In Praise of Cohort Studies

Helen Pearson has just written a fascinating Nature News article about a British cohort study – the National Survey of Health and Development, run by the Medical Research Council – that’s been following more than 5,000 subjects since their births in 1946. Cohort studies take groups of people who share a common characteristic, such as being born at a particular time or place or sharing a particular job, and follow them over a span of time. (In prospective cohort studies, the study period begins with enrollment; in retrospective cohort studies, medical records are used to study subjects’ pasts.) The participants in this British study are part of the world’s longest-running birth-cohort study, and they’ve contributed reams of data to the study. Information on the participants includes information on their diet and exercise habits, results of tests on heart and lung function, and reports of health conditions like cancer and heart disease. Pearson describes some of researchers’ findings:

It has shown that the heaviest babies were most at risk of breast cancer decades later; that children born into lower social classes were more likely to gain weight as adults; that women with higher IQ reached menopause later in life; and that young children who spent more than a week in hospital were more likely to suffer behaviour and education problems later on.

… In 1985, [former study leader Michael] Wadsworth and his team reported that cohort members whose birth weight had been low had higher blood pressure as adults1. It was an early hint that fetal and infant growth shape adult health, a link that became known as the Barker hypothesis after David Barker, an epidemiologist at the University of Southampton, UK, who published a 1989 analysis of birth weight and health in a different cohort2. He found that babies with the lowest birth weights had the highest risk of heart disease as adults.

Study after study from the 1946 cohort supported the link, showing a tangle of connections between infant and child growth or development and adult traits from cognitive ability to frailty, diabetes, obesity, cancer and schizophrenia risk.

The US National Heart Institute (now the National Heart, Lung, and Blood Institute, part of the National Institutes of Health) launched a similarly ambitious cohort study in 1948 to learn about the causes of cardiovascular disease. More than 5,000 residents of the Framingham, Massachusetts enrolled in the study, which involved extensive checkups every two years, multiple lab tests, and lifestyle questionnaires. In 1971, the study enrolled more than 5,000 of the original participants’ adult children and their spouses, and in 2005 more than 4,000 third-generation subjects completed initial examinations. Some of the milestone findings from the Framingham Heart Study include:

  • 1960: Cigarette smoking found to increase the risk of heart disease
  • 1961: Cholesterol level, blood pressure, and electrocardiogram abnormalities found to increase the risk of heart disease
  • 1967: Physical activity found to reduce the risk of heart disease and obesity to increase the risk of heart disease
  • 1970: High blood pressure found to increase the risk of stroke
  • 1976: Menopause found to increase the risk of heart disease
  • 1978: Psychosocial factors found to affect heart disease
  • 1988: High levels of HDL cholesterol found to reduce risk of death
  • 1999: Lifetime risk at age 40 years of developing coronary heart disease is one in two for men and one in three for women
  • 2001: High-normal blood pressure is associated with an increased risk of cardiovascular disease, emphasizing the need to determine whether lowering high-normal blood pressure can reduce the risk of cardiovascular disease.
  • 2002: Obesity is a risk factor for heart failure.
  • 2007: Based on evaluation of a densely interconnected social network of 12,067 people assessed as part of the Framingham Heart Study, network phenomena appear to be relevant to the biologic and behavioral trait of obesity, and obesity appears to spread through social ties.
  • 2008: Based on analysis of a social network of 12,067 people participating in the Framingham Heart Study (FHS), researchers discover that social networks exert key influences on decision to quit smoking.

It may be hard to remember a time when we didn’t know that smoking increased the risk of heart disease and physical activity decreased it, but it really does take extensive, well-funded studies to find the evidence that informs health recommendations.

There are other long-running cohort studies in progress. The US has some occupation-specific studies, including the Nurses’ Health Study and the Agricultural Health Study (which includes pesticide applicators and their spouses). The National Children’s Study will soon start enrolling participants and will examine a range of environmental factors, from air and water to community and cultural influences. In Britain, additional birth cohorts were launched in 1958, 1970, and 2000, and the new Birth Cohort Facility Study aims to enroll 90,000 British newborns in 2012. (I’m sure other countries have interesting cohort studies going on, too – please let me know about others in the comments.)

Running these studies is an enormous undertaking, and in a sidebar to the Nature News article Pearson writes about the importance of study leaders building strong relationships with participants. Diana Kuh, the 1946 survey director, cites personal communication between researchers and participants, including birthday cards to participants every year, as one of the reasons they’ve kept an average of 80% of the original cohort in the study – an amazingly high retention rate.

This is just one example of the kind of science that no one besides the government is going to fund. It’s hugely expensive and isn’t likely to lead directly to big profits for a particular drug or device manufacturer, although cholesterol-lowering drugs probably wouldn’t be selling so well right now if not for the Framingham findings on cholesterol and heart disease. But the benefits to public health can be enormous. After the Framingham study and other research described the risk factors for heart disease, US age-adjusted death rates from cardiovascular disease dropped by 60% in the second half of the 20th century – and the Centers for Disease Control and Prevention deemed this one of the most important public health achievements of that century.

But government staff and funding isn’t sufficient by itself. For this kind of research, we also need thousands of volunteers willing to spend time and endure some discomfort in order to benefit the rest of us. So thank you, study volunteers, for helping us all learn more about how to improve our health!

Comments

  1. #1 KeithB
    March 1, 2011

    “It’s hugely expensive and isn’t likely to lead directly to big profits for a particular drug or device manufacturer, although cholesterol-lowering drugs probably wouldn’t be selling so well right now if not for the Framingham findings on cholesterol and heart disease.”

    Note only that, but it may blow up in their face. What if one of the Tobacco companies had funded a study like this? They certainly would not have liked the results!

  2. #2 Liz Borkowski
    March 2, 2011

    I shudder to think of a tobacco company running a cohort study to inform our understanding of cardiovascular health. When confronted with these kinds of results, they might figure out a way to re-analyze the data in a way that would produce findings showing smoking to be safe. (See Celeste’s post on the chromium industry’s approach to cohort studies for an example of how this might work.)

  3. #3 Janet
    March 5, 2011

    Really great article, thanks!

    I’ve read about some cohort studies around food allergies in Europe and Scandinavia. Here’s one: http://www.ncbi.nlm.nih.gov/pubmed/21257644

    And there’s EuroPrevall: http://www.europrevall.org/Summary/Summary.html; not sure what the current status of this work is.

Current ye@r *