A researcher is attempting to understand the prevalence of hypertension in a small island population. The researcher enrolls approximately 15,000 patients and contemplates which study design is best suited, to determine what (if any) exposures result in a high or low prevalence of hypertension at a single point in time. Which of the following types of studies is best suited to answer this researcher’s question?

Cross sectional study

Last updated: November 01, 2022

Foundational SciencesBiostatistics, epidemiology, population health, and interpretation of the medical literatureStudy design, types, and selection of studies

USMLE® Step 1 ReviewBiostatistics, epidemiology, population health, and interpretation of the medical literatureStudy design, types, and selection of studies

ClerkshipsInternal medicineBack to the Basic SciencesCore competenciesInformation mastery

USMLE® Step 2 ReviewBiostatistics, epidemiology, population health, and interpretation of the medical literatureBack to the Basic SciencesStudy design, types, and selection of studies

Definitions & Key takeaways

A cross-sectional study is a type of observational study that assesses the relationship between exposures and outcomes at a single point in time. It can be used to study factors associated with an outcome. In a cross-sectional study, data are collected from a group of participants who have been followed for a certain length of time. Data are then analyzed to see if there is an association between the different exposures and the outcome measure. Cross-sectional are usually quick and cheap studies that help to sample large groups of people.

Watch video only

A cross- sectional study is a study design where an exposure and an outcome are measured at the same time. For example, let’s say you want to figure out if people who are obese – which means having a body mass index or BMI of 30 or higher - have higher serum cholesterol levels compared to people who are not obese – having a body mass index below 30.

To do this, you might look at the medical records of 100 people to see who has high cholesterol levels and who has low cholesterol levels and compare that to how many people in each group are obese or not obese.

You can think about a cross-sectional study like a snapshot of the population at a certain point in time. Since you can only collect the information you see in that one moment, you don’t know what happens before or after the snapshot was taken.

So, we can only collect information on prevalence – the proportion of exposures or outcomes that already exist at a certain time – and not incidence – the proportion of new exposures or outcomes that occur in a certain time period.

In terms of prevalence, there’s an outcome prevalence and an exposure prevalence. The outcome prevalence is the proportion of people who have an outcome in the exposed group and the non-exposed group.

In a cross-sectional study this can be organized in a 2 by 2 table, with the exposure – obesity or no obesity – on the side and the outcome – high or low cholesterol levels – on the top, and each box is labeled a, b, c, or d.

Cell a includes individuals who have high cholesterol and who are obese; cell b includes individuals who have low cholesterol and who are obese; cell c includes individuals who have high cholesterol and who are not obese; and cell d includes individuals who have low cholesterol and who are not obese.

We can calculate an outcome prevalence to figure out if high cholesterol – the outcome – is more prevalent for people who are obese or not obese – the exposure.

Let’s say that there are 50 people in cell a, 15 in cell b, 5 in cell c, and 30 in cell d. To find the outcome prevalence, we first calculate the proportion of people who have high cholesterol in the group that is obese – so “a” divided by “a plus b”, or 50 divided by 65, which rounds to 0.77.

Then we compare it with the proportion of people who have high cholesterol in the group who are not obese – so “c” divided by “c plus d”, or 5 divided by 35, which rounds to 0.14.

And, 0.77 divided by 0.14 equals 5.5, which means that the people who are obese have about 5.5 times the prevalence of high cholesterol compared to people who are not obese.

We can similarly calculate the exposure prevalence, which is a comparison of the proportion of people who have an exposure in the outcome group and the no outcome groups.

For example, we might calculate the exposure prevalence to figure out if obesity – the exposure – is more prevalent for people who have high cholesterol or low cholesterol – the outcome.

To find the exposure prevalence, we first calculate the proportion of people who are obese in the group that has high cholesterol– so “a” divided by “a plus c”, or 50 divided by 55, which is 0.91.

We then compare it to the proportion of people who are obese in the group that has low cholesterol – so “b” divided by “b plus d”, or 15 divided by 45, which is 0.33.

And 0.91 divided by 0.33 equals 2.72, which means that the people with high-cholesterol have about 2.72 times the prevalence of obesity compared to people with low cholesterol.

Notice that the exposure and outcome prevalence are not the same number, so you can’t assume that they will estimate each other.

Cross-sectional studies are often a cheap, quick, and easy way to collect information about a large numbers of participants, since all of the information is collected at one time, and typically they are done through surveys.

Also, a lot of information can be collected from each participant, so cross-sectional studies are especially useful for looking at relationships between multiple diseases and multiple outcomes.

For example, in addition to looking at obesity, you could also compare the prevalence of diabetes or heart disease in people with high or low cholesterol levels.

But cross-sectional studies also have some major downsides, particularly because they only measure information about prevalence, and not about incidence.

Since you don’t know anything about what happened to the individuals before they joined the study, it’s impossible to tell if the exposure or the outcome happened first.

For example, you don’t know if people were obese before they got high cholesterol, or if people had high cholesterol before they became obese.

Because of this, the results from a cross-sectional study can only suggest that obesity is a possible risk factor of high cholesterol, but they don’t provide enough evidence to say that obesity actually causes high cholesterol.

It’s even possible that some other factor – like eating fast food more than three times a week - might be influencing both the exposure and the outcome.

But, since we only see a snapshot, we won’t necessarily know if there are other variables that may be confounding or distorting the relationship.

Another downside of cross-sectional studies is that they don’t capture people who have already died from the exposure or the outcome in question, so they’re not very helpful for studying acute diseases with high mortality rates, like myocardial infarctions or heart attacks.

For example, it would be difficult to use a cross-sectional study to figure out the prevalence of high cholesterol for people that have heart attacks, since some people can die from heart attacks right away.

So, the people included in the study will only be those who have survived the heart attack, and they have different characteristics from those who died.

This is a problem because the results of the study will only apply to the sample population – individuals that had heart attacks who were included in the study – and not to the target population – anyone in the general population that has a heart attack.

Alright, as a quick recap, cross-sectional studies measure both an exposure and an outcome at the same time, and they are often a cheap, quick, and convenient way to sample large groups of people.

Often times they are done through surveys, which offer a snapshot of a sample population. The results from cross-sectional studies can be used to calculate the outcome prevalence or exposure prevalence.

But, since they measure prevalence instead of incidence, cross-sectional studies can’t be used to determine a causal relationship between the exposure and the outcome and they can’t study diseases with short durations.