Tilted writing after stroke, a sign of biased verticality representation

Dear Editor. Handwriting is a complex task involving motor, linguistic, perceptual, and attentional skills predominantly controlled by the left hemisphere but requiring a spatial organization that depends on the right hemisphere . Handwriting is often affected after right-hemispheric lesions, with a spectrum of signs related to the spatial layout of the written language , constituting spatial dysgraphia/agraphia . These signs are multi-faceted . Most deal with spatial compression or mental rotation of the space and are related to spatial neglect: omitting the left half of the paper, overwriting or compressing some words, and omitting and substituting letters or graphemes. Some deal with the visual processing of complex material to write (simultagnosia) . Others such as tilted writing remain to be understood.

Here we present a case suggesting that the tilted handwriting after stroke might be due to a counterclockwise tilt in representing the vertical, transposed on the sheet of paper referring to top and bottom.

JW was a 75-year-old right-handed male who had a right fronto-parietal hematoma ( Fig. 1 A and B) causing left hemiplegia with hemianesthesia, left hemianopsia and spatial neglect (US National Institutes of Health Stroke Scale score 13). At entry in rehabilitation (day 13 after the stroke), JW showed spatial neglect and pusher syndrome interpreted as a result of an extreme bias in the internal model of verticality . He behaved as if he implicitly aligned his body posture onto a representation of the vertical, tilted counterclockwise. His spontaneous erect posture in standing was tilted about 12° leftward, before falling ( Fig. 1 C). JW also presented a consistent and global counterclockwise tilt of his written production ( Fig. 1 E–G): drawing (−12°) and writing (left margin −9° with respect to the vertical, lines −11° with respect to the horizontal, most ascending or descending characters −10° with respect to the vertical). He wrote without any space compression, deletions or omissions, or simultagnosia.

JW agreed to a further clinical investigation of these troubles to guide their rehabilitation and also signed informed consent to be enrolled in the cohort DOBRAS (ClinicalTrials.gov: NCT03203109 ). Several domains of spatial cognition were assessed at 2, 3 and 9 months post-stroke (M2, M3 and M9): spatial neglect by means of a battery of tests including the representation of the subjective straight ahead (SSA) , representation of the vertical [visual vertical (VV) and postural vertical (PV)] , and drawing and writing on blank and cued (lined) paper as well as after a modulation of the verticality bias. Assessments involved using validated tests, devices, and protocols, all routinely used (details in additional material), except for a novel procedure to test handwriting. The SSA (10 trials) was compared to the actual straight ahead (0°), whereas deviations affecting the VV and PV (10 trials each) were compared to published normal values .

To interpret JW’s drawing and writing, we tested 12 right-handed healthy individuals (5 males), matched in age [mean (SD) 75.2 (2.8) years] and sociocultural level. All participants were tested in accordance with the Helsinki Declaration, with the approval of the local ethics committee (Comité d’Ethique pour les Recherches de l’Université Grenoble Alpes; 2019-04-09-1) after giving their written consent.

Statistical analysis involved using SPSS v23 (IBM Corp., Armonk, NY, USA). The existence of a systematic deviation (tilt) by comparison to the reference was tested by one-sample t tests. Negative values indicated a counterclockwise tilt (leftward, upward). Amplitudes and the significance of JW’s tilts were analyzed with Z-scores (VV, PV) or T-scores (drawing, writing) calculated with control data. Conditions were compared by Wilcoxon test for paired samples or with Friedman’s ANOVA. Significance was set according to Bonferroni corrections (bilateral tests). Data are presented as mean (SD).

JW’s spatial neglect was severe at M2, affecting body and non-body spaces ( Table 1 ), without any sign of spatial alexia. Only a few neglect signs persisted at M3 but no longer at M9, when the representation of the SSA, not tested earlier, was also normal (−0.7° [2.8], P = 0.739).

Table 1

Spatial neglect in JW after right hemispheric stroke at 2, 3 and 9 months (M2, M3, M9) post-stroke.

	M2	M3	M9
Body neglect scores
Bisiach test (0 to 3, cut)	1*	3*	0
Comb test (% neglect index, cut)	13*	0	9
Razor test (% neglect index, cut)	21*	0	0
Fluff	3*	2	0
Non-body neglect scores
Bells cancellation test (total omissions, cut-off ≥ 6)	1	1	1
Line bisection (ipsilesional deviation in mm, cut-off ≥ 7 mm)	20*	5.5	−2
Landscape copying (omissions 0 to 5, cut-off > 0)	0	0	0
Text reading (omissions, cut-off > 0)	0	0	0
Overlapping figures test (omissions, cut-off > 0)	0	0	0
Neglect in daily life score
Catherine Bergego Scale	6.3	4.4	0