Associations between parental bonding and health-related quality of life in a clinical sample of youths aged 14–35 years in Singapore

Abstract

Introduction:

Individuals with psychiatric disorders tend to report having poorer bonds with their parents during their early years. These individuals often experience lower quality of life as well. This study investigated the associations between aspects of parental bonding and health-related quality of life (HRQOL) in a clinical sample of youths. It was hypothesised that high parental care and low parental overprotectiveness would be associated with higher levels of HRQOL.

Methods:

Data were obtained from a larger cross-sectional study. The sample consisted of 400 psychiatric outpatients: 191 patients aged 14–21 (mean ± standard deviation 18.1 ± 2.22) years and 209 patients aged 22–35 (28.0 ± 4.33) years. The Parental Bonding Instrument was used to measure parental care and overprotectiveness. Short Form-12 measured physical health (physical component summary [PCS]) and mental health (mental component summary [MCS]) components of HRQOL, and the 8-item Patient Health Questionnaire assessed depressive symptoms. These scales and a sociodemographic form were self-administered. Multivariable linear regression was used for analysis.

Results:

About half of the sample reported affectionless control for mothers (46.6%) and fathers (45.9%). After controlling for sociodemographic variables, no significant relationship was found between aspects of parental bonding and PCS scores. Maternal care was associated with MCS scores (β = 0.32, P < 0.01) and PHQ-8 scores (β = -0.12, P < 0.05).

Conclusion:

Our results suggest that youths who experienced quality care from their mothers exhibit better mental health functioning despite their clinical diagnoses, which suggests that early maternal care exerts an overall long-term protective effect. Early parental education that promotes positive parenting practices could improve the overall HRQOL of individuals in adulthood despite their clinical diagnoses.

INTRODUCTION

Decades of research have identified different aspects of parenting and sought to elucidate their impact on child outcomes and development. The contextual model of parenting posited parenting style as a context that is influenced by parental goals and values, and as a key influence on the child’s receptiveness to parents.[1] Both parenting style and the child’s receptiveness impact the child’s development indirectly by moderating the effect of parenting practices. The parent–child bond also exerts mediating effects on parenting style and child outcomes.[2] Bonding patterns set the emotional climate for parent–child interactions, and evidence shows that parental bonding can have long-term effects on adolescents’ and young adults’ later well-being.[3,4]

SUMMARY BOX

What is known?

Parental bonding has a long-term impact on the later well-being of individuals in the general population.

What is new?

Affectionless control was the most common maternal and paternal bonding pattern in the clinical sample. Youths who experienced early maternal care show better mental health functioning despite their psychiatric condition.

What is the impact?

Findings can be useful to clinical practitioners and parent educators to identify families with non-optimal bonds and intervene early to mitigate the negative effects of poor bonds on quality of life and psychopathology.

Parental bonding is defined as “the parental behaviours and attitudes that contribute to the parent–child bond”.[5] It includes a range of key parenting characteristics that affect the formation and quality of the child’s attachment and the parent–child relationship. These characteristics, such as care, affection, encouragement of autonomy and independence, as well as indifference, rejection, psychological and behavioural control, and punitive actions, are conceptualised onto two principal dimensions: care versus rejection/indifference and overprotection versus encouragement of autonomy. Bonding experiences are thus classified into four different types: optimal bonding, affectionate constraint, affectionless control and absent or weak bonding [Figure 1].

Figure 1.

Dimensions and types of parental bonding.

Parental bonding is related to but differs from parenting style. Baumrind’s parenting styles (i.e. authoritarian, authoritative, and permissive parenting) were conceptualised based on the extent of parental control over the child.[6] These parenting styles were later reformed to include a second principle of responsiveness, thus deriving four types of parenting styles: authoritative, authoritarian, indulgent and neglectful parenting.[7] Although parental bonding and parenting style use similar dimensions where care and overprotection correspond to responsiveness and demandingness, respectively,[8] they refer to different facets of parenting.[2] For example, parents who are caring and do not enforce much control (i.e. show warmth and grant the child more autonomy) fit the criteria of indulgent parenting style, but these parental behaviours are more likely to contribute to more positive bonding experiences with their child. However, the parent–child bond is more likely to be characterised as affectionless control when parents practise authoritarian parenting (i.e. indifferent and highly controlling).

Among the multitude of parenting characteristics that contribute to the formation of the parent–child bond, parental affection and encouragement of autonomy are protective factors, while emotional neglect and overprotection are risk factors for psychopathology.[9] Poor bonding experiences increase adolescents’ vulnerability to psychiatric symptoms such as depression and anxiety.[10,11] Indeed, clinical populations generally report different parental bonding patterns as compared to those without psychiatric conditions. A recent systematic review found that individuals with mood and anxiety-related disorders recalled lower parental care and higher overprotection.[12,13] Similar bonding patterns were also observed among individuals with other conditions such as eating disorders,[14] psychotic disorders[15,16,17] and personality disorders.[18] In fact, affectionless control appears to be consistent across different diagnoses.[19,20]

Optimal parental bonding has a long-term positive influence on mental well-being.[21,22] However, those with authoritarian parents, whose parental bonds are characterised as affectionless control, reported lower quality of life (QOL).[23] In a national study involving more than 14,000 adolescents, family dimensions identified as the most important determinants of health-related quality of life (HRQOL) were encouragement of autonomy, family activities and particularly, parental affection.[24] A study that compared patients with chronic regional pain and patients with mental illness found that parents, especially mothers of the latter group were less caring and protective, and that patients with mental illness scored worse in multiple QOL domains such as physical health, mental health and personal relationships.[25] Although optimal bonding is typically linked to better well-being, studies suggest that the influence of parental care is independent of parental control.[21,22] In other words, at both low and high levels of overprotection, QOL is higher when parental care is present. Interestingly, several studies also reported that only the lack of parental care is significantly associated with psychopathology.[20,26]

Literature seems to suggest different effects of paternal and maternal care.[21] Individuals who experienced optimal maternal bonding reported better general health and higher QOL in areas such as social relationships, psychological health and physical health.[27,28] Yet in another study, the quality of paternal care together with childhood maltreatment and depressive symptoms were stronger predictors of QOL instead.[29] A local study also found differences in the effects of maternal and paternal care, whereby only the latter interacted with childhood externalising problems and predicted externalising symptoms in early adulthood.[30] To the best of our knowledge, there are mixed results on this topic and no conclusive evidence has been drawn regarding whether paternal or maternal care exerts greater influence on the child’s functioning. There is little evidence on how different parental bonding types contribute to the lower QOL observed among the clinical population too. Furthermore, parental bonding studies that specifically examine HRQOL are especially limited; this is a gap the present study aims to explore.

In summary, studies have found that children have higher QOL and better relationships with parents who show care and encourage autonomy,[23,27] whereas children whose parents are more controlling or neglectful have lower QOL and increased risk for psychopathology.[31] Yet, despite the strong association between psychopathology and lower QOL,[32] few studies have directly examined the role of parental bonding in a clinical population. Therefore, this study aimed to examine the associations between aspects of parental bonding and HRQOL among youths diagnosed with a mental illness in Singapore. We hypothesised that youths who experienced higher parental care and lower parental overprotection would report better physical and mental health functioning.

METHODS

Data for this study were taken from a cross-sectional study examining the prevalence of deliberate self-harm in a clinical sample of youths aged 14–35 years.[33] The age range was determined to accommodate the definitions of a young person (aged 14–16 years) and a youth (aged 15–35 years) as stated by the Children and Young Persons Act[34] and the National Youth Council,[35] respectively. Participants were outpatients from the Institute of Mental Health, a tertiary psychiatric hospital in Singapore. Using a convenience sampling strategy, participant recruitment was conducted between October 2015 and June 2016 through clinician referrals and poster advertisements. Clinicians were informed of the study, and they referred eligible patients to the study team. Patients could also participate by responding to the posters put up in the outpatient clinics. Individuals who were willing to participate were eligible for the study if they were outpatients aged between 14 and 35 years, citizens or permanent residents of Singapore, literate in English language and had a psychiatric condition. Individuals with an intellectual disability diagnosis or who were unable to read and speak English were excluded. All participants provided informed consent and were screened for eligibility by the study team before they completed a sociodemographic form and a self-administered survey questionnaire. Parental consent for those aged 14–20 years was waived in the original study due to sensitivity concerning the reporting of non-suicidal self-injury behaviours. Participants’ primary clinical diagnoses were later retrieved from their medical records. Upon completion of the survey, participants were given an inconvenience fee for their participation in the study. Ethics approval was obtained from the institutional ethics committee, the National Healthcare Group Domain Specific Review Board (No. 2014/01099). The study was carried out in accordance with the Declaration of Helsinki.

The Parental Bonding Instrument (PBI) measures parental characteristics as perceived by the child in their first 16 years of life.[5] The instrument comprises two scales: care (12 items) and overprotection/control (13 items). Each item is rated on a 4-point scale from 0 (very like my parent) to 3 (very unlike my parent). Participants answered all items for each parent. Care and protection cutoff scores established in the original study were based on a general population sample (mothers: 27.0 and 13.5 and fathers: 24.0 and 12.5, respectively).[36,37] The PBI has been used in previous local studies,[15,30,38] and its scales showed good internal consistency scores of 0.92, 0.83, 0.92, and 0.85 for maternal care, maternal overprotection, paternal care and paternal overprotection respectively, in this study.

Short Form-12 (SF-12) is an HRQOL measure that assesses eight health domains: physical functioning, physical role, bodily pain, general health, vitality, social functioning, emotional role and mental health.[39] The first four domains form the physical component summary (PCS) scale and the other four domains form the mental component summary (MCS) scale. Both summary scores range from 0 to 100, with higher score indicating better health. The SF-12 is well-validated[40,41] with the PCS and MCS scales showing acceptable internal consistency of 0.79 and 0.82, respectively, in this study.

Patient Health Questionnaire-8 (PHQ-8) assesses the severity of self-reported depressive symptoms.[42,43] Participants were asked to rate how often they were bothered by each symptom in the past 2 weeks on a 4-point scale from 0 (not at all) to 3 (nearly every day). A score of 5 and above indicates presence of depressive symptoms, with scores 5–9, 10–14, 15–19 and 20–24 indicating mild, moderate, moderately severe and severe depression, respectively. The PHQ-8 has been used in the local context,[44] and it showed an excellent internal consistency of 0.91 in this study.

Sociodemographic information collected and used in this analysis included age, gender, ethnicity, marital status, sexual orientation, highest level of education, employment status and housing type (a socioeconomic status indicator).

Data were analysed using STATA version 15 (StataCorp LP, College Station, TX, USA) with a two-sided test and at a statistical significance level of P < 0.05. Descriptive statistics were performed on the overall sample and the two age groups (14–21 and 22–35 years). Categorical variables were represented as frequencies and percentages. Continuous variables were represented as mean and standard deviation (SD). A series of multivariable linear regression analyses were performed to examine the association between PBI factors with PCS, MCS and PHQ-8. In separate models, analyses were performed individually with PCS, MCS and PHQ-8 scores as the dependent variables. Sociodemographic variables and clinical diagnosis were included as control variables.[45] All assumptions of linear regression were not severely violated, and multicollinearity was absent according to the variance inflation factor analysis.[46]

RESULTS

A total of 400 outpatients were recruited for the study (age: mean ± SD 23.3 ± 6.04 years). The 14–21 years age group consisted of 191 participants (18.1 ± 2.22 years), while the 22–35 years age group consisted of 209 participants (28.0 ± 4.33 years). Overall, the sample comprised almost equal representation of both genders (51.2% males); the participants were mainly of Chinese ethnicity (71.3%), grew up in a dual parent household (76.5%) and were of heterosexual orientation (81.2%). Majority of the participants had a primary diagnosis of mood disorder (34.3%), followed by psychotic disorder (22.8%), adjustment disorder (16.3%), anxiety disorder (15.3%), childhood disorder (5.3%) and other disorder (6.0%). The sociodemographic variables of the sample are presented in Table 1.

Table 1.

Sociodemographic variables of the clinical sample (N=400).

Variable	n (%)
	Overall	Aged 14–21 yr	Aged 22–35 yr
Sex
Female	195 (48.8)	97 (50.8)	98 (46.9)
Male	205 (51.2)	94 (49.2)	111 (53.1)
Ethnicity
Chinese	284 (71.0)	143 (74.9)	141 (67.5)
Malay	71 (17.8)	30 (15.7)	41 (19.6)
Indian	34 (8.5)	12 (6.3)	22 (10.5)
Others	11 (2.7)	6 (3.1)	5 (2.4)
Marital status
Single, separated or divorced	365 (91.5)	191 (100)	174 (83.3)
Currently married	35 (8.5)	0 (0)	35 (16.7)
Sexual orientation
Heterosexual	325 (81.2)	146 (76.4)	179 (85.7)
Non-heterosexual	75 (18.8)	45 (23.6)	30 (14.3)
Education
Below tertiary	250 (62.5)	157 (82.2)	93 (44.5)
Tertiary and above	150 (37.5)	34 (17.8)	116 (55.5)
Employment statusa
Working	197 (49.3)	118 (63.8)	79 (38.0)
Not working	196 (49.0)	67 (36.2)	129 (62.0)

^aMissing data for employment status: n=7

Table 2 shows the descriptive statistics of the study variables. The PBI scores were similar between the two age groups. Overall, maternal care and paternal care scores were lower than the cutoffs (maternal = 27.0, paternal = 24.0), while maternal overprotection and paternal overprotection scores were higher than the cutoffs (maternal = 13.5, paternal = 12.5). In terms of maternal bonding type, 76 (20.6%) participants recalled optimal bonding, 47 (12.7%) recalled affectionate constraint, 172 (46.6%) recalled affectionless control and 74 (20.1%) recalled absent or weak bonding. For paternal bonding type, 74 (21.6%) participants recalled optimal bonding, 31 (9.1%) recalled affectionate constraint, 157 (45.9%) recalled affectionless control and 80 (23.4%) recalled absent or weak bonding. Mean scores of PCS and MCS were 50.6±8.56 and 36.6±11.74, respectively. The mean PHQ-8 score fell within the cutoff range of 10–14, corresponding to depressive symptoms of moderate severity.

Table 2.

Descriptive statistics of the study variables.

Variable	Mean±SD
	Overall	Aged 14–21 yr	Aged 22–35 yr
Parental Bonding Instrument (PBI)a
Maternal care	22.2±8.41	22.2±8.79	22.3±8.06
Maternal overprotection	15.4±7.13	15.7±7.31	15.1±6.97
Paternal care	19.3±8.90	19.5±9.00	19.2±8.83
Paternal overprotection	13.3±7.45	13.1±7.15	13.6±7.75
Short Form-12
Physical Component Summary	50.6±8.56	51.0±8.72	50.3±8.42
Mental Component Summary	36.6±11.74	35.1±12.11	38.0±11.24
Patient Healthcare Questionnaire-8	10.5±7.07	11.8±7.09	9.3±6.86

^aMissing data – PBI-maternal: n=31; PBI-paternal: n=58. SD: standard deviation

Results from the multivariable linear regression analyses are found in Table 3. Among the PBI variables, none of the components were significantly associated with PCS, whereas only maternal care was positively associated with MCS (β = 0.32, 95% confidence interval [CI] 0.16 to 0.49). Maternal care was also associated with PHQ-8 (β = -0.12, 95% CI -0.23 to -0.02); as maternal care decreases, the severity of depressive symptoms increases. Age group and sex were not associated with PCS and MCS scores.

Table 3.

Multivariable linear regression for association of Parental Bonding Instrument with HRQOL and PHQ-8.

Independent variable	PCS β (95% CI)	P	MCS β (95% CI)	P	PHQ-8 β (95% CI)	P
Parental Bonding Instrument
Maternal care	0.05 (–0.09 to 0.19)	0.481	0.32 (0.16 to 0.49)	<0.001†	–0.12 (–0.23 to –0.02)	0.023*
Maternal overprotection	0.08 (–0.09 to 0.27)	0.357	–0.07 (–0.27 to 0.13)	0.522	0.04 (–0.08 to 0.16)	0.536
Paternal care	–0.09 (–0.22 to 0.04)	0.170	0.09 (–0.06 to 0.25)	0.236	0.02 (–0.07 to 0.12)	0.668
Paternal overprotection	–0.13 (–0.28 to 0.03)	0.120	0.06 (–0.14 to 0.25)	0.572	0.01 (–0.10 to 0.13)	0.821

Controlled for age, ethnicity, sex, employment status, education level, marital status, type of housing and clinical diagnosis. *P<0.05, ^†P<0.01. CI: confidence interval, HRQOL: health-related quality of life, MCS: mental component summary, PCS: physical component summary, PHQ-8: Patient Healthcare Questionnaire-8

DISCUSSION

This study assessed the relationship between different aspects of parental bonding and HRQOL among clinically diagnosed youths. Overall, like other clinical populations, affectionless control was the most prevalent parental bonding type for both parent figures in this sample.[12,15,19,47] Maternal scores were generally higher than paternal scores, a pattern similar to that of another study conducted in Singapore by Ong et al.[30] However, unlike the study by Ong et al., the present study sample reported overall lower care and higher overprotection scores for both parents. The sample also had moderate severity of depressive symptoms,[30] and their physical and mental health functioning levels resembled those with severe mental illness[48] more than that of the local general population.[49] Our hypothesis was partially supported: among the parental bonding aspects, only maternal care was associated with the mental health functioning component of HRQOL.[27,28]

In a clinical population like our study sample, individuals who experienced more care from their mothers during their early years showed better mental health functioning.[27,28] This finding seems to suggest that the positive effect of early maternal care is not limited to just the general population, but rather, it exerts an overall long-term protective effect for those who were later diagnosed with a mental illness as well. It also hints at the significance of the maternal figure in the child’s early years, which could be explained by the traditional role mothers have in the family unit, where she would spend more time with her child.[20] Naturally, the type of bonding experience between the mother and child would, therefore, exert a greater influence on the child. Although paternal care was not significantly associated with QOL in this study, previous research has shown that the child’s relationship with both parents is nonetheless central to the child’s well-being and is an important predictor of life satisfaction among adolescents.[3,4]

While maternal care seems to be a protective factor, it is also plausible that poor parental bonding contributes to the development of psychiatric symptoms. Parental bonding plays a small yet significant role in psychopathology,[31] having a modest effect of about 1%–5% variance in the presence of mental disorders in adulthood.[20,50] Our study showed that individuals who experienced lower levels of maternal care had worse depressive symptoms.[10,11] Some studies found that autonomy from mothers was lowest among those with anxiety disorders and paternal care was lowest among those with depressive disorders,[13] whereas other studies concluded that overall poor parental bonding is a risk factor for lifetime clinical diagnoses.[51] Regardless, the combination of psychopathology and a lack of parental care could result in even poorer mental health functioning later in life.

The lack of associations between parental overprotection and HRQOL may be attributed to cultural differences in perceiving parental behaviours and attitudes. Parental bonding patterns in the Asian context vary from those of Western countries.[52] For instance, parental overprotective behaviours may be perceived as showing care and concern, rather than as a form of control and impediment of autonomy.[53] As such, instead of a two-factor structure, some studies found that a three-factor structure was more suitable, in which the care factor was retained and the overprotection factor was separated into overprotection and authoritarianism/autonomy as the third factor.[13] A similar factor structure was also found in a local sample of individuals at ultra-high risk for psychosis and controls.[15] The study additionally reported that individuals with higher paternal overprotection showed worse clinical symptoms and poorer social, occupational and psychological functioning. However, no significant associations were found for parental overprotection with mental health functioning of HRQOL in this study.

In this study, parental bonding was not significantly related to the physical health aspect of HRQOL. This was a surprising finding, given that lack of parental warmth or parental overprotection is linked to poorer physical HRQOL.[25] Furthermore, psychiatric patients tend to develop chronic health problems and have poorer physical health.[48,54] A likely reason could be the younger age of the study participants, who would thus have better physical health functioning.

The current study has several limitations. First, causality cannot be determined due to the cross-sectional study design. The use of convenience sampling in a general clinical outpatient setting also limited the generalisability of our findings, as the sample may not be representative of all outpatients or youths with mental illness. Second, PBI is a retrospective measure and only self-reports were used in the study, possibly resulting in self-report and recall biases.[55] Although influences of mood state and psychopathology-related memory biases are plausible as well,[56] evidence has proven otherwise as parental bonding measures were valid and stable even after 20 years.[57,58] Furthermore, the self-report and parent report of bonding experiences were found to be similar between individuals with a psychotic disorder and their parents.[16] Third, parenting does not happen independently in the real-world context; other variables like parental psychopathology and the child’s characteristics such as temperament can influence the parent–child relationship. Future research can, therefore, adopt a bidirectional or transactional approach and study parental bonding and QOL longitudinally. Lastly, it is valuable to note that the study was conducted in an Asian context. Cultural context influences the way parents interact with their child and how the child interprets their parents’ parenting styles, which in turn, shapes the parent–child bond. Therefore, our findings should be interpreted in the context of a multiethnic Southeast Asian setting, whereby parenting styles and practices are strongly influenced by Eastern values, and they should not be directly compared to those of studies done in the West without consideration of such cultural influences.

Findings from this study have practical implications. While a reduction in psychiatric symptoms can improve QOL, psychopathology tends to persist into adulthood. In addition, the effects of early parental bonding on mental health functioning continue across different life stages.[51] Therefore, understanding the relationship between parental bonding and QOL in a clinical population has high clinical importance, as it opens possibilities of utilising parental bonding as a therapy target to improve one’s overall functioning and QOL,[29,47] and this can be especially beneficial for treatment-resistant patients.[59] Early identification of families with non-optimal bonding patterns, especially between the mother and the child, would thus allow intervention programmes to equip parents with knowledge of positive parenting practices that can improve the bonding experiences as well as mitigate any negative effect of poor bonding patterns on later QOL and psychopathology. Ultimately, the parent–child relationship sets the emotional climate for the child’s developmental processes, which is vital to children’s emotional development[30] and well-being.[3,4]

In conclusion, our findings suggest that early maternal care promotes better QOL in later life; individuals report better mental health functioning when parents showed care during their early years despite later clinical diagnosis. Early parenting classes can emphasise the importance of care and encouragement of autonomy to enhance the quality of parental bonding.[3] As for interventions, clinicians and community workers can consider incorporating parental education and family sessions that strengthen positive parenting practices for improving the parent–child relationship.

Financial support and sponsorship

This research was supported by the Singapore Ministry of Health’s National Medical Research Council under the Centre Grant Programme (Grant No.: NMRC/CG/004/2014).

Conflicts of interest

There are no conflicts of interest.