dm2 x alz
alice
09 dec 2020
dm2 <- read_excel("dm2_r.xlsx")
surf <- read_csv("freesurfer_data.csv")
merged <- merge(dm2,surf,by="RID")
codes <- unname(unlist(read_excel("aseg_codes.xlsx", col_names = FALSE)))
names <- unname(unlist(read_excel("aseg_names.xlsx", col_names = FALSE,range = cell_cols("B"))))
brains <- data.frame(codes,names)
keep <- c("RID","VISCODE.y","COLPROT.x","COLPROT.y","DX_bl","AGE","PTGENDER","APOE4","MMSE")
df <- subset(merged, select=c(keep,codes))Overview of Data
From our hospital record data, we have 152 subjects with a diagnosis of Diabetes type 2.
When we merge these with freesurfer data, our count is reduced to 32 observations of patients for whom we have freesurfer data and have a diagnosis of DM2.
Summary Table: DM2+ hosp records only (n = 152)
Here, we look at summary statistics for the 152 records of patients documented has having diabetes.
tbl_summary(dm2[,5:9], by= DX_bl)%>%
add_p() %>%
modify_spanning_header(everything() ~ "**SUMMARY OF PT DATA BY DIAGNOSIS**") %>%
bold_labels()| SUMMARY OF PT DATA BY DIAGNOSIS | ||||
|---|---|---|---|---|
| Characteristic | AD, N = 301 | CN, N = 361 | MCI, N = 861 | p-value2 |
| AGE | 76 (70, 80) | 76 (72, 78) | 72 (68, 77) | 0.040 |
| PTGENDER | 0.5 | |||
| Female | 12 (40%) | 11 (31%) | 25 (29%) | |
| Male | 18 (60%) | 25 (69%) | 61 (71%) | |
| APOE4 | <0.001 | |||
| 0 | 6 (21%) | 30 (83%) | 45 (53%) | |
| 1 | 19 (68%) | 5 (14%) | 33 (39%) | |
| 2 | 3 (11%) | 1 (2.8%) | 7 (8.2%) | |
| Unknown | 2 | 0 | 1 | |
| MMSE | 24.00 (22.00, 25.00) | 29.00 (29.00, 30.00) | 28.00 (26.00, 29.00) | <0.001 |
|
1
Statistics presented: median (IQR); n (%)
2
Statistical tests performed: Kruskal-Wallis test; chi-square test of independence; Fisher's exact test
|
||||
Data Viz: DM2+ hosp records only (n = 152)
Here, we look at plots briefly showing the general distribution of these 152 patients
Summary Table: merged data (n = 32)
Next, the 152 pulled Diabetes patient records are joined across the available FreeSurfer data. The overlap between the two sets yielded 32 rows. This is insufficient for analysis, but I go through the steps below to establish the method.
I believe something was wrongly interpretted at one point along the way, because none of the FreeSurfer patients have been labeled “AD.” I am hoping we will look into this and resolve some issue, that will then increase the sample size to a suitable number.
tbl_summary(merged[,5:9], by= DX_bl)%>%
add_p() %>%
modify_spanning_header(everything() ~ "**SUMMARY OF MERGED PT DATA BY DIAGNOSIS**") %>%
bold_labels()## Warning in 'add_p()' for variable 'AGE':
## simpleWarning in wilcox.test.default(x = c(72.6, 70.6, 70.6, 76.9, 66.5, 62.3, : cannot compute exact p-value with ties| SUMMARY OF MERGED PT DATA BY DIAGNOSIS | |||
|---|---|---|---|
| Characteristic | CN, N = 91 | MCI, N = 231 | p-value2 |
| AGE | 66.8 (65.9, 70.6) | 64.4 (61.9, 69.7) | 0.11 |
| PTGENDER | 0.4 | ||
| Female | 1 (11%) | 7 (30%) | |
| Male | 8 (89%) | 16 (70%) | |
| APOE4 | 0.3 | ||
| 0 | 8 (89%) | 17 (74%) | |
| 1 | 0 (0%) | 5 (22%) | |
| 2 | 1 (11%) | 1 (4.3%) | |
| MMSE | 0.4 | ||
| 26 | 0 (0%) | 3 (13%) | |
| 27 | 0 (0%) | 3 (13%) | |
| 28 | 0 (0%) | 3 (13%) | |
| 29 | 6 (67%) | 8 (35%) | |
| 30 | 3 (33%) | 6 (26%) | |
|
1
Statistics presented: median (IQR); n (%)
2
Statistical tests performed: Wilcoxon rank-sum test; Fisher's exact test
|
|||
Data Viz: merged data (n = 32)
Brief Stats: merged data (n = 32)
m <- cor(df[,c(6,8:9)],df[,10:38])
rownames(m) <- c("$AGE","$APOE4","$MMSE")
corrplot(m,bg="midnightblue", tl.cex=2, cl.cex=2)
corrplot(m,method="number",number.cex = 1.4,tl.cex=2,cl.cex=2, bg="midnightblue")
kbl(brains) %>%
kable_styling(bootstrap_options = c("striped", "hover", "condensed"), full_width = F, font_size = 16, fixed_thead = list(enabled = T, background = "lightcoral"))| codes | names |
|---|---|
| ST101SV | Subcortical Volume (aseg.stats) of RightPallidum |
| ST112SV | Subcortical Volume (aseg.stats) of RightPutamen |
| ST11SV | Subcortical Volume (aseg.stats) of LeftAccumbensArea |
| ST120SV | Subcortical Volume (aseg.stats) of RightThalamus |
| ST12SV | Subcortical Volume (aseg.stats) of LeftAmygdala |
| ST16SV | Subcortical Volume (aseg.stats) of LeftCaudate |
| ST17SV | Subcortical Volume (aseg.stats) of LeftCerebellumCortex |
| ST18SV | Subcortical Volume (aseg.stats) of LeftCerebellumWM |
| ST1SV | Subcortical Volume (aseg.stats) of Brainstem |
| ST28SA | Surface Area (aparc.stats) of LeftHemisphereWM |
| ST29SV | Subcortical Volume (aseg.stats) of LeftHippocampus |
| ST42SV | Subcortical Volume (aseg.stats) of LeftPallidum |
| ST53SV | Subcortical Volume (aseg.stats) of LeftPutamen |
| ST61SV | Subcortical Volume (aseg.stats) of LeftThalamus |
| ST70SV | Subcortical Volume (aseg.stats) of RightAccumbensArea |
| ST71SV | Subcortical Volume (aseg.stats) of RightAmygdala |
| ST75SV | Subcortical Volume (aseg.stats) of RightCaudate |
| ST76SV | Subcortical Volume (aseg.stats) of RightCerebellumCortex |
| ST77SV | Subcortical Volume (aseg.stats) of RightCerebellumWM |
| ST87SA | Surface Area (aparc.stats) of RightHemisphereWM |
| ST88SV | Subcortical Volume (aseg.stats) of RightHippocampus |
| ST147SV | Subcortical Volume (aseg.stats) of LeftCorticalGM |
| ST148SV | Subcortical Volume (aseg.stats) of RightCorticalGM |
| ST149SV | Subcortical Volume (aseg.stats) of CorticalGM |
| ST150SV | Subcortical Volume (aseg.stats) of LeftCorticalWM |
| ST151SV | Subcortical Volume (aseg.stats) of RightCorticalWM |
| ST152SV | Subcortical Volume (aseg.stats) of CorticalWM |
| ST153SV | Subcortical Volume (aseg.stats) of SubcorticalGM |
| ST154SV | Subcortical Volume (aseg.stats) of TotalGM |
