Cross-posted on nevernotrunning.com
In JACC, a study was recently published which concluded that sitting in front of the TV, computer, game console, etc. is associated with higher risk of death and heart attacks. This is concerning, but not a novel finding. This study does have a couple of unique features that I wanted to write about.
The study is an observational cohort of self reported data. This means that 1) no one controlled the variable of "screen time" and subjects could watch as much or little as they wanted, 2) all the data was self-reported, ie: no one set up a timer on the TV to measure how much TV they watch, and therefore the data are limited by "recall bias" (how well can you remember just how much TV you watch?), and 3) the study is drawn from a database of previously collected data and therefore you cannot prove a causal relationship, just a correlation. So while the study has limits, this type of study design is great because it is easy and cheap to do, and can help bring up ideas about where people should do more research in the future.
One of the findings that was particularly concerning is that the researchers also had access to the subjects' report of how much exercise they did. That means they could "control" for that variable and take it into account. When they did, the risk of death or heart attack was still higher, even with regular exercise. The authors do point out that screen time is sometimes underreported, which is a limitation. One thing they failed to point out or account for is overestimation of physical exercise. Many of my patients try to find some way to say yes when I inquire about exercise, even when it seems highly unlikely that they exercise. I do not mean to imply that people are lying, just that they know what the right answer is (ie: daily exercise) and they do not want to admit that they fail to meet that standard. It is the same phenomenon that leads people to underreport alcohol use. Similarly, people also very frequently tell me that they are "quitting" smoking, or cutting back, but few have either quit or openly admit they smoke.
Another important observation about the study is that it only accounted for recreational screen time, not work screen time. The exact question from the survey was: "Thinking of weekdays, how much time on average day do you spend watching TV or another type of screen such as a computer, or video game? Please do not include any time spent in front of a screen while at school, college, or work." This seems to me a major limitation because it does not account for the 40+ weekly hours when subjects might have been computer programming or doing manual labor on a construction site, both of which would seemingly have markedly different effects on cardiovascular risk.
The other interesting data that this research group had access to was blood tests, such as c-reactive protein and cholesterol. They estimate that these account for 25% of the risk associated with screen time. This may play into my prior point that recreational screen time is only part of the equation and perhaps being immobile at work causes higher cholesterol and inflammation which then blurs the effect of recreational exercise. In any case, the risk of death and heart attacks is complex, and cannot be simply explained by sitting and watching TV or by going out running.
The study confirms prior research that recreational screen time is unhealthy and is associated with risk or cardiovascular events. Second, it suggests that exercise does not modulate this risk appreciably. This conclusion, however, does not fit with the vast majority of other studies which do show a protective effect of exercise, and I do not think that the methods used are adequate to draw this conclusion.
Additional commentary here and here.