Journal of the American College of Nutrition, Vol. 18, No. 2, 201-205 (1999)
Published by the American College of Nutrition
Knee Height as a Predictor of Recumbent Length for Individuals with Mobility-Impaired Cerebral Palsy
S. Eileen Hogan, Ph.D.
School of Nutrition and Food Science, Acadia University, Wolfville, Nova Scotia, CANADA
Address reprint requests to: S. Eileen Hogan, PhD, Associate Professor, School of Nutrition and Food Science, Acadia University, Wolfville, Nova Scotia, CANADA. BOP 1XO
 |
ABSTRACT
|
|---|
Background: Because accurate measures of recumbent length are essential to assess growth and energy requirements of mobility-impaired individuals with cerebral palsy (CP), a reliable and simple method of estimating recumbent length is required. Prediction of recumbent length from knee height in this population has not yet been investigated.
Objectives: i) To correlate direct measures of recumbent length in mobility-impaired individuals having lower leg extremity cerebral palsy (LECP) involvement with indirect measures of recumbent length calculated using knee-height prediction equations and ii) to determine if knee height is a reliable predictor of recumbent length in this population.
Methods: Subjects (n=34; 15F, 19M), aged 6 to 30 years, were participants in a six-month nutrition rehabilitation program. All subjects had varying degrees of LECP involvement. Recumbent length to the nearest 0.5 cm was measured by standardised techniques. Knee height was measured to the nearest 0.5 centimetre using sliding callipers. Equations based on normal, healthy individuals with application to mobility-impaired or handicapped individuals were used to predict recumbent length from knee height.
Results: Direct measures of recumbent length of subjects significantly correlated with indirect measures calculated using knee height prediction equations (R=0.88, p
0.0001). In addition, knee height of these subjects was a reliable predictor of recumbent length (R2=0.78, p<0.0001).
Conclusions: Results suggest that knee height may be a reliable predictor for recumbent length in this population.
Key words: knee height, recumbent length, lower leg extremity cerebral palsy (LECP) involvement
|
INTRODUCTION
|
|---|
Because of fixed joint contractures, scoliosis and/or involuntary movements, it is difficult to obtain reliable anthropometric measurements from individuals who are severely handicapped, particularly those with contractures who cannot stand and/or those with involuntary movements [1–4]. To assess linear growth in children with severe developmental disabilities, Cameron (1986) recommended either upper-arm or lower-arm measurements [5]. Upper-arm length is preferred to lower-arm length because the acromion is easier to palpate than the ulnar station. Spender et al. (1989) [6] and Han and Lean, (1996) [7] recommended lower-leg length. Because the standardised measurement of knee height is easy to perform and provides a high degree of precision, knee height has been recommended as routine anthropometry of both children and adults without lower-leg extremity cerebral palsy (LECP) involvement [8–9]. Furthermore, knee-height growth charts for normal children are now available [10].
Prediction of recumbent length from knee height in individuals with LECP involvement has not been investigated. In addition, the reliability of knee height as a predictor of recumbent length in this population has not been assessed [11]. Since both standardised techniques for measurement of knee height and recumbent length [12] and combined reference data from the NHANES 1 (1971–1974) and NHANES II (1976–1980) [13] were used in the present study for comparative purposes, the results should provide more reliable estimates of recumbent length for this population. The purpose of the present study is: (a) to correlate direct measures of recumbent length with indirect estimates of recumbent length in both children and adults having cerebral palsy (CP) with varying degrees of lower leg extremity involvement and ii) to determine if knee height is a reliable predictor of recumbent length in this population. Indirect estimates of recumbent length were calculated using knee-height prediction equations based on healthy, normal individuals, but developed for use with mobility-impaired individuals such as those with LECP involvement [11].
|
METHODS
|
|---|
A six-month nutrition rehabilitation program was conducted with a group of severely developmentally-handicapped individuals in a chronic-care facility and an auxiliary group home in Southern Ontario, Canada. All study procedures conformed to the standards decreed by the Human Ethics Committee at the University of Guelph. The project received approval of the board of directors and the administration of the facility. All parents of potential subjects received a letter describing the study in detail and requesting consent for their daughters or sons to participate. The consent form was signed by the parents of all participants prior to commencement of the study. Thirty-four severely physically-handicapped residents (15F, 19M), aged 6 to 30 years, participated in this study. Although 56 residents with CP participated in the nutrition rehabilitation program, only 50 residents had LECP involvement. However, based on the degree of LECP involvement, knee height could be assessed for only 34 of these residents. The major medical diagnoses of these participants included cerebral palsy or cerebral palsy with spastic quadriplegia. All subjects had profound mental retardation. Table 1 describes other characteristics of the 34 residents. The nutrition rehabilitation program, which was previously reported in detail [14], consisted of individualised diets containing optimum nutrients for age and sex as suggested by the Canadian Recommended Nutrient Intake (RNI) [15].
Anthropometric measurements consisted of weight, recumbent length and knee height. A detailed description of the dietary and anthropometric measurement procedures was also reported earlier, together with procedures followed to calculate the technical error of measurements (Tem) and coefficient of variation (CR) [16]. Each recumbent anthropometric measurement was taken at the beginning and end of the nutrition rehabilitation program. However, for the purpose on this study, baseline recumbent length measurements only are reported and the entire sample is represented as a single group. The project director (SEH) took measurements three times on the same limb side and the mean of the three measurements was calculated [16]. When subjects were very upset and uncooperative during the measurement process, repeated measurements were taken at another time within a one-week period. All measurements were taken on the left side of the body unless body contractures or deformities were more prevalent on the left side. Recumbent length (to the nearest 0.5 cm) was measured with the subject lying on a flat surface with a board placed at the top of the head. With the subject’s head in the frankfurt plane and the help of an assistant, measurement was taken from the top of the head to the bottom of the heel of the longest leg [1,12,17]. When the subject became tactile defensive because she or he was restricted or held down, only the anthropometrist (SEH) measured the subject. Knee height (to nearest 0.5 cm) was measured as the supine subject bent the left knee to the 90° angle. One blade of the Mediform sliding calliper (Anthropometric calliper, Beaverton, Oregon) was placed under the heel of the left foot and the other blade was placed over the anterior surface of the femoral condyles of the left thigh, adjacent to the patella [12,17]. The shaft of the calliper remained parallel to the long axis of the tibia. Gentle pressure was applied to compress the tissue [12]. In this study, knee height prediction equations previously used to estimate stature in ambulatory, healthy individuals, but with application to non-ambulatory individuals, were used to predict recumbent length [11].
Reliability refers to the variation between measurements taken over time [18]. To estimate measurement reliability, mean measurements were assessed on ten residents, from which the technical error of measurement (Tem) and the coefficient of reliability (CR) were calculated using the equations recommended by Frisancho (1990) [13] and Chumlea et al. (1990) [18] (Table 2).
Statistical Analysis
Data were analyzed using SYSTAT 7.0 Statistics software package [19]. For all analysis, significance level was set at p0.05. Tests of significance were all two-tailed. Analysis included descriptive statistics, correlations and regression analysis. Pearson product-correlation coefficient analysis was used to determine the relationship between direct measures of recumbent length (RL) and indirect estimates of RL calculated using knee-height prediction equations. Regression analysis evaluated the significance of knee height as a predictor of recumbent length with age as an independent variable.
|
RESULTS
|
|---|
The Tem and the coefficients of reliability (CR) of the recumbent anthropometric measurements are shown on Table 2. The Tem was used as an index of the error of imprecision of recording and measurement techniques. To determine the level of accuracy in a measure, it was necessary to compare the Tem values obtained from the study anthropometrist (SEH) to the reference values of a well-trained examiner. The technical error of measurement of recumbent length was slightly greater than recommended. This was attributed to the difficulty of precise measurement inherent in this population due to fear, body contractures, tactile defensiveness, involuntary muscle movement and blindness. To date there is no published intra-examiner mean-reference value for knee height. The CR and 1-CR indicate that both knee height and recumbent length were reliable measurements (Table 2).
Table 3 shows the mean direct measure of recumbent length of subjects compared to the mean predicted values. Direct measures of recumbent length significantly correlated with predicted recumbent length (r=.88, p0.0001). Furthermore, knee height was found to be a significant predictor variable for recumbent length in these subjects (Table 4). Age was included as an independent variable because of the loss of recumbent length that occurs with age, especially in women and those with increased severity of disease.
View this table:
[in this window]
[in a new window]
|
Table 3. Mean (±SD*) Direct Measures of Recumbent Length (RL) and Mean (±SD*) Predicted RL of Subjects (Calculated Using Knee Height Prediction Equations) and Correlation of Direct Measures of RL with Predicted Measures of Recumbent Length
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Regression Analysis of Knee Height as a Predictor of Recumbent Length of Subjects When Age Is Included as an Independent Variable
|
|
Chumlea et al. (1994) state that the use of recumbent anthropometric data may increase the errors of prediction of anthropometric indices over those reported on normal, healthy individuals [11]. The mean standard error for predicting recumbent length for these handicapped subjects was greater than the standard error for predicting stature of normal, healthy individuals [11] (Tables 5 and 6). The known error of these prediction equations provides the health professional with an estimate of error that must be considered in predicting stature and setting population-based nutrition recommendations [11].
View this table:
[in this window]
[in a new window]
|
Table 5. Measures of Performance of the Equation Models for Predicted Recumbent Length of Male and Female Subjects in the Validation* and Cross-Validation** Groups
|
|
The difference between predicted and direct measurement of recumbent length for each subject is the residual. The residuals had a mean of zero. Using the procedure of Chumlea et al. 1994, the subjects were divided into two groups, the validation group and the cross-validation group. The accuracy of the estimation equation for recumbent length from the validation group was tested by applying the equation to the cross-validation group [11]. The pure error was assessed. Table 5 shows the measures of performance of the equation models in both groups. There is no systematic difference between stature and recumbent length in normal individuals [20]. As well, the prediction sum of squares (PRESS) selection criterion showed that the residuals did not have a systematic bias [21]. Table 5 shows the root mean squared error (RMSE) and the pure error of two of these models. The pure error is the mean of the squared differences between the measured and estimated recumbent length or the level of accuracy of the measure [11]. When applied to individuals with LECP involvement, the estimate of recumbent length from these prediction equations is reliable, but shows a larger error than when applied to normal, healthy individuals [11] (Tables 5 and 6). Table 6 compares the reliability of the prediction equations in this study to previously published prediction equations used to determine stature of normal, healthy individuals [11].
|
DISCUSSION
|
|---|
Because direct measures of height are not possible in most individuals with CP, with or without LECP involvement [8], recumbent length can assess stature. Interobserver reliability as well as means and standard deviations of measurements taken in either a standing or recumbent position have been reported to be of similar value [20] and similar to those reported for standard anthropometric techniques [11]. Comparison of the coefficient of variation and reliability for recumbent length measurements indicates that more accurate measurements would be collected if the individual were placed in a recumbent position [12]. Furthermore, measurement errors for recumbent anthropometry are slightly less than similar errors for standing anthropometry, but the differences are insignificant [20].
Recumbent anthropometric data, particularly recumbent length, was difficult to obtain from this population due to body contractures, stature deformities, tactile defensiveness, spastic movements, blindness and ensuing fear and isolation [1,4]. More severely handicapped individuals with CP such as the subjects in this study cannot be measured reliably even in the supine position [8]. In addition, most health care professionals do not have access to a supine measuring board long enough to measure individuals over three years of age. Because of the heterogeneous effect of disease on growth parameters in the mobility-impaired, greater caution must be taken in the interpretation of recumbent length [8].
Knee height has been suggested as an alternative measure of stature in the severely physically handicapped [12] and in individuals with cerebral palsy [8–9]. In mobility-impaired individuals, accurate assessment of recumbent length versus height is important, not only to assess growth, but also to develop equations to estimate basal and total energy requirements in the assessment of over- and undernutrition. Knee height provides a reliable and affordable surrogate for assessing recumbent length in mobility-impaired individuals with CP [8]. The knee height calliper is commercially available for about $250, so health care professionals can use it in their practises.
Earlier studies reported the use of knee-height equations based on healthy, normal individuals to estimate height in ambulatory individuals with cerebral palsy [8–9]. These equa-tions served as a method of comparing height measurements of these individuals with an alternate method of assessing stature [9]. Johnson and Ferrara, (1991) showed that knee height correlated with height especially in females over age 18 and in men over age 18 without LECP involvement [9]. Stevenson (1995) validated the clinical usefulness of knee height as a "proxy" for stature in children with CP up to twelve years of age (r=0.98; R2=94) [8]. Furthermore, Stevenson (1995) found no differences in the correlation of knee height with height resulting from age, gender or race [8].
In this study, direct measures of recumbent length of subjects with LECP involvement were found to significantly correlate with predicted recumbent length (Table 3). Residents in this study varied in age, gender and physical disability. Results show that, regardless of age, gender or physical disability, knee height is a strong predictor of recumbent length in non-ambulatory individuals with CP. Race differences could not be investigated since all subjects with the exception of one Portuguese child were Caucasian.
In conclusion, because of the characteristic short stature of this population, energy requirements are based on height-for-age or skeletal age rather than weight-for-age [3]. However, accurate measures of recumbent length are difficult to obtain especially in mobility-impaired individuals and those with LECP involvement. Using the knee height calliper and these prediction equations [11], most health-care professionals can obtain a reasonably reliable measure of recumbent length (with known errors) in mobility-impaired children and adults with CP. Recumbent length calculated using these equations can be compared to standard National Centre for Health Statistics growth charts [22].
|
ACKNOWLEDGMENTS
|
|---|
I thank the residents who participated in the study and their families as well as the Board of Directors and Staff of the Sunbeam Residential Development Centre.
Received July 1, 1997.
Accepted November 1, 1998.
|
REFERENCES
|
|---|
- Tobis JS, Saturen P, Larios G, Posniak A: Study of growth patterns in cerebral palsy.
Arch Phys Med Rehab
June:
475–481,
1961.
- Ruby DO, Matheny WD: Comments on growth of cerebral palsied children.
J Amer Diet Assoc
40:
525–527,
1962.
- Culley W, Middleton T: Caloric requirements of mentally retarded children with and without motor dysfunction.
J Pediat
75:
380–384,
1969.[Medline]
- Patrick J, Boland M, Stolski D, Murray GE: Rapid correction of wasting in children with cerebral palsy.
Devel Med Child Neurol
28:
734–739,
1986.[Medline]
- Cameron N: The methods of auxological anthropology. In Falkner F, Tanner JM (eds):
"Human Growth: A Comprehensive Treatise." New York: Plenum Press,
1986.
- Spender QW, Cronk CE, Charney EB, Stallings VA: Assessment of linear growth of children with cerebral palsy: Use of alternative measures to height or length.
Dev Med Child Neurol
31:
206–214,
1989.[Medline]
- Han TS, Lean ME: Lower leg length as an index of stature in adults.
Int J Obes Relat Metab Disord
20:
21–27,
1996.[Medline]
- Stevenson RD: Use of segmental measures to estimate stature in children with cerebral palsy.
Arch Paediatr Adolesc Med
149:
658–652,
1995.[Abstract]
- Johnson RK, Ferrara MS: Estimating stature from knee height for persons with cerebral palsy: An evaluation of estimation equations.
J Amer Diet Assoc
91:
1283–1284,
1991.
- Ekvall SW (ed.): Appendix 8—Skinfold grids—Children: Other anthropometry standards. In
"Pediatric Nutrition in Chronic Diseases and Developmental Disorders: Prevention, Assessment and Treatment." New York, NY: Oxford University Press, pp
489–491,
1993.
- Chumlea WC, Guo SS, Steinbaugh ML: Prediction of stature from knee height for black and white adults and children with application to mobility-impaired or handicapped persons.
J Am Diet Assoc
94:
1385–1391,
1994.[Medline]
- Chumlea WC: Methods of Assessing Body Composition in Nonambulatory Persons. In
"Body Composition Assessment in Youths and Adults." Report of the Sixth Ross Conference on Medical Research. Columbus, OH: Ross Laboratories, pp
86–90,
1984.
- Frisancho AR:
"Anthropometric Standards for the Assessment of Growth and Nutritional Status." Ann Arbor, MI: The University of Michigan Press,
1990.
- Hogan SE, Evers SE: A nutrition rehabilitation program for individuals with severe physical and developmental disabilities.
J Am Diet Assoc
97:
162–166,
1997.[Medline]
- Health and Welfare Canada:
"Nutrition Recommendations: The Report of the Scientific Review Committee." Ottawa: Health and Welfare,
1990.
- Hogan SE, Evers SE: Dietary intake and anthropometric status of persons with severe physical and developmental disabilities.
Nutr Res
14:
1473–1489,
1994.
- Lohman TG, Roche AF, Martorell R:
"Anthropometric Standardization Reference Manual." Champaign, IL: Human Kinetics Books,
1988.
- Chumlea WC, Guo S, Kuczmarski RJ, Johnson CL, Leahy CK: Reliability of anthropometric measurements in the Hispanic Health and Nutrition Examination Survey (HHANES 1982–1984).
Am J Clin Nutr
51:
902S–907S,
1990.[Abstract/Free Full Text]
- SPSS Inc:
"SYSTAT 7.0 Statistics Software", Chicago, IL,
1997.
- Chumlea WC, Roche AF: Nutritional anthropometric assessment of nonambulatory persons using recumbent techniques.
Am J Phys Anthropol
63:
146–151,
1984.
- Neter J, Wasserman W, Kuter MH:
Applied Linear Statistical Models. Regression, Analysis of Variance, and Experimental Design. Homewood, IL: IRWIN, pp
450–451,
1990.
- Hamill PVV, Drizd TA, Johnson CL, Reed RB, Roche AF, Moore WM: Physical growth: National Center for Health Statistics percentiles.
Am J Clin Nutr
32:
607–629,
1979.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
K. W. Cullen, T. Baranowski, L. Rittenberry, and N. Olvera
Social-environmental influences on children's diets: results from focus groups with African-, Euro- and Mexican-American children and their parents
Health Educ. Res.,
October 1, 2000;
15(5):
581 - 590.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
