Hi everyone,
I am working on case control genetic data. I have data for SNP with genotypes AA, AG and GG. I want to check the effect of these individual genotypes on the disease outcome, which in my case is diabetes. Now I want to calculate unadjusted and adjusted (for age and gender) Odds ratio in SPSS. I calculated unadjusted Odds ratio by using multinomial regression (is this suitable for my data?). But do not know how to calculate Odds ratio after adjustment for age and geneder.
Secondly, How can I apply dominant, recessive and co-dominant model to check which fits more.
I would be highly obliged if you provide me the flowchart of commands like
Spss--> analyze--> regression--> .........
Thanks,
Misbah
How is your outcome coded? If it is dichotomous you can perform binary logistic regression. If it has more than two unordered levels than multinomial logistic regression suits.To obtain adjusted ORs you need to include those variables like age and gender into the model. The ORs obtained from that model will be adjusted ones. Model selection requires some other techniques.
Thank you Mehmet Sinan Lyisoy for answering my question
Genotypes are coded as 1, 2 and 3. while disease is coded as 0 and 1 for absence and presence of diabetes.
To obtain adjusted ORs, do I need to put age and gender in convariate box along with independent variable?
In binary logistic regression dialog box in SPSS, put age,gender and other independent variables into covariates box. But do not forget to indicate all categorical variables like gender inside categorical box.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS