The Autonomic Symptom Profile

Questionnaire design.

The questionnaire was designed to address the following criteria:

• 1. Items should reflect domains or systems (i.e., orthostatic, secretomotor, gastrointestinal) that are relevant to patients with autonomic disorders.

• 2. Summary quantitative scores should be provided by domain and overall for each person.

• 3. The scale scores should be clinically meaningful and accurate.

• 4. The questionnaire should demonstrate face, content, and criterion validity.

Questions.

We asked questions used routinely in practice. We specifically asked questions about onset and evolution of symptoms as well as their frequency and severity, exacerbating and relieving factors, and possible relationship to meals, standing, exercise, and other factors influencing autonomic function.

Questionnaire.

The questionnaire was 169 items, including demographic questions. The items were generated by seeing patients with autonomic disorders, having experience in the design of other questionnaires, engaging in discussion with clinicians specializing in autonomic disorders, reviewing other questionnaires, and conducting extensive and unstructured interviews with patients with autonomic disorders and peripheral neuropathy.1 From this set of questions, 73 items emerged as the most important and frequently asked questions about autonomic symptoms. This profile is an expansion of an earlier questionnaire used to evaluate symptoms of orthostatic hypotension.2 These 73 questions assess the following nine domains of autonomic symptoms: orthostatic (9 items); secretomotor, including sudomotor symptoms (8 items); male sexual dysfunction (8 items); urinary (3 items); gastrointestinal, including gastroparesis, diarrhea, and constipation (14 items); pupillomotor, including visual symptoms (7 items); vasomotor (11 items); reflex syncope (5 items); and sleep function (8 items). Some categories consisted of primary questions followed by conditional modifying questions. This approach was helpful in saving time for those patients with few or no symptoms. The nine domains were chosen because they are commonly affected in autonomic disorders and because they have the potential to aid in recognizing patterns or systems involved with possible diagnostic implications. The following is a sample question from the orthostatic domain: “In the past year, have you ever felt faint, dizzy, or ‘goofy’ or had difficulty thinking soon after standing up from a sitting or lying down position?” A group of 12 questions aimed at detecting psychosomatic indices (6 items) and understatement indices (6 items) were mixed with the autonomic items. An example of such an understatement question is: “Have you ever in your adult life had difficulty keeping your mind on your job or task?”

Validation.

Assessing the degree of criterion validity was accomplished by correlating the COMPASS scores with the scores of the Composite Autonomic Scoring Scale (CASS).3 We selected the CASS, which objectively quantitates autonomic deficits derived from the Autonomic Reflex Screen (ARS), as our criterion “gold standard” assessment. The ARS consists of a comprehensive battery of noninvasive autonomic tests that are quantitative, sensitive, specific, reproducible, and standardized.4 Autonomic tests were performed as previously described,5 and the elements of the ARS include the Quantitative Sudomotor Axon Reflex Test, which evaluates one upper and three lower limb sites; orthostatic blood pressure and heart rate response to tilt; heart rate response to deep breathing; the Valsalva ratio; and beat-to-beat blood pressure measurements during the Valsalva maneuver, tilt, and deep breathing. The CASS is a 10-point scoring system of autonomic function. The scale rates adrenergic deficits by one of four grades and sudomotor and cardiovagal deficits each by one of three grades. Each score is normalized for the compounding effects of age and sex. This score ranges from 0 (no deficit) to 10 (maximal deficit).

Study subjects.

Three subject groups were studied: normal controls (NMLs), patients with peripheral neuropathy (PN), and patients with neurogenic autonomic failure (NAF). We consecutively recruited patients referred for evaluation of PN and autonomic dysfunction. Normal subjects were recruited from the community and healthy controls from a longitudinal population-based study (Rochester Diabetic Neuropathy Study). All NMLs received medical care at Mayo Clinic Rochester, and their medical records were reviewed to exclude those with neurologic or autonomic disorders. None of the control subjects were taking medications known to influence the results of autonomic tests. When medically permissible, we asked patients in the PN and NAF groups to discontinue medications that could alter the results of autonomic tests for 24 to 48 hours before testing.5

A total of 113 subjects completed a self-administered questionnaire before undergoing autonomic tests. To ensure compliance and minimize missing items, the questionnaire was assessed for completeness by the study coordinator before autonomic studies. The diagnoses of patients in the PN and NAF groups were determined before analyzing the responses to the questionnaire.

Statistical analysis.

Nonparametric methods were used for data analysis because of skewness in the overall COMPASS score and subscores. The Wilcoxon rank sum test was used to make pairwise comparisons between the three subject groups and between the sexes. Univariate and age- and sex-adjusted Spearman rank correlations were used to describe the overall association between the CASS score and the COMPASS subscores. Stepwise regression using Kendall’s τ_b from the correlation matrix6 was used to assess the association of the COMPASS scores, age, and sex with the CASS. Statistical significance was defined as a two-tailed p value ≤ 0.05.

General observations.

The questionnaire was well received and was completed in approximately 20 minutes in most cases.

Clinical features of groups.

Demographic details of the subjects are summarized in table 2. Individuals in the NML group tended to be younger than those in PN and NAF groups. The age and sex distribution of patients in the PN and NAF groups were not significantly different. The etiology of the nonautonomic neuropathies was as follows: distal small fiber neuropathy (10), monoclonal gammopathy-associated neuropathy (5), diabetic neuropathy (3), chronic inflammatory demyelinating polyradiculoneuropathy (3), hereditary motor and sensory neuropathy (2), multiple mononeuropathies and immune-mediated neuropathies (6), neuropathy associated with osteosclerotic myeloma (2), idiopathic sensory neuropathy (2). The causes of generalized autonomic failure in 39 patients of the NAF group were as follows: idiopathic autonomic neuropathy (17), pure autonomic failure (9), parkinsonism-plus (4), diabetic autonomic neuropathy (3), multiple system atrophy/Shy Drager syndrome (2), amyloidosis (2), system degeneration (1), chronic neuropathy (1). The clinical characteristics of patients with idiopathic autonomic neuropathy have been recently reported.7

Score profiles in healthy subjects and in patients with peripheral neuropathy and neurogenic autonomic failure.

Patients in the NAF group had significantly higher overall COMPASS scores (52.3 ± 24.2 [mean ± SD]) than did NMLs (9.8 ± 9) and PN patients (25.9 ± 17.9). The figure illustrates the distribution of COMPASS scores in the three groups and by sex. Table 3 shows the COMPASS scores for the different domains in the three groups. The orthostatic domain included symptoms such as feeling dizzy or “goofy” and difficulty thinking upon standing. There were additional questions aimed at identifying the presence of aggravating factors such as relationship with meals, prolonged standing, exercise, and increased body temperature. Orthostatic intolerance scores were higher in NAF and PN patients than they were in NMLs. There was also a statistically significant difference for orthostatic intolerance between the NAF and PN groups. The secretomotor domain included symptoms characteristic of thermoregulatory impairment such as heat intolerance, changes in general and specific areas of body sweating, and dry eyes and mouth. These scores were also consistently higher in PN and NAF patients than they were in NMLs and statistically significant when compared among the different groups. Symptoms of sexual difficulty were more common in NAF patients than they were in any other group. This subscore of the questionnaire applied only to 60 men because there were no questions addressing sexual function in women. There was no statistical difference between NAF and PN patients.

• Download figure
• Open in new tab
• Download powerpoint

Figure. The distribution of Composite Autonomic Symptom Scale (COMPASS) scores by sex, control, nonautonomic peripheral neuropathy (PN), and neurogenic autonomic failure (NAF). The box and whisker figures represent the extremes and the 25th, 50th, and 75th percentile values. ○ = mean value.

Scores for urinary symptoms consisted of data regarding difficulty voiding, urinary retention, and loss of control of bladder function. These scores were higher in the NAF and PN groups than they were in NMLs and were statis-tically significant between NAF and PN patients. The domain for gastrointestinal symptoms included gastroparesis, diarrhea, and constipation as the most frequently reported gastrointestinal symptoms in NAF and PN patients. COMPASS subscores assessing pupillomotor and visual symptoms such as blurred vision, photophobia, trouble focusing, and difficulty seeing at night were higher in the NAF group than they were in the PN and NML groups. Vasomotor symptoms subscores, such as color changes in the skin, were not significantly different between NAF and PN patients but were more frequent in both of these groups than they were in NMLs. Scores for reflex syncope were obtained from questions such as, “Have you ever fainted while passing urine or coughing?” This mean domain score was significantly higher in the NAF group compared with the NML group and marginally significantly different between NAF and PN groups. Items assessing the sleep domain included questions to identify symptoms associated with obstructive sleep apnea, narcolepsy, and abnormal or disordered sleep patterns. Sleep domain scores were significantly higher in NAF and PN patients than they were in NMLs, but the difference between NAF and PN patients was not statistically significant.

The understatement index, for which high scores would be indicative of a more stoic personality, included questions such as, “Have you ever in your adult life had difficulty keeping your mind on your job or task?” There was a higher mean score in NAF and PN patients than there was in NMLs. The difference was not statistically significant between NAF and PN subjects and was borderline between the NAF and NML groups. The psychosomatic index for which higher scores would be reflective of neurotic, hypochondriac, or suggestible respondents included questions such as “In the past 5 years, how would you rate the amount of trouble, if any, you have had with feelings of complete weakness all over the body?” and “in the past 5 years, how would you rate the amount of trouble, if any, you have had with oversensitive hearing?” The mean score was slightly higher in NAF patients than it was in the PN and NML groups, but there was no statistically significant difference between NAF and PN groups.

Severity of autonomic failure.

We evaluated the presence and severity of autonomic deficits in all groups using the CASS. The CASS was significantly increased in the NAF group (6.9 ± 2; p < 0.001) and in the PN group (3.1 ± 2.3; p < 0.001) over NMLs (mean = 0). The mean overall score for NAF patients was more than two times higher than the score for patients in the PN group.

Validity.

Criterion validity was assessed by the degree of association between the COMPASS overall and domain-specific scores with the CASS overall score. Table 4 presents the Spearman rank correlation coefficients and the partial Spearman rank correlation, adjusting for age and sex, of CASS with COMPASS subscores, in which the independent variable (x) was the COMPASS total score and the dependent variable (y) was the CASS score. For the overall total COMPASS score, the rank correlation was 0.67. A good correlation was observed with the orthostatic intolerance and secretomotor subscores. A modest correlation was found with pupillomotor, male sexual dysfunction, gastroparesis, reflex syncope, urinary symptoms, vasomotor, diarrhea, and constipation domain-specific subscores. No correlation was noted between the sleep disorder COMPASS score and the CASS score. In addition, the two validity scores for psychosomatic and understatement indices were not found to be associated with the CASS score.

We performed stepwise linear regression with Kendall’s τ_b correlations to examine the relationship between CASS and the domain-specific COMPASS scores. Combining all three subject groups, the COMPASS domain-specific scores for orthostatic intolerance (p < 0.001) and secretomotor symptoms (p < 0.001) were the variables that in combination best predicted the CASS. Overall, the analysis demonstrated an acceptable level of criterion validity of the questionnaire compared with the standard instrument, the ARS.