Neurodevelopmental and School Performance of Very Low-Birth-Weight Infants: A Seven-Year Longitudinal Study

(1)

Neurodevelopmental

and School

Performance

of Very

Low-Birth-Weight

Infants:

A

Seven-Year

Longitudinal

Study

Betty

R.

Vohr, MD, and Cynthia T. Garcia CoIl, PhD

From the Department of Pediatrics, Women and Infants Hospital of Rhode Island, and Section of Pediatrics, Brown University Program in Medicine, Providence

ABSTRACT. The changing patterns of neurologic and

developmental functioning between 1 and 7 years of age were studied in very low-birth-weight infants (birth weight 1,500 g). Subjects included 42 infants born in

1975 who were followed for 7 years. Based on the 1-year neurologic assessment, 22 infants were classified as nor-mal, 12 as suspect, and eight as abnormal. The three groups did not differ in birth weight, gestational age, sex, or Hollingshead socioeconomic status (SES) score. The neurologic findings at 7 years of age were significantly related to the neurologic examination findings at 1 year of age. Seventy-seven percent of the normal group, 58% of the suspect group, and 100% of the abnormal group remained in the same neurologic category at 7 years of age. Children in the abnormal group had the greatest improvement in cognitive functioning between 1 and 7 years of age but did not achieve the IQ level of children in the normal group. Forty-five percent of the normal group, 75% of the suspect group, and 100% of the abnor-mal group had poor visual-motor integration. Fifty-eight percent of the suspect group and 87% of the abnormal group were reading below age level. Of the total sample, 54% required special education or resource help at 7 years of age, and the three groups differed significantly in their need for a special educational plan (P < .05). These data indicate that a neurologic classification at 1 year of age provides a guide for monitoring very low-birth-weight infants and can be helpful in alerting school personnel to potential needs. Pediatrics 1985;76:345-350; neurologic performance, developmental performance, low-birth-weight infant, school performance.

Although the incidence of major neurologic and

developmental sequelae in very low-birth-weight

(VLBW) survivors has gradually declined, the

in-cidence of minor neurodevelopmental sequelae,

Received for publication Aug 21, 1984; accepted Dec 20, 1984. Reprint requests to (B.R.V.) Women and Infants Hospital of Rhode Island, 50 Maude St, Providence, RI 02908.

learning disabilities, and needs for special education

have continued to be an area of concern.7 The

purpose of this study was to assess the changing patterns of neunologic and developmental

perform-ance of VLBW infants (birth weight 1,500 g) for

the first 7 years of life. Our hypotheses were that: (1) infants who are abnormal on suspect neurologi-ca!!y at 1 year of age will be at greater risk of

persistent atypical neurodevelopmental findings

and subsequently will have greater needs for special education services; and (2) “soft” neunologic signs

and subtle developmental abnormalities will be

prevalent at 7 years of age in VLBW survivors.

MATERIALS AND METHODS

The data base consisted of 121 infants with a

birth weight of 1,500 g who were treated in the

Special Cane Nursery at Women and Infants

Hos-pita! of Rhode Island during 1975. Of 121 infants, 70 (57%) survived the neonatal period and eight infants subsequently died before their first birth-day. All survivors were prospectively enrolled in the

Neonatal Follow-up Program for longitudinal

com-prehensive assessments of neurologic and develop-mental performance. Of the 62 survivors, 42 (67%)

were followed neunodevelopmentally at 3 to 4, 5,

and 7 years of age and constitute the sample of this study. The reasons for not including the remaining 20 children in this report are as follows: an evalu-ation was done at 1 year of age only for 16 infants (seven normal, five suspect, and four abnormal); an

evaluation was done only at 7 years of age for two

children, both normal; and two children were lost

to follow-up. The medical records of the 42 study

children were reviewed, and a modified Hobel8 risk

score composed of prenatal, intrapartum, and

neo-natal risk factors was completed on each subject. The Hollingshead#{176} Four Factor Index was

(2)

Birth wt (g) Gestation (wk) Sex (M/F)

Hollingshead socioeconomic status Hobel risk score

(range)

C Valuesaremeans ± SD. Classifications are defined asfollows: normal, no neurologic abnormality; suspect, deviations of tone, posture, movement patterns, reflexes, cranial nerves, or head growth; abnormal, cerebral palsy, blind, deaf.

t P < .05 v values for normal group.

TABLE 1. Clinical Characteristics of Study Infants*

paternal educational and occupational level. Panen-tal consent was obtained for all participants in the

study.

A neunologic assessment was completed by one

of the authors (B.R.V.) at the time of each yearly visit. During the first year of life, the infants were classified as normal (no neunologic abnormality), suspect (deviations oftone, posture, movement pat-terns, reflexes, cranial nerves, on head growth), on abnormal (cerebral palsy, blind, deaf). The

first-year neurologic examination was based on the

com-bined methods of Pnechtel’#{176} and those used in the Collaborative Perinatal Study” and provided the

basis for establishing the three study groups.

Twenty-two infants were classified as normal, 12

as

suspect, and eight as abnormal. Intrauterine growth retardation defined as birth weight <10%

for gestation was present in two of eight in the

abnormal group, four of 12 in the suspect group,

and three of 22 in the normal group.

As shown in Table 1, the three study groups did

not differ in birth weight, gestation, sex distnibu-tion, and Hollingshead index. There was a definite

trend, however, for the abnormal and suspect

groups to have a higher Hobel risk score than the

normal group. The difference was significant

be-tween normal and suspect groups (P < .05). Developmental testing at 1 year of age (connected for gestational age) consisted of the Bayley Scale

of Infant Development’2 (mental and motor scones)

and, at 3 to 5 years of age, the Stanford-Binet Intelligence Scale.’3 IQ scores of the blind children

were obtained from evaluations completed at the

special education facilities that they attended. These evaluations included the Psychological

Stim-ulus Response Evaluation (PSR)’4 and the verbal

portion of the Wechslen Preschool and Primary

Scale of Intelligence (WPPSI).’5

At 7 years of age, the evaluation procedure was

expanded as follows. The neurologic performance of the study subjects was determined using standard neunologic examination to assess tone; reflexes; and

Normal

Group (n=22)

1,286 ± 168

30.6 ± 2.0

8/14

39.2 ± 12.0

118 ± 23 (76-166)

Suspect Group (n = 12)

1,172 ± 205

30.6 ± 2.0

7/5 28.8 ± 9.0

146 ± 45t

(80-232)

Abnormal Group

(n=8)

1,253 ± 169 30.0 ± 2.0

4/4 37.5 ± 18.0

142 ± 48

(92-240)

motor, sensory, and verbal performance in

conjunc-tion with the Riley’6 motor examination, which

consists

of

specific oral motor, fine motor, and gross motor components. The

psychological-developmen-tal evaluations included the Stanford-Binet Intel-ligence Scale,13 the Beery Visual-Motor Integration

Test,’7 the Wide Range Achievement Test

(WRAT)’8 for reading, and the subscales of picture completion (perception) and block design (percep-tual integrative) from the Wechsler Intelligence

Scale for Children-Revised (WISC-R))5 IQ scores

for the two blind children were again obtained from

tests

performed at the special education facilities these children attend. Based on these neurodevel-opmental assessments, the children were classified at 7 years of age as normal (no neurologic or

devel-opmental handicap and an IQ 80); suspect (fine

on gross motor inefficiency, language inefficiency, on seizure disorder); or abnormal (cerebral palsy,

blind, deaf, or hydrocephalic). A questionnaire on

resource needs, special education placement, and

academic performance was completed by personnel

at the children’s schools.

During the interim between 1 and 7 years of age,

primary care was provided by the subjects’

com-munity physicians. Referrals for early intervention programs on preschool special education placement

were recommended either by the follow-up team or

the primary physician.

A set of three-way analysis of variance (ANOVA) was used to compare the outcome variables for the three study groups. Unpaired Student’s ttests were

used to identify significant differences between

groups, and x2 analysis was used to compare the

distribution of the groups on certain variables.

Be-cause of incomplete developmental assessments at

2 and 3 years of age, we analyzed the data for 1, 3

to

4, 5, and 7 years of age only.

RESULTS

Neunologic status of the three study groups at 1,

(3)

TABLE 2. Neurologic Classification*

1 Year of Age 3-4 Years

NSA

0 0 8

5 Years

N S A

18 0

7 2 0

0 0 8

Normal (n = 22) 18 0

Suspect (n = 12) 6 3

Abnormal (n = 8) 0 0

Children seen: No./(%) 35 (83%) 35 (83%) 42 (100%)

7 Years

N S A

17 5 0 5 7 0

0 0 8

* Classifications (normal [N], suspect [SJ, abnormal [A]) are defined in Table 1 footnote.

* Classifications (normal, suspect, abnormal) are defined in Table 1 footnote. Shifting of neurologic classification of the suspect

and normal groups occurred between 1 and 5 years

of age but was most apparent at 7 years of age when the neurologic examination was expanded to enable specific quantification of subtle motor

abnormali-ties.

All eight infants classified as abnormal at 1

year of age remained abnormal throughout this

period of the study. Specific neurologic diagnoses of the total group are seen in Table 3. Severity of

gross motor handicap diminished in seven of eight

children in the abnormal group who were

ambula-tory by 7 years of age, although seizure

abnormali-ties had developed in three of these children. In

contrast, “soft” neurologic findings (fine and gross

motor inefficiencies) were present in five of 22

(23%) children classified as normal at 1 year of age.

The 7-year motor performance of 40 children

(two blind children were not tested) was evaluated

further with the Riley motor examination. In this

test, the higher the score, the greater the motor

abnormality. Motor scores of the three groups are

shown in Table 4. An increasing degree of

abnor-mality in motor performance was identified from

the normal group to the suspect group to the

ab-normal group. Gross motor performance of the

sus-pect group was significantly less proficient than

that of the normal group. However, the overall

motor performance score was evaluated for all three

groups because all three groups’ mean scores were

above three (the median score for the standardiza-tion sample of the test).

Mean mental development index and IQ scones

at 1, 3 to 4, 5, and 7 years of age are shown in Table 5. Mean developmental test scores of the children rated as neurologically normal, suspect, or

abnor-mal at 1 year of age consistently have similar

de-velopmental quotient (DQ) or IQ scores from 1 to

7 years of age. Scores of children in the suspect group show more variability and are significantly

lower than scores of the normal group only at 3 to

4 years of age. In the abnormal group, six of the

eight children have DQs 80 at 1 year of age

compared with only two of eight children at 7 years

of age, demonstrating a pattern of improving

cog-nitive performance (developmental catch-up) over time. Despite this, the mean IQ score of the

abnor-ma! group remains significantly lower than that of

the normal group at 7 years of age.

IQ scores, visual-motor integration (Beery test

scones), and reading ability (Wide Range

Achieve-TABLE 3. Neurologic Diagnosis*

Status at 1 Year of Age Normal (n = 22)

Suspect (n = 12)

Abnormal (n = 8)

3 with spastic diplegia 2 with right hemiplegia

1 with spastic diplegia with retrolental fibroplasia and hydrocephalus 1 with spastic diplegia

1 with spastic diplegia and retrolental fibroplasia

Outcome at 7 Years Normal (n = 17)

Suspect (n = 5)

3 with gross motor inefficiency 2 with fine motor inefficiency Normal (n = 5)

Suspect (n = 7)

4 with gross motor inefficiency 1 with fine motor inefficiency 1 with controlled seizure disorder 1 with controlled seizure disorder and

gross motor inefficiency Abnormal (n = 8)

Improving and ambulatory Improving and ambulatory

Severely involved, deaf, with seizures

(4)

Bayley

1 Year of Age

Stanford-Binet

3-4 Years 5 Years 7 Years

Normal(N) (22) Suspect(S) (12) Abnormal(A) (8)

101 ± 15 (20) 95 ± 21

(11)

68.2 ± 18

(8)

101 ± 13 (18) 84 ± 15

(9)

79.2 ± 15

(7)

105 ± 16 (17) 90 ± 16

(9)

80.6 ± 23

(5)

105 ± 15 (22) 97 ± 18

(12)

86 ± 20

(8)

Statistical significance NvS

N v A SvA

NS P < .001

NS

P<.01 P < .005

NS

NS P < .025

NS

NS P < .05

NS

*Includes Psychological Stimulus Response Evaluation and verbal portion of Wechsler Preschool and Primary Scale of Intelligence for two blind children at 7 years of age. Values are means ± SD. Number of study subjects is shown in parentheses. Classifications (normal, suspect, abnormal) are defined in Table 1 footnote.

IQ

BEERY

(%)

wR AT

(%)

TABLE 4. Riley Motor Examination: Findings and Significance*

No. in Oral Motor Fine Motor Gross Motor Total Motor

Group

Normal(N) 22 2.18 ± 0.85 2.13 ± 1.35 2.04 ± 2.29 6.36 ± 3.28

Suspect(S) 12 2.58 ± 1.6 2.33 ± 1.66 4.58 ± 1.28 9.50 ± 4.50

Abnormal(A) 6 3.66 ± 0.81 4.50 ± 0.83 6.13 ± 0.98 15.33 ± 2.16

Statistical significance

NvS NS NS P<.001 NS

NvA P<.025 P<.005 P<.001 P<.001

SvA NS P<.005 NS P<.001

* Values are means ± SD. Classification (normal, suspect, abnormal) are defined in Table

1 footnote.

TABLE 5. Developmental Test Scores: Findings and Significance*

ment Test scones) of the three groups at 7 years of

age are shown in the Figure. A high percentage of

children in the normal group (10/22, 45%) had

visual-motor integration scones that fell below the 50th percentile, although the majority (20/22) had age-appropriate reading skills. Nine of 12 (75%) of

children in the suspect group demonstrated less

efficient visual-motor skills (below the 50th

per-centile) and 7/12 (58%) were reading below the

50th percentile. In the abnormal group, 4/5 children had reading abilities below the 50th percentile. All five children tested had inefficient visual perceptual motor skills.

There were no significant differences, however,

in age-equivalent performance in the WISC-R

pic-tune completion (normal group, 7.6 ± 1.4 years of

age; suspect group, 7.1 ± 1.1 years of age; and

abnormal group, 6.8 ± 0.5 years of age) or in the

WISC-R block design (normal group, 7.5 ± 1.4

years of age; suspect group, 7.3 ± 1.7 years of age; and abnormal group, 6.6 ± 8.0 years of age). There

was a

trend, however, for the suspect and abnormal groups to achieve lower mean scores, and small sample size may be a limiting factor in identifying

differences with this test. Three children in the

abnormal group were untestable on these subtests

because of blindness on low functioning.

0

10 0 cD

50.1

of

too-I

%c5b

0100 cPp p

p(o.o5 O% 4’#{176}cP%% 0 1 0o,, 0 _MtSD O 0 c

o”:aP I oo:

po.o5

‘ 0cb

#{176}#{176}fb#{225}’O

I

I 0cIb#{176} ‘#{176}‘#{176}#{176}‘

NORMAL SUSPECT ABNORMAL

n.22 n-12 n.8

Figure. Individual and mean scores on Stanford-Binet Intelligence Scale (IQ), Beery Visual Motor Integration Test, and Wide Range Achievement Test (WRAT). Anal-ysis of variance scores are: IQ, P < .05 (normal v abnor-mal); Beery, P < .05 (normal v abnormal); WRAT, P <

.001 (normal v suspect and normal v abnormal).

School placement for all children in the study at 7 years of age is shown in Table 6. Whereas 72% of the children in the normal group were appropriately placed in either first or second grade, 6/12 (50%)

of the suspect group and only 1/8 (12%) of the

abnormal group were in a regular first or second

grade. x2 analysis (2 x 3) indicated a significant

difference between the three groups of children in

their need for an alternate educational plan (P <

(5)

Status at 1 Year of Age

Age-Appropriate School Placement

Yes Special Educa-tion

Other Needs

Special Resource

Speech and Language

Remedial

Mathe-matics

Physical

Ther-apy

Occupa-tional Therapy

Normal (n = 22) Suspect (n = 12)

Abnormal (n = 8)

16 (72%) 0

6 (50%) 3

1 (12%) 5

6

3 2t

3

3 1

3

0 . . .

0

1 8

0

0 4

x2 P<.05

*Classifications (normal, suspect, abnormal) are defined in Table 1 footnote. Resource, Individualized educational programming with up to three 45-minute segments with a diagnostic prescriptive teacher.

t One child received resource during physical therapy. TABLE 6. School Placement*

therapy and occupational therapy requirements. The mean Hollingshead SES score, which was updated at 7 years of age, did not differ significantly between children in the neurologically suspect and normal categories at 7 years of age. It did, however, differ significantly between children receiving com-prehensive special educational services (23.5 ± 7.0), children receiving educational resource help (41.4

± 13.0), and children normally placed in school

(36.1 ± 11.0) (special education v special resource: P < .005) (special education v normal: P < .025).

DISCUSSION

The purpose of this study was to assess the

patterns of neurologic and developmental change

of VLBW infants from 1 to 7 years of age, and to determine whether neurologic classification at 1 year of age was a marker for subsequent

neurode-velopmental status and school performance.

Al-though all of the children who were classified as abnormal neurologically during the first year of age continued to manifest abnormal neurologic findings at 7 years of age (a 100% correlation), severity of abnormality had lessened in six of the eight chil-dren, and they had become ambulatory. Fifty-eight percent of the children who were classified as sus-pect at 1 year of age continued to have soft neuro-logic findings, although 41% had completely re-solved atypical neurologic findings and were consid-ered normal at 7 years of age. Theis phenomenon of resolving or improving neurologic findings in VLBW infants has been observed previously.’9 On

the other hand, five of 22 children who were

con-sidered normal at 1 year of age had subtle fine and gross motor coordination difficulties at 7 years of age. The Riley motor testing used in conjunction with the neurologic examination confirmed coon-dination inefficiencies, specifically, the gross motor finding of the inability to balance on one foot in a manner appropriate for their age. Balancing inef-ficiencies have previously been identified in VLBW

survivors7’20 and raise the possibility of a cerebellar

insult. Although this shift from the normal to the

suspect category is somewhat disturbing, it should be emphasized that the neurologic findings in these five children at 7years of age were relatively minor and did not interfere with everyday activities. It appears that neurologic status does fluctuate and change over time in VLBW survivors, and, although the neurologic assessment in the first year is appro-pniate for identifying the abnormal infant and iden-tifying candidates for early intervention, repetitive evaluations are necessary to identify children with more subtle problems.

Mean developmental test scores using the Bayley

scale at 1 year of age and the Stanford-Binet test

at 3, 4, 5, and 7 years of age reveal three differing developmental patterns. The children considered neurologically normal at 1 year of age perform consistently better on standardized testing at all ages with scores at 7 years of age within the normal range. Children considered suspect have more var-iability within the group with a small percentage demonstrating lower cognitive abilities at 7 years of age. The abnormal group actually demonstrates fairly dramatic improvement in mean test scores between 1 and 7 years of age with the mean score, however, remaining low because of the consistently low test scores of the two blind children. Despite an IQ in the normal range (>80) at 7 years of age in all but four of 42 children, there was an increased incidence of learning problems, especially visual-motor integration difficulties, which were identified in all three groups of children. Visual-motor inte-gration inefficiencies were previously observed by Weiner et a121 in 8- to 10-year-old children who had had very low birth weight. Reading difficulties were

also encountered in 66% of the suspect group and

80% of the abnormal group. The study of Nickel et al7 (survivors 1,000 g) also identified reading levels below age expectation. These findings

under-score the importance of a comprehensive

(6)

evaluate

developmental outcome.

The pre-intensive-care era (referring to babies

born between 1948 and 1956) long-term follow-up

study of Dnillien’ revealed a high incidence (50%) of children requiring placement in special schools for the handicapped because of mental or physical

handicap and another 25% who were considered

low functioning and requiring special education

services. In contrast, the Stewart et a15 school-age follow-up of 50 infants born between 1966 and 1976 (early intensive-care era) determined that 76% of

the children 8 years of age or older were placed

appropriately in school, a substantial improvement

oven the previous study. Therefore, the high

per-centage of children in need of an alternate educa-tion plan at 7 years of age was not expected. How-ever, because of the enactment of the Education for

All Handicapped Children Act, Public Law 94-142

(1975) and the ready availability of appropriate screening and special education services within the

Rhode Island education system, we make the

as-sumption that the percentage of children receiving alternate educational services is probably valid.

This included 27% of children normal at 1 year of

age, 50% of children suspect at 1 year of age, and

87% ofchildren abnormal at 1 year ofage, an overall incidence of 54%. Nickel et a17 also recently iden-tified that 64% of children with birth weight 1,000 g or less born between 1960 and 1972 were in special

education placement at a mean age of 10.6 years.

These findings seem to indicate that, although the

continuing improved survival of VLBW infants has

been associated with a leveling off of the incidence

of major neurologic sequelae, the incidence of

de-velopmental inefficiencies and academic difficulties may be increasing. A limiting factor in interpreting

the outcome variables is the limited number of

children in this long-term follow-up study and the

lack of a control group. Future collaborative studies are needed to clarify the issues surrounding

long-term outcome. Also, follow-up studies of even

longer duration will clarify whether these children eventually “outgrow” their academic difficulties.

Social and environmental status did not have an

apparent influence on changing neurologic outcome

(fine and gross motor function), although it did

relate to school placement. Children from the fam-ilies with the lowest socioeconomic status families were most likely to require comprehensive special education service.

In summary, there was a close relationship

be-tween abnormal neurologic findings at 1 year of age

in

VLBW infants and neurologic, developmental,

and school performance at 7 years of age. The

relationship between suspect and normal

neuro-logic status and outcome is less clear. Although some neurologic problems are resolved, others emerge in conjunction with specific learning diffi-culties (visual-motor integration deficit). Our data

suggest that a neurologic classification of VLBW

infants at 1 year of age provides a guide for ongoing monitoring of high-risk infants.

ACKNOWLEDGMENTS

We thank Marie Preneta, RN, for technical assistance and Janice Brusini for typing the manuscript.

REFERENCES

1. Drillien CM: The incidence of mental and physical handi-caps in school age children of very low birth weight. Pedi-atrics 1961;27:452-464

2. Lubchenco LO, Homer, FA, Reed LH, et al: Sequelae of premature birth. Am J Dis Child 1963;106:101

3. Rawlings, G, Reynolds EOR, Stewart A, et al: Changing prognosis for infants of very low birth weight. Lancet

1971;1:516

4. Francis-Williams J, Davies PA: Very low birth weight and later intelligence. Dev Med Child Neurol 1974;16:709-729 5. Stewart A, Turcan D, Rawlings G, et al: Outcome for infants

at high risk of major handicaps. Ciba Found Symp 1978;59:151-164

6. Hack M, Fanaroff AA, Merkatz IR: The low birth weight infant: Evolution of changing outlook. N Engi J Med

1979;301:1 162-1165

7. Nickel RE, Bennett FC, Lamson FN: School performance of children with birth weights of 1000 g or less. Am J Dis

Child 1982;136:105

8. Hobel CJ, Hyvarinen MA, Okada DM, Oh W: Prenatal and intrapartum high-risk screening. Am J Obstet Gynecol

1973;117:1

9. Hollingshead AB: Four Factor Index of Social Status, work-ing paper. Yale University, New Haven, CT, 1975

10. Prechtel H, Beintema D: The Neurological Examination of

the Full Term Newborn Infant. London, William G Heine-man Ltd, 1964

11. Niswander KR, Gordon M: The Collaborative Perinatal Study: The Women and Their Pregnancies. Philadelphia, WB Saunders Co, 1972

12. Bayley Scale of Infant Development. New York, Psychologi-cal Corp, 1969

13. Terman LM, Merrill MA: Stanford-Binet Intelligence Scale. Boston, Houghton-Mifflin Co, 1973

14. Mullen E: Psychological Stimulus Response Evaluation for Severely Multiply Handicapped Children. Providence, RI, T.O.T.A.L. Child mc, 1977

15. Wechsler D: Manual for the Wechsler Intelligence Scale for Children-Revised. New York, Psychological Corp, 1974 16. Riley GD: Riley Motor Problems Inventory. Los Angeles,

Western Psychological Services, 1976

17. Beery KE, Buktenica NA: Developmental Test of Visual-Motor Integration. Chicago, Follett Publishing Co, 1967 18. Jastak JF, Jastak 5: The Wide Range Achievement Test:

Manual of Instructions. Wilmington, DE, Jastak Assoc mc,

1978

19. Nelson KB, Ellenberg JH: Children who ‘outgrew’ cerebral palsy. Pediatrics 1982;69:529

20. Black B, Brown C, Thomas D: A follow-up of 58 preschool children less than 1500 grams birthweight. Aust Paediatr J

1977;13:265-270

21. Weiner G, Rider RV, Oppel WC, et al: Correlates of low birth weight: Psychological status at 8 to 10 years of age.

(7)

1985;76;345

Pediatrics

Betty R. Vohr and Cynthia T. Garcia Coll

Seven-Year Longitudinal Study