Variables.Rmd

---
title: "Variables"
author: "Bryon Langford, Matthew Hoctor"
date: "7/12/2021"
output:
  html_document:
    number_sections: no
    theme: lumen
    toc: yes
    toc_float:
      collapsed: yes
      smooth_scroll: no
  pdf_document:
    toc: yes
  word_document:
    toc: yes
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

if (Sys.info()["sysname"] == "Windows") 
  knitr::opts_chunk$set(engine.path = list(stata = "C:/Program Files/Stata17/StataBE-64")) 
if ((Sys.info()["sysname"] == "Darwin"))
  knitr::opts_chunk$set(engine.path = list(stata = "/Applications/Stata/StataIC"))
```

```{r libraries, include=FALSE}
library(Statamarkdown)
```

# BP categories

## Median SBP

```{stata create median SBP}
use "data\NHANES0518.dta", clear

egen NmissSBP = rowmiss(bpxsy1 bpxsy2 bpxsy3 bpxsy4)
egen MaxSBP = rowmax(bpxsy1 bpxsy2 bpxsy3 bpxsy4)

egen MedSBP = rowmedian(bpxsy1 bpxsy2 bpxsy3 bpxsy4)
  replace  MedSBP=. if NmissSBP>2
  replace  MedSBP=. if NmissSBP==2 & MedSBP!=MaxSBP

save "data\NHANES0518_new.dta",replace
```

```{stata exploring missing SBP values}
use "data\NHANES0518.dta", clear

gen miss_SBP1=missing(bpxsy1)
gen miss_SBP2=missing(bpxsy2)
gen miss_SBP3=missing(bpxsy3)
gen miss_SBP4=missing(bpxsy4)

ta miss_SBP1 miss_SBP2
ta miss_SBP3 miss_SBP4

egen NmissSBP = rowmiss(bpxsy1 bpxsy2 bpxsy3 bpxsy4)

ta  NmissSBP miss_SBP1

#save "data\NHANES0518_new.dta",replace
```

## Median DBP

```{stata create median DBP}
use "data\NHANES0518_new.dta", clear

egen NmissDBP = rowmiss(bpxdi1 bpxdi2 bpxdi3 bpxdi4)
egen MaxDBP = rowmax(bpxdi1 bpxdi2 bpxdi3 bpxdi4)

egen MedDBP = rowmedian(bpxdi1 bpxdi2 bpxdi3 bpxdi4)
  replace  MedDBP=. if NmissDBP>2
  replace  MedDBP=. if NmissDBP==2 & MedDBP!=MaxDBP

save "data\NHANES0518_new.dta",replace
```

```{stata exploring missing DBP values}
use "data\NHANES0518.dta", clear

gen miss_DBP1=missing(bpxdi1)
gen miss_DBP2=missing(bpxdi2)
gen miss_DBP3=missing(bpxdi3)
gen miss_DBP4=missing(bpxdi4)

ta miss_DBP1 miss_DBP2
ta miss_DBP3 miss_DBP4

egen NmissDBP = rowmiss(bpxdi1 bpxdi2 bpxdi3 bpxdi4)

ta  NmissDBP miss_DBP1

egen NmissSBP = rowmiss(bpxsy1 bpxsy2 bpxsy3 bpxsy4)
ta NmissSBP NmissDBP

#save "data\NHANES0518_new.dta",replace
```

## BP Categories

```{stata BP categories}
use "data\NHANES0518_new.dta", clear

gen BP_cat=0
#  replace BP_cat=0 if MedSBP<120 & MedDBP<80
  replace BP_cat=1 if MedSBP>=120
  replace BP_cat=2 if MedSBP>=130 | MedDBP>=80
  replace BP_cat=3 if MedSBP>=140 | MedDBP>=90
  replace BP_cat=. if MedSBP==. | MedDBP==.
  
ta BP_cat

label define BP_cat 0 "Normal" 1 "Elevated" 2 "Stage 1 HTN" 3 "Stage 2 HTN"
label values BP_cat BP_cat
label variable BP_cat "BP Category"
ta BP_cat,m

save,replace
```

## Generate Missing BP Variable

```{stata}
use "data\NHANES0518_new.dta", clear

gen BP_miss=0
  replace BP_miss=1 if BP_cat==.

label define BP_miss 0 "Not Missing BP" 1 "Missing BP"
label values BP_miss BP_miss
label variable BP_miss "Meets BP Inclusion Criteria"

ta BP_miss,m

save,replace
```

## Elevated BP

```{stata}
use "data\NHANES0518_new.dta", clear

recode BP_cat 0=0 1=1 2=1 3=1 .=., gen(BP_abn)

label define BP_abn 0 "Normal BP" 1 "Abnormal BP"
label values BP_abn BP_abn
label variable BP_abn "Abnormal/Normal BP"

ta BP_cat BP_abn,m

save,replace
```

# Cannabis Use Categories

## MJ Use Categories

```{stata exploring data}
use "data\NHANES0518_new.dta", clear

ta duq200
ta duq220q
ta duq220u
```

```{stata MJ categories}
use "data\NHANES0518_new.dta", clear

recode duq200 2=0 1=1 7=. 9=., gen(MJ_cat)
  replace MJ_cat=2 if duq230<77

label define MJ_cat 0 "Never Use" 1 "Past Use" 2 "Current Use"
label values MJ_cat MJ_cat
label variable MJ_cat "Cannabis Use Category"
ta MJ_cat,m

ta MJ_cat duq200,m

save,replace
```

## MJ Frequency Categories

```{stata MJ Frequency Categories}
use "data\NHANES0518_new.dta", clear

recode duq230 min/10=0 11/20=1 21/30=2 31/max=., gen(MJ_freq)

label define MJ_freq 0 "1-10 days" 1 "11-20 days" 2 "21-30 days"
label values MJ_freq MJ_freq
label variable MJ_freq "Cannabis Use Frequency In the Past 30 Days"
ta MJ_freq,m

ta MJ_freq duq200,m

save,replace
```

## Generate Single MJ Exposure Variable

```{stata unitary MJ variable}
use "data\NHANES0518_new.dta", clear

gen MJ=MJ_cat
  replace MJ=3 if MJ_freq==1
  replace MJ=4 if MJ_freq==2

label define MJ 0 "Never Use" 1 "Past Use" 2 "1-10 times in past 30 days" 3 "11-20 times in past 30 days" 4 "21-30 times in past 30 days"
label values MJ MJ
label variable MJ "Cannabis Use"

ta MJ_cat MJ_freq,m
ta MJ,m
ta MJ_freq MJ,m

save,replace
```

## Generate Missing MJ Variable

```{stata}
use "data\NHANES0518_new.dta", clear

gen MJ_miss=0
  replace MJ_miss=1 if MJ_cat==.

label define MJ_miss 0 "Not Missing MJ" 1 "Missing MJ"
label values MJ_miss MJ_miss
label variable MJ_miss "Meets MJ Inclusion Criteria"

ta MJ_miss,m
ta MJ_miss BP_miss,m

save,replace
```

## Collapsed MJ variable

```{stata collapsed MJ variable}
use "data\NHANES0518_new.dta", clear

recode MJ 0=0 1=0 2=1 3=1 4=2 .=., gen(MJ2)

label define MJ2 0 "Never/Past Use" 1 "Light-Moderate 1-20 times in past 30 days" 2 "Frequent 21-30 times in past 30 days"
label values MJ2 MJ2
label variable MJ2 "Cannabis Use Collapsed"

ta MJ MJ2,m

save,replace
```

# Inclusion criteria

## MEC Examined

```{stata}
use "data\NHANES0518_new.dta", clear
gen MEC_response=1 
  replace MEC_response=0 if wtmec2yr==0

label define MEC_response 0 "Did not respond" 1 "Responded"
label values MEC_response MEC_response
label variable MEC_response "MEC Response"

ta MEC_response,m
ta MEC_response BP_miss,m
ta MEC_response MJ_miss,m

save,replace
```

## Age Inclusion Criteria, 20-59 Years

```{stata}
use "data\NHANES0518_new.dta", clear

recode ridageyr min/19=1 20/59=0 60/85=1 86/max=., gen(AGE_exclude)

label define AGE_exclude 0 "Age 20-59" 1 "Age Excluded"
label values AGE_exclude AGE_exclude
label variable AGE_exclude "Meets Age Inclusion Criteria"

ta AGE_exclude,m
ta AGE_exclude BP_miss,m
ta AGE_exclude MJ_miss,m

save,replace
```

## Overall Inclusion, Age, BP, MJ

```{stata}
use "data\NHANES0518_new.dta", clear

gen include=0
  replace include=1 if BP_miss==0 & MJ_miss==0 & AGE_exclude==0

label define include 0 "Excluded" 1 "Included"
label values include include
label variable include "Meets Overall Inclusion Criteria"

ta include,m
ta include BP_miss,m
ta include MJ_miss,m
ta include AGE_exclude,m

save,replace
```

# Covariate Coding

## Race/Ethnicity

```{stata}
use "data\NHANES0518_new.dta", clear

#ta ridreth1
recode ridreth1 1=3 2=4 3=1 4=2 5=4, gen(race_eth)

label define race_eth 1 "White-NH" 2 "Black_NH" 3 "Mex Am" 4 "Oth"
label values race_eth race_eth
label variable race_eth "Recoded Race & Ethnicity"

ta race_eth
ta ridreth1 race_eth,m

save,replace
```

## Smoking

```{stata}
use "data\NHANES0518_new.dta", clear

recode smd650 min/10=2 11/20=3 21/95=4 96/max=., gen(SMK_cat)
  replace SMK_cat=1 if smq040==3
  replace SMK_cat=0 if smq020==2

label define SMK_cat 0 "Never" 1 "Past Smoker" 2 "Light, 1-10 cif/day" 3 "Moderate, 11-20 cig/day" 4 "Heavy, 21+ cig/day"
label values SMK_cat SMK_cat
label variable SMK_cat "Smoking Category"

ta SMK_cat,m
# ta SMK_cat smd650,m
# ta SMK_cat BP_miss,m
# ta SMK_cat MJ_miss,m

save,replace
```

## Gender

```{stata}
use "data\NHANES0518_new.dta", clear

recode riagendr 2=0 1=1, gen(gndr)

label define gndr 0 "Female" 1 "Male"
label values gndr gndr
label variable gndr "Gender"

ta gndr,m
# ta gndr riagendr,m

save,replace
```

## Alcohol

```{stata}
use "data\NHANES0518_new.dta", clear

recode alq130 min/2=0 3/4=1 5/95=2 99/max=., gen(AL_cat)
  replace AL_cat=0 if alq101==2 | alq111==2 | alq121==0
  replace AL_cat=1 if alq130==2 & riagendr==2
  replace AL_cat=2 if alq130==4 & riagendr==2

label define AL_cat 0 "None-Light" 1 "Moderate" 2 "Heavy"
label values AL_cat AL_cat
label variable AL_cat "Alcohol Use Category"

ta AL_cat,m
# ta AL_cat BP_miss,m
# ta AL_cat MJ_miss,m

save,replace
```

## Healthy Eating Index

```{stata}
use "data\NHANES0518_new.dta", clear

gen hei2015=hei2015_total_score
  replace hei2015=. if dr1tkcal==. & dr2tkcal==.
  replace hei2015=. if dr1tkcal<500 & dr1tkcal!=.
  replace hei2015=. if dr1tkcal>5000 & dr1tkcal!=.
  replace hei2015=. if dr2tkcal<500 & dr2tkcal!=.
  replace hei2015=. if dr2tkcal>5000 & dr2tkcal!=.

# replace hei2015_total_score=. if dr1tkcal<500
# replace hei2015_total_score=. if dr1tkcal>5000
# replace hei2015_total_score=. if dr2tkcal<500
# replace hei2015_total_score=. if dr2tkcal>5000

label variable hei2015 "Healthy Eating Index"

gen HEI_miss=0
  replace HEI_miss=1 if hei2015==.
ta include HEI_miss

save,replace
```

## Education Category

```{stata}
use "data\NHANES0518_new.dta", clear

recode dmdeduc2 1=1 2=2 3=3 4=4 5=5 7=. 9=., gen(EDUC_cat)
label define EDUC_cat 1 "Less than 9th Grade" 2 "Less than High School" 3 "High School/GED" 4 "Some College" 5 "College Graduate"
label values EDUC_cat EDUC_cat
label variable EDUC_cat "Recode Education Level"

ta dmdeduc2 EDUC_cat,m

save,replace
```

## Survey Year

```{stata}
use "data\NHANES0518_new.dta", clear

label define sddsrvyr 4 "2005-2006" 5 "2007-2008" 6 "2009-2010" 7 "2011-2012" 8 "2013-2014" 9 "2015-2016" 10 "2017-2018"
label values sddsrvyr sddsrvyr
label variable sddsrvyr "Survey Year"

ta sddsrvyr,m

save,replace
```

## BMI implausible values

```{stata}
use "data\NHANES0518_new.dta", clear

gen BMI_miss=0
  replace BMI_miss=1 if bmxbmi==.
ta BMI_miss include,m

replace bmxbmi=. if bmxbmi>65

replace BMI_miss=1 if bmxbmi==.
ta BMI_miss include,m

save,replace
```


## MEC weight

```{stata}
use "data\NHANES0518_new.dta", clear
gen wtmec12yr = (1/7) * wtmec2yr
save,replace
```

# Cross tabulation of missing variables for T1

```{stata}
use "data\NHANES0518_new.dta", clear
ta EDUC_cat include,m
ta gndr include,m
ta SMK_cat include,m
ta AL_cat include,m

gen AGE_miss =0
  replace AGE_miss=1 if ridageyr==.
ta AGE_miss include,m

# gen BMI_miss=0
#   replace BMI_miss=1 if bmxbmi==.
ta BMI_miss include,m

gen PIR_miss=0
  replace PIR_miss=1 if indfmpir==.
ta PIR_miss include,m

# gen HEI_miss=0
#   replace HEI_miss=1 if hei2015==.
ta HEI_miss include,m
```

# Cross tabulation of missing variables for T1, by MJ_cat

```{stata}
use "data\NHANES0518_new.dta", clear
svyset sdmvpsu [pw=wtmec12yr], strata(sdmvstra)

ta EDUC_cat MJ_cat if include==1,m
ta gndr MJ_cat if include==1,m
ta SMK_cat MJ_cat if include==1,m
ta AL_cat MJ_cat if include==1,m

gen AGE_miss =0
  replace AGE_miss=1 if ridageyr==.
ta AGE_miss MJ_cat if include==1,m

# gen BMI_miss=0
#   replace BMI_miss=1 if bmxbmi==.
ta BMI_miss MJ_cat if include==1,m

gen PIR_miss=0
  replace PIR_miss=1 if indfmpir==.
ta PIR_miss MJ_cat if include==1,m

# gen HEI_miss=0
#   replace HEI_miss=1 if hei2015_total_score==.
ta HEI_miss MJ_cat if include==1,m
svy, subpop(if include==1): ta HEI_miss MJ_cat, col percent
```

# Create .zip file for github upload

```{r}
zip("data/NHANES0518_new.zip","data/NHANES0518_new.dta")
```