Natural History and Relationship to Motor and Functional Recovery
Background and Purpose— Patients with hemiparetic stroke have impaired balance control. Some patients (“pushers”) are resistant to accepting weight on and actively “push” away from the nonparetic side. This research identified pushers from stroke patients with moderate to severe hemiparesis and examined longitudinal changes in symptoms, level of impairment, and functional independence.
Methods— Prospective sample of hemiparetic stroke patients (n=65) located in Toronto, Canada. Detailed clinical assessments were performed within 10 days postonset, at 6 weeks, and at 3 months.
Results— At 1 week after stroke, 63% of patients demonstrated features of pushing. In 62% of pushers, symptoms resolved by 6 weeks, whereas in 21%, pushing symptoms persisted at 3 months. Motor recovery and functional abilities at 3 months were significantly lower among the pushers compared with the nonpushers. Pushers also had a significantly longer hospital length of stay (89 days versus 57 days). It is noteworthy that motor and functional recovery improved significantly over the 3-month study period for both pushers and nonpushers.
Conclusions— Identification of stroke patients with pushing symptoms has prognostic implications for recovery. In light of this potential recovery, rehabilitation specialists need to refine treatment approaches for the pushers to further improve functional outcome.
Balance and postural control problems after stroke are common, because balance requires the integration of intact motor and sensory processes1 and 88% of all patients experience hemiparesis and 53% have sensory deficits.2,3 Steadiness, symmetry, and dynamic stability (components of balance) are all affected after stroke4 and contribute to overall disability. Control of stability is critical for the performance of activities of daily living5 and is correlated with long-term functional outcome.6 Profound limitations in balance control, such as an inability to sit independently, often preclude patients from being accepted to poststroke rehabilitation programs. A subset of stroke patients with such balance problems are distinguished by resistance to supporting weight on their nonparetic side: a phenomenon historically referred to as “pushing,”7–14 “listing,”15–17 or “asymmetrical trunk posture.”18 Pushing is clinically characterized as a tendency to adopt postures aligned toward the affected side (as opposed to the unaffected side, as is typical in nonpushing stroke patients). In addition, pushers resist attempts by the clinician to move the body toward the midline. Such behavior can be evident in seated and standing positions and predisposes individuals to a high risk of instability and falling.
Despite the frequent reference to such behavior, there have been few studies that have explored the prevalence or natural history. Pedersen et al8 reported a prevalence of 10.4% (34/327). Two other studies highlighted the challenges in reporting prevalence, with estimates ranging from 5% to 50%.10,14 Although it has been suggested that symptoms of pushing are transient,7 there exists only 1 study focused on a small subset of stroke subjects noting symptoms were nearly gone by 6 months after stroke.19
Parallel to the limited information available regarding the recovery of pushing, there is limited understanding of the relationship between the occurrence of pushing and functional and sensorimotor recovery. Karnath et al19 reported significant improvement in the recovery of strength in the hemiparetic leg but not in the arm at 6 months. Taylor et al18 found no significant difference between groups in the scores for motor recovery of the arm, the leg, or the trunk by 6-week follow-up, though they did reveal that pushers have significantly worse overall function scores.
In light of the potential impact of pushing symptoms on treatment strategies, this study prospectively investigated the proportion of pushing and the rate of recovery of pushing during the first 3 months after stroke among patients with a moderate or severe hemiparesis. In addition, it sought to establish the association between symptoms of pushing and other indices of recovery as well as measures of institutional care. The present study tested the following hypotheses: (1) the proportion of symptoms of pushing will be >10% reported by Pedersen et al8; (2) the symptoms of pushing will resolve over the first 3 months as reflected by pushers’ ability to bear weight on the less-affected limb and to no longer resist passive shifting of their weight from affected side to unaffected side in sitting or in standing; and (3) at the 3-month time period, patients with symptoms of pushing will demonstrate a lower rate and degree of recovery from impairment and disability than the nonpushers.
This was a descriptive observational study conducted within the framework of a double-blind randomized placebo-controlled trial described below. The local institutional research ethics boards approved the protocol.
Sixty-five admitted stroke patients who experienced moderate to severe hemiparesis due to a hemispheric stroke were enrolled over a 3-year period (January 2000 to December 2002). Stroke severity was evaluated using the Fugl-Meyer (FM) motor scale (details below). A score of ≤36 on the FM motor scale delineated the severe stroke patients from the moderate group based on similar cutoff points used by Duncan et al.20 The subjects were referred by the rehabilitation therapists at each of the 5 participating acute care hospitals and were screened by either the study neurologist or study nurse. Eligible patients had to be medically able to participate in rehabilitation and were expected to survive at least 3 months. Patients with brain stem or cerebellar strokes, significant premorbid disability, unstable cardiac disease, or history of dementia were excluded. Informed consent was obtained from the subject or the substitute decision maker. Three people were lost to follow-up by 6 weeks after stroke (2 deaths, 1 with medical complications), and 1 person died before 3-month follow-up. Therefore, the 1- and 6-week calculations are based on a sample of 62 people and the 3-month calculations, on a sample of 61. The mean age of the group was 68 (SD 14) years. A summary of the group’s characteristics is provided in Table.
All patients currently enrolled were involved in a pharmacotherapy trial where they were randomly allocated to dextroamphetamine 10 mg or placebo in combination with 1 extra hour of physiotherapy twice per week for 5 weeks. This trial found no benefit in the amphetamine-treated patients versus placebo in terms of lower extremity motor impairment recovery, ambulation, or overall FM motor score. Forty-nine percent of the pushing subgroup and 39% of the nonpushing subgroup were represented in the treatment group of the amphetamine trial. Given these results, it is unlikely that the administration of amphetamine confounded the results of this study.
Description of Physiotherapy Treatment
Each participant received 10 extra 1-hour physiotherapy treatment sessions per the protocol of the amphetamine study. These sessions did not differ significantly from or interfere with the normal course of therapy provided by the therapists in the acute care or rehabilitation facilities. The focus of treatment sessions varied from patient to patient and from session to session according to patient need. The treatment of the pushers differed from the treatment of the nonpushers in that the pushers needed guidance and practice shifting their weight to the unaffected side. All treatment sessions were individualized based on patient abilities and were “hands-on,” goal-oriented, and required the active participation of each patient.
Patients were assessed using a standardized battery focusing on 4 domains: (1) pushing, (2) motor impairment, (3) functional ability, and (4) general neurological deficits. All measures were taken at 3 time points after stroke: 1 week (mean 8±2 days, range 4 to 11), 6 weeks (mean 45±3.6 days, range 37 to 56), and 3 months (mean 90±2 days, range 74 to 108). Assessment of motor impairment using the FM was done twice per week during the 5 weeks of treatment, and in a subgroup of 33 patients, pushing was also assessed during the treatment period. The general neurological assessments were completed by a trained stroke nurse or stroke neurologist. The same research physiotherapist completed all other measures at all time points.
The Scale for Contraversive Pushing (SCP)21 is scored 0 to 6: the higher the score, the greater the severity of pushing. There are a total of 3 domains (posture, extension, and resistance) that are assessed for both sitting and standing positions (total of 6 scored items). Patients were identified as pushers if they scored >0 on any of the 3 domains (ie, SCP>0).
The 100-point motor domain of the FM Assessment was used for the assessment of motor impairment.22
There were 2 measures used that assessed the patient’s ability for general bed mobility, transfers, and ambulation: the Chedoke-McMaster Disability Inventory (CMDI)23 and the Physiotherapy Clinical Outcome Variables (COVS).24 The Functional Independence Measure (FIM) was used to assess the amount of assistance each patient needed for basic activities of daily living.25
General Neurological Findings
The presence or absence of sensory deficits, visual field deficits, neglect, and aphasia was assessed as part of the National Institutes of Health Stroke Scale (NIHSS).26
All measures (with the exception of the SCP) have been shown to be valid and reliable for use in the stroke population. The SCP is the first standardized tool to assess pushing behavior, and there has been no work published on the measurement properties. However, Karnath et al (2000a) used the original clinical characterization of Davies (1985) to develop the instrument.
The primary outcome used to indicate the presence and severity of pushing was the SCP score (SCP>0=pusher). Frequency analysis (χ2 test) was used to test the first hypothesis regarding the proportion of pushing symptoms compared with an expected frequency of 10.4% as reported in the Copenhagen Stroke Study.8 Tests of the change over time were conducted using repeated measures ANOVA, including all individuals who initially presented with symptoms of pushing. Two-way repeated measures ANOVA (Factor 1: time; Factor 2: group) were used to contrast the rate of change in impairment and disability, comparing between pushers and nonpushers. ANOVA was used to compare the number of days spent in hospital by the pushers and nonpushers. In all cases, statistical significance was accepted when the alpha level was ≤0.05.
The majority of patients (63%, 39/62) initially exhibited symptoms of pushing (SCP score>0). This frequency was greater than the 10% reported in previous work8 (χ2(1)=183, P<0.0001). Eighty-seven percent of the 39 pushers had a score of ≥3 on the SCP. Pushers were more likely to have right hemisphere lesions (59% versus 30%, P<0.05), severe hemiparesis (82% versus 35%, P<0.0001), and hemispatial neglect (62% versus 17%, P=0.001). The details of the clinical characteristics of these groups are summarized in the Table.
Pushing, as measured by the SCP, did decrease significantly (F(2,74)=221, P<0.0001) from 1 week to 3 months (Figure 1A). Within-subjects contrasts showed that the decrease in pushing symptoms was significant from 1 week to 6 weeks (F(1,37)=219, P<0.0001) and from 6 weeks to 3 months after stroke (F(1,37)=223, P<0.0001). A more detailed perspective of recovery is revealed from SCP scores for a group of pushers (n=20) whose data were collected 2× per week for 5 weeks (Figure 1B). The average data reveals a progressive change in SCP scores over the 5-week duration. The individual recovery profiles revealed steady improvement over time in all but 3 individuals in whom the SCP score rose transiently before decreasing (see Figure 1B). It is noteworthy that the mean SCP scores at 1 week, 6 weeks, and 3 months for the pushers assessed less frequently (n=19) were slightly greater at each time point but showed the same rate of change over time.
In 62% of the pushing group (24/39), the SCP was equal to 0 by 6 weeks. Of those individuals with SCP>0 at 6 weeks, the SCP scores ranged from 0.5 to 6. At 3 months after stroke, only 21% (8/38) continued to have symptoms with SCP scores ranging from 0.25 to 3.25. The difficulties that the 8 pushers encountered at 3 months were in the standing posture; all but 1 person had recovered control of midline in sitting.
On the basis of the time course of recovery (time when SCP=0), the pushers were grouped into 4 categories: (A) before 24 days, (B) 25 to 45 days, (C) 46 to 90 days, and (D) >3 months (Table). It is noteworthy that the duration of the symptoms of pushing was longer among those with higher initial SCP scores (mean 1-week SCP score for each subgroup gets larger as the length of time increases, eg, group D all scored 6 initially). In addition, individuals having a longer duration of pushing symptoms were more likely to have hemispatial neglect. This is illustrated by the frequency of neglect in Groups A through D: Group A, 25%; Group B, 50%; Group C, 83%; and Group D, 100%. Five of the 8 people in Group D (still pushing at 3 months) also showed neglect of the paretic side at the 3-month follow-up visit.
The pushing group and the nonpushing group both made statistically significant motor and functional recovery from 1 week to 3 months (Figure 2A and 2B). Although the rate of change was comparable between groups, the absolute level was lower among pushers (Group effect: FM, F(2,118)=186, P<0.0001 and CMDI, F(2,118)=197, P<0.0001; Group-by-Time interaction: FM, F(2,118)=2.0, P=0.141 and CMDI, F(2,118)=2.7, P=0.068).
There did not appear to be a within-subject association between recovery of motor control in the arm and leg and symptoms of pushing. For example, many pushers’ SCP scores changed during the 5 weeks of biweekly measurement, whereas their FM motor scores did not. (Figure 3 highlights 2 examples: FM leg scores stay constant, whereas SCP scores improve). Furthermore, at 3 months after stroke, the symptoms of pushing had resolved (SCP=0) for most people, yet there were still considerable variations in the motor recovery scores (range 4 to 96 of a possible 100). Relevant to the impact that symptoms of pushing may have on care, the pushers had significantly longer length of stay (LOS) in acute care hospital (F(4,56)=7.5, P<0.0001) and inpatient rehabilitation facilities (F(4,52)=5.4, P=0.001) compared with the nonpushers. In addition, increased severity of pushing was also associated with greater LOS (Groups C and D versus nonpushers of Groups A and B, P≤0.001; Figure 4).
Three main observations from the present study are (1) the proportion of pushing symptoms is much higher than previously reported; (2) greater detail of the time course of recovery revealed new insight into the natural history; and (3) pushers, although characterized by lower functional outcomes and longer lengths of stay, did demonstrate significant recovery over time.
In stroke patients with moderate to severe hemiparesis, 63% initially demonstrated signs of pushing. This frequency of pushing is much greater than 8% to 10% as reported by Pedersen et al8 (n=327) and Bohannon et al17 (n=172). Differences between the present work and these 2 previous studies are (1) measurement/classification of pushing, (2) stroke severity of patients evaluated, and (3) time of initial assessment. A likely explanation for the difference between studies is the sensitivity of the standardized measurement. In this study, patients were classified as pushers if they had an SCP>0. Pedersen et al8 classified subjects as pushers using criteria set out by Davies,7 Bohannon et al17 identified people as pushers (“listers”) using the descriptions of Bobath,27 so it is possible that patients with mild to moderate severity of pushing symptoms were not included. Certainly, potential differences in prevalence that may have occurred using different cutoff scores from the SCP would have had an impact on reported prevalence and recovery profile. However, had we used a cutoff of 6 (most severe pushers), the current study would have still featured a frequency of 23%. We believe, however, that the cutoff currently selected is appropriate for 2 reasons. First, a cutoff of “0” was chosen because any score >0 indicates a postural orientation toward the affected side or a reaction to being passively positioned toward midline. Such observations are considered clinically meaningful as described by Davies (1985). Second, it is also argued that a score of 0 may actually underestimate persisting symptoms of pushing. A score of 0 only indicates that individuals approached or reached midline. As a result, the measure does not indicate a persisting bias to stand/bear weight on the paretic side. As noted, an alternative explanation for the differences between the studies could include the severity of strokes (or of pushing) and the time of initial evaluation. This is currently difficult to establish, because equivalent measures were not used and details of measurement timing were not always provided.
Another significant observation from this study comes from the detailed assessment of the time course of recovery from pushing symptoms. The current results show that pushing resolved in 62% (24/39) by 6 weeks after stroke and 79% (30/38) by 3 months. The profile of the change revealed that the symptoms in many began to resolve as early as 7 days after stroke onset, which is much sooner than Karnath et al19 were able to show. It should be noted, however, that 8 individuals (21%) continued to exhibit symptoms of pushing at 3 months after stroke. All of these persisting pushers presented with a maximum SCP score (most severe) when initially assessed.
Although the pushers’ degree of motor recovery as measured by the FM motor scale was lower at 3 months, it was noteworthy that the mean rate of change (from 1 week to 3 months) was similar for pushers and nonpushers. The most important difference appeared to be the lower initial stroke severity scores in pushers as opposed to nonpushers, reflecting more severe neurological impairment at start (Figure 2A). This finding that the stroke severity impairment measures were always significantly lower in the pushers is comparable with previous work.18 The differences in recovery between pushers and nonpushers (lower overall scores but parallel rates of recovery) were also observed for functional ability measures (Figure 2B). The determinants of the differences in extent of recovery between pushers and nonpushers may relate to differences in the initial stroke severity with other associated deficits or to the effects of pushing or both.
The current work has shown that the recovery of pushing is not strongly associated with the recovery of motor control, suggesting that the problem arises from factors other than just motor dyscontrol. It is not believed that pushing arises from disrupted primary sensory processing.21,13 Rather, our study reinforces the concept that there is an association between symptoms of pushing and high-level sensory processing/perceptual disorders, such as visuospatial neglect.13,18,21
The association between pushing and longer lengths of stay (63% longer) in inpatient facilities supports previous work.8,19 This is most likely associated with the overall severity of the stroke.28,29 However, it is possible that initial disposition (specifically linked to symptoms of severe pushing) may influence the LOS in some individuals because of potential challenges in participating in therapy.
This article suggests that pushing may have been underrepresented in previous studies, reinforcing the need for well-defined criteria for pushing, such as the SCP measure. The measured improvement in pushing supports the argument that the presence of pushing should not limit opportunities for rehabilitation. Although the pushers did have greater lengths of stay in acute care and rehabilitation facilities, they were discharged home with similar frequency to the nonpushers. Further studies are needed to improve understanding of the underlying etiology to design specific rehabilitation approaches to improve recovery of this disabling stroke syndrome.
This study was supported by funding from the Heart and Stroke Foundation of Ontario and the Canadian Institutes of Health Research.
- Received September 17, 2004.
- Accepted September 21, 2004.
Gresham GE, Duncan PW, Stason WB, Adams HP, Adelman AM, Alexander DN, Bishop DS, Diller L, Donaldson NE, Granger CV, Holland AL, Kelly-Hayes M, McDowell FH, Myers L, Phipps MA, Roth EJ, Siebens HC, Tarvin GA, Trombly CA. Post-stroke Rehabilitation: Clinical Practice Guidelines. Rockville, Md: U.S. Department of Health and Human Services; 1995.
Bonita R, Beaglehole R. Recovery of motor function after stroke. Stroke. 1988; 19: 1497–1500.
Nichols DS. Balance retraining after stroke using force platform biofeedback. Physical Therapy. 1997; 77: 553–558.
Davies PM. Steps to Follow: A Guide to the Treatment of Adult Hemiplegia. New York, NY: Springer-Verlag; 1985.
Bateman A, Riddoch MJ. Neuropsychological perspectives on “pusher syndrome.” European Journal of Physical Medicine and Rehabilitation. 1996; 6: 93–96.
Reding M, David A, Volpe B. Neuroimaging study of the pusher syndrome post stroke. J Neurol Sci. 1997; 150: S129. Abstract.
Karnath HO, Broetz D. Understanding and treating “pusher syndrome.” Phys Ther. 2003; 83: 1119–1125.
Beevor CE. Remarks on paralysis of movement of the trunk in hemiplegia. BMJ. 1909: 881–885.
Brunnstrom S. Movement Therapy in Hemiplegia: A Neurophysiological Approach. New York, NY: Harper & Row Publishers Inc; 1970.
Bohannon RW, Cook AC, Larkin PA, Dubuc WE, Smith MB, Horton MG, Hypes C. The listing phenomenon of hemiplegic patients. Neurology Report. 1986; 10: 43–44.
Taylor D, Ashburn A, Ward CD. Asymmetrical trunk posture, unilateral neglect and motor performance following stroke. Clin Rehabil. 1994; 8: 48–53.
Duncan PW, Goldstein LB, Horner RD, Landsman PB, Samsa GP, Matchar DP. Similar motor recovery of upper and lower extremities after stroke. Stroke. 1994; 25: 1181–1188.
Karnath HO, Ferber S, Dichgans J. The origin of contraversive pushing. Neurology. 2000a; 55: 1298–1304.
Gowland C, VanHullenaar S, Torresin W, Moreland J, Vanspall B, Barrecca S, Ward M, Huijbregts M, Stratford P, Barclay-Goddard R. Chedoke-McMaster Stroke Assessment: Development, Validation, and Administration manual. Hamilton, Ontario, Canada: Chedoke-McMaster Hospitals and McMaster University; 1995.
Seaby L, Torrance G. Reliability of a physiotherapy functional assessment used in a rehabilitation setting. Physiother Can. 1989; 41: 264–271.
Data Management Service, The Centre for Functional Assessment Research. Guide for Use of the Uniform Data Set for Medical Rehabilitation. Buffalo, NY: State University of New York at Buffalo; 1990.
Lyden P, Brott T, Tilley B, Welch KM, Mascha EJ, Levine S, Haley EC, Grotta J, Marler J. Improved reliability of the NIH Stroke Scale using video training. NIINDS TPA Stroke Study Group. Stroke. 1994; 25: 2220–2226.
Bobath B. Adult Hemiplegia: Evaluation and Treatment. Oxford, UK: Butterworth-Heinemann Ltd; 1990.