In a regression model with maltreatment as an explanatory variable and executive functions as the outcome variable while controlling for participant’s age, education in years and trauma load, we found a significant association between maltreatment and executive functions (see Table 2 and Fig. 1). The regression model explained 20% of the variability in executive functions. In addition, trauma load and maltreatment significantly correlated with mental health problems (see Table 2 and Fig. 1). This model explained 26% of the variance of mental health problems. To investigate whether the association between maltreatment and executive functions was mediated by mental health problems, we conducted a simple mediation analysis with maltreatment as an independent variable and executive functions as dependent variable (Fig. 1). When the composite score of mental health problems was added as a mediator variable, the indirect effect of child maltreatment via mental health problems was not significant (see Table 2, Bootstrap results: B = 0.011, SE = 0.023, 95%-CI: -0.036, 0.059).
4.2. Associations between maltreatment, mental health, and working memory
After controlling for age, years of education, and trauma load, maltreatment was significantly associated with working memory (Table 3 and Fig. 2). The regression model explained 4% of the variability in working memory. Furthermore, maltreatment and trauma load were significantly associated with mental health problems (Table 3 and Fig. 2). The regression model explained 26% of the variations of mental health problems. To investigate whether the association between maltreatment and working memory is mediated by mental health problems, we conducted a simple mediation analysis with maltreatment as independent variable, working memory as dependent variable (Fig. 2). When the mental health composite score was added as a mediator variable, the indirect effect of child maltreatment via mental health problems was not significant (see Table 3, Bootstrap results: B = 0.005, SE = 0.008, 95%-CI: − 0.010, 0.021).
Table 3: Association between child maltreatment, mental health and Executive functions
|
|
B
|
SE of B
|
β
|
t
|
95% CI
|
Mental Health problems a
|
|
|
|
|
LCI
|
UCI
|
Age
|
-0.017
|
0.061
|
-.032
|
-0.285
|
-0.138
|
0.103
|
Education in years
|
0.112
|
0.080
|
.155
|
1.392
|
-0.047
|
0.270
|
Trauma
|
0.276
|
0.059
|
.292
|
4.715***
|
0.161
|
0.392
|
Maltreatment
|
0.074
|
0.015
|
.297
|
4.828***
|
.044
|
.104
|
Executive functions b
|
|
|
|
|
|
|
Age
|
0.436
|
0.282
|
.179
|
1.550
|
-0.119
|
0.991
|
Education in years
|
-0.366
|
0.371
|
-.114
|
-0.987
|
-1.097
|
.365
|
Trauma
|
0.427
|
0.271
|
.101
|
1.577
|
-0.106
|
0.960
|
Maltreatment
|
-0.541
|
0.071
|
-.487
|
-7.657***
|
-0.680
|
-0.402
|
Mediation Model
|
|
|
|
|
|
|
Executive functions c
|
|
|
|
|
|
|
Age
|
0.439
|
0.282
|
.180
|
1.556
|
-0.117
|
0.995
|
Education in years
|
-0.383
|
0.373
|
-.119
|
-1.027
|
-1.119
|
0.352
|
Trauma
|
0.385
|
0.284
|
.091
|
1.354
|
-0.175
|
0.944
|
Maltreatment d
|
-0.552
|
0.074
|
-.497
|
-7.428***
|
-0.698
|
-0.405
|
Mental health problems
|
0.152
|
0.308
|
.034
|
0.494
|
-0.455
|
0.759
|
Indirect Effect e
|
B
|
SE of B
|
β
|
Z
|
LCI
|
UCI
|
Maltreatment → Mental Health problems
|
0.011
|
0.023
|
.001
|
0.481
|
-0.036
|
0.059
|
Note: B: Unstandardized regression weight, SE: standard error, β: standardized regression weight, t: t-test statistic, CI: Confidence interval at 0.05 level of significance, *** p < .001
a Adj. R2= .257; F (4,225) = 20.759, p < .001
b Adj. R2= .205; F (4,225) = 15.799, p < .001
c Adj. R2= .203; F (5,224) = 12.646, p < .001
d Total effect of maltreatment on executive functions: B= -0.541, SE of B:0.071, β =-0.498, t = -7.657***, LCI: -0.680, UCI: -0.402.
e Sobel test result for indirect effect of maltreatment on executive functions via the mediator mental health problems
Table 4. Association between child maltreatment, working memory and mental health problems
|
|
|
B
|
SE of B
|
β
|
t
|
95% CI
|
|
Mental Health problems a
|
|
|
|
|
LCI
|
UCI
|
Age
|
-0.017
|
0.061
|
-.032
|
-0.285
|
-0.138
|
0.103
|
Education in years
|
0.112
|
0.080
|
.155
|
1.392
|
-0.047
|
0.270
|
Trauma
|
0.276
|
0.059
|
.292
|
4.715***
|
0.161
|
0.392
|
Maltreatment
|
0.074
|
0.015
|
.297
|
4.828***
|
0.044
|
0.104
|
Working memory b
|
|
|
|
|
|
|
Age
|
-0.003
|
0.095
|
-.004
|
-0.031
|
-0.190
|
0.184
|
Education in years
|
0.111
|
0.125
|
.111
|
0.884
|
-0.136
|
0.358
|
Trauma
|
0.019
|
0.091
|
.014
|
0.206
|
-.0161
|
0.199
|
Maltreatment
|
-0.083
|
0.024
|
-.242
|
-3.482***
|
-0.130
|
-0.036
|
Mediation Model
|
|
|
|
|
|
|
Working memory c
|
|
|
|
|
|
|
Age
|
-0.002
|
0.095
|
-.002
|
-0.020
|
-0.189
|
0.186
|
Education in years
|
0.104
|
0.126
|
.104
|
0.822
|
-0.145
|
0.352
|
Trauma
|
0.001
|
0.096
|
.001
|
0.012
|
-0.188
|
0.190
|
Maltreatment d
|
-0.088
|
0.025
|
-.256
|
-3.498***
|
-0.137
|
-0.038
|
Mental health problems
|
0.064
|
0.104
|
.046
|
0.614
|
-0.141
|
0.269
|
Indirect Effect e
|
B
|
SE of B
|
β
|
Z
|
LCI
|
UCI
|
Maltreatment → Mental Health problems
|
0.005
|
0.008
|
.012
|
0.597
|
-0.010
|
0.021
|
Note: B: Unstandardized regression weight, SE: standard error, β: standardized regression weight, t: t statistic, CI: Confidence interval at 0.05 level of significance, *** p < .001
a Adj. R2= .257; F (4,225) = 20.759, p < .001
b Adj. R2= .047; F (4,225) = 3.833, p = .005
c Adj. R2= .044; F (5,224) = 3.124, p = .01
d Total effect of maltreatment on working memory: B= -0.830, SE of B:0.024, β =-0.268, t = -3.482***, LCI: -0.130, UCI: -0.036
e Sobel test result for indirect effect of maltreatment on working memory via the mediator mental health problems