In 2013, roughly six million children were referred to child welfare agencies in the United States (
1). Children in contact with child welfare services are at distinct risk for mental health and behavioral concerns. A child welfare referral is a notable marker for risks in the home environment that can increase the likelihood of mental health problems. A substantial percentage (24%−48%) of children who remain at home after a child welfare referral requires mental health treatment (
2,
3). Yet only one-quarter of those with mental health issues in the clinical range receive services, according to some estimates (
2).
Reasons for low rates of service use among children who remain at home after contact with child welfare services may be partially related to the fact that families from different racial-ethnic backgrounds have unique experiences and may interact differently with social service systems. Among families in contact with child welfare, race-ethnicity is a common predictor of mental health service use (
4,
5).
In 2000, Latino children represented 14.2% of confirmed cases of maltreatment (
6), and by 2013, the proportion had risen to 22.4% (
1). Although this shift may be consistent with changes in the U.S. population, Latino families may have greater risk of exposure to particular stressors related to the immigration experience that may increase need for services. For children in such families, exposure to acculturative stress, chronic poverty, neighborhood disadvantage, discrimination, and social isolation may elevate the need for mental health services. Among families with undocumented members, the fear of discovery may cause significant stress and insecurity for children (
7,
8). Immigrant status has been associated with resilience, traditional coping mechanisms, and family ties that seem to be protective of mental health (
9), but this may be less true for the subset of families that experience child maltreatment referrals.
Although immigrant families—particularly those involved with child welfare—often encounter stressors that could increase need for services, overall they may have poorer access to services. Studies have connected parental nativity and legal status to deep inequalities in receipt of general medical and mental health services in the general population (
9–
11), and growing evidence suggests that this may be true among Latino families in the child welfare system as well (
8,
12–
14). Reasons for inhibited use of services by immigrants in child-serving systems have been theorized to include unfamiliarity with available services, difficulty navigating service systems and identifying culturally or linguistically congruent service providers, provider bias, and cultural differences in the understanding of symptomatology and help-seeking behaviors (
8,
12–
21). For children with undocumented parents, restrictive social policies that limit eligibility for government support and parent fear of exposure to public authorities, even when children are eligible for services, may serve as additional barriers to service use (
10,
11). Latino families referred to child welfare may be especially likely to have such experiences, given estimates that over two-thirds of children involved with child welfare who have immigrant parents are Latino (
21).
This study examined how factors related to the immigration experience may influence mental health needs and service utilization among children in Latino families under investigation by child welfare. We analyzed how characteristics of the immigration experience influence mental health service utilization through the lens of Andersen’s help-seeking model, which characterizes service use as driven by a combination of family predisposing characteristics as well as enabling conditions, needs, and contextual factors (
22–
24).
Predisposing characteristics, such as gender and age, are unmalleable. We treated parent nativity as a predisposing characteristic because we were interested in whether children in families with caregivers born in the United States have different patterns of service need and use compared with children of parents born elsewhere. Enabling conditions are factors subject to change that may enable or inhibit service receipt, such as income and insurance status. We considered parent legal status as an enabling condition because we were interested in knowing whether a parent’s undocumented legal status inhibits child mental health service receipt.
Methods
Sample and Study Design
This study used data from the second round of the National Survey of Child and Adolescent Well-Being (NSCAW II), a national probability study of families investigated by child welfare agencies for maltreatment (
25). The NSCAW II cohort included children from the ages of birth to 17.5 years at the time of sampling who had contact with the child welfare system during a 15-month period beginning in February 2008 (N=5,872). This research investigated a subset of 390 Latino children ages three to 17 who remained in the home with a biological parent at the time of the initial interview and who had complete data for all predictors used in multivariate models.
The NSCAW II sample was drawn using a two-stage cluster approach in which children (one per family) were sampled from 81 child welfare jurisdictions in 30 states. Face-to-face interviews were conducted with children, caregivers, and child welfare workers. Baseline interviews were completed approximately four to six months after the onset of the child welfare investigation. Informed consent was obtained by trained field representatives at the time of the baseline interview. Wave 2 follow-up interviews were conducted on average 18 months after the onset of the child welfare investigation. Further details on the study design and procedures are available elsewhere (
25). Approval for this study was obtained from the Institutional Review Board at the University of Southern California.
Measures
Predisposing characteristics.
Sociodemographic characteristics (age and gender) of children were derived from questions in child, caregiver, and caseworker interviews. Immigration-related variables (child nativity, parent nativity, and parent years residing in the United States) were obtained from caregiver interviews.
Enabling conditions.
Parent characteristics, including legal status, education, single parenthood, language, income, and perceptions of economic hardship, as well as the child’s insurance type were obtained from the caregiver interview at baseline. Children whose caregivers reported at both baseline and follow-up interviews that they were uninsured were considered to be uninsured in multivariate models.
Child mental health need.
Need was assessed by caregivers at baseline by using the Child Behavior Checklist (CBCL). The CBCL is a widely used measure of behavior problems that has been standardized by age and gender with large child populations from different socioeconomic backgrounds (
26). The CBCL has well-established reliability and validity for use across cultures and racial-ethnic groups (
27,
28). The internalizing and externalizing scales have strong concurrent validity for clinically diagnosed disorders among Latino youths (
29), and convergent and divergent validity has been established in multiethnic samples of American children (
30). Similar to other NSCAW studies, for this sample of Latino children, internal consistency of the CBCL scale was high (α=.96) (
31).
Two forms of the CBCL were used: one for children ages three to five and another for children ages six to 18. Differences in groups by parent nativity and legal status were tested using a dichotomous CBCL clinical cutoff variable (<64 vs. ≥64), a three-category variable including the borderline range score (60–64), as well as each of the CBCL subscales (internalizing and externalizing). No differences were found by parent nativity or immigration status in any of these groupings of the measure. For the sake of simplicity and to avoid inflating estimates of need, therefore, children were considered in need of mental health services if they scored in the clinical range (≥64) of the CBCL.
Contextual child welfare characteristics.
The most serious type of maltreatment alleged in the investigation was obtained from the caseworker interview. Level of severity of risk to the child was also determined by the caseworker, who identified maltreatment risk as none, mild, moderate, or severe. Drawing from the caseworker interview, we created a variable to represent whether the family received child welfare services, such as parent training, respite, child care, family preservation, or other services to prevent out-of-home placement.
Use of mental health services.
Caregivers completed questions from the Child and Adolescent Services Assessment (CASA) to measure children’s use of mental health services in the following settings at baseline and follow up: specialty mental health services (inpatient psychiatric treatment, residential treatment, day treatment, community mental health center treatment, and treatment from private professionals), medical mental health services (from a hospital or family doctor), and school-based mental health services. A dichotomous variable was created to indicate whether children received mental health services any time during the period in question, ranging from five months before to approximately 18 months after the initiation of the child welfare investigation.
Data Analysis
Bivariate, weighted chi square analyses were conducted to examine associations between mental health service use and relevant predisposing characteristics and enabling, need, and contextual factors based on previous literature (
32) and to estimate whether rates of need, service use, and use of specific types of services differed depending on the nativity and legal status of the parent. Variables with associations approaching significance (p<.20) were included in multivariate weighted logistic regression analyses. Multivariate models evaluated relationships between parent nativity and legal status and mental health service receipt, after adjusting for covariates. Parent report of financial need, in lieu of income levels, was used in multivariate models because a substantial proportion of undocumented parents were missing data on family income. All analyses used Stata 12 to adjust for the complex sampling design; a value of p<.05 was considered statistically significant.
Results
Sample Characteristics
Table 1 displays the sample’s weighted data. Over half of the children in the sample were female (52.6%). The mean±SD age was 8.48±7.31 years. Although 91.1% of children were U.S. born, more than one-third (36.0%) had a foreign-born primary caregiver. Of the parents, 14.3% had been living in the United States for more than 20 years, but 8.1% had resided in the United States for ten years or less. Most parents (75.0%) were U.S. citizens, 12.1% were legal residents, and 13.0% were undocumented. Just over half of caregivers (52.8%) had a high school degree. A language other than English was spoken in 60.2% of households. Nearly half of parents reported struggling financially, with over half of families below the poverty level, and 17.1% of children had no health insurance at baseline.
Neglect was the primary allegation of maltreatment for over one-third (35.5%) of children, followed by physical abuse (16.9%), domestic violence (13.9%), and substance use issues (12.4%). Roughly one-fifth (21.3%) of children fell within the clinical range for a behavioral or mental health problem. Differences in child mental health problems by parent nativity and citizenship or legal status were nonsignificant, although the percentage of children with clinical mental health problems was lower among those with immigrant parents compared with U.S.-born parents and lower still among those with undocumented parents (
Table 2).
Mental Health Service Use
Overall, 29.1% of the sample used mental health services (
Table 2). A gap between overall clinical need (10.3%) and service receipt (4.1%) was identified only for children with undocumented parents. A significantly lower percentage of children of immigrants received mental health services compared with children of U.S.-born parents (17.9% versus 35.4%, F=4.03, df=1 and 69, p=.04), although there were no significant differences between the groups in terms of mental health need. The percentage of children who received services was much lower among children of an undocumented parent (4.1%) versus a U.S. citizen (33.6%) or a legal resident (28.2%) (F=5.64, df=1.96 and 135.27, p=.004). Over one-third (37.1%) of children whose parent had at least a high school degree received services, compared with just 20.2% of those whose parent did not complete high school (F=7.58, df=1 and 69, p=.007). As expected, the percentage of children who received services was much higher among children who scored in the clinical range for a behavioral or mental health problem compared with those who did not have a clinically significant mental health issue (60.7% versus 20.6%, F=41.25, df=1 and 69, p<.001).
Table 2 presents the proportions of children who received mental health services by the nativity and legal status of the primary caregiver. Notably, 13.1% of children of immigrants received specialized mental health services compared with 23.5% of children of U.S.-born parents (F=3.84, df=1 and 69, p=.05). Rates of receipt of both specialized and school-based services were much lower among children with undocumented parents compared with children of U.S. citizens and legal residents (specialized services, F=3.46, df=1.87 and 129.31, p=.03; school-based services, F=8.07, df=1.30 and 89.62, p=.002).
Multivariate Analyses
Multivariate models examined the influence of parent nativity and legal status on child service use (
Table 3). After adjusting for covariates, model 1 showed that the odds of receiving services were significantly lower for children of immigrant parents (odds ratio [OR]=.35, p=.05) compared with children of U.S.-born parents. Follow-up analyses that adjusted for length of caregivers’ residence in the United States found that this finding applied only to children whose parents had resided in the United States for ten years or less. Model 2, which accounted for parent legal status, revealed that the odds of receiving services were significantly lower for children whose parents were undocumented compared with those whose parents were U.S. citizens (OR=.09, p<.01).
Discussion
This study provides new information concerning the roles of parent nativity and legal status in meeting the mental health service needs of Latino children involved with the child welfare system. Although rates of clinical need were not statistically different between children of immigrants and children of U.S.-born parents at the bivariate level, children of immigrants had significantly lower rates of mental health service receipt. This finding is consistent with previous literature examining unmet mental health needs among children of immigrants (
8). However, although parent nativity and legal status rates were not significantly associated with mental health need, it is possible that the rates of lower mental health need among children of immigrants versus U.S.-born children may have been substantial enough to explain the reduced odds of service receipt among children of immigrants. Separate multivariate analyses that identified a positive association between years of residence in the United States and service receipt suggest that barriers related to acculturation may partly explain the reduced service receipt among Latino children. Alternatively, such findings might also reflect provider bias in assessment or understanding of needs or in beliefs about assisting children of immigrants, which may influence access to services (
33,
34).
Results of this study advance previous study of the relationship between parent nativity and child mental health service use by identifying a gap in rates of clinical need and service receipt among children with undocumented parents (
8). Multivariate results showed that after the analyses adjusted for need and other relevant variables, children of undocumented immigrants had the lowest odds of service receipt. There are several possible explanations. Given that a lack of health insurance was not predictive of mental health service receipt, it appears that ineligibility for services does not explain reduced service use among children of undocumented immigrants. Rather, data concord with arguments that the effects of stigma and fear of engaging with public agencies due to caregivers’ vulnerable legal status may deter parents from seeking services for their children (
10,
11,
13,
14,
18–
21). It is possible that such children receive referrals for services, but parents are too fearful to pursue treatment.
Findings further revealed that children with undocumented caregivers were less likely to receive school-based services compared with other children—a significant finding considering that schools typically serve children who may not have the opportunity to receive interventions in other settings. This finding suggests that the needs of children with undocumented caregivers remain under the radar of school officials or that the hesitation of undocumented caregivers to engage in government services extends to school-based settings.
This study represented a unique opportunity to explore the influence of parent nativity and legal status on patterns of mental health service need in a nationally representative sample of Latino children who had been reported to child welfare. It also offers potential explanations for disparate use of mental health services among children of immigrants. Despite its strengths, the study was not without limitations. It is possible that some of the most vulnerable immigrant families included in the sampling frame may have been more likely to decline participation in a national survey, potentially causing underrepresentation in the sample. In addition, although the CBCL has been validated with diverse populations, it is possible that the measure does not fully address cultural presentations of mental health symptoms.
The study did not account for all possible factors contributing to service receipt. For example, the study was unable to explicitly examine the effects of acculturation on service receipt, given that the information was not collected. Although years of residence in the United States is a crude proxy for acculturation, it alone does not represent the cultural factors that could be at play. For instance, it is unknown whether families sought more culturally congruent or traditional forms of assistance for mental health issues. Finally, the term “Latino” is a broad term representative of an ethnicity that contains much diversity. In this study, the majority of the Latino sample was of Mexican origin. Future research is needed to examine varying linguistic, cultural, and legal mechanisms within Latino subgroups.
Conclusions
The results of this study contribute to growing discourse on Latino family needs and child mental health within the child welfare system. Analyses support previous findings regarding the effects of parent nativity on mental health service use by children and advance the literature by identifying parent legal status as a unique inhibitor of service receipt. Data suggest that the mental health needs of children of immigrants, especially those with undocumented parents, are not adequately met and that disparities in service utilization in the Latino population may be driven by immigrant families. That is especially concerning at a time when an increasing number of children living in the United States have parents who are immigrants and when immigrant families face unique stressors that contribute to risk for mental health issues. These stressors also exclude them from the benefits of social welfare policies aimed at supporting vulnerable families and children.
For immigrant families, a child welfare referral may provide a rare window through which to evaluate the mental health needs of children who may not access services in other circumstances. Unfortunately, this study suggests that the child welfare system and provider bias against immigrants cannot be eliminated as potential barriers to the receipt of needed services among children of immigrant families. Findings also present a clear message that parent legal status, with its associated fears and stigma, may be a prominent explanation for inhibited service use among Latino children, the vast majority of whom are U.S. born. Yet, it is also possible that immigrant families opt to engage in more traditional methods for addressing mental health issues, turning to religious leaders or folk healers in lieu of prescribed counseling or talk therapy. Understanding reasons for such barriers to service receipt is a necessary step toward providing better quality services to Latino children and families.