Temporal and spatial gait characteristics of children with Hurler syndrome after umbilical cord blood transplantation.

Hurler syndrome is the most severe form of mucopolysaccharidosis type I (MPS I) and has an incidence of approximately 1.19 per 100,000 live births. (1,2) Children diagnosed with Hurler syndrome typically have retarded growth, coarse facial features, enlarged tongues, dysostosis multiplex, limitations in joint range of motion, thickening of cardiac valves, hernias, deafness, liver and spleen enlargement, corneal clouding, and abnormal hair growth. Life expectancy for children with Hurler syndrome is typically less than 10 years. Death usually results from airway obstruction, respiratory infection, or cardiac complications. (3) In a previously published case series, including one of the subjects included in this study, we documented gross motor abilities that were below average as early as 10 months of age. (4) Children with untreated Hurler syndrome may never gain the ability to ambulate or may lose this ability as their symptoms progress.

Recent medical advances have increased the number of children with Hurler syndrome who are surviving. In 2003, enzyme replacement therapy (ERT) was approved for the management of MPS I. However, the enzyme does not cross the blood brain barrier, limiting the effect of ERT on the central nervous system. (5) Hematopoietic stem cell transplantation (HSCT) using bone marrow (also referred to as "bone marrow transplantation" [BMT]) has been used successfully to treat children with Hurler syndrome at less than 24 months of age with minimal cognitive impairment. (6) However, BMT does not appear to alter the natural history of musculoskeletal impairments in this population. (7) Umbilical cord blood transplantation (UCBT), a type of HSCT, is a relatively new intervention that has been associated with normalization of growth, kyphosis stabilization or reduction, and cognitive gains. (8) Finally, a combination of HSCT and ERT has recently been investigated. (9)

Although HSCT with and without ERT show promise in improving the somatic impairments, in increasing cognitive functioning, and in increasing the life expectancy of children with Hurler syndrome, little is known about the gross motor and gait abilities of this population after any of these interventions. The few studies that have reported motor abilities indicate that children with Hurler syndrome continue to have delays after BMT. (10,11) Additional research conducted by our team, including 17 of the subjects in this study, indicates that children with Hurler syndrome gain abilities at varying rates in different gross motor domains after UCBT (Dr Dusing, unpublished data, 2005).

Limitations in ambulatory ability, including walking at a slow speed or needing an assistive device, are related to decreased physical activity, limited independence with community activities, and reduced self-esteem in children. (12-14) In addition, the inability to independently ambulate may decrease a child's ability to participate in active, unstructured, and independent play. This type of play appears to be important for social, emotional, and cognitive development. (15) Many children with Hurler syndrome have limitations in range of motion and gross motor deficits, which increase their risk of immature or inefficient gait patterns and limitations in performing functional activities and community participation. Research conducted on individuals 5 to 29 years of age with a less severe form of MPS I, Hurler-Scheie syndrome, documented reduced gait efficiency and functional abilities. (16,17) The presence of these limitations in children with Hurler-Scheie syndrome supports our hypothesis that children with Hurler syndrome will have immature or delayed gait maturation.

The development and increasing utilization of HSCT and ERT to treat children with Hurler syndrome have increased the likelihood that children with Hurler syndrome will have longer lives. Increased life expectancy and improved medical status increase the likelihood that children with Hurler syndrome will be able to ambulate after UCBT. Documentation of temporal and spatial gait characteristics of children with Hurler syndrome after UCBT will enhance our ability to assess the efficacy of orthopedic, physical therapy, and orthotic interventions that may be needed to improve or preserve ambulation. The purpose of this study was to describe the spatial and temporal gait characteristics of children with Hurler syndrome after UCBT using gait characteristics of children with typical development as a reference.

Method

Subjects

This study included a group of 18 children with Hurler syndrome, aged 19.6 to 96.8 months, who received UCBT and were referred to the Neurodevelopmental Function in Rare Disorders (NFRD) Program at the Center for the Study of Development and Learning at the University of North Carolina (Tab. 1). Each of the children with Hurler syndrome participated in 1 to 4 gait assessments longitudinally. Assessment frequency varied and was based on the frequency of clinical follow-up assessments with the NFRD Program. A sample of 438 children who were developing typically, aged 14.4 to 131.8 months, was recruited from local elementary schools, preschools, day care facilities, and the community during a previous study (18) and used as a comparison group (Tab. 1). The comparison group participated in 1 gait assessment. Both studies were approved by the Biomedical Institutional Review Board at the University of North Carolina at Chapel Hill and met the requirements of the Health Insurance Portability and Accountability Act (HIPAA).

Procedure

All children with Hurler syndrome who were examined in the NFRD Program between December 2002 and May 2005 and could walk 15.24 m (50 ft) without an assistive device were included in this study. Each child's date of birth, date of UCBT, date of assessment, height, weight, clinical characteristics, temporal and spatial gait characteristics, and frequency of physical therapy intervention were recorded at each assessment.

Temporal and spatial gait parameters for both groups of children were assessed using a standard clinical protocol and the GAITRite electronic walkway. * The total distance for each walk was 7.66 m, the middle 3.66 m of which was on the GAITRite walkway. The GAITRite walkway was connected to a laptop computer that utilized the GAITRite Gold software (version 3.4). * When necessary, a parent stood at the end of the walkway to encourage the child to walk toward him or her. Each child completed 2 walks (7.66 m each) at his or her self-selected pace with at least 4 footfalls on the GAITRite walkway. Gait speed, cadence, and step length were calculated using GAITRite software. (18) The repeatability of the GAITRite assessment protocol and the calculation included in this study were reported for 57 children aged 1.3 to 10.9 years in a previous publication. (19) Intraclass correlation coefficients (ICC [1,11]) ranged from .62 to .89 for gait speed, cadence, and step length for children under 8 years of age and from .40 to .93 for children 8 to 10.9 years of age. (19) The reader is referred to the article by Thorpe et al (19) for further details.

Data Analysis

Gait speed, cadence, and step length were normalized using nondimensional normalization procedures described by Hof (20) and Stansfield et al (21) to reduce the effect of body stature because children with Hurler syndrome are significantly shorter than children of the same age who are developing typically. The following formulas were used for nondimensional normalization, where g is the acceleration of gravity (9.81 m/[s.sup.2]) and SQRT is square root:

   Normalized step length=step length/
    height

  Normalized gait speed=gait speed/
   SQRT(g x height)

  Normalized cadence = cadence/SQRT
   (g/height)

Hierarchical linear models (HLMs) were fit using SAS Proc Mixed (version 9.01). ([dagger]) Hierarchical linear models allow for variability in the number and timing of assessments while accounting for correlations in the data from repeated measures of individual subjects. (22) The first set of HLMs was used to compare the gait parameters of children with Hurler syndrome with those of the comparison group of children who were developing typically at similar ages. All subjects were considered to be independent of one another, and restricted maximum likelihood estimation was used to estimate the fixed- and random-effect parameters. A separate regression model was fit for each of the 3 measures of gait. The regression models used group membership, age at the time of the assessment modeled up to a quadratic effect, and group x age and group x age squared as interaction terms to predict normalized gait speed, normalized cadence, and normalized step length. Each model was reduced to include only those variables that were significant. Ninety-five percent confidence intervals were calculated and depicted to enhance the reader's ability to visualize group differences.

The results of the planned analyses indicated the need for a post hoc analysis to investigate other possible predictors of these specific gait parameters. An HLM was performed with age at the time of UCBT and the time after UCBT as predictors of the specific normalized gait parameters.

Results

The HLMs included the data of all children in both the group of children with Hurler syndrome and the comparison group. According to parental report, most of the children with Hurler syndrome (n = 13) were receiving physical therapy services at the time of at least one of their assessments. Of those children who were receiving physical therapy services, 75% were receiving the services at least weekly. The intervention setting, type of physical therapy intervention, and goals for therapy were not consistently reported by parents. Two of the children with Hurler syndrome were diagnosed around the time of birth, received UCBT prior to presenting clinical symptoms, and were each assessed a single time. Statistical models were not significantly different with the exclusion of the 2 children who were asymptomatic; therefore, their data were retained for all analyses.

Table 2 presents the mean normalized gait speed, normalized cadence, and normalized step length for the entire study sample and for each group. Gait development of the children with Hurler syndrome lagged behind that of the children in the comparison group with respect to normalized gait speed and normalized step length (Tab. 3, Figs. 1 and 2). Children with Hurler syndrome had slower normalized gait speeds than the comparison group at both 24 and 36 months of age (1.09 [+ or -] 0.03, P<.001 and 0.47 [+ or -] 0.17, P=.008, respectively); however, the differences gradually decreased, and the groups were similar by 48 months of age (0.06 [+ or -] 0.16, P=.70). Children with Hurler syndrome also had lower normalized step lengths at 24 and 36 months of age (0.10 [+ or -] 0.02, P>.001 and 0.05 [+ or -] 0.01, P>.001, respectively), but increased their normalized step length quickly to approximate the normalized step lengths of the children in the comparison group by 48 months (0.01 [+ or -] 0.01, P=.27). There were no between-group differences in normalized cadence, and this parameter decreased linearly for both groups at a similar rate (Tab. 3, Fig. 3). Post hoc analysis revealed that time after UCBT, but not age at the time of UCBT, was positively associated with normalized gait speed and normalized step length (Tab. 4).

[FIGURES 1-3 OMITTED]

Discussion

At 24 and 36 months of age, the children with Hurler syndrome after UCBT walked with a slower speed and shorter step length than the children who were developing typically after controlling for body stature. However, by 48 months of age, their normalized gait speed and normalized step length increased rapidly to reach age-appropriate values. No group differences were observed in normalized cadence at any age, as the change in normalized cadence with gait maturation was small in the children who were developing typically as well as in the children with Hurler syndrome.

Children with Hurler syndrome have been documented to have delayed gross motor abilities both before (4) and immediately after UCBT (Dr Dusing, unpublished data, 2005). Locomotor abilities such as transitioning between positions, creeping, walking, running, and jumping have been the most delayed based on previous research using the Peabody Developmental Motor Scales, second edition. As a result of delayed locomotor abilities, children with Hurler syndrome walk independently at an older age than children who are developing typically. In this study, the youngest children were 19.6 and 14.4 months of age for the group with Hurler syndrome and the comparison group, respectively. Later onset of independent walking, limited practice ambulating while hospitalized, orthopedic abnormalities, and balance deficits that may result from vestibular dysfunction or hydrocephalus could contribute to the immature gait ob served in the children with Hurler syndrome at 24 and 36 months of age. (23-25)

Follow-up analysis provided evidence that a child's age at the time of UCBT had a limited effect on the temporal and spatial gait parameters. However, time after UCBT was a predictor of normalized gait speed and normalized step length. The mean age at UCBT was 18.8 months, an age when most children with typical development are newly independent walkers. Prolonged hospitalization with limited opportunities to practice ambulation may have contributed to diminished endurance and muscle strength (force-generating capacity), which may have affected the children's ability to initiate independent ambulation during this critical period of gait development. However, following a recovery period, the children with Hurler syndrome developed age-appropriate normalized gait speed and normalized step length.

There were several limitations to this study. The use of a comparison group of children who were developing typically rather than a group of children with untreated Hurler syndrome limited our ability to describe the effect of UCBT on temporal and spatial gait parameters in the children with Hurler syndrome. In addition, without a comparison group of children who had similar medical procedures (UCBT), it is unclear what portion of the observed gait deficits were a result of the intervention (UCBT) rather than the syndrome. However, the data included represent the gait of this relatively large sample of children with Hurler syndrome after UCBT and can be used for comparison with future samples to determine the efficacy of interventions such as physical therapy, orthotic device application, or orthopedic interventions. Although the group of children with Hurler syndrome and the comparison group were similar in sex, it is unclear whether the samples were of similar racial or ethnic backgrounds. The comparison group was 74.7% Caucasian, and all subjects in this group resided in 1 of 3 counties in North Carolina. Racial or ethnic data were not collected for the children with Hurler syndrome, and their primary residences were scattered throughout the United States. Socioeconomic data were not available for either group. The majority of children with Hurler syndrome included in this study reported receiving physical therapy services. However, the frequency, setting (school, early intervention, or private), and type of physical therapy services were not consistent or well described by parental report, limiting our ability to include physical therapy services in our analysis.

Joint range of motion could not be reliably assessed in the children with Hurler syndrome, limiting our ability to describe the effect of range of motion on gait in this sample. In addition, specific balance and strength measures were not included in the clinical protocol, limiting our ability to assess relationships among strength, balance, and gait parameters. Analysis of hearing, vestibular function, and hydrocephalus also should be included in future studies.

The results of this study describe the temporal and spatial gait parameters of children with Hurler syndrome after UCBT in reference to a population of children who are developing typically. The findings provide preliminary evidence that 4- to 6-year-old children with Hurler syndrome who received UCBT between 2.1 and 43.9 months of age ambulated with nearly normal gait speed, cadence, and step length for their height. There also was evidence that these children were able to maintain nearly normal temporal and spatial gait parameters through 6 years of age. Although children who are developing typically achieve and maintain adult-like gait at 7 to 10 years of age, children with Hurler syndrome after UCBT may or may not have similar gait outcomes. Orthopedic deformities such as genu valgum and hip dysplasia have been documented in children with Hurler syndrome after BMT and also may be present after UCBT. (7) Further research is needed to assess the presence and effects of orthopedic deformities after UCBT and to document qualitative aspects of gait. As children with Hurler syndrome age, they may be less fearful of range-of-motion assessments, thereby increasing the number of assessments that can be completed reliably. Longitudinal post-UCBT assessment of joint range of motion in these children will improve the ability to analyze the effects of this variable.

Conclusion

The results of this study provide baseline temporal and spatial gait data for children with Hurler syndrome after UCBT. The study also provides evidence that, after controlling for body stature, children with Hurler syndrome after UCBT have less mature gait at 24 and 36 months of age than children who are developing typically. The positive relationship between time after UCBT, gait speed, and step length provides evidence that the more time post-UCBT, the more typical the child's temporal and spatial gait parameters. Recovery from the UCBT process, increased opportunities for ambulation, and increased strength or range of motion each may have contributed to the maturation of gait documented in the study. Further research is needed on the benefits of orthopedic procedures, physical therapy intervention, use of orthotic devices, strengthening and range-of-motion programs, and out-of-bed activities to facilitate gait maturation in children with Hurler syndrome after UCBT.

The authors acknowledge the contribution of each member of the Neurodevelopmental Function in Rare Disorders (NFRD) Program who participated in assessments. They thank Matt Gribin for his assistance with SAS programming.

This work was funded by the following awards to the first author (SCD): Promotion of Doctoral

Studies Award, Foundation for Physical Therapy; Smith Research Grant, Graduate School at the University of North Carolina at Chapel Hill; and Leadership Education in Neurodevelopmental and Related Disabilities Fellowship, The Maternal and Child Health Bureau, North Carolina.

This article was a portion of Dr Dusing's doctoral dissertation.

This article was submitted July 11, 2006, and was accepted April 3, 2007.

DOI: 10.2522/ptj.20060196

References

(1) Poorthuis BJ, Wevers RA, Kleijer WJ, et al. The frequency of lysosomal storage diseases in the Netherlands. Hum Genet. 1999;105:151-156.

(2) Neufeld EF. Muenzer J. The mucopolysaccharidoses. In: Scriver CR, Beaudet AM, Sly WS, Valle D, eds. The Metabolic and Molecular Basis of Inherited Disease. Vol 3. 8th ed. New York, NY: McGraw-Hill; 2001:3421-3452.

(3) Muenzer J. Mucopolysaccharidoses. Adv Pediatr. 1986;33:269-302.

(4) Dusing SC, Thorpe DE, Rosenberg AE, et al. Gross motor abilities in children with Hurler syndrome: a case series. Dev med Child Neurol. 2006;48:927-930.

(5) Shull RM, Kakkis ED, McEntee MF, et al. Enzyme replacement in a canine model of Hurler syndrome. Proc Natl Acad Sci U S A. 1994;91:12937-12941.

(6) Peters C, Shapiro EG, Anderson J, et al, for the Storage Disease Collaborative Study Group. Hurler syndrome, II: outcome of HLA-genotypically identical sibling and HLA-haploidentical related donor bone marrow transplantation in fifty-four children. Blood. 1998;91:2601-2608.

(7) Weisstein JS, Delgado E, Steinbach LS, et al. Musculoskeletal manifestations of Hurler syndrome: long-term follow-up after bone marrow transplantation. J Pediatr Orthop. 2004;24:97-101.

(8) Staba SL, Escolar ML, Poe MD, et al. Cordblood transplants from unrelated donors in patients with Hurler's syndrome. N Engl J Med. 2004;350:1960-1969.

(9) Cox-Brinkman J, Boelens JJ, Wraith JE, et al. Haematopoietic cell transplantation (HCT) in combination with enzyme replacement therapy (ERT) in patients with Hurler syndrome. Bone Marrow Transplant. 2006;38:17-21.

(10) Bjoraker KJ, Delaney K, Peters C, et al. Long-term outcomes of adaptive functions for children with mucopolysaccharldosis I (Hurler syndrome) treated with hematopoietic stem cell transplantation. J Dev Behav Pediatr. 2006;27:290-296.

(11) Hugh-Jones K. Psychomotor development of children with mucopolysaccharidosis type 1-H following bone marrow transplantation. Birth Defects Orig Artic Ser. 1986;22:25-29.

(12) Lancioni GE, O'Reilly MF. A review of research on physical exercise with people with severe and profound developmental disabilities. Res Dev Disabil. 1998;19: 477-492.

(13) Dykens EM, Rosner BA, Butterbaugh G. Exercise and sports in children and adolescents with developmental disabilities: positive physical and psychosocial effects. Child Adolesc Psychiatr Clin N Am. 1998;7:757-771.

(14) Lepage C, Noreau L, Bernard PM. Association between characteristics of locomotion and accomplishment of life habits in children with cerebral palsy. Phys Ther. 1998;78:458-469.

(15) Burdette HL, Whitaker RC. Resurrecting free play in young children: looking beyond fitness and fatness to attention, affiliation, and affect. Arch Pediatr Adolesc Med. 2005;159:46-50.

(16) Haley SM, Fragala Pinkham MA, Dumas HM, et al. A physical performance measure for individuals with mucopolysaccharidosis type I. Dev Med Child Neurol. 2006;48:576-581.

(17) Dumas HM, Fragala MA, Haley SM, et al. Physical performance testing in mucopolysaccharidosis, I: a pilot study. Pediatr Rehabil. 2004;7:125-131.

(18) Dusing SC, Thorpe DE. A normative sample of temporal and spatial gait parameters in children and adults using the GAITRite electronic walkway. Gait Posture. 2007; 25:135-139.

(19) Thorpe DE, Dusing SC, Moore CG. Repeatability of temporospatial gait measures in children using the GAITRite electronic walkway. Arch Phys Med Rehabil. 2005;86:2342-2346.

(20) Hof A. Scaling gait data to body size. Gait Posture. 1996;4:222-223.

(21) Stansfield BW, Hillman SJ, Hazlewood ME, et al. Normalisation of gait data in children. Gait Posture. 2003;17:81-87.

(22) Singer J, Willett J. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York, NY: Oxford University Press; 2003.

(23) Pine RM, Lindeblad S, Donovan P, et al. Balance and motor skills in young children with sensorineural hearing impairment: a preliminary study. Pediatric Physical Therapy. 1996;8(2):55-61.

(24) Blomsterwall E, Bilting M, Stephensen H, Wikkelso C. Gait abnormality is not the only motor disturbance in normal pressure hydrocephalus. Scand J Rehabil Med. 1995;27:205-209.

(25) Cowan JA, McGirt MJ, Woodworth G, et al. The syndrome of hydrocephalus in young and middle-aged adults (SHYMA). Neurol Res. 2005;27:540-547.

* CIR Systems, MAP/CIR Inc, 1625 E Darby Rd, Havertown, PA 19083.

([dagger]) SAS Institute Inc, PO Box 8000, Cary, NC 27511.

Dr Dusing, Dr Thorpe, Dr Mercer, and Dr Rosenberg provided concept/idea/research design. Dr Dusing, Dr Thorpe, and Dr Mercer provided writing. Dr Dusing provided data collection, and Dr Dusing and Dr Poe provided data analysis. Dr Dusing and Dr Escolar provided project management. Dr Dusing, Dr Rosenberg, and Dr Escolar provided fund procurement. Dr Thorpe, Dr Rosenberg, and Dr Escolar provided institutional liaisons. Dr Escolar provided subjects and clerical support. Dr Thorpe, Dr Mercer, Dr Rosenberg, Dr Poe, and Dr Escolar provided consultation (including review of manuscript before submission).

SC Dusing, PT, PhD, is Assistant Professor, Department of Physical Therapy, School of Allied Health Professions, Virginia Commonwealth University, 1200 E Broad St, PO Box 980224, Richmond, VA 23298-0224 (USA). Address all correspondence to Dr Dusing at: [e-mail].

DE Thorpe, PT, PhD, PCS, is Associate Professor, Center for Human Movement Science, Division of Physical Therapy, University of North Carolina, Chapel Hill, NC.

VS Mercer, PT, PhD, is Associate Professor, Center for Human Movement Science, Division of Physical Therapy, University of North Carolina.

AE Rosenberg, PT, DrPH, is Assistant Professor, Center for Human Movement Science, Division of Physical Therapy, University of North Carolina, and Training Director, Center for the Study of Development and Learning, University of North Carolina.

MD Poe, PhD, is Investigator, Frank Porter Graham Child Development Institute, University of North Carolina.

ML Escolar, MD, is Director, Program for the Study of Neurodevelopmental Function and Rare Disorders, University of North Carolina.

[Dusing SC, Thorpe DE, Mercer VS, et al. Temporal and spatial gait characteristics of children with Hurler syndrome after umbilical cord blood transplantation. Phys Ther. 2007;87:978-985.]

Table 1.
Subject Information

                       Entire Sample           Children With
                                               Hurler Syndrome

No. of subjects        456                     18
No. of observations    477                     39
Female (%)              44.7                   44.4
Age (a) ([bar.X]        79.10 [+ or -] 31.23   48.62 [+ or -] 17.67
  [+ or -] SD)
Tage (b) ([bar.X]      NA (d)                  18.75 [+ or -] 8.52
  [+ or -] SD)
TPT (c) ([bar.X]       NA                      29.87 [+ or -] 2.91
  [+ or -] SD)
Height (cm) ([bar.X]   117.77 [+ or -] 19.13   95.15 [+ or -] 9.72
  [+ or -] SD)
Weight (kg) ([bar.X]    25.67 [+ or -] 11.65   17.07 [+ or -] 3.90
  [+ or -] SD)

                       Comparison Group

No. of subjects        438
No. of observations    438
Female (%)              44.7
Age (a) ([bar.X]        81.82 [+ or -] 30.74
  [+ or -] SD)
Tage (b) ([bar.X]      NA
  [+ or -] SD)
TPT (c) ([bar.X]       NA
  [+ or -] SD)
Height (cm) ([bar.X]   119.79 [+ or -] 18.46
  [+ or -] SD)
Weight (kg) ([bar.X]    26.44 [+ or -] 11.80
  [+ or -] SD)

(a) Age=age (in months) at the time of assessment.

(b) Tage=age (in months) at the time of umbilical cord blood
transplantation.

(c) TPT=time (in months) after umbilical cord blood transplantation.

(d) NA = not applicable.

Table 2.
Normalized Gait Parameters

                              Entire Sample

Normalized gait speed      3.49 [+ or -] 0.77
  ([bar.X] [+ or -] SD)
Normalized cadence        501.64 [+ or -] 73.89
  ([bar.X] [+ or -] SD)
Normalized step length      0.41 [+ or -] 0.06
  ([bar.X] [+ or -] SD)

                              Children With
                             Hurler Syndrome

Normalized gait speed      3.20 [+ or -] 1.10
  ([bar.X] [+ or -] SD)
Normalized cadence        511.54 [+ or -] 119.22
  ([bar.X] [+ or -] SD)
Normalized step length      0.37 [+ or -] 0.08
  ([bar.X] [+ or -] SD)

                               Comparison
                                  Group

Normalized gait speed      3.51 [+ or -] 0.73
  ([bar.X] [+ or -] SD)
Normalized cadence        500.75 [+ or -] 668.57
  ([bar.X] [+ or -] SD)
Normalized step length     0.42 [+ or -] 0.05
  ([bar.X] [+ or -] SD)

Table 3.
Hierarchical Linear Models With Group and Age as Predictors of
Normalized Gait Parameters

                   Normalized Gait Speed

                   Estimate [+ or -] SE      P

Intercept           3.45 [+ or -] 0.16     <.001
Group               0.06 [+ or -] 0.16      .70
Age (a) linear      0.03 [+ or -] 0.008     .0001
Age quadratic    -0.0009 [+ or -] 0.0003    .005
Age X group        -0.03 [+ or -] 0.009     .004
Group X age       0.0007 [+ or -] 0.0003    .02
  quadratic

                  Normalized Cadence

                 Estimate [+ or -] SE      P

Intercept        511.86 [+ or -] 11.59   <.001
Group              6.12 [+ or -] 12.65     .63
Age (a) linear    -0.51 [+ or -] 0.11    <.001
Age quadratic    NA (b)                  NA
Age X group      NA                      NA
Group X age      NA                      NA
  quadratic

                  Normalized Step Length

                   Estimate [+ or -] SE        P

Intercept            0.39 [+ or -] 0.01      <.001
Group                0.01 [+ or -] 0.01       .27
Age (a) linear      0.004 [+ or -] 0.0005    <.001
Age quadratic    -0.00007 [+ or -] 0.00002    .0004
Age X group        -0.002 [+ or -] 0.0006    <.001
Group X age       0.00006 [+ or -] 0.00002    .004
  quadratic

(a) Age = age (in months) at the time of assessment.

(b) NA = not applicable.

Table 4.
Post Hoc Hierarchical Linear Models for Age at Time of Umbilical
Cord Blood Transplantation (UCBT) and Time After UCBT as
Predictors of Normalized Gait Parameters

                        Normalized Gait Speed

                        Estimate [+ or -] SE          P

Intercept                 3.56 [+ or -] 0.22       <.001
Tage (a)                -0.008 [+ or -] 0.04        .83
TPT (b) linear           0.041 [+ or -] 0.01        .002
TPT (b) quadratic       -0.001 [+ or -] 0.0005      .02
Tage (a) x TPT (b)     -0.0008 [+ or -] 0.002       .66
  linear
Tage (a) x TPT (b)   0.00000219 [+ or -] 0.00006    .97
  quadratic

                      Normalized Cadence

                     Estimate [+ or -] SE       P

Intercept            543.47 [+ or -] 24.54   <.001
Tage (a)              -3.80 [+ or -] 3.95     .34
TPT (b) linear         2.22 [+ or -] 1.39     .12
TPT (b) quadratic     -0.11 [+ or -] 0.05     .05
Tage (a) x TPT (b)    -0.01 [+ or -] 0.19     .94
  linear
Tage (a) x TPT (b)   0.0006 [+ or -] 0.007    .93
  quadratic
                       Normalized Step Length

                        Estimate [+ or -] SE            P

Intercept                0.39 [+ or -] 0.01           <.001
Tage (a)                0.002 [+ or -] 0.002           .44
TPT (b) linear          0.004 [+ or -] 0.0008         <.001
TPT (b) quadratic     -0.00008 [+ or -] 0.00003        .01
Tage (a) x TPT (b)     -0.0001 [+ or -] 0.0001         .36
  linear
Tage (a) x TPT (b)   0.000000631 [+ or -] 0.0000038    .87
  quadratic

(a) Tage=age (in months) at the time of umbilical cord blood
transplantation.

(b) TPT=time (in months) after umbilical cord blood transplantation.