imaging data pulled: 2019-12-04
clinical data pulled: 2019-12-04
code written: 2019-09-30
last ran: 2019-12-08
website: http://rpubs.com/navona/NM_LC_demo
github: https://github.com/navonacalarco/CurAge/blob/master/scripts/analysis/04_NM_ROI_LC_demoAnalysis.Rmd

Here we conduct a preliminary group-wise analysis of neuromelanin in the locus coeruleus (LC). We use data from the N=31 participants with neuromelanin scans as of 2019-12-08, including 16 patients with late-life depression (LLD) and 15 healthy controls (HC). I have not omitted participant SEN015 (that participant was excluded in earlier pilot analyses on the basis of an outlier value in the SN.)

NM by group

We calculated the primary NM outcome variables reported in Chen (2014) (i.e., volume and CNR) in all slices in which the LC was visually apparent. Volume is a simple sum across all slices. CNR is a weighted average across all slices by volume.

LLD, n=16 HC, n=15 p
Slice count 1.81 (0.83) 2.20 (0.86) .213
Volume 16.25 (7.88) 15.79 (13.61) .908
CNR 5.14 (0.51) 4.88 (0.35) .106

NM by group distribution

The null results above are not an upshot of outliers or non-normalcy in slice count, or the two main outcome variables.

Regression analysis: sex

We want to look at the differences in volume / CNR (outcome variable) between sexes (predictor variable), as this will indicate if sex should be covaried for in the above analyses. If there are no significant differences in volume or CNR, sex should not be covaried for.

## 
## Call:
## lm(formula = totalVolume ~ Sex, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -17.305  -5.219  -2.105   0.781  37.495 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   18.105      2.453   7.381 3.93e-08 ***
## SexM          -5.372      3.943  -1.363    0.184    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 10.69 on 29 degrees of freedom
## Multiple R-squared:  0.06017,    Adjusted R-squared:  0.02776 
## F-statistic: 1.856 on 1 and 29 DF,  p-value: 0.1835

Above, we see that the average volume for females is 18.11 and for males is slightly lower at 12.73. The p value for the dummy variable SexM is not significant, suggesting that there is no statistical evidence of a difference in volume between sexes.

## 
## Call:
## lm(formula = weightedCNR ~ Sex, data = df)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.87463 -0.20897 -0.08194  0.27036  1.31601 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   5.1208     0.1013  50.550   <2e-16 ***
## SexM         -0.2796     0.1628  -1.717   0.0966 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4416 on 29 degrees of freedom
## Multiple R-squared:  0.0923, Adjusted R-squared:  0.061 
## F-statistic: 2.949 on 1 and 29 DF,  p-value: 0.09661

Above, we see that the average CNR for females is 5.12 and for males is slightly lower at 4.84. The p value for the dummy variable SexM is not significant, suggesting that there is no statistical evidence of a difference in CNR between sexes.

Regression analysis: age

We want to look at the differences in volume / CNR (outcome variable) between age (predictor variable), as this will indicate if age should be covaried for in the above analyses. As there are no significant differences in volume or CNR, age should not be covaried for.

model <- lm(totalVolume ~ Age, data = df)
summary(model)
## 
## Call:
## lm(formula = totalVolume ~ Age, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -16.261  -5.541  -3.713   3.573  34.528 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)
## (Intercept) -15.0271    18.6350  -0.806    0.427
## Age           0.4457     0.2661   1.675    0.105
## 
## Residual standard error: 10.53 on 29 degrees of freedom
## Multiple R-squared:  0.08821,    Adjusted R-squared:  0.05677 
## F-statistic: 2.806 on 1 and 29 DF,  p-value: 0.1047
model <- lm(weightedCNR ~ Age, data = df)
summary(model)
## 
## Call:
## lm(formula = weightedCNR ~ Age, data = df)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.79481 -0.33131 -0.02052  0.24945  1.27726 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  6.19276    0.78994   7.840  1.2e-08 ***
## Age         -0.01694    0.01128  -1.502    0.144    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4464 on 29 degrees of freedom
## Multiple R-squared:  0.07216,    Adjusted R-squared:  0.04016 
## F-statistic: 2.255 on 1 and 29 DF,  p-value: 0.144

Sex-by-diagnosis interaction

We want to know if the relationship between sex and SN volume / CNR differs between LLD and HC groups.

totalVolume ~ Sex * Diagnosis

#first, look at interaction with LLD as reference
summary(lm(data=df, totalVolume ~ Sex * Diagnosis))
## 
## Call:
## lm(formula = totalVolume ~ Sex * Diagnosis, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -19.667  -5.974  -2.467   0.759  35.133 
## 
## Coefficients:
##                  Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        17.015      3.051   5.578 6.48e-06 ***
## SexM               -4.082      7.045  -0.579    0.567    
## DiagnosisHC         3.451      5.428   0.636    0.530    
## SexM:DiagnosisHC   -3.718      9.123  -0.408    0.687    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 11 on 27 degrees of freedom
## Multiple R-squared:  0.07407,    Adjusted R-squared:  -0.02881 
## F-statistic:  0.72 on 3 and 27 DF,  p-value: 0.5488
#now, look at interaction with HC as reference
df$Diagnosis_recode <- fct_rev(df$Diagnosis)
summary(lm(data=df, totalVolume ~ Sex * Diagnosis_recode))
## 
## Call:
## lm(formula = totalVolume ~ Sex * Diagnosis_recode, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -19.667  -5.974  -2.467   0.759  35.133 
## 
## Coefficients:
##                          Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                20.467      4.490   4.558 9.99e-05 ***
## SexM                       -7.800      5.797  -1.346    0.190    
## Diagnosis_recodeLLD        -3.451      5.428  -0.636    0.530    
## SexM:Diagnosis_recodeLLD    3.718      9.123   0.408    0.687    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 11 on 27 degrees of freedom
## Multiple R-squared:  0.07407,    Adjusted R-squared:  -0.02881 
## F-statistic:  0.72 on 3 and 27 DF,  p-value: 0.5488

weightedCNR ~ Sex * Diagnosis

#first, look at interaction with LLD as reference
summary(lm(data=df, weightedCNR ~ Sex * Diagnosis))
## 
## Call:
## lm(formula = weightedCNR ~ Sex * Diagnosis, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -0.9495 -0.2313 -0.1130  0.2324  1.2411 
## 
## Coefficients:
##                  Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        5.1957     0.1241  41.861   <2e-16 ***
## SexM              -0.2908     0.2866  -1.014    0.319    
## DiagnosisHC       -0.2372     0.2209  -1.074    0.292    
## SexM:DiagnosisHC   0.1522     0.3712   0.410    0.685    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4475 on 27 degrees of freedom
## Multiple R-squared:  0.132,  Adjusted R-squared:  0.03553 
## F-statistic: 1.368 on 3 and 27 DF,  p-value: 0.2736
#now, look at interaction with HC as reference
summary(lm(data=df, weightedCNR ~ Sex * Diagnosis_recode))
## 
## Call:
## lm(formula = weightedCNR ~ Sex * Diagnosis_recode, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -0.9495 -0.2313 -0.1130  0.2324  1.2411 
## 
## Coefficients:
##                          Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                4.9585     0.1827  27.141   <2e-16 ***
## SexM                      -0.1386     0.2359  -0.587    0.562    
## Diagnosis_recodeLLD        0.2372     0.2209   1.074    0.292    
## SexM:Diagnosis_recodeLLD  -0.1522     0.3712  -0.410    0.685    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4475 on 27 degrees of freedom
## Multiple R-squared:  0.132,  Adjusted R-squared:  0.03553 
## F-statistic: 1.368 on 3 and 27 DF,  p-value: 0.2736

Age-by-diagnosis interaction

totalVolume ~ Age * Diagnosis

#first, look at interaction with LLD as reference
summary(lm(data=df, totalVolume ~ Age * Diagnosis))
## 
## Call:
## lm(formula = totalVolume ~ Age * Diagnosis, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -16.611  -5.772  -1.631   2.239  33.789 
## 
## Coefficients:
##                 Estimate Std. Error t value Pr(>|t|)
## (Intercept)      -0.2500    31.2592  -0.008    0.994
## Age               0.2444     0.4614   0.530    0.601
## DiagnosisHC     -32.1620    40.8631  -0.787    0.438
## Age:DiagnosisHC   0.4250     0.5874   0.724    0.476
## 
## Residual standard error: 10.72 on 27 degrees of freedom
## Multiple R-squared:  0.1201, Adjusted R-squared:  0.02237 
## F-statistic: 1.229 on 3 and 27 DF,  p-value: 0.3184
#now, look at interaction with HC as reference
summary(lm(data=df, totalVolume ~ Age * Diagnosis_recode))
## 
## Call:
## lm(formula = totalVolume ~ Age * Diagnosis_recode, data = df)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -16.611  -5.772  -1.631   2.239  33.789 
## 
## Coefficients:
##                         Estimate Std. Error t value Pr(>|t|)  
## (Intercept)             -32.4120    26.3183  -1.232   0.2287  
## Age                       0.6694     0.3635   1.842   0.0765 .
## Diagnosis_recodeLLD      32.1620    40.8631   0.787   0.4381  
## Age:Diagnosis_recodeLLD  -0.4250     0.5874  -0.724   0.4756  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 10.72 on 27 degrees of freedom
## Multiple R-squared:  0.1201, Adjusted R-squared:  0.02237 
## F-statistic: 1.229 on 3 and 27 DF,  p-value: 0.3184

weightedCNR ~ Age * Diagnosis

#first, look at interaction with LLD as reference
summary(lm(data=df, weightedCNR ~ Age * Diagnosis))
## 
## Call:
## lm(formula = weightedCNR ~ Age * Diagnosis, data = df)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.87498 -0.24062 -0.08542  0.31495  1.03534 
## 
## Coefficients:
##                 Estimate Std. Error t value Pr(>|t|)    
## (Intercept)      7.84419    1.22860   6.385 7.73e-07 ***
## Age             -0.04004    0.01813  -2.208   0.0359 *  
## DiagnosisHC     -3.32847    1.60606  -2.072   0.0479 *  
## Age:DiagnosisHC  0.04504    0.02309   1.951   0.0615 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4214 on 27 degrees of freedom
## Multiple R-squared:  0.2303, Adjusted R-squared:  0.1448 
## F-statistic: 2.693 on 3 and 27 DF,  p-value: 0.06599
#now, look at interaction with HC as reference
summary(lm(data=df, weightedCNR ~ Age * Diagnosis_recode))
## 
## Call:
## lm(formula = weightedCNR ~ Age * Diagnosis_recode, data = df)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.87498 -0.24062 -0.08542  0.31495  1.03534 
## 
## Coefficients:
##                          Estimate Std. Error t value Pr(>|t|)    
## (Intercept)              4.515718   1.034400   4.366 0.000167 ***
## Age                      0.004995   0.014287   0.350 0.729320    
## Diagnosis_recodeLLD      3.328471   1.606061   2.072 0.047897 *  
## Age:Diagnosis_recodeLLD -0.045040   0.023086  -1.951 0.061512 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.4214 on 27 degrees of freedom
## Multiple R-squared:  0.2303, Adjusted R-squared:  0.1448 
## F-statistic: 2.693 on 3 and 27 DF,  p-value: 0.06599

NM correlations

Volume. Here, we plot volume against continous demographic and clinical variables of interest (Pearson’s r).

The same data is plotted below, but with a unique regression line per group.

CNR. Here, we plot CNR against continous demographic and clinical variables of interest (Pearson’s r).

The same data is plotted below, but with a unique regression line per group.

Correlation values LC volume

Combined, n=31 LLD, n=16 HC, n=15
Age 0.297 0.186 0.388
Education -0.093 0.128 -0.259
PHQ9 0.013 -0.040 0.043
MADRS 0.004 0.053 -0.116
CIRS-G 0.039 0.370 -0.122
RBANS total 0.023 0.153 -0.036
RBANS immediate memory -0.059 -0.211 0.045
RBANS visuospatial -0.008 0.366 -0.232
RBANS language -0.085 0.099 -0.210
RBANS attention 0.072 0.074 0.079
RBANS delayed memory 0.148 0.075 0.182

Correlation values LC CNR

Combined, n=31 LLD, n=16 HC, n=15
Age -0.269 -0.470 0.111
Education -0.127 -0.158 0.222
PHQ9 0.319 0.261 -0.530
MADRS 0.310 0.262 -0.287
CIRS-G 0.291 0.200 0.243
RBANS total 0.061 -0.185 0.371
RBANS immediate memory 0.008 -0.181 0.488
RBANS visuospatial -0.103 -0.156 0.039
RBANS language -0.219 -0.354 0.003
RBANS attention 0.051 -0.091 0.363
RBANS delayed memory 0.206 -0.028 0.371