| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Anthropology (T.W.McD.), Northwestern University, Evanston, IL; Department of Anthropology (J.F.S., C.M.W.), Emory University, Atlanta, GA; Center for Developmental Epidemiology (A.A., E.J.C.), Duke University Medical Center, Durham, NC; Department of Social and Behavioral Sciences (M.B.), Arizona State University West, Phoenix, AZ; Department of Psychology (J.T.C.), University of Chicago, Chicago, IL; and Department of Molecular Virology, Immunology and Medical Genetics (R.G.), Ohio State University Medical Center, Columbus, OH.
Address reprint requests to: Thomas W. McDade, PhD, Northwestern University, Department of Anthropology, 1810 Hinman Ave, Evanston, IL 60208. Email: t-mcdade{at}nwu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: The EBV antibody method was evaluated through analysis of precision, reliability, stability, and comparisons with plasma and indirect immunofluorescence methods. The effects of life events on p18-VCA antibody level were considered in a subsample of 9, 11, and 13 year-old children participating in the Great Smoky Mountains Study in North Carolina. The subsample was stratified by age, sex, and degree of overall life strain.
RESULTS: Dried blood spots provided a convenient, sensitive, precise, and reliable method for measuring EBV p18-VCA antibody titer. Life events were positively associated with p18-VCA antibodies in girls but not in boys.
CONCLUSIONS: The validity of the blood spot EBV p18-VCA antibody assay, as well as the ease of sample collection, storage, and transportation, may provide an opportunity for psychoneuroimmunology to explore a wider range of stress models in larger, community-based studies.
Key Words: psychoneuroimmunology • cell-mediated immunity • stress • methods • lifeevents • sex difference
Abbreviations: EBV = Epstein-Barr virus; VCA = viral capsid antigen; IgG = immunoglobulin G; IF = indirect immunofluorescence; ELISA = enzyme-linked immunosorbent assay; CV = coefficientof variation; GSMS = Great Smoky Mountains Study; CAPA =Child and Adolescent Psychiatric Assessment.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
In this article, we introduce and validate a minimally invasive, indirect method for measuring one aspect of cell-mediated immune function: antibodies against EBV in whole blood spot samples. Previous work has shown that EBV antibody level is a valid marker of cell-mediated immunity that responds to a range of psychosocial stressors (1–6). While these studies have used serum or plasma from venipuncture, we have validated an EBV-VCA antibody assay for whole blood samples from finger pricks. The method minimizes the inconvenience and burden imposed on participants by blood sampling and eases the requirements for sample collection, storage, and transportation before analysis. Such features allow for the collection of samples away from the research laboratory or clinic and provide access to individuals for whom venipuncture may be problematic, such as children or the elderly. In this article, we first present this method in detail, then demonstrate its utility for psychoneuroimmunology with pilot data from a large, community-based study of traumatic life events and immune function in children and adolescents from North Carolina.
|
STUDY 1: VALIDATION OF AN INDIRECT METHOD |
|---|
|
TOPABSTRACTINTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Methods: EBV Antibody Assay
Assay Principle.
The method used here is an ELISA for measuring p18-VCA antibodies in whole blood spots. Most studies have used IF to measure EBV antibody levels, in which substrates expressing EBV antigens are incubated with serial dilutions of subject sera. Using IF, EBV antibody titers are determined by the final dilution at which IF-positive cells are no longer detectable. ELISA provides another measure of EBV antibody titer, correlates well with IF (see below), and has been shown to be a more sensitive, objective, and less labor-intensive method (6, 7). Although previous work has demonstrated that dried blood spots are a convenient and accurate medium for measuring specific antibodies (8, 9) as well as hormones such as cortisol, testosterone, estradiol, and gonadotropins (10, 11), this study is the first to use a blood spot method to measure EBV antibodies.
The method is an adaptation of a commercially available kit for measuring p18-VCA IgG antibodies in serum (No. 7590, DiaSorin Corporation, Stillwater, MN). The assay measures antibodies against one VCA protein, the p18 polypeptide, which is a marker protein containing immunodominant epitopes of the viral capsid antigen complex. Antigen-antibody complexes form between p18-VCA IgG antibodies present in the sample and synthetic p18 peptide bound to the surface of microtiter wells. Horseradish peroxidase–labeled antihuman IgG in the presence of a chromogen substrate reacts with the antigen-antibody complex resulting in color development. The concentration of p18-VCA IgG antibody is directly related to the absorbance of the solution measured at 450 nm.
Standards and Controls.
The manufacturer provides four levels of p18-VCA IgG standard in stabilized human sera. Controls consist of three levels of manufacturer-supplied control sera and four levels of blood spot controls made from whole blood pools spotted onto filter paper and stored at -23°C. Standard concentrations are provided by the manufacturer and are used to convert absorbance values to standard ELISA scores. Blood spot control ranges are derived from values obtained over multiple assays.
Elution of Whole Blood Spots.
The day before an assay is to be performed, blood spot samples and controls are removed from the freezer, and a small hole punch (available from local office supply stores) is used to punch out a 2.5-mm diameter disk of whole blood. The disk is transferred with tweezers to a labeled 12 x 75 glass tube, and 250 µL of diluent buffer (supplied with kit) is added. The sample is incubated overnight at room temperature. The following day, the assay is performed as per the kit instructions, with 100 µL of the blood spot eluate pipetted directly into microtiter wells in duplicate, along with 100 µL of each standard and diluted control sera.
Assay Validation
Plasma/Blood Spot Correlation.
Complex antigen-antibody binding dynamics preclude direct evaluation of linearity and accuracy by manipulating EBV antibody concentration. Therefore, blood spot antibody methods rely on comparisons with plasma values for validation (12, 13). Matched plasma and whole blood samples were obtained from 40 adult subjects. From each whole blood sample (EDTA-anticoagulated), a portion of whole blood was spotted onto filter paper, and another portion was centrifuged and the plasma withdrawn. Plasma p18-VCA antibody values were obtained following kit instructions and compared with matched blood spot values. Correlation between blood spot and plasma values is linear and high (Pearson correlation R = 0.97) (Figure 1).
|
According to kit specifications, a serum value 
20 ELISA units indicates the presence of p18-VCA antibodies. A serum value <20 ELISA units indicates nondetectable levels of p18-VCA antibodies, and these individuals are thus presumed to be seronegative for EBV. This cutoff was established by the kit manufacturer through analysis of samples from individuals before and after becoming ill with infectious mononucleosis. To establish a seropositivity cutoff value for the blood spot method, 19 individuals with low titers of EBV antibodies were identified, and matched plasma and whole blood spot samples were run together in one assay. According to this analysis, the plasma cutoff of 20 corresponds to a blood spot cutoff of 18. If the plasma cutoff of 20 is considered to be the standard, then the blood spot cutoff of 18 resulted in one false-negative and one false-positive. Increasing the blood spot cutoff to 20 resulted in two false-negatives and zero false-positives, thereby increasing specificity at a cost to sensitivity. Further increases in the blood spot cutoff resulted in more false-negatives. Although the number of individuals in this analysis is small, it provides justification for a blood spot seropositivity cutoff of 18 or 20. We have elected to use the more conservative cutoff of 20 because this corresponds to the kit specifications and limits the number of false-positives. As a consequence of this cutoff, a small number of false-negative individuals with low levels of EBV p18-VCA antibodies may be eliminated from analysis.
Comparison With IF.
Indirect immunofluorescence is the current gold standard for assessing EBV antibody titers and has been used in numerous studies linking psychosocial stress to immune function (1, 4, 15). To compare the blood spot ELISA EBV-VCA method with plasma IF, 111 undergraduate students (mean age, 19.2 years; range, 18–24) were recruited at Ohio State University. Blood samples were collected with a 20-gauge indwelling angio-catheter, which was inserted into an antecubital vein on each subject’s dominant arm. At least 30 minutes after insertion, a 20-mL blood sample was drawn into a syringe, and blood spots were made for each participant by placing five drops of whole blood onto filter paper. Blood spots were allowed to dry at room temperature overnight and were then frozen until delivery to the Laboratory for Comparative Human Biology at Emory University. Blood spot samples were stored at -23°C until ELISA analysis. The remaining blood sample was injected into heparinized collection tubes and centrifuged, and the plasma was stored at -70°C at Ohio State until IF analysis.
The indirect IF test was used to measure EBV-VCA IgG antibody titers in plasma as previously described (16). Briefly, acetone-fixed HR-1 cells (a cell line latently infected with EBV) prepared on glass coverslips were adsorbed with two-fold dilutions of plasma in phosphate-buffered saline. Approximately 5% of these cells were early antigen/VCA positive by IF. The highest dilution of plasma that resulted in detection of at least 1% IF positive cells was considered the end point. All slides were blind coded and read by examining 300 to 400 cells.
Before statistical analysis, ELISA and IF results were log-transformed to normalize the distributions. There was a significant positive relationship between the two measures (Pearson R = 0.77; p < .01) (Figure 2). The strength of this relationship is consistent with previous work comparing IF and ELISA methods for serum EBV antibody titers (7). A perfect correlation between IF and ELISA was not expected because our blood spot ELISA method only detects antibodies against p-18, the immunodominant VCA epitope. In contrast, the less-specific IF method detects antibodies to multiple EBV proteins. Individual differences in the heterogeneity of antibody responses to VCA can thus result in different EBV antibody levels across these methods.
|
|
Cross-Reactivity.
This method is reported by the kit manufacturer not to be cross-reactive with Herpes Simplex virus I and II, Cytomegalovirus, Varicella Zoster virus, Toxoplasma, and Rheumatoid factor.
Stability.
The stability of p18-VCA antibodies in blood spot samples was determined over an 8-week period in which blood spot cards made from a single venipuncture were exposed to one of three temperature conditions (4°C, room temperature, and 37°C) for varying lengths of time up to 8 weeks. A sample at the high end of the assay range was used to maximize the likelihood of detecting any deterioration. In addition, the potential impact of humidity was considered by sealing one set of cards in plastic bags with approximately 30 g of desiccant (CaSO4). The other set of cards was not exposed to desiccant. Samples were considered to be stable as long as values remained within a 10% CV range of the initial values. Results for the samples with desiccant are presented in Figure 3, although no difference in the pattern of deterioration was found with or without desiccant. Samples remained stable for at least 8 weeks at 4°C and room temperature, but began to deteriorate after 1 week at 37°C.
|
Overall, these results indicate a high degree of EBV antibody stability in whole blood spots. Although prompt freezer storage is always desirable, samples can be collected in the field and stored at room temperature for up to 2 months. In settings where temperatures are likely to exceed 22°C to 23°C, attempts should be made to protect samples from excessive heat.
Freeze/Thaw.
The stability of p18-VCA antibodies to repeated cycles of freezing and thawing was evaluated for four blood spot controls that represented the entire assay range. Blood spot cards were removed from the freezer each afternoon, left at room temperature for 2 to 3 hours, and returned to the freezer. This protocol was repeated for six consecutive days. Samples were considered stable if they remained within a 10% CV range of the sample that remained frozen 24 hours after preparation. No sample deterioration was detected for up to six freeze/thaw cycles.
|
STUDY 2: STRESS AND EBV ANTIBODIES IN NORTH CAROLINA CHILDREN AND ADOLESCENTS |
|---|
|
TOPABSTRACTINTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Methods
Participants.
This study focused on a subsample of 9-, 11-, and 13-year-old children participating in the ongoing Great Smoky Mountains Study of adolescent psychopathology and service use. This large community-based study relates symptoms of child psychopathology to comprehensive measures of life strain, socioeconomic status, family environment, and morphological and endocrine measures of pubertal status.
Methodological details for the GSMS are presented elsewhere (24). Briefly, a multistep sampling procedure was used to select among the 12,000 9-, 11-, and 13-year-old children listed in the public school database of the southern Appalachian mountain region of North Carolina. Within each age category, children were selected on an equiprobability basis to generate a total screening sample of 4500 children. A screening questionnaire was administered to parents of 3896 children to identify children with a high probability of psychiatric symptoms. All children with scores above a predefined cutoff point were recruited for the main study. A 1 in 10 random sample of the remainder was also recruited. Of the 1346 children recruited for the study, 1071 (79.6%) were enrolled. Children and their parents were interviewed as close as possible to the birthday on which children became 9, 11, or 13 years of age.
In this population, family dysfunction, negative life events, and poverty are unsurprisingly and strongly associated with psychiatric symptoms (24). Previous findings have reported the development of a significant sex difference in the relationship between stressful life circumstances and symptoms of depression. For girls, advancement through puberty is associated with a strengthening of the relationship between life strains and depression, whereas puberty acts as a protective factor for boys, actually eliminating the correlation between hardship and depression (25).
Accordingly, in this study we sought to evaluate the immunological consequences of stressful life circumstances and hypothesized that a similar pattern of sex difference would emerge in puberty. Therefore, we selected a subsample of children stratified by sex (male/female), age (9, 11, and 13 years), and degree of overall life strain (low/high), yielding 12 groups. Life strain scores were constructed from the following: welfare dependency, parental unemployment, low parental education, family size (four or more children), the presence of domestic violence, parental drug use, parental arrest, maternal depression, and diagnosed psychopathology among family members. Scores ranged from 0 to 32 in the main sample, from which two subsets of participants were drawn: individuals with scores <10 comprised the low-strain group, whereas those with scores >14 comprised the high-strain group. An attempt was made to assign 30 individuals to each of the 12 groups, but this was not possible, largely due to higher rates of refusal to give blood among girls. No group had fewer than 16 individuals, and the final sample size for this pilot study was 256, representing 23.9% of the overall GSMS sample.
Measures of Psychosocial Stressors.
The CAPA protocol was used in parent and child interviews to gather information on psychopathology and life circumstances and experiences (26). This structured protocol elicits information regarding symptoms for a wide range of diagnoses, functional impairment, and the quality of family relationships.
In addition to overall life strain, we considered the number of traumatic life events, defined as death of a close family member and/or friend, sexual and/or physical abuse, and exposure to other traumatic experiences (eg, extreme violence or accident) endured over a child’s lifetime. Children had experienced on average 1.0 traumatic life events (SD = 1.2), with a range of 0 to 6. Children in the high-strain group experienced significantly more life events than children in the low-strain group (1.42 vs. 0.65, p < .001). Traumatic life events were not included in our summary measure of life strain.
Sample Collection.
Blood spot samples were procured as described above. At least two drops of capillary whole blood were collected on filter paper, dried overnight at room temperature, and placed in a refrigerator for no more than 3 weeks before shipment to the laboratory for storage at -23°C until analysis. The ELISA method described above was used to determine the EBV p18-VCA antibody level for each individual.
Statistical Analysis.
Statistical analyses were performed for all EBV-positive individuals using Statistical Analysis Software (Version 6.12, SAS Institute Inc., Cary, NC). Before analyses, EBV antibody scores were logarithmically transformed to normalize the distribution. We set alpha = 0.05 as our criterion for statistical significance.
Results
Of the initial 256 study participants sampled, 80.1% were seropositive for EBV according to ELISA detection of p-18 VCA antibodies, yielding a final sample of 205 for our analyses. Table 2 presents the untransformed mean EBV antibody levels according to our sampling strategy and the distribution of participants across the sampled groups. Boys were overrepresented in our sample because girls were more likely to refuse to give blood, but boys were no more likely than girls to be seropositive (
2 = 0.22, NS).
|
Although neither age nor degree of life strain were significantly related to p18-VCA antibodies, we included these variables in subsequent regression analyses because they comprised our sampling criteria. The addition of traumatic life events resulted in a significant model for girls, with traumatic life events positively related to p18-VCA antibody titer, whereas the sampling variables, including life strain, were nonsignificant (Table 3). The partial correlation coefficient between number of life events and p18-VCA antibody titer—indicating the strength of the relationship between these two variables, controlling for the effects of age and life strain—was R = 0.31 in girls (p < .01). In contrast, life events were not significantly related to p18-VCA for boys (partial R = 0.007, NS).
|
1 event, 117.4 ± 16.4 ELISA units; no events, 64.6 ± 9.8 ELISA units). In contrast, there was no difference in antibody titers for girls under conditions of low life strain, regardless of the number of life events (1 event, 68.2 ± 14.5 ELISA units; no events, 70.5 ± 6.8 ELISA units).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Studies of stress-induced immunosuppression have reliably demonstrated the utility of the EBV model (28). Academic exam stress in medical students has been linked to increases in EBV antibody titers and concurrent reductions in EBV-specific memory T-cell proliferation and cytotoxic T-cell killing of infected cells (3, 16). Similar results are reported for the chronic stress associated with caring for a family member with Alzheimer’s disease (29) or involvement in a poor quality marriage (30, 31). Additionally, loneliness, defensiveness, and anxiety have all been positively associated with EBV antibodies (1, 32). Recently, EBV reactivation has also been associated with the anticipation of space flight by astronauts (33). Conversely, stress management interventions and disclosure of previously repressed trauma are associated with decreases in EBV antibodies (6, 34). These studies validate the EBV model as an indirect measure of stress-induced cell-mediated immune suppression such that higher stress burdens are reflected in higher EBV antibody titers.
Assay validation studies reported here uniformly demonstrate that our blood spot method of EBV antibody determination represents a sensitive, reliable, and stable measure. Results from a pilot study of stress and EBV VCA antibody level in children and adolescents in North Carolina furthermore demonstrate the field utility of our blood spot method. The ease of sample collection, storage, and transportation facilitates the acquisition of samples outside of the confines of the clinic or laboratory.
In addition to minimizing participant burden and maximizing researcher convenience, this method may provide an opportunity for psychoneuroimmunology to explore a wider range of stress models in larger community-based samples. Such an epidemiological approach allows one to recruit larger, more representative samples that can overcome some of the limitations of clinic-based studies. The minimal pain and inconvenience imposed by finger pricks also facilitates more frequent blood sampling and even enables participants to collect samples from themselves at home (35). Field studies of stress and immune function in diverse ecological contexts are also feasible (36). Blood spot methods for gonadal and adrenal steroid hormones have been used successfully in cross-cultural studies of reproductive ecology around the world, including Papua New Guinea, East Africa, Nepal, and North Carolina (11).
Although the results of this study show that the use of a commercially available ELISA kit using one EBV polypeptide can be useful, it is also important to note that measuring differences in antibodies to only one viral peptide may provide aberrant results. If the peptide being measured is not synthesized in vivo (as a result of a change in the steady state expression of the latent EBV genome) then one would not see a change in antibody titers to that protein (2).
Analysis of a subsample of participants in an ongoing study of puberty and psychopathology reveals that, among the sample of seropositive girls, traumatic life events are significantly related to increased levels of EBV p18-VCA antibodies, suggesting lower levels of cell-mediated immune function. Girls in situations of high life strain stand out as the group that manifests immunosuppression associated with life events.
The significant effect of traumatic life events is consistent with previous findings linking life events to increased EBV antibody titers in adults (28, 37). In children, adverse life events have previously been associated with reduced NK cell activity (19), whereas parental separation or divorce has been correlated with reduced bacterial killing (21). The association between life events and immune function under conditions of high life strain may be analogous to recent work reporting that background life stress increases one’s sensitivity to the immunomodulating effects of acute stressors (38, 39). Alternatively, life strain and traumatic life events in this study may represent two interdependent measures of stressful life circumstances, with additive effects on p18-VCA antibody level.
The sex-specific pattern of results is consistent with previous work in this population that demonstrated the development of a sex-differentiated relationship between life stress and symptoms of depression (25). Few psychoneuroimmunology studies report sex differences in stress-induced immunosuppression, although marital discord and negative interactions have recently been shown to have greater physiological impact on women than men (5, 40). Our findings may indicate a similar sex difference in sensitivity to social stress in late childhood and adolescence.
To date, few studies have considered the relationships between stress and immune function in children and adolescents, and fewer still have addressed these issues in large, community-based samples. We suggest that attention to gender, developmental status, and social context may enrich future psychoneuroimmunological studies, whereas blood spot methods can facilitate the collection of samples in future epidemiological studies that complement clinical and experimental designs.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication February 22, 1999.
Revision received November 17, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSTUDY 1: VALIDATION OF...STUDY 2: STRESS AND...DISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  W. E. Funk, S. Waidyanatha, S. H. Chaing, and S. M. Rappaport Hemoglobin Adducts of Benzene Oxide in Neonatal and Adult Dried Blood Spots Cancer Epidemiol. Biomarkers Prev., August 1, 2008; 17(8): 1896 - 1901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Fachiroh, P. R. Prasetyanti, D. K. Paramita, A. T. Prasetyawati, D. W. Anggrahini, S. M. Haryana, and J. M. Middeldorp Dried-Blood Sampling for Epstein-Barr Virus Immunoglobulin G (IgG) and IgA Serology in Nasopharyngeal Carcinoma Screening J. Clin. Microbiol., April 1, 2008; 46(4): 1374 - 1380. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. B. Dowd, M. N. Haan, L. Blythe, K. Moore, and A. E. Aiello Socioeconomic Gradients in Immune Response to Latent Infection Am. J. Epidemiol., January 1, 2008; 167(1): 112 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Opstelten, A. M. van Loon, M. Schuller, A. J. M. van Wijck, G. A. van Essen, K. G. M. Moons, and T. J. M. Verheij Clinical Diagnosis of Herpes Zoster in Family Practice Ann. Fam. Med, July 1, 2007; 5(4): 305 - 309. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. W. McDade, L. C. Hawkley, and J. T. Cacioppo Psychosocial and Behavioral Predictors of Inflammation in Middle-Aged and Older Adults: The Chicago Health, Aging, and Social Relations Study Psychosom Med, May 1, 2006; 68(3): 376 - 381. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. W. McDade, J. Burhop, and J. Dohnal High-Sensitivity Enzyme Immunoassay for C-Reactive Protein in Dried Blood Spots Clin. Chem., March 1, 2004; 50(3): 652 - 654. [Full Text] [PDF]
|
|
|
|
 |
|
 T. W. McDade and B. Shell-Duncan Whole Blood Collected on Filter Paper Provides a Minimally Invasive Method for Assessing Human Transferrin Receptor Level J. Nutr., December 1, 2002; 132(12): 3760 - 3763. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Kiecolt-Glaser, L. McGuire, T. F. Robles, and R. Glaser Psychoneuroimmunology and Psychosomatic Medicine: Back to the Future Psychosom Med, January 1, 2002; 64(1): 15 - 28. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
