overweight or class I and II obesity in men but not women. Increasing age is associated with greater prevalence of sarcopenia, which may explain why waist circumference was more strongly associated with mortality than BMI. Survivorship bias may also influence the findings, as individuals with higher BMI may be less likely to survive to older ages. Additionally, chronic diseases and cancer can result in lower BMI at baseline. Therefore, these findings should be interpreted with caution, particularly given the 14-year follow-up and inclusion of participants aged 85 years or older."

"The data reported here are from one of the first studies to be underpinned by the 2025 Lancet Commission on Clinical Obesity framework by looking closely at the relationship between body mass index (BMI) and waist circumference (WC) in all cause risk of death amongst older adults residing in the US. A real strength is that it followed participants across a14-year period. This provided a far more accurate, moving picture of how body changes actively influence mortality.

Interestingly, the authors found that the males in their study who were overweight and obese—even when carrying excess fat around the waist—had an overall lower risk of dying compared to healthy weight men. This same protective effect was seen in overweight women but not obese women. Now, I want to point out that these results are not a reason to dismiss maintaining a healthy weight.

There are a few fundamental flaws with this study that should be pointed out.

Firstly, even though waistlines were measured by trained professionals, the participants' height and weight were completely self-reported and thus may not be truly indicative of the ‘real’ values. Older adults lose height naturally as they age due to spinal compression, but they rarely update their known height. If a person reports being taller than they actually are, their calculated BMI will look artificially lower.

This misclassification skews the numbers and weakens the study’s reliability. Secondly, men typically preserve more lean muscle as they get older, which acts like a protective ‘metabolic booster’ compared to women of the same age. Conversely, older people with a "normal" BMI might be experiencing muscle wasting and frailty. Ultimately, despite its strengths, the study’s findings are not proof that excess body weight is biologically protective; rather, they reflect BMI’s inability to measure healthy muscle alone, which clinicians already knew."

"It’s great to see that this review of 7000 individuals aged 65yo and older in an American database, has generated so much interest and discussion. Are the findings from this observational study unexpected-not really. We know that BMI is the least helpful measure of adiposity (carry excess body fat) and it definitely doesn’t tell you how “sick” or at-risk the person is. 

In January 2025, the Lancet Commission published on this very point, making it clear that it is the location of the excess body fat that matters most. Waist circumference is often used as a surrogate marker for visceral adipose (fat that is tucked in and around the vital organs such as the heart, liver, kidneys etc which is already known to be harmful to one’s health). 

Measuring waist circumference however can’t distinguish between fat under the skin (subcutaneous) vs visceral fat (explained above); this distinction is important especially since various ethnic populations are known to carry more visceral fat for the same waist circumference or BMI eg South Asians, East Asians. Conversely, people from African ancestry carry more subcutaneous fat for the same waist size/BMI.

To make things even more interesting, we have this paradox in people from Maori and Pacific Island populations who (i)have high visceral fat/ internal fat storage but (ii)they are generally leaner with greater skeletal muscle mass and higher bone mineral density (ie more muscle & heavier stronger bones) than Europeans, and so overall have lower “total body” fat.

Bottom line, body composition analysis is more meaningful. It’s where the fat is located, ie visceral vs subcutaneous AND the cumulative lifetime exposure to harmful visceral fat..ie a person who starts living with overweight obesity from age 7, has a different life trajectory than someone who starts living with obesity, from the age of 40… 

This study only looks at one snapshot in time, and didn’t provide that nuanced information, to help put things into context.

The important take home message/call to action:  as part of routine medical assessments ie your usual health check up which includes Blood pressure check, and blood test review etc… measurement of body size (preferably body composition analysis eg Bio-Impedance Analysis [BIA] scale or dual energy x-ray absorptiometry (DXA scan); at bare minimum one of the measures as displayed here on the image PLUS, most importantly and assessment of how the adiposity (excess harmful fat) is affecting the person’s health (physical, emotional) function and wellness."