1 Introduction

Parental separation has become a major concern as regards child experiences, health and well-being, especially in low- and middle-income countries (LMICs) where formal child welfare services are not well-developed (Amoah, 2020b; Ansell, 2016; Ruiz-Casares et al., 2017). In the context of this study, parental separation from a child describes absence of one or both parents for a significant period due to forced and voluntary reasons such as death, incarceration, divorce, travel (internal or international migration) (Amoah, 2020a; Murray et al., 2008; Vezzetti, 2016). In some situations, separation of parent(s) can improve the well-being of children, as in the case of removal/absence of an abusive parent (Turney & Lanuza, 2017). However, in most cases, parental separation—either forced or voluntary—can have adverse consequences for child development.

Ghana has the third highest incidence of children living in households with single parents in West Africa (Better Care Network, 2015). According to the 2022 Demographic and Health Survey, only 51% of children lived with both biological parents, while 16% did not live with biological parent(s), and as many as 8% were orphans (Ghana Statistical Service & ICF, 2024). Children separated from their parent(s) in Ghana are at risk of poverty, engaging in economic activity at a younger age, being physically and sexually abused by guardians (usually, informal foster care through extended family members) and being trafficked (Abdullah et al., 2020; Amoah, 2020b; MoGCSP & UNICEF, 2015). This is why the well-being of children without a full complement of their parents has become a priority for the Government of Ghana, as shown in initiatives to enrol more children in school (e.g., School-feeding programme) and address poverty (e.g., the Livelihood Empowerment Against Poverty, LEAP), and health concerns of children (e.g. free insurance premium for children under the National Health Insurance Scheme) (MoGCSP & UNICEF, 2015). Despite numerous formal and informal efforts to protect children without one or both parents, significant knowledge gaps remain in how such situation affects their health-related well-being (Bohman et al., 2017; Raley & Sweeney, 2020).

While the type or cause of separation from a parent(s) can have unique impact on children (Cebotari et al., 2018; Murray et al., 2008), short- to medium-term consequences of being separated from a parent such as low school participation and poor academic performance, child poverty, stigmatisation, behavioural issues due to maladjustment, and social isolation are consistent across places (Amato & DeBoer, 2001; Bianchi, 1994). Absence of a parent is also a risk factor for child neglect—a situation where a child has inadequate supervision and basic needs (e.g., clothing, housing and nutrition) (Ruiz-Casares et al., 2012). These short- and medium-term consequences can trigger unwelcome long-term outcomes such as low educational attainment and poor health outcomes among the children. Nevertheless, the extent of effects of separation from a parent(s) on a child varies depending on the reason for separation, the duration of separation, the timing of separation, the existing conditions before the separation and the prevailing child welfare policies, although in-depth study of these variables is not well-documented (Turney, 2023). For instance, a longitudinal study among children in transnational families in Ghana and Nigeria found that while the absence of a parent(s) does not necessarily have a negative influence on a child’s well-being, although being female or living in a divorced home increases the risk of ill-being (Cebotari et al., 2018). To better understand the dynamics of the associations among these short- and long-term consequences, a natural step is to research the situation of children without one or both parents and support such families to manage short- and long-term implications of their situation (e.g., educational attainment and health outcomes). Research of this nature is also a means to avert potential transgenerational transmission of voluntary child-parent separation and its ill consequences, which are common among people who have experienced separation from their parent(s) (Amato & DeBoer, 2001).

Extant evidence suggests that the effects of single-parent homes, such as monetary poverty, low school participation, social isolation, low social support, and stigmatisation, can adversely affect the health outcomes of children (Amato, 2010; Raley & Sweeney, 2020). However, such adverse effects are often worse among children with single parents (Amato & Anthony, 2014). The influence of these factors on health-related outcomes among children separated from parent(s) children can be explained by the fundamental cause theory of health disparities (Phelan et al., 2010). The fundamental cause theory argues that “individual risk factors must be contextualised, by examining what put people at risk of risks…[since] social factors such as socioeconomic status and social support are likely ‘fundamental causes’ of diseases because they embody access to important resources, affect multiple disease outcomes through multiple mechanisms, and consequently maintain an association with disease even when intervening mechanisms change” (Link & Phelan, 1995, p. 80). This paper aims to extend this theory and the debates on health and well-being in the context of child-parent separation in Ghana by examining the associations of some ‘fundamental causes’ of health and their interrelations with the health-related well-being over time.

1.1 Fundamental Causes of Health-Related Outcomes of Children Separated from a Parent Across Time

The intensity of the effects of parental separation is time-sensitive (Turney, 2023). Children are the most affected during and immediately after parental separation (Darbeda et al., 2018; Wen et al., 2021). However, long-term separation can be devastating to the health of children and even persist in adulthood (Bohman et al., 2017; Xu et al., 2018). Even prolonged separation due to voluntary reasons (e.g. migration) disrupts parent–child relations and occasions psychosocial difficulties (Zhao et al., 2018), generating anxiety and depression among children (Fan et al., 2010). Others argue that long separation rather enables children to adjust better to their situation, which averts health-related difficulties (Dreby, 2007).

While extant evidence is divergent, an important question is the consistency of factors associated with health-related well-being of children separated from a parent in the short- to medium- and long-term and the mechanism that links various determinants to the health-related well-being. Therefore, this study will explore how some fundamental causes of health shape the health outcomes of children (see Fig. 1). The specific objectives of the study are:

  1. 1.

    To examine the associations of health literacy, access to money for children (conceived as experience of poverty), family support, and experience of social isolation with health-related well-being of children separated from a parent and the influence of duration of separation.

  2. 2.

    To investigate the extent to which health literacy explains the associations of experience of poverty, family support, and experience of social isolation with health-related well-being of children separated from a parent.

  3. 3.

    To examine the extent to which an explanatory role of health literacy in the association of experience of social isolation, family support, and experience of poverty with health-related well-being differ by duration of parental separation.

Fig. 1
figure 1

An analytical framework

Full size image

These fundamental determinants of health are selected for this study as they provide individual (e.g., health literacy skills) and contextual (poverty, family support, and isolation) insights into determinants of health outcomes of children without parent(s). The study is also an opportunity to combine the influence of contemporary (i.e., health literacy) and conventional (i.e., experience of poverty, family support) causes of health-related well-being among these children in the context of Ghana.

Household poverty affects children in single-parent households in two significant ways. First, it can be the cause of separation in the form of divorce or migration of a parent to seek employment, which can aid in lowering poverty levels as is often the intention (Amoah, 2020b; Zhao et al., 2018). While parental migration for work can improve the living standards of children, evidence in rural China, Ghana, Nigeria and Angola shows that such separation can also be detrimental to the psychological well-being and poor behaviour of the children (Mazzucato et al., 2015; Murphy et al., 2016; Zhao et al., 2018). Second, monetary poverty caused by separation from a parent, especially if the breadwinner is the one that leaves (e.g. through divorce or death) can lead to financial challenges (Andreß et al., 2006). For children in Kumasi in such a situation, their economic conditions are likely to be extenuating due to rising monetary poverty in the city (Cobbinah et al., 2020; Poku-Boansi et al., 2020). Historically, a significant proportion of residents in the study area, Kumasi, engage in unstable small-scale informal economic activities that keep households in monetary poverty (or at high risk) (Cobbinah et al., 2020). With the expansion of the city, the enclaves of poverty have been expanding due to lack of economic opportunities and critical infrastructure (Poku-Boansi et al., 2020). Increased financial challenges are linked to health problems as it makes it difficult to access essential and quality health services, nutritious food and safe environment (Davidson, 2019; Marmot et al., 2001). Among children separated from a parent, poverty can also evolve in the form of medical neglect and poor health outcomes, especially those living with lone migrant mothers (Murphy et al., 2016; Wen et al., 2021). Moreover, based on the theory of relative deprivation, children separated from a parent may experience poor health outcomes as a result of consistent comparison with their peers in better socioeconomic conditions causing unhappiness and distress, especially when the separation is forced one (e.g. death) (Walker et al., 2002).

For many children separated from a parent, their short- and long-term outcomes are influenced by social support—the emotional and practical support (e.g., money, and information)—accessible to them. Social and related family support is an important fabric of people in Kumasi and its environs and has critical influence on their health and well-being (Amoah, 2018; Amoah & Adjei, 2023; Manful & Cudjoe, 2018). Having appropriate and adequate forms of social support promote better mental and physical health, lower rates of chronic diseases, and faster recovery times from illness or injury (Løseth et al., 2022). Access to social support (including family support) can mitigate against adverse consequences of parental separation such as monetary poverty and offer support in seeking healthcare and improving health knowledge (Berkman & Krishna, 2014; Kawachi & Berkman, 2014). However, children separated from parents because of divorce or family breakdown tend to have low social support (Condry, 2012) just as in the case of those whose parent(s) migrate, which have negative impact on mental well-being (Zhao et al., 2017). Children who lose parent(s) tend to have low social support and seek more support from their social networks in Ghana (Salifu Yendork & Somhlaba, 2015).

Moreover, parental separation can impact on social participation of a child which an manifest in the form of social isolation and loneliness (Zhao et al., 2018). For instance, parental separation as a result of divorce or incarceration may cause stigmatisation from their peers and others, leading social isolation and consequently poor health outcomes (Condry, 2012). Children whose parent(s) have migrated also are likely to feel lonely and have a high risk of social anxiety (Chai et al., 2019; Murphy et al., 2016). In times of crisis, social isolation contributes to poor mental health of children (Loades et al., 2020). Evidence suggests that the loss of a parent has a long-term impact in the form of social isolation and loneliness as a result of low social support, which culminates into poor health-related well-being in the long term (Ellis et al., 2013; Salifu Yendork & Somhlaba, 2015).

The influence of family support, social isolation and poverty on health-related outcomes can be partly explained by another fundamental social determinant of health: health literacy (Amoah et al., 2022a; Paasche-Orlow & Wolf, 2007). Health literacy encompasses the ability of individuals to acquire, analyse, and comprehend basic health-related information to make informed health decisions (Nutbeam, 2008). Individuals with low health literacy often lack knowledge about their health conditions and needs, are less proactive about preventive diseases, often rely on others for health information, communicate poorly with health providers and ultimately have poor health outcomes (Berkman et al., 2011; Sørensen et al., 2015). Health literacy equips people to mitigate adverse effects of other social determinants of health, such as stigmatisation, low social support and low socioeconomic conditions (Cui et al., 2021; Nutbeam, 2008). Health literacy enables children in difficult circumstances (e.g. street-involved children) to maintain good health outcomes (Amoah et al., 2017). Unfortunately, children in difficult situations, as is so often the case of children in single-parent households in LMICs, tend to have low health literacy, as observed among children in Kumasi (Amoah et al., 2017, 2018). Low socioeconomic conditions (e.g. poverty levels and social support) are linked to health literacy (Amoah et al., 2021; Svendsen et al., 2020). People with low socioeconomic status often have low health literacy, which informs deleterious health behaviours (Stewart et al., 2014). Adequate social support, such as that from family, improves health literacy, which is known to impact health outcomes positively among young people (Amoah, 2019). Also, social isolation limits people's ability and even willingness to access and utilise health information, receives poorer social support for health, and has lower appraisal of health information among young people and communities (Agho et al., 2011; Vasan et al., 2023).

Thus, health literacy is a plausible explanation for how other fundamental determinants of health, such as poverty, family support and social isolation, influence health-related well-being. Indeed, the model of causal pathway linking health literacy to health positions health literacy as a mediator between socioeconomic conditions (e.g. social status and income levels) and health outcomes (Paasche-Orlow & Wolf, 2007). A multi-country study has also offered evidence of health literacy as not only a mediator of other determinants of health (e.g. age, gender, education, social status and financial status) (Pelikan et al., 2018). However, empirical evidence in support of this claim, particularly those studies that focus on children without parent(s) is in paucity. It is hypothesised that experience of poverty (i.e. access to money for children) and social isolation will negatively affect health-related well-being as opposed to health literacy and social support, which will be positively associated with their health-related well-being. Additionally, health literacy will mediate the association of experience with poverty, social isolation, and family support with health-related well-being.

2 Methodology

2.1 Study Design

This article emerged from a larger study which explored how separation from parents affected the well-being of children in Ghana. The data for this article is based on a cross-sectional data gathered from October 2018 to December 2018 in the Ashanti region of Ghana as part of the larger study. Ashanti region is one of the top three regions with children living in single-parent households due to parental separation, divorce and death (Better Care Network, 2015). Kumasi is the capital city of Ashanti region and the largest metropolitan area in the region (Cobbinah et al., 2020). According to the latest census data prior to this data collection, there were approximately 211,336 single-parent households in the Kumasi Metropolitan Area (KMA) prior to carving six municipal areas from KMA (Cobbinah et al., 2020; Ghana Statistical Service, 2014). With an average household size of 3.8, it could be estimated that there were about 422,672 children in single-parent households (i.e., an average of two children per household). However, data on specific areas with such children are unavailable. Hence, the data for this study was gathered from three municipal areas, which were previously part of the Kumasi Metropolitan Area. The three municipal areas included Tafo, Asawase and Nhyiaeso. The research team consulted relevant staff at the Kumasi Metropolitan Area, who recommended the three municipalities and specific areas within them, where children without one or both parents are predominant based on their experience working with communities and schools in the Kumasi area. These areas were also selected because of variations in social amenities and standard of living (KMA, 2013) to have a balanced sample. Within these areas, children in private and public schools with children from single-parent homes, as advised by the municipal education officers, were recruited to participate in the study. The study obtained ethical approval from the Committee on Human Research, Publication and Ethics of the School of Medical Sciences/Komfo Anokye Teaching Hospital, Kwame Nkrumah University of Science and Technology, Ghana (Ref. CHRPE/AP/495/18). All children also provided verbal consent to participate in the study.

2.2 Sampling and Data Collection

Selection of participants comprised a combination of purposive sampling and convenience sampling techniques to ensure that adequate and relevant sample were recruited. The criteria for selecting respondents included being aged 8 to 17 and having been separated from one parent for at least three months. A minimum of 3 months of separation from one parent was required to include a child in the study to ensure that the separation had been long enough to affect a child’s life. The convenience sampling entailed recruitment of respondents from schools. In the identified schools, children who met the inclusion criteria were given consent forms to be endorsed by their parents/guardians. Children whose parent/guardian endorsed the form were included in the survey. As the research team worked with teachers, the refusal/non-response rate to our invitation was about 4%. The teachers and the research team assisted the children to complete the survey. In total, 418 children were surveyed in the broader study (aged 8 to 17 years), of which 337 who were aged 11–17 have been included in the present article. This study used the age range of 11–17 to ensure that the children could at least read at 6th grade level, which the level often required for health information and messages to be prepared and conveyed were included in the analyses (Nielsen-Bohlman et al., 2004). The data collected entailed four parts covering demographics, financial status, social relationships and family support, health literacy, well-being (happiness) and health status.

2.3 Measures

All the variables were measured based on subjective assessment of participants.

Dependent Variable

The health-related well-being of the children was measured using a one-item scale. They were asked to respond to this question: “In general, would you say your health is excellent, very good, good, fair, or poor in the past one month (four weeks)?”. This scale has been used across contexts and among different population groups and has shown to be reliable in measuring the health status of people, including children (Amoah et al., 2017; Moriarty et al., 2003; Woan et al., 2013).

Independent Variables

The three main independent variables were derived from a self-developed 25-item scale to measure socioeconomic and academic condition of children separated from a parent. The instrument was developed based on qualitative research as part of the larger study. The qualitative research explored the experiences and life circumstances of the children during the period of separation from a parent as part of the broader study. The items below emerged from the broader scale:

  1. i.

    Access to money for children (conceived as experience of poverty) was measured as: This was measured using a one-item scale. The children responded to this statement: “The absence of my parent is affecting my ability to get money to buy what I want (e.g., toffee, biscuit)”. This approach allowed for the children to estimate their monetary poverty based on their experience and views as Lau et al. (2019) advocates.

  2. ii.

    Family support (a form of social support) was measured by exploring their perception of support (especially emotional support) that they received from other family members using a single-item scale. They answered to this statement: “I have received love and care from other family members after the separation of my parent”.

  3. iii.

    Experience of social isolation: This aspect assessed the extent to which the social lives of the children were affected by absence of one or more of their parents. They responded to this statement: “Other people avoid me because of the absence of my parent”.

For all three statements above, the children responded using this Likert scale from “1 = Strongly disagree” to “5 = Strongly agree”. The single-item scales were preferred as they were easier for the children to understand and answer.

2.4 Mediating Variable: Health Literacy

The study used the Swedish Functional Health Literacy Scale to measure the health literacy (Wångdahl & Mårtensson, 2015). It comprises five items with five options for each item, beginning with “1 = Never”, “2 = Seldom”, “3 = Sometimes”, “4 = Often” and “5 = Always”. Respondents answered items such as “Do you think that it takes a long time to read health information?”, “Do you ever ask someone else to read or explain health information?” and “Do you think that it is difficult to understand words or numbers in health information?”. This scale was preferred because of its adaptability for use among different population groups, including asylum seekers (Wangdahl et al., 2015), children and youth (Amoah et al., 2018), and older adults (Amoah, 2019). The instrument has adequate reliability (Cronbach Alpha = 0.75). The responses were summed to have a score for health literacy of the children with a minimum of 5 (indicating low health literacy) and a maximum of 25 (indicating sufficient health literacy).

Covariates

Data on child age, sex (male/female), reason for parental separation (travelled, death, incarceration, divorce), education institution attended (public or private), sex of available parent (male/female), employment status of available parent (employed, unemployed) and place of residence (urban/rural) were collected.

2.5 Data Analysis

The data analysis entailed three parts. The first part was descriptive statistics to gauge the characteristics and patterns of responses across different variables measured. Based on the distribution of the responses, the data was split into two according to the length of separation from a parent for a comparative and in-depth understanding of their situation. The two categories were children separated from a parent for five years or less and those over five years (6 years +). The first part also included a correlation analysis (see Appendix I, Table 6) to understand the relationship between the variables. Most variables with significant correlations only showed a weak relationship, showing they measured different phenomena. The variable “School grade” of children was excluded from associational analyses due to a strong correlation with the age of respondents to avoid multicollinearity.

The second part of the analysis involved a binary logistic analysis in SPSS, which was conducted to explore the association of experience of poverty, health literacy, experience of social isolation and family support with health-related well-being of the children between those separated five years or less and more than five years. Binary Logistic regression was employed to ensure that each category of the dependent variable had enough cases for a robust analysis. Hence, responses to “poor and fair” were categorised as 0—“Poor health-related well-being” and “good, very good and excellent” were categorised as 1—“Good health-related well-being” the binary logistic analysis was done separately from a subsequent mediation analysis to ensure that the results were independent of the mediation analysis. Omnibus test (x2 = 53.22, p < 0.000) and Homer and Lemeshow tests (x2 = 9.92, p < 0.271) showed the overall model was a good fit for the data. The model also showed high sensitivity to the dependent variable by explaining 93.9% of the variance. Using the recommended formula for determining overfitted models; 50 + 8n (n = the number of predictors in a model), the minimum sample required for the analysis was 138 for each of the two categories of children (based on duration of separation). Tables 1 shows 197 and 140 responses for those separated for five years or less and more than five years, respectively, indicating sample adequacy for the analysis (Tabachnick & Fidell, 1996). Two further analyses were conducted to assess the robustness of the binary logistic regression analysis results. The first was a binary logistic regression analysis involving only the four primary independent variables (Appendix I, Table 7). The second analysis involved an ordinary least square regression technique (OLS) to analyse the data (Appendix I, Table 8). Postestimation tests of the OLS including Ramsey test statistic (F = 0.51, p = 0.675), and Link test (B = 0.255, p = 0.788) showed that model specification was correct, and no variable was omitted. Results of both analyses were largely like that of the logistic regression, demonstrating the robustness of the main findings. Moreover, due to potential bi-directional causality between the independent variables and dependent variable is possible, we conducted instrumental variable analysis using the approach suggested by Lewbel (2012). Details of this analysis and instruments can be found in Appendix IB, Appendix I (Table 9) and Appendix I (Table 10). The aim was to decide whether to adopt the binary logistic analysis and our follow-up ordinary least square regressions. Based on our results across the three kinds of analysis (i.e. binary logistics, OLS and the instrumental variable analysis) and with support of theoretical perspectives on the association between the independent variables and dependent variables as explained in Appendix IB, we opt to use the results of our binary logistic regression and the robustness check using the OLS instead of the instrumental variable results. Given the significance of our analysis to expanding understanding of social determinants of health-related well-being of children with separated parent(s), particularly in a context like Ghana where little studies have taken place, it is prudent to maintain our analysis to support future studies and policies aimed at promoting the well-being of such children (see McCloskey & Ziliak, 1996).

Table 1 Characteristics of variables in the study
Full size table

The third part of the analysis involved a mediation analysis, which employed structural equation modelling (SEM) technique based on Fig. 1. After the mediation analysis, the significant paths (ad, bd, and cd) in Fig. 1 were tested for their differences relative to two specified durations of separation from a parent. The study used the moderated mediation software offered by Gaskin (2021) to explore how the mediational role of health literacy differed according to the two durations. The SEM was conducted using SPSS AMOS with a bootstrap set at 5000. The model showed good-fit with CMIN/df = 1.973 (p-value = 0.160), GFI = 0.999, AGFI = 0.924, CFI = 0.995, IFI = 0.997, and RMSEA = 0.050 (Byrne, 2013). Based on latest recommendation (Kenny, 2024), a the Bollen-Stine bootstrap test was conducted and it showed that the model was a correct and good fit of the data (p = 0.130).

3 Results

As shown in Table 1, most respondents were female (53%) and mainly were aged 13.5 (Standard Deviation, SD = 1.7) years. Most of them had been without at least one parent for five years or less (58.5%). The available parent were often their mothers (83.1%). Most of the children were separated from a parent due to divorce (40.9%) or migration (34.4%) but others also had parents who were incarcerated (6.5%) or death (18.1%). Almost half (48.1%) of them had experienced monetary poverty due to absence of a parent. Fortunately, most of the children (62.3%) had received various forms of family support from other family members and friends. A significant proportion (18.1%) of the children sometimes felt socially isolated because of the absence of a parent. The majority (31.2%) of the children had been separated from one parent between two and five years, while many others (30%) had also been separated between 6 and 9 years. Most (94.9%) of the available parents were employed. On average, children whose parents had left for five years or less had slightly better health-related well-being and health literacy (albeit insignificant difference) than those whose parents had left for more than five years. Those who had separated from a parent for relatively shorter duration had better health-related outcomes than those separated for longer period.

Table 2 shows that health literacy (Odds Ratio, OR, = 1.086, p < 0.05) and social isolation (OR, = 0.584, p < 0.05) were positively and negatively associated with the health-related well-being of children who had been separated from a parent for five year or less respectively. family support (OR = 1.592, p < 0.05) was positively associated with the health-related well-being of children separated from a parent for more than five years respectively. All the four independent variables showed significant association with health-related well-being as was also observed in the robustness check (Appendix I, Table 8). However, the duration of separation was not associated with the health-related well-being of the children according to the overall model (Table 2). These results were assessed for sensitivity and robustness.

Table 2 Factors associated with health-related well-being of Children separated from a parent by Binary Logistic Regression
Full size table

Further analysis (see Fig. 2) through SEM showed that the association of experience of poverty (β = -0.012, p < 0.05) and experience of social isolation (β = -0.010, p < 0.10) were associated with health-related well-being through their health literacy, as shown in Tables 3, 4. Table 3 provides information on the direct and total effects of the SEM.

Fig. 2
figure 2

Mediational role of experience of poverty, family support and social connectedness with health-related status of children separated from a parent by SEM. Notes: Coefficients are based on standardised estimates. **p < 0.01; *p < 0.05

Full size image
Table 3 Direct, indirect, and total effects of mediation paths examined in this study
Full size table
Table 4 Indirect effects of experience of poverty, Family support and social connectedness on health-related well-being of children separated from parents(s) through health literacy
Full size table

As shown in Table 5, the significant explanatory role of health literacy in the associations of experience with poverty and experience with social isolation with health-related well-being were not different for children separated from a parent five years or less or those above five years.

Table 5 Differences in duration of separation from a parent in the observed indirect effects of health literacy in associations of experiences of poverty and isolation on health-related well-being of older persons
Full size table

4 Discussion

This study aimed to provide nuanced perspectives into conventional and contemporary fundamental social determinants of health-related well-being of children separated from their parents in Ghana, where few studies have been conducted.

The first notable finding was that many of the children did not have both parents because of divorce and migration. This finding iterates recent developments that divorce rates in Ghana and many other countries in the region are on the rise and can have implications for children’s health-related well-being (Amato & Anthony, 2014; Vezzetti, 2016). Additionally, the high rate of separation due to migration (either internal or international) can be predominantly connected to the economic pursuits of parents, especially male parents, as appears to be the case in this study, partly due to the patriarchal nature of family systems in Ghana (Cebotari et al., 2018). The phenomenon of leaving children behind is common in situations of economic difficulty, as has variously been observed in rural China (Fan et al., 2010; Xu et al., 2018; Zhao et al., 2018). Given the likely adverse influence of parent–child separation on child well-being, mechanisms to ensure that the partner that leaves continues to offer emotional and instrumental support to such children are not strongly enforced (Issahaku, 2019). The findings also showed that the duration of separation was not directly associated with the health-related well-being of children. This could be due to the timing of the separation as opposed to the duration itself. Turney (2023) has observed a significant impact (e.g., strength of relationship with separated parent) on children separated from a parent at a young age compared to older children. Notwithstanding, children whose parents had left for five years or less in this study had slightly better health-related well-being and health literacy (although not significant) than those whose parents had left for more than five years. Hence, longer period away from a parent due to separation (for whatever reason) can adversely affect the health and well-being of children involved (Fan et al., 2010; Vezzetti, 2016). Such children are more likely to engage in deleterious health behaviours (e.g., smoking) and have low self-esteem and locus of control, all of which negatively affect health-related outcomes compared to children with both parents (Amato & Anthony, 2014). Further mechanisms of how separation from a parent affects health-related well-being are discussed below. Consistent with the hypothesis and the fundamental cause theory, all four independent variables of interest to this study were associated with health-related well-being of the children.

Among children who had been separated for five years or less from a parent, health literacy positively predicted their health-related well-being. This finding is consistent with several others that show that health literacy is critical to the health outcomes of different groups of young people (Amoah et al., 2017, 2018) and general populations (Berkman et al., 2011). People with sufficient health literacy have the skills to prevent, promote and protect themselves from poor health outcomes (Sorensen et al., 2012). This finding shows that in the short- to medium-term after separation from a parent (i.e., within the first five years as specified in this study), health literacy can be crucial to the health and well-being of children, but its effects may not be entirely critical in the longer term. Nonetheless, as Ghana’s child and family welfare policies aim to empower children and young people to be “agents of change in their capacity” (MoGCSP & UNICEF, 2015, p. 14), health literacy can be one of the avenues to promote the well-being of children separated from a parent. Health literacy is a form of empowerment as it enables individuals to take charge of their health (Nutbeam, 2008). Future child welfare policies in Ghana and places alike can thus incorporate tenets of health literacy into efforts to empower children at risk of poor health. Such research can investigate the viability of incorporating health literacy into education curriculum, for example, as part of existing relevant subjects on health education in especially public schools where most children with single parents (e.g., see Table 1) obtain their education. This proposal is consistent with those in places such as Germany (Schulenkorf et al., 2021).

Children tend to suffer social isolation in the short term after separation from a parent. Being isolated or unable to make social connections tends to cause stress. According to the psychosocial hypothesis of health and illness, differences in social status (e.g. considering oneself as lower on a social pedestal than others due to the absence of a parent) can induce stress and anxiety that can cause mental and physical health problems (Davidson, 2019; Marmot et al., 2001). This is likely the connection between the experience of social isolation and health-related well-being these children encounter in the first few years of separation from a parent. However, a more concrete mechanism linking the experience of social isolation to health-related outcomes of children separated from a parent is through health literacy, according to this study. One could argue that having sufficient health literacy improves interpersonal relationships as one key component of the phenomenon involves effective interaction and communication with others on matters about health and, in so doing, limits the adverse effects of isolation on health (Amoah, 2018; Cui et al., 2021; World Health Organization, 2015). This explains why health literacy is sometimes the social resources required for individuals and communities to gain access, comprehend, evaluate, utilise and communicate about information and services to make health-related decisions in context of low and middle income countries (World Health Organization, 2015).

Furthermore, experience of social isolation, even if it is mere perception, can limit critical social connections that could be source of important material support and caution against health risks (Agho et al., 2011; Marmot et al., 2001). This explanation is supported by other findings in this study, which suggests that for children separated from a parent for more than five years, having adequate family support was instrumental to good health-related outcomes, as also documented in numerous studies (Amoah, 2020b; Wu et al., 2014). In the long run, adequate family support could allow some children to overcome social isolation, poverty (expanded below) and a sense of relative deprivation, all of which adversely affect health and well-being (Davidson, 2019; Su et al., 2017). For the children in this study, it is plausible that family support positively affected health-related well-being by serving as source of resources (e.g., financial support) to enable them to meet basic needs and access needed healthcare (Amoah & Nyamekye, 2022) and chart positive life outlook (Luther, 2015) in view of their significant experience with poverty (i.e., access to money).

Indeed, it was evident that the experience of poverty among the children negatively influenced their health-related well-being after five years. These findings echo the neo-materialist hypothesis that factors such as economic and social resources a person possesses individually and through their social networks (e.g., remaining parent and other family) do have an effect on their health and well-being (Davidson, 2019). Without these resources, children with one parent who are economically unstable—a phenomenon common among women after partner separation (Andreß et al., 2006), are likely to have poor health outcomes in the long term due to economic poverty. Many children in this study lived with their mothers, and hence, the propensity of facing the vagaries of monetary poverty on health was inevitable for most of them. Sadly, the current child and family welfare policy in Ghana appears to focus more on facilitating the roles of other institutions (including informal ones such as communities and families) and connecting families to broader welfare programmes instead of tailor-made and evidenced-based programmes and strategies to support children at risk of distress (Issahaku, 2019; MoGCSP & UNICEF, 2015). Given this finding, more concrete policies must be designed to identify and support children at risk of poverty-induced poor health, such as those separated from a parent. The above findings and discussions strengthen claims that factors such as poverty, health literacy, family support and connections are fundamental causes of health as they shape outcomes through multiple mechanisms (Phelan et al., 2010). This study adds that, like many other social determinants of health, they interact in their effects on health, and this is also the case among children with a separated parent.

4.1 Limitations

The study adds to the discourse about the health and well-being of children separated from a parent. Despite its contribution to the literature, the findings should be interpreted with some limitations in mind. First, the study is based on a self-reported cross-sectional survey using primarily single-item instruments to measure most variables. The full depth of the concepts measured may not have been captured. Additionally, there is a possible bi-directional relationship between the dependent and independent variables (e.g. health literacy and health outcomes (Amoah et al., 2022b)). Therefore, the study does not assume a uni-directional causal effect, although the robustness checks and tests generally show consistency in the relationships explored in this study. Our observations are instructive for policy and practice on issues relating to the well-being of children in complicated family situations. Secondly, the study relied on non-probability sampling techniques, which put the generalisability of the findings into question. Nonetheless, the process entailed meticulous efforts to recruit qualified participants for the study. Third, the perspectives and experiences of children with both parents, which could have provided a sharper perspective on the situation of those without both parents, were not considered. Fourth, potential, influential variables such as the health insurance coverage for the children were not controlled in the analyses. However, other socioeconomic variables, such as employment status of available parents, were included in the analyses, which partly offers a window into their ability to afford healthcare. Another limitation could be that traditional healers are reported to have a substantial impact on families’ health-seeking behaviours, as 70% of healthcare services are provided by traditional healers (Amoah et al., 2014), which further studies could examine. Despite these limitations in the data and given the limited research on this group of people in Ghana and the selected region, the results provide important insight into the lives of children without all parents physically in their lives.

5 Conclusion

Children separated from a parent, who so often live in single-parent homes, are at significant risk of poor health and well-being. Grounded in the fundamental cause theory, this study sought to provide empirical analysis of factors directly and indirectly associated with health-related well-being of children without one parent in Ghana. The findings lend support to existing evidence that factors such as experience of poverty, experience of social isolation, social support (i.e., family support) and health literacy were critical to the health-related well-being of these children. Moreover, experience of poverty (specifically, access to money) and social isolation indirectly influenced their health-related well-being through health literacy.

Amidst calls for countries in West Africa to empower families to strengthen their care responsibilities and children to exercise their rights (Issahaku, 2019), this study offers a window into potential areas where families and children can be supported to promote and protect the health and well-being of children who has been separated from a parent. From this study, strengthening mechanism for poverty alleviation of children at risk of poor health, creating more opportunities for children and families to diversify their social networks and strengthen current policies aimed at empowering informal and community-based institutions (e.g., families and traditional groups and norms) to support children in potentially tricky situations better. Additionally, policies and programmes specific to the needs of children at risk of poor health due to poverty, low knowledge about health and social connections in Ghana must be designed instead of relying on ad hoc and generic welfare programmes as is the case of the current Child and Family Welfare Policy (Issahaku, 2019; MoGCSP & UNICEF, 2015). These mechanisms will also allow for a sustainable conduit to enhance the health literacy of children and their families in a bid to promote their health. For instance, more robust social and informal networks (e.g., extended family members, religious leaders and, community leaders and neighbours) can serve as avenues to circulate proper health information to these children and their families.