OHDSI Home | Forums | Wiki | Github

Chinese ethnicity

Whatever you guys like. What about the Koreans who have come during the 20th century? What about the Koreans running from the North right now? Look: As you probably noticed I have no preference what you store and how. We know that the biological difference is negligible, and the cultural difference after living >100 years in China has probably subsided. :slight_smile:

@Christian_Reich agreed, that is why I withdrew my post :wink:


Actually these four are also inside China.Maybe it ’s just my translation wrong.
I give you chinese name,then let the vocabulary group help?

Though I don’t want to be a racist, but many Korean thinks that Chinese Korean (or Korean Chinese)朝鲜族 is a different ethnic group from Korean-Korean (韓民族).

PS. personally, I think ‘ethnicity’ is a kind of fiction (I’m not sure that Korean, Chinese and Japanese are really different ‘ethnic’ group biologically).

I agree too.so the “english name” we can discuss

Oh boy. I already stressed my opinion in the other debate. Total fiction. Even Japanese, Chinese and Koreans - I would argue the differences are due to small founding populations in the past with random disequilibrium of some genes. But people keep saying “If you study outcomes you will find huge differences in the races and ethnicities, so there must be something”. Yes, it is socio-economical and cultural.

The world’s largest randomized trial happened in Germany, where one part of a biologically homogeneous population (in the West) was exposed to different socio-economical factors than the other (in the East). I used to be a subject in that study. Guess what! The outcomes were different. The Koreans are still running that same experiment, and despite identical gene pool the differences couldn’t be more drastic. The Northerners are smaller, live a lot shorter lives and have more diseases like tuberculosis and intestinal parasite infections. But hey, not all is bad in the North. Almost no obesity (except the Supreme Leader) and diabetes.

All right, I stop pontificating.

@Andrea: Please state the scientific use cases, and please provide how you want these ethnicities laid out, and we see what we can do.


@Christian_Reich .
As you said, I do believe that ethnicity is more cultural than biological. And it matters (Another my belief is that socio-economic or cultural difference is bigger than genetic difference across the world). Chinese-Korean(朝鲜族) is different from Korean because they are minority both in China and Korea. And it means something (This is really important feature for PLE or PLP!)

Again, because the ethnicity is fiction originated from history and culture, I think we need to admit ethnic groups if the history or culture divides it. :slight_smile:

Fair enough. And you seem to have mentioned a use case (“PLE or PLP”). So, now I need a list of those “ethnicities” you guys can live with. And btw, before you get too creative: Make sure that list is also supported by the data. If the data cannot distinguish between Chinese-Koreans and Korean-Koreans (and maybe Korean-Chinese) then don’t give them to me.

1 Different ethnic groups have different genes, and some susceptible genes may be found.
2 Different ethnic groups have different rates of disease
3 Migrant population of different ethnicities can reflect the current status of medical insurance participation and its influencing factors
This is what I personally can think of so far


Now I need the list of ethnicities to support 1-3 in Asia (and elsewhere, but you may know better the Asian situation). Which means the above list is a good start, but you somehow need to figure out the “migrant” status, I guess. The Chinese-Koreans, Korean-Koreans, Chinese-Chinese and Korean-Chinese. (Sounds funnier than it is but I think by these you mean members of an ethnic group living as a minority inside another ethnic group, and the Koreans and Chinese have this situation both ways, if I am not mistaken).

Maybe Ethnicity belongs in the Observation table.

Has a WG been created? As far as I can see, it will not just require an expansion of the ethnicity vocabulary. To do it right will require hierarchies, ancestors, and relationships.

Do people belong so a single ethnicity? How often will ethnicities change? Who will be responsible for maintaining it?

The longer I think about it, the bigger the questions get.

1-2 Han Zhuang Hui Manchu Uyghur Miao
3 you are not mistaken.Because I’m not sure what the ethnic group of the floating population is so I need all 56 ethnic groups

sorry, my poor english :joy:may be it’s better to say floating instead of migrant

Summary of my points from the other debate:

  • Ethnicity is most usefully defined as a self-identified cultural concept that is distinct from and complementary to race.
  • It is useful because identification with an ethnic group can be associated with attitudes, beliefs, and behaviors that affect health.
  • It is usually less reliably observed than race because self-identification, beliefs, and attitudes are inherently subjective and the definition of an ethnic group may change over time and across regions. This variability and subjectivity might make the OBSERVATION table a better fit than the PERSON table.
  • There is an OBO-compliant Ethnicity Ontology (EO ) that might be worth investigation to assess its coverage, hierarchic structure, extent of adoption by researchers, EHR vendors, etc.
1 Like

Thank you @Andrew for nice summary! And I do agree that we need to look at OBO-compliant Ethnicity Ontology.

Still, I do not agree with that ethnicity in observation table or domain.

First, even though ethnicity is perceived one, it’s hardly changed along one’s life.
Second, current OMOP convention and OHDSI method library supports ethnicity in person table domain.
Third, we’ve already experienced how important the harmony between structure and method library is.

Hi @Andrew
thanks for pointing that out.
Unfortunately CSV file is unavailable for download. I don’t want to spend time parsing OWL or XML file.
Do you have this CSV (or other plain text) file?

A simple approach to have such ethnic groups is to look at the patient volume distribution. Except Han, the minor ethnicities with the top largest volumes or specific interest in your research can be displayed respectively. and then combine the left small volume ethnicities with no strong meaningfulness together.

You may use TopBraid to do a SPAQL query for EO and export the result in csv you wanted. Also, please explore protégé plug-in: https://github.com/protegeproject/csv-export-plugin, which may help.
Also, please see my comments about EO in this thread: Race and Ethnicity in the OMOP CDM
May give you some ideas.

Hi all,

Please see this: https://bioportal.bioontology.org/ontologies/SNOMED-Ethnic-Group/?p=classes&conceptid=root
Please note that no subgroups for Chinese class (under Asian - ethnic group) yet. It maybe possible to just add @Andrea’s subgroups there.

thanks for the link.then,how to add it?

Let’s continue this here and close this post.