The Body's Quiet Inflection Point — What a 47-Year Swedish Study Actually Tells Us About Fitness and Ageing

TL;DR

• A 47-year longitudinal study from Sweden's Karolinska Institutet — one of the longest-running fitness studies ever conducted — found that physical capacity (strength, endurance, cardiorespiratory fitness) begins declining around age 35, and the decline accelerates with each passing decade.
• The same individuals were tested repeatedly from age 16 to 63. This is not cross-sectional guesswork. It is the same bodies, measured over nearly half a century.
• The finding that matters more: adults who became physically active later in life — after years of inactivity — improved their physical performance by 5–10%. The body responds. It is not a one-way street.
• Published 15 May 2026 in the Journal of Cachexia, Sarcopenia and Muscle. Lead author: Maria Westerståhl, Karolinska Institutet.
• The cohort will be tested again next year at age 68. The study is not finished.

What Happened

On 15 May, a team at Sweden's Karolinska Institutet published findings from the Swedish Physical Activity and Fitness study (SPAF) — a project that has been running since 1979. Several hundred randomly selected Swedish men and women were first tested at age 16. The same individuals were brought back for repeated measurements of fitness, muscular strength, and endurance across 47 years, most recently at age 63.

The headline finding: physical capacity starts declining around age 35, and the slope steepens with age. This held across people with different training backgrounds. Fitness, strength, and endurance all followed the same downward arc.

But the finding the researchers themselves emphasised — and which most coverage underplayed — is the reversibility. Participants who became active during adulthood, even after years of low activity, improved their physical capacity by 5–10%. The body at 50 is not the body at 25, but it is also not a closed system. It responds to stimulus.

The study was published in the Journal of Cachexia, Sarcopenia and Muscle (DOI: 10.1002/jcsm.70134). The researchers plan to test the same cohort again in 2027, when participants will be 68.

What It Actually Means

The 35-year-old inflection point is not new. Exercise physiologists have known for decades that VO₂ max, muscle mass, and power output begin declining in the mid-30s. What makes this study different is the methodology — and what the methodology permits us to say with confidence.

Most ageing-and-fitness research is cross-sectional: researchers compare 30-year-olds to 60-year-olds at a single point in time and infer a decline curve. The problem is that cross-sectional data confuses ageing with cohort effects. The 60-year-olds in a 2026 study grew up in a different nutritional, medical, and occupational environment than the 30-year-olds. You are not measuring pure ageing; you are measuring ageing plus history.

SPAF eliminates this problem by following the same people. When a 63-year-old tests lower than they did at 35, you know it is their decline, not a generational artefact. That is what makes the 35-year inflection point credible in a way most previous estimates were not.

But the study's real contribution is the reversibility data. A 5–10% improvement from late-life activity adoption is not trivial. To put it in context: a 10% improvement in VO₂ max is associated with a roughly 15% reduction in all-cause mortality in epidemiological studies. For someone who has been sedentary for two decades, that is the difference between functional independence at 75 and needing assistance with stairs.

The mechanism question — why does everyone peak at 35, and why does exercise slow but not halt decline — remains open. Westerståhl's team is now investigating the biological pathways. The 2027 follow-up at age 68 will be the next major data point.

Hype Deconstruction

Some coverage has framed this as "your body starts falling apart at 35." That is not what the study says.

The decline is gradual, not cliff-like. A 36-year-old is not meaningfully different from a 34-year-old. The finding is about the onset of a population-level trend, not an individual switch that flips on a birthday.

Also worth noting: the study measured physical capacity — what the body can do under test conditions — not health outcomes. A decline in maximum bench press does not mean you are sick. It means you are ageing. The two are related but not identical.

Finally, the 5–10% improvement figure applies to people who became active. It does not mean a lifelong athlete can add another 10% by training harder at 60. The ceiling moves down with age. What you can recover is some of the gap between your current state and your age-adjusted potential — not the gap between your current state and your 25-year-old peak.

Stakeholder Landscape

Who this matters for directly:

• Adults aged 30–50. You are inside the window where the decline begins. The study is not a warning — it is a permission structure. Starting now is better than starting at 55, but starting at 55 still works.
• Public health agencies. The reversibility finding strengthens the case for midlife physical activity interventions. If a 50-year-old sedentary adult can recover 5–10% of physical capacity, the cost-benefit of community exercise programmes shifts.
• Clinicians and physiotherapists. The data provides a reference curve for what "normal" decline looks like in a general population — useful for identifying patients whose decline is pathological rather than age-related.

Who benefits from the noise:

• The anti-ageing supplement industry, which will use "decline begins at 35" as a marketing hook while ignoring the part where exercise — not pills — produced the improvement.
• Boutique fitness brands targeting the 35+ demographic with urgency-based messaging. The study does not support urgency. It supports consistency.

Cross-Layer Implications

The longevity science layer. This study lands in the same week as the BBC/WaPo amplification of Ulf Ekelund's Lancet paper showing that five extra minutes of daily moderate activity prevents 10% of premature deaths. Together, they form a coherent 2026 message from exercise science: the dose required for meaningful benefit is smaller than most people think, and the window for starting is wider than most people assume. The public-health implication is that we should stop selling exercise as an all-or-nothing proposition and start selling it as a marginal-gains proposition.

The wearable-tech layer. SPAF used lab-based testing (cycle ergometers, strength dynamometers). But the next generation of longitudinal fitness data will come from wearables — Apple Watch, Oura, Whoop — which already track VO₂ max estimates and heart rate variability continuously. The question is whether consumer-grade data can replicate the precision of lab measures over decades. If it can, the sample sizes will make SPAF's several hundred participants look quaint.

The policy layer. Most national physical activity guidelines (WHO, CDC, NHS) are built on cross-sectional evidence. SPAF provides longitudinal evidence that could strengthen the case for midlife screening and intervention programmes. The 2027 follow-up at age 68 will be particularly relevant for pension-age policy.

What This Means for You

If you are under 35: you are likely near your physical peak. The study does not tell you to panic. It tells you that the habits you build now determine the slope of your decline curve. A higher peak means a higher absolute level at 60, even if the percentage decline is similar.

If you are 35–55 and inactive: the study's most important sentence is the one about the 5–10% improvement. You have not missed the window. Start. Walking, cycling, resistance training — anything that raises heart rate and loads muscle. The study does not prescribe a specific protocol, but the broader literature supports 150 minutes of moderate activity plus two resistance sessions per week.

If you are 55+: the improvement effect still holds, though the absolute ceiling is lower. The goal shifts from "improve" to "slow the decline." Resistance training becomes disproportionately important because sarcopenia (age-related muscle loss) accelerates after 60.

If you are a clinician: the SPAF reference curves — once published in full — will give you a population norm against which to assess individual patients. A 55-year-old whose decline is steeper than the SPAF curve warrants investigation.

Uncertainty Ledger

• Mechanism unknown. The study documents that decline happens and that exercise helps, but not why the peak is at 35 or why exercise cannot fully halt decline. Westerståhl's team is investigating.
• Sample size and generalisability. Several hundred Swedish participants. The findings may not generalise perfectly to populations with different genetic backgrounds, diets, or occupational physical activity patterns.
• Self-selection in the "became active" group. People who took up exercise in adulthood may differ systematically from those who did not — in motivation, health status, or socioeconomic factors. The improvement cannot be attributed purely to exercise without accounting for these confounders.
• The 2027 follow-up. The next testing round at age 68 will be critical. If the decline accelerates sharply between 63 and 68, the reversibility window may be narrower than the current data suggest.

Bottom Line

The 47-year SPAF study confirms what physiologists have long suspected — physical decline begins around 35 — but it does so with a methodological rigour that cross-sectional studies cannot match. The finding that deserves to change behaviour is not the decline itself but the reversibility: adults who become active later in life improve by 5–10%. That is not a rounding error. It is the difference between ageing and ageing well. The study does not tell you to fear 35. It tells you that 50 is not too late.

Sources:

• Westerståhl M, Jörnåker G, Jansson E, et al. "Rise and Fall of Physical Capacity in a General Population: A 47-Year Longitudinal Study." Journal of Cachexia, Sarcopenia and Muscle, 2025; 16(6). DOI: 10.1002/jcsm.70134. [Tier 1 — peer-reviewed journal]
• Karolinska Institutet / ScienceDaily. "A 47-year study reveals when strength and fitness start to fade." 15 May 2026. [Tier 2 — university press release with direct researcher quotes]
• BBC Future. "Five minutes of exercise a day could help millions of people live longer." 18 May 2026. [Tier 1 — established outlet, contextual sourcing]
• The Washington Post. "Exercise in middle age linked to longer, healthier life, study shows." 17 May 2026. [Tier 1 — established outlet]