Background

Child maltreatment involves physical maltreatment, psychological and sexual abuse and negelct inflicted upon children. The World Health Organization (WHO) estimated that globally one out of two children aged 2–17 years experiences maltreatment each year [1]. A meta-analysis on child maltreatment also indicates a lifetime prevalence of 22.6% and 36.3% worldwide for physical and psychological maltreatment respectively [2]. It is evident that any forms of maltreatment adversely affect the child’s health, rights and development, which has long been considered an international public health issue [3, 4]. In particular, childhood maltreatment is observed as associated with increased emotional and behavioral difficulties in children compared to physical health affect [5,6,7]. A meta-analysis estimated that worldwide, 46% of maltreated children are suffering from chronic depression, and over 57% of population with bipolar disorder revealed that they have experienced maltreatment during their childhoods [8]. A plethora of studies identified that victimized children, either physically, psychologically or neglected, extensively suffer from anxiety, sleeping and eating disorders, difficulties concentrating, withdrawal symptoms, antisocial and borderline personality disorders, and suicidal symptoms [6, 9,10,11,12,13,14]. These adverse outcomes persist from childhood [7] and can lead to severe psychosocial difficulties in long run [15, 16]. Notably, among the different groups of children, child laborers have seen as highly susceptible to be subjected to maltreatment [17].

Child labor itself considered as a serious global public health and social concern. Currently 152 million children are engaged in labor with 70 million working in hazardous conditions [18]. Evidently, in addition to occupational physical injuries caused by child labor, there is also evidence that child laborers are more likely to experience emotional difficulties than children who do not participate in labor [19, 20]. There is evidence that children who are employed are more likely to suffer from internalized psychosocial impairments such as social isolation, hopelessness, low self-esteem, decreased coping efficacy, and externalized psycho-somatic problems including aggressive behavior, a hostile attitude, and academic failure [10, 21,22,23].

The extent of maltreatment of child laborers is widespread. According to a study exploring the extent of child abuse among working children in Turkey, approximately 62% of child worker apprentices have been abused at work [24]. A number of rigorous studies conducted in the South Asian region have concluded that child labor abuse is endemic on the continent [25,26,27]. Approximately 46.6% and 40.77% of child laborers in Nepal were found to be victims of intentional physical and psychological abuse, respectively [26], while over 83% were found to be victimized in India [27]. Child laborers in Bangladesh are not free from this curse, as Hadi (2000) noted a substantial proportion of working children are routinely maltreated. Of note, their abusive experiences both at workplaces and home further escalate the risk of being socially and psychologically impaired [17, 26]. This study focused in particular on this complex issue.

In spite of these high spikes of violence against child laborers, very few studies have attempted to assess the impact of maltreatment on the mental health impact of these children [16, 17, 26, 27]. Given that it is already established that child labor has a negative effect on child’s emotional health, the impact is further compounded by violence they endure while in the workplace. Studies on the maltreatment of child laborers have also documented an intensified emotional and behavioral impairments [26,27,28,29]. The early detection and treatment of any psychological impairment of children is a primary prevention strategy in any country. However, the emotional well-being of victimized children tends to be neglected in many developing countries, such as Bangladesh, and continues to be ignored by low socio-economic families [30]. Therefore, there is a need, therefore, to gain a comprehensive understanding of the impact of maltreatment of child laborers on their emotional and behavioural well-being.

The purpose of this study is to examine the impact of maltreatment on psycho-social difficulties among child laborers from the perspective of their parents. In particular, this study aimed (i) to identify the psycho-social factor structure using the Paediatric Symptom Checklist (PSC) for child laborers, and (ii) to assess the relationship between maltreatment experiences and psycho-social impairments among child labourers.

Methods and materials

Study setting

This cross-sectional study has been conducted in the three Upazilas (sub-district) namely Bishwambarpur, Dharmapasha and Doarabazar of Sunamganj district of Bangladesh. The purposive selection of the three Upazilas was also based on the criterion such as higher illiteracy, poverty rate and agricultural holdings compared with other Upazilas of this district [31, 32].

Participants

A total of 100 parents of child laborers, aged 10 to 17 years, were selected as a representative population using a snowball sampling method. Snowball sampling were followed since the participants are marginalised group of population, and it would be difficult to reach them randomly [33]. As the parents were asked about the maltreatment history of their children as laborers, the sample size of parents was estimated based on the proportion of child maltreatment (82.4%) in Bangladesh found in a prior study [34], 95% confidence interval and an absolute precision of 7% as per the method proposed by Lwanga & Lemeshow [35]. This resulted in 114 parents with an expected dropout rate of 20%. Later, the study was able to recruit 100 parents, with an 88% response rate.

Data collection tools

ISPCAN child abuse Screening Tool- parent version (ICAST-P)

The ICAST-P questionnaire was adopted to measure the extent of physical, psychological maltreatment and neglect. It measures disciplinary maltreatment practices directed at children [36]. The ICAST-P was found adequate to assess child maltreatment at work and at home [37]. This tool is found validated and reliable (Cronbach alpha: 0.60–0.87) in cultural contexts of South Asian countries [38, 39]. Children’s maltreatment experiences were measured by sixteen physical maltreatment items, eight psychological items, and three neglect items. Items of both physical and psychological maltreatment included six response categories for the past year prevalence of child maltreatment including never, yes but not in the past year, once or twice, 3 to 5 times, 6 to 10 times, and ≥ 10 times and weighted values were 1, 2, 3, 4, 5, and 6 respectively. A mean score was calculated for each form of maltreatment. The mean score ranged from between 0 and 6 for each module of maltreatment [36].

Paediatric symptom checklist (PSC)

Typically, standardized instruments are used to estimate psycho-social problems among victimized children worldwide. Mental health professionals use a number of clinically validated tools for screening and diagnosing psychological disorders. However, most of the existing measurement tools are not conducive to apply in all paediatric settings as these are time-consuming, lack standardized norms, require professional training to administer [14, 40]. Additionally, the younger children may not possess sufficient cognitive understanding to comprehend some psycho-matric items [14]. In contrast to these measurement tools, PSC is found as widely used screening inventory, which is easy to administer as only 5 min take to complete scoring and is simple to interpret [40,41,42]. It provides opportunity to capture the parents’ impressions of their children’s psycho-social conditions [40]. It has been validated for use in clinical, educational, and public health, low-income and minority settings [40, 43]. This tool has been identified as a reliable screening measure (Cronbach alpha: 0.86) for assessing children’s psychosocial problems [44]. In this study, the PSC tool captured the parents’ recognition of the cognitive, emotional and behavioral problems associated with child maltreatment. A total of 35 items were listed in PSC with response items weighted as 0, 1 and 2 for Never, Sometimes and Often [45]. The range of possible scores is between 0 and 70 points, which were calculated by adding the individual values for each item. The threshold score of 28 or above indicates psychological impairment in child laborers [42].

For comparison, the English questionnaire, which included both the ICAST-P and PSC (Parents’ version) tools, was translated into Bangla and back-translated. In the first stage, three academics with expertise in socio-epidemiology and health science research, including one paediatrician with a doctorate and one biostatistician with a doctorate, as well as one registered nurse with a doctorate, translated the English versions of the ICAST tools and PSC tools into Bangla.

Two academics were knowledgeable about the scales, their terminologies, as well as the theoretical construction of the questionnaire. One academic (academic and paediatrician) was familiar with child health related terminologies. He knew colloquial phrases, idioms, and idiomatic expressions in English. In collaboration with the academic experts, we translated the original English version of questionnaires into Bangla.

Two additional academic experts (who were also registered nurses with PhD degrees and experienced in scale validation) reviewed the transcription process with the authors of the study and resolved concerns. In a consensus meeting composed of all the experts (translators) and authors of the study, we analysed and discussed discrepancies in words, meaning, or phrases, and resolved them. As a result, we were able to develop Bangla versions of all questionnaires. Additionally, the Bengali version of the questionnaire was tested on three parents of child laborers to determine whether it was understandable for parents with low literacy levels. We found the translated versions were comprehensible.

In the second stage, an academic expert (Biostatistician), whose native language is Bangla and English is a second language, and one registered nurse with a PhD degree, whose first language is English, but who can also speak Bangla performed a back-translation of the translated Bangla scales into English. Both translators were experienced in international translation. These translated versions were combined to produce one single translated version for each scale. Additionally, we consulted with an academic with expertise in child protection research, whose native language is English and Bangla is a second language, in order to compare and evaluate the translated versions in terms of similarities and conceptual equivalence compared to the original English version and the first Bangla translation. As a result, the Bangla versions of the scales were back translated into English.

Collection of data

The study conducted face to face interviews with parents of child laborers aged 10 to 17 years. To conduct interviews, the study recruited three skilled investigators, one of who had experience of collecting clinical data. Considering the scale items in mind, they were given a three-day training session to familiar with on the subject matter of the study, building rapport, obtaining consent, maintaining confidentiality, avoiding conflict, and responding appropriately. Data were collected in between April to June 2021.

Data analysis

Prior to exploring the relationship between ICAST-P and PSC, the study conducted the exploratory factor analysis (EFA) and Confirmatory factor analysis to compress the eligible items of PSC into a small number of factors [46]. The application of EFA is identified as appropriate to identify factor structure of psycho-social impairment among vulnerable children [47].

The study measured Kaiser-Meyer-Olkin (KMO) and the Bartlett’s tests of sphericity to consider whether principal component analysis is appropriate in EFA [46]. The study considered a fixed number of factors instead of considering eigenvalues. Evidently, studies on the factor analysis of PSC reported by parents of children retained a three-factor structure including internalising, externalising and attention factors [41, 48]. Therefore, a varimax rotation approach is applied constrained to these three factors [46]. A loadings value of greater than 0.30 is considered in order to evaluate strong factor loadings [40, 48]. Later, the mean value of each latent constructs with retained PSC items were estimated.

Further, a CFA was conducted to verify the validity of the latent constructs that were derived from the EFA. To assess whether the data are accurately fit to the model developed in EFA, PCMIN/df, Root mean square error of approximation (RMSEA), Comparative fit index (CFI) and PCFI were taken into account. PCMIN/df > 3.0, RMSEA > 0.08, PCFI > 0.5 and CFI > 0.9 are the standard criteria for determining model fitness [49]. The multivariable linear regression analysis was employed to assess the relationship between forms of maltreatment and emerged psycho-social factors (internalizing factors, externalizing factors and attention). A 95% confidence level and significance level of less than 0.05 were set as criteria for determining significant relationship. The overall analysis was performed using SPSS version 26, Stata version 16.1 and R version 4.2.1.

Ethical approval

To conduct the study and interview with parents of child labourers aged 10 to 17 years, a formal ethical approval was obtained from the Human Research Ethics Committee of the University where the research is being conducted. Additionally, authorization to conduct interviews were also taken from a University of Bangladesh.

Data collectors shared the study description and purpose of the research to parents and ensured that they provided verbal consent to be interviewed. It was explained to them that they could withdraw from giving an answer during the interview session if they feel uncomfortable or reluctant to continue. Participants were assured of confidentiality and compensated $2.5AUD (157 Taka) for their participation in the survey. Researchers further assured them that if they became physically or psychologically ill during the interview session, they would assist them in obtaining medical treatment. Nevertheless, no participants found suffered from any health problems during interviews. In view of the fact that many of them may not understand the questions, the data collector filled in the form on their behalf.

Results

Construct validity of PSC tool using EFA and CFA

The fixed three-factor solution resulted in 36.6% of the total variance being explained in the PSC-35 items. In the rotated factor pattern, the estimated factor loadings for items related to psycho-social difficulties were above 0.4, as shown in Table 1.

Table 1 Factor loadings for the paediatric symptom checklist (PSC)
Full size table

The EFA resulted that Factor 1 (internalizing subscale) is consisted of a total of ten items related to internalized psycho-social symptoms among child laborers. Factor 2 is an externalized psycho-social subscale which combined a total of eight items related to the externalized psycho-social problems of child laborers. The third identified factor is termed as attention subscale which consisted of four attention-oriented psycho-social difficulty items. The measured Cronbach alpha demonstrate that internalizing items are highly reliable (α = 0.81) followed by moderately consistent attention subscale (α = 0.71), and externalizing factor (α = 0.63). Based on CFA, the three-factor model appears adequately fitted to the data, as most fit criteria are acceptable. The Bartlett’s test of sphericity (χ2 = 2390.8, p < 0.001) identified the eligibility of conducting a principal component analysis and KMO of 0.6 indicated an adequate sample size for exploratory factor analysis.

Psycho-social impairments and prevalence of maltreatment of child laborers

The overall mean score of the PSC was 16.4. Specifically, 43% of child laborers exhibited an elevated score (28 or above) on PSC, which indicates a startling episode of psycho-social problems among child laborers. The results in Table 2 indicate the mean of psycho-social factors. This apparently reflect that the victimized child laborers were highly susceptible to exhibit internalized psycho-social symptoms followed by externalized psycho-social symptoms and attention difficulties.

Table 2 Descriptive statistics of the identified psycho-social factors
Full size table

The estimated prevalence of physical, psychological maltreatment and neglect of child laborers from the items of ICAST-P is presented in Tables 4, 5 and 6 (included in Appendices). The prevalence estimation indicates that child laborers are highly susceptible to psychological maltreatment (2.8) followed by physical maltreatment (2.6), and neglect (1.8).

Multivariable regression analysis of psycho-social factors and child maltreatment reported by parents of child laborers

Data reported by parents indicate that child laborers who have been victimized were significantly screened for internalized and attention-related psychosocial problems, presented in Table 3.

Table 3 Multivariate regression analysis of the psycho-social impairments and maltreatment of child laborers
Full size table

On adjustment of factors in the multivariable modelling, the study found that child laborers who have experienced physical maltreatment are more likely to be screened for internalized psychosocial symptoms by 0.12 units (β = 0.12, 95% CI = 0.03, 0.21, p = 0.01). The Table 3 further provides evidence that child laborers were more likely to be screened for internalized psycho-social difficulties by 0.11 units when they were psychologically victimized (β= 0.11, 95% CI = 0.01, 0.19, p = 0.02). Of note, no particular type of maltreatment was found to have had a significant impact on child laborers’ externalized psychosocial difficulties. Neglect is also observed as not being significantly related to psychosocial problems experienced by child laborers. In addition, as illustrated in Table 3, child laborers who have experienced physical maltreatment displayed an increase of 0.22 units (β = 0.22, 95% CI = 0.09, 0.36, p < 0.01) in attention-related psychosocial difficulties. Additionally, psychologically maltreated child laborers were more likely to be screened for attention-related psychosocial difficulties by 0.27 units (β = 0.27, 95% CI = 0.14, 0.42, p < 0.001), regardless of other forms of maltreatment.

Discussion

It has been documented that child maltreatment has serious negative effects on the psychological health of children. Among the vicimtized children, socially and economically vulnerable groups of children are highly susceptible to expereicne a variety of psycho-traumatic symptoms [19, 21, 27, 50]. In spite of the fact, the impact of maltreatment of children as laborers remains under-researched [20]. This study attempted to survey the children’spsycho-social difficulties utilizing parents version of PSC and investigate the relationships between psycho-social factors and different forms of maltreatment.

In this study, we found that child laborers, as a vulnerable group of children, show psychosocial impairments associated with internalizing, externalizing, and attention-oriented symptoms, which are also typical psychosocial symptoms of acutely ill children. The retained items derived from factor analysis, under these three psycho-social factors were maintained concordance to the three factors models developed for children with acute illnesses in earlier studies [51,52,53]. The items included in internalized factors apparently depicting depression, anxiety, withdrawal behavior and other internal psycho-somatic symptoms. Externalizing items reflect aggressive, destructive behavior or conduct disorders, while items in the attention subscale apparently implying the effects of cognitive difficulties among child laborers [40, 48, 54].

The estimate of the parents’ reported data revealed that the victimized children, as laborers, are highly prone to experiencing psychosocial difficulties compared with experiences of different disadvantaged and physically challenged groups of children observed in prior studies [40, 41, 47, 48]. The extent of psycho-social difficulties indicates the need for additional research to determine the long-term psycho-social impacts in the child labor population. Specifically, the study resulted that the victimized child laborers are largely exhibiting internalized psycho-social symptoms followed by externalized problems and attention deficit disorders which were parallelly observed in earlier studies [41, 47, 48]. Similar to the findings of current study, Badrand colleagues found that nearly half of all children who are maltreated exhibit internalized symptoms; depression, stress, and anxiety [54]. These internalized symptoms are highly responsible to disrupt normal brain development in early childhood [54,55,56]. On the flip side, a study outlined that externalized psycho-social problems and attention difficulties outweigh internalized problems among victims [6].

Apart from an overall impact of child labor maltreatment, the study specifically measured the psycho-social consequences of each type of maltreatment. Consistent with previous studies, the multivariable regression model further identified that child laborers who have been subjected to physical and psychological maltreatment exhibit greater internalizing symptomatology and attention deficit symptoms [15, 56,57,58]. Similar to the prior studies on child labor maltreatment, the current study also noticed that child laborers who are physically victimized are more prone to exhibit internal emotional difficulties, and psychological maltreatment explicitly resulted in internalized and attention specific psychosocial difficulties [26, 27]. Furthermore, while the current study estimated an insignificant relationship between all forms of maltreatment and externalized psycho-social symptoms, the prior studies identified that maltreated children are more likely to encounter externalized defiant disorders [13, 27, 56]. Child neglect has also been categorized as a potential predicator of psycho-social symptoms. There is ample evidence that neglected children have elevated rates of internalized symptoms (i.e., major depression) and attention difficulties (i.e., ADHD) [56, 59]. However, this study, based on the reports of parents of child laborers observed an insignificant relationship between neglect and psycho-social difficulties among child laborers.

The findings indicate that exposure to physical or psychological maltreatment are significantly more likely to develop internalized and attention-related disorders among child laborers. Therefore, efforts should be made to eliminate all forms of psychological bullying as well as corporal punishment from all settings, although in Bangladesh they are only enforced in educational environments [60]. Specifically, this study shows that victimized child laborers are most likely to experience internalized psychosocial problems, which are often difficult to identify. Therefore, the importance of early detection of psychosocial impairments cannot be overstated. A comprehensive system of early intervention and prevention should be offered by public child protection services in Bangladesh. Additionally, child welfare workers may be able to assist in the assessment of psychosocial symptoms associated with maltreatment, as well as providing information on prevention strategies. Furthermore, this study draws attention to the urgent need for further research into the relationship between child laborers’ experiences of maltreatment and the potential risks of psychosocial difficulties.

Limitation

The study had a number of limitations. The PSC tools were mostly analysed in the context of physically vulnerable children, whereas this study focused on the maltreatment of child laborers drawing on the views of parents of these children. This resulted in a constraint in bridging the psycho-social difficulties of children in general and those of victims of child labor in particular. Furthermore, the data were collected during the Covid-19 pandemic, limiting the population size compared to prior studies of psycho-social impairments in children.

Conclusion

Despite of limitations, the study clearly identified that the maltreated child laborers primarily exhibit internalized, externalized and attention related psycho-social problems. The mental health of children engaged in labor, particularly those who have been physically and psychologically maltreated is compromised. It is therefore necessary to conduct further surveys of the psycho-social problems of these vulnerable groups. Ideally, the findings of this study will provide prospective social science and clinical researchers with insight into the psychological health of children in labor.