Heart and lungs

Television heart risk needs more study

“Watching TV for four hours a day doubles the risk of a heart attack,” The Sun has reported. “The reason is thought to be that simply sitting for so long causes coronary problems,” the article added.

The story is based on a study that surveyed 4,512 people to estimate their television viewing and physical activity, comparing their habits with their risk of death or cardiovascular disease over the next four years. Those viewing TV and video games for four hours or more per day were 48% more likely to die (due to any cause) and 125% more likely to have a cardiovascular-related event (such as a heart attack or stroke) than those who watched less than two hours. The relationship was independent of smoking, social class and how much physical activity people did.

This well-conducted study suggests that lengthy periods of recreational viewing may have harmful effects on the cardiovascular system, increasing the risk of heart attacks, strokes and early death. However, the study had some limitations, such as not accounting for the influence of diet or time sitting in front of a computer at work. This initial research is of interest, but there is now a need for larger, longer studies to verify the relationship.

Where did the story come from?

The study was carried out by researchers from University College London, the University of Queensland, Brisbane and Edith Cowan University and the Heart and Diabetes Institute, Melbourne. The researchers were financially supported by the UK’s National Institute for Health Research, the British Heart Foundation and the Victorian Health Promotion Foundation Public, Australia.

The study was published in the peer-reviewed Journal of the American College of Cardiology. It was reported accurately, but uncritically, by newspapers.

What kind of research was this?

This was a cohort study which aimed to find out if TV watching and other screen-based entertainment had any associations with the risk of cardiovascular disease (CVD) and early death due to any cause. It also investigated whether markers such as body mass index (BMI), cholesterol levels and C-reactive protein (a protein found in the blood, levels of which rise in response to inflammation) had a role in any associations between screen viewing time and CVD-related events.

The researchers say that some studies have suggested that prolonged sitting is linked to the development of risk factors for cardiovascular disease independent of the physical activity one performs. However, they point out that there is little hard evidence so far.

This type of study, which can follow people over a number of years, is useful in looking at possible associations between certain events (such as TV watching) and health outcomes (such as heart attack). However, there can be limitations to this study design including, in this case, that all lifestyle factors were self-reported, and that some possible confounding factors such as diet were not taken into account.

What did the research involve?

The researchers recruited a total of 4,512 people aged 35 and over who had taken part in a large, household-based health survey in Scotland in 2003. These survey results were linked to official data on hospital admissions and deaths.

The participants reported how long and how often they watched TV or other types of screen-based entertainment (such as computer or video games), both on weekdays and weekends. These estimates did not include any time spent at the screen while at school, college or work. The researchers divided participants into four groups, ranging from those who watched less than two hours a day to those who watched four hours a day or more. Researchers also asked about frequency of physical activity such as walking, housework and gardening and leisure time exercise.

Trained interviewers asked participants about their health, lifestyle and socioeconomic status and in a separate visit, trained nurses collected blood samples. These were used to look at levels of C-reactive protein (CRP) and cholesterol levels.

The researchers followed up participants to December 2007. They looked at deaths due to any cause among the sample and CVD-related hospital episodes such as heart attack, coronary bypass, stroke and heart failure. 

They used standard statistical techniques to analyse any association between how much TV people watched, premature death and CVD-related events. They adjusted their findings to account for the influence of physical activity along with risk factors including BMI, social class, smoking habits and long-standing illness. They also performed analyses to see whether biological risk factors such as cholesterol levels played a role in any association.

The researchers’ analysis excluded 340 participants who had been admitted to hospital for CVD-related events either before the initial testing or in the first two years of follow-up, to reduce the risk of reverse causality (i.e. that pre-existing cardiovascular problems may have actually been the reason why some participants spent more time watching TV).

It should be noted that of the 6,353 potentially eligible adults in the health survey, 5,814 (91.5%) consented to their records being linked to records of mortality and hospital episodes; and that of these 5,814 participants, 1,302 (22.4%) were lost to follow-up. The researchers found that, compared with those who did not consent to their records being linked, those who consented were older, reported less physical activity and greater viewing time. Those who were lost to follow-up were younger and did more physical activity, among other characteristics.

What were the basic results?

The researchers report that among the 4,512 participants analysed there were 215 CVD-related events and 325 deaths from any cause among participants during the follow-up period. They found that, after adjusting for physical activity and other risk factors, people who spent four hours or more watching TV and playing games:

  • were nearly 50% more likely to die of any cause than people who watched less than two hours a day (hazard ratio [HR] 1.48, 95% confidence interval [CI] 1.04 to 2.13)
  • were over twice as likely to experience a CVD-related event than those watched less than two hours per day (HR 2.25, 95% CI 1.30 to 3.89)

Excluding people who had a CVD-related event during the first two years of follow-up did not change the results appreciably.

Levels of C-reactive protein and cholesterol, as well as body mass index, were reported to account for 25% of the association between viewing time and cardiovascular events.

How did the researchers interpret the results?

The researchers conclude that recreational sitting, as reflected by television/screen viewing time, is associated with raised mortality and CVD risk, regardless of the amount of physical activity one does.

They say that although the pathways linking prolonged sitting to cardiovascular disease are unclear, it is possible that sitting has an effect on risk factors such as lipid levels and inflammation.  Public health recommendations should include guidelines on sedentary behaviour, they conclude.


This well-conducted study examining a large, representative sample of the population suggests that lengthy periods of recreational viewing time increase the risk of cardiovascular disease and early death. However, it does have its limitations:

  • People had to self-report their viewing time, in addition to other lifestyle factors, which increases the possibility of inaccurate reporting.
  • It took no account of diet as a risk factor, although watching TV is linked to excess calorie consumption. However, the researchers did find that the observed associations were independent of BMI.
  • The study’s follow-up period was on average just over four years. Had it followed up the health of participants for longer, the results may have been different.
  • It is possible that being unwell caused people to spend more time watching TV, rather than vice versa (which is called reverse causality), although the researchers took measures to minimise this possibility.

Crucially, the researchers only looked at recreational screen time, not time spent on the screen at work. Since a large proportion of the population spend much of their day sitting in front of computers, the exclusion of this activity somewhat limits the study’s applicability to everyday life. Similarly, the researchers did not look at time spent sitting engaged in other sedentary activities, such as reading or sewing.

Further research, using objective measures of activity and time spent sitting, is required to recognise the possible association between sitting for long periods and adverse health events.

NHS Attribution