Hair cortisol concentrations among urban and rural-dwelling mother–child dyads, La Romana, Dominican Republic

Abstract

Hair cortisol concentrations (HCC) were studied in mother–child dyads of La Romana, Dominican Republic (DR), a low-income city, and of the surrounding bateyes, sugarcane plantation villages with inhabitants frequently of Haitian descent. Populations of low socioeconomic status (SES) experience hypothalamic-pituitary-adrenal (HPA) axis dysregulation. Urban communities may be increasingly exposed to stressors such as crime and concentrated poverty whereas rural communities may be devoid of important community resources. As a result, the experience of stress in poverty may differ by place of residence. The goal of this study was to examine differences in HCC among urban and rural-dwelling mother–child dyads in socioeconomically disadvantaged communities surrounding La Romana, DR. Forty-five mother/child dyads were enrolled in La Romana and 45 at several bateyes surrounding La Romana. Mothers were ≥18 years and children were between 7 and 14 years. Mothers self-reported perceived stress and demographic factors. Hair samples were collected from mothers and children, and HCC was assessed using enzyme-linked immunosorbent assays. General linear models examined associations between socioeconomic factors and HCC, and between maternal and child HCC. HCC were measured in 88 maternal and 87 child samples (N = 175). Mothers living in a batey had higher HCC than those living in La Romana (p = 0.001). HCC was positively associated among maternal–child dyads (p = 0.001). Further, Haitian-born mothers as a population who frequently live in a rural batey experienced higher HCC (p = 0.025) that may partially be explained by discriminatory practices in the DR. This research helps to elucidate the impact of urban and rural environmental settings on HCC.

Lay summary

• This study focuses on chronic stress, measured by hair cortisol levels, among a low-income population of Dominican and Haitian mother–child pairs who live in urban and rural settings. We found that Haitian-born mothers, who frequently live in a rural batey, had higher hair cortisol levels than Dominican born mothers. Hair cortisol levels between mothers and their children were positively associated. This study addresses the impact of urban and rural environments on the stress response among socioeconomically disadvantaged persons living in an upper middle income country who bear an excessive burden of psychosocial stress.

Keywords:

• Chronic stress
• hair cortisol
• urban
• rural
• HPA-axis
• Dominican Republic

Introduction

The Dominican Republic (DR) is classified as an upper middle income country by the World Bank though in 2020, it is estimated that 21% of its residents live in poverty and up to 13.8% of its residents earn less than US $5.50 per day (WorldBank, 2020). The city of La Romana ranks among the highest for homicide (20 per 100,000) and reported robbery (Osac.gov, 2019). The batey (i.e. remote sugarcane plantation villages) inhabitants around La Romana are impoverished and have limited access to medical care (Crouse et al., 2010). The batey residents are a marginalized, rural population of undocumented Haitian migrants or self-identified Dominicans frequently of Haitian descent. Batey residents frequently fall victim to discriminatory practices in the Dominican Republic and given the remote nature of their rural residence, they suffer from lack of medical, educational, and community resources.

Living standards such as low socioeconomic status (SES), are associated with chronic stress and hypothalamic-pituitary-adrenal (HPA) axis dysregulation (Clearfield et al., 2014; Hajat et al., 2010; Sheridan et al., 2013; Vaghri et al., 2013). The role of the HPA axis is to facilitate the neuroendocrine adaptation of the stress response. Populations of low SES experience disproportionate exposure to chronic stress driven by environmental and social factors such as poor/inadequate nutrition, lifestyle and healthcare limitations, and increased exposure to discrimination, violence, toxins, pollutants, and noise (G. W. Evans, 2006). These stressors have been demonstrated to elicit a negative emotional response and a pronounced inflammatory response in low SES populations (Chen et al., 2003; Chen et al., 2006). A strong association also exists between parental SES and indicators of child health (Clearfield et al., 2014; G. W. Evans & English, 2002; G. W. Evans & Kim, 2010; Vliegenthart et al., 2016). Moreover, children and adolescents in families of low SES have higher levels of cortisol measured by hair cortisol concentration (HCC) (Vliegenthart et al., 2016) likely mediated by chronic stress. A strongly heritable component of HCC has been demonstrated in twin studies (Rietschel et al., 2017), and a positive association between maternal and child HCC exists (Olstad et al., 2016) in populations of low SES (Hollenbach et al., 2019; Liu et al., 2017), as maternal stressors are often associated with HPA dysregulation in offspring (Dunkel Schetter, 2011).

Exposure to chronic stress leads to adverse physical and mental health outcomes. Commonly known as the stress hormone, cortisol is a downstream effector of the HPA axis, released as an adaptive response to a challenging environment. Cortisol is an end product of the HPA axis and can therefore serve as a measure of HPA dysregulation. HPA dysregulation has been linked to unfavorable, long-term health outcomes such as depression, increased cardiovascular risk, cancer and metabolic diseases (Volden & Conzen, 2013), increased vulnerability to asthma exacerbations (Sandberg et al., 2004; Sandberg et al., 2000), and immune system suppression (Clemens et al., 2020; Doom & Gunnar, 2013). Cortisol can be measured in blood serum, urine, saliva and hair.

Blood serum, urine, and saliva are acute/short term measures of cortisol. Therefore, they are subject to large fluctuations due to time of day, caffeine intake, menstrual cycles, and other factors, whereas HCC reflects cumulative exposure and are not influenced by these daily variations. HCC can be used as a retrospective biomarker of cumulative cortisol exposure over a period of months rather than at a single time point (Stalder et al., 2017; Staufenbiel et al., 2013). Furthermore, HCC has been shown to correlate with 24-h urinary cortisol (Sauve et al., 2007) and salivary cortisol across a month (Short et al., 2016). Therefore, HCC is favored over other biomarkers of the stress response as a painless, noninvasive, unique way to measure dysregulation of the HPA axis (Petimar et al., 2020).

Traditionally, research addressing poverty and stress focuses on the urban poor. However, recent evidence exists regarding the geography of poverty and suggests the stress response may be dysregulated in both urban and rural disadvantaged populations through different means. For example, urban communities may be increasingly exposed to stressors such as noise, toxin exposure, and violent crime triggering the stress response system whereas rural populations may suffer from limited availability and accessibility of family and community resources leading to activation of the stress response system (Miller et al., 2019). As a result, the experience of stress in poverty may differ by place of residence.

This study aimed to fill a gap in knowledge by examining how HCC differs among economically disadvantaged families living in rural versus urban environments. Previous studies have examined socioeconomic determinants of HCC in disadvantaged populations (Hollenbach et al., 2019; Ursache et al., 2017), among child and adolescent populations (Gray et al., 2018), and in populations from a low/middle income country (Lembcke et al., 2020). To our knowledge, there is one other study that has examined how HCC levels may vary across urban and rural areas though this was in a population of predominantly high SES (Evans et al., 2019).

The results of this study will further assess 1) differences in HCC among urban and rural-dwelling mother–child dyads and 2) the association between child and maternal HCC in socioeconomically disadvantaged communities surrounding La Romana, DR. We hypothesize increased HCC among participants in rural bateyes compared to an urban environment and a positive correlation between maternal and child HCC.

Methods

Participants

Mothers/guardians and their children who presented to Good Samaritan Hospital in La Romana, DR, or who lived in several of the bateyes surrounding La Romana, were invited to participate in the study. Mothers/guardians were eligible if they were ≥18 years of age and had at least one child between 7 and 14 years. Dyads were excluded if the mother and/or her child reported oral corticosteroid use within four weeks prior to enrollment or had hair length shorter than 3 cm. All participants provided informed consent and children provided assent. The goal was to enroll 45 dyads from La Romana, and 45 from the bateyes. The study was approved by the Quinnipiac University Institutional Review Board in May 2018 in partnership with Hospital El Buen Samaritano in La Romana, DR. Data collection was performed between June and July 2018.

Participants enrolled in the urban setting of La Romana and the rural setting of the bateyes were recruited with the same strategy. Participants from La Romana were recruited in the waiting area of Hospital El Buen Samaritano. Participants from the bateyes were recruited during medical missions to the villages organized by Hospital El Buen Samaritano. All participants were randomly approached while children and/or guardians were waiting to see healthcare providers for either well or sick visits. No participant was recruited in an emergency setting.

Demographic data were collected on all dyads including nationality, language, place of birth, age, sex, education, household income, and BMI (kg/m²). Height and weight were obtained during the clinic visit using a scale and tape measure. A list of all current medications taken by the child was also collected. For all interactions, a facilitator was present who could speak both Spanish and Kreyòl.

Short Form Perceived Stress Scale (PSS-10)

The Short Form Perceived Stress Scale (PSS-10) measures the degree to which respondents perceive their lives over the past 4 weeks as unpredictable, uncontrollable, and overburdened (Cohen & Williamson, 1988). The PSS-10 was developed for community samples with at least an eighth grade reading level. The 10-item version was administered during the single study visit. Items were answered on a 5-point Likert scale (0 = never to 4 = very often) based on how often a participant felt in the past month. Responses were summed, with four positively worded items reverse-coded, for a total score ranging from 0 to 40.

Hair cortisol

Hair samples consisting of ∼100 hairs were taken from the posterior vertex of the scalp. The hair sample was tied off near the scalp end before cutting to indicate the scalp end and then stored in an envelope and plastic bag at room temperature until analysis. The proximal three cm of hair, reflecting the most recent three months of hair growth and stress exposure, was cut and weighed on an electronic scale and transferred to disposable screw-top glass vials. Hair samples were washed three times with isopropanol and allowed to dry completely. One mL of HPLC-grade methanol (MeOH) was added to each vial and cortisol was extracted overnight at room temperature. After the overnight extraction, MeOH was decanted into disposable glass tubes and extracts were dried at 37 °C in a vacuum centrifuge (Sorvall, Thermo Fisher Scientific, Waltham, MA) for at least 1.5 hours. Extracts were reconstituted in 250 μL of phosphate buffered saline (PBS), pH 8.0 and vortexed.

Cortisol quantification was conducted in duplicate along with standards and quality controls using the ELISA salivary cortisol kit according to the manufacturer’s instructions (Arbor Assays, Ann Arbor, MI). The interpolated concentrations were converted to picograms (pg) cortisol and then corrected to the mass (mg) of hair analyzed (Hollenbach et al., 2019). As is typical for cortisol data (Staufenbiel et al., 2015), raw HCC were not normally distributed and thus raw HCC were transformed to reduce skewness and kurtosis by taking the natural log of the raw HCC plus one. The ln-transformed values were used in all analyses. Intra-assay coefficient of variation for this assay is <25% and inter-assay coefficient of variation for this assay are <16%. Ninety mothers and 90 children provided cortisol data. However, only 88 of the maternal cortisol data and 87 of the child cortisol data were included in analyses due to cortisol concentrations being below the assay’s standard curve after two analysis attempts.

Data analysis

First, univariate analyses were conducted to describe the demographics of the participants. Pearson’s chi-square test was used to compare the categorical demographic variables between La Romana and batey groups. A list of demographic variables was generated after review of the literature which determined which socioeconomic variables may confound HCC. Hair bleach, age, sex, BMI (Staufenbiel et al., 2015), education (Hollenbach et al., 2019), and ethnicity (Ursache et al., 2017) have all demonstrated an association with maternal and/or child HCC. Mother’s birth place and nationality were included as these factors have implications for discriminatory practices in the DR. Fisher’s exact test (two-tailed) was used for variables with cell counts less than five. Student’s independent t-test was used to compare continuous demographic variables between groups.

Bivariate general linear models were used to examine the association between maternal socioeconomic factors and PSS with maternal HCC and child HCC. Variables remained in the subsequent adjusted model if they were statistically significant in the bivariate analysis (p ≤ 0.05) with maternal HCC. Third, a multivariable general linear model was used to examine associations between maternal HCC and child HCC along with significant socioeconomic variables. We report the results in terms of Exponentiated Regression Coefficients (Exp(B)), since the dependent variable HCC was log-transformed, and this which can be interpreted as the multiplicative change in HCC for a one unit increase in the corresponding predictor variable (holding all other variables constant). An association with a p-value smaller than 5% was considered statistically significant. Analyses were completed using the Statistical Package for the Social Sciences (SPSS) version 26.

Results

The target enrollment was achieved with 90 mother/child dyads enrolled between June and July 2018; 50% from La Romana and 50% from the bateyes surrounding La Romana. A total of 180 participants were enrolled in this study. Hair cortisol levels were measured in 88 maternal and 87 child samples due to cortisol levels being below the assay’s standard curve (N = 175). Most caregiver/parents were the biological mother (93%) and Spanish-speaking (57%). The mean age among mothers was 35 (±9 SD) years and among children was 10 (±2 SD) years old; 98% of children were female and 94% were Dominican (Table 1). Sixty-nine percent of mothers were born in the DR and 60% identified as Dominican. Fifty percent of mothers received an education beyond primary school. The average household income was $126 per month. Half of participants report being in a relationship and living together (54%). On average, mothers were overweight with a BMI of 27.3 (±5.17 SD). Children had an average BMI of 17.2 (±3.2 SD). Mothers reported being moderately stressed with an average PSS score of 18.40 (±5.65 SD). The median HCC for mothers was 5.39 pg/mg (IQR = 3.04 to 11.68 pg/mg) and for children was 4.03 pg/mg (IQR = 2.25 to 8.79 pg/mg).

Table 1. Demographic characteristics of parental respondent and child, N = 180.

Parental respondent demographics (n = 90)	All	La Romana	Batey	p Value
	n = 90, n (%)	n = 45, n (%)	n = 45, n (%)
Age (years, mean ± SD^a)	35 ± 9	36 ± 9	35 ± 9	0.585
Relationship to child				0.434
Natural mother	83 (93%)	43 (96%)	40 (91%)
Step/guardian	6 (7%)	2 (4.4%)	4 (9.1%)
Language				<0.001
Spanish	51 (57%)	44 (98%)	7 (16%)
Kreyòl	24 (27%)	1 (2%)	23 (51%)
Spanish and Kreyòl	15 (33%)	0 (0%)	15 (33%)
Place of birth				<0.001
Haiti	28 (31%)	1 (2%)	27 (60%)
DR	62 (69%)	44 (98%)	18 (40%)
Nationality				<0.001
Haitian	27 (30%)	1 (2%)	26 (58%)
Haitian and Dominican	9 (10%)	3 (7%)	6 (13%)
Dominican	54 (60%)	41 (91%)	13 (29%)
Education level				<0.001
No school/primary (grade 1–8)	45 (50%)	7 (16%)	38 (84%)
Secondary/university (grade >8)	45 (50%)	38 (84%)	7 (16%)
Household income median (IQR)/month	$126 (45 to 252)	$243 (180 to 365)	$72 (54 to 99)	<0.001
Marital status				0.017
Married	23 (26%)	14 (31%)	9 (20%)
Relationship living together	49 (54%)	18 (40%)	31 (69%)
Single/divorced/living a part	18 (20%)	13 (29%)	5 (11%)
Body mass index (BMI, mean ± SD)	27.3 ± 5.17	27.9 ± 5.12	26.8 ± 5.23	0.332
PSS-10 (mean ± SD)	18.4 ± 5.65	18.4 ± 7.09	18.3 ± 3.86	0.905
ln HCC (pg CORT per mg hair), median (IQR^b), n = 88^c	1.68 (1.11 to 2.46)	1.35 (0.91 to 1.85)	2.05 (1.41 to 2.86)	0.001
Child demographics (n = 90)
Age (years, mean ± SD)	10 ± 2	10 ± 2	10 ± 3	0.474
Sex, female	88 (98%)	43 (96%)	45 (100%)	0.153
BMI (mean ± SD)	17.2 ± 3.2	17.9 ± 3.45	16.6 ± 2.75	0.049
Race/ethnicity				0.242
Haitian	3 (3%)	0 (0%)	3 (7%)
Dominican	85 (94%)	43 (96%)	42 (93%)
ln HCC (pg CORT per mg hair), median (IQR^b), n = 87^c	1.21 (1.17 to 2.28)	1.62 (1.18 to 2.35)	1.62 (1.10 to 2.28)	0.954

Bolded type denotes significance.

^aSD: standard deviation.

^bIQR: interquartile range.

^cDue to cortisol levels being below the assay’s standard curve.

Mothers from the batey were more likely to be Kreyòl speaking (p < 0.001), Haitian-born (p < 0.001), identify as Haitian (p < 0.001), have a primary school education or less (p < 0.001), have a lower household income per month (p < 0.001), and be in a relationship living together (p = 0.017) than mothers in La Romana (Table 1). Mothers living in a rural batey had higher cortisol levels than those living in the city of La Romana (p = 0.001). Children from the batey had a lower BMI than children from La Romana (p = 0.049). There were no differences between mothers from the bateyes as compared to La Romana with respect to maternal age, relationship to child, BMI, or PSS score. There were no differences between children from the bateyes as compared to La Romana with respect to child age, sex, race/ethnicity, or cortisol level (Table 1).

Unadjusted, bivariate analyses assessing the association between demographic factors and maternal HCC demonstrate that mothers who bleached their hair had lower cortisol levels (p = 0.023) (Table 2). Mothers born in Haiti had 2.4 times higher cortisol levels than those born in the DR (p < 0.001). Mothers who identified as Haitian had 2.1 times higher cortisol levels than those who identified as only Dominican (p = 0.001). Mothers living in a rural batey had 2.0 times higher cortisol levels than those living in the city of La Romana (p < 0.001). Mothers receiving more education and with a higher income had lower cortisol levels (p = 0.002; p = 0.021, respectively). A positive association between a mother’s and child’s HCC was observed (p = 0.006). Marital status was not significantly associated with maternal HCC (p = 0.088). A mother’s BMI and perceived stress score was not significantly associated with her cortisol levels.

Table 2. Associations of demographic variables, perceived stress, and child HCC with maternal HCC.

	Unadjusted				Adjusted
Covariates	b	95% CI	Exp (b)	p Value	b	95% CI	Exp (b)	p Value
Mother’s age	0.02	(−0.00,0.40)	1.02	0.105
Bleached hair, yes	−0.47	(−0.87, −0.06)	0.63	0.023	−0.12	(−0.51, 0.30)	0.90	0.605
Place of birth
DR	REF				REF
Haiti	0.88	(0.48, 1.28)	2.40	<0.001	0.67	(0.08–1.26)	1.95	0.025
Nationality
DR	REF
Haitian and DR	0.52	(−0.12, 1.15)	1.68	0.109
Haitian	0.74	(0.32, 1.17)	2.11	0.001
Residence
Urban/La Romana	REF	REF
Rural/Batey	0.69	(0.31, 1.06)	1.99	<0.001	0.43	(−0.07, 0.93)	1.54	0.09
Education level	−0.07	(−0.11, −0.03)	0.94	0.002	−0.01	(−0.07, 0.05)	0.99	0.832
Income (1,000 per unit)	−0.02	(−0.03, −0.00)	0.98	0.021	−0.00	(−0.02, 0.01)	1.00	0.712
Marital status				0.088
Single	REF
Relationship living a part	2.35	(−0.65, 5.34)	10.44	0.125
Relationship living together	0.25	(−0.15, 0.66)	1.29	0.221
Divorced	1.04	(0.10, 1.98)	2.83	0.029
Married	0.50	(−0.09, 1.09)	1.65	0.094
BMI	0.01	(−0.03, 0.05)	1.01	0.665
PSS-10	−0.02	(−0.05, 0.02)	0.98	0.347
Child ln HCC	0.35	(0.10, 0.60)	1.42	0.006	0.41	(0.17, 0.65)	1.50	0.001

Bolded type denotes significance.

Coefficients presented in the table are unstandardized coefficients.

Variables remained in the multivariate model if they were statistically significant in the bivariate analysis (p ≤ 0.05) with child HCC.

Analysis of the relationship between maternal and child HCC demonstrated a significant correlation (r(85) = 0.284, 95% bootstrap CI: 0.01–0.50). In the multiple linear regression, variables that were significant in bivariate analyses and did not significantly linearly correlate with one another as measured by multicollinearity studies were included. We determined that place of birth and nationality significantly correlated with one another and as a result, we adjusted for place of birth in the multivariate model. Adjusting for mother’s income, education, use of bleach, place of birth, and place of residence, mothers born in Haiti had a twofold higher hair cortisol level than mothers born in the DR (p = 0.025). Additionally, mother’s and children’s HCC were significantly and positively associated (p = 0.001) (Table 2) in the adjusted model.

Bivariate analyses assessing stress and social determinants of child HCC demonstrate that children with older mothers (p = 0.010) and mothers with higher cortisol levels (p = 0.006) had higher cortisol levels (Table 3). This association persisted in adjusted multiple linear regression analysis.

Table 3. Stress and social determinants of child HCC.

Covariates	Unadjusted				Adjusted
	b	95% CI	Exp (b)	p Value	b	95% CI	Exp (b)	p Value
Mother’s age	0.02	(0.01,0.04)	1.02	0.010	0.02	(0.01, 0.04)	1.02	0.012
Place of birth
DR	REF
Haiti	0.11	(−0.24, 0.47)	1.12	0.542
Nationality
DR	REF
Haitian and DR	−0.04	(−0.60, 0.51)	0.96	0.876
Haitian	0.10	(−0.27, 0.47)	1.11	0.589
Residence
Urban/La Romana	REF
Rural/Batey	−0.01	(−0.34, 0.32)	1.01	0.953
Education level	0.001	(−0.12, 0.12)	1.00	0.987
Income (1,000 per unit)	−0.002	(−0.02, 0.01)		0.769
Marital status
Single	REF
Relationship living a part	0.76	(−0.37, 1.89)	2.14	0.186
Relationship living together	0.03	(−0.42, 0.49)	1.03	0.889
Divorced	0.28	(−0.85, 1.41)	1.33	0.624
Married	0.37	(−0.15, 0.89)	1.45	0.160
BMI	0.02	(−0.01, 0.05)	1.02	0.207
PSS-10	0.01	(−0.02, 0.04)	1.01	0.440
Maternal ln HCC	0.23	(0.06, 0.39)	1.26	0.006	0.19	(0.03, 0.35)	1.21	0.020

Bolded type denotes significance.

Coefficients presented in the table are unstandardized coefficients.

Variables remained in the multivariate model if they were statistically significant in the bivariate analysis (p ≤ 0.05) with child HCC.

Discussion

In this study of maternal/child dyads from socioeconomically disadvantaged communities, we did not observe a significant association between place of residence and HCC. However, we observed that Haitian-born mothers, a population who frequently reside in rural bateyes, demonstrate higher hair cortisol levels. We also observed a positive association between maternal and child HCC regardless of urban or rural residence. Maternal stressors are often associated with HPA dysregulation in offspring (Dunkel Schetter, 2011), and in close pairs, such as a mother and child, stressors experienced by one member may influence the other (Hollenbach et al., 2019; Waters et al., 2014). Some studies examining maternal–child HCC demonstrate conflicting results (Boeckel et al., 2017; Olstad et al., 2016; Ursache et al., 2017). However, a recent study among an urban-dwelling, minority sample demonstrated a strong, positive association between maternal and child HCC (Hollenbach et al., 2019) likely through a combination of genetic and environmental influences. Our results support this conclusion and demonstrate this association persists in an understudied, low income, urban and rural-dwelling population from an upper middle income country. Future research should address the clinical significance of this association in children, however, our research may help to provide physiologic evidence to explain the intergenerational correlation of economic status. Poverty is associated with increased stress and repeated stress exposure has deleterious cognitive effects (Aizer et al., 2016) which has negative implications for adult economic status and achievement. As a result, HPA dysregulation among mothers and their offspring can potentially be one biologic mechanism to explain the persistence of poverty often seen among generations in concert with structural and social barriers. Barriers of significant importance in the context of our population include lack of educational and economic opportunity, discrimination, resource shortages, and life adversities to name a few.

We examined a sample of Haitian and Dominican born participants and observed higher HCC among Haitian born participants. The majority of these participants reside in an impoverished, rural environment. The island of Hispaniola is shared by the DR and Haiti. Agriculture is an important sector of the DR’s economy and sugarcane is the principal agriculture product. In order to harvest sugarcane, the DR recruits Haitian workers to work and live in agricultural shantytowns known as bateyes. In some cases these workers earn less than $1/day and work 14 hours/day. This economic situation may explain why Haitian participants in our sample demonstrated higher HCC than their Dominican counterparts. Moreover, the circumstance of persons of Haitian-descent is further complicated by discriminatory practices in the DR. Neural processing of stress is altered in stress regulatory areas, such as the ventral striatum and and perigenual ACC, among ethnic minorities who experience higher levels of perceived discrimination (Tost et al., 2015). Moreover, it is documented that individuals who experience chronic discrimination demonstrate higher levels of chronic stress, psychopathology (Cohen et al., 2007), and higher HCC (O'Brien et al., 2017).

Anti-haitianismo (Anti-Haitian) is as old as the inception of the DR (Paulino, 2006), and maintains a culture of discriminatory practices rooted in race, nationality and class (Keys et al., 2019). In 2013, La Sentencia stripped rights of citizenship to 200,000 Dominicans of Haitian-descent. This limits their access to healthcare, education, jobs and amplifies discrimination throughout the country. Perceived discrimination, measured by Estimated Discrimination Score (EDS), is common among batey residents. Additionally, the EDS score was strongly associated with Haitian birth and descent (Keys et al., 2019). Most Haitian-born persons live in the east of the DR (Keys et al., 2019) and the bateyes visited in this study were located in the east. Perceived discrimination contributes to a chronically activated stress response (Cohen et al., 2007) which is supported by our finding that Haitian-born persons experience higher HCC. Though the majority of this population lives in a rural batey, more research is needed regarding the influence of rural and urban living on the stress response. In our study, being Haitian born was more predictive of increased hair cortisol levels than living in a rural environment.

Our study demonstrates higher child hair cortisol levels among children with older mothers. A systematic review assessing determinants of hair cortisol concentration in children does not report an association between maternal age and child hair cortisol levels (Gray et al., 2018). However, research shows that older maternal age may be associated with an increased risk of stress, depression, and anxiety in female offspring (Tearne et al., 2016). The population in this study is mostly female and as a result, the increased hair cortisol levels seen in offspring of older mothers are consistent with previous research in this area.

There are several strengths to this study. First, this study is the first to compare the hair cortisol levels of socioeconomically disadvantaged persons living in both rural and urban environments in a country where they bear a disproportionate burden of psychosocial stress. Second, this study measures chronic cortisol release to demonstrate long-term, retrospective stress exposure.

Limitations include the inability to infer directionality or causality since this was a cross-sectional study. The bateyes that were chosen were closer to major roads, therefore, the results may not be generalizable to persons living in more remote bateyes. Because the enrollments were conducted in both Spanish and Haitian Kreyòl there is potential for measurement bias in how questions were asked, however, the same interpreter was used to minimize this bias. Parenting practices and family related factors that may moderate a child’s psychopathological response to adversity were not assessed (Lembcke et al., 2020). The study included mostly females due to length of hair and cannot account for differences seen in gender. Lastly, this study did not assess child’s pubertal stage which can have implications on the maturation of the HPA axis.

In conclusion, this study demonstrates that HCC is positively associated among socioeconomically disadvantaged maternal–child dyads living in both urban and rural environments likely through a combination of genetic and environmental influences. Further, Haitian-born participants, a population who frequently reside in a rural environment, experience higher HCC that may be explained by discriminatory practices and perceived discrimination in the DR. This research helps to elucidate the impact of urban and rural environmental settings on hair cortisol concentrations among populations of low SES.

Future studies in this population should assess how HCC correlate with adverse childhood health outcomes. They should focus on the correlation of HCC with other diseases in which this population carries a high burden of disease such as asthma, hypertension, atopic dermatitis/eczema, gastrointestinal, fungal and skin infections. This research will help elucidate the physiological mechanisms underlying the disease burden carried by populations experiencing low SES so that effective strategies, which may differ by place of residence, can be implemented to prevent and mitigate adverse clinical outcomes.

Acknowledgements

Thank you to all the mothers and their children who contributed to this research. We would also like to thank Hospital El Buen Samaritano for allowing us to enroll patients at their hospital and their medical missions to the bateyes. A special thank you to Moises Sifren, Mario Sifren, and Maggie for their support that made it possible to conduct this study.

Disclosure statement

The authors report no conflicts of interest.

Additional information

Funding

This research was supported by a grant from the Frank H. Netter MD School of Medicine Summer Research Fellowship.

Notes on contributors

Nicole Gherlone

Nicole Gherlone Conceptualization, Methodology, Validation, Formal Analysis, Investigation, Resources, Data Curation, Writing, Visualization.

David R. Hill

David R. Hill Administration, Conceptualization, Visualization, Resources, Writing - Review & Editing, Methodology.

Richard Feinn

Richard Feinn: Formal Analysis, Data Curation, Writing - Review & Editing, Software, Methodology.

Jessica P. Hollenbach

Jessica P. Hollenbach Conceptualization, Visualization, Supervision, Data Curation, Formal Analysis, Writing - Review & Editing, Resources.