riggs
H. K. Kumpulainen S1369431
Master Thesis Clinical Psychology Supervisor: Dr. D. S. Sieh
Institute of Psychology Universiteit Leiden 18-08-2017
Depression in Adolescents of Depressed or Nondepressed,
Somatically Ill Parents: The Role of Marital Relationship
Abstract
The aim of this study was to investigate depression in adolescents of depressed or nondepressed, somatically ill parents and to examine the role of marital relationship and parent attachment in the relationship between parental depression and adolescent depressive behaviour. The sample consisted of 51 adolescents (age M = 14.49) and 62 parents (age M = 46.25) in the target group, and 54 adolescents (age M = 14.37) and 72 parents (age M = 47.23) in the control group. Parental somatic illnesses included for example: multiple
sclerosis and rheumatoid arthritis. Parental depression in the target group was mild in 83.9%, moderate in 12.9%, and severe in 3.2% of the cases, while no depression was found in the control group. The data were collected with self-report questionnaires. Descriptive statistics, Pearson correlations, and PROCESS technique by Hayes (2012) were used. No significant relationship between parental depression and adolescent depressive behaviour was found, therefore no mediation could be found either. However, depressed parents had a lower quality of marital relationship than nondepressed parents (p < .001) and families with parental depression had lower quality of parent attachment (p < .001), which was linked to more depressive behaviour in adolescents regardless of parental depression (p < .001 in the target group and p = .019 in the control group). Based on these results we suggest that professionals provide adequate help for families with somatically and mentally ill parents in forming intrafamily relations.
Keywords: parental depression, adolescent depression, marital relationship, parent attachment, somatic illness, chronic medical condition, parents, adolescents
Depression in Adolescents of Depressed or Nondepressed, Somatically Ill Parents: The Role of Marital Relationship and Parent Attachment
Depression is a familiar disorder with heritability of about 32 to 42% (Sullivan, Neale, & Kendler, 2000). Furthermore, maternal postnatal depression with later episodes has been associated with an increased risk of depression in adolescent offspring (Halligan, Murray, Martins, & Cooper, 2007). The risk of depression is elevated in those with a chronic medical condition (CMC), a form of somatic illness;for example, in people suffering from multiple sclerosis (MS) the 12-month prevalence of depression was 15.7% compared to 7.4% in the general population (Patten, Beck, Williams, Barbui, & Metz, 2003). Another study looked at the prevalence of depressive disorders in people undergoing dialysis for their chronic kidney disease and found that 78.5% of those patients had some form of depressive disorder (Kokoszka et al., 2016). Not only is parental depression assumed to be linked to offspring depression through genetic factors (Sullivan et al., 2000), but also through mediators such as: marital relationship (Amato & Keith, 1991) and parent attachment (Murray et al., 2011). However, not all adolescents of depressed parents end up with depression; even though they are affected by it.
The quality of the marital relationship of the parents affects their children greatly. High levels of conflict between the parents have been associated with poor psychological adjustment (depression, anxiety, and self-esteem) in adolescents (Noller, Feeney, Sheehan, Darlington, & Rogers, 2008). Furthermore, a link was found between perceived quality of parent attachment and depression in children (Chesmore, Weiler, Trump, Landers, & Taussig, 2016). Child functioning is also affected by parental CMC (Armistead, Klein, & Forehand, 1995). In their literature search Armistead et al. (1995) also found that this relationship is mediated by parental depression, relationship conflict, parental divorce, and disrupted
parenting. These findings show the importance of mediators between parental depression and depression in children: they can serve as protective factors in children’s psychological well-being. Thus, this thesis will further investigate marital relationship and parent attachment as mediators for depressive behaviour in adolescents of somatically ill parents, comparing those with parental depression to those with no parental depression.
The Effect of Marital Relationship on Adolescent Functioning
Marital relationship is assumed to affect the psychological well-being of the offspring. A meta-analysis, based on over 13,000 children, found that children of divorced parents show
lower levels of well-being than children living in intact families (Amato & Keith, 1991). Furthermore, a twin study found that environmental influences of divorce are associated with higher rates of psychopathology in children (D’Onofrio et al., 2005). Not only is the marital relationship of parents associated with depression in children, but also with depression in the parents themselves, as discordant spouses show more depression (Whisman, Robustelli, Beach, Snyder, & Harper, 2015). Another study looked at 231 couples with husbands with cardiovascular disease and found that lower marital satisfaction was associated with depression and lower health satisfaction in husbands (Novak, Sandberg, & Harper, 2014).
Emotional security theory (EST) states that conflict between the parents increases children’s vulnerability to adjustment problems by strengthening their insecurity in
interparental relationship (Cummings & Davies, 1996; Davies & Cummings, 1994). A more recent finding by Suh et al. (2016) found support for the EST, showing that the theory still applies to this day; this is important since the data for our research was collected in 2008-2010. Furthermore, Davies, Harold, Goeke-Morey, and Cummings (2002) confirmed EST in four studies. In their first study they found that in response to destructive conflict between the parents the children express fear, avoidance and involvement. The second study found that emotional insecurity in children mediated the association between parental conflict and child maladjustment. The findings of the third study were the same as the findings of the second study, however, the researchers controlled for other parenting processes. The fourth study showed that the mediational pathways between parental conflict, child insecurity, and maladjustment were strengthened by family instability, parenting difficulties, and attachment insecurity.
Parent Attachment
Attachment theory states that attachment to the caregiver develops from early on. Infants need constant nurturing from the caregiver in order to develop into healthy adults (Bowlby, 1969, 1973, 1980). Attachment relationship is closely related to the quality of perceived parent attachment, which will be further studied in this thesis. Poor parent attachment is closely related to development of depressive symptoms (Moran, Bailey, & DeOliveira, 2008; Reitman & Asseff, 2010). Morever, Bradford, Burningham, Sandberg, and Johnson (2016), studying married couples, found that insecure attachment mediated the link between poor parent-child relationship and depressive symptoms. Additionally, Jiang, You, Zheng, and Lin (2017) found that poor attachment relationships with parents increased the
likelihood of adolescents engaging in nonsuicidal self-injury, which is often related to
depression and other mental health problems (Nock, Joiner Jr., Gordon, Lloyd-Richardson, & Prinstein, 2006). Parent attachment is not only related to the psychological well-being of the children, but it is affected by the psychological well-being of the parents; research has found that the development of a good attachment relationship is negatively influenced by parental depression because depression in parents diminishes their ability of being emotionally available and supportive to the children (Leinonen, Solantaus, & Punamäki, 2003).
Due to the fact that depression is more likely to come about with a life altering stressor, such as CMC (Patten et al., 2003), and that parental depression is closely linked to offspring depression (Sullivan et al., 2000), it can be assumed that parental CMC is affecting adolescents in terms of depressive symptoms. This was indeed found in a study by Evans, Keenan, and Shipton (2007) with 55 children with maternal chronic pain and 48 children in the control group; children of mothers with chronic pain showed more internalizing problems (such as depression) and more insecure attachment. Together this suggests that parental CMC is also linked to lower quality parent attachment.
Link between Marital Relationship and Parent Attachment
It is assumed that lower marital relationship is linked to poorer parent attachment in children. This is called the spillover hypothesis: negative interaction between the spouses spills over to predict negative interaction with the child (Cox, Paley, & Harter, 2001). A meta-analysis found that effects of a divorce have a significant negative effect on parent-child relationships (Amato & Keith, 1991). This would indicate that poor marital relationship is related to poorer parent-child relationship. Marital intimacy was associated with a better father-child relationship in depressed fathers (Engle, 2011). However, the results are
contradictory: Engle (2011) found that higher levels of marital intimacy were correlated with poorer mother-child relationship in depressed mothers. Taraban et al. (2017) obtained similar results concerning depressed mothers and they suggest that depressed mothers in high quality marital relationships are more willing and able to give out the parenting responsibility to their spouses, making the mothers less skilful and experienced in parenting. It is also important to note that depression might moderate the relationship between the quality of marital
relationship and the quality of parent attachment, as mothers’ and fathers’ depressive symptoms moderated the relationship between marital quality and the other parent’s relationship quality with their child (Kouros, Papp, Goeke-Morey, & Cummings, 2014).
However, some findings support the contrary: spillover effects were found not to be affected by parents’ neuroticism, but are more about daily fluctuations in negative mood (Sears, Repetti, Reynolds, Robles, & Krull, 2016). Moreover, Cassé, Oosterman, and Schuengel (2016) found no direct effect of partner dissatisfaction on attachment quality. Nonetheless, they found parenting self-efficacy, the conviction that one is able to parent successfully, to be a confounder: for high parenting self-efficacy, partner dissatisfaction increased the risk for avoidant attachment in children.
Sitnick, Masyn, Ontai, and Conger (2016) found that spouses can serve as barriers between depression and parenting, where parenting was a measure of mothers’ child rearing practices as perceived by the child. They also found that mothers’ poor physical health was associated with more disrupted parenting. This is in line with the longitudinal research of Conger and Conger (2002), who found that marital conflict mediates the relationship between parental depression and disrupted parenting. This would suggest that marital relationship and parent attachment indeed are linked together, but as in research of Sitnick et al. (2016) just having a spouse could already be helpful against the effects of depression on parenting. Peltz, Rogge, Rogosch, Cicchetti, and Toth (2015) investigated this association in the opposite direction. They wanted to find out whether child-parent psychotherapy (CPP) for depressed mothers would also improve marital relationship and they did indeed find improvement in marital satisfaction as a result of CPP, demonstrating a link between marital relationship and parent attachment.
Aims, Research Questions and Hypotheses
This research will look into marital relationship and parent attachment as mediators for depressive behaviour in adolescents of depressed or nondepressed parents, when at least one of the parents has CMC. It will provide more information on the likelihood of developing adolescent depression with parental depression (biological factor) along with CMC. CMC is a stressor that might trigger depression in the parent, as well as in the adolescent; therefore, the probability of adolescent depression is likely to be more prevalent in a sample with CMC than in a sample without. This can be useful information for healthcare professionals when looking for risk factors for adolescent depression. In addition, the spillover hypothesis is tested, and since the findings so far are conflicting, this research will provide more information on the topic.
This thesis has two aims, namely to examine whether: (1) families with parental depression have lower quality marital relationships and parent attachment and more
depressive behaviour in adolescents than families with no parental depression, and whether (2) high quality of marital relationship and parent attachment are linked to less depressive behaviour in adolescents when the parent has CMC and depression. The research questions for this study are: “Do the families with a depressed parent have lower quality marital relationships, parent attachment, and more depressive behaviour in adolescents than those with two nondepressed parents?” and “When a parent has depression and CMC: is higher quality of marital relationship and parent attachment correlated with less depressive
behaviour in adolescents?”. Hypotheses are: (H1) parental depression is positively correlated with depressive behaviour in adolescents, (H2) depressed parents have a lower quality of marital relationship than nondepressed parents, (H3) a higher quality of marital relationship is correlated with less depressive behaviour in adolescents, (H4) adolescents of depressed parents have a lower quality of parent attachment than adolescents with nondepressed parents, (H5) a higher quality of parent attachment is correlated with less depressive behaviour in adolescents, and (H6) the two mediators, marital relationship and parent attachment, are positively correlated.
Method Participants
The participants were 67 families with 134 parents and 105 adolescents. One of the parents in each family had CMC. Single parent families (16 parents with 21 adolescents) were also recruited as part of the general population, but they were deleted before the analysis. The CMC found in the sample were multiple sclerosis, rheumatoid arthritis, neuromuscular disease, traumatic brain injury, spinal cord injury, inflammatory bowel
disease, Parkinson disease, and diabetes type I with physical complications. The families with a parent with CMC and depression served as the target group (31 families), and the control group consisted of families with a parent with CMC but no depression (36 families).
Exclusion criteria for adolescents were insufficient level of Dutch, residency outside the Netherlands, severe physical illnesses, and cognitive disabilities. Six participants met the exclusion criteria and six participants decided not to participate. In the control group the families where the non-CMC partner showed depression were removed (18 parents with 15
adolescents in total). Four families were deleted because the data from their partners or adolescents were missing (5 parents and 4 adolescents); two other families were deleted because both of the parents had CMC (4 parents and their 4 adolescents). One family (2 adults, 1 adolescent) was deleted due to the inability of characterising them as part of the target group or the control group because of missing data on parental depression. Instruments
Demographic variables. Demographic questions for the parents included age (measured from their birthday), gender (male or female), education level (highest level of education achieved), and income (an estimate of monthly net income). Also, their
relationship status (including living situation), illness type (diagnosis) and duration (when it started), and nationality were recorded. Demographic questions for the adolescents consisted of: age, gender, nationality, and educational level.
Parental depression. Depression in the parents was measured with the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), which consisted of 21 items on depressive symptoms. The items were answered on a scale from 0 to 3, for example: 0 = I don’t feel disappointed in myself, 1 = I am disappointed in myself, 2 = I am disgusted with myself, and 3 = I hate myself. Scores of 0 to 9 indicated minimal
depression, 10 to 18 indicated mild depression, 19 to 29 indicated moderate depression, and 30 to 63 indicated severe depression. For the analysis, the mean of the sum scores of both partners were used. The reliability of this questionnaire was α = .75 for the target group and α = .47 for the control group. Due to the low reliability in the control group special attention should be given to the interpretation of the regression analysis.
Adolescent depressive behavior. Depressive behavior in adolescents was measured with the Youth Self-Report (YSR; Achenbach, 1991). The questions were answered on a scale from 0 (not true) to 2 (very/often true). A new scale was composed of 8 items
measuring depressive behavior taken from the depressive/anxious subscale of the YSR. The items were I feel lonely, I cry a lot, I deliberately try to hurt or kill myself, I feel that no one loves me, I feel worthless or inferior, I feel too guilty, I think about killing myself, and I am unhappy, sad, or depressed. The scores ranged from 0 to 16 points. The reliability of the scale was α = .82 for the target group and α = .67 for the control group. For both groups the reliability would have been slightly higher if the item about suicidality was deleted, but the change was minor and the item’s content matched the scale well, so we made the decision not to delete it.
Marital relationship. The quality of marital relationship was measured with the Interactional Problem Solving Inventory (IPSI; Lange, 1983). It had 17 items which were answered on a five-point scale ranging from 1 = exactly applicable to me/my partner to 5 = absolutely not applicable to me/my partner. Higher scores indicated a well-functioning marriage and a higher problem solving ability, meaning that the respondents are able to solve their problems through communication and pragmatic solutions. This questionnaire was only available in Dutch, so the following example item was freely translated: in our relationship there are a lot of problems that are not solvable. Both of the parents answered this
questionnaire. The sum scores of the answers were computed and the total score was
calculated as the mean of the two sums. The possible scores form IPSI ranged from 17 to 85 points. The reliability of the scale for the target group was α = .94 and for the control group α = .91.
Parent attachment. Parent attachment was measured from the viewpoint of the adolescent and assessed with the Inventory of Parent and Peer Attachment (IPPA; Armsden & Greenberg, 1987). IPPA measured the quality of the relationship with the mother (12 items; α = .89) and with the father (12 items; α = .86). The items were answered on a scale from 1 (almost never or never true) to 4 (almost always or always true). IPPA included questions such as: I trust my mother and my father respects my feelings. The items were the same for maternal and paternal attachment, only the word mother was replaced by father. It had three scales: communication with the parent, alienation from the parent, and confidence in the parent. Higher scores indicated better communication, more confidence, and more alienation. The total score was calculated as a mean of the two scales measuring the quality of the relationship with the mother and the father. The maximum score was 48 points and the minimum was 12 points. The reliability of the scale was α = .92 for the target group and α = .88 for the control group.
Procedure
Participants were recruited from the Netherlands in rehabilitation centres, hospitals, schools, community centres, intercultural institutions, general health practitioners’ offices, and public libraries by poster and brochures. Professionals in the participating institutes provided additional information to those who were interested. Commercials were also posted online on patient organizations. Potential participants could reach the research manager by post or call in order to receive further instructions and to request an informed consent form.
After the written informed consent was received from the parents and the adolescents, trained research assistants made appointments with the families to administer the questionnaires at their homes. The research assistants had been trained by the research manager to follow a research protocol designed by the research manager. As compensation for participation the participants received a gift voucher, a movie ticket, or a cell phone cover. After the
participation, the families were informed about the status of the research project in four occasions through a newsletter. The questionnaires were administered between September 2008 and April 2010. The study was approved by the ethical commission of the research institute of Child Development and Education of the University of Amsterdam (Sieh, Visser-Meily, & Meijer, 2012a).
Statistical Analysis
Data were collected between 2008 and 2011 as part of the dissertational research of Sieh (2012). Gender was dummy-coded as 1 = male and 2 = female. Age and duration of illness were measured in years. Descriptive statistics and t-tests were used to present the means and the distributions of the scores and to describe the group characteristics and differences between the target group and the control group. Age and gender of the
adolescents, and illness duration were tested as covariates. Little’s Missing Completely at Random test was performed; it was found to be significant for parents with CMC (p = .009), so missing values were not replaced. Less than 5% of the cases had missing data. Outliers were identified with face validity by creating box plots for the relationships between marital relationship, parent attachment, parental depression, and adolescent depressive behaviour as shown in Figure 1. As a result, 5 families were deleted from the analyses (10 adults and their 11 children).
The main analyses were conducted according to Figure 1. Parental depression acted as an independent variable and the adolescent depressive behaviour as a dependent variable. Marital relationship and parent attachment were mediators. The independent variable was multicategorical and all of the other variables were measured at interval level. Assumptions were tested before running the analyses. Normality was assumed due to the central limit theorem, which suggests that in samples with n larger than 30, the distribution tends to be normal (Field, 2013). Multicollinearity was checked with variance inflation factor and the result was under 10 for each predictor and thus, it was not violated (Field, 2013).
violated while linearity was not. The assumption of independence is also violated in this sample due to family connections. The correlations between all of the variables were tested separately with Pearson correlation, with the exception of correlations with gender, which were tested with Spearman correlation.
To test the mediation model, bootstrapping PROCESS technique by Hayes (2012) was used on behalf of a syntax available on www.processmacro.org. This bootstrapping method provided increased power and control over type I error (Briggs, 2006; Hayes & Preacher, 2014; Williams & MacKinnon, 2008). First, the total effect model of parental depression on adolescent depressive behaviour was tested by running a linear analysis in order to check for the correlation between the variables (H1). After that, marital relationship and parent attachment were entered into the model as mediators in the relationship between adolescent depressive behaviour and parental depression (H2-H6). In the end, the covariates of age, gender, and illness duration were added. For all models, a 95% confidence interval (CI) was used. 5000 bootstrapping samples were created (Hayes & Preacher, 2014). Analyses were conducted on IBM SPSS version 24 (IBM corp, 2016). Cohen’s d effect sizes of all the outcome scores were calculated to indicate the standardized difference between two means. Effect sizes of d = 0.20 were considered small, d = 0.50 medium, and d = 0.80 large (Cohen, 1992). This study had the power to detect medium effects (Sieh et al., 2012a).
Figure 1. The hypothesised mediation model.
Results
Descriptive analyses revealed that there were more mothers with CMC in both the target group (71%) and the control group (69.1%) than fathers. See Table 1 for all descriptive statistics. The CMCs found in the group were multiple sclerosis (28.4%), rheumatoid arthritis (20.9%), neuromuscular disease (17.9%), traumatic brain injury (7.5%), spinal cord injury (6.0%), inflammatory bowel disease (6.0%), Parkinson disease (7.5%), and diabetes type I with physical complications (1.5%). Parental depression in the target group was mild in
83.9%, moderate in 12.9%, and severe in 3.2% of the cases. The parents were middle-aged and most had completed university of applied sciences or a lower education. Most families’ net income in the target group was 2500 to 2999 euros per month (26.7%), while most families in the control group had a net income of 4000 euros and more per month (27.8%; t (64) = -1.45). Illness duration in the target group was nearly half of the illness duration in the control group. Depression was only found in the target group (Cohen’s d = 2.55, 95% CI [7.65, 11.15]). The quality of marital relationships were lower in the target group than in the control group (Cohen’s d = 1.22, 95% CI [-18.15, -7.66]).
The adolescents were on average 14.43 years old and most of them were in high school, with most of the target group being in the lowest level of high school (25.5%) and most of the control group being in the highest level of high school (27.8%). All of them had a Dutch nationality. Only slightly more depressive behaviour was found in the target group (Cohen’s d = 0.36, 95% CI [-1.55, 0.12]). A lower quality of parent attachment was found in the target group (Cohen’s d = 0.76, 95% CI [1.99, 6.08]).
Parental depression was found to negatively correlate with marital relationship in both target group and control group, and with parent attachment in the target group. Adolescent depressive behaviour correlated negatively with parent attachment in both the target group and the control group, and positively with gender in the target group. More depressive behaviour was found in girls than in boys in target group (M = 2.81, SD = 3.06; M = 0.96, SD = 1.37 respectively) t (49) = -2.76, p = .008. Parent attachment correlated negatively with age in the target group. See Table 2 for all Pearson and Spearman correlations.
PROCESS technique by Hayes (2012) was used to test the mediation model, see Figure 1. In order to test the hypotheses, the analysis was run separately for the target group and control group. Three models were created for both groups with marital relationship (Model 1), parent attachment (Model 2), and adolescent depressive behaviour (Model 3) as outcome variables, see Table 3 for all of the variables. All of the models were found to be significant for the target group with F (4, 22) = 3.40, p = .026 for Model 1, F (5, 21) = 5.75, p = .002 for Model 2, and F (6, 20) = 3.51, p = .016 for Model 3. See Table 3 for all beta and t-scores. The models were also tested exploratively for the control group and only Model 1 was significant (F (4, 28) = 3.81, p = .01) with the only significant predictor being parental depression (b = -2.23, t (28) = -3.68, p = .001).
To answer the hypotheses, multiple tests were conducted. Hypothesis 1; parental depression is positively correlated with depressive behaviour in adolescents, was rejected (r = -.01, n = 31, p = .974 in target group and r = -.09, n = 36, p = .592 in control group).
Hypothesis 2; depressed parents have a lower quality of marital relationship than nondepressed parents, was confirmed (t (61) = -4.92, p < .001). Hypothesis 3; a higher quality of marital relationship is correlated with less depressive behaviour in adolescents, was rejected (r = .01, n = 28, p = .968 on target group and r = -.18, n = 35, p = .305 in control group). Hypothesis 4; adolescents of depressed parents have a lower quality of parent attachment than adolescents with nondepressed parents, was confirmed (t (102) = 3.19, p < .001). Hypothesis 5; a higher quality of parent attachment is correlated with less depressive behaviour in adolescents, was confirmed (r = .53, n = 50, p < .001 in target group and r = -.32, n = 53, p = .019 in control group). Hypothesis 6; the two mediators, marital relationship and parent attachment, are positively correlated, was rejected (r = .13, n = 28, p = .515 in target group and r =.27, n = 35, p = .112 in control group).
Lastly, explorative analyses were conducted. Two new scales were computed from the YSR (Achenbach, 1991), namely internalizing problem scale (INT) and anxious symptom scale (ANX). INT had a reliability of α = .91, with a minima of 0 and maxima of 62 points. ANX had a reliability of α = .81, with a minima of 0 and maxima of 16 points. Pearson correlation tests were conducted in order to explore the correlations between parental
depression and INT and ANX but no significant correlations were found (r = .20, n = 30, p = .296; r = .14, n = 31, p = .455, in target group and r = -.05, n = 36, p = .795; r = -.09, n = 36, p = .591, in control group, respectively). Finally, linear regression was used to test for an interaction effect between marital relationship and parent attachment. However, no interaction effect was found (b = 1.61, t (59) = 1.20, p = .236).
Table 1.
Descriptive Statistics for Target Group and Comparison Group.
Note. Parent variables are measured per couple, not per person.
a Parental and adolescent educational level is measured from 1 (primary school) to 12 (university). b Income is only measured in the healthy partner as an estimate of net
family income from 1 (less than 1000 euros) to 8 (more than 4000 euros), 5 = 2500 to 2999 euros. *p < .05. **p < .01. ***p < .001.
Target Group n Control Group n t(df) Cohen’s d
Adolescents (n) 51 54
Female 26 (51.0%) 51 29 (53.7%) 54 0.28 (103) 0.06
Mean Age (SD) 14.49 (2.27) 51 14.37 (2.41) 54 -0.26 (103) 0.05 Mean Education Level (SD)a 6.47 (3.36) 51 6.80 (3.42) 54 0.49 (103) 0.09
Mean Depressive Behaviour (SD) 1.90 (2.54) 51 1.19 (1.17) 53 -1.70 (102) 0.36 Mean Parent Attachment (SD) 37.24 (5.96) 50 41.28 (4.51) 54 3.19 (102) 0.76***
Parents (couples, n) 31 36
Female, CMC 22 (71%) 31 25 (69.4%) 36 0.13 (65) 0.04
Mean Age, CMC (SD) 46.25 (5.56) 31 47.23 (5.09) 36 -0.75 (65) 0.18 Mean Educational Level, CMC
(SD)a
4.32 (1.49) 31 4.33 (1.41) 36 -0.03 (65) 0.01 Mean Income (SD)b 5.00 (1.60) 30 5.64 (1.92) 36 -1.45 (64) 0.36
Mean Illness Duration (SD) 8.51 (7.67) 29 15.14 (12.13) 34 -2.54 (61) 0.65* Mean Depression (SD) 14.89 (4.80) 31 5.49 (2.03) 36 10.71 (65) 2.55*** Mean Marital Relationship (SD) 57.98 (12.55) 28 70.89 (8.19) 35 -4.92 (61) 1.22***
Table 2.
Pearson Correlations for Target Group and Control Group.
1. n 2. n 3. n 4. n 5. n 6. n
1. Parental Depression -
2. Adol. Depr. Behv. -.01/-.09 31/36 -
3. Marital Relationship -.55**/-.58*** 28/35 .01/-.18 28/35 -
4. Parent Attachment -.42*/-.01 30/36 -.53***/-.32* 50/53 .13/.27 28/35 -
5. Age -.29/-.06 31/36 .22/.17 51/53 .18/-.05 28/35 -.41*/-.11 50/54 -
6. Gendera .05/-.13 31/36 .40**/.24 51/53 .08/.15 28/35 -.08/.10 50/54 .04/.16 51/54 -
7. Illness Duration .18/.04 29/34 -.28/-.02 29/34 -.09/-.09 27/33 .06/.26 28/34 -.26/-.14 29/34 -.30/.17 29/34
Note. The results are displayed target group first and control group second. Adol. Depr. Behv. stands for adolescent depressive behaviour. Significant correlations are in boldface.
a Correlations with gender were calculated with Spearman correlation.
*p < .05. **p < .01. ***p < .001.
Table 3.
Mediation Models for Target Group.
Note. Adol. Depr. Behv. stands for adolescent depressive behaviour. n = 27. *p < .05. **p < .01. ***p < .001.
Model 1 (Marital Relationship) Model 2 (Parent Attachment) Model 3 (Adol. Depr. Behv.)
b t(22) p b t(21) p b t(20) p Constant 62.17 3.33 ** 73.18 8.41 *** 15.96 1.70 Parental Depression -1.46 -3.32 ** -0.74 -3.63 ** -0.24 -1.74 Marital Relationship -0.04 -0.53 -0.04 -1.00 Parent Attachment -0.30 -2.61 * Age 0.64 0.66 -1.51 -4.06 *** -0.02 -0.09 Gender 4.44 0.99 -0.04 -0.03 2.26 2.51 * Illness Duration 0.16 0.6 0.00 0.02 -0.01 -0.14
Discussion
This study looked at depression in adolescents of depressed or nondepressed, somatically ill parents. Marital relationship and parent attachment were looked at as
mediators in the relationship between parental and adolescent depression. We did not find a significant relationship between parental depression and adolescent depressive behaviour. This relationship was required for the mediation, so no mediation could be found. However, we did find that depressed parents had a lower quality of marital relationship than
nondepressed parents, but this was not correlated with adolescent depressive behaviour. Adolescents of depressed parents were found to have a lower quality of parent attachment, which was linked to depressive behaviour, too. Marital relationship and parent attachment were not found to be connected. Model testing by PROCESS revealed that the overall model was a good fit for the target group, meaning that the predictors of parental depression, marital relationship, and parent attachment predicted adolescent depressive behaviour well. On the other hand, in line with our expectations, the model did not fit the control group as well, probably because it was designed for families with parental depression.
It is surprising that no correlation was found between parental depression and adolescent depressive behaviour considering that Sieh, Visser-Meily, and Meijer (2012b) found borderline significant correlation between parental depression and internalizing problem behaviour in adolescents on the same data set. Due to this finding, we conducted explorative analyses on correlations between internalizing problem behaviour as well as anxious symptoms and parental depression. No significant correlations were found, so the difference in findings was probably due to differently selected samples, specifically, in the study of Sieh et al. (2012b) parental depression was not a criterion for the selection of the target group. It should also be noted that adolescent depressive behaviour was measured with a scale that was composed for this study alone. There is no previous research on the
reliability of a depressive symptoms scale combined from the items of the internalizing problem scale of YSR (Achenbach, 1991), but due to its high reliability we decided to run the analysis with it. This provided us with a better understanding of the depressive symptoms than the standard internalizing problem scale would have.
The mediation analysis revealed that depressed parents indeed had a lower quality of marital relationship than nondepressed parents, but marital relationship was not connected to adolescent depressive behaviour. This could be explained by the research of Henderson,
Sayger, and Horne (2003) where they found that lower level of marital satisfaction was correlated with behavioural problems in children. Our research did not look at externalizing problem behaviour, so this could explain the contradicting findings. However, a lot of previous research concludes that marital relationship is linked to depression in children (D’Onofrio et al., 2005; Amato & Keith, 1991), which suggests that there is a link which just was not found in our sample.
Adolescents of depressed parents experienced a lower quality of parent attachment than those of nondepressed parents, and parent attachment was linked to more depressive behaviour as well. This was according to the hypotheses. It should be noted that lower quality of parent attachment was correlated with adolescent depressive behaviour also in the control group, which makes parent attachment an important risk factor for depression in adolescents. This is in line with the research of Milan, Snow, and Belay (2009) as they found no direct link between parental depression and adolescent depression, but found that those children with insecure (low quality) attachment to their parents were at risk for depression. The two mediators were not found to be connected. This could be explained by the findings of
Summers, Forehand, Armistead, and Tannenbaum (1998), which suggest that marital divorce, the marker of low quality marital relationship, was not connected to parent-child relationship, but to the security of attachment to the romantic partners of the adolescent, meaning that the children of parents with a low quality of marital relationship did not show a lower quality of parent attachment, but showed a lower quality of attachment to their partners in adolescence. An explorative analysis was also conducted revealing no interaction effect between marital relationship and parent attachment.
Three covariates were included in the analyses, namely age, gender, and illness duration. More depression was found in girls than in boys in the target group, which could indicate that girls are more vulnerable to parental depression than boys are. By the age of 15 girls are twice as likely to experience depression as boys are (Cyranowski, Frank, & Young, 2000). For girls the perception of support from their parents is especially important during adolescence and has been associated with lower self-esteem and depression (Bámaca-Colbert, Tilghman-Osborne, Calderón-López, & Moore, 2017). One study found that maternal
depression correlated significantly with externalizing problem behaviour and not with internalizing problem behaviour in boys (Henderson et al., 2003), which could explain why the correlation was less strong for boys in our study. Parent attachment correlated negatively with age in the target group, meaning that older children showed lower quality of parent
attachment, but only when at least one of their parents had depression. Prior research also suggests that self-reported attachment security to parents decreases with pubertal maturity (Papini, Roggman, & Anderson, 1991), explaining why older children report lower levels of parent attachment compared to younger ones.
Illness duration did not correlate significantly with any of the variables tested here, which would suggest that CMC has the same effects on people at its early stage as at a later stage. However, it was the biggest demographical difference between the target group and the control group, since the illness duration in the target group was only half of that in the control group (see Table 1 for all demographical differences). Research has found that child reported illness severity and the stress that it causes are better predictors of depressive symptoms in children than parental illness duration, stage and prognosis (Compas et al., 1994). However, a meta-analysis found that internalizing problem behaviour in children is linked to longer illness duration in their parents (Sieh, Meijer, Oort, Visser-Meily, & van der Leij, 2010), which could explain the low scores of adolescent depressive behaviour in the target group. The illnesses that were looked at in this meta-analysis were different and often more serious than the illnesses present in our study, for example HIV and cancer. Taking together these findings, we argue that illness duration on its own is not so important for adolescent depressive behaviour, but as Compas et al. (1994) suggest, the severity and stress it causes are better predictors of the outcome in the adolescents.
This study was not without limitations. For a cleaner sample, those families, where only the partner of the parent with CMC had depression, were excluded. This provided us with the possibility to compare the effects of depression and no depression, but in real life it is not always the case that the parent with CMC is the one with depression and not their partner. This type of control group also did not allow us to examine the effects of CMC itself, since a control group without CMC would have been needed. Also, the sample consisted of people with higher than average SES and education level, so the results might not be generalizable to the general Dutch population. Some of the participants were recruited from general health practitioners’ offices and possibly the sample included more adolescents who already got help for their problems than the general population does. This could explain why no link between parental depression and adolescent depressive behaviour was found. The data was collected with self-reports only, which could have led to untruthful answers from the participants. Therefore, it would have been better to use a combination of self-reports and interviews. It is also important to note that family members are never completely independent
of each other. This is called the family cluster effect, which was tested by Sieh et al. (2012a) who found that 41% of the variability of the internalizing problem behaviour was explained by it. Therefore, future research should consider using multilevel analysis. Another
assumption, namely homoscedasticity, was also violated making the probability of false positives more likely. Therefore, the results should be viewed with caution. Furthermore, the external validity was limited, due to the fact that only depression and one CMC per person was taken into account when measuring illnesses. Therefore, no data on comorbid illnesses were available. The YSR (Achenbach, 1991) only predicts DSM-IV disorders moderately (Ferdinand, 2008), so no psychiatric conclusions could be drawn from this sample. Similarly, no clinical interviews were conducted on the parents to diagnose depression, so the
depression scores were solely based on self-reports. Additionally, the reliability for BDI (Beck et al., 1961) in the control group was quite low (α = .47). We expected it to be slightly but not this much lower than in the target group due to the fact that the depressed participants were removed before the conduction of the reliability analysis. Lastly, the data is solely correlational, so no causal links can be drawn from it.
Future research should consider adding a second control group of families with no CMC, in order to provide data on the effects of CMC and not only depression on adolescent depressive behaviour. Using another measurement tool to measure adolescent depressive behaviour is advised as well as using clinical interviews to measure both adolescent and parental depression. Since this study had a correlational design, future research is advised to look toward a longitudinal design in order to assess the causal directions. In addition, based on our findings and previous research (e.g. Henderson et al., 2003) it seems like children of depressed, chronically ill parents do not only show depression, but both internalizing and externalizing problem behaviours should be looked at.
In conclusion, based on our results, it is important for professionals taking care of patients with CMC to note that the quality of marital relationship is correlated with CMC. Married couples with CMC should be provided with the possibility of getting help for marital problems. Since parental depression was also found to be correlated with parent attachment, which was further correlated with adolescent depressive behaviour, families with CMC and parental depression should be provided with support on establishing good parent-child relationships. A family-cantered approach is recommended. Poor parental attachment was correlated with adolescent depressive behaviour, not only when the parent was depressed, but also without parental depression. Therefore, in order to prevent depression in adolescence it
is important to support the parent-child relationships, especially in families with mental or physical illness. These parents might not have the same means to provide the social and emotional support for their children as healthy parents could have.
References
Achenbach, T. M. (1991). Manual for the youth self-report and 1991 profile. Burlington: University of Vermont, Department of Psychiatry.
Amato, P. R., & Keith, B. (1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin, 110, 26-46. doi:10.1037/0033-2909.110.1.26 Armistead, L., Klein, K., & Forehand, R. (1995). Parental physical illness and child
functioning. Clinical Psychology Review, 15, 409-422. doi:10.1016/0272-7358(95)00023-I
Armsden, G. C., & Greenberg, M. T. (1987). The Inventory of Parent and Peer Attachment: Individual differences and their relationship to psychological well-being in
adolescence. Journal of Youth and Adolescence, 16, 427-453. doi:10.1007/BF02202939
Bámaca-Colbert, M. Y., Tilghman-Osborne, E. M., Calderón-López, S., & Moore, A. M. (2017). Perceptions of support from mothers, fathers, and friends: Direct and indirect associations with the psychological adjustment of Mexican-origin girls. Journal of Latina/o Psychology, 5, 142-156. doi:10.1037/lat0000070
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression (BDI). Archives of General Psychiatry, 4, 561-571.
doi:10.1001/archpsyc.1961.01710120031004
Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York, NY: Basic Books. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New York,
NY: Basic Books.
Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression. New York, NY: Basic Books.
Bradford, A. B., Burningham, K. L., Sandberg, J. G., & Johnson, L. N. (2016). The association between the parent-child relationship and symptoms of anxiety and depression: The roles of attachment and perceived spouse attachment behaviours. Journal of Marital and Family Therapy, 1, 1-17. doi:10.1111/jmft.12190
Briggs, N. E. (2006). Estimation of the standard error and confidence interval of the indirect effect in multiple mediator models. Dissertation Abstracts International,37, 4755B. Cassé, J. F. H., Oosterman, M., & Schuengel, C. (2016). Parenting self-efficacy moderates
linkage between partner relationship dissatisfaction and avoidant infant–mother attachment: A Dutch study. Journal of Family Psychology, 30, 935-943. doi:10.1037/fam0000247
Chesmore, A. A., Weiler, L. M., Trump, L. J., Landers, A. L., & Taussig, H. N. (2016). Maltreated children in out-of-home care: The relation between attachment quality and internalizing symptoms. Journal of Child and Family Studies, 26, 381-392.
doi:10.1007/s10826-016-0567-6
Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159.
Compas, B. E., Worsham, N. L., Epping-Jordan, J. E., Grant, K. E., Mireault, G., Howell, D. C., & Malcarne, V. L. (1994). When mom or dad has cancer: Markers of
psychological distress in cancer patients, spouses, and children. Health Psychology, 13, 507–515. doi:10.1037/0278-6133.13.6.507
Conger, R. D., & Conger, K. J. (2002). Resilience in Midwestern families: Selected findings from the first decade of a prospective longitudinal study. Journal of Marriage and Family, 64, 361-373. doi:10.1111/j.1741-3737.2002.00361.x
Cox, M. J., Paley, B., & Harter, K. (2001). Interparental conflict and parent–child
relationships. In J. H. Grych & F. D. Fincham (Eds.), Interparental conflict and child development: Theory, research, and application (pp. 249–272). New York, NY: Cambridge University Press.
Cummings, E. M., & Davies, P. T. (1996). Emotional security as a regulatory process in normal development and the development of psychopathology. Development and Psychopathology, 8, 123–139. doi:10.1017/S0954579400007008
Cyranowski, J. M., Frank, E., & Young, E. (2000). Adolescent onset of the gender difference in lifetime rates of major depression: A theoretical model. Archives of General Psychology, 57, 21-27. doi:10.1001/archpsyc.57.1.21
Davies, P. T., & Cummings, E. M. (1994). Marital conflict and child adjustment: An emotional security hypothesis. Psychological Bulletin, 116, 387–411.
Davies, P. T., Harold, G. T., Goeke-Morey, M. C., & Cummings, E. M. (2002). Child
emotional security and interparental conflict. Monographs of the Society for Research in Child Development, 67, 1-26. doi:10.1111/1540-5834.00206
D’Onofrio, B. M., Turkheimer, E., Emery, R. E., Slutske, W. S., Heath, A. C., Madden, P. A., & Martin, N. G. (2005). A genetically informed study of marital instability and its association with offspring psychopathology. Journal of Abnormal Psychology, 114, 570-586. doi:10.1037/0021-843X.114.4.570
Engle, J. M. (2011). Parental depressive symptoms and marital intimacy as predictors of parent-child interaction: A family systems perspective. Dissertation Abstracts International, 73(5-B), 3299.
Evans, S., Keenan, T. R., & Shipton, E. A. (2007). Psychosocial adjustment and physical health of children living with maternal chronic pain. Journal of Paediatrics and Child Health, 43, 262–270. doi:10.1111/j.1440-1754.2007.01057.x
Ferdinand, R. F. (2008). Validity of the CBCL/YSR DSM-IV scales anxiety problems and affective problems. Journal of Anxiety Disorders, 22, 126–134.
doi:10.1016/j.janxdis.2007.01.008
Field, A. P. (2013). Discovering statistics using IBM SPSS statistics:And sex and drugs and rock 'n' roll (4th ed.). London: Sage publications.
Halligan, S. L., Murray, L., Martins, C., & Cooper, P. J. (2007). Maternal depression and psychiatric outcomes in adolescent offspring: A 13-year longitudinal study. Journal of Affective Disorders, 97, 145-154. doi:10.1016/j.jad.2006.06.010
Hayes, A. F. (2012). PROCESS: a versatile computational tool for observed variable mediation, moderation, and conditional process modeling. Retrieved on from http://www.afhayes.com
Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical and Statistical Psychology, 67, 451-470. doi:10.1111/bmsp.12028
Henderson, A. D., Sayger, T. V., & Horne, A. M. (2003). Mothers and sons: A look at the relationship between child behavior problems, marital satisfaction, maternal depression, and family cohesion. The Family Journal: Counseling and Therapy for Couples and Families, 11, 33-41.doi:10.1177/1066480702238469
IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.
Jiang, Y., You, J., Zheng, X., & Lin, M.-P. (2017). The gualities of attachment with significant others and compassion protect adolescents from non suicidal self-injury. School Psychology Quarterly, 32, 143-155. doi:10.1037/spq0000187
Kokoszka, A., Leszczyńska, K., Radzio, R., Daniewska, D., Łukasiewicz, A., Orzechowski, W. M., … Gellert, M. (2016). Prevalence of depressive and anxiety disorders in dialysis patients with chronic kidney disease. Archives of Psychiatry and Psychotherapy, 1, 8-13. doi:10.12740/APP/61977
Kouros, C. D., Papp, L. M., Goeke-Morey, M C., & Cummings, E. M. (2014). Spillover between marital quality and parent–child relationship quality: Parental depressive symptoms as moderators. Journal of Family Psychology, 28, 315-325.
doi:10.1037/t05644-000
Lange, A. (1983). Interactionele Probleem Oplossings Vragenlijst (IPOV) [Interactional Problem Solving Inventory, IPSI]. Deventer, Netherlands: Van Loghum Slaterus. Leinonen, J. A., Solantaus, T. S., & Punamäki, R. (2003). Parental mental health and
children’s adjustment: The quality of marital interaction and parenting as mediating factors. Journal of Child Psychology and Psychiatry, 44, 227–241. doi:10.1111/1469-7610.t01-1-00116
Milan, S., Snow, S., & Belay, S. (2009). Depressive symptoms in mothers and children: Preschool attachment as a moderator of risk. Developmental Psychology, 45, 1019-1033. doi:10.1037/a0016164
Moran, G., Bailey, H. N., & DeOliveira, C. A. (2008). The roots of depression in early attachment experiences. In K. S. Dobson & D. J. A. Dozois (Eds.), Risk factors in depression (pp. 289–316). United States: Elsevier Ltd.
Murray, L., Arteche, A., Fearon, P., Halligan, S., Goodyer, I., & Cooper, P. (2011). Maternal postnatal depression and the development of depression in offspring up to 16 years of age. Journal of the American Academy of Child and Adolescent Psychiatry, 50, 460-470. doi:10.1037/t01756-000
Nock, M. K., Joiner Jr., T. E., Gordon, K. H., Lloyd-Richardson, E., & Prinstein, M. J. (2006). Non-suicidal self-injury among adolescents: Diagnostic correlates and relation to suicide attempts. Psychiatry Research, 144, 65-72.
doi:10.1016/j.psychres.2006.05.010
Noller, P., Feeney, J. A., Sheehan, G., Darlington, Y., & Rogers, C. (2008). Conflict in divorcing and continuously married families: A study of marital, parent–child and sibling relationships. Journal of Divorce and Remarriage, 49, 1-24.
doi:10.1037/t06793-000
Novak, J. R., Sandberg, J. G., & Harper, J. M. (2014). Older couples with and without cardiovascular disease: Testing associations between and among affective
communication, marital satisfaction, physical and mental health. Families, Systems, and Health, 32, 186-197. doi:10.1037/fsh0000015
Papini, D. R., Roggman, L. A., & Anderson, J. (1991). Early adolescent perceptions of attachment to mother and father. A test of the emotional distancing and buffering hypotheses. Journal of Early Adolescence, 11, 258 – 275.
doi:10.1177/0272431691112006
Patten, S. B., Beck, C. A., Williams, J. V., Barbui, C., & Metz, L. M. (2003). Major depression in multiple sclerosis: A population-based perspective. Neurology, 61, 1524–1527. doi:10.1212/01.WNL.0000095964.34294.B4
Peltz, J. S., Rogge, R. D., Rogosch, F. A., Cicchetti, D., & Toth, S. L. (2015). The benefits of child-parent psychotherapy to marital satisfaction. Families, Systems, and Health, 33, 372-382. doi:1091-7527/15/$12.00
Reitman, D., & Asseff, J. (2010). Parenting practices and their relation to anxiety in young adulthood. Journal of Anxiety Disorders, 24, 565–572.
doi:10.1016/j.janxdis.2010.03.016.
Sears, M. S., Repetti, R. L., Reynolds, B. M., Robles, T. F., & Krull, J. L. (2016). Spillover in the home: The effects of family conflict on parents’ behaviour. Journal of Marriage and Family, 78, 127-141. doi:10.1037/t41354-000
Sieh, D. S. (2012). The impact of parents’ chronic medical condition on children (Unpublished doctoral dissertation). University of Amsterdam, Amsterdam, the Netherlands.
Sieh, D. S., Meijer, A. M., Oort, F. J., Visser-Meily, J. M. A., & Van der Leij, D. A. V. (2010). Problem behavior in children with a chronically ill parent: A meta-analysis. Clinical Child and Family Psychology Review, 3, 384–397. doi:10.1007/s10567-010-0074-z
Sieh, D. S., Visser-Meily, J. M. A., & Meijer, A. M. (2012a). Differential outcomes of adolescents with chronically ill and healthy parents. Journal of Child and Family Studies, 22, 209-218. doi:10.1007/s10826-012-9570-8
Sieh, D. S., Visser-Meily, J. M. A., & Meijer, A. M. (2012b). Parental depression and adolescent functioning in families with a chronically ill parent and single parent families compared to intact families. Manuscript submitted for publication.
Sitnick, S. L., Masyn, K., Ontai, L. L., & Conger, K. J. (2016). Mothers’ physical illness in one- and two-parent families. Journal of Family Issues, 37, 902-920.
doi:10.1177/0192513X14536563
Suh, G. W., Fabricus, W. V., Stevenson, M. M., Parke, R. D., Cookston, J. T., Braver, S. L., & Saenz, D. S. (2016). Effects of the interparental relationship on adolescents’ emotional security and adjustment: The important role of fathers. Developmental Psychology, 52, 1666-1678. doi:10.1037/dev0000204
Sullivan, P. F., Neale, M. C., & Kendler, K. S. (2000). Genetic epidemiology of major depression: Review and meta-analysis. American Journal of Psychiatry, 157, 1552-1562. doi:10.1176/appi.ajp.157.10.1552
Summers, P., Forehand, R., Armistead, L., & Tannenbaum, L. (1998). Parental divorce during early adolescence in Caucasian families: The role of family process variables in predicting the long-term consequences for early adult psychosocial adjustment. Journal of Consulting and Clinical Psychology, 66, 327-336.
doi:0022-006X/98/53.00
Taraban, L., Shaw, D. S., Leve, L. D., Wilson, M. N., Dishion, T. J., Natsuaki, M. N., … Reiss, D. (2017). Maternal depression and parenting in early childhood: Contextual
influence of marital quality and social support in two samples. Developmental Psychology, 53, 436-449. doi:10.1037/dev0000261
Whisman, M. A., Robustelli, B. L., Beach, S. R. H., Snyder, D. K., & Harper, J. M. (2015). Marital discord and depression in middle-aged and older couples: Is taxon status associated with depression? Journal of Social and Personal Relationships, 37, 967-973. doi:10.1177/0265407514554519
Williams, J., & MacKinnon, D. P. (2008). Resampling and distribution of the product methods for testing indirect effects in complex models. Structural Equation Modelling, 15, 23–51. doi:10.1080/10705510701758166
