13.2 million runners in France, and an awkward meta-analysis about running with others

The Observatoire du Running presented on 10 April 2026 by the Union Sport et Cycle and the French Athletics Federation counts 13.2 million runners in France, forty-three per cent of whom say they run in a group. In the same year, the largest meta-analysis ever published on the subject concludes that group programmes do not produce more physical activity than individual ones. Those two sentences do not contradict each other, but the second has to be read with a great deal of care before anything is drawn from it.

How many people in France run regularly?

Thirteen point two million runners is a million more than five years ago, and sixty-four per cent of them run at least once a week. The newcomers are young and mostly women: sixty-three per cent of new runners are women, and forty-four per cent are under twenty-four.

One clarification is needed, though, because it is systematically left out. The same observatory’s historical series gave 13.4 million runners in 2021, coming out of the lockdowns. The 2026 figure is therefore high, it is up over five years, but it is not an all-time record. And the survey methodology changed between editions, which rules out calculating a year-on-year change.

On the licensing side, the French Athletics Federation passed 321,000 members by mid-January 2025, spread across 2,535 clubs, with 48.6 per cent women. Around eighteen hundred of those clubs are running-oriented and account for thirty-five per cent of the membership.

Do runners in France run alone or in groups?

Forty-three per cent of runners in a group sounds like a lot. The same observatory also reports that twenty-eight per cent describe themselves as members of a club or group, which probably covers a form of supervision rather than formal membership. Sixty-two per cent follow a training plan, forty-one per cent through an app, and only seven per cent with a coach.

The INJEP and CRÉDOC barometer conducted in July 2024 among 4,057 people gives a wider frame: forty-seven per cent of French people do their main physical activity alone, fourteen per cent as a couple, fourteen per cent with friends, twelve per cent with family. That barometer, however, lumps walking and running together, and so cannot isolate a figure specific to running.

The thing to take away is this: surveys of running produce shares of collective practice that swing from one extreme to the other depending on the population questioned and the wording of the question. None of them measures quite the same thing, and they should not be added up in your head.

What does the research actually say about running in a group?

On 15 April 2026, Nature Human Behaviour published a meta-analysis of seventy-one studies and 523 effect sizes comparing individual and group interventions. On the volume of physical activity produced, the difference is nil and the equivalence test is significant: across more than twenty-two thousand participants, the group format does not get people moving more than the individual one.

Three clarifications make that result far less final than it looks.

First, the equivalence holds only for physical activity. On psychosocial outcomes the equivalence test is not significant: the result is undetermined, which is not the same as null. The point estimate there is in fact more than three times higher. Writing that science has shown a group adds nothing, not even socially, would be a complete misreading.

Second, on functional capacity, meaning strength, balance and mobility, the group is significantly superior once outliers are removed.

Third, adherence was not analysed as a distinct outcome, it was folded into physical activity behaviour. And recreational running does not appear among the populations studied, which are mostly clinical or elderly. Transposing that result to the Sunday runner is an extrapolation, not a reading.

What separates a real group from a simple crowd?

A much older meta-analysis, published in 2006 from forty-four studies and 4,578 participants, clarifies the picture decisively, because it does not ask the same question.

It does not compare one programme format with another, it compares qualities of group. And there the gaps are large. A true group, where people know one another and matter to one another, produces a strong effect compared with a mere collective of people doing the same thing in the same place, with an effect size of zero point seven three, and zero point seven four on adherence itself. Conversely, a home-based programme with no human contact at all is indistinguishable from no programme, and a collective does no better than home-based support with regular contact.

The dividing line, then, is not between alone and together. It is between being surrounded and being expected. Running next to someone changes almost nothing. Running with someone who notices your absence changes a great deal.

Why does belonging matter more than the running itself?

A study published in the Journal of Sports Sciences among 289 participants in free weekly runs, combining questionnaires with objective participation data, reaches a result that deserves to be better known. Identification with the group strongly predicts cohesion, exercise satisfaction and, further down the chain, life satisfaction. Simply taking part often, on the other hand, does not predict life satisfaction at all: the coefficient is null and not significant.

Another study, following 548 new participants over six months, measures an average gain in life satisfaction of a quarter of a point on a ten-point scale, with the clearest gain going to the least active people at the outset. It has no control group, which sharply limits how far the conclusion can be taken.

Both pieces of work are cross-sectional or uncontrolled, so they do not establish causality. But they converge on a simple point: what seems to count is not the number of outings, it is the feeling of belonging.

Is group running right for everyone?

It would be dishonest to stop there. An Irish survey released as a preprint in June 2025 among 4,304 responding runners shows that 69.4 per cent interact with someone during their weekly run, 45.9 per cent with friends and 44.1 per cent with strangers, and that 48.2 per cent feel part of their community. But 18.5 per cent report that the event has left them feeling isolated, and that proportion is significantly higher among women, 20.9 per cent against 15.4 per cent among men. That figure does not measure loneliness in general, it measures a feeling of isolation within the running group itself. Given the survey’s very low response rate, it is likely to be a floor rather than a ceiling.

A Scottish analysis of 20,191 first-time participants also observes that 64.18 per cent come back at least once, men slightly more than women, and that the slowest runners and the largest events both retain markedly less well.

An experiment run with 236 women runners adds one useful result: activating their identity as a runner rather than as a woman significantly increases their sense of safety. Belonging is not only a comfort, it has a measurable effect on the willingness to go out.

Can you take part in a parkrun in France?

Since almost all the available research on collective running rests on parkrun, one clarification is due: those free Saturday morning runs are not available in France. Activity there was suspended in July 2022 and the suspension was confirmed with no end date on 16 May 2024, for reasons tied to French law and the associated legal risk. The model exists in more than twenty countries and has registered more than ten million people. It is not accessible here, and any article that lets you think otherwise has not bothered to check.

How do you find a running group near you?

Do not look for a group, look for two or three people who run at your pace, near where you live, at a time that suits you. The data on who gets people started points the same way: it is family, partners and colleagues, each at around ten per cent, well ahead of clubs and social networks.

Set a standing appointment rather than a target. One slot kept every week beats a twelve-week plan nobody will reread. And accept that a group probably will not make you run more, but that it will very likely make you run for more years of your life, because you will have a place in it.

That leaves the part statistics do not solve: finding those two or three people. That is exactly what Smatchy does, by putting you in touch with runners at your level and in your area, for Sunday’s outing rather than for the whole season.