Abstract
Occupational therapists working with families of young children may find it helpful to assess occupational balance (OB) and address OB-related issues to facilitate better family quality of life (FQOL).
Occupational balance (OB) historically developed from a simple idea of balancing work, play, rest, and sleep to promote health, but it later evolved as a multidimensional concept characterizing a satisfactory amount and variety of occupations in daily life (Wagman et al., 2012). Variety and balance are desired among those occupations in which people work with others versus alone; obligatory and voluntary occupations; energy-giving and energy-consuming occupations; restful and strenuous occupations (Wilcock, 2006); physical, social, mental, and rest occupations (Wilcock et al., 1997); occupations performed to satisfy and care for self versus others; and challenging versus relaxing occupations (Stamm et al., 2009). Several studies have highlighted the relationships between OB and overall health and well-being (Günal et al., 2021; Wagman et al., 2020; Yu et al., 2018).
Achieving OB is challenging for parents of children with disabilities (PCWD) because of the high demands of the disability-related caregiving role (McGuire et al., 2004). Luijkx et al. (2017) compared parents of typically developing children (PTDC) with PCWD on time use and reported that PCWD have 1.5 hr less free time per day compared with PTDC. In this context, it would be meaningful to compare OB between the two groups. Previous studies of OB have focused only on the association between OB and individual quality-of-life indicators, but none have explored associations between OB and an indicator of family quality of life (FQOL), which seems important in the context of family and caregiving occupations (Dhas & Wagman, 2020). FQOL is defined as “a dynamic sense of well-being of the family, collectively and subjectively defined and informed by its members in which individual and family-level needs interact” (Zuna et al., 2010, p. 262). Therefore, the primary objective of the study was to compare OB in PCWD and PTDC. The secondary objectives were to explore the relationship between OB and FQOL and to examine the potential of selected personal factors in predicting OB.
Method
A cross-sectional group-comparison design with convenience sampling was used. PCWD were recruited from the outpatient occupational therapy clinics in the Hamad Medical Corporation in Doha, Qatar. Simultaneously, PTDC were recruited from among the nonclinical staff from these hospitals and their family members. All potential participants were informed about the study by the primary investigator (Brightlin Nithis Dhas). Those who agreed to participate provided written informed consent and completed the study questionnaires. All data collection forms were made available in English and Arabic. Data collection was conducted between October 2020 and February 2021. The study was approved by the ethical committees of the Hamad Medical Corporation (MRC No. 01-18-465), where the participants were recruited, and Nova Southeastern University (IRB No. 2019-45).
Participants
Parents with at least one child younger than age 14 yr were recruited with predefined inclusion and exclusion criteria. The inclusion criterion for PCWD was that the parents have one or more child with a disability (CWD) in the family; for PTDC, it was that they have at least one typically developing child and no CWD in the family. Additional inclusion criteria for all participants were that they lived with their children for at least 1 mo at the time of recruitment and were able to read English or Arabic. For PTDC, professionals involved in direct clinical care in the recruiting hospitals were excluded.
Measures
Occupational Balance
OB was the main outcome variable and was measured using the revised 11-item Occupational Balance Questionnaire (OBQ11; Håkansson et al., 2020). The OBQ11 is a self-rated questionnaire designed to assess a person’s perception of having the right amount and variation of occupations in everyday life. Each item in the OBQ11 is rated on a 4-point Likert scale ranging from 0 (strongly disagree) to 3 (strongly agree). The total score ranges from 0 to 33, with higher scores indicating higher OB. The internal construct validity of the OBQ11 was found to be good on the basis of Rasch measurement theory (Håkansson et al., 2020). The original 13-item Occupational Balance Questionnaire was reported to have good internal consistency (Cronbach’s α = .94) and test–retest reliability (Spearman’s ρ = .93; Wagman & Håkansson, 2014). The Arabic version of the OBQ11 was translated for the purpose of this study and was found to have good internal validity (Cronbach’s α = .864) and good content validity on the basis of cognitive debriefing procedures (Dhas et al., 2021).
Family Quality of Life
The Family Quality of Life Survey–2006, short version (FQOL–2006), was used to measure FQOL (Isaacs et al., 2007). The FQOL–2006 has nine domains: Health of the Family, Financial Well-Being, Family Relationships, Support From Other People, Support From Disability-Related Services, Influence of Values, Careers and Preparing for Careers, Leisure and Recreation, and Community Interaction. Each domain is measured by six dimensions: Importance, Opportunities, Initiatives, Attainment, Stability, and Satisfaction. Only Attainment and Satisfaction are considered outcome dimensions that measure, respectively, the degree to which the family is able to accomplish their needs and the family’s overall perception of satisfaction in the aspect of family life pertaining to the domain (Isaacs et al., 2007). These dimensions are measured on a 5-point ordinal scale ranging from lowest (1) to highest (5). The total FQOL score is calculated by adding the Attainment and Satisfaction ratings from the nine domains (Samuel et al., 2016). The total FQOL score for this study was calculated from eight domains because the PTDC did not fill in the Support From Disability-Related Services domain. In addition, the FQOL–2006 has a global FQOL score, which is the average of two items on the overall FQOL measured using a 5-point ordinal scale (Samuel et al., 2016). For the global and total FQOL scores, Cronbach’s αs were .85 and .83, respectively, which was indicative of good internal consistency (Samuel et al., 2016). The FQOL–2006 was translated to Arabic by Neikrug et al. (2011) and has been successfully used in other studies (Neikrug et al., 2014; Roth & Brown, 2017).
Independent Variables
The independent variables were measured using an investigator-developed demographic survey. The nationality of the participants was grouped as the Middle East and North Africa (MENA), South Asia, the Philippines, and others, following the United Nations Children’s Fund’s (n.d.) example of grouping countries under MENA and South Asia. Role satisfaction was measured on a 7-point Likert scale ranging from 1 (not at all) to 7 (very much) on the eight predetermined roles from the Role Checklist (Oakley et al., 1986); namely, student, worker, volunteer, caregiver, home maintainer, friend, family member, and religious participant. To indicate any additional roles, an “Other” option was provided. Percentage scores were calculated for each role rated by the participant (a score of 7 = 100% satisfied, and a score of 1 = 0% satisfied), and a total role satisfaction score was calculated by averaging all individual item scores.
Spousal support was measured with the statement “Please rate the helpfulness of your partner” on a 5-point Likert scale ranging from not at all helpful (1) to extremely helpful (5) (Warfield, 2005). Family support in child care and family support in housework were measured using two questions: “Do other members in the family assist you in caring for your child?” and “Do other members in the family assist you in household chores?” Participants responded on a 3-point response scale on which 1 = no, 2 = to some extent, and 3 = yes (Riyahi et al., 2017). Difficulty finding reliable child care was measured using the statement “It is hard to find a reliable person to take care of my child” on a 5-point Likert scale ranging from extremely hard (1) to very easy (5; Warfield, 2005). Satisfaction with health care services was measured using the question “Overall, how satisfied are you with the quality of care received from health and rehabilitation services in Qatar?” (Liu et al., 2008). Participants responded on a 5-point Likert scale ranging from very dissatisfied (1) to very satisfied (5). The other independent variables were age, gender, educational status, marital status, employment status, family income, number of children, number of children younger than age 5 yr, and availability of paid help.
Sample Size
A preliminary analysis with the first 30 participants (15 PCWD and 15 PTDC) served as a pilot phase to estimate the sample size. It showed a statistically significant difference between the PCWD and PTDC in OBQ11 scores, with an effect size of .76. With this effect size, a sample of 27 would be adequate to assess the primary study objective, with an α of .05 and a power of .80. However, the secondary objectives of the study required a larger sample. Therefore, the sample size requirements for two main modifiable factors, role satisfaction and spousal support, were calculated using the same parameters, and the required samples were found to be 32 and 178, respectively. The higher estimate of 178 (89 per group) was used in the study.
Data Analyses
Statistical analyses were performed with IBM SPSS Statistics (Version 27). Continuous variables were summarized by mean and standard deviation, and categorical variables were summarized by frequency count and percentage. Role satisfaction (M = 82; Mdn = 81) was the only variable with comparatively large missing values, accounting for 12.3% (7% among PTDC and 5% among PCWD), for which mean substitution was done. Missing values on other variables were less than 3% and left unattended.
Using chi-square analyses, we explored group differences in age, gender, education, employment, nationality, income, availability of paid help, number of children, and number of children younger than age 5 yr between PTDC and PCWD before proceeding to the comparison of OBQ11 scores. Groupwise correlation analyses were conducted between the OBQ11 and the global and total FQOL scores for PTDC and PCWD.
Finally, we performed a regression analysis consistent with the steps suggested by Kellar and Kelvin (2013) in the combined sample (PTDC and PCWD participants), with the presence of CWD as a potential predictor. There were 16 other potential predictor variables. Marital status was dropped from the analysis because of the extremely low variance (99% married, 1% divorced). Nationality was dichotomized on the basis of language: Arabic and non-Arabic. Family support in child care and family support in housework were strongly intercorrelated (r = .73); hence, they were combined into one single binary variable called family support. Family support was deemed to be present for those participants who responded “yes” or “to some extent” to at least one of the two questions on family support in child care and family support in housework; otherwise, family support was deemed absent.
In the bivariate analysis with each of the resulting 15 variables, 6 variables (presence of CWD in the family, number of children younger than age 5 yr, role satisfaction, spousal support, family support, and difficulty finding reliable child care) were significantly associated with OBQ11 scores, and another two variables—gender and satisfaction with health care—had borderline significance. Standard multiple regression was performed with these eight variables and language as predictors. Partial regression plots and a plot of studentized residuals against the predicted values were suggestive of linearity. The Durbin-Watson statistic was 1.81, which indicated independence of residuals. Visual inspection of studentized residuals versus an unstandardized predicted values plot confirmed homoscedasticity. There was no multicollinearity, as evidenced by tolerance values greater than 0.1 for all the variables included in the model. A Q–Q plot confirmed the normality assumption. All statistical tests were two-sided and tested at a 5% level of significance.
Results
A total of 178 parents (89 PCWD, 89 PTDC) took part in the study. More mothers (58%) than fathers (42%) participated. The majority of participants were residents (expatriates living in Qatar; 92%), 29% of whom came from other MENA countries; 34%, from South Asia; 17%, from the Philippines; and 12%, from other countries. Only a small number were Qatari citizens (8%). Sixty-seven participants (38%) completed the surveys in Arabic, and the remaining 111 (62%) completed them in English. The mean age of CWD of the PCWD group was 4.5 yr (SD = 2.47), and diagnoses included autism spectrum disorder (44%), developmental delay (16%), cerebral palsy (12%), Down’s syndrome (5%), and other (12%; 12% did not report). Thirty-eight percent of CWD were reported to require assistance in all self-care activities by their parents. Thirty-one percent required assistance for most, but not all, self-care activities; 17% required assistance for only some self-care activities; and only 2% did not require assistance for self-care activities. In terms of communication abilities, 38% reported very little meaningful communication; 28% were able to communicate basic needs and wants; 14% were able to communicate needs, wants, and some ideas in a meaningful way; 6% were able to communicate within a limited range of topics in a meaningful way; and only 2% were able to communicate about a wide variety of topics in a meaningful way (12% did not report).
Differences Between Parents of Typically Developing Children and Parents of Children With Disabilities
Chi-square tests (Table 1) showed that PTDC and PCWD in the sample differed significantly on two factors: family income and the number of children younger than age 5 yr. The results from the independent t test (Table 2) showed that PCWD scored significantly lower on the OBQ11, compared with PTDC. The effect size was medium (Cohen’s d = 0.36).
Participant Characteristics: Parents of Typically Developing Children and Parents of Children With Disabilities
Note. Due to missing values, certain cells (Income in PTDC, PCWD, and Total; No. of children <5 yr old in PCWD and Total; and Availability of paid help in PTDC and Total) do not sum to n values. Percentages are rounded to the nearest value. MENA = Middle East and North Africa; PCWD = parents of children with disabilities; PTDC = parents of typically developing children.
Independent t Test Results for OBQ11 Means
Note. M difference = 1.87; 95% confidence interval [0.331, 3.339]; Cohen’s d = 0.36, p = .02. OBQ11 = revised 11-item Occupational Balance Questionnaire; PCWD = parents of children with disabilities; PTDC = parents of typically developing children.
Correlations Between the 11-Item Occupational Balance Questionnaire and the Family Quality of Life–2006
A statistically significant moderate correlation was found between the OBQ11 and total FQOL–2006 scores (n = 89; r = .52, p = .001), as well as between the OBQ11 and global FQOL–2006 scores (n = 89; r = .57, p = .001) among PCWD. Similarly, statistically significant correlations were found between the OBQ11 and total FQOL–2006 scores (n = 89; r = .48, p = .001) and between the OBQ11 and global FQOL–2006 scores (n = 89; r = .31, p = .003) for PTDC.
Predictors of Occupational Balance
The regression model for predicting OB included nine variables (Table 3). The model was statistically significant and showed good fit to the data, F(9, 160) = 6.46, p < .001, which accounted for approximately 27% of the variance in the OBQ11 scores (R 2 = .27, adjusted R 2 = .23; root-mean-square error = 4.57). Six variables (presence of CWD in the family, number of children younger than age 5 yr, role satisfaction, spousal support, difficulty finding reliable child care, and satisfaction with health care) contributed statistically significantly to the prediction (p < .05). Regression coefficients and standard errors are shown in Table 4.
Bivariate Regression Analysis for Total OBQ11 Scores
Note. N = 178. CWD = children with disabilities; OBQ11 = revised 11-item Occupational Balance Questionnaire; ref. = reference.
Measured in percentage scores.
Measured on a 5-point Likert scale.
p < .05.
Predictors of OBQ11 Scores and Multiple Linear Regression
Note. Reference groups are shown in parentheses. R 2 = 0.267; adjusted R 2 = 0.225. CI = confidence interval; CWD = children with disabilities; NA = not applicable; OBQ11 = revised 11-item Occupational Balance Questionnaire.
Measured in percentage scores.
Measured on a 5-point Likert scale.
p < .05.
Discussion
The study explored the difference in OB between PTDC and PCWD. Although there was a statistically significant difference in their OB, the difference was small, with a modest effect size of .3. In the regression model combining both groups, the presence of CWD in the family emerged as a significant predictor, but its predictive strength (β = 0.165) was not notably different from that of other predictors in the model. In other words, the findings suggest that parenting affects OB regardless of the disability status of the children. The number of children younger than age 5 yr was the strongest predictor of OB (β = 0.21) in the regression model. This indicates that parenting in general—and parenting young children (who require more care and supervision than older children) in particular—is difficult and could affect OB (Håkansson et al., 2019). In addition, FQOL was statistically significantly associated with both PTDC and PCWD. This suggests a need for occupational therapists to work with parents of young children in addressing OB.
A few insights were gained from the regression analysis. First, the independent variables used in the regression model were taken from previous theories or research that proposed them as personal factors influencing OB. The R 2 obtained in the model was small but statistically significant, accounting for only 27% variance in OB, which reinforces the notion that OB is complexly determined (Backman, 2010), and other factors that were not assessed in the present study likely influence the perception of OB.
Second, some potential targets for occupational therapy intervention to promote OB were identified. In the nine-variable regression model, role satisfaction (β = 0.20) and spousal support (β = 0.20) were among the strongest predictors. Occupational therapists can work with parents to select and adapt roles that positively influence OB (Moyers, 2005). Spousal support could be another target of occupational therapy intervention. In keeping with the associations found by Håkansson et al. (2019) between satisfaction with the division of domestic work and high OB in parent couples, intervention focusing on cooperation and mutual support between both parents might improve OB.
Strengths and Limitations
This study contributes to the sparse research evidence with regard to OB and parenting. Moreover, the geographical context of the study adds to the significance of findings, because no OB studies have been reported so far from Qatar or the Gulf region. This study has several limitations. One major issue is that data were collected in two languages, which may be nonequivalent. However, we included language in the regression to check its influence on OB, and we did not find it to be a significant predictor.
Convenience sampling is nonrepresentative and consequently affects the external validity of the findings. Accessing a representative population and random sampling was not feasible, because it would be extremely difficult and time consuming. Replication studies on the same topic in the future could contribute to generalizability.
In predictive studies, the strength and relevance of one predictor could change significantly in the presence or absence of another predictor. Although all possible personal factors associated with OB found in the literature were included, other variables unknown to us could influence associations between variables. The influence of child characteristics such as disability severity and developmental status was not examined in this study. In addition, only longitudinal studies can truly identify predictors; the cross-sectional design used in the study cannot determine the direction of associations.
Western perspectives of OB were used in this study. The understanding and interpretation of OB could be different among families living in Qatar. This requires further exploration.
The timing of the study during the global coronavirus disease 2019 pandemic was beyond the control of the investigators. However, data collection was postponed until November 2020, when the first wave ended in Qatar and life returned to nearly normal, and it was completed by February 2021, well before the end of March 2021, when restrictions because of the second wave began. Although these restrictions were in place for everyone, it is possible that PCWD, being a risk group, may have felt the consequences differently, thus affecting the validity of the findings.
Implications for Occupational Therapy Practice
The findings of this study may have the following implications for occupational therapy practice: ▪ Parenting might affect OB, and it is useful for occupational therapists working with families of young children to assess OB and address OB-related issues, given their association with FQOL. ▪ Detailed assessment of and intervention for role satisfaction and spousal support could be a starting point for addressing OB among parents, regardless of whether they have children with disabilities.
Conclusions
The objectives of the study were to compare OB ratings between PTDC and PCWD, identify possible predictors of OB, and examine the relationship between OB and FQOL. PCWD reported lower OB than PTDC, and OB was statistically significantly associated with FQOL for both PTDC and PCWD. Therefore, OB among parents is an important factor to be considered in pediatric occupational therapy practice.
Footnotes
Acknowledgments
We thank Sultan Salim Hammam Al-Abdalla, Chief of Occupational Therapy, and Alvin Rol Carpio, Pediatric Occupational Therapy Supervisor, at the Hamad Medical Corporation. We also thank all of the participants.