November 6, 2013. Marta Wegorzewska, PhD candidate, in collaboration with Dr. Seth Bokser, MD. Edited by Dr. Darya Rose, PhD.
All Scientific Studies Are NOT Created Equal
If a scientific study claims eating garlic daily during pregnancy shortens labor, should pregnant women start eating garlic every day?
The answer depends on the type of scientific study. All studies are not created equal. Some studies prove, and others suggest.
Studies that suggest may be right, but they may also be wrong. For example, Facebook suggests people we may know. Some of them we do know, and we request their friendship. Some of the suggestions are wrong, and we ignore them.
Which studies prove and which suggest?
Studies that Prove
The best study would ask, “Does garlic shorten labor?”
To answer this question, women would be asked to volunteer for the study. They would have just found out they are pregnant so they could eat garlic throughout nine months of pregnancy (prospective).
Hundreds to thousands of women would be involved in the study to ensure any observations are not due to chance. The women would be randomly split into two groups (randomized). The due dates for the two groups would be similar. One group would eat garlic throughout pregnancy and the other would not (control).
The number of hours the women are in labor would be recorded. To prevent unintentional bias, the individual recording the time would not know whether the woman ate garlic or not. He would be “blinded” to that information.
Women who require a drug to speed labor would be excluded from the study. Excluding these cases ensures the effect on labor is from the garlic and not from the drug (confounding factor).
This type of study is called a prospective randomized control study.
The strength of this type of study is that it proves if one factor (garlic) causes another (shorter labor). If the number of hours of labor is shorter for women who ate garlic during pregnancy compared to those that didn’t, the study proved garlic shortens labor!
Why aren’t all studies designed as a prospective randomized control study?
Imagine if scientists wondered if women who are obese during pregnancy give birth to sick babies.
A prospective randomized control study would ask women to voluntarily become obese during pregnancy to see if their babies would be born sick. No woman would volunteer for this study! It would be unethical to conduct a study that would potentially harm mom and baby.
Studies that Suggest
Other studies would provide clues.
A case control study would look at already sick babies. This study would compare the weight of pregnant women delivering sick babies to women delivering healthy babies.
A cohort study would look at already obese pregnant women. The health of their babies would be compared to the health of babies of non-obese women.
These studies would not prove that obesity causes sick babies. They would suggest that being obese during pregnancy may be associated with (linked to) sick babies.
Case control and cohort studies suggest instead of prove because they have limitations. They cannot eliminate other factors that may contribute to sick babies. For example, maybe obese pregnant women eat more fast food because they are struggling financially, working longer hours and are more stressed. Stress may be another factor that could impact a baby’s health. A prospective randomized control study would avoid this problem by randomly splitting women into groups. Stress would affect both groups equally and cancel out. Any differences seen between the groups would be due to obesity.
That being said, are you going to start eating garlic every day?
If a prospective randomized control study proved garlic causes shorter labor, I would try to eat garlic even if I hated it.
If a case control or cohort study suggested garlic is linked to shorter labor, I might eat garlic if I didn’t mind it, but I wouldn’t view garlic as a guarantee my labor would be shorter. If I hated garlic, I would ignore this suggestion.
This work is funded in part by the Graduate Student Internships for Career Exploration (GSICE) program at UCSF