Introduction

Sub-Saharan Africa (SSA) is home to approximately 53.6 million orphaned children, a majority of whom live in extreme poverty [1], often with relatives or guardians of limited means, and in households with many other dependent children [2, 3]. Grandparents are commonly their primary caretakers [4]. Children who live in poor households often do not have access to basic necessities such as shelter, food, clean water, health care, and education [2, 5–7]). In Kenya, there are over 20 million children living in impoverished families that survive on less than 1 US dollar per day. To achieve the goals of the United Nations Children’s Fund (UNICEF) related to children’s rights and development, and the targets of the United Nation’s Sustainable Development Goal 3 to ensure healthy lives and promote well-being for all people at all ages, appropriate socioeconomic policies, services and interventions are required to support poor and vulnerable children in Kenya and across Africa.

Poverty constitutes an important risk factor for adverse mental health outcomes among children, and those who live in the poorest households are more likely to experience negative psychological outcomes [8–10]. Poverty and mental health are inextricably linked (Evans & Cassells, 2014), and poverty in earlier years of life particularly affects cognitive and psychosocial development [11]. To address poverty and its adverse child health and developmental consequences, economic incentives are increasingly used in African countries. A number of studies have shown that cash transfers (CTs), which involve the provision of cash assistance with the objective of increasing a household’s real income [12], can reduce poverty and improve health outcomes of orphaned children [13–15]. CT studies in African countries suggest that economic incentives play an important role in influencing adolescents' sexual risk-taking intentions [16], and that savings-related interventions can improve self-rated health of orphaned adolescents in impoverished settings [17]. Self-rated health is an important correlate of actual future health [18]. Recent data from Malawi demonstrated that unconditional CTs to poor households improved health outcomes for vulnerable children aged 6–17 years [13]. In Kenya, a reduction in transactional sex and increased school enrolment were observed among women who received CTs [14]. These studies support the hypothesis that economic incentives such as CTs may be effective for improving the health outcomes of orphaned and vulnerable children.

The Kenyan government has expanded unconditional CTs to households caring for orphaned and vulnerable children, increasing the number of such households receiving cash assistance from 7,500 in 2007 to 85,891 by December 2010. As of January 2014, the number of households receiving CTs had reached 170, 000 [19]. The program benefits more than 1,800 households in Uasin Gishu county in western Kenya, where the CT program began in 2007. Enrolled households receive a cash payment of 1,500 Kenya shillings (approximately $20 USD) per month, which is equivalent to about 20% of average monthly household expenditure in the region. Given the rapid expansions in the program, it is critical to evaluate the impact of unconditional government CTs on health and wellbeing in this geographic and epidemiologic context.

Although there is evidence that providing financial help may be an effective way to support orphaned children in resource-limited settings [13, 15, 20, 21], few studies have examined the relationship between direct financial support and psychological wellbeing among OVA [22]. Indeed, researchers have recently made appeals for more data that address the complexity of psychosocial wellbeing among OVA across a wide range of outcome indicators [23, 24]. It is important to understand the potential impact of financial incentives on the psychological states of OVA because mental health conditions are among the leading causes of mortality among adolescents living in poverty [25, 26]. In SSA, orphaned adolescents experience a greater risk for adverse psychological outcomes compared to their non-orphaned counterparts [27]. Recent qualitative evidence suggests that CT programs in SSA may have a positive impact on psychosocial well-being, including improvements in educational performance, participation in social life, and empowerment for decision making [22]. However, only one study has examined the effect of CT programs on self-reported depression among young people aged 15–24 years in Kenya, finding a reduction in depression among those in CT households [9]. The effect of CT was significant only for males, especially those between 20–24 years old and those who were orphans. Effects on other key mental health outcomes such as anxiety, post-traumatic stress symptoms (PTSS), and future outlook were not reported. Furthermore, the study only included data from a part of western Kenya that has higher rates of poverty and HIV relative to the rest of the region [28, 29].

Accordingly, we aimed to investigate the association between unconditional CTs and depression as well as three other mental health outcomes, namely PTSS, anxiety, and future outlook among younger adolescents (10–18 years old) in a different part of western Kenya. We hypothesized that OVA living in CT households would perform better across the four measures of psychological wellbeing compared to their counterparts in non-CT households.

Methods

Results

The mean age (standard deviation) of study participants was 14.0 (2.4) years, and 50% of them were female. On average, the number of children in both CT and non-CT households was similar, with 5.4 (1.9) and 5.1 (1.9) children, respectively. Table 1 shows descriptive characteristics and tests of group differences. There were more female participants in CT households (57.6% vs 47.0%). Although majority of participants (71%) were single orphans, orphan status did not differ between the two CT groups. Most participants (89.5%) reported having at least one pair of shoes (92.2% in CT-households compared to 87.9% non-CT households). Children in CT households were more likely to have deceased siblings (14.4% vs 10.4%). Majority of participants (88.1%) viewed religion as important in their life, with no difference in this proportion between the two groups. Additionally, adolescents in CT households were less likely to have engaged in transactional sex or to have been sexually abused (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic, caregiver, household, and psychological characteristics of OVA aged 10–18 by CT status.

https://doi.org/10.1371/journal.pone.0178076.t001

The mean age of caregivers was 49.6 (14.1) years, and CT household caregivers were slightly older than those non-CT household caregivers (52.4 [13.6] vs 47.9 [14.1] years). Approximately half of participants lived with their mothers (45.3%). Most caregivers had no formal education or attained only a primary school level (77.4%), and most OVA (68.8%) had lived for more than five years with their current caregiver. Only about 21.2% of households were considered food secure, including 27.7% of CT households relative to 17.7% of non-CT households. We compared the distribution of several proxies for socioeconomic status between CT and non-CT households and found no statistically significant differences (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Comparison of proxies for socioeconomic status or poverty level among CT and non-CT households contributing participants to the study.

https://doi.org/10.1371/journal.pone.0178076.t002

Table 3 shows the results of multivariable logistic regression modelling. Adolescents in CT households had higher odds of positive future outlook (adjusted odds ratio [AOR] 1.47, 95% CI 1.08, 1.99), were less likely to be anxious AOR 0.57 (0.42, 0.78), and less likely to have symptoms of post-traumatic stress AOR 0.50 (0.29, 0.89). Depression was not statistically significantly associated with whether an OVA was living in a CT or non-CT household AOR 0.77 (0.47, 1.27). Higher age was associated with higher odds of PTSS, with an AOR of 1.26 (1.12, 1.42), and females were more likely to report positive future outlook than males AOR 1.62 (1.14, 2.31). OVAs with a history of sexual abuse had higher odds of depression AOR 5.22 (2.65, 10.27) and PTSS (AOR 1.80 [1.07, 3.03]). Sensitivity analysis using Poisson regression with robust error variance yielded similar results, with no material differences (S1 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Adjusted odds ratios (AOR) and 95% confidence intervals for associations between psychological outcomes and cash transfer among OVA population.

https://doi.org/10.1371/journal.pone.0178076.t003

Discussion

We found that among OVA in western Kenya, living in a household receiving unconditional government CTs was associated with better psychological wellbeing compared to living in a non-CT-receiving household. OVA in CT households were less likely to have anxiety and PTSS, and more likely to report perceptions of positive future outlook compared to their counterparts in households not receiving CTs. Although not attaining statistical significance, we also found that OVAs in CT living households were on average 23% less likely to have depressive symptoms. This finding aligns with the estimates of Kilburn and colleagues who reported that Kenyan youth living in households receiving unconditional cash transfers were 24% less likely to have depressive symptoms compared to those not receiving cash transfers [9]. Few studies have assessed the impact of unconditional government cash transfers on the mental wellbeing of orphaned and vulnerable adolescents in sub-Saharan Africa. A recent review of evaluations of cash transfer programmes concluded that more evidence is needed on the effectiveness of poverty alleviation programs for improvement of mental health [22, 23]. Our study contributes towards filling this gap in knowledge.

These results support our primary hypothesis that in a setting like rural Kenya, CT programs appropriately targeted at poor households may confer psychological benefits to children and adolescents. Our findings are also consistent with previous studies conducted in Uganda which showed that interventions aiming to provide social protection and economic empowerment contributed to improved mental functioning and decreased levels of depression and hopelessness among AIDS-orphaned children [15, 17, 46] Additionally, our findings are also in agreement with those of Baird and colleagues, who found that girls receiving unconditional CT experienced a reduction in psychological distress [47].

This study builds on the growing body of literature on how access to financial resources affects the psychological wellbeing of orphans and vulnerable children, especially in Africa where mental health care infrastructure and services remain highly inadequate. The existing literature indicates that in sub-Saharan Africa, CT programs benefit children’s nutrition, physical growth, help reduce risky sexual behaviors and improve school attendance [8, 9, 13, 14, 17, 20, 48]. Our research contributes to this field by showing that such poverty alleviation programs can play a key role in improving the psychological well-being of OVA. This is especially significant because past evaluations of social cash transfers have generally focused on outcomes related to food security, schooling, sexual risk behaviours, and physical health. Poverty and mental health are mutually linked and increase vulnerability to each other, yet comparatively little attention has been paid to the mental health consequences of poverty. CT programs could be effective in improving psychological outcomes of young and vulnerable people in SSA, where most of these outcomes often go undiagnosed and untreated [26].

Psychological well-being is widely linked to many desirable outcomes, ranging from individual and collective economic productivity, peace, security and general physical health [49]. Our findings may provide a basis for how governmental and non-governmental organizations focus their poverty reduction efforts towards implementing well designed CT programs, including making the improvement of psychological wellbeing one of their core goals. Consequently, governments and non-governmental partners should, whenever possible, collect data on the levels and trends in indicators of psychological wellbeing among OVA in their catchment areas, and monitor whether their interventions affect these outcomes.

By enabling households to meet basic needs, cash transfer program could increase OVA participation in school and other productive community activities, and reduce social isolation and stigma [24]. For instance, adolescents have opportunities to build more interactive and beneficial peer networks [50], which may then improve social wellbeing and general health, potentially strengthening coping strategies against anxiety, depression, and PTSS.

A causal association between participating in an unconditional CT program and reduction in depressive symptoms among young men in Kenya was recently reported [9]. We add to this nascent literature by describing associations between CT and additional indicators of mental health, namely anxiety, PTSS and future outlook. A strength of this study is that the selection of both cash transfer and non-cash transfer households was random, using a well-structured sampling frame [30], therefore reducing potential selection bias. Our sampling scheme enabled us to draw direct comparisons between households caring for OVA that received or did not receive CTs. Secondly, the study comprised adolescents, an important high-risk population that is often understudied because much research focusses on younger or older groups. Evidence-based insight into the health needs and potential preventive and therapeutic interventions for this special population are required. Lastly, we used community health workers to collect household level data, which may help limit some biases associated with participation in the study, such as social desirability bias. Due to their relationships of trust and in-depth knowledge of the local communities, participants are more likely to give them honest responses to our questions. On the other hand, this is a cross-sectional analysis relying on self-reported data, therefore we cannot infer causality. Although all psychological measures had strong internal reliability, the measures were not developed for our specific study population and might have cultural, linguistic, or ecological limitations. In addition, we cannot rule out residual confounding or measurement error, which may have played a role, for example, in the negative association between ownership of a pair of shoes and two of the outcomes of interest. Although results do not change when shoe ownership is excluded from the models, this may be a sign of residual confounding or mismeasured covariates, problems that we are addressing in forthcoming work that includes more detailed quantitative measurement as well as qualitative work.

There is need for further research to understand the emotional and psychological needs of caregivers. It is likely that orphans’ physical, mental, and interpersonal well-being is influenced by their relationships with caregivers, and their community context. Furthermore, additional research to find out whether mental health outcomes differ by gender and age, and whether certain adolescent age brackets are more susceptible to long-term mental health consequences is also warranted. As global economic growth slows down, high poverty rates in the least developed countries persist, and the number of orphaned and vulnerable children and adolescents rises, the need for evidence of the effectiveness of direct financial incentives on physical and mental health outcomes becomes more urgent.

In conclusion, we found that OVA living in households receiving CTs in western Kenya reported better psychological wellbeing relative to their counterparts in households not receiving CT. Our findings complement other studies that point to the positive role of CT programs in promoting health. As suggested by recent finding from Kenya, CTs may be most effective as part of a combination of health promotion strategies for orphaned and vulnerable children [9, 14, 48]. As evidence mounts for a cause and effect relationship between CT programs and mental health outcomes among children in developing countries like Kenya, governmental and non-governmental agencies should be encouraged to expand CT programs to reach a wider pool of eligible households caring for orphaned and vulnerable children, and to make the measurement and evaluation how CT interventions impact these outcomes a central component of their health promotion strategies.

Supporting information

Acknowledgments

We wish to gratefully acknowledge all the households and individuals who participated in this research. We also thank all the residents of the Uasin Gishu County especially those caring for vulnerable members of their community, notably orphaned and separated children, children with physical and mental disabilities, and the elderly. Also we wish to thank the County and District Children’s Officers, for their dedication to protecting the children of Uasin Gishu County. Also we would like to thank the staff and volunteers of the OSCAR project for their efforts and engagement with this project. This work was supported by the Brown/Tufts Fogarty AIDS International Training and Research program, grant number D43-TW000237 from the NIH Fogarty International Center and award Number R01HD060478 and R24HD077976 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Eunice Kennedy Shriver National Institute of Child Health and Human Development or the National Institutes of Health.