Introduction

Despite ongoing interest in adopting from abroad, the number of intercountry adoptions to the United States (US) precipitously dropped during the past two decades from an all-time high of 22,891 in fiscal year (FY) 2004 to 1,622 in FY 2020 (U.S. Department of State-Bureau of Consular Affairs, 2022). The reasons for this decline are complex and include implementation of the Hague Convention on Intercountry Adoption; economic downturns; political tensions, more restrictive requirements, improved support for single parents and emerging domestic adoption programs in many countries; the COVID-19 pandemic; as well as perceived or documented fraud and abuse within the system (Mignot, 2015; Selman, 2009). Haiti remains among the few countries who view intercountry adoption as an option within their child protection program. Recognizing that either restavèk (child slavery) or long-term institutional care were the fates most likely for their orphaned children, Haiti placed over 7,653 children worldwide for intercountry adoption from 2003 to 2009, principally in France (3880), the United States (1968), Canada (924), and Netherlands (382) (Selman, 2015).

Incremental improvements in economic and living conditions in Haiti suffered a severe setback following the magnitude 7 earthquake on January 12, 2010 which destroyed Port-au-Prince and surrounding areas. Following reports of the destruction and mortality, interest in adopting orphaned children from Haiti surged. On January 18, 2010, the Secretary of Homeland Security, Janet Napolitano, announced that the United States would institute a special program of humanitarian parole permitting certain children who qualified to enter the U.S. for adoption. From that date until the U.S. Citizenship and Immigration Services stopped accepting humanitarian parole applications on April 14, 2010, over 1150 children from Haiti were approved for placement with U.S. families (Reitz, 2011).

Despite massive disruption in government infrastructure, on January 31, 2011, Haiti’s adoption authority announced that they were accepting new adoption applications for Haitian children (U.S. Department of State-Bureau of Consular Affairs, 2011). Subsequently, the Haitian parliament approved ratification of the Hague Convention on Intercountry Adoption and established the required central authority to ensure international best practices. Adoptions from Haiti under the Hague Convention began on April 1, 2014 and have proceeded unabated since then (U.S. Department of State-Bureau of Consular Affairs, 2022). From FY 2013 through 2020, Haiti ranked 5th among US placing countries with 1792 children, exceeded only by China, Ethiopia, Ukraine, and South Korea (U.S. Department of State-Bureau of Consular Affairs, 2022).

The traumatic consequences of the January 2010 earthquake and ensuing decade of recurrent natural disasters, political turmoil, and social unrest on the health of Haiti’s children have been well documented (Dube et al., 2018). However, the health of children adopted from Haiti (CAF-H) who arrived in the U.S. after the 2010 earthquake has not been reported. The purpose of this study is to evaluate the health status of CAF-H on arrival to the U.S. to provide evidence-based information for adoption agencies who assist in setting adoption expectations and for adoptive families and healthcare providers who will care for these newly placed children. The findings from CAF-H were compared to children adopted from Latin America (CAF-LA) and children adopted from Asia (CAF-A) to determine if CAF-H present unique medical challenges.

Methods

Study Design and Participants

This study is a retrospective chart review of CAF-H compared with age and sex matched non-Haitian adoptees, all seen at the University of Minnesota Adoption Medicine Clinic (AMC) between January 2010 and November 2019. The Minnesota Department of Health (MDH) supplied data on adoptees seen at the AMC prior to February 2017. Data after this point were obtained through the Clinical and Translational Science Institute-Information Exchange (CTSI-IE). Some children seen in clinic were excluded due to research opt-out status.

Adoptees from primarily Latin America and Asia were considered when creating our matched control dataset due to lack of sample size in other regions. Each adoptee from Haiti (N = 87) was age and sex matched with one adoptee from the Asian group and one adoptee from the Latin American group, resulting in a total sample size of 261 patients seen between January 2010 and November 2019. The University of Minnesota Institutional Review Board determined that this study falls into the category of secondary research for which consent is not required because information was recorded in a manner that the identity of human subjects cannot be readily identified directly or through the identifiers linked to each subject and because research involved information collection and analysis on retrospective patient records for the sole purpose of research. Therefore, this study met the criteria for exemption from review.

Measures

Physical and Social History Examination

Length/height, weight, and body mass index (BMI) were obtained, and z-scores were calculated using the Center for Disease Control’s age and sex specific values (Centers for Disease Control and Prevention, 2022). Records provided by the adoptive parents were reviewed to obtain data related to numbers of transitions and months in pre-adoptive care prior to adoption. Parents reported a variety of caregiving environments prior to adoption but rarely had information on each specific placement their child experienced. Since only 120 government supported family-based foster homes had been established in Haiti by 2020 versus 754 orphanages, we assumed that children were cared for principally within institutional care settings (France 24, 2020). Based on a broad survey of Haiti’s NGO-funded orphanages, they are likely comparable to institutional care settings of the matched controls (Vernaelde & Guillaume, 2017).

Laboratory Screenings

Laboratory screenings according to recommended American Academy of Pediatrics (AAP) guidelines were conducted for all adoptees during their intake appointment and processed through M Health Fairview Medical Diagnostic Laboratory. Immune subjects were defined as only those with serological evidence of immunity excluding subjects with a history of immunization but no serologic data. Tuberculosis (TB) infection was determined either by a history of TB treatment, a positive tuberculin skin test of greater than 10 mm, or a positive interferon-gamma release assay. Tables 1, 2, 3, 4 and 5 in the supplemental section define each health outcome of interest and how it was measured.

Table 1 Demographic Characteristics of Haitian and Non-Haitian Cohorts summarized by N (%) or mean (SD).
Full size table

Statistical Analysis

Data were summarized by mean and standard deviation (SD) or by N and percent for continuous and categorical variables respectively. Differences between groups were determined by ANOVA F-tests for continuous variables and Fisher’s exact tests for categorical variables. Unadjusted and adjusted models assessed the effect of region of origin on months in pre-adoptive care and number of transitions (via linear regression); on high lead, tuberculosis, syphilis, low hemoglobin, low vitamin D, high TSH, polio, and diphtheria/tetanus (via odds ratios (OR) with logistic regression); and on hepatitis A, hepatitis B, parasites, and T4 levels (via OR with multinomial regression). Adjusted analyses accounted for age, sex, and height z-score due to residual differences between the cohorts after matching. BMI z-score could not be used since it is undefined for patients under two years of age. Observations with missing data were excluded from their respective analyses. Significance was assessed at the p = 0.05 level and all analyses were conducted in R version 3.5.1.

Results

Patient Characteristics

The group of adoptees from Asia were primarily from China (72.4%), with a smaller percentage from India, the Philippines, Thailand, Marshall Islands, and Nepal. The group of adoptees of Latin American origin were from Colombia (55.6%) or Guatemala (23.0%), with a minority from Costa Rica, Dominican Republic, Ecuador, Grenada, Guyana, and Honduras.

Demographic variables were summarized in Table 1. The average age of adoptees at the time of evaluation was 6.3 years old for Haiti, 7.33 years old for Latin America, and 6.28 years old for Asia (p = 0.127). The racial demographic makeup of Haiti’s cohort was 43.7% Black in comparison to 0% in both Latin America and Asia (p < 0.001), although 48.3% of Haitians did not have a race listed in their medical chart. Duration of time spent in pre-adoptive care was similar between all cohorts (Haiti = 45.1 months, Latin America = 36.7, Asia = 48.2, p = 0.053). Auxology showed that CAF-H had greater height z-scores (-1.28 vs. -1.82 vs. -2.13) and weight z-scores (-0.32 vs. -0.57 vs. -1.57) than their counterparts from Asia and Latin America (both p < 0.001).

Table 2 summarizes immunization status and incidence of vaccine preventable diseases. Only 34.5% of CAF-H were immune to hepatitis B in comparison to 62.1% of CAF-LA and 64.4% of CAF-A. Hepatitis B infections were present in 6.9% of CAF-H in comparison to 0.0% of CAF-LA and 5.7% of CAF-A. Immune status to hepatitis A was found in 66.7% of CAF-H, 32.2% of CAF-LA and 42.5% of CAF-A. Only 32.2% of CAF-H were immune to polio in comparison to 42.5% of CAF-LA and 50.6% of CAF-A. No Haitian child tested positive for hepatitis C while 5.7% of both CAF-LA and CAF-A tested positive (p = 0.051).

Table 2 Immunization Status for Haitian and Non-Haitian Cohorts
Full size table

Table 3 summarizes laboratory screening results. Enteric/cutaneous pathogens were more prevalent in the Haiti cohort. A higher percentage of CAF-LA (65.5%) and CAF-A (85.1%) had no parasites in comparison to 31.0% of CAF-H (p < 0.001). The proportion of CAF-H with low hemoglobin levels was 14.9% compared to 1.1% in CAF-LA and 2.3% in CAF-A (p < 0.001). The proportion of CAF-H presenting with lead levels of ≥ 5 ug/dL or above was 39.1% in comparison to 2.3% of CAF-LA and 9.2% of CAF-A (p < 0.001). Low vitamin D levels were found in 62.1% of CAF-A compared to 40.2% of CAF-H and 42.5% of CAF-LA (p = 0.008). CAF-H, CAF-LA, and CAF-A had similar percentages of elevated TSH levels (12.6% vs. 11.5% vs. 13.8%, p = 0.968). No Haitian or non-Haitian adoptee tested positive for HIV.

Table 3 Laboratory Screenings for Haitian and Non-Haitian Cohorts
Full size table

Model Results

Unadjusted and adjusted analyses were conducted where adjusted results account for age, sex, and height z-score (Table 4) to model the relationship of medical indicators with the Haiti and non-Haiti cohorts. The immune group is defined two ways, with and without expanding to all that could have immunity by including those with a history of immunization but no serologic data. However, redefining which patients had immunity had minimal effect on the results (Table 4).

Table 4 Results of Regression Analyses
Full size table

Vaccine Preventable Diseases

CAF-H demonstrated increased odds of lack of immunity to polio compared to CAF-LA (adjusted OR: 2.61, 95% CI: 1.16–5.84, p = 0.020), but no statistical difference was found between CAF-H vs. CAF-A. Statistical tests on immunity to Diphtheria and Tetanus failed to meet statistical significance.

Geographic origin had a significant effect on hepatitis A and hepatitis B immunity (p < 0.001). CAF-H demonstrated significantly decreased odds of being non-immune to hepatitis A compared to CAF-LA (adjusted OR:0.38,95%CI:0.17–0.86,p = 0.019) and CAF-A (adjusted OR:0.30,95%CI:0.14–0.65,p = 0.002). However, when expanding immunity to all subjects including a history of immunization, only the relationship between CAF-H and CAF-A remained statistically significant (adjusted OR: 0.36, 95% CI: 0.17–0.75, p = 0.007). The odds of lack of immunity to hepatitis B were significantly increased for CAF-H compared to CAF-LA (adjusted OR: 5.89, 95% CI: 2.80–12.40, p < 0.001) and CAF-A (adjusted OR: 6.87, 95% CI: 3.11–15.19, p < 0.001).

Contagious Diseases

Geographic region impacted the odds of tuberculosis infection (p = 0.052) and parasite prevalence (p < 0.001). CAF-H lacked statistical significance of an increased odds of tuberculosis infection compared to CAF-LA (adjusted OR: 2.71, 95% CI: 0.98–7.45, p = 0.054), but demonstrated statistical significance of an increased odds of tuberculosis compared to CAF-A (adjusted OR: .16, 95% CI: 1.08–9.31, p = 0.036). The odds of having one parasite infection was significantly greater in CAF-H compared to CAF-LA (adjusted OR: 3.32, 95% CI: 1.55–7.08, p = 0.002) and CAF-A (adjusted OR: 10.44, 95% CI: 4.15–26.29, p < 0.001) and even greater when comparing the odds of 2 or more parasites (CAF-H vs. CAF-LA: adjusted OR: 8.43, 95% CI: 3.38-21.00, p < 0.001; CAF-H vs. CAF-A: adjusted OR: 38.48, 95% CI: 9.97–148.40, p < 0.001).

Other Medical Conditions

Country of origin significantly affected the risk of an adoptee having elevated lead levels (p < 0.001), low vitamin D levels (p = 0.013), and low hemoglobin levels (p < 0.001). CAF-H were at significantly increased odds of higher lead levels compared to CAF-LA and CAF-A (both p < 0.001), but large standard errors prevent an accurate assessment of the magnitude of this association.

CAF-H demonstrated significantly lower odds of low vitamin D levels compared to CAF-A (adjusted OR: 0.40, 95% CI: 0.20–0.79, p = 0.008), but no statistical difference was found between CAF-H versus CAF-LA (p = 0.762).

CAF-H had significantly increased odds of lower hemoglobin levels compared to CAF-LA (adjusted OR: 15.25, 95% CI: 1.88-123.57, p = 0.011) and CAF-A (adjusted OR: 9.18, 95% CI:1.81–46.60, p = 0.007). Additionally, the odds of anemia in children with elevated lead levels in all three groups combined was significantly higher compared to children with normal lead levels (adjusted OR: 4.89, 95% CI: 1.55 to15.45, p = 0.007; results not shown).

Discussion

We found significant medical differences between adoptees from Haiti and non-Haitian adoptees in immunity to polio, hepatitis A, and hepatitis B, and in tuberculosis, enteric/cutaneous pathogen prevalence, lead levels, anemia, and height/weight. This is the first study to report the health profile of CAF-H following the 2010 earthquake and demonstrates that there are significant differences in the health of CAF-H compared to adoptees from other regions of the world. The purpose of this research is not to stigmatize children adopted from Haiti, but to help identify health needs that will help providers better care for this vulnerable patient population.

Immediately after the 2010 earthquake, accomplishments were made in Haiti’s rural and urban health settings due to the huge influx of foreign aid, but shrinking resources, civil unrest, and government instability have caused progress to either stagnate or reverse. The resulting deficits in public health infrastructure and sanitation systems can be directly or indirectly tied to the differences in health conditions in CAF-H noted above.

CAF-H demonstrated a significant lack of viral immunity, with 34.5% of Haitians non-immune to polio and 54% non-immune to hepatitis B. CAF-H have 2.61 times the odds of lack of immunity to polio compared to CAF-LA after adjustment (p = 0.020). Lack of immunity may be attributed to chronic under-vaccination in Haiti that existed even prior to the earthquake. Previous research on Haitian children found that only 40.4% had received all eight recommended routine vaccination (Rainey et al., 2012). Early infant vaccinations in Haiti are high, but many children fail to complete the full vaccination series due to decreased awareness, difficulty accessing vaccinations, and childhood illness. Gaps in the vaccination program were undoubtedly exacerbated by the 2010 earthquake. Research has found that successful strategies to increase vaccine coverage in Haiti include increased vaccine cold chain storage and effective vaccine management (Tohme et al., 2017).

Haitian children’s lack of immunity to hepatitis B may also be attributed to the lack of a hepatitis B birth dose requirement in Haiti’s immunization schedule. Haiti was the last country in the Americas to introduce the hepatitis B vaccine in combination with a pentavalent vaccine in 2012 and is classified as an intermediate endemic hepatitis B country by the WHO, with a surface antigen prevalence of over 7%, almost identical to the percentage observed in our group of CAF-H (6.9%) (Childs et al., 2019). Previous studies found that Haitian children who tested positive for the hepatitis B surface antigen had no documented hepatitis B vaccine (Childs et al., 2019). When compared to non-Haitian adoptees, CAF-H were more than five times as likely to be non-immune to hepatitis B after adjustment.

Haiti has the highest prevalence of tuberculosis in the Americas (Delva et al., 2016) and the situation is known to be even more dire in Haitian orphanages. In a study of 445 children in six Haitian orphanages, prevalence was reported as 1123 in 100,000, nine times greater than their figure for Haiti’s general population (Francis et al., 2002). We found that 16.1% of CAF-H had evidence of TB infection compared to 8.0% of CAF-LA and CAF-A each. In our adjusted analysis, CAF-H had over 3 times the odds of tuberculosis compared to CAF-A (p = 0.036) while CAF-H had 2.71 times the odds of tuberculosis compared to CAF-LA, but failed to meet statistical significance (p = 0.054). Adoptees are known to be at higher risk of tuberculosis, likely due to origin from countries with a high prevalence of tuberculosis and residence in institutional settings that favor M tuberculosis transmission, including close contact with orphanage workers, who are seldom screened for tuberculosis (Mandalakas et al., 2007). Previous research on CAF-LA provides a situational comparison between CAF-LA versus Haiti. A closer look at Guatemalan international adoptees displays lower rates of TB (7%), lead exposure (3%), and parasitic infections (8%) (Miller, 2005). Prior to the earthquake, there was not accurate TB case detection, and the earthquake further exacerbated the struggles in the healthcare system due to the loss of Haiti’s National TB program office and treatment center for multidrug-resistant TB (Koenig et al., 2015). Our findings demonstrate the importance of initial and annual tuberculosis screening for both employees and adoptees in institutional care settings.

Poor sanitation in Haiti likely contributed to our findings of 68.9% of CAF-H testing positive for one or more enteric/cutaneous pathogens compared to only 34.4% of CAF-LA and 14.9% of CAF-A. CAF-H had 8.43 times the odds of having 2 or more parasites compared to 0 versus CAF-LA and 38.48 times the odds versus CAF-A. The following parasites were detected most frequently among CAF-H: Giardia duodenalis (44.8%), Blastocystis hominis (33.3%), and Hymenolepis nana (11.5%). Parasite prevalence is strongly associated with socioeconomic factors like sanitation and safe drinking water, both of which are chronically deficient in Haiti (Hailegebriel, 2017). Previous studies in Ethiopia have also found an association with increased parasitic infections in areas of poor sanitation and unsafe drinking water, indicating the need for Haiti to scale up public sanitation and water treatment programs (Hajare et al., 2022).

Since immunization for hepatitis A is not routinely provided in Haiti, the lack of public health infrastructure and poor sanitation likely played a role in our finding of increased odds of hepatitis A immunity among CAF-H (p = 0.001). 66% of CAF-H were immune to hepatitis A in comparison to 32.2% of CAF-LA and 42.5% of CAF-A. The incidence of hepatitis A is strongly associated with socioeconomic factors like clean sanitation and safe drinking water (Franco et al., 2012). In terms of the above infectious diseases, not only are the consequences for the adopted child significant but transmission of hepatitis A and B as well as TB to adoptive family members have been documented and child-parent associations for Giardia are well described (Abdulla et al., 2010; Curtis et al., 1999; Heusinkveld et al., 2016; Sokal et al., 1995).

Haiti’s poor public health infrastructure also likely contributed to elevated lead levels in 39.1% of CAF-H in comparison to only 2.3% of CAF-LA and 9.2% of CAF-A. After adjustment, CAF-H had 23.79 times the odds of high lead levels compared to CAF-LA and 7.04 times the odds compared to CAF-A, but the confidence intervals were wide. Lead is a highly toxic metal that can impact human development, hearing, and puberty, and increase morbidity and mortality. Increased lead levels can significantly affect the central nervous system leading to hyperactivity, behavioral disorders, impaired cognitive and motor function, and reduced intelligence (Sanders et al., 2009). Children absorb nearly five times as much lead compared to adults due to increased hand-to-mouth contact, and lead is especially toxic in undernourished children who absorb increased lead when lacking in other nutrients (World Health Organization, 2021). Despite Haiti phasing out leaded gasoline in 1998, a 2015 study found that over half of Haitian children had lead levels greater than ≥ 5 µg/dL (Carpenter et al., 2019). Exposure to improperly discarded car and household batteries appeared to be a primary driver of increased lead levels with over half the children reporting walking past discarded batteries on their way home from school. In addition, half of parents interviewed were unaware of their child’s exposure to lead based paint (Carpenter et al., 2019).

We found that 14.9% of CAF-H had low hemoglobin levels and were at 15.25 times the odds of low hemoglobin levels compared to CAF-LA and 9.18 times the odds compared to CAF-A (p < 0.001). Anemia is a major global childhood public health problem that impacts growth, cognitive and motor development, and mortality. The prevalence of anemia in Haitian children is found to be up to 40% (Ayoya et al., 2013). Research on the types of anemia prevalent in Haitian pediatric populations is limited, however, Haiti’s primary intervention approach to combat anemia is iron-folic acid supplementation (Heidkamp et al., 2013). Other factors such as genetics, diet, and environment may also affect the prevalence of anemia. A causal relationship between elevated blood lead level and iron deficiency must be considered as lead competitively blocks iron absorption from the gut and inhibits hemoglobin biosynthesis at several points of the process (Hegazy, Zaher, Abd El-Hafez, Morsy, & Saleh, 2010). In our study, children with elevated lead levels in all three groups combined were 4.89 times more likely to have anemia compared to adoptees with normal lead levels (p = 0.007).

One unexpected difference between Haitian and non-Haitian adoptees was that stature and weight in CAF-H were significantly greater than their counterparts CAF-LA and CAF-A. Rather than experiencing the expected severe growth failure versus peers noted in institutionalized children worldwide (van IJzendoorn et al., 2020), percentages of CAF-H who were stunted (25.3%) or wasted (9.2%) were similar to Haitian children in general (21.9% and 3.7% respectively) (World Bank). Growth is one of the most sensitive measures of the pre-adoptive caretaking environment being influenced not only by nutrition, but also the quality of care and number of disruptions in the caregiving environment (Johnson & Gunnar, 2011; Johnson et al., 2018). One factor contributing to this finding of better growth in CAF-H could be that under Haitian law, prospective adoptive parents who are not adopting relatives must adopt from the small number of crèche/orphanages that Haiti’s Central Authority Institut du Bien-Être Social et de Recherches has licensed and rated a “green light” facility (U.S. Department of State-Bureau of Consular Affairs, 2022). Another reason may be that Haiti is an outlier in terms of orphanage funding. In 2017, Lumos Foundation estimated that the approximately 30,000 children in Haiti’s over 700 mostly private orphanages received upward of $100 million in annual funding primarily from US faith-based donors (Vernaelde & Guillaume, 2017). This remarkable level of support was equivalent to nearly one-half of all U.S. foreign aid to Haiti planned for 2017. Better than expected growth suggests that generous NGO support of Haiti’s best orphanages may have a positive effect on children in out-of-home placement, at least in the area of growth. Nevertheless, individuals and organizations involved in out-of-home care in Haiti should also have the resources to address at least some of the deficiencies that lead to the problems identified above.

This study was limited by a number of factors. The study was retrospective in design, which constrained the ability to uncover cause and effect relationships. Unmeasured confounding likely exists since there are a variety of factors over an adoptee’s life that can influence medical outcomes. While our study set out to capture the health of CAF-H following the 2010 earthquake during a ten-year period, we cannot disregard the possibility of time-based differences between the health of children adopted at the beginning versus the end of the decade. To preserve power, however, we did not account for time differences in our analysis, which may have biased our results. Only CAF-LA and CAF-A were included as controls, meaning that generalizations between Haitian and non-Haitian international adoptees cannot be made. In addition, some variables including hepatitis A, vitamin D, and T4 had significant amounts of missingness (10–25%) and may have biased our results. There is also missingness for race/ethnicity which relies on data collected in the medical chart provided by foster or adoptive parents. Lastly, despite using a matched cohort to increase precision, we may have failed to detect additional differences between CAF-H, CAF-LA, and CAF-A due to being underpowered.

Conclusion

There are numerous health conditions that require careful screening by primary health care providers once patients have joined their adoptive families. With the exception of growth impairment, this study highlights significant differences in the magnitude of common medical problems between CAF-H, CAF-LA, and CAF-A. Providers should ensure that screening tests are done for vaccine immunity (irrespective of the child’s immunization records), anemia, tuberculosis, parasites, and lead, all of which are particularly important for incoming CAF-H. It is important that providers, both prior to and after adoption, are aware of the prevalence of these specific medical conditions so that appropriate care and treatment can be provided for Haitian children that join families through adoption.