Midlife and Menopause
Understanding Research Results
As we learn more about menopause, we may read and hear many competing claims and statistics. Understanding the different types of research studies behind these numbers can help us evaluate their results and make informed health care decisions.
There are three major types of research studies10:
- In a cohort study, researchers follow a group, or cohort, of women for a period of time. The researchers determine whether the women experience a particular exposure, such as whether they take a drug or supplement, exercise, smoke, or eat certain foods. The researchers follow all women in the cohort (both those who do and those who do not experience the exposure of interest) to see whether the women's experience a particular outcome (for example, development of a certain disease) The researchers then calculate the risk of developing the outcome (called the incidence rate).
Cohort studies can detect relationships between exposure and outcome but they cannot deﬁnitively claim that the exposure or treatment causes the outcome.
- In a case-control study, researchers investigate two groups, one group of women who have a certain outcome, such as a disease, and one group of women who do not. The group of women with the disease is called the case group, and the others are known as the control or comparison group. Researchers determine whether or not the women experienced speciﬁc exposures of interest (such as diet or exercise). Identifying case and control groups based on outcome is the hallmark of case-control studies, which are common when the outcome of interest is rare (such as certain cancers) or when it is not possible for practical or ethical reasons to assign different groups to different treatments (such as smoking or experiencing abuse). Like cohort studies, case-control studies show a relationship between treatment and outcome but they cannot deﬁnitively claim that the relationship is causal.
- In a randomized controlled trial, researchers recruit women to participate in a study. Once the women agree to participate, they are assigned randomly to receive either the treatment being tested or a placebo (such as a sugar pill). The study is called randomized because there is no rhyme or reason to whether a woman ends up getting the treatment or the placebo. If neither the women participating in the study nor the health professionals working with them know which group each woman is in, the study is called double-blind. The placebo helps with the blinding, so that the woman and her doctor can’t tell by looking at the pill she’s taking whether it’s an active treatment or a sugar pill. Randomized, double-blind placebo-controlled trials are thought of as the “gold standard” in research. They attempt to test whether treatment X causes outcome Y, and they can come the closest to claiming causation because of the blinding, the randomization, and the use of a placebo. The idea is that everything is held constant, except whether women are taking the placebo or the actual treatment.
Understanding Relative versus Absolute Risk
Statistics about health can be confusing because risk can be measured in different ways. Risk indicates whether or not a treatment or behavior (an exposure) is associated with an increased likelihood of developing a disease or condition (an outcome).
Relative risk is an indicator of the strength of the association between exposure and outcome. It is used to assess the importance of a particular factor or treatment in the development of a disease or condition. Relative risk is often expressed as a percentage. For example, a study might conclude that women who took a speciﬁc medicine had, on average, a 25 percent decreased risk of developing a particular disease than women who did not take the drug. Relative risk is always an average for a group.
Absolute risk describes the effect of an exposure on an outcome in the general population (as opposed to comparing speciﬁc groups). In the population overall, how harmful is an exposure? What is the likelihood of developing a particular condition? Absolute risk provides answers to these questions. For example, women in a certain country or city might have a one in ten chance of developing a certain condition during their lifetimes. (This would be the same as a risk of 10 percent.) Like relative risk, absolute risk estimates represent averages for a population, not the proscribed fate of any individual.
Sometimes studies report what sound like dramatic changes, for example a 50 percent decrease in risk for developing a certain disease. But if this is a decrease in the relative risk, and the absolute risk of developing the disease is small, the number of women affected will be low. For example, if the relative risk of developing a disease is cut in half by taking a certain drug, but only one in 100,000 women who do not take the medicine develop that disease each year, only one in 200,000 women per year would avoid developing the disease if the whole group took the medicine—and all of the other women would be at risk of experiencing its unwanted “side” effects.
Understanding what research results tell us about our risk of developing a certain disease or condition is an important part of making health care decisions. It’s also important to remember that there is enormous variation among individuals and that our decisions are informed by our values and preferences as well as scientiﬁc evidence.
End of excerpt.
Excerpted from Chapter 2: Making Health Care Decisions in Our Bodies, Ourselves: Menopause © 2006 Boston Women's Health Book Collective.
For more information, see Understanding Research Results: The Difference between Prevalence and Incidence.
10. Kay Dickersin. “Behind the Numbers,” MAMM, June 2003, accessed at www.findarticles.com/p/articles/mi_kmmam/is_200306/ai_kepm41109 on October 5, 2005; see also Charles H. Hennekens, Julie E. Buring, Sherry L. Mayrent, eds., Epidemiology in Medicine (Boston: Lippincott Williams & Wilkins, 1987). [back to text]
Excerpted from Our Bodies, Ourselves: Menopause, © 2006, Boston Women's Health Book Collective.
< Return to Chapter 26 Overview