Pathophysiology

There are two cardinal pathophysiologic features of sickle cell disease: chronic hemolytic anemia and vasoocclusion. The polymerization of the hemoglobin S molecule (Hb S) within the red blood cells upon deoxygenation causes the red blood cells to change from the usual biconcave disc to an irregular sickled or crescent shape. Upon reoxygenation, the red cell initially resumes a normal configuration but after repeated cycles, the erythrocyte is damaged permanently, resulting in red cell dehydration and erythrocyte destruction. Sickled red blood cells also have a propensity to adhere to the walls of blood vessels and are susceptible to hemolysis, causing chronic anemia (Ashley-Koch et al., 2000). The deformed red blood cells cause microcirculatory obstruction and prevent normal blood flow and decreased delivery of oxygen to organs and tissues resulting in the vasoocclusive crisis. However, information summarized by Stuart and Nagel (2004) indicates that the actual mechanism is more complicated.

One of the factors complicating the pathophysiology is cell heterogeneity. Sickle cells vary in their density and deformity because cation homeostasis is impaired in some cells. The amount of hemolysis is related to the number of irreversibly sickled cells and dense cells (Steinberg and Rodgers, 2001). Another factor that varies is fetal hemoglobin (Hb F) concentrations. Vasoocclusive events depend on the interaction of features intrinsic to the sickled erythrocyte, including degree of polymer formation and cellular damage, interacting with other factors in the cells environment such as endothelial cells and leukocytes (Steinberg and Rodgers, 2001). Other potentially contributing factors include neutrophil transmigration that “adds to the increased inflammation in the microvascularture” and “disregulation of vasomotor tone by perturbations in vasodilator mediators such as nitrous oxide (NO)” (Stuart and Nagel, 2004, p. 1345). The abnormal cation homeostasis contributes to dehydrated dense sickle cells which in turn contributes to anemia and hemolysis.

The recognition that the adherence of sickled erythrocytes to the endo-thelium correlated with disease severity focused attention on the mechanisms involved (Stuart and Nagel, 2004). As a barrier between blood and tissue, endothelial cells have a number of functions that may contribute to the vascular pathology of sickle cell disease and “genetic differences are likely to cause different responses among patients” (Steinberg and Rodgers, 2001, p. 300). One of the functions of endothelial cells is to control vascular tone by elaborating vasoconstrictors and vasodilators. Endothelial cells also express genes adhesion molecules for blood cells and proteins (Steinberg and Rodgers, 2001). Endothelial cell activators are generated by a number of factors such as hypoxia, thrombin, and infection (Steinberg and Rodgers, 2001). Other extra-erythrocyte related pathophysiological factors include leukocyte size, rigidity, and adhesive characteristics and coagulation activation, with thrombin hypothesized as potentially providing a crucial link between coagulation activation and adhesion (Stuart and Nagel, 2004). Of particular interest is the finding that laminin bonds strongly to sickle erythrocytes via the protein that carries Lutheran blood-group antigens (B-CAM/ Lu) and epinephrine increases this adhesion. “Since stress is a potential initiation factor for vasoocclusion, epinephrine modulation of adhesion provides a powerful biological link between intraerythrocytic signaling pathways and the external milieu” (Stuart and Nagel, 2004, p. 1346).

Clinical Manifestations

Two primary consequences of hypoxia secondary to vasoocclusive crisis are pain and damage of organ systems. The organs at greatest risk are those where blood flow is slow, such as the spleen and bone marrow, or those with a limited terminal arterial blood supply, including the eye and the head of the femur and humerus, and lung as the recipient of deoxygenated sickle cells that escape the spleen or bone marrow. Major clinical manifestations of sickle cell disease include painful events, acute chest syndrome, splenic dysfunction, and cerebrovascular accidents.

Painful events occur as a result of ischemic tissue injury and can be precipitated by hypoxia, dehydration, and extreme cold. The frequency and severity of painful events are varied. Musculoskeletal pain is the most common, followed by abdominal pain, and low back pain. Painful events typically last 4-6 days. Transduction is the process whereby noxious inflammatory mediators that are generated by tissue damage in turn activate nocioceptors to chemical or mechanical forms of energy to an electrochemical impulse, which is transmitted along the spirothalamic tract to the thalamus which in turn transmits the signal to the brain where it is perceived as pain (Ballas, 2001a). Descending fibers in the midbrain can inhibit the transmission of painful stimuli via endogenous endorphins and communi cations through the limbic system can modulate the emotional response to pain and thereby enhance or inhibit the intensity of the perception of pain (Ballas, 2001a).

Acute chest syndrome involves chest pain, fever, increased leukocytosis, hypoxemia, and pneumonia-like symptoms. Typical causes include infection and pulmonary infarction (Ballas, 2001b). This acute illness can be self-limiting or can rapidly progress and may be fatal. “Risk factors include HB SS genotype, low HB F concentrations and high steady state leukocyte and HB concentrations” (Stuart and Nagel, 2004, p. 1350).

Splenic dysfunction develops during infancy and predisposes the infant to overwhelming infection from encapsulated bacteria, particularly streptococcus pneumonia and haemophilus influenza. Between the ages of 5 months and 2 years, children with sickle cell anemia are at risk for sudden intrasplenic pooling of vast amounts of blood, known as splenic sequestration. The hemoglobin level can drop precipitously, causing hypovolemic shock and death. High concentrations of Hb F serve as a protection factor (Stuart and Nagel, 2004).

Stroke affects 6-12% of patients with sickle cell disease. In children, the most common cause of stroke is cerebral infarction; intracerebral hemorrhages become increasingly common with age. Recurrent stroke causes progressive impairment of cognitive functioning. “Risk factors include the HB SS phenotype, previous transient ischemic attacks, low steady state HB concentrations, high leukocyte counts, raised systolic blood pressure, and previous acute chest syndrome” (Stuart and Nagel, 2004, p. 1351). Silent brain lesions have been evidenced on magnetic resonance imaging (MRI) accompanied by neurocognitive deficits (Armstrong et al., 1996).

The efforts to enhance clinical care are focusing on increasing understanding of the pathophysiology of sickle cell disease to enable a precise prognosis and individualized treatment. What is required is knowledge about which genes are associated with the hemolytic and vascular complications of SCD and “how variants of these genes interact among themselves and with their environment” (Steinberg, 2005, p. 465).

Genetic Modulation of Disease Severity

Individual differences occur in part through differences in the order and pattern of gene expression (i.e., variations in the regulatory sequence of the genome, referred to as promoters). A promoter is a special sequence of bases usually found immediately upstream of the gene itself. A gene is expressed or transcribed into messenger RNA by the binding of a protein called a transcription factor to a promoter. The binding of a transcription factor and the expression of a gene can be altered by experience (Ridley, 2004). An example is the elevation of cortisol that occurs upon appraisal of a situation as stressful, which in turn alters gene expression in the immune system by reducing the expression of interleukin II and turning down the activity, number, and life span of lymphocytes (Ridley, 2004).

Two broad molecular genetic strategies have been employed to identify the role of genes (de Gues, 2002). One strategy involves whole genome scans through linkage analysis. The advantage of this approach is that all relevant genes are examined but the disadvantage is that it requires large samples of genetically related subjects. A second approach is an allelic association or candidate gene studies. Associations with known functional candidate genes are investigated, for example “genes suspected to influence neurotransmission in the brain because they code for protein constituents of receptors, transporters, or enzymes involved in neurotransmitter synthesis and degradation (Plomin and Crabbe, 2000)” (de Gues, 2002, p. 4). The advantage of this approach is the ability to use smaller samples of unrelated subjects but the disadvantage is that some genetic influences are missed because they are not among the candidate genes studied.

It is easier to identify the effect of the gene on a more elementary trait than on a complex one. The strategy is to identify an endophenotype that is upstream of the more complex effect, determine the amount of variance that the gene explains in the endophenotype, and then determine the variance explained in the disease outcome by the endophenotype. Identifying allelic candidate genes is a matter of looking for genes that are part of a system known to influence the disease. The genes influence the disease by influencing the concentration of a protein or its functionality or efficiency or responsiveness to the environment.

Ridley (2004) maintains that “Diversity in the human population is starting to be explained at least as much by variations in the number of repeats of a genetic phrase in the regulatory region of the gene as by single-nucleotide polymorphisms” (p. 97). “Varying the number of repeats of a phase has a much subtler effect on gene function then does changing a single nucleotide in a codon, which tends to shut a gene down” (p. 97).

Steinberg (2005) views the use of Bayesian networks as a promising approach for the discovery of the genetic basis of complex traits in large association studies and describes a Bayesian network that was developed to analyze 235 single nucleotide polymorphisms (SNPs) in 80 candidate genes in 1398 unrelated patients with sickle cell anemia. The findings indicated that “SNP’s on 11 genes and four clinical variables, including α-thalassemia and Hb F, interacted in a complex network of dependency to modulate the risk of stroke. This network of intersections included three genes, BMP6, TGFBR2, and TGFBR3 with a functional role in the TGF-β [transforming growth factor-β pathway and one gene (SELP) associated with stroke in the general population” (Steinberg, 2005, p. 472). Subsequently, this model was validated by predicting the occurrence of stroke in a different popula tion with a true positive rate of 100%; a true negative rate of 98.14%; and an overall predictive accuracy of 98.2% (Sabastiani et al., 2005). In his comprehensive review of predictors of SCD complications, Steinberg (2005) considers both established predictors, including fetal hemoglobin and α-thalassemia, and potential predictors.

Fetal Hemoglobin

Fetal hemoglobin (Hb F) inhibits Hb S polymerization and higher levels are associated with a reduction of most vasoocclusive complications of sickle cell anemia (Steinberg, 2005). However, Hb F concentrations vary among patients with sickle cell anemia, ranging from 0.1% to 30%, and there is considerable variability in severity of complications among patients with similar concentration levels.

Typical levels of Hb F vary across the four major β-globin haplotypes. The highest Hb F level and mildest clinical course is found in carriers of the Hb S gene on the Senegal or Arab-India haplotype, intermediate levels and severity on the Benin haplotype, and the lowest levels and most severity on the Bantu (Central African Republic) haplotype (Steinberg, 2005).

Fetal hemoglobin expression is a quantitative trait and investigations are addressing complex interactions among transcription factors, genes modulating erythropoiesis, and elements linked to the β-globin cluster. In addition, similar genetic analyses are being undertaken in an effort to predict responsiveness to hydroxyurea, which is used to treat the complications of SCD and is thought to work by increasing Hb F levels (Steinberg, 2005).

α-Thalassemia

Alpha thalassemia is the result of the deletion of one of two α-globin genes from a chromosome (Nagel and Steinberg, 2001). Coincidental α-thalassemia occurs in approximately 30% of patients with sickle cell anemia and affects the phenotype of sickle cell anemia by reducing the concentration of Hb S polymerization (Steinberg, 2005). The presence of α-thalassemia with sickle cell anemia is also associated with less hemolysis, higher concentration of hemoglobin (Nagel and Steinberg, 2001) and higher packed cell volume (PCV), and lower mean corpuscular volume and reticulocyte counts (Steinberg, 2005). However, the clinical effects of co-existing α-thalassemia are mixed. Benificial effects are generally found with vasoocclusive events that are dependent on PCV, such as stroke and leg ulcer, whereas deleterious effects are associated with complications that are dependent on blood viscosity, such as painful episodes and acute chest syndrome (Steinberg, 2005).

Since the diversity of sickle cell anemia cannot be explained entirely by Hb F and α-globin gene-linked modulation, attention is being directed to epistatic or modifying genes that act independently of Hb S polymerization. The genes that potentially could modulate the phenotype of sickle cell anemia include: “mediators of inflammation, oxidant injury, NO biology, vasoregulation, cell-cell interaction, blood coagulation, haemostasis, growth factors, cytokine and receptors and transcriptional regulators” (Steinberg, 2005, p. 470). However, studies of candidate genes, seeking associations of SNP with phenotypes, are in the beginning stages and present many interpretative challenges (Steinberg, 2005).

Psychological Adjustment

The findings with regard to the psychological adjustment of children with sickle cell disease are consistent with those for children with chronic illnesses in general (Thompson and Gustafson, 1996). The risk of psychological adjustment problems in children with chronic illness is 1.5 to 3 times as high as with their healthy peers (Thompson and Gustafson, 1996). In addition to determining the type and frequencies of adjustment problems, effort has been directed to identifying the mediating and moderating role of illness parameters, typically disease severity, and psychological and social processes to adjustment to the stress of chronic illness. The transactional stress and coping model (Thompson and Gustafson, 1996; Thompson et al., 1992) has proven to be a useful conceptual framework for these investigations and psychological adjustment was the target of a number of studies done through the Duke University of North Carolina Sickle Cell Center.

Psychological adjustment was assessed in a study of 50 children, age 7 to 17 years of age with sickle cell disease, (Hb SS 60%; Hb SC 12%; sickle β-thalassemia syndromes 16%). In terms of mother reported behavioral problems, 64% of the children were classified with poor adjustment, primarily of the internal behavior problem type. In terms of child self-report, as assessed through a semi-structured diagnostic interview, 50% reported symptoms that met the criteria for one or more DSM-III diagnosis. Internalizing problems reflected in anxiety, phobic, and obsessive-compulsive diagnoses were most frequent. In contrast, externalizing problems reflected in conduct disorder and oppositional disorder were relatively infrequent. Hierarchal multiple regression analysis was utilized to assess the increment in psychological adjustment accounted for by maternal psychological adjustment and children’s cognitive processes and pain coping strategies over and above that accounted for by demographic parameters and illness severity parameters, including type of sickle cell disease, pain frequency, pain severity, and number of complications. In terms of the variance in mother-reported internalizing behavioral problems, the demographic variables of gender, socioeconomic status, and age accounted for 8% and the illness parameters of pain frequency and type of sickle cell disease accounted for another 9% and 8%, respectively. Maternal anxiety accounted for 16% of the variance in mother-reported internalizing behavioral problems and 33% in mother-reported externalizing behavioral problems. In terms of child-reported total symptom score, sickle cell type did not account for any of the variance, the number of illness complications accounted for 2%, and pain frequency accounted for 1%. The demographic variables of socioeconomic status and gender only accounted for 6% of the variance. However, children’s pain coping strategies characterized by negative thinking accounted for a 21% increment in child reported total symptom score.

Psychological adjustment over time was assessed at 3 points across 2 years with a sample of 50 children with sickle cell disease (Hb SS, 54%; Hb SC, 34%; sickle β-thalassemia, 12%; males, 64%; females, 36%). In terms of child-reported symptoms, 12% met diagnostic criteria for a DSM-III diagnosis across all three time points whereas 17% consistently demonstrated good adjustment. The variability in report of symptoms meeting diagnostic criteria over time is also reflected by the percentage of children who had 1 (49%) or 2 (27%) changes in adjustment classification over the three-time periods. In terms of specific diagnoses, internalizing disorders were most frequent at each time but there was very little consistency in specific diagnoses across time. In terms of mother-reported behavioral problems, 47% met the criteria for poor adjustment and 19% for good adjustment across all three assessment points. One change in classification occurred for 25% and two-changes occurred for 4% (Thompson et al., 1999a).

Maternal psychological adjustment was assessed in a study of 78 mothers of children and adolescence, 7 to 17 years of age, with sickle cell disease (Hb SS, 62%; Hb SC 23%; and sickle β-thalassemia syndromes; 15%) (Thompson et al., 1993b). In terms of self-reported symptoms of psychological distress, 36% of mothers’ met criteria for poor psychological adjust ment. None of the illness or demographic parameters accounted for significant amounts of variance in mothers’ symptoms scores. Over and above the 7% of variance in adjustment accounted for by illness parameters and 2% by demographic parameters, a 46% increment in variance was accounted for by three psychosocial processes: mother-reported use of palliative coping in relation to active coping (30%); stress associated with daily hassles (13%); and family functioning characterized by an emphasis on control (3%) (Thompson et al., 1993a).

In a study of maternal adjustment across three assessment points over 2 years, 43% of mothers of children with sickle cell disease consistently met the criteria for poor adjustment (Thompson et al., 1999b). Mothers with stable good adjustment differed significantly from those with stable poor adjustment in terms of lower levels of daily stress and use of palliative coping methods in relation to adaptive coping and lower levels of illness-related stress.

The psychological adjustment of adults with sickle cell disease was assessed in a sample of 109 patients (female, 55%; male, 45%;) ranging in age from 18-68 years (Hb SS, 77%; Hb SC, 12%; and sickle β-thalassemia syndromes, 11%). The criteria for poor adjustment in terms of self-reported symptoms of psychological distress was met by 56% of the patients with 40% demonstrating elevations into the clinical range of distress on depression and 32% on anxiety. In terms of illness parameters, type of sickle cell disease and number of complications accounted for no significant increment in reported psychological distress and pain frequency only accounted for a 2% increment. Similarly, the demographic parameters of socioeconomic status, gender, and age only accounted for an additional 9% of the variance. In contrast, with these variables controlled, daily stress accounted for an additional 35% of the variance in reported psychological distress and pain coping strategies characterized by negative thinking accounted for an additional 4% (Thompson et al., 1992).

The stability of psychological adjustment across three time periods spanning 20 months was assessed in a study of 59 African American adults with sickle cell anemia (Thompson et al., 1996). In terms of self-reported symptoms of psychological distress, consistently poor adjustment was demonstrated by 32% and consistently good adjustment by 25% of patients. Variability in adjustment at the individual level was also reflected in 26% of the patients changing classifications once and 17% changing twice over the three assessment points. With adjustment at the 20-month follow-up period as the outcome measure, the illness parameters of complications and pain frequency at baseline did not account for any significant increment in variance and the demographic parameter of socioeconomic status only accounted for 6%. With illness and demographic parameters controlled, baseline levels of daily stress accounted for a 29% increment in psychological distress at 20-month follow-up and illness-related stress accounted for another 8% increment.

A multisite Cooperative Study of Sickle Cell Disease (Farber et al., 1985; Gaston and Rosse, 1982) provided an opportunity to examine the independent and combined contributions of family functioning and neurocognitive functioning to behavioral problems in children with sickle cell disease. In an initial cross-sectional study of 289 children (Hb SS, 68%; Hb SC, 32%; males, 52%; females, 48%) 5.9 to 15.5 years of age completed a neuropsychological evaluation, brain MRI and mothers completed the child behavior checklist and family environment scale (Thompson et al., 1999a). Mother-reported behavior problems occurred with 30% of the patients. The subgroup with behavior problems had significantly lower verbal IQ, reading, and math scores and lower levels of family support and higher levels of family conflict. The rate of behavioral problems did not vary across the three subgroups formed on the basis of MRI status (normal, clinically apparent cerebral infarction, and silent infarction). Demographic parameters of child age and gender and mother age and education and the biomedical parameters of hematocrit level and type of SCD each accounted for only 2% of the variance in behavioral problems. However, family functioning characterized as conflicted, reflecting both high levels of conflict and a lack of organization and support, accounted for a 19% increment in variance in behavioral problems.

The relationship of behavioral problems, intellectual functioning, and family functioning was assessed longitudinally in a follow-up prospective study of 222 children with at least two complete sets of measures obtained across four assessment points over the study period of nine years. The findings indicated that overall 60% of the children were consistently classified in terms of behavioral problems (9%) or good adjustment (51%) based on at least three measures across four assessment points. The risk of consistent behavior problems was not related to MRI classification, gender, education level of the mother, or age of the child but significantly increased with higher baseline levels of family conflict and decreased with higher baseline full-scale IQ. More importantly, an increase in behavioral problems was associated with a reported increase in family conflict but was not related to change in intellectual functioning. There was a decline in neurocognitive functioning over time. On average, full-scale IQ decreased 1.2 points per year with age and compared with a child with a normal MRI, was 3.8 points lower for a child with silent infarction and 14.4 points lower for a child with stroke.

In summary, the findings across a number of studies indicate an increased risk for psychological adjustment problems in children and adolescents with sickle cell disease and their mothers and adults with sickle cell disease. However, there is considerable variability in adjustment over time and good adjustment is the norm. Biomedical indicators of disease severity, including type of sickle cell disease and number of complications and frequency of pain episodes, account for very little variance in psychological adjustment. Similarly, demographic factors of gender, age, and socioeconomic status also account for very little variance in adjustment. In contrast pain coping strategies characterized by negative thinking and passive adherence account for a significant portion of variance in children and adolescent psychological adjustment. Stress processing variables account for a significant portion of variance in the adjustment of adult patients with sickle cell disease and mothers of children and adolescents with sickle cell disease. More specifically appraisals of stress, especially daily stress, use of palliative coping strategies, and family functioning characterized by low levels of supportiveness and high levels of conflict account for significant increments in adjustment variance over and above that accounted for by illness and demographic factors.

Neurocognitive Functioning

The cooperative study of sickle cell disease included neuropsychological and MRI assessment of children 6-12 years of age (Armstrong et al., 1996). For children with Hb SS disease, 6.6% had a clinical CVA and 15.6% had a silent infarct. For children with Hb SC disease, none had evidence of CVA and 5.1% demonstrated a silent infarct. For children with Hb SS disease, those with a history of stroke had a significantly lower verbal, performance, and full-scale IQ scores and math achievement scores than children without MRI abnormalities and significantly lower performance and full-scale IQ scores than with children with silent infarcts. In turn, children with silent infarcts had significantly lower verbal scale IQ scores than children without MRI abnormalities.

The independent and combined contribution of biomedical risk and parenting risk to child neurocognitive functioning was assessed in a study of young children with sickle cell disease through 3 years of age (Thompson et al., 2002). The study sample included 89 African American children with sickle cell disease (Hb SS, N = 55; Hb SC, N = 27; and other, N = 7). Measures of cognitive and psychomotor development were obtained at 6, 12, 24, and 36 months of age. There was no significance decrease in psychomotor functioning (PDI) over time but cognitive functioning (MDI) declined, with a significant decrease occurring between the 12- and 24-month assessment points. By 24 months of age, 29% of the children have MDI scores and 24% had PDI scores more than one standard deviation below the mean for the normative group. There were no significant differences in MDI or PDI scores at any assessment time as a function of type of sickle cell disease. However, multiple regression analyses of developmental outcome at 24 months of age indicated that maternal learned helplessness attributional style accounted for 20% of the variance in MDI followed by type of sickle cell disease which accounted for another 22% increment in variance. The findings indicated that developmental functioning at 24 months of age was associated with both Hb SS phenotype and maternal learned helplessness attributional style, with parenting processes as the hypothesized mechanism of effect.

Sickle Cell Disease Pain

Painful episodes or crises are a cardinal aspect of sickle cell disease. The onset of pain is not predictable and the duration of pain is highly variable from a few hours to several days. Sickle cell disease pain has been associated with increased utilization of health care services, decreased social activities, and increased frequencies of psychological distress in children, adolescents, and adults with sickle cell disease (Gil et al., 1991). A number of studies have addressed the relationship of disease severity, demographic parameters, and pain coping strategies on pain associated with sickle cell disease (Gil et al., 1991).

In a study of 72 children and adolescents, ranging in age from 7 to 17 years, the relationship of reported pain coping strategies was assessed with three outcome measures: health care utilization, reflected in the medical record; psychological adjustment, as assessed through a structured diagnostic interview; and reduction in activities reported by parents (Gil et al., 1991). The Coping Strategies Questionnaire (Rosenstiel and Keefe, 1983) was used to assess pain coping strategies and three major patterns of coping were identified. Negative thinking is a pattern of coping in which children engage in catastrophizing and self-statements of fear and anger. Passive adherence is a pattern of coping in which children relied on concrete, passive strategies, such as resting. Coping attempts is a pattern in which the patient used multiple cognitive and behavioral strategies to deal with pain, such as diverting attention and calming self-statements. The sample included patients with Hb SS disease, Hb SC disease, and sickle β-thalassemia syndromes. Children and adolescents high on the negative thinking and passive adherence patterns were less active in school and social activities, had higher levels of psychological distress during painful episodes, and had higher levels of health care service utilization in comparison to those low on these patterns. Children and adolescents high on coping attempts were more active and required less frequent health care services. These coping strategy patterns accounted for significant portions of variance in household, school, and social activity reduction and emergency room (ER) visits and psychological distress even after controlling for the effects of age and frequency of painful episodes.

The extent to which pain coping strategies measured at baseline predict subsequent adjustment in children and adolescents with sickle cell disease was assessed with 70 patients ranging in age from 7 to 18 years (Hb SS, 58%; Hb SC, 13%; and sickle β-thalassemia syndromes, 29%). With age and pain frequency controlled, baseline levels of pain coping strategies characterized by coping attempts were associated with higher levels of school, household, and social activity during painful episodes. In contrast, baseline patterns of pain coping characterized by passive adherence were associated with more frequent health care contacts. Furthermore, increases in pain coping strategies characterized by negative thinking were associated with further increases in health care contacts and those with less negative thinking over time decreased their health care contacts (Gil et al., 1993).

In a study of 79 adults with sickle cell disease, pain coping strategies characterized by negative thinking and passive adherence were associated with more severe pain episodes, less activity during painful episodes, more frequent hospitalization and ER visits and higher levels of self-reported psychological distress (Gil et al., 1989). In a related study, the relationship between stress, coping, and psychological adjustment was assessed in 109 patients with sickle cell disease (Hb SS, 77%; Hb SC disease, 12%; and sickle β-thalassemia, 11%) ranging in age from 18 to 68 years (Thompson et al., 1992). With self-reported levels of psychological distress as the outcome variable, the demographic parameters of socioeconomic status, gender, and age accounted for 9% of the variance, and illness parameters of pain frequency, number of complications, and type of sickle cell disease accounted for 2% of the variance. Over and above the contribution of these variables, daily stress accounted for a 35% increment in psychological distress and pain coping strategies characterized by negative thinking accounted for an additional 4%.

The relationship between stress and pain was examined in a study of 53 adults ranging in age from 18 to 58 years with sickle cell disease (Hb SS, 85%; Hb SC, 8%; sickle β-thalassemia syndrome, 8%) (Porter et al., 1998). Patients completed daily pain ratings for a 2-week period. Stress was assessed in terms of daily hassles (Kanner et al., 1981). Activity reduction, pain occurrence and intensity, reported medication use, and health care use were not significantly related to type of sickle cell disease or the number of sickle cell disease-related complications. However, higher pain intensity ratings were associated with greater health care use in terms of ER visits, hospitalizations, physician visits, and phone calls. Intensity ratings of daily stress were significantly related to pain intensity levels, and reductions in housework and social activities, even after controlling for pain intensity. Thus higher levels of daily stress were related to greater pain and greater functional impairment (Porter et al., 1998).

This examination of the role of behavioral and social factors in sickle cell disease indicates the limitations of the approaches utilized and thus our state of knowledge. These studies were driven by biopsychosocial conceptual models. However, the specific studies predominantly focus on psychological adjustment as the outcome variable. Demographic and illness parameters, including type of sickle cell disease, pain frequency and severity, and number of complications have little effect on psychological adjustment but stress processing variables including appraisals, coping methods, and social support have a large effect. There are a significant number of studies that have addressed pain as an outcome measure and an emerging research literature on neurocognitive functioning. The findings with regard to pain are similar to those with psychological adjustment. Demographic and illness parameters account for relatively little variance in reported pain but stress appraisal and pain coping strategies account for significant amounts of variance. With regard to neurocognitive functioning, both phenotype and parenting process account for significant amounts of variance. There is a notable lack of studies in which physiological measures of illness severity or complications are included as outcome variables. Furthermore, genetic markers, other than type of sickle cell disease, are for the most part not included in behavioral and psychosocial studies. Although the field has not yet actualized the potential of a full biopsychosocial model, the pathway is discernable. Studies that assess the impact of candidate genes on multiple measures of pathophysiology need to be conjoined with studies that assess the impact of behavioral and social processes on stress and stress processing.

Stress Activation of the Sympathetic Nervous System (SNS)

Activation of the SNS in response to stress results in increased secretions of catecholamines, epinephrine and norepinephrine, and higher levels of catecholamines lead to increase in blood pressure and heart rate and more oxygenated blood glucose is required (Cruess et al., 2004). There is evidence that increased SNS activity is a mechanism for atherogenesis, ventricular hypertrophy, and hypertension (Cruess et al., 2004). There is a large body of evidence that perceived stress, personality characteristics, and specific emotion states, including hostility and depression, are linked to decreases in the neurotransmitter serotonin in particular, and depression may have a link to coronary heart disease (CHD) through the serotonergic system (Cruess et al., 2004). Depressive symptoms are often associated with CHD and there are indications that proinflammatory cytokines mediate this relationship (Cruess et al., 2004).

Stress Activation of the HPA Axis

Stress can also have an effect through over activation of the HPA axis. Psychological stressors elicit a physiological response by activating specific cognitive and affective processes and their central nervous system underpinnings (Dickerson and Kemeny, 2004). Sensory information is integrated and the significance of environmental stimuli is appraised through the thalamus and frontal lobes. These cognitive appraisals can elicit emotional responses through the connections from the prefrontal cortex to structures of the limbic system including the amygdala and hippocampus which connect to the hypothalamus and serve as a pathway for activating the HPA axis (Dickerson and Kemeny, 2004). Activation of the HPA axis is initiated by the paraventricular nucleus of the hypothalamus releasing corticotropin releasing hormone, which in turn stimulates the anterior pituitary to secrete adrenocorticotropin hormone (ACTH) which in turn triggers the adrenal cortex to release the glucocorticoid (GC) cortisol into the bloodstream (Dickerson and Kemeny, 2004). GCs act to restore homeostasis. Cortisol affects metabolism by mobilizing energy resources by elevating blood glucose levels; surpresses the immune system by inhibiting proteins that play a central role in regulating inflammation; and affects the cardiovascular system through the catecholamines and other sympathetic products that induce vasoconstriction (Dickerson and Kemeny, 2004; Herman et al., 2003). “Although the effects of catecholamines are almost immediate and transient, cortisol is slower acting and more likely to influence blood flow and glucose production during prolonged stress responses” (Cruess et al., 2004, p. 43). Prolonged cortisol activation brought about by failure to shut down this response after stressor termination or by frequent exposure to stressors is associated with a number of negative health consequences including immune system suppression, for example decreased lymphocyte proliferation and cytokine production, damage to the hippocampus, and hypertension (Dickerson and Kemeny, 2004). The vasoconstrictive and immunological impact of the activation of the HPA axis is of relevance for sickle cell disease.

The effect of cortisol on tissues is mediated by the glucocorticoid receptor (GR) through direct binding to hormone-responsive elements in the RNA or by interactions with, and modulation of, other transcription factors (Wüst et al., 2004b). The response of a cell to cortisol is a function of the level of the steroid and its GC sensitivity. Variants of the GR gene (located on chromosome 5, locus 5q31) affect sensitivity (Wüst et al., 2004b). Support has been provided for the hypothesis that common polymorphisms in the GR gene may have modulating effects on the HPA response to psychological stress. In a recent study, the impact of three GR gene polymorphisms (BclI RFLP, N363S, and ER22/23EK) on cortisol and ACTH responses to psychological stress and pharmacological stimulation was assessed (Wüst et al., 2004b). In comparison to subjects with two wild-type alleles, 363S carriers showed a significant increased salivary cortisol response to stress whereas the cortisol response of the BclI homozygotes was diminished. This study provides evidence that common polymorphisms of a single gene impact HPA regulation and contribute to the individual variability in response to psychological stress. The impact of genetic factors on HPA axis activity was reported from findings of twin studies and association studies with polymorphisms in the GR gene (Wüst et al., 2004a). In addition, a number of polymorphisms were identified as good candidate genes for future studies (Wüst et al., 2004a).

Evidence suggests that the GCs act through genetic mechanisms, to modify transcription of key regulatory proteins, and by non-genetic mechanisms on cell signaling processes that have a more rapid impact on homeostatic regulation (Herman et al., 2003). The HPA mediated response to stressful stimuli differ depending upon whether the threat to homeostasis is “real” or “predicted.” By real stressor is meant stimuli that are recognized by somatic, visceral, or circumventricular sensory pathways as a challenge to homeostasis. These stimuli include hormonal signals, such as renin-angiotensin, visceral or somatic pain, or humoral inflammatory signals such as blood-borne cytokines signaling infection (Herman et al., 2003). In addition to these “reactive” responses, GC responses can occur in “anticipation” of homeostatic disruption under situations in which threat may be predicted or associated with learned experience. The anticipatory responses are under the control of limbic regions such as the hippocampus, amygdale, and prefrontal cortex (Herman et al., 2003). These two systems act together in an integrated, hierarchal manner. The reactive pathway evokes direct PVN activation whereas the anticipatory pathway involves forebrain processing of polysensorial and associational input that also mediate reactive responses. “The resultant hierarchal organization of stress-responsive neurocircuitries is capable of comparing information from multiple limbic sources with internally generated and peripherally sensed information, thereby tuning the relative activity of the adrenal cortex” (Herman et al., 2003, p. 151). Both genetics and early life experiences can modulate response characteristics of the HPA axis (Herman et al., 2003). Changes in limbic system integration patterns as a function of experience are hypothesized to play a role in HPA axis dysfunction (Herman et al., 2003).

The importance of psychological stress processing for the understanding of the psychobiological stress response is becoming increasingly clear (Gaab et al., 2005). Conceptualizations of stress have moved from that of a stimulus or response to “A relationship between the person and the environment that is appraised by the person as taxing or exceeding his or her resources and endangering his or her well being” (Lazarus and Folkman, 1984, p. 19). A recent study provided support for the role of anticipatory cognitive appraisal, but not general personality factors or retrospective stress appraisal, in the salivary cortisol response to psychological stress (Gaab et al., 2005).

Whereas there is evidence that psychological stressors are capable of activating the HPA axis, the effects are highly variable. For example, several aspects of perceived chronic stress, more specifically worries, social stress, and lack of social recognition, were found to be significantly associated with increased cortisol awakening response (Wüst et al., 2000). To evaluate the characteristics of psychological stressors that evoke a cortisol response, a meta-analysis of 208 empirical studies was undertaken (Dickerson and Kemeny, 2004). The findings indicated that psychosocial stressors that involved social evaluative threat and uncontrollability were significantly associated with increased cortisol response. The findings were also similar for ACTH response. However, psychological distress in and of itself was not associated with increased cortisol response. The findings indicate that only those threats to central goals, such as physical self-preservation or preservation of the social self, and not having control over these situations, triggers cortisol activation. Sickle cell disease provides just such a situation of threat to self-preservation and social evaluative threat and the negative self-appraisals generated under these conditions rather than emotional stress in general could constitute psychological stressors that impact the HPA axis.

Stress and the Immune System

In understanding the relationship of psychosocial stressors to the immune system, Segerstrom and Miller (2004) maintain that it is useful to distinguish between natural and specific immunity. Natural immunity involves cells that do not provide a defense against a particular pathogen but operate broadly in a short time frame. These cells include the granulocytes, both neutrophil and macrophage, which releases cytokines such as interleukin, and natural killer (NK) cells. Specific immunity involves cellular response to intracellular pathogens and humoral responses to extracellular pathogens. Lymphocytes have receptor sites that respond to a specific antigen and when activated divide to create a population of cells in a process referred to as colonal proliferation (Segerstrom and Miller, 2004).

The immune system is of importance in sickle cell disease and one way of examining the impact of genetic, behavioral, and psychosocial processes on the immune system is through the impact of stress and stress processing. There are several ways that stress can affect the immune response (Segerstrom and Miller, 2004). The immune system is regulated both by neural inputs from the sensory, sympathetic, and parasympathetic system as well as by circulating catecholamines and GCs (McEwen, 2000). The substances released through the action of the nervous system bind to specific receptors on white blood cells and have a regulatory effect on their distribution and function (Segerstrom and Miller, 2004). More specifically, sympathetic fibers release substances that bind to receptors on lymphocytes, and “the hypothalamic-pituitary-adrenal axis, the sympathetic-adrenal-medullary axis, and the hypothalamic-pituitary-ovarian axis secrete the adrenal hormones, epinephrine, norepinephrine, and cortisol; the pituitary hormones prolactin and growth hormone; and the brain peptides melatonin, β-endorphin, and enkephalin” (Segerstrom and Miller, 2004, p. 604). Under acute stress, elevations of stress hormones (catecholamines and GCs) facilitate the movement of immune cells, lymphocytes, monocytes, and NK cells which are reduced in other tissues where other mediators of immune function activation become involved. For example, interferon gamma “is known to induce expression of antigen-presenting and cell-adhesion molecules on endothelia cells and macrophages and cell adhesion molecules on leukocytes” (McEwen, 2000, p. 175).

Stress also affects the immune system through behaviors, such as changes of sleep patterns, that could modify immune system processes (Segerstrom and Miller, 2004). Another association of the immune system with stress arises through the immunological activation of “sickness behavior” which refers to a constellation of behavioral changes that accompany infection that include a “reduction in activity, social interaction, and sexual activity, as well as increased responsiveness to pain, anorexia, and depressed mood” (Segerstrom and Miller, 2004, p. 604).

Support for the relationship of psychological stress and immune system response was provided through a meta-analysis of more than 300 empirical studies (Segerstrom and Miller, 2004). The findings across these studies indicated that acute stressors were associated with upregulation of natural immunity parameters and downregulation of specific immunity functions (Segerstrom and Miller, 2004). Acute stressors were associated with an increase in the number of NK cells, neutrophils, and large granular lymphocytes in peripheral blood, increased production of proinflammatory cytokines and cytokines, and decrease in colonal proliferation response (Segerstrom and Miller, 2004). Chronic stressors were associated with suppression of both cellular and humoral responses (Segerstrom and Miller, 2004). Furthermore, stress appraisal was found to be associated with a reduction in NK cell cytotoxicity (Segerstrom and Miller, 2004).

Chronic stress leading to sustained levels of stress hormones can also affect the immune system (Cruess et al., 2004). A proinflammatory cytokine, interleukin-6 (IL-6) is elevated under stress and stimulates SNS and HPA activation (Cruess et al., 2004). Furthermore, inflammation is critical in the development and progression of atherosclerosis which is associated with the rupture of plaque that can block blood flow (Cruess et al., 2004). Low-density lipoprotein cholesterol retained in the cell wall undergoes oxidative modification and the “resultant modified lipids can induce the expression of adhesion molecules and proinflammatory cytokines as mediators of inflammation in macrophages and vascular cell walls” (Cruess et al., 2004, p. 40). Psychological factors such as depression and stress have been associated with decrements in lymphocyte proliferative response and lower NK cell cytotoxicity (Cruess et al., 2004). Thus, alterations in neuroendocrine functioning affect the immune system and neurohormonal changes have been linked to a number of psychosocial factors including cognitive appraisals, coping responses, perceived loss of control, attributions of helplessness, and feelings of hopelessness, low self-efficacy, passive coping strategies, and lack of social support (Cruess et al., 2004).

Stress and Erythrocyte Adhesion

The vasoocclusive process in sickle cell disease is complex and increasing attention is focused on the role of the adhesion of sickle erythrocytes (SS RBCs) to endothelial cells (ECs). A direct relationship between the rating of vasoocclusive pain and biological markers of erythrocytes/EC adhesion has been reported (Dampier et al., 2004). In addition, there is evidence that the stress hormone epinephrine enhances adhesion of sickle erythrocytes (SS RBCs), but not normal RBCs, to ECs (Zennadi et al., 2004). Febrile episodes are frequently associated with vasoocclusive pain episodes in sickle cell anemia and are hypothesized to be viral in origin. Support was provided for the hypothesis that viruses, through double-stranded RNA, can induce sickle erythrocytes adherence to ECs through alpha4beta1-VCAM-1-mediated adhesion (Smolinski et al., 1995). A recent review summarizes the increasing knowledge about how membrane structures contribute to cell adhesion (Telen, 2005).

Stress and Neurocognitive Functioning

Chronic high levels of stress hormones and GCs contribute to impairment of cognitive function through effects on the hippocampus (McEwen, 2000). The hippocampus has two types of adrenal steroid receptors, type 1 (mineralocoiticoid), and type 2 (glucocorticoid), that mediate hormone effects on gene expression (McEwen, 2000). It is the combined action of circulating GCs and catecholamines interacting with local tissue mediators, such as cytokines, that affect the immune system and the excitatory amino acids, such as glutamate, and neurotransmitters, particularly serotonin, that affect the brain and cognitive functioning (McEwen, 2000). Brain atrophy has been shown to occur, particularly of the hippocampus, as a result of elevated GCs and severe stress and declines in hippocampally related cognitive functions such as episodic memory are correlated with increases in HPA activity (McEwen, 2000).

Adrenocortical stress responses to ordinary daily stress is sufficient to produce atrophy of hippocampal structures (McEwen, 2000). However, individual differences in stress responsiveness also play a role (McEwen, 2000). “Individuals with a more reactive stress hormone profile will expose themselves to more cortisol and experience more stress-related neural activity, than other people who can more easily habituate to psychosocial challenges” (McEwen, 2000, p. 183).

In assessing the impact of stress, it is useful to have multiple physiological measures within the same study. In a study of monozygotic and dizygotic female twin pairs, genetic and environmental effects on autonomic reactivity to a psychologically stressful situation was examined for both single physiological variables and functional combinations of seven of these variables (Lensvelt-Mulders and Hettema, 2001). The findings supported the hypothesis that autonomic response profiles would yield larger genetic effects than single autonomic measures and that the idiosyncratic relationship of a person and his/her environment is a heritable trait. Up to 80% of the variance in the functional profiles were accounted for by differences in individual genotypes. The authors comment, “there are at least two ways people physiologically respond to a situation: Directly, by making people more genetically liable to express a certain trait, and indirectly by influencing idiosyncratic interactions between a person and his environment” (Lensvelt-Mulders and Hettema, 2001, p. 38).

Stress and Cardiovascular and Renal Response

It has been hypothesized that exaggerated cardiovascular response to stress is a mechanism in the pathogenesis of essential hypertension and CHD. Snieder et al. (2002) have developed a biobehavioral model of stress-induced hypertension to explain how repeated exposure to stress, in combination with genetic susceptibility, could lead to the development of hypertension. This model is useful to consider, not only because of the cardiovascular problems in sickle cell disease, but because the biobehavioral model enables a systems perspective. The biobehavioral model focuses on the complex interrelationship of three underlying physiological systems that mediate the stress response of the heart, vasculature, and kidney: the SNS; the renin-angiotensin-aldosterone system (RAAS), and the endothelial system (ES). In support of this model, evidence is reviewed for a genetic influence on the two major intermediate phenotypes of the model: cardiovascular reactivity to psychological stress and the renal stress response in terms of stress-induced sodium retention. The data reviewed were from twin and family studies and a limited number of candidate gene association studies. The authors acknowledge that other biological systems, such as the HPA axis, parasympathetic autonomic reactivity, and serotonin functioning in the central nervous system may mediate the influence of stress on the development of the essential hypertension, and the importance of genetic variation of these systems has been demonstrated as well. The biobehavioral model of stress-induced essential hypertension proposes that in response to stress there is an increased central nervous system activity that in turn results in the release of catecholamines, norepinephine and epinephrine, which in turn increases heart rate. In addition, norepinephrine causes vaso-constriction and epinephrine causes vasoconstriction in some vessels and vasodilation in others (Snieder et al., 2002).

The ES influences the control of vascular smooth muscle function through the production of nitric oxide (NO), a vasodilator, and endothelin-1 (ET-1), a vasoconstrictor. SNS arousal potentiates the release of these vasoactive substances. Under stress there is evidence of increased release of ET-1 and decreased production of NO resulting in increased vasoconstictive tone (Snieder et al., 2002). The RAAS is activated by both the activity of the ES and SNS arousal. This results in further vasoconstriction and an increase in sodium retention enhances the vasoconstrictive effects of norepinephrine on peripheral vasculture (Snieder et al., 2002).

A complex interaction of these three systems contributes to increase total peripheral resistance in response to stress and repeated exposure leads to disregulation in appropriately activating, and/or turning off, cardiovascular function (Sneider et al., 2002). The responses to stress result in increases in cardiac and vascular wall tension and intravascular shear stress that leads to secondary renal damage and cardiovascular remodeling, including diminished endothelium-dependent arterial dilation to reactive hyperemia (Snieder et al., 2002). Another manifestation of vascular remodeling is increased arterial stiffness which in turn is associated with stroke, renal failure, and coronary artery disease and left ventricular hypertrophy, which is a strong predictor of cardiovascular morbidity and mortality (Snieder et al., 2002).

Snieder et al. (2002) also examined the evidence for the role of specific candidate genes on cardiovascular response to stress. Since the β2-adrenergic receptor mediates peripheral vasodilation, polymorphic variation in this gene may influence response to stress. Evidence has been provided for an association between Arg16Gly polymorphism β2-adrenergic receptor gene (ADRB2) and the Arg389Gly and Arg16Gly polymorphisms in the β1-adrenergic receptor gene (ADRB1) were associated with blood pressure at rest and reactivity to stress. The Gln27Glu polymorphism of the β2-adrenergic receptor gene also showed significantly higher levels of blood pressure at rest and stress but interestingly, no associations were found between these polymorphisms and cardiovascular reactivity for African Americans (Snieder et al., 2002). It should also be noted that an increase in cardiovascular response to stress has also been associated with a promoter polymorphism of the serotonin transporter gene (5HTTLPR) through higher levels of serotonin (Williams et al., 2001).

Snieder et al. (2002) suggested that future studies investigating genetic influences on cardiovascular and renal stress should employ measures of polymorphic variation in candidate genes that underlie the SNS, the ES, and the RAAS. They argue that rather than studying the effects of candidate genes in isolation that the biobehavioral model provides a framework for describing the interrelated physiological network underlying blood pressure regulation in response to stress. More specifically, Snieder et al. (2002) suggest the following candidate genes for the respective systems. SNS: “the α1- and α2-adrenergic receptor gene (ADRA1, ADRA2) and the β1- and β2-adrenergic receptor genes (ADRB1, ADRB2)”; RAAS: “the genes for angiotensin converting enzyme, and the angiotensin II type-1 receptor (AGTR1), aldosterone synthase (CYP11B2) and angiotensinogen”; ES: “the ET-1 gene (EDN1), the gene for ET-1 receptor A (EDNRA) and the genes for the three types of nitric oxide synthase (NOS1, NOS2, NOS3)” (Snieder et al., 2002, p. 87).