Visuospatial Deficits Are Associated with Pisa Syndrome and not Camptocormia in Parkinson's Disease

ABSTRACT

Background

Pisa syndrome (PS) and camptocormia (CC) are postural abnormalities frequently associated with Parkinson's disease (PD). Their pathophysiology remains unclear, but the role of cognitive deficits has been postulated.

Objectives

To identify differences in the neuropsychological functioning of patients with PD with PS or CC compared with matched patients with PD without postural abnormalities.

Methods

We performed a case‐control study including 57 patients with PD with PS (PS+) or CC (CC+) and 57 PD controls without postural abnormalities matched for sex, age, PD duration, phenotype, and stage. Patients were divided into four groups: PS+ (n = 32), PS+ controls (PS−, n = 32), CC+ (n = 25), and CC+ controls (CC−, n = 25). We compared PS+ versus PS− and CC+ versus CC− using a neuropsychological battery assessing memory, attention, executive functions, visuospatial abilities, and language. Subjective visual vertical (SVV) perception was assessed by the Bucket test as a sign of vestibular function; the misperception of trunk position, defined as a mismatch between the objective versus subjective evaluation of the trunk bending angle >5°, was evaluated in PS+ and CC+.

Results

PS+ showed significantly worse visuospatial performances (P = 0.025) and SVV perception (P = 0.038) than their controls, whereas CC+ did not show significant differences compared with their control group. Reduced awareness of postural abnormality was observed in >60% of patients with PS or CC.

Conclusions

Low visuospatial performances and vestibular tone imbalance are significantly associated with PS but not with CC. These findings suggest different pathophysiology for the two main postural abnormalities associated with PD and can foster adequate therapeutic and prevention strategies.

Pisa syndrome (PS) and camptocormia (CC) are disabling trunk postural abnormalities that might be observed in more than 20% of patients with Parkinson's disease (PD) and have been associated with higher disability and worse quality of life. ¹ , ²

The pathophysiology of PS and CC still needs to be elucidated, and many hypotheses have been proposed, including dystonia, alterations of proprioception, body schema perception, or vestibular function. ¹ , ³

Prior studies have suggested that PS and CC may be associated with cognitive alterations involving visuospatial and attentive functions. ⁴ , ⁵ , ⁶ , ⁷ However, such an association has not been robustly confirmed because the available evidence stems from a few single‐center studies with small sample sizes, often with inadequate methodologies to test this hypothesis. Confirming a relationship between postural abnormalities and cognitive dysfunction would be important not only for a better understanding of PS and CC pathophysiology but also for early recognition of patients at risk for developing postural alterations and for designing specific rehabilitation programs that might combine cognitive and physical therapy.

In this multicenter, case‐control study, we therefore aimed to identify differences in the neuropsychological functioning of patients with PD with PS and CC compared with patients with PD without postural abnormalities matched for the main demographic and clinical characteristics. As secondary aims, we explored the awareness of the trunk postural alteration in patients with PS and CC and further compared their perception of the subjective visual vertical (SVV) against patients without postural abnormalities.

Materials and Methods

Study Population

Consecutive consenting patients were enrolled in 7 Italian movement disorders centers and 1 German movement disorders center with a harmonized protocol reported in the section ‘Outcome Measures’ and according to the following inclusion criteria: diagnosis of idiopathic PD according to the Movement Disorders Society (MDS) clinical diagnostic criteria ⁸ ; the presence of PS, defined as a lateral flexion of the trunk of at least 10°, almost completely reverted by passive mobilization or supine positioning, ¹ or the presence of CC, defined as an anterior flexion ≥30° of the lumbar spine (lower CC) or ≥45° of the thoracic spine (upper CC), which disappears in the supine position. ⁹

We excluded patients with concomitant neurologic diseases known to negatively affect posture (ie, myopathy, motor neuron disease, myasthenia), history of major spinal surgery or muscle and/or skeletal diseases, diagnosis of severe dementia or depression rendering impossible to accomplish the neuropsychological assessments, treatment with drugs potentially able to induce abnormal postures (ie, typical and atypical neuroleptics, tricyclic antidepressants, cholinesterase inhibitors, antiemetic drugs, lithium carbonate) in the 6 months before enrollment, ¹⁰ and treatment of posture alterations with botulinum toxin injections in the 6 months before enrollment.

For every enrolled patient with PS and CC (“cases”), a control with PD and without PS, CC, or other postural alterations was enrolled, matching for the following criteria: sex, age (±3 years), PD duration (±3 years), PD motor phenotype (rigid‐akinetic, tremor dominant, or mixed type), ¹¹ and Hoehn and Yahr stage. ¹²

All centers assessed the patients with the same protocol reported in the section ‘Outcome Measures’, encompassing the clinical and neuropsychological evaluation, and did not rely on historical data. A standard protocol was prepared and agreed among all involved centers to evaluate patients in the morning, during daily on therapeutic conditions, and with shared rules for the harmonization of the evaluations, including the neuropsychological tests and their scoring system and the methods for posture assessment by means of pictures taken with patients undressed and with a fixed camera position and distance from the patients.

The enrollment and assessment period was June 2019 to December 2021. However, a break of variable duration in the recruitment occurred in each center because of the coronavirus disease 2019 pandemic between March 2020 and January 2021.

Outcome Measures

Demographic and Clinical Data

We collected age, education, age at PD onset, PD duration, MDS–Unified Parkinson's Disease Rating Scale (MDS‐UPDRS), Hoehn and Yahr stage, PD motor phenotype, ongoing pharmacological therapies, levodopa equivalent daily dose, ¹³ number of falls in the previous month, ¹⁴ the Charlson Comorbidity Index, ¹⁵ quality of life by means of the Parkinson's Disease Questionnaire–8 (PDQ‐8), and the presence of back or neck pain on a Visual Analog Scale from 0 to 10. ¹⁶

The PD motor phenotype was assigned according to a validated formula: the mean of MDS‐UPDRS items 2.10, 3.15a, 3.15b, 3.16a, 3.16b, 3.17a, 3.17b, 3.17c, 3.17d, 3.17e, and 3.18 was divided by the mean of MDS‐UPDRS items 2.12, 2.13, 3.10, 3.11, and 3.12. If the resultant ratio was >1.15, the patient was classified as tremor dominant; if the ratio was <0.90, as postural instability/gait disorder; and if the ratio was between 0.90 and 1.15, as mixed. ¹¹

Neuropsychological Assessment

We performed a cognitive screening test, the Montreal Cognitive Assessment (MoCA), ¹⁷ and a comprehensive neuropsychological test battery assessing the following 5 cognitive domains: memory, attention and working memory, executive functions, visuospatial abilities, and language.

Memory was assessed by means of the Bi‐Syllabic Word Repetition Test and Rey Auditory Verbal Learning Test ¹⁸ , ¹⁹ ; attention and working memory by means of the Digit Cancellation Test (DCT), ¹⁸ Trail Making Test A (TMT‐A), and Digit Span Backward ²⁰ , ²¹ ; executive functions by means of the Frontal Assessment Battery ²² and Trail Making Test B (TMT‐B) and Trail Making Test B‐A ²⁰ ; visuospatial abilities by means of the Constructional Apraxia Test—a copying drawing test—and Benton Judgment of Line Orientation (BJLO) ¹⁸ , ²³ ; and language by means of the Phonemic Verbal Fluency and Category Verbal Fluency. ¹⁸ , ²⁴

The raw scores of each neuropsychological test were converted into Z scores using published normative data, corrected for age and education when possible, ²⁵ and each center used the normative values of its own country. Finally, a global Z score for each cognitive domain was calculated using the average Z scores of single subtests to obtain a cognitive composite score for memory, attention and working memory, executive functions, visuospatial abilities, and language. ²⁶

Postural Abnormalities Evaluation

PS was measured using standardized photos by means of ImageJ free software (National Institutes of Health, Bethesda, MD). ²⁷ The angle of PS was be measured as the angle between (1) the vertical axis and (2) a line connecting the fulcrum of the bent spine with the seventh spinous process. ²⁸ Latency to develop PS after PD onset (months), PS duration (years), and PS direction (right or left) were also collected from each patient with PS.

CC was measured using standardized photos, according to the recent consensus method for measuring CC, accounting for 2 CC angles: the total CC angle and the upper CC angle. ⁹ The following clinical data on CC were collected: latency to develop CC after PD onset (months), CC duration (years), fluctuation of CC during the day, and pattern of CC onset (acute, <1 month; subacute, ≥1 month to <3 months; chronic, ≥3 months).

Patients with a diagnosis of combined postural abnormalities (eg, PS and CC together or other severe postural deformities) were excluded from the analysis.

Bucket Test for Measuring the Perception of Body Verticality

To determine the presence of alteration in the SVV, a sensitive sign of peripheral or central subcortical vestibular tone imbalance, all patients performed the bucket test. ²⁹ Patients sat upright and looked into a translucent plastic bucket; their visual field was covered completely by the rim of the bucket. On the bottom, inside the bucket, there is a dark, straight, diametric line. On the bottom outside there is a perpendicular line originating from the center point of a quadrant divided into degrees with the zero line adjusted to the dark line inside. For measurement, the bucket was randomly rotated right or left by the examiner (to exclude haptic clues) to various end positions and then slowly rotated back to the zero degrees position. Patients were asked to signal when they estimated the inside bottom line to be truly vertical by saying “stop.” Degrees were read off on the outside scale by the examiner. The procedure was repeated 10 times (5 clockwise and 5 counterclockwise rotations), and the mean of the 10 scores was used as SVV score.

Analysis of Awareness of Postural Abnormalities

The analysis of awareness was performed according to a methodology already used in previous studies. ⁷ , ³⁰ For the evaluation of awareness of lateral trunk bending in patients with PS, they were asked to maintain a standing position in front of a wall goniometer with eyes closed and say whether they felt they were leaning to one side. If so, keeping eyes closed, the trunk was passively moved from the initial position to 0° on the coronal plane in a time‐frame of about 5 seconds. During the maneuver, the patients were asked to report the exact moment when they perceived to have reached the vertical position. In that moment, the degrees of the lateral trunk bending were recorded by the evaluator based on the difference between the virtual position of C7 spinous process on the wall goniometer and the vertical (0°). This maneuver was repeated 3 times, and the mean value and standard deviation of the 3 scores obtained were calculated.

For the evaluation of awareness of anterior trunk bending in patients with CC, they were asked to maintain a standing position in front of a wall goniometer with eyes closed and say whether they felt they were leaning forward. If so, keeping eyes closed, the trunk of the patients was passively moved at a speed of about 5 seconds from the initial position to 0° on the lateral plane. During the maneuver, the patients were asked to report when they perceived to have reached the vertical position. In that moment, the degrees of the anterior trunk bending were recorded by the evaluator based on the difference between the virtual position of C7 spinous process on the wall goniometer and the vertical (0°). This maneuver was repeated 3 times, and mean value and standard deviation of the 3 scores obtained were calculated.

Finally, the reduced awareness was calculated as the degree of mismatch between the vertical and the subjective perception of having reached the vertical position with eyes closed. A mismatch ≥5° was considered clinically meaningful. ²⁸ , ³¹

Statistical Analysis

Sample Size Calculation

According to our preliminary study on PS, ⁷ we performed a scenario analysis for the sample size calculation based on possible differences on 5 cognitive domains. We adopted an adjustment procedure for test multiplicity based on Holm's correction to ensure a familywise error rate associated with the null hypothesis of no difference in cognitive ability in the 2 groups at a level of 0.05. Differences between groups were tested with a Wilcoxon 2‐tailed test for independent data. According to the analysis, we estimated that 120 patients with full neuropsychological data, 60 patients with PD with PS or CC, and 60 control patients with PD was sufficient to address the primary endpoint with 80% power at 5% level of significance with a possible exclusion rate of 15% due to missing data. To avoid bias, we decided to exclude from the analyses those patients for whom data on sex, age, PD duration, PD motor phenotype, Hoehn and Yahr stage, posture, or neuropsychological tests received from all involved centers were missing or incomplete.

Data Analysis

Data collected from all involved centers were centralized and analyzed by the University of Torino center. Patients who proved to fulfill the criteria for both CC and PS (n = 7) were excluded from the analyses.

Demographic and clinical features were summarized as mean ± standard deviation or percentages, as appropriate. The nonparametric Kruskal–Wallis test or Fisher exact test was used to compare the demographic and clinical features of the different groups: patients with PD and PS (PS+) versus matched patients with PD without postural abnormalities (PS−) and patients with PD and CC (CC+) versus matched patients with PD without postural abnormalities (CC−).

The Kruskal–Wallis test (and post hoc analysis in case of significant differences) was used to compare the main demographic and clinical features of patients and confirm the good matching between cases and controls. The Z score of each cognitive domain between PS+ versus PS− and CC+ versus CC− as well as the results of the bucket test were compared between groups by means of the Mann–Whitney test. Given the clear hypothesis behind the study design and the presence of previous preliminary studies for which this study has the aim of being confirmatory, we did not perform multiple comparison adjustment tests, which would reduce type I error while increasing type II error. Moreover, the sample size calculation was based on the comparison of the 5 cognitive domains to obtain an adequate sample size considering 5 levels of comparison.

All P values reported are 2‐tailed, and a P < 0.05 is considered statistically significant. Data were analyzed using the Statistical Package for the Social Sciences (SPSS version 22 for Mac [IBM Corp., Armonk, NY]) and R software version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria). Data collected and used for the study are available upon reasonable request.

All patients involved in the project provided written informed consent for participation, and the local ethic committee of “A.O.U. Città della Salute e della Scienza di Torino ‐ A.O. Ordine Mauriziano ‐ A.S.L. Città di Torino” approved the study (protocol number: 0041780/2019).

Results

We enrolled a total of 114 patients with PD (32 with PS, 25 with CC, and 57 PD controls without postural abnormalities [32 matched with PS+ and 25 matched with CC+]) from 8 movement disorders centers between June 2019 and June 2021. Demographic and clinical features were reported in Table 1 and did not show any significant difference between groups, with the exception of a higher degree of back pain in PS+ and CC+ versus both PS− and CC− (P < 0.001) and a worse quality of life, as per higher PDQ‐8 scores, in PS+ and PS− in comparison with CC+ and their controls (P = 0.014).

TABLE 1.

Demographic and clinical data of the entire sample, cases, and controls

Demographic/clinical data	Entire Sample	PS+	PS−	CC+	CC−	P Value
Sex, male/female	76/38	20/12	20/12	18/7	18/7	0.703
Age, years	73.4 ± 6.2 (57–85)	73.8 ± 6.5 (57–85)	73.3 ± 6.1 (59–82)	73.3 ± 6.1 (62–83)	73.3 ± 6.6 (62–85)	0.974
Age at PD onset, years	65 ± 7.1 (46–79)	65.4 ± 6.9 (46–79)	65.5 ± 6.4 (51–77)	64.6 ± 7.8 (50–76)	64.4 ± 7.6 (46–76)	0.971
PD duration, years	8.5 ± 4 (1–18)	8.3 ± 3.9 (1–18)	8.2 ± 4.4 (2–17)	8.6 ± 4.3 (3–17)	9 ± 3.8 (2–17)	0.859
PD phenotype, tremor/akinetic‐rigid/mixed	85/18/11	23/4/5	21/6/5	21/3/1	20/4/1	0.308
Education, years	10.9 ± 4.5 (3–22)	10.1 ± 4.2 (3–18)	11 ± 4.5 (3–22)	11.2 ± 4.6 (5–18)	11.4 ± 5 (3–22)	0.682
MoCA, raw score	23.23 ± 3.1 (14–28)	22.9 ± 3 (16–27)	23.1 ± 3.1 (16–28)	22.9 ± 3.3 (14–28)	24.1 ± 2.9 (16–28)	0.839
MDS‐UPDRS 1	11.8 ± 5.5 (0–27)	12.2 ± 5.1 (5–25)	12 ± 5.9 (0–27)	12.8 ± 5.5 (3–26)	9.7 ± 5.2 (0–18)	0.288
MDS‐UPDRS 2	14.2 ± 6.7 (2–33)	14.9 ± 7 (2–33)	13.3 ± 6.7 (3–33)	14.6 ± 5.6 (3–26)	13.1 ± 5.8 (5–25)	0.346
MDS‐UPDRS 3	33.9 ± 14.3 (10–69)	36.5 ± 14.9 (11–64)	30.5 ± 13.6 (10–62)	38.8 ± 14.8 (16–69)	29.7 ± 12.2 (11–55)	0.060
MDS‐UPDRS 4	3.1 ± 3.8 (0–14)	2.2 ± 2.6 (0–7)	3 ± 4.2 (0–14)	4 ± 4.2 (0–14)	3.6 ± 4.2 (0–13)	0.585
Hoehn and Yahr stage	2.9 ± 2.7 (2–4)	3.8 ± 5 (2–4)	2.5 ± 0.6 (2–4)	2.7 ± 0.6 (2–4)	2.5 ± 0.6 (2–4)	0.276
Number of falls	0.6 ± 1.4 (0–12)	0.7 ± 1.1 (0–4)	0.5 ± 1 (0–3)	0.5 ± 0.9 (0–3)	0.7 ± 2.4 (0–12)	0.615
VAS back	4 ± 3 (0–10)	5.6 ± 2.6 (0–10)	2.6 ± 2.7 (0–8)	5.4 ± 2.9 (0–10)	2.4 ± 2.5 (0–8)	<0.001
VAS neck	2 ± 2.7 (0–10)	2.9 ± 3.2 (0–8)	1.7 ± 2.3 (0–8)	2.4 ± 3 (0–10)	0.9 ± 1.4 (0–5)	0.137
PDQ‐8	19.8 ± 13.6 (0–59.4)	25.4 ± 14.1 (6–59.4)	21.3 ± 13.2 (0–46.9)	15 ± 11.6 (4–53.1)	16 ± 13.2 (0–43.8)	0.014
CCI	1.9 ± 1.8 (0–11)	1.7 ± 1.8 (0–6)	2.3 ± 2.4 (0–11)	1.9 ± 1.2 (0–4)	1.8 ± 1.6 (0–6)	0.714
LEDD, mg	739.6 ± 311.3 (225–1631)	730.1 ± 308.1 (250–1605)	665.2 ± 269.1 (250–1386)	798.7 ± 392.8 (225–1631)	788 ± 260.6 (400–1365)	0.425
LEDD levodopa, mg	601.8 ± 290.4 (0–1600)	594.4 ± 315 (150–1600)	559.4 ± 264.1 (150–1206)	638 ± 338 (0–1431)	628 ± 238.1 (300–1050)	0.737
LEDD DA, mg	94.8 ± 110.1 (0–500)	110.8 ± 116.6 (0–500)	78.7 ± 116.4 (0–390)	82.4 ± 96.5 (0–310)	106.5 ± 109.9 (0–320)	0.438
Trunk flexion angle, °	NA	17.6 ± 9.2 (10–40)	NA	44.6 ± 10.8 (30–67)	NA	NA

Note: All data are reported as mean ± standard deviation (range), with the exception of sex and phenotype. Bold indicates statistically significant difference.

Abbreviations: CC−, patients with Parkinson's disease and no postural abnormalities matched with CC+ cases; CC+, patients with Parkinson's disease and camptocormia; CCI, Charlson Comorbidity Index; DA, dopamine agonists; LEDD, levodopa equivalent daily dosage; MDS‐UPDRS, Movement Disorders Society–Unified Parkinson's Disease Rating Scale; MoCA, Montreal Cognitive Assessment; NA, not applicable; PD, Parkinson's disease; PDQ‐8, Parkinson's Disease Questionnaire–8; PS−, patients with Parkinson's disease and no postural abnormalities matched with PS+ cases; PS+, patients with Parkinson's disease and Pisa syndrome; VAS, Visual Analog Scale.

Differences in Neuropsychological Functions

PS+ Versus PS−

The global cognitive assessment did not show a significant difference between PS+ and PS− (MoCA raw score 23.1 ± 2.9 for PS+ and 23.3 ± 3 for PS−; P = 0.787).

Regarding cognitive domains, a significantly worse performance was found in the visuospatial abilities of PS+ (Z scores −1 ± 1.1 for PS+ and −0.5 ± 0.9 for PS−; P = 0.025). No differences were found for memory (Z scores 0.4 ± 1.2 for PS+ and −0.2 ± 0.9 for PS−; P = 0.344), attention and working memory (Z scores 0 ± 1 for PS+ and 0.3 ± 0.7 for PS−; P = 0.295), executive functions (Z scores −0.2 ± 1.3 for PS+ and 0 ± 0.8 for PS−; P = 0.995), and language (Z scores 0.5 ± 1.2 for PS+ and 0.5 ± 1.3 for PS−; P = 0.984) (Fig. 1).

FIG. 1.

Comparison of the 5 cognitive domains scores between patients with Pisa syndrome and matched patients without postural abnormalities. PS+, patients with Parkinson's disease and Pisa syndrome; PS−, patients with Parkinson's disease and no postural abnormalities matched with PS+ cases; Columns represent the mean values of the cognitive index, and bars indicate the standard deviation. *Statistically significant difference.

Comparing Z scores for each test, no differences were found between PS+ and PS− with the exception of a trend toward a low performance of PS+ in the BJLO (P = 0.071), a test pertaining to visuospatial abilities, and in the DCT (P = 0.056), a test pertaining to the attention domain. The raw scores of each test confirmed these trends and showed that PS+ performed worse than PS− on the BJLO (P = 0.032), DCT (P = 0.056), and TMT‐A (P = 0.048), another attention domain test (Table 2).

TABLE 2.

Neuropsychological test scores

Neuropsychological tests	PS+	PS−	P Value	CC+	CC−	P Value
MoCA score	23.1 ± 2.9 (16–27)	23.3 ± 3 (16–28)	0.787	23.3 ± 2.9 (14–28)	24.4 ± 3.2 (16–30)	0.160
BWR, raw score	4 ± 0.9 (1–5)	4 ± 0.7 (2–5)	0.526	4 ± 0.7 (3–5)	3.9 ± 0.9 (2–6)	0.632
BWR, Z score	0.1 ± 1.2 (−4 to 1.7)	0 ± 0.9 (−2.7 to 1.7)	0.525	−0 ± 1 (−2.1 to 1.7)	−0.1 ± 1.1 (−2.4 to 2.2)	0.548
AVL‐immediate recall, raw score	31 ± 11 (14–57)	31.1 ± 7.6 (16–46)	0.752	32 ± 10 (14–57)	31.3 ± 10.4 (12–65)	0.907
AVL‐immediate recall, Z score	−0.5 ± 1.5 (−2.9 to 3.1)	−0.4 ± 1.1 (−3 to 1.7)	0.464	−0.1 ± 1.2 (−1.9 to 2)	−0.4 ± 1.3 (−4 to 2.3)	0.449
AVL‐delayed recall, raw score	4.8 ± 4 (0–15)	5.6 ± 2.8 (1–12)	0.108	5.7 ± 3.4 (0–13)	4.7 ± 2.7 (0–12)	0.324
AVL‐delayed recall, Z score	−0.7 ± 1.9 (−4 to 4)	−0.3 ± 1.2 (−3.1 to 2.7)	0.224	−0.4 ± 1.5 (−4 to 1.5)	−0.8 ± 1 (−3.6 to 1.1)	0.290
DCT, raw score	43.3 ± 9.2 (15–56)	47.8 ± 6.8 (32–58)	0.056	43.6 ± 9.4 (23–60)	46.8 ± 10.1 (22–59)	0.189
DCT, Z score	−0.2 ± 1.1 (−3.9 to 1.5)	0.3 ± 0.9 (−1.8 to 1.7)	0.056	−0.2 ± 1.2 (−3.7 to 1.4)	0.3 ± 1 (−1.7 to 2.3)	0.193
TMT‐A, raw score	81.5 ± 38.8 (31–176)	63.3 ± 27.1 (28–141)	0.048	75.2 ± 35.2 (24–150)	60.5 ± 27.5 (28–125)	0.138
TMT‐A, Z score	0.2 ± 1.6 (−3.9 to 2.9)	0.5 ± 1.2 (−2.4 to 2.4)	0.722	0.2 ± 1.4 (−2.8 to 2.6)	0.7 ± 1.1 (−2 to 2.3)	0.118
DSB, raw score	4 ± 1.3 (0–6)	4.1 ± 0.9 (3–6)	0.900	4.1 ± 1.2 (3–8)	4.4 ± 1.4 (0–7)	0.146
DSB, Z score	0.1 ± 1.3 (−2.4 to 3.6)	0 ± 0.9 (−1.3 to 2.1)	0.745	0.1 ± 1.3 (−1.3 to 4)	0.2 ± 1.5 (−4 to 3.6)	0.716
TMT‐B, raw score	202.8 ± 96 (87–445)	163.6 ± 73 (68–361)	0.130	209.8 ± 119 (70–466)	164.1 ± 125 (57–600)	0.075
TMT‐B, Z score	0 ± 1.3 (−4 to 1.3)	0.4 ± 0.9 (−2 to 1.3)	0.299	0 ± 1.4 (−4 to 1.3)	0.5 ± 1.2 (−4 to 1.3)	0.037
TMT B‐A, raw score	132.3 ± 78.5 (29–352)	104 ± 64.1 (−8 to 245)	0.157	140 ± 98.1 (29–373)	107.7 ± 105.9 (20–494)	0.156
TMT B‐A, Z score	0.1 ± 1.3 (−4 to 1.2)	0.3 ± 0.9 (−1.85 to 1.44)	0.684	0.1 ± 1.3 (−4 to 1.2)	0.3 ± 1.2 (−4 to 1.3)	0.242
FAB, raw score	13.8 ± 3.7 (4–18)	14.6 ± 2.4 (8–18)	0.723	14.1 ± 3.4 (7–18)	15 ± 2.2 (9–18)	0.590
FAB, Z score	−0.7 ± 1.7 (−4 to 2.1)	−0.7 ± 1.5 (−4 to 1.2)	0.864	−0.7 ± 2.1 (−4 to 3.9)	−0.5 ± 1.4 (−3.9 to 2.1)	0.810
CA, raw score	11.3 ± 2.5 (5–14)	12.2 ± 2 (7–14)	0.174	12.1 ± 2.3 (6–14)	12.7 ± 1.4 (9–14)	0.533
CA, Z score	−0.2 ± 1.3 (−4 to 1.2)	0.1 ± 1.2 (−3.1 to 1.2)	0.356	0 ± 1.3 (−3.5 to 1.2)	0.4 ± 0.5 (−0.8 to 1.2)	0.417
BJLO, raw score	15.3 ± 9.3 (0–30)	20 ± 7 (0–29)	0.032	18 ± 9.2 (0–28)	19.5 ± 7 (0–29)	0.806
BJLO, Z score	−2 ± 1.7 (−4 to 1.1)	−1.2 ± 1.3 (−4 to 0.8)	0.071	−1.5 ± 1.8 (−4 to 2)	−1.5 ± 1.6 (−4 to 0.9)	0.991
PVF, raw score	29.3 ± 11.2 (2–48)	28.7 ± 12.2 (8–55)	0.554	28.2 ± 10.8 (10–53)	31.4 ± 9 (15–51)	0.225
PVF, Z score	0.6 ± 1.2 (−2 to 4)	0.1 ± 2.1 (−4 to 3.7)	0.437	0.3 ± 1.1 (−1.4 to 2.5)	0.7 ± 2 (−4 to 4)	0.200
CVF, raw score	18 ± 8.7 (6–50)	20.9 ± 9 (9.5–58)	0.101	22.2 ± 10.3 (8–51)	19.8 ± 8.4 (11.5–50)	0.312
CVF, Z score	0.6 ± 1.9 (−2.3 to 4)	1 ± 1.4 (−1.1–4)	0.153	0.3 ± 1.1 (−1.2 to 4)	0.8 ± 1.4 (−1.1 to 4)	0.285

Note: Data are reported as mean ± standard deviation (range). Bold indicates statistically significant difference.

Abbreviations: AVL, Rey Auditory Verbal Learning Test; BJLO, Benton Judgment of Line Orientation; BWR, Bi‐Syllabic Word Repetition Test; CA, Constructional Apraxia Test; CC−, patients with Parkinson's disease and no postural abnormalities matched with CC+ cases; CC+, patients with Parkinson's disease and camptocormia; CVF, Category Verbal Fluency; DCT, Digit Cancellation Test; DSB, Digit Span Backward; FAB, Frontal Assessment Battery; MoCA, Montreal Cognitive Assessment; PS−, patients with Parkinson's disease and no postural abnormalities matched with PS+ cases; PS+, patients with Parkinson's disease and Pisa syndrome; PVF, Phonemic Verbal Fluency; TMT B‐A, Trail Making Test B‐A; TMT‐A, Trail Making Test A; TMT‐B, Trail Making Test B.

CC+ Versus CC−

The global cognitive assessment did not show a significant difference between CC+ and CC− (MoCA raw scores 23.3 ± 2.9 for CC+ and 24.4 ± 3.2 for CC−; P = 0.160).

Regarding cognitive domains, there were no differences for visuospatial (Z scores 0.7 ± 1.2 for CC+ and 0.5 ± 0.9 for CC−; P = 0.900), memory (Z scores −0.2 ± 0.9 for CC+ and −0.4 ± 0.7 for CC−; P = 0.244), attention and working memory (Z scores 0 ± 1 for CC+ and 0.4 ± 0.9 for CC−; P = 0.187), executive functions (Z scores −0.2 ± 1.3 for CC+ and 0.1 ± 1 for CC−; P = 0.410), and language (Z scores 0.8 ± 1.1 for CC+ and 0.8 ± 1.4 for CC−; P = 0.691) domains (Fig. 2).

FIG. 2.

Comparison of the 5 cognitive domains scores between patients with camptocormia and matched patients without postural abnormalities. CC+, patients with Parkinson's disease and camptocormia; CC−, patients with Parkinson's disease and no postural abnormalities matched with CC+ cases. Columns represent the mean values of the cognitive index, and bars indicate the standard deviation.

Comparing Z scores for each test, we found worse performance of CC+ in the TMT‐B (P = 0.037), a test pertaining to the frontal executive functions, with the trend confirmed by the comparison of the raw scores of the test between CC+ and CC− (P = 0.075). No other differences emerged between the 2 groups (Table 2).

Perception of the SVV

The bucket test showed that patients with PS+ had a significantly higher SVV (ie, misperception of body verticality) than patients with PS− (degrees of deviation from the vertical: 5.7 ± 14.2 for PS+ and 1.9 ± 2.2 for PS−; P = 0.038), whereas patients with CC+ did not show a significant difference from patients with CC− (3.5 ± 4.2 for CC+ vs. 2.4 ± 5.1 for CC−; P = 0.133).

Awareness of the Trunk Posture Abnormality

A total of 21 patients with PS+ and 16 patients with CC+ correctly completed this task and were included in the analysis. Among patients with PS+, 66.7% (n = 14/21) had a significant misperception of their trunk misalignment (ie, >5°), with a mean group misperception of 10.2° ± 2.2. Among patients with CC+, 62.5% (n = 10/16) had a significant misperception, with a mean group misperception of 12.3° ± 2.7.

Discussion

In this multicenter, case‐control study, we demonstrated that patients with PD with PS have worse visuospatial performances when compared with matched patients with PD without postural abnormalities; conversely, patients with PD with CC showed no significant cognitive difference when compared with their PD controls. Also, we found a difference in the SVV perception, a sensitive sign of peripheral or central subcortical vestibular tone imbalance, between PS+ and their controls, but not between CC+ and their controls. Finally, a reduced awareness of the postural abnormality was observed in >60% of patients with both PS or CC.

Starting from preliminary evidence, we performed for the first time a case‐control study with adequate sample size to verify the hypothesis of selective cognitive deficits in patients with PD with PS and CC. Indeed, 2 single‐center studies with limited sample sizes suggested that specific alterations in visuospatial and attention domains may be present in patients with PS. ⁵ , ⁷ Another study evaluated the performance in the executive functions of 59 patients with PD (Hoehn and Yahr stage 2–3.5) divided according to the presence of a postural abnormality as per the posture item of the UPDRS part III; the authors observed worse executive performances in patients with postural abnormalities after controlling for age. ⁴ Finally, a study on 23 Chinese patients with PD with CC found that compared with the entire group of 263 patients with PD, the presence of CC was associated with the following: male sex, a higher UPDRS part III score, a higher sexual dysfunction score, and a lower orientation score on the MoCA subscore. ⁶

Our study partly confirmed these preliminary findings, indicating a clue for the pathophysiology of PS and its risk factors. Moreover, we found that the involvement of visuospatial domain functioning is specifically associated with PS and not with CC despite a similar level of unawareness of the trunk flexion in both postural abnormalities. Interestingly, the alteration of vestibular information already reported both in patients with PS ³² and CC ³⁰ was confirmed in our case‐control study only for patients with PS, who proved to have a significantly more altered SVV than patients without postural abnormalities (5.7° vs. 1.9°). Although a worse performance on the bucket test might suggest an alteration of high‐level vestibular information processing, which has been recently reported as associated with cholinergic deficits in PD, ³³ this test cannot differentiate between peripheral and central vestibular deficits. In a recent study investigating factors contributing to the severity of PS, the authors found that both the severity of SVV alteration and unilateral ear canal paresis can be correlated with the side and severity of PS, ³² suggesting a role for peripheral and central vestibular information processing in the pathogenesis of the lateral trunk flexion. However, only a longitudinal prospective study can definitely clarify the role of the vestibular system as a risk factor for the onset of PS.

When we explored the differences in single cognitive tests between PS+ and controls, we found that specifically BJLO, a widely used measure of visuospatial judgment measuring the accuracy of angular orientation, was impaired in patients with PS. Also, DCT and TMT‐A showed borderline results, possibly indicating that impaired attention could also be associated with PS pathophysiology. On the contrary, patients with CC performed similarly to their controls on the BJLO, DCT, and TMT‐A and also on the global assessment of the 5 cognitive domains. The only difference disclosed between CC+ and their controls was a worse performance in TMT‐B, a test pertaining to the frontal executive functions. Noteworthy, the Trail Making Test, divided into parts A and B, can be globally considered a test of visual attention and task switching during a visuo‐motor task. Although part A, showing worse performance in patients with PS, requires adequate visual processing skills, recognition of numbers, knowledge and reproduction of numerical sequences, and motor speed, part B specifically requests cognitive flexibility and shifting ability.

Other interesting findings emerged from our study. Back pain was significantly higher in patients with both PS and CC than controls, as expected. ² , ³ Similarly, a higher degree of motor impairment (albeit not reaching the significant threshold) was found in patients with PS and patients with CC. The finding was already reported in previous studies ² and can be referred to the higher posture score obtained at the MDS‐UPDRS part III evaluation and to the higher axial motor impairment associated with a severe posture abnormality. ² , ³⁴ On the contrary, only patients with PS showed a worse quality of life, obtaining significantly higher scores on the PDQ‐8. Worse quality of life in patients with PD with postural abnormalities was reported in large cross‐sectional studies ² ; however, direct comparisons in case‐control studies, separating patients with PS from those with CC, is a new finding, suggesting a differential impact on patients and caregivers according to the type of severe postural abnormality.

The robustness of our findings is provided by the use of an adequate sample size to test our central hypothesis, by a comprehensive neurological and neuropsychological evaluation, and by strict case‐control matching. The main limitation of the study is due to its cross‐sectional nature, which does not allow to ensure the direction of the observed associations. Another caveat is related to the exclusion of patients with severe dementia. A high load of axial symptoms is typically associated with more severe disease, including the presence of dementia and other disability milestones. ³⁵ Our cohort included patients with a mean of 3 years of severe postural alterations and the absence of severe dementia; a clinical and neuropsychological follow‐up would add significant insights on the possible different progression of both motor and cognitive symptoms between patients with PS or CC and their controls.

These limitations notwithstanding, our study showed that (1) prominent alterations in the visuospatial cognitive domain are associated with PS, (2) there are no significant cognitive differences between patients with and without CC, (3) most patients with both PS and CC may be at least partially unaware of their postural alteration, and (4) only patients with PS have a significant misperception of body verticality.

Although adding a meaningful contribution to the investigation of the pathophysiology of such disabling axial symptoms, we believe that these findings can serve to prompt adequate therapeutic and prevention strategies, encompassing multidisciplinary and personalized rehabilitation programs focused on proprioception and perception, and can be associated with tailored cognitive rehabilitation.

Author Roles

(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript: A. Writing of the First Draft, B. Review and Critique.

C.A.A.: 1A, 1B, 1C; 2A, 2C, 3A.

E.M.: 1A, 1C, 2C, 3B.

R.E.: 1A, 1C, 2C, 3B.

N.M.: 1A, 1C, 2C, 3B.

C.G.: 1C, 3B.

A. Pilotto: 1C, 3B.

L.M.: 1C, 3B.

F.S.: 1C, 3B.

A.M.: 1C, 3B.

L.S.: 1C, 3B.

S.C.: 1C, 3B.

M.S.: 1C, 3B.

M.R.: 1C, 3B.

B.C.: 1C, 3B.

P. Berchialla: 2A, 2B, 3B.

B.M.: 1C, 3B.

B.V.: 1C, 3B.

M.Z.: 1C, 3B.

A.M.R.: 1B, 3B.

P. Barone: 1B, 3B.

C.C.: 1B, 3B.

A. Padovani: 1B, 3B.

M.T.: 1B, 3B.

L.L.: 1A, 1B, 3B.

Disclosures

Ethical Compliance Statement: The Ethical Committee of the “A.O.U. Città della Salute e della Scienza” of Turin approved the study (protocol number: 0041780/2019). Patients provided written informed consent to participate in the study. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflicts of Interest: No specific funding was received for this work. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months: Carlo Alberto Artusi has received honoraria from Ralpharma and Zambon for scientific support and speaker honoraria from BIAL. Roberto Erro receives royalties from publication of Case Studies in Movement Disorders—Common and Uncommon Presentations (Cambridge University Press, 2017) and Paroxysmal Movement Disorders (Springer, 2020). He has received consultancies from Sanofi and honoraria for speaking from the International Parkinson's Disease and Movement Disorders Society. Nils Margraf has received travel grants from Angelini Pharma and Eisai Pharma; advisory board grants from GW Pharma and Angelini Pharma; and fees for lectures from Angelini Pharma; support for projects from UCB Pharma, Desitin Pharma, LivaNova, Angelini Pharma, Eisai Pharma, and GW Pharma. Andrea Pilotto received speaker honoraria and travel grants from AbbVie Pharmaceuticals, BioMarin Pharmaceutical, Chiesi Pharmaceuticals, Nutricia Pharmaceuticals, Roche Pharmaceuticals, UCB Pharma, and Zambon Pharmaceuticals and is supported by competitive grants of the Italian Ministry of Health (PRIN (Progetti di Ricerca di Interesse Nazionale) 2019), Airalzh, LIMPE (Lega Italiana per la lotta contro la malattia di Parkinson e le sindromi extrapiramidali e le demenze) Foundation, and H2020‐IMI (Innovative Medicine Initiative) initiative (IMI2‐2018‐15‐06). Francesca Spagnolo has received honoraria for lecturing and travel grants from Sanofi, Zambon, and Medtronic. Alberto Marchet has received honoraria for travel grants from Zambon and lecturing from BIAL. Maurizio Zibetti has received honoraria for lecturing and travel grants from Medtronic, BIAL Pharma, and AbbVie. Augusto Maria Rini has received honoraria for lecturing and travel grants from Sanofi, Medtronic, Bristol, and Biogen. Paolo Barone received consultancies as a member of the advisory boards for Zambon, Lundbeck, UCB, Chiesi, AbbVie, and Acorda. Cristoforo Comi has received travel grants from Zambon and BIAL. Alessandro Padovani is a consultant and served on the scientific advisory boards of GE Healthcare, Eli‐Lilly, and Actelion Ltd Pharmaceuticals; received speaker honoraria from Nutricia, PIAM, Lansgstone Technology, GE Healthcare, Lilly, UCB Pharma, and Chiesi Pharmaceuticals; and is supported by competitive grants of the Italian Ministry of Health (PRIN 2020) and Cariplo and H2020‐IMI initiative (IMI2‐2018‐15‐06). Leonardo Lopiano has received honoraria for lecturing and travel grants from Medtronic, UCB Pharma, and AbbVie. Elisa Montanaro, Christian Geroin, Luca Magistrelli, Lidia Sarro, Sofia Cuoco, Marta Sacchetti, Marianna Riello, Barbara Capellero, Paola Berchialla, Bettina Moeller, Beeke Vullriede, and Michele Tinazzi have no financial disclosures to report.