Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Patrice L. Engle and Henry N. Ricciuti
Abstract
The quality of psychosocial care provided the young child is reflected in the caregiver's responsiveness, warmth and affection, involvement with the child, and encouragement of autonomy and exploration. First, research was examined linking the quality of psychosocial care to a child's development of mental abilities, and to his or her growth and nutrition status. There is considerable correlational evidence and some experimental evidence for this linkage. Second, the barriers to adequate psychosocial care were explored, including maternal beliefs and confidence, stress and depression, social support, and autonomy and control of resources. Third, a number of strategies to enhance psychosocial care as a mechanism for increasing the child's nutrition status are described, at the level of the child, the caregiver, the child-caregiver relationship, and the community. All of the work was interpreted in terms of an interactive or transactional model of child development.
Relevance of psychosocial care
It is widely recognized that the normal growth and development of infants and young children requires care that adequately meets their basic physical needs (nutrition, health, clothing, housing, sanitation, protection from dangers, etc.) as well as their socio-emotional or psychological needs. The latter includes the availability of caring, loving adults with primary responsibility for the child's well-being, with whom the child can form an affectionate, stable, and trusting relationship, and who provide opportunities for positive social interactions, play, and learning appropriate to the cultural context in which families live. Characteristics of psychosocial care that meets these needs include the caregiver's responsiveness and sensitivity, affection and warmth, psychological involvement with the child, and encouragement of learning and development.
This paper presents an overview of the role of psychosocial care in promoting child growth and development, with special reference to linkages between psychosocial care and nutrition. One of the major questions to be considered is whether enhancement of psychosocial care can improve children's nutrition status in low-income populations facing significant resource constraints. Consideration is given to what is already known, significant gaps in the present knowledge base, and strategies for protecting and promoting good psychosocial care along with good nutritional care.
To illustrate the relevance of psychosocial factors in nutritional contexts early in life, we begin by presenting four vignettes:
1. Jorge's mother is away at work, but he is being cared for by his aunt. She prepares lunch of rice and an egg sauce for Jorge (age two years) and her two children (age four and seven years). She spreads a mat on the ground outside the door of their small wood frame home, directs the three children to sit in a row, and places a bowl in front of each child. She enters the house to continue her housework, leaving them to eat. They giggle and laugh together, eating spoonfuls and playing with the food.
2. Eighteen-month-old Esmeralda is still carried on her mother's back most of the time. At 11:00 in the morning she stands, leaning on her mother's knee for support, and watches her two older sisters eat freshly cooked kernels of corn they have been spooning out of the cooking pot. The two room house is cool and dark even at this hour, with a dirt floor and the fire burning in the corner. She reaches out for the corn her sister cradles in the large spoon, but the three-year-old sister twists away, continuing to eat the kernels one by one. Her mother is talking to a visitor, paying no attention to this silent drama. Esmeralda reaches out again and grunts (she has no language yet), but her sister quickly moves to the other side of the room. Finally Esmeralda scoots herself on her knees (she cannot yet walk alone) over to the other side of the room, next to the cooking pot, and reaches in for a handful of corn. She sits contentedly beside the pot eating kernel after kernel.
3. It is time for the main meal of the day, and the family sits in a circle waiting for the food. The older sister brings a single pot, and each family member uses a piece of bread to dip into the bowl. At 20 months of age, Mani has difficulty reaching the food, but she receives no help. They eat enthusiastically, reaching into the pot until the food is gone. The family assumes that Mani will eat what she wants, and in any case she is still being breastfed. The meal is millet with a sauce that contains a small amount of meat. The piece of meat has been reserved for the father in the household.
4. Two-year-old Pedro's mother greets a visitor at the door. Before she sits down on the simple chair to talk to her friend, she enters the house and reaches for a basket containing a sturdy glass jar, a number of small pebbles, and a plastic bottle with a small neck. She places the basket in front of her child, asking him, "How many of these can you put in the bottle?" He eagerly grasps a pebble and drops it into the bottle, fascinated with the sound. He continues to play in absorption for some time while his mother sits nearby talking to her friend. Occasionally she looks over at him and comments, "Look what you are doing. You are putting the pebbles in the bottle."
These examples from Guatemala, Nicaragua, and West Africa illustrate the importance of psychosocial variables for children's nutrition. In the first example, the caregiver is not an active feeder and is not even present at the meal, but eating with other family members encourages the child to pay attention to eating for a longer time. In the second example, the child's snacking (an extremely important source of nutrients for young children) is dependent on the child's ability to signal hunger to either her mother or to her older sisters, and the willingness of the siblings to comply with those requests. Perhaps if she had developed language, she would have been able to get the attention of her mother or the cooperation of one of her sisters to obtain more food. In the third example, the amount of food that the child receives is largely dependent on her ability to reach for it, and on her appetite. She receives no encouragement to eat; if she is not hungry, or does not like the food presented, she will eat little. In comparison with the first three, the fourth example illustrates good psychosocial care in a non-feeding situation (providing stimulating materials and comments to the child).
Definition of psychosocial care
A substantial body of research indicates that children's early mental, motor, and social development is significantly influenced by variations in their early experiences. This research can form the basis for understanding the possible influences of psychosocial care on child nutrition. Early experiences have been assessed using the nature of children's home environments and the quality of care provided by their parents or other caregivers [1, 2].
Children's early home environments have commonly been characterized in terms of various demographic or "structural" characteristics (family income, parent education or occupation, family size and composition), or in terms of various "distal" features of the physical or social environment in the home (housing quality, crowding, noise, birth order, presence of newspapers, radio). In addition, various parental characteristics, such as age, child-rearing knowledge and attitudes, and mental and physical health, have sometimes been employed as "environmental" measures. Since many of these characteristics tend to be related to child developmental outcomes and may significantly influence the quality of care parents provide, they are often utilized as "proxies" for variations in early experience and child care.
We are particularly concerned with the child's functional environment, as reflected in more direct or proximal assessments of children's day-to-day experiences in their natural settings, including the quality of child care or rearing provided by parental or other caregivers. Although a variety of specific features of child care or parent-child interaction have been investigated and found to be related to behavioural development, the following broadly defined qualities of early psychosocial care are considered particularly important, and have generally been found to be associated with positive developmental outcomes in children in a variety of cultural settings:
- responsiveness, sensitivity, and consistency in responding to the child's cues or needs, versus lack of responsiveness, insensitivity, and inconsistency (the extent to which the caregiver is aware of the infant's signals and needs, interprets them accurately, and responds to them promptly, appropriately, and consistently);
- warmth, affection, and acceptance, versus rejection and hostility (reflects balance of positive and negative feelings toward the infant, including accepting characteristics of the infant that might be seen as undesirable);
- involvement with the child, versus lack of involvement and detachment (reflects the degree to which the caregiver or parent cares about the infant's well-being on a day-to-day basis, takes appropriate actions on the infant's behalf, and spends time with the infant when possible);
- encouragement of autonomy, exploration, and learning, versus restrictiveness and interference (extent to which the caregiver provides opportunities for motor, mental, and social development, including independence and self-confidence, through appropriate activities and verbal and social interactions).
Although researchers are inclined to distinguish among these broad aspects of child care for conceptual and analytic purposes, in the "real world" they tend to be interrelated to a considerable degree. All four dimensions are likely to be reflected in the extent to which caregiver and child engage in mutually rewarding, age-appropriate, reciprocal interactions. Also, cultural factors will influence the particular ways in which the characteristics of psychosocial care just outlined are expressed, as well as the extent to which particular child-care practices may be valued or discouraged, in the light of their adaptive value in a given setting [3-5]. This is likely to be particularly true in the case of discipline and socialization practices, which are not discussed specifically here. Finally, there are a variety of social and economic factors in the ecology of the family and community that may either support or limit the primary caregiver's ability to provide good psychosocial care. These factors will be discussed later.
FIG. 1. Factors that affect the relationship of the caregiver and the child
Conceptual model
The conceptual model (fig. 1) outlines the major pathways through which care may influence a child's growth and development. First, we suggest that psychosocial care and nutritional care are closely interconnected in the routine caregiving of everyday life, and thus should be considered holistically. Moreover, the quality of care a child receives will reflect, and to a considerable extent be influenced by, the nature of the overall child-caregiver relationship, a unique pattern of behaviours that develops over time between a parent or caregiver and a particular child (see the box in the centre of fig. 1). As indicated by the solid arrows, it is well known that the quality of psychosocial care has a direct influence on children's behavioural development (linguistic and cognitive development, growth of social skills, psychological adjustment and adaptation, and development of motor skills). Also, the nutritional and health care children receive has a direct effect on their nutrition status and physical growth.
However, of particular interest is the hypothesis that the quality of psychosocial care may influence the child's nutrition status and growth indirectly, through its linkage with, and possible effect on, the quality of nutritional and health care. At the same time, as suggested by the dotted feedback arrows in figure 1, a poorly nourished, less responsive infant may significantly affect a caregiver's behaviour in regard to both psychosocial and nutritional care. Similarly, the child's developmental status and behaviour may also influence the quality of care received in both the psychosocial and the nutritional areas. Finally, the model indicates that the child's behavioural development may be directly influenced by his or her health and nutrition status. Thus it is apparent that the relationships among care, nutrition, growth, and behavioural development may be viewed as interactive or bidirectional.
Since the predisposing factors affecting development tend to interact with one another, we should not expect to find a single causal link between a particular input (e.g., psychosocial care) and an output (e.g., child growth). We are beginning to understand that the development of a child is better represented by a series of feedback loops, with each change in the child influencing changes in the environment, and vice versa, in a continuous series of interactions. This ongoing pattern of interactions has been called transactional. There is not much known about these influences, particularly in situations with substantial resource constraints.
It is important to recognize that the nature of the caregiver-child relationship and the quality of psychosocial care or nutritional care that parents or other caregivers are able to provide for children are very much affected by several major sources of influence. These are 1) the characteristics of the child (age, developmental level, gender, irritability); 2) the characteristics of the relationship between child and caregiver; 3) the characteristics of the caregiver (attitudes and beliefs about caregiving, motivation and physical energy, emotional state, sense of personal efficacy, physical and mental health); 4) the support system (nature of available informational and emotional support); and 5) resource constraints (adverse environmental conditions affecting disease prevention, limited disposable income and unstable food supply, excessive demands on caregiver's time for other tasks). The first two will be discussed briefly here, and the last three will be discussed later.
Characteristics of the child
Children bring very different strengths and weaknesses to the caregiving environment. Many of these are genetically determined (temperament, genetic predisposition for disease), but some are associated with nutritional conditions. For example, severely malnourished infants display a range of symptoms such as apathy, listlessness, and irritability that can make caregiving more difficult. Motor development may be more advanced at 24 months in children receiving relatively small amounts of nutritional supplementation [6].
One of the most important characteristics of the child is developmental level. Psychosocial development involves dramatic changes over the first five years of a child's life, and caring requirements vary tremendously with the age or developmental status of the child. The possible risks to the child and the responsibilities of the caregiver vary with each developmental period: prenatal, early infancy (first 6 months), late infancy (6 to 12 months), early toddler (12 to 24 months), late toddler (24 to 36 months), and preschool (3 to 5 years).
Children may also differ in characteristics that affect how they are valued by their families. In Guatemala, children with lighter-coloured skin or children who resemble a particular parent may be seen as special. There is ample evidence in many societies for gender bias favouring males [7, 8].
Characteristics of the child associated particularly with nutrition include sucking ability, appetite and hunger, temperament, and energy and motor skill to self-feed during critical early years. All aspects of the feeding situation vary according to the developmental status of the child.
Characteristics of the relationship
The most important variable from a psychosocial point of view is the nature of the child-caregiver relationship. This will be strongly influenced by the characteristics of the child and caregiver, but over time it forms a unique pattern of behaviours between two individuals. Thus, within one family, a parent will have different relationships with different children. These relationships may be relatively durable and resistant to change. They are also synergistic and dynamic. A poorly nourished infant may be less responsive to the caregiver, who in turn may believe that the child requires less attention than others, and she may direct her scarce time to other problems. Thus the early perturbations in the relationship will amplify over time, as with any dynamic or chaotic system. The caregiver's responsiveness to the child's cues is essential for timely feeding, particularly between meals. This relationship is important during breastfeeding as well as during supplementary feeding, and lays the groundwork for good care and nutrition during the early years of life.
The model as outlined in figure 1 provides guidance as we ask the critical questions related to care: can high-quality psychosocial care enhance the use of health resources, or compensate for insufficient resources, to result in improved nutrition status of children? If so, can caregivers be trained, encouraged, allowed, or facilitated to provide this high-quality care?
Quality of psychosocial care is associated with more favourable developmental outcomes in children
Findings supporting the generalization that young children's behavioural development is significantly influenced by the nature of their early experiences and the quality of child care received are illustrated below. It is particularly important to note that variations in experience and child care associated with children's development are found not only among major socio-economic status (SES) groups, but within low-income populations as well.
Care and behavioural and development outcomes: Industrialized countries
Psychosocial care and mental development
Many studies find that measures of the caregiver's psychosocial care early in life are associated with later cognitive or emotional development of the child. These include observational or interview measures of care and the "HOME" scale, which has been widely used both in the United States and in third world settings (with appropriate adaptations), and is based on interviewing the caregiver at home as well as observing the child with the caregiver [9]. It provides an overall score and sub-scores on six scales: 1) emotional and verbal responsivity of caregiver, 2) avoidance of restriction and punishment (acceptance), 3) organization of the physical and temporal environment, 4) provision of appropriate play materials, 5) caregiver's involvement with child, and 6) opportunities for variety in daily stimulation.
Studies employing the HOME scale for assessing the quality of the home environment in the first few years of life have consistently reported significant associations between these measures and children's intellectual development [1, 10, 11]. The HOME scale was used in a longitudinal study of a large, varying SES sample, and the investigators found that children's Binet IQ at 4 years of age was significantly associated with the quality of the home environment (total HOME score) at 24 months of age (r=.60) and even at 12 months (r= .43) [11]. Another measure of psychosocial care-mother's positive, encouraging teaching style on specific tasks observed at 24 months-correlated significantly (r= .53) with IQ at 4 years of age. The correlations were generally equivalent for the higher and lower SES groups when examined separately.
Several studies of premature infants in low-income families also provide evidence of the predictive value of the HOME scales. In a sample of low-income African-American mothers, verbal/emotional responsiveness measured on the HOME scale at 9 months was significantly correlated with Binet IQ at 3 years of age (r= .41) [12]. Also, maternal responsiveness at 20 months predicted social competence at 3 years, as rated by teachers in a 3-week summer camp (r = .48). In a more recent study of children of rural, low-income families, half of whom were premature, the quality of the home environment (total HOME) assessed when the infants were as young as 5 months old (corrected for gestational age) was found to be substantially associated with Binet IQ at 30 months (r = .61).
Other measures of psychosocial care have also been associated with cognitive outcomes. Qualitative ratings of maternal sensitivity and acceptance/ warmth derived from a number of lengthy home observations of children 9 to 12 months of age were correlated with Griffiths Developmental Quotients (DQ) at 1 year (r=.46, .45) [13]. Ratings of the amount of "floor freedom" allowed these pre-toddlers also correlated with Griffiths 1-year DQ (r=.46). Although these findings were based on a middleclass white sample, similar correlations were obtained with a low-income black group. Systematic observations of specific positive features of mother infant interaction, including play and teaching situations made at various points during the first 2 years of life, were significantly associated with intellectual competence at age 5 (Binet IQ) among premature infants [14].
Some investigators have been able to link particular features of the verbal interaction between mothers and infants to the children's language development. In one study, vocabulary progress during the second year of life was related to frequent, responsive mother-child language interchanges, even when family SES and maternal education were controlled [15]. Increasing research emphasis is currently being placed on the need to link specific features of young children's early environment and experience to particular dimensions of behavioural development, and to determine whether such environmental effects vary as a function of characteristics of the child, such as age or developmental status [16].
Motor development has also been found to be associated with measures of the early environment. Black toddlers in Washington, DC, who were reared in a "permissive, accepting"' rather than a "rigid" home environment (and thus had more opportunities for "floor freedom") tended to have higher Gesell Motor DQ scores at 18 months of age [17]. This relationship was observed in both middle- and lower-class families.
In addition to these correlational studies, there have been some experimental studies that have altered the early experiences of children through random assignment, and have shown significant effects on cognitive development. The most striking results come from the Carolina Abecedarian Project [18], which placed poor children in an educationally oriented day-care programme for 8 hours a day, 5 days a week, 50 weeks a year from age 4 months to 5 years. Significant effects on intellectual development and academic achievement were maintained through age 12, although the effects were greatest at the conclusion of the day-care experience. The authors conclude that intensive intervention during the first 5 years of life increases the likelihood of long-term cognitive effects.
Psychosocial care and the development of mother-child attachment
The relationship of the child to its caregiver has a significant effect on the quality of care the child receives. One aspect of this relationship is the formation in the child of an enduring, focused, affectionate relationship with the mother or other primary caregiver, referred to as child-caregiver attachment. A secure and trusting attachment relationship with the mother or primary caregiver is viewed as providing a very important base for the child's subsequent social and personality development, as well as adaptive behaviour more generally [19].
During the first 18 months of life, the characteristics of psychosocial care mentioned at the outset (particularly maternal responsiveness or sensitivity, warmth, and involvement) play a significant role in facilitating the development of a secure attachment relationship between infant and mother (or primary caregiver). Empirical evidence of the importance of maternal sensitivity and responsiveness comes from a variety of studies in which "security of attachment" was measured experimentally on the basis of the child's observed behaviour when left alone or with a stranger in a standard "strange situation" paradigm. Such measures of attachment security in 12- to 24-month-old children have been found to be significantly associated with maternal sensitivity and responsiveness based on home or laboratory observations [20, 21].
The kinds of maternal sensitivity, warmth, and involvement that promote mental development, as well as a secure attachment relationship, begin to play an important role very early in life, and are often observable in the feeding situation. Moreover, variations in such maternal characteristics may well have significant implications for breastfeeding and the quality of early feeding practices.
Care and behavioural and development outcomes: Developing countries
Early psychosocial care has also been shown to be associated with subsequent behavioural development in a variety of third world populations exposed to significant nutritional and health risks. One of the earliest of these involved the use of an adaptation of the HOME scale in a prospective longitudinal study of a cohort of children growing up in a Mexican village [22]. Children who came from home environments characterized as relatively less favourable in the first 2 years of life tended to have lower mental development scores at 5 years of age (and they also were more likely to be identified as clinically malnourished early in childhood).
Several more recent studies support these early findings from Mexico. In a study of 196 children from 10 rural villages in northern India, for example, the HOME scale correlated positively with children's IQ (r = .51) as well as Gesell DQ (r = .46) at 36 months [23]. Moreover, in both instances HOME scores explained significant variance in IQ/DQ even after controlling for SES and nutrition status. Significant correlations were also found between HOME and the motor development sub-scale of the Gesell.
In a study of 5- to 6-year-old Philippine children of varying nutrition status, interviews were used to measure variety of stimulation and parental involvement in child care in the home [24]. Both measures were significantly correlated with a composite measure of intellectual competence (r = .35, .38). In addition, the level of parental involvement added unique variance to the prediction of intellectual competence beyond that provided by mother's IQ, parental education, and the quality of the physical environment in the home.
Similarly, detailed and extended home observations of rural Kenyan toddlers indicated that children who performed better on the Bayley Mental Scale at 24 to 30 months of age tended to have caregivers who had more frequently talked to them, responded to their vocalizations, and engaged them in sustained social interactions (r = .19 to .29) [25]. The latter two caregiver behaviours were also positively associated with children's smiling at 30 months (r = .24, .34).
Specific features of the child-care environment have been associated with particular aspects of infant and toddler development in a detailed observational study of Egyptian village toddlers [26]. The frequency of toddlers' distress was significantly associated with the caregivers' lack of responsiveness to the child's fussiness (r=.34, .48), whereas the amount of toddlers' positive vocalization was related to the frequency of caregivers' vocal stimulation (r=.58, .43). Caregiver behaviour in this instance was not related significantly to measures of cognition.
One study in Jamaica found evidence of the importance of early psychosocial care in the behavioural development of nutritionally at-risk children [27]. Stunted toddlers enrolled in a home visiting programme aimed at encouraging mothers to engage in play and interaction avoided a potential decline of about 7 to 8 points in Griffiths DO scores, about the same benefit as that provided by nutritional supplementation alone. (Additional benefits were gained when psychosocial and nutritional intervention were combined.) A second study compared previously hospitalized malnourished children receiving a similar home visit programme with malnourished control children and with previously hospitalized but not malnourished children, using a case-control design [28]. Positive mental development effects were found for the home visit group, which continued to be seen at a 14-year follow-up [29].
Quality psychosocial care is also associated with more favourable nutrition, health, and growth outcomes
There is considerable evidence to suggest that good psychosocial care is not only associated with positive behavioural outcomes in poor children, but also with more favourable nutritional and growth out comes, even within populations with nutritional and health risks. Conversely, where such positive features of psychosocial care are greatly reduced or lacking, children from poor families are also likely to be at increased risk of malnutrition [3, 30, 31].
Some research strategies have focused on identifying which family or demographic factors place some low-income children or families at increased nutritional risk [3234]. One method recently developed tries to identify maternal, family, or child-care variables characterizing families whose children tend to be adequately nourished although living in "high risk" nutritional environments: the "positive deviance" studies [5]. This strategy has resulted in many important findings [35]. However, the variables that make some children least likely and others most likely to be malnourished may not be the same [36].
A similar distinction has been made in psychological conceptualizations between "protective" factors, which prevent or minimize the likelihood of adverse developmental outcomes in children otherwise at high risk, and independently defined "risk" factors, which place some children at heightened likelihood of mal-adaptive behavioural outcomes [37, 38]. Psychologists tend to prefer these terms to the more statistically derived "positive deviance" concept. In both the psychology and the nutrition literature, increasing emphasis is being placed on the need for greater understanding of the positive features of parental child-care practices, coping strengths, and adaptive strategies that may protect children in "high-risk" environments from adverse nutritional or developmental outcomes, rather than focusing primarily on parental or family risk factors [5, 31, 38, 39].
In considering the potential linkages between psychosocial and nutritional care, one of the principal assumptions would be that caregivers who are minimally involved and show little affection for the baby, who are insensitive in responding to the child's needs and signals, and who fail to encourage exploration and learning, are also likely to provide relatively poor nutritional, feeding, and physical care. Thus, it is important to examine the quality of child care as an integrated whole, since psychosocial and nutritional or physical care are likely to be rather closely linked.
Psychosocial care and nutrition status: Industrialized countries
Although there is little resource-shortage malnutrition in industrialized countries, some useful evidence about the role of psychosocial care can be found in a number of studies of "non-organic failure to thrive" (FTT), which have suggested a breakdown in parenting functions [40]. These "parenting breakdowns" are often attributed to major family and socio-economic stresses, including having a difficult, sickly, or low-birth-weight infant. Unfavourable maternal mental health, personality characteristics, or attitudes may also contribute to the diminished quality of care received by such infants.
The homes of 23 low-income FTT children in Little Rock, Arkansas, USA, were rated lower on the HOME scale overall, but especially on scales indicating less maternal responsiveness to and acceptance of the child, as well as a more disorganized environment [41]. Further evidence of the link between inadequate psychosocial and nutritional care comes from a large-scale study of five-year-old children of low-income families in Baltimore, Maryland, USA [42], which indicated that those coming from homes characterized by generally inadequate overall "mothering" or child care tended to be at greater nutritional risk (based on dietary intake and biochemical measures).
One experimental study also suggested that favourable psychosocial care may ameliorate the negative effects of foetal malnutrition. Low-SES infants were placed in an intensive 8-hour per day child care programme, and controls who stayed home with their mothers were identified [43]. A small number of infants in each group were foetally malnourished (low ponderal index). There were no differences in maternal involvement or 18-month test performance between foetally malnourished and normal children in the day-care treatment group, but the foetally malnourished children who were cared for at home had significantly lower maternal involvement and test scores than their home controls, as well as both day-care groups.
Psychosocial care and nutrition status: Developing countries
A number of correlational studies in developing countries have also linked inadequate psychosocial care and poor nutrition or health status. In one longitudinal study [32], children who were clinically malnourished some time in their first three years tended to come from homes scoring lower on the HOME scales as early as the first year of life. Mothers of these children were also found to be more passive, less responsive and sensitive to the child's needs, and less open to "modernization" [22]. Similar characteristics reflecting relatively poor "mothering competence" for mothers of children who had been clinically malnourished were reported from Jamaica [44]; these mothers tended to be rather passive, have low self-esteem, and be socially isolated. Data suggesting that children hospitalized for clinical malnutrition early in life may have experienced less than optimal psychosocial caregiving have been reported from Jamaica [45] and Barbados [46].
The transactional nature of the child-caregiver linkage is illustrated by a number of studies; poor psychosocial care is often a response to a child's poor health or low nutrition status. Observations of chronically undernourished 7- to 18-month-old children and their mothers in West Bengal [47] revealed less maternal responsiveness and mutual interaction. Although these differences were not replicated in a parallel study in Nepal [48], in both studies the undernourished children tended to show a greater need to maintain close proximity to the mother, which might be viewed as a sign of attachment insecurity. Similarly, mothers of undernourished 5- to 11-month-old Chilean babies were observed to show significantly less positive non-verbal emotional responsiveness (through facial expression and touch) than mothers of adequately nourished children, both in a feeding situation and when asked to talk about their child [49].
Consistent with the above findings suggesting attachment insecurity, a recent study [50] reported that Chilean toddlers from low-income homes who had been chronically underweight showed an insecure attachment relationship with their mothers, as measured with the "strange situation" paradigm. It is interesting to note that the potential role of inadequate mother-infant attachment in the aetiology of clinical malnutrition was also suggested in a nutritional/anthropological study of 20 severely malnourished children under three years of age in an East African village [51]. Although no observations of mother-child interaction were reported, the investigators hypothesized that the basic risk factor was a failure in "bonding" (or attachment), since more malnourished than comparison children were cared for by someone other than the mother, were illegitimate, were premature, and were weaned before the end of the first year. Similar family background factors differentiating young Ugandan children with kwashiorkor from control children have been reported [52].
Several recent studies of mild-to-moderate undernutrition and toddler development in Egypt and Kenya also suggest that a number of specific caregiver behaviours are influenced to some extent by children's nutrition status, as indexed by dietary intake. Based on extensive and detailed home observations of toddlers' rearing experiences with mothers or other caregivers over a period of some 15 months, Sigman et al. [53] found that Kenyan village children with reduced food intake were more frequently held, carried, or given physical care (consistent with earlier findings [54]); these factors also tended to be negatively associated with behavioural development. However, food intake was unrelated to the child's being touched or talked to [55]. Using similar extended observations of Egyptian toddlers, modest relationships were found between toddlers' dietary intake and specific caregiver behaviours, primarily for girls. Comparisons with equivalent Kenyan data suggested similar patterns of relationships for some measures, but not for others, presumably because of differences in cultural practices and/or in toddler dietary intake in the two settings. A child's nutrition status was found to be associated with both the frequency of interactions and whom the child interacted with [56].
Maternal caregiving behaviour may be influenced by the nutrition or growth status of infants as young as 3 to 6 months. For example, rural Indian mothers' behaviour toward low-birth-weight (<2.5 kg) children was compared to their behaviour toward high birth-weight (>3.5 kg) children, observed systematically over an 8-hour period [57]. The findings indicated that smaller babies received fewer shows of affection, smiling, talking to, and stimulation (though more feeding) than larger infants. A study of infants from semi-rural Egyptian households reported that infant weight at 6 months was positively related (r = .53) to number of vocalizations received from caregivers, but not to amount of physical contact, identity or proximity of caregiver, or caregiver responses to infant vocalization or distress [58]. Finally, mothers of six-month-old Kenyan infants of lower weight and length were reported to hold and feed their babies more frequently [59].
Thus, considerable correlational and longitudinal evidence suggests that variations in psychosocial care may be a function of the health or nutrition status of the child, and that poor nutrition status arises in conditions where psychosocial care is less adequate. Although the direction of influence is not always clear in such studies, the findings suggest that it is important to conduct experimental studies to determine whether enhancing the quality of psychosocial care might result in improvements in nutritional care and thus in nutrition status as well. One of the few experimental studies is the analysis by Super et al. of the Bogota experiment that measured the combined effects of a home visiting programme and food supplementation on children's growth and cognitive development [60]. The authors report that three years after the study ended, children who had been in the home visiting programmes only (not the food supplementation group) were better nourished than those who had not received the home visiting. The authors suspect that the parents in the home visiting programmes increased their attention to the nutrition and development of their children, and channeled more resources into food for the children. Other mechanisms could also have been possible.
Risk and protective factors influencing quality of psychosocial care
Three factors influence significantly the quality of psychosocial care: the characteristics of the caregiver, the support system for the caregiver, and resource constraints.
Parenting, risk, and protective factors
A number of strategies have been devised for defining and categorizing these three classes of factors that can be associated with good psychosocial care. Since the most common caregivers are parents, the literature often discusses factors associated with parenting. However, we have used a more inclusive concept of psychosocial care, since a great deal of care is actually provided by siblings and other family members in the third world [61], and also since we will discuss fathers separately from mothers. Most literature suggests that fathers in many developing countries have relatively little contact with young children [62].
Belsky's model of parenting in the United States identifies three primary determinants of parenting: the personal and psychological resources of the parent, the characteristics of the child, and the broader context in which parent-child relationships evolve [63]. Following Bronfenbrenner's ecological model of development, Okagaki and Divecha [64] make a distinction between contextual variables located in the family, such as the nature of the marital relationship, and those located outside the home, including the extended family, neighbourhoods, livelihood system, and parental social support networks.
Several concepts have been used to describe possible threats to adequate psychosocial care. As noted previously, a risk factor is an environmental characteristic that has been associated with and may contribute to poor outcomes for a significant percentage of children, although not all. Examples might be prenatal and perinatal complications, poverty, or lack of social support. Cumulative risk occurs when more than one risk factor is present in a child's life, and may increase the chances for negative outcomes more than one might expect with an additive model.
Protective factors are those characteristics of the caregiving environment or the child that minimize the potential effects of risk (analogous to the positive deviance concept). "We need to understand not only how various factors combine in a cumulative fashion to exacerbate risk but also how they might interact so as to substantially attenuate or minimize potential risks" [31, p. 189]. Examples might be the presence of a supportive grandmother or the development of a positive marital relationship [65]. These factors are not simply an absence of risk factors but may reflect particular strengths despite risk. Finally, some children manage to have positive outcomes despite high levels of cumulative risk in their family and broader environments. These resilient children can teach us a lot about the kinds of interventions that might be effective even if major reductions in environmental risk factors are not possible [66, 67].
The number of possible risk factors investigated in these categories is quite large. We will focus on a few that could have special relevance for nutrition care. At the level of the caregiver, these are beliefs about parenting and parenting efficacy, and depression and anxiety. At the level of the family, we will examine alternative care systems in the home, including the father, and autonomy and income control for the woman in the family. At the level of the broader environment, we will consider alternative child care and social support.
