I am conducting research to assess the impact of variables on Alzheimer's risk. I am looking to use a cox proportional hazards model Library(survival). My current dataframe has the following column names:
[1] "Autopsy_ID" "Postcode"
[3] "Age" "Gender"
[5] "AD"
[7] "Diabetes_Mellitus" "Stroke"
[9] "Hypercholestrolaemia" "Cerebrovascular_disease"
[11] "IHD" "Hypertension"
[13] "Parkinsons_disease" "Lewy_body_dementia"
[15] "VD" "Cardiac_disease"
[17] "Dementia" "Allele_2"
[19] "Allele_3" "Allele_4"
[21] "ozone" "pm10"
[23] "pm2.5" "nox"
[25] "imd_score"
I am trying to predict risk of developing Alzhiemers from all these variables. I currently don't have a time variable but would it be plausible to use age as a time variable and formulate my cox proportional hazards model as the following code?
cox_model= coxph(Surv(Age, AD) ~ + Gender + Diabetes_Mellitus + Stroke + Hypercholestrolaemia + Cerebrovascular_disease + IHD + Hypertension + Lewy_body_dementia + VD + Cardiac_disease + Dementia + pm10 + Allele_2Allele_3 + Allele_2 * Allele_4 + Allele_3Allele_4 + Allele_4, data = spatial_join)