Epidemiology has distributions and fluctuations (Easy Epidemiology Explanation 9)

We are delivering the "Easy Epidemiology Explanation" series, which allows you to naturally become well-versed in "epidemiology" and understand its importance and how to apply it. You can check all the articles in this magazine.

In any epidemiological research design, the population and measurement methods are important. We have already explained the "population," and in the previous note, I conveyed that it is very important to appropriately set what kind of people should be studied in your research. We also explained the concept of "population representativeness."

So, this time we will move on to the content of "measurement." When conducting epidemiological research, how should you measure the items you want to measure? You will find that deciding this is also very profound (Reference 1).

● Important populations and measurements in epidemiological research

Epidemiological research is conducted using various methods (research designs), as already explained. I will show the figure used at that time (Figure 1) again below.

On the left, it is shown in the figure that "the population and measurement methods are important in any epidemiological research." These two are factors that are always involved in any epidemiological research design, which is why they are written in this position and with this layout.

● It becomes science because it is measured

And this time, it is an explanation of "measurement." In a previous note article, I explained the definition of epidemiology, and in it, I conveyed that epidemiology is a discipline that "clarifies the frequency and distribution of various health-related events and the factors that influence them." " Clarifying the frequency and distribution " of health-related events is, in short, measurement, isn't it? For example, you will measure various things such as blood pressure values, height and weight, and dietary intake. And you will explain the results using numerical values that everyone can recognize in the same way, or using precise terminology.

In this way, it was necessary in the world of science to express things so that everyone can perceive them in the same way. So, what kind of characteristics does 'measurement' in epidemiology have?

● Values have distributions (inter-individual variation)

First, epidemiology has the characteristic of measuring a 'population'. We must measure many people living their daily lives and describe the state of the whole. And, for each individual, that value is different. This state is called having a 'distribution'. To show a distribution in a way that is easy to understand visually, it is good to use a graph called a histogram. For example, you can measure daily salt intake using a method called 24-hour urine collection, and the result will look like the diagram in Figure 2 (Reference 2).

For salt intake, there are people who consume less than 4.8 g/day, and there are people who consume 21.3 g/day or more. The most common amount seems to be 11–13 g/day. Since everyone consumes a different amount of salt, it means that 'there are people with various values'. Even if you list all these amounts in a table, it is just a row of numbers, and it is difficult to understand what the characteristics are. Therefore, these measured values are summarized and expressed using statistical values such as 'mean', 'median', and 'maximum value'. By doing so, you can begin to understand the situation of the values in that population. Showing them with statistical values is convenient for clearly showing the values of a population.

In this way, measurement in epidemiology results in values that have a distribution, and it is necessary to devise ways to think about which statistical figures should be used to explain them clearly. In epidemiological terminology, the state where values differ from person to person is called having 'inter-individual variation'. This is the first characteristic.

● Values fluctuate (intra-individual variation)

Next, health-related events measured in epidemiology also have the characteristic that values are prone to change, and it is not known if a single measurement is an accurate value, as well.

For example, even when you measure your weight at home, many people have probably experienced that the value measured the first time is different from the value measured when you step off the scale and measure again. Also, regarding blood pressure, don't you feel that it is rarer for the first measurement and the second measurement to be exactly the same? In this way, even for the same person, the value differs depending on the time, place, and conditions of measurement. This state is called having 'intra-individual variation'. This is the second characteristic.

As you can see, measuring the daily state of people living their daily lives is quite difficult. For that reason, it must be decided in advance under what conditions the values used for measurement are obtained. Measurement methods for values used in disease diagnosis are established so that the method is almost the same regardless of which medical institution performs the measurement. Deciding on a measurement method is an important thing.

● Summary

In epidemiological research, we perform 'measurement' to scientifically demonstrate various health-related events. At that time, because of the epidemiological characteristic of 'measuring people in a population', there are characteristics to the measurement as well. One is that many people have different values, which is inter-individual variation, and the other is that even when measuring a single person, the value changes depending on the measurement situation, which is intra-individual variation. To be able to recognize that there is inter-individual variation, population values are expressed not as a list of the values themselves, but as statistical values such as the population mean or median. Also, because there is inter-individual variation, measurements must be performed using a predetermined method so that everyone can be measured using the same measurement method.

Furthermore, deciding on that 'measurement method' is also difficult. To be continued next time.

Click here for the first installment of the 'Easy Epidemiology Explanation' series

Click here for the continuation of the 'Easy Epidemiology Explanation' series

[Email Magazine]
10 points to determine if food information is reliable We are sharing this. Please register!
https://hers-m-and-s.com/p/r/sPWrxMBU

[Online Salon]
You can watch course videos and materials. We also accept questions about nutritional epidemiology. Please join us!
https://yoor.jp/room/kobayan_nutrepi

[References]
1. Satoshi Sasaki. Easy-to-understand EBN and Nutritional Epidemiology. Dobunshoin. 2005.
2. Asakura K, et al. Br J Nutr 2014; 112: 1195-205.

Toward a society where all 100-year-olds can choose and eat their own meals.

We will provide assistance to help make the 100,000 meals you eat in a lifetime a time for building better eating habits.

Please come back and read again.
If you liked the article, please give it a "Like" reaction!
It encourages me!

[Tips for discerning food and health nutrition information]

With these 5 steps, you can easily judge whether food and health information is reliable.