«Early Family Trauma and the Ontogeny of Glucocorticoid Stress Response in the Human Child: Grandmother as a Secure Base Mark V. Flinn Department of ...»
02--Flinn & Leone--31-68 8/7/06 12:48 PM Page 31
Early Family Trauma and the Ontogeny of
Glucocorticoid Stress Response in the Human
Child: Grandmother as a Secure Base
Mark V. Flinn
Department of Anthropology and
Department of Psychological Sciences
University of Missouri
David V. Leone
Department of Anthropology
University of North Carolina
Abstract: Loss of a parent by death or divorce is among the most traumatic experiences
faced by a human child. Exposure to early family trauma (EFT) can have long-term effects on the limbic hypothalamic-anterior pituitary-adrenal cortex (HPA) axis and other components of neuroendocrine stress response. Because the HPA system has important immunomodulatory functions, children that have endured difficult family social conditions during development may be at higher risk for immune dysfunction and other health problems.
Here we investigate the mediating effects of grandparental relationships, mostly maternal grandmothers, on the ontogeny of HPA stress response among EFT children.
Sample and methods: A longitudinal eighteen-year study (1988–2006) of child health in a rural community on the island of Dominica provides sequential biyearly data on salivary cortisol, morbidity, growth, and social environment (N= 282 children, 26,738 saliva samples from children + 5,470 saliva samples from their parents and other kin, 89,109 health observations, 46,788 anthropometric measurements, 42 months of residence in the community studying social environment). Patterns of cortisol response, extent of care from grandparents, morbidity, growth, and immune function are examined for associations with trauma at different periods during development.
Results: Grandparents are a common and important mediator of child wellbeing in this community. EFT children with extensive grandparental care are more likely to recover normal HPA function than are EFT children with little or no grandparental care. Lengthy informal discussions with children and their caregivers over the past 18 years suggest that grandparents are important components of their social worlds, including roles as temporary and permanent replacements for parents. EFT children with low contact rates with grandparents have more erratic cortisol profiles, higher morbidity, more growth disruptions, and more behavioral problems than EFT children with extensive g
Summary and conclusions: Grandparents appear to have significant mediating effects on child development in this population. Intensive grandparental care is positively associated with lower average cortisol, normal growth, and lower morbidity. No significant associations were found between grandparental care and parasite loads or average fluctuating asymmetry. Overall these results suggest the importance of grandparents, especially maternal grandmothers, for children who have experienced early trauma in their family environments.
Introduction: Grandparenting as an Evolved Human Adaptation “I love my granbaby more than anything God put here on this earth”... “I love Mama”1... mutual hug and laugh. (Bwa Mawego2 grandmother and her granddaughter, from MVF fieldnotes, Oct. 10, 1992).
Grandparents and grandoffspring share 25% of their genes identical by descent, a significant opportunity for kin selection. Few species, however, live in groups with multiple overlapping generations of kin. Fewer still have significant social relationships among individuals two or more generations apart—such as elephants (Lee, 1987), some cetaceans (Mesnick, Evans, Taylor, Hyde, Escorza-Trevino, & Dizon, 2003; Whitehead, 2003), and some primates (Goodall, 1986; Lancaster & King, 1992).
Humans appear rather exceptional in this regard. Grandparenting is cross-culturally ubiquitous and pervasive (Murdock, 1967; e.g., Clarke, 1957; Sear, Mace, & McGregor, 2000). Our life histories allow for significant generational overlaps, including an apparent extended post-reproductive stage facilitated by the unique human physiological adaptation of menopause (Alexander, 1974, 1987; Hawkes, 2003; Williams, 1957).
The neuroendocrinological mechanisms guiding attachment processes in grandrelationships—such as the intense affiliation described in the anecdote above—are uncertain. The maternal neuropeptide oxytocin is a likely candidate. Regardless, the significance of emotional bonding between grandparents and grandchildren is beyond doubt. The evolved functions are uncertain, but likely involve the extraordinary importance of long-term extensive and intensive investment for the human child. The emotional and cognitive processes that guide grand-relationships evolved because they enhanced survival and eventual reproductive success of grandchildren.
Throughout human evolution, most children were likely to have benefited from the extra care provided by grandparents in addition to the investment by the immediate family. Children that for whatever reasons were not receiving much care from one or both parents were at risk, and, therefore, especially dependent upon help from grandparents. In addition to the physical basics of food, protection, and hygienic care, development of the human child is strongly influenced by the dynamics of the social environment (Dunn, 2004; Hetherington, 2003a, 2003b; Hinde & Stevenson-Hinde, 1987; Konner, 1991). Monitoring neuroendocrine stress response systems can provide useful insights into how children are coping with social challenges. Here we first reThis three year-old child referred to her grandmother as “Mama,” and her mother by her given name.
2. All place names and research participant names used here are pseudonyms to protect privacy rights.
02--Flinn & Leone--31-68 8/7/06 12:48 PM Page 33
view the potential functions of physiological stress response to stimuli in the social environment. We then investigate the effects of grandparents on the ontogeny of stress response among children that have endured the loss of a parent by death or separation.
The Paradox of Psychosocial Stress Humans elevate levels of the stress hormone cortisol in response to social challenges (Dickerson & Kemeny, 2004; Gunnar, Bruce, & Donzella, 2000). The magnitude and temporal pattern of cortisol release by the limbic hypothalamic-anterior pituitary-adrenal cortex (HPA) system can be altered by traumatic early experiences (Heim, Newport, Heit, Graham, Wilcox, Bonsall, Miller, & Nemeroff, 2000; Mirescu, Peters, & Gould, 2004; Weaver, Cervoni, Champagne, D’Alessio, Sharma, Seckl, Dymov, Szyf, & Meaney, 2004). High levels of the primary stress hormones cortisol, corticotropin releasing hormone (CRH), and epinephrine may be toxic to developing dendrites, neurons, and receptors in the hippocampus and other brain regions that are involved in the negative feedback loop for cortisol regulation (Chen, Bender, Brunson, Pomper, Grigoriadis, Wurst, & Bararn, 2004; Sapolsky, Romero, & Munck, 2000). High levels of cortisol may also affect sensitivity of fear responses mediated by the amygdala. Hence early trauma may lead to permanent HPA-cortisol dysregulation. Because the HPA system has important immuno-modulatory functions (Ader, Felten, & Cohen, 2006; Sapolsky, 2005), children that have endured difficult social conditions during development may be at higher risk for immune dysfunction and other health problems.
The sensitivity of the HPA system to social stimuli presents an evolutionary paradox: why release cortisol if it has negative effects on health? We do not have good explanations for why natural selection favored links between the neuropsychological mechanisms involved with assessment of the social environment and the neuroendocrine mechanisms that regulate stress hormones. We also do not understand why these links are modifiable during development, such that early traumatic experiences may permanently alter hormone response to social challenges. And finally, we do not understand what factors, such as social support, may subsequently ameliorate the negative effects of early trauma on stress response (Caldji, Tannenbaum, Sharma, Francis, Plotsky, & Meaney, 1998; Cameron, Champagne, Parent, Fish, Ozaki-Kuroda, & Meaney, 2005; Francis, Diorio, Plotsky, & Meaney, 2002; Lupien, Fiocco, Wan, Maheu, Lord, Schramek, & Thanh Tu, 2005).
We approach these questions from an evolutionary paradigm that integrates proximate physiological explanations with ontogeny, phylogeny, and adaptive function (Tinbergen, 1963). Here we first briefly review the idea that childhood is an adaptation for coping with an increasingly complex and dynamic social and cultural environment. We then explore relations between physiological stress response and the ontogeny of social competencies. We posit that one of the important functions of the stress response system, in connection with emotional states such as fear or anxiety, is to manage the direction of mental processes to solving specific problems (Ademec, Blundell, & Burton, 2005; Flinn, 2006b, 2006d; Huether, 1998; Kaiser & Sachser, 02--Flinn & Leone--31-68 8/7/06 12:48 PM Page 34
34 MARK V. FLINN AND DAVID V. LEONE
2005; Meaney, 2001; Pitman, 1989; Rodriguez Manzanares, Isoari, Carrer, & Molina, 2005). For example, in the relatively straightforward prey-predator situation of a gazelle smelling a lion, a “freeze” response, focused sensory acuities, and neurological circuits for escape behaviors may be enabled. The human child may face more cognitively challenging problems that use more information processing capacity, such as complex social interactions (Roth & Dicke, 2005). For example, when dealing with the challenge of making friends on the first day of school, a child needs to allocate her cognitive efforts to the tasks at hand: prepare for immediate contingencies by recalling salient information, enhancing relevant sensory input, and pre-activating circuits for appropriate actions. Stress hormones may enable not only the acute responses to such challenges, but facilitate their modification during development as well. Parents and grandparents may play an important role in this dynamic interplay between stress response to social challenges and the ontogeny of a child’s coping mechanisms by providing both security and information.
Hypotheses are evaluated with analyses of data from an 18-year study of child stress in a rural community on the island of Dominica. The longitudinal depth, large sample size (30,122 salivary cortisol measures from 282 children and their caregivers), and naturalistic paradigm provide a unique research design for investigating relations between social environment and ontogeny of stress response. Empirical analysis is complicated by the pleiotropic nature of the key stress hormone cortisol.
Moreover, the Dominica study does not have neurological data, hence direct or strong demonstrations of causal links among stress response, neural plasticity, family environment, grandparental care, and ontogeny of social competencies are not possible.
Our initial objective here is to review a plausible model that links stress response to the neural plasticity that enables adaptation to the dynamic human social environment. We then assess the effects of early family trauma on the ontogeny of stress response, and the potentially ameliorative effects of grandparental care. Understanding these relations may have significant consequences for child development and public health because it could provide new insights into associations among stress response, social disparities, and perinatal programming, among other outcomes (Barker, 1998;
Dressler, Oths, & Gravely, 2005; Heim & Nemeroff, 2001; Maccari, Darnaudery, Morley-Fletcher, Zuena, Cinque, & Van Reeth, 2003; Marmot, 2004).
Ordered down from the tree, Danny headed quickly home, head bowed in apparent numb submission (vignette from MVF field notes, July 14, 1994). Danny’s salivary cortisol level rose from 2.2 to 3.8 µg/dl in little more than an hour. That afternoon, his secretory immunoglobulin-A levels dropped from 5.70 to 3.83 mg/dl. Three days later he had common cold symptoms: runny nose, headache, and fever. His two companions resumed their morning play, exhibiting a normal circadian decline in cortisol, and remained healthy over the next two weeks (Figure 1).
Danny had lost his father in a fishing accident when he was 3 years old. His relationships with his grandparents were not close; he visited them less than one hour a week. His friend James also did not have a co-resident father, as a consequence of a conjugal separation, but in contrast with Danny’s situation, James’ relationship with his maternal grandmother was very close, as illustrated by the following vignette.
James was sitting contentedly sorting rice with Ma Tee-Jean, his maternal grandmother. He startled when I (MVF) kicked a rock on the path to forewarn them of my approach into the household yard, and he inched closer to grandma, who put her arm around him. James was dressed and ready for school. He had left his mother’s house when an argument had broken out between his mother and stepfather last night, walking the 200 meters up the hill in the dark with his sister (MVF field notes, Sept 14, 1992). He was accustomed to this situation; James slept 44 out of 62 nights there in this summer of 1992, splitting his time between both households for most of the past
four years, since his father had left when James was three years old and his stepfather had slowly moved in. James had always been close to his grandmother—whom he called “Mama”—as well as his mother—whom he called by her given name Lily.
James’ cortisol levels were often elevated by family conflicts, but the solace he usually found at his grandmother’s often resulted in a return to normal levels.
These anecdotal case examples contribute to a common pattern. Children in this rural Dominican community are more than twice as likely to become ill during the week following a stressful event than during a week when they had not recently experienced any significant stressors (Flinn & England, 2003). Chronic elevation of cortisol levels may have negative consequences for health (Cohen, Doyle, Turner, Alper & Skoner, 2003; Maier, Watkins, & Fleschner, 1994; Marmot & Wilkinson, 1999;