Low Weight, Morbidity, and Mortality in Children WithCerebral Palsy: New Clinical Growth Charts

WHAT’S KNOWN ON THIS SUBJECT: Weight-for-age percentiles ofchildren with cerebral palsy are lower than in the generalpopulation. This is especially true in children with more severemotor dysfunction. Poor growth, loosely defined, is associatedwith increased hospitalization and school absences.

WHAT THIS STUDY ADDS: This article reports evidence-basedthresholds for low weight and provides estimates of associatedincreases in mortality risk. These estimates are illustrated onnew clinical growth charts for children with cerebral palsy,stratified according to gender and Gross Motor FunctionClassification System levels.

abstractOBJECTIVE: To determine the percentiles of weight for age in cerebralpalsy according to gender and Gross Motor Function ClassificationSystem (GMFCS) level and to identify weights associated with negativehealth outcomes.

PATIENTS AND METHODS: This study consists of a total of 102 163measurements of weight from 25 545 children with cerebral palsy whowere clients of the California Department of Developmental Servicesfrom 1988 through 2002. Percentiles were estimated using generalizedadditive models for location, scale, and shape. Numbers of comorbidi-ties were compared using t tests. The effect of low weight on mortalitywas estimated with proportional hazards regression.

RESULTS: Weight-for-age percentiles in children with cerebral palsyvaried with gender and GMFCS level. Comorbidities were more com-mon among those with weights below the 20th percentile in GMFCSlevels I through IV and level V without feeding tubes (P � .01). ForGMFCS levels I and II, weights below the 5th percentile were associatedwith a hazard ratio of 2.2 (95% confidence interval: 1.3–3.7). For chil-dren in GMFCS levels III through V, weights below the 20th percentilewere associated with a mortality hazard ratio of 1.5 (95% confidenceinterval: 1.4–1.7).

CONCLUSIONS: Children with cerebral palsy who have very lowweights have moremajor medical conditions and are at increased riskof death. The weight-for-age charts presented here may assist in theearly detection of nutritional issues or other health risks in thesechildren. Pediatrics 2011;128:e299–e307

AUTHORS: Jordan Brooks, MPH,a,b Steven Day, PhD,a

Robert Shavelle, PhD,a and David Strauss, PhD,a

aLife Expectancy Project, San Francisco, California; andbDepartment of Biostatistics, University of California, Berkeley,Berkeley, California

KEY WORDSgrowth charts, cerebral palsy, mortality, morbidity

ABBREVIATIONSCDC—Centers for Disease Control and PreventionGMFCS—Gross Motor Function Classification SystemCDER—Client Development Evaluation Report

www.pediatrics.org/cgi/doi/10.1542/peds.2010-2801

doi:10.1542/peds.2010-2801

Accepted for publication Apr 7, 2011

Address correspondence to Jordan Brooks, MPH, LifeExpectancy Project, 1439 17th Ave, San Francisco, CA 94122.E-mail: brooks@lifeexpectancy.org

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2011 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: The authors have indicated they haveno personal financial relationships relevant to this article todisclose.

COMPANION PAPER: A companion to this article can be found onpage e436 and online at www.pediatrics.org/cgi/doi/10.1542/peds.2011-1472.

ARTICLES

PEDIATRICS Volume 128, Number 2, August 2011 e299by guest on January 22, 2016Downloaded from

Growth charts are standard tools formonitoring pediatric growth, develop-ment, and overall health. They containestimated weight-for-age percentilesbased on a reference population. If achild’s weight falls well outside agenorms, it may raise clinical concern.The standard charts in pediatric prac-tices are those of the Centers for Dis-ease Control and Prevention (CDC) forboys and girls in the US general popu-lation.1 These charts may not be help-ful for children with cerebral palsy,whose growth patterns may be mark-edly different from those of the gen-eral pediatric population.2–15

Krick et al2 produced the first cerebralpalsy–specific growth charts based onthe weight and stature of children withsevere quadriplegia. The North Ameri-can Growth in Cerebral Palsy ResearchCollaboration has produced curves forseveral other growth parameters, in-cluding weight, knee height, upper-arm length, mid–upper arm musclearea, triceps skinfold, and subscapu-lar skinfold.3 Recently, Day et al4 con-structed a series of height, weight, andBMI charts stratified by motor andfeeding skills.

Some researchers and practitionershave raised concerns over the useful-ness of growth charts as diagnostic orprognostic tools. One concern is thatexisting charts are descriptive refer-ences rather than prescriptive stan-dards, showing how a particular groupof children grew rather than how aparticular child should grow. Recently,the World Health Organization at-tempted to address this concern byconstructing growth charts based ona select sample of “healthy childrenliving under conditions likely to favorachievement of their full geneticgrowth potential [and whose moth-ers] engaged in fundamental health-promoting practices, namely breast-feeding and not smoking.”16 Whetherthe resulting World Health Organiza-

tion charts are truly prescriptive orin any sense more useful than thedescriptive CDC reference curves isan open question.17

Whether such a select sample forcerebral palsy growth curves wouldbe helpful is far from clear. Cerebralpalsy growth patterns are depen-dent on the severity of disabilities,4

and children with more severe dis-abilities are likely to have signifi-cant comorbidities. Thus, defining a“healthy” cerebral palsy populationbecomes a difficult and somewhatarbitrary task. Perhaps a more rea-sonable approach to growth-chartconstruction is to begin with a clini-cally appropriate reference popula-tion to the construct charts thenanalyze empirical data to determinegrowth thresholds that are associ-ated with good or bad health out-comes in that population. This ap-proach was taken by Stevensonet al3 and Samson-Fang et al,5 whoshowed that poor growth, measuredby a combination of weight andother parameters, was associatedwith increased health care useand decreased social-participationoutcomes.

The following were the goals of thepresent study:

1. Estimate reference weight-for-agepercentiles for children with cere-bral palsy at each Gross MotorFunction Classification System(GMFCS) level.

2. Test for associations betweenweight for age and morbidity andmortality and quantify those thatare significant.

3. Construct cerebral palsy growthcharts that clearly illustrate poten-tially unhealthy low weights.

4. Design the charts to mimic the CDCcharts so that they may easily beintegrated into existing clinicalpractice.

METHODS

Inclusion and Exclusion Criteria

The study population included childrenwith cerebral palsy who were clientsof the California Department of Devel-opmental Services between January1988 and December 2002. Clients of theDepartment of Developmental Ser-vices are assessed annually with theClient Development Evaluation Report(CDER).18 This report contains over 200medical, functional, behavioral, andcognitive items. For each client, a teamheaded by a pediatric neurologistmakes medical diagnoses, includingthe assessment of cerebral palsy,whereas functional items (crawling,walking, and feeding, etc) may beassessed by other professionalsfamiliar with that aspect of the client’sdevelopment.

Children who had a CDER with an Inter-national Classification of Disease,Ninth Revision19 code for any of severaldegenerative conditions or condi-tions acquired after infancy wereexcluded from all analyses. Theinclusion-exclusion algorithm isshown in Fig 1.

Gross Motor Classification

Growth patterns in children with cere-bral palsy vary with motor and feedingabilities.4 The classification system formotor disability in children with cere-bral palsy usedmost commonly in clin-ical and research settings is the 5-levelGMFCS20:

I. Walks without limitations

II. Walks with limitations

III. Walks using a hand-held mobilitydevice

IV. Self-mobility with limitations, mayuse powered mobility

V. Transported in a manual wheelchair

The specific criteria for each level areage dependent and were developedwith the intent that children would

e300 BROOKS et alby guest on January 22, 2016Downloaded from

maintain the same GMFCS levelthroughout childhood and adoles-cence. Wood and Rosenbaum21 docu-mented the reliability of GMFCS fromthe age of 2 to 12 years to be 0.79.

For the present study, the age-specificGMFCS criteria were approximatedwith functional items from the CDERbased on the classification algorithmused in Krach et al.25 Functional-itemdata have been independently validat-ed22–24 and have interrater reliabilityexceeding 0.85.24 Because the pres-ence of a feeding tube may affectgrowth, GMFCS level V was subdividedinto children who fed orally without afeeding tube (GMFCS V-NT) and thosewho had a feeding tube (GMFCS V-TF).The vast majority of feeding tubes(well over 90%) are gastrostomytubes. In the United States, nasogas-tric feeding is rarely used for ex-tended periods.

Some children gained or lost abilitiesand were represented in 1 or moreGMFCS levels over the course of the

study. A relatively small number ofchildren (�1%) were not assigned toany GMFCS level because they hadmissing functional assessments or be-cause they had rare combinations ofabilities and disabilities. These chil-dren were excluded from additionalanalysis.

Weight-for-Age Growth Curves

Weight measures for the CDER weretaken directly or, in some cases, re-ported by a parent or other caregiver.Discrepancies between weights re-corded on the CDER and those in anindividual’s actual medical recordswere found in 9% of a random sample,but these were small enough to be ig-nored as immaterial.22

For approximately one-third of theassessments, weight values werecarried over from a previous CDER.Because such observations do notaccurately represent age-specificweights, we excluded them from ad-ditional analysis. Few individuals had

recorded weights that were wellabove or below biologically plausiblelimits. In addition, some assess-ments suggested extreme rates ofweight change; for example, a 5-year-old child gaining 50 pounds during a1-year period. Together, all suchdoubtful observations made up lessthan 0.1% of our study sample andwere excluded from additionalconsideration.

Gender- and GMFCS-specific referencepercentiles (growth curves) were esti-mated for children with cerebral palsywho were aged 2 to 20 years (data onchildren aged 1 to 25 years were usedto improve the precision of weight per-centiles at ages 2 and 20). This agerange was selected to match the stan-dard CDC charts. Percentiles were es-timatedwith generalized additivemod-els for location, scale, and shape(GAMLSS), with a Box-Cox power expo-nential distribution. This is a semipa-rametric statistical-modeling tech-nique that allows estimation ofage-specific percentiles and z scores.26

Models were fit in accordance withWorld Health Organization methodol-ogy using cubic smoothing splines.Model selection was based on penal-ized maximum likelihood.27

Morbidity

Separately for each GMFCS level, themean number of chronic major medi-cal conditions was calculated withinweight-for-age quintiles. According tothe Department of Developmental Ser-vices, chronic major medical condi-tions are “major, chronic medicalproblems that limit or impede the cli-ent or significantly impact the provi-sion of service” and “include, but arenot limited to, diabetesmellitus, hyper-tension, congenital or arterioscleroticheart disease, upper respiratory infec-tions, etc.”18 Differences in the meannumber of chronic major medical con-ditions, for people in the extreme

Persons with CP, 1988–200248 447

No exclusion conditions40 413

Age 1–25 y29 463

Classified by GMFCS29 264

Weight percentile estimation29 246

Morbidity and mortality analysis25 545

Exclusion conditions: 8034Chromosomal anomalies: 818

Degenerative conditions: 4295Traumatic brain injury,

motor vehicle accident, near drowning, or other acquired injury: 2034

Autism: 532Cancer: 355

Age <1 or >25 y10 950

Missing motor function or not classified by GMFCS

199

Missing or implausible weight values18

Age < 2 or Age > 203701

FIGURE 1Study population inclusion-exclusion algorithm. CP indicates cerebral palsy.

ARTICLES

PEDIATRICS Volume 128, Number 2, August 2011 e301by guest on January 22, 2016Downloaded from

weight quintiles versus those in the 3middle quintiles were assessed with ttests.

Mortality

Electronic death records were ob-tained from the California Departmentof Health Services. Individuals surviv-ing 3 or more years after their lastweight measure were censored at 3years. All individuals surviving to De-cember 31, 2002, were administra-tively censored at that date.

We used Cox proportional hazards re-gression analysis with time-varying co-variates28 to relate survival time toweight percentiles. This enabled us tocontrol for other variables, such asfeeding skills, that might confound ormodify the effect of low weight on mor-tality. Separate models were fit forGMFCS levels I and II and GMFCS levelsIII through V because children in thesegroups tend to be different with re-spect to functional skills beyond grossmotor function, feeding and cognition,age-specific weight values, age-specific mortality patterns, and secu-lar trends. Low-weight cutoffs were se-lected on the basis of maximumlikelihood. Data were managed in SASversion 9.12,29 and analyzed by using Rversion 2.9.30

RESULTS

Descriptive Statistics

The study population included 25 545children (56% male, aged 2–20 years)who contributed 102 163 weight mea-surements (Table 1). Age, gender, pre-maturity or low birth weight, and cal-endar year of CDER did not varysignificantly by GMFCS level. The mostfrequent level in our study populationwas GMFCS level II (31%). This was fol-lowed by levels IV (24%), V (17%), I(14%), and III (14%). The proportionwith severe feeding and cognitive dis-abilities increased with increasingGMFCS level. For example, 2% of chil-dren in GMFCS level I were either tubefed or orally fed by others comparedwith 42% of children in GMFCS level IVand 90% inGFMCS level V. Eleven percentof children in GMFCS level I had severe orprofound mental retardation comparedwith 50% of children in GMFCS level IVand 73% in GFMCS level V.

Weight-for-Age

In all but the most severe group(GMFCS level V), weight-for-age dataexhibited nonlinear dependence onage, with a visible growth spurt be-tween ages 9 and 13 years and plateauin late adolescence. For each GMFCSlevel, weight-for-age percentiles for

boys and girls were similar up to aboutthe age of 15 years. Girls plateauedearlier than boys, and between theages of 15 and 20 years boys tended toweigh more than girls. Gender differ-ences were smaller in the more se-verely affected groups. For example, atage 20 years the difference in medianweights for boys and girls in GMFCSlevel I was 7.3 kg; the difference wasonly 1.8 kg in the GMFCS V-TF group.Figure 2A showsa scatter plot ofweight-for-age data in boys from GMFCS level I,alongwith estimatedweight-for-age per-centiles and the CDCpercentiles for boysin the general population. The 90th per-centile in GMFCS level I closely trackedthat of the general population. The me-dianwas lower, and thedifference inme-dians increased with age. The 10th per-centile was markedly lower at all ages.Children in GMFCS level V exhibitedmorelinear growth patterns (ie, no growthspurt), with a plateau in late adoles-cence (Fig 2B).

Morbidity

The mean number of chronic majormedical conditions increased mod-estly with GMFCS level. The most strik-ing marker for chronic medical condi-tions was the presence of a feedingtube. For example, children in GMFCSV-TF had, on average, twice as many

TABLE 1 Study Population

GMFCS Level

I II III IV V-NTa V-TFb

No. of assessments 14 030 31 808 13 994 24 744 11 919 5668Male, % 61 57 55 57 54 54Age, median (interquartile range), y 4.5 (1.7–6.8) 4.5 (2.7–6.4) 4.6 (3.2–6.5) 4.4 (2.6–6.4) 3.9 (2.2–6.3) 4.1 (2.4–6.4)Has a feeding tube, % 1 2 5 9 0 100Orally fed by others, % 1 2 11 33 85 0Has severe (IQ 20–34) or profound(IQ� 20) mental retardation, %

11 22 34 50 68 84

Low birth weight (�2500 g) or pretermlabor (�37 wk), %

25 28 35 31 23 22

Weight, median (interquartile range), kg 28 (16–50) 27 (19–44) 26 (18–39) 21 (15–32) 18 (13–26) 23 (16–31)

Observations are 102 163 CDERs from 25 545 subjects with cerebral palsy, who received services from the California Department of Developmental Services between 1988 and 2002. Somechildren contributed observations to more than 1 group.a Is fed orally.b Dependent on a feeding tube.

e302 BROOKS et alby guest on January 22, 2016Downloaded from

major medical conditions as those inGMFCS V-NT (Fig 3). Among children inthe GMFCS levels I through IV and thelevel V-NT groups, those with weightsbelow the 20th percentile had moremajor medical conditions than chil-dren whose weights fell in the middle60% (P� .01). In contrast, children inGMFCS V-TF who had weights below the20th percentile had fewermajormedical

conditions than the middle 60% (P �.0001). The mean number of major med-ical conditions for children with weightsabove the 80th percentile was not signif-icantly different from that of childrenwith weights in the middle 60%.

Mortality

Study participants contributed a totalof 166 327 person-years of follow-up

time. There were 1496 deaths, for anoverall mortality rate of 9 deaths per1000 person-years. For GMFCS levels IIIthrough V, children with weight for agebelow the 20th percentile had signifi-cantly higher mortality rates com-pared with children with weight forage in the 20th to 80th percentile range(P� .01) (Fig 4). The excess death ratein this lowest quintile increased steadilywith GMFCS level (0.3 per 1000 person-years [GMFCS level I] up to 26 per 1000person-years [GMFCS V-TF]). Weightabove the 80th percentile was not asso-ciated with differential mortality.

Because mortality rates in childrenwith cerebral palsy vary strongly withthe severity of disabilities, for model-ing purposes the data were dividedinto 2 groups: mild to moderate(GMFCS levels I and II) and severe(GFMCS levels III through V). Withineach group, we fit unadjusted Cox pro-portional hazard regressionmodels andalsomore complexmodelswith baselinehazard functions stratified by GMFCSlevel and adjusted for time-varying cova-riates, including age, gender, mobility,feeding, mental retardation, low birthweight or prematurity, and calendaryear. Unadjusted and adjusted hazardratios from the models are given in Ta-bles 2 and 3. For GMFCS levels I and II,weight below the 5th percentile wasassociated with an adjusted hazard ra-tio of 2.2 (95% confidence interval: 1.3–3.7). For GMFCS levels III through V,weight below the 20th percentile wasassociated with increased mortality(adjusted hazard ratio: 1.5 [95% confi-dence interval: 1.4–1.7]). The relativemortality risk associated with lowweight did not vary with gender, age,or calendar year. Sensitivity analysesconfirmed that the pattern of missingage-specificweightswere noninforma-tive with respect to survival and there-fore did not influence these results.

These mortality risk research findingsare illustrated on newly developed

FIGURE 2Weight-for-age data and fitted percentiles.

FIGURE 3Mean number of chronic major medical conditions according to weight quintile. a Significant differ-ence from the middle 3 quintiles (P� .01).

ARTICLES

PEDIATRICS Volume 128, Number 2, August 2011 e303by guest on January 22, 2016Downloaded from

growth charts with shaded weight-for-age values where mortality risk issignificantly increased. Fig 5 showsweight-for-age charts for girls inGMFCS level IV and boys in GMFCS levelV who are tube fed. The new charts arestyled after the standard CDC chartsand include designated areas to re-cord patient name, dates, parentalheight and weight, and general notes.The full set of growth charts is

available at www.lifeexpectancy.org/articles/newgrowthcharts.shtml.

DISCUSSION

Among children in GMFCS levels Ithrough IV, and level V who are nottube fed, low weight was, as expected,associated with an increase in thenumber of concurrent chronic majormedical conditions. Why very lowweight is associated with fewer major

medical conditions for children in theGMFCS V-TF group is unclear. It may bethat some very-low-weight childrenhave feeding tubes placed strictly toaddress weight issues even in theabsence of comorbidities, whereasheavier children have feeding tubes toreduce risks from aspiration pneumo-nias or to address other medical is-sues. Additional research is necessaryto fully understand this.

The concept of failure to thrive is usedfrequently in general pediatric prac-tice without much evidence regardingits associations with health out-comes.31 It is interesting to note thatour evidence-based GMFCS levels I andII low-weight threshold (ie, the 5th per-centile) is broadly consistent with an-thropometric failure-to-thrive crite-ria.32 This threshold also is consistentwith studies of the general populationthat have found the 10th percentile ofadult BMI to be associated with mod-estly increased mortality.33,34 That thelow-weight percentile threshold forGMFCS levels I and II is lower than thatfor GMFCS levels III through V (5th ver-sus 20th percentile) reflects the factthat children in GMFCS levels I and IIweighmore than those in GMFCS levelsIII through V.

It may seem counterintuitive thathigh weights were not associatedwith increased mortality or morbid-ity, particularly because obese chil-dren may be subject to additionalcomorbidities and may require mod-ified care regimes. The most likelyexplanation may be that the effectsof overweight or obesity do not no-ticeably increase mortality risk untiladulthood. The impact of childhoodobesity on adult outcomes in peoplewith developmental disabilities re-mains an open question.

The proper clinical interpretation ofthe risks discussed here deserves ad-ditional comment. A practicing clini-cian may ask, “Do these risks apply to

FIGURE 4Crudemortality rates according to weight quintile. a Significant difference from themiddle 3 quintiles(P� .05).

TABLE 2 Cox Regression Results for Children in GMFCS Levels I and II

Hazard Ratio for Death and95% Confidence Interval

Unadjusted Adjusteda

Weight below the 5th percentileb 3.2 (1.9–5.3) 2.2 (1.3–3.7)

Based on 45 838 evaluations of 13 118 individuals in GMFCS levels I or II. The cohort experienced 125 deaths over 76 733person-years of follow-up.a Adjusted for gender, age, stair climbing ability, mental retardation, feeding, and low birth weight or prematurity.b GMFCS- and age-specific 5th percentile.

TABLE 3 Cox Regression Results for Children in GMFCS Levels III Through V

Hazard Ratio for Death and95% Confidence Interval

Unadjusted Adjusteda

Weight below the 20th percentileb 1.6 (1.4–3.8) 1.5 (1.4–1.7)

Both models account for functional skills that vary over time (ie, time-varying covariates). The baseline hazard functionswere stratified by GMFCS level. Based on 56 325 evaluations of 14 688 individuals in GMFCS levels III through V. The cohortexperienced 1371 deaths over 89 594 person-years of follow-up.a Adjusted for gender, age, head-lifting ability, feeding, mental retardation, low birth weight or prematurity, and calendaryear.b GMFCS- and age-specific 20th percentile.

e304 BROOKS et alby guest on January 22, 2016Downloaded from

my patient? And, if so, for how long?”The evidence presented here isgenerally applicable to all childrenwith cerebral palsy, but additionalpatient-specific features should al-ways be considered. One potentiallybenign reason for low weight maysimply be small parental stature. Incases where benign etiology hasbeen ruled out, the excess risks as-sociated with low weight should beinterpreted as persistent for as longas the child remains in the low-weight category. On the other hand,for the reasons stated above clini-cians should not take the findings ofthis study to infer that overweight isnot a significant health risk in chil-dren with cerebral palsy.

The primary limitation of the study isthe lack of information regarding theetiology of low weight. Low weight is aknown marker for nutritional deficitsand general frailty, which is a reason-able mechanism for increased mor-bidity andmortality. On the other hand,children may lose weight or have trou-ble gaining it as a result of chronic oracute illness. A secondary limitation isthe apparent underrepresentation ofGMFCS level I (14%) in our study popu-lation. In other population-based cere-bral palsy registries, the proportion ofthose in level I ranged from 30% to40%.35–37 It may be that the most mildlyaffected individuals in California dis-proportionately choose not to seeklong-term services from the Depart-

ment of Developmental Services be-cause of a perceived lack of need.Thus, our GMFCS level I findingsmay bevalid only for children with ongoingneeds for services.

Thestudyhasanumberof strengths. Thefindings represent the first evidence-based link between lowweight andmor-tality risk in childrenwith cerebral palsy.The large sample size allowed percentileestimates that are robust to modelingassumptions. For example, the chartspresented here have estimated weight-for-age percentiles that are consistentwith those in Day et al.4 The GAMLSSgrowth chart methodology used here isconsistent with that of both the CDCand World Health Organization. It al-

FIGURE 5Clinical growth charts for children with cerebral palsy.

ARTICLES

PEDIATRICS Volume 128, Number 2, August 2011 e305by guest on January 22, 2016Downloaded from

lows for the calculation of both per-centiles and z scores, which have be-come popular in both the researchand clinical communities. Finally, theuse of a simple and reliable mea-sure, weight, may have practical ben-efits over using a more detailed butpossibly unreliable combination ofmeasures, for example stature orskinfold thickness, in children withcerebral palsy.

CONCLUSIONS

Evidence-based decision-making iscrucial in clinical and care-planningsettings. Without sound empirical evi-

dence to rely on, clinicians may beforced to make important treat-ment decisions on the basis of subjec-tive impressions. The extent to whichtoday’s clinicians can practiceevidence-based medicine dependslargely on the availability of toolsdesigned with these principles inmind.

The new cerebral palsy growthcharts presented here are the first togive a visual indication of potentiallyunhealthy weights. GMFCS is rela-tively stable throughout childhoodand adolescence and thus provides auseful stratification scheme from

which to monitor a particular child’sgrowth. To facilitate integration intocurrent clinical practice, our growthcharts are styled in accordance withthose of the CDC and include desig-nated areas to record patient char-acteristics and clinical notes. Ulti-mately, the utility of the charts willbecome more apparent as they areused in clinical practice.

ACKNOWLEDGMENTSProvision of data from the CaliforniaDepartments of Developmental Dis-abilities and Health Services is grate-fully acknowledged.

REFERENCES

1. Kuczmarski RJ, Ogden CL, Grummer-StrawnLM, et al. CDC growth charts: United States.Adv Data. 2000;314(314):1–27

2. Krick J, Murphy-Miller P, Zeger S, Wright E.Pattern of growth in children with cerebralpalsy. J Am Diet Assoc. 1996;96(7):680–685

3. Stevenson RD, Conaway M, Chumlea WC, etal. Growth and health in children withmoderate-to-severe cerebral palsy. Pediat-rics. 2006;118(3):1010–1018

4. Day SM, Strauss DJ, Vachon PJ, RosenbloomL, Shavelle RM, Wu YW. Growth patterns in apopulation of children and adolescentswith cerebral palsy. Dev Med Child Neurol.2007;49(3):167–171

5. Samson-Fang L, Fung E, Stallings VA, et al.Relationship of nutritional status to healthand societal participation in children withcerebral palsy. J Pediatr. 2002;141(5):637–643

6. Tobis JS, Saturen P, Larios G, Posniak AO.Study of growth patterns in cerebral palsy.Arch Phys Med Rehabil. 1961;42:475–481

7. Krick J, Van Duyn MA. The relationship be-tween oral-motor involvement and growth:a pilot study in a pediatric population withcerebral palsy. J Am Diet Assoc. 1984;84(5):555–559

8. Stevenson RD, Roberts CD, Vogtle L. The ef-fects of non-nutritional factors on growth incerebral palsy. Dev Med Child Neurol. 1995;37(2):124–130

9. Motion S, Northstone K, Emond A, Stucke S,Golding J. Early feeding problems in chil-dren with cerebral palsy: weight and neuro-developmental outcomes. Dev Med ChildNeurol. 2002;44(1):40–43

10. Fung EB, Samson-Fang L, Stallings VA, et al.

Feeding dysfunction is associated with poorgrowth and health status in children withcerebral palsy. J Am Diet Assoc. 2002;102(3):361–373

11. Shapiro BK, Green P, Krick J, Allen D, CaputeAJ. Growth of severely impaired children:neurological versus nutritional factors. DevMed Child Neurol. 1986;28(6):729–733

12. Stallings VA, Charney EB, Davies JC, CronkCE. Nutritional status and growth of chil-dren with diplegic or hemiplegic cerebralpalsy. Dev Med Child Neurol. 1993;35(11):997–1006

13. Stallings VA, Charney EB, Davies JC, CronkCE. Nutrition-related growth failure ofchildren with quadriplegic cerebralpalsy. Dev Med Child Neurol. 1993;35(2):126 –138

14. Zainah SH, Ong LC, Sofiah A, Poh BK, HussainIH. Determinants of linear growth in Malay-sian children with cerebral palsy. J Paedi-atr Child Health. 2001;37(4):376–381

15. Shim ML, Moshang T Jr, Oppenheim WL,Cohen P. Is treatment with growth hor-mone effective in children with cerebralpalsy? Dev Med Child Neurol. 2004;46:569 –571

16. de Onis M, Onyango A, Borghi E, et al, eds.WHO Child Growth Standards: Length/Height-for-Age, Weight-for-Age, Weight-for-Length, Weight-for-Height and Body MassIndex-for-Age: Methods and Development.Geneva, Switzerland: World Health Organi-zation Press; 2006

17. Mei Z, Ogden CL, Flegal KM, Grummer-Strawn L. Comparison of the prevalence ofshortness, underweight, and overweightamong US children aged 0 to 59 months byusing the CDC 2000 and the WHO 2006

growth charts. J Pediatr. 2008;153(5):592–594

18. California Department of DevelopmentalServices. Client Development EvaluationReport (CDER). Sacramento, CA: Califor-n ia Department of Deve lopmenta lServices; 1986

19. World Health Organization. InternationalStatistical Classification of Diseases, Inju-ries and Causes of Death, Ninth Revision.Geneva, Switzer land: World HealthOrganization; 1977

20. Palisano R, Rosenbaum P, Walter S, Rus-sell D, Wood E, Galuppi B. Developmentand reliability of a system to classifygross motor function in children with ce-rebral palsy. Dev Med Child Neurol. 1997;39(4):214 –223

21. Wood EP, Rosenbaum PL. The Gross MotorFunction Classification System for cerebralpalsy. Dev Med Child Neurol. 2000;42(5):292–296

22. Citygate Associates, LLC. Independent Eval-uation of the Department of DevelopmentalServices’ Community Placement Practices:Final Technical Report. Sacramento, CA:Citygate Associates, LLC; 1998

23. Arias M, Ito E, Takagi N. Concurrent valid-ity of the client development and evalua-tion report. In: Silverstein AB, ed. PacificState Archives VIII. Pomona, CA: Universityof California at Los Angeles, Developmen-tal Disabilities Immersion Program; 1983:28 –33

24. Harris CW, Eyman RK, Mayeda T. An Interra-ter Reliability Study of the Client Develop-ment Evaluation Report: Final Report to theCalifornia Department of DevelopmentalServices. UCLA Mental Retardation Re-

e306 BROOKS et alby guest on January 22, 2016Downloaded from

search Center. Pomona, CA: LantermanState Hospital; 1983

25. Krach LE, Kriel RL, Day SM, Strauss DJ.Survival of individuals with cerebral palsyrece i v i ng con t i nuous in t ra theca lbaclofen treatment: a matched-cohortstudy. Dev Med Child Neurol. 2010;52(7):672– 676

26. Stasinopoulos DM, Rigby RA. Generalizedadditive models for location scale andshape (GAMLSS) in R. Stat Softw 2007; 23(7):1–46

27. Borghi E, de Onis M, Garza C, et al. Construc-tion of the World Health Organization childgrowth standards: selection of methods forattained growth curves. Stat Med. 2006;25(2):247–265

28. Cox DR, Oakes D. Analysis of Survival Data.

London, United Kingdom: Chapman and Hall;1984

29. SAS/STAT Software [computer program].Release 9.1. Cary, NC: SAS Institute; 2001

30. R Development Core Team. R: A Languageand Environment for Statistical Computing.Vienna, Austria: R Foundation for StatisticalComputing; 2008

31. Wilcox WD, Nieburg P, Miller DS. Failure tothrive: a continuing problem of definition.Clin Pediatr (Phila). 1989;28(9):391–394

32. Olsen EM, Petersen J, Skovgaard AM, et al.Failure to thrive: the prevalence and con-currence of anthropometric criteria in ageneral infant population. Arch Dis Child.2007;92(2):109–114

33. Calle EE, Thun MJ, Petrelli JM, Rodriguez C,Heath CW Jr. Body-mass index andmortality

in a prospective cohort of U.S. adults. N EnglJ Med. 1999;341(15):1097–1105

34. Fontaine KR, Redden DT, Wang C, Westfall AO,Allison DB. Years of life lost due to obesity.JAMA. 2003;289(2):187–193

35. Nordmark E, Hagglund G, Lagergren J. Cere-bral palsy in southern Sweden II: gross mo-tor function and disabilities. Acta Paediatr.2001;90(11):1277–1282

36. Howard J, Soo B, Graham HK, et al. Cerebralpalsy in Victoria: motor types, topographyand gross motor function. J Paediatr ChildHealth. 2005;41(9–10):479–483

37. Gorter JW, Rosenbaum PL, Hanna SE, et al.Limb distribution, motor impairment, andfunctional classification of cerebralpalsy. Dev Med Child Neurol. 2004;46(7):461– 467

ARTICLES

PEDIATRICS Volume 128, Number 2, August 2011 e307by guest on January 22, 2016Downloaded from

DOI: 10.1542/peds.2010-2801; originally published online July 18, 2011; 2011;128;e299Pediatrics

Jordan Brooks, Steven Day, Robert Shavelle and David StraussClinical Growth Charts

Low Weight, Morbidity, and Mortality in Children With Cerebral Palsy: New  

ServicesUpdated Information &

/content/128/2/e299.full.htmlincluding high resolution figures, can be found at:

References

/content/128/2/e299.full.html#ref-list-1at:This article cites 28 articles, 3 of which can be accessed free

Citations /content/128/2/e299.full.html#related-urls

This article has been cited by 6 HighWire-hosted articles:

Subspecialty Collections

/cgi/collection/neurologic_disorders_subNeurologic Disorders

/cgi/collection/neurology_subNeurologythe following collection(s):This article, along with others on similar topics, appears in

Permissions & Licensing

/site/misc/Permissions.xhtmltables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints /site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2011 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on January 22, 2016Downloaded from

DOI: 10.1542/peds.2010-2801; originally published online July 18, 2011; 2011;128;e299Pediatrics

Jordan Brooks, Steven Day, Robert Shavelle and David StraussClinical Growth Charts

Low Weight, Morbidity, and Mortality in Children With Cerebral Palsy: New  

  /content/128/2/e299.full.html

located on the World Wide Web at: The online version of this article, along with updated information and services, is

 

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2011 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on January 22, 2016Downloaded from