Many people complain they do not get enough sleep, and it seems they are right to be concerned.
Researchers have found that adults under the age of 65 who get five or fewer hours of sleep for seven days a week have a higher risk of death than those who consistently get six or seven hours of shut-eye.
However, the effect of short sleeps over a few days may be countered by a later lie-in.
The research found that individuals who managed just a few hours' sleep each day during the week but then had a long snooze at weekends had no raised mortality risk, compared with those who consistently stuck to six or seven hours a night.
"Sleep duration is important for longevity," said Dr Torbjorn Akerstedt, first author of the study, at the Stress Research Institute at Stockholm University, and Karolinska Institute, also in the Swedish capital.
The study, published in the Journal of Sleep Research, is based on data from more than 38,000 adults, collected during a lifestyle and medical survey conducted throughout Sweden in 1997. The fate of participants was followed for up to 13 years, using a national death register.
Dr Akerstedt said researchers had previously looked at links between sleep duration and mortality but had focused on sleep during the working week. "I suspected there might be some modification if you included also weekend sleep, or day-off sleep."
Once factors such as gender, body mass index, smoking, physical activity and shift work, were taken into account, the results revealed that those under the age of 65 who got five hours of sleep or under that amount seven days a week had a 65 per cent higher mortality rate than those getting six or seven hours' sleep every day.
But there was no increased risk of death for those who slept five or fewer hours during the week but then managed eight or more hours' sleep on weekend days.
"The assumption in this is that weekend sleep is a catch-up sleep," said Dr Akerstedt, though he noted the study did not prove that to be the case.
However, people who slept for eight or more hours, seven days a week, were found to have a 25 per cent higher mortality rate compared with those who kept to six or seven hours a day.
The study also found that the link between sleep patterns and mortality disappeared for those aged 65 or older. That, Dr Akerstedt said, was perhaps because older individuals got the sleep they needed.
Average sleep duration at weekends and the percentage of those saying they did not feel rested at waking, did fall with age. Across the week, older people had more consistent and more often sleep over a shorter time span.
While the study did not investigate the link between sleep patterns and mortality rates, Dr Akerstedt said it was possible little sleep had a negative effect on the body, while consistently lengthy sleep could be a sign of underlying health problems.
The study had limitations; participants were only asked about their sleep patterns at one point in time.
But Dr Stuart Peirson, an expert on the human "body clock" but who was not involved in the research, said the study offered a more nuanced view than previous research, which had suggested that both very little or a lot of sleep was bad for health and longevity.
"It fits with what we do know about sleep - that sleep is regulated by the body clock but also regulated by what is called a homeostatic process, which means the longer you are awake the more you need to sleep."
Dr Peirson, based at the University of Oxford, noted that sleep requirements varied from person to person according to genetics, but he added that "sleep debt" needed to be "paid off".
He said: "You can't keep burning the candle at both ends. Well, you can, but you won't live as long."