Individuals with myasthenia gravis (MG) are required to cope with a chronic illness characterized by a variable course and uncertain prognosis. As in other chronic and unpredictable neuroimmune diseases,
1 the prevalence of major depression among MG patients appears to be significantly elevated compared with healthy control subjects.
2 However, relatively few studies have formally examined the psychiatric manifestations of MG, and studies that have focused on this issue have yielded somewhat conflicting results.
3–6Unfortunately, these previous studies are limited by small sample sizes and/or the use of depression measures that confound symptoms of disease activity with common features of depression. Nyenhuis et al.
7 identified the latter problem for multiple sclerosis (MS) patients. Individuals with MS frequently present with illness-related symptoms that are characteristically similar to depression (e.g., fatigue). To circumvent this confound, Nyenhuis and co-workers
8,9 developed the Chicago Multiscale Depression Inventory (CMDI) to assess depression. This 42-item self-report inventory consists of three subscales, each of which assesses either the vegetative, evaluative, or mood symptoms of depression. Studies of the CMDI with MS patients reveal significant elevations on the vegetative scale and total score, but not on the mood or the evaluative scales.
8In the present study, we administered the CMDI to a sample of MG patients and a group of healthy control subjects to determine if the pattern of ratings reported by MG patients would be similar to the pattern previously reported in MS. We predicted that MG patients would endorse significantly more items than control subjects on the vegetative subscale and the total score of the CMDI, but not on the mood and evaluative subscales.
METHODS
A total of 29 individuals with MG were recruited from local chapters of the National Myasthenia Gravis Foundation (NMGF). Diagnoses of MG were confirmed by a positive edrophonium chloride (Tensilon) test, the presence of anti-acetylcholine receptor antibodies in serum, or electromyographic findings consistent with MG. Patients were excluded if they reported a history of alcohol or drug abuse, major psychiatric disorder (e.g., schizophrenia, personality disorder), or neurological disorder other than MG, or if they were taking more than 60 mg of prednisone daily.
Disease severity was assessed with a modified version of disease classification for MG.
10 Severity was rated on a three-point scale (1=mild, 3=severe) based on the degree of interference with completion of daily activities despite medications. Most patients were taking regular doses of pyridostigmine (
n=22), azathioprine (
n=12), or prednisone (
n=15; median daily dose=16.0 mg). Patients averaged 56.8±13.78 (SD) years of age, 15.27±2.76 years of education, 7.57±6.67 years since diagnosis, and 1.62±0.67 on the disease severity scale (mild-moderate).
The 34 control subjects were either people recruited from the community by fliers (n=4), hospital volunteers (n=24), or relatives of patients (n=6). Exclusion criteria were the same as for patients. Control subjects averaged 47.63±16.65 years of age and 15.51±2.36 years of education.
All participants provided written informed consent after receiving a comprehensive description of the study. Participants from both groups volunteered to complete the study without compensation.
All subjects were administered the CMDI.
9 The CMDI consists of three subscales that separately measure the mood (e.g., sad), evaluative (e.g., useless), and vegetative (e.g., tired) symptoms of depression. Each scale consists of 14 items. For review of subscale items and principal component analysis loadings, see Nyenhuis et al.
9 The three subscales and the overall total score have been found to be reliable and valid and to correlate very highly with clinical measures of depression. Importantly, the mood subscale has been shown to be a sensitive measure of mood disturbance among individuals with major depression.
8RESULTS
There was no significant group difference in education (F<2, P>0.05), but control subjects were significantly younger than the patient group (F=5.59, df=1,61, P<0.05).
Between-group analyses of variance were computed to examine scores on each of the subscales and the total score of the CMDI between the patient group and the control group (
Table 1). In agreement with our predictions, results of these analyses revealed significantly higher scores among MG patients on the vegetative subscale (
F=25.57, df=1,61,
P<0.05) and the total score on the CMDI (
F=10.28, df=1,61,
P<0.05); group means were not statistically different on the mood or evaluative subscales (
F<2).
Subsequent analyses were conducted to examine the frequency of depression in MG compared with the control sample, as measured by each of the subscales and the total score of the CMDI. Clinically meaningful depression was defined as scores greater than 1.5 standard deviations from the means obtained from the control sample. Results of these analyses revealed that the frequency of depression was 17% in the MG group and 8% in the control sample when depression was measured by using only the mood subscale. A Mantel-Haenzsel chi-square analysis (correction for small cell sizes) revealed that this difference was not statistically significant (P>0.05). Similarly, when depression was assessed with the evaluative subscale, 17% of MG patients and 11% of control subjects met criteria for depression. Again, this group difference was not statistically significant. Measuring the frequency of depression with the vegetative scale produced very large group differences: 41% of the MG patients met criteria for depression, and only 8% of the control sample were classified as depressed. This group difference was statistically significant (χ2=8.99, df=1, P>0.01). Similarly, when the total score of the CMDI was used to assess depression, 24% of MG patients and 5% of the control sample fell in the significant range. Again, this group difference was statistically significant (χ2=4.19, df=1, P<0.05).
Regression analyses revealed that MG disease severity significantly predicted scores on only the vegetative subscale (F=4.23, R2=0.585) and the total score of the CMDI (F=5.87, R2=0.348). Disease duration and average daily dose of prednisone did not predict scores on any of the scales of depression.
DISCUSSION
The findings from the current study are consistent with previous studies examining depression in MS
8 and suggest that patients with neuroimmune diseases experience significant illness-related symptoms characteristically similar to the vegetative signs of affective disturbance. Moreover, the findings provide further support for observations reported by Nyenhuis and co- workers
7–9 that self-report measures of depression that do not independently account for similarities between typical symptoms of neuroimmune diseases and vegetative symptoms of depression may in fact overestimate the prevalence of depression.
Several limitations of the study warrant discussion. First, mood was examined in patients who were recruited from NMGF chapters and who did not have severe disease activity. Individuals with more advanced disease who do not participate in support groups may present to clinicians with more severe mood disturbance. Second, the overall prevalence of depression in MG cannot be concluded from the present study given the limited sample size. Finally, the absence of a clinical psychiatric interview and of a comparison group with diagnoses of depression but not MG limits the scope of conclusions that can be drawn from the current study.
Despite the concerns noted above, the critical message of this paper, that physical symptoms of MG may artificially inflate the prevalence of depression in MG, is retained. This issue is important because between 20% and 30% of MG patients are initially misdiagnosed with a psychiatric illness.
11 Perhaps even more clinically significant, symptoms of depression among individuals already diagnosed with MG may be dismissed as symptoms of the neurologic disease, thus delaying appropriate treatment of the mood disturbance. Results from this study, as well as previous studies of MS patients,
8 suggest that the CMDI may serve as a useful screening measure to rule out depression in clinic patients.
ACKNOWLEDGMENTS
This work was funded by National Institute of Neurological Disorders and Stroke Training Grant F32 NS10909-01.