1.1 The Nurturing Care Framework

The Nurturing Care Framework (NCF), developed by the World Health Organization (WHO) and its partners in 2018, sums up decades of early child development (ECD) research in a comprehensive approach to promoting children’s optimal development and well-being [7]. The NCF focuses on five necessary components of early care: 1) adequate nutrition, 2) good health, 3) security and safety, 4) responsive caregiving, and 5) opportunities for early learning. Adequate nutrition involves vitamin supplementation during pregnancy, exclusive breastfeeding to at least age six months when possible, and a diverse diet free of toxins and parasites. Good health involves a caregiver protecting a child’s physical and emotional health, protecting a child from environmental dangers, having good hygiene practices, and providing appropriate medical care when needed. Security and safety involve helping young children feel safe, even when they misbehave, and protecting them from physical danger and emotional distress by preventing accidents such as choking or falling and avoiding physical punishment. The caregivers’ responsibilities involve an awareness that a young child can experience intense fear when beaten or threatened with abandonment, and that this fear can interfere with a child’s ability to develop trusting relationships in the future.

The NCF’s emphasis on responsive caregiving distinguishes it from other early childhood models. Responsive caregiving, defined in the model as “observing and responding to children’s movements, sounds and gestures, and verbal requests”, begins early and provides the foundation for neurodevelopment and learning. Responsive care is a buffer for children experiencing early adversity [8,9]. This form of caregiving is particularly relevant in settings where children are impacted by adversity in daily life: A as a growing body of longitudinal research demonstrates that responsive caregiving plays a role in mediating the effect of poverty on child development [10]. For example, in 2019, Luby et al followed preschoolers for 5–10 years to assess how poverty in pre-kindergarten impacted brain development in elementary school. They found that poverty was associated with smaller white and cortical gray matter and hippocampal and amygdala volumes, the areas of the brain collectively responsible for translating emotion into long-term memory. Parental caregiving support, measured observationally during a preschool visit, mediated this relationship [10].

Closely related to responsive caregiving is early learning. Infants need visual, auditory, and tactile stimulation through light, sounds, visual patterns, objects, and tactile experiences in the environment [11,12]. Most of this learning will occur in the context of consistent and affectionate interactions with adult caregivers. Together with adults, infants will explore their surroundings, seek out visual and tactile stimulation, share attention on objects with others, orient towards and listen to voices and singing. As they grow older, they may also participate in preschool or social interaction with other young children in stimulating environments [13]. Given their importance, researchers have integrated early learning and responsive caregiving into parent and caregiver/educator coaching interventions in settings of poverty and seen positive ECD outcomes [14].

The NCF employs a life course perspective within which the critical period of development is from pregnancy to age five [13,15]. This is based on research showing that, during this time, a ‘window of opportunity’ allows the brain to construct vital internal structures. The brain develops thousands of neural connections, and its plasticity allows children to absorb vast amounts of new information and experience rapid brain growth [16]. Around age five, weaker neural connections are discarded, and stronger ones are strengthened to allow for more efficient brain function in a process called ‘pruning.’ During the window of rapid growth, it is crucial for children to develop basic brain architecture and processes that they will then strengthen, expand, and build upon in primary school and into adulthood [16].

Further work to understand responsive caregiving and opportunities for early learning in the DR, therefore, could help ease the impact of poverty on children’s ability to learn and thrive. This study aims to develop knowledge about how mothers living in poverty in the DR provide their children with nurturing care, according to its conceptualization within the Nurturing Care Framework. In particular, we explore caregivers’ beliefs and practices regarding responsive caregiving, opportunities for early learning, and safety and security in the home.

1.2 Culture in ECD research

In 2022, The Lancet Child and Adolescent Health published a response to a paper by McCoy et al claiming almost 75% of children in lower- and middle-income countries were not receiving minimal nurturing care [17]. The response, titled “Different is Not Deficient: Respecting Diversity in Child Development,” called attention to the importance of appreciating difference in cultural childrearing practices [18]. It points out that NCF indicators used in large-scale assessments may be missing facilitators to early development in low- and middle-income countries and thereby inadvertently stigmatizing these families. For instance, McCoy et al determined that early learning is present if the family owns a book or toy, and the child is attending an early child education program [17]. These indicators don’t include other early learning opportunities, such as playing with older children, singing, playing with everyday household items or recycled toys, and reading non-print material such as signs. McCoy et al point out that the NCF is a useful framework and should be interpreted within each cultural setting with care. Likewise, ECD assessment tools fall short of defining domains that reflect daily life in settings with high concentrations of poverty such as the Dominican Republic [19]. Understanding these domains using culture-specific data will allow for feasible action steps toward achieving goals such as ensuring lifelong learning and eradicating poverty set by the United Nations as part of the 2030 Sustainable Development Goals [20].

1.3 ECD in the Dominican Republic

The DR was chosen for this study due to the high prevalence of poverty and other socioeconomic hardships impacting neurodevelopment in children. The Dominican Republic (DR), a Spanish-speaking Caribbean nation on the island of Hispaniola, has a population of 11.2 million people with 49.3% of children living in poverty, according to a 2019 government report [21]. In the DR, the prevalence of neurodevelopmental delay, defined as not reaching full developmental potential by age five, is high at about 14.5% [22], compared to 4.6% in the United States [23]. This reality is associated with school achievement: based on the 2014 and 2019 Multiple Indicator Cluster Studies (MICS) data, primary students in the DR perform among the lowest in Latin America on math and language evaluations [19]. Further, in 2022, Sánchez-Vincitore and Castro found that socioeconomic gradient predicted ECD outcomes among children 36–59 months of age and that a psychosocial model, including opportunities for early learning components, mediated this relationship [24].

Since 2014, the government provides low-income children with early stimulation, home visits, and preschool, and women with high-risk pregnancies accompaniment to local hospital visits, identification registration, and government-sponsored social programs providing a minimum monthly salary, vaccinations, and food baskets through the Programa Integral a la Primera Infancia de Base Familiar y Comunitaria [25]. The Instituto Nacional de Atención Integral a la Primera Infancia (INAIPI) provides preschool services, including preschool for children ages 45 days to five years, similar to the Head Start program in the United States. In 2021, INAIPI reached 215,933 children ages zero to five years [26].

2.1 Recruitment

Eligibility criteria included having a child under five years of age, reporting an income less than $3 a day, and living in and around Santo Domingo, the capital of the DR. We recruited participants from a contact list of women who had participated in the Pregnant Women-Centered Care in the Dominican Republic Project (Project MAC). This was a Tulane study funded by UNICEF that identified obstetric violence during childbirth through birth observation in two public hospitals in Santo Domingo and a neighboring province [28]. All participants had agreed to be contacted by phone for future studies. For the UNICEF study, birth observers were present for all births during a two-month period in October and November 2019 without any other exclusion criteria. Approximately three years later, the first author, together with two research assistants, sorted the 169 addresses of women who lived in Santo Domingo by distance from the Laboratorio Laboratorio de Neurocognición y Psicofisiología (NeuroLab) at the Universidad Iberoamericana, where the team was stationed. The research team called phone numbers beginning with those who lived closest and working out. Approximately 112 phone numbers were out of service or no longer belonged to the previous owner, 36 did not respond to contact attempts (one text and two phone calls for each number), seven were initially interested but did not respond to further requests to set a date, one said she did not have time for the study, and two had lost their child since birth and had no other young children. Thirteen women from the previous study agreed to participate in interviews, all from the province of Santo Domingo. To augment our interviews and provide contrast, we used snowball sampling to recruit an additional 12 women from a town called Yamasá, located about an hour from Santo Domingo in the province of Monte Plata. Ultimately, we conducted interviews with 25 women who met eligibility criteria, 10 women were from Los Mina in the province of Santo Domingo, three from Boca Chica in Santo Domingo, and 12 from Yamasá.

2.2 Participant consent

All participants consented to take part in the study and authorized the utilization of their de-identified data for research purposes. Those who joined the study via Zoom provided verbal consent, which was captured and recorded during the interview. The researcher obtained consent signatures from these participants by asking them to permit and witness an electronic rendering of their name on the consent form. We recorded interviews using the voice recorder application on the researcher’s laptop or phone. One participant preferred not to have their interview recorded, and the interviewer took notes instead. The two research assistants transcribed the interviews verbatim in Spanish, removing all identifying information and designating a pseudonym for confidentiality.

2.3 Data collection

Because most children under five in the DR are at home with a female caregiver during the day [21], we focused the interviews on the home environment. We asked about responsive caregiving and opportunities for early learning in-depth and intentionally allowed conversations about security and safety to emerge organically during the interviews. We chose to target these topics and not good health and adequate nutrition because the former have yet to be explored in the literature among this population and evidence suggests these areas may buffer the impact of poverty on child development [8,9]. In contrast, national-level data are available about the health and nutrition of this population [21]. Finally, responsive caregiving and opportunities for early learning are well-suited for integration as parent coaching modules in existing public health interventions and of interest among our partners at INAIPI.

2.4 Interview guide

We used semi-structured interview guides, with open-ended questions taken from the Working Model of the Child Interview and informed by the Nurturing Care Framework. The Working Model of the Child Interview is a structured interview designed to identify a mother’s internal representations, or working models, of the relationship with their child [29] (S1 File). We used nine of ten questions from the WMC (numbers 1 (simplified), 2, 3, 4, 5 (simplified), 6 (simplified), 7, 9, and ten). We simplified several questions and removed questions soliciting specific words. We then added in additional questions related to responsive care and opportunities for early learning. For each of these domains, we inquired about 1) how mothers interact with their child when they are alone, 2) their perception of what experiences and behaviors promote development, and 3) how they perceive the child’s environment and interactions to impact their child’s development. Before the interview, we administered a short structured sociodemographic survey to gather information about number and ages of children. We also included one question from the Household Food Insecurity Assess Scale that asked how often they had worried about not being able to afford food for their family in the past month, with three Likert scale response options (1- Rarely or 1–2 times in the past month, 2- A few times, or 3–10 times in the past month, and 3- Many times, or more than 10 times in the past month) [30]. We conducted data collection between August 13, 2022 and February 15, 2023.

The first author (principal investigator) trained two Dominican psychologists research assistants to conduct qualitative interviews. Research assistants observed five interviews conducted by the first author, then the research assistants conducted three interviews with the first author present. Research assistants conducted a total of seven interviews. We ensured all interviewees could speak privately during the interview. We conducted all interviews in Spanish, in-person (N = 21) and virtually (four women had moved to distant locations after participating in the original study). We conducted all interviews in quiet, private locations. In Santo Domingo, we held in-person interviews at public meeting places selected by participants (e.g., restaurants, shopping malls). In Yamasá, we conducted interviews in the patio or yard of participants’ homes. The participants received a stipend in consideration for their time.

2.5 Data analysis

We used deductive thematic analysis to analyze transcriptions based on themes from the Nurturing Care Framework and the Working Model of the Child, but allowed for additional themes to emerge from the data [27]. Literature indicates that parent age is associated with parenting behavior [31,32], so we examined divergence and converge among data from adolescent versus older mothers. First, the first author and the research assistants conducted open coding with a pre-developed codebook in Dedoose [33]. We structured the codebook to cover the Nurturing Care Framework’s themes of responsive care and early opportunities for learning and allowed subthemes to emerge from the interviews. When it was unclear which theme an excerpt fit in, we consulted the Nurturing Care Framework’s summary. For example, this quote involves both responsive caregiving and early opportunities for learning: “sometimes I help her, I sing. And I teach her the vowels, I give her a pencil and I say, ‘look, write here, this is the letter A. Look, these are the numbers, count them’. She starts to count.” We coded this excerpt as responsive caregiving because the action subject is the mother and her interaction with the child. The child’s response is secondary. Because participants simultaneously reported on safety and security topics, we included that as a theme in the codebook. Our final themes included: 1) responsive caregiving (subthemes: relevance for development, parent interactive behavior, perceived role as mother), 2) opportunities for early learning (subthemes: beliefs about how children learn, intentional teaching, preschool/ early stimulation), and 3) safety and security (subthemes: managing difficult behavior, use of corporal punishment). All analysts individually coded five interviews while blind to others’ coding and then compared answers. After five interviews the three analysts reached 80% agreement and we proceeded to code individually, checking our coding every five interviews to avoid drift. Once all interviews were coded, we wrote summaries of each theme, and organized the themes into a logical narrative [34]. The first author and one of the assistants separately translated Spanish excerpts into English for this report and discrepancies in translation were resolved by the senior author, a native Spanish speaker with 10 years’ experience conducting maternal and child health research in the Dominican Republic.

2.6 Ethics review

We received approval from the Universidad Iberoamericana Human Research Protection Program (# CEI2022-13) and the Tulane University Research Ethics Committee (# 2019–2375).

3.1 Demographics

Participants’ mean current age was 27 years “Table 1”. Seven participants became mothers before reaching the age of 18 and mean age of motherhood was 19.5 years (median = 18 years). They had a mean of two children and the youngest child was two years old. Three participants had children that did not live with them, in addition to those who were living with them, and two women were caring for dependents or children in their home who were not their biological children. Eleven had at least a high school degree, 16 were employed, and three women stayed home exclusively with children. Eighteen participants reported food insecurity, defined as being worried about having enough food for their family in the past four weeks.

Download:

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

Table 1. Demographic data, n = 25.

https://doi.org/10.1371/journal.pone.0314432.t001

3.2 Results summary

With respect to the relevance of responsive caregiving for development, few women explicitly identified mother/ child interaction or responsive caregiving as important for their child’s development. However, many spontaneously described taking part in this behavior. With regards to their role as mothers, many expressed intentions to raise well-behaved children. Some participants, particularly adolescents, demonstrated low confidence in their mothering skills. Speaking about opportunities for early learning, participants mostly expressed confidence in tools external to what they could provide themselves, such as preschool or educational videos. A few mothers took part in intentional teaching, often together with educational cartoons or games on the tablet with material related to colors or numbers. Mothers voiced confidence in their child learning from interactions with children of all ages, such as in a preschool or informal family setting. When addressing issues of security and safety, participants reported struggling with their child’s behavior and almost half of mothers reported using corporal punishment, a phenomenon more common among adolescent mothers. See “Table 2” for additional themes and selected quotes not presented in the text.

Download:

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

Table 2. Select excerpts.

https://doi.org/10.1371/journal.pone.0314432.t002

3.3 Responsive caregiving

3.4 Opportunities for learning

3.5 Safety and security

4.1 Summary

While most mothers did show affection to their young children, on average they underestimated or did not recognize their ability to promote their development. We found mothers provide responsive caregiving but do not view their own caregiving interactions and activities as an explicit mechanism or agent for promoting their child’s learning and development. For adolescent mothers, this may be because they reported low confidence in their own parenting skills. Mothers predominantly described relying on factors external to themselves, such as tablets, to provide early learning opportunities. The reported use of physical punishment was high, particularly among adolescent mothers, which aligns with findings that over half of DR adolescents have experienced physical abuse from their parents or guardians [35].

4.2 Responsive caregiving

We found a divergence in findings between adolescent and adult mothers’ confidence in their skills as mothers. Adolescent mothers from our sample reported greater deficits in childrearing skills compared to older mothers. The portion of adolescent mothers in our cohort is consistent with the same proportion among the poorest quintile of mothers in the Dominican Republic (33.9%) [21]. However, while 68% in our study were under the age of 20 and 36% were under the age of 18 at first birth, all mothers in our cohort were young, with a mean age of 19.5 years and a median age of 18 years at first birth. Our findings suggest young mothers face doubt in their motherhood skills, and that those under the age of 20 experience heightened feelings of ill preparedness. Similar findings have been reported in Brazil within the context of Brazil’s Criança Feliz early childhood program [36] and among adolescent mothers in shelters [37], and in the United States [38]. This confirms the importance of targeting this demographic for programs that promote parenting skills for child development and address socioeconomic inequities in income, access to education, and gender norms (e.g. gender-based social norms and sexual abuse), all contributors to adolescent pregnancy in the DR [39].

Improving parent awareness of how the activities they are already doing improve their child’s brain development during this early stage of development, and encouraging more similar behavior could be another step toward improving the quality of stimulation children receive in the home. In particular, articulating the broad-reaching implications of critical periods in child development and how to positively influence them especially when incorporated into daily activities, such as feeding, a diaper change, a bath, or a walk to the market, could help caretakers understand the importance of their roles and increase its use in this community. Our study also suggests the importance of opportunities to deepen parents’ recognition of developmental delays and ways to remediate these (i.e., improving early relational abilities in a child with aversion to touch and eye contact). Parents can benefit from access to information about stages of cognitive and language development in their children and how to recognize and promote progress in these areas for their children.

Ways to improve parent education in responsive caregiving in low-income areas have proven successful at small to medium scale, most notably through community-based home visiting programs [40]. One innovative video tool which demonstrated potential in Peru, Universal Baby, uses local footage of parent-child interactions to educate parents on brain development and responsive caregiving [41]. These studies suggest that parents would likely be receptive to additional ways to build brain power in their child (e.g., how reading aloud with a child from a book can foster attention and language; how encouraging a child to move beyond rote recitation of numbers to gain a deeper understanding of early numerical concepts such as single, few and many; or how parents can help a child build organizational and categorization skills that form the foundations of higher order thinking).

4.3 Intentional teaching

Few mothers took part in intentional teaching when interacting with their child, and when they did it was mostly focused on traditional recitation of letters and numbers or moral teachings such as songs and stories from church or learning the difference between right and wrong. There are very few studies investigating early learning within the Nurturing Care Framework in low- and middle-income countries, however, our results echo findings from a cross-sectional study examining knowledge and attitudes of nurturing care in a low-income cohort in Turkey, which found that knowledge and practices associated with opportunities for early learning are not strong and suggest that mothers miss critical periods for learning because they are unfamiliar with developmental milestones [42]. Likewise, one review of policy documents in Kenya found little support for early opportunities for learning among government-subsidized programs [43]. Another study among 5,570 Brazilian municipalities demonstrated the largest regional inequities in the early opportunities for learning domain, likely reflecting corresponding socioeconomic inequalities [44].

Most children this age did not attend preschool, despite INAIPI centers being available in many of their neighborhoods in Santo Domingo. This was either because they weren’t aware of the program or because the program was full. Caregivers who lived in neighborhoods perceived as dangerous preferred to have their child taken care of, and sometimes even live, with a family member, often a grandmother. This may demonstrate the tension young mothers experience with competing requests for their time as students, wage earners, and childcare providers. According to the 2019 MICS report, 63% of children ages 2–4 years received some stimulating activity from a caregiver other than their parents in the three days prior to the survey, 44% participated in these activities with their mother, and 10% with their father [21]. Studies involving the nurturing care framework should include conversations with other family members, particularly grandmothers who act as primary caregivers, and explore further the role of fathers in child interactions.

Caregivers had difficulty articulating goals and dreams for their child, and how they might help them achieve these goals as caregivers. Responses centered around child behaviors such as being obedient and avoiding trouble. This reflects the structures of poverty and limited opportunities available to low-income families to help their children reach their potential. On the other hand, it also suggests that caregivers see themselves as relatively powerless in this setting. Given that child well-being is a community norm and that institutions such as school and health clinics exist in the community, as do public spaces for recreation and play, even if limited, opportunities do exist for community members to play an active role in improving opportunities and services.

Women looked primarily to external props, distal factors, and activities outside the home for learning opportunities, such as tablets and videos, saving money for their child’s future education, and social interactions with other children.

Parents promotion of screens and assumptions about the educational value of screen exposure were concerning in light of growing evidence about the ways that childhood screen time may disrupt the development of healthy social skills, attention and concentration, and emotional self-regulation. The maximum recommended amount of time for a three-year-old to watch a screen is one hour per day [45,46], and excessive screen use in children ages 3–4 years is associated with lower psychosocial well-being [47]. While we did not collect data on the exact time of screentime use per day, our qualitative data suggest that most children spend several hours per day on a tablet or phone. Since 2021, the Dominican government has provided all children in elementary public schooling with a tablet, and many children in this study had access to a sibling’s tablet for educational games. Some interacted with adults while playing these games, which may be less detrimental to development than unsupervised screentime [48], however, it is unclear how much other activities, such as watching videos, using social media, or surfing the internet, are conducted on the tablets. The dos Santos et al study found that caregivers use screens to calm their children during tantrums [36]. More research is needed to understand whether children in the DR are supervised while using screens, the material’s quality, and adult participation in these contexts.

4.4 Security and safety

Although frequent in the Dominican Republic, the use of physical punishment has been shown to negatively impact child behavior and development both immediately and longitudinally [49].

We found a high use of physical punishment for disciplinary purposes, in almost half of the participants, who spontaneously reported using physical punishment at least once with their child. The 2019 MICS survey reports that an average of 8.8% of Dominican women or caregivers of young children agree with the use of physical punishment, with a higher use among people in the lowest wealth quintile (12.6%), the lowest education level (15.0%) and among people under 25 years of age (9.3%) and over 50 years of age (9.1%) [21].

The use of physical punishment (mostly hitting with a shoe or pole) among mothers in this cohort may be augmented by the perceived helplessness and alienation that comes with adolescent motherhood. It was difficult to assess correlations with the small sample, however, 10 of the 11 mothers reporting physical punishment were under the age of 20 at first birth. These mothers reported little faith in their parenting abilities and attributed this lack of parenting ability to their own compulsive behavior. Most women who used physical punishment mentioned losing their temper and then regretting it later, while a few believed physical punishment was the best way for children to learn right from wrong.

Physical punishment is more common in low-income areas worldwide. One study in Uganda investigating attitudes about physical punishment by showing hypothetical scenes of misbehavior and asking mothers to indicate their response strategy, found that 2/3 of women report they would use physical punishment to deal with bad behavior [50]. Similarly, in a baseline assessment of women participating in a clinical trial in Afghanistan, 71.8% reported using physical punishment within the past month. Within the Brazilian sample studied by dos Santos et al, women reported using swearing and shouting to control children when they felt unable to manage difficult behavior [36], and connected this response to feeling out of control. Physical punishment can have a lasting negative impact on development. A metanalysis based on 69 surveys worldwide found that children experiencing physical punishment were on average 24% less likely to be developmentally on track [51]. Parents everywhere should have access to global data about the negative impacts that accompany corporal punishment and harsh parenting.

One objective of this paper was to deepen findings from Sánchez-Vincitore and Castro that stimulating caregiving and opportunities for early learning mediated the relationship between socioeconomic gradient and ECD outcomes among children 36–59 months of age in the Dominican Republic [24]. Our findings contribute to current knowledge of how physical punishment is used among young mothers, a population more inclined to believe physical punishment is an appropriate disciplinary technique in the MICS data, compared to older mothers. Likewise, we also show how opportunities for early learning could be improved among families in this setting. While we didn’t measure the specific indicators used in Sánchez-Vincitore and Castro (2022) to predict early child development (such as number of children’s books at home, stimulating activities at home, and attendance to an early childhood education), our findings suggest that the use of tablets and cell phones may be excessive and may be impeding other opportunities for learning in the home. However, among our cohort, social interaction, particularly with other children, was quite common.

5. Limitations

This was a small qualitative study focused on understanding how responsive caregiving, opportunities for early learning, and safety domains as recommended by the NCF are experienced by children from low-income families as reported by their mothers in and near Santo Domingo, Dominican Republic. The sample size was small; however, given the limited scope of our interviews, we reached saturation on all our topic areas of interest. We could not assess factors requiring home observation and did not cover other determinants of development, such as nutrition and childhood illness.