Abstract
Objective: To examine the relationship between stressful life events and psychological distress, and the subsequent development of gadolinium-enhancing (Gd+) brain lesions.
Background: It has long been speculated that stressful life events and psychological distress are associated with disease exacerbation in MS. This is the first prospective longitudinal study of the relationship between stressful life events, psychological distress, and disease activity as measured by Gd+ brain MRI.
Methods: Thirty-six patients (mean age, 44.4 years; 22 women, 14 men) with relapsing forms of MS were assessed once every 4 weeks for 28 to 100 weeks. Assessments included Gd+ MRI, the Social Readjustment Rating Scale (SRRS), the Hassles Scale, and the Profile of Mood States. The SRRS was altered in the following manner: 1) three items that confounded with MS were eliminated, 2) endorsed items were rated for intensity, and 3) the scale was divided into three subscales: major negative events, conflict and disruption in routine, and positive life events. Data were analyzed using mixed-effects logistic regression to account for intrasubject correlations. Stress and distress measures were used to predict concurrent and future MRI activity.
Results: For the total sample of patients, increased conflict and disruption in routine was followed by increased odds of developing new Gd+ brain lesions 8 weeks later (odds ratio, 1.64; p = 0.00083). There was no strong evidence of a relationship between psychological stress or distress and clinical exacerbation.
Conclusions: These data provide support for the notion that conflict and disruption in routine are related to subsequent disease activity in MS. However, this relationship is not sufficiently robust to predict clinical exacerbations reliably in individual patients.
The notion that psychological stress may trigger disease activity was first considered by Charcot,1 who speculated that grief, vexation, and adverse changes in social circumstance were related to the onset of MS. Since that time, the potential relationship between stressful life events, psychological distress, and clinical exacerbations has been examined in several retrospective case–control studies.2 Although early case–control studies failed to detect a reliable relationship between stressful life events, psychological distress, and first symptom or clinical exacerbations of MS,3,4 more recent studies have suggested such a relationship. Patients with MS are more likely to report stressful life events before the first identified symptom compared with patients experiencing the first symptom of other neurologic disorders or rheumatoid arthritis5 or healthy control subjects.6 MS patients experiencing a clinical exacerbation are also more likely to report stressful life events during the preceding 3 months, compared with MS patients who are clinically stable.7
Retrospective case–control studies of the relationships between stressful life events, psychological distress, and clinical exacerbations are prone to several methodologic limitations. First, retrospective reports are more likely to be inaccurate as time increases between the time of the report and the time of the event. Second, patient beliefs may bias reporting of past events. For example, patients who believe stress affects the frequency or severity of clinical exacerbations may be more likely to report stressful life events before clinical exacerbations. Third, the affective state may influence recall by facilitating access to memories of events with similar affective states.8 Thus, the stress resulting from an MS exacerbation9 may promote recall of previous stressful events.
Longitudinal prospective studies, particularly those in which data are collected at frequent intervals, may be less prone to such pitfalls as poor memory, patient beliefs, and affective state. There have been two longitudinal studies that have examined the relationships between stressful life events, psychological distress, and clinical exacerbations. Franklin et al.10 observed that patients were more likely to report stressful life events and psychological distress before clinical exacerbations than during periods of clinical stability. However, their results must be interpreted cautiously because the evaluation of stress “before exacerbation” was undertaken 1 to 5 days after the onset of symptoms.10 Thus, patient recall may have been influenced by the stress of the exacerbation itself. Sibley11 avoided this pitfall by recording patients’ stressful life events at regular 3-month intervals for a median observation period of 5.3 years. A small but significant relationship between patient-reported marital or job stress and subsequent clinical exacerbation was detected (p < 0.02).
Data also suggest that different types of stress may have differential effects. Although Sibley11 found that marital and job-related stress was followed by clinical exacerbation, major negative life events, such as a death in the family, were not. Patients with MS in a clinical trial in Israel showed a decreased exacerbation rate while under actual or threatened missile attack during the Persian Gulf War and for the 2-month period thereafter.12 This suggests that it may be important to differentiate between major and moderate life stressors when examining the relationship between stress and disease activity.
A recent report by the American Academy of Neurology, which considered the relationship between psychological stress and MS exacerbation, stated that “. . . the strength of the favorable evidence is tempered by the lack at present of a clear biologic model,”2 and called for a study using serial gadolinium-enhanced (Gd+) MRI.
The goals of the present study were 1) to extend earlier observations by examining the relationships between monthly standardized measures of stressful life events, psychological distress, and the appearance of Gd+ MRI lesions; and 2) to understand better the biology of any relationships that were detected. We chose to measure disease activity as the presence of new Gd+ lesions on monthly MR images for three reasons. First, the appearance of new Gd+ lesions is a reliable imaging marker of disease activity, whereas the occurrence of clinical exacerbations is often a matter of clinical judgment. Second, the appearance of new Gd+ lesions is widely accepted to be a very early event in the formation of most new MS lesions. Third, new Gd+ MRI lesions are detected five to 10 times more frequently than clinical exacerbations.13
Methods.
Patients.
Enrollment was offered to patients at the University of California at San Francisco/Mt. Zion MS Center participating in ongoing studies that included monthly MRI and clinical evaluations. These studies included a natural history study of early relapsing–remitting (RR) MS14 and a clinical trial of interferon beta 1b (IFN-β1b) in patients with secondary progressive (SP) MS. Other inclusion criteria were an Expanded Disability Status Scale (EDSS) score15 of less than 7.0 points and the appearance of at least one Gd+ lesion during the first 24 weeks of enrollment.
Overview of study procedures.
All patients provided informed consent to participate in this study. Patients completed Gd+ MRI sessions and standardized neurologic evaluations every 4 weeks for 28 to 96 weeks. Standardized measures of stressful life events and psychological distress were administered by telephone 24 hours before each monthly MRI session. Length of time in the study was dependent on remaining duration of enrollment in the parent study at the time stress assessments were initiated. Length of time was independent of any other patient variable (e.g., not resulting from attrition).
Neurologic evaluations.
Standardized neurologic evaluations were completed within 72 hours of each monthly MRI session. An EDSS score was derived from each monthly standardized neurologic evaluation by an examining physician who was held constant for each patient for the duration of the study. A clinical exacerbation was defined conservatively as an increase of 1.0 point on the EDSS from the previous examination. Clinical exacerbations that resulted in loss of ability to perform activities of daily living were treated with methylprednisolone (500 to 1,000 mg administered IV for 3 to 5 days). To minimize the known effect of corticosteroids on MRI measures of interest, imaging and assessment procedures were delayed until 14 days after methylprednisolone was discontinued.16
MRI methods.
MR data for 17 patients with RRMS participating in the natural history study were acquired on a General Electric Signa 1.5-T system (Milwaukee, WI) using a quadrature head coil. Before MR image acquisition, the B0 homogeneity over the entire head was optimized using an automated routine supplied by the manufacturer. Two-dimensional gradient echo scout images, acquired in three planes, were used for positioning of subsequent MR images. All subsequent MR images were collected with full-brain coverage using a 256 × 256 matrix collected over a 220 × 220-mm field of view and contiguous slices that were angled parallel to an imaginary line connecting the anterior commissure to the posterior commissure. The nominal in-plane resolution was 1 mm2. An axial fast spin-echo scan (repetition time [TR], 2,500 msec; echo time [TE], 16 and 80 msec) was acquired with a 3-mm slice thickness, and an axial conversion spin-echo scan (TR, 2,000 msec; TE, 30 msec) was acquired with a 5-mm slice thickness. A T1-weighted spin-echo MR image (TR, 550 msec; TE, 13 msec; number of excitations, 2) was acquired before T1 weighting and 10 minutes after T1-weighted Gd+ IV injection of gadodiamide (Omniscan, Nycomed, Inc., Princeton, NJ; 0.1 mmol/kg). All monthly scans were coregistered to the baseline proton density (PD) image using the algorithm by Woods et al.17
MRI data for 19 patients with SPMS participating in the clinical trial of IFN-β1b were acquired on a Siemens Vision 1.5-T system (Erlangen, Germany) using a quadrature head coil. Prior to MR acquisition, the B0 homogeneity over the entire head was optimized using an automated routine supplied by the manufacturer. Two-dimensional gradient echo scout images, acquired in three planes, were used for positioning of subsequent MR images. All subsequent MR images were collected with full-brain coverage using a 192 × 256 matrix collected over a 180 × 240-mm field of view and 3-mm contiguous slices that were angulated parallel to an imaginary line connecting the anterior commissure to the posterior commissure. The nominal in-plane resolution was 1 mm2. During each monthly MRI session, full-brain image sets were acquired in the following order: PD images (TR, 2,500 msec; TE, 20msec) and T2-weighted images (TR, 2,500 msec; TE, 80 msec) were acquired using a double-spin-echo sequence. A T1-weighted MR image (TR, 600 msec; TE, 17 msec) was acquired before T1 weighting and 10 minutes after T1-weighted Gd+ IV injection of gadodiamide (Omniscan; 0.1 mmol/kg). All MR images were coregistered to the PD MR image collected during the first MRI session using the algorithm of Woods et al.17
New lesions appearing on all images were identified by visual inspection of the T2-weighted and T1-weighted Gd+ images by a single radiologist who was masked to patients’ clinical information. New disease activity was defined as the presence of one or more Gd+ lesions not visible on the previous MR image.
Measures of stressful life events and psychological distress.
All measures of stress and distress were administered over the telephone. Although formal validation for telephone administration has not been performed with these measures specifically, there are many studies indicating that telephone administration of similar psychological and neuropsychological assessments is equivalent to face-to-face administration.18-26 These measures have also been administered by telephone in previous MS studies.27,28
The Holmes and Rahe Social Readjustment Rating Scale (SRRS).
Stressful life events were rated using the Holmes and Rahe Social Readjustment Rating Scale (SRRS),29 which requests the patient to endorse 43 life events that do not occur on a regular basis. Several a priori modifications were made to the SRRS. First, two items that could be confounded with MS and its symptoms were eliminated—“major illness or injury” and “sexual difficulties.” A third item, “pregnancy,” was also eliminated because it is known to affect disease activity.30 Second, because the SRRS has been criticized for not including a rating of experienced intensity of stress, we added a 3-point Likert rating of intensity to endorsed items.28,31 Third, the SRRS has been criticized for not differentiating between different types of stress (e.g., including both positive and negative items).31 We therefore divided the SRRS items into positive and negative events. Because major negative life events may have neutral11 or inhibitory12 effects on disease exacerbation whereas moderate stressors may be associated with increased exacerbation,7,11 negative items were further divided into major and moderate life stressors. This resulted in three subscales: major negative stress (e.g., death of family member), conflict and disruption in routine (e.g., family or job conflict, changes in routine), and positive stress (e.g., outstanding personal achievement).
Hassles Scale.
In contrast to major stressful life events, the Hassles Scale32 assesses common events that are irritating or mildly stressful. Patients are asked to endorse 117 common stressful events. Events that occurred are rated for stress intensity on a 3-point Likert scale.
Profile of Mood States (POMS).
Psychological distress was measured using the POMS.33 The POMS consists of 65 adjectives describing mood states that are rated on a 5-point Likert scale. The POMS produces six subscales; however, these subscales are highly intercorrelated and therefore are not reported here.
Statistical analysis.
To assess the relationship between stress and MRI activity within all patients and by disease type (e.g., RRMS versus SPMS), measures of stressful life events and psychological distress were used to predict the appearance of new Gd+ MRI lesions. Mixed-effects logistic regression34 was used to account for intrasubject correlation. This provides more interpretable effect estimates than the common, generalized estimating equations approach.35 To normalize differences in the scaling of the various psychological measures, odds ratios (OR) were calculated corresponding to an increase in the predictor equal to its SD observed in this sample (table 1): OR = exp(SD * beta). Because the scaling of psychological measures varied, with a range of 4 to 176 points, this facilitates comparison of different predictors’ OR. To assist interpretation, 95% CI are also provided for significant OR. The relationship between clinical exacerbation and psychological stress and distress was analyzed in an identical fashion. All reported p values are two-tailed.
Means, ranges, and SD of the psychological stress and distress assessments
All analyses were preplanned using predictors measured concurrently, and 4, 8, and 12 weeks before each outcome.
Results.
Patients.
Fifty-two patients were available to participate in this study (22 RRMS and 30 SPMS patients). Of these patients, 36 were enrolled (17 RRMS and 19 SPMS patients). Eleven patients declined participation in the monthly stress assessments and five patients were excluded because they did not have at least one new Gd+ MRI lesion in the first 6 months of potential enrollment. Of the 36 patients who received stress assessments, 17 in the natural history study had clinically definite RRMS and 19 patients in the clinical trial had clinically definite SPMS.36,37 SPMS patients were older (mean ± SD, 48.1 ± 10.8 years) than RRMS patients (mean, 39.8 ± 6.8 years; t[34] = 2.86, p = 0.0084). There was no strong evidence of differences in gender between SPMS and RRMS (p = 0.37). The mean EDSS score at study enrollment was higher for patients with SPMS (mean, 5.0 points; range, 2.5 to 6.5 points) compared with patients with RRMS (mean, 1.6 points; range, 0 to 3.0 points; t[34]) = 8.0, p < 0.0001). There was no strong evidence that patients with RRMS had more visits with new Gd+ lesions (53%) compared with patients with SPMS (39%; p = 0.61) or that the clinical exacerbation rate differed by MS type (11% versus 4% respectively; p = 0.87). Patients with SPMS were enrolled longer (mean, 77.6 weeks; range, 28 to 96 weeks) than patients with RRMS (mean, 50.13 weeks; range, 28 to 78 weeks; t[34] = 3.59, p = 0.0012), reflecting differences in the protocols of the parent studies. The means and SD for the measures of stressful life events and psychological stress are provided in table 1.
Relationship between measures of stressful life events, psychological stress, and new Gd+ lesions.
The relationship between psychological stress or distress variables and the development of new Gd+ lesions for all patients and for patients with RRMS and SPMS is displayed in table 2.
OR (p values) for psychological stress and distress assessments and presence of Gd enhancement for the total sample and for patients with relapsing–remitting MS (RRMS) and secondary progressive MS (SPMS) separately
Concurrent analyses.
Concurrent analyses revealed no strong evidence that stressful life events or psychological stress influence the odds of experiencing new brain lesions for the total sample (all, p > 0.15). However, there was an interaction effect for MS clinical type on the relationship between conflict and disruption in routine and new Gd+ lesions (p = 0.046). Conflict and disruption in routine was related to increased odds of concurrently developing new Gd+ lesions among patients with SPMS (OR = 1.96; 95% CI, 1.08 to 3.55; p = 0.024). There was also an interaction effect for MS clinical type and the relationship between daily hassles and Gd+ lesions (p = 0.044), although there was no strong evidence of a relationship between daily hassles and Gd+ lesions within patients with RRMS (p = 0.33) or patients with SPMS (p = 0.053).
Four-week lagged analyses.
There was no strong evidence that stressful life events or psychological stress influenced the odds of experiencing new brain lesions 4 weeks later for the total sample (all, p > 0.11). However, there was an interaction effect for MS clinical type on conflict and disruption in routine and Gd+ lesions (p = 0.025). Among patients with SPMS the odds of experiencing new Gd+ lesions increased after increases in conflict and disruption in routine (OR = 1.95; 95% CI, 1.12 to 3.38; p = 0.015).
Eight-week lagged analyses.
Conflict and disruption in routine was related to increased odds of the appearance of new Gd+ lesions 8 weeks later in the total patient sample (OR = 1.64; 95% CI, 1.22 to 2.20; p = 0.00083). This relationship was present in both the RRMS patients (OR = 1.48; 95% CI, 1.03 to 2.15; p = 0.032) and patients with SPMS (OR = 1.75; 95% CI, 1.00 to 3.06; p = 0.047). Daily hassles were related marginally to increased odds of developing new Gd+ lesions 8 weeks later for the total patient sample (OR = 1.35, p = 0.056) but not within each of the RRMS and SPMS subgroups (both, p > 0.19). No other psychological measures showed strong evidence of a relationship to the odds of developing new lesions in the 8-week lagged analyses (all, p > 0.15).
Twelve-week lagged analyses.
There was no strong evidence that stressful life events or psychological distress were related to the appearance of new Gd+ lesions 12 weeks later (all, p > 0.10). There was, however, an interaction with MS clinical type and daily hassles. Among patients with SPMS, increased odds of subsequently developing Gd+ lesions were related to daily hassles (OR = 2.16; 95% CI, 1.23 to 3.77; p = 0.0059) and greater conflict and disruption in routine (OR = 1.78; 95% CI, 1.06 to 3.01; p = 0.026).
Relationship between measures of stressful life events, psychological stress, and clinical exacerbations (EDSS).
Concurrent analyses.
No measures of stress or distress variables were related to concurrent clinical exacerbation (all, p > 0.091).
Lagged analyses.
No stressful life events or psychological distress measures were related to clinical exacerbation in 4-, 8-, or 12-week lagged analyses (all, p > 0.27).
Discussion.
Many patients with MS believe that their clinical exacerbations are related to stressful life events. Although stressful life events have been associated with clinical exacerbation in both case–control5-7 and longitudinal studies,10,11 methodologic problems have not permitted a definitive conclusion.2 These problems include the limitations of memory, biases due to patient beliefs and affective state, and unreliability in the definition and assessment of clinical exacerbation. These problems were addressed in this study through monthly standardized neurologic examinations from which an EDSS score was derived, standardized prospective measures of stressful life events and psychological distress, and presence of new Gd+ MRI lesions. We observed a significant relationship between moderately stressful life events characterized by conflict and disrupted daily routines, and the appearance of new Gd+ lesions 4 to 8 weeks later.
The relationship between stressful life events and the appearance of new Gd+ lesions varied with the type and severity of the stressor. Although stressful life events resulting in conflict and disruption of daily routine were related to increased odds of developing new Gd+ lesions, major negative stressors were not. A similar observation was made by Sibley,11 who reported that medically confirmed exacerbations were related to conflict in the job or family but were not related to major stressors such as death in the family. Consistent data are reported from studies of experimental autoimmune encephalomyelitis (EAE) in which chronic, varied stressors exacerbate38 but severe stressors suppress clinical manifestations of disease.39 Thus, the relationship between stressful life events and disease activity, whether measured as clinical exacerbations or new Gd+ MRI lesions, is not straightforward and appears to depend on many factors including chronicity, severity, and type of stress as well as individual patient characteristics such as temperament, coping skills, level of social support, and psychopathology.40-42 Little is known about how these factors, individually or in combination, are related to clinical exacerbations and the appearance of new Gd+ lesions, and additional work in this area is warranted.
The mechanisms by which stress might lead to MS disease activity are unknown. It has been well documented that stress reduces lymphocyte counts and proinflammatory cytokine production, which then increase 4 to 5 weeks thereafter.43,44 When stress in EAE mice is eliminated by sympathectomy, the mice experience fivefold increases in T cell independent antibody responses, increased expression of β2-adrenergic receptors on T cells, and increased severity of EAE.45 In patients with MS, β2-adrenergic receptor expression is increased on CD8 cells in patients with progressive disease, is relatively normal in patients who are clinically stable, and varies in relation to MRI activity.44,45 These observations have led to the notion that stress may be accompanied by immune suppression and may be followed by immune activation. Our observations are consistent with this hypothesis and suggest this relationship may be mediated in part by an effect on the blood–brain barrier or vascular permeability.
Although patients with stressful life events that produced conflict or disrupted daily routines were more likely to experience new Gd+ lesions, we found no strong evidence of a relationship between stressful life events or measures of psychological distress and clinical exacerbations. Thus, the relationship between stress and MRI activity in our study was not sufficiently robust to be clinically evident, and our results do not confirm the results reported in earlier longitudinal studies.10,11 It is possible that this discrepancy is the result of differences in the measurement of clinical exacerbation. Previous investigators defined clinical exacerbations as new or worsening symptoms associated with a confirmatory change on neurologic examination lasting at least 48 hours in the absence of fever. We additionally required that change on neurologic examination result in a change of 1.0 point or more on the EDSS. This stringent definition made exacerbations too rare to be studied definitively with this number of patients (for example, the OR of 1.76 for the association of conflict and disruption of routine and clinical exacerbation 8 weeks later had a CI of 0.72 to 4.49, consistent with either no influence or a very important influence). Nonetheless, we believe our definition of clinical exacerbation is less likely to overemphasize minor fluctuations associated with nonspecific symptoms or measurement error in a standardized neurologic examination.
We recognize certain limitations, none of which invalidates the results of this study. First, the relationship we observed was between stressful life events and visually detected new Gd+ MRI lesions. MRI activity can be detected by quantitative MRI approximately 1 month before new Gd+ lesions become visible.14,47 Thus it is possible the relationship we report could be detected within 4 weeks instead of between 4 and 8 weeks. This issue will have to be addressed in future studies. Second, the majority of patients with SPMS were participating in a placebo-controlled clinical trial of IFN-β1b. The treatment assignment of these patients is not yet known. Because IFN-β can reduce the number of new Gd+ lesions, we think our results are probably a conservative estimate of the relationship between stressful life events and this measure of MRI activity in this group. Third, patients with RRMS and SPMS came from two separate studies, with different protocols and different patient samples. Patients entering a natural history study may have fewer concerns about their level of experienced disease activity compared with patients willing to enter a placebo-controlled clinical trial study. Because of confounding between study and type of MS, we have drawn no conclusions about differences between the disease types. Fourth, we undertook numerous statistical comparisons and acknowledge that more restrictive p values might be used to define the significance of the relationship between stressful life events and new Gd+ lesions. For example, p values less than 0.0025 would be required for significance in a study in which a conservative Bonferroni adjustment was made for 20 (five measures × four time points) independent comparisons. If such an adjustment was made for our data, the relationship between the appearance of new Gd+ lesions 4 to 8 weeks after stressful life events would remain highly significant.
- Received October 18, 1999.
- Accepted in final form March 22, 2000.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Use of Whole-Genome Sequencing for Mitochondrial Disease Diagnosis
Dr. Robert Pitceathly and Dr. William Macken