Inside Science: Cross-sectional versus longitudinal study design
July 18, 2022 - 6 minutes readThe key to quality scientific research is asking the right questions and then making accurate interpretations of the data that you’ve collected to answer those questions. Often scientists will address the same research question from multiple perspectives. One example of this is the use of both cross-sectional and longitudinal approaches. The study design of the Dog Aging Project allows our scientists to do both!
What’s the difference between cross-sectional and longitudinal research?
I’m glad you asked! This is an important distinction, which has a strong influence on how we interpret our data. Let’s dive in!
- Cross-sectional research discovers things that are happening at the same time, but cannot determine whether or not one thing caused the other.
- Longitudinal research follows specific individuals over a period of time, to determine what happens to them in order so that cause and effect can be identified.
For instance, consider the following imaginary cross-sectional study:
A researcher studies the bedtime and coffee-drinking habits of a group of people. When she looks at the results, the data show some very interesting patterns. Some of the people in the study report drinking coffee in the afternoon and also staying up late at night. Other people in the same study reported that they do not drink coffee in the afternoon and go to bed earlier.
What does it all mean? All we know is that both things are true of the same people: they drink coffee AND stay up, or they don’t. From this study we cannot determine cause and effect. Maybe people who drink coffee in the afternoon can’t fall asleep, so they stay up late. Or maybe people who have to stay up late for work or school drink coffee in the afternoon to help them stay awake.
In a cross-sectional study, we can’t know which came first. For that reason, it would be wrong to tell people in that study to stop drinking coffee and assume that means they will go to bed earlier. If they still have late evening work or school responsibilities, they will stay up just as late but without the coffee.
Does this mean cross-sectional studies aren’t useful?
Definitely not! Often scientists use cross-sectional studies to help them design future studies that do allow them to identify cause and effect.
Going back to our imaginary coffee study. The researcher could follow a group of people over time, maybe starting the research when they are young and don’t drink coffee. If she collects data on bedtime and also the introduction of coffee drinking habits, she may learn that when people start drinking coffee in the afternoon their bedtimes shift later. Or she might identify that there are “morning” people and “night” people but whether or not they drink coffee in the afternoon doesn’t seem to make a difference. (If she found this result, we’d definitely want her to talk to the genetics team!)
What do you do at the Dog Aging Project?
Our team conducts both cross-sectional and longitudinal research.
Recently we published data from a cross-sectional study of feeding frequency. We found that dogs who are fed once daily are often dogs with fewer health concerns. Because this is a cross-sectional study, it is not possible to know which thing caused the other. It is not necessarily true that feeding less often makes dogs have fewer health concerns. It is equally likely that dogs with health concerns don’t have consistent appetites, and their owners feed them more frequently to encourage them to eat.
The results of this study are fascinating, but they don’t tell us whether the health status of the dog caused the feeding frequency or whether feeding frequency influenced health status. More research will be necessary to determine which way the causal arrows point.
FYI, when this paper hit the news media, there were lots of lazy headlines claiming that if people only fed their dogs once a day, their dogs would be in better health. This is not a valid interpretation of the data. So we repeat the advice we shared elsewhere: do not change your dog’s feeding regime based on our study! Talk with your primary care veterinarian if you have questions about your dog’s feeding management.
And stay tuned for more on this topic from our team!
Dr. Kate Creevy
Chief Veterinary Officer