← Longevity & Supplement Guides
Fitness Peaks at 35: Why the 47-Year Study Is Better News Than It Sounds
This week a 47-year study made the rounds with a sentence designed to ruin your morning: your fitness and strength start to decline at about 35, even if you exercise. The press release came from Karolinska Institutet, the coverage followed, and somewhere a person who had just renewed a gym membership at 41 read it on the train.
I read the paper because the headline felt like a verdict. It is not. The study is careful and genuinely rare: the same people, tested again and again, from age 16 to age 63. But the most useful thing it found is not the age of the peak. It is what exercise did to the shape of the curve, and what it did not do.
The short version: exercise did not stop the decline, and it did not obviously slow it. It lifted the whole line. That sounds like a consolation prize. Read to the end and you may decide it is the better deal.
What the Swedes actually measured
The study, led by Maria Westerståhl and colleagues in the Journal of Cachexia, Sarcopenia and Muscle, comes from the Swedish Physical Activity and Fitness cohort. In 1974, 427 randomly chosen 16-year-olds born in 1958 were tested at school. The researchers came back at ages 27, 34, 52 and 63. By the last visit, 8% of the group had died; the models still used everyone's data.
Four things were measured. Aerobic capacity, estimated from a submaximal cycling test. Muscular endurance, as bench-press repetitions at a fixed rhythm with a light bar (20 kg for men, 12 kg for women). Power, as a standing jump. And leisure-time activity, asked as a single yes-or-no question.
That last detail matters later. "Are you physically active in your free time?" is a blunt instrument. It cannot tell a weekly walk from five hard runs, which means the study can see that activity matters but not how much of what kind.
The slope is gentle, and then it is not
Aerobic capacity and muscular endurance peaked between 26 and 36. Absolute aerobic capacity topped out at 35 in men and 36 in women, which is where the "35" headline comes from. After the peak, the loss is almost invisible at first: around 0.3 to 0.6% a year. Around age 40, absolute aerobic capacity was falling by about half a percent a year. By 63, the rate had climbed to roughly 2 to 2.5% a year on every measure.
So the honest summary is not "decline starts at 35". It is "decline starts quietly at 35 and gets louder after 50". From peak to 63, the group lost about a third of its aerobic capacity and bench-press endurance, and more than 40% of its jump height. Men and women declined at the same rate.
None of this is new to physiologists. The Baltimore Longitudinal Study of Aging, which followed 810 healthy adults with repeated treadmill tests, found the same acceleration in 2005: a loss of 3 to 6% per decade in the 20s and 30s, rising to more than 20% per decade from the 70s. What the Swedish study adds is the long view of one group of ordinary people, starting in a school gym.
47 years, one cohort
How much was lost from peak by age 63
Bar length = percentage lost from peak to age 63, on a shared 0–50% scale. Values from Westerståhl et al., J Cachexia Sarcopenia Muscle 2025 (Table 2). Power peaked earliest and fell furthest.
Exercise did not bend the curve. It lifted it.
Here is the part the headlines compressed into "even if you exercise". People who were active at 16 had higher capacity on all four measures for the rest of their lives. People who became active as adults gained roughly 6 to 7% in aerobic capacity, 11% in bench-press endurance and 4% in jump height, compared with staying inactive.
But when the researchers tested whether activity changed the rate of decline with age, the answer in their own statistics was no: the interaction was not significant. The authors write that the later decline appears modifiable by activity. Their numbers show something more precise. Being active moved the whole curve up, and the curve then fell at about the same speed.
Baltimore found exactly this twenty years earlier. Regular exercisers declined as fast as everyone else, but because they started higher, they stayed higher at every age. The lead author admitted to being a little disappointed. I understand the feeling, and I think it is misplaced.
Why height beats slope: the floor
A slope you cannot change is not a sentence. It is a timeline. What matters is where you are on it when you reach the level at which ordinary life starts to feel like exercise.
The physiologist Roy Shephard put numbers on that level in a review in the British Journal of Sports Medicine: independent living tends to become hard when aerobic capacity falls below about 18 ml/kg/min in men and 15 in women, typically somewhere in the early 80s. Climbing stairs, carrying shopping and walking uphill all draw on the same reserve. His argument was that regular aerobic training could keep people above that line for something like ten more years.
That is the real arithmetic of the Swedish study. If two people lose capacity at the same rate and one starts higher, the higher one reaches the floor later. The lift you build at 45 is not cancelled by the slope; it is carried down it.
Schematic
Same slope, different height: who reaches the floor first
Schematic, not plotted data: the shape follows the accelerating decline reported by Fleg et al. (2005) and Westerståhl et al. (2025), with activity shifting the curve up rather than changing its slope. The floor follows Shephard (2009). Exact ages differ from person to person.
Reading 1: “Exercise can’t stop ageing”
True, and nobody serious claimed it could. Active people in both studies still lost capacity every decade after their peak.
Reading 2: “So why bother after 35?”
Because adults who became active gained capacity in this study, and in Baltimore the higher starting point was kept at every later age. The lift is carried forward.
Reading 3: “The slope is fixed, the height is yours”
The closest fit to both datasets. With a single yes-or-no activity question, the Swedish study cannot rule out that harder or more regular training changes the slope a little. It simply did not show it.
At 63, your birth year explains less than you think
The most interesting number in the paper got almost no coverage. Between age 16 and age 63, the spread between people grew enormously: about 25-fold for aerobic capacity per kilogram, about 5-fold for jump height, and about 3-fold for bench-press endurance.
At 16, everyone born in 1958 was roughly similar. At 63, the same birth year held people far below the average for their age and people far above it. Age sets the slope. Four decades of habits, body weight, health and luck set the height, and by the sixties the height is the bigger story.
This is also why population averages make poor personal forecasts. You are not the average curve. You are one line in a fan that keeps opening.
Power goes first, so train it on purpose
Jump height peaked earliest, at 27 in men and 19 in women, and fell the furthest: 41% and 48% by 63. Power is force produced fast, and it is what you use to stand up quickly, catch a stumble or take stairs two at a time. Slow, steady exercise does little for it.
The practical answer is not dramatic. Do some movements with intent: rise from a chair briskly, push the lifting phase of a squat fast and lower it slowly, climb a flight of stairs at a quick pace. We wrote about this distinction in muscle power versus strength, and the Swedish data are a good reason to take it seriously from your forties onward.
What to do this week
First, stop reading "35" as a deadline. The decline at 40 is about half a percent a year. That is small enough for a few months of training to outrun for a while, and adults in this cohort who became active gained measurable capacity.
Second, build height on both axes. Aerobic work two or three times a week, including some harder efforts (our guide to VO₂max and longevity covers how), plus two sessions of strength training with a few fast repetitions in each.
Third, watch your own line, not the average one. A timed stair climb, a chair-stand count or your heart rate at a familiar walking pace, checked once a month, says more about you than any cohort curve. The Agen Band tracks resting heart rate continuously, and a slow downward drift over weeks of training is one of the clearer signs that the lift is happening. A wearable's VO₂max number is an estimate; read it as a trend, not a score.
If you have a heart condition or have been inactive for a long time, talk to your doctor before adding hard efforts.
The bottom line
The Swedish study is good science and a slightly misleading headline. Fitness does peak in the late twenties to mid-thirties, the loss starts quietly and speeds up after 50, and exercise does not stop it. What exercise did, in this cohort and in Baltimore before it, was raise the whole line, at any age people started.
That is the deal on offer: you cannot choose the slope, but you can choose the height you carry down it. Measured in years above the floor, it is a generous one.
