Family functioning in depressed and non-clinical control families

Abstract

Background:

Existing studies of depression and family functioning have used western samples to examine how depression and impaired family functioning are related, and to explore levels of discrepancy between depressed patients and their family members. The purpose of the current study is to explore these questions in a Chinese sample.

Method:

This study examined the association between family functioning and depression in a Chinese sample of 60 depressed patients and their family members and 60 non-clinical controls and their family members. The study evaluated levels of agreement between family members on a self-report measure of family functioning (Family Assessment Device) using reliability statistics. It also compared families’ self-reported family functioning to that of a trained observer using an observational rating scale (McMaster Clinical Rating Scale).

Results:

Results indicate poorer family functioning among Chinese families with a depressed family member, high to moderate agreement between patients and family members, moderate to low agreement between non-clinical participants and their family members, and moderate correlations between subjective and objective ratings of family functioning in a mainland Chinese sample.

Conclusions:

As in other cultures, depression is associated with impaired family functioning in Chinese families. There is good agreement between family members and a trained evaluator about the family impairments. The Family Assessment Device and the McMaster Clinical Rating Scale are useful for assessing family functioning in Chinese families.

Keywords

Family functioning depression Family Assessment Device McMaster Clinical Rating Scale

Background

Depression is a very common mental disorder. According to the World Health Organization (WHO), depression can lead to serious adverse outcomes including a variety of disruptive psychological symptoms, lost productivity and significant economic costs (Gladstone & Beardslee, 2009; Wang, Simon, & Kessler, 2003). Both genetic and environmental factors are implicated in the etiology and course of mood disorders (Goodyer, Herbert, Tamplin, & Altham, 2000; Silberg, Maes, & Eaves, 2010). Among environmental factors, family environment plays a crucial part in the behaviours and emotions of children and adults. Family functioning has a major influence on emotional development during childhood (Cuffe, McKeown, Addy, & Garrison, 2005), and may predict the onset of mood disorders in adults (Yap, Allen, & Sheeber, 2007). Family functioning also has implications for treatment for and recovery from depression.

Families with a depressed family member experience significant impairment of family functioning (Heru & Ryan, 2004; Weinstock, Keitner, Ryan, Solomon, & Miller, 2006). Family impairment can be defined as a family’s inability to accomplish tasks that are important for their well-being, inability to adapt to changing circumstances and inability to balance the needs of individuals in the family with that of the family system (Keitner, Heru, & Glick, 2010). Previous studies have found that family functioning of patients with depression is dysfunctional during an acute episode of depression (Friedmann et al., 1997; Keitner & Miller, 1990; Keitner, Miller, Epstein, Bishop, & Fruzzetti, 1987; Miller, Kabacoff, Keitner, Epstein, & Bishop, 1986). Evidence has also accumulated to show that families with depressed patients continue to report poorer family functioning than non-clinical families after recovery from depression (Keitner et al., 1987; Keitner & Miller, 1990; Weinstock et al., 2006). Additionally, poor family functioning is one predictor of poor illness outcome in patients with depression (Keitner & Miller, 1990; Keitner, Miller et al., 1987; Keitner, Ryan et al., 1995). In summary, family impairment is an important concomitant of major depression and influences the course and long-term outcome of the depression (Weinstock et al., 2006). There is a reciprocal interactive loop between family functioning and depression, whereby depression affects the quality of family functioning, and family functioning, in turn, affects the course of depression (Keitner & Miller, 1990; Restifo & Bogels, 2009). However, the treatment of depression including family interventions could improve the prognosis of the depressed patients and the quality of life among them and their family members. For example, the treatment of parents with depression could decrease the risk of depressive symptoms in children by improving family functioning (Silberg, Maes, & Eaves, 2010). In addition, the improvement of family and social functioning in families with depressed adolescents was found to mediate the prognosis of patients with depression, which suggests that healthy family functioning is helpful to improve the clinical course of depressed patients (Sanford et al., 2006).

However, although there has been considerable research about the relationship between depression and family functioning, it is still not clear to what extent family impairments in patients with depression are specific to the depressive illness and to what extent the impairments are a function of the sociocultural norms that influence the family. Few studies have conducted both subjective and objective methods to assess family functioning among families with depressed patients. One way to better understand the relative contributions of these factors is to study family functioning of depressed patients in different cultures. Studies to date have used mainly western samples.

Some research has demonstrated that different family members could perceive family functioning differently. For example, the adolescents with depression perceived worse family functioning than their parents (Tamplin & Goodyer, 2001). This is the first study that we are aware of that studied mainland Chinese families (both depressed and non-clinical) using both subjective and objective measures of family functioning. The present study uses reliability statistics to assess agreement between family members on a self-report measure of family functioning. Studies have suggested that there are differences in the perception of family functioning between patients and their family members (Koyama, Akiyama, Miyake, & Kurita, 2004; Miller et al., 1986; Tamplin, Goodyer, & Herbert, 1998). In particular, patients’ assessment of family functioning could be distorted by their negative mood (Koyama et al., 2004). Although some studies have explored the discrepancy between patients and family members, to date no studies have explored agreement between family members using reliability statistics.

The present study is also the first of its kind to employ both subjective and objective measures of family functioning in a mainland Chinese sample. It is important to obtain an objective perspective on the functioning of depressed families in order to determine the degree to which subjective perceptions of family functioning may be influenced by the altered mood state. An objective family assessment by a trained rater using a systematized rating scale can help to more clearly define the ways in which families cope with depression and provide useful clinical information to help guide potential interventions. Previous studies evaluating family functioning have relied mainly on patient and family self-report of their family’s functioning. Objective assessments of family functioning can differentiate actual dysfunction from potential negative biases of depressed family members (Tamplin et al., 1998).

In summary, the aims of this study were: (1) to understand differences in family functioning between families with a depressed adult and non-clinical families in a mainland Chinese population; (2) to evaluate differences in perceptions of family functioning between depressed and non-depressed family members in the same families using reliability statistics; and (3) to compare subjective perceptions of family functioning in depressed families with an interviewer-based objective evaluation of the family.

Methods

Participants

The sample comprised 60 depressed Chinese patients who met DSM-IV criteria (American Psychiatric Association, 2000) for major depressive disorder and their family members, and 60 non-clinical comparison individuals and their family members. The participants were patients (15 inpatients and 45 outpatients) from the Tongji Hospital affiliated with Tongji University School of Medicine in Shanghai, China. Patients diagnosed with major depressive disorder, if accompanied with one or more family member, were invited and their family members were randomly invited to participate in the study. At least one family member participated in the study. Patients with spouses lived together and spent about 100 hours with each other every week. In parent/adult child dyads and sibling/sibling dyads, they lived close to one another and spent about 10 hours together every week. The non-clinical sample was obtained in the community through a neighbourhood committee named Daning Community. The community controls were recruited by flyers in the community committee. At least two family members were recruited randomly in each family. As in the patient sample, controls with spouses lived together and spent about 100 hours with each other every week. In parent/adult child dyads and sibling/sibling dyads, they lived close to one another and spent about 10 hours together every week. Exclusion criteria for all participants included other mental disorders, neurological disorders, severe physical problems and active substance abuse or dependence within the three months prior to the study. Non-clinical subjects had no past history of any psychiatric disorder. This study was approved by the local Institutional Review Board of Tongji University School of Medicine. Written informed consent was obtained from all voluntary subjects after an explanation about the study.

Measures

Demographic information consisting of age, gender and level of education was collected. Patients with major depressive disorder were diagnosed by the treating psychiatrist using DSM-IV criteria. The diagnosis was confirmed by the study psychiatrist (JW) according to DSM- IV criteria. The study psychiatrist also evaluated the patients’ families using the McMaster Clinical Rating Scale (MCRS; Miller et al., 1994).

The McMaster Model of Family Functioning (MMFF; Ryan, Epstein, Keitner, Miller, & Bishop, 2005) offers a comprehensive and integrated model of family assessment and treatment. It provides normative descriptions of family functioning, focuses on multiple areas of family life and has validated instruments to assess functioning from subjective (Family Assessment Device (FAD); Epstein, Baldwin, & Bishop, 1983) and objective (MCRS) perspectives. The MMFF also has a treatment component, the Problem Centered Systems Therapy of the Family that has been described in manual form and tested in treatment outcome studies (Keitner, Archambault, Ryan, & Miller, 2003). As well as overall general functioning, the MMFF (Ryan et al., 2005) assesses six dimensions of family functioning: (1) problem-solving (the ability of the family to resolve problems to a level that maintains effective family functioning); (2) communication (how family members exchange information with each other); (3) roles (how the family allocates responsibilities to ensure fulfillment of family functions); (4) affective responsiveness (whether the family members experience and respond with a full spectrum of feelings experienced by human beings); (5) affective involvement (the family’s ability to care about and be interested in each other); and (6) behaviour control (rules that the family adopts to handle dangerous situations to meet psychobiological needs and interpersonal socializing behaviour within and outside the family).

Subjective perceptions of family functioning were assessed using the Chinese version of the FAD. The FAD (Epstein et al., 1983) is a 60-item self-report inventory that measures the perception of family members about various aspects of their functioning according to the MMFF. The FAD was translated into Chinese (Mandarin) and back-translated into English. The back-translation was reviewed by the originators of the FAD and items were re-translated to ensure that they reflected the original meaning. The validity and reliability of the Chinese FAD has been demonstrated (Shek, 2001, 2002). The test-retest reliability is 0.53–0.81, and coefficient α ranges from 0.53 to 0.94. Health/pathology cut-off scores have been established for the FAD for each dimension of family functioning (Miller et al., 1994), with scores above the health/pathology cut-off indicating unhealthy functioning. The current study uses a mainland Chinese version of the FAD (Liu & He, 1999); the internal consistency of all dimensions is fair to excellent (α = 0.55–0.92) and the test-retest reliability is acceptable (ICC = 0.56–0.85). All patients, their family members and the non-clinical families completed the FAD independent of each other.

The objective evaluation of the family was done using the MCRS (Miller et al., 1994). The MCRS is an observational assessment scale designed to be used during a comprehensive evaluation of family functioning by a clinician trained in the MMFF. The MCRS outlines anchor points for the six dimensions of the MMFF and general overall functioning. Subscale scores on the MCRS range from 1 to 7, with lower scores (1–4) representing poorer family functioning and higher scores (5–7) optimal family functioning).

Families with a depressed patient were interviewed by the study psychiatrist (JW), who received extensive training in the MMFF and the use of the MCRS at the Family Research Program, Rhode Island Hospital/Brown University. The study psychiatrist achieved reliable rates of agreement using the MCRS during her training. The inter-rater reliability of MCRS with the criterion rater was 0.92 in the current study. During the process of study, the researcher (JW) received the supervision of one of originators of the MCRS (Dr Gabor Keitner).

Previous research has demonstrated positive correlations between FAD and MCRS scores (Miller et al., 1994; Weinstock et al., 2006) in a western sample. The MCRS assessment and rating were completed prior to the subjects filling out the FAD so that the objective rating of the family was not influenced by knowledge of how the family viewed its own functioning.

Statistical analysis

χ² analyses were used to analyse significant differences between groups for categorical variables. It is unclear whether there were differences in FAD scores between patients with depression and their family members as well as between community controls and their family members depending on relationship role, and intra-class correlation coefficients (ICCs) were used to assess the level of agreement between depressed patients and their family members as well as between family members in the non-clinical group. The ICC was used to assess agreement between family members for this study. For the purposes of this study, low agreement corresponds to ICC = 0–0.3; moderate agreement to ICC = 0.3–0.7; and high agreement to ICC = 0.7–1. Mean differences between patients with depression and community controls were assessed using independent t-tests. The frequencies of scores above and below established FAD health/pathology cut-off scores in families with patients and controls were analysed and compared. Pearson correlations were used to determine the correlations between subjectively assessed (FAD) and interviewer-assessed (MCRS) family functioning in the families with a depressed patient.

Results

Demographic characteristics

The depressed patients from 60 families had the following characteristics: 19 (31.7%) were male and 41(68.3%) were female; mean age was 42.6 years (SD = 14.6); mean years of schooling was 10.6 years (SD = 4.1); 45 (75.0%) had 12 years of schooling and no university degree; 13 (21.7%) had a university degree; and 2 (3.3%) had a graduate degree. The severity of depression was assessed by the Clinical Global Impression (CGI-clinician) and patients used the Visual Analog Mood Scale (VAMS) of 1–10 with 1 being ‘not at all depressed’ and 10 being ‘most depressed’. The mean CGI score was 4 (SD = 0.64; range = 3–6). The mean VAMS score was 7.28 (SD = 1.45; range = 4–10).

The community controls from 60 non-clinical community families had the following characteristics: 18 (30.0%) were male and 42 (70%) were female; mean age was 38.0 years (SD = 12.9); mean years of schooling was 13.6 (SD = 3.8); 28 (46.7%) had 12 years of schooling and no university degree; 24 (40.0%) had a university degree; and 8 (13.3%) had a graduate degree.

There were no significant differences between the two groups in age, education and gender. The patients with depression and the controls came from a similar social class, and from the same district in Shanghai. Both patients and community families included two family members. In the depressed families there were 29 husband/wife dyads, 25 parent/adult child dyads, 2 sibling/sibling dyads, 2 grandparent/grandchild dyads and 2 patient/other relative dyads. In the community families there were 37 husband/wife dyads, 19 parent/adult child dyads and 4 sibling/sibling dyads. All subjects were over the age of 18.

Comparison of family functioning between depressed families and community families

A mean FAD score was computed for each family that participated. Table 1 lists the comparison of family functioning between depressed families and community families. Families with a depressed patient reported problems in all dimensions of their family’s functioning, while non-clinical families reported problems only on the behavioural control dimension. Mean FAD scores were significantly worse for depressed families on all dimensions except affective involvement, which suggests that those six dimensions (roles, problem-solving, communication, behaviour control, affective responsiveness and general functioning) may be particularly problematic for depressed families in China, while the behaviour control dimension was perceived as problematic for both depressed families and control families.

Table 1.

FAD family means, standard deviations and comparisons between depressed and control families.

	Depressed (n = 60)	Control (n = 60)	t (d.f. = 118)
PS	2.39 (0.32)	1.98 (0.29)	7.50**
CM	2.35 (0.30)	2.05 (0.22)	6.27**
RL	2.38 (0.24)	2.18 (0.21)	4.65**
AR	2.38 (0.31)	2.02 (0.31)	6.52**
AI	2.10 (0.40)	2.08 (0.26)	0.33
BC	2.60 (0.20)	2.14 (0.25)	11.28**
GF	2.28 (0.27)	1.88 (0.25)	8.41**

Note: PS = problem-solving; CM = communication; RL = roles; AR = affective responsiveness; AI = affective involvement; BC = behaviour control; GF = general functioning. Scores above cut-off are in italics. Higher scores indicate worse family functioning.

p < .01.

Agreement between family members

Table 2 presents mean values for depressed patients and their family members, as well as non-clinical controls and their family members on each dimension of the FAD. In depressed families, patients’ means were in the unhealthy range on all dimensions, and their family members’ means were in the unhealthy range on all dimensions except for affective involvement. Mean family scores for community controls were in the healthy range on all dimensions, except for behaviour control, indicating overall healthy family functioning.

Table 2.

FAD mean scores for patients and their family members and healthy controls and their family members.

	Patient M (n = 60)	Patient family member M (n = 60)	Patient and family ICC	Control M (n = 60)	Control family member M (n = 60)	Control and family ICC (n = 60)
PS	2.47 (0.36)	2.31 (0.36)	0.54**	1.97 (0.34)	1.97 (0.35)	0.41**
CM	2.41 (0.35)	2.29 (0.33)	0.54**	2.02 (0.28)	2.06 (0.26)	0.29*
RL	2.39 (0.29)	2.36 (0.27)	0.44**	2.16 (0.26)	2.19 (0.25)	0.39*
AR	2.44 (0.40)	2.32 (0.31)	0.55**	1.96 (0.36)	2.06 (0.36)	0.50**
AI	2.14 (0.46)	2.06 (0.39)	0.74**	2.07 (0.30)	2.09 (0.34)	0.28*
BC	2.62 (0.27)	2.58 (0.23)	0.28*	2.13 (0.29)	2.13 (0.27)	0.57**
GF	2.34 (0.25)	2.22 (0.28)	0.47**	1.88 (0.31)	1.87 (0.27)	0.51**

p < .05, ** p < .01.

Overall, Table 2 shows that there were higher rates of agreement between depressed patients and their family members than between controls and their family members. Depressed families evidenced high agreement on affective involvement, moderate agreement on problem-solving, communication, roles, affective responsiveness and general functioning, and low agreement on behaviour control. Control families evidenced moderate agreement on problem-solving, roles, affective responsiveness, behaviour control and general functioning, and low agreement on communication and affective involvement.

Previous research found that there were some differences of perceived family functioning. For example, the adolescents with depression perceived worse family functioning than their parents (Tamplin & Goodyer, 2001). Therefore, the hypothesis would be that the perceived family functioning by different dyads would be different. To test the hypothesis, intra-class correlations among husband/wife dyads, parent/adult child dyads, sibling/sibling dyads and other dyads were conducted. In order to better understand agreement between family members, we explored whether level of agreement (Table 3) differed depending on relationship role (i.e. spouse, mother, father, child, sibling, grandparent). In depressed families, agreement between depressed patients and their children was high on communication, affective involvement and general functioning, moderate on problem-solving, roles and affective responsiveness, and low on behaviour control. By contrast, agreement between depressed patients and their parents was in the low range on problem-solving, roles, affective responsiveness and behaviour control, in the moderate range on general functioning, and in the high range on affective involvement. Agreement between depressed patients and their spouses was in the moderate range on all scales except affective involvement, which was in the high range. Agreement between depressed patients and their siblings was in the high range on all scales, although the sample size for this comparison group was low. Of note, all family dyads in the depressed group evidenced high levels of agreement on affective involvement.

Table 3.

ICC of FAD mean scores for patients and their family members and healthy controls and their family members.

	Patient and child ICC (n = 12)	Patient and parent ICC (n = 13)	Patient and spouse ICC (n = 29)	Patient and sibling ICC (n = 2)	Patient and family ICC (n = 60)	Control and child ICC (n = 8)	Control and parent ICC (n = 11)	Control and spouse ICC (n = 37)	Control and sibling ICC (n = 4)	Control and family ICC (n = 60)
PS	0.69**	0.27	0.57**	0.92**	0.54**	0.50**	0.09	0.39*	0.09	0.41**
CM	0.82**	0.07	0.45**	0.93**	0.54**	0.32*	0.42**	0.33*	0.33*	0.29*
RL	0.63**	0.03	0.36**	0.98**	0.44**	0.49**	0.31*	0.27	0.15	0.39*
AR	0.62**	0.10	0.48**	0.86**	0.55**	0.82**	0.08	0.45**	0.09	0.50**
AI	0.72**	0.82**	0.73**	0.82**	0.74**	0.47**	0.19	0.31*	0.40*	0.28*
BC	0.08	0.10	0.32*	0.96**	0.28*	0.95**	0.01	0.53**	0.32	0.57**
GF	0.73**	0.35	0.45**	0.97**	0.47**	0.23	0.63**	0.56**	0.46**	0.51**

Note: PS = problem-solving; CM = communication; RL = roles; AR = affective responsiveness; AI = affective involvement; BC = behaviour control; GF = general functioning.

p < .05, ** p < .01.

Levels of agreement between specific dyadic constellations of control families evidenced a different pattern, with lower levels of agreement on most scales in most dyads. Controls and their children had high levels of agreement on behaviour control, low agreement on general functioning and moderate agreement on all other scales. Controls and their parents had low levels of agreement on problem-solving, affective responsiveness, affective involvement and behaviour control, and moderate agreement on communication, roles and general functioning. Controls and their spouses had moderate agreement on all scales except roles, which was in the low range. Controls and their siblings had low levels of agreement on problem-solving, roles and affective responsiveness, and moderate levels of agreement on communication, affective involvement, behaviour control and general functioning.

In summary, depressed family dyads had higher levels of agreement overall than control family dyads. All depressed family dyads evidenced high levels of agreement on affective involvement. Patients and controls and their children endorsed higher levels of agreement than did patients and controls and their parents.

Proportion of depressed vs control families endorsing unhealthy family functioning

In order to compare the depressed and control groups in terms of health/pathology cut-off scores, we calculated frequencies of the number of depressed and control families above and below established FAD health/pathology cut-off scores. Scores above the healthy/pathology cut-off score indicate unhealthy functioning. χ² tests were used to assess the differences between the frequencies of the number of depressed and control families above and below established FAD cut-off scores. Table 4 lists FAD health/pathology cut-off scores (Miller et al., 1994), along with proportions of depressed and control families in the healthy and unhealthy range on FAD dimensions. On all dimensions except affective involvement, significantly more depressed than control families were in the unhealthy range. In addition, 75% of control group families reported the behaviour control dimension as unhealthy.

Table 4.

Frequency table depicting number of families in healthy and unhealthy range on the FAD by diagnostic condition.

	Cut-off score^a	Number of depressed families in the unhealthy range (%)	Number of depressed families in the healthy range (%)	Number of control families in the unhealthy range (%)	Number of control families in the healthy range (%)	p
PS	2.20	41 (68.3)	19 (31.7)	10 (16.7)	50 (83.3)	< .0001
CM	2.20	43 (71.7)	17 (28.3)	13 (21.7)	47 (78.3)	< .0001
RL	2.30	37 (61.7)	23 (38.3)	17 (28.3)	43 (71.7)	< .0001
AR	2.20	40 (66.7)	20 (33.3)	14 (23.3)	46 (76.7)	< .0001
AI	2.10	31 (51.7)	29 (48.3)	27 (45.0)	33 (55.0)	.465
BC	1.90	60 (100.0)	0 (0)	48 (80.0)	12 (20.0)	< .0001
GF	2.00	49 (81.7)	11 (18.3)	13 (21.7)	47 (78.3)	< .0001

Miller, Epstein, Bishop, & Keitner (1985)

Note: PS = problem-solving; CM = communication; RL = roles; AR = affective responsiveness; AI = affective involvement; BC = behaviour control; GF = general functioning.

Subjective (FAD) and objective (MCRS) assessment of depressed families

A comparison of means and standard deviations for self-report ratings of family functioning (FAD) versus clinician ratings of family functioning using the MCRS is presented in Table 5. Average scores on all measures fell within the unhealthy range on both clinician-rated and self-report measures (Miller et al., 1994). Correlations between the FAD and the MCRS were moderate on all scales except problem-solving and affective responsiveness; subjective and objective ratings were not significantly correlated on these scales.

Table 5.

Comparison between subjective (FAD) and objective (MCRS) ratings of family functioning.

	Depressed FAD family score M (SD)	Depressed MCRS family score M (SD)	r
PS	2.39 (.32)	3.58 (.87)	−0.195
CM	2.35 (.30)	3.77 (.74)	−0.595**
RL	2.38 (.24)	3.68 (.75)	−0.447**
AR	2.38 (.31)	3.60 (.53)	−0.094
AI	2.10 (.40)	3.97 (.82)	−0.288**
BC	2.60 (.20)	4.00 (.66)	−0.343**
GF	2.28 (.27)	3.72 (.55)	−0.400**

p<.01.

Correlation between severity of depression and family functioning

Table 6 shows the correlations between severity of depression and family functioning perceived by depressed patients. There were positive correlations between CGI scores and FAD scores on all dimensions but for affective responsive and affective involvement, and there were positive correlations between VAMS scores and FAD scores on all dimensions but for affective responsive, indicating that depression severity may be positively associated with family dysfunction in the patients with depression.

Table 6.

Correlations between severity of depression and family functioning.

	PS	CM	RL	AR	AI	BC	GF
CGI (Clinician)	0.281**	0.302**	0.276**	0.191	0.157	0.297**	0.301**
VAMS (Patient)	0.276**	0.298**	0.284**	0.185	0.279**	0.297**	0.288**

Note: PS = problem-solving; CM = communication; RL = roles; AR = affective responsiveness; AI = affective involvement; BC = behaviour control; GF = general functioning; CGI = Clinical Global Impression; VAMS = Visual Analog Mood Scale.

p < .05, ** p < .01.

Discussion

Previous research has shown that families with a depressed patient experience family impairment (Keitner & Miller, 1990; Weinstock et al., 2006). Consistent with this, the present study in a mainland Chinese sample found significantly worse perceptions of family functioning among depressed families as compared to control families on all dimensions except affective involvement. Communication and problem-solving have been demonstrated to be problematic in depressed western families (Keitner & Miller, 1990). Taken together with other cross-cultural studies, our findings indicate a consistent association between depression and problematic family functioning (Keitner et al., 1991; Koyama et al., 2004; Saeki, Asukai, Miyake, Miguchi, & Yamawaki, 2002). However, the specific areas of impaired family functioning may vary by culture.

Our findings indicate that in a mainland Chinese sample, depressed patients and their family members reported worse functioning in many domains of family life in comparison to non-clinical individuals and their family members. These findings are similar to those found in other countries, confirming that depression has a negative impact on family functioning in all cultures studied (Keitner et al., 1991; Koyama et al., 2004; Restifo & Bogels, 2009). These findings were consistent using both subjective and objective measures. Patients’ perceptions of family functioning do not appear to be distorted by their depression, as evidenced by the level of agreement between patients and their families. However, the results showed that 21% of the patients had a university background and 40% in the control group had a university background, which could have an influence on the comparison of family functioning between patient group and control group. In Chinese Shanghai, the percentage of people with a university background is about 8.8% in the general population. The target population in our study came from an economically developed district, and the educational level was higher than the normal level in Shanghai. The percentage of people with a university degree in rural districts is lower than that in Shanghai. The educational level could influence perceived family functioning. Therefore, our results only represent the relationship between family functioning and depression in Shanghai.

Chinese families with a depressed member reported particular concerns about their ability to solve problems, to communicate effectively with each other, to arrange roles, to experience a satisfactory level of emotional responsiveness with each other, and to establish satisfactory expectations and rules in the family. The behaviour control dimension in non-clinical families was also in the unhealthy range, suggesting that even non-clinical families are dissatisfied with rules and expectations in their families. There are two possible explanations for elevated behaviour control scores in both groups in this study. It may be that setting clear and consistent expectations for rules in the family is an issue for both clinical and non-clinical Chinese families. Perhaps the nature of authority in Chinese culture, which emphasizes strong external authority and conformity to rules, makes it difficult for families to develop internal authority structures for themselves. Alternatively, the unhealthy ratings on the behaviour control dimension may be an artifact of the scale items, several of which emphasize dealing with emergency conditions. In Chinese culture, little emphasis is put on coping with these conditions. As such, elevated scores on this scale may not indicate family impairment, but rather a cultural pattern of not talking about how to respond to emergency situations (Shek, 2002). The findings of behaviour control being similarly worse in both cases and controls might need further elaboration on Chinese concepts of ‘control’ and ‘care’, which could be culturally different from the western concept of ‘care’ and ‘autonomy’. It would also suggest that the cultural validity of this instrument for Chinese families would need further review and exploration.

With regard to agreement between family members, two important patterns emerged: there was greater agreement between family members in depressed families than in control families, and there was greater agreement between family members and their adult children than between family members and their parents. These findings are consistent with previously published results in western families (e.g. Tamplin et al., 1998). The results also show that the agreement between adult patients and their parents is much less than that between patients and their adult children. One explanation could be that when children experience depression, their parents do not know why they are depressed. On the flip side, when parents experience depression, their children also perceive their problem, and may therefore have a better understanding of it. Another explanation could be that the patients did not live with their parents and siblings, and they may not totally understand the patients’ perceived family functioning. Controls and their children had good agreement on family functioning, suggesting that they have consistent opinions about their family’s functioning. Patients and their spouses had moderate agreement, while controls and their spouses had poor to moderate agreement. Perhaps when a married person is depressed, his/her spouse notices the problem and they therefore perceive family functioning similarly. Likewise, a depressed person and his or her siblings may be more attuned to each other due to the depression.

The good agreement on affective involvement in depressed families suggests that depressed patients and their family members perceive affective involvement similarly, while non-clinical families endorsed less agreement. Perhaps in normal couples, spouses feel less involved with each other, due perhaps to more of a focus on their own needs. When there is a depressed patient in the family, family members may pull together, and therefore perceive functioning similarly, but lacking such pressure, non-clinical family members may not perceive functioning similarly.

Also, findings from the present study suggest that self-reports of family functioning using the FAD are reliable, as evidenced by significant moderate correlations between objective and subjective measures of family functioning. Patients, family members and objective clinical raters perceived family functioning similarly in the present sample, which indicates that patient concerns about family functioning cannot be dismissed as merely reflecting a distorted perception of their social situation.

Finally, the results also indicated that severity of depression was positively correlated with family dysfunction in most domains of perceived family functioning for patients with depression and family functioning assessed by clinicians, indicating that severity of depression may influence the degree of unhealthy family functioning. This reminds clinicians about the importance of family interventions when treating patients with depression. The findings were consistent with previous research that has indicated correlations between depression severity and most domains of family functioning in depressed patients (Febres, Rossi, Gaudiano, & Miller, 2011).

Findings from the present study suggest that when treating Chinese depressed patients, clinicians should be particularly sensitive to concerns about perceived problems between family members. Therefore, the treatment of depression should also address family concerns in addition to medical treatment. Combining subjective with objective family assessments can help clinicians to understand family functioning more comprehensively.

Given the problematic scores on behaviour control, depressed patients and their families may benefit from the chance to discuss whether they want to change their pattern of setting rules and expectations in their family.

Limitations

Limitations of the present study include the FAD norms being based on western samples. In addition, a high percentage of the target population in the study had a university degree and education level could play a role in family functioning. Only 60 families of patients and 60 families of controls were recruited in the study, and there were only a few sibling/sibling dyads among them. This study did not include a structured clinical interview for diagnosis of depression. In addition, data were not collected on co-occurring psychiatric illnesses in this sample. Finally, the objective evaluation of family functioning was conducted by the psychiatrist who was aware of the status of the participant as cases. This could lead to interviewer bias that might tend to rate the depressed families as worse functioning.

Conclusion

Despite the aforementioned limitations, the present study is the first of its kind to assess family functioning in a mainland Chinese sample using both subjective and objective assessments. Furthermore, it is the first study of its kind to assess agreement between family members on the FAD using ICCs.

Future studies should explore the relationship between family functioning and depression including the population in rural areas. Future research should replicate these findings in a larger sample, especially the sibling/sibling dyads. Replication studies are needed to determine whether the present findings are consistent, a structured diagnostic interview should be used, and there should be a control for co-occurring psychiatric diagnoses. Future studies should also explore different ways that families in other cultures are impacted by depression and also explore the moderating effects of social support, co-morbid illnesses and stressful life events. Longer-term follow-up studies could also evaluate how family functioning influences the course and outcome of depressive episodes.

The FAD and the MCRS (Barney & Max, 2005; Weinstock et al., 2006) appear to be useful instruments, even in China, in helping clinicians to acquire a more comprehensive understanding of family functioning. The FAD may be a cost-effective screening tool as it can be filled out by family members prior to meeting with a clinician, and can identify specific dimensions of family life that are perceived as being problematic so as to focus further clinical exploration and intervention.

Footnotes

Acknowledgements

This work has been supported by a grant from the Foundation of the Ministry of National Science and Technology (2009BAI77B05).

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders, text revision, 4th edn. Washington, DC: American Psychiatric Association.

Barney

M. C.

Max

J. E.

(2005). The McMaster family assessment device and clinical rating scale: Questionnaire vs interview in childhood traumatic brain injury. Brain Injury, 19, 801–809.

Cuffe

S. P.

McKeown

R. E.

Addy

C. L.

Garrison

C. Z.

(2005). Family and psychosocial risk factors in a longitudinal epidemiological study of adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 44, 121–129.

Epstein

N. B.

Baldwin

L. M.

Bishop

D. S.

(1983). The McMaster Family Assessment Device. Journal of Marital and Family Therapy, 9, 171–180.

Febres

Rossi

Gaudiano

B. A.

Miller

I. W.

(2011). Differential relationship between depression severity and patients’ perceived family functioning in women versus in men. Journal of Nervous and Mental Disease, 199, 449–453.

Friedmann

M. S.

McDermut

W. H.

Solomon

D. A.

Ryan

C. E.

Keitner

G. I.

Miller

I. W.

(1997). Family functioning and mental illness: A comparison of psychiatric and nonclinical families. Family Process, 36, 357–367.

Gladstone

T. R.

Beardslee

W. R.

(2009). The prevention of depression in children and adolescents: A review. Canadian Journal of Psychiatry, 54, 212–221.

Goodyer

I. M.

Herbert

Tamplin

Altham

P. M.

(2000). Recent life events, cortisol, dehydroepiandrosterone and the onset of major depression in high-risk adolescents. British Journal of Psychiatry, 177, 499–504.

Heru

A. M.

Ryan

C. E.

(2004). Burden, reward and family functioning of caregivers for relatives with mood disorders: 1-year follow-up. Journal of Affective Disorders, 83, 221–225.

10.

Keitner

G. I.

Archambault

Ryan

C. E.

Miller

I. W.

(2003). Family therapy and chronic depression. Journal of Clinical Psychology, 59, 873–884.

11.

Keitner

G. I.

Fodor

Ryan

C. E.

Miller

I. W.

Bishop

D. S.

Epstein

N. B.

(1991). A cross-cultural study of major depression and family functioning. Canadian Journal of Psychiatry, 36, 254–259.

12.

Keitner

G. I.

Heru

A. M.

Glick

I. D.

(2010). Clinical manual of couples and family therapy. Washington, DC: American Psychiatric Publishing Incorporation.

13.

Keitner

G. I.

Miller

I. W.

(1990). Family functioning and major depression: An overview. American Journal of Psychiatry, 147, 1128–1137.

14.

Keitner

G. I.

Miller

I. W.

Epstein

N. B.

Bishop

D. S.

Fruzzetti

A. E.

(1987). Family functioning and the course of major depression. Comprehensive Psychiatry, 28, 54–64.

15.

Keitner

G. I.

Ryan

C. E.

Miller

I. W.

Kohn

Bishop

D. S.

Epstein

N. B.

(1995). Role of the family in recovery and major depression. American Journal of Psychiatry, 152, 1002–1008.

16.

Koyama

Akiyama

Miyake

Kurita

(2004). Family functioning perceived by patients and their family members in three Diagnostic and Statistical Manual-IV diagnostic groups. Psychiatry and Clinical Neurosciences, 58, 495–500.

17.

Liu

P. Y.

M. T.

(1999). Family Assessment Device, FAD. Chinese Mental Health Journal, Supplement, 149–151. [in Chinese]

18.

Miller

I. W.

Epstein

N. B.

Bishop

D. S.

Keitner

G. I.

(1985). The McMaster Family Assessment Device: Reliability and validity. Journal of Marital Family Therapy, 11, 345–356.

19.

Miller

I. W.

Kabacoff

R. I.

Epstein

N. B.

Bishop

D. S.

Keitner

G. I.

Baldwin

L. M.

van der Spuy

H. I.

(1994). The development of a clinical rating scale for the McMaster model of family functioning. Family Process, 33, 53–69.

20.

Miller

I. W.

Kabacoff

R. I.

Keitner

G. I.

Epstein

N. B.

Bishop

D. S.

(1986). Family functioning in the families of psychiatric patients. Comprehensive Psychiatry, 27, 302–312.

21.

Restifo

Bogels

(2009). Family processes in the development of youth depression: Translating the evidence to treatment. Clinical Psychological Review, 29, 294–316.

22.

Ryan

C. E.

Epstein

N. B.

Keitner

G. I.

Miller

I. W.

Bishop

D. S.

(2005). Evaluating and treating families: The McMaster approach. New York: Routledge.

23.

Saeki

Asukai

Miyake

Miguchi

Yamawaki

(2002). Characteristics of family functioning in patients with endogenous monopolar depression. Hiroshima Journal of Medical Science, 51, 55–62.

24.

Sanford

Boyle

McCleary

Steele

Duku

Offord

(2006). A pilot study of adjunctive family psychoeducation in adolescent major depression: Feasibility and treatment effect. Journal of the American Academy of Child and Adolescent Psychiatry, 45, 386–495.

25.

Shek

D. T.

(2001). The General Functioning Scale of the Family Assessment Device: Does it work with Chinese adolescents? Journal of Clinical Psychology, 57, 1503–1516.

26.

Shek

D. T.

(2002). Assessment of family functioning in Chinese adolescents: The Chinese version of the Family Assessment Device. Research Social Work Practice, 12, 502–524.

27.

Silberg

J. L.

Maes

Eaves

L. J.

(2010). Genetic and environmental influences on the transmission of parental depression to children’s depression and conduct disturbance: An extended Children of Twins study. Journal of Child Psychology and Psychiatry, 51, 734–744.

28.

Tamplin

Goodyer

I. M.

(2001). Family functioning in adolescents at high and low risk for major depressive disorder. European Child and Adolescent Psychiatry, 10, 170–179.

29.

Tamplin

Goodyer

I. M.

Herbert

(1998). Family functioning and parent general health in families of adolescents with major depressive disorder. Journal of Affective Disorders, 48, 1–13.

30.

Wang

P. S.

Simon

Kessler

R. C.

(2003). The economic burden of depression and the cost-effectiveness of treatment. International Journal of Methods in Psychiatric Research, 12, 22–33.

31.

Weinstock

L. M.

Keitner

G. I.

Ryan

C. E.

Solomon

D. A.

Miller

I. W.

(2006). Family functioning and mood disorders: A comparison between patients with major depressive disorder and bipolar I disorder. Journal of Consulting and Clinical Psychology, 74, 1192–1202.

32.

Yap

M. B.

Allen

N. B.

Sheeber

(2007). Using an emotion regulation framework to understand the role of temperament and family processes in risk for adolescent depressive disorders. Clinical Child and Family Psychology Review, 10, 180–196.