Glabellar and palmomental reflexes in parkinsonian disorders

Methods.

The subjects were recruited consecutively from the population of patients, family members, friends, and others attending the Parkinson’s Disease Center and Movement Disorders Clinic (PDCMDC), Baylor College of Medicine. The study was approved by the Institutional Review Board for Baylor College of Medicine and Affiliated Hospitals. After signing the consent subjects were given the Mini-Mental State Examination (MMSE) and were rated on the Unified PD Rating Scale (UPDRS). The test subjects were then taken to the PDCMDC laboratory where they were seated and videotaped according to a structured video protocol (see the video on the Neurology Web site at www.neurology.org). For the GR the video screen was framed so as to only show the eyes and forehead region, but no other parts of the body. Subjects were instructed to look straight ahead and the glabellar region was tapped with a neurologic hammer 10 times at a frequency of approximately 1 per second. The stimulus was administered from outside the field of vision to prevent an avoidance response.4,9⇓ For the PMR the video screen was framed to show only the chin region, but no other part of the body. Subjects were instructed to look straight ahead while a stimulus was applied with a sharp stroke of the end of a neurologic hammer from proximal to distal aspect of the thenar eminence, to the left then right hand. The stimulus, intended to cause discomfort but no pain, was administered 10 times at a frequency of 1 per second, and each time across the same location to accurately judge habituation or lack of habituation of the reflex.7 As the video focused on the mentalis muscles, each stroke was accompanied by a verbal cue of 1 to 10 to indicate when the stimulus was being applied. The videos were subsequently viewed and rated blindly by a board-certified neurologist and movement disorders specialist (M.T.) who was not aware of the diagnosis and was not involved in either the recruitment or the testing. The GR was rated on a scale of 0 to 5: 0 = no reflex blinking upon stimulation, 1 = reflex blinking habituates after < 5 taps, 2 = concomitant blinking for most but not all stimulus taps, 3 = concomitant blinking persists for entire stimulus period (10 taps), 4 = excessive reflex blinking in addition to the 10 stimulus taps, and 5 = excessively strong and persistent blinking and/or closure (blepharospasm). The PMR was rated on a scale of 0 to 5: 0 = no reflex movement of mentalis muscles, 1 = weak ipsilateral mentalis muscle contraction, 2 = strong ipsilateral mentalis muscle contraction, 3 = weak bilateral mentalis muscle contraction, 4 = strong bilateral mentalis muscle contraction, 5 = persistent bilateral muscle contraction throughout stimulus period.

Parametric and nonparametric analyses were employed on raw and dichotomized scores: χ² tests, Kruskal-Wallis tests, Mann-Whitney U test, Spearman rho correlations, and analysis of variance. Using dichotomized data allows for clear group definitions, but at the cost of lower statistical power. The PR were analyzed using both non-dichotomized scores (PR rating scale of 0 to 5) and dichotomized scores, with a rating of 0 equaling reflex absence and 1 to 5 equaling reflex presence. Biographical data were analyzed using descriptive statistics. All analyses were performed using SPSS version 10.0 (SPSS, Inc., Chicago, IL). To help minimize type 1 and type 2 errors with multiple comparisons, a more conservative statistical significance level was set, α < 0.01.

Results.

A total of 100 subjects (55 men) were enrolled in the study; the mean age was 65.7 ± 9.0 years (range, 44.3 to 84.5) (table 1). Compared to normal controls, parkinsonian patients as a whole scored lower on the MMSE scale (p < 0.007) (see table 1). Both GR and PMR were more likely to be present in patients with PD as compared to normal controls (p < 0.01 and p < 0.015). GR (p < 0.018) but not PMR (p = 0.331) was significantly more frequent in patients with PSP as compared to controls. Furthermore, the PR did not differentiate between normal controls and patients with MSA (table 2A). Additionally, there was no statistical difference among the three parkinsonian disorders with respect to GR and PMR presence, indicating lack of specificity. As compared to normal age-matched controls, PMR ratings were higher among patients with PD (p < 0.006) and GR ratings were higher among patients with PD, PSP, and MSA (ps < 0.007) (table 2B). Spearman correlative analysis showed that higher GR ratings correlated with higher PMR ratings (ρ = 0.27, p < 0.007) across all four groups, but there was no correlation between MMSE scores and either GR (ρ = −0.08, p = 0.45) or PMR (ρ = 0.08, p = 0.43).

Based on videotaped rating, we found GR to be a sensitive sign of parkinsonian disorders (83.3%) with modest positive predictive value (70.4%), but it lacks sufficient specificity (47.5%) to serve as a reliable diagnostic marker of a particular parkinsonian disorder. In contrast, PMR has a poor sensitivity (33.3%), but relatively high specificity (90.0%) and high positive predictive value (83.3%) (table 3).

Discussion.

Although PRs have been part of neurologic examination for more than a century, their specific diagnostic utility in parkinsonian disorders have not been fully evaluated. While both GR and PMR were more frequently present among parkinsonian patients than in controls, only PD patients differed significantly from controls with respect to both GR and PMR. PSP patients differed from controls with respect to only GR and MSA patients were not statistically different from controls for either PR. This apparent inconsistency may be related to the small sample size of both the PSP and the MSA groups (12 and 7). On the other hand, it may indicate that PRs are relatively useful indicators for only PD as opposed to all parkinsonian disorders.

PR have been found more frequently with basal ganglia disorders and they may be reversed with levodopa, but the relationship between PR and underlying dopaminergic system has not been clarified.6,10⇓ One study showed that GR is frequently present in PD, but it does not correlate with any particular sign of PD, such as tremor, rigidity, bradykinesia, loss of balance, micrographia, or shuffling gait.9

The other goal of this study was to examine the correlation between the presence of PR and cognitive decline. Higher GR scores correlated with higher PMR scores across all four groups, but there was no correlation between the MMSE scores and either PMR or GR. The lack of apparent relationship between cognitive decline, based on MMSE, and the presence of PR in our study is consistent with the findings from other studies.2,5,6⇓⇓ Despite the high association of GR and PMR with PD as compared to controls, shown in our study, the PR should not be considered specific parkinsonian signs as they may be elicited in a variety of neurodegenerative disorders. We, therefore, conclude that the chief utility of these reflexes resides in their value as indicators of underlying neurodegenerative process, rather than any specific parkinsonian disorder. Larger sample sizes in all parkinsonian groups will be needed in future studies designed to test the feasibility of using PR as a diagnostic tool.