"If you want to stave off death for as long as possible, you might want to reach for a tennis racquet," The Guardian reports.
A study looking at the impact of individual sports on mortality found racquet sports reduced the risk of death by around 47%.
Researchers also found reduced risks of death for people who took part in cycling, swimming and aerobics.
They didn't find such effects for people who took part in rugby, football or running – although this unexpected finding may be explained by the low number of deaths, which may have skewed the statistics. The smaller the data set, the bigger the chance of the data being influenced by chance.
While the researchers found taking part in some sports reduced the risk of death compared to not taking part, they did not directly compare the benefits of different sports. That means we can't say which sport is "best" for health.
What is clear from the study is that any sort of regular physical activity is likely to help us stay healthier and live longer.
The study was carried out by researchers from the UKK Institute in Finland, University of Edinburgh, University of Oxford, Loughborough University and University of Exeter in the UK, Victoria University and University of Sydney in Australia, and University of Graz in Austria. No information about funding was provided.
Most of the UK media reported that tennis and badminton were the "best" exercise, because people participating in these sports had the biggest reductions in risk of death compared to people not taking part.
However, these headlines ignore the fact that the effects of football and running were probably underestimated.
This was a cohort study using information from eight health surveys in England and three surveys in Scotland, linked to data about deaths.
Cohort studies can spot links between factors such as taking part in exercise and length of life, but they can't prove that one factor causes another.
Researchers analysed questionnaires from 80,306 people. These people (average age 52, more than half women) were followed up for an average of nine years, and any deaths recorded.
After adjusting their figures to account for factors such as age, smoking and weight, researchers looked for links between how long people lived and whether they took part in a sport.
The questionnaires came from two big annual surveys, the Health Survey for England and the Scottish Health Survey. They used questionnaires from 11 years between 1994 and 2008. People were asked if they had taken part in any of the following sports during the past four weeks:
For each of the sports included, researchers compared the chances of being alive at the end of the study, between people who said they took part in them with people who didn't take part in them.
They tried to account for the seasonal nature of sports like football and rugby by spreading the questionnaires year-round, but this may have missed some participants.
In addition to age, smoking and weight, researchers took account of how much other physical activity (outside of the named sports) people did, as well as the following confounders:
Of the 80,306 people studied, 8,790 (10.9%) died during the average nine years of follow-up.
After adjusting their figures for confounding factors, researchers found that people who took part in sports had the following reduced chances of death during the study:
They did not find a statistically significant reduced chance of death for people taking part in running or football.
They found reduced chances of death from heart disease or stroke for swimming, racquet sports and aerobics, but not for running, cycling or football.
The researchers said their results "demonstrate that participation in specific sports may have significant benefits for public health". They said they had found "robust evidence" that swimming, racquet sports, cycling and aerobics were linked to reduced chance of death.
They acknowledge their findings on running were "surprising" in light of four big studies conducted previously. They suggest the low number of deaths among people who went running (68 of 4,012 runners, or 1.6%) could have prevented the statistical model from reaching statistical significance.
They also say that asking people about their participation in running during the past four weeks could have been misleading, so that those who jogged occasionally were included among those who ran regularly, year round. They say their result should be seen as adding to the body of evidence supporting running, rather than contradicting it.
Similarly for football, they say the results were "somewhat unexpected" and may reflect only the low numbers of people in the study who said they played football.
The overall conclusion we can take from this study is that taking part in sport or fitness activities is linked to a lower chance of death in a given period.
It's encouraging to see that a wide range of popular activities, including swimming, aerobics and cycling, are likely to be beneficial.
But we should be wary about comparing the types of different sports against each other. They weren't directly compared in the study and there may be reasons why results for some activities, such as football and running, were found to be statistically non-significant (potentially down to chance).
Statistician Professor David Spiegelhalter said that making a distinction between the sports was "simply not valid" and the differing results only reflected the small number of deaths among football players and runners.
The statistical uncertainty may have come about because of the way in which the results were adjusted to take account of confounding factors. For example, runners are likely to be non-smokers, younger, do more exercise overall and be leaner, compared to people who don't run – all of which will reduce their chances of death.
Once you've taken these factors into account, the additional impact of running may be hard to measure.
Professor Spiegelhalter points out that because this is an observational study, we can't really tell whether taking part in those sports where researchers did find a statistically significant result actually caused the lower death rate among participants.
He said it was "equally plausible" that "those at increased risk of death over the next few years are less likely to be healthy enough to play active sports now."
So what should people do as a result of the study?
The sensible advice seems to be to find a physical activity you enjoy – whether that's swimming, tennis, dancing, football or anything else that gets you out of breath – and take part. The more you enjoy an activity the greater the possibility that you will carry on doing it on a long-term basis.
While we can't say that one sport is better than others at helping you to live longer, evidence shows that physical exercise is likely to keep us fitter, healthier and happier for longer.