| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLE |
From the Departments of Medical Psychology and Psychotherapy (C.S., A.L., G.R., E.C., G.S.) and Internal Medicine (P.K.), University Hospital Innsbruck, and Institute of Medical Chemistry and Biochemistry (D.F.), University Innsbruck, Innsbruck, Austria.
Address reprint requests to: Christian Schubert, MD, Department of Medical Psychology and Psychotherapy, University Hospital Innsbruck, Sonnenburgstrasse 9, A-6020 Innsbruck, Austria. Email: Christian.Schubert{at}uibk.ac.at
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: A 40-year-old woman with SLE (last flare-up September 1995) was interviewed initially to determine major life events and difficulties (using the Life Events and Difficulties Schedule) in the previous 2 years. She was then observed for 63 days. Urine neopterin, an immunological parameter demonstrated to parallel disease activity in SLE patients, was measured in daily overnight urine. Daily incidents were identified weekly by the Incidents and Hassles Inventory and independently rated. Intervening factors, including infections, medication, and lifestyle, were controlled.
RESULTS: Retrospectively, data obtained from the Life Events and Difficulties Schedule indicated that major life events and difficulties had preceded the patient’s last flare-up in 1995. Although there were no clinical signs of SLE during this prospective study of 63 days, cross-correlational analyses revealed that "moderately" stressful incidents associated with higher levels of emotional irritation (lag 0: +0.271, p < .05) predicted an increase in urine neopterin the following day (lag 1: +0.441, p < .05). Moreover, a 7-day cyclicity in neopterin levels that corresponded to the weekly examinations and interviews was found.
CONCLUSIONS: This study showed a causal relationship between psychosocial stressors and urine neopterin concentrations that may be related to SLE disease activity. Furthermore, the workability of an integrative approach using single-case design and time-series analysis in psychoneuroimmunology was demonstrated for the first time.
Key Words: psychoneuroimmunology • stress • single-case design • time-series analysis • systemic lupus erythematosus • neopterin
Abbreviations: EWL = Eigenschaftswörterliste ("list ofadjectives"); HAWIE-R = Hamburg-Wechsler-Intelligenz-Test fürErwachsere; SLAM = Systemic Lupus Activity Measure; SLE =systemic lupus erythematosus; ARIMA = auto regressive integratedmoving average.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
According to new concepts in stress research, differences among individuals no longer need be considered a nuisance and need not be obscured by averaging (7, 8). New stress concepts are based on the reconceptualization of health and disease that has occurred in the last decade (9, 10). According to this view, stress-mediated disease results from a perturbation of dynamic functions that is represented by changes in form, frequency, and amplitude of any variable that behaves rhythmically (7). These dynamic processes are unique. A variable’s dynamics consist of characteristic longitudinal patterns of interdependent time points (11). This same uniqueness holds true for the complex interaction of many variables over time. When we average individual dynamics or interactions in groups, these unique structures are extinguished or denaturalized. Therefore, to analyze the dynamic interdependencies between psychosocial, emotional, and physiological factors, it is necessary to consider individuals on a single basis.
This new perspective in stress research is in line with the present investigation, the aim of which was to gather evidence, using a single-case design, about the temporal relationship between daily psychosocial stress and immunological functioning in SLE (12). To investigate the dynamic interactions among psychosocial, emotional, and physiological variables in a patient with SLE, time-series analysis as proposed by Box and Jenkins (11) was applied. This method requires at least 50 to 100 equidistant (in this study, daily) measurements for statistical description of the internal structure of a variable’s series. To guarantee the validity of the data, it was of utmost importance to interfere as little as possible in the patient’s normal routine. Daily overnight urine samples were thus collected by the patient herself, and urine neopterin concentrations were then measured as the indicator of immune system activity. Large amounts of neopterin are released by macrophages during T-cell-dependent activation (13) and show a circadian rhythm peaking in the early morning (14). In several studies comparing nonspecific as well as specific SLE parameters, SLE activity was found to be paralleled by fluctuating neopterin concentrations in serum and urine. Similarly, neopterin concentrations have been identified as one of the best single indicators for the serial determination of SLE activity (15–18). Daily psychosocial incidents and hassles were identified in this study in weekly semistructured interviews, allowing a wide variety of patient responses (19). Furthermore, various lifestyle variables that can intervene with cross-correlations between stressors and immune functions were identified (20).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The patient was then observed daily for 63 days, from December 13, 1996, to February 13, 1997. Late each afternoon, the patient answered questions about her emotional state, daily lifestyle, SLE, and unrelated symptoms. Beginning at 8 PM each evening, the patient collected her urine for the next 12 hours and froze it at a temperature of -20°C. These urine samples were brought to the laboratory weekly in a cold box and kept frozen at -70°C until further analysis. Each week, the patient was examined by a senior internist (P.K.) to control for any clinical symptoms of disease activity and was interviewed by a trained psychotherapist (A.L.) to identify the past week’s stressful incidents and hassles. At the end of the study, these incidents were assessed by independent raters. The interviewer as well as the independent raters were blind to the physiological data.
Because this study was designed to gather evidence about biopsychosocial interactions over time, the following inclusion criteria were necessary: disease remission, no use of steroidal antiphlogistics or immunosuppressants, no psychiatric disease (based on the third revised edition of the Diagnostic and Statistical Manual of Mental Disorders) or attention and memory deficits (HAWIE-R), and no renal dysfunction.
The patient gave informed consent to participate, and the protocol was approved by the hospital’s internal review board.
Patient
The patient is a 40-year-old white woman who was diagnosed with chronic discoid lupus erythematosus in 1979. In 1991, she developed erythema, photosensitivity, and arthralgia (small joints) and exhibited decreased complement C3. According to American Rheumatism Association criteria, she was thus diagnosed with SLE. In addition, the patient had proteinuria (102 mg/dl). A kidney biopsy for histological examination was refused by the patient. Tests for antinuclear anti–double-stranded DNA antibodies were negative. The patient was then treated with steroid bolus therapy and a daily maintenance dose of 10 mg of methylprednisolone for 1 year. After this treatment, neither proteinuria nor pathological urine sediment could be detected. The patient’s last flare-up occurred in September 1995.
The patient is married and has a 21-year-old son who left home in November 1995. She is a hair dresser but has been on disability pension since December 1995 because of her illness. She is a moderately heavy smoker (30 cigarettes per day). In May 1993, the patient had a hysterectomy.
MEDICAL EXAMINATION
Systemic Lupus Activity Measure.
The SLAM (21) is used to assess SLE activity by measuring 24 clinical manifestations and 8 laboratory parameters. Immune function parameters are not included. The weekly medical examination using the SLAM required approximately 30 minutes.
PSYCHOLOGICAL EXAMINATION
Assessment of life events, chronic difficulties, and incidents.
According to Brown and Harris (22), incidents occur at a discrete point in time and introduce changes that require an adaptive response. Events are incidents that introduce changes above a certain agreed level of seriousness. Severe events introduce what most people in equivalent biographical circumstances would see as even more serious and upsetting changes. Difficulties are ongoing problems that last at least 4 weeks and may or may not be made up of events and/or incidents. The following two interviews were used to assess these variables.
Life Event and Difficulties Schedule.
Life events and chronic difficulties were assessed for the period 2 years preceding the interview. This was necessary 1) to investigate retrospectively whether highly stressful life events and chronic difficulties might be capable of provoking severe disease exacerbations and 2) to have more information about patient’s psychosocial background when rating the daily incidents. Contextual threat rating of major life events (1 = marked, 2 = moderate, 3 = some, 4 = little/no) and chronic difficulties (1 = high marked, 2 = low marked, 3 = high moderate, 4 = low moderate, 5 = mild, 6 = very mild, 7 = no longer a difficulty) were scored by an independent panel of blind raters as previously described (22).
Incidents and Hassles Inventory.
In the semistructured Incidents and Hassles Inventory (G.W. Brown and T.O. Harris, 1996, unpublished), comprising 39 items, the proband is asked about incidents and hassles that occurred during the previous week. Additionally, at the end of the interview, daily notes made by the proband are included. The duration of incidents and the emotional response to them is documented. Afterward, an independent panel of blind raters assesses each incident according to its severity on a three-point scale (1 = marked, 2 = moderate, 3 = some). The length of the weekly interview was approximately 60 minutes.
3-skalen Version der Eigenschafswörterliste.
The 3-scale-EWL ("list of adjectives"; Ref. 23) is a paper-and-pencil test used to measure the proband’s emotional state (mood, mental energy levels, irritation) using 28 adjectives. Emotional state is assessed using a four-point system. Use of the 3-scale-EWL in longitudinal designs is recommended. The proband needs about 5 minutes to complete the test.
Assessment of daily lifestyle factors and subjective estimation of SLE activity.
The patient records daily cigarette use, alcohol and coffee consumption, drug use, and sleep. Using visual analog scales, the proband rates the degree of physical activity, joint pain, weakness, fatigue, and overall disease activity. Moreover, the proband indicates body temperature and whether any symptoms unrelated to SLE are present. The proband needs about 5 minutes for these questions.
Biochemical Analysis
Daily urine neopterin levels were determined by high-performance liquid chromatography (model LC 550, Varian Associates, Palo Alto, CA) as previously described (24). To avoid interassay variability, all 63 urine aliquots were measured in one single run within 3 months after all urine samples had been collected and stored at -70°C. For each of three independent determinations, a new aliquot was used. Urine neopterin levels are expressed as the micromolar concentration of neopterin per molar concentration of creatinine to compensate for variations in urine density (15, 16, 24).
Statistical Analysis
Statistical analyses were conducted using SPSS-Trends (25). In this study, the time-series were cross-correlated both at zero lag (ie, concurrent correlation) and at higher lags (until ±7) to determine whether one variable significantly preceded and hence predicted the other during the following days. Statistically significant cross-correlations reached the p < .05 criterion.
Time-series values are generated successively, and each value tends to (auto)correlate with the preceding value. Therefore, according to Box and Jenkins (11), each time-series is a function of two main factors, internal serial dependencies, such as autoregression and moving average components, and the independent, standard distributed residuals. Because of serial dependencies, cross-correlational analysis between two time-series may lead to false-positive or false-negative correlations. Therefore, adjusted cross-correlational analysis using only time-series residuals was applied in this study (26). Residuals that are free of serial dependencies were generated by ARIMA modeling. The residuals’ serial independency was controlled for using the Durbin-Watson statistical method, ranging from 0 to 4. A value near 2 indicates nonautocorrelation (27). Because modeling can overadjust and thus cancel out true covariance, unadjusted cross-correlations with premodeled series were also calculated for comparison.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
|
During the 63 days, 59 incidents were identified using recorded interviews. Of these, 11 (19%) were independently rated as having "some" severity, and 8 (14%) were rated as having "moderate" severity. None were rated as having "marked" severity. This study proceeded on the basis of a threshold model, assuming that only incidents of a certain severity are able to interfere with the course of urine neopterin. Therefore, only the eight incidents rated as moderate are listed chronologically in Table 1. Additionally, short descriptions of all moderate incidents are given. Days featuring a moderate incident were coded as "1," and the rest were coded as "0." The resulting time-series of moderate incidents is described by a (1,0,0) ARIMA model.
|
|
Cross-correlational analyses including all variables measured in this study revealed no other significant one-lagged correlation with urine neopterin. Alcohol had a suppressive effect on urine neopterin (lag 0: -0.407(1.17)/-0.269, p < .05).
As to the clinical features during the study period, weekly examinations revealed no evidence of increased SLE activity according to SLAM criteria. From days 28 to 36, the patient had an elevated temperature (37.2–37.7°) without evidence of infection. During the last 11 days (days 53–63), the patient reported having a cold accompanied by an elevated temperature. This was confirmed by the weekly medical examinations (data not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The current investigation also provided evidence about the specificity of stressors. Four of the eight moderate incidents, start of the study, argument with sister, son’s departure, and filthy student’s club were followed by substantial urine neopterin increases 1 day later. In contrast, forgotten house key, argument with husband and son, trouble with computer, and skidding car (Table 1) were not associated with neopterin increases. We believe that this discrepancy was due to the fact that those incidents that triggered urine neopterin increases were characterized by emotional irritation and high interpersonal stress, whereas the others were characterized by mental activity and lower interpersonal stress. To illustrate the varying meaning of interpersonal stress to the patient, we can look more closely at 1) argument with sister and 2) argument with husband and son, both of which deal with conflicts between the patient and persons close to her. The argument with her sister lasted considerably longer: In the interviews, the patient said that the argument with her husband and son had ceased to be a problem for her as early as half an hour after its occurrence. The argument with her sister, however, was not settled until the next day and remained on the patient’s mind for several days. In regard to content, the argument with her sister was clearly more significant: Whereas the argument with her husband and son was an everyday argument in an otherwise good-functioning marriage and loving relationship with her son, the argument with her sister was an argument with biographically significant content within the framework of the patient’s tense relationship with her sister, a problem that had been an issue for years.
When looking at the 7-day cyclicity of urine neopterin in the first 4 weeks of the study, one might conclude that our patient also experienced interpersonal stress due to the weekly interviews and examinations each Friday (Figure 2). The increase in neopterin levels was likely due to anticipation of these visits; the decrease, in turn, probably resulted from the relief that followed such visits. This phenomenon became weaker from week to week. This was confirmed by the patient’s own statements in the interviews. In this patient, interpersonally meaningful stressors have also been associated with both the onset of chronic discoid lupus erythematosus, which occurred while she was on her honeymoon, 8 months after her father’s death, and her last SLE flare-up in September 1995 (Figure 1). This, again, is in agreement with previous results demonstrating retrospectively that particularly significant crises in interpersonal relations (eg, death, divorce, feared loss of loved one) precede the onset of SLE or SLE exacerbations (3–5).
The results of the current investigation contrast with the findings of Wekking et al. (6), who concluded that no clear relationship exists between laboratory data and subjective ratings of physical and psychosocial status in SLE patients. Unfortunately, the authors investigated their topic’s dynamic aspects by static means, making their conclusions questionable: They determined both daily stress parameters and laboratory data at large 6-week intervals, averaged the longitudinal data of 21 probands, and neglected the data’s serial dependency. Although the approach of the current investigation may be more valid, it can certainly be refined in future studies. As a posteriori analysis of Figure 3 revealed, the interview process, the identification of incidents, and severity ratings should be improved to achieve a higher differentiation of psychosocial stressors.
As to the clinical relevance for the patient of this study, our results emphasize the importance of recognizing and controlling for psychosocial stressors with a special focus on stress in interpersonal relationships. For example, if our patient should experience a highly stressful major life event, closer monitoring of the disease course (eg, weekly urine neopterin determinations) would be recommended so that pharmaceutic as well as psychotherapeutic intervention could be made as soon as possible. This might prevent severe SLE exacerbations and therefore significantly decrease the risk of serious complications.
Nevertheless, particularly in light of the heterogeneity of SLE, a generalization of the results and conclusions of this study can only occur after numerous other single cases have been investigated. We may then be able to draw conclusions about possible systematic patterns, which may then form the basis for cohort studies. In this sense, the single-case design proposed in this study demonstrated the feasibility of collecting data on life elements, which, until now, have been neglected in biomedical research. It seems possible to gather data about the dynamics and interaction of biological as well as psychosocial variables. This approach is in line with new concepts in stress research (7, 8). Furthermore, the unique design of this study, based on consecutive weekly interviews while preserving the patient’s normal routine, allowed closer access to the meaning of stress under natural conditions on both the biochemical and psychosocial levels. This type of integrative single-case research is therefore also in accordance with George Engel’s biopsychosocial model (29, 30).
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
We thank the patient of this study for her participation. We also thank Gerhard Lücke for his valuable help. Finally, we thank Tirril Harris and George Brown for both their training in conducting and rating the Life Event and Difficulty Schedule and the Incidents and Hassles Inventory and their contributions to this article.
Received for publication July 21, 1998.
Revision received May 13, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C T Chou and C M Hwang Changes in the clinical and laboratory features of lupus patients after the big earthquake in Taiwan Lupus, February 1, 2002; 11(2): 109 - 113. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
