Toe
Walking and Language Development
Pasquale Accardo, M.D.* Jill Morrow, M.D.* M. Susan Heaney, M.D.* Barbara Whitman, Ph.D.* Terry Tomazic, Ph. D.**
Summary: Neurodevelopmental markers that are present early in childhood may identify children at risk for later developmental disabilities. This paper attempts to clarify the relationship between one such proposed marker, toe walking, and language development in a general pediatric population. One hundred sixty-three children being seen for well-child visits were included in the study. Information from each child’s caretaker was obtained for language development and a history of toe walking; observation of toe walking during the visit was also included. The frequency of toe walking was 24%. Language quotients were calculated and compared for toe walkers (n=39) and non-toe walkers (n=127). The mean language quotient for toe walkers tended to be consistently lower than that for non-toe walkers. The specificity of toe walking for low language scores was 85% but had a sensitivity of only 32%. Although an association between toe walking and language delay is supported by the present data, the association does not appear to be clinically significant. Toe walking,
in the absence cerebral palsy JL. or other neuromuscular disorders, is generally described as a normal variant in the evolution of infant gait. 1,2 The presence of toe walking for over three months is considered persistent and has been characterized by orthopedic surgeons as either congenital short
of autism
*
**
or
tendo calcaneus or habitual toewalking.3.4 Both are considered to be benign conditions without longterm
neurodevelopmental implica-
tions. However, clinical impression and experience suggest a link between benign toe walking and a delay in language development. Both language delay and toe walking are characteristic of autistic
Department of Pediatrics, St. Louis University School of Medicine Department of Research Methodology, St. Louis University This
study was funded in part by Columbian Charities of Missouri.
Presented at the Society for Pediatric Research meeting May 7-11, 1990, in Anaheim, California. Address correspondence to: Pasquale Accardo, M.D., Knights of Columbus Developmental Center, Cardinal Glennon Children’s Hospital, 1465 S. Grand Blvd., St. Louis, M0 63104-1095.
158
Downloaded from cpj.sagepub.com at UNIVERSITE DE MONTREAL on June 25, 2015
children. A previous retrospective study of 799 developmentally disabled children showed that toe walking was prominent not only in children with autism but also in children with less severe language and communication disorders.5 Moreover, the prevalence of toe walking was higher in those diagnostic subgroups with more severe language delay. Interpreting autistic disorder from one perspective as the paradigm for severe language problems, it was hypothesized that normal children who toe walk might have lowered language abilities. The present study was undertaken to investigate the strength of association between toe walking and language delay in an unselected
ambulatory pediatric population.
Subjects and Methods The general pediatric population studied at Cardinal Glennon Children’s Hospital is 53% nonwhite, with an equal gender ratio. Subjects were recruited from the ambulatory pediatric clinic of one practitioner (MSH) at the time of routine well-child visits. Recruitment occurred during the 18month period between September 1988 and March 1990. The children in the study ranged in age from eight months to 10 years. All
subjects were normal children who walking at the time of the visit; normality was defined by ongoing routine pediatric developmental screening and the absence of any medical diagnosis that might indicate an interference with gait or language. A total of 186 children were
were
recruited, from which 23 were
then excluded. Exclusions included any child with a diagnosis of cerebral palsy, mental retardation
(n=2), or previously diagnosed language disorder (n=9). Also excluded
were
children who had
a
history of any condition known to affect central nervous system function, such as meningitis (n=3), and any child who had re-
peated a grade in school because of academic problems (n=5). Two children were excluded because the historians were judged to be poor informants. The study group did include five children (3.1%) who had been premature infants of between 32 and 35 weeks gestational age; two of these were toe walkers. Study instruments included an objective measure of language development either the Clinical Linguistic and Auditory Milestones Scale (CLAMS) for children less than two yearS6,1 or the Mecham Language Scale for children two years and older’ and a structured interview regard-
-
Figure 1.
Mean
language quotients for toe walkers and non-toe walkers in three age groups.
ing toe-walking history, speech and language problems, and school performance. Toe walking scored as present if it was observed during the visit (no duration criterion) or if a history of previous toe walking of more than one month’s duration was obtained from the caretaker; otherwise it was scored as absent. (The orthopedic criterion of three months’ duration was not used.) The history was obtained by one author, while group assignment and scoring of language scales was determined by another. Children were placed in age groups on the basis of which language scale was used (CLAMS for those 0 to 2 years of age) and school attendance was
(school-age versus preschool). Results
frequency of toe walkers all age groups was 24%. By age groups, toe walking was present or noted previously by the mothers in 35% of the infants (0 to 2 years), 19% of the pre-schoolers (2 to 6 years), and 22% of the The
across
school-age children (6 to 10 years) . The 163 children included in the study were divided into six groups based on their age and toe-walking history. Mean language quotients (language age divided by chronological age) were calculated for toe walkers and non-toe walkers in each of the three age subgroups: infants, preschoolers, and school-age children. The mean language quotient for toe walkers was lower than that for non-toe walkers for all age subgroups (Figure 1). The discrepancies among the language quotients for the three age subgroups may indicate true differences in the language abilities of the members of each subgroup but may also reflect the varying sensitivities of the instruments used to measure language skills at different ages. An odds-ratio was calculated for each age subgroup, using a language quotient cut-off of 85 or below as suspect for the presence of a mild language delay or the diagnosis of a language disorder ; the odds-ratio was calculated for the number of subjects above and below the cut-off and
Downloaded from cpj.sagepub.com at UNIVERSITE DE MONTREAL on June 25, 2015
159
not
the was
for their
language scores. For preschoolers, the odds-ratio
2.46. The odds-ratio for the
school-age group was 9.7, but was only 1.79 for the infant group. An odds-ratio of greater than 2.0 supports the presence of a meaningful association. For the entire study population the specificity was 85%, while the sensitivity was 32%.
Discussion Toe walking has previously been linked to language delay in a population of developmentally disabled children. The incidence of toe walking and its relationship to language development has not been studied in the general pediatric population. Our results indicate that even in children without a diagnosis of a neurodevelopmental problem, toe walking is associated with lower mean language scores. This trend was present across all age groups and was most striking in
preschoolers.
160
The association of toe walking and language diagnoses in our previous study and the trend toward lower language scores in normal child toe walkers in the present study both suggest that toe walking may not be an entirely benign incidental finding, even though the association does not appear to be sensitive enough to achieve a specific clinical diagnosis. A history (or the presence) of toe walking does not necessarily indicate a serious language delay. The potential utility of an early nonlinguistic marker for language problems in young children nevertheless remains sufficiently promising to warrant further investigation, with controls for both minor motor dysfunction and general cognition. The possibility that the weak association between toe walking and language delay may represent a mild form of subclinical diplegia remains to be investigated. A prospective study of the evolution of equinus gait and language is indicated.
Downloaded from cpj.sagepub.com at UNIVERSITE DE MONTREAL on June 25, 2015
REFERENCES 1.
Green M: Pediatric
2.
phia, PA: WB Saunders; 1980. Illingworth R. Common Symptoms of Dis-
3.
Diagnosis.
Philadel-
ease in Children. Oxford, UK: Blackwell Scientific Publications; 1975. Griffm P, Wheelhouse W, Shiavi R, Bass W. Habitual toe walkers: A clinical and
electromyographic gait analysis. J Bone Joint Surg. 1977;59A:97-101. 4.
Lovell W, Price C, Meehan P. The foot. In: Lovell W, Winter R, eds. Pediatric Orthopedics. Vol. 2. Philadelphia, PA: JB Lippincott; 1986:918-919.
5.
Accardo a
P, Whitman B. Toe walking as neurodevelopmental marker for lan-
guage 6.
disorders.
Clin
Pediatr.
1989;28:347-350. Capute A, Palmer F, Shapiro B,
et
al.
Clinical
linguistic and auditory milestone scale: prediction of cognition in infancy. Dev Med Child Neurol. 1986;28:762-771. 7.
Capute A, Shapiro B,
Wachtel
RC,
Gunther VA, Palmer FB. The clinical linguistic and auditory milestone scale (CLAMS): identification of cognitive defects in motor-delayed children. Am J
8.
Dis Child. 1986;140:694-698. Mecham J. Verbal Language Development. Los Angeles, CA: Western Psychological Services, 1971.
Walking and Language Development
Pasquale Accardo, M.D.* Jill Morrow, M.D.* M. Susan Heaney, M.D.* Barbara Whitman, Ph.D.* Terry Tomazic, Ph. D.**
Summary: Neurodevelopmental markers that are present early in childhood may identify children at risk for later developmental disabilities. This paper attempts to clarify the relationship between one such proposed marker, toe walking, and language development in a general pediatric population. One hundred sixty-three children being seen for well-child visits were included in the study. Information from each child’s caretaker was obtained for language development and a history of toe walking; observation of toe walking during the visit was also included. The frequency of toe walking was 24%. Language quotients were calculated and compared for toe walkers (n=39) and non-toe walkers (n=127). The mean language quotient for toe walkers tended to be consistently lower than that for non-toe walkers. The specificity of toe walking for low language scores was 85% but had a sensitivity of only 32%. Although an association between toe walking and language delay is supported by the present data, the association does not appear to be clinically significant. Toe walking,
in the absence cerebral palsy JL. or other neuromuscular disorders, is generally described as a normal variant in the evolution of infant gait. 1,2 The presence of toe walking for over three months is considered persistent and has been characterized by orthopedic surgeons as either congenital short
of autism
*
**
or
tendo calcaneus or habitual toewalking.3.4 Both are considered to be benign conditions without longterm
neurodevelopmental implica-
tions. However, clinical impression and experience suggest a link between benign toe walking and a delay in language development. Both language delay and toe walking are characteristic of autistic
Department of Pediatrics, St. Louis University School of Medicine Department of Research Methodology, St. Louis University This
study was funded in part by Columbian Charities of Missouri.
Presented at the Society for Pediatric Research meeting May 7-11, 1990, in Anaheim, California. Address correspondence to: Pasquale Accardo, M.D., Knights of Columbus Developmental Center, Cardinal Glennon Children’s Hospital, 1465 S. Grand Blvd., St. Louis, M0 63104-1095.
158
Downloaded from cpj.sagepub.com at UNIVERSITE DE MONTREAL on June 25, 2015
children. A previous retrospective study of 799 developmentally disabled children showed that toe walking was prominent not only in children with autism but also in children with less severe language and communication disorders.5 Moreover, the prevalence of toe walking was higher in those diagnostic subgroups with more severe language delay. Interpreting autistic disorder from one perspective as the paradigm for severe language problems, it was hypothesized that normal children who toe walk might have lowered language abilities. The present study was undertaken to investigate the strength of association between toe walking and language delay in an unselected
ambulatory pediatric population.
Subjects and Methods The general pediatric population studied at Cardinal Glennon Children’s Hospital is 53% nonwhite, with an equal gender ratio. Subjects were recruited from the ambulatory pediatric clinic of one practitioner (MSH) at the time of routine well-child visits. Recruitment occurred during the 18month period between September 1988 and March 1990. The children in the study ranged in age from eight months to 10 years. All
subjects were normal children who walking at the time of the visit; normality was defined by ongoing routine pediatric developmental screening and the absence of any medical diagnosis that might indicate an interference with gait or language. A total of 186 children were
were
recruited, from which 23 were
then excluded. Exclusions included any child with a diagnosis of cerebral palsy, mental retardation
(n=2), or previously diagnosed language disorder (n=9). Also excluded
were
children who had
a
history of any condition known to affect central nervous system function, such as meningitis (n=3), and any child who had re-
peated a grade in school because of academic problems (n=5). Two children were excluded because the historians were judged to be poor informants. The study group did include five children (3.1%) who had been premature infants of between 32 and 35 weeks gestational age; two of these were toe walkers. Study instruments included an objective measure of language development either the Clinical Linguistic and Auditory Milestones Scale (CLAMS) for children less than two yearS6,1 or the Mecham Language Scale for children two years and older’ and a structured interview regard-
-
Figure 1.
Mean
language quotients for toe walkers and non-toe walkers in three age groups.
ing toe-walking history, speech and language problems, and school performance. Toe walking scored as present if it was observed during the visit (no duration criterion) or if a history of previous toe walking of more than one month’s duration was obtained from the caretaker; otherwise it was scored as absent. (The orthopedic criterion of three months’ duration was not used.) The history was obtained by one author, while group assignment and scoring of language scales was determined by another. Children were placed in age groups on the basis of which language scale was used (CLAMS for those 0 to 2 years of age) and school attendance was
(school-age versus preschool). Results
frequency of toe walkers all age groups was 24%. By age groups, toe walking was present or noted previously by the mothers in 35% of the infants (0 to 2 years), 19% of the pre-schoolers (2 to 6 years), and 22% of the The
across
school-age children (6 to 10 years) . The 163 children included in the study were divided into six groups based on their age and toe-walking history. Mean language quotients (language age divided by chronological age) were calculated for toe walkers and non-toe walkers in each of the three age subgroups: infants, preschoolers, and school-age children. The mean language quotient for toe walkers was lower than that for non-toe walkers for all age subgroups (Figure 1). The discrepancies among the language quotients for the three age subgroups may indicate true differences in the language abilities of the members of each subgroup but may also reflect the varying sensitivities of the instruments used to measure language skills at different ages. An odds-ratio was calculated for each age subgroup, using a language quotient cut-off of 85 or below as suspect for the presence of a mild language delay or the diagnosis of a language disorder ; the odds-ratio was calculated for the number of subjects above and below the cut-off and
Downloaded from cpj.sagepub.com at UNIVERSITE DE MONTREAL on June 25, 2015
159
not
the was
for their
language scores. For preschoolers, the odds-ratio
2.46. The odds-ratio for the
school-age group was 9.7, but was only 1.79 for the infant group. An odds-ratio of greater than 2.0 supports the presence of a meaningful association. For the entire study population the specificity was 85%, while the sensitivity was 32%.
Discussion Toe walking has previously been linked to language delay in a population of developmentally disabled children. The incidence of toe walking and its relationship to language development has not been studied in the general pediatric population. Our results indicate that even in children without a diagnosis of a neurodevelopmental problem, toe walking is associated with lower mean language scores. This trend was present across all age groups and was most striking in
preschoolers.
160
The association of toe walking and language diagnoses in our previous study and the trend toward lower language scores in normal child toe walkers in the present study both suggest that toe walking may not be an entirely benign incidental finding, even though the association does not appear to be sensitive enough to achieve a specific clinical diagnosis. A history (or the presence) of toe walking does not necessarily indicate a serious language delay. The potential utility of an early nonlinguistic marker for language problems in young children nevertheless remains sufficiently promising to warrant further investigation, with controls for both minor motor dysfunction and general cognition. The possibility that the weak association between toe walking and language delay may represent a mild form of subclinical diplegia remains to be investigated. A prospective study of the evolution of equinus gait and language is indicated.
Downloaded from cpj.sagepub.com at UNIVERSITE DE MONTREAL on June 25, 2015
REFERENCES 1.
Green M: Pediatric
2.
phia, PA: WB Saunders; 1980. Illingworth R. Common Symptoms of Dis-
3.
Diagnosis.
Philadel-
ease in Children. Oxford, UK: Blackwell Scientific Publications; 1975. Griffm P, Wheelhouse W, Shiavi R, Bass W. Habitual toe walkers: A clinical and
electromyographic gait analysis. J Bone Joint Surg. 1977;59A:97-101. 4.
Lovell W, Price C, Meehan P. The foot. In: Lovell W, Winter R, eds. Pediatric Orthopedics. Vol. 2. Philadelphia, PA: JB Lippincott; 1986:918-919.
5.
Accardo a
P, Whitman B. Toe walking as neurodevelopmental marker for lan-
guage 6.
disorders.
Clin
Pediatr.
1989;28:347-350. Capute A, Palmer F, Shapiro B,
et
al.
Clinical
linguistic and auditory milestone scale: prediction of cognition in infancy. Dev Med Child Neurol. 1986;28:762-771. 7.
Capute A, Shapiro B,
Wachtel
RC,
Gunther VA, Palmer FB. The clinical linguistic and auditory milestone scale (CLAMS): identification of cognitive defects in motor-delayed children. Am J
8.
Dis Child. 1986;140:694-698. Mecham J. Verbal Language Development. Los Angeles, CA: Western Psychological Services, 1971.
