Being unhappy or stressed will not kill, says study
Being miserable or stressed will not increase your risk of dying, according to the UK's Million Women Study.
It had been thought that being unhappy was bad for health - particularly for the heart.
But the decade-long analysis, published in the Lancet, said previous studies had just confused cause and effect.
However, experts argued that unhappiness in childhood may still have a lasting impact.
A series of studies had shown that how happy people are, strongly predicts how long they are going to live.
Ideas included detrimental changes in stress hormones or the immune system resulting in a higher risk of death.
But the research team in the UK and Australia said those studies failed to deal with reverse causality - namely, that people who are ill are not very happy.
Participants in the Million Women Study were asked to regularly rate their health, happiness and levels of stress.
The results showed that whether people were "never", "usually" or "mostly" happy had no impact on their odds of dying during the duration of the study once other factors such as health or whether they smoked were taken into account.
Dr Bette Liu, one of the researchers at the University of New South Wales in Australia, said: "Illness makes you unhappy, but unhappiness itself doesn't make you ill.
"We found no direct effect of unhappiness or stress on mortality, even in a 10-year study of a million women."
Co-author Prof Sir Richard Peto, from the University of Oxford, said light smokers had double the risk of an early death and regular smokers had three times the risk of dying during the study period, but that happiness was "irrelevant".
He said it could have indirect effects if people started consuming large amounts of alcohol or massively overeating, but happiness itself "does not have any material, direct, effect on mortality".
But he warned the myth may be too entrenched to shake off: "People will still believe stress causes heart attacks after this story has been and gone.
"It isn't true, but it suits people to believe it."
In a commentary, Dr Philipe de Souto Barreto and professor Yves Rolland from the University Hospital of Toulouse in France, said: "Further research from a lifecourse perspective is needed since happiness during critical periods, such as childhood, could have important consequences on health in adulthood."