Distressedneighborhoods andchild disability rates:analyses of 157 000school-age children

Michael E Msall* MD, Comer and LaRabida Children’sHospitals, Chicago, Illinois;Roger C Avery PhD, Population Studies and Training Center,Brown University, Providence, Rhode Island;Emily R Msall, Fellow Citizen Schools of Boston, Mattapan,Massachusetts;Dennis P Hogan PhD, Population Studies and TrainingCenter, Brown University, Providence, Rhode Island, USA.

*Correspondence to first author at University of ChicagoPritzker School of Medicine, 5841 S. Maryland AvenueMC0900, Chicago, IL 60637, USA.E-mail: mmsall@peds.bsd.uchicago.edu

The aim of this study was to assess rates of childhooddisability as indicated by functional limitation of motor,sensory, or self-care skills in children living in severely-distressed neighborhoods. For a neighborhood in the USCensus Track to be considered severely distressed, three of thefollowing four characteristics need to be present: >27% ofchildren live in poverty, >23% high school drop-out rate,>34% male unemployment rate, and >37% of householdsheaded by females alone. In the 2000 US Census, 157 000children between ages 5 and 15 years resided in the State ofRhode Island. Severely-distressed neighborhoods were foundin 12.6% of Rhode Island Census Tracks. These areasaccounted for 14.5% of the school-age population, 25% ofchildren with motor disabilities, 29% of children with self-care disabilities, and 14% of children with sensorydisabilities. For each increasing level of neighborhooddistress, rates of child disability increased. Child disabilityrates in moderately distressed neighborhoods were 3.7%,compared with 1.1% in advantaged neighborhoods. Childrenin distressed neighborhoods had disproportionately high ratesof disability, especially in motor and self-care functioning.Comprehensive interventions aimed at children in distressedneighborhoods are crucial to reduce health disparities forvulnerable children.

Several federal and state initiatives took place in the UnitedStates (US) during the 1990s that aimed to improve children’saccess to health insurance, and to hold schools accountablefor learning outcomes. The goal of these policies was to opti-mize children’s access to health services1 and to provide chil-dren with educational services that would ensure theirparticipation as continual learners with the skills necessary tocompete economically and to contribute as citizens.2,3 At thesame time, the US economy and employment grew atunprecedented rates. It was assumed that this economicgrowth would lessen discrepancies between rich and poorand positively impact on indicators of child disability.4

Childhood poverty and neighborhood disadvantage areinversely related to children’s health and developmental well-being.5–8 Studies by Duncan and Brooks-Gunn demonstratedthat ongoing poverty and lack of community supports have acumulative adverse impact on children’s health, mentalhealth, and social well-being.9 In their analysis of availablechild development literature, Shonkoff and Phillips conclud-ed that neighborhood context is critical for determiningwhether children live in caring and secure environments thathave adults at home and in the community who are able tomentor and nurture them.3 Keating and Hertzman demon-strated that societies with a high gradient of disparity betweenrich and poor have higher rates of adverse outcomes at allstages of human development: infancy, childhood, adoles-cence, adulthood, older age.10 However, there are limited dataat population-level to analyze this theme in middle child-hood.11 The 2000 US Census was specifically designed toreflect traditional measures of household composition, familyhealth, housing, and economic well-being; in addition, it wasthe first US Census to investigate child disability specifically.

The purpose of this article is to address the followingquestions: What is the relationship between indicators ofneighborhood distress and schoolchildren’s rates of motor,sensory, and self-care disability? How do children living inseverely-distressed neighborhoods fare in comparison withchildren living in less distressed neighborhoods? The overallhypothesis was that child disability rates would be higheramong schoolchildren living in distressed neighborhoods.This assumption is based on the complex relationship betweenneighborhood disadvantage, family poverty, preterm birth,access to health service, and injury.

Researchers selected the State of Rhode Island to examinebecause of this state’s efforts to regionalize programs for chil-dren with special needs and its commitment that all children,US citizens or immigrants, can obtain health insurance andspecial educational services. In addition, Rhode Island ranksin the top 10% for children receiving early interventiondevelopmental services from birth to 3 years of age.

MethodsChild disability indicators of distressed areas in Rhode Islandwere compared with those of the rest of the state. For thisanalysis, a domain of disability was defined as an area in whicha child has functional limitations compared with peers.12–14

To classify disability, two approaches were used. First, the fre-quency of children with either a sensory or motor or self-caredisability was enumerated. Second, the total number of chil-dren with any disability (whether in a single domain or withchallenges in multiple domains) were counted.

Survey questions from the 2000 US Census were used to

814 Developmental Medicine & Child Neurology 2007, 49: 814–817

Distressed Neighborhoods and Child Disability Rates Michael E Msall et al. 815

assess rates of sensory, motor, and self-care disability for chil-dren living in Rhode Island. Census data were investigated toassess specifically whether child disability among schoolchild-ren is concentrated in distressed neighborhoods. There wereseveral reasons to use the 2000 Census: (1) this was the firstUS Census in which child disability was queried in a uniformway for school-age children; (2) the short form is the coreCensus form and includes age, ethnicity, parental education,income, and household composition; (3) the long form includesparental reports of physical, developmental, and behavioralimpairments of children and was given to approximately20% of households; and (4) the 2000 Census provides a largeenough sample to allow analysis at neighborhood level.

In the US Census, four key questions address child disabil-ities. The first question focuses on sensory disability: ‘Doesthe child have blindness, deafness, or a severe vision or hearingimpairment?’. Second, the Census asks: ‘Does the child havea condition that limits one or more basic physical activities,such as walking, climbing stairs, reaching, lifting, or carryingan item?’. This question is used as an indicator of motor dis-ability. The third question: ‘Does the child have difficulty withdressing, bathing, or getting around inside the house?’, isdesigned to identify self-care disability. The fourth questionaddresses issues of attention, learning, and memory. Theresearchers’ initial analysis with this question revealed that thisquestion was not interpreted similarly across cultural, ethnic,and socioeconomic groups. Parents reported low rates oflearning, attention, and memory problems among Hispanicand African-American children in very poor neighborhoods,suggesting that in these children, specific and mild learningdisabilities were probably not well recognized, given the for-mat of this question. This is important because most childrenidentified in the Census as having disabilities only reportedattention, learning, and memory disorders. Because of these

concerns, the fourth question was not included in the cur-rent analysis.

This study used the definition of a severely distressedneighborhood as specified by the Population ReferenceBureau’s KIDS COUNT project, a report on the status andwell-being of children in the US.15 Distressed neighborhoodswere defined as having a Census Track (i.e. a geographicunit/neighborhood containing about 3000–4000 people,as defined by the US Census Bureau) with more than 27% ofchildren living in poverty, a high school drop-out rateexceeding 23%, male unemployment rate exceeding 34%,and households headed by females alone exceeding 37%.Each of Rhode Island’s 233 Census Tracks were analyzed todetermine which ones were distressed. Twenty-eight areaswere identified that met the criteria of a distressed neighbor-hood. In addition, five Census Tracks averaged 1 standarddeviation above the population mean for neighborhood dis-advantage using all four of the criteria of the PopulationReference Bureau’s KIDS COUNT report. These 33 CensusTracks were distributed as follows: 23 in Providence, four inWoonsocket, two in Pawtucket, two in Central Falls, one inCranston, and one in Newport. All neighborhoods exceptWoonsocket and Newport are contiguous with Providence,the capital city of the state.

In addition, this study examined correlations betweenrates of child disability and the degree of neighborhood dis-tress. The total number of distress factors were calculated,using summary z codes, for each child and rated neighbor-hoods as having: (1) no distress (≤–1z); (2) minimal distress(–0.99 to –0.5z); (3) some distress (–0.5 to 0.5z); (4) moder-ate distress (0.5 to 0.99z); or (5) high distress (≥1z). Crosstabulations of rates of child disability by degree of neighbor-hood distress were used to display the results. In addition,Pearson’s correlation coefficient was used to measure the

Table I: Concentration of child disability in distressedneighborhoods

Functional Ratio of disability Proportion of disabled

limitation in distressed areas to children in distressed

non-distressed areas neighborhoods, %

Self-care 2.05 28.8Motor 1.71 24.9Sensory 0.94 14.4Any disability 1.67 24.2All children n=157 000 14.5

Table II: Correlations between child disability rates andneighborhood distress

Functional limitation Child disability and neighborhood

distress correlation

Any disability 0.349Self-care 0.274Motor 0.267Sensory 0.151

Figure 1: Child disability rates per 1000 children and

neighborhood distress.

None = no distress (≤–1z)

Min = minimum distress (–0.99z to –0.51z)

Some = some distress (–0.5z to 0.5z)

Mod = moderate distress (0.51z to 0.99z)

High = high distress (≥1z)

None

N/1

000

40

35

30

25

20

15

10

5

0

Level of neighborhood distress

Min Some Mod High

association between degrees of neighborhood distress andrates of child disability.

Ethical approval for the study was granted by theInstitutional Review Board at Brown University.

ResultsAmong the 157 000 children aged 5 to 15 years living in RhodeIsland in 2000, 21 per 1000 had some functional disability.These limitations included sensory (9 per 1000), self-care (9per 1000), and motor disabilities (10 per 1000). The overallrate of these disabilities was less than the sum of each type ofdisability because some children had multiple disabilities.

Table I compares rates of child disability in distressedneighborhoods with those of the rest of the state for motor,self-care, and sensory limitations. Rates of disability were sig-nificantly higher in distressed neighborhoods. The total childdisability rate in all distressed neighborhoods combined was35 per 1000. In distressed neighborhoods, self-care disabilityoccurred in 18 per 1000, motor disability occurred in 17 per1000, and sensory disability occurred in 9 per 1000. Overallchild disability was 67% higher in distressed neighborhoodsthan non-distressed neighborhoods.

Analysis revealed that self-care disability and motor dis-ability were substantially higher in distressed neighborhoodscompared with non-distressed neighborhoods. Of all childrenin Rhode Island, one in seven lived in distressed neighbor-hoods, yet one in four children with motor disability and twoin seven with self-care disability lived in those neighborhoods.This translates into 71% more motor and 105% more self-caredisability in children living in distressed neighborhoods.

Table II shows the correlation between rates of child disabil-ity and summary z score measures of neighborhood distress.Pearson’s correlation coefficients between type of disabilityand neighborhood distress were highest for self-care disabil-ity. As self-care disability reflects more severe functional chal-lenges, this indicates that severe disability is concentrated indistressed neighborhoods. In addition, the high correlationbetween any disability, which includes children with sensory,motor, and/or self-care limitation, reflects that overall dis-ability is more common in distressed neighborhoods.

Figure 1 describes the total rate of child disability per 1000children in relation to levels of neighborhood distress. Ateach level of neighborhood distress, rates of child disabilityincreased so that children in the most distressed neighbor-hoods had a rate of 37 children with disability per 1000 chil-dren compared with a rate of 11 per 1000 in advantagedneighborhoods.

If the conditions of neighborhood distress were improvedso that the number of children living with severe disadvan-tage decreased, the overall rates of child disability woulddecrease as well; there potentially would be 50% less childdisability. This suggests that population-based interventionscould substantially decrease rates of child disability if theywere able to affect children who were the most vulnerable.

DiscussionO’Hare and Mather demonstrated that between 1990 and2000, there was a decrease in the number of children living inneighborhoods with high poverty rates.15 However, during thesame period the number of children living in severely-dis-tressed neighborhoods increased 18%, from 4.7 million to 5.6million. The current study aimed to investigate the relationship

of these statistics to child disability rates. The State of RhodeIsland was chosen as the initial cohort because of the state’spolicies in providing all children, regardless of their citizenshipstatus, with health insurance and ensuring that all children canattend state-funded school and receive special education ser-vices. Although these policies aim to eliminate health dispari-ties between neighborhoods, this study found that childrenwith motor and sensory disability were disproportionatelyresiding in distressed neighborhoods. This is consistent withfindings of other investigators who used a different frameworkof analysis and explored potential mechanisms of how povertyadversely affects child health, well-being, and disability;16–18

these analyses examined home environments, social supports,neighborhood crime, race, and ethnicity.

The strength of the current data is that both the neighbor-hood indicators and child disability survey questions havebeen used in previous studies.12,18–20 A weakness of this studyis that it lacked descriptions of positive outcomes of childrenliving in adversity. Also, the current indications for learningand behavior disability were not included in this study becausethey are not similarly interpreted by social class, ethnicity,and community professionals.

Results present several challenges to current policy. First,passing legislation in support of childhood health insuranceand educational accountability does not immediately counter-act the adversities children encounter while living in dis-tressed neighborhoods. Of the 5.6 million children growingup in distressed neighborhoods across the US, 55% areAfrican-American and 29% are Hispanic. Overall, two in sevenAfrican-American children and one in seven Hispanic childrenlive in severely-distressed neighborhoods, compared with onein 100 non-Hispanic, white children.15 If the US is to achieveits goal of lessening health disparities, attention needs to begiven to children in distressed neighborhoods – the imple-mentation of a child safety net for comprehensive health,housing, education, and income support is required.

Several historic interventions have demonstrated the impor-tance of comprehensive preschool support for children at riskfor disability. These include the Infant Health and Develop-mental Program,21 The Perry Preschool Program,22 and theBrookline Early Education Program.23 It has been arguedthat we should think of the period of early childhood as anopportunity to invest in skill formation.24 This investmentstrategy leads to biopsychosocial interventions that are devel-opmentally and functionally enhancing.25

ConclusionFailure to address systematically the challenges of the largenumber of children living in severely-distressed neighbor-hoods results in children having suboptimal physical anddevelopmental health. This substantially reduces their chancesfor participation in the requirements of higher educationand competitive employment. In addition, given recent trendsin health policy with respect to children with special health-care needs, children with the greatest needs are least likely toget the minimum support required for full citizenship.Unless we can address this challenge, we cannot claim thatwe wish all children to be healthy, ready to learn, and trainedfor the challenges of adulthood in the 21st century.

From a public health policy perspective, we need toreduce the absolute number of children with disabilities.18,19

However, from a medical system perspective, we also need to

816 Developmental Medicine & Child Neurology 2007, 49: 814–817

Distressed Neighborhoods and Child Disability Rates Michael E Msall et al. 817

provide comprehensive interventions that optimize healthand functioning and have major impacts on long-termdependency and community participation.26 Health profes-sionals can promote support that builds on a child’sstrengths while ensuring that medical, educational, andrehabilitation interventions prevent regression, optimizelearning, and maximize the chances for independent living.

Accepted for publication 22nd May 2007.

Acknowledgments

This research was supported in part by grant 1U01 HD37614entitled ‘NICHD Family and Child Well Being Network: ChildDisability.’ This manuscript is dedicated to Elizabeth Burke Bryantfor her commitment to biopsychosocial interventions on behalf ofvulnerable populations. Larry Gray, Peter Smith, Ron Espinal, TomBlondis, Jennifer Park, Patricia Flanagan, Mary Sullivan, Betty Vohr,Ellen Gurney, and Diana Havill Ryan provided critical feedback.Herb Abelson, Paula Jaudes, and Steve AN Goldstein provided thesupports for a shared vision that combines the best of science andadvocacy. This article was presented in part at the AmericanAcademy of Cerebral Palsy and Developmental Medicine AnnualScientific Meeting, Los Angeles, California, October 1, 2003.

References1. Szilagyi PG, Shone LP, Klein JD, Bajorska A, Dick AW. (2007)

Improved health care among children with special health careneeds after enrollment into the State Children’s HealthInsurance Program. Ambul Pediatr 7: 10–17.

2. National Research Council and Institute of Medicine. (2004)Children’s Health, the Nation’s Wealth: Assessing and Improving

Child Health. Committee on Evaluation of Children’s Health.Board on Children, Youth, and Families, Division of Behavioraland Social Sciences and Education. Washington, DC: NationalAcademy Press.

3. National Research Council and Institute of Medicine. (2000)From Neurons to Neighborhoods: The Science of Early Childhood

Development. Committee on Integrating the Science of EarlyChildhood Development. Shonkoff JP, Phillips DA, editors. Boardon Children, Youth, and Families, Commission on Behavioral andSocial Sciences and Education. Washington, DC: NationalAcademy Press.

4. Plotnick, RD. (1997) Child poverty can be reduced. Future Child

7: 72–87.5. Huston AC. (1991) Children in Poverty. New York, NY:

Cambridge University Press.6. Kawachi I, Berkman LF. (2003) Neighborhoods and Health. New

York, NY: Oxford University Press. 7. Spencer N. (1996) Poverty and Child Health. New York, NY:

Radcliffe Medical Press.8. Cagney KA. (2006) Neighborhood age structure and its

implications for health. J Urban Health 83: 827–834.9. Duncan GJ, Brooks-Gunn J. (2000) Family poverty, welfare

reform, and child development. Child Dev 1: 188–196.

10. Keating DP, Hertzman C. (1999) Developmental Health and the

Wealth of Nation: Social, Biological, and Educational

Dynamics. New York, NY: The Guilford Press.11. Horn IB, Beal AC. (2004) Child health disparities: framing a

research agenda. Ambul Pediatr 4: 269–275. 12. Hogan DP, Msall ME, Rogers ML, Avery RC. (1997) Improved

disability population estimates of functional limitation amongAmerican children aged 5–17. Matern Child Health J

1: 203–216. 13. Hogan DP, Rogers ML, Msall ME. (2000) Children with disability

in the 1994 and 1995 National Health Interview Surveys onDisability: functional limitations and key indicators of well-being. Arch Pediatric Adolesc Med 154: 1042–1048.

14. Stein REK, Johnson Silver E. (1999) Operationalizing aconceptually based noncategorical definition: a first look at US children with chronic conditions. Arch Pediatr Adolesc Med

153: 68–74.15. O’Hare W, Mather M. (2003) The growing number of kids in

severely distressed neighborhoods: evidence from the 2000Census. KIDS Count PRP Report. www.kidscount.org (Annie ECasey Foundation) and www.ameristat.org (PopulationReference Bureau Ameristat).

16. Bradley RH, Corwyn RF, McAdoo HP, Coll CG. (2001) The homeenvironments of children in the United States part I: variationsby age, ethnicity, and poverty status. Child Dev 6: 1844–1867.

17. Ceballo R, McLoyd VC. (2002) Social support and parenting inpoor, dangerous neighborhoods. Child Dev 4: 1310–1321.

18. Newacheck PW, Hung Y, Wright KK. (2002) Racial and ethnicdisparities in access to care for children with special health careneeds. Ambul Pediatr 2: 247–254.

19. Starfield B, Robertson J, Riley AW. (2002) Social class gradientsand health in childhood. Ambul Pediatr 2: 238–246.

20. Waterston T, Alperstein G, Brown SS. (2004) Social capital: a keyfactor in child health inequalities. Arch Dis Child

89: 456–459.21. Infant Health and Development Program. (1990) Enhancing the

outcomes of low birthweight, premature infants: a multisiterandomized trial. JAMA 263: 3035–3042.

22. Barnett WS. (1996) Lives in the balance: benefit-cost analysis of

the Perry Preschool program through age 27. (Monographs ofthe High/Scope Educational Research Foundation, No. 11).Ypsilanti, MI: High/Scope Press.

23. Palfrey JS, Hauser-Cram P, Bronson MB, Warfield ME, Sirin S,Chan E. (2005) The Brookline Early Education Project: a 25-yearfollow-up study of a family-centered early health anddevelopment intervention. Pediatrics 116: 144–152.

24. Heckman JJ. (2006) Skill formation and the economics ofinvesting in disadvantaged children. Science

312: 1900–1902. 25. Knudsen EI, Heckman JJ, Cameron JL, Shonkoff JP. (2006)

Economic, neurobiological, and behavioral perspectives onbuilding America’s future workforce. Proc Natl Acad Sci U S A

103: 10155–10162. 26. World Health Organization. (2001) ICF: International

Classification of Functioning, Disability and Health. Geneva:World Health Organization.