Generation 100 trial

Generation 100 is a Norwegian trial that assigned nearly 1,600 mostly healthy adults in their seventies to five years of structured exercise, and the only trial to test whether training healthy people lowers the death rate. It found no clear difference overall: every group, the comparison group included, held its fitness steady, and the trial was too small to settle mortality.

Read it as "not shown," not "shown not to work": hard interval training had the fewest deaths of the three groups, but by a margin the trial could not confirm. It is good evidence that fitness can be held steady through your seventies, with hard intervals keeping it slightly higher, and no evidence either way for younger adults.

What the trial did

Of 6,966 people in Trondheim, Norway, born between 1936 and 1942, 1,567 took part. Aged 70 to 77 and about half of them women, they were randomised to one of three groups for five years:^[1]

High-intensity intervals: two sessions a week at about 90% of peak heart rate (400 people).
Moderate continuous training: two sessions a week at about 70% of peak heart rate (387 people).
Control: follow the national physical-activity guidelines on your own (780 people).

The two training groups were offered supervised group sessions, but only about half kept to the programme. The main question was whether training lowered deaths from all causes.

What it found

Deaths from all causes did not differ between the two training groups combined and the control group.

Split by group, the picture is more interesting but no more certain. About 4.7% of the controls died, against 3.0% in the interval group: roughly a third fewer deaths, but with a range of plausible effects running from about two-thirds fewer to about a fifth more, so chance cannot be ruled out. The moderate group did slightly worse than the controls, also within the range of chance. Interval training against moderate training came closest to a clear difference, about half the death rate, and still did not reach it.

Fitness did not rise so much as hold. Peak oxygen uptake did not decline in any group over the five years, the controls included, so the expected age-related fall was prevented across the board. The interval group ended slightly fitter than the controls, by 0.7 mL/kg/min. Heart disease and cancer did not differ between the groups.

Why the result is hard to read

Three features blunt the comparison.

The control group was not sedentary. At the start, 87.5% of participants called their health good and 80% reported medium or high activity. Over the trial, the controls chose to do more of their activity as hard intervals than the moderate group did, so their exercise ended up between the two training groups in intensity. The authors also point to healthy-volunteer bias: people who sign up for a five-year exercise trial are unusually well.

It was too small for the question. With deaths in the low single-digit percentages over five years, even a one-third reduction falls inside the margin of error.

It tested one age group. Everyone was in their seventies and mostly well. Nobody has run the equivalent trial in younger adults, so the trial says nothing about whether raising a 40-year-old's fitness adds years.

How it fits with the rest of the evidence

Observational studies find that fitter people live markedly longer, and VO₂ max covers that evidence. Generation 100 is the closest anyone has come to testing the step those studies cannot: whether training for fitness extends life. The other trials that came near it were in people already ill. In heart failure, training made essentially no difference to deaths;^[2] a weight-loss and activity programme in type 2 diabetes raised fitness but was stopped early because heart attacks and strokes were not falling;^[3] and pooled cardiac-rehabilitation trials show a likely small drop in deaths from all causes that fades with longer follow-up, though deaths from heart disease fall by around 40% over the long term.^[4] That is a gap in the evidence, not a case against training.

Safety

The trial excluded people with conditions that make hard exercise risky, such as uncontrolled high blood pressure, unstable angina, heart failure and severe heart-rhythm problems. No heart events occurred during supervised sessions; three participants broke bones slipping during outdoor training.

What to do

Train at any age. Fitness held steady in people starting in their seventies; the case for exercise does not rest on this one mortality result.
Include some hard intervals if your health allows, and check with a clinician first if you have unstable heart disease. The interval group had the fewest deaths and the highest fitness, even if the trial could not confirm the difference. How to build them is in Training VO₂ max.
Don't read the null as permission to skip it. The comparison group was active too; the trial compared exercise with exercise.