Scientists challenge decades-old assumption that ageing is a steady decline 

A perspective published this month in Nature Reviews Genetics by ARMI researchers argues that the body ages in sharp, concentrated shifts rather than steady decline, and that the statistical tools scientists have built to study it were designed around the wrong assumption. The consequences for how clinical trials are designed and timed could be significant. 

Researchers from the Australian Regenerative Medicine Institute (ARMI) have challenged one of the oldest assumptions in biology: that ageing is a smooth, gradual process. 

The finding is significant because most tools used to study ageing, including many of the molecular ‘clocks’ that estimate biological age from blood samples, are built on the assumption of steady, linear change. If ageing actually happens in sharp concentrated bursts rather than as a linear decline, then an intervention, drug, lifestyle change or screening test may only work within a specific window before or during one of those shifts. 

In a perspective published August 2026 in Nature Reviews Genetics, Dr Robin Grolaux and Professor Nir Eynon from ARMI and Professor Andrew Teschendorff from the Shanghai Institute of Nutrition and Health argue that the human body does not steadily decline with age. Instead, their analysis published earlier this year in Genome Biology found there are several points across a lifespan where the body changes sharply rather than gradually: the first weeks after conception, birth, puberty, the mid-forties to fifties, and a significant change at age 70 to 75. 

Dr Grolaux, who led the research, said the assumption of steady decline had shaped not just the conclusions scientists had historically reached but the instruments they built to reach them. 

“For years scientists have measured ageing the way you might measure a candle burning down at a constant rate,” he said. “What we are increasingly finding is that the body behaves more like a building under accumulating stress. Nothing happens for a long time, and then a support fails and the whole structure shifts.” 

What the evidence shows 

Why systems fail suddenly 

To explain why biological systems can absorb constant damage yet still fail suddenly, Dr Grolaux drew on an analogy from structural engineering. 

“A suspension bridge can lose individual cables for years with apparent minimal effect, because the load is redistributed,” he said. “It is only when that spare capacity runs out that you see a collapse. We think something similar happens at the cellular level. Genetic damage accumulates at a fairly constant rate throughout life, but the body carries spare capacity that absorbs it. What we see as a sudden shift in biology is just the point where that capacity runs out. The damage rate hasn’t changed. The body’s ability to hide it has.” 

Implications for treatment 

That distinction has practical implications for medicine. If ageing-related decline accelerates at specific points rather than progressing evenly, then treatments timed to those windows could be more effective than the same treatment given at a random point in life. Animal studies already suggest that lifespan-extending interventions only work when delivered during certain life stages. 

Dr Grolaux said this should change how researchers think about timing in clinical trials and preventative medicine. “If a drug only works during a narrow window, and you test it across people of all ages without accounting for that, you could conclude the drug doesn’t work,” he said. “We may be discarding treatments that would succeed if only we gave them at the right moment.” 

Better tools needed 

The authors argue that detecting those transition points requires different statistical tools than ageing researchers currently rely on.  

The standard approach draws a straight line through measurements taken across many people of different ages and reads the slope as the rate of decline. The problem with this approach is that a straight line will either miss a sharp transition or flatten it into a misleading average. In that kind of analysis, a spike in biological change at age 45 looks like a slightly steeper gradient across middle age. 

The paper reviews a range of alternatives. Some methods let the model bend with the data rather than forcing a straight line through it. Others hunt explicitly for the point where a trend changes direction, dividing a life course into segments with different slopes. A third approach tracks not the level of a biological measure but how fast it is changing, which can reveal acceleration in ageing that a snapshot comparison between age groups would miss. More complex dynamic models treat frailty as a kind of queue, where damage arrives continuously but the body’s capacity to process and repair it dwindles with age, until the queue backs up. 

Dr Grolaux said the field now needed to commit to methods capable of detecting what linear models had been smoothing away. “Ageing transition points show up consistently across independent studies and biological systems,” he said. “The priority now is validating the tools to detect them reliably, so we can start asking when in a person’s life an intervention will actually work, rather than assuming the answer is the same at any age.” 

This research was funded by Wallonia-Brussels International,the National Health and Medical Research Council of Australia, the Australian Research Council, the National Science Foundation of China, Hevolution and the American Federation for Aging Research. 

Media Contact:
 
Penny Fannin, Communications Consultant, ARMI
 
M: +61 (0) 417125700 
E: penny.fannin@monash.edu